(PDF) Population studies of horse mackerel (Trachurus trachurus (L.)) and herring (Clupea harengus L.) using parasites as biological tags.

The use of parasites as biological tags is a well used tool for fish stock identification. Recently, questions have been raised as to the appropriateness of stock boundaries in the Atlantic horse mackerel (Trachurus trachurus) and the herring (Clupea harengus) in European waters. Studies of the parasite fauna of these fishes could contribute useful information to resolve these issues. To this end, nineteen hundred and nineteen horse mackerel were examined from samples taken from up to nineteen sites around Europe in 2000 and 2001. These fish were subject to a complete parasitological examination. This resulted in the discovery of 45 species of parasites. Of these, 11 appear to be new host records, and one may be a new species. A number of these parasites may be suitable for use as biological tags and their application is discussed. Results showed a significant degree of separation was possible between fish from the North Sea and Western horse mackerel stocks, based on the relative abundance of the nematodes Anisakis spp. and Hysterothylacium sp.. differentiation between Western and Southern stocks and Southern and African stocks is less clear. Results from the Mediterranean Sea suggest three stocks, confined to the western, central and eastern basins. These findings compare well with results from otolith and body shape morphometry, and a range of genetic markers, which were applied to the same fishes, confirming both the usefulness of investigating parasites as biological tags and the currently applied stock management system. A neural network approach to classification based on parasitological data was evaluated, and found to be successful in over 90% of cases. This compares favourably with the results obtained by more traditional discriminant analysis methods. This will be a useful tool for further parasitological studies. The endoparasitic fauna of four thousand and thirty-three herring at various life stages (juvenile, spawning and non-spawning) from sites to the west of the British Isles and outlier samples from the North Sea, Baltic Sea and northern Norway, were examined to obtain information on stock identity, mixing and recruitment patterns in these areas. Results revealed that substantial stock mixing takes place to the west of the British Isles, with fish from the Irish Sea, west of Ireland and west of Scotland being found together in non-spawning aggregations to the west of Scotland. This area is also home to two different spawning populations of fish which recruited from nursery grounds in the eastern North Sea as well as from the west of Scotland, at different times of year. The implications of these findings for fisheries management are discussed. The ribosomal small sub-unit cytochrome oxidase (I) gene of the parasitic nematode worm, Anisakis simplex s.s., was sequenced from specimens taken from a number of herring hosts at sites to the west of Scotland, the Irish Sea, south and northwest coasts of Ireland and the Baltic Sea, to investigate any link between parasite population genetics and host population structure. Results revealed the COI gene to be highly variable, with around 50% of worms sequenced having unique haplotypes. There were no population structures evident from principal component analysis of genetic distances. This lack of genetic difference, even between the isolated population from the Baltic Sea and the other sites suggests that genetic diversity of the parasite population is controlled to a much higher degree by the movements of the marine mammals which act as final hosts than by any degree of population structure present in herring.

Population studies of horse mackerel (Trachurus trachurus (L.)) and herring (Clupea harengus L.) using parasites as biological tags. Neil Campbell B.Sc.(hons) (University of Newcastle, 1999) M.Sc. (University of Aberdeen, 2000) This thesis is submitted in fulfilment of the requirements of the degree of Doctor of Philosophy and may not be referred to without prior permission of the author or supervisors. Index Contents i List of figures viii List of tables xii List of appendices xiv Abstract xv Declaration xvii Acknowledgements xvii Chapter 1. An introduction to stock identification. 1 1.1 Introducing Fisheries 2 1.1.2 Why manage fisheries? 3 1.1.3 The process of fisheries management 5 The Stock Concept 8 1.2.1 Definition of a stock 8 1.2.2 History of stock identification 10 Principles of stock identification 11 1.3.1 Stock identifications techniques 13 1.2 1.3. 1.3.1.1 Distribution and abundance of fish, eggs and larvae 14 1.3.1.2 Morphological or meristic variations 18 1.3.1.2a Meristic variations 18 1.3.1.2b Analysis of body morphometry 19 1.3.1.2c Analysis of otolith morphometry 20 1.3.1.3 Variations in life history traits 23 1.3.1.3a Growth rate parameters 23 1.3.1.3b Reproductive parameters 24 1.3.1.4 Artificial tags 25 i 1.4 1.3.1.5 Variation in chemical composition 27 1.3.1.6 Genetic differences between populations 29 1.3.1.6a Multilocus allozyme electrophoresis (MAE) 30 1.3.1.6b Mitochondrial DNA sequence data 31 1.3.1.6c Microsatellite DNA 32 1.3.1.6d Single strand conformation polymorphism 34 1.3.1.7 Parasites as biological tags 35 1.3.1.7a Parasite genetics 40 1.3.2 Multivariate statistical methods of stock identification 41 1.3.3 Stock identification and management advice 43 Conclusions 47 Chapter 2. The biology and parasite fauna of the Atlantic horse mackerel, Trachurus trachurus (L.) 50 2.1 Introduction 51 2.1.1 Taxonomic status and distribution 51 2.1.2 Stock identity of the horse mackerel 53 2.1.3 Prey and predation 55 2.1.4 Reproduction, growth and migration 56 2.1.5 Fisheries 58 2.1.6 Parasites of horse mackerel 62 2.2 Methods 66 2.3 Results 71 2.3.1 List of species recorded 71 2.3.2 Descriptions 73 2.3.2.1 Apicomplexa 73 2.3.2.2 Myxosporea 75 2.3.2.3 Monogenea 82 2.3.2.4 Digenea 87 ii 2.4 2.5 2.3.2.5 Acanthocephala 97 2.3.2.6 Cestoda 101 2.3.2.7 Copepoda 104 2.3.2.9 Nematoda 107 Discussion 111 2.4.1 Apicomplexa 111 2.4.2 Myxosporea 112 2.4.3 Monogenea 116 2.4.4 Digenea 118 2.4.5 Cestoda 120 2.4.6 Acanthocephala 122 2.4.7 Crustacea 123 2.4.8 Nematoda 124 Conclusions 128 2.5.1 Ecological Implications 128 2.5.2 Nestedness 130 2.5.3 Conclusions 131 Chapter 3. Parasites as biological tags of stock identity in the horse mackerel, 134 Trachurus trachurus (L.). 3.1 3.2 3.3 Introduction 134 3.1.1 Parasites as biological tags for stock identification studies 135 3.1.2 Criteria for selection of tag species 137 3.1.3 Stock identity of horse mackerel 139 Methods 143 3.2.1 Sampling design 143 3.2.2 Dissection 148 3.2.3 Analysis of data 150 Results 151 3.3.1 Age and length information 151 iii 3.3.2 Parasite prevalence data 154 3.3.3 Selection of biological tag species 161 3.3.3.1 Apicomplexa 161 3.3.3.2 Myxosporea 163 3.3.3.3 Monogenea 165 3.3.3.4 Digenea 168 3.3.3.5 Cestoda 170 3.3.3.6 Acanthocephala 171 3.3.3.7 Nematoda 171 3.3.4 Parasites as biological tags 3.4 173 3.3.4.1 “Western” vs “North Sea” stocks 173 3.3.4.2 “Western” vs “Southern” stocks 175 3.3.4.3 Mediterranean stocks 176 Discussion 179 3.4.1 Parasites as tags 179 3.4.2 Comparison with other stock identification techniques 183 3.4.3 Implications for fishery management 185 Chapter 4. A comparison of neural network and discriminant analysis for 187 classification of horse mackerel stocks based on parasitological data. 4.1 Introduction 188 4.1.1 Stock identification 188 4.1.2 The horse mackerel 189 4.1.3 Statistical methodology in biological tag studies 192 4.1.4 Neural networks 193 4.2 Methods 199 4.3 Results 201 4.3.1 Selection of species for neural network analysis 201 4.3.2 Data exploration 202 iv 4.4 4.5 4.3.4 Neural network design 205 4.3.5 Discriminant analysis results 207 4.3.6 Neural network results 209 Discussion 213 4.4.1 Stock identification 213 4.4.2 Norwegian non-spawning sample 215 4.4.3 La Coruña Spawning Sample 216 Conclusions 217 Chapter 5. Parasites as biological tags for identification of herring stocks, and 218 patterns of recruitment and mixing of herring (Clupea harengus L.) to the west of the British Isles. 5.1 5.2 5.3 Introduction 219 5.1.1 Taxonomic status and distribution 220 5.1.2 Prey and predation 222 5.1.3 Stock Assessment of herring to the west of the British Isles 223 5.1.4 Spawning and Reproduction 225 5.1.4a Division VIIa (N) (Irish Sea) 226 5.1.4b Division VIIj and VIIa(S) (Celtic Sea) 226 5.1.4c Divisions VIIb and VIa (S) (North and West of Ireland) 227 5.1.4d Division VIa (N) (West of Scotland) 227 5.1.5 Parasites of Herring 228 5.1.6 Herring stock identity 231 5.1.7 Parasite tag studies of herring 234 5.1.8 The WESTHER project 235 Methods 236 5.2.1 Sample collection 236 5.2.2 Parasitological examination 245 5.2.3 Analysis of parasitological data 246 Results 248 v 5.4 5.3.1 Parasitology of herring 248 5.3.2 Long-term temporal variability 254 5.3.3 Parasites as biological tags 255 5.3.3.1 Short-term variability 258 5.3.3.2 Linking up life-history stages 262 5.3.3.3 Neural network analysis 265 Discussion 266 5.4.1 Parasites of herring 266 5.4.1.1 Temporal stability 268 5.4.1.2 Spatial variability 270 5.4.3 Conclusions 272 5.4.4 Implications for management 274 Chapter 6. Is cytochrome oxidase I (CO1) mtDNA of Anisakis simplex sensu stricto an indicator of host population biology? 276 6.1 Introduction 277 6.1.1 Biology and Taxonomy of the Anisakidae 277 6.1.2 Studies of parasite population genetics 278 Methods 279 6.2.1 Collection of specimens 279 6.2.2 Extraction, sequencing and analysis of CO1 region 280 6.3 Results 283 6.4 Discussion 287 6.2 Chapter 7. Incorporating the use of parasites as biological tags of horse mackerel and herring into fisheries management systems. 289 7.1 The management process 290 7.2 Informing the management of horse mackerel 292 7.2.1 292 North Sea vs. Western stocks vi 7.3 7.4 7.5 7.2.2 Western vs. Southern stocks 293 7.2.3 Extent of the Southern stock 294 7.2.4 Migratory behaviour in the Atlantic Ocean 294 7.2.5 Stock structure in the Mediterranean Sea 296 Informing the management of herring fisheries 298 7.3.1 Juvenile herring 298 7.3.2 Spawning aggregations of herring 299 7.3.3 Mixed stock aggregations 300 7.3.4 Management implications 301 Novel Techniques 302 7.4.1 Neural network analysis 302 7.4.2 Parasite population genetics 302 Conclusions 303 Bibliography 304 vii List of Figures Chapter 1 Page No. 1.1 Global marine fish production for the period 1950-2003. 2 1.2 A simple representation of the processes acting on a fish stock. 5 1.3 Historical trends in stock identification publication numbers. 11 1.4 Catches of horse mackerel in the North East Atlantic during the first and third quarters of 2001 14 1.5 Distribution of Herring in the North Sea and west of Scotland. 16 1.6 Representation of methods used to derive data from otolith shape. 21 Chapter 2 2.1 The Atlantic Horse Mackerel, Trachurus trachurus (L.). 52 2.2 Stock definitions of horse mackerel in the north east Atlantic. 55 2.3 ICES records of horse mackerel catches in the Atlantic from 1973 to 2004. 59 2.4 Horse mackerel catches in the Mediterranean Sea, as recorded by the General Council on Fisheries in the Mediterranean. 61 2.5 Location of sampling sites in northern Atlantic and Mediterranean waters. 67 2.6 Dissection of a horse mackerel. 69 2.7.1 Alataspora serenum in lateral view. 75 2.7.2 Alataspora solomoni in lateral view. 76 2.7.3 Kudoa sp. in apical and transverse views. 77 2.7.4 Kudoa nova in transverse and apical views. 78 2.7.5 Myxobolus spinacurvatura in lateral and transverse views. 79 2.7.6 Photomicrograph of A. solomoni. 80 2.7.7 Photomicrograph of A. serenum in lateral view. 80 2.7.8 Photomicrograph of A. serenum in apical view. 80 2.7.9 Photomicrograph of K. nova in apical and lateral view. 81 2.7.10 Photomicrograph of M. spinacurvatura. 81 2.7.11 Illustration of Gastrocotyle trachuri. 82 2.7.12 Photomicrograph of G. trachuri. 83 viii 2.7.13 Photomicrograph of Heteraxinoides atlanticus. 83 2.7.14 Illustration of Pseudaxine trachuri. 84 2.7.15 Photomicrograph of P. trachuri. 84 2.7.16 Photomicrograph of Paradiplectanotrema trachuri. 86 2.7.17a Photomicrograph of Derogenes varicus. 87 2.7.17b Illustration of D. varicus. 87 2.7.18 Illustration of Ectenurus lepidus. 88 2.7.19 Photomicrograph of E. lepidus. 89 2.7.20 Photomicrograph of Hemiuris communis. 90 2.7.21 Photomicrograph of Lasiotocus typicus. 91 2.7.22 Photomicrograph of Lecithocladium excisum. 92 2.7.23 Photomicrographs of Monascus filiformis. 93 2.7.24 Photomicrograph of Opechona bacillaris. 93 2.7.25 Illustration of Prodistomum polonii. 95 2.7.26 Photomicrograph of P. polonii. 96 2.7.27 Photomicrograph of Tergestia laticollis. 96 2.7.28 Photomicrograph of Corynosoma strumosum. 98 2.7.29 Scanning Electron Micrographs of Corynosoma wegeneri. 98 2.7.30 Photomicrograph of C. wegeneri. 99 2.7.31 Photomicrograph of Rhadinorhynchus cadenati. 99 2.7.32 Photomicrograph of Grillotia erinaceus. 102 2.7.33 Photomicrograph of Nybellinia lingualis. 103 2.7.34 Photomicrograph of Ceratothoa oestroides. 105 2.7.35 Photomicrograph of Anisakis sp.. 107 2.7.36 Photomicrograph of Hysterothylacium aduncum. 108 2.7.37 Photomicrograph of Pseudanisakis sp.. 110 2.8 Presence/absence matrix of species from horse mackerel samples. 130 Chapter 3 3.1 Current ICES stock boundaries for horse mackerel. 141 3.2 Site of HOMSIR sampling positions. 145 ix 3.3 Incisions made during horse mackerel dissection. 150 3.4 Length statistics for horse mackerel samples. 151 3.5 Age statistics for horse mackerel samples. 153 3.6 Prevalence of Goussia cruciata. 161 3.7 Relationship between G. cruciata prevalence and mean length. 162 3.8 Prevalence of Alataspora spp.. 164 3.9.1 Prevalence of Pseudaxine trachuri. 166 3.9.2 Mean abundance of P. trachuri. 166 3.10.1 Prevalence of Gastrocotyle trachuri. 167 3.10.2 Mean abundance of G. trachuri. 167 3.11 Relationship between length and infection with G. trachuri. 168 3.12 Prevalence of Paradiplectanotrema trachuri. 168 3.13 Prevalence of Tergestia laticollis. 169 3.14 Prevalence of Ectenurus lepidus. 170 3.15 Mean abundance of Anisakis spp. vs Hysterothylacium aduncum, 2000. 173 3.16 Mean abundance of Anisakis spp. vs H. aduncum, 2001. 174 3.17 Mean abundance of monogeneans in Atlantic samples. 175 3.18 Mean abundance of Anisakis spp. vs H. aduncum, Mediterranean Sea. 177 3.19 Relationship between mean intensity of Anisakis spp. and mean length in Mediterranean samples. 178 Chapter 4 4.1 A graphical representation of the distribution of horse mackerel stocks in the north east Atlantic. 190 4.2 Sigmoid function of neural networks. 195 4.3 Graphical representation of the structure of a neural network. 197 4.4 Data exploration and transformation. 203 4.5 Estimation of the optimum number of units in the hidden layer. 206 4.6 Median successful classification with discriminant analysis. 207 4.7 Graphical representation of discriminant analysis results. 208 4.8 Neural network classifications. 210 4.9 Norwegian sample neural network classification. 211 x 4.10 La Coruna sample neural network classification. 212 Chapter 5 5.1 The Atlantic herring, Clupea harengus L. 221 5.2 Representation of the distribution of Clupea harengus. 222 5.3 Catches of herring to the west of the British Isles, 1973-2003. 224 5.4 Distribution of gravel beds to the west of the British Isles. 225 5.5 Idealised sampling stations for the WESTHER project. 237 5.6 Realised position of spawning herring samples. 240 5.7 Realised sampling positions for juvenile herring samples 241 5.8 Realised sampling positions for mixed stock aggregation samples. 242 5.9 Relationship between anisakis infection and fish age. 246 5.10 Prevalence and mean intensity of Hemiuris luehei. 250 5.11 Plot of log transformed Cercaria doricha abundances against herring age. 253 5.12 Plot of log transformed C. pythionike abundances against herring age. 253 5.13 Temporal variation in prevalences of Anisakis simplex, C. pythionike and C. doricha. 261 5.14 Plerocercoid of Lacistorhynchus tenuis. 266 Chapter 6 6.1 Location of spawning sample sites from where Anisakis simplex specimens were extracted and sequenced. 280 6.2 Percentage frequency of shared haplotypes at six spawning sites. 284 6.2 Principal component analysis of Tamura & Nei genetic distance between anisakids from 6 sampling sites. 286 xi List of Tables Chapter 2 2.1 Distribution of parasites recorded by Gaevskaya & Kovaleva (1980). 64 2.2 Location, timing and sample size of horse mackerel examined for parasites at the University of Aberdeen. 66 Chapter 3 3.1 Location of horse mackerel sample collection sites. 146 3.2 Sample size and details of examiner. 148 3.3.1 Mean abundances of parasites recorded from samples collected in 2000. 155 3.3.2 Prevalences of parasites recorded from samples collected in 2000. 156 3.3.3 Mean abundances of parasites recorded from samples collected in 2001. 158 3.3.4 Prevalences of parasites recorded from samples collected in 2000. 159 Chapter 4 4.1 Location and size of samples collected in 2001 for stock identification and used in the neural network analysis. 200 4.2 Commonly encountered parasite species used for stock identification analysis. 201 Chapter 5 5.1 Parasites which are known to infect herring (MacKenzie, 1985). 228 5.2 Details of herring sample collection and sample size. 243 5.3 Numbers of herring viscera in each sample examined for parasites. 249 5.4 List of parasites infecting the viscera of herring caught west of the British Isles. 251 5.5 Comparison of prevalences (%) of Cercaria doricha and Cercaria pythionike at several sites between MacKenzie’s study (1985) and the present study. 255 5.6 Mean age of sample, prevalence and mean intensity of the four selected biological tag species from samples of spawning herring. 256 5.7 Prevalence and mean intensity of the four selected biological tag species from samples of juvenile herring. 256 5.8. Mean age of sample, prevalence and mean intensity of the four selected biological tag species from samples of non-spawning herring aggregations. 256 5.9 Prevalence and mean intensity of the four selected biological tag species from samples of spawning (X01) and juvenile (X02-04) herring outliers. 256 xii 5.10 Significance of differences in prevalence between equivalent samples taken in different years. 258 5.11 Significance of differences in prevalence of four tag parasites between juvenile samples 259 collected at the same sites in different years. 5.12 Significant differences in prevalence between mixed stock aggregations collected in different years at the same sites 262 5.13 Differences in prevalences of tag species between mixed stock samples and spawning samples collected in the same functional unit. 263 5.14 Median percentage classification of spawning herring to temporally pooled spawning samples using neural network, over 1000 repetitions of analysis. 265 Chapter 6 6.1 Sample sizes, haplotype diversity and nucleotide diversity at six sampling sites. 283 6.2 Results of AMOVA test. 285 xiii List of Appendices Appendix I Translation from Russian into English by the author “Ecogeographical features of the parasitofauna of the common Atlantic Horse Mackerel” (1980) by Gaevskaya & Kovaleva. Appendix II “The myxosporean parasitofauna of the Atlantic horse mackerel, Trachurus trachurus” published in Acta Parasitologica. Appendix III Horse mackerel discriminant analysis “R” code Appendix IV Horse mackerel neural network analysis “R” code. Appendix V “Spatial and temporal variation of parasitic helminth fauna and infracommunities of herring to the west of the British Isles: implications for fisheries management”. (publication in press). Appendix VI Haplotypes of Anisakis simplex s.s. found in herring to the west of the British Isles and in the Baltic Sea. Appendix VII Multidimensional scaling plot of Tamura & Nei distances between Anisakis haplotypes “R” code. Appendix VIII Abstract of presentation to International Council for the Exploration of the Seas Annual Science Conference, Aberdeen, September 2004. Appendix IX “Identification of fish stocks through neural network analysis of parasitological data” published in “Analysing Ecological Data” (eds. A.Zuur, E. Ieno & G.M. Smith) xiv Abstract The use of parasites as biological tags is a well used tool for fish stock identification. Recently, questions have been raised as to the appropriateness of stock boundaries in the Atlantic horse mackerel (Trachurus trachurus) and the herring (Clupea harengus) in European waters. Studies of the parasite fauna of these fishes could contribute useful information to resolve these issues. To this end, nineteen hundred and nineteen horse mackerel were examined from samples taken from up to nineteen sites around Europe in 2000 and 2001. These fish were subject to a complete parasitological examination. This resulted in the discovery of 45 species of parasites. Of these, 11 appear to be new host records, and one may be a new species. A number of these parasites may be suitable for use as biological tags and their application is discussed. Results showed a significant degree of separation was possible between fish from the North Sea and Western horse mackerel stocks, based on the relative abundance of the nematodes Anisakis spp. and Hysterothylacium sp.. Differentiation between Western and Southern stocks and Southern and African stocks is less clear. Results from the Mediterranean Sea suggest three stocks, confined to the western, central and eastern basins. These findings compare well with results from otolith and body shape morphometry, and a range of genetic markers, which were applied to the same fishes, confirming both the usefulness of investigating parasites as biological tags and the currently applied stock management system. A neural network approach to classification based on parasitological data was evaluated, and found to be successful in over 90% of cases. This compares favourably with the results obtained by more traditional discriminant analysis methods. This will be a useful tool for further parasitological studies. The endoparasitic fauna of four thousand and thirty-three herring at various life stages (juvenile, spawning and non-spawning) from sites to the west of the British Isles and outlier samples from the North Sea, Baltic Sea and northern Norway, were examined to obtain information on stock identity, mixing and recruitment patterns in these areas. Results revealed that substantial stock mixing takes place to the west of the British Isles, with fish from the Irish Sea, west of Ireland and west of Scotland being found together in non-spawning aggregations to the west of Scotland. This area is also home to two different spawning populations of fish which recruited from nursery grounds in the eastern North Sea as well as from the west of Scotland, at xv different times of year. The implications of these findings for fisheries management are discussed. The ribosomal small sub-unit cytochrome oxidase (I) gene of the parasitic nematode worm, Anisakis simplex s.s., was sequenced from specimens taken from a number of herring hosts at sites to the west of Scotland, the Irish Sea, south and northwest coasts of Ireland and the Baltic Sea, to investigate any link between parasite population genetics and host population structure. Results revealed the COI gene to be highly variable, with around 50% of worms sequenced having unique haplotypes. There were no population structures evident from principal component analysis of genetic distances. This lack of genetic difference, even between the isolated population from the Baltic Sea and the other sites suggests that genetic diversity of the parasite population is controlled to a much higher degree by the movements of the marine mammals which act as final hosts than by any degree of population structure present in herring. These results are discussed in light of the implications they have for fishery management. xvi Author’s Declaration This thesis was prepared whilst I worked full-time on the two EU funded multidisciplinary stock identification projects which it refers to, and as such, my work inevitably incorporates that of other people. Further details of the division of labour in terms of fish dissections are given in chapters 3 and 5. Fish ages were determined by unknown otolith readers at Bfa Fischerei, Hamburg, IEO Santander and the FRS Marine Laboratory, Aberdeen. Dr K MacKenzie identified many of the more obscure parasites and assisted with dissections. Some approaches to statistical analyses were suggested by our colleagues at meetings. Dr Catherine Collins and the good will of technical staff of the Molecular Genetics group at the FRS Marine Laboratory were crucial to obtaining and interpreting the Anisakis CO1 sequences. Other work, including the text contained herein, is my own. This work has not been accepted in any previous application for a degree. All quotations have been distinguished by quotation marks and the sources of information specifically acknowledged. Publications contained in the appendices are either entirely or substantially my own work. Signed: Neil Campbell Acknowledgements There are numerous people without whom this thesis would not have been written, and I extend my thanks to them all. People deserving of particular mentions are given here, in no particular order but starting at the beginning, and my utmost apologies to anyone I have forgotten. Thanks go to the staff of Dame Allan’s Schools, Newcastle upon Tyne, particularly Messrs. Wildsmith, Downey, Bisset, Blundred and Dr Gillis. At Aberdeen University I’d like to thank Graham Pierce, for getting me started on this road, and also Les Chappell and Les Noble, my official supervisors, for keeping me on the straight and narrow as regards university paperwork. Pablo Abaunza and Emma Hatfield coordinated the HOMSIR and WESTHER projects respectively, without whom this work would have never come together, and who I’m deeply grateful to. Numerous nameless individuals who collected the samples used in both of these projects deserve great credit for carrying out tedious work in often extreme conditions. Marcus Cross, Catherine Collins, Nessie Fraser, John Smith and Jimmy Chubb have all been, in their own ways, responsible for shaping the way I have approached this work, and I would like to thank all of them for their patience and friendship. Thanks also to Dr Sam Martin and Prof. Harford Williams, my examiners, for their constuctive comments. To my parents, Val and Angus Campbell, who made numerous sacrifices over the years to give me the opportunity to test myself, I am eternally grateful. Thanks also to Catherine, my wife, for love and understanding she has shown over the past five years while I have worked on this, without whom I am sure I would not have stayed the course. Some credit is also due to my son, Owen, for putting off his arrival for a few hours whist I attended my viva! Finally, I am deeply indebted to Dr Ken MacKenzie for taking a chance on me as a student, his careful supervision, introducing me to jazz, and, I am proud to say, his friendship. xvii Chapter 1. An introduction to stock identification. 1. Introduction 1.1. Introducing Fisheries The rational and scientific management of fisheries depends upon a fundamental understanding of all aspects of the biology of fishes (Pitcher & Hart, 1982). Global marine production from capture fisheries and aquaculture is currently around 140 million tonnes per year (FAO, 2003) (fig.1.1) – a five fold increase over the past fifty years. This is significant, both economically, with a first sale value of US$50 billion, and as a dietary source of protein in many parts of the world. In order to ensure the continued health and production of marine ecosystems, adequate management strategies must be devised which address the socioeconomic needs of fishers and the dietary requirements of the population which consume the fish, 150 while maintaining the ecological integrity of the stock. 100 50 0 Production (Million Tonnes) Total Production Capture Fisheries Aquaculutre 1950 1960 1970 1980 1990 2000 Fig.1.1. Global marine fish production for the period 1950-2003. Capture fisheries seem to have peaked at around 90 million tonnes per year in the late 1990s; however aquaculture production is still increasing rapidly (FAO, 2003). 2 2 1. Introduction 1.1.2 Why Manage Fisheries? Fish, as a renewable natural resource, have been generally considered unowned or at least common property, until captured. Any fish which is caught by a fisherman becomes unavailable to others. The fortunes of every fisher are therefore dependent on the actions of the fishing community as a whole. This interdependence between fishermen is inevitable and forms the basis for their collective responsibility for the health of the fish population, and its related ecosystem, as a whole. The concept of “freedom of the seas” has endured since the seventeenth century, when the Dutch merchant and politician, Hugo Grotius, defended Holland's trading in the Indian Ocean, with the argument of “mare librum”, based on the idea that fish stocks were so abundant that there could be no possible benefit obtained by claiming national jurisdiction over large areas of sea. His arguments prevailed, and freedom of the seas became synonymous with the freedom to fish (FAO, 1993). Countries such as Scotland had claimed exclusive rights to fishing in inshore waters as early as the fifteenth century, but there was no formal consensus as to how far off shore these areas extended. This ad hoc situation was codified by legislation in the 1930 Hague Convention on International Law. However, only Chile and Peru claimed more than a few miles of territorial waters. This system endured until the 1970s, when it became apparent to nations with large fishery resources that “their” stocks were being overexploited by non-local fishers. The 1973 UN Conference on Law of the Sea allowed 200 nautical mile national limits, which were immediately claimed by a number of countries. This caused much international friction, such as the “cod wars” between Iceland and the UK and Germany, who had operated distant water fleets in areas that were now Icelandic Exclusive Economic Zone (EEZ) 3 3 1. Introduction waters. At the same time, the degree of control over fishing effort which this legislation provided meant that for the first time, scientifically based, responsive management measures could be applied, rather than just studying a stock in order to attempt to predict future population size. Fisheries, therefore, are currently managed in order to prevent the undesirable outcomes of uncontrolled fishing occurring. An undesirable impact is defined by the stakeholder community which it affects. These consequences are variable, but include fishery collapse, economic inefficiency and loss of employment, loss of biodiversity, loss of amenity use and the destruction of important habitat sites. This, in turn, can lead to conflicts of interest between stakeholder groups. Clark (1985) divided these into four main areas of impact: biological, social, economic and political. He showed that a management measure can address two or more objectives simultaneously. Given the intensely political nature of fisheries management and the effort needed to take any management concept through to application into policy it is doubtful whether any management objective can be considered as having no political impact. In developed countries management strategies are usually imposed by an external regulator rather than arrived at by the fishers themselves. Contemporary fisheries management in developed countries aims to satisfy an increasingly diverse range of objectives through an equally diverse range of measures. Support from the fishing community and their compliance with the management measures taken are essential for positive outcomes of management measures. To this end, effective management can only be achieved by a straightforward and transparent decision making process supported by the highest quality of scientific information possible, in order to build confidence of the stakeholder communities in the decisions being 4 4 1. Introduction taken on their behalf. 1.1.3 The process of fisheries management Fisheries biologists contribute to the management of fisheries in two main ways. Firstly, through the study of the basic biological parameters of a fishery, such as distribution, abundance, growth rate and fecundity of the species being fished, and secondly through applying this biological knowledge to studies of the population dynamics of a species in order to make predictive models of how changes in fishing activity are likely to affect future populations. The forces acting on a simplified fish population were first identified by Russell (1931) (fig. 1.2). Figure 1.2. A simple representation of the processes acting on a fish stock. Biomass is increased through growth and reproduction (blue) and decreased by mortality, either natural or through fishing (red). Note that this model represents a “closed population” where there is no immigration or emigration. 5 5 1. Introduction This is summarised in the central axiom of fisheries science: B i+1 = Bi +(R+G)-(F+M) Where: B Biomass i A point in time R Recruitment G Growth F Mortality caused by fishing M Mortality caused by other sources One of the first things that is apparent from this simple model is that there are no parameters accounting for migration into or out of the stock. Most stock assessment methods model the dynamics of a closed population (Haddon, 2001) and assume homogeneity of life history for all individuals in the population (Walters & Martel, 2002). Misleading results can be produced if several closed populations or a portion of a population are the components modelled after a closed population has been assumed. Disregarding stock structure in a fishery leads to changes in biological attributes, productivity rates and genetic diversity, as well as over-fishing and depletion of less productive stocks (Begg et al., 1999). While it has been recognized that this assumption of a “closed population” is complicated by migrations and mixing between management units, disproportionately little attention had been paid to these complications, relative to other factors which affect the dynamics of a fished population (Stephenson, 1999). Whether the resource being managed consists of a single homogeneous population or two or more functionally discrete populations is one of the most 6 6 1. Introduction fundamental pieces of information available to a fishery manager. This knowledge allows decisions to be made on allocation of resources to fleets, distribution of quotas and policy on fisheries shared between two neighbouring countries. The process of stock identification deals with two principles; the concept of stock structure, and the methodology of identifying fish from different stocks. 7 7 1. Introduction 1.2. The Stock Concept All species have geographic limits to their distribution, which are determined by their tolerance to environmental conditions, and their ability to compete successfully with other species. In marine environments this may be less evident than on land because there are fewer topographical boundaries, however, discontinuities still exist, produced for example by mesoscale and sub-mesoscale circulations that minimize long-distance dispersal of fish larvae (Cowen et al., 2000). For fishes, it is rare for an individual to reproduce randomly with conspecifics within the entirety of this biological range (Pawson & Jennings, 1996). There is a tendency to form a structured series of discrete populations which have a degree of reproductive isolation from each other in space, in time, or in both. This isolation is reflected in the development between sub-populations of genetic differences, morphological variations in response to selective pressures and exposure to different chemical regimes and parasitic species. In theory, subpopulations respond to fishing in such a way that fishing on one population appears to have no effect on the population dynamics of a neighbouring population. 1.2.1 Definition of a “stock” The idea of a stock of fishes has been variously defined to suit the aims of the definer and the resolution of the methods being used to study them (e.g. Booke, 1981; Dizon et al., 1992; Carvalho & Hauser, 1994). Murray (1961, cited in Booke, 1999) demonstrates a range of meanings for the word, and shows that the meaning has changed over time, from the 14th century, when stock referred to an ancestor or “the source of a line of decent”, to the more recent past, when “stock” was used as a term for what we now think of as a species, to the present day, when a stock is a unit 8 1. Introduction that can be thought of as a sub-population, with some associated notion of reproductive isolation and genetic integrity. The currently accepted definition of a stock in fisheries science is that of Begg et al. (1999), “… [A “stock”] describes characteristics of semi-discrete groups of fish with some definable attributes which are of interest to fishery managers.” This definition may seem rather vague, but can be applied to the broad spectrum of methods used to identify stocks. Put more simply, this definition allows a stock to be considered as a population of interest, living in a definable area, and for which birth, growth and death are more significant to the population dynamics than immigration and emigration parameters, thereby conforming to the structure espoused in the central axiom of fisheries science (Russell, 1931). A requirement for fisheries scientists to report uncertainties in their assessment methods is found in article 7.5 of the FAO Code of Conduct of Responsible Fisheries (Anon., 1995), a related agreement to the United Nations Convention on the Laws of the Sea (1994), which the United Kingdom has ratified into law. Treating an assessment of a population and the consequent level of catches which such a population would support as an exact value would lead to management actions driven by methodical errors or the natural variability in population levels (Peterson et al., 2001). It is therefore necessary to look at the assumptions which population models make and attempt to quantify their inherent inaccuracies, such as estimates of mixing, or natural mortality. In order to prevent stocks being fished to extinction by the application of inappropriate fisheries management measures, knowledge of stock distribution and population dynamics is an essential first step in creating a sustainable plan for the exploitation of a species. Any study of the dynamics of a commercially exploited 9 1. Introduction population should begin by determining the stock identity and migratory dynamics of that population in relation to any neighbouring populations of the same species before any further decisions concerning the development of a fishery (Cushing, 1968; Secor, 1999). 1.2.2 History of Stock Identification The first record of the observation of differences in the biology of fishes from two different seas was by the Korean naturalist, Yak-Jun Jung (1816, cited in Secor, 1999), who found Pacific herring (Clupea pallasi) from the Yellow Sea possess fewer vertebrae than those from the Sea of Japan. Heinke (1898) is credited with the first modern use of the stock concept to refer to different groups of herring (Clupea harengus) in the North Sea, defined by small but measurable differences in morphology. In recent years, stock identification has been increasingly seen as an important base on which to carry out stock modelling. This is reflected in the increased number of peer-reviewed and other publications relating to stock identification. The Aquatic Science and Fisheries Abstract database shows around a three-fold increase in the number of stock identification publications over the past 20 years (see fig.1.3). Much of this increase has been driven by the increased ease and lowered cost of genetic studies, although these are frequently carried out in association with one or more other methods. 10 1. Introduction 140 120 Publications 100 80 60 40 20 2002 2000 1998 1996 1994 1992 1990 1988 1986 1984 1982 0 Year Figure 1.3. Numbers of publications per year, recorded in the ASFA database in which “fish” and “stock identification” are given as keywords, over the period 1982-2002. 1.3. Principles of Stock Identification Stock identification is an integral component of modern fisheries assessments, fisheries management and the management of endangered species (Begg et al., 1999). In order to define and separate stocks, characteristics which can distinguish one fish from one population from those from another need to be identified. Stock identification research can be classed as an observational study of a wild population, the main characteristic of sampling being that there is no opportunity to repeat the process under similar conditions. Approaches to stock identification range from simple counts of commercial catch to complex genetic assessments; there is no single correct method of determining stock distributions, each having its own set of advantages and 11 1. Introduction drawbacks. Multidisciplinary studies have the advantage over single method studies in that comparisons can be made between the results obtained from different techniques, which have been carried out on exactly the same fish. While multidisciplinary projects have been becoming more popular in recent years, often the analysis of results have been interpreted on a method by method basis, rather than a true interdisciplinary approach taking variables from all methods and analysing them in combination. Creating a meaningful combination of variables is a challenge, given that each variable addresses a slightly different aspect of the biology of the fish, for example, data on otolith microchemistry relates a fish to the chemical environment in which it developed as a larva, data on internal parasites is indicative of where a fish has been feeding and genetic indices are representative of spawning isolation over successive generations. Such multidisciplinary analyses may be of more use in assigning proportions of mixed fisheries to discrete or semidiscrete spawning populations than simple stock delineation studies. Multidisciplinary studies have been, or are currently being, applied to a range of problems in European fisheries management issues, such as confirming from which area fish have been caught in order to prosecute illegal fishers (e.g. the CODTRACE project 1 ), confirmation of stock boundaries (e.g. the HOMSIR project 2 ) and investigation of composition of mixed fisheries (e.g. the WESTHER project 3 ). 1 http://www.ucd.ie/codtrace/index.htm http://www.homsir.com 3 http://www.clupea.net/westher 2 12 1. Introduction 1.3.1 Stock Identification Techniques The field of stock identification methodology is undergoing rapid change due to the possibilities presented by genetic advances, increased processing power of computer and miniaturisation of electronic components in data storage tags (Sissenwine, 2005). This does not, however, render existing methods outdated, but adds tools which can be called upon to answer particular questions. Stock identification techniques can be classified into a number of complimentary categories: o distribution of fish o distribution of eggs and larvae o variation in morphological and meristic characteristics o variation in life-history traits o variation in chemical composition o genetic differences between populations o artificial tagging o natural tags The application of these techniques, and the life-history stage to which they are relevant will be addressed in the following sections. Multidisciplinary studies applying complex multivariate analysis methods need to carefully choose ranges of methods and variables which provide complimentary information. 13 1. Introduction 1.3.1.1 Distribution and abundance of fish, eggs and larvae One of the simplest and most frequently recorded pieces of data in a fishery is where, when and what quantity of a species of fish were caught by a particular vessel. This data is collected routinely to assess the status of fisheries and to check compliance with fishing regulations. Integrating this data from the whole fleet over time allows a seasonal picture of population distribution and migration to be developed. If combined with a measure of the effort used to catch this quantity of fish a crude method of abundance can be obtained. Fig. 1.4a (left) and 1.4b (right). Catches of horse mackerel in the North East Atlantic during the first and third quarters of 2001. After Abaunza et al., (2004). An example of this can be seen in seasonal changes in the fishery for horse mackerel in the North Atlantic (Abaunza et al., 2004). Fig 1.4a shows catches of horse mackerel in each ICES statistical rectangle (an area of water 30 nautical miles 14 1. Introduction by 30 nautical miles) taken in the first quarter of 2001. Fig 1.4b shows the equivalent data from the third quarter. In fig. 1.4a, catches are concentrated along the edge of the Atlantic shelf, while in fig. 1.4b catches are much more dispersed, with significant quantities of fish being taken in the North Sea and around the southern coasts of Norway. Note that there is very little seasonal change in the fishery around the Iberian Peninsula and Bay of Biscay. While this data provides information on the distribution of horse mackerel, it tells us nothing about the migratory routes which are taken, or the mixing of the population within these areas. The seasonal progression of landing in a fishery has been used to track general migrations of mackerel (Lockwood, 1988) and herring (Ford, 1933), and is useful in the initial recognition and development of hypotheses on stocks. A drawback with using fishery data to define stocks is that the values it provides are a reflection of human behaviour, rather than the true distribution of fishes. One particular difficulty is problem of misreporting of catches. This occurs when the skipper of fishing boat intentionally records misleading position values in their logbooks, usually in an attempt to disguise illegal fishing activity. This introduces a source of uncertainty into the data which is difficult to quantify. Another confounding factor is changing behaviour of fishers, for example, an increase in fuel costs, advances in fishing technology or a drop in fish prices, which can influence the picture which is obtained by this method, and not represent a true change in stock distribution (Greenstreet et al., 1999). Catch data can be reinforced by fishery-independent surveys carried out by research vessels, sampling fish populations with standardised methods at random positions to estimate the population size in an area. A drawback to this method is that stock identification is rarely the main aim of these distribution and abundance 15 1. Introduction studies; surveys are too infrequent or geographically imprecise to provide good evidence of stock separation, and can provide no evidence at all in areas where two stocks overlap (Jennings et al., 2001). Methods such as acoustic surveys or, in blue waters for suitable species, aerial visual census, also generate information about the distribution of a fish species (fig. 1.5). Again, this technique can be used to investigate stock identity, but it does not provide information about population composition in areas where two stocks overlap. 62 61 60 59 58 57 56 55 54 53 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 Figure 1.5 Distribution of Herring in the North Sea and west of Scotland, as revealed from the 2002 acoustic survey of the North Sea and west of Scotland. This figure supports the idea of a centre of abundance to the east of the Shetland Islands, with a number of small populations to the north and west of Scotland. With the growth in methods available to fisheries science, all life stages of fish have been investigated to aid stock identification. Eggs can be sorted from plankton samples to identify spawning grounds and seasons of different stocks (Thompson & Harrop, 1987). Alternatively, larvae can be collected from plankton 16 1. Introduction samples to investigate movement from a spawning area to a nursery ground (Jennings & Pawson, 1992). Areas may be separated by wide or deep expanses of water which adult fish would not cross, but may be connected by dispersal of planktonic larvae. An example of this is the Blue stripe snapper, Lutjanus kasmira, which was deliberately introduced to Hawaii in the early 1970s to encourage development of a fishery. During the following ten years this species spread throughout all the islands of the archipelago (Oda & Parrish, 1981). Despite evidence from physical tagging revealing that although adult fish are capable of moving long distances between patches of reefs, they are non-migratory between islands (Freidlander et al., 2002); these populations can be considered to represent a single stock, due to their larval connectedness. Egg and larval surveys are useful in stock identification studies as the definition of a stock includes an idea of reproductive isolation, even if stocks are mixed at later life stages. Evidence for numerous separate spawning stocks of herring within the overall northeast and northwest Atlantic populations has been provided by larval surveys (Blaxter & Hunter, 1982). A drawback to this method is the huge uncertainty inherent in larval survival and drift. The presence of eggs or larvae in an area can be taken as an indicator that an adult population has spawned nearby, but is no guarantee that any fish will recruit from that area in the future. 1.3.1.2 Morphological or meristic variations Analysis of the shape of fish or details of their bones or organs is one of the 17 1. Introduction oldest means of identifying stocks of fish (Heinke, 1898). The basis of this approach is the fact that evolutionary changes in reproductively discrete populations or the effects of different environmental regimes as a selective pressure, or on larval development, will result in changes in body morphology or meristic characteristics which can be measured and compared between populations. Early morphological studies involved univariate analysis, often removing the flesh of fish and measuring the length of various bones, usually those which comprise the skull. Later studies looked at bivariate analysis of relative growth, in order to investigate ontogenetic changes. The development of computer-aided image analysis and digital photography has greatly facilitated progress in morphometric studies (MacLeod, 1990) and even three dimensional images can now be captured and analysed (Dean, 1996). 1.3.1.2a Meristic differences Differences in enumerable meristic characteristics, such as counts of gill rakers, fin rays or pyloric caeca, are the simplest measures of morphological variation which can be collected. Differences between stocks represent genetic variation between stock, environmental influence on development, or a combination of the two. Although the technique involves measuring very simple variables, complex statistical techniques can be applied to these data to investigate stock identity (Groeger, 2000). This method has been applied to a wide range of fish species (e.g. Hotta et al., 1999; Flores, 2001, Smith et al., 2002) from both larval and adult life stages (Miller, 2002). This technique is cheap to implement due to the ease of routine collection of samples at fish markets or at sea. 18 1. Introduction 1.3.1.2b Analysis of body morphometry The body shape of a fish is the result of the interaction between genetics and the environment (Swain and Foote, 1999), especially during early life stages. Therefore, morphological changes during growth, that vary geographically, may be useful for stock discrimination (Cadrin, 2000). Phenotypic characters can be considered to be as valid as or even more valid than genetic ones for the identification of stocks (Swain and Foote, 1999). While just a small level of gene flow between two different stocks may prevent the detection of genetic differences, a consistent difference in morphological variables may exist, due to the influence of different selective pressures of the environment which the stock inhabits, indicating the existence of different fisheries management units. As morphological differences can be more dependent on environmental than genetic factors, substantial mixing of fish from separate stocks can be hidden by the noise of the system. Morphometric studies have been revolutionised by the advent of digital photography and cheap computers (Cadrin, 2005). Traditional morphometric studies involve making careful measurements of consistent parts of the fish, such as ocular diameter or length of the jawbone using callipers, and then comparing variables between stocks on a univariate or multivariate basis. A more modern approach involves digitising an image of the fish using a number of landmark features on the body, such as the position of the pelvic fin or hindmost edge of the eye. Landmark features on the outline of the fish are used to create a network of distances between all points. These are analysed by principal component or discriminant function analysis, and different groups identified. This method generates large quantities of data, for instance, a 12 point network involves 77 different distances, so clearly this 19 1. Introduction approach is computer intensive. Due to the limitations of multivariate analyses, as the number of variables included in the study increases, so must the number of fish in each sample. To limit the number of variables included, various derived measures have been proposed, such as box-truss or triangle-truss networks, which operate on trapezoidal or triangular cells, formed from adjacent landmarks (Straus and Bookstein, 1982; Bookstein, 1991). 1.3.1.2c Analysis of otolith morphometry The study of otolith shape variation has advanced from measuring simple linear distances or otolith masses to deriving complex geometric variables from otolith outlines (Cadrin & Friedland, 2005). Otoliths are now routinely photographed digitally and prepared through the application of an image processing algorithm. The mathematical formulae which describe this outline becomes the means to distinguish between stocks. 20 1. Introduction Figure 1.6. Representation of methods used to derive data from otolith shape. Otolith is photographed, the outline extracted and radial distances calculated from either the geometric or physical centroid of the otolith. Distances can be plotted as a function of radial angle, θ. Outline methods routinely transform the otolith shape into a series of radial distances, spaced by an equally stepped angle of rotation. This allows the otolith to be “unrolled” and considered by a variety of statistical methods (Fig. 1.6). The most popular methodology currently applied in stock discrimination studies is Fourier analysis, which attempts to find harmonic frequencies of the radial angle, θ, which describe the radial distances. Fourier analysis of otolith morphometry has been recently successful in discriminating stocks of king mackerel around Florida, and redfish (Sebastes mentella) in the North Atlantic (De Vries et al., 2002; Stransky, 2004). Although all otolith bones have been investigated to determine stock identity, 21 1. Introduction Campana and Casselman (1993) demonstrated that the saggital otoliths are the most useful. Little variation has been found between the right and left saggita within an individual fish (Castonguay et al., 1991). While this approach has been most successful when applied to otolith morphometry, some workers have enjoyed success when applying it to analysis of other hard structures, for instance with scales of striped bass (Ross & Pickard, 1990) or analysis of mussel shell morphology (Ferson et al., 1985). 22 1. Introduction 1.3.1.3 Variations in life history traits Life history parameters are the consequences of life history strategies to which a population of fish have evolved, reflecting the underlying dynamics of the fish stock (Begg, 2005). Typically, data on vital population or life history parameters are collected as part of routine baseline fishery surveys. While this is not collected with the purpose of discriminating stocks, if collected with an appropriate spatial resolution this data can provide an initial indication of stock boundaries. 1.3.1.3a Growth rate parameters Growth can be defined as the balance between the inputs provided by the consumption of food, and the outputs represented by energetic losses associated with metabolic maintenance (Wooton, 1998). This equation could be expressed in terms of the energy budget, but most studies of growth in fishes deal with the length or the weight of the fish at particular ages (Ricker, 1979). The Von Bertalanffy Growth Function (VBGF) has been by far the most studied and most used of all length-age models in fish biology (Quinn and Deriso, 1999). In the analysis of the VBGF parameter estimates obtained by different authors, it is important to take into consideration sampling design, material used to infer the age of the fish, its preparation, the age-reading method employed and whether it has been validated for the species in question (Abaunza et al., 2003). This means that comparisons between different growth rate studies are difficult to perform. Differences in growth rate have been used to discriminate between a range of stocks of demersal and small pelagic fishes (Berg & Albert, 2003; Ojaveer et al., 2004). 1.3.1.3b Reproductive parameters 23 1. Introduction Lengths at maturity, age at maturity or oocyte density are commonly recorded variables in a wide range of fisheries science studies, particularly those looking at the reproductive potential of a population. One of the key factors in defining a biological stock is a notion of reproductive isolation and selfsustainability (Ihssen et al., 1981). Knowledge of the temporal and spatial extent of spawning can provide information on intraspecific variation in life history parameters that can be used to discriminate separate stocks (Schaefer, 1987). Differences in timing and location of spawning provide particularly valuable information because they can result in reproductive isolation among stocks by restricting gene flow to a level that effectively isolates a spawning stock (Bailey et al., 1999). They have been used widely to discriminate between populations, and also to link particular cohorts of recruits to particular spawning populations, in areas where stocks spawning at different times of the year co-occur. Several genetically distinct stocks of herring (Clupea harengus) in the northwest Atlantic Ocean have been determined on the basis of the number of geographically stable spawning and nursery grounds (Iles & Sinclair, 1982; Stephenson, 1991). 24 1. Introduction 1.3.1.4 Artificial Tags Artificial tags are one of the most commonly used techniques for stock identification, and, in terms of the data generated, one of the most trusted by fishery managers. Tag-recapture methods have been used to determine stock identity for over seventy years (Needler, 1930, Schroeder, 1942). This technique involves applying a tag, tattoo or otherwise identifiable mark to individuals, then releasing them back into the population. In this way, it is similar to the “capture-markrecapture” (CMR) technique common in terrestrial ecology. It differs from this technique in terms of the data generated. CMR studies focus on estimating population size, dynamics or the effects of changing environmental conditions (e.g. Lima et al., 2003; Yloenen et al., 2003). In fisheries, these parameters are estimated mainly from catch data, such as recorded landings or catches, estimating population indirectly by calculations on catch per unit effort. If tagging and recapture are separated by a significant amount of time, which can be months or even years, conventional tagging can provide information on stocks identity, movements, migration rates and routes, abundance, growth and mortality (Metcalfe et al., 2002). Investigations on mackerel are considered to be a good example of how tagging can be used to obtain necessary information for management purposes (Uriarte et al., 2001; Uriarte and Lucio, 2001). There are a number of types of artificial tags, with their own advantages and drawbacks. The most commonly used tag is the “T-bar” design. It consists of a stout piece of plastic as the crossbar of the T, with a longer, flexible piece of plastic attached to this. This is usually highly coloured. The crossbar is embedded in the flesh or body cavity, while the flexible tag is left to protrude from the body. Tagging is a relatively straightforward procedure, which allows a large number of fish to be 25 1. Introduction tagged in a short time. This procedure is obviously traumatic to the fish, and survival is variable. In some studies, little reference is made to post tagging mortality or the changes which they can induce in behaviour. Physical tag studies have been limited in the areas to which they can be applied by the need to obtain live fish and return them to the water relatively unharmed. The difficulty of collecting deepwater fishes makes the use of physical tags impossible, as the fish are usually dead before they make it to the surface. Recently, automated tagging machines, mounted in the belly of pelagic or demersal trawls have had some success with in situ tagging of redfish (Sebastes sp.) in the North Atlantic (Jacobsen & Hansen, 2005). The advantage of this approach is that it provides direct evidence of fish movements. Fish are able to be individually identified from a serial number on the tag, which can relate them to the location and date where they were tagged, the location and date where they were recaptured and the distance and time between these two points. One of the drawbacks of this approach is that large numbers of the species being studied need to be tagged to ensure a statistically meaningful rate of return of recaptured tags. This can be achieved through the mechanisation of the screening process. A number of tagging programs have used individually numbered magnetic tags inserted into herring. Large electromagnets over conveyor belts at herring processing facilities have been used to screen thousands of fish per hour and extract tagged individuals for further examination (Schweigert et al., 2001). Physical tags are also limited in the data which they generate – they are generally not informative about the movements which a fish makes other than that they have moved from the point of release to the point of capture in a specified time – any other migratory 26 1. Introduction information is not available. Data storage tags which are capable of recording extra information such as salinity, depth or temperature, are prohibitively expensive to use for large scale stock identification tagging studies. The tagging process needs to be carefully managed. The tags themselves need to be recognizable to the final captor. There are several ways of achieving this, either the tag is applied to the outside of the fish, or the tag is inserted into the fish, and it is marked in some way to make it recognizable (Brennan et al., 2001). Either method results in a change in the energetics of the tagged fish, and alters the mortality rates of the tagged portion of the population. This needs to be measured by maintaining a population of tagged fish in a captive situation to quantify the changes in mortality. 1.3.1.5 Variation in chemical composition Fish spend their whole lives in a medium which is chemically complex and variable. This variability is reflected in the levels of numerous elements incorporated into their calcified structures. As early as 1967, preliminary studies had suggested that quantitative analysis of trace elements in otoliths, vertebrae and scales could provide information on population structure and movements of individual fish in a population. (Anon., 1967, cited in Mulligan et al., 1983). To date, 37 elements have been identified as varying in concentration in calcified structures between fish stocks or sample sites (Thresher, 1999). However, most of these were reported from a single study (Campana & Gagne, 1995), which reported 28 elements differing between otoliths from different cod stocks. Most studies report reliable variations in between five and ten elements. Thresher (1999) reviewed the use of otolith microchemistry in stock 27 1. Introduction identification and found that this method was applied mainly to three sets of problems: - discriminating between marine and freshwater populations - determining links between natal rivers or nursery areas and adult stocks - assessing population structure and migration in marine fishes. Chemical composition need not necessarily utilize variations in naturally occurring marker elements. Chou et al. measured the contamination of lobsters (Homarus americanus) with heavy metals to trace migrations of the lobster population in the Bay of Fundy, Canada (2002). During the winter they examined lobsters from three sites which were known to be contaminated with heavy metals (Ag, Cd, Cu, Mn and Zn) and modelled the effects of further contamination, and ability to excrete heavy metals from the body, then analysed samples taken from fisheries at the same sites in the summer months. They found a high level of mixing between two of the three sites, and overall movement of lobsters into the inner reaches of the Bay of Fundy. Variation in naturally occurring biochemicals has also been used for stock delineation. The fatty acid profile of fishes is highly variable, with differences evident between individuals, ages, maturities, and tissues sampled (Grahl-Nielsen, 2005). Diet is the most important factor influencing the composition of tissue fatty acids, and differences in diet between stocks the underlying force behind the utility of this technique. Principal component analysis of the entire fatty acid profile of fish has been a useful technique in distinguishing between North Sea and AtlantoScandian herring (Grahl-Nielsen & Ulvund, 1990). 28 1. Introduction 1.3.1.6 Genetic differences between populations The concept of stock can be considered from two genetic perspectives: the “operational” approach, whereby the stock is a population of interest to fishery biologists, with definable characteristics, and the “genetic” approach where a stock is a discrete breeding population which is genetically distinct from neighbouring stocks of the same species (Gulland, 1971; Smith et al., 1990; Carvalho and Hauser, 1994). Genetic variation between stocks can provide a direct basis for stock structure, but can prove inadequate in situations where there even low levels of mixing between two stocks (Begg et al., 1999). Recently some emphasis has been given to stock as a genetic and evolutionary unit, identified by the genetic stock assessment technique (GSA) (Utter, 1991). In spite of the fact that GSA has been shown to be essential for the conservation and rational management of fisheries resources, genetic structure of exploited fish stock units has been ignored until recently, when the use of molecular genetic approaches (e.g. multilocus allozyme electrophoresis, microsatellite DNA, mitochondrial DNA sequencing, RAPDs, etc.) have provided genetic markers which are cheap to test for and reliable in terms of the results which they can provide to fish stock assessment (Ward, 2000). Marine fishes in general tend to show little genetic subdivision between geographically separated populations (Gyllensten 1985; Ward et al. 1994). Gene flow, mediated by the passive transport of pelagic eggs or larvae or by adult migration over large distances, contributes to maintaining the genetic homogeneity of populations at a broad geographic scale. The slow pace of genetic drift in 29 1. Introduction populations of large and stable effective size may also contribute to the maintenance of genetic homogeneity, even in the extreme case of geographic isolation - provided this is recent enough. The earliest attempts to identify stocks based on genetic differences were carried out in the 1930s and tried to separate herring stocks based on differences in blood group frequencies (de Ligny, 1969). These studies demonstrated the utility of a genetic approach to stock identification, but this serological method failed to be widely adopted and is rarely applied today. 1.3.1.6a Multilocus Allozyme Electrophoresis (MAE) One of the oldest “genetic stock identification” techniques, Multilocus Allozyme Electrophoresis (MAE) involves extracting common “housekeeping” enzymes from fish tissue, then running this extract through a starch gel in order to identify variable gene loci through the proteins which they express. Inheritance studies have shown allozymes to follow Mendellian inheritance patterns, with fish receiving one copy of the gene in question from each parent. This means the genetic structure of a population is relatively stable over generations (Kornfield et al., 1981). The proportion of each haplotype occurring in a given population can be measured and compared to identify genetically discrete populations. MAE is a methodology which is able to provide a large number of genetic markers and has been shown to provide reliable results in stock assessment of demersal fishes (Rossi et al., 1998). With respect to many other genetic approaches multilocus electrophoresis is easier to set up and able to provide reliable results which can be easily compared with a large volume of data in the literature (Carvalho & Pitcher, 1995). Use of 30 1. Introduction MAE allows the analysis of a large number of specimens in a relatively short time, providing reliable results on the genetic partitioning within the range of a species distribution, degree of gene flow between each subdivision and the levels of genetic variability which are observed. As one of the older genetic techniques, it has been applied in a wide range of areas, but has been particularly useful in determining stock identity in mixed stock fisheries of Pacific salmonids (Beacham et al., 1985, Grant et al., 1980, Seeb & Crane, 1999). 1.3.1.6b Mitochondrial DNA Sequence Data Mitochondrial DNA (mtDNA) is a small circular piece of DNA found within the mitochondria of the cell, and which codes for products which are vital for respiration. Because of its ease of isolation, absence of recombinations, taxonomic homology and its inheritance through the maternal line, the mtDNA genome is an excellent system for population analysis (Levin, 2000). Mitochondrial DNA appears to evolve significantly more rapidly than nuclear DNA. Non-transcribed sites of mitochondrial protein-coding genes and the non-transcribed control region have been shown to be particularly useful for analyzing relationships of recently diverged taxa, such as among populations (Levin, 2000). Direct measurement of mtDNA sequence variation has become feasible by using polymerase chain reaction (PCR) amplification. Specific gene segments can be amplified via PCR by using a pair of primers in quantities sufficient for direct sequencing (Meyer 1994; Hillis et al. 1996; Palumbi 1996). This technique can detect single base pair changes and, therefore, provides a large number of characters for comparative purposes. These features 31 1. Introduction mean this technique is applicable to acute taxonomic and stock identification problems (Kocher & Stepien, 1997). This technique can be applied to the identification of stocks, and also used to estimate levels of gene flow between populations, the effective size of populations, patterns of historical biogeography and phylogenetic relationships (Carvalho & Pitcher 1994; Kocher & Stepien 1997). Usually, marine fishes display weak phylogeographic structuring, and intraspecific mitochondrial DNA clades are shallow compared with freshwater fishes (Shulman & Bermingham 1995; Avise et al. 1998; Avise 2000). However, intraspecific genetic differentiation using mitochondrial DNA have been found in several marine fishes, mainly coral-reef fishes, near shore fishes, and fish species distributed both sides of 'invisible barriers' or oceanographic fronts (Grant & Bowen 1997; Avise 2000), including pelagic and migratory fishes (Alvarado-Bremer et al., 1996; Naciri et al., 1999; Nesbo et al 2000;). 1.3.1.6c Microsatellite DNA The term microsatellite DNA refers to regions of genomic DNA consisting of tandem repeats of units of 2-5 bp. The total length of microsatellite arrays is usually less than 300 bp. The microsatellite loci are usually classified in three categories: perfect, when they contain an uninterrupted stretch of repeat units; imperfect when they contain one or more bases interrupting the repeat units; and compound when they contain repeats of different types, e.g. (CT)n and (GT)n (Levin, 2000). The use of microsatellite DNA is found in increasingly diverse applications as genetic markers for intraspecific population structure studies (e.g. Bentzen et al., 1996; Ruzzante et al., 1997, Rico et al., 1997, O’Connel et al., 1998). They are 32 1. Introduction highly polymorphic, because of their increased mutation rate; they can be relatively easily “scored” (the number of repeat units measured) through PCR amplification and electrophoresis; they can be scored from small amounts of tissue or partially degraded DNA (Levin, 2000). The use of microsatellite DNA markers in marine species can uncover subtle population genetic structuring, even in cases where other types of markers failed to do so (Shaw et al., 1999). The critical step in using microsatellite DNA markers for a population study is the development of appropriate primers for PCR amplification for the species being considered. This is done by sequencing the regions flanking the microsatellite core, which are usually evolutionary conserved at a species level, and designing primers of between 15-20 bp long. The development of primers is relatively difficult, but primers which have been designed for a given species can work also in closely related species, which simplifies the process. Once the appropriate primers have been developed and by using modern automated equipment, hundreds of individuals can be scored within a day. 1.3.1.6d Single Strand Conformation Polymorphism (SSCP) The different levels of mobility of the sense and anti-sense strands of DNA, when denatured and run single stranded in an acrylamide gel, has been described for some time. With the advent of PCR, acrylamide gels were found to be capable of producing differential mobility in short, single stranded, PCR fragments differing only by single base substitutions. This method is called Single Strand Conformation Polymorphism (SSCP) analysis (Orita, 1989). This technique was quickly applied to 33 1. Introduction the task of detecting unknown point mutations. The electrophoretic mobility of ssDNA is dependent on several factors. Firstly, electrostatic charge of the DNA strand, which remains constant in molecules of the same size and chemical composition. Secondly, and more importantly, is the shape or configuration which the ssDNA fragment adopts. The ssDNA is free to form weak intramolecular associations (“secondary structure”) between regions where the sequence is selfcomplimentary. The configuration or shape adopted by the ssDNA, and hence the electrophoretic mobility, is highly variable and sequence dependent. Any sequence change modifying regions of self complementarities is likely to alter the configuration and produce a detectable shift in the mobility of the DNA strand on the gel (Rhebein et al., 1999; Sunnucks et al., 2000). Advantages associated with this technique are relatively low cost, compared to other genetic methods, the simplicity of the procedure, which allows a good degree of reproducibility, the speed at which it can be applied, and the fact that, theoretically, this technique can detect single base substitutions (Sunnucks, 2000). 34 1. Introduction 1.3.1.7 Parasites as biological tags An important applied aspect of marine parasitology is the use of parasites as biological tags to investigate the population structure of marine organisms (MacKenzie & Abaunza, 2005). The basic principle behind this method is that the host can become infected with a particular parasite species only within the endemic area of that parasite, namely, the area where environmental and ecological conditions are suitable for the completion of the life cycle of the parasite. If a host is found infected with a parasite, outside that species’ endemic area, it can be inferred that the host has been within the endemic area at some stage of its life. The use of parasites as biological tags is a long established technique for discriminating between fish stocks. It was first applied by Herrington et al., (1939) to stocks of redfish off the coast of the United States of America. Although the technique was developed for use in fish, it is just as applicable to invertebrates (e.g. lobsters - Uzmann, 1970; prawns - Owens, 1985; cephalopods - Pascual & Hochberg, 1996). In a major review of the subject, MacKenzie (1987) defined three types of study for which parasites are informative: - Stock separation studies which aim to identify intraspecific groups within a host population, distinguished by different patterns of behaviour at certain stages of their life history. - Recruitment studies which follow the migrations of juveniles when they leave their nursery grounds to join adult stocks. - Seasonal migration studies which aim to follow different populations of fish to and from spawning or feeding grounds. MacKenzie (1987) and MacKenzie & Abaunza (2005) also proposed criteria 35 1. Introduction to assess the probable value of a parasite as a biological tag for any type of fish population study. Developments in analytical methods mean that these criteria should be taken as guidelines, rather than hard and fast rules. o The parasite should have significantly different levels of infection in the host in different parts of the study area o No ectoparasite which is easily detached, leaving no evidence of its presence on the host should be considered, as these can easily be lost in capture or transport. o The method of examination should involve a minimum of dissection. A high degree of site specificity on the part of the parasite is an advantage to avoid time limiting the number of specimens examined. o The parasite should be easily detectable and identifiable, otherwise the chance of missing an infection is too great. o The parasite should have a lifespan (or remain in an identifiable form inside the host) for a period of time which is relevant to the aims of the investigation. o The parasite should have no marked pathological effects on the host. A highly pathogenic parasite may cause selective mortalities, or behavioural changes which 36 1. Introduction reduce its value as a tag. o Parasites with direct, single-host life cycles are simplest to use; those with complex life cycles with numerous stages in different hosts are more of a challenge as knowledge of the distribution of all potential hosts is required to interpret the information which is obtained. o There should be no significant variations in levels of infection from year to year or season to season. The effects of interannual variation can be minimised through examination of single cohorts. A methodology for use in studies of parasites as biological tags for stock discrimination studies was published by MacKenzie and Abaunza (1998, 2005). They defined two main approaches to parasitological tag studies. Firstly, selecting a small number of parasites which meet the defined criteria for use as tags, and examining a large number of hosts for their presence. The advantage of this approach is the statistical simplicity and clarity of the output of these studies. Treating the variation in each suitable species of parasite as an indicator of host population biology allows a complex picture of stock identity to be built up. The drawbacks to this approach are the preliminary work which needs to be carried out in order to show that a parasite species is a suitable tag, the large numbers of hosts which need to be examined in order to statistically test a hypothesis of stock identity, and the inherent uncertainty of basing stock identities on different levels of a single or small group of variables. Nevertheless, this is the most widely applied 37 1. Introduction form of “parasites as tags” study. It has been applied to stocks of small pelagic fishes (e.g. Moser & Hsieh, 1992; Mogrovejo & Santos, 2002) demersal fishes (eg.Oliva & Ballón, 2002; Larsen et al., 1997) and deep water fishes (Szuks, 1980; Zubchenko, 1985). The second approach involves analysis of the entire parasite assemblage using more sophisticated statistical techniques such as discriminant analysis. This approach is particularly useful for studying large, valuable species, where availability of samples may be limited. Two examples of the use of this approach are the studies by Lester et al., (2001) on Spanish mackerel (Scomberomorus commersoni) around the coast of Australia, and of Lester et al., (1985) on Skipjack Tuna (Katsuwonis pelamydis) from the western Pacific Ocean. Both studied large species of extremely high commercial value, and obtaining large numbers of fish to examine would have been prohibitively expensive. Where suitable parasites are available this can be a very powerful technique for resolving stock structure. Successful classification using this technique has been reported as averaging around 80%. Power et al. (2005) applied a range of parametric (linear and quadratic discriminant analysis) and non-parametric (k-nearest neighbour joining tree and feed forward neural network) statistical classification techniques to the same set of data on parasite fauna of the bogue (Boops boops) from around the coast of the Iberian peninsula. They found high rates of successful classification with all of these methods, with k-nearest neighbour joining being the most successful, with up to 94% of fish successfully reclassified. MacKenzie et al. (2005) proposed specific guidelines for the use of parasites as tags in studies of small pelagic fishes. Small pelagic fishes are generally infected with fewer parasite species than demersal fish, and within the group, piscivorous 38 1. Introduction species and those with relatively diverse diets have richer parasite faunas than planktophagous species (Polyanski, 1961). The parasite fauna of small pelagic fishes tends to be dominated by larval forms, particularly helminths, which mature in the predatory species that prey on them (MacKenzie et al., 2005). Parasites have been used as tags for population studies in a range of taxa, other than fish. Owens (1983, 1985) and Thompson and Margolis (1987) used metacestodes and digenean metacercariae to distinguish between stocks of commercially important crustaceans. Pascual and Hochberg (1996) reviewed the specific features of biological tag studies in cephalopods. Their trophic position has led the authors to recommend the use of genetic studies of anisakid nematodes (Pascual, et al., 1996). Parasites have also been used as biological tags of marine mammals, as reviewed by Balbuena et al. (1995). In the few studies carried out to date, acanthocephalans have been shown to be particularly useful. The practical application of parasites as biological tags was verified by Mosquera et al., (2000). They used a simple model of parasite-host interaction to evaluate the potential usefulness of tag parasites. They demonstrated mathematically that the parasite community of a migratory stock, moving through an area which is home to a non-migratory stock of the same species would take a considerable period of time to come to equilibrium with the parasite fauna of that stock. Their model is limited by the fact that they assume a finite parasitic lifespan, and parasite induced mortality of heavily infected hosts. These conditions represent parasites which fishery parasitologists would avoid as tags under most circumstances due to difficulties in quantifying these parameters in the real world. The information which can be obtained from parasite studies is determined by the choice of parasite, for example if infection occurs through feeding, 39 1. Introduction differences in infection rates reflect different feeding locations, if infection occurs through the skin of juveniles, infection data reflects different nursery areas, and so on. It becomes apparent that the question which a study is addressing needs to be carefully framed before selection of parasites to be examined takes place. 1.3.1.7a Parasite genetics Enzyme electrophoresis of anisakid nematodes was first proposed as an indicator of fish populations over 25 years ago (Beverley-Burton, 1978). It has been recently used to identify, to species level, larval stages of anisakid nematodes that lack morphological characters adequate for identification (Nascetti et al., 1986; Mattiucci et al., 1997). These species have different life cycles, and consequently different geographic distributions. Their specific identification, in conjunction with a larger “biological tag” study, can greatly enhance the usefulness of anisakids for determining stock structure (Mattiucci et al., 1997). 40 1. Introduction 1.3.2 Multivariate statistical methods of stock identification Typically, these stock identification methods generate a number of variables relating to each individual. The statistical power of any study is improved by the consideration of these variables in a multidisciplinary method, rather than treating each in a univariate manner. Many statistical methods have been proposed for stock composition studies; however, the most common are discriminant analysis, logistic regression and artificial neural networks. Discriminant analysis, in its linear form (LDA), has the longest history of use in stock composition analysis (Hill, 1959). The fundamental assumptions of LDA and its related method, quadratic discriminant analysis (QDA), is that observed characteristics follow a multivariate normal distribution, and that these have a common variance structure between stocks. An alternative form of discriminant analysis, logistic discriminant analysis (LGA), makes no distributional assumptions (Prager & Fabrizio, 1990), however it's adoption has been poor in fisheries science, compared to LDA and QDA (Pella & Masuda, 2005). Most of the applications of discriminant analysis in fisheries are concerned with the estimation of source population proportions to be found within a mixed population. Unless mixed population measurements form distinct distributions, this sort of classification is reliant upon some prior knowledge of the measured characteristics of the source populations which mix. This can limit the usefulness of discriminant analysis where one is unsure of the identity of source populations, for example, if not all populations which may be present in a mixed population are present in the a priori measured source populations. Logistic regression is a type of generalised linear model (McCullagh & Nedler, 1989), which was suggested for use in stock identification studies by Prager 41 1. Introduction and Fabrizio (1990). The main advantage of this method is that it assumes neither multivariate normality of data or equal variances between stocks. It can also handle continuous, categorical data or a mixture of both. It is most useful in situations with a binary outcome (i.e. does this fish belong to this stock or not), however it has been applied to problems with more than two stocks (Waldman et al., 1997). In logistic regression, prior information is not provided, however the data is split into training and classification sets. A logistic regression estimator is derived from the distributions observed in the training set. This is biased towards the stock composition which is encountered in the training set, and the regression model will become less accurate as the composition of the classification set moves away from this make-up. Artificial neural networks (ANNs) refer to a group of computational algorithms which combine many non-linear logistic regression models to produce an optimum classification method. ANNs make no assumptions about the distribution of data with which they are presented, which is of major advantage when dealing with data which would violate the assumptions of other methods. Neural network classification was first used in stock composition analysis by Prager (1988). Because of the specialist nature of neural network software, methods are not standardised, and it may not be possible to duplicate existing results unless exactly the same software, data set and starting parameters are used. Estimation of stock composition by neural networks is also conditional on the composition of the training sample. 42 1. Introduction 1.3.3 Stock Identification and Management Advice It is clear from considering the above methods that there are a number of valid approaches to the characterisation or definition of a stock. The question remains: is this of any use to decision makers? The answer is, as so many things in fisheries science, rather ambiguous. Boundaries between stocks are diffuse, and stocks have a tendency to expand their distribution. Naturally, this leads to mixing of stocks and exchange of genes. Roque et al. (2002) characterised marine fish species as having very weak or no obvious genetic population structures over large geographic areas, and attributed this to large effective population sizes, high potential for dispersal and weak physical barriers to gene flow. The range of descriptions of what “stocks” actually are (e.g. Ihssen et al., 1981; Carvalho & Hauser, 1994) bears little relationship to the information required by those making decisions on the exploitation of the fishery. To a biologist, the core of the stock concept is a notion of a discrete population or metapopulations, to which one can assign exploitation rates and patterns to each population, with the assumption that there is a stock specific sustainable yield (Hammer & Zimmerman, 2005). This notion is less attractive to managers and fishers, who often define stocks as groups of fish in a particular area, being harvested by a particular method (Carvalho & Hauser, 1994). This confusion has occurred because, at the development of modern European fisheries, stock sizes were relatively large and different stock concepts were held by different groups without causing confusion. As a result, semantic differences of what each party meant when talking about a particular stock became ingrained, and we are left with the current situation where the biological notion of a stock is not necessarily the one which has management 43 1. Introduction actions assigned to it. Hammer and Zimmerman (2005) identified four main areas where new information on stock identity can be most useful: Under-informed Fisheries – For example, deep sea stocks in the North Atlantic show a huge disparity between knowledge of the stocks and their rate of exploitation. Deepwater fish stocks are highly vulnerable to exploitation and can become reproductively impaired very quickly (Merrett & Haedrich, 1997; Koslow et al., 2000). These fisheries are managed on the basis of politically convenient areas, and any information on the identity of stocks within them is welcomed. A problem with this is that these fisheries often occur in international waters, making regulation of management actions problematic. Fisheries where data collection has been compromised – This occurs, for instance, where a fishery crashes and is closed to commercial fishing, or effort in a fishery declines for other, principally economic reasons. It becomes no longer possible to measure stock integrity because of the paucity of data available. This has been seen in the North Sea herring fishery. Originally consisting of three separately assessed components, “Bank” (spawning on Dogger Bank in early autumn), “Buchan” (spawning close to the east coast of Scotland in late summer) and “Downs” (spawning in the English Channel in winter), it was concluded in the 1970s that a year 44 1. Introduction class of a particular stock retains its spawning integrity, even though stocks may have mixed during feeding. Following the decline of the fishery in the English Channel, the collection of biological data was considerably reduced. Having little substantive information, the stock assessment for Downs herring was eventually merged into the overall North Sea herring assessment, even though a healthy herring population still exists in the Downs area. This is convenient for management purposes, as these fish are simply herring which live in the North Sea; however, from a biological point of view, this is less appropriate. This stock forms dense spawning aggregations which could potentially be heavily fished, and also represents one of the southern-most herring stocks. Potentially, this situation could easily lead to depletion of the Downs herring, with loss of genetic diversity, and severe implications for future adaptability to environmental change. Fish stocks in which biological properties change – The fundamental dilemma for fisheries managers is that what may be true today, in terms of the biological parameters (e.g. growth rate, reproductive output), is not necessarily what will be true next year. Biological properties of stocks change in response to climatic effects, predator-prey relationships or the impacts of the fishery on the ecosystem or stock. This has been seen in cod stocks (Gadus morhua) in the Arctic Ocean. Heavy fishing has selected for early maturing individuals; however, the offspring of young fish are less 45 1. Introduction viable. As a result, the stock size becomes more variable, and the stock more vulnerable to collapse due to the shorter age structure. Changing biological properties, however, can be masked by the mixing of two separate stocks, therefore proper understanding of stock dynamics is essential to monitor changing biological parameters. Fisheries where biological information is underutilized – Some fisheries where it is apparent that stock separation occurs, are still managed as a single unit. This is true of the Baltic cod stocks. Two stocks have been shown to exist, one eastern, the other western, which are separated by the island of Bornholm. Despite evidence of stock separation, with quantified levels of migration of juvenile cod from west to east, and a commensurate spawning migration in the opposite direction, cod in the Baltic Sea are assigned TACs (Total Allowable Catches) on the basis of a single stock unit. A change of management units is now considered very urgent as cod populations in the eastern and central Baltic are declining, leading to fishing effort being transferred to the western stock component. This is an example of management actions being taken in the face of available stock identity information; however, additional information increases the political pressure on managers to adopt a more sensible and precautionary management strategy. Evidently, there is a marked difference between the perception and application of the stock concept between fisheries biologists and fisheries managers. The TAC 46 1. Introduction suggested by fisheries biologists refers to a biological entity, while the TAC adopted by managers in most cases refers to a management unit. The two are not necessarily the same. The idea of separation and management of biological entities would only be possible if methods exist to assign individuals to stocks in catches of mixed aggregations. This will only be possible with far more basic research and routine data collection on spawning stocks and stock specific parameters. 1.4. Conclusions It is apparent that identification of discrete fish stocks and the quantification of any degree of mixing between them should be the key first step in the rational management of a fishery resource. Whether this first step is adopted into management practice is another matter. There are a wide range of different approaches which can be applied to investigate the identity of stocks, the choice of which is determined by the concept of stock which is being investigated, the species concerned and existing hypotheses of its stock identity. The application of a range of techniques to a sample of fish is often a more robust way of investigating identity of stocks than the application of a single method. Within this range of techniques, the evaluation of parasitic infection parameters (prevalence, mean intensity and abundance) has been recognised as one of the most promising and under-applied techniques (Hammer & Zimmerman, 2005). It is known that there are gaps in the knowledge of stock identity, behaviour and distribution of both herring to the west of the British Isles and the horse mackerel throughout its geographic range. The application of parasites as biological tags would be a useful contribution to the knowledge of these two species. This is particularly true of the horse mackerel, which is a relatively poorly studied species, 47 1. Introduction compared to the herring. Improved understanding of stock identity in herring is important as the species is known to follow boom-bust cycles, as seen in the North Sea in the late 1970's when the fishery was closed completely. Improved knowledge of the distribution and mixing of spawning populations would allow more biologically relevant fishery regimes to be implemented, going some way to avoid, or at least give prior warning of, population crashes. The herring is fished for both in spawning aggregations, and as non-spawning mixed stock fisheries. Going some way to quantifying stock mixing in the herring would make it more likely that TACs could be set on a more biologically centred basis than they are currently. The horse mackerel, in contrast, has much more stable population levels, characterised by occasional peaks in recruitment, which seem unable to reproduce themselves. Understanding how these exceptional year classes integrate with the existing populations would allow some predictions to be made on the likely extent of distribution of any increase in abundance, improving the quality of fisheries management advice that can be deployed in response to changing biological parameters. Both species are small pelagic fishes that play an important role in terms of energy flow from planktonic organisms to higher predators in the food webs which they inhabit. This makes the comparative aspects of their parasitology and the relative importance of parasites as biological tags in informing multidisciplinary studies on their stock identity additional useful information which can be taken from this work. The aims of this study are fourfold: 48 1. Introduction - To study the parasite fauna of the horse mackerel throughout its geographical range in order to identify parasites which are suitable biological tags. - To develop and apply analytical methods to the data generated in order to study stock structure of horse mackerel. - To study the parasite fauna of the herring in waters to the west of the British Isles in order to investigate stock structure and linkages between juveniles and spawning stocks, and between spawning stocks and mixed feeding aggregations. - To develop novel tools to investigate the link between stock identity of herring and population genetics of their parasites, with a view to applying this to stock discrimination problems. These aims will be addressed with the intention of obtaining results which can be translated into meaningful advice for fisheries management. 49 Chapter 2. The biology and parasite fauna of the Atlantic horse mackerel, Trachurus trachurus (L.) 2.1 Introduction 2.1.1 Taxonomic Status and Distribution The horse mackerel, Trachurus trachurus (L.) (fig. 2.1) - a member of the carangid family, has a worldwide distribution and includes many important commercial species. The name is slightly misleading as the true mackerel-like fishes are members of the Scombridae. There are currently 15 species recognised as valid within the genus Trachurus. o Trachurus aleevi Rytov & Razumovskaya, 1984 o Trachurus capensis Castelnau, 1861 (Cape horse mackerel) o Trachurus declivis (Jenyns, 1841) (Greenback horse mackerel) o Trachurus delagoa Nekrasov, 1970 (African scad) o Trachurus indicus Necrasov, 1966 (Arabian scad) o Trachurus japonicus (Temminck & Schlegel, 1844) (Japanese jack mackerel) o Trachurus lathami Nichols, 1920 (Rough scad) o Trachurus longimanus (Norman, 1935) (Crozet scad) o Trachurus mediterraneus(Steindachner, 1868) (Mediterranean horse mackerel) o Trachurus murphyi Nichols, 1920 (Inca scad) o Trachurus novaezelandiae Richardson, 1843 (Yellowtail horse mackerel) o Trachurus picturatus (Bowdich, 1825) (Blue jack mackerel) o Trachurus symmetricus (Ayres, 1855) (Pacific jack mackerel) o Trachurus trachurus (Linnaeus, 1758) (Atlantic horse mackerel) o Trachurus trecae Cadenat, 1950 (Cunene horse mackerel) 51 T. trachurus is a small pelagic fish, with a maximum theoretical size (Linf), based upon observed growth rates, of around 50cm (Borges, 1991), although Murta et al. found considerably larger fish of 59cm in the field (1993). It is the most northerly distributed member of this family, and is found feeding in the Norwegian Sea during the summer months. Three species from the genus Trachurus are present in European waters, T. trachurus, T. mediterraneus and T. picturatus (Smith-Vaniz, 1986). The most obvious feature which distinguishes these species is the extent of the accessory lateral line. In T. trachurus this organ extends to below the posterior end of the second dorsal fin, between rays 23-31. In T. mediterraneus it ends below the area between the first and second dorsal fins, while T. picturatus shows an intermediate length of accessory lateral line. Although these species can be found in mixed shoals in the Mediterranean, genetic studies showed three clear groups corresponding to each species, with no intermediate principal component scores, ruling out the possibility of hybridization (A. San-Juan, University of Vigo, Pers. Comm.; Anon., 1998). Figure 2.1. The Atlantic Horse Mackerel, Trachurus trachurus (L.). The taxonomic position of the Cape Horse Mackerel or “Meerbanker”, which is found along the south western African coast, is unclear, with some workers concluding that it is a separate species, Trachurus capensis, and others relegating it to the level of subspecies, Trachurus trachurus capensis. For the purposes of this work, the more common use of T. capensis will be followed. This subspecies is found along 52 the West African coast from Mauritania to the Cape. It forms mixed shoals with T. trachurus (Davies, 1957) and hosts many of the same parasites (Gaevskaya & Kovaleva, 1980). Hereafter, the vernacular name “horse mackerel” or “scad” will be considered as a term referring only to T. trachurus. All references to other Trachurus species will be noted by the use of the Latin binomial. Horse mackerel is a long lived species. Cohort analysis shows that the 1982 year class has been extraordinarily strong in the north east Atlantic and supported the fishery for over twenty years. For assessment purposes, these fish are amalgamated into a 15+ year class. 2.1.2 Stock Identity of the Horse Mackerel In the northeast Atlantic, it is now assumed that there are three distinct spawning stocks of T. trachurus (ICES, 2004) (Fig 2.2): – North Sea Horse Mackerel (NSHM) Divisions IIIA (Excluding Western Skagerrak), IVbc and VIId – Western Horse Mackerel (WHM) (Divisions IIa, IIIa (Western Part), IVa, Vb, VIa, VIIa–c, VIIe–k, and VIIIa,b,d,e – Southern Horse Mackerel (SHM) Division IXA A number of methods have been used to provide evidence for the existence of true stocks of horse mackerel, with limited success. (Polonsky & Baydalinov, 1964 Nevedov et al., 1978, Gaevskaya, 1978, Kompotilski, 1975). The recently completed multidisciplinary stock identification project, HOMSIR (Horse Mackerel Stock Identification Research) (EU Contract no. QLK5- Ct1999-01438), has been 53 incorporated into ICES stock definitions, and the results of this project have broadly supported this notion of stock identity, although some workers have supported the division of the North Sea stock into a northern and southern component (Rückert et al., 2002). An attempt to discriminate horse mackerel (Trachurus trachurus) stocks in Iberian and North African Atlantic waters was carried out by Murta (2000). For this study, samples were collected at 4 sampling sites in the Portuguese and Spanish coasts, to the south of Cape Finisterre, and in 1 site off the Moroccan coast. Fourteen morphometric measurements were recorded, along with 5 meristic characters. The main conclusions were that, in this case, meristic measurements did not provide useful information regarding the discrimination of stocks. The morphometric measurements showed that the areas in the Portuguese coast were more closely related to each other morphologically than to fish from Morocco, and all these were set apart from fish from the Spanish coast, off Cadiz. However, this latter area was less sampled than the others, and the sampled fish were of smaller size. 54 Figure 2.2. Stock definitions of horse mackerel in the north east Atlantic. 2.1.3 Prey and Predation Horse mackerel are a significant trophic link between planktonic organisms and higher predators, and feed at a slightly higher trophic level than many other small pelagic fishes (Loc’h & Christian, 2005). In the Atlantic Ocean, they prey heavily upon zooplanktonic crustaceans, particularly Nyctiphanes couchi and Meganyctiphanes norvegica, but have also been recorded as feeding on gastropod and cephalopod molluscs, decapods and their larvae and marine worms (Cabral and Murta, 2002). Horse mackerel also feed to a lesser extent on fish larvae. Fish from the North Sea predate more extensively on fishes, and from an earlier age, than those from adjoining Atlantic waters. In the Mediterranean Sea horse mackerel also feed upon zooplanktonic crustaceans such as N. couchi and Euphausia krohni, but up to 55 one third of the diet consists of small fishes, particularly Maurolicus muelleri and Gadiculus argentus (Jardas et al., 2004). Horse mackerel are reported to show significant diurnal patterns in their feeding behaviour, with both Boely et al. (1973) and Olasao et al. (1999) finding that feeding was most intense overnight, and at its lowest ebb during the late afternoon. Jardas et al. (2004) and Cabral and Murta (2002) did not find significant seasonal variations in dietary composition in the Mediterranean or Atlantic, although Cabral and Murta found significantly more empty stomachs during winter months. Horse mackerel appear to stop feeding at low temperatures (Anon., 1998). From a parasitological perspective, this means they are unlikely to receive new endoparasitic infections during the first and last quarters of the year (October – March) in colder more northerly waters. Horse mackerel are preyed upon by a diverse range of predators. They have been recorded from the stomachs of the Giant Squid, Architeuthis dux (Lordan et al., 1998), whales (Fernandez et al., 2004), seals (Pierce and Santos, 2003) seabirds (Granadeiro et al., 1998; Granadeiro et al., 2002) and predatory fishes (Deudero and Morales-Nin, 2000; Zidowitz et al., 2002). This diverse diet and range of predators makes the horse mackerel a “crossroads” species for many parasites, which goes some way to explain the diverse larval parasitofauna which it plays host to. 2.1.4 Reproduction, growth and migration North Sea and Western stocks have distinct spawning areas which have been identified by a series of egg surveys. However, horse mackerel is a highly migratory species and, outside the spawning season, distributions of Western and North Sea stocks may overlap. This is believed to be the result of fish moving into the North Sea 56 from the Western stock, mainly over the north of Scotland, but also by migration through the English Channel (Anon., 1992). Horse mackerel are asynchronous spawners, with the spawning period in some regions extending up to 8 months (Abaunza et al., 2003). In the North Sea and English Channel, the spawning period is considered to be a four month period from May to August (Polonsky & Tormosova, 1969; Macer, 1974). Eltink (1992) found peak spawning in the North Sea to be temperature dependent, with high temperature resulting in earlier spawning. In the Western stock area, duration of spawning is roughly similar, although spawning begins earlier, with fish preferentially spawning in the springtime (Lucio & Martin, 1989). Migration of the Western stock to and from its overwintering areas appears to be driven by the position of the 10°C isotherm, with cessation of feeding occurring at 9°C (Abaunza et al., 2003). They appear to show complete avoidance of water of 8°C or less (Polonsky, 1965). In the Southern stock area, egg surveys have not been successful in identifying discrete spawning areas, either spatially or temporally (Anon., 1992). This means that there is some uncertainty in the demarcation between spawning areas of Western and Southern stocks. Peak spawning period is similar to that of fish in the Western stock, although the spawning duration can be considerably longer, with some workers finding spawning from February to December, with a peak from April to June (Sola et al., 1990). Horse mackerel is a long-lived species, with a maximum age of well over 30 years (Eltink & Kuiter, 1989). The horse mackerel or at least the western stock area thereof, is capable of producing infrequent, massive, recruitment events. The last of these was spawned in 1982 and led to a four-fold increase in the biomass of the stock. Currently the factors which contribute to such spectacular reproductive success are 57 unknown. This poses certain problems to the assessment of the species, as the 1982 year class was not capable of reproducing itself, suggesting that the recruitment success of this species is primarily driven by environmental factors, and relatively independent of parental stock size. Spawning of horse mackerel in the Mediterranean Sea is not well studied. Fish spawn in late winter or early spring (Alegria-Hernandez, 1994) in a similar location to their Atlantic counterparts, along the shelf edge, in around 200m of water. The Mediterranean shelf is generally narrow, with few exceptions, such as the Adriatic Sea. It is likely that fish spawn along the shelf edge throughout the Mediterranean (Lleonart, 2004), and more widely dispersed in places like the Adriatic Sea (AlegriaHenrnandez, 1984). This is supported by the distribution of horse mackerel fisheries, which operate mainly on juvenile fish throughout the Mediterranean. 2.1.5 Fisheries Horse mackerel in Atlantic waters have supported catches of over 500,000 tonnes per annum within the last decade (Fig.2.3). The growth in catches is particularly attributed to the very strong 1982 year class. This year class is almost entirely responsible for the massive growth in catches over the following fifteen years. These fish are now being fished out of the stock, leading to a decline in population numbers. Effort in this fishery has also declined in recent years, especially in the Western stock area. 58 500 300 200 0 100 Catches ('000 Tonnes) 400 Western North Sea Southern 1975 1980 1985 1990 1995 2000 2005 Year Fig. 2.3. ICES records of catches in the Atlantic from 1973 to 2004. The vast 1982 year class of the Western stock supported a major expansion of the fishery over the following 15 years, with total catches reaching over 500,000 tonnes. Horse mackerel fisheries are assessed on a stock by stock basis. Fisheries on the Western stock take place in areas west of Ireland or at the Bay of Biscay continental slopes. On occasion, larger individuals of the Western stock reach as far north as Trondheim, on the south-western coast of Norway, in July and August, when a seasonal, non-spawning fishery develops. Fisheries on the North Sea stock operate mainly in the southern North Sea and eastern English Channel. The North Sea stock appears in April at the southern Dutch and English coasts (Meek, 1916), and reaches the western Jutland coast and southern Norwegian coast by August (ICES, 2004). 59 The fishery in the Atlantic is mainly prosecuted in North Sea and Western stock areas by pelagic trawlers, both for human consumption and for industrial purposes. In the Southern stock area the fishery is more dependent upon small artesanal fishers (ICES, 2004). In the Mediterranean sea, catches are split between a range of pelagic and demersal fleets, but the generally used mesh size of 40mm means that the fishery operates mainly on juvenile fishes (Lleonart, 2004) with relatively few surviving to adulthood and becoming inaccessible to most fishers (Caddy, 1990). There are no robust records of the distribution or size of catches in the Mediterranean, however the species is a commercially significant one, with catches throughout the sea varying between 5,155 and 21998 tonnes per year between 1989 and 1998 (FAO, 2000) (Fig. 2.4). This is an order of magnitude smaller than catches from Atlantic waters over the same period. 60 30 25 20 15 10 0 5 Catch ('000 Tonnes) 1970 1975 1980 1985 1990 1995 2000 Year Figure 2.4. Horse mackerel catches in the Mediterranean Sea, as recorded by the General Council on Fisheries in the Mediterranean. Catch levels are very low, relative to the same species in the Atlantic. The southern stock has large overlaps between spawning and feeding areas and is fished on throughout the year in the same locations. The Western stock gathers to spawn on the shelf edge, and then spreads out along the Atlantic shelf waters to feed (Abaunza et al., 2003). Catches follow the distribution and abundance patterns of the stock. This is in part due to horse mackerel forming mixed shoals with mackerel (Scomber scombrus L.) (Van Marlen, 2000), which is the primary target of the fleets. 61 2.1.6 Parasitology of horse mackerel Although a highly commercially significant fish, studies on the parasite fauna of the horse mackerel have been relatively uncommon. Previous studies have either concentrated on the distribution (or more commonly, the discovery) of a single species of parasite within horse mackerel from a particular geographical area (Cruz et al., 2003), related differences in the distribution of a single species to management objectives (Abaunza et al., 1995), or addressed public health issues (Adroher et al., 1996). Numerous species have been reported to infect the horse mackerel, starting with the work of Nicoll in the early twentieth century, who described a number of species of digeneans from horse mackerel off the Scottish coast (1909, 1910). Llewellyn (1959, 1962) studied the monogenean parasites of horse mackerel in the English Channel, focussing on seasonal changes in their prevalence and abundance. Llewellyn (1959) showed that larvae of Gastrocotyle trachuri and Pseudaxine trachuri are most common in May – only adults are found in July or August. Llewellyn (1962) went on to show that G. trachuri and P. trachuri are most common on young fish and less frequent on 2 and 3 year old fish and rarer still on older fish. Llewellyn proposed that chances of monogeneans successfully infecting their hosts are enhanced by synchronisation of parasite reproduction with suitable host behaviour, thus G. trachuri and P. trachuri infect their host near sea bottom early in the year. Later in the summer, fish move up in the water column to feed on plankton, a change to which the parasites are adapted to by ceasing to produce larvae (Llewellyn, 1962). As for infections with other parasites, Kabata (1979) records 3 species of copepods which are known to infect the horse mackerel in British waters, belonging to the genus Caligus. These are C. diaphanus, C. elongatus and C. pelamydis. 62 Records of acanthocephalans infecting horse mackerel include Pseudoechinorhynchus clavula in British waters by Southwell and Macfee (1925) (a somewhat doubtful identification, given that this is a parasite of freshwater fishes (Valtonen, 1979; Choudhury & Dick, 1998)) and Rhadinorhynchus cadenati by Golvan (1969) from the Senegalese coast, where it also infected the albacore (Neothunnus albacore), the bogue (Boops boops), the sardinella (Sardinella aurata) and another species of carangid, Trachurus trecae. While individual aspects of the parasitology of the horse mackerel, such as Llewellyn’s (1962, 1969) studies of seasonality of monogenean infections, have been investigated in depth, only a single publication examines geographical variations in the parasite fauna of horse mackerel throughout its range in the Atlantic Ocean (Gaevskaya & Kovaleva, 1980), while the distribution of parasites in the Mediterranean Sea remains an unknown quantity. 63 Species North Sea English Channel Myxosporea Celtic Sea Kudoa quadratum Digenea Ancylocoelium typicum Aphanurus stossichi Derogenes varicus Ectenurus lepidus √ √ √ √ √ √ Hemiurus communis Hemiurus lueheii Lecithaster confusus Lecithaster gibbosus Monascus filiformis √ √ Tergestia laticollis √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ Zoogonus rubellus Monogenea Cemocotyle trachuri √ Pseudaxine trachuri √ √ Cestoda Anthobothrium cornucopia Christianella minuta Grillotia erinaceus Lacistorhynchus tenuis Nybellinia lingualis Phyllobothrium sp. Scolex pleuronectis Tetrarhynchid larvae √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ Acanthocephala Anisakis simplex Contracaecum aduncum Crustacea Caligus curtus Caligus elongatus √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ Rhadinorhynchus cadenati Nematoda √ √ √ Diplectanotrema trachuri Heteraxinoides atlanticus √ √ √ Opechona magnibursta Pseudopecoeloides chloroscombri √ √ √ Neopechona pyriforme Sahara √ √ Kudoa nova Goussia cruciata Gibraltar √ Alataspora serenum Apicomplexa Bay of Biscay √ √ √ √ √ √ √ √ √ √ √ √ Table 2.1. Distribution of parasites of horse mackerel recorded by Gaevskaya & Kovaleva (1980) Caligus pelamydis Lernanthropus trachuri 64 Gaevskaya and Kovaleva (1980) examined 5600 individual horse mackerel on a number of research voyages in the northern and central Atlantic Ocean between 1973 and 1976 A translation of this work can be found in appendix I. They collected material from the North and Celtic seas, on the Porcupine Bank, the English Channel, Bay of Biscay, the Strait of Gibraltar and the coast of Mauritania. They recorded 37 parasite taxa, 11 of which were new records for the horse mackerel, and 3 of which were new to science. Their findings of parasite distributions are summarised in table 2.1. This study found the horse mackerel to be host to a wide range of adult stage digenean parasites. This is a reflection of the diverse planktonic diet which it consumes. The horse mackerel was also found to be a host of a diverse range of post larval cestodes and host to large numbers of L3 stage Anisakis spp., indicating the importance of this species in the transfer of these parasites to their final hosts, predatory fishes and marine mammals. The study identified a number of species as potential biological tags, but did not make hypotheses regarding stock structure. The parasitology of the Black Sea horse mackerel has been relatively well studied (Radulescu, 1979), and shows numerous similarities with the parasitofauna of the Atlantic horse mackerel reported by Gaevskaya and Kovaleva (1980), including the digeneans Tergestia laticollis, Ancylocoelium typicum, Lepocreadium sp., Ectenurus lepidus, Stephanostomum sp., Prodistomum polonii and the cestode complex Scolex pleuronectes (Vlasenko, 1931; Osmanov, 1940). Of the other carangid species, Trachurus japonicus and Trachurus declivis have the best studied parasite fauna. Ichihara (1968) found Anisakis sp., Bolbosoma sp., Hysterothylacium sp, Tergestia laticollis and Aponurus tsugunri in T. japonicus in Japanese coastal waters. 65 2.2 Methods Mature horse mackerel were collected from a number of sites throughout European waters during 2000 and 2001 by a combination of research cruise and market sampling, ideally during the spawning period (table 2.2, figure 2.5). These fish were individually frozen to -18°C upon capture and transported to Aberdeen for parasitological examination. Year Sample ID No. Latitude Longitude Month Sample Size 2000 01-00 59.00°N 03.00°E October 50 02-00 52.00°N 11.20°W June 50 03-00 49.30°N 10.15°W July 50 05-00 55.45°N 06.50°E June-July 53 06-00 48.15°N 05.51°W March 50 07-00 43.25°N 08.30°W March 50 12-00 39.24°N 02.28°E May 50 15-00 38.15°N 20.42°E June 22 20-00 40.43°N 01.30°E June 50 21-00 44.00°N 01.38°W June 50 01-01 57.41°N 05.10°E October 50 02-01 52.50°N 12.00°W March 36 03-01 54.45°N 06.00°E April 50 05-01 48.45°N 09.29°W May 50 06-01 51.35°N 11.06°E April 50 07-01 43.35°N 08.25°E March 50 11-01 19.58°N 17.28°W January 50 12-01 39.24°N 02.28°E April-May 52 15-01 38.15°N 20.42°E June 44 16-01 40.28°N 24.55°W May 50 20-01 42.07°N 03.18°E December 49 21-01 44.00°N 01.38°W June 50 2001 Table 2.2. Location, timing and sample size of horse mackerel examined for parasites at the University of Aberdeen. 66 Fish were defrosted at room temperature, up to the point where they became pliable. Total length was measured where possible. In cases where the tail of the fish had been damaged, fork length was measured and total length approximated from this. Figure 2.5. Location of sampling sites in northern Atlantic and Mediterranean waters. Red circles represent a site sampled in 2000, blue represents samples collected in 2001 and yellow represents sites sampled in both years. A further sample was collected from the Atlantic coast of North Africa, although it is not shown here. Fish were examined externally for ectoparasites on the skin. The pectoral and pelvic fins were removed and examined under a dissecting microscope. Fish were opened by an incision from behind the base of the left pelvic, to the right of the base of the pectoral fin to the top of the left operculum. A further incision was made from above the base of the left pectoral fin, along the body cavity, ending above the vent. A final incision was made along the ventral side of the fish, from the pelvic fin to just in 67 front of the vent (fig.2.6a). The opened body cavity wall was then folded back, revealing the left side of the viscera. Scissors were inserted forwards in front of the liver and the oesophagus was cut. A further cut was made at the hindmost point of the intestine. The viscera were removed as a single unit and allowed to defrost completely (fig. 2.6b). The viscera was divided into stomach, pyloric caeca, intestine, gonad, gall bladder and liver. Each organ was placed in individual Petri dishes and irrigated with physiological saline (fig. 2.6c). The left operculum was removed and placed in a Petri dish (fig. 2.6d), followed by the individual gill arches. These were irrigated with physiological saline and examined using a dissecting microscope for the presence of monogeneans. The process was repeated for the opposite opercula and gill arches. The stomach was examined externally for nematode and cestode larvae under a dissecting microscope. It was then opened and the contents examined for the presence of Monogenea and Digenea. The pyloric caeca were examined for nematode larvae. They were then macerated and the contents examined for the presence of digeneans. The intestine was also examined for nematode and cestode larvae before being opened longitudinally and the lumen examined for the presence of adult nematodes and digeneans. Finally, nematode larvae on the surface of the gonads were removed. Smears of liver tissue and bile from the gall bladder were examined for the presence of protozoan parasites, using a Zeiss photo-microscope II set for phase contrast microscopy, at a magnification of 325X. Finally, the fish was filleted and the fillets candled to reveal parasites encysted in the musculature. 68 Figures 2.6 a-d. Clockwise from top left, a) the wall of the body cavity is removed to reveal the internal organs, b) the viscera can be removed as a single body, as they are only partially defrosted – this helps retain the contents of the gall bladder c) the viscera are separated and placed in individual Petri dishes for examination, d) the right operculum is removed, to reveal the gill arches beneath. All parasites were removed from the organs and placed into watch-glasses containing physiological saline maintained on ice. Anisakis spp. and Hysterothylacium aduncum were placed into separate dishes. All other parasites were placed into a single dish for identification and enumeration. For descriptive purposes, acanthocephalans, monogeneans, digeneans and cestodes were stained with borax carmine solution, dehydrated with a series of ethanol solutions, cleared with xylene or beechwood creosote, and mounted in DPX, before being examined at magnifications of up to 800x, under a Zeiss Photomicroscope II. Nematode heads and tails were removed, cleared and mounted in glycerine jelly for microscopic examination. Copepods were dehydrated in ethanol 69 series and cleared in beechwood creosote before being mounted in DPX for examination. Permanent mounts of myxosporeans were prepared by air-drying smears before fixing in methanol and staining with Giemsa stain. Temporary mounts of myxosporeans were prepared as agar monolayer slides, which were examined using oil immersion microscopy at magnifications of up to 2000x. All photomicrographs were taken using a Nikon Coolpix 4500 digital camera, mounted on a Brunell Microscopes 10x eyepiece adaptor. Measurements were taken using an eyepiece graticule. All measurements are given in micrometers (μm) as ranges, and where appropriate, followed in parentheses by means plus or minus one standard deviation. Sites and dates of capture are given for the location where the parasite was most prevalent. 70 2.3 Results 2.3.1 List of species recorded A list of the parasites found throughout the area of study is presented below. Apicomplexa Goussia cruciata (Thélohan, 1892) Myxosporea Alataspora serenum Gaevskaya & Kovaleva, 1979 Alataspora solomoni Yurakhno, 1988 Kudoa nova Naidenova, 1975 Kudoa sp.* Myxobolus spinacurvatura Maeno, Sorimachi, Ogawa & Egusa, 1990* Monogenea Cemocotyle trachuri Dillon & Hargis, 1965 Gastrocotyle trachuri van Beneden & Hesse, 1863 Heteraxinoides atlanticus Gaevskaya & Kovaleva, 1979 Pseudaxine trachuri Parona & Perugia, 1889 Unidentified polyopisthocotylean monogenean* Paradiplectanotrema trachuri (Kovaleva, 1970) Digenea Derogenes varicus (Müller, 1784) Ectenurus lepidus Looss, 1907 Hemiurus communis Odhner, 1905 Lasiotocus tropicus (Nicoll, 1912) Lasiotocus typicum (Nicoll, 1912) 71 Lecithocladium excisum (Rudolphi, 1819) Monascus filiformis (Rudolphi, 1819) Opechona bacillaris (Molin, 1859) Pseudopecoeloides chloroscombri (Fischal & Thomas, 1970) Prodistomum polonii (Molin, 1859) Tergestia laticollis (Rudolphi, 1819) Cestoda (all postlarvae) Grillotia erinaceus (van Beneden, 1858) Nybelinia lingualis Cuvier, 1817 Pseudophyllidean plerocercoids* Scolex pleuronectis (Müller, 1788) Acanthocephala Corynosoma strumosum (Rudolphi, 1802) juveniles* Corynosoma wegeneri Heinze, 1934 juveniles* Rhadinorhynchus cadenati (Golvan & Houin, 1964) Nematoda Anisakis spp. (Rudolphi, 1809) larvae Hysterothylacium aduncum (Rudolphi, 1802) larvae & adults Pseudanisakis sp. larvae* Pseudoterranova decipiens (Krabbe, 1878) larvae * Crustacea Caligus elongatus Nordmann, 1832 Ceratothoa oestroides (Risso, 1826) Argulus vittatus (Risso, 1826)* Parasites which are new host records are highlighted with an asterix. 72 2.3.2 Descriptions Although it was possible to identify all parasites to the genus level, with the exception of the pseudophyllidean plerocercoids and the polyopisthocotylean monogenean, the fact that the material being examined had been frozen meant that it was not always possible to obtain suitably preserved or sufficient quantities of specimen material for comprehensive descriptions of all parasite taxa recovered. Descriptions of the more significant or frequently encountered parasites as well as any associated pathological changes are given below. 2.3.2.1 Apicomplexa Only one species of apicomplexan was recorded in this study, Goussia cruciata Thélohan 1890. This infected the liver, and appeared to cause no histological changes. Goussia cruciata (Thélohan, 1892) Family: Eimeriidae Schnieder 1875 Genus: Goussia Labbé 1906 Site of Infection: Liver; gonad (heavy infection) Location: Portuguese coast, approx 41.50º N, 08.50º W Date: February 2000 Host length range: 21.6-31.0cm Prevalence: 98% (50/51) Description: Pathological changes not observed, there are some reports of liver necrosis in “very heavily infected individuals” (pers. comm. S. Mattiucci, Universitá 73 di Roma “La Sapienza”, Rome) although these changes were not observed in fish examined by the author and cannot be commented upon. Oocyst spherical, thickwalled (approx.1-2μm). Dimensions, based on 50 measurements, oocyst diameter 1622 (18.4±1.35), oocyst contains four ellipsoidal sporocysts, in a characteristic and symmetrical tetrahedral arrangement with lesser ellipsoidal end innermost, length of sporocysts 8-9 (8.46±0.23), width of sporocyst 6-7 (6.55±0.40). 2.3.2.2 Myxosporeans Five myxosporean parasites from three families were recorded infecting the horse mackerel in the Atlantic Ocean and Mediterranean Sea. Alataspora serenum, Alataspora solomoni and Kudoa sp. were found to be coelozoic in the gall bladder. Kudoa nova was found in the red musculature. Myxobolus spp. was found in the liver of only one fish from the Adriatic Sea. A publication based on the myxosporeans found in this study can be found in appendix II. Alataspora serenum Gaevskaya et Kovaleva, 1979 (fig. 2.7.1 & 2.7.6) Order: Bivalvulida Shulman, 1959 Family: Alatasporidae Schulman, Gaevskaya et Kovaleva, 1979 Genus: Alataspora Schulman, Gaevskaya et Kovaleva,, 1979 Site of infection: Gall Bladder Location of capture: West of Ireland, 52º 52.8’ N 12º 03.6’ W Date: 25th March 2001 Prevalence: 6/25 (24%) 74 Host Length Range 19.6-43.1cm Figure 2.7.1. Alataspora serenum in lateral view. Scale bar represents 5μm. Description: No pathological changes in host noted. Trophozoite stage not observed. Spores transparent, crescent shaped, anterior end convex, posterior end concave. Spore valves equally sized. Suture straight and distinct. Triangular body containing polar capsules, with alate projections extending laterally from the anterior of the body. Polar capsules spherical. Spore dimensions, based on observations of 50 spores: length 3.8-7.7 (5.80±0.83), width 11.5-19.2 (14.40±1.56). Spore length:width ratio, 1:1.67-3.50. Polar capsules equally sized, opening to anterior of spore. Polar capsule dimensions, based on 40 measurements, diameter 1.3-2.6 (1.87±0.39). Alataspora solomoni Yurakhno, 1988 (fig. 2.7.2, 2.7.7 & 2.7.8) Order Bivalvulida Shulman, 1959 Family: Alatasporidae Schulman, Gaevskaya et Kovaleva, 1979 Genus: Alataspora Schulman, Gaevskaya et Kovaleva,, 1979 75 Figure 2.7.2. Alataspora solomoni in lateral view. Scale bar represents 10μm. Site of infection: Gall Bladder Location of capture: Ionian Sea 40º 28’ N 24º 55' E Date: 7th June 2001 Prevalence: 11/50 (22.0%) Host Length Range (cm): 12.4-30.3 Description: No pathological changes noted. Vegetative stages not observed. Spores transparent, crescent shaped, anterior end convex, posterior end concave. Spore valves equally sized. Suture straight and distinct. Triangular body, with slight thickening around suture, containing polar capsules. Polar capsules opening onto different sides of anterior side of the spore. Spore dimensions (μm), based on observations of 50 spores: length 5.1-9.0 (6.7±0.8), width 17.9-33.3 (24.3±3.2). Spore length:width ratio, 1:2.67 - 1:5.75. Polar capsules spherical, unequal in size. Dimensions of polar capsules, based on 40 observations, diameter of larger polar capsule, 1.6-2.88 (2.24±0.39), diameter of smaller polar capsule 0.96-2.56 (1.94±0.4). 76 Kudoa sp. (fig. 2.7.3) Order: Multivalvulida Shulman, 1959 Family: Kudoidae Meglitsch, 1960 Genus: Kudoa Meglitsch, 1947 Site of infection: Coelozoic in the Gall bladder Location of capture: South of Ireland, 48º45’N 09º29’W Date: April 2001 Prevalence: 3/46 (6.5%) Host Length Range (cm): 26.4 – 31.8 Figure 2.7.3. Kudoa sp. in apical (left) and transverse (right) view. Scale bar represents 2μm. Description: No pathological changes noted. Vegetative stages not observed. Spore transparent, subquadrate in apical view, with rounded valves. Sutural line thin and indistinct. Anterior side of spore convex, posterior side “bell” shaped. Polar capsules large, pyriform. Spore dimensions, based on 17 measurements: length 5.1-7.7 (5.97±0.79), spore width 5.3-7.3 (6.11±0.53) polar capsule length 2.6-3.8 (3.10±0.45). 77 Kudoa nova Naidenova, 1975 (fig. 2.7.4, fig 2.7.9a & b) Order: Multivalvulida Shulman, 1959 Family: Kudoidae Meglitsch, 1960 Genus: Kudoa Meglitsch, 1947 Site of infection: Dorsal, ventral and lateral red muscle Location of capture: North African Coast, 19º58’N 17º28’W Date: January 2001 Prevalence: 3/46 (6.5%) Host Length Range: 21.3-24.6cm Figure 2.7.4. Kudoa nova in transverse (left) and apical (right) views. Scale bar represents 3μm. Description: Pseudocysts macroscopic, white, spherical, up to 3mm in diameter, found in clusters in the dorsal, ventral and lateral red muscles. Vegetative state of spore not observed. Spore subquadrate in apical view, with rounded valves. Sutural line thin and indistinct. Slightly elongate, ventrally flattened, in lateral view. Polar capsules large, pyriform. Spore dimensions based on 30 spores: length 5.1-7.7 (6.2±0.9), width 5.1-7.7 (6.2±0.9), polar capsule length 1.3-2.6 (1.8±0.4). 78 Myxobolus spinacurvatura Maeno, Sorimachi, Ogawa et Egusa, 1990 (fig. 2.7.5, fig.2.7.10) Order: Bivalvulida Shulman, 1959 Family: Myxobolidae Thélohan, 1892 Genus: Myxobolus Butschli, 1882 Site of infection: Liver Location of capture: 40º 28’ N 24º 55' E Date: 7th June 2001 Prevalence: 1/43 (2.3%) Host Length Range (cm): 23.8 Figure 2.7.5. Myxobolus spinacurvatura in lateral (left) and transverse (right) view. Scale bar represents 5μm. 79 Description: No pathological changes noted in host. Spore oviform in lateral view, flattened perpendicular to sutural line. Two polar capsules, opening separately on the apical side of spore. Polar filaments coiled in three or four turns. Spore dimensions, based on measurements of 40 spores: length 8.9-11.5 (9.97±0.90), thickness 3.8-6.4 (4.92±0.82), width 7.1-9.2 (8.56±0.61). Length of polar capsule 3.8-5.1 (4.24±0.60), width of polar capsule 2.3-2.9 (2.78±0.26). This species has been comprehensively described by Maeno et al. (1990) and Bahri and Marques (1996) Figure 2.7.6. Alataspora solomoni Gaevskaya & Kovaleva,1979. (x800) Scale bar represents 5μm. Figure 2.7.7 & 2.7.8. Alataspora serenum Yurakhno, 1988, in lateral (left) and apical (right) view, showing the orientation of the polar capsules (x800). Scale bar represents 10μm. 80 Figure 2.7.9a & b. Kudoa nova Naidenova, 1975 in apical (left) and lateral (right) views (x800). Scale bar represents 5μm. Figure 2.7.10. Myxobolus spinacurvatura. (x800). Scale bar represents 10 μm. 81 2.3.2.3 Monogenea Four monogeneans were found infecting the gills of horse mackerel, Gastrocotyle trachuri, Pseudaxine trachuri, Heteraxinoides atlanticus and a single specimen of an unidentified polyopisthocotylean monogenean. A number of specimens of Paradiplectanotrema trachuri were also found. This species is noteworthy as it is an endo- or mesoparasitic monogenean, being found in the stomach or on the pharynx. Gastrocotyle trachuri van Beneden et Hesse, 1863 (fig. 2.7.11, fig. 2.7.12) Order: Gastrocotylinae Sproston, 1946 Family: Gastrocotylidae Price, 1943 Genus: Gastrocotyle van Beneden et Hesse, 1863 Site of infection: Gill filaments Location of capture: southern North Sea Date: 7th June 2001 Prevalence: 62% (31/50) Host Length Range (cm): 20.7-30.4 Description: Terminal lappet with 1-3 pairs of anchors, opisthaptor developed along one edge of the body as a marginal frill which bears numerous clamps (22-40). Vitellaria extending throughout length of body. Dimensions, based on 20 measurements; length 260-420, width 96-118. Figure 2.7.11. Gastrocotyle trachuri. 82 Figure 2.7.12. Gastrocotyle trachuri stained with borax carmine (x125). Heteraxinoides atlanticus Gaevskaya & Kovaleva, 1979 (fig. 2.7.13) Family: Axinidae Unnithan, 1957 Genus: Heteraxinoides Yamaguti, 1963 Location of capture: Norwegian coast, 57.41ºN 05.10ºE Date of capture: October 2001 Prevalence:2% (1/50) Host length range:32.7cm Figure 2.7.13. Heteraxinoides atlanticus (x125) Description: H. atlanticus was rarely encountered during this study, and none of the material recovered was in a good enough condition to base a description on. 83 Pseudaxine trachuri Parona et Parugia, 1890 (fig. 2.7.14, fig.2.7.15) Family: Gastrocotylidae Price, 1943 Subfamily: Gastrocotylinae Sproston, 1946 Genus: Pseudaxine Parona et Perugia, 1890 Site of infection: Gill filaments Location of capture: 38.15°N 20.42 Date of capture: June 2000 Prevalence:27% (8/22) Host length range:17.9-22.9cm Figure 2.7.14. Pseudaxine trachuri. Description: Terminal lappet with one pair of anchors. Haptor asymmetric, unilateral, with 24-28 hooks, of similar structure to those of G. trachuri. Vitellaria distributed between base of haptor and genital pore. Figure 2.7.15. Pseudaxine trachuri stained with borax carmine (x125). 84 Unidentified polyopisthocotylean monogenean Subclass: Polyopisthocotylea Family: incertae cedis Genus: incertae cedis Site of infection: Gill filaments Location of capture: Norwegian coast, 57.41ºN 05.10ºE Date of capture: October 2001 Prevalence:2% (1/50) Host length range:34.4cm Description: A single specimen of an unidentified polyopisthocotylean monogenean was recorded from a fish captured from the Norwegian coast. This monogenean was not sufficiently well preserved to describe or identify to species level. Paradiplectanotrema trachuri (Kovaleva, 1970) (fig. 2.7.16) Order: Monopisthocotylea Odhner, 1912 Family:Dactylogyridae Bychowsky, 1933 Genus: Paradiplectanotrema (Johnston et Tiegs, 1922) Site of infection: Pharynx, Stomach Location of capture: south of Corsica, 39.00ºN 9.09ºE Date of Capture: March 2000 Prevalence:52% (11/21) Host length range:17.1 – 37.8cm 85 Description: body elongate, cylindrical, of equal width. Front end is rounded with 2 groups of glands and 2 pairs of eyes. Pharynx is round, 0.1mm in diameter. The alimentary tract consists of 2 branches, which do not discharge. The alimentary canal does not reach past the lower boundary of vitellaria. Length varies from 1.3 to 2.2mm and width from 0.3-0.47mm. Attachment disk is armed with 2 pairs of large hooks. Ovary is ovoid-shaped and located around one third of the way along the body. Figure 2.7.16. Paradiplectanotrema trachuri stained with borax carmine (x125). 86 2.3.2.4 Digenea A diverse range of digeneans were found, inhabiting all parts of the digestive tract. Derogenes varicus (Müller, 1784) (fig. 2.7.17 a-b) Family:Derogenidae Nicoll, 1910 Genus: Derogenes (Lühe, 1900) Site of infection: Stomach Location of capture: southern North Sea, 54.45ºN 06.00ºE Date: May 2001 Prevalence:26% (13/50) Host Length Range (cm):22.8-29.3cm Figure 2.7.17a & b Derogenes varicus stained with borax carmine (x125) Description: Small, vermiform body, found in stomach. Small, subterminal oral sucker, with small pharynx. Ventral sucker larger, positioned behind mid-line of body. Testes located at posterior end of body, behind ovary. Vitellaria distributed 87 throughout. This species is one of the most cosmopolitan of all fish parasites, and shows little host specificity. Ectenurus lepidus Looss, 1907 (figs 2.7.18-19) Family:Hemiuridae Looss, 1899 Genus: Ectenurus Looss, 1907 Site of infection: Stomach Location of capture: south of Corsica, 37.00N 08.30W Date: February 2000 Prevalence: 41% (25/61) Host Length Range (cm): 17.9-23.0cm Description:Although the body surface has been reported to have pliations (Gibson, 2002), these were not visible in all the frozen specimens examined. Oral sucker small, terminal, followed by very small pharynx. Ventral sucker large, projects from body when viewed laterally. Testes and ovary round, located behind ventral sucker. Ecsoma well developed. This species is specific to the genus Trachurus. Figure 2.7.18 Ectenurus lepidus 88 Figure 2.7.19 Ectenurus lepidus stained with borax carmine (x125). Hemiurus communis Odhner, 1905 (fig. 2.7.20) Family Hemiuridae Looss, 1899 Genus Hemiurus Rudolphi 1809 Site of Infection: Stomach Location of capture: Gulf of Lyon, 40.43N 01.30E Date: June 2000 Prevalence: 18% (9/50) Host length range:17.5-20.0cm Description: Found in the stomach, H. communis is between 0.5-2.5mm in length. Body cylindrical, striated, ecsoma well developed. Oral sucker small, sub-terminal, ventral sucker larger, located just behind the oral sucker. Testes sit in tandem just behind the ovary. 89 Figure 2.7.20. Hemiuris communis stained with acid alum carmine (x125) Lasiotocus typicum (Nicoll, 1912) (fig. 2.7.21) Family: Monorchiidae Odhner 1911 Genus Lasiotocus Looss in Odhner, 1911 Site of Infection: Posterior intestine Location of capture:Gulf of Lyon, 40.43N 01.30E Date: June 2000 Prevalence:6% (3/50) Host length range:18.6-20.7cm Description: Originally described from horse mackerel from off the coast of Aberdeen, Lasiotocus typicum (syn. Ancylocoelium typicum) has recently been reviewed and redescribed by Bartoli and Bray (2004) who concluded that this species 90 was not distinct from the type-species of the genus Lasiotocus Looss in Odhner, 1911, L. mulli (Stossich, 1883), at the generic level. Body cylindrical, although tegument is reported to be spinous (Bartoli & Bray, 2004), this was not observed, oral sucker terminal, small. Short pre-pharynx, pharynx small, oesophagus long . Ventral sucker small, laterally flattened. Figure 2.7.21. Lasiotocus typicus stained with borax carmine (x125). Lecithocladium excisum (Rudolphi, 1819) (fig. 2.7.22) Family Hemiuridae Looss, 1899 Genus Lecithocladium Lühe, 1901 Site of Infection: Location of capture: Norwegian coast, 59.00N 03.00E Date: October 2000 Prevalence:2% (1/50) Host length range:32.4cm Description: Body cylindrical, prominent ecsoma 2.8-4.8mm in length. Oral sucker terminal, large, ventral sucker smaller. No prepharynx. Pharynx large, elongate. Testes oval, tandem, position depends upon eversion of ecsoma 91 Figure 2.7.22. Lecithocladium excisum stained with borax carmine (x125). Monascus filiformis (Rudolphi, 1819) (fig. 2.7.23) Family: Fellodistomidae, Nicoll, 1909. Subfamily: Monascinae Dollfus, 1947 Genus: Monascus Looss, 1907 Site of infection: Intestine Location of capture: Gulf of Lyon, 40.43N 01.30E Date: June 2000 Prevalence:12% (6/50) Host Length Range (cm):18.6-20.7cm Description: Body elongate, cylindrical, length 1.8-4.5mm. Oral sucker slightly elongate, terminal, slightly larger than ventral sucker, which is found between one third and one quarter of the way along body. Two rounded testes lie in tandem in hind 92 body, rounded ovary lies in front of testes. Vitellaria lie in two bilaterally symmetrical bands between ventral sucker and hindmost testicle. Figure 2.7.23 (left) Monascus filiformis stained with borax carmine (x125). Figure 2.7.24 (right) Opechona bacillaris stained with acid alum carmine (x125). 93 Opechona bacillaris (Molin, 1859) Dollfus 1927 (fig. 2.7.24) Family: Lepocreadiidae Odhner 1905 Genus: Opechona Looss, 1907 Site of Infection: Location of capture: Gulf of Lyon, 40.43N 01.30E Date: June 2000 Prevalence: 24% (12/50) Host length range:17.5-19.9cm Description: Body cylindrical, elongate, 1.8-4.2mm in length. Oral sucker large, terminal, pharynx large, no pre-pharynx. Ventral sucker slightly smaller, located roughly one quarter of the body length behind the oral sucker. Vitellaria arranged in two bands behind the ventral sucker, merging in the hind body. Testes arranged in tandem in the rear body, behind ovary. Ecsoma absent. Pseudopecoeloides chloroscombri (Fischal & Thomas, 1970) Bartoli, Bray et Gibson 2003 Family Opecoelidae Ozaki 1925 Genus Pseudopecoeloides Yamaguti 1940 Site of infection: Intestine Location of capture: Bay of Biscay, 44.00ºN 01.38ºW Date: June 2001 Prevalence:22% (11/50) 94 Description: Body slender, elongate, length 3-5mm. Oral sucker terminal, large pharynx located immediately behind oral sucker. Ventral sucker large, mounted on conspicuous peduncle, emerging slightly posterior to pharynx. Ovary rounded, smooth, located in mid-body. Testes ellipsoid, smooth, arranged in tandem in postovarian region. Prodistomum polonii (Molin, 1859) (figs 2.7.25-26) Family: Lepocreadiidae Odhner, 1905 Genus: Prodistomum (Molin, 1859) Site of infection: Intestine Location of capture: Balearic Islands, 39.24N 02.28E Date: June 2000 Prevalence:16% (8/49) Host Length Range (cm): 14.3-18.0cm Figure 2.7.25. Prodistomum polonii. Description: Body elongate, oral sucker small, pharynx of similar size, set behind long pre-pharynx. Ventral sucker small, located on the mid-line of the body. Testes rounded, set in tandem, at posterior end of body, ovary smaller, off-set slightly to one side. Vitellaria distributed in two lateral concentrations, reaching from the pharynx to the excretory pore, which is located terminally. This species is specific to the genus Trachurus. 95 Figure 2.7.26 (left) Prodistomum polonii and Figure 2.7.27 (right) Tergestia laticollis both stained with borax carmine (x125). Tergestia laticollis (Rudolphi, 1819) Stossich, 1899 (fig. 2.7.27) Family:Fellodistomidae Nicoll, 1909 Subfamily: Tergestiinae Skrjabin et Koval 1957 Genus: Tergestia Stossich 1899 Site of infection: Intestine Location of capture: southern North Sea, 54.45ºN 06.00ºE Date: May 2001 Prevalence:26% (13/50) Host Length Range (cm):26.5-29.7cm Description: Body cylindrical, elongate, length 1.6-3.6mm, large oral sucker, surrounded by ring of “spine-like” projections. Pharynx large, flanked by continuation of spines along “neck”. Ventral sucker large, located approximately one third of the body length behind oral sucker. Testes located in rear third of body, just behind ovary. Numerous eggs throughout body. This species is specific to the genus Trachurus. 96 2.3.2.5 Acanthocephala Two species of acanthocephalans were found in this study, belonging to the genus Corynosoma. A follow up study by MacDonald (2005) found Rhadinorhynchus cadenati to be a common parasite of horse mackerel from the coast of Morrocco, so this species is also described here. Corynosoma strumosum (Rudolphi,1802) Luhe, 1904 (fig 2.7.28) Family: Polymorphidae Meyer 1932 Genus: Corynosoma Lühe 1904 Site of infection: Intestinal wall Location of capture: Celtic Sea, 55.45N 06.50E Date: June 2000 Prevalence:2% (1/50) Host length range: 38.7cm Corynosoma wegeneri Heinze 1934 (figs. 2.7.29a-b & 2.7.30) Family: Polymorphidae Meyer 1932 Genus: Corynosoma Lühe 1904 Site of infection: Intestinal wall Location of capture: southern North Sea, 51.35N 11.06W Date: April 2001 Prevalence:1% (1/99) Host length range: 31.2cm 97 Description: Both Corynosoma specimens found in this study were in the acanthor stage, found encysted in the mucosal membranes of the intestine. Body yellow/white, approx. 5mm in length, found encysted on the external wall of the intestine. Everted probosces and the extent of body spination allow differentiation of the two species. C. wegeneri has 15-16 rows of 10-11 hooks on its proboscis, while C. stumosum has 18 rows of 10-11 hooks. In both species body spination is restricted to the ventral surface of the anterior body, however genital spination, present in C. strumosum, is absent in C. wegeneri. Figure 2.7.28 Corynosoma strumosum with proboscis retracted, stained with borax carmine (x125). Figure 2.7.29a (right) and 2.7.29b (left) Scanning electron micrographs of the body (left) and proboscis (right) of Corynosoma wegeneri. Scale bars represent 1mm and 0.1mm respectively. (from Campbell, 2000). 98 Figure 2.7.30. Corynosoma wegeneri with proboscis everted, stained with borax carmine (x125). Figure 2.7.31. Everted proboscis of Rhadinorhynchus cadenati (x300). 99 Rhadinorhynchus cadenati (Golvan et Houin 1964) (fig. 2.7.31) Family:Rhadinorhynchidae Travassos 1923 Genus: Rhadinorhynchus Lühe 1911 Site of infection: body cavity and intestine Location of capture: southern Portuguese waters, 37.00N 08.30W Date: February 2000 Prevalence: 3% (2/66) Host length range: 19.9-22.7cm Description: Body, white, cylindrical, between 20mm and 35mm in length. Proboscis armed with 16 rows of 24-26 hooks. The taxonomy of the genus Rhadinorhynchus is poorly defined and somewhat confused (Golvan, 1969), however, these specimens have significantly fewer hooks on their proboscides than the other species reported to infect carangids, Rhadinorhynchus carangis Yamaguti 1939. This, combined with the southern distribution of the species coinciding with the area from which it was described, and the reports of this species by Gaevskaya & Kovaleva (1979) allow us to identify this species with some confidence. 100 2.3.2.6 Cestoda Three species were found, all of them metacestodes. None was common, the most frequently encountered being Grillotia sp., found in 14 fish from a number of samples in the northern Atlantic part of the study area. Another trypanorhynch, Nybelinia lingualis, was found in eight fish from the same region and in one from the Mediterranean coast of France. Fourteen fish in the two Greek samples from 2001 were infected with pseudophyllidean plerocercoids. Grillotia erinaceus (van Beneden, 1858) (fig. 2.7.32) Family: Grillotiidae Dollfus 1969 Genus: Grillotia Guiart, 1927 Site of Infection: Visceral cavity Location of capture: Celtic Sea, 51.35N 11.06W Date: April 2001 Prevalence:11% (11/99) Host length range:20.0-30.9cm Description: Body small, 1.5-2mm in length, contained in thick walled cyst, in intestinal wall. Excysted, the scolex is broader than it is long. Proboscides could not be everted so there are no hook measurements. 101 Figure 2.7.32. Grillotia erinaceus stained with acid alum carmine (x125). Nybelinia lingualis (Cuvier, 1817) Poche, 1926 (fig. 2.7.33) Family: Tentaculariidae Poce 1926 Genus: Nybelinia Poche, 1926 Site of Infection: Visceral cavity Location of capture: southern North Sea, 54.45N 06.00E Date: May 2001 Prevalence: 2% (1/50) Host length range: 29.3cm Description: Length of body 3.3mm, width 1.7mm, scolex length 1.9mm. Proboscides could not be everted so these were not measured and there are no hook measurements. Great variations in the dimensions of the scolex are mentioned by Stunkard (1977). 102 The monophyly of the genus Nybelinia is disputed, with a number of changes and revisions proposed (Palm & Walter, 2000; Pereira & Boeger, 2005). Figure 2.7.33. Nybelinia lingualis stained with borax carmine (x125). 103 2.3.2.7 Copepoda Caligus elongatus Nordmann, 1832 Family: Caligidae Müller, 1785 Genus: Caligus Müller, 1785 Site of Infection: Gills Location of capture: Ionean Sea, 38.15N 20.42E Date: May 2001 Prevalence:2.7% (1/44) Host length range:24.0cm Description: A single specimen of C. elongatus was recovered from a fish in the Ionean Sea, and another off the Norwegian coast. Neither had survived the freezing process in good condition, so identification is tentative. An excellent description can be found in Kabata (1979). 104 2.3.2.8 Isopoda Ceratothoa oestroides (Risso, 1816) (fig. 2.7.34) Family: Cymothoidae Leach, 1818 Genus: Ceratothoa Dana, 1852 Figure 2.7.34. Ceratothoa oestroides cleared with beechwood creosote (X50). Site of Infection: Buccal cavity Location of capture: Ionean Sea, 38.15N 20.42E Date: May 2001 Prevalence: 2% (1/44) Host length range:28.0cm Description: Body divided into one head segment, seven peraeon segments and six pleon segments. Length approx. 1cm. Two large eye spots on head segment. Six pairs of limbs with three articulations. Lower limb segments 'sickle shaped', pointed. Two 105 antennae on head, with six segments on each. Telson rounded, with two pairs of tail appendages bearing numerous hairs. Argulus vittatus (Raffinesque-Schmaltz, 1814) Family: Argulidae Leach 1819 Genus: Argulus Müller, 1784 Site of Infection: Skin Location of capture: Alboran Sea, 36.40N 04.00W Date: May 2000 Prevalence:2% (1/50) Host length range: 22.4cm Description: Body divided into one head segment, seven peraeon segments and six pleon segments. First maxillae consists of two suction cups. Two eye spots on head segment. Length approx. 2cm. Body has a distinctive purple colour, from with the name of the junior synonym, A. purpureus, is derived. 106 2.3.2.9 Nematoda Eight species of nematode were found. Hysterothylacium aduncum was most commonly found in larval form but a few adult worms were also recovered. The other seven species were present as larvae only. Five were members of the genus Anisakis and were identifiable to species level only by the application of multilocus allozyme electrophoresis, as described in Mattiucci et al. (in press). Anisakis spp. (fig. 2.7.35) - Anisakis pegreffii Campana-Rouget & Biocca, 1955 - Anisakis physeteris (Baylis, 1923) - Anisakis simplex sensu stricto(Rudolphi, 1819) - Anisakis typica (Diesing, 1860) - Anisakis ziphidarum Paggi et al., 1988 Family: Anisakidae Skrjabin et Karokhin, 1945 Genus: Anisakis Dujardin, 1845 Figure 2.7.35. Unstained Anisakis sp. photographed under dissection microscope (x30). 107 Site of Infection: Visceral cavity Location of capture: Ubiquitous in many sites. Samples from both Atlantic and Mediterranean have prevalence of 100% Description: Anisakis sp. were identified on the basis of their size and characteristic “watchspring coil” shape. The species of Anisakis identified by multilocus allozyme electrophoresis were: A. simplex sensu stricto, A. pegreffii, A. typica, A. physeteris, A. ziphidarum and Anisakis paggiae (sp.n.). The geographical distribution of each species is described in Mattiucci et al. (in press). Hysterothylacium aduncum (Rudolphi, 1802) (fig. 2.7.36) Family: Anisakidae Skrjabin et Karokhin, 1945 Genus: Hysterothylacium Ward et Magath, 1917 Figure 2.7.36. Unstained Hysterothylacium aduncum larvae, photographed under dissection microscope (x30). 108 Site of Infection: Visceral cavity, intestinal lumen Location of capture:54.45N 06.00E Date: May 2001 Prevalence:100% (50/50) Host length range:20.7-33.7cm Description: Larval and mature Hysterothylacium aduncum were found encysted in the body cavity or free-living in the intestine respectively. Worms were cylindrical, between 0.8cm and 3cm in length. When present in very high densities (over 1000 larval worms were found in some fishes) they formed a dense mat in the viscera. Partial or total melanisation of some worms was noted, suggesting that these worms can persist inside the body cavity for some time. Pseudanisakis sp. (figs. 2.7.37a-b) Family: Acanthocheilidae Wülker, 1929 Genus: Pseudanisakis (Layman et Borovkova, 1926) Mozgovoi, 1951 Site of Infection: Musculature Location of capture: 40.28°N 24.55°W Date: May 2001 Prevalence: 2% (1/50) Host length range (cm): 26.8 Description: A single specimen of a large nematode was found encysted under the skin of the left flank of a fish from the Aegean Sea. This was tentatively identified as a larval Psuedanisakis sp.. This genus parasitizes elasmobranchs in the adult form, 109 particularly rays and skates, and although flatfish and gadoids have been reported as second intermediate hosts, no records of Pseudanisakis infections of horse mackerel have been recorded (Marcogliese, 2002). Figure 2.7.37a & b Mouthparts (left) and anus (right) of Pseudanisakis sp. cleared with glycerine (x125). Pseudoterranova decipiens sensu lato (Krabbe, 1878) Class: Nematoda Incertae sedis Genus: Pseudoterranova Deardoff et Overstreet, 1981 Site of Infection: Musculature Location of capture: 52.50°N 12.00°W Date: March 2001 Prevalence: 3% (1/36) Host length range: 39.0 Description: Pseudoterranova decipiens is easily identifiable by its large size, the presence of a caecum, the absence of an appendix, and its characteristic red colour (Palm et al. 1994). 110 2.4 Discussion 2.4.1 Apicomplexa Coccidian parasites are obligate intracellular Apicomplexans, which infect many marine fish species. Some fish coccidians have considerable pathogenic potential, of particular economic concern in mariculture (Lom & Dykova, 1992), but also in wild fisheries, where infections may reduce the market value of fish (Kent & Hedrick 1985). Little information is available on this group of parasites from European Atlantic waters (Costa & MacKenzie 1994). Small pelagic fishes in European waters are parasitized by a number of Goussia species, most notably Goussia clupearum (Thélohan, 1894) and G. cruciata. Infections of G. clupearum have been reported from the liver of the garfish, Belone belone (L.), blue whiting Micromesistius poutassou (Risso), sardine Sardina pilchardus (Walbaum) and mackerel, Scomber scombrus L. (Dykova & Lom 1981; Daoudi 1987; Lom & Dykova 1992; Azevedo 2001). Conversely, G. cruciata has only been recorded from three trachurid species, T. trachurus, T. mediterraneus and T. picturatus (Lom & Dykova, 1992). Little is known of the epidemiology and pathology of fish Coccidia, however, some species of Goussia are known to be serious fish pathogens (Dykova & Lom, 1981; Kent & Hedrick 1985; Costa & MacKenzie 1994). MacKenzie (1981) found that heavy infections of G. clupearum caused severe emaciation, macroscopic lesions and reduction of the liver in Micromesistius poutassou. In the present study, macroscopic lesions were not observed in livers infected by G. cruciata. Nevertheless, there are reports of heavy infections in which there were large agglomerations of oocysts, with a corresponding decrease and eventual destruction of functional parenchyma and development of an inflammatory reaction (Gestal & Azvedo, 2005). 111 2.4.2 Myxosporea The myxosporeans are microscopic parasites of fishes which are capable of causing significant economic harm, both in cultured and wild fish populations, through host mortality (e.g. Tetracapsuloides bryosalmonae) or product spoilage (e.g. Kudoa thrysites). Recently they have been shown to have alternate life stages as actinosporeans, which infect invertebrates and had previously been thought of as separate taxa. This complex life cycle pattern is poorly understood, and actinosporean stages and their alternative hosts have only been described for a small number of economically significant species, such as Myxobolus cerebralis, which causes “whirling disease” in salmonids, and has an actinosporean stage which requires the oligochaete worm Tubifex tubifex as a host. Life cycles of marine myxosporeans are poorly resolved, with actinosporean stages of a handful of species being found in invertebrate hosts, and no links made to fish hosts (Køie, 2000). The actinosporean stages of the myxosporean species which infect horse mackerel are not known, however this development in our understanding of the life cycles of this group increases their value as biological tags, as their distribution will be restricted to areas where both the fish and invertebrate host co-occur. This is the first comprehensive study of the myxosporean parasites of T. trachurus throughout its geographic range, and reveals a diverse community. Several of the myxosporean species that infect horse mackerel show potential for use as biological tags. Alataspora serenum was originally described from T. trachurus in the Celtic Sea, infecting around 15% of fish (Gaevskaya & Kovaleva, 1979a). It was recorded in the present study in three samples, one site to the north west of Ireland, one in the Celtic Sea and one in the western English Channel, in both 2000 and 2001. Prevalence in all samples was around 15-20%. It was not recorded in samples from the adjacent 112 waters of the North Sea and Bay of Biscay. This finding supports the current management system by suggesting that movements between the Western stock and the Southern and North Sea stocks are limited. A. serenum was also recorded from a single fish (1/46) from the Atlantic coast of North Africa (19º58’N 17º28’W), suggesting that some horse mackerel may be highly migratory. This species was not recorded in the Mediterranean Sea. It could be useful as a tag, indicating mixing of stocks in the north east Atlantic. Alataspora solomoni was originally described infecting the Mediterranean horse mackerel, T mediterraneus, from Quarantine Bay, near Sevastopol, in the Black Sea (Yurakhno, 1988). This is the first record of A. solomoni infecting the Atlantic horse mackerel (T. trachurus), and the first record of infection outside the area of the Black Sea. A. solomoni was only recorded in fish from the samples taken in the eastern part of the Mediterranean Sea. It was not recorded in any of the samples taken in the western part of the Mediterranean, and suggests the existence of discrete stocks in this body of water. Over 450 species of Myxobolus have been described from fish hosts. Of these, the majority have been described from freshwater fishes (Landsberg and Lom, 1991). This is the first record of Myxobolus spinacurvatura infecting T. trachurus, and the first record of this species from the European coast of the Mediterranean Sea. The rarity of this finding suggests that this was an accidental infection. This species does not provide any useful information on stock identity or distribution. An unidentified Myxobolus sp. has been previously recorded in the Adriatic Sea, infecting the livers of mullet, Mugil cephalus, in aquaculture facilities in Italy (Fioravanti et al., 2001). It is unknown whether this was the same species which has been recorded here, but the geographical proximity of this record to our finding and 113 the fact that this species was first described from an infection of M. cephalus (Maeno et al.,1990) would suggest that this is possible. The genus Kudoa Meglitsch, 1947 (Myxozoa: Myxosporea) comprises species which are typically histozoic parasites (Moran et al., 1999). This is only the third report of a Kudoa species being found free in the gall bladder, the other two being Kudoa tachysurae Sarkar & Mazumder, 1983, reported from the three-spined catfish, Arius tenuispinis Day, 1877, and Kudoa haridasae Sarkar & Ghosh, 1991, from the gold-spot mullet, Liza parisa (Hamilton, 1822), both from the Bay of Bengal (Sarkar & Mazumder, 1983; Sarkar & Ghosh, 1991). This species was found infecting horse mackerel from one site to the south-west of Ireland only. Where present, infections were of a very low intensity, which made a detailed description or electron microscopy impossible. It is likely that this finding represents the discovery of a previously undescribed species. This species needs further study and description before any useful information can be extracted from its distribution. Kudoa nova has been reported from T. trachurus, T. mediterraneus and T. capensis, from the Atlantic Ocean, Black Sea and Mediterranean (Kovaleva et al., 1979). In this study, it was most frequently recorded in the sample collected from the African coast, at a prevalence of 6.5%. This agrees with the southerly distribution proposed by Kovaleva et al. (1979) who found infected T. trachurus only along the Morocco-Saharan coast. This species could prove a useful biological tag to study migration of fish from African waters into European seas. A single fish from a sample collected off the Galician coast was found to by infected with K. nova. Again, this supports the idea that some horse mackerel are highly migratory. Despite examining samples from the Bay of Biscay, where it had previously been recorded (Gaevskaya & Kovaleva, 1980), no fish infected with Kudoa quadratum were encountered in this study. 114 Cruz et al. (2003) reported a number of horse mackerel from the coast of Portugal as being infected with a Kudoa sp.. Further sampling should be carried out in this area to determine the identity of this species. If found to be K. nova, it would support the findings of Murta (2000), who proposed a degree of mixing between North African and European horse mackerel stocks. A. serenum, A. solomoni and K. nova all show potential as useful biological tags for stock identification in different areas. A. serenum is most commonly found in fish from the Western stock, A. solomoni has only been found in fish from the eastern Mediterranean, and K. nova shows a more southerly distribution. It is a simple procedure to examine a fish for infections with these species, and examination of greater numbers or of further samples from areas that have not been covered in this study would be an simple and useful contribution to understanding the stock distribution of this species. 115 2.4.3 Monogenea The Monogenea are macroscopic parasites of the external surfaces of fish (although some species, such as Paradiplectanotrema trachuri are found internally) which are capable of causing significant host mortality due to homeostatic imbalances caused by the loss of electrolytes from the attachment points of the monogeneans hooks (Scott & Anderson, 1984). They have a direct life-cycle, which means they are transmitted most easily when fish are found in dense shoals, such as when gathering to spawn or feed (Trouvé et al., 1998). Pseudaxine trachuri is reported to be a widely dispersed species (Radujkovic & Euzet, 1989) It has been reported from other species of Trachurus (Radujkovic, 1986) and also from the Bogue (Boops boops) from the coast of Granada and the Ligurian Sea (Lopez-Roman, 1973). Gastrocotyle trachuri is also widely distributed and in this study was generally more abundant than P. trachuri. Jones (1933) noted that infection with monogeneans and copepods was essentially mutually exclusive. That finding could not be supported by this study due to the low level of copepod infections that were found. Llewellyn (1959, 1962) found evidence for seasonality and changes in infection with age. His data showed infections of G. trachuri and P. trachuri were more common in the spring, that adult helminths produced fewer larvae in AugustSeptember when the behaviour of horse mackerel alters in a way which does not facilitate transmission of parasites, and that infections were more common on younger fish, with very low levels of infection in fish over three years of age. Radha (1971) suggests that the change in dynamics of gastrocotylid infection with age of fish is not due to the clamps being optimally suited to a single size of gill filament but rather due to some acquired immunity which takes time to develop. 116 Paradiplectanotrema trachuri is an interesting monogenean due to its way of life as a mesoparasite. This species was found in the pharynx and occasionally the stomach of horse mackerel. The distribution of P. trachuri is restricted to the Mediterranean Sea, and the centre of abundance is the western samples, particularly the southern Ligurian Sea, although it was also recorded in fish from the Gulf of Lyon. The direct life cycle of this species would make it a useful tag, although its relative rarity restricts its utility. 117 2.4.4 Digenea The Digenea is a subclass within the Platyhelminthes consisting of parasitic flatworms with a syncytial tegument and, usually, two suckers, one ventral and one oral. They are particularly common in the digestive tract, but occur throughout the organ systems of all classes of vertebrates. Once thought to be related to the Monogenea, it is now recognised that they are closest to the Aspidogastrea and that the Monogenea are more closely allied with the Cestoda. Around 6000 species have been described to date. In general, the life-cycles may have two, three, or four obligate hosts, sometimes with transport or paratenic hosts in between. The three host life-cycle is probably the most common. In most species the first host in the life-cycle is a mollusc. This has lead to the inference that the ancestral digenean was a mollusc parasite and that subsequent hosts have been added by terminal addition (Cribb et al., 2003). In all digenean life cycles, alternation of generations is an important feature. This phenomenon involves the presence of several discrete generations in one lifecycle. The genus Monascus Looss 1907 formerly contained numerous species of similar morphology, recognised solely on the basis of their host species. These have been synonymised by Bray and Gibson (1980). M. filiformis appears to be cosmopolitan (Gaevskaya & Kovaleva, 1980; Gaevskaya et al., 1985), and favours carangid fishes (Gaevskaya, 1990). Martorelli and Cremonte (1998) proposed a three host life cycle for this species, involving a mollusc and a chaetognath. These were species found in the south-west Atlantic ocean, however a similar pattern is likely to hold true throughout the range of the species. Tergestia laticollis is a parasite predominantly of T. trachurus, but has also been recorded from the Atlantic mackerel, Scomber scombrus (Bray & Gibson, 1980). Its life cycle is not well recorded, with sporocysts recorded in benthic bivalve 118 molluscs (Angel, 1960), and cercaria being found free-living in the water column (Dubois et al., 1952). The only record of a metacercaria in the genus Tergestia is of T. agnostomi, reported from the gastrovascular system of the ctenophore, Pleurobrachia pileus (Boyle, 1966). In this study, Pseudopecoeloides chloroscombri was only found at site 21 in the south eastern corner of the Bay of Biscay. Although the species has been well described, details of its life cycle are absent. Its restriction to the one area in this study is somewhat surprising, given the wide range it has been recorded from. The species was described from a single specimen recovered from the coast of Ghana (Fischal & Thomas, 1970) and has been recorded from the Mediterranean Sea, around Corsica (Bartoli et al., 2003), although a record from the coast of Libya is doubtful (Al Bassel, 2000). In sample 21, the prevalence of this species was 22%. The description of the species based on a single specimen from the African coast would support the hypothesis that some horse mackerel are highly migratory. Lecithocladium excisum was first recorded by Rudolphi (1819) from the mackerel, Scomber scomber, in the Mediterranean Sea. It has since been reported from a range of fishes from the Mediterranean and Black Seas and the northeast Atlantic region, as far north as the Faroe Islands and as far south as Namibia (Gibson & Bray, 1986) . A number of intermediate hosts have been described, including copepods of the genera Acartia, Paracalanus, Pseudocalanus, Eurytemora and Oithona, the ctenophore Pleurobrachia pileus and the holoplanktonic polychaete Tomopteris helgolandica (Koie, 1991). This diversity of hosts and wide geographic spread suggests that this species is unlikely to be a useful biological tag. Derogenes varicus is widely acknowledged as being one of the most cosmopolitan of all parasites (Køie, 1979). It has a three host life cycle involving a mollusc, a copepod and a teleost fish. The great range of host species from which this 119 parasite has been recorded, and it’s global distribution, suggest that it is unlikely to be a useful tag species. Ectenurus lepidus is another widely distributed parasite of carangid fishes, reported from the Mediterranean Sea, north and south Atlantic Ocean and the Indian Ocean (Bartoli et al., 2005; Manter, 1947; Daporte et al., 2006; Bray, 1990). Again, the lack of host specificity and wide geographic range limit the usefulness of this species as a biological tag. 120 2.4.5 Cestoda Larval cestodes, due to their mode of infection and the general absence of a pathogenic response from the host, make excellent biological tags for a wide range of hosts. The plerocercus stage of Grillotia erinaceus has been proposed as a biological tag for both the haddock (Lubieniecki, 1977) in the northern North Sea and north Atlantic Ocean, and whiting (Shotter, 1973) in the Irish Sea, while plerocercoids of Nybelinia spp. have been used as tags for halibut in the northern Pacific ocean (Blaylock et al., 2003), arrow squid around New Zealand and Pacific salmon, Onchorhynchus keta, from South Korea (Kim et al., 2005). This study found a much less diverse community of cestodes infecting the horse mackerel than that reported by Gaevskaya and Kovaleva (1979). Many cestodes infect elasmobrachs as their final host (Keeney & Campbell, 2001). The loss of cestode diversity in horse mackerel since the work of Gaevskaya and Kovaleva in the late 1970s may be a reflection of the general downward trend observed in the abundances of most elasmobranchs in ICES waters over the intervening period, with some species being placed on the International Union for the Conservation of Nature and Natural Resources (IUCN) Red List (Anon., 2006). 121 2.4.6 Acanthocephala The Acanthocephala is a phylum of parasitic worms, characterised by the presence of an evertable proboscis, armed with spines, which it uses to pierce and hold the gut wall of its host. Acanthocephalans typically have complex life cycles, involving a number of hosts, including a crustacean first host, fishes, amphibians, birds, and mammals. About 1150 species have been described (Perez-Ponce de Lyon, 2000). Acanthocephalans are highly adapted to a parasitic mode of life, and have lost many organs and structures through evolutionary processes, including any mouth or alimentary canal – a feature they share with the Cestoda, although the two groups are not closely related. Adult stages live in the intestines of their host and uptake nutrients which have been digested by the host directly through their body surface. Their degenerative nature makes determining relationships with other higher taxa through morphological comparison problematic. Phylogenetic analysis of the 18S ribosomal gene has revealed that the Acanthocephala are most closely related to the rotifers or may even belong in that phylum (Garey et al., 1996; 1998). In this study, a small number of acanthocephalan species were found infecting horse mackerel, although within the study area these were rare. A follow up study by MacDonald (2005) found the acanthocephalan, Rhadinorhynchus cadenati, to be a useful tag for discriminating the southern boundary of the putative southern stock, off the Atlantic coast of north Africa. This was the only acanthocephalan recorded by Gaevskaya and Kovaleva (1980), who proposed it as a useful biological tag. Other species of Corynosoma have been reported from other Trachurus species, such as Corynosoma australis and Corynosoma obtuscens from Trachurus murphyi on the Pacific coast of South America (Oliva, 1999; Montalvo, 1977). This is the first record of both C. strumosum and C. wegeneri from T. trachurus. 122 Oguz and Kvatch (2006) reported infections of T. trachurus with the pomphorhynchid, Longicollum pagrosomi on the Turkish coast of the Mediterranean basin. Their specimens from the Sea of Marmara were smaller than those described by Yamaguti (1935) from Trachurus japonicus in Japanese waters. This is an unusual finding as this species has not been recorded elsewhere in the range of T. trachurus. 2.4.7 Crustacea The Crustacea are a subphylum of arthropods containing around 52,000 extant species (Martin & Davies, 2001). The vast majority of crustaceans are free-living and motile, but a number of taxa are parasitic, some of which are highly degenerate to the point that their taxonomic relationships become obscured, particularly in their adult stages. Of particular interest are the class Isopoda and the subclass Copepoda. In this study, both parasitic isopods and copepods were rare. This contrasts somewhat with the findings of Gaevskaya and Kovaleva (1980) who found a number of horse mackerel parasitised by various species of Lernaocera in the North Sea. Gaevskaya and Kovaleva did not record any infections with isopods, and these were also very uncommon in this study. 123 2.4.8 Nematoda The nematodes represent a large group of metazoan invertebrate animals noted for the presence of a secondary body cavity (pseudocoel) and a complete or incomplete digestive system in all stages of development, differing in this respect fundamentally from related groups such as the Nematomorpha. The body is elongate, vermiform, cylindrical, covered with a well developed cuticle without cilia, protonephridia, respiratory organs or blood system. The nematodes show a wide range of ecological adaptation and most are free living. Of around 16,000 described species, around 40% are animal parasites (Moravec, 1994). Anisakis simplex is one of the most heavily studied of all fish parasites, partly due to its geographical ubiquity and partly due to its human health implications. It reaches maturity in the intestines of cetaceans, and can infect humans easily through the consumption of raw or undercooked seafood. Furthermore, it releases characteristic proteins into the flesh of fish which are capable of producing severe allergic reactions in people sensitive to nematodes (Audicana et al., 2002). Smith (1983) concluded that euphausids in the north-east Atlantic and northern North Sea, and perhaps universally, were the major intermediate hosts of A. simplex. The distribution of A. simplex therefore, was thought of as strongly influenced by the distribution of euphausids, for instance, they are not found in the Baltic Sea, and herring infected with A. simplex in this reason represent a migratory sub-population which feeds outside the Baltic and returns to spawn (Grabda, 1974). Klimpel et al. (2004), however, found that the oceanic life cycle of A. simplex in the Norwegian Deep to be very different in terms of the hosts and proposed life cycle patterns of A. simplex to that which has been proposed from other regions. In contrast to earlier suggestions, euphausids were not required for the successful transmission of A. simplex in the Norwegian Deep. This demonstrates that this 124 nematode is able to utilise different host species depending on the locality, apparently having a very low level of host specificity. This could explain the wide range of different hosts that have been recorded for this species, and can be seen as the reason for the success of this parasite in reaching its marine mammal final hosts in an oceanic environment. It is in the final cetacean host that host specificity becomes important for Anisakis spp.. Mattiucci and Nascetti (2006) summarised their applications of genetic techniques, resulting in the recognition of nine species of nematodes belonging to genus Anisakis, namely A. pegreffii, A. simplex s.s., A. simplex C, A. typica, A. ziphidarum, Anisakis sp., A. physeteris, A. brevispiculata and A. paggiae. These are, to a greater or lesser extent, specific to certain groups of cetaceans. For example, A. ziphidarum is specific to the beaked whales (Ziphiidae) (Paggi et al., 1998). Genetic differentiation and phylogenetic relationships inferred from allozyme and mitochondrial (cytochrome oxidase I) markers, are phylogenetically consistent and depict two main clusters, one encompassing the species A. pegreffii, A. simplex s.s., A. simplex C, A. typica, A. ziphidarum and Anisakis sp. and the second including A. physeteris, A. brevispiculata and A. paggiae (Valentini, 2006). Comparison of phylogenetic relationships among Anisakis spp. with those currently available for their cetacean definitive hosts suggests parallel evolution and cospeciation between host and parasite (Mattiucci & Nascetti, 2006). Anisakis spp. have previously been reported to infect Trachurus trachurus from a number of locations in the North Atlantic Ocean (Adroher et al., 1996; Abaunza et al., 1995), and the Mediterranean Sea (Manfredi et al., 2000), and also from T. picturatus in the South Atlantic Ocean (Mattiucci et al., 2002) and T. murphyi (Oliva, 1999), T. novozelandiae (Hurst, 1984) and T. japonicus (Itagaki & Ishimaru, 1967) in the Pacific Ocean. 125 Gaevskaya and Kovaleva (1980) found A. simplex at all of their stations, and did not propose it as a useful tag species. However, Abaunza et al., (1995) noted variations in abundance of A. simplex from samples collected around the Iberian Peninsula, and suggested that this species would be a useful biological tag. Hysterothylacium aduncum lives as a larva in various marine invertebrates and fishes. Its adult stage resides in the digestive tract of marine teleosts. H. aduncum is very common in the viscera of fishes in the North Atlantic and adjacent seas (Koie, 1993). At least one intermediate host, a crustacean, is obligatory for the transmission of H. aduncum, and although polychaetes and gastropods have been experimentally infected, this is unlikely to occur in the wild (Popova & Valter, 1968). The third stage larva is infective to fish, and can either moult and mature to an adult stage, or penetrate the intestinal wall and remain an L3 larvae, which is capable of considerable growth (Koie, 1993). Infections of a number of species of Trachurus with H. aduncum have been reported on a number of occasions, such as and T. murphyi (Cabrera and SuarezOgnio, 2002; González et al., 2006). Gaevskaya and Kovaleva (1980) found this nematode in horse mackerel at all of their stations, but did not propose it as a biological tag. Pseudoterranova decipiens is a cosmopolitan species, occurring in many aquatic hosts, from boreal seas of the northern hemisphere to the Antarctic Ocean (Kerstan 1991; Palm et al. 1994, Palm, 1999). P. decipiens matures in seals, and the first intermediate hosts are crustaceans, such as copepods, amphipods, shrimps, and isopods (Marcogliese 1996). Larval worms can infect fish, which serve as second intermediate hosts (McClelland 1995), once they reach a size of 2mm (Kerstan 1991). This size is attained only in macroinvertebrates, such as mysids (Jackson et al. 1997), which are 126 an essential step in the life cycle of P. decipiens (McClelland 1995). P. decipiens is considered to follow a benthic life cycle, with larvae not being able to swim (Palm, 1999), and infection of horse mackerel with P. decipiens reflects a certain amount of benthic feeding in their diet. Pseudanisakis spp. are predominantly parasites of elasmobranchs. Their lifecycle is poorly studied, but Gibson (1973) notes that related species such as Anisakis simplex and Contracaecum aduncum utilise a crustacean first intermediate host and a teleost second intermediate host. Gibson goes on to say that the lack of an authenticated record of a third stage Pseudanisakis spp. larvae from a teleost fish suggests that a teleost host is not involved in the life-cycle. This finding marks the third time a specimen of Pseudanisakis spp. has been recorded from a teleost, the others being from a tub gurnard (Trigla hirundo) off the Belgian coast (Punt, 1941) and from a blue and yellow grouper (Epinephelus flavocaeruleus) from Kenyan waters (Morgans, 1964). The paucity of such records suggests that these have all been accidental infections. 127 2.5 Conclusion The parasite fauna of the horse mackerel is informative on a number of aspects of its life history, its ecological niche and marine food-webs in general. 2.5.1 Ecological Implications Marcogliese (2002) considered helminth parasites of fish in marine systems to be generalists, lacking host specificity for both intermediate and definitive hosts. He pointed out that many parasites in the marine environment possess life-cycles which involve long lived larval stages utilising intermediate and paratenic hosts, and believed these to be adaptations to the long food chains and the low densities of organisms distributed over broad spatial scales that are characteristic of open-ocean marine systems. These features are evident in the parasitofauna of the horse mackerel, evinced by the large number of generalist species reported from numerous other hosts, and the large abundances of larval nematodes using the horse mackerel as a paratenic host. Marcogliese (2002) predicted that such features of helminth parasites should lead to the homogenization of parasite communities within a fish species, yet pointed out that there can be significant differences in the parasite fauna of a species on a spatial scale. Parasite distributions are superimposed on distributional patterns of freeliving animals that participate as hosts in parasite life cycles (Marcogliese, 2002). This superimposition occurs as a result of the lateral partitioning of a host population by patches of unsuitable habitat or barriers to free movement, leading to the segregation of parasites among fish hosts. Within each suitable habitat patch, resource partitioning, density dependent effects and dietary preferences further contribute to the establishment of distinct parasite assemblages (Arneberg et al., 1998). 128 Patterns of transmission in marine systems should be interpreted in the context of food web structure. Fish parasite communities are influenced by, and influence in return, food web structure. Some parasites limit the abundance of their host species through induced mortality of heavily infected individuals (Lester, 1984). Others facilitate their transmission through infection-induced changes in behaviour resulting in increased predation of infected individuals (Lafferty & Morris, 1996). The parasite fauna of the horse mackerel examined in this study is numerically dominated by larval stages of endoparasitic nematodes, encysted within the body cavity. Anisakis spp., the most numerous parasite recorded, matures in marine mammals. This, combined with the presence of other species which mature in other marine mammals such as Corynosoma spp., suggests that the horse mackerel plays an important role in food webs involving marine mammals. Generally, there was a relative absence of larval stages of parasitic helminths which reach maturity in piscivorous fishes. The exception to this rule were samples collected in the southern North Sea, which had very low abundances of Anisakis spp.. This could be related to the scarcity of cetaceans in the North Sea, however Anisakis spp. is known to also use seals as a definitive host (Brattey & Stenson, 1993), which are abundant in the area (McConnell et al., 1999), or to the absence of euphausids. owever, as Klimpel et al.(2004) showed, these are not always necessary for the successful completion of the anisakid lifecycle. The rarity of anisakids in these samples, coupled with high abundances of larval Hysterothylacium aduncum, which matures in piscivorous fishes, suggests that horse mackerel in the North Sea occupy a very different niche to that which they occupy in the wider Atlantic ecosystem. 129 2.5.2 Nestedness Nestedness is a measure of order in an ecological system, referring to the order in which the number of species is related to area or other factors. The more a system is "nested" the more it is organised (Atmar & Patterson, 1993). In a nested system, the fauna recorded in depauperate communities is a subset of the fauna recorded in Species relatively more diverse communities. Sample Figure 2.8 Presence/absence matrix of species from horse mackerel samples. A black mark represents a species being found in a sample. Results reveal a small group of core species found in almost all samples, and a larger number of species found in a few samples. The community is far from a nested distribution (dashed line) which suggests that community composition is essentially random. 130 The geographical variation in parasite community structure among populations of the same host species remains one of the least understood aspects of parasite community ecology, however, a parasite community which displays nestedness implies that the factors acting to influence the structure of the community are not random (Timi & Poulin, 2003). The degree of nestedness in the community composition of the horse mackerel examined in this study is shown in figure 2.5. In a nested system, the upper quadrant would be black, while the lower one would be white. This is clearly not the case here. There are a small number of species recorded in nearly all samples, while the occurrence of other species seems to be randomly distributed. There is no evidence of co-occurrence or mutual exclusion of certain pairs or groups of species. This pattern of community composition suggests that the factors which influence its composition are essentially random. The lack of depauperate communities towards the geographical extremities of the range of T. trachurus could be seen as evidence of their highly migratory nature, and also could be related to their ecological replacement at the extremes its distribution by other members of the genus Trachurus, which presumably play host to very similar parasite faunas. 2.5.2 Conclusions The horse mackerel is evidently host to a diverse parasite fauna, which, when compared with the findings of Gaevskaya and Kovaleva (1980), remains temporally stable in its distribution. Few of the species recorded are known to be pathogenic to their hosts, and no pathogenic changes with infection were noted here. Most species show significant differences in their abundance or prevalence, which fulfils the most basic criterion for the selection of species as biological tags. The distinct regional differences in parasite fauna suggests that the application of parasites as biological tags to problems in stock identity of the horse mackerel will be informative. 131 The parasite fauna of the horse mackerel is informative on a number of other aspects of its life history. The large number of larval taxa, including cestodes, nematodes and acanthocephalans, suggest that this species plays an important role in the transmission of these groups to their final hosts – piscivorous fishes or marine mammals (Marcogliese, 2002). The regional differences observed in larval parasite faunas may therefore represent guilds of parasites transmitted in different regional food-webs to the most likely predator of horse mackerel in each particular area – for example, Anisakis spp. to cetaceans in the north east Atlantic area, Hysterothylacium aduncum and Corynosoma spp. to fishes and seals in the southern North Sea and pseudophyllideans and Pseudanisakis spp. to elasmobranchs in the eastern Mediterranean Sea. Numerous studies suggest that parasite diversity is related to the diversity of organisms in the host’s diet (Campbell et al., 1980; Scott, 1981; Campbell, 1983; Hemmingsen & MacKenzie, 2001). The diverse parasite fauna of sixty-eight species found in a relatively poorly studied species such as the horse mackerel therefore suggests that the horse mackerel is a generalist feeder. This agrees with the findings of Murta et al. (1993), Olaso et al., (1999), Murta and Cabral (2002) and Jardas et al. (2004) who all found a diverse range of items in the diet of horse mackerel, as well as the observations of stomach contents from this study, which was dominated by planktonic crustaceans, but also included, amongst other things, benthic gastropods, fish eggs and larvae, Ammodytes spp., juvenile horse mackerel and various cephalopods. The complete absence of nestedness is an interesting feature, and may be related to the distribution of the horse mackerel, in relation to congeneric species. The horse mackerel is ecologically replaced at its southern and eastern extremities by other members of the genus Trachurus. Numerous studies have suggested these species host a parasite fauna very similar to that recorded from the horse mackerel 132 (Hewitt & Hine, 1972; Piaseki, 1982; Gaevskaya & Kovaleva, 1985). Therefore, the geographical range of the horse mackerel is not necessarily the boundary of the species which parasitise it, and a search for nestedness may produce different results if samples from other members of the genus Trachurus were included. 133 Chapter 3. Parasites as biological tags of stock identity in the horse mackerel, Trachurus trachurus (L.). 3.1 Introduction Fisheries science is based on the notion of an idealised “unit stock”, a discrete entity with its own demographic parameters, which are independent of surrounding conspecific populations (Waldman, 2005). Fish stock assessment is the science of estimating certain key population dynamic parameters to enable better management strategies to be applied to the unit stock. If rates of growth, natural mortality, reproduction or fishing mortality of a closed population can be estimated, their effects can be modelled to predict future population structure under various fishing regimes. An understanding of stock structure is necessary for designing appropriate management regulations in fisheries where multiple stocks are differentially exploited (Ricker, 1979). This study of horse mackerel stock identity was carried out under the aegis of the HOMSIR project. This project attempted to apply a range of techniques to the same set of fish. These included a range of genetic markers, a study of the potential for parasites to act as biological tags, investigation of body and otolith morphometry and life history traits of the same set of fish, as well as some physical tagging. 3.1.1 Parasites as biological tags for stock identification studies In the century since the first application of the term “stock” in the sense that we understand it today (Heinke, 1898), the toolkit for identifying what is and is not a stock has grown considerably larger, more technical, and specific to particular life stages relevant to the stock identity question being asked, as evinced by the growth in publications dealing with the topic. The inherent individual variability of the components of a population is an important factor to bear in mind when carrying out comparative studies in stock identity. Some uncertainty can be eliminated if all the 135 techniques are applied on the same fish. Using this design, the results from different techniques are comparable, and if one technique finds a specimen which differs from the rest of the specimens in a sampling unit, limited migration from one stock to another could be hypothesised. This result could be totally misleading if the different techniques are not applied to the same material. The use of parasites as tags for the identification of stocks is a technique which has been applied in fisheries science for over sixty years (Herrington et al., 1939), and many original works and reviews of the field have been published since (e.g. Kabata, 1963; MacKenzie, 1983; 1987; Lester, 1990; Williams et al., 1992; etc.) establishing this as one of the most widely applicable methods of stock discrimination. Like other methods, it has a number of advantages and disadvantages: Low cost - Unlike genetic or micro-chemical analyses, parasitology does not require expensive consumable materials. The main cost associated with the method is staff-time, both to examine the material being studied, and also to gain sufficient expertise in parasite identification. Elimination of behavioural changes – Fish tagged with artificial tags are thought to behave differently from their untagged peers. Tagged fish also need to be released and recaptured, with a low rate of recapture greatly reducing the sample size. The use of parasites as tags circumvents this problem as each fish examined is a valid observation. Resistance to gene flow – Parasite infestation levels are determined by the presence and/or relative abundance of the organisms required to successfully complete the life-cycle of the species in question. The integration of these 136 factors mean that parasite prevalences and abundances are likely to be temporally stable in a given area, while a mixing rate of 2% between population is sufficient to obscure genetic differentiation. Lack of information on parasite ecology – As mentioned previously, parasite infestation levels are determined by the ecological parameters of the host species required for successful completion of the life-cycle. These are often unknown, making hypothetical predictions rather difficult. As our knowledge of the biology and ecology of marine parasites increases, their use will become more efficient (MacKenzie & Abaunza, 2005). Mortality of study specimens – It is difficult to conduct a parasitological examination without inducing the mortality of, or at least significant stress to, the organism being studied. This makes the use of parasites as biological tags unsuitable for studies of stock identity in populations whose existence is threatened such as endangered species, or studies of sub-populations who exist on an extremely small scale, when non-lethal methods, such as the collection of genetic material or of morphometric variables are more suitable. 3.1.2 Criteria for selection of tag species A number of criteria have been proposed for the selection of suitable parasite tags (e.g. Kabata, 1963; Sindermann, 1983; MacKenzie, 1987; Williams et al., 1992). The most recent are proposed by MacKenzie and Abaunza (2005), and acknowledge that parasites which fulfil all of these criteria are rarely encountered, so usually a compromise has to be made. MacKenzie and Abaunza’s proposals were as follows: 137 1. A parasite should have significantly different levels of infection in the subject host between different parts of the study area. Infection data, such as prevalence, intensity and abundance of infection, should be measured in terms as defined by Bush et al. (1997). 2. A parasite tag should persist in the host for a period of time relevant to the nature of the study. 3. Parasites with single-host life cycles (monogenetic trematodes, parasitic crustaceans, etc.) are the simplest tags to use. Parasites with complex lifecycles, such as cestodes or digeneans, tend to produce results which are more difficult to interpret due to the fact that they utilise more than one host, therefore variations in their levels of infection may relate to more than the biotic and abiotic factors which influence transmission of the parasite to the host of interest. 4. The level of infection should remain relatively temporally stable (assuming the host population is also remaining temporally stable), although year to year variations can be overcome by considering parasitic infections of single year classes as they make their way through the cohorts. 5. Parasites should be readily identifiable, with the minimum of dissection, to prevent time from becoming a limiting factor in the study. 6. Pathogenic parasites should be avoided as they are likely to be selected out of a population, as are those which can be shown to alter host behaviour. 138 A number of statistical treatments have been applied to these data sets, ranging from the identification of single species which vary in prevalence, abundance or intensity between two putative stock to more complex assessments treating each fish as a “habitat” and considering the parasite fauna of each fish as an ecological assemblage of species, and applying a more complex method, such as discriminant analysis (Lester et al. 1985), or investigating numerous samples, both temporally and spatially, throughout the geographical range of the species, such as the study of narrow banded Spanish mackerel (Scomberomorus commerson) in Australian waters by Lester et al. (2001). 3.1.3 Stock identity of horse mackerel Fisheries on small and medium-sized pelagic fish are, in terms of catch volume, the world’s most important fisheries, providing food supply, employment, and an exportable resource on a global scale (FAO, 2003). Amongst these small and medium sized pelagic fisheries, those operating on the horse mackerels (Carangidae) are second-most in importance after those exploiting the herring-like clupeids (Bailey, 1992). Catches of Atlantic horse mackerel, Trachurus trachurus, in the north-east Atlantic have been over 500,000t per annum in recent years (FAO, 2003). In contrast with the clupeids, their natural pattern of variation in abundance is characterised by a “steady state” situation, so stocks which are exploited in a sustainable fashion produce catch levels without significant variation from year to year (Caddy and Gulland, 1983; Bass, 1995). This steady-state property can change if the pattern of fishing or pattern of recruitment changes, as can be observed in the historical development of the 1982 year-class of the “Western stock” in the Northeast Atlantic horse mackerel (Anon., 139 2004), which was abnormally large and has supported the fishery for over 20 years. To properly assess the fished resource, a comprehensive knowledge of the stock structure is required (Gulland, 1971; Hilborn & Walters, 1992). In the north east Atlantic area, horse mackerel stocks have been defined by ICES (Anon., 1992), mainly on the basis of observation of egg distributions in plankton surveys over the spawning period, resulting in the stock being assessed as three separate stocks; the Western stock (North-east continental shelf of Europe, from France to Norway); the North Sea stock (central and southern North Sea) and the Southern stock (the Atlantic waters of the Iberian peninsula) (See figure 3.1). Particular problems have been encountered when assessing the Southern stock. Due to uncertainties in the range of the stock, it has been difficult to aggregate catch data and survey information to produce the tuning data on which an analytical assessment is based (Anon., 2004). Horse mackerel are a migratory species and whilst the distribution of the Western and North Sea stocks overlap partly during over-wintering in the English Channel (Macer, 1977), their spawning and feeding areas seem to be separated (Eltink, 1992). Information on stock identity and mixing at the boundary between the Western and Southern stocks is scarce. The southern limit of the Southern stock is arbitrarily defined as the southern boundary of EU Atlantic waters. 140 Figure 3.1.- Horse mackerel (Trachurus trachurus) stock boundaries in the Atlantic ocean, as defined by ICES in 1992 (Anon. 1998), 200 m depth contour drawn. (pers. comm.. C. Zimmermann, Bundesforschungsansalt fur Fisherei, Hamburg). In the Mediterranean Sea, biological information on stock identity is rare. Some workers have suggested that there is little useful information to be garnered from landings statistics from these areas (Csirke, 1995), however Caddy (1998) suggested a separation of stocks based on areas defined from commercial landings data. These have not been officially adopted and there is currently no clear separation or assessment of horse mackerel stocks in the Mediterranean. Although the General Council for Fisheries in the Mediterranean (GCFM) maintains an interest in Mediterranean fisheries, it does not take an active role in the assessment of horse 141 mackerel populations or the management of their fisheries. In the Central Eastern Atlantic, there are considered to be three stocks of T. trachurus, extending down the African coast as far as the Cape Verde islands, where the species is replaced ecologically by the Cape horse mackerel, Trachurus capensis. The largest and, from a fisheries perspective, most important of these is the SaharaMauritanian stock (Maxim, 1995), found along the coasts of Morocco and Mauritania and around the Canary Islands. It has long been recognised that doubts over the current stock definitions, uncertainties in migratory behaviour and the lack of biological information to support such stock units are significant factors hindering the sustainable long term exploitation of horse mackerel (Anon., 1992; 1998). Publications dealing with the stock structure of horse mackerel have been very few and far between, and have addressed only a small part of the stock distribution, or the information provided has been so scant that it is not possible to adequately delineate subpopulations. Some workers found differences between areas in the North-east Atlantic using multilocus allozyme electrophoresis (Nefedov et al, 1978) whilst others applying the same approach did not (Borges et al., 1993). In the putative “Southern stock” there are some works dealing with the analysis of differences in infestation with anisakid nematodes (Abaunza et al., 1995; Murta et al., 1995). Information on the fisheries, general biology and physiology of the horse mackerel is much more abundant, and forms the basis for the interpretation of the results produced by this study. Publications on distribution and fisheries include Borges et al., 1996; Villamor et al., 1997 and Iversen et al., 2002. A comprehensive review of growth and reproduction is found in Abaunza et al., 2003 (and references therein). Publications on diet include those of Ben-Salem (1988), Olaso et al., (1999) and Cabral and Murta (2002); other aspects of their physiology can be found in Macer (1977), Kerstan (1988) and Temming and Herrmann (2001a, b). 142 3.2 Methods 3.2.1 Sampling design The design of sample collection for a stock identity study has tremendous influence on the ability of the study to discriminate stocks. One of the key problems to be considered in the design of a study is that of pseudoreplication, that is, given that sample collection removes fish from the system, and that the fish sampled are in constant motion it is not possible to collect truly independent samples. Annual replication of sample collection aims to solve this problem by collecting multiple samples from the same site in different years. This relies on a certain amount of fidelity in migratory and spawning behaviour, and stock mixing in a sampling area can produce very variable results, however this is the best approach to gain meaningful information on stability of stock structures from year to year. Sampling for this project was designed to investigate horse mackerel stock structure throughout its entire geographical range, using a range of stock identification techniques on the same specimens, including an investigation of parasites as biological tags. The sampling programme covered most of the whole distribution area of Trachurus trachurus, only omitting the middle-eastern sector of the Mediterranean Sea, the Mediterranean coasts of North Africa and the Atlantic coast of central West Africa. To obtain a better discrimination of stocks the sampling programme was constructed on the basis of a priori information on stock units, distribution area and biology. Some areas in which there was almost no prior information available were sampled, such as the Atlantic waters of North Africa. In general, sampling focussed on western European waters and the European coasts of the Mediterranean Sea, partly due to the interest in improving knowledge in areas where active programmes for assessing and managing horse mackerel fisheries are implemented, and partly due to the political difficulties involved in obtaining samples 143 from African and middle-eastern waters. Samples were collected at twenty sites throughout the north east Atlantic and Mediterranean Sea (see figure 3.2 and table 3.1). Besides consideration of adequate geographical coverage, critical aspects to consider in the construction of a sampling programme to meet the objectives of the study were: the number of specimens taken at a sampling site, the temporal replication of sample collection, and the selection of an appropriate season to collect specimens which is relevant to their spawning behaviour. One hundred fish per-sampling site, per-year, during each of the two years (2000 and 2001) considered by the study was deemed an adequate sampling strategy to obtain significant results with most of the techniques used in the project as a whole. Due to limitations on time taken to carry out a parasitological examination, only the first fifty of these fish were used for the parasitological study. The aim was to carry out sampling during the local spawning season, on the basis of prior information available about the spawning time. This meant that it was possible to relate the sampled specimens to their spawning grounds and to ensure that the stocks we were investigating were discrete spawning entities. The temporal replication of samples allowed us to explore the possibility of temporal changes in migratory or spawning behaviour. 144 Figure 3.2. Realised sampling site positions for the EU-project HOMSIR in 2000 (circles) and 2001 (triangles). (pers. comm. C. Zimmermann, Bundesforschungsansalt fur Seefisherei, Hamburg). The seasonality of the fleets working in ICES division IVc, the southern North Sea, meant that it was not possible to sample the eastern English Channel/southern North Sea boundary. To attempt to fill this gap, sampling site 06, initially intended to be collected in the Bay of Biscay, was moved to the southern side of the English Channel, as close as possible to division IV. This reallocation meant that there was a large area in the Bay of Biscay without sampling coverage. An additional sampling site was therefore added later in the study in the southern Bay of Biscay and designated as sample 21. 145 Sampling site 18, initially located off the north coast of Algeria, was moved to the south coast of Sardinia, partly due to the difficulties in maintaining contacts with the appropriate Algerian authorities, which implied a risk committing resources to sample collection, then losing the sampling opportunity at the last minute due to political difficulties, partly due to concerns about the availability of Algerian on-shore cold storage facilities and partly due to difficulties in finding a reliable courier service to transport the specimens out of the country. Area Code 1 2 3 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Location (2000) Month 59°00’N 03°00’E October 52°00’N 11°20’W June 49°30’N 10°15’W July 55°45’N 06°50’E June 48°15’N 05°51’W March 43°25’N 08°30’W March 41°00’N 08°50’W February 38°30’N 09°20’W February 37°00’N 08°30’W February 39°24’N 02°28’E June 42°00’N 11°38’E May 51°15’N 17°00’E July 38°15’N 20°42’E March 36°40’N 04°00’W May 39°00’N 09°09’E August 37°25’N 12°10’E August 40°43’N 01°30’E June 44°00’N 01°38’E June Location (2001) Month Notes 59°41’N 05°10’E 52°53’N 12°03’W 51°35’N 11°05’W 54°45’N 06°00’E 48°45’N 09°29’W 43°35’N 08°52’W 41°00’N 08°50’W 38°30’N 09°20’W 37°00’N 08°30’W 19°58’N 17°28’W 39°24’N 02°28’E 42°00’N 11°38’E 51°15’N 17°00’E 38°15’N 20°42’E 40°28’N 24°55’E 36°40’N 04°00’W 39°00’N 09°09’E 37°25’N 12°10’E 42°07’N 03°18’E 44°00’N 01°38’E October March April July March March February February February January April June September May June May June August December June N/S S S S S S A/P, S A/P, S A/P, S S P/S P/S P/S S S S S P/S P/S S Table 3.1. Location of sample collection sites. S – spawning, P/D – partly spawning, N/S – non spawning, A/P – approximate position. Due to difficulties in obtaining personnel to conduct sampling, site 11, off the Atlantic coast of North Africa, was not collected in 2000. Sample 16 was collected in 2000, however, after being transported from Crete, the fish were held, unrefrigerated at the height of summer, for several days Heathrow airport during the final stages of an outbreak of infectious salmon anaemia (ISA) in Scotland whilst confirmation was 146 sought that these fish did not carry the ISA virus (Stagg et al., 2001). When these fish finally arrived in Aberdeen, they had decayed to a state in which they could not be meaningfully examined, and were incinerated by Health and Safety Executive officials as a biohazard. In the north east Atlantic area, during the 2001 sampling activity, some of the samples collected did not exactly match with the planned locations. The area coded and sampled as 06 in 2001 was located much closer to the initially planned area 03. For the purposes of analysis, samples 03-01 and 06-01 were merged. In some sampling sites it was impossible to get a whole sample of 100 fish in spawning stages, although they were captured in what was supposed to be the spawning season. When there were difficulties in catching sufficient spawning fish, non-spawning fish were retained to make the sample size up to 100. This was the case in some of the Mediterranean areas. In area 01 (Northern North Sea) horse mackerel are available only in autumn and winter, after spawning activity is over (see table 2.1 and figure 2.1). These fish were sampled out of interest to see which spawning stock they belonged to, or whether they were a mixture of fish from the North Sea and Western stocks. 147 Area Code 1 2 3 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Sample size (2000) Examined by Sample size (2001) Examined by 50 NC 50 NC 50 NC 34 NC 50 NC 99 NC 53 NC 50 NC 50 NC (see 3) - 50 NC 50 NC 51 ALP 100 ALP 56 ALP 52 ALP 62 ALP 100 ALP - - 50 NC 50 NC 52 NC 42 SM 49 SM - - 45 SM 22 NC 44 NC - - 50 NC 50 ALP/NC 50 ALP 21 SM 49 SM 50 SM 50 SM 50 NC 49 NC 50 NC 50 NC Table 3.2. Sample size and details of examiner. NC – Neil Campbell, University of Aberdeen, ALP – Ana Luisa Pereira, Instituto de Investigação das Pescas e do Mar, Lisbon; SM – Simmonetta Mattiucci, Universitá di Roma “La Sapienza”, Rome. 3.2.2 Dissection On collection, horse mackerel were labelled with a tag around the wrist of the tail or through the gills, and stored individually in sealed, labelled plastic bags. Fish were frozen and held at -18°C until required. Fish were shipped by courier to the University of Aberdeen, the Instituto de Investigação das Pescas e do Mar (IPIMAR) in Lisbon or the Universitá di Roma “La Sapienza”, in Rome (see table 3.2 for examination details and sample sizes). Fish were defrosted at room temperature over 148 several hours, to the point where they became pliable. Fish were placed on individual plastic laminated sheets of graph paper; their fins pinned out and were photographed using a 35mm Leica camera. These images were developed and passed to colleagues at IPIMAR who were carrying out the morphometric aspects of the study. Total length of each fish was measured to the nearest millimetre and recorded. Where this was not possible due to damage to the caudal fin, a measurement of fork length was taken and the following formula applied to convert this value to total length (Froese & Pauly. 2006). TL= 1.079x FL The pectoral and pelvic fins were removed, irrigated with physiological saline and examined under a Hunt dissecting microscope (x6-x30) for the presence of parasitic crustaceans or monogeneans. The body cavity was opened with a vertical incision, upwards, from the rear of the pelvic fin to flank of the body, level with the top edge of the eye, a second incision laterally from above the pectoral fin along the length of the body cavity allowed the internal organs to be removed without damage to the intestine (figure 3.3). The viscera were separated into intestine, pyloric caeca, stomach and gonads, and each was placed in a separate Petri dish and irrigated with physiological saline. The liver and gall bladder were also retained. The left operculum was removed, then each gill arch removed individually, these were also covered in saline and the procedure repeated on the opposite side. Examination of viscera for the presence of parasites was carried out using a Hunt dissection microscope at magnifications of between 6 and 50X, examinations of smears of liver and gall bladder were carried out using a Zeiss Photomicroscope II at magnifications of around 300X. 149 Incision 2 Incision 3 Incision 1 Figure 3.3. Incisions made in the body of a horse mackerel to remove the visceral organs as a single unit. 3.2.3 Analysis of data The measures of parasitic infection and ecological terms used in this study are as defined by Bush et al. (1997). Linear regression was used to investigate relationships between infection and host age. The chi-squared test was used as a test of statistical significance between sample prevalences. The non-parametric KruskalWallis test was used to test for significant differences in mean intensity of infection between samples 150 3.3 Results As shown in the previous chapter, 45 taxa of parasites were recorded in this study. Many of these species were recorded in single fish in a single year and may represent accidental infections (tables 3.2.1-4). These species are of no use for stock identification studies. 3.3.1 Age and Length Information As many parasite infections can be related to fish length (figs. 3.4) or age (figs. 3.5), it is important to bear these factors in mind when interpreting the results of a biological tag study. 151 0 100 200 300 400 A 1 2 3 5 6 7 8 9 10 12 13 15 17 19 20 21 100 150 200 250 300 350 400 B 1 2 3 5 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Fig 3.4 a & b Length statistics (mm) for 2000 (a) and 2001 (b). Boxes represent 75th and 25th quartiles, whiskers represent 5th and 95th quartiles, outliers are shown as individual points, medians are shown by a line and mean lengths are shown by a red dot. 152 20 0 5 10 15 A 2 3 5 6 7 8 9 10 12 13 15 17 18 19 20 20 1 21 0 5 10 15 B 1 2 3 5 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Fig 3.5a & b Age statistics for 2000 (a) and 2001 (b). Boxes represent 75th and 25th quartiles, whiskers represent 5th and 95th quartiles, outliers are shown as individual points, medians are shown by a line and mean ages are shown by a red dot. Mean ages of fish from the North Atlantic and North Sea are noticeably higher than those of fish from the Atlantic coast of the Iberian Peninsula and North Africa. 153 This is strongly influenced by the presence of the 1982 year class. In the Mediterranean Sea, mean ages are generally lower than those from Atlantic waters, although mean lengths are roughly the same, reflecting faster growth rates in warmer waters. 3.3.2 Parasite prevalence data Mean abundances and prevalences of species recorded from samples collected in 2000 are presented in table 3.2.1 and 3.2.3 respectively. The same information for samples collected in 2001 is shown in table 3.2.2 and 3.2.4. 154 Species Anisakis spp. Hysterothylacium aduncum larvae Hysterothylacium aduncum adults Prodistomum polonii Tergestia laticollis Derogenes varicus Ectenurus lepidus Pseudaxine trachuri Gastrocotyle trachuri Paradiplectanotrema trachuri Cemocotyle trachuri Nybellnia lingualis Corynosoma strumosum Rhadinorhynchus cadenati Peniculus fistula Scolex pleuronectis Lecithocladium excisum Lasiotocus typicus Monascus filiformis Hemiuris communis Opechona bacillaris Pseudopecoeloides chloroscombri Argulus purpureus Lerneocera trachuri Bathycreadium elongatum Isopod praniza Area 1 2 3 5 6 21 7 8 55.32 96.56 61.02 21.81 7.46 124.46 17.64 16.24 10.44 11.34 52.64 196.55 54.26 67.58 11.06 0.35 0.02 0.28 0.06 0.83 0.00 0.18 0.00 0.00 0.00 0.00 0.00 0.13 0.00 0.00 0.06 0.00 0.10 0.08 0.00 0.53 0.04 0.04 0.00 0.00 0.04 0.00 0.00 1.06 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.55 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.15 0.00 0.00 0.00 0.04 0.22 0.00 0.04 0.58 0.00 0.02 0.00 1.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.04 0.00 0.11 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.04 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 9 2.05 0.04 0.00 0.00 0.00 0.00 0.27 0.13 1.38 0.00 0.00 0.00 0.00 0.00 0.02 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 10 0.74 0.05 0.00 0.00 0.00 0.00 0.71 0.37 3.58 0.00 0.00 0.00 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 12 0.00 1.27 0.00 0.31 0.06 0.00 0.04 0.00 0.00 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 20 13 0.00 50.91 0.74 0.09 0.00 0.00 0.04 0.00 0.00 0.00 0.00 0.00 0.16 0.00 0.16 0.00 1.62 0.18 0.00 0.00 0.00 0.00 0.02 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.12 0.00 0.14 0.00 0.22 0.00 0.52 0.00 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.21 0.00 0.00 15 2.18 0.00 0.00 0.00 0.00 0.00 0.00 0.32 0.09 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 17 18 19 1.30 81.67 95.68 0.00 0.00 0.12 0.00 0.00 0.00 0.14 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.56 0.00 0.00 0.00 0.00 0.00 0.24 1.90 0.30 0.00 0.76 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.05 0.08 0.00 0.00 0.02 0.00 0.14 0.00 Table 3.3.1. Mean abundances of parasites recorded from samples collected in 2000. 155 Species 1 2 3 5 Anisakis sp. 84.5 127.2 37.7 11.5 Hysterothylacium aduncum larvae 10.2 46.8 38.2 323.3 Hysterothylacium aduncum adults 0.0 0.0 0.0 0.2 Pseudoterranova decipiens 0.0 0.0 0.0 0.0 Prodistomum polonii 0.0 0.0 0.0 0.0 Tergestia latticolis 0.0 0.8 0.3 0.9 Derogenes varicus 0.0 0.0 0.0 0.7 Ectenurus lepidus 0.0 0.0 0.0 0.6 Lasiotocus tropicus 0.0 0.0 0.0 0.0 Pseudaxine trachuri 0.0 0.0 0.0 0.3 Gastrocotyle trachuri 0.0 0.0 0.1 1.9 Heteraxinoides atlanticus 0.0 0.0 0.0 0.0 Paradiplectanotrema trachuri 0.0 0.0 0.0 0.0 Cemocotyle trachuri 0.0 0.0 0.0 0.0 Nybelinia lingualis 0.0 0.0 0.0 0.0 Grillotia sp. 0.0 0.0 0.2 0.0 Corynosoma wegeneri 0.0 0.0 0.0 0.0 Corynosoma strumosum 0.0 0.0 0.0 0.0 Rhadinorhynchus cadenati 0.0 0.0 0.0 0.0 Peniculus fistula 0.0 0.0 0.0 0.0 Scolex pleuronectis 0.0 0.0 0.0 0.0 Lecithocladium excisum 0.0 0.0 0.0 0.0 Ancylocoelium typicum 0.0 0.1 0.1 0.0 Monascus filiformis 0.0 0.0 0.0 0.3 Hemiuris communis 0.0 0.0 0.0 1.0 Opechona bacillaris 0.0 0.0 0.0 0.0 Pseudopecoeloides chloroscombri 0.0 0.0 0.0 0.0 Argulus purpureus 0.0 0.0 0.0 0.0 21 7 8 9 10 11 Area 12 20 13 5.6 10.1 NA NA NA 0.1 0.0 0.0 72.8 14 15 16 153.3 217.5 119.1 17 18 19 NA 80.7 25.3 3.5 6.9 NA NA NA 0.3 0.5 0.2 0.1 0.3 0.1 0.6 NA 0.1 0.0 0.0 0.0 NA NA NA 0.0 0.0 0.0 0.0 0.0 0.0 0.0 NA 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.5 0.0 0.0 0.0 0.0 0.1 0.0 0.1 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.2 0.4 0.7 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.3 0.4 2.9 6.9 5.8 0.7 0.0 2.0 0.0 0.4 0.3 0.8 0.4 1.2 0.9 0.0 0.0 0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.1 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Table 3.3.2. Mean abundances of parasites recorded from samples collected in 2001. 156 Species Lernanthropus trachuri Bathycreadium elongatum Isopod Praniza Ceratothoa oestroides Caligus elongatus Caligus pelamydis Pseudophylidean plerocercoids Pseudanisakis sp. 1 0.0 0.0 0.0 0.0 0.0 0.0 2 0.0 0.0 0.0 0.0 0.0 0.0 3 0.0 0.0 0.0 0.0 0.0 0.0 5 0.0 0.0 0.0 0.0 0.0 0.0 21 0.0 0.0 0.0 0.0 0.0 0.0 7 0.0 0.0 0.0 0.0 0.0 0.0 8 0.0 0.0 0.0 0.0 0.0 0.0 9 0.0 0.0 0.0 0.0 0.0 0.0 10 0.0 0.0 0.0 0.0 0.0 0.0 Area 11 0.0 0.0 0.0 0.0 0.0 0.0 12 0.0 0.0 0.0 0.0 0.0 0.0 20 0.0 0.0 0.0 0.0 0.0 0.0 13 0.0 0.0 0.0 0.0 0.0 0.0 14 0.0 0.0 0.0 0.0 0.0 0.0 15 0.0 0.0 0.0 0.0 0.0 0.0 16 0.0 0.0 0.0 0.0 0.0 0.0 17 0.0 0.0 0.0 0.0 0.0 0.0 18 0.0 0.0 0.0 0.0 0.0 0.0 19 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.3 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Table 3.3.2 (cont.). Mean abundances of parasites recorded from samples collected in 2001 (cont.). 157 Species Alataspora sp. Goussia cruciata Anisakis spp. Hysterothylacium aduncum larvae Hysterothylacium aduncum adults Prodistomum polonii Tergestia laticollis Derogenes varicus Ectenurus lepidus Pseudaxine trachuri Gastrocotyle trachuri Paradiplectanotrema trachuri Cemocotyle trachuri Nybelinia lingualis Corynosoma strumosum Rhadinorhynchus cadenati Peniculus fistula Scolex pleuronectis Lethicocladium excisum Lasiotocus typicus Monascus filiformis Hemiuris communis Opechona bacillaris Pseudopecoeloides chloroscombri Argulus purpureus Lerneocera trachuri Bathycreadium elongatum Isopod praniza 1 0.00 0.10 1.00 0.84 0.02 0.00 0.08 0.04 0.00 0.02 0.14 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.02 0.02 0.02 0.02 0.00 0.00 0.00 0.00 0.00 0.00 2 0.00 0.38 1.00 0.66 0.16 0.00 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 3 0.00 0.64 0.90 0.90 0.04 0.00 0.00 0.00 0.02 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 5 0.02 0.96 0.70 1.00 0.30 0.02 0.19 0.13 0.11 0.11 0.23 0.00 0.00 0.08 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 6 0.00 0.68 0.64 1.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 21 0.00 0.66 1.00 0.90 0.10 0.00 0.04 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 7 0.00 0.50 0.96 0.92 0.00 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Area 8 0.00 0.98 0.96 0.22 0.00 0.00 0.00 0.00 0.00 0.02 0.31 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 9 0.00 0.96 0.73 0.04 0.00 0.00 0.00 0.00 0.21 0.11 0.54 0.00 0.00 0.00 0.00 0.00 0.02 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 10 0.00 0.74 0.48 0.05 0.00 0.00 0.00 0.00 0.42 0.29 0.73 0.00 0.00 0.00 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 12 0.00 0.06 0.00 0.60 0.00 0.16 0.04 0.00 0.04 0.00 0.00 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 20 0.04 0.04 0.00 0.42 0.00 0.04 0.00 0.00 0.14 0.16 0.54 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.06 0.12 0.20 0.26 0.02 0.00 0.00 0.00 0.00 13 0.00 0.57 0.79 0.07 0.00 0.00 0.00 0.00 0.00 0.00 0.14 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.07 0.00 15 0.00 0.00 0.77 0.00 0.00 0.00 0.00 0.00 0.00 0.27 0.09 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 17 0.00 0.36 0.24 0.00 0.00 0.12 0.00 0.00 0.32 0.00 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.12 0.00 0.00 0.00 0.02 0.00 0.00 0.00 18 0.00 0.95 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.57 0.52 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.05 0.00 0.05 19 0.00 0.22 1.00 0.12 0.00 0.00 0.00 0.00 0.00 0.00 0.20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.08 0.02 0.00 Table 3.3.3. Prevalences of parasites recorded from samples collected in 2000. 158 Species Alataspora solomoni Alataspora serenum Goussia cruciata Kudoa nova Kudoa sp. Myxobolus spinacurvatura Anisakis sp. Hysterothylacium aduncum larvae Hysterothylacium aduncum adults Pseudoterranova decipiens Prodistomum polonii Tergestia latticolis Derogenes varicus Ectenurus lepidus Lasiotocus tropicus Pseudaxine trachuri Gastrocotyle trachuri Heteraxinoides atlanticus Paradiplectanotrema trachuri Cemocotyle trachuri Nybelinia lingualis Grillotia sp. Corynosoma wegeneri Corynosoma strumosum Rhadinorhynchus cadenati 1 2 3 5 21 7 8 Area 9 10 11 12 20 13 14 15 16 17 18 19 0.00 0.00 0.02 0.00 0.00 0.00 1.00 0.78 0.02 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.21 0.53 0.00 0.00 0.00 0.94 0.97 0.00 0.03 0.00 0.12 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.15 0.77 0.00 0.06 0.00 0.95 0.95 0.02 0.00 0.00 0.09 0.00 0.00 0.00 0.01 0.06 0.00 0.00 0.00 0.00 0.11 0.01 0.00 0.00 0.00 0.00 0.90 0.00 0.00 0.00 0.90 1.00 0.12 0.00 0.00 0.26 0.26 0.06 0.00 0.18 0.62 0.00 0.00 0.00 0.02 0.02 0.00 0.00 0.00 0.00 0.00 0.48 0.00 0.00 0.00 0.78 0.64 0.00 0.00 0.02 0.04 0.04 0.14 0.00 0.04 0.24 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.52 0.02 0.00 0.00 0.88 0.68 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.02 0.26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 NA NA 0.86 0.00 0.00 0.00 NA NA NA 0.00 0.00 0.00 0.00 0.00 0.00 0.12 0.62 0.10 0.00 0.00 0.00 0.00 0.00 0.00 0.00 NA NA 0.90 0.00 0.00 0.00 NA NA NA 0.00 0.00 0.00 0.00 0.00 0.00 0.27 0.79 0.08 0.00 0.02 0.00 0.00 0.00 0.00 0.00 NA NA 0.85 0.00 0.00 0.00 NA NA NA 0.00 0.00 0.00 0.00 0.00 0.00 0.34 0.86 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.06 0.00 0.00 0.04 0.16 0.00 0.00 0.00 0.00 0.00 0.06 0.00 0.04 0.40 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.33 0.00 0.00 0.00 0.02 0.21 0.00 0.02 0.02 0.06 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.22 0.00 0.00 0.00 0.02 0.22 0.00 0.00 0.02 0.02 0.00 0.08 0.00 0.04 0.47 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 NA NA 0.59 0.00 0.00 0.00 1.00 0.12 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.04 0.00 0.00 0.02 0.00 0.00 0.00 NA NA 0.18 0.00 0.00 0.00 1.00 0.20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.29 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.14 0.00 0.52 0.00 0.00 0.00 1.00 0.02 0.00 0.00 0.02 0.00 0.00 0.07 0.00 0.02 0.18 0.00 0.07 0.00 0.00 0.00 0.00 0.00 0.00 0.22 0.00 0.36 0.00 0.00 0.02 0.92 0.22 0.00 0.00 0.00 0.02 0.00 0.04 0.00 0.00 0.36 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 NA NA 0.28 0.00 0.00 0.00 NA NA NA 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 NA NA 0.76 0.00 0.00 0.00 1.00 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.45 0.00 0.06 0.00 0.00 0.00 0.00 0.00 0.00 NA NA 0.50 0.00 0.00 0.00 0.98 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.38 0.00 0.32 0.00 0.00 0.00 0.00 0.00 0.00 Table 3.3.4. Prevalences of parasites recorded from samples collected in 2001 159 Species Peniculus fistula Scolex pleuronectis Lecithocladium excisum Ancylocoelium typicum Monascus filiformis Hemiuris communis Opechona bacillaris Pseudopecoeloides chloroscombri Argulus vittatus Lernanthropus trachuri Bathycreadium elongatum Isopod Pranziae Ceratothoa oestroides Caligus elongatus Caligus pelamydis Pseudophylidean plerocercoids Pseudanisakis sp. 1 2 3 5 21 7 8 Area 9 10 11 12 20 13 14 15 16 17 18 19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.06 0.12 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.02 0.00 0.00 0.22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.08 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.02 0.00 0.20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.10 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Table 3.3.4 (cont.). Prevalences of parasites recorded from samples collected in 2001. 160 3.3.3 Selection of biological tags 3.3.3.1 Apicomplexa. One species belonging to the phylum Protozoa, Goussia cruciata, occurred commonly in most samples (Figs. 3.6a and 3.6b). Prevalence of infection varied greatly and ranged from zero to 98% in individual samples. In the Atlantic region prevalence was over 50% in most samples except for those at the northern and southern extremes of the study area (01 and 11); in the Mediterranean prevalence was highest in the central part. Prevalence was not stable from year to year, suggesting that infections with G. cruciata are of limited duration, possibly on a very short term basis. Sample 01, non-spawning fish collected off the Norwegian coast, shows consistently low levels of infection in both years, which might suggest prevalence increases as transmission becomes easier when fish congregate to spawn. The fish are collected later in the year than any samples of spawning fish in north Atlantic waters, suggesting that the infection can be resolved in a short time-span. Due to these associated uncertainties, this species is unlikely to 0.2 0.4 0.6 0.8 A 0.0 Prevalence 1.0 be an informative tag species. 1 2 3 5 6 7 8 9 10 12 13 15 17 18 19 20 21 0.2 0.4 0.6 0.8 B 0.0 Prevalence 1.0 2000 1 2 3 5 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 2001 Figure 3.6 Prevalence of Goussia cruciata in 2000 (a) and 2001 (b) samples. 161 There was no relationship between the prevalence of G. cruciata and mean length of a sample, suggesting that infection is not related to age or growth (r2= 0.6 0.4 0.0 0.2 Prevalence 0.8 1.0 0.003259, p=0.846) (fig. 3.7). 150 200 250 300 350 400 Mean Length (mm) Figure 3.7. Relationship between Goussia cruciata prevalence and mean length of a sample. 162 3.3.3.2 Myxosporea. Five species of myxosporeans were found, the most common being two species of the genus Alataspora infecting the gall bladders (Fig. 3.8a and 3.8b). In 2000 samples Alataspora infections were only identified to a genus level. In 2001 Alataspora serenum was found mainly in samples 02.01, 03.01, and 06.01 to the south and west of the British Isles, but was also recorded in one fish from sample 11.01 off Mauritania. The gall bladders of fish from samples 08.01, 09.01 and 10.01 from off the coast of Portugal could not be examined, and no infected fish were found in samples 07.01 and 21.01 from off the north coast of Spain. Alataspora solomoni was found only in samples 15.01 and 16.01 from the eastern Mediterranean. One fish from each of samples 05.00 and 20.00 was found to be infected with an Alataspora species. Unfortunately these were not identified to a species level. Four fish, from stations 07.01 and 11.01, were infected with Kudoa nova, and three fish in sample 03.01 were infected with an unknown species of Kudoa. The liver of one fish, from sample 16.01 in the eastern Mediterranean, was infected with Myxobolus spinacurvatura. The myxosporeans, particularly the two species of Alataspora, show considerable potential for use as biological tags to identify populations. Gaevskaya and Kovaleva (1979) proposed a southern distribution for K. nova. That hypothesis is supported by these findings, and this species may be indicative of mixing between Southern and African stocks. 163 1.0 Prevalence 0.0 0.2 0.4 0.6 0.8 A 1 2 3 5 6 7 8 9 10 12 13 15 17 18 19 20 21 0.2 0.4 0.6 0.8 B 0.0 Prevalence 1.0 2000 1 2 3 5 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 2001 Figure 3.8a. Prevalence of Alataspora spp. from 2000 and Figure 3.8b. Prevalence of A. solomoni (dark grey) and A. serenum (light grey) in 2001 samples. 164 3.3.3.3 Monogenea. Five species of monogeneans were found, four of them parasitic on the gill filaments, primarily Pseudaxine trachuri (figures 3.9.1a-b, 3.9.2a-b) and Gastrocotyle trachuri (figures 3.10.1a-b, 3.10.2a-b), and one in on the pharynx and in the stomach. The most commonly encountered species was Gastrocotyle trachuri, prevalence of which ranged from zero to 86% in individual samples, with intensities of up to 46 worms per fish. The heaviest infections were seen in smaller fish (fig. 3.11). This is in line with the findings of Llewellyn (1959, 1962) who observed a relationship between intensity of G. trachuri and P. trachuri and age of horse mackerel collected off the south coast of England. Llewellyn concluded that infection peaked at around two years old due to the size of the gill filaments being the optimal size for the clamps of the monogeneans. Pseudaxine trachuri was less common, with prevalence ranging from zero to 34% and with intensities of up to seven per fish. Both species are found throughout the Atlantic and Mediterranean Sea samples, although they appear to be more common in Atlantic waters, and show a pattern of increasing abundance and prevalence along the Portuguese coast. Of the remaining two gill parasites, Heteraxinoides atlanticus was found in fish from seven stations, all in the Atlantic region except for two infected fish in sample 17.00 from the extreme western part of the Mediterranean, while single infections of Cemocotyle trachuri were found in one fish from sample 01.00 and in one from sample 09.01. The endoparasitic monogenean Paradiplectanotrema trachuri was found in Mediterranean samples only and was most common in fish from stations 18 and 19 (Figure 3.12a-b). Intensity of infection was low, with infection ranging from 1 to 4 parasites per individual. The distribution of infections does not appear consistent from year to year, limiting its usefulness as a biological tag. 165 1.0 Prevalence 0.0 0.2 0.4 0.6 0.8 A 1 2 3 5 6 7 8 9 10 12 13 15 17 18 19 20 21 0.2 0.4 0.6 0.8 B 0.0 Prevalence 1.0 2000 1 2 3 5 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 2001 2 4 6 8 A 0 Mean Abundance 10 Figure 3.9.1a-b. Prevalence of Pseudaxine trachuri in 2000 and 2001. 1 2 3 5 6 7 8 9 10 12 13 15 17 18 19 20 21 2 4 6 8 B 0 Mean Abundance 10 2000 1 2 3 5 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 2001 Figure 3.9.2a-b. Mean abundance of Pseudaxine trachuri 166 1.0 Prevalence 0.0 0.2 0.4 0.6 0.8 A 1 2 3 5 6 7 8 9 10 12 13 15 17 18 19 20 21 0.2 0.4 0.6 0.8 B 0.0 Prevalence 1.0 2000 1 2 3 5 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 2001 2 4 6 8 A 0 Mean Abundance 10 Figure 3.10.1 a-b. Prevalence of Gastrocotyle trachuri in 2000 and 2001 1 2 3 5 6 7 8 9 10 12 13 15 17 18 19 20 21 2 4 6 8 B 0 Mean Abundance 10 2000 1 2 3 5 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 2001 Figure 3.10.2.a-b. Mean abundance of Gastrocotyle trachuri in 2000 and 2001. 167 50 40 30 20 Number of G. trachuri 10 0 0 100 200 300 400 Length (mm) Figure 3.11 Infection with G. trachuri peaks at around 20cm, in accordance with the findings of 0.0 0.2 0.4 0.6 0.8 1.0 Llewellyn (1959, 1962) (black dots – 2000, red dots – 2001). 2 3 5 6 7 8 9 10 11 12 12 13 15 17 18 19 20 21 0.0 0.2 0.4 0.6 0.8 1.0 1 1 2 3 5 7 8 9 10 13 14 15 16 17 18 19 20 21 Figure. 3.12a (top) & 3.12b (bottom) Prevalence of Paradiplectanotrema trachuri in 2000 and 2001. 168 3.3.3.4. Digenea. Twelve species of digeneans were found, all of them adult forms. None was particularly common, the most frequently encountered species being Tergestia laticollis (fig. 3.13a-b) and Ectenurus lepidus (fig. 3.14a-b), with sample prevalences of up to 20% and 42% respectively. Bathycreadium elongatum is a new host record for T. trachurus and was found only in fish from stations 13 and 19 in the central Mediterranean. Pseudopecoeloides chloroscombri was found only in fish from station 21 off the north coast of Spain in both years. None of the other digeneans showed any 1.0 clear regional distribution. 0.0 0.2 0.4 0.6 0.8 A 2 3 5 6 7 8 9 10 12 13 15 17 18 19 20 1.0 1 21 0.0 0.2 0.4 0.6 0.8 B 1 2 3 5 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Figure 3.13a-b. Prevalence of Tergesia laticollis in 2000 (a) and 2001 (b) shows consistent patterns of distribution. 169 1.0 0.0 0.2 0.4 0.6 0.8 A 2 3 5 6 7 8 9 10 12 13 15 17 18 19 20 1.0 1 21 0.0 0.2 0.4 0.6 0.8 B 1 2 3 5 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Figure 3.14a-b. Prevalence of Ectenurus lepidus in 2000 (a) and 2001 (b) shows a higher level of interannual variation in its distribution. 3.3.3.5 Cestoda. Four species of cestodes were found, all of them postlarval forms. None was common, the most frequent being Grillotia sp., found in 18 fish from samples from the Western stock area. Another trypanorhynch, Nybelinia lingualis, was found in seven fish from the same region and in one from sample 20.00 in the Mediterranean. Fourteen fish in samples 15.01 and 16.01 from the eastern Mediterranean were infected with pseudophyllidean plerocercoids, while Scolex pleuronectis was found in only one fish from station 09.00 and two from 21.00, both stations around the Iberian Peninsula. Due to the very low prevalence and abundance of these parasites they are unlikely to provide any useful information on the stock identity of the horse mackerel. 170 3.3.3.5 Acanthocephala. Three species of acanthocephalan were found. One fish in sample 05.00 was infected with two juvenile specimens of Corynosoma strumosum, and one fish from sample 03.01 had one juvenile specimen of C. wegeneri. Two fish from sample 10.00 had single adult specimens of Rhadinorhynchus cadenati in their intestines. Sylianteng (2004) showed this was a characteristic parasite of horse mackerel from Portuguese waters. A follow up study by MacDonald (2005) found R. cadenati to be an extremely important tag species for delimiting the southern boundary of the Southern stock, and placed this on the Atlantic coast of Morocco, between the straights of Gibraltar and the town of Larache. 3.3.3.6 Nematoda. Eight species of nematodes were found. Hysterothylacium aduncum was present mostly as larvae but a few adult worms were also found. The other seven species were present as larvae only. Five were members of the genus Anisakis and were identifiable only by the application of multilocus allozyme electrophoresis, as described in Nascetti et al. (1986) and Mattiucci et al. (1997). Genetic markers obtained from these studies enabled the identification to specific level of 2005 larvae by other workers in the HOMSIR project (Mattiucci et al., in press). The species of Anisakis identified were: A. simplex sensu stricto, A. pegreffii (a sibling species of the A. simplex complex), A. typica, A. physeteris and Anisakis sp. Anisakis physeteris was identified only from fish caught at station 13 in the Tyrrhenian Sea. A single specimen of A. typica was identified from station 16 in the eastern Mediterranean, and a single representative of a new genotype, recently studied and characterized from other fish hosts and from the cetacean Mesoplodon mirus 171 (Mattiucci et al., in press), was identified from station 05 in the North Sea. This has provisionally been named Anisakis paggiae (pers. comm., S. Mattiucci, Universita di Roma). For the main data analyses these congeneric species are grouped together as Anisakis spp. In some samples, most notably the Western stock samples (sites 02 and 03) and the eastern Mediterranean samples (sites 15 and 16), several hundred Anisakis larvae were found in individual fish, which led to the estimation of the numbers present in the most heavily infected fish of the first year’s samples. However, when the potential of this species was realised, exact numbers were recorded. The most heavily Anisakis-infected fish, with 795 nematodes, was from sample 15-01 in the eastern Mediterranean. The other common nematode was H. aduncum, larvae of which were found in most samples, but which were particularly abundant in fish from the North Sea station 05, where the heaviest single infection of 1083 larvae was recorded in 2001. In the Mediterranean, H. aduncum was present in significant numbers only in samples from stations 12 and 20 in the western part. 172 3.3.4 Parasites as biological tags The significant differences in parasite prevalence and abundance observed in this study allow us to reject the null hypothesis that horse mackerel exist in one single mixed population in European waters. Parasites can therefore be used to examine the relationship between individual putative stock units. 3.3.4.1 “Western” vs. “North Sea” stocks The mean abundances of Anisakis spp. and H. aduncum in the Atlantic area provides information on stock identity (fig 3.15). Mean A bundanc e (Hy s terothy lac ium) 250 5 200 150 100 3 50 21 6 2 7 8 1 0 0 20 40 60 80 100 120 Mean A bundanc e (A nis akis ) Figure 3.15a. Mean abundance of H. aduncum and Anisakis spp. in Atlantic samples from 2000, ± 95% confidence limits. 173 600 M ea n Abundance (Hys te rothylac ium ) 500 5 400 300 200 100 2 7 21 1 3 0 0 10 20 30 40 50 60 70 80 90 Mean Abundance (Anisakis) Figure 3.15b. Mean abundance of H. aduncum and Anisakis spp. in Atlantic samples from 2001, ± 95% confidence limits. Samples from station 05 in both years were characterised by heavy infections of several hundred H. aduncum and light infections of Anisakis spp.. In contrast, fish from stations 02 and 03 to the west and south-west of Ireland were characterised by heavy infections of Anisakis and light infections of H. aduncum. Fish sampled in 2000 from station 21 in the southern Bay of Biscay had a similar pattern of nematode infection to those from stations 02 and 03, but in 2001 they were very different and had a pattern almost identical to those from station 07 off north-west Spain, with light infections of both nematodes. Some anomalous individual fish could be clearly distinguished from all other fish in the same sample by their markedly different patterns of nematode infection. In sample 05-00, three fish with heavy infections of Anisakis (>300 worms) stood out starkly from all the other lightly infected fish. 174 3.3.4.2 “Western” vs. “Southern” stocks The relative abundances of Anisakis spp. and Hysterothylacium aduncum are less useful for distinguishing between the Southern and Western stocks, with the exception of the two samples collected at area 21 in 2000 and 2001. These have patterns of nematode infection consistent with the Western stock in 2000, then the Southern stock in 2001. This could suggest this area is a site of stock mixing, but that individual groups retain their distinct identity within it, leading to the collection of A 0 2 4 6 8 Prevalence (%) very different samples. 3 5 7 8 9 10 11 60 B 2 3 5 7 8 9 10 11 10 20 30 1 C 0 Prevalence (%) 2 0 20 Prevalence (%) 1 1 2 3 5 7 8 9 10 11 Figure 3.16a-c. Prevalence of monogeneans in Atlantic samples, 2001. A – Heteraxinoides atlanticus, B – Gastrocotyle trachuri, C – Pseudaxine trachuri. 175 Perhaps the strongest evidence of stock identity in the Southern stock comes from the monogeneans. The prevalences of the three common monogeneans, H. atlanticus, G. trachuri and P. trachuri are noticeably higher in Southern stock samples (08, 09, and 10) than in Western stock samples (02, 03, and 07) (fig. 3.16ac). This would support the placement of the division between the putative Southern and Western stocks between sites 07 and 08. 3.3.4.3 Mediterranean Stocks The relative absence of H. aduncum in the Mediterranean area means the comparison of relative abundances of H. aduncum and Anisakis spp. is less useful for the distinguishment of stocks (fig. 3.17), and most variation in Anisakis abundances between stocks was due to differences in mean age (fig. 3.18). The myxosporean Alataspora solomoni and the pseudophyllidean plerocercoids were found only in the eastern Mediterranean samples, although we have no information on their extent along the middle-eastern coasts, and gall bladders were not examined in the fish taken to Rome. Fish from the central basin of the Mediterranean Sea are host to a range of species either absent or present at much lower intensities elsewhere. These include the monogenean Paradiplectanotrema trachuri (fig. 3.12a & 3.12b), the copepod Lernanthropus trachuri and the digenean Bathycreadium elongatum. Fish examined from areas 12 and 20 in both years have a very low mean age, with most fish being aged 1 and have a mean length below 20cm. It is difficult to interpret the parasite fauna of these fish, as they are likely to consume a very different range of prey items to adults, and therefore acquire different parasites, but they appear to host a diverse digenean parasitofauna. Fish from sample 17 have a parasite fauna very similar to the samples 9 and 176 10, from the Southern stock. This would suggest a linkage of the Southern stock with 150 100 50 0 Mean Abundance (Anisakis) 200 horse mackerel populations within the Mediterranean. 0 2 4 6 8 10 Mean Abundance (Hysterothylacium) Figure 3.17. Variation of mean abundance of Anisakis spp. and Hysterothylacium aduncum in Mediterranean samples (2001), ±1 S.E. 177 200 150 100 0 50 Anisakis Mean Intensity 1 2 3 4 5 6 7 8 Mean Age Figure 3.18. Relationship between mean intensity of Anisakis spp. and mean age of samples in the Mediterranean. There was a significant relationship between mean age and intensity (r2=0.49, p<0.05). 178 3.4. Discussion 3.4.1 Parasites as tags The significant difference in the prevalences and abundances of parasites between the various sites support the hypothesis that horse mackerel exist in a number of spatially separated sub-stocks. Prevalence of the coccidian Goussia cruciata differed significantly between samples within both the Atlantic and Mediterranean regions. Stations 01 and 11 had significantly lower levels of infection than all other Atlantic samples, while stations 15 and 20 (in both years) had significantly lower levels than all other Mediterranean samples. Prevalence of infection does not appear to be related to host length or age (fig. 3.7), but as stations 1 and 11 are collected at different times to the other Atlantic samples, levels of infection may vary seasonally, so this parasite can only be considered as potential biological tag pending further information on its life cycle and ecology. Given more information on their geographical distributions at different times of the year, the two myxosporean species Alataspora serenum and A. solomoni could be used to follow migrations of T. trachurus and to estimate the extent of mixing between stocks. Alataspora serenum was found only in Atlantic samples, mainly from the Celtic Sea area, while A. solomoni was found only in the eastern Mediterranean. Alataspora solomoni was originally described from T. mediterraneus ponticus in the Black Sea by Yurakhno (1988), who recorded a prevalence of 42% in this host near Sevastopol. We have no information on the occurrence of either parasite in the central and eastern Mediterranean and off the coast of Portugal. Examinations of samples from these areas would provide the information needed to use these myxosporeans as biological tags. The most effective biological tags to emerge from this study are the larval 179 nematodes Anisakis spp. and Hysterothylacium aduncum. The distinctive pattern of infection with these nematodes observed in samples from the North Sea station 05 clearly distinguishes it from the nearest stations in the Western stock, 01, 02, 03 and 06, and supports the current management strategy which assesses the North Sea population as a separate stock. The three anomalous fish in sample 05-00, identified by infections more characteristic of fish from the putative ‘Western’ stock area, indicate that some migration does occur from Western areas into the North Sea and provides a means of estimating the extent of such migration. The fact that these were three of the oldest fish in the sample may also be significant. The distinction between the ‘Western’ and ‘Southern’ stocks is less clear. The significant differences in levels and patterns of nematode infection between the two samples collected from station 21 suggests that the southern Bay of Biscay is an area of mixing between these two stocks. Sample 21-00 had a similar pattern of nematode infection to samples 02-00 and 03-00, whereas sample 21-01 was markedly different from 02-01 and 03-01 but almost indistinguishable from 07-01 from northwest Spain. Further information has been elucidated from molecular investigation of Anisakis sibling species (Mattiucci et al., in press). The monogenean Heteraxinoides atlanticus was found in over 5% of fish from stations 08, 09 and 10 taken off the coast of Portugal and was also found occasionally in fish from stations 01, 03, 17 and 21. This monogenean is a characteristic parasite of Trachurus spp. caught off the west coast of Africa to the south of the present study area. Gaevskaya and Kovaleva (1979c) reported 20-40% prevalences of H. atlanticus in T. trachurus caught off an area of West Africa to the south of our southernmost station 11. The same authors (Gaevskaya & Kovaleva, 1980) failed to find this species in samples of the same host caught further north as far as the North Sea. Gaevskaya and Kovaleva (1985) reported it from T. picturatus caught off Western 180 Sahara but not from around the Azores. The records of H. atlanticus from the present study are suggestive of migrations of T. trachurus from West Africa northwards as far as southwest Norway and into the extreme western part of the Mediterranean. This is supported by the findings of MacDonald (2005) who suggested similar patterns based on the observation of Rhadinorhynchus trachuri distributions on the African coast. The monogenean Cemocotyle trachuri is also more characteristic of Trachurus spp. caught outside the HOMSIR study area. Gaevskaya and Kovaleva (1979c) reported it from 24% of T. trachurus caught off Western Sahara and in 1-2% of T. capensis caught off Namibia. The same authors (Gaevskaya & Kovaleva, 1980) also reported a single specimen in a T. trachurus caught in the north-western North Sea. Gaevskaya and Kovaleva (1985) found C. trachuri in T. picturatus caught off Western Sahara and at the Azores. Our records of this monogenean come from off the coast of southwest Norway and off the coast of Portugal and are further evidence of extensive migrations of T. trachurus from West Africa into European waters. Gaevskaya and Kovaleva (1980) reported the occurrence of Rhadinorhynchus cadenati in T. trachurus caught off the coast of West Africa, and from T. picturatus caught off Western Sahara and at the Azores (1985). R. cadenati also parasitizes a wide range of other teleost fish species off West Africa (Golvan, 1969). Two specimens of the acanthocephalan Rhadinorhynchus cadenati found in fish from sample 10-00 provide further evidence of northward migrations of T. trachurus from West Africa. This agrees with the findings of MacDonald (2005) who found high prevalences in fish from the Moroccan coast, just south of the present study area. The copepod Lernanthropus trachuri was found only in fish from stations 18 and 19 in the central Mediterranean. It was originally described from T. trachurus caught in the Ligurian Sea off the coast of Italy (Brian, 1903), but has also been reported from T. picturatus caught off Western Sahara and at the Azores by 181 Gaevskaya and Kovaleva (1985), and from T. capensis caught off Namibia by Piasecki (1982). The endoparasitic monogenean Paradiplectanotrema trachuri had a slightly wider distribution in the Mediterranean, but was most common in fish from stations 18 and 19. Its occurrence in T. trachurus and T. mediterraneus was reported by Kovaleva (1970), while Gaevskaya and Kovaleva (1980; 1985) reported its occurrence in T. trachurus and T. picturatus in the Atlantic from Gibraltar southwards along the coast of Africa. They found no infections in samples of T. trachurus taken north of Gibraltar. The digenean Bathycreadium elongatum was found only in fish from stations 13 and 19 in the central Mediterranean. This is the first record of this species from T. trachurus, and its occurrence in this host only in the central Mediterranean probably reflects differences in the feeding habits of horse mackerel in this area. The myxosporean Alataspora solomoni and the pseudophyllidean plerocercoids were found only in samples 15 and 16 from the eastern Mediterranean. The distributions of these parasites suggest the existence of three distinct subpopulations of T. trachurus – western (which has not been characterised due to the small size and young age of fish examined from areas 12 and 20, but appears to contain a diverse digenean fauna), central (characterised by the presence of Paradiplectanotrema trachuri, Lernanthropus trachuri and Bathycreadium elongatum) and eastern (characterised by high abundances of Anisakis spp., Alataspora solomoni infections and the presence of pseudophyllidean plerocercoids) – in the Mediterranean. Fish from sample 17 have a parasite fauna very similar to the samples from the Southern stock samples. It has been suggested that the boundary between the Mediterranean and Atlantic environments occurs not at the Straights of Gibraltar, but at the mouth of the Gulf of Cadiz. 182 3.4.2 Comparison with other stock identification techniques There are two main conclusions which can be taken from the application of parasites as biological tags to problems in stock identity of the horse mackerel. Firstly, in the part of the Atlantic Ocean which was sampled, there is strong evidence that the North Sea population of T. trachurus should continue to be treated as a separate stock, but there is also evidence of some migration from areas to the west of the British Isles into the North Sea, possibly limited to older fish. The distinction between the putative “Western”, “Southern” and “Sahara-Mauritanian” stocks is less clear, with evidence of considerable mixing between populations. These findings are consistent with the findings of the other techniques applied to the same set of fishes. Hierarchical clustering of the results of otolith shape morphometry found a number of groupings. The North Sea samples, combined across years, fell into a clade with those from the putative “Western” stock; however, they were the most distinct member of this clade. This technique found good separation between the putative “Southern” (08, 09, 10) and “Sahara-Mauritanian” (11) samples and those from the “Western” stock (pers. comm., C. Zimmerman, Bundesforschungsansalt fur Seefischerei, Hamburg). Analysis of growth rates found that the fish collected at site 01 were shown to be very fast growing, much more so than the other Western or North Sea stock samples (pers. comm., P. Abaunza, Instituto Oceanograffia Espanol, Santander). This rate was more consistent with fish from the Southern stock area. It was hypothesised that the fastest growing fish are in a better condition and are capable of migrating further north in search of food. This hypothesis is supported by our finding of characteristically “Southern” parasites such as Cemocotyle trachuri or Lasiotocus typicum in this area. In the Mediterranean part of the study area populations of T. trachurus appear 183 to comprise three main stocks – western (12, 20), central (13, 18, 19) and eastern (15, 16). There is also evidence of a migration of fish from Atlantic populations into the extreme western part of the Mediterranean, area 17, evinced by the presence of Anisakis spp., Goussia cruciata and Ectenurus lepidus and absence of Hysterothylacium aduncum larvae, very different from the fauna of areas 12 and 20. This existence of three stocks in the Mediterranean Sea is supported by the results of the body shape morphometry (pers. comm., A. Murta, Instituto de Investigação das Pescas e do Mar, Lisbon). Discriminant analyses of Procrustes transformed variables revealed three groups, a central one, corresponding to areas 13, 14, 18 and 19, one that includes the eastern areas (15 and 16) and one corresponding to the western ones (12, 17 and 20). Otolith shape morphometric analysis revealed no structure in the Mediterranean Sea, however they showed that the western Mediterranean areas 12, 17 and 20 exhibited more similarity to the southern Atlantic areas 10 and 11 than to the eastern Mediterranean, suggesting interbreeding between these areas. These findings are also supported by the results obtained from single strand conformation polymorphism (SSCP) (pers. comm., R. Quinta, , Instituto de Investigação das Pescas e do Mar, Lisbon) which found a division between the eastern and western Mediterranean samples, and also placed a division between the Southern and Western stocks between areas 7 and 8. Other genetic techniques did not find any population structures in either the Atlantic or Mediterranean samples, and in fact could not distinguish between the two. 184 3.4.3 Implications for fisheries management The International Council for the Exploration of the Sea (ICES), advices on the scientific and technical aspects of international research and assessment aimed at promoting the rational and sustainable harvesting of living resources, or on guidelines for environmental management, through working groups which focus on particular species or aspects of stocks. For many years the ICES Working Group on Mackerel, Horse Mackerel and Sardine (WGMHMS) has considered the horse mackerel in the north east Atlantic as separated into three management stocks: the North Sea, the Southern and the Western stocks (Anon., 1992). Since little biological information from research has been available until now, this separation was based on the observed egg distributions from planktonic egg surveys, and the patterns of temporal and spatial distribution of the fishery. Western horse mackerel are thought to follow broadly similar migration patterns to the Western mackerel stock. The egg surveys have demonstrated that it is difficult to determine a realistic border between a Western and southern spawning area (Anon., 1998). The dynamics of the Southern stock has presented particular assessment problems to the WGMHMS, due to difficulties in aggregating catch and survey information into coherent tuning series to be used in the assessment because of uncertainties over the extent of the stock. The combined results of the HOMSIR project support the delineation between the Western and Southern stocks being placed on the Gallician coast. The results of the HOMSIR project support the current management of the North Sea stock, which was not clear at all with the previous available information. Furthermore, the characteristically different patterns of nematode infection between the Western and North Sea stocks has provided useful, and relatively easy to use, biological markers to identify the extent of the North Sea stock. 185 In the Mediterranean, the definition of stock units has been largely absent. The General Fisheries Commission for the Mediterranean (GCFM) has established management areas based on political and statistical considerations rather than biological ones (Lleonart and Maynou, 2003). These took almost no information on horse mackerel stocks in the Mediterranean Sea into account. The results of HOMSIR support the existence of sub-structuring in horse mackerel populations. A number of techniques, including the use of parasites as biological tags, support the existence of at least three main areas, corresponding roughly to the Eastern (Greek waters), Central (around Sardinia) and Western (Iberian coast) Mediterranean Sea. This information will be useful to the GCFM as it reduces uncertainty in stock structure, provides a knowledge base for further studies and allows improved management advice on this resource in the Mediterranean Sea. 186 Chapter 4. A comparison of neural network and discriminant analysis for classification of horse mackerel stocks based on parasitological data. 4.1 Introduction 4.1.1 Stock Identification The main aim of fisheries science is to interpret relevant information on the biology of the species in question, records of fishing effort and size of catches, in order to predict the future size of the population under different fishing regimes, allowing fisheries managers to make decisions on future fishing effort. The common approaches to evaluation, modelling and management of fish stocks assume discrete populations for which birth and death are the significant factors in determining population size, and immigration and emigration are not (Haddon 2001). Consequently, for successful management of fisheries, it is vital that populations which conform to these assumptions are identified. In areas where two stocks mix, it is useful to be able to quantify this so that catches can be assigned to spawning populations in the correct proportions. A number of techniques have been used to identify discrete fish stocks and quantify their mixing, such as physical tags, microchemistry of hard parts and a range of genetic markers (see Cadrin et al., 2005, for a comprehensive review). There is no single “correct” approach to stock identification, the trend being towards multidisciplinary studies which apply a range of methods to the same set of fish to allow cross-validation of findings. One of the more popular methods involves the use of parasites as biological tags, and has been used for over sixty years (Herrington et al., 1939). This technique has a number of advantages over other methods, such as low cost, suitability for delicate species and straightforward sampling procedures. Its main disadvantage is the limited knowledge available on the life-cycles and ecology of many marine parasites, but as research in these areas results in more and more information becoming available, the efficiency of the method increases accordingly 188 (MacKenzie, 1983; Lester, 1990). The theory behind this technique is that geographical variations in the conditions that a parasite needs to successfully complete its life cycle varies between areas and so between fish stocks (distribution of obligatory hosts in the life cycle, environmental conditions or host feeding behaviour). This leads to differences in parasite prevalence (the proportion of a host population infected with a particular parasite species), abundance (the average number of a particular parasite species found per host) or intensity (the average number of parasites found in infected individuals) between areas. A “classical” parasites-as-tags study involves carrying out a preliminary study to identify parasite species which vary in prevalence, abundance or intensity within the study area, followed by the collection of data from a larger number of fish over a number of years to produce conclusive evidence of a lack of mixing between different parts within the study area. (Note that the absence of a difference in parasite prevalence, abundance or intensity between samples is not necessarily indicative of homogeneous mixing.) This approach allows migrations, recruitment of juveniles or mixing of different spawning populations to be observed and quantified. The practical application of this method was modelled and verified by Mosquera et al., (2000). It is particularly useful in areas where a small but significant degree of mixing between two populations occurs, obscuring genetic differences between populations. 4.1.2 The Horse Mackerel The Atlantic horse mackerel (Trachurus trachurus) is a small pelagic species of fish, with a maximum size of about 40cm. They are the most northerly distributed species of the jack-mackerels (family Carangidae) (FAO 2000), and support a 189 sizeable fishery in the northeast Atlantic, both for human consumption and for industrial processing. Catches in the region have been over 500,000 tonnes per year in recent times. They feed at a slightly higher trophic level than many small pelagic fishes, their diet consisting of planktonic copepods, small fishes and benthic invertebrates. This diverse diet is reflected in a diverse parasite community, and 68 taxa have been reported to infect T. trachurus (MacKenzie et al. 2004). Figure 4.1. A graphical representation of the distribution of horse mackerel stock in the north east Atlantic. 1. North Sea Stock. 2. Western Stock. 3. Southern Stock. 4. African Stocks, after ICES (1992, 2004). Approximate locations of samples collected for verification of stock identity are marked with filled circles and those used to investigate stock mixing with empty circles. There has been uncertainty over the identity of stocks in the northeast Atlantic 190 for over a decade. In this area, the International Council for the Exploration of the Seas (ICES) issues management advice for the horse mackerel which assumes the existence of three stocks. These are a (i) Western stock, (ii) a North Sea stock and (iii) a Southern stock (fig. 4.1). These stocks have been defined mainly from observations of the distribution of eggs in regular plankton surveys, and on historical records of the distribution of catches (Eltink 1992; ICES 1992). Until recently, publications dealing with the definition of stock structure in horse mackerel were rare and covered only a small part of the species distribution. In the southern stock there are some works dealing with differences in anisakid infestation levels (Abaunza et al. 1995; Murta et al. 1995). Whilst using allozymes, some authors found differences between areas in the Northeast Atlantic (Nefedov et al. 1978) whereas others did not (Borges et al. 1993). One of the problems with stock definition for T. trachurus is that it is a highly migratory species, spawning along the edge of the continental shelf, in water around 200 metres deep, then dispersing to feed over a wider area. It is thought that the Western and North Sea stocks overlap at certain seasons in the English Channel (Macer 1977), which may cause some degree of mixing between these stocks. Mixing between the Western and Southern stocks remains an unknown quantity, and there has been particular concern about the boundary between these stocks. 191 4.1.3 Statistical methodology in Biological Tag Studies A number of more recent studies have taken a more complex approach to the statistical treatment of parasites as tags of their host populations. These studies have considered each fish as a “habitat” and treated the entire parasite fauna in that individual as a “community”. Discriminant analysis (DA) is applied to the parasite abundance data of the community, in order to identify groups of similar fishes (Lester et al., 1985). This approach is particularly useful in investigations of larger, rarer or more valuable fish species, where the material available for examination by scientists may be limited. Moore et al. (2003) had some success with the application of discriminant analysis to the parasite fauna of narrow-barred Spanish mackerel (Scomberomorus commersonii) around the coast of Australia, in order to quantify movement and mixing of stocks. Discriminant analysis, however, makes a number of assumptions which make it less suitable for this sort of analysis. Firstly, for testing of hypotheses, DA assumes independent variables (in this case, parasite abundance) are both normally distributed and have homoscedasticity (equal variances between sample groups). Secondly, DA works best with roughly equal sample sizes, and requires the number of independent variables to be less than the smallest sample size minus two. Lastly, DA is not appropriate when the standard deviation of one independent variable is zero for one sample, limiting the use of certain variables. In ecological terms, if one species was absent from a sample area, it would have a mean abundance and standard deviation of zero. In a stock identification study it is quite possible for a parasite which is absent from part of the study area to be highly informative, but such a parasite would have to be ignored by a discriminant analysis. While the studies which have been carried out using this technique have applied data transformations, such as the addition of a small 192 random value to each datum, to deal with these assumptions, these transformations can obscure important trends. 4.1.4 Neural Networks The original “neural network” model was proposed in the 1940s (McCulloch & Pitts 1943), although it is only with the advent of cheap, powerful computers that this technique has begun to be applied widely. There has been a great deal of hype surrounding neural networks. They are, however, simply a non-linear statistical approach to classification. One of the attractions of neural networks to their users is that they promise to avoid the need to learn other more complex methods - in short, they promise to avoid the need for statistics! But this is a misconception: for example, extreme outliers should be removed, co-linearity of variables should be investigated before training neural networks, and it would be foolish to ignore obvious features of the data distributions and summaries such as the mean or standard error. The neural network promise of "easy statistics" is, however, partly true. Neural networks do not have implicit assumptions of linearity, normality or homogeneity, as many statistical methods do, and the sigmoid functions which they contain appear to be much more resistant to the effects of extreme values than regression based methods. Many of the claims made about neural networks are exaggerated, but they are proving to be a useful tool and have solved many problems where other methods have failed. The name “neural network” derives from the fact that it was initially conceived of as a model of the functioning of neurons in the brain - the components of the network represent nerve cells and the connections between them, synapses, with the output of the nerve switching from 0 to 1 when the synapses linking to it reach a “threshold value”. 193 For the purposes of this chapter, a neural network can be thought of as a classification model of a real world system, which is constructed from the processing units (“neurons”) and fitted by training a set of parameters, or weights, which describe a model that forms a mapping from a set of given values known as inputs to an associated set of values known as outputs (Saila, 2005). The weights are trained by passing sets of input-output pairs through the model and adjusting the weights to minimize the error between the answer provided by the network and the true answer. A problem can occur if the number of training iterations, or “epochs” is too large. This reduction in classification success of the data not used in training is known as over-fitting. Once the weights have been set by a suitable training procedure, the model is able to provide output predictions for inputs not included in the training set. The neural network takes all the input variables presented in the data and linearly combines them into a derived value, in a so-called “hidden layer object” or node (Smith 1993). It then performs a nonlinear transformation of this derived value (figure 4.2). The use of multiple hidden layer objects in a neural network allows different non-linear transforms of data, with each neuron (node) having its own linear combination, increasing the classifying power of the network. Originally, the neuron was activated with a step function (represented as the dashed line in figure 4.2), when the combined input values exceeded a certain value, however, this more flexible sigmoid function allows differentiation and least squares fitting, leading to the back propagation algorithm, making it possible to tune the weights more finely. 194 1.0 0.8 0.6 0.0 0.2 0.4 Y -10 -5 0 5 10 X Figure 4.2. An example of the sigmoid function which hidden layer units typically apply to linear combinations of the input data. The original, sharper activation curve used in the first neural network models is shown in the dashed line. A possible example is shown in figure 4.3. In this case, we are interested in knowing the stock composition of a mixed sample of fish, and we have count data on six species of parasites from these fish. These counts are treated as the six input variables to our network. The network has four units in the hidden layer. The neural network in such an example would need to have been trained with data from fish which we knew belonged to stocks X and Y beforehand if it was to work successfully. This type of neural network is referred to in the literature by a number of names, such as feed-forward network, multilayer perceptron, or simply vanilla neural network, named for the generic ice-cream flavour. The number of units in a hidden layer is variable. Problems can arise if too few or too many units are used. If the network has too few units, it will not be flexible 195 enough to correctly classify the data it is presented with. On the other hand, if it has too many units, a problem known as over-parameterisation occurs, which reduces the chances of successful classification. Having many hidden layer objects also increases the computing power required to run the function. Often, a trial-and-error approach is used to determine the optimum number of hidden layer units for a particular data set, however there are other methods which take a more considered approach, such as cross-validation, bootstrapping, and early stop. Neural networks differ in philosophy from most statistical methods in several ways. A network typically has many more inputs than a typical regression model. Because these are so numerous, and because so many combinations of parameters result in similar predictions, the parameters can quickly become difficult to interpret and the network is most simply considered as a classifying “black box”. This means that areas where a neural network approach can be applied in ecology are widespread. They are less useful when used to investigate or explain the physical process which generated the data in the first place. In general, the difficulty in interpreting what the functions contained within these networks mean has limited their usefulness in fields like medicine, where the interpretation of the model is vital. 196 Figure 4.3. Graphical representation of the structure of a neural network. This one has six inputs (P1 to P6), one hidden layer with four neurons or units (A to D) and two output neurons into which it will classify the data (stocks X and Y) There have been a number of uses of neural networks in fisheries science (Huse and Grøjsæter 1999; Maravelias et al. 2003; Engelhaard and Heino 2004), however, these have mainly attempted to predict changes in fish abundance, recruitment or distribution, based on environmental and ecological inputs. The use of neural networks in recognising fish stocks is a relatively new development, and has been restricted to analyses of morphometry or of otolith microchemistry (Murta 2000; Hanson et al. 2004). These were reviewed and summarised by Saila (2005). A more specific introduction to neural network architectures can be found in Dayhoff (1990) and Smith (1993). These techniques produce continuously distributed values, such as the distance between two points on the body of a fish, or the quantity of a particular element in an otolith. Parasitological studies, on the other hand, are characterised by relatively large numbers of fish with low numbers of parasites, and a small percentage of observations with high numbers. This tends to give problems with classical statistical techniques that require normality assumptions, such as classical discriminant analysis. 197 Furthermore, some forms of parasite, such as metazoan species, are too numerous to count, therefore a fish is either classed as infected or uninfected. Neural networks are able to cope with both forms of numeric data, as well as presence/absence data, in any combination. They are also robust to missing values and skewed distributions. Note that the question which is addressed by a neural network approach to parasite data is not “do these fish belong to different stocks”, but, “based on the parasitological data available, is it possible to successfully assign these fish to a stock”. The difference is subtle, but it should be apparent that low levels of successful classification between two sets of observations would not suggest that they belong to two different stocks, whereas high success would support a hypothesis that the samples were drawn from different populations. 198 4.2 Methods This work is based on samples of T. trachurus collected as part of the HOMSIR stock identification project (see previous chapter for a detailed explanation of the theory behind sample collection). Fish were collected with a pelagic trawl by a number of research vessels at a number of locations in the north-east Atlantic (see fig. 4.1) in 2000 and 2001 and were immediately frozen and returned to the laboratory for examination. Samples were returned to either the University of Aberdeen and examined by the author, or the Instituto do Investigação das Pescas e do Mar (IPIMAR), Lisbon, and examined by Portuguese colleagues. Between 34 and 100 fish were collected from each site. The number of specimens in each sample and their division between Aberdeen and Lisbon is noted in table 4.1. Details of examination procedures are as described in the previous chapter. To investigate spawning stock identity, three samples each from the Western and Southern stocks, and one each from the North Sea and African stocks, were examined. For estimation of stock mixing, one sample from a non-spawning seasonal fishery from the Norwegian coast, and one spawning sample from the boundary between the Western and Southern samples were examined. 199 Stock Lat. Long. Examined Sample by… Size North Sea 54.45N 06.00E NC 50 Western 52.53N 12.03W NC 34 48.45N 09.29W NC 50 51.35N 11.06W NC 50 44.00N 01.38W NC 50 41.00N 08.50W ALP 100 38.30N 09.20W ALP 52 37.00N 08.30W ALP 100 19.58N 17.28W NC 50 Norwegian 57.41N 05.10E NC 50 La Coruna 43.35N 08.52W NC 50 Southern African Mixed Samples Table 4.1. Location and size of samples collected in 2001 for stock identification and used in the neural network analysis. Examined by: NC – Neil Campbell, University of Aberdeen; ALP – Ana Luisa Pereira, IPIMAR, Lisbon. The equivalent western and North Sea samples from 2000 were also used in the discriminant analysis. 200 4.3 Results 4.3.1 Selection of species for neural network analysis A total of 636 fish from eleven sites were examined. Parasites which infected less than 2% of fish were deemed to represent either rare species or “accidental” infections and were discounted. Eleven species of parasites were found to be commonly present (table 4.2). Class Species Location Data Type Gall Bladder Presence/Absence Liver Presence/ Absence Anisakis spp. Body Cavity Abundance Hysterothylacium sp. (larval forms) Body Cavity Abundance Hysterothylacium sp. (adult forms) Intestine Abundance Tergestia laticollis (Rudolphi, 1819) Intestine Abundance Derogenes varicus (Müller, 1784) Stomach Abundance Ectenurus lepidus (Loos, 1909) Stomach Abundance Pseudaxine trachuri (Parona & Perugia, 1889) Gills Abundance Gastrocotyle trachuri Gills Abundance Gills Abundance Myxosporea Alataspora serenum (Gaevskaya & Kovaleva, 1979) Apicomplexa Goussia cruciata (Theolan, 1892) Nematoda Digenea Monogenea (van Beneden & Hesse, 1863) Heteraxinoides atlanticus (Gaevskaya & Kovaleva, 1979) Table 4.2. Commonly encountered parasite species used for stock identification analysis. 201 4.3.2 Data Exploration Exploration of any set of data is an essential first step in carrying out an appropriate analysis. One of the problems with this sort of study is that because fish vary in size, age or sex between samples, the examinations cannot be thought of strictly as replications of observations on a homogeneous population. It is therefore important to look for variation caused by these factors and remove it from further analysis. Plots were made of the abundance of different parasites against fish length, sex and age. No significant relationships were found, with the exception of the nematode, Anisakis spp. (figures 4.4.a-d). This species encysts in the body cavity of the horse mackerel before maturing to its adult stage when its fish host is eaten by a marine mammal. The abundance of this species was found to have a logistic relationship to the length of its host. A ln (n+1) transformation was performed on the Anisakis abundance values, and a linear regression carried out against fish length. The residual values of this regression were then taken forwards for use in the later analysis. This eliminates bias caused by differing lengths between samples 202 B 400 300 0 150 100 200 Number of Anisakis 300 250 200 Length (mm) 350 500 A AFRICAN NORTH SOUTH WEST ERN 150 200 250 300 350 3 Length(mm) C D 1 0 -1 Residual 4 3 -2 2 -4 0 -3 1 Log(Anisakis +1) 5 2 6 Stock 150 200 250 300 350 Length (mm) AFRICAN NORT H SOUT H WEST ERN Stock Figure 4.4a-d. A: Fish lengths vary significantly between stocks. B: There is a significant relationship between fish length and Anisakis spp. abundance. C: Logarithm transformation of Anisakis abundance produces a linear relationship. D: The residuals of this relationship are taken forward for use in the analysis after removal of length dependency. The values for the southern stock are generally lower as the parasite occurred less frequently in this area. 4.3.3 Discriminant Analysis To provide a benchmark against which to test the usefulness of a neural network approach, a function was created in the R statistical software environment v2.1.1 (R Development Core Team, 2005), using the LDA and predict.LDA functions from the MASS library (Venables & Ripley, 2006), to apply the discriminant analysis method proposed by Lester et al. (1985) to a similar set of data. This function selected 203 a sample of half of the fish from each stock at random for use as a training set, then attempted to reclassify the remaining fish to a stock level. In order to account for variability induced by the random selection of a training set and to provide a measure of the magnitude of error in the analysis, the function was repeated 1000 times and the values ranked, the 50th, 500th and 950th values being used once more as median and 95% confidence limits. Discriminant analysis is unable to handle the binary nature of the presence/absence data presented by Alataspora solomoni and Goussia cruciata, so these species have been excluded from use in this method. In order to produce a near normal distribution from overdispersed parasite abundance data, a logarithmic transformation was applied to abundance values. To allow for matrix inversion in the discriminant analysis, a small (in the order of 10-3), and random value was added to each point of data. This produces variables which have a non-zero standard deviation, but does not significantly affect stock classification. Because Anisakis spp. and Hysterothylacium aduncum were not counted separately in some samples from the southern stock, and because discriminant analysis is unable to deal effectively with missing values, the southern stock was excluded and the analysis focused on distinguishing between the western and North Sea stocks. To increase the power of the analysis, the 2000 and 2001 samples collected in the western and North Sea areas were pooled. The species used for discriminant analysis were Anisakis spp., Hysterothylacium aduncum larvae, Tergestia laticollis, Ectenurus lepidus, Derogenes varicus, Pseudaxine trachuri and Gastrocotyle trachuri. 204 4.3.4 Neural Network Design The first problem in a neural network approach to a classification problem is to select an appropriate structure for the network. This is often done on an ad-hoc basis. A single hidden-layer, feed forward neural network function was constructed using the nnet function (Venables & Ripley, 2002) in R. This function is designed to select half the fish from all samples on a random basis, for use as a training set, then to reclassify the remaining fish to a stock. In the first instance, a single hidden layer object was used, and, to remove chance results caused by selection of fish at random, the process was repeated 100 times with different random training sets. The mean percentage of successful reclassifications over all simulations is then taken and stored. The function then increases the number of hidden layer units and repeats the process. Finally, mean successful reclassification is plotted against number of hidden layer units (figure 4.5). This allows an educated guess to be made as to the most suitable number of hidden layer objects, which is sufficiently flexible to reclassify data successfully, but not over-parameterised to such an extent that over-fitting occurs and excessive processing time is required. 205 100 80 60 40 20 0 Percentage Successful Classification 0 10 20 30 40 50 Number of Units In Hidden Layer Figure 4.5. Estimation of the optimum number of units in the hidden layer. Mean successful classification (±1 standard deviation) reaches over 90% with 8 nodes. Further increases in the number of nodes cause a decrease in success, and an in-crease in variability, at the cost of increased computing time From these results it would appear that the optimum number of hidden layer units is around eight. Eight hidden layer units for the later neural networks were used in the analysis, as there was some decrease in performance of the network at higher values. To investigate stock identity, the same method of selecting half of the fish in a sample as a training set and reclassifying remaining fish was used. The percentage of fish correctly classified was recorded, along with the numbers from each stock misclassified, and the stock to which they were assigned. This network had eleven inputs, eight objects in the single hidden layer, and four outputs. To obtain a measure of the error inherent in selecting a training sample at random, and allowing the 206 starting weights of the network to be selected at random, the process was repeated 1000 times. Once outcomes were sorted, median successful reclassification was represented by the 500th value and 95% confidence limits by the 50th and 950th values. 4.3.5 Discriminant Analysis Results The discriminant analysis proved a useful means of distinguishing between the western and North Sea stocks (fig. 4.6). Median successful reclassification using 1000 different training and test sets for the North Sea was 69.3%, and for fish from the 0.8 0.6 0.4 0.2 0.0 Propotion Successfully Classified 1.0 western stock was 95.7% North Sea Western Figure 4.6. Median successful classification of fish from the 2000 and 2001 North Sea and Western stock samples. Error bars represent the 50th and 950th quartiles of the 1000 repetitions. A graphical representation of the discriminant analysis is presented in figure 4.7. 207 8 NS-00 6 NS-00 4 NS-01 NS-00 NS-00 NS-00 2 LD2 NS-00 NS-01 NS-01 NS-00 NS-01 NS-01 NS-01NS-01 0 NS-01 NS-01 NS-01 -2 NS-01 NS-01 NS-00 NS-00NS-01 NS-01 NS-00 NS-00 NS-00 NS-00 W-01 NS-00 NS-01 W-01 NS-00 W-00 W-00 NS-00 W-00 W-00 W-01 NS-00 W-01 W-01 W-01 NS-00 NS-00 W-01 W-01 NS-01 W-00 W-01 NS-00 W-01 W-00 W-00 W-00 W-00 NS-00 W-01 W-00 W-01 W-01 W-01 W-01 W-01 NS-00 NS-00 W-01 W-00 NS-01 W-01 W-01 W-01 NS-00 W-01 W-01 W-00 W-00 W-00 W-00 W-00 W-00 NS-00 NS-00 W-00 W-00 NS-01W-01 W-01 W-01 W-01 W-01 W-01 W-01 W-01 W-01 NS-00 NS-00 NS-00 NS-00 NS-01 W-01 W-01 W-00 W-01 W-01 W-01 W-01 W-01 W-01 NS-00 W-01 W-01 W-01 W-00 W-00 NS-01 W-01 W-00 W-00 W-00 W-01 W-01 W-01 W-01 W-01 NS-00 W-01 W-01 W-01 W-01 NS-01 W-01 W-01 W-01 W-01 NS-00 W-00 W-00 W-00 W-00 W-00 W-01 W-00 W-01 W-01 W-01 W-01 NS-01 W-00 W-00 W-01 W-00 W-00 W-00 W-01 W-01 W-01 W-01 NS-01 W-01 W-01 NS-01 W-01 NS-00 NS-00 W-00 W-00 W-00 W-01 W-00 W-00 W-00 W-01 NS-01 NS-00 W-01 W-01 W-01 W-00 W-01 W-00 W-00 W-01 W-01 NS-00 W-00 W-00 W-00 W-00 W-00 W-01 W-00 W-00 W-00 W-00 W-01 W-01 W-00 W-01 W-00 W-01 W-01 NS-01 W-00 W-00 W-01 NS-00NS-01 NS-01 W-00 W-00 W-01 W-01 W-01 W-01 W-00 W-01 W-01 W-01 NS-01 NS-01 NS-01 W-01 NS-01 W-01 W-00 W-01 W-01 W-00 W-01 NS-00 W-01 NS-01 W-00 W-01 W-01 NS-01 W-01W-00 NS-00 W-01 W-00 W-01 W-01 NS-00 NS-00 NS-01 W-01 NS-00 W-01 NS-01 W-00 NS-00 W-01 W-00 W-00 W-00 NS-00 W-01 W-01 W-00 NS-00 W-00 W-01 W-00 NS-01 W-01 NS-00 W-01 NS-01 NS-01 NS-01 W-01 W-00 W-01 NS-01 NS-01 W-01 NS-00 NS-01 NS-01 NS-01 W-00 NS-01 NS-00 NS-01 NS-00 NS-01 -4 NS-01 -8 -6 -4 -2 0 2 4 6 LD1 Figure 4.7. Discriminant analysis results for 2000 and 2001 North Sea and western stock samples. 208 The first and second discriminant functions account for 84% and 11% of the variance. Whilst the results from the western stock fishes are tightly clustered, those from the North Sea stock are more widely dispersed. This may have been a product of the different sample sizes considered. 4.3.6 Neural Network Results The outputs of the neural network analysis shows that the neural network is able to correctly classify fish to a stock with a high degree of accuracy – median successful reclassification for the Southern stock is over 95%, other stocks show successful classifications of 80-90% (fig. 4.8). Misclassification tends towards neighbouring stocks, for instance, when a fish from the southern stock is misclassified, it tends to be placed in the African stock. This could suggest that the network is picking up elements of stock mixing that are not apparent from initial examination of the data, or that the network is recognising regional features of a parasite fauna. The difficulties in interpreting the significance of the functions contained within the network mean that it is not possible to examine these features further. These findings support the ICES stock definitions as they are currently applied, and suggest that the application of this neural network to mixed stock analysis will give an accurate picture of stock composition. For investigation of the two mixed stock samples, the whole of the spawning data set was used to train the neural network. This was then used to reclassify the mixed data in question. The neural network was allowed to choose random starting weights; hence outputs are still variable, even when considering the same set of data. Consequently, this process was also repeated 1000 times in order to produce an 209 estimate of inherent variability. Figure 4.8. Percentage of sample assigned to each stock by the neural network; Top Left: fish from the North Sea. Top Right: fish from African waters. Bottom Left: fish from the Southern stock. Bottom Right: fish collected in the Western stock area. The Norwegian seasonal fishery shows a more mixed composition than any of the spawning samples, suggesting it may be made up of fish from more than one area. The neural network classifies around 65% of fish as belonging to the Western stock. The remaining 35% are a mixture of Southern and African stocks. Very few fish are assigned to the North Sea stock (fig. 4.9). 210 The spawning sample from the area of stock uncertainty to the north of Spain is much less conclusive. The neural network assigns around 40% of the sample to the Western stock, 30% to the Southern stock and 20% to the African stock (fig. 4.10). Figure 4.9. Neural network assignment of stock membership of horse mackerel from a mixed, nonspawning seasonal fishery which develops in the summer months off the Norwegian coast. 211 Figure 4.10. Reclassification of horse mackerel in spawning sample taken from area of stock uncertainty on the North West coast of Spain. 212 4.4 Discussion 4.4.1 Stock Identification This study represents the first successful attempt to validate current understanding of fish stocks used for assessment purposes through the application of a neural network to parasite abundance data. This technique successfully reclassifies over 90% of fish from Western and Southern stocks. Success in reclassifying fish from the North Sea and African stocks is lower, although still over 80%. The neural network is very good at distinguishing between members of the Western and Southern stocks. Median misclassification of fish from the Western stock to the Southern stock is 2%, and from the Southern stock to the Western is 0%. Using parasites as biological tags and multivariate analysis of morphometric distances, it is possible to distinguish between fish from the western and North Sea stocks, but it has previously been impossible to conclusively distinguish between fish from the western and southern areas using any method (ICES, 2004). Although neural networks are robust enough to deal with differences in sample size, it is apparent that the areas with larger sample sizes (western and southern) have higher success rates than the two areas with only 50 fish (North Sea and African stocks). The confidence ranges for these two areas are also much wider than for the larger samples. It is interesting to note that where misclassification occurs, it tends to be towards stocks which are already believed to mix, rather than to those with which mixing is not regarded as possible. This might suggest that a small number of “alien” fish from adjacent stocks are present in our “discrete” spawning samples. These fish would produce such misclassifications, although the difficulties in investigating the processes that go on inside the neural network make this impossible to verify. 213 It has been suggested that an element of stock mixing can take place between the Western and North Sea stocks while fish overwinter in the English Channel (Macer, 1977). This has been supported by recent studies (MacKenzie et al., in press). If stock mixing is occurring in this area, it is likely that this effect is being reflected in the results obtained from the neural-network, and that there are a number of fish which belong to the Western stock in the North Sea, and vice-versa. This will slightly confuse the picture which the neural network gives, and could explain the tendency of North Sea fish to be misclassified into the Western stock. A degree of mixing has also been proposed between the Southern and African stocks (Murta, 2000; MacDonald, 2005). This is supported by our findings. Although successful reclassification of fish from the Southern stock is over 95%, the neural network classifies around 20% of fish from the African sample as belong to the Southern stock. Very few fish from the southern samples are classified as belonging to the African stock. This could suggest that mixing between these areas is a one way process. Power et al. (2005) evaluated a range of statistical classifiers to assign bogue (Boops boops) to one of four fisheries in which they were harvested, around the coast of the Iberian Peninsula. They obtained their best results with k-nearest neighbour clustering. This method was applied to our results after the samples collected in 2000 had been examined, however results were not encouraging and this method was not used further (Campbell et al., 2002). Power et al. (2005) also had success with neural networks and linear discriminant analysis, and found optimal reclassification values of 94% for a feed-forward neural network and 93% for discriminant analysis, similar to the values seen in this study. 214 4.4.2 Norwegian non-spawning sample Having established the utility of a neural network approach to assign fish to spawning stocks based on host-parasite data, it is a straightforward matter to apply this to a non-spawning stock to investigate its composition. These findings suggest that fish in the Norwegian sample do not come from a single stock, but rather are drawn mainly from the Western stock, with a sizeable proportion from much more southerly stocks. This finding is in line with work by Abaunza et al. (in press) who measured growth rates of horse mackerel, and found variability from stock to stock, with fishes from warmer waters growing more quickly than their more northerly conspecifics. When examining fish from the Norwegian area, growth rates were noticeably higher than that in neighbouring fisheries, to the west of Ireland and the southern North Sea. Abaunza proposed the existence of a highly migratory “infrastock” which spawned and overwintered in the Southern stock area, then migrated northwards to feed in Norwegian waters. Evidence supporting this hypothesis came from the discovery of characteristically “southern” parasites, such as Cemocotyle trachuri, in fish from the Norwegian area (MacKenzie et al., in press). The neural network classified few fish from this sample as belonging to the North Sea stock. This is interesting when considering how close these two stocks are, spatially, and the previous suggestions of mixing between western and North Sea stocks (Eltink, 1992). These results suggest that very little mixing of North Sea and western horse mackerel occurs in this area. This is an important finding for fisheries managers to consider when allocating catches of fish from the Norwegian fishery to particular spawning populations. 215 4.4.3. La Coruña Spawning Sample The neural network was not able to classify fish from the North West coast of Spain to a particular stock with any great certainty. No one stock appears to dominate this area and the 95% confidence limits are relatively small, at around 10%. The boundary between the Western and Southern stocks was recently moved from the north to the west coast of the Iberian Peninsula (ICES, 2004). These findings suggest that this change was appropriate, in that the sample is composed of fish more classed as “Western” than “Southern”, but also suggest that a high degree of mixing takes place in this area, and that more intensive sampling in this area would be a worthwhile contribution to stock identification of this horse mackerel. 216 4.5 Conclusions It is apparent that a neural network approach to classifying individuals into presupposed groups is a powerful tool for problems such as this. One of the more frustrating aspects of this analysis is that it is not evident why the network has made classifications in a particular way, so it is not possible to determine if mixed samples really are composed of fish from different areas, or whether the network is simply unable to consistently determine stock membership for a particular sample. The ease with which it is possible to use neural networks, their lack of restrictive assumptions and their ability to cope with combinations of different types of data make them extremely useful for dealing with problems of classification in ecology. A number of biological tag studies have used discriminant analysis. Whilst our results showed this method could be used to classify individuals to a stock, the difference in successful classification between Western and North Sea stocks highlights the problems caused by different sample sizes. Whilst a fish is classified into the Western stock with a certainty of over 95%, we are less that 70% sure that a classification to the North Sea stock is correct. The high level of success seen when classifying with neural networks, combined with their ease of use and lack of restrictive assumptions about data make them ideal for this sort of application. 217 Chapter 5. Parasites as biological tags for identification of herring stocks, and patterns of recruitment and mixing of herring (Clupea harengus L.) to the west of the British Isles. 5.1 Introduction The identification of appropriate boundaries of discrete fish stocks is one of the most fundamental pieces of information in the development of appropriate management and exploitation strategies for commercial fishes. The application of management measures that assume erroneous population boundaries can lead to depletion and extinction of local sub-stocks (Butterworth & Penney, 2004). Parasites have been widely used as indicators of stock boundaries, recruitment, migration and mixing of commercially important species of marine, anadromous and freshwater fish. Since their first practical application in 1939 (Herrington et al., 1939), numerous reviews and guidelines for their use in this field have been published (e.g. Sindermann (1961, 1983), MacKenzie (1983, 1987b, 2002, 2004), Lester (1990), Moser (1991), Williams et al. (1992), Arthur (1997) and MacKenzie & Abaunza (1998, 2005)). Although infection statistics of a single parasite species can be enough to determine the origin of fish and to identify stocks, to build up a more detailed picture of host population structure it is preferable to use either combined infection data from several parasite species (MacKenzie, 1985; MacKenzie & Longshaw, 1995; Larsen et al., 1997) or to analyse entire parasite assemblages (Groot et al., 1989; Arthur & Albert, 1993; Speare, 1994, 1995; George-Nascimento, 2000; Lester et al., 2001; Moore et al., 2003; Sardella & Timi, 2004; Timi et al., 2005). The advantage of the latter approach is that the analysis is less sensitive to sampling biases and the discriminant power of the statistical techniques can be assessed by resampling simulations (Groot et al., 1989; Balbuena et al., 1995). The accuracy obtained by these methods in establishing the capture location of fish can be over 95% (Arthur & Albert, 1993; Power et al., 2005). 219 Up to now most biological tag studies have relied on traditional parasitological methods, using light microscopy, for morphological diagnosis of species. The development of molecular techniques for identifying and studying parasite populations brings a new dimension to the use of parasites as biological tags, particularly where different but morphologically indistinguishable species of parasites occur in the same host species. In this situation, populations of the cryptic parasite species can be used as biological tags if the population has sufficiently distinct genetic characteristics (Abollo et al., 2001; Klimpel et al., 2004; Mattiucci et al., 2004; 2005). In this study morphological and genetic these approaches have been brought to bear on the problem of herring stock identity to the west of the British Isles. 5.1.1 Taxonomy and Distribution of Herring The clupeoid or herring-like fishes are commonly reported to be the most abundant fish in the world (Foot, 2001). There are currently considered to be only two species in the genus Clupea, the Atlantic and Pacific herrings, divided into five subspecies (Froese & Pauly, 2004): o Clupea harengus Linnaeus, 1758 (Atlantic Herring) o Clupea harengus membras Linnaeus, 1761 (Baltic Herring) o Clupea pallasii Valenciennes, 1847 (Pacific Herring) o Clupea pallasi marisalbi Berg, 1923 (White Sea Herring) o Clupea pallasii suworowi Rabinerson, 1927 (Chosa Herring) 220 Hereafter, “herring” will be considered to refer to the Atlantic Herring, Clupea harengus, L.1758, Clupeiformes: Actinopterygii (Fig. 5.1). And reference to any other species or subspecies will be noted by the use of the Latin binomial. Figure 5.1. The Atlantic herring, Clupea harengus L. The herring has a maximum recorded length of 45cm and weight of 1050g (Bigelow et al., 1963). They are distributed throughout the north Atlantic, on the eastern side from the Bay of Biscay northward to Iceland, on the western side from South Carolina to Greenland and Labrador, and throughout the Arctic Ocean as far as Nova Zemlya and Spitsbergen (Fig. 5.2) (Froese & Pauly, 2004). 221 Fig. 5.2. Representation of the distribution of Clupea harengus. Based on Froese & Pauly, 2004. 5.1.2 Prey and predation Herring change their feeding behaviour depending on their life-stage and what food is available. In the northeast Atlantic Ocean, herring larvae feed on small copepods, rotifers and eggs (Thiel et al., 1996). Adult herring feed mainly upon planktonic copepods, such as Calanus finmarchius, decapod larvae and fish larvae (Ammodytes, gadoids and clupeoids) and post-larval juvenile Ammodytes (Rice, 1963). In more turbid coastal waters, herring still feed on planktonic copepods, but polychaete and gastropod larvae make up a significant proportion of the diet (Fox et al., 1999). Herring are also known to feed heavily at times on benthic organisms (Dalpadado et al., 1996). Adult herring make daily vertical migrations in response to movements in the water column of plankton, and to avoid predation by seabirds (Cardinale et al., 2003). 222 As one of the most abundant species on the planet, herring are a significant food source for many species of higher predators. They are preyed upon by a wide range of marine mammals, sea birds and predatory fishes (Pierce & Santos, 2003; Lindstrøm et al., 2002; Nøttestad, 2002). As significant links in what can be very short food chains between plankton and large predatory organisms, herring are often host to larval parasites which mature in marine mammals or larger fish, such as the nematodes Anisakis spp., or cestode larvae. 5.1.3 Stock Assessment of Herring to the west of the British Isles Seasonal fisheries for herring take place in many areas to the west of Britain and Ireland. Herring stocks in this area were first considered for assessment by ICES in 1969 (Anon., 1970). The identity of stocks to the west of the British Isles was reviewed by ICES in 1994 (Anon., 1994). Historically, assessments were carried out on two stocks; “West of Scotland”, and “Celtic Sea”, ignoring catches of herring in the Irish Sea. The management areas for these stocks coincided with ICES division VIa to the north and west of Ireland, and VIIg-h and VIIa below 52°30'N for the Celtic Sea stock. The Irish Sea stock was first assessed in 1972, and was managed on the basis that two stocks, the “Manx” and “Mourne” stocks, coexisted in this area. In 1982 it was decided that it was not practical to manage these as separate stocks, and a single assessment was carried out. A population crash in the late 1970s caused the fishery throughout the western area to be closed for several years (Fig.5.3). 223 300 250 200 150 100 0 50 Landings (tonnes) 1975 1980 1985 1990 1995 2000 2005 Figure 5.3. Catches of herring to the west of the British Isles (ICES areas VIa, VIIa, b, g and j), 1973-2003. The majority of these catches come from VIa(N). Herring are assessed by ICES on the basis of a number of functional management units. These units are not necessarily based on discrete spawning stock structure, given the complex pattern of spawning behaviour of the herring, and the fact that fisheries tend to operate on non-spawning mixed stock fisheries. These management units correspond to ICES areas VIIa (S), VIIg and VIIj (the southern part of the Irish Sea, the Celtic Sea and southwest of Ireland), VIIa (N) (the northern part of the Irish Sea), VIa(S) and VIIb (north and west of Ireland) and VIa (N), to the west 224 of Scotland. The fisheries in the Clyde were formerly assessed as a separate stock, however the herring seems to have disappeared from the Clyde in recent years and this population is no longer assessed. 5.1.4 Spawning and Reproduction Herring in waters around the British Isles choose to spawn over gravel beds. A map of the distribution of suitable sediments is presented in figure 5.4 in pink. Figure 5.4. Distribution of gravel beds to the west of the British Isles (from British Geological Survey, NERC, 2005). 225 5.1.4a Division VIIa (N) (Irish Sea) Identification of spawning populations in the Irish Sea are mainly estimates, based on the presence of fish with stage VI ovaries in catches. Some areas, such as Douglas Bank, have been confirmed as spawning sites via the discovery of herring eggs in benthic grab samples. The main spawning ground for the Manx stock is the Douglas Bank, to the east of the Isle of Man. This was confirmed by Bowers (1969) who collected herring eggs in benthic grab samples. Other spawning areas of lesser importance are found in the Solway Firth, Luce Bay, the west coast of the Isle of Man and the north coast of Wales. The Mourne stock spawns off the coast of County Down, on an area known as “Kilkeel Hard” (Anon., 1979). Both the Manx and Mourne stocks spawn from September to November; however, there is some evidence for occasional winter spawning in this area (Anon., 1994). 5.1.4b Division VIIj and VIIa(S) (Celtic Sea) Herring in the Celtic Sea are thought to spawn along the coasts of Counties Wexford, Waterford and Cork, although the exact location of their spawning sites is not known. In division VIIj, herring spawn at Baltimore and in Bantry Bay. These grounds are also only known from the presence of stage VI fish in catches. Spawning in the Celtic Sea is very prolonged, and can take place from as early as September until February. In division VIIj the spawning period is shorter, with a peak in late October; however, some spawning appears to take place in the winter months off the Kerry coast (Anon., 1994). 226 5.1.4c Divisions VIIb and VIa (S) (North and West of Ireland) Again, the precise location of spawning in these two areas is not known; however, fisheries for stage VI fish take place around the mouth of the River Shannon, in Galway Bay and at Aran Island. In area VIa (south), spawning fisheries take place along the Donegal coast and at Aranmore Island (Anon., 1994). Timing and duration of spawning in these areas is variable. Spawning starts in late September in the northern part of VIa (S), through to late November for the southern part of VIIb. A spring spawning population has been reported in these areas on a number of occasions (Farran, 1946; Saville et al. 1966). 5.1.4d Division VIa (N) (West of Scotland) Spawning grounds in this area were described by Woods (1968) and Rankine (1986), based on a combination of distribution of recently hatched larvae, known gravel deposits and maturity stages of adult fish. There appears to be a number of spawning sites spread over a wide area, from Cape Wrath, the Minch, the west of the Hebrides, and at St. Kilda. It has been suggested that at the time of Rankine's 1986 study, spawning was restricted to east of the Minch, while previously extensive spawning had occurred to the west of this body of water (Anon., 1994). This area is mainly occupied by autumn spawners, spawning in September and October, considerably earlier than their neighbours in VIa(S). Historical reports refer to spring spawning fish in this area, and it may be the case that spawning time has switched in the past (Anon., 1994). 227 5.1.5 Parasites of herring The most recent and comprehensive review of the parasites of herring was carried out by MacKenzie (1985). This review expanded upon the checklist of herring parasites presented by Arthur and Arai (1984), listing over 80 species. (See table 5.1). Taxa Protozoa Species Ceratomyxa acadiensis Mavor, 1915 Ceratomyxa auerbachi Kabata, 1962 Ceratomyxa orientalis (Doigel, 1948) Eimeria nishin Fujita, 1934 Eimeria sardinae (Theolan, 1890) Goussia clupearum (Theolan, 1894) Kudoa clupeidae (Hahn, 1917) Trichodina ploveri Zhukov, 1964 Monogenea Gyrodactyloides andriaschewi Bychowski & Polyanski, 1953 Gyrodactyloides baueri Kulchkova 1970 Gyrodactyloides pertuschewskii Bychowski, 1947 Gyrodactylus cyclopteri Stsiborskaya, 1948 Gyrodactylus flesi Malmberg, 1957 Gyrodactylus gerdi Bychowski, 1948 Gyrodactylus groenlandicus Levinsen, 1881 Gyrodactylus harengi Malmberg, 1964 Gyrodactylus pterygialis Bychowski & Polanksi, 1953 Gyrodactylus pungitii Malmberg, 1964 Gyrodactylus robustus Malmberg, 1957 Gyrodactylus unicopa Glukhova, 1955 Laminiscus dogieli (Zhukov, 1960) Mazocraeoides georgi Price, 1936 Mazocraes harengi (van Beneden & Hesse, 1963) Digenea Cercaria doricha Rothschild, 1935 (metacercariae) Cercaria pythionike Rothschild, 1935 Cryptocotyle lingua (Creplin, 1825) Diplostomum spathaceum (Rudolphi, 1819) Galactosomum phalacrocoracis Yamaguti, 1939 Mesorchis denticulatus (Rudolphi, 1802) Prosorhynchoides basargini (Layman, 1930) Prosorhynchoides gracilescens (Rudolphi, 1819) Digenea Brachyphallus crenatus (Rudolphi, 1802) (adult worms) Derogenes varicus (Muller, 1784) Hemiurus appendiculatus (Rudolphi, 1802) Hemiurus levinseni Odhner, 1905 Hemiurus luehei Odhner, 1905 Lecithaster confusus Odhner, 1905 Lecithaster gibbosus (Rudolphi, 1802) Opechona bacillaris (Molin, 1859) Parahemiurus merus (Linton, 1910) Podocotyle atomon (Rudolphi, 1802) 228 Digenea Podocotyle reflexa (Creplin, 1825) (cont.) Pronoprymna petrowi (Layman, 1930) Cestoda Abothrium gadi van Beneden, 1870 (metacestodes) Grillotia erinaceus (van Beneden, 1858) Lacistorhynchus tenuis (van Beneden, 1858) Nybelinia surmenicola Okada in Dollfus, 1929 Scolex pleuronectis Muller, 1788 Cestoda Bothriocephalus scorpii (Muller, 1776) (adult worms) Eubothrium crassum (Bloch, 1779) Nematoda Anisakis simplex (Rudolphi, 1809) (larvae) Contracaecum osculatum (Rudolphi, 1802) Desmidocercella numidica (Seurat, 1920) Hysterothylacium aduncum (Rudolphi, 1802) Pseudoterranova decipiens (Krabbe, 1878) Nematoda Ascarophis pacificus Shukhov, 1960 (adult worms) Cystidicola farionis Fischer, 1798 Hysterothylacium aduncum (Rudolphi, 1802) Acanthocephala (juveniles) Corynosoma semerme (Forssell, 1904) Corynosoma strumosum (Rudolphi, 1802) Corynosoma villosum Van Cleave, 1953 Acanthocephala Echinorhynchus gadi Zoega in Muller, 1776 Echinorhynchus salmonis Muller, 1780 Neoechinorhynchus rutili (Muller, 1780) Pomphorhynchus kostylewi Petrochenko, 1956 Pomphorhynchus laevis (Zoega in Muller, 1776) Rhadinorhynchus trachuri Harada, 1935 Hirudinea Calliobdella vivida (Verrill, 1872) Branchiura Argulus alosae Gould, 1841 Argulus coregoni Thorell, 1864 Copepoda Bomolochus cuneatus Fraser, 1920 Caligus clemensi Parker and Margolis, 1964 Caligus elongatus Nordmann, 1832 Ergasilus sieboldi Nordmann, 1832 Lepeophtherius pollachius Bassett-Smith, 1896 Lernaeenicus sprattae (Sowerby, 1806) Table 5.1. Parasites which are known to infect herring (MacKenzie, 1985). 229 The herring clearly has a diverse parasite community. This is in part a reflection of the trophic position which the species occupies, which facilitates the transmission of parasites to higher predators, and in part a reflection of the huge commercial importance of the species, historically. This has led to all aspects of the biology of the herring being studied in detail for over a century. 230 5.1.6 Herring stock identity A considerable amount of research has been carried out on the complex of commercially important Atlantic herring stocks (Clupea harengus L.) to the west of Great Britain and around Ireland. Despite this intensive research, levels of mixing, recruitment patterns and genetic interactions within the complex are still poorly understood (see Anon., 1994, for a comprehensive review). The understanding of herring stocks to the west of the British Isles has changed considerably over the past forty years. Parrish and Saville (1965) considered the herring stocks to the west of the British Isles to consist of two main components, the ‘oceanic’ and ‘shelf’ populations. The oceanic populations consisted of a Scottish west coast winter/spring spawning stock and a southern Irish winter/spring spawning stock. The boundary between these two stocks was considered to be the central to southern west coast of Ireland. The other boundary was in the Irish Sea. Parrish and Saville (1965) proposed two stocks of ‘shelf spawners’, the Scottish west coast Minch stock and Irish Sea stock. Mixing between the principal North Sea stocks and west of Scotland spawners was suggested. The shelf spawners in the Irish Sea were subdivided into a northern Irish Sea stock, a southern Irish Sea autumn/winter spawning stock, a Clyde winter/spring spawning stock and a Minch summer/autumn spawning stock. ICES (Anon., 1979) examined age composition data from several of these spawning components to determine possible associations between fish subpopulations. They concluded that the south Minch and the northwest of Ireland were areas in which a complex stock mixing took place. Morrison and Bruce (1981) found evidence of mixing between Clyde, Irish Sea and Manx herring from the results of a physical tagging study. This was supported by King (1985) who examined 231 morphometric characteristics of fish from nine of the west coast spawning grounds and proposed a number of associations. Firstly, the Irish Sea spawning components (Manx and Mourne) were shown to be closely related with a certain degree of intermingling. The association between the Clyde stock and Irish Sea stocks indicated that the autumn spawners present in the Clyde could migrate out of the Clyde to spawn in these regions. There was a clear separation of the Dunmore East stock from the other three stocks on the west coast of Ireland, west Cork, Kerry and Galway. These were shown to be closely related suggesting they could be considered as one cohesive unit. Links between the western and North Sea herring populations were confirmed by the Bløden tagging experiement (Anon., 1975), which found a number of herring which had been tagged as juveniles on the eastern coasts of the North Sea, in spawning aggregations off the Hebridean coasts. A number of studies have demonstrated a drift of herring larvae away from Cape Wrath, along the north coast of Scotland, around the Orkney Islands and into the North Sea (Heath & MacLachlan, 1987; Heath & Rankine, 1988). Heath et al., (1987) reported that larvae of herring spawning to the west of the Hebrides are entrained into the Scottish coastal current and carried towards the North Sea, while those spawning in other areas are inshore of the main current flow, and the transport route of these larvae has not been convincingly identified. This current-driven transport of larval herring and evidence of spawning migrations in the opposite direction points to some degree of mixing between herring in the North Sea and to the west of Scotland. In summary, Parrish and Saville (1965), ICES (Anon., 1979) and King (1985) all suggested a mixing of herring stocks in areas VIa and VIIa North, with the Celtic Sea stock as a separate group. There were no conclusive agreements on the 232 boundaries of west coast of Ireland populations. However, it did appear that the separation between putative stocks probably occurred on the west coast of Ireland and in the southern Irish Sea. Herring to the west of the British Isles are currently managed on the basis of five different management areas; the west of Scotland (ICES Division VIa (North)); the Firth of Clyde; the area to the north and west of Ireland (ICES Divisions VIa (South) and VIIb, c); the Irish Sea to the north of 52o 30’N (ICES Division VIIa (North)), and finally, the Celtic Sea and Division VIIj, to the south of Ireland. Current management strategies for the exploitation of herring stocks in this area is by a combination of total allowable catch limits (TACs), nursery area and spawning site closures, closed seasons and week by week quota allocation to vessels. To the west of the British Isles, herring fisheries are prosecuted by vessels from France, Germany, Ireland, the Netherlands, Northern Ireland and Scotland. Assessments carried out on these stocks show a great degree of uncertainty as to the current level of exploitation (Anon. 2006b). This uncertainty may be compounded by incorrect division of the herring into currently applied stock units. Whilst fisheries are managed in terms of “management units”, the herring stock complex may not observe similar boundaries, or conversely, current boundaries may contain more than one single discrete population. For the development of proper management strategies, it is vital that the interactions between the currently recognised functional stocks are understood, both in terms of the mixing of adult fishes and the recruitment of juveniles. The application of management regimes based on erroneous concepts of population structures can lead to the overexploitation and extinction of component populations. 233 Two opposing theories of linkages between these different spawning components of Atlantic herring have been developed; the discrete population concept (Iles and Sinclair, 1982) and the dynamic balance concept (Smith and Jamieson, 1986). Current opinion (Anon, 2006b) is that neither of these concepts adequately explains herring population structure and dynamics and that they are better described using the metapopulation concept (see McQuinn, 1997), where population structure in a given area can be considered as an complex of local subpopulations linked by variable degrees of gene flow. 5.1.7 Parasite tag studies in herring Small pelagic marine fish species have been the subjects of many biological tag studies (see Williams et al., 1992 and MacKenzie, 2002), a number of which have focused on the clupeids, the Atlantic herring, Clupea harengus and the Pacific herring, C. pallasi (see MacKenzie 1987b for studies up to 1985; Moser & Hsieh, 1992). Anisakid nematode larvae have been the most frequently used tag parasites, but other groups used as tags include larval and adult digeneans, cestode plerocercoids, monogeneans and myxosporeans. MacKenzie (1985) used two species of renicolid metacercariae, Cercaria doricha and C. pythionike, and a cestode plerocercoid, Lacistorhynchus tenuis, in his study of herring recruitment around the Scottish coasts. The value of the metacercariae as tags derives from three features of the host-parasite relationship: (1) herring are only susceptible to infection in their first year of life and no further infection occurs thereafter; (2) the parasites have life spans in herring extending to several years and possibly as long as that of the host itself; (3) levels of infection vary significantly between different nursery areas. This means that herring are effectively tagged for life as juveniles and these parasites can be used to 234 trace samples of adult herring to their nursery grounds of origin. The plerocercoid of Lacistorhynchus tenuis is also known to infect herring as juveniles, but its value as a tag is constrained by the fact that further infection of herring as adults is possible. Because the study area of MacKenzie (1985) overlapped that of the present study, it was anticipated that these three parasite species would prove to be useful tags once again. 5.1.8 The WESTHER project This work was carried out as part of a wider stock assessment project on herring, applying a range of techniques to the same sets of fishes and using multivariate analyses to combine the findings of different work packages into a single overarching picture of herring stock identity. This was known as the WESTHER project (the full title of the project was “WESTHER – A Multidisciplinary Approach To The Identification of Herring (Clupea harengus L.) Stock Components West Of The British Isles Using Biological Tags and Genetic Markers”, EU Contract no. Q5RS-2002-01056). WESTHER aimed to provide information supporting an effective separation or aggregation of herring stocks in western European waters and a clear association between nursery areas and adult stocks. WESTHER consisted of seven separate work packages. The collection of samples was treated as the first component of the project, followed by morphometric analyses, parasites as biological tags, microsatellite genetics, otolith microstructure and otolith microchemical analysis. The final workpackage was a multivariate statistical analysis of the combined results of the individual components. 235 5.2 Methods 5.2.1 Sample collection To deliver the aims of a stock identification project, a sampling scheme which is appropriate to the biology of the species being studied, and appropriate to the aims of the study, needs to be devised. The WESTHER project aimed to identify discrete spawning populations, investigate the composition of non-spawning aggregations and to link recruits on nursery grounds to spawning adult populations. Because of these requirements, samples of all life stages were collected. In order to get an idea of the scale of differences within and between the contiguous populations being sampled, it was necessary to collect samples from populations which are known not to mix with populations to the west of the British Isles. Consequently, samples of spawning herring were collected in the Baltic Sea, near Rügen, and juveniles from the north coast of Norway. To investigate possible mixing of recruits from the North Sea with west-coast spawners, juveniles were collected from the east and west coasts of the North Sea. These juvenile outlier samples were used only in the parasitological workpackage. Samples were collected and processed at sea onboard research vessels, or were caught by commercial vessels and processed that day after the catch was landed. A schematic of the idealised sampling collection is shown in figure 5.1 (pers. comm., C. Zimmerman, Bundesforschungs Ansalt für Fischerei, Hamburg). One hundred fish were collected for spawning and juvenile samples, five hundred were collected from mixed stock aggregations. The number actually processed by each workpackage was variable, ranging between 30 and 500. Samples were collected at a variety of times between late 2002 and early 2005. 236 Figure 5.5. A map of the idealised sampling stations for spawning herring (red circles), juveniles (green squares) and mixed aggregations (blue triangles). Locations of outlier samples are shown on the inset with yellow inverted triangles (pers. comm., C Zimmerman, Bundesforschungsansalt fur Seefishcerei, Hamburg) 237 Overall, sample collection was very successful over the duration of the project. Eight of ten planned spawning areas were sampled, five of them at least twice. Those that were missed were due to a combination of factors such as herring not being available in the expected time and place, and previously occurring fisheries not taking place in the duration of the project, such that three sites; sample 7, off Barra Head, sample 8, to the west of the Hebrides and sample 9, north of the Butt of Lewis, were not sampled at all. A new site was included, off the coast of Skye, which added some information on Scottish coastal spawners that might otherwise have been lost. Six of the seven juvenile areas envisaged were sampled. All four mixed fishery areas were sampled, three of them completely, and most of the outgroup samples were also taken. In all, 5966 herring were collected (1377 spawning adults, 1716 juveniles, 2349 non-spawning adults and 524 outgroup herring). Realised location of sampling sites are shown for spawners, juveniles and mixed stocks in figures 5.2, 5.3 and 5.4 respectively. The details of sample collection are shown in table 5.1. It was the intention at the start of the project that each spawning site would be sampled in two separate years, but this could not be carried out for the Clyde, Rosamhil or Dingle spawning grounds due to a lack of the expected fishery or a shortage of experienced samplers. Similar problems were encountered trying to obtain some juvenile samples. On capture, fish were stored in iced sea water to prevent deformation caused by crushing, when stored in boxes, or caused by freezing when stored on ice. Basic biological data such as length and weight were recorded before each fish was pinned out and individually photographed for morphological analysis. One of the pectoral 238 fins was removed and stored in an Eppendorf tube with 80-100% ethanol for extraction of genetic material. The fish were then opened and maturity stage of the gonad was recorded. The viscera were removed and preserved individually in vials of 80-100% ethanol for parasitological study. Finally, otoliths were removed with ceramic forceps and placed in Eppendorf tubes padded with polymer-wool, all of which had been washed with nitric acid to prevent elemental contamination, for otolith morphometry and microchemistry, and for aging purposes. 239 Figure 5.6. Realised sampling positions for spawning samples. 240 Figure 5.7. Realised sampling positions for juvenile herring samples. 241 Figure 5.8. Realised sampling positions for mixed stock aggregation samples. 242 Sample Area Period Status No. Fish 3-S01B 3-S02A 3-S03A 3-S04A 3-S04B 3-S05A 3-S06A 3-S07A 3-S08A 3-S09A 3-S10A 3-S10B 3-M01A 3-M02A 3-M03A 3-M04A 2-J05B 3-J01A 3-J02A 3-J03A 3-J03B 3-J04A 3-J04B 3-J05A 3-J05B 3-J06A 3-J07A 3-X01A 3-X02A 3-X03A 3-X04A 4-S01A 4-S02A 4-S03A 4-S04A 4-S04B 4-S05A 4-S06A 4-S07A 4-S08A 4-S09A 4-S10A 4-S10B 4-M01A 4-M02A 4-M03A 4-M04A 4-J01A 4-J02A 4-J03A 4-J03B 4-J04A 4-J04B 4-J05A 4-J05B 4-J06A 4-J07A 4-X01A 4-X02A 4-X03A 4-X04A Celtic Sea SW Ireland Rosamhil Donegal Donegal Clyde Irish Sea Barra Head Western Hebrides Butt of Lewis Cape Wrath Cape Wrath VIIa(S),VIIb,c VIa(N) Irish Sea Celtic Sea Scottish Sea Lochs Baltimore Donegal Bay Irish Sea East Irish Sea East Irish Sea West Irish Sea West Scottish Sea Lochs Scottish Sea Lochs Minches Stanton Bank Western Baltic E. North Sea W. North Sea N. Norway Celtic Sea SW Ireland Rosamhil Donegal Donegal Clyde Irish Sea Barra Head Western Hebrides Butt of Lewis Cape Wrath Cape Wrath VIIa(S),VIIb,c VIa(N) Irish Sea Celtic Sea Baltimore Donegal Bay Irish Sea East Irish Sea East Irish Sea West Irish Sea West Scottish Sea Scottish Sea Minches Stanton Bank Western Baltic E. North Sea W. North Sea N. Norway Dec-03 Oct - Dec 03 Oct - Dec 03 Feb-03 Oct-03 Mar-03 Oct-03 Aug-03 Mar-03 Aug-03 Mar-03 Aug-03 Jul - Aug-03 Jul-03 Sep-03 Jul-03 Nov-02 Sep-03 Sep-03 Mar-03 Oct-03 Mar-03 Oct-03 Jan-03 Nov-03 Jul-03 Mar-03 Apr-03 Jul-03 Jul-03 Nov-03 Jan-Feb-04 Nov-04 Mar-04 Feb-04 Nov-04 Mar-04 Oct-04 Sep-04 Nov-04 Aug-04 Mar-04 Aug-04 Feb-04 Jul-04 Sep-04 Aug-04 Sep-04 Sep-04 Mar-04 Oct-04 Mar-04 Oct-04 Jan-04 Nov-04 Jul-04 Mar-04 Apr-04 Jul-04 Jan-04 Nov-04 Collected Not collected Not collected Not collected Collected Collected Collected Not collected Not collected Not collected Not collected Collected Not collected Collected Collected Collected Collected Not collected Not collected Not collected Collected Not collected Collected Collected Collected Collected Collected Collected Collected Not collected Collected Collected Collected Collected Collected Not collected Not collected Collected Not collected Not collected Not collected Collected Collected Collected Collected Collected Collected Collected Not collected Not collected Collected Collected Collected Collected Collected Collected Collected Collected Collected Collected Collected 100 NA NA NA 105 44 120 NA NA NA NA 102 NA 480 120 350 100 NA NA NA 78 NA 114 100 100 85 110 104 50 NA 50 105 105 105 105 NA NA 140 NA NA NA 101 101 161 445 153 485 105 NA NA 87 122 120 105 105 50 120 102 118 50 50 Table 5.2. Details of herring sample collection and sample size. 243 Sample Area Period Status No. Fish 5-S04A Donegal Feb-05 Collected 105 5-S10A Cape Wrath Mar-05 Collected 39 5-J03A Irish Sea East Mar-05 Collected 105 5-J04A Irish Sea West Mar-05 Collected 110 5-M03A Irish Sea Sep-05 Collected 155 5-X03A W. North Sea Jan-05 Collected 50 Table 5.2(cont). Details of herring sample collection and sample size. 244 5.2.2 Parasitological Examination Due to the constraints of sample collection, and the need to have material preserved in ethanol so that the calcareous corpuscles in the bodies of C. dorica and C. pythionike did not dissolve, ectoparasitic species were not recovered. To increase the number of specimens examined, to ensure comparable analyses and as a contingency against accidental losses, samples were divided upon collection and half of each sample was examined by the author at the University of Aberdeen. The remainder were examined by Dr Marcus Cross at the School of Biological Sciences, University of Liverpool. In the laboratory, the visceral organs of herring were removed from their ethanol vials and placed in tap water for several minutes to allow the tissue to soften, aiding examination. Viscera were then examined for helminth macroparasites under a dissecting microscope at magnifications of from x60 to x120, numbers of pyloric caeca counted and recorded, and smears of liver, spleen and gonad were examined for microparasites (myxozoans, protozoans, fungi) under a research microscope at x300 to x400. All parasites present were identified to species level, recorded, and all macroparasites counted. Samples of larval Anisakis sp. nematodes and renicolid metacercariae were preserved in 96% ethanol for development of molecular markers. The number of pyloric caeca in each herring was counted and recorded in the course of the examinations and used for morphological analysis. This procedure can be carried out much more quickly for juvenile than for adult herring, and more quickly for herring with smaller numbers of parasites than for those with heavy infections. The number of individuals of each of these parasite species in every fish has to be counted, which is obviously time-consuming. This is compensated for, however, when examining less heavily infected samples. As mixed 245 stock samples were generally heavily infected with Anisakis spp., time became a limiting factor, and so only 100 of the 500 fish collected were examined for the parasitological workpackage. Morphological identification of parasites was carried out by referring to scientific literature, including descriptions of parasites previously reported from herring. Permanent mounts of representative specimens of each helminth parasite species found, with the exception of Anisakis sp. nematode larvae, were stained with acetocarmine and mounted in DePeX. All specimens of Anisakis spp. larvae were preserved in ethanol and retained for investigation of molecular genetics, together with several hundred specimens of each species of renicolid metacercariae. 5.2.3 Analysis of parasitological data The measures of parasitic infection and ecological terms used in this study are as defined by Bush et al. (1997). Linear regression was used to determine relationships between infection and host age. The chi-squared test was used as a test of statistical significance between sample prevalences. The non-parametric KruskalWallis test was used to test for significant differences in mean intensity of infection between samples. A neural network analysis of spawning herring stock identity based on parasite faunal data was performed using the “nnet” package (Venables & Ripley, 2002) in the R statistical software environment (R Development Core Team, 2005). The methods used are based on those contained in Campbell et al. (2007) and chapter 4 of this volume. For this analysis, samples of spawning herring collected in the same functional unit were combined across years to increase sample sizes. A significant relationship was found between intensity of infection with Anisakis spp. and herring 246 length (fig. 5.5). Anisakis spp. numbers were logarithmically transformed in order to remove bias and to produce a linear relationship with length. A linear regression was performed on these transformed values, and the residuals of this regression taken forwards as a variable for the analysis. A data frame containing abundances of C. pythionike, C. doricha, H. aduncum and residuals of Ln(Anisakis +1) was created. A random selection of 50% of each spawning sample was selected and used as a training set to train the neural network, which was then used to classify the remaining 50%. 3 2 0 1 Ln (Anisakis+1) 4 5 This process was repeated 1000 times using different random sets. 22 24 26 28 30 32 Length Figure 5.9. Relationship between anisakis infection and fish age in combined samples. The residuals of this regression were used for the neural network analysis. 5.3 Results 5.3.1 Parasitology of herring A total of 3981 herring viscera samples were examined for parasitic infections. This comprised 1651 juvenile herring, 1314 spawning herring, 715 mixed adults and 300 outliers (table 5.3). A total of 14 parasite species were found in the herring examined (table 5.4). RFPL analysis of a sub-sample of anisakid nematode larvae from herring caught off the North West coast of Scotland positively identified the 247 specimens as Anisakis simplex sensu stricto (pers.comm. M.Cross, University of Liverpool). Only four of these species were commonly recorded, these were the nematodes Anisakis simplex, the digenean Hemiuris luehei and the renicolid metacercariae, Cercaria doricha and Cercaria pythionike. 248 Number of herring examined Sample Code 2002 2003 2004 2005 S01A - - 95 - 95 S01B - 104 - - 104 Total S02A - - 92 - 92 S03A - - 97 - 97 S04A - - 94 85 179 S04B - 103 - - 103 S05A - 43 - - 43 S06A - 104 138 - 242 S10A - - 124 37 161 S10B - 100 99 - 199 1315 Total 0 454 739 122 J01A - - 97 - 97 J03A - - - 104 104 J03B - 70 86 - 156 J04A - - 121 109 230 207 J04B - 87 120 - J05A - 100 103 - 203 J05B 100 99 101 - 300 132 J06A - 83 49 - J07A - 108 112 - 220 Total 100 547 789 213 1649 M01A - - 116 M02A - 110 98 - 208 322 116 M03A - 99 125 98 M04A - - 69 - 69 Total 0 209 408 98 715 X01A X02A - 104 13 102 50 - 206 50 X03A - - 49 44 93 X04A - 24 21 - 45 Total 0 128 172 94 394 200 1338 2068 527 4033 Grand Total Table 5.3 Numbers of herring viscera in each sample examined for parasites. Totals are in bold. Although they were common and showed significant differences between samples, Hemiuris luehei was discounted as a tag species as it showed a strong seasonal influence in its prevalence and abundance (figs. 5.11a & 5.11b) 249 40 30 20 0 10 Prevalence (%) 2 4 6 8 10 12 Month 0.5 0.0 Mean Intensity 1.0 1.5 Figure 5.10a. Prevalence of H. luehei in all samples from month to month. 2 4 6 8 10 12 Month Figure 5.10b. Mean intensity of H. luehei in all samples from month to month. 250 Liver smears revealed a number of fish infected with Goussia clupearum. The hardening of tissue caused by preservation in ethanol meant that creating smears was difficult or impossible for some fish, and this part of the examination was abandoned later in the project. The dehydrating effect of the ethanol meant that gall bladders could not be found in all specimens, and made the preparation of smears difficult. This part of the examination was also abandoned later in the project. Consequently, differences in prevalence of Goussia clupearum and Ceratomyxa auerbachi can not be used as informative tag species. The gall bladder of a single fish was found to be infected with a species of Myxobolus. This species was present at a very low intensity, so it was not possible to prepare permanent mounts or to identify the parasite to specific level. Parasite group Protozoa Myxosporea Digenea Nematoda Cestoda Acanthocephala Parasite species Goussia clupearum Ceratomyxa auerbachi Myxobolus sp. Cercaria pythionike Cercaria doricha Hemiurus luehei Derogenes varicus Brachyphallus crenatus Lecithaster confusus Pronoprymna ventricosa Anisakis simplex sensu stricto Hysterothylacium aduncum Lacistorhynchus tenuis Echinorhynchus gadi Site of infection Liver Gall bladder Gall bladder Visceral cavity Visceral cavity Stomach Stomach Stomach Intestine Pyloric caeca Visceral cavity Visceral cavity Visceral cavity, stomach & intestine Intestine Table 5.4. List of parasites infecting the viscera of herring caught west of the British Isles. The renicolid metacercariae C. pythionike and C. doricha were found encysted in the mesenterial tissue between the pyloric caeca. Intensity of infection ranged from zero to several hundred individuals per fish. The characteristic differences in shape, 251 size and cyst thickness meant that morphological identification to cercarial type was straightforward. The adults of Renicola spp. are extremely difficult to identify due to the calcareous corpuscles which occupy most of the body. Sequences of the Internal Transcribed Spacer (ITS) rDNA, a section of DNA commonly used for distinguishing between species, have been obtained from C. doricha and C. pythionike from herring, and from adult Renicola spp. from one fulmar Fulmaris glacialis, two common guillemots Uria aalge and one puffin Fratercula arctica (Campbell et al., 2004). Sequences from C. doricha and C. pythionike showed 88% similarity, indicating they belong to separate species. Sequences obtained from Cercaria pythionike were identical to those taken from all the adult renicolids, which were identical from all three seabird species, indicating that all the seabirds were infected with the same species. These have been previously described as separate species, however this molecular evidence suggests they should be synonymised, and the taxonomy of the genus Renicola reviewed in light of modern techniques. In order to confirm that renicolid infections remain constant for the life of the fish, abundances of C. pythionike and C. doricha were log transformed, equivalent to normalising the data, and the transformed value plotted against herring age (figs. 5.12 and 5.13). Linear regression revealed no significant relationship between age and abundance of either C. pythionike (r2=0.0021, p=0.55) or C. doricha (r2=0.016, p=0.09). This indicates that further infection with these parasites does not occur in adult herring, that heavily infected individuals are not subject to increased mortality, and that infections are therefore strongly indicative of nursery ground of origin. 252 5 4 3 2 0 1 ln(Cercaria doricha + 1) 2 4 6 8 10 12 Age Figure 5.11. Plot of log transformed Cercaria doricha abundances against herring age. There is no 3 2 0 1 ln(Cercaria doricha + 1) 4 5 relationship between the two variables (r2=0.016, p=0.09). 2 4 6 8 10 12 Age Figure 5.12. Plot of log transformed Cercaria pythionike abundances against herring age. There is no relationship between the two variables (r2=0.0021, p=0.55). 253 5.3.2 Long-term temporal variability Two of the areas in which juvenile herring were sampled between 1973 and 1982 as part of MacKenzie’s (1985) study were sampled again during this study. It was possible to compare prevalences of infection of C. doricha and C. pythionike in samples taken from Stanton Bank and the Scottish west coast sea lochs with those taken from those areas around 30 years previously (table 5.5). There were no significant differences between years in terms of prevalence of either Cercaria doricha (p=0.059) or Cercaria pythionike (p=0.282) at Stanton Bank (J-07), or in the eastern North Sea outlier (X-02). Although there were significant differences between years at the Scottish Sea Lochs (J-05) position (C. doricha p=0.002, C. pythionike p=0.039), prevalences were still similar. In contrast, prevalences of both parasites, but most noticeably that of C. doricha, had increased dramatically (p>0.001) between the 1973-1982 period and the time of the present study in the western North Sea juvenile outlier taken from the Moray Firth (X-03). This could be due to significant changes in the seabird community around the Moray Firth, and any seabirds whose numbers have changed dramatically in the area in recent years could be investigated as potential final hosts of the currently “orphaned” C. doricha. Although there are no data covering the intervening period and it is impossible to rule out wild fluctuations during this time, it would appear that prevalences of C. doricha and C. pythionike remain relatively stable in the short as well as long term. 254 Sampling area Sampling date Prevalence(%) C. pythionike C. doricha J05 (Scottish Sea lochs) 1973-1982 2002 2003 2004 52 60 55 72 78 90 89 83 J07 (Stanton Bank) 1973-1982 2003 2004 23 27 13 50 48 36 X03 (Moray Firth) 1973-1982 2004 2005 68 96 100 2 53 84 X02 (eastern North Sea) 1973-1982 2004 <1 0 0 0 Table 5.5. Comparison of prevalences (%) of Cercaria doricha and Cercaria pythionike at several sites between MacKenzie’s study (1985) and the present study. 5.3.3 Parasites as biological tags Due to the permanent nature of their infections, and the aspects of their life history which make the renicolid metacercariae such excellent tag species, A. simplex, C. pythionike, C. doricha and L. tenuis were selected as biological tags; their prevalence and mean intensities for spawning fish, juveniles, mixed stock aggregations and outlier samples are shown in tables 5.6, 5.7, 5.8 and 5.9. Sample code Mean age of herring 3-S01B 3-S04B 3-S05A 3-S06A 3-S10B 4-S01A 4-S02A 4-S03A 4-S04A 4-S06A 4-S10A 4-S10B 5-S04A 5-S10A 2.43 3.56 4.39 4.15 3.02 3.74 2.35 3.40 3.60 3.18 3.40 4.57 4.55 3.03 A. simplex C. pythionike C. doricha L. tenuis Pr(%) MI Pr(%) MI Pr(%) MI Pr(%) MI 68.6 72.6 72.1 81.7 94.0 68.1 89.1 75.3 64.9 60.2 64.5 98.0 71.8 91.9 6.8 11.3 9.7 19.5 12.9 7.4 9.3 8.5 4.9 11.2 10.7 17.0 6.5 14.0 34.3 45.1 7.0 0 0 55.3 10.9 29.9 58.5 0 58.1 0 36.5 21.6 10.0 8.0 1.0 0 0 16.0 8.3 12.8 10.5 0 7.8 0 14.0 6.4 4.8 21.6 0 0 0 6.4 9.8 12.4 19.1 0 25.0 0 14.1 5.4 40.4 4.9 0 0 0 1.8 24.4 19.7 4.7 0 8.8 0 13.3 8.0 0 0 0 8.7 0 0 1.1 0 0 4.3 0 0 0 0 0 0 0 1.6 0 0 1.0 0 0 1.2 0 0 0 0 Table 5.6. Mean age of sample, prevalence and mean intensity of the four selected biological tag species from samples of spawning herring. 255 Sample code A. simplex Pr(%) MI C. pythionike Pr(%) MI C. doricha Pr(%) MI L. tenuis Pr(%) MI 2-J05B 3-J03B 3-J04B 3-J05A 3-J05B 3-J06A 3-J07A 4-J01A 4-J03B 4-J04A 4-J04B 4-J05A 4-J05B 4-J06A 4-J07A 5-J03A 5-J04A 5.0 1.4 7.2 4.0 7.2 38.6 39.8 48.0 11.6 16.5 21.7 6.8 0 8.2 42.9 5.9 19.4 1.2 1.0 1.2 1.0 1.1 1.9 1.6 2.0 1.3 1.4 1.6 1.0 0 1.0 2.4 1.8 1.6 90.0 15.5 42.9 81.0 97.0 42.2 48.1 1.0 9.3 27.3 11.7 80.6 85.2 32.7 35.7 12.8 37.0 13.0 2.8 43.0 7.9 16.0 6.3 11.3 24.0 31.0 27.6 16.8 12.7 5.8 4.2 6.4 41.6 18.2 60.0 0 17.9 46.0 64.0 22.9 26.8 0 2.3 5.8 2.5 64.1 79.2 20.4 12.5 3.9 8.3 4.4 0 11.9 6.2 18.9 7.8 15.3 0 10.0 1.6 2.3 12.5 6.0 1.8 3.4 1.8 2.7 0 1.4 2.3 0 0 0 0 0 0 0 0 0 0 0 0 0 0.9 0 1.0 1.0 0 0 0 0 0 0 0 0 0 0 0 0 0 1.0 Table 5.7. Prevalence and mean intensity of the four selected biological tag species from samples of juvenile herring. Sample code 3-M02A 3-M03A 4-M01A 4-M02A 4-M03A 4-M04A 5-M03A Mean age of herring A. simplex C. pythionike L. tenuis C. doricha Pr(%) MI Pr(%) MI Pr(%) MI Pr(%) MI 90.9 80.0 80.2 95.9 49.6 75.4 48.0 12.2 9.0 6.1 17.6 5.5 7.5 6.6 9.1 5.0 33.6 1.0 9.6 10.2 15.3 5.4 11.6 7.3 42.0 10.9 14.9 12.1 4.6 1.0 15.5 1.0 4.0 5.8 6.1 4.0 1.0 5.2 27.0 12.0 4.5 21.3 1.8 2.0 0.9 1.0 3.2 0 5.1 1.0 1.0 1.0 1.0 1.0 0 1.2 3.8 4.4 3.8 4.2 2.0 3.4 3.4 Table 5.8. Mean age of sample, prevalence and mean intensity of the four selected biological tag species from samples of non-spawning herring aggregations. Sample code A. simplex C. pythionike L. tenuis C. doricha Pr(%) MI Pr(%) MI Pr(%) MI Pr(%) MI 3-X01A 4-X01A 4-X02A 96.2 100.0 2.0 21.4 37.2 1.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4-X03A 5-X03A 3-X04A 4-X04A 53.1 45.5 66.7 76.2 2.4 2.8 2.1 1.8 95.9 100 0 0 33.2 31.5 0 0 53.1 84.1 0 0 10.4 9.1 0 0 0 0 0 0 0 0 0 0 Table 5.9. Prevalence and mean intensity of the four selected biological tag species from samples of spawning (X01) and juvenile (X02-04) herring outliers. One of the first things to be apparent in the data from spawning samples is the difference between the 4 and 5-S10A samples and the 3 and 4-S10B samples, both 256 collected to the northwest of Scotland. The A sample was collected off the coast of Skye and represents Scottish west coast spring spawning herring populations, while the B sample off Cape Wrath is representative of Scottish west coast autumn spawners. The S10A samples show a moderate prevalence of A. simplex, between 60% and 90%, together with C. doricha and C. pythionike. The S10B samples have a prevalence of A. simplex above 90% in both years, and both renicolid metacercariae are absent. This immediately suggests that the autumn spawning herring on the Scottish west coast are not recruited from nursery grounds on the west coast of the British Isles, in all of which renicolid metacercariae are found, with the exception of the southern-most sample, off the coast of Dingle. The high prevalence of these metacercariae in the juvenile outlier sample collected in the Moray Firth would suggest that these fish have not recruited from the western North Sea. It could be suggested that these fish have recruited from juveniles near the eastern coast of the North Sea, or otherwise from a population which has not been sampled. The Irish Sea spawning samples (S06A) are the only other samples to have no infections with C. doricha or C. pythionike. These samples all show some infections with the cestode L. tenuis. The juvenile samples collected in the Irish Sea are the only ones which contain fish infected with this parasite, which, on its own, would support the idea that juvenile fish in the Irish Sea are recruited to the Irish Sea spawning stock. This assumption, however, is contradicted by the observation of both C. doricha and C. pythionike, at a prevalence of up to 42.9% and mean intensity of 43 parasites per fish in Irish Sea juveniles. The complete absence of these parasites in adults in this area, coupled with what is known about the nature of infection with these metacercariae, could only suggest that Irish Sea spawners can not be recruited directly from Irish Sea juveniles. 257 5.3.4.1 Temporal combination of samples A number of spawning sites were sampled in multiple years. In order to have confidence in pooling samples collected at different times it was necessary to compare samples from the same site taken in different periods (table 5.10). There were no significant differences in prevalence or abundance of any of the four tag species between 3-S04B, collected off Donegal in October 2003, and 4-S04A, collected at the same point in February 2004; therefore we are confident that these samples were taken from a single population spawning in this area over an extended period. Samples 3-S04B and 4-S04A were therefore combined for comparison with 5-S04A. There were no significant differences in mean intensity of any of the four species at any site across the two periods. The S10A samples were significantly different from the S10B sample collected in the same period, as previously mentioned, so these samples were compared separately. One S06 sample was taken in each period. Period Sample code Pr(%) P Pr(%) P Pr(%) P Pr(%) P ’03-‘04 ’04-‘05 3+4-S04 5-S04A 69.0 71.8 >0.05 51.3 36.5 <0.05 20.3 14.1 >0.05 0 0 NA ’03-‘04 ’04-‘05 3-S10B 4-S10B 94.0 98.0 >0.05 0 0 NA 0 0 NA 0 0 NA ’03-‘04 ’04-‘05 4-S10A 5-S10A 64.5 91.9 <0.001 58.1 21.6 <0.001 25.0 5.4 <0.01 0 0 NA ’03-‘04 ’04-‘05 3-S06A 4-S06A 81.7 60.2 >0.05 0 0 NA 0 0 NA 8.7 4.3 >0.05 A. simplex C. pythionike L. tenuis C. doricha Table 5.10. Significance of differences in prevalence between equivalent samples taken in different years The two S04 samples differed only in prevalence of C. pythionike, which was only just significant. The S10A samples, however, differed significantly in prevalences of 258 three of the four parasites. This would support the idea that the North West coast of Scotland is an area of stock mixing for herring. Seven of the sampling sites for juvenile herring were sampled more than once during the study. Prevalences of the four selected tag parasites were compared for temporal stability (table 5.11). The site showing the greatest variability was the western Irish Sea samples (J-04), where prevalences of three of the four parasites showed significant variation. J05 (Scottish Sea Lochs) showed highly significant variations in prevalence of two parasites, X03 showed a highly significant variation in one parasite, while J03 (Irish Sea East) and J07 (Stanton Bank) showed small but slightly significant variations in one parasite. Sample code Pr(%) 3-J03B 4-J03B 5-J03A 1.4 11.6 5.9 3-J04B 4-J04B 4-J04A 5-J04A 7.2 21.7 16.5 19.4 2-J05B 3-J05B 4-J05B 3-J05A 4-J05A 5.0 7.2 0 4.0 6.8 3-J06A 4-J06A 38.6 8.2 <0.001 42.2 32.7 >0.05 22.9 20.4 >0.05 0 0 >0.05 3-J07A 4-J07A 39.8 42.9 >0.05 48.1 35.7 >0.05 26.8 12.5 <0.05 0 0 >0.05 4-X03A 5-X03A 53.1 45.5 >0.05 95.9 100.0 >0.05 53.1 84.1 <0.01 0 0 >0.05 3-X04A 4-X04A 96.2 100.0 >0.05 0 0 >0.05 0 0 >0.05 0 0 >0.05 A. simplex C. pythionike P Pr(%) <0.05 15.5 9.3 12.8 <0.05 >0.05 42.9 11.7 27.3 37.0 90.0 97.0 85.2 81.0 80.6 C. doricha P Pr(%) >0.05 0 2.3 3.9 <0.001 <0.001 17.9 2.5 5.8 8.3 60.0 64.0 79.2 46.0 64.1 L. tenuis P Pr(%) P >0.05 1.4 0 0 >0.05 <0.001 <0.001 2.3 0 0 0.9 0 0 0 0 0 >0.05 >0.05 Table 5.11. Significance of differences in prevalence of four tag parasites between juvenile samples collected at the same sites in different years. 259 Although there are significant differences in prevalence between samples collected in different years at the same sites, it is apparent from figure 5.15 that the same general trends hold true for each area, for example, juveniles collected at site J05 (western Scotland sea lochs) are characterised by low prevalences of Anisakis simplex, prevalences of C. pythionike of over 80% and intermediate prevalences of C. doricha. For site J-04 it is obvious from the values in table 5.15 that the significant variation comes from the differences between the two B samples and that prevalences in the two A samples are very similar. Another interesting point is the almost identical infection levels of C. pythionike and C. doricha in samples 4-J03B and 4-J04B. From these results it appears that for further analysis the three J-03 samples can be combined across years, as could the two J-07, the two J-04A and the two X-04 samples. There are certain features of the parasite infections that can be used to characterize each sampling site. A. simplex is much more common in the more offshore sites J-06 and J-07, whilst Cercaria pythionike and Cercaria doricha have significantly higher prevalences in J-05 than elsewhere, and Lacistorhynchus tenuis only infects juvenile herring in the Irish Sea sampling sites J-03 and J-04. 260 100 80 60 Prevalence (%) 40 20 0 Anisakis C. pythionike C. doricha Figure 5.13. Temporal variation in prevalences of Anisakis simplex, C. pythionike and C. doricha. Red = J-03, Green = J-04, Blue = J-05, Cyan = J-06, Magenta = J-07, Yellow = X-03, Grey = X-04. Two of the sampling sites for mixed stock aggregations of herring were sampled in more than one sampling period – M-02A (VIa North) twice and M-03A (Irish Sea) three times – so again, prevalences of the four selected tag parasites have been compared for temporal stability (table 5.16). M-02A shows some stability, with only C. pythionike prevalences in the two samples being different, at the limits of significance. In M-03A prevalence of A. simplex in the 2003 sample was significantly greater than in the 2004 and 2005 samples, which were almost identical. This could be related to the higher mean age of this first sample (4.4 versus 2.0 and 3.4 respectively). These results suggest that in further analyses the two M02 samples can be considered as one, as could samples 4M03A and 5-M03A, or if A. simplex is removed from consideration there is no difference between any of the three M03A samples 261 Sample code A. simplex C. pythionike C. doricha L. tenuis Pr(%) P Pr(%) P Pr(%) P Pr(%) P 3-M02A 4-M02A 90.9 95.9 >0.05 9.1 1.0 <0.05 4.6 1.0 >0.05 1.8 1.0 >0.05 3-M03A 4-M03A 5-M03A 80.0 49.6 48.0 <0.001 5.0 9.6 15.3 >0.05 1.0 4.0 6.1 >0.05 1.0 1.0 1.2 >0.05 Table 5.12. Significant differences in prevalence between mixed stock aggregations collected in different years at the same sites. 5.3.3.2 Linking up life history stages One of the aims of the WESTHER project is to find markers linking fish on nursery grounds with spawning populations and non-spawning mixed stock aggregations. MacKenzie (1985) showed that C. doricha and C. pythionike are excellent tags for tracing recruitment patterns. A. simplex infections are not informative on recruitment, since these parasites are cumulative with age. The prevalences of the three remaining tag species were not significantly different between spawning samples collected in functional unit VIa(S) – VIIb-c (samples S-03 and S-04) and juveniles collected at Stanton Bank. As previously noted, there are obvious significant differences between the fish spawning in the Irish Sea and the juveniles which recruit there. The picture is reversed in fish from the Celtic Sea/VIIj functional unit, with significant differences between spawners and juvenile fishes, however this time juvenile fish have a significantly lower prevalence of C. doricha and C. pythionike than the spawning adults. The proximity of these two functional units could mean that there is a drift of larvae from the Celtic Sea into the Irish Sea, which recruit and migrate back to spawn. This would be supported by the finding of L. tenuis in a spawning sample to the south of Ireland – the only spawning fish infected with this parasite outside of the Irish Sea. There could also be 262 recruitment of juveniles from the Celtic Sea, with their very low prevalences of C. pythionike and C. doricha, to the Irish Sea spawning population. It is clear, however, that the southern Irish Sea is an area of stock mixing between populations with very different parasite prevalences. Spawners from pooled S-10B samples were almost identical to juveniles from the North Sea and Atlanto-Scandian outliers X-02A and X-04A, in that almost all the herring in these samples were infected with A. simplex to the exclusion of the other three tag parasites. This is supportive of a migration of herring from the northern North Sea to the north of Scotland for spawning Parasite tags are not suitable for assigning individuals from mixed stock aggregations to spawning stocks in this instance, due to the low number of suitable tag species found in this study, however, general patterns can be observed. Mixed stock aggregations were compared with spawner samples taken in the same functional unit and pooled across sampling periods on the basis of the findings presented in section 5.3.4.1 (fig. 5.17). Functional Unit VIa(S) / VIIb,c VIa North Parasite prevalences Sample Period 1 Period 2 As Cp Cd Lt Mixed 80.2 33.6 15.5 0.9 ’03-’04 spawners ’04-’05 spawners 70.7 44.6 17.7 0 03-’04 Mixed ’03-’04 spawners 90.9 76.8 9.1 28.1 4.6 11.6 1.8 0 ’04-’05 Mixed ’04-’05 spawners Irish Sea Celtic Sea / VIIj 03-’04 Mixed ’03-’04 spawners ’04-’05 Mixed ’04-’05 spawners Mixed ’03-’04 spawners ’04-’05 spawners 80.0 81.7 68.3 5.0 0 44.7 1.0 0 5.5 As Cp Cd Lt 71.8 36.5 14.1 0 95.9 96.3 1.0 21.3 1.0 5.1 1.0 0 48.9 60.2 12.1 0 4.9 0 4.0 4.3 75.4 10.2 5.8 0 89.1 10.9 9.8 1.1 2.0 8.7 0 Table 5.13. Differences in prevalences of tag species between mixed stock samples and spawning samples collected in the same functional unit. 263 One mixed sample was taken in each period from functional unit VIa North. These two samples were previously shown to differ only in prevalence of C. pythionike, at just outside the 95% confidence limits. These prevalences are similar to the prevalences obtained from the pooled spawning samples taken in this area. Given the large difference in prevalences between the S-10A and S-10B samples, this would suggest that this mixed stock aggregation is a combination of these two spawning populations, comprising fish recruiting from the west of Scotland and another area outside our study, but most likely the northern North Sea. Prevalences of all four tag parasites in the only mixed sample taken from areas VIaS and VIIb,c were very similar to those in spawner samples taken in both periods, the only slightly significant difference (p=0.046) being between prevalences of C. pythionike in 2003-04. This would suggest that this stock aggregation is formed from fish spawning to the north and west of Ireland. Only one mixed sample was taken in area Celtic Sea and VIIj, in 2004-05. It differed very little from the spawner sample taken in the same period, with only prevalences of A. simplex in the two samples showing a slightly significant difference (p=0.032), but a comparison with the spawner samples taken from the same area in 2003-04 showed a highly significant difference (p<0.001) between them in prevalences of C. pythionike. Mixed samples collected in the Irish Sea differ significantly from Irish Sea spawners in terms of C. doricha and C. pythionike. As previously mentioned, the recruitment patterns of fish in this area are not well understood, so this is perhaps to be expected. 264 5.3.3.3 Neural Network Analysis The parasites used for the creation of the data frame were C. pythionike, C. doricha, H. aduncum and residuals of Ln(Anisakis +1). The median percentage of each sample correctly classified over 1000 loops of the neural network function is shown in table 5.18. This shows that successful classification levels are low. Only in samples collected at sites S04B and S10A have more than half of fish examined correctly classified. The network is unable to correctly identify fish belonging to the outlier sample, which is known to come from a stock that does not mix with the western stock complex. This inability to successfully classify spawning fish correctly with any confidence means that this method is not suitable for classification of mixed stock aggregation samples. The relationship between A. simplex infection and length/age means that this approach will produce misleading results if it is used to classify juveniles. The results are encouraging, however, as the network reclassifies better than random chance would suggest, and it is likely that the incorporation of parasite data into a neural network analysis of multiple workpackages would be a productive approach to take. Classified as Sample classified S01A S01A S01B S02A S03A S04B S06A S10A S10B X01A 26.2 0.4 0.2 0 30 12 2.6 21.2 7.4 10 5.2 0.8 0 27.5 17.1 1.7 27.3 10.4 S02A 3 1.2 6.6 0.6 16 8.6 1.8 47.4 14.8 S03A 2.4 0.8 0.4 5.2 46.2 14.8 1 21.8 7.4 S04B 2.7 0.4 0.1 0.5 64 12 0.8 13.2 6.4 S06A 0 0.1 0.2 0.3 14.6 38.6 1.2 28.7 16.4 S10A 4.8 1 0.1 0.3 46.9 14.1 10 13 9.8 S10B 0 0 0 0 10.1 17.6 0.8 52 19.5 X01A 0.1 0 0 0.6 13.6 20.4 0.4 39.2 25.7 S01B Table 5.18. Median percentage classification of spawning herring to temporally pooled spawning samples, over 1000 repetitions of analysis. Successful classification is highlighted on the diagonal. 265 5.4 Discussion 5.4.1 Parasites of herring The finding of a Myxobolus spp. free living in the gall bladder is relatively uncommon. Over 450 species of Myxobolus have been described from fish hosts. Of these, most have been described from freshwater fishes (Landsberg & Lom 1991). This is the first record of a Myxobolus spp. infecting herring. The rarity and low intensity of this finding suggests that this was an accidental infection. This species does not provide any useful information on stock identity or distribution. The finding of Pronoprymna ventricosa is a new host record for herring, which has previously been reported from a range of small pelagic, mainly clupeid, fishes, and most frequently from shads, Alosa spp.. Bray & Gibson (1980) reported and described specimens from Alosa alosa and A. fallax caught in the Celtic Sea and the River Severn. The other adult digeneans found in the digestive tract are generalist species that have previously been reported from a wide range of species. The cestode plerocercoid, L. tenuis (fig. 5.19), reaches maturity in elasmobranchs, particularly spurdogs (Squalus acanthias) and thresher sharks (Alopias vulpinus). It is known to infect herring as juveniles, but its value as a tag is constrained by the fact that further infection of herring as adults is possible (MacKenzie, 1985). 266 Figure 5.18. Plerocercoid of Lacistorhynchus tenuis stained with acid alum carmine (x125). MacKenzie (1987c) recorded a decline in the prevalence of L. tenuis in herring from the North Sea. Between 1973 and 1978 prevalences remained at a stable level of around 10%, declining to a stable level of less than 1% in the years after 1978. MacKenzie (1987c) attributed this to decline to the change in the distribution of an unknown copepod, required as an intermediate host (Mudrey & Daily, 1971) associated with the end of the “great salinity anomaly” (Dickson et al., 1988). The prevalence of this parasite in this study remains about the level that MacKenzie (1987b) reported from 1978 onwards, however this could now be associated with the long-term decline in abundance of the spurdog (S. acanthias), the final host of this species (ICES, 2006a). Anisakis simplex is one of the most heavily studied of all fish parasites, and has been used as a tag species in herring a number of times (e.g. Grabda, 1974). It reaches maturity in the intestines of cetaceans, uses a wide range of intermediate hosts 267 and has a certain element of plasticity in its life cycle (Smith, 1983; Klimpel et al., 2004). The results obtained in this project support the use of biological tags for the separation of populations and following recruitment. They are less successful in assigning fish from mixed population to spawning stocks. This is due to the similarities in mean abundance of parasites that are observed in populations with a large number of individuals and a short life-span such as the herring. 5.4.2.1 Temporal stability The close similarity between prevalences of infection with C. pythionike and C. doricha in samples of juvenile herring taken from the same Scottish west coast sites over 30 years previously would suggest a the relationship between host and parasite is in equilibrium in this area. However, as we have no data for the intervening period, any conclusion drawn from these figures can only be tentative. On a shorter time-scale, samples of spawning, juvenile and mixed stock aggregations of herring were taken in the present study from the same sites in two, or in some cases, three consecutive years. In spawning herring the greatest temporal variations in prevalence of the four selected tag parasites were found in ICES area VIa North between the sampling sites off Skye and at Cape Wrath within the same year, and also between the two samples taken from off Skye in two successive years. Heath & Baird (1983) and MacKenzie (1985) both concluded that recruitment to the inshore fisheries in the Minch is from a variety of sources, including Scottish coastal nurseries and the eastern North Sea. Different proportions of recruits from different sources in different years could account for the variations in prevalence of infection observed in 268 our two samples from off Skye. The absence of three of our four tag parasites from the Cape Wrath samples is a clear indication of recruitment from outside Scottish coastal waters and is similar to the infections reported from the same area by MacKenzie (1985). Recruitment to adult herring populations in this area are therefore likely to be entirely from the eastern North Sea, where juvenile herring are not infected with either C. pythionike or C. doricha. This area is clearly one of intense stock mixing. These results are consistent with the otolith and body shape morphometry workpackage of the WESTHER project (pers. comm., S. Jansen, Bundesforschungsansalt fur Seefischerei, Hamburg), which found significant differences in the morphometry of herring from these two spawning samples. When these samples were temporally combined, there was significant misclassification between samples collected at the same site and season in different sampling periods. This would support the idea that area VIa North is home to two distinct spawning populations, exhibiting a degree of site fidelity in their spawning behaviour. In juvenile herring the most significant temporal variations were between samples collected in the western Irish Sea and also in the Scottish Sea Lochs. Most of the variation in Irish Sea West was between the two B samples taken in October in successive years. These results suggest that the population of juvenile herring present in the western Irish Sea in October is different from that present in March, when the A samples were taken. One of the Scottish Sea Lochs samples was taken from a different sea loch to the others (Loch Nevis, as opposed to Loch Hourn), which may account for some of the temporal variation observed in these samples. There was little variation between the two mixed stock samples taken from the VIa North area in successive years, suggesting a degree of stability in stock 269 compositions. In three samples taken from the mixed stock site in the Irish Sea, prevalence of A. simplex was significantly higher in the 2003 sample than in the following two years, but prevalences of the other three tag parasites did not vary significantly between the three samples. This difference is almost certainly due to the difference in age structure of the three samples: herring in 3-M03A had a mean age of 4.4 compared with 2.0 and 3.4 for 4-M03A and 5-M03A. 5.4.2.2 Spatial Variability A comparison of infections in spawning herring from different ICES areas showed clear differences between areas. A consistent feature was the marked difference between infections in Skye and Cape Wrath spawners, both taken within area VIa North at different times of year. Mixed stock and juvenile samples also showed markedly different patterns of infection between areas. Comparisons between prevalences of infection in spawner and juvenile samples showed close similarities between Donegal spawners and Stanton Bank juveniles in both temporal periods, and between Irish Sea spawners and Celtic Sea juveniles in the 2004-05 period. Unfortunately, relationships based solely on similarities between prevalences of infection can be biologically and hydrographically highly unlikely and almost certainly purely coincidental, such as those between Cape Wrath and Baltic spawners and juveniles of the two outlier samples from North Norway. In area VIa South/VIIb,c, spawning samples from Donegal and Rosamhil, mixed stock and juvenile (Stanton Bank) samples were similar in their parasite prevalences, and consistent in both sampling periods. In area VIa North, in contrast, there were significant variations in patterns of infection between spawners, mixed stock and juvenile samples. In the Irish Sea, the patterns of infection in juvenile 270 samples differed greatly from both spawning and mixed stock samples, which differed from one another to a lesser extent. In area Celtic Sea and VIIj, the single mixed stock sample, for which parasite data were available, showed a similar pattern of infection to Dingle spawners but was significantly different to Celtic Sea spawners, while the juvenile sample was significantly different to both. The presence of L. tenuis in both the VIa North and Irish Sea mixed samples suggests that adults which recruit from or spawn in the Irish Sea are present in both areas. Historical knowledge would suggest that this mixing is not unusual (Morrison and Bruce, 1981). An unlikely similarity was found between the VIa North mixed sample and Baltimore juveniles. 5.4.3 Conclusions The results of the work investigating the use of parasites as biological tags as part of the WESTHER project clearly support the use of this technique. The finding of significant differences in parasite prevalences and abundances between sampling sites allows us to reject the null hypothesis of this project: that the herring stock to the west of the British Isles constitute a single, well mixed population. Having rejected this hypothesis, differences in parasite distributions and prevalences allow a number of alternative population structures to be proposed. There were significant differences in prevalences of renicolid metacercariae between the Celtic Sea, Dingle, Irish Sea and Clyde spawners. This suggests that these spawning adults have retained the different early infection characteristics that can only have been picked up in different nursery areas. The significant differences between the adults therefore suggest that they are all separate stock components. The Rosamhil and Donegal samples were much closer, suggesting a these samples 271 constitute a single spawning population, although similar levels of infection do not necessarily mean that they belong to the same stock. There were also significant differences in infection between the Cape Wrath (S10B) and Skye (S10A) samples suggesting that they represent different stock components. The Cape Wrath spawners shared parasite characteristics with juveniles sampled in the North Sea suggesting that they spent some of their life history there; the two Skye samples were quite distinct from each other suggesting the area may be home to numerous stock components. Both the parasites as biological tags and otolith core microchemistry workpackages provide useful information on the nursery grounds in which herring developed. There is a clear distinction of many of the different juvenile samples, based on parasite prevalences. There was little temporal variation of parasite prevalence within most sampling areas, with the exception of the Irish Sea west (J04) samples. The wide scatter of sampling positions within this area probably accounted for much of this variation (fig. 5.3). Anisakis simplex is much more common in the more offshore juvenile samples in the Minch (J06) and Stanton Bank (J07) whereas C. pythionike and C. doricha have significantly higher prevalences in the Scottish sea lochs (J05) than elsewhere. Levels of A. simplex infection in the two Minch (J06) samples were significantly different. These samples were again taken from positions scattered over a very wide area, which would explain some of this variation. Comparing these nursery area samples to the spawning adults enables an assessment of recruitment of juveniles to different spawning components to be made, and the following hypotheses suggested. Firstly, that Stanton Bank (J-07 samples) is a nursery ground which provides recruits to both the north-west of Ireland and western Scotland spawning populations. Secondly, the Irish Sea contains juveniles which arise 272 from spawning in two different areas, most likely the Irish Sea and the Celtic Sea. The relationship between juveniles and spawners at the boundary between the Celtic Sea and west of Ireland is not clear. These hypotheses are supported by the findings of the otolith core microchemistry workpackage (pers. comm., A. Geffen, Institute of Marine Research, Bergen, Norway), which showed that the Irish Sea contains a mixture of juveniles spawned in the Irish Sea, the Celtic Sea and probably the Clyde, and that Stanton Bank and the Minches likely act as nursery areas for fish spawned to the northwest of Ireland and around Skye. Parasite data from “mixed” feeding aggregation samples suggests that fish from different spawning components do not remain distinct on the feeding grounds, as there is evidence to suggest mixing of adults from separate spawning components, especially in VIa North. There is also evidence to suggest that the Celtic Sea and VIIj spawning ground fish do not mix with the more northerly spawners to the same extent that the more northerly spawners mix with each other. The VIa South/VIIb, c mixed stock sample (M-01) was very similar in terms of its parasite prevalences to the Rosamhil (S-03) and Donegal (S-04) spawners. The occurrence of fish in this area infected with L. tenuis suggested a component of fish recruiting from either Irish Sea or eastern North Sea nursery grounds (MacKenzie, 1985). Samples from VIa North (M-02) were very similar to Cape Wrath spawners (S-10B), but the occurrence of some fish infected with C. pythionike and C. doricha indicated some mixing with Skye (S-10A) or Donegal spawners (S-04). Again, the occurrence of fish infected with L. tenuis suggested a component of recruits from either the Irish Sea or the eastern North Sea. Potentially then, fish originating from 273 spawning grounds in VIa North, in VIa South, VIIb, c and the Irish Sea could all be caught in mixed fisheries over the Malin Shelf. The Irish Sea mixed samples (M-03) were closest to Irish Sea spawners, but a small component infected with C. pythionike and C. doricha suggests the presence of some Celtic Sea spawners (S-01). The Celtic Sea/VIIj mixed sample (M-04) was most similar to the Dingle Bay spawners (4-S02A), but with slightly lower prevalences of all the tag parasites. The absence of L. tenuis indicated that no component of Irish Sea spawners could be detected with this method in the Celtic Sea and VIIj. The analysis of the otolith core microchemistry and body shape morphometry workpackages supported these findings. Otolith microchemistry results suggested that fish in VIa North (M-02) contained fish which had recruited from the Irish Sea, Scottish sea lochs, Minches and Stanton Bank, that 70% of the herring sampled in the Irish Sea (M-03) had recruited as Irish Sea juveniles. The Celtic Sea and VIIj samples (M-04) showed a mix of fish that had spent their juvenile period in both the Irish Sea and the Celtic Sea/VIIj (pers. comm., A. Geffen, Institute of Marine Research, Bergen, Norway). Body shape morphometry results suggested that fish sampled in VIa North (M-02) comprised a mixture of fish from all spawning samples analysed. Fish sampled in the Irish Sea (M03) showed a mixture of morphological features (pers. comm., S. Jansen, Bundesforschungsansalt fur Seefischerei, Hamburg). 5.4.4 Implications for management These results suggest that what are currently defined as separate stocks (VIa North, VIa South/VIIb, c and Irish Sea) mix at various periods in their life history and represent a metapopulation. One of the implications of these results would be the merging of these three stocks for assessment purposes. It is likely that any assessment 274 of this “super-stock” will be governed by the dynamics of the VIa North stock component as this is the largest, in terms of spawning stock biomass and abundance, by a factor of around ten (Anon., 2006). There is little evidence of mixing between these northern stocks and the population in the Celtic Sea/VIIj area, other than at a juvenile stage, which would support the continued assessment of the population in this area as a separate stock. Assessing the VIa North, VIa South/VIIb, c and Irish Sea herring as a single metapopulation should not be seen as a green light to freely transfer fishing effort from one component to another. There is a continued need to conserve the genetic diversity of the northern herring stocks. There are still numerous distinct spawning grounds in the area. Each of these spawning grounds is populated by a group of herring which, in general, have a certain level of site fidelity (Morrison & Bruce, 1981). To maintain the diversity it is necessary to ensure that single spawning populations are not targeted and fished to the level of extinction. Current theory suggests that if a population utilising a spawning ground is driven to extinction, with an increased level of the overall metapopulation biomass, the spawning ground will eventually be recolonised. The time-scale over which this process occurs is variable and unpredictable. In addition the replacement population will not have the same genetic characteristics as the previous population (Ruzzante et al., 2006). For these reasons, ICES has recently advised in situations where a large management area combines a number of possible separate population components, that a sensible management programme should “prevent local depletion of any unit”. This approach appears to be useful for the future management of the herring resources west of the British Isles. 275 Chapter 6. Is cytochrome oxidase I (CO1) mtDNA of Anisakis simplex sensu stricto an indicator of host population biology? 6.1 Introduction 6.1.1 Biology and Taxonomy of the Anisakidae The Anisakidae are a family of ubiquitous marine parasites, with a life cycle that includes free-living stages, euphausiids, marine fish and cetaceans (Smith, 1983; Klimpel et al., 2004). They are a group of interest due to their economic affect on fish prices (Adroher et al., 1996), their effect on human health, either through infection leading the disease anisakiasis or their inducement of an anaphylactic response in people with IgE sensitivity (Audicana et al., 2002), and also as biological tags for various aspects of fish life history studies. Recent developments in molecular biology, particularly multilocus allozyme electrophoresis (MAE), have allowed a revision of the taxonomy of the Anisakidae (Mattiucci & Nascetti, 2006). Type I and Type II morphotypes of Anisakis simplex are now recognised to comprise a complex of morphologically similar sibling species, each maturing in a particular cetacean final host. Using this approach, it has been demonstrated that the larval morphotype Anisakis Type I (sensu Berland, 1961) comprises the species: A. typica, A. simplex s.s., A. pegreffii, A. simplex C and A. ziphidarum (Mattiucci et al., 1997), whereas Anisakis Type II (sensu Berland, 1961) includes A. physeteris and A. brevispiculata (Mattiucci et al., 1986, 2001). Routine identification of Anisakis larvae by allozyme markers is particularly useful when different Anisakis species occur sympatrically in the same host and/or the same geographical area. Adult anisakids are known to be infective to 53 cetaceans and pinnipeds world wide while around 200 species of pelagic marine fish and 25 cephalopod species are known intermediate or paratenic hosts of larval stages for this parasitic nematode (Klimpel et al. (2004)). Fish generally acquire infections through the 277 consumption of infected euphausiids (Smith, 1983), although alternative lifehistories have been proposed (Klimpel et al., 2004). Herring (Clupea harengus) are thought to be paratenic hosts as nematodes undergo no development during infection until subsequent ingestion by the definitive host, a marine mammal. In fishes, anisakid larvae are most commonly found in the internal viscera of the host but may also be found in the musculature and gonads. Anisakids have been used as tags of marine fishes on numerous occasions (e.g. Beverley-Burton, 1978; Horbowy & Poldoska, 2001; Mattiucci et al., 2001, 2004, in press; MacKenzie et al., in press). 6.1.2 Studies of parasite population genetics Recent molecular studies on parasitic organisms in relation to the natural history of their hosts have substantially improved our understanding of the population structure of a number of animals (e.g. McCoy et al., 2003; Godfrey et al., 2006). The ability to measure genetic variability in parasite populations is important for studying the population biology and its underling mechanisms of the parasite and its host. Recent investigations have shown that ability of parasites to indicate subtle population structure of its host where analysis of said host revealed little or none (Criscione et al., 2006). This is due to the higher mutation rates of parasites compared to their hosts. The majority of work carried out on population genetics of parasitic nematodes has focussed on species which infect domestic and livestock animals or which pose a potential human health risk (Blouin et al., 1992, 1995; Anderson et al., 1998). Many genetic studies on fish parasites have focused the biogeographic distribution of closely related and cryptic species to infer relationships between host 278 population structures (e.g. Jousson et al., 2000; Sterud et al., 2002). Mitochondrial DNA has not previously been used to investigate the haplotype diversity in a marine nematode and their potential to discriminate between their host populations. 6.2 Methods 6.2.1 Collection of specimens Herring were collected as detailed in section 5.2.1. For genetic analysis, samples of Anisakis simplex were taken from spawning herring samples collected at the six most spatially distinct sites. These were Baltic Sea, Cape Wrath, Clyde, Irish Sea, Donegal and Celtic Sea (fig. 6.1). Visceral samples were preserved in 100% ethanol to preserve the structural integrity of Cercaria doricha and Cercaria pythionike, and to preserve genetic material contained within any parasites present in each fish. Anisakids were removed when the viscera was examined, pooled for each fish and placed in labelled Eppendorf tubes containing 100% ethanol for later DNA extraction and processing. 279 Figure 6.1. Location of spawning sample sites from where Anisakis simplex specimens were extracted and sequenced. 1. Celtic Sea, 2. Donegal, 3. Clyde, 4. Irish Sea, 5. Cape Wrath, 6. Baltic. 6.2.2 Extraction, sequencing and analysis of CO1 region Heads and tails of individual worms were removed and stored to allow morphological identification if any aberrant sequences were recorded. DNA was extracted from the mid-sections of individual Anisakis specimens using the DNEasy DNA extraction kit (Qiagen) as per manufacturers’ instructions, and DNA was eluted in a final volume of 200µl. Initially, isolation of the cytochrome c oxidase I (COI) region from Anisakis was performed using the universal primers (LCO1490 and HCO2198) and methods as described in Folmer et al., (1994). However, it was discovered that these primers were, in many cases, preferentially amplifying contaminating herring DNA. New primers were designed, AnCO1-9.10.27.28-2F (5’-ATTTGGTCTTTGATCTGGTATGG-3’) and AnCO1-9.10.27.28-2R (5’-TGGCAGAAATAACATCCAA- 280 ACTAG-3’). These primers were designed from COI nucleotide sequence which was relatively conserved between a range of nematode species. Secondly, the Anisakis COI sequence was aligned with a number of Clupeid COI sequences and primers were designed which displayed nucleotide variation between nematode and Clupeid sequences at important primer binding positions. The PCR reaction mix contained 1x Accuzyme Buffer (Bioline), 2.5U Accuzyme Taq (Bioline), 0.5µM each of primers AnCO1-9.10.27.28-2F and AnCO1-9.10.27.28-2R, 100µM dNTPs, 0.8mM MgCl2, 3µl template DNA (Qiagen purified), and dH20, to give a total reaction volume of 50µl. The cycling conditions were one cycle of 94ºC for 4 minutes, followed by 40 cycles of 94ºC for 1 minute, 64ºC for 1 minute and 72ºC for 2.5 minutes, and a final cycle of 72ºC for 5 minutes. PCR products were excised from the gel and purified using QiaQuick Gel Purification Kit (Qiagen). Purified products were visualised on a 1.5% agarose gel alongside a mass marker (Low DNA mass marker, Life technologies), to determine concentration of purified product. Extracted PCR products were sequenced using the Big Dye Terminator Cycle Sequencing Kit (PE Applied Biosystems) and electrophoresis on an AB1377 automated sequencer (Applied Biosystems). Sequence for both strands was obtained using the original PCR primers as forward and reverse primers in the case of PCR products. Template amounts were within the range suggested by the manufacturer for Big Dye Terminator Cycle Sequencing Kit. Cycle sequencing conditions were: 35 x [96°C for 30 sec, 50°C for 20 sec, 60°C for 4 min] and a final extension of 72°C for 5 minutes. Sequences were analysed using Sequencher (Gene Codes Corporation). The sequences obtained were compared with databases of known sequences, using 281 BLAST. Alignment of isolated sequences was carried out using CLUSTALW (Thompson et al., 1997). Intraspecific analysis was performed using Arlequin ver. 2.001 (Schneider et al. 2001) to calculate haplotype and nucleotide diversities, and molecular variance (AMOVA). Samples were pooled within sample areas and differences were measured using pairwise FSTs. In total 385 sequences were generated from six spawning sampling areas: Celtic Sea (S-01); Donegal (S04); Clyde (S-05); Irish Sea (S-06); Cape Wrath (S-10B) and Baltic Sea (X01) (table 6.1). Principal component analysis was carried out on Tamura and Nei genetic distance between individuals, and an analysis of similarity (ANOSIM) test carried out on the resulting groupings using the R statistical software environment the “vegan” package and the “Genetics” package (R development core team, 2005; Oskanen et al., 2006; Warnes & Leisch, 2004). 282 6.3 Results A PCR product of 1030 nucleotides was obtained following amplification of COI mtDNA from A. simplex. This represents a significant increase in the amount of sequence data currently available for Anisakis spp. COI, the only other available sequence consisting of approximately 500 nucleotides from A. simplex (Genbank accession No. AF096226), homologous to the 3'-end of the sequence isolated here. This new sequence has been submitted to the Genbank database under accession no. DQ489705. Sampling Area N Haplotype diversity (h) Nucleotide diversity (π) (x103) Clyde 11 0.9455 ± 0.065 9.233 ± 6.047 Celtic Sea 34 0.7736 ± 0.073 6.207 ± 4.129 Irish Sea 78 0.8288 ± 0.043 6.918 ± 4.415 Baltic Sea 40 0.8000 ± 0.068 6.289 ± 4.153 Cape Wrath 145 0.9321 ± 0.017 8.712 ± 5.270 Donegal 77 0.8712 ± 0.035 7.223 ± 4.567 Table 6.1. Sample sizes, haplotype diversity and nucleotide diversity at six sampling sites. Due to difficulties in generating good quality sequences for the whole length of the COI product from many individuals, comparisons have been carried using only the 280 b.p. region of the sequence which was common to all sequences. The COI region is highly variable, with a high proportion (ca. 50 %) of haplotypes unique to each sampling site. Recorded haplotypes are presented in appendix VIa. 283 Examination of the sequences revealed differences in presence and absence of haplotypes, and frequency of shared haplotypes, between the parasite populations (fig. 6.2). However, at present it is not possible to differentiate among distinct parasite populations (as delineated by spawning host populations), with the present data set. ARLEQUIN v. 2.000 software (Schneider et al., 2000) was used to calculate standard diversity indices (table 6.1) and pairwise FST values between samples, as well as a hierarchical analysis of molecular variance (AMOVA); the AMOVA partitioned the level of genetic differentiation among various groups based on geographical location and possible migration routes of host populations, including Baltic vs Atlantic Ocean (Celtic Sea, Donegal, Cape Wrath) vs Irish Sea (Irish Sea and Clyde), and Baltic vs Celtic Sea vs (Irish Sea and Clyde) vs (Donegal and Cape Wrath). 100% Rare AP AO AN AL AK AB AA 90% Percentage Frequency 80% 70% 60% 50% 40% 30% 20% 10% 0% Celtic Sea Donegal Clyde Irish Sea Cape Wrath Baltic Sea Figure 6.2. Percentage frequency of shared haplotypes at six spawning sites. Unique haplotypes (white) make up more than 50% of those recorded at some sites. 284 Source of Variation Percentage of Variation p Among groups (FCT) 0.41 0.206 -0.47 0.741 100.06 0.564 Among populations with groups (FSC) Within populations (FST) Table 6.2. Results of AMOVA. Most variation occurs within samples. There are no significant differences between groups. Nucleotide diversity (π) within populations was low (0.006-0.009), while haplotype diversity (h) was high (0.77-0.94), a pattern that is characteristic of recent population expansion. Pairwise FST values between sampling sites were not significant, indicating no genetically distinct populations. No significant genetic structure was found between groups of samples or among samples within groups indicating that most variation in the Anisakis CO1 was found within host populations rather than between host populations or groupings of host populations. A principal component analysis was carried out Tamura and Nei genetic distance between individual sequences (fig. 6.3). An ANOSIM test (Oskanen et al., 2006) was carried out on the resulting groupings. This revealed there were no significant differences between groupings of any combination of samples. 285 Figure 6.3. Principal component analysis of Tamura & Nei genetic distance between anisakids from 6 sampling sites. 1. Celtic Sea. 2. Donegal. 3. Clyde. 4. Irish Sea. 5. Cape Wrath. 6. Baltic. These preliminary results may not reflect the true structure of the anisakid populations. A significant (r2=0.976 p<0.001), positive correlation was found between number of different haplotypes and sample size, suggesting that too few parasites per sample were analysed to attain a reliable estimation of genetic variation within, and therefore between, each population. It is likely that the lifecycles of both parasite and host affect a low level of genetic differences between anisakid populations, thereby limiting their usefulness for fine-scale differentiation of herring host populations. 286 6.4. Discussion These results show that the mtDNA cytochrome oxidase 1 region of the Anisakis genome does not vary in ways which relate to the population biology of the herring. This method was applied in an attempt to find a marker with a finer resolution than the sibling species approach used by Mattiucci et al. (2004, in press), which applied multilocus allozyme electrophoresis to anisakids on a basinwide scale. Parasitic nematode populations are generally characterised by high haplotype and low nucleotide diversity (Blouin et al., 1992; Brant & Orti, 2003). High within-population diversities have been attributed to the rapid evolution of mtDNA, high rates of gene flow and large effective population sizes and this could be the case with A. simplex - cetacean final hosts have migratory routes which cross oceans (Kenney et al., 2001) and their guts play host to populations of many thousand of anisakids, giving large effective population sizes and allowing mating more or less at random. Significant host population structure has been observed using the mtDNA of parasites in species which have limited mobility, direct life cycles and small effective population sizes, such as the parasitic nematode of soildwelling insects Heterorhabditis marelatus (Blouin et al., 1999) or the bovine lungworm (Hu et al., 2002). These are not features that can be associated with the life-cycle of A. simplex. The results of the AMOVA on combined samples showed that all variation was observed within a population with only a small percentage among groups (table 6.2). These results are not unexpected due, in part, to the complex life history of A. simplex which results in large panmictic populations. 287 The finding that only a single species of Anisakis - A. simplex sensu stricto (pers. comm., M. Cross, University of Liverpool) - was found to infect herring in the study area on one hand simplified the interpretation of infection data for this nematode, but on the other limits the opportunity of using the nematodes as tags that the presence of other species would have afforded. It had been anticipated that one other species - A. pegreffii – might be found infecting herring in the Celtic Sea or Donegal parts of our study area, because Mattiucci et al. (in press) reported the finding of larvae of this species in horse mackerel Trachurus trachurus caught to the southwest of Ireland. This variation is repeated in studies of temporal variation of the Anisakis CO1 gene collected from fish at the same site in multiple years (Cross et al., in press) and analysis of 163 individual Anisakis CO1 sequences obtained from a single fish revealed a ratio of almost 1:1 between worms sequenced and haplotypes discovered (pers. comm., M. Cross, University of Liverpool). Investigation of other genetic regions, such as the ND4 region and microsatelites, commonly used as indicators of population biology in other species, also revealed no significant groupings of Anisakis simplex (pers. comm., C. Collins, FRS Marine Laboratory, Aberdeen). To conclude, the high motility of the hosts of A. simplex at all stages of their life cycle serve to maintain the great genetic diversity of the CO1 gene observed in this study. 288 Chapter 7. Incorporating the use of parasites as biological tags of horse mackerel and herring into fisheries management systems. 7.1 The Process of management. The diverse approaches taken to identification of fish stocks reflect the complex nature of the biological units which they represent. Whilst numerous definitions of what is meant by the term “unit stock” were given in chapter one of this work, it is often the case that fisheries managers and fisheries scientists mean different things when talking about a “stock” (Hammer & Zimmerman, 2005). This presents an immediate problem – how to translate the scientific findings and associated uncertainties generated by a stock identification programme into meaningful and practical fisheries policies in the real world. In the North Atlantic and associated seas, advice on fish stocks, including both herring and horse mackerel, is provided to decision-makers, such as the European Commission and the fisheries ministers of individual member states, by the International Council for the Exploration of the Seas (ICES). Stocks are assessed, in terms of population abundance and sustainability of harvest levels, on an annual or multi-annual basis by ICES working groups, their recommendations reviewed by further committees and then delivered into the hands of decision-makers for incorporation into policy. For research to be effectively considered when formulating management advice, it needs to be presented by scientists from diverse fields to ICES working groups, consisting typically of scientists with a mathematical background working to tightly defined terms of reference, in a form which is relevant to the outcomes which they are trying to achieve. In its most useful form, this compises evidence for or against changes in the way landings figures, market and discard sampling and fishery-independent data are aggregated spatially. The boundaries of regions used for this spatial aggregation have often been set on the basis of the historical operation of certain fisheries exploiting particular 290 species, and as such, may not reflect current status of fisheries and/or exploited populations, and furthermore lack the flexibility to do so (Hammer & Zimmerman, 2005). The process of obtaining stock identification information to provide to managers was framed by Kutkuhn (1981). Firstly, indications of possible stock structure, upon which testable hypotheses (and the null hypothesis of a lack of stock structure) can be made, need to be observed. Next, a well constructed survey, relevant to the needs of managers and the extent of the species being managed, needs to be carried out. Lastly, a robust discriminatory technique needs to be applied to the collected specimens. The question of what advice to pass to managers if the results do not allow the null hypothesis to be rejected is a difficult one. It may be that the particular technique being applied lacks the resolving power to support the existence of stocks, whilst another technique would find differences. Waldmann (2005a) states that non-existence of stocks cannot be proved in a formal sense. The lack of evidence for a separation of stocks given by a particular technique should not be seen as an implication that the null hypothesis is correct and be translated into management practice, only that it has not been falsified. For the successful incorporation of the results obtained by the use of parasites as biological tags into the management of horse mackerel and herring stocks, they need to be expressed in terms of current management practices, and whether they support the status quo, or changes in the ways stocks are assessed. 291 7.2 Informing Management of Horse Mackerel Until recently, the ICES Working Group on Mackerel, Horse Mackerel and Sardine (WGMHMS) faced considerable uncertainty in the distribution and extent of stocks (Anon., 1992; Anon., 2004). Management advice was given which assumed the existence of three putative stocks, western, North Sea and southern, on the basis of observed distribution of fertilised eggs and landings data. Evidence for or against this stock structure was lacking in many areas, as were well defined definitions of the boundaries between them. The information presented in this study helped in several ways. 7.2.1 North Sea vs. Western Stocks The parasitological data is a strong argument in favour of the delineation of stocks between the western region and the southern North Sea. The most effective biological tags to emerge from this study are the larval nematodes Anisakis spp. and Hysterothylacium aduncum. The distinctive pattern of infection with these nematodes observed in samples from the North Sea station 05 clearly distinguishes it from the nearest stations in the western stock, 01, 02, 03 and 06 and supports the current management strategy which assesses the North Sea population as a separate stock. Further sampling in the English Channel would be a useful contribution to the management of this species, both to investigate the position and nature of the boundary between stocks and to investigate stock mixing in non-spawning aggregations. The three anomalous fish in sample 05-00, identified by infections more characteristic of fish from the putative ‘western’ stock area, indicate that some migration does occur from western areas into the North Sea and provides a means of 292 estimating the extent of such migration. The fact that these were three of the oldest fish in the sample may also be significant, suggesting that migration is more likely with age. No fish could be identified in the western samples which had a characteristically “North Sea” pattern of infection. Uncertainty about the boundary between the North Sea and western stocks in the northern North Sea remains. Currently, WGMHMS assigns catches from the eastern Skaggerak to the western stock, and those from the western Skaggerak to the North Sea stock. This is related to the timing of the fisheries in these areas coinciding with fisheries elsewhere off the south-west coast of Norway and in the northern North Sea, thought to take place on fish from the western and North Sea stocks, respectively. Parasitological investigation of fish from the Skaggerak fishery would help to define the relationship between fish caught in this area and their parent stocks. 7.2.2 Western vs Southern Stocks The distinction between the ‘western’ and ‘southern’ stocks is less clear – the nematode fauna of fish from both putative stocks are dominated by Anisakis spp., and whilst the mean intensity of infection of fish from the southern stock is lower, it falls within the range of fish from western stock samples. The significant differences in levels and patterns of nematode infection between the two samples collected from station 21 in different years suggests that the southern Bay of Biscay is an area of mixing between these two stocks. Sample 21-00 had a similar pattern of nematode infection to samples 02-00 and 03-00, whereas sample 21-01 was markedly different from 02-01 and 03-01 but almost indistinguishable from 07-01 from northwest Spain. The parasite fauna of fish from station 07 is very similar to those of fish from station 08. This would suggest that the 293 horse mackerel found along the Gallician coast, ICES subdivision VIIId, belong to the southern rather than western stocks. ICES has agreed with this finding, which was reported by a number of workpackages in the HOMSIR project, and have revised their stock boundaries accordingly (Anon., 2005). 7.2.3 Extent of the Southern Stock Although there are significant differences between the parasite fauna of fish from sample 11-01 and those from samples in the southern stock, the results of this project shed little light on the position of the boundary between these stocks. The work of Sylianteng (2004) and MacDonald (2005) was useful in delineating stocks in this area, placing the boundary between southern and African stocks north of the town of Larache on the Morroccan coast. The similarities in parasite fauna between fish from site 17, in the Alboran Straight and those from the Portuguese coast, combined with the differences seen between this site and sites 12 and 20, in the Gulf of Lyon and suggests that fish in this area belong to the southern stock. 7.2.4 Migratory Behaviour in the Atlantic The monogenean Heteraxinoides atlanticus was found in over 5% of fish from stations 08, 09 and 10 taken off the coast of Portugal and was also found occasionally in fish from stations 01, 03, 17 and 21. This monogenean is a characteristic parasite of Trachurus spp. caught off the west coast of Africa to the south of the present study area. Gaevskaya and Kovaleva (1979c) reported 20-40% prevalences of H. atlanticus in T. trachurus caught off an area of West Africa to the south of our southernmost station 11-01. The same authors (Gaevskaya & Kovaleva, 1980) failed to find this species in samples of the same host caught further north as far as the North Sea. The 294 records of H. atlanticus from the present study are suggestive of migrations of T. trachurus from West Africa northwards as far as southwest Norway and into the extreme western part of the Mediterranean. The monogenean Cemocotyle trachuri is also more characteristic of Trachurus spp. caught outside the HOMSIR study area. Gaevskaya and Kovaleva (1979c) reported it from 24% of T. trachurus caught off Western Sahara and in 1-2% of T. capensis caught off Namibia. The same authors (Gaevskaya & Kovaleva, 1980) also reported a single specimen in a T. trachurus caught in the north-western North Sea. Gaevskaya and Kovaleva (1985) found C. trachuri in T. picturatus caught off Western Sahara and at the Azores. Our records of this monogenean come from off the coast of southwest Norway and off the coast of Portugal and are further evidence of extensive migrations of T. trachurus from West Africa into European waters. This is supported by the findings of MacDonald (2005), who compared the parasite fauna of fish off the African coast with the results obtained in this study and suggested similar migrations based on the observation of Rhadinorhynchus trachuri distributions on the African coast. Gaevskaya and Kovaleva (1980) reported the occurrence of R. cadenati in T. trachurus caught off the coast of West Africa coast, and from T. picturatus caught off Western Sahara and at the Azores (1985). R. cadenati also parasitizes a wide range of other teleost fish species off West Africa (Golvan, 1969). Two specimens of the acanthocephalan R. cadenati found in fish from sample 10-00 provide further evidence of northward migrations of T. trachurus from West Africa. Taken in combination, these results suggest that horse mackerel can be highly migratory, with strong evindence for northward migration from southern stocks over a distance of several thousand miles. This is important, not so much for formulating 295 management, as stock assessment techniques are generally insensitive to immigration and emigration, but for the interpretation of the results of other workpackages in the HOMSIR project. Small levels of migration between populations are often enough to prevent the development of genetic differences between them (Avise, 2000). The apparent lack of success in the genetic techniques in finding stock structure may be explained by the significant level of gene flow caused by highly migratory individuals travelling between stocks. This allows us to say with confidence that the lack of evidence of stock structure which these techniques provide should not be taken as a sign that horse mackerel exist in a single, mixed population. 7.2.5 Stock structure in the Mediterranean Sea In the Mediterranean, the definition of horse mackerel stock units has been largely absent. The General Fisheries Commission for the Mediterranean (GCFM) has established management areas based on political and statistical considerations rather than biological (Lleonart and Maynou, 2003). These take little biological information other than catch distribution into account. The results of the HOMSIR support the existence of sub-structuring in horse mackerel populations, in particular, the existence of three stocks located in the western, central and eastern basins. Were the GCFM to produce management advice for horse mackerel in the Mediterranean, this information will be useful. The myxosporean Alataspora solomoni and the pseudophyllidean plerocercoids were found only in samples 15 and 16 from the eastern Mediterranean. Alataspora solomoni was found only in the eastern Mediterranean. It was originally described by Yurakhno (1988) from T. mediterraneus ponticus in the Black Sea, who 296 recorded a prevalence of 42% in this host near Sevastopol. Unfortunately, we have no information on the occurrence of this species in the central and eastern Mediterranean. The copepod Lernanthropus trachuri was found only in fish from stations 18 and 19 in the central Mediterranean, where it was originally described from T. trachurus caught in the Ligurian Sea off the coast of Italy (Brian, 1903). The endoparasitic monogenean Paradiplectanotrema trachuri had a slightly wider distribution in the Mediterranean, but was most common in fish from stations 18 and 19, whilst the digenean Bathycreadium elongatum was found only in fish from stations 13 and 19 in the central Mediterranean. This is the first record of this species from T. trachurus, and its occurrence in this host only in the central Mediterranean probably reflects differences in the feeding habits of horse mackerel in this area. The parasite fauna of the putative western Mediterranean subpopulation of T. trachurus has not been characterised by this study due to the small size and young age of fish examined from areas 12 and 20, but appears to consist of a diverse digenean fauna. If there is further interest in stock definitions in the Mediterranean, obtaining larger specimens of horse mackerel from this area would be a priority. 297 7.3 Informing the management of herring fisheries A considerable amount of research has been carried out on the complex of commercially important Atlantic herring stocks (Clupea harengus L.) to the west of Great Britain and around Ireland. Despite this intensive research, levels of mixing, recruitment patterns and genetic interactions within the complex are still poorly understood (Anon., 1994). The results of the work investigating the use of parasites as biological tags as part of the WESTHER show clear differences in parasite fauna, prevalence and intensity of infections. The finding of significant differences in parasite prevalences and abundances between sampling sites allows us to reject the null hypothesis of this project: that the herring stock to the west of the British Isles constitute a single, well mixed population. Having rejected this hypothesis, differences in parasite distributions and prevalences allow a number of alternative population structures to be proposed. 7.3.1 Juvenile herring Juvenile herring show significant differences in parasite fauna, most significantly, in species which are known to be good tag species such as Cercaria doricha and Cercaria pythionike. Anisakis simplex is much more common in the more offshore juvenile samples in the Minch (J06) and Stanton Bank (J07) whereas Cercaria pythionike and Cercaria doricha have significantly higher prevalences in the Scottish sea lochs (J05) than elsewhere. Comparing these nursery area samples to the spawning adults enables an assessment of recruitment of juveniles to different spawning components to be made, and the following hypotheses suggested to managers. Firstly, that Stanton Bank is a nursery ground which provides recruits to both the north-west of Ireland and western 298 Scotland spawning populations. Secondly, the Irish Sea contains juveniles which arise from spawning in two different areas, most likely the Irish Sea and the Celtic Sea. The relationship between juveniles and spawners at the boundary between the Celtic Sea and west of Ireland area is not clear. There was little temporal variation of parasite prevalence within most sampling areas, with the exception of the Irish Sea west (J04) samples. The wide scatter of sampling positions within this area probably accounted for much of this variation. 7.3.2 Spawning aggregations of herring There were significant differences in prevalences of renicolid metacercaria between the Celtic Sea, Dingle, Irish Sea and Clyde spawners. This suggests that these spawning adults have retained the different early infection characteristics that can only have been picked up in different nursery areas. The significant differences between the adults therefore suggest that they are all separate stock components. The parasite fauna of Rosamhil and Donegal samples were much closer, suggesting a these samples constitute a single spawning population, although similar levels of infection do not necessarily mean that they belong to the same stock. There were also significant differences in infection between the Cape Wrath (S10B) and Skye (S10A) samples suggesting that they represent different stock components. The Cape Wrath spawners shared parasite characteristics with juveniles sampled in the North Sea suggesting that they spent some of their life history there; the two Skye samples were quite distinct from each other suggesting the area may be home to numerous stock components. 299 7.3.3 Mixed stock aggregations Parasite data from “mixed” feeding aggregation samples suggests that fish from different spawning components do not remain distinct on the feeding grounds, as there is evidence from tag parasite species to suggest the mixing of adults from separate spawning components, especially in VIa North. There is also evidence to suggest that the Celtic Sea and VIIj spawning ground fish do not mix with the more northerly spawners to the same extent that the more northerly spawners mix with each other. The Irish Sea mixed samples have a similar parasite fauna to Irish Sea spawners, but the presence of a small component infected with Cercaria pythionike and Cercaria doricha suggests the presence of some fish recruiting from juvenile grounds elsewhere, most likely Celtic Sea spawners. The parasite fauna of Celtic Sea/VIIj mixed sample was most similar to that of the Dingle Bay spawners, but with slightly lower prevalences of all the tag parasites. The absence of Lacistorhynchus tenuis indicated that no component of Irish Sea spawners could be detected with this method in the Celtic Sea and VIIj. This suggests that the Celtic Sea population is relatively isolated and in terms of inward migrations. The VIa South/VIIb, c mixed stock sample was very similar in terms of its parasite prevalences to the Rosamhil and Donegal spawners, also in VIa South/VIIb, c. The occurrence of fish in this area infected with Lacistorhynchus tenuis suggested a component of fish recruiting from either Irish Sea or eastern North Sea nursery grounds (MacKenzie, 1985). Samples from VIa North (M-02) were very similar to Cape Wrath spawners (S-10B), but the occurrence of some fish infected with Cercaria pythionike and Cercaria doricha indicated some mixing with Skye (S-10A) or Donegal spawners (S-04). Again, the occurrence of fish infected with 300 Lacistorhynchus tenuis suggested a component of recruits from either the Irish Sea or the eastern North Sea. Potentially then, fish originating from spawning grounds in VIa North, in VIa South, VIIb, c and the Irish Sea could all be caught in mixed fisheries over the Malin Shelf. This is an important finding, and raises questions on how catches in this area are assigned by the working group to a spawning population. 7.3.4 Management implications These results suggest that what are currently defined as separate stocks (VIa North, VIa South/VIIb, c and Irish Sea) mix at various periods in their life history and represent a metapopulation. One of the implications of these results would be the merging of these three stocks for assessment purposes. It is likely that any assessment of this “super-stock” will be governed by the dynamics of the VIa North stock component as this is the largest, in terms of spawning stock biomass and abundance, by a factor of around ten (Anon., 2006). There is little evidence of mixing between these northern stocks and the population in the Celtic Sea/VIIj area, other than at a juvenile stage, which would support the continued assessment of the population in this area as a separate stock. The connection between populations in VIa North and the northern North Sea remains unresolved, and future research should focus on this region. 301 7.4 Novel Techniques Two new techiques were applied to the use of parasites as biological tags in these studies, namely neural network analysis and the investigation of parasite genetic variation as an indicator of host population biology. While the application of neural networks can be seen as a success, the use of Anisakis CO1 variation can not be recommended 7.4.1 Neural network analysis Neural network analysis is a statistical classification tool which is particularly robust when faced with normal and non-normal data, factors and combinations of the three. The inherent overdispersal of parasite data makes the application of this approach to problems in parasitology a worthwhile line of future study. The drawback of this method is the difficulty in interpreting the way in which the neural network operates. Networks with a number of “hidden layer units” become increasingly complex to the point where they become “classifying black boxes”, where no information about how the network arrived at the classifications it did can be extracted. In roles such as the assignment of fish to stock on the basis of their parasites, this is not a major drawback, and the technique can be used with confidence. 7.4.2 Parasite population genetics The investigation of Anisakis simplex CO1 mtDNA as an indicator of herring population biology was not successful, in as far as it did not reveal meaningful population structures, however that is not to say that the technique was a failure and should not be tried again. Anisakis simplex has a complex life-cycle, using many other 302 fish hosts as well as herring. The cetacean final hosts are highly migratory, which is likely to supress the development of genetic differences on the scale of herring stock units. It could have been the case that investigating the population genetics of a more herring-specific parasite such as the renicolid metacercariae, Cercaria pythionike and Cercaria doricha, would have provided more meaningful information. At the outset of this study it was unclear how widespread these two species would be, whereas we were confident of finding A. simplex in nearly every sample. With hindsight, this may have been a more worthwhile path to follow. 7.4 Conclusions The success of these projects in introducing information from parasites as biological tag studies into management advice and policy underlines the usefulness of this technique in stock discrimination studies. It becomes an even more useful technique when used as a component of a multidisciplinary study which applies a range of approaches to a stock to produce information which cross-validates and strengthens the picture of population structure which is obtained. Recent advances in analytical methods such as neural networks and multiple logistic regression allow comparisons between parasite faunal composition of large samples of fish, resulting in powerful discrimination techniques now being applied to numerous species. The results of these projects also show that the use of parasites as biological tags is scalable, revealing population structure on oceanic, basin and coastal scales. 303 Bibliography 7. Bibliography Bibliography Abaunza, P., Villamor, B. and Pérez, J.R. (1995). Infestation by larvae of Anisakis simplex (Nematoda: Ascaridata) in horse mackerel, Trachurus trachurus and Atlantic mackerel, Scomber scombrus, in ICES Divisions VIIIb, VIIIc and IXa (N - NW of Spain). Scientia. Marina 59: 223-233. Abaunza, P., Garcia, M.T., Murta, A., Ramos, C.P.R., dos Santos, A.M.T., Pereira, A., Quinta, R., Molloy, J., Nascetti, R., Cimmaruta, R., Mattiucci, S., Magoulas, A., Kotoulas, G., San Juan-Lopez, A., Comesana, S., MacKenzie, K., Campbell, N., Gordo, L, Hammer, C., Zimmerman, C. Stransky, C. (1999) A multidisciplinary approach using genetic markers and biological tags in horse mackerel (Trachurus trachurus) stock structure analysis. Final Report of the HOMSIR Project to the European Commission, QLK5-Ct1999-01438. Abaunza, P., Gordo, L., Karlou-Riga, C., Murta, A., Eltink, A.T.G.W., García Santamaría, M.T., Zimmermann, C., Hammer, C., Lucio, P., Iversen, S.A., Molloy, J. & Gallo, E. (2003) Growth and reproduction of horse mackerel, Trachurus trachurus (Carangidae). Reviews in Fish Biology and Fisheries 13:2761. Abaunza, P., Murta, A., Campbell, N., Cimmaruta, R., Comesaña, S., Dahle, G., Gallo, E., Gordo, L., Iversen, S., MacKenzie, K., Magoulas, A., Mattiucci, S., Molloy, J., Nascetti, G., Pinto, A.L., Quinta, R., Ramos, P., Ruggi, A., Sanjuan, A., Santamaría, M.T., Santos, A.T. Stransky, C. and Zimmermann, C. Considerations on sampling strategies for an holistic approach to stock identification: the example of the HOMSIR project. Fisheries Research. (in press). Abollo, E., Calvo, M. & Pascual, S. (2001) Hepatic coccidiosis of the blue whiting, Micromesistius poutassou (Risso), and horse mackerel, Trachurus trachurus (L.), from Galician waters. Journal of Fish Diseases 24:335-343. Adroher, F.J., Valero, A., Ruiz-Valero, J. & Iglesias, L. (1996) Larval anisakids (Nematoda:Ascaridoidea) in horse mackerel (Trachurus trachurus) from the fish market in Granada, Spain. Parasitology Research 82:319-322. Al-Bassel, D.A.M. (2000) On Myorhynchus pritchardae and Podocotyloides chloroscombri (Digenea: Trematoda) described from new hosts from the Mediterranean Sea in Libya. Veterinary Medical Journal of Giza 48:247–252. Alegria-Hernandez, V. (1984) Some aspects of horse mackerel (Trachurus trachurus L.) biology in the middle Adriatic. FAO Fisheries Report 290:123–125. Alegria-Hernandez, V. (1994) Reproductive cycle and changes in conditions of the horse mackerel (Trachurus trachurus L.) from the Adriatic Sea. Acta Adriatica 35:59-67. Alvarado-Bremmer, J.R., Baker, A.J. & Mejuto J. (1996). Mitochondrial DNA control region sequences indicate extensive mixing of swordfish (Xiphias gladius) populations in the Atlantic Ocean. Canadian Journal of Fisheries and Aquatic Science 52:1720-1732. 305 7. Bibliography Anderson, T.J., Blouin, M.S. & Beech, R.N. (1998) Population biology of parasitic nematodes: applications of genetic markers. Advances in Parasitology 41: 219-283. Angel, L.M. (1960) Cercaria haswelli Dollfus, 1927; a re-examination of Haswell’s material, with a discussion of the genus Tergestia. In: Libro Homenaje al Dr Eduardo Caballero y Caballero, Jubileo 1930-1960 Secretaria de Educacion Publica, Instituto Politechnico Nacional, Mexico City, Mexico. pp.95-111. Anon. (1967) Study on identification of the Pacific salmon by neutron activation analysis. Survey and Research Division, Fisheries Agency of Japan, Tokyo. Anon. (1970) Report on the state of the herring fisheries north-west of Ireland and west of Scotland, December 1969. ICES Li:6. Anon. (1975). Report of the ICES working group on the Bløden tagging experiment, 1969/70. ICES Cooperative Research Report Vol 47. 136pp. Anon. (1979) Final Report of the Cross Border Herring Study. Fisheries Development Ltd. London. 160pp. Anon. (1992) Report of the Study Group on Stock Identity of Mackerel and Horse Mackerel. ICES C.M. 1992/H:4. Anon. (1994) Report of the Study Group on Herring Assessment and Biology in the Irish Sea and Adjacent Waters. ICES C.M. 1994/H:5. Anon. (1995) Guidelines on the precautionary approach to capture fisheries and species introduction. FAO Fisheries Technical Paper 350(1): 52 pp. FAO, Rome, Italy. Anon. (1998) Working Group on the Assessment of Mackerel, Horse Mackerel, Sardine and Anchovy. ICES C.M. 1998/ACFM:6. Anon. (1998b) Consumption Rares of Predatory Fish Relevant for multispecies assessment in the North Sea and the Atlantic off Spain and Portugal (CORMA). – final report. FAIR-CT 95-0604. 92pp. Anon. (2002) FAO Yearbook of fishery Statistics, “Catches and Landings”. Vol.80 ftp://ftp.fao.org/fi/stat/windows/fishplus/fstat.zip. Anon. (2004) ICES Working Group on Horse Mackerel, Sardine and Anchovy Report. ICES C.M. 2004/ACFM:8 pp. 176-195. Anon. (2006a) Report of the Working Group on Elasmobranch Fishes (WGEF). ICES C.M. 2006/ACFM:31. Anon. (2006b) Report of the Herring Assessment Working Group for the Area South of 62ºN (HAWG). ICES C.M. 2006/ACFM:20. 306 7. Bibliography Arai, H.P. (1989) Acanthocephala, p.1-90. In Margolis L. & Kabata Z. [ed.] Guide to the parasites of fishes in Canada. Part III. Canadian Special Publication of Fisheries and Aquatic Sciences 107. 95pp. Arneberg, P., Skorping, A., Grenfell, B & Read, A.F. (1998) Host densities as determinants of abundance in parasite communities. Proceedings of the Royal Society B. 265:1283-1289. Arthur, J.R. (1997). Recent advances in the use of parasites as biological tags for marine fish. In Diseases in Asian Aquaculture III (Flegel, T.W & MacRae, I.H., eds.), pp.141-154. Fish Health Section, Asian Fisheries Society, Manila, Phillipines. Arthur, J.R. & Albert, E. (1993) Use of parasites for separating stocks of Greenland halibut (Reinhardtius hippoglossoides) in the Canadian Northwest Atlantic. Canadian Journal of Fisheries and Aquatic Science 50:2175-2181. Arthur, J.R. & Arai, H.P. (1980) Studies on the parasites of Pacific herring (Clupea harengus pallasi Valenciennes): survey results. Canadian Journal of Zoology 58 64-70. Arthur, J.R. & Arai, H.P. (1984) Annotated checklist and bibliography of parasites of herring (Clupea harengus L.). Canadian Special Publication of Fisheries and Aquatic Sciences 70 26pp. Atmar, W. & Patterson, B.D. (1993) The measure of order and disorder in the distribution of species in fragmented habitat. Oecologia 96:373-382. Audicana, M.T., Ansotegui, I.J., de Corres L.F. & Kennedy, M.W. (2002) Anisakis simplex:dangerous dead or alive. Trends in Parasitology 18:20-25. Avise, J.C., Walker, D. & Johns, G.C. (1998) Speciation durations and Pleistocene effects on vertebrate phylogeography. Proc R Soc Lond Ser B 265:1707–1712. Avise, J.C. (2000) Phylogeography. The history and formation of species. Harvard University Press, Cambridge, Massachusetts, USA. 464pp. ISBN 0674666380. Bailey, K.M, Quinn, T.J., Bentzen, P. & Grant, W.S. (1999) Population structure and dynamics of walleye pollock, Theregra chalcogramma. Advances in Marine Biology 37: 179-255. Bailey, R. S. (1992) The global pelagic fish resource and its biolgical potential. In: (J.R. Burt, R. Hardy and K.J. Whittle (eds.)): Pelagic Fish. The resource and its exploitation pp. 1-20. Fishing News Books, Univ. Press, Cambridge. Bahri, S. & Marques, A. 1996. Myxosporean parasites of the genus Myxobolus from Mugil cephalus in Ichkeul lagoon, Tunisia: Description of two new species. Diseases of Aquatic Organisms 27(2): 115-122. 307 7. Bibliography Balbuena J.F., Anzar, F.J., Fernandez, M. & Raga, J.A. (1995) Parasites as indicators of social structure and stock identity of marine mammals. In Whales, Seals, Fish and Man: Developments in Marine Biology eds. A.S.Blix, L. Walloe & O Ulltang. pp133-140. Elsevier, Amsterdam, Holland. Bartoli, P. & Bray, R.A. (2004) Ancylocoelium typicum Nicoll, 1912 (Digenea: Monorchiidae), a poorly known parasite of Trachurus spp. (Teleostei: Carangidae) from the western Mediterranean and north-eastern Atlantic, and observations on its taxonomic position. Systematic Parasitology 58:23-39. Bartoli, P., Bray, R.A. & Gibson, D.I. (2003) Opecoelidae (Digenea) from western Mediterranean fishes: three rare species. Systematic Parasitology 55: 81– 95. Bartoli, P., Gibson, D.I. & Bray, R.A. (2005) Digenean species diversity in teleost fish from a nature reserve off Corsica, France (Western Mediterranean), and a comparison with other Mediterranean regions. Journal of Natural History 39:4770. Bass, C. (1995). Ecological structures: expansion and replacement. Scientia. Marina 59: 373-380. Beacham, T.D., Withler, R.E. & Gould, A.P. (1985) Biochemical genetic stock identification of chum salmon (Oncorhynchus keta) in southern British Columbia. Canadian Journal of Fisheries and Aquatic Sciences. 42:437-448. Begg, G.A. (2005) Life History Parameters. In: (eds. Cadrin, S.X., Friedland, K.D. & Waldman, J.R.) Stock Identification Methods – Applications in Fishery Science. Elsevier Academic Press, Amsterdam. pp. 119-150. Begg, G.A., Friedland, K.D. & Pearce, J.B. (1999) Stock identification and its role in stock assessment and fisheries management: an overview. Fisheries Research 43:1-8. Ben Salem, M. (1988). Régime alimentaire de Trachurus trachurus (Linnaeus, 1758) et de T. mediterraneus (Steindachner, 1868), (Poissons, Téléostéens, carangidae) de la province Atlantico-Mediterranéenne. Cybium 12:247-253. Bentzen, P., Taggart, C.T., Ruzzante, D.E. & Cook, D. (1996) Microsatellite polymorphism and the population structure of Atlantic cod (Gadus morhua) in the Northwest Atlantic. Canadian Journal of Fisheries and Aquatic Science 53:27062721. Berg, E. & Albert, O.T. (2003) Cod in fjords and coastal waters of North Norway: distribution and variation in length and maturity at age. ICES Journal of Marine Science 60:787-797. 308 7. Bibliography Beverley-Burton, M. (1978) Population genetics of Anisakis simplex (Nematoda: Ascaridoidea) in Atlantic salmon (Salmo salar) and their use as biological indicators of host stocks. Environmental Biology of Fishes 3:369-378. Bigelow, H.B., Bradbury, M.G., Dymond, J.R., Greeley, J.R., Hildebrand, S.F., Mead, G.W., Miller, R.R., Rivas, L.R., Schroeder, W.L., Suttkus R.D., and Vladykov, V.D., (1963) Fishes of the western North Atlantic. Sears Foundation for Marine. Research, New Haven, USA. 630pp. Blaxter, J.H.S. & Hunter, J.R. (1982) The biology of clupeoid fishes. Advances in Marine Biology. 20:1-223. Blaylock, R.B., Margolis, L. & Holmes, J. (2003) The use of parasites in discriminating stocks of Pacific halibut (Hippoglossus stenolepis) in the northeast Pacific. Fishery Bulletin 101:1-9. Blouin, M.S., Dame, J.B., Tarrant, C.A. & Courtney, C.H. (1992) Unusual population genetics of a parasitic nematode: mtDNA variation within and among populations. Evolution 46:470-476. Blouin, M.S., Yowell, C.A., Courtney, C.H. & Dame, J.B. (1995) Host movement and genetic structure of popualtions of parasitic nematodes. Genetics 141:10071014. Blouin, M.S., Liu, J. & Berry, R.E. (1999) Life cycle variation and the genetic structure of nematode populations. Heredity 83:253-259. Boely, T., Wysokinski, A. & Elwertowski, A. (1973) Les chinchards des côtes Sénégalaises et Mauritaniennes. Office de la Recherche Scientifique et Technique, Outre-Mer. DSP:46. Booke, H.E. (1981) The conundrum of the stock concept - are nature and nurture definable in fishery science? Canadian Journal of Fisheries and Aquatic Science 38:1479-1480. Booke, H.E. (1999) The stock concept revisited: perspectives on its history in fisheries. Fisheries Research 43:9-11. Bookstein, F.L. (1991) Morphometric Tools for Landmark Data. Cambridge University Press, Cambridge, UK. 435pp. Borges M.F. (1991) Biannual cohorts, growth parameters and migration effects of horse mackerel (Trachurus trachurus) in western Iberian waters, using length frequency analysis. ICES C.M. 1991/H:52. Borges, M.F., Turner, R.J., Casey, J. (1993) Plasma transferrin polymorphisms in scad (Trachurus trachurus L.) populations from the north-east Atlantic. ICES Journal of Marine Science 50:299-301. 309 7. Bibliography Borges, M.F., Silva, A., Abaunza, P., Eltink, A., Lucio, P.,Walsh, M., Poulard, J.C., Casey, J., Iversen, S.A., Skagen, D. & Sanchez, F. (1996) Changes of horse mackerel (Trachurus trachurus) catch distribution in the North-east Atlantic. ICES C.M. 1996/S:26. Bowers, A.B. (1969) Spawning beds of Manx Autumn Herring. Journal of Fish Biology 1:355-359. Bowman, R.E., Stillwell, C.E., Michaels, W.L. & Grosslein, M.D., (2000). Food of northwest Atlantic fishes and two common species of squid. NOAA Technical. Memo. NMFS-NE 155, 138 pp. Boyle, M.S. (1966) Trematode and nematode parasites of Pleurobrachia pileus Müller, in New Zealand waters. Transactions of the Royal Zoological Society of New Zealand 8(6):51-62. Brant, S.V. & Orti, G. (2003) Evidence for gene flow in parasitic nematodes between two host species of shrews. Molecular Ecology 12:2853-2859. Brattey, J. & Stenson, G.B. (1993) Host specificity and abundance of parasitic nematodes (Ascaridoidea) from the stomachs of five phocid species from Newfoundland and Labrador. Canadian Journal of Zoology 71:2156-2166. Bray, R.A. (1990) Hemiuridae (Digenea) from marine fishes of the southern Indian Ocean: Dinurinae, Elytrophallinae, Glomericirrinae and Plerurinae. Systematic Parasitology 17:187 -217. Bray, R., & Gibson, D. (1980) The Fellodistomidae (Digenea) of fishes from the northeast Atlantic. Bulletin British Museum Natural History (Zoology) 37:1-199. Brennan, NP; De Bruler, R; Blankenship, HL & Leber, KM (2001) Coded-wire tag and visible implant elastomer tag retention in juvenile red snapper Lutjanus campechanus. Aquaculture 2001: Book of Abstracts. 77 pp. World Aquaculture Society, Baton Rouge, USA. Brian, A. (1903). Sui Copepodi parassiti di pesci marini dell’isola d’Elba (4a nota). Bollettino dei Musei di Zoologia e Anatomia Comparata della Reale Università di Genova 121:1-8. Bush, A.O., Lafferty, K.D., Lotz, J.M. & Shostak, A.W. (1997) Parasitology meets ecology on its own terms: Margolis et al. revisited. Parasitology 83:575-583. Butterworth, D.S. & Penney, A.J. (2004) Allocation in high seas fisheries: avoiding meltdown. pp. 165-189. in Management of Shared Fish Stocks (eds. A.I.L. Payne, C.M. O'Brien & S.I. Rodgers). Blackwell Publishing Ltd., Oxford, UK.367pp. ISBN 1405106174. Cabral H.N. & Murta, A.G. (2002) The diet of blue whiting, hake, horse mackerel and mackerel off Portugal. Journal of Applied Ichthyology 18:14-23. 310 7. Bibliography Cabrera, R. & Suarez-Ognio, L. (2002) Probable emergencia de anisakiosis por larvas de Anisakis physeteris durante el fenómeno El Niño 1997-98 en la costa peruana. Parasitología Latinoamericana 57:3-4. Caddy, J.F. (1990) Some future perspectives for assessment and management of Mediterranean fisheries. Scientia Marina 57:121-130. Caddy, J.F. (1998) Issues in Mediterranean fisheries management: geographical units and effort control. FAO Study Review. 70 56 pp. Caddy, J.F., Gulland, J.A. (1983) Historical patterns of fish stocks. Marine Policy October 1983: 267-278. Cadrin, S.X. (2000). Advances in morphometric identification of fishery stocks. Reviews in Fish Biology and Fisheries 10:91-112. Cadrin, S.X. (2005) Morphmetric Landmarks. In: (eds. Cadrin, S.X., Friedland, K.D. & Waldman, J.R.) Stock Identification Methods – Applications in Fishery Science. Elsevier Academic Press, Amsterdam. pp. 153-172 Cadrin, S.X. & Friedland, K.D. (2005) Morphometric Outlines. In: (eds. Cadrin, S.X., Friedland, K.D. & Waldman, J.R.) Stock Identification Methods – Applications in Fishery Science. Elsevier Academic Press, Amsterdam. pp 247-269. Cadrin, S.X., Friedland, K.D., Waldman, J.R. eds. (2005) Stock Identification Methods : Applications in Fishery Science. Academic Press, Amsterdam.717pp. ISBN 012154351X. Campana, S.E. & Casselman, J.M (1993) Stock discrimination using otolith shape analysis. Canadian Journal of Fisheries and Aquatic Science 50:162-183. Campana, S.E. & Gagne, J.A. (1995) Cod stock discrimination using ICPMS elemental assays of otoliths. In: Secor, D.H., Dean, J.M. & Campana, S.E. (eds). Recent Developments in Fish Otolith Research. pp 671-691. Campbell, N. (2000) Some notes on Corynosoma wegeneri, an acanthocephalan parasite of saithe, Pollachius virens, in Scottish waters. M.Sc. Thesis, University of Aberdeen. 44pp. Campbell, N., MacKenzie, K., Mattiucci, S., Ramos, P., Pereira, A.L. & Abaunza, P. (2002) Parasites as tags for stock identification of Horse Mackerel (Trachurus trachurus (L.)) – results of the HOMSIR project. Poster presentation, Scottish Association of Marine Sciences (Spring Meeting), Oban. Campbell, N., Collins, C., Cross, M. A., Chubb, J. C., Cunningham, C. O., MacKenzie, K. & Hatfield, E.M.C. (2004) Life cycle studies on Renicola spp. (Digenea: Renicolidae) infecting sea birds: DNA tools provide an update. Poster presentation to European Multicolloquium of Parasitology, Valencia, Spain, July 2004. 311 7. Bibliography Campbell, N., MacKenzie, K., Zuur, A.F., Ieno, E.N. and Smith, G.M. (2007) Fish stock identification through neural network analysis of parasite fauna. In: Ecological Analysis (eds. Zuur, A.F., Ieno, E.N. & Smith, G.M.) Springer Verlag, New York. 749pp. Cardinale, M., Casini, M., Arrhenius, F. & Håkansson, N. (2003) Diel spatial distribution and feeding activity of herring (Clupea harengus) and sprat (Sprattus sprattus) in the Baltic Sea. Aquatic Living Resources 16:283-292. Carvalho G.R. & Hauser, L. (1994) Molecular genetics and the stock concept in fisheries. Reviews in Fish Biology and Fisheries 4:326-350. Carvalho G.R. & Pitcher T.J. (1995). Molecular genetics in fisheries. Chapman & Hall, Padstow, UK 131pp. Castonguay, M., Simard, P. & Gagnon, P. (1991) Usefulness of Fourier analysis of otolith shape for Atlantic mackerel (Scomber scombrus) stock discrimination. Canadian Journal of Fisheries and Aquatic Science 48:296-302. Chou, C.L., Paon, L.A., Moffatt, J.D. (2002) Metal contaminants for modelling lobster (Homarus americanus) migration patterns in the Inner Bay of Fundy, Atlantic Canada. Marine Pollution Bulletin 44:134-141. Clark, C.W. (1985) Bioeconomic Modelling and Fisheries Management. John Wiley & Sons, New York. 291pp. Choudhury, A. & Dick, T.A. (1998) Patterns and determinants of helminth communities in the Acipenseridae (Actinopterygii:Chondrostei), with special reference to the lake sturgeon, Acipenser fulvescens. Canadian Journal of Zoology 76:330-349. Costa, G. & MacKenzie, K. (1994) Histopathology of Goussia clupearum (Protozoa: Apicomplexa: Coccidia) in some marine fish from Scottish waters. Diseases of aquatic organisms. 18:195-205. Cowen, R.K., Lwiza, K.M.M., Sponaugle, S., Paris, C.B., Olson, D.B. (2000) Connectivity of marine populations: Open or closed ? Science 287: 857-859. Cribb, T.H., Bray, R.A., Olson, P.D. & Littlewood D.T. (2003) Life cycle evolution in the digenea: a new perspective from phylogeny. Advances in Parasitology 54:197-254 Criscione, C.D., Cooper, B & Blouin, M.S. (2006) Parasite genotypes identify source populations of migratory fish more accurately than fish genotypes. Ecology 87:823-828. Cruz, C., Vas. A & Saraiva, A. 2003. Occurrence of Kudoa sp. (Myxozoa) in Trachurus trachurus L. (Osteichthyes) in Portugal. Parasite 10(2), 69-79 Csirke, J. (1995) Fluctuations in abundance of small and mid-size pelagics. Scientia. Marina 59:481-490. 312 7. Bibliography Cushing, D.H. (1968) Fisheries Biology - a study in population dynamics. Univ. Wisconsin Press, Wisconsin, 200pp. Dalpadado, P., Melle, W., Ellertsen, B., & Dommasnes, A. (1996) Food and feeding conditions of herring Clupea harengus in the Norwegian Sea. ICES CM 1996/L: 20. Daponte, M.C., Gil de Pertierra, A.A., Palmieril, M.A. & Ostrowski de Nunez, M.O. (2006) Parasite infections (Trematoda, Digenea) of Sagitta friderici (Chaetognatha) from the southwestern Atlantic Ocean: prevalence and distribution. Diseases of Aquatic Organisms 71:225-232. Davies, D.H. (1957) The South African pilchard (Sardinops ocellata) and Maasbanker (Trachurus trachurus) young fish survey, 1955-56. Division of Fisheries Investigational Report 29. The Government Printer, Pretoria.36pp. Dayhoff, J.E. (1990) Neural Network Architectures: An Introduction. Van Nostrand Reinhold, New York. 259pp. Dean, D. (1996) Three-dimensional data capture and visualization. In: Marcus, L.F., Corti, M., Loy., Naylor, G.J.P. and Slice, D.E. eds. Advances in Morphometrics. NATO ASI Series A: Life Sciences 284:53-70. De Ligny, W. (1969) Serological and biochemical studies in populations.Annual Review in Oceanography and Marine Biology 7:411-513. fish De Vries, D.A., Grimes, C.B. & Prager, M.H. (2002) Using otolith shape analysis to distinguish eastern Gulf of Mexico and Atlantic Ocean stocks of king mackerel. Fisheries Research 57:51-62 Deudero, S. & Morales-Nin, B. (2000) Occurrence of Polyprion americanus under floating object in western Mediterranean oceanic waters: inference from stomach content analysis. Journal of the Marine Biological Association of the United Kingdon 80:751-752. Dickson, R.D., Meincke, J., Malmberg, S-A. & Lee, A.J. (1988) The ”great salinity anomaly“ in the Northern North Atlantic 1968-1982. Progress in Oceanography. 20:103-151. Dizon, A.E., Lockyer, C., Perrin, W.F., Demaster, D.P. & Sisson, J. (1994) Rethinking the stock concept: a phylogeographic approach. Conservation Biology 6:24-36. Dubois, G., Baer, J.G. & Euzet, L. (1952) Une nouvelle cercaire du plankton matin de Sarte, Cercaria mathiasi n. sp. Revue Suisse de Zoologie 59:503-510. Dykova, I. & Lom, J. (1981) Fish coccidian: critical notes on life cycles, classification and pathogenicity. Journal of Fish Diseases 4:487-505. 313 7. Bibliography Eltink, A.T.G.W (1992) Horse mackerel egg production and spawning stock size in the North Sea in 1991. ICES C.M. 1992/H:21. Eltink, A.T.G.W & Kuiter, C. (1989) Validation of aging techniques on otoliths of horse mackerel (Trachurus trachurus (L.)) ICES C.M. 1989/H:43 . Engelhard, G.H., Heino, M. (2004) Maturity changes in Norwegian springspawning herring before, during, and after a major population collapse. Fisheries Research 66:299-310. FAO (1993) Marine fisheries and the law of the sea: a decade of change. FAO Fisheries Circular 853. FAO (2003) FAO Yearbook, Fishery Statistics, 2003, Vol.97. United Nations Food and Agriculture Organisation, Rome. pp 713. Farran, G.P. (1946) The herring fishery in Eire, 1921-1941. Department of Agriculture Journal 2. Fernández, M., Aznar, F.J., Montero, F.E., Georgiev, B.B. & Raga, J.A. (2004) Gastrointestinal helminths of Cuvier’s Beaked Whales, Ziphius cavirostris, from the western Mediterranean. Journal of Parasitology 90:418-420. Ferson, S., Rohlf, F.J. & Koehn, R.K. (1985) Measuring shape variation of two dimensional outlines. Systematic Zoology 34:59-68. Fioravanti, M.L., Caffara, M., Delgado, M.L., Florio, D., Marcer, F., Quaglio, F. and Restani, R. 2001. Survey on the diffusion of Myxosporeans (Myxozoa) in marine fish farmed in Italy. Bollettino Societa Italiana di Patologia Ittica. 13:12-25 (in Italian). Fischthal, J.H. & Thomas, J.D. (1970) Digenetic trematodes of marine fishes from Ghana: Family Opecoelidae. Proceedings of the Helminthological Society of Washington 37, 129–141. Flores, I. (2001) Morphometric and meristic counts of the Sailfish Istiophorus platypterus in waters near to Mazatlan, Sinaloa, Mexico. Ciencias y Marinas 5:5762 Folmer, O., Black, M., Hoeh, W., Lutz, R. & Vrijenhoek, R. (1994) DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Molecular Marine Biology and Biotechnology 3:294 299. Foot, T. (2001) Guinness Book of World Records 2001 Guinness World Records Ltd., London. 284pp. Ford, E. (1933) An account of the herring investigations conducted at Plymouth during the years 1924 to 1933. Journal of the Marine Biological Association of the United Kingdom. 19:305-384. 314 7. Bibliography Fox, C.J., Harrop, R. & Wimpenny, A. (1999) Feeding ecology of herring (Clupea harengus) larvae in the turbid Blackwater Estuary. Marine Biology. 134:353-365. Froese, R. and D. Pauly. (2004) FishBase. A World Wide Web electronic publication. www.fishbase.org, version (06/2004). Freidlander, A.M., Parrish, J.D. & Defelice, R.C. (2002) Ecology of the introduced snapper Lutjanus kasmira (Forsskal) in the reef fish assemblage of a Hawaiian bay. Journal of Fish Biology 60:28-48. Gaevskaya, A.V. (1990). Some comments on trematodes of the genus Monascus Looss, 1907 (Fellodistomidae). In Parasites of animals and plants. Ed. B.Y. Lebedev. Academy of Sciences of the USSR, Far-East Branch, Vladivostok. pp. 69–73. Gaevskaya, A.V. & Kovaleva, A.A. 1979a. Novye e Redko Vestrechanshesya Veydu Myxsosporidiye Roub Keltskogo Morya [New and Rarely Encountered Forms of Myxosporidia from Fishes of the Celtic Sea] Parazitologia 13(2),159-165 (in Russian). Gaevskaya, A.V. & Kovaleva, A.A. 1979b. Dva novykh vida miksosporidij ot kapskoj stavridy iz Yugo-Vostochnoj Atlantiki [Two new species of myxosporidia from horse mackerel in the south eastern Atlantic.] Biologiya Morya. 3, 80-83 (in Russian). Gaevskaya, A.V. & Kovaleva, A.A. (1979c) Monogenea ot Atlanticheski Scombrii Loshadi Trachurus. [Monogeneans from the Atlantic horse mackerels of the genus Trachurus.] Zoologiski Zhurnal 58:1110-1116. (in Russian). Gayevskaya A. V. & Kovaleva A. A. (1980) Ekologo-Geograficheski osobenosti parazitofauni obknovenoy stravdi Altanticheskogo Okeana. [Eco-geographical features of the parasitofauna of the common Atlantic Horse Mackerel] . Studies of the biological resources of the Atlantic Ocean. Kaliningrad, AtlantNIRO pp18-24 (in Russian). Gaevskaya, A.V. & Kovaleva, A.A. (1985). Parazitofauni ot stravdiv okeani, Trachurus picturatus picturatus e ekologo-geografeski osbenosti. [The parasite fauna of the "oceanic horse mackerel" Trachurus picturatus picturatus and ecogeographical characteristics of its formation.]. Ekologiya Morya 20:80-84. (In Russian). Gaevskaya, A.V., Kovaleva, A.A., & Rodjuk, G.N. (1985) Parasitofauna of the fishes of the Falkland–Patagonian region. NOAA Technical Report NMFS25 pp. 25–28 . Garey, J.G., Schmidt-Rhaesa, A., Near, T.J & Nadler, S.A. (1998) The evolutionary relationships of rotifers and acanthocephalans. Hydrobiologia 287:8391. 315 7. Bibliography Garey, J.G., Near, T.J., Nonnemacher, M.R. & Nadler, S.A. (1996) Molecular evidence for Acanthocephala as a subtaxon of Rotifera. Journal of Molecular Evolution 43:287-282. George-Nascimento, M. (2000) Geographical variations in the jack mackerel Trachurus symmetricus murphyi populations in the southeastern Pacific Ocean as evidenced from the associated parasite communities. Journal of Parasitology 86:929-932. Gestal, C. and Azvedo, C. (2005) Ultrastructure of Goussia cruciata (Apicomplexa: Coccidia) infecting the liver of horse mackerel, Trachurus trachurus (L.), from Ibero-Atlantic waters. Journal of Fish Diseases 28:125-132. Gibson D.I. (1973) The genus Pseudanisakis Layman & Borovkova, 1926 (Nematoda: Ascaridida). Journal of Natural History 7: 319 – 340. Gibson, D.I. (2002) Family Hemiuridae Loos,1899. pp299-340. In, Keys to the Trematoda, Volume I. eds. Gibson, D.I., Jones, A. & Bray, R.A. CABI Publishing, Oxford, U.K. 521pp. ISBN 0951995470. Gibson, D.I. & Bray, R.A. (1986) The Hemiuridae (Digenea) of fishes from the north-east Atlantic. British Museum (Natural History), London. pp.125 ISBN: 0565050230. Godfrey, S.S., Bull, C.M., Murray, K & Gardner, M.G. (2006) Transmission mode and fistribution of parasites among groups of the social lizard Egernia stokesii. Parasitology Research 99:223-230. Golvan, Y.J. (1969) Systematique des Acanthocephales (Acanthocephala Rudolphi 1801), L'Ordre des Palaeacanthocephala Meyer 1931, La Super-Famille Des Echinorhynchoidea (Cobbold 1876) Golvan et Houin 1963. Memoires du Museum National D'Histoire Naturelle, LVII. Editions Du Museum, Paris, France. 373pp. González, M.T., Barrientos, C. & Moreno, C.A. (2006) Biogeographical patterns in endoparasite communities of a marine fish (Sebastes capensis Gmelin) with extended range in the Southern Hemisphere. Journal of Biogeography 36:10861095. Grahl-Nielsen, O. (2005) Fatty Acid Profiles as Natural Marks for Stock Identification. In: (eds. Cadrin, S.X., Friedland, K.D. & Waldman, J.R.) Stock Identification Methods – Applications in Fishery Science. Elsevier Academic Press, Amsterdam. pp 247-269. Grahl-Nielsen, O. & Ulvund, K.A. (1990) Distinguishing populations of herring by chemometry of fatty acids. American Fisheries Society Symposium 7:566-571. Granadeiro, J.P., Monteiro, L.R. & Furness, R.W. (1998) Diet and feeding ecology of Cory's shearwater Calonectris diomedea in the Azores, north-east Atlantic. Marine Ecology Progress Series 166:267-276. 316 7. Bibliography Granadeiro, J.P., Monteiro, L.R., Silva, M.C. & Furness, R.W. (2002) Diet of Common Terns in the Azores, Northeast Atlantic. Waterbirds 25:149-155. Grabda J. (1974). The dynamics of the nematode larvae, Anisakis simplex (Rud.) invasion in the South Western Baltic herring (Clupea harengus L.). Acta Ichthyologia et Piscatoria. 4 :3-21. Grant, W.S., Milner, G.B., Krasnowski, P. & Utter, F.M. (1980) Use of biochemical agenetic variants for identification of sockeye salmon (Oncorhynchus nerka) stocks in Cook Inlet, Alaska. Canadian Journal of Fisheries and Aquatic Sciences 37:1236-1247 Grant W.S. & Bowen B.W. (1998) Shallow population histories in deep evolutionary lineages of marine fishes: Insights from sardines and anchovies and lessons for conservation. Journal of Heredity 89:415-426. Greenstreet, S.P.R., Spence, F.B., Shanks, A.M. & McMillan, J.A. (1999) Fishing effects in northeast Atlantic shelf seas: patterns in fishing effort, diversity and community structure. II. Trends in fishing effort in the North Sea by UK registered vessels landing in Scotland. Fisheries Research 40:107-124 Groeger, J. (2000) Separation of two herring stocks in the transition zone between Baltic and North Sea, based on logistic regression and meristic characters Archiv fur Meeresforshung 48:161-174 Groot, C., Bailey, R.E., Margolis, L. & Cooke, C. (1989) Migratory patterns of sockeye salmon (Oncorhynchus nerka) smolts in the Strait of Georgia, British Columbia, as determined by analysis of parasite assemblages. Canadian Journal of Fisheries and Aquatic Science 41:400-408. Gulland, J. A. (1971). Ecological aspects of fishery research. Advances in. Ecological. Research 7:115-176. Gyllensten U. (1985) The genetic structure of fish: differences in the intraspecific distribution of biochemical genetic variation between marine, anadromous, and freshwater species. Journal of Fish Biology 26:691-699 Haddon, M. (2001) Modelling and Quantitative Methods in Fisheries. Chapman Hall/CRC Press, Boca Raton, USA. 424pp. Hammer, C. & Zimmerman, C. (2005) The role of stock identification in formulating fisheries management advice. In: (eds. Cadrin, S.X., Friedland, K.D. & Waldman, J.R.) Stock Identification Methods – Applications in Fishery Science. Elsevier Academic Press, Amsterdam. pp 631-658. Hanson, P.J., Koenig, C.C. & Zdanowicz, V.S. (2004) Elemental composition of otoliths used to trace estuarine habitats of juvenile gag, Mycteroperca microlepis, along the west coast of Florida. Marine Ecology Progress Series 267:253-265 Heath, M.R. & Baird, D.L. (1983) Recruitment and seasonal migrations of herring to the west of Scotland. ICES CM 1983/H:51 12 pp 317 7. Bibliography Heath, M.R. & MacLachlan, P. (1987) Dispersion and mortality of yolk-sac herring (Clupea harengus L.) larvae from a spawning ground to the west of the Outer Hebrides. Journal of Plankton Research. 9:613-630. Heath, M. & Rankine, P. (1988). Growth and Advectrion of Larval Herring (Clupea harengus L.) in the Vicinity of the Orkney Isles. Estuarine, Coastal and Shelf Science. 27:547-565. Heath, M.R., MacLachlan, P.M. & Martin, J.H.A. (1987). Inshore circulation and transport of herring larvae off the north coast of Scotland. Marine Ecology Progress Series. 40:11-23. Heinke, F. (1898) Naturgeschichte des herings I. Die lokalformen und die Wanderungen des Herings in den europäischen Meeren. Abhandlung der Deutschen Seefishereivereins (in German), Salle, Berlin, 217pp. Henault, M. & Fortin, R. (1989) Comparison of morphometric and meristic characters among spring and fall spawning ecotypes of Cisco (Coregonus artedii) in southern Quebec, Canada. Canadian Journal of Fisheries and Aquatic Sciences 46:166-173. Herrington, W.C., Bearse, H.M. & Firth, F.E. (1939) Observations on the life history, occurrence and distribution of the redfish parasite Sphyrion lumpi. United States Bureau of Fisheries Special Report 5: 1-18. Hewitt, G.C. & Hine, P.M. (1972) Checklist of parasites of New Zealand fishes and of their hosts. New Zealand Journal of Marine and Freshwater Research 6:69114. Hill, D.R. (1959) Some uses of statistical analysis in classifying races of American Shad (Alosa sapidissima). Fisheries Bulletin 59:269-286. Hilborn, R. & Walters, C.J. (1992) Quantitative fisheries stock assessment. Choice, dynamics & uncertainty. Chapman & Hall. New York, 570 pp. Hillis, DM, C. Moritz & B. K. Mable (1996) Molecular Systematics, 2nd edn. Sinauer Assoc., Sunderland, Massachusetts, USA. 655pp. ISBN 0878932828 . Horbowy J. & Poldoska, M. (2001) Modelling infection of Baltic herring (Clupea harengus membras) by larval Anisakis simplex. ICES Journal of Marine Science 58:321-330. Hotta, T., Matsuishi, T., Sakano, J. & Kanno, Y. (1999) Population structure of Pacific herring Clupea pallasii in the eastern Hokkaido waters. Bulletin of the Japanese Society of Scientific Fisheries 65:655-660. Hu, M., Hoglund, H., Chilton, N.B., Zhu, X & Gasser, R.B. (2002) Mutation scanning analysis of mitochondrial cytochrome c oxidase subunit 1 reveals limited gene flow among bovine lungworm subpopulations in Sweden. Electrophoresis 23:3357-3363. 318 7. Bibliography Hurst, R. J. (1984) Identification and description of larval Anisakis simplex and Pseudoterranova decipiens (Anisakidae: Nematoda) from New Zealand waters. New Zealand Journal of Marine and Freshwater Research 18:177-186. Huse, G. & Gjøsæter, H. (1999) A neural network approach for predicting stock abundance of the Barents Sea capelin. Sarsia 84:457-464. Ichihara, A. (1968) On the parasitic helminths of marine fish in Sagami Bay – I. On horse mackerel, flasher, butterfish, hashikinme, frigate mackerel, barracuda and alfonsino. Bulletin of the Japanese Society of Scientific Fisheries 34:365-377. Ihssen, P.E., Booke, H.E., Casselman, J.M., McGlade, J.M., Payne, N.R. & Utter, F.M. (1981) Stock Identification: materials and methods. Canadian Journal of Fisheries and Aquatic Sciences 38:1481-1855. Iles, T.D. & Sinclair, M. (1982) Atlantic herring: stock discreteness and abundance. Science 215:627-633. Itagaki, H., & Ishimaru. S. (1967) Studies on Anisakinae larvae in horse mackerels, Trachurus japonicus. Japanese Journal of. Parasitology 16:288-289. Iversen, S.A., Skogen, M.D. & Svendsen, E. (2002) Availability of horse mackerel (Trachurus trachurus) in the north-eastern North Sea, predicted by the transport of Atlantic water. Fisheries Oceangraphy 11:245-250. Jackson, C.J., Marcogliese, D.J. & Burt, M.D.B. (1997) Role of hyperbenthic crustaceans in the transmission of marine helminth parasites. Canadian Journal of Fisheries and Aquatic Science. 54:815-820. Jacobsen, J.A. & Hansen, L.P. (2005) Internal and External Tags. In: (eds. Cadrin, S.X., Friedland, K.D. & Waldman, J.R.) Stock Identification Methods – Applications in Fishery Science. Elsevier Academic Press, Amsterdam. pp 415433. Jardas I., Šantić M. & Pallaoro, A. (2004) Diet composition and feeding intensity of horse mackerel, Trachurus trachurus (Osteichthyes: Carangidae) in the eastern Adriatic. Marine Biology 144:1051-1056. Jennings, S. & Pawson, M.G. (1992) The origin and recruitment of bass, Dicentrarchus labrax, larvae to nursery areas. Journal of the Marine Biological Association of the United Kingdom 72:199-212. Jennings, S., Kaiser, M.J. & Reynolds, J.D. (2001) Marine Fisheries Ecology Blackwell Science, Oxon. 417pp. Jones, E.I. (1933) Studies on the monogenea of Plymouth. Gastrocotyle trachuri v.Ben and Hesse, 1863. Journal of the Marine Biological Association of the United Kingdom 19:227-232. 319 7. Bibliography Jousson, O., Bartoli, P. & Pawlowski, J. (2000) Cryptic speciation among intestinal parasites (Trematoda: Digenea) infecting sympatric host fishes (Sparidae). Journal of Evolutionary Biology 13:778-785. Jung, Y.J. (1816) Chasan Oba: Fisheries in Huk-San Island (in Korean, translated by Chyung, M.K., 1992) Ja Si San Eop Sa, Seoul, 216pp. Kabata, Z. (1979) Parasitic copepoda of British Fishes. The Ray Society, London, United Kingdom. 2018pp. ISBN0903874059. Keeney, D.B. & Campbell, R.A. (2001) Grillotia borealis sp.n. (Cestoda: Trypanorhynca) from five species of Bathyraja (Rajaformes:Arhynchobatidae) in the North Pacific Ocean with comments on parasite distributions. Folia Parasitologica 48:21-29. Kenney, R.D., Mayo, C.A. & Winn, H.E. (2001) Migration and foragin strategies at varying spatial scales in western North Atlantic right whales: a review of hypotheses. Journal of Cetacean Research and Management 2:251-260. Kent M.L. & Hedrick R.P. (1985) The biology and associated pathology of Goussia carpelli (Léger and Stankovitch) in goldfish Carassius auratus (Linnaeus). Fish Pathology 20:485–494. Kerstan, M. (1985) Age, growth, maturity, and mortality estimates of horse mackerel (Trachurus trachurus) from the waters west of Great Britain and Ireland in 1984. Archiv fur Fischereiwissenschaft 36:115-154. Kerstan S.L. (1991) Der Befall von Fischen aus dem Wattenmeer und dem Nordatlantik 1988-1990 mit Nematodenlarven und eine Bibliographie uber parasitische Nematoden in Fischen und Seesäugern. Berichte aus dem Institut fuer Meereskunde an der Christian-Albrechts-Universitaet Kiel 219: 1-205. Kim, J.-H., Lee, C.H. & Lee C.S. (2005) Preliminary Studies on Metazoan Parasites of Chum Salmon (Oncorhynchus keta) in Korea. NPAFC Doc. 885 5pp. Faculty of Marine Bioscience & Technology, Kangnung Nat'l Univ., Kangnung 210-702, Republic of Korea. King, D.P.F. (1985) Morphological and meristic differences among spawning aggregations of north-east Atlantic herring Clupea harengus L. Journal of Fish Biology 26:591-607. Klimpel, S., Palm, H.W., Ruckert, S. & Piatkowski, U. (2004) The life cycle of Anisakis simplex in the Norwegian Deep (northern North Sea).Parasitology Research 94:1-9. Køie, M (1991) Aspects of the morphology and life cycle of Lecithocladium excisum (Digenea, Hemiuridae), a parasite of Scomber spp. International Journal for Parasitology 21:597-602. 320 7. Bibliography Køie, M. (1993) Aspects of the life cycle and morphology of Hysterothylacium aduncum (Rudolphi, 1802) (Nematoda, Ascaroidea, Anisakidae) Canadian Journal of Zoology 711289-1296. Køie, M. (2000) First Record of an Actinosporean (Myxozoa) in a Marine Polychaete Annelid. Journal of Parasitology 86:871-872. Kompotilski, A. (1975) The intraspecific geographical variability of horse mackerel Trachurus trachurus L. in the West African Shelf waters. Acta Icthiologica et Piscatoria 5:13-29. Kornfield, I., Beland, K.F., Moring, J.R. & Kircheis, F.W. (1981) Genetic similarity among endemic Arctic Char (Salvelinus alpinus) and implications for their management. Canadian Journal of Fisheries and Aquatic Sciences 38:32-39. Koslow, J.A., Boehlert, G., Gordon, J.D.M., Haedrich, R.L., Lorance, P. & Parin, N. (2000) Continental slope and deep sea fisheries: implications for a fragile ecosystem. ICES Journal of Marine Science 57:548-557. Kovaleva, A.A. (1970). Gelmitofauna ot loshadi Trachurus (Carangidae, Perciformes) Basseuna Atlanticheskogo (Helminth fauna of mackerel of the genus Trachurus (Carangidae, Perciformes) in the Atlantic basin.] Biologiya Morya 20: 37-66. (In Russian). Kovaleva, A.A., Shulman, S.S. & Yakovlev, V.N. (1979). Myxsosporidiye Roda Kudoa (Myxsosporidia, Multivalvulida) Basseuna Atlanticheskogo Okeana. [Myxosporidia of the genus Kudoa (Myxosporidia: Multivalvulida) of the Atlantic Ocean basin]. Trudy Zoologicheskiye Institut 87: 42-64 (in Russian). Kutkuhn, J.H. (1981) Stock definition as a necessary besis for cooperatieve management of Great Lakes fish resources. Canadian Journal of Fisheries and Aquatic Sciences. 38:1476-1478. Lafferty, K.D. & Morris, A.K. (1996) Altered behavior of parasitized killifishes increases susceptibility to predation by bird final hosts. Ecology 77:1390-1397. Landsberg, J.H. and Lom, J. (1991). Taxonomy of the genera of the Myxobolus/Myxosoma group (Myxobolidae:Myxosporea), current listing of species and revision of synonyms. Systematic Parasitology 18:165-186. Larkin, P.A. (1977) An epitaph for the concept of maximum sustainable yield. Transactions of the American Fisheries Society 106:1-11.. Larsen, G., Hemmingsen, W., MacKenzie, K. & Lysne, D.A. (1997) A population study of cod, Gadus morhua L., in northern Norway using otolith structure and parasite tags. Fisheries Research 32:13-20. Levin, B. (2000) Genes VII. Oxford University Press. Oxford, UK. 989pp. Lester, R.J.G. (1990) Reappraisal of the use of parasites for fish stock identification. Australian Journal of Marine and Freshwater Research 41:855-864. 321 7. Bibliography Lester, R.J.G., Barnes, A., & Habib, G. (1985) Parasites of Skipjack Tuna: Fishery Implications. Fishery Bulletin 83:343-356. Lester, R.J.G., Thompson, C., Moss, H., & Barker, S.C. (2001) Movement and stock structure of narrow-barred Spanish mackerel as indicated by parasites. Journal of Fish Biology 59:833-842. Lima, M., Stenseth, N.C., Leirs, H. & Jaksic, F.M. (2003) Population dynamics of small mammals in semi-arid regions: a comparative study of demographic variability in two rodent species. Proceedings of the Royal Society of London, Series B: Biological Sciences. 270:1997-2007. Lindsey, C.C. (1981) Stocks are chameleons: plasticity in gill rakers of coregonid fishes. Canadian. Journal of Fisheries and Aquatic Sciences 38:1497-1506. Lindsey, C.C. (1988) Factors controlling meristic variation. Fish Physiology 11: 197-274. Lleonart J. (2004) A review of Mediterranean shared stocks, assessment and management. In: Management of Shared Fish Stocks (eds. A.I.L. Payne, C.M. O'Brien & S.I. Rodgers). Blackwell Publishing Ltd., Oxford, UK.367pp. ISBN 1405106174. Lleonart, J. & Maynou, F. (2003). Fish stock assessments in the Mediterranean: state of the art. Scientia Marina 67(Suppl. 1): 37-49. Llewellyn, J. (1959) The larval development of two species of gastrocotylid trematode parasites from the gills of Trachurus trachurus. Journal of the Marine Biological Association of the United Kingdom 38:461-467. Llewellyn, J. (1962) The life histories and population dynamics of monogenean gill parasites of Trachurus trachurus (L.) Journal of the Marine Biological Association of the United Kingdom 42:587-600. Loc'h F.L. & Christian H. (2005) Stable carbon and nitrogen isotope analysis of Nephrops norvegicus/Merluccius merluccius fishing grounds in the Bay of Biscay (Northeast Atlantic). Canadian Journal of Fisheries and Aquatic Sciences 62:123132. Lockwood, S.J. (1988) The Mackerel. Fishing News Books, Farnham, UK. 188pp. Lom, J. & Arthur, J.R. (1989) A guideline for the preparation of species descriptions in the Myxosporea. Journal of Fish Diseases 12:151-156 Lom, J. & Dykova, I. (1992) Protozoan Parasites of Fishes. Elsevier Academic Press, Amsterdam. 316pp. Lom, J. & Laird, M. (1969) Parasitic protozoa from marine and euryhaline fish of Newfoundland and New Brunswick. I. Peritrichous ciliates. Canadian Journal of Zoology 47:1367-1380. 322 7. Bibliography Looss, A. (1899) Weiterer beiträge zur Kenntniss der Trematoden-Fauna Aegyptens, zugleich Versuch einer natürlichen Gliederung des Genus Distomum Retzius geleitet haben. Zentralblatt für Bakteriologie, Parasitenkunde, Infektionskranheiten (und Hygiene) 29:191-210. (in German). Looss, A. (1907) Zur Kenntnis der Distomenfamilie Hemiuridae. Zoologischer Anzeiger 31:585-620. Lopez-Roman, R. (1973) Nuevos datos sobre la morfologia y la anatomia de Axine belones Abildgaard, 1794 (Monogenea). Revista Ibero Parasitologia 33:557-571. Lordan, C., Collins, M.A. & Perales-Raya, C. (1998) Observations on morphology, age and diet of three Architeuthis caught off the west coast of Ireland in 1995. Journal of the Marine Biological Association of the United Kingdom 78:903-917. Lucio, P. & Martin, I. (1989) Biological aspects of horse mackerel (Trachurus trachurus L. 1758) in the Bay of Biscay in 1987 and 1988. ICES C.M. 1989/H:28. Macer, C.T. (1974) The reproductive biology of the horse mackerel (Trachurus trachurus (L.) in the North Sea and English Channel. Journal of Fish Biology 6:415-438 . Macer, C.T. (1977) Some aspects of the biology of the horse mackerel (Trachurus trachurus L.) in waters around Britain. Journal of Fish Biology 10:51-62 McClelland, G. (1995) Experimental infections of fish with larval sealworm, Pseudoterranova decipiens (Nematoda, Anisakidae), transmitted by amphipods. Canadian Journal of Fisheries and Aquatic Science. 52:140-155. McConnell, B.J., Fedak, M.A., Lovell, P & Hammond, P.S. (1999) Movements and foraging areas of grey seals in the North Sea. Journal of Applied Ecology 36:573-590. McQuinn, I.H. (1997) Metapopulations and the Atlantic Herring. Reviews in Fish Biology and Fisheries 7:297-329. MacDonald, P. (2005) Stock identification of the horse mackerel Trachurus trachurus L. from the North-west African coast using parasites as biological tags. M.Sc. thesis. University of Aberdeen. MacKenzie, K. (1983) Parasites as Biological Tags in Fish Population Studies. Advances in Applied Biology 7:251-331. MacKenzie, K. (1985) Relationships between the Herring, Clupea harengus L., and its Parasites. Advances in Marine Biology 24:263-319. MacKenzie, K. (1987a). Parasites as indicators of host populations. International Journal for Parasitology 17:345-352. 323 7. Bibliography MacKenzie, K. (1987b). Relationships between the herring, Clupea harengus L., and its parasites. Advances in Marine Biology 24:263-319. MacKenzie, K. (1987c). Long-term changes in the prevalence of two helminth parasites (Cestoda: Trypanorhyncha) infecting marine fish. Journal of Fish Biology 31: 83-87. MacKenzie, K. (2002). Parasites as biological tags in population studies of marine organisms: an update. Parasitology 124:S153-S163. MacKenzie, K. (2004). Parasites as biological tags for marine fish populations. Biologist 51:86-90. MacKenzie, K. (2005) Parasites as Biological Tags. In, Marine Parasitology eds. K. Rohde. CABI Publishing, Oxford, UK. 563pp. MacKenzie, K. & Abaunza, P. (1998) Parasites as biological tags for stock discrimination of marine fish: a guide to procedures and methods. Fisheries Research 38:45-56. MacKenzie, K & Abaunza, P. (2005). Parasites as biological tags. In Stock Identification Methods: Applications in Fisheries Science (Cadrin, S.X.,Friedland, K.D. & Waldman, J.R., eds.), pp. 211-226. Elsevier Academic Press. MacKenzie, K and Longshaw, M. 1995. Parasites of the hakes Merluccius australis and M. hubbsi in the waters around the Falkland Islands, southern Chile, and Argentina, with an assessment of their potential value as biological tags. Canadian Journal of Fisheries and Aquatic Science 52 (Suppl. 1):213-224. MacKenzie, K., Campbell, N., Mattiucci, S., Ramos, P., Pereira, A.L. & Abaunza, P. (2004). A checklist of the protozoan and metazoan parasites reported from the Atlantic Horse Mackerel, Trachurus trachurus L. Bulletin of the European Association of Fish Pathologists 24(4):180-184. MacKenzie, K., Abaunza, P. & Campbell, N. (2005) The use of parasites as biological tags in multidisciplinary stock identification studies of small pelagic fish. ICES C.M. 2005/K:17. MacKenzie, K., Campbell, N., Mattiucci, S., Ramos, P., Pinto, A.L. & Abaunza, P. (In Press) Parasites as biological tags for stock identification of Atlantic horse mackerel Trachurus trachurus L. Fisheries Research . Macleod, N., (1990). Digital images and automated image analysis systems. In: Rohlf, F.J., Bookstein, F.L. (eds.), Proceedings of the Michigan Morphometrics Workshop, vol. 2, pp 21-36. University of Michigan Museum of Zoology Special Publications. Maeno, Y., Sorimachi, M., Ogawa, K. & Egusa, S. (1990). Myxobolus spinacurvatura sp. n. (Myxosporea: Bilvalvulida) parasitic in deformed mullet, Mugil cephalus . Fish Pathology. 25: 37-41. 324 7. Bibliography Manfredi, M.T., Crosa, G., Galli, P. Ganduglia, S. (2000) Distribution of Anisakis simplex in fish caught in the Ligurian Sea Parasitology Research 86:551553. Manter, H.W. (1947) The digenetic trematodes of marine fishes of Tortugas, Florida. The American Midland Naturalist 38:252-416. Maravelias, C.D., Haralabous, J. & Papaconstantinou, C. (2003) Predicting demersal fish species distributions in the Mediterranean Sea using artificial neural networks. Marine Ecology Progress Series 25:249-258. Marcogliese, D.J. (2002) Food webs and the transmission of parasites to marine fish. Parasitology 124:S83-S89. Martin, J. W. & Davis, G. E. (2001) An Updated Classification of the Recent Crustacea. Natural History Museum of Los Angeles County, Los Angeles, United States of America. 132 pp.. Martorelli, S.R. & Cremonte, F. (1998) A proposed three-host life history of Monascus filiformis (Rudolphi, 1819) (Digenea: Fellodistomidae) in the southwest Atlantic Ocean Canadian Journal of Zoology 76:1198–1203. Mattiucci, S., Nascetti, G., Cianchi, R., Paggi, L., Arduino, P., Margolis, L., Brattey, J., Webb, S., D´Amelio, S., Orecchia, P. & Bullini, L. (1997). Genetic and ecological data on the Anisakis simplex complex, with evidence for a new species (Nematoda, Ascaridoidea, Anisakidae) . Journal of Parasitology 83:401416. Mattiucci, S., Paggi, L., Nascetti, G., Portes-Santos, C., Costa, G., Di Beneditto, A.P., Ramos, R., Argyrou, M., Cianchi, R. & Bullini, L. (2002) Genetic markers in the study of Anisakis typica (Diesing, 1860): larval identification and genetic relationships with other species of Anisakis Dujardin, 1845 (Nematoda: Anisakidae) Systematic Parasitology 51:159-170. Mattiucci, S., Abaunza, P., Ramadori, L. & Nascetti, G. (2004) Genetic identification of Anisakis larvae in European hake from Atlantic and Mediterranean waters for stock recognition. Journal of Fish Biology 65:495-510. Mattiucci, S., Nascetti, G., Dailey, M., Webb, S.C., Barros, N.B., Cianchi, R. & Bullini, L. (2005) Evidence for a new species of Anisakis Dujardin, 1845: morphological description and genetic relationships between congeners (Nematoda : Anisakidae). Systematic Parasitology 61:157-171. Mattiucci, S., Farina, V., Campbell, N., Mackenzie, K., Ramos, P., Pinto, A.L., Abaunza, P. & Nascetti, G. (in press) Anisakis spp. larvae (Nematoda: Anisakidae) from Atlantic horse mackerel: their genetic identification and use as biological tags for host stock identification. Fisheries Research. 325 7. Bibliography Mattiucci, S. & Nascetti, G. (2006) Molecular systematics, phylogeny and ecology of anisakid nematodes of the genus Anisakis Dujardin, 1845: an update. Parasite 13:99-113 . Maxim, C. (1995) Horse mackerel and false scad stock assessment and catch projections, CECAF Divisions 34.1.3 and 34.3.1. Scientia Marina 59:611-627. McCoy, K.D., Tirard, C & Michalkis, Y. (2003) Spatial genetic structure of the ectoparasite Ixodes uriae within breeding cliffs of its colonial seabird host. Heredity 91:422-429. McCullagh, P. & Nedler, J.A. (1989) Generalized Linear Models, 2nd Ed. Chapman & Hall, London, UK. 532pp. McCulloch, W.S. & Pitts, W.H. (1943) A logical calculus of the ideas immanent in nervous activity. Bulletin of Mathematical Biophysics 5:115-133 . McHugh, J.L. (1951) Meristic variations and populations of northern anchovy (Engraulis mordax mordax). Bulletin of the Scripps Institute of Oceanography 6: 123-160. Meek, A. (1916) The Migrations of Fish. E. Arnold Publishers, London, UK. 427 pp. Metcalfe, J., Arnold, G. & McDowall, R. (2002). Migration. Handbook of Fish Biology and Fisheries. Volume I. eds. J.B. Hart and J.D. Reynolds pp: 175-199. Blackwell Publishing, London, UK. Meyer, A. (1994) Molecular phylogenetic studies of fish. In: Genetics and evolution of aquatic organisms ed. A. R. Beaumont, Chapman & Hall, London, pp. 219-249. Merrett, N.R. & Haedrich, R.L. (1997) Deep Sea Demersal Fisheries Chapman & Hall, London, 282pp. Miller, M.J. (2002) The distribution and ecology of Ariosoma balearicum (Congridae) Leptocephali in the Western North Atlantic. Marine Ecology Progress Series 227:97-107. Mogrovejo, C. & Santos, C.P. (2002) Biological Tags and Environmental factors in Southwestern Atlantic Waters. Proceedings of the Tenth International Congress of Parasitology. pp 229-235. Monduzzi Editori. Rome. Montalvo, A. (1977) Helmintos parásitos en jurel Trachurus symmetricus murphyi Nichols, 1920 desembarcado en la Caleta Santa Rosa, Lambayeque. Tesis Licenciado Biología Pesquera. Universidad Nacional “Pedro Ruíz Gallo”, Lambayeque, Peru. (in Spanish). Moore, B.R., Buckworth, R.C., Moss, H. & Lester, R.J.G. (2003) Stock discrimination and movements of narrow-barred Spanish mackerel across northern Australia as indicated by parasites. Journal of Fish Biology 63:765-779 . 326 7. Bibliography Moran, J.D.W., Whitaker, D.J. & Kent, M.L. (1999). A review of the myxosporean genus Kudoa Meglitsch, 1947, and its impact on the international aquaculture industry and commercial fisheries. Aquaculture 172:163-196. Moravec, F. (1994) Parasitic Nematodes of Freshwater Fishes of Europe Kluwer Academic Publishers, Dordrecht, the Netherlands. 477pp. ISBN 0792321723. Morgans, J.F.C. (1964) A preliminary survey of bottom fishing on the north Kenyan banks. Fishery Publications of the Colonial Office No. 21. Morrison, J.A. & Bruce, T. (1981) Scottish herring tagging experiments in the Firth of Clyde 1975-1979, and evidence of affinity between Clyde herring and those in adjacent areas. ICES CM 1981/H:53. Moser, M. (1991). Parasites as biological tags. Parasitology Today 7:82-185. Moser, M. & Hsieh, J. (1992) Biological tags for stock separation in Pacific Herring Clupea harengus pallasi in California. Journal of Parasitology 78:54-60. Mosquera, J., Gomez-Gesteira, M. & Perez-Villar, V. (2000) Using parasites as Biological Tags of Fish Populations: A Dynamical Model Bulletin of Mathematical Biology 62: 87-99. Mulligan, T.J.; Lapi, L., Kieser, R., Yamada, S.B. & Deuwer, D.L. (1983) Salmon stock identification based on elemental composition of vertebrae. Canadian Journal of Fisheries and Aquatic Sciences. 40:215-229. Mudrey, D.R. & Daily, M.D. (1971). Postembryonic development of certain tetraphyllidean and trypanorhynchan cestodes with a possible alternative life cycle for the order Trypanorhyncha. Canadian Journal of Zoology. 49:1249-1253. Murta, A. (2000). Morphological variation of horse mackerel (Trachurus trachurus) in the Iberian and North African Atlantic: implications for stock identification. ICES Journal of Marine Science 57(4):1240-1248. Murta, A.G., Borges, M.F. & Cabral H. (1993) Analysis of stomach contents of some economically important marine fishes caught around Portuguese waters (Division IXa) 1990-1992. ICES CM 1993/ H:39. Murta, A.G., Borges, M.F. & Silveiro, M.L. (1995) Infestation of the horse mackerel, Trachurus trachurus (L.) by Anisakis simplex (Rudolphi, 1809) (Nematoda, Ascaridida) larvae in Portuguese waters (ICES Div. Ixa). ICES C.M. 1995/H:19. Naciri, M., Lemaire, C., Borsa, P. & Bonhomme, F. (1999). Genetic study of the Atlantic/Mediterranean transition in sea bass (Dicentrarchus labrax). Journal of Heredity 90:591-596. 327 7. Bibliography Nahhas, F.M. & Powell, E.C. (1971) Digenetic trematodes of marine fishes from the Floridian northern Gulf of Mexico. Tulane Studies in Zoology and Botany 17:19 Nascetti, G., Paggi, L., Orecchia, P., Smith, J.W., Mattiucci, S., & Bullini, L. (1986) Electrophoretic studies on the Anisakis simplex complex (Ascaridida: Anisakidae) from the Mediterranean and North-East Atlantic. International Journal for Parasitology 16:633-640. Nasir, P. & Gomez, Y. (1977) Digenetic trematodes from Venezuelan marine fishes. Rivista di Parassitologia 38:53-73. Needler, A.W.H. (1930) The migrations of haddock and the interrelationships of haddock populations in North American waters. Contributions in Canadian Biology 6:241-313. Nevedov, G.N., Alferova, N.M. & Chuksin, U.V. (1978) Polymorphic esterases of horse mackerel in the north-east Atlantic. Biologia Morya 2:64-74. Nesbo, C.L., Rueness, E.K., Iversen, S.A., Skagen, D.W. & Jakobsen, K.S. (2000) Phylogeography and population history of Atlantic mackerel Scomber scombrus L: a genealogical approach reveals genetic structuring among the eastern Atlantic stocks. Proc. R. Soc. Lond., B 267:281-292. Nicoll, W. (1909). Studies on the structure and classification of the digenetic trematodes. Quarterly Journal of Microscopical Science 53: 391-487. Nicoll, W. (1910). On the entozoa of fishes of the Firth of Clyde. Parasitology 3:322-359. Nøttestad, L. (2002). Killer whales (Orcinus orca L.) and saithe (Pollachius virens L.) trap herring (Clupea harengus L.) in shallow water by taking advantage of steep bottom topography. ICES Theme Session on Environmental Influences of Trophic Interactions.ICES CM 2002/N:20. O’Connel, M., Dillon, M.C., Wright, J.M., Bentzen, P., Merkouris, S. & Seeb, J. (1998) Genetic structuring among Alaskan Pacific herring (Clupea pallasi) populations identified using microsatellite variability. Journal of Fish Biology 53:150-163 . Oda, D.K. & Parrish, J.D. (1981) Ecology of commercial snappers and groupers introduced to Hawaiian reefs. The reef and man. Proceedings of the fourth international coral reef symposium, Manilla, Philippines. pp.59-68. Odhner, T. (1905) Die trematoden de arktischen Gebietes. Fauna Arctica 4:289372. (in German). Oguz M. Ç. & Kvach, Y. (2006) Occurrence of acanthocephalans in Teleost fishes of Gemlik Bay, Sea of Marmara, Turkey Helminthologia 43:103-108. 328 7. Bibliography Ojaveer, E., Raid, T. & Suursaar, U. (2004) On the assessment and management of local herring stocks in the Baltic Sea. pp.240-250 in Management of Shared Fish Stocks (eds. A.I.L. Payne, C.M. O'Brien & S.I. Rodgers). Blackwell Publishing Ltd., Oxford, UK.367pp. ISBN 1405106174. Olassao, I., Cenderero, O. & Abaunza, P. (1999) The diet of horse mackerel, Trachurus trachurus (Linneaus, 1758), in the Cantabrian Sea (north Spain). Journal of Applied Ichthyology 15:13-198. Oliva, M.E. (1999) Metazoan parasites of the jack mackerel Trachurus murphyi (Teleostei, Carangidae) in a latitudinal gradient from South America (Chile and Peru). Parasite 6:223-230. Oliva, M.E. & Ballón, I. (2002) Metazoan parasites of the Chilean hake Merluccius gayi gayi as a tool for stock discrimination. Fisheries Research 56:313320. Orita, M., (1989) Detection of Polymorphisms of Human DNA by Gel Electrophoresis as SSCPs. Proceedings of the National Academy of Sciences of the United States of America 86:2766-70. Oksanen, J., Kindt, R., Legendre, P. & O'Hara, R.B. (2006) vegan: Community Ecology Package version 1.8-2. http://cran.r-project.org/. Osmanov, S.U. (1940) Uchnoenye Zapinski Leningrad 30:187. Owens, L. (1985) Polypocephalus sp. (Cestoda:Lecanicephalidae) as a biological marker for the banana prawn, Penaeus merguiensis de Man. Australian Journal of Marine and Freshwater Research 36:291-299. Øresland, V. (1986) Parasites in the chaetognath Sagitta setosa in the western English Channel. Marine Biology 92:87-91. Paggi, L., Nascetti, G., Webb, S.C., Mattiucci, S., Cianchi, R. & Bullini, L. (1998) A new species of Anisakis Dujardin, 1845 (Nematoda, Anisakidae) from beaked whales (Ziphiidae): allozyme and morphological evidence. Systematic Parasitology 40:161-174. Palm H.W. (1999) Ecology of Pseudoterranova decipiens (Krabbe, 1878) (Nematoda: Anisakidae) from Antarctic waters. Parasitology Research 85:638-646 Palm, H.W., Andersen, K., Klöser, H. & Plötz, J. (1994) Occurrence of Pseudoterranova decipiens (Nematoda) in fish from the south-eastern Weddell Sea (Antarctic). Polar Biology 14:539-544. Palm H.W. & Walter T. (2000) Tentaculariid cestodes (Trypanorhyncha) from the Muséum national d’Hitoire Naturelle, Paris. Zoosystema 22:641-666. Palumbi, S. R. (1996). Nucleic acids II: The polymerase chain reaction. In: Molecular Systematics ed. D.M. Hillis. Sinauer Association., Sunderland, Massachusetts, USA. pp. 205-247. 329 7. Bibliography Parrish, B. B. & Saville A. (1965) The biology of the north-east Atlantic herring populations. Oceanography and Marine Biololgy Annual Revue 3:323-373. Pascual, S. & Hochberg, F.G. (1996) Marine parasites as biological tags of cephalopod hosts. Parasitology Today 12:324-327. Pascual, S., Gonzalez, A., Arias, C. & Guerra, A. (1996) Biotic relations of Illex coindetti and Todaropsis eblanae (Cephalopoda, Ommastrephidae) in the northeast Atlantic: evidence from parasites. Sarsia 81:265-274. Patterson, K; Cook, R; Darby, C; Gavaris, S; Kell, L; Lewy, P; Mesnil, B; Punt, A; Restrepo, V; Skagen, DW & Stefansson, G (2001) Estimating uncertainty in fish stock assessment and forecasting Fish and Fisheries 2(2):125157. Pawson, M.G. & Jennings, S. (1996) A critique of methods for stock identification in marine capture fisheries. Fisheries Research 25:203-217. Pella, J. & Masuda, M. (2005) Classical discriminant analysis, classification of individuals, and source population composition of mixtures. In: (eds. Cadrin, S.X., Friedland, K.D. & Waldman, J.R.) Stock Identification Methods – Applications in Fishery Science. Elsevier Academic Press, Amsterdam. pp. 517-552. Pereira J. Jr & Boeger W.A. (2005) Larval tapeworms (Platyhelminthes, Cestoda) from sciaenid fishes of the southern coast of Brazil. Zoosystema 27:5-25. Pérez-Ponce de León, G., de la Torre, P., Cummings, M.P., Sarma, S.S.S. & Laclette, J.P. (2000) Phylogenetic Relationships of Acanthocephala Based on Analysis of 18S Ribosomal RNA Gene Sequences Journal of Molecular Evolution 50:523-540. Piaseki, W. (1982). Parasitofauna of cape horse mackerel, Trachurus trachurus capensis Castelnau, 1861. Acta Ichthyologica et Piscatoria 12:43-56. Pierce, G.J. & Santos, M.B. (2003) Diet of Harbour Seals (Phoca vitulina) in Mull and Skye (Inner Hebrides, Western Scotland). Journal of the Marine Biological Association of the United Kingdom 83:647-650. Pitcher T.J. & Hart, J.B. (1982) Fisheries Ecology. Chapman & Hall, London, 414pp. Polonsky, A.S. (1965) The horse mackerel of the Eastern Atlantic and its fishery. Rybnoe Khozyaistvo 41(6):8-10 (in Russian). Polonsky, A.S. & Baydalinov A.P. (1964) On the position of the horse mackerel in the systematics of the English Channel, the Bay of Biscay and the Celtic and North Seas. Trudy AtlantNIRO 13:60-65. 330 7. Bibliography Polonsky, A.S. & Tormosova, I.D. (1969) Growth, age and maturation of the horse mackerel (Trachurus trachurus Linne) in the north-east Atlantic. Trudy AltantNIRO. 23:49-60. (in Russian). Polyanski, Y.I. (1961) Ecology of Parasites of Marine Fishes. In: (eds. Dogiel, V.A., Petruschevsky, G.K. & Polyanski, Y.I.) Parasitology of Fishes. Oliver & Boyd, Edinburgh, UK. pp 48-83. Popova, T.I., & Valter, E.D. (1965) On the elucidation of the life cycle of the fish nematode Contracecum aduncum (Rudolphi 1802) Bayliss, 1920 (Ascaridata). Materialy Nauchnoi Konferenstsii Vsesoyuznogo Obshchestve Gel’mintologov 1:175-178 (in Russian) (English translation by L. Margolis. Fisheries Research Board of Canada Translation Series No. 1797). Power, A.M., Balbuena, J.A. & Raga, J.A. (2005) Parasite infracommunities as predictors of harvest location of bogue (Boops boops L.): a pilot study using statistical classifiers. Fisheries Research 72:229-239. Prager, M.H. (1988) Group method of data handling: a new method for stock identification. Transactions of the American Fisheries Society 117:290-296. Prager, M.H. & Fabrizio, M.C. (1990) Comparison of logistic regression and discriminant analyses for stock identification of anadromous fish, with application to striped bass (Morone saxatilis) and American shad (Alosa sapidissima). Canadian Journal of Fisheries and Aquatic Science 47:1570-1577. Punt, A. (1941) Recherches sur quelques nematodes parasites de poissons de la mere du nord. Bulletin du Museum d'Histoire Naturelle de Belgique 98:1-110. Quinn, J.T. & Deriso, R.B. (1999) Quantitative Fish Dynamics. Oxford University Press, New York, 542 pp. R Development Core Team (2005) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. http://www.Rproject.org. Radujkovic, B.M. & Euzet, L. (1989) Parasites des poissons marins du Montenegro: Monogenes. Acta Adriatica 30(1):51-135. Radulescu, I. (1979) La parasitofaune des chincards de la Mer Noir. In La Chincard de la Mer Noir (ed. Pora, E.) Institut Roumain de Recherches Marines, Constanta, Romania, pp.615-641. Rahda, E. (1971) Some notes on the population dynamics of the monogenean gill parasite Gastrocotyle indica. Marine Biology 8:213-219. Rankine, P.W., (1986) Herring spawning grounds around the Scottish coast. ICES CM/H:15. 331 7. Bibliography Rhebein, H., Gonzalez-Sotelo, C., Perez-Martin, R., Quinteiro, J., ReyMendez, M., Pryde, S., Mackie, I.M. & Santos, A.T. (1999). Differentiation of sturgeon caviar by single strand conformation polymorphism (PCR-SSCP) analysis. Archiv für Lebernsmilfelhygiene, 50:13-17. Rice, A.L. (1963) The Food of Irish Sea Herring in 1961 and 1962. Journal du Conseil International pour l’Exploration de la Mer. 28:188-200. Ricker, W.E. (1979) Growth rates and models. in Fish Physiology. Vol.VIII, eds. Hoar, W.S., Randall, D.J. & Brett, J.R. Academic Press, London, pp. 677-743. Rico, C., Ibrahim, K.M., Rico, I. & Hewitt, G.M. (1997) Stock composition in North Atlantic populations of whiting using microsatellite markers. Journal of Fish Biology. 51:462-475. Roque, S., Sevigny, J.-M., & Bernatchez, L. (2002) Genetic structure of deepwater redfish, Sebastes mentella, populations across the North Atlantic. Marine Biology 140:297-307. Ross, W.R. & Pickard, A. (1990) Use of scale pattern and shape as discriminators of wild and hatchery striped bass stocks in California. In: Fish Marking Techniques (eds. Parker, N.C., Giorgi, A.E., Heidinger, R.C., Jester, B.D., Prince, E.D. & Winans, G.A.)American Fisheries Society, Bethseda, USA. pp.71-77. Rossi A.R., Capula M., Crosetti D., Sola L. & Campton D.E. (1998) Allozyme variation in global populations of striped mullet, Mugil cephalus (Pisces, Mugilidae) .Marine Biology 131:203-212. Rounsefell, G.A. (1930) Contribution to the biology of the Pacific herring, Clupea pallasi, and the condition of the fishery in Alaska. Bulletin of the United States Bureau Fisheries 45:227-320. Rückert, C., Floeter, J. & Temming, A. (2002) An estimate of horse mackerel biomass in the North Sea, 1991–1997. ICES J. Mar. Sci. 59:120–130. Rudolphi, C.A. (1819). Entozoorum synopsis cui accedunt mantissa duplex et indices locu pletissimi. August Rücker, Berolini, 811pp. Russell, E.S. (1931) Some theoretical considerations on the “overfishing” problem. Journal du Conseil International pour l’Exploration de la Mer 6:3-20. Ruzzante, D.E., Taggart, C.T., Cook, D. & Goddart, S.V. (1997) Genetic differentiation between inshore and offshore Atlantic code (Gadus morhua) off Newfoundland: a test and evidence of temporal stability. Canadian Journal of Fisheries and Aquatic Science 54:2700-2708. 332 7. Bibliography Ruzzante, D.E., Mariani, S., Bekkevold, D., Andre, C., Mosegaard, H., Clausen, L.A.W., Dahlgren, T.G., Hutchinson, W.H, Hatfield, E.M.C., Tortensen, E., Brigham, J., Simmonds, E.J., Laikre, L., Larsson, L.C., Stet, R.J.S., Ryman, N. & Carvahlo, G.R. (2006) Biocomplexity in a highly migratory marine fish, Atlantic herring. Proceedings of the Royal Society, series B. 273: 1459-1464. Saila, S.B. (2005) Neural networks used in classification with emphasis on biological populations. In: (eds. Cadrin, S.X., Friedland, K.D. & Waldman, J.R.) Stock Identification Methods – Applications in Fishery Science. Elsevier Academic Press, Amsterdam. pp 553-570. Sardella, N.H. & Timi, J.T. (2004) Parasites of Argentine hake in the Argentine Sea: population and infracommunity structure as evidence for host stock discrimination. Journal of Fish Biology 65:1472-1488. Sarkar, N. & Ghosh, S. (1991). Two new coelozoic Myxosporidia (Myxosporea: Myxosporea) from estuarine teleost fishes (Mugilidae) of West Bengal, India. Proceedings of the Zoological Society of Calcutta. 44:131-135. Sarkar, N. & Mazumder, S.K. (1983). Studies on myxosporean parasites (Myxozoa: Myxosporea) from marine fishes in West Bengal, India: I. Description of three new species from Tachysurus spp. Archiv fur Protistenkunde. 127:59-63 Savi, P. (1997) Anisakis, divagazione di mezza estate. Il pesce 5:65-68. Saville, A., Parrish, B.B., & Baxter, I.G. (1966) Review of the herring fisheries and stocks off the west coast of Scotland. ICES C.M. 1966/H:1 Schaefer, K.M. (1987) Reproductive biology of black skipjack, Euthynnus lineatus, and eaestern Pacific tuna. Inter American Tropical Tuna Commision Bulletin 19 260pp. Schneider, S., Roessli, D.& Excoffier, L. (2000) Arlequin Version 2.0: a software for population genetics data analysis. Genetics and Biometry laboratory, Department of Anthropology University of Geneva, Switzerland. Schroeder, W.C. (1942) Results of haddock taggings in the Gulf of Maine from 1923 to 1932. Journal of Marine Research 5:1-19 Schweigert, J.F. (1990) Comparison of morphometric and meristic data against truss network for describing Pacific herring stocks. American Fisheries Society Symposium 7:47-62. Schweigert, J., Flostrand, L., Slotte, A. & Tallman, D. (2001) Application of coded wire tagging technology in Pacific herring to investigate stock structure and migration. ICES C.M. 2001/O:12. Scott, M.E. & Anderson, R.M. (1984) The population dynamics of Gyrodactylus bullatarudis (Monogenea) within laboratory populations of the fish host Poecilia reticulata. Parasitology 89:159-194. 333 7. Bibliography Secor, D.H. (1999) Specifying divergent migrations in the concept of stock: the contingent hypothesis. Fisheries Research 43:13-34 Seeb, L.W. & Crane, P.A. (1999) Allozymes and mitochondrial DNA discriminate Asian and North American populations of chum salmon in mixed stock fisheries along the south coast of the Alaska Peninsula. Transactions of the American Fisheries Society 128:88-103. Sindermann, C.J. (1961) Parasite tags for marine fish. Journal of Wildlife Management 25:41-47. Sindermann, C.J. (1983) Parasites as natural tags for marine fish: a review. NAFO Science Council Studies 6:63-71. Sissenwine, M. (2005) Foreword. In: (eds. Cadrin, S.X., Friedland, K.D. & Waldman, J.R.) Stock Identification Methods – Applications in Fishery Science. Elsevier Academic Press, Amsterdam. pp xi-xii. Schaefer, K.M. (1991) Geographic variation in morphometric characters and gillraker counts of yellowfin tuna, Thunnus albacares, from the Pacific Ocean. Fisheries Bulletin 89:289-297. Shaw, P.W., Pierce, G.J. & Boyle, P.R., (1999) Subtle population structuring within a highly vagile marine invertebrate, the veined squid Loligo forbesi, demonstrated by microsatellite DNA markers. Molecular Ecology 8:407-417. Shotter, R.A. (1973) Changes in the parasite fauna of whiting Odontogadus merlangus L. with age and sex of host, season and from different areas in the vicinity of the Isle of Man. Journal of Fish Biology 5:559-573. Shulman, M.J. & Bermingham, E. (1995) Early life histories, ocean currents, and the population genetics of Caribbean reef fishes. Evolution 49:897-910. Shulman, S.S. & Shulman-Albova, R.E. (1953) Parasites of Fishes of the White Sea (in Russian). Akademiya Nauk SSSR, Moscow, 198pp. Skirnisson, K., Eydal, E., Gunnarsson, E. & Hersteinsson, P. (1993) Parasites of the arctic fox (Alopex lagopus) in Iceland. Journal of Wildlife Diseases 29 (3): 440446. Smith, J.W. (1983) Anisakis simplex (Rudolphi, 1809, det. Krabbe, 1878) (Nematoda: Ascaridoidea): morphology and morphometry of larvae from euphausiids and fish, and a review of the life-history and ecology. Journal of Helminthology 57:205-224. Smith, M. (1993) Neural Networks for Statistical Modeling John Wiley & Sons, Inc. New York. 256pp. 334 7. Bibliography Smith, P.J., Roberts, P.E. & Hurst, R.J. (1981) Evidence for two species of arrow squid in the New Zealand fishery. New Zealand Journal of Marine and Freshwater Research 15: 247-253. Smith, P.J., McMillan, P.J., Bull, B., McVeagh, S.M., Gaffney, P.M. & Chow, S. (2002) Genetic and meristic variation in black and smooth oreos in the New Zealand Exclusive Economic Zone. New Zealand Journal of Marine and Freshwater Research 36:737-750. Smith W.G. & Jamieson A. (1986) Stock discreteness in herring: a conceptual revolution. Fisheries Research 4:223–234. Smith-Vaniz, W.F. 1986. The Carangidae. In: “Fishes of the North-Eastern Atlantic and the Mediterranean, vol. II” (Whitehead, P.J.P., Bauchot, M. -L., Hureau, J.-C., Nielsen, J. and Tortonese, E., Eds), pp.815-844. Unesco, Paris. ISBN 9230023086. Sola, A., Motos, K., Franco, C. & Lago de Lanos, A. (1990) Seasonal occurrence of pelagic fish eggs and larvae in the Cantabrian Sea (VIIIc) and Galicia (IXa) from 1987 to 1989. ICES C.M. 1990/H:25. Speare, P. (1994) Relationships among black marlin, Makaira indica, in eastern Australian coastal waters, inferred from parasites. Australian Journal of Marine and Freshwater Research 45:535-550. Speare, P. (1995) Parasites as biological tags for sailfish Istiophorus platypterus, from east coast Australian waters. Marine Ecology Progress Series 118:43 50. Southwell, T. & Macfee, N. (1925) On a collection of Acanthocephala in the Liverpool School of Tropical Medicine. Annals of Tropical Medicine and Parasitology 19:141-184. Stephenson, R.L. (1991) Stock discreteness in Atlantic herring: a revies of the arguments for and against. In: “Proceedings of the International Herring Symposium, Achorage, Alaska, October 23-25, 1990” (eds. Wespestad, V., Collie, J. & Collie, E.). University of Alaska, Fairbanks. pp. 659-666. Stephenson, R.L. (1999) Stock complexity in fisheries management: a perspective of emerging issues related to population sub-units. Fisheries Research 43:247-249. Stepien, C. A. & T. D. Kocher (1997) Molecules and morphology in studies of fish evolution. In: Molecular systematics of fishes, eds. T. D. Kocher & C. A. Stepien, Academic Press, San Diego, California, USA. pp. 1-11. Sterud, E., Mo, T.A., Collins, C.M. & Cunningham, C.O. (2002) The use of host specificity, pathogenicity and molecular markers to differentiate between Gyrodactylus salaris Malmberg, 1957 and G. thymalli Zitnan, 1960 (Monogenea: Gryodactylidae). Parasitology 124:203-213. 335 7. Bibliography Stransky, C. (2004) Stock separation and growth of redfish (genus Sebastes) in the North Atlantic by menas of shape and elemental analysis of otoliths. PhD dissertation. University of Hamburg, 132pp. (http://www.sub.uni-hamburg.de/opus/volltexte/2004/2173) Straus, R.E. & Bookstein, F.L. (1982) The truss: body form reconstructions in morphometrics. Systematic Zoology 31:113-135. Stunkard, H.W. (1977) Studies on tetraphyllidean and tetrarhynchidean metacestodes from squids taken on the New England coast. Biological Bulletin 153:387-412. Sunnucks, P., (2000) Efficient genetic markers for population biology. Trends in Ecology & Evolution, 15:199–203. Sunnucks, P., Wilson, A.C.C., Beheregaray, L.B., Zenger, K., French, J. & Taylor, A.C. (2000) SSCP is not so difficult: the application and utility of singlestranded conformation polymorphism in evolutionary biology and molecular ecology. Molecular Ecology 9:1699-1710. Swain, D.P. & Foote, C.J. (1999) Stocks and chameleons: the use of phenotypic variation in stock identification Fisheries Research 43:113 – 128. Sylianteng, C.E. (2004) Using parasites as biological tags to distinguish between two populations of Atlantic Horse Mackerel from two sites off the Portuguese Coast. M.Sc. Thesis, University of Aberdeen. Szuks, H. (1980) Die verwendbarkeit von parasiten zur Gruppentrennung bein Grenadierfish Macrorurus rupestris. Angewandte Parasitologie 21:211-214. Temming, A. & Herrmann J.-P. (2001a) Gastric evacuation of horse mackerel. I. The effects of meal size, temperature and predator weight. Journal of Fish Biology. 58:1230-1245. Temming, A. & Herrmann, J.-P. (2001b) Gastric evacuation in horse mackerel. II. The effects of different prey types on the evacuation models. Journal of Fish Biology 58:1246-1256. Timi, J.T. & Poulin, R. (2003) Parasite community structure within and across host populations of a marine pelagic fish: how repeatable is it? International Journal for Parasitology 33:1353-1362. Thiel, R., Mehner, T., Köpcke, B. & Kafemann, R. (1996) Diet niche relationship among early life stages of fish in German estuaries. Marine and Freshwater Research 47:123-136 Thompson, B.M. & Harrop, R.T. (1987) The distribution and abundance of bass (Dicentrarchus labrax) eggs and larvae in the English Channel and southern North Sea. Journal of the Marine Biological Association of the United Kingdom 67:263274. 336 7. Bibliography Thompson, J.D., Higgins, D.G. & Gibson, T.J. (1994) CLUSTALW: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Research 22:4673-4680. Thresher, R.E. (1999). Elemental composition of otoliths as a stock delineator in fishes. Fisheries Research 43:165-204. Trouvé, S., Sasal, P., Jourdane, J., Renaud, F. & Morand, S. (1998) The evolution of life-history traits in parasitic and free-living platyhelminthes: a new perspective Oecologia 115:370-378. Utter, F. (1991). Biochemical genetics and fishery management: an historical perspective. Journal of Fish Biology 39A:1-20. Uriarte, A., Alvarez, P., Iversen, S., Molloy, J., Villamor, B., Martins, M.M. & Myklevoll, S. (2001). Spatial pattern of migration and recruitment of north east Atlantic mackerel. ICES CM 2001 O:17 36 pp. Uriarte, A. & Lucio, P. (2001). Migration of adult mackerel along the Atlantic European shelf edge from a tagging experiment in the south of the Bay of Biscay in 1994. Fisheries Research 50:129-139. Uzmann, J.R. (1970) Use of parasites in identifying lobster stocks. Journal of Parasitology 56:349. Valentini, A., Mattiucci, S., Bondanelli, P., Webb, S.C., Mignucci-Giannone, A.A., Colom-Llavina, M.M. & Nascetti, G. (2006) Genetic relationships among Anisakis species (Nematoda: Anisakidae) inferred from mitochondrial cox2 sequences, and comparison with allozyme data. Journal of Parasitology 92:156-66. Valtonen, E.T. (1979) Neoechinorhynchus rutili (muller, 1780) (Acanthocephala) in the whitefish Coregonus nasus (Pallas) sensu Svardson from the Bay of Bothnia. Journal of Fish Diseases 2:99-103. Van Cleave H.J. (1953) A preliminary analysis of the acanthocephalan genus Corynosoma in mammals of North America. Journal of Parasitology 39:1-13. Van Marlen, B. (2000) Technical modifications to reduce the bycatches and impacts of bottom-fishing gears. In: Kaiser, M.J. & De Groot, S.J. (eds.), Effects of Fishing on Non-target Species and Habitats. Blackwell Science, London, UK. pp. 253-268. Venables, W.N. & Ripley, B.D. (2002) Modern Applied Statistics with S. Fourth Edition. Springer Verlag, New York. 495pp. ISBN 0387954570. Villamor, B., Abaunza, P., Lucio, P. & Porteiro, C. (1997) Distribution and age structure of mackerel (Scomber scombrus, L.) and horse mackerel (Trachurus trachurus, L.) in the northern coast of Spain, 1989-1994. Scientia Marina 61:345366. 337 7. Bibliography Vlasenko, P.V. (1931) Trudy Karadagskoj Biologiceskoj Sancii 4(4): 88 Waldman, J.R., Richards, R.A., Schill, W.B., Wirgin, I. & Fabrizio, M.C. (1997) An empirical comparison of stock identification techniques applied to striped bass. Transactions of the American Fisheries Society 126:369-385 Waldman, J.R. (2005a) Definition of Stocks: An Evolving Concept In: (eds. Cadrin, S.X., Friedland, K.D. & Waldman, J.R.) Stock Identification Methods – Applications in Fishery Science. Elsevier Academic Press, Amsterdam. pp 7-16. Waldman, J.R. (2005b) Meristics in Stock Identification Methods: Applications in Fishery Science. In: (eds. Cadrin, S.X., Friedland, K.D. & Waldman, J.R.) Stock Identification Methods – Applications in Fishery Science. Elsevier Academic Press, Amsterdam. pp 197-210. Walters, C. & Martell, S.J.D. Stock assessment needs for sustainable fisheries management .Bulletin of Marine Science 70 (2):629-638. Ward, R.D., Woodwark, M. & Skibinski, D.O.F. (1994). A comparison of genetic diversity levels in marine, freshwater, and anadromous fishes. Journal of Fish Biology 44:213-232. Ward, R.D. (2000). Genetics in fisheries management. Hydrobiologia 420:191201. Warnes, G & Leisch, F. The “genetics” Package rproject.org. v1.1.2 http://www.cran- Williams, H.H., MacKenzie, K. & McCarthy, A.M. (1992). Parasites as biological indicators of the population biology, migrations, diet, and phylogenetics of fish. Reviews in Fish Biology and Fisheries 2:144-176. Woods, R.J. (1968) Autumn-spawning grounds of herring to the west of Scotland. Symposium on the biology of early stages and recruitment mechanisms of herring. ICES CM 1968 Document 8. Wooten, R.J. (1998) Ecology of teleost fishes. Second edition. Kluwer Academic Publishers, Dordrecht, 386 pp. Yamaguti, S. (1935) Studies on the helminth fauna of Japan. Part 8. Acanthocephala I. Japanese Journal of Zoology 6:247-278. Yloenen, H; Jacob, J; Runcie, MJ; Singleton, GR (2003) Is reproduction of the Australian house mouse (Mus domesticus) constrained by food? A large-scale field experiment. Oecologia 135: 372-377. Yurakhno, V.M. (1988). Novye O Myxsosporidiye Roub Chernogo Morya. [On New Myxosporidia of Black Sea fishes.] Parazitologia 22:521-524 (in Russian). 338 7. Bibliography Zidowitz, H., Fock, H.O. & von Westernhagen H. (2002) The role of Zenopsis spp. as a predator in seamount and shelf habitats. ICES Theme Session on Oceanography and Ecology of Seamounts ICES CM 2002/M:28. Zubchenko, A.V. (1985) Use of parasitological data in studies of local groupings of rock grenadier, Coryphanoides rupestris. In Parasitology and Pathology of Marine Organisms of the World Ocean.(eds. Hargis, W.J.) pp. 19-23. NOAA Technical Report NMFS 25. 339 Appendix I – Translation of Gaevsakya & Kovaleva (1980) from Russian into English. Eco-geographical features of the parasitofauna of the common Atlantic Horse Mackerel (Ekologo-Geogeficheskie Ocobennocti Parazitofauni Obuknovennou Stavride Atlanticheskogo Okeana) Studies of the biological resources of the Atlantic Ocean. (Issledovaniya Biologicheskich Resoursov Atlantichskogo Okeana) Kaliningrad, AtlantNIRO, 1980, pp18-24 Gaevskaya A. & Kovaleva A. Introduction The common horse mackerel Trachurus trachurus L., which dwells in the eastern part of the Atlantic Ocean, forms two subspecies. The area of one of them, Trachurus trachurus, covers an extensive territory from the North Sea to the middle of the African continent. The second subtype. T. trachurus capensis Castelnau, is located on the south-western coast of Africa. The present communication is based on the results of the examination 5600 individual horse mackerel, from which 275 individuals were investigated by the method of complete parasitological dissections. Material was taken from the North and Celtic seas, on the southwest of the Irish shelf, in the English Channel, Bay of Biscay, the Strait of Gibraltar, also, on the coast of the Western Sahara. In all in this material we recorded 37 forms of the parasites, 11 of which are noted for the first time, 3 of which are new to science. Taking into account its own and literature data on horse mackerel it is host to 39 species of parasite, the geographical distribution of which represented in the table. Brief information about the parasites either in cases of either new host records, or new geographical records, is given below. Protozoa We discovered 3 forms of microsporidian and 1 coccidian. Alataspora serenum Gayevskaya And Kovalzhova, 1979 is found in 15% of horse mackerel in the Celtic sea and Bay of Biscay. Kudoa quadratum (Theolan, 1895) is recorded in the Bay of Biscay. On the north-western shores of Africa it is replaced by K. nova Nadienova, 1975. Eirneria cruciata 1 (Theolan, 1892) was found in 26/50 fishes in the North and Celtic seas, the Bay of Biscay; it reaches the Western Sahara (Reimer, 1976). Trematoda There are 14 species of trematode. Monascus filiformis (Rudolphi, 1819) is recorded throughout the north-eastern Atlantic infecting 1213% of horse mackerels. 1 Goussia cruciata Parasite Alataspora serenum Kudoa quadratum Kudoa nova Eimeria cruciata Ancylocoelium typicum Tergestia laticollis Monascus filiformis Zoogonus rubellus *Opechona magnibursta Neopechyuona to puriforme *Podocotyloides chloroscombri Hemiurus communis Hemiuris luyuei Aphanurus stossichi Ectenurus lepidus Lecithaster gibbosus L. confusus Derogenes varicus Diplectanotrema trachuri Pseudaxine trachuri *Heteraxinoides atlanticus Cemocotyle trachuri *Lacistorynchus tenuis *Ghristianella minuta *Grillotia erinaceus *Nybelinia lingualis *Anthobothrium cornucopia *Phyllobothrium sp. Scolex pleuronectis Tetrarhynchid larvaye Rhadinorhynchus cadenati Anisakis simplex Contracaecum aduncum *Caligus elongatus *Caligus curtus Caligus pelamydis Lernanthropus trachuri Total North Sea + + + + + + + + + + + + + + + + + + + 20 English ChaCeltic Sea Biscay + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + 15 15 17 Gibraltar + + + + + + + + + + + + + 15 Sahara + + + + + + + + + + + + + + 17 * new host records. Tergestia laticollis (Rudolphi. 1819) is found for the first time in the Celtic sea and the Bay of Biscay. Depending on region the prevalence varied from 32 to 60%, the intensity of infection did not exceed 510. This species is not found south of the Strait of Gibraltar. Opechona magnibursata Gayevskaya and Kovalyova sp.n., is described as a parasite of horse mackerel from the Bay of Biscay. Podocotyloides chloroschombri Fisehthal and Thomas, 1970 is found in 18% of horse mackerel in the shelf waters of the Western Sahara. It is encountered singly. It was first described from one specimen from Chloroscombrus chrysurus off Ghana (Fisenthal & Thomas, 1970). This is the first record of the parasite in Trachurus. 20 - 22° N seems to be the region of highest abundance. Aphanurus stossichi (Montichelli, 1891) is reported for the first time in the Celtic sea. Lecithaster confusus Odyuner, 1905 is recorded for the first time in horse mackerel. 1 example is found. Apparently a random parasite. Monogenea 6 forms were found. Diplectanotrema trachuri 2 Kovaleva, 1970. Not found north of the Straits of Gibraltar, in the south it reaches to 23° N. Gastrocotyle trachuri Ben and Hesse, 1863. Not found in the Celtic Sea and the Bay of Biscay. Depending on region and season of study the prevalence varied from 6 to 75%, the intensity of infection was 1.18 per individual. Cemocotyle trachuri Dillon and Hargis, 1965 is reported for the first time in the North Sea. Heteraxinoides atlanticus Kovaleva And Gaevskaya, 1979 was found in horse mackerels along the African coast. It is found singly. Cestoda The cestode fauna comprises of larval forms of 8 species, the majority of which are new host records. In all cestodes found, the final hosts are cartilaginous fishes. Acanthocephala Rhadinorynchus cadenati Golvan, 1969, which usually parasitizes horse mackerel, was recorded on the coast of Africa, where it infects 15-18% of fishes. During January 1976 it was found in the Bay of Biscay. Nematoda We recovered 2 forms of larval forms the nematode, Contracaecum 3 spp. and Anisakis spp., that invade horse mackerel in the entire area. Copepoda We found 4 forms. Caligus curtus Muller, 1785, and C. elongatus Nordmann, 1832 were found in horse mackerel for the first time. Lemanthropus trachuri Brian, 1903. A specific parasite of horse mackerel, not found north of the Strait of Gibraltar. The analysis of parasitofauna of horse mackerel revealed the close connection of species composition, range and intensity of infection of the parasitofauna with the biological and ecological features of the host. 2 3 Now Paradiplectanotrema trachuri (Kovaleva, 1970) Now Hysterothylacium spp. Infection by parasites with a direct cycle of development (eg. coccidia, myxosporidians and monogeneans) occurs in places of high accumulation of fishes, revealed by horse mackerel coccidia and monogenea that are specific to Carangidae. High infection with these groups (50-60% in some seasons) testifies about the formation in a number of regions of dense seasonal concentrations of these fishes. The food spectrum of horse mackerel is very rich. In the north-eastern Atlantic in summer it is predominantly pelagic organisms (copepods, shrimps, roe, larvae and young of fishes), in winter an important place is occupied by benthic crabs and molluscs (Letaconnoux, 1951). On the north-western coast of Africa the food of horse mackerel consists of chaetognaths, copepods, euphausids, myxids, decapods and fishes (Lipaskaya, 1972). Such diverse range of food items causes a significant variation in the parasite fauna of horse mackerel. A large part of these parasites make use of intermediate hosts in their life cycle. Trematodes of the genus Hemiurus use copepods for this purpose, Aphanurus stossici - copepods and chaetognaths, Zoogonus rubellus - polychaetes, Opechona spp.- chaetognaths, jellyfish and ctenophores, Ectenurus lepidus - chaetognaths and jellyfish, Derogenes varicus chaetognaths and copepods (Solenko, 1966; Dollfus, 1963; Koie, 1975; Lebour, 1917). Despite the fact that the jellyfish and ctenophores were not found in the prey items of horse mackerel, on the basis the presence of the above-indicated helminths it is possible to speculate that they feed upon these invertebrates. Cestodes develop in euphausids, copepods and molluscs. Detection in the horse mackerel of the larval forms of 8 different cestodes testifies about its diverse feeding behaviour on invertebrate fauna. For the acanthocephalan, R. cadenati, the larval cestodes and the nematodes, horse mackerel are intermediate hosts. In this case cestodes complete their development in the sharks, and nematodes the sea mammalian (Anisakis simplex) or large predatory fishes ( Contracaecum aduncum ). Thus, the analysis of the parasitological interrelations shows that the horse mackerel is an important link in the food chain, occupying the trophic niche of trophic level IV, and it is the important connecting link between small planktivores of trophic level III, and large predators - bony fishes, sharks and marine mammals, to complete the food chain. The geographical analysis of the parasites of horse mackerel makes it possible to reveal division into two distinct groups; those found throughout the entire area of study and those tied to specific regions. The first group is extensive: M.filiformis, E. lepidus , all Monogenea (with the exception of D . trachuri), cestodes and nematodes. Infection by these parasites varies from one region of study to another. Infection by M. filiformis and A. simplex decreases southwards; the first of them is encountered in 13% of fishes in the North Sea and the Bay of Biscay, 28% of fishes in the region of the Strait of Gibraltar and 3% in Western Sahara region, A. simplex is found in 50.7%, 8% and 24% respectively. At the same time occurrence in the horse mackerel of E. lepidus, Gastrocotyle, Pseudaxine and Cemocotyle decreases towards the north. E. lepidus is found in 42% of fishes in the Western Sahara and 6% of fishes in Bay of Biscay. G. trachuri in 63% and 30%, C. trachuri in 24% and 6% respectively. The second group of parasites is less numerous, but it characterizes the heterogeneity of the stocks of horse mackerel. It is clearly demonstrated by the distribution of separate forms of myxosporidia to specific regions (see table 1). Trematodes like Z. rubellus are found only in the region of the Strait of Gibraltar, Podocotyloides chloroscombri and R. cadenati - Western Sahara. Diplectantorema trachuri and Lernanthropus trachuri extend south of the Straits of Gibraltar. Not less interesting is the fact that parasites in separate regions noted above infect horse mackerel differently. For example, P. chloroscombri in 23.27° N region is encountered in 6% of fishes, and on 20.22° N. in 32%, E. lepidus 12 and 75%, D. trachuri 23.5 and 6.2% respectively. Thus, parasitological data shows, that the horse mackerel forms continuous separate local groupings from the North Sea up to 20° N, which is confirmed by a study of the morphology of horse mackerel in these regions (Kompotilski, 1975). However, these stocks are not in a strict isolation. Horse mackerel from Western Sahara region probably migrate into the Bay of Biscay. As confirmation of, this during January 1976, R. cadenati (a characteristic parasitise fish in subtropical waters) were found in horse mackerel in this region. In the Bay of Biscay horse mackerel were found infected with Kudoa quardratum and A. stossici, recorded from horse mackerel from the Mediterranean (Parukhin, 1976; Theolan, 1892) and absent in the horse mackerel that dwell along the Atlantic coast of Africa. All these forms are not characteristic of the fauna of the Bay of Biscay and, most probably, they could be brought into this region only by their hosts migrating to this location. The analysis of parasitofauna of horse mackerel showed that it consists of several groups: 1.) Those specific to Trachurus trachurus, 2.) Those parasitizing in several fishes in the Carangidae, 3.) Characteristic of fishes which occupy a similar ecological niche to the horse mackerel. Alatospora serenum, would seem characteristic for this host (Gaevskaya & Kovaleva, 1979), however, the weak host-specificity of this group of parasites does not make it possible to make a concrete conclusion about the degree of their specificity. The first group of parasites are characterised by D. trachuri, P. chloroscombri, K. quadratum, A. typicum, in this case the propagation of each of the species indicated is limited by a comparatively small water area. The second group comprises the forms that represent the general parasitofauna of the Carangidae. G. trachuri, P. trachuri, C. trachuri, T. laticollis, M. filiformis are all found in carangids in different regions of the worlds oceans. The third group of the parasitofauna of horse mackerel are forms common with many other fishes that lead similar ways of life. As a rule, these are parasites with a compound cycle of development; trematodes, cestodes and nematodes. The basic factor that facilitates the infection of fishes with these parasites, is trophic transmission. For example, in the north-eastern Atlantic, Hemiurus, Lecithaster, Derogenes, etc., which use copepods and chaetognaths as first hosts become a component part of the parasitofauna of horse mackerel that feed upon these prey items. In the same region the parasitic copepods Caligus curtus and C. elongatus infect fishes which are typical planktivores. The proximity of horse mackerel with the usual hosts of these crustaceans creates favourable possibilities for their infection. Thus, parasitofauna of horse mackerel represents the stable state of complex biotic relations. Conclusions 1. In the horse mackerel, investigated in 1973 - 76. from the North Sea to the coast of the Western Sahara, we found 39 forms of the parasites, 11 of which new host records, 3 forms (Alataspora serenum, Opecona magnibursata, Heteraxinoides atlanticus) are new for the science. 2. Ecological analysis of the parasitofauna of the horse mackerel reveals the close connection between the composition, range and intensity of infection of parasite species with the biological and ecological features of the host. 3. It is established that there is an uneven spatial distribution of the parasites of horse mackerel, some of which are extended over entire area, others are tied to separate regions. 4. The possibility of using the parasites during population studies of horse mackerel is shown. References 1. Gayevskaya A. V., Kovaleva A. A. New and rarely encountered forms of myxosporidians from the fishes of the Celtic sea. Parazitologiya (1979) 13;2: 159-165. 2. Solonchenko A. I. Metacercariae of the trematode Aphanurus stossici (Montichelli, 1891) in the mesenteries of Acartia clausi. In the book: Helminthic fauna of animal South seas. Kiev. Naukova Dumka , (1966), pp140.141. 3. Lipskaya N. Some data about growth and feeding of horse mackerel. Trachurus trachurus L. on the West coast of Africa. Trudy NIRO - - All Union Scientific Research Institute of Sea Fisheries and Oceanography. (1972) 77;2: 186-196. 4. Parukhin A. M. The parasitic worms of commercial fishes of southern Baltic Sea, Naukova Dumka, (1976), pp.98. 5. Dollfus R.-Ph. Liste des Coelenteres marins, palearctiques et indicus ou ont ete trouves des Trematodes digenetiques. Bull. L'inst. Peches Marit (1963) 9-10:33-57. 6. Fiscktal I.H., Thomas J. D. Digenetic trematodes of marine fishes from Ghana: family Opecoelidae.Proc. Helminthol. Soc. Wash. (1970). 37;2: 129-141. 7. Koie М. On the morphology and life-history of Opechona bacillarie (Moller, 1859) Looss, 1907 (Trematoda, Lepocreadiidae). Ophelia, (1975). 13: 63-86. 8. Kompotilski A. The intraspecific geographical variability of horse mackerel Trachurus trachurus (L.) in the West African Shelf waters. Acta ichthyol. et piscatol. (1975) 5:13-29. 9. Lebour М. Some parasites of Sagitta bipunctata. J. Mar. biol. Ass., U. K. (1917) 2 : 202-206. 10. Letaconnoux R. Contribution a l'etude des especes du genre Trachurus et. specialement du Trachurus trachurus (Linne, 1758). Mem. off. sci tech. Pech. marit. (1951) 15: 67 11. Reimer L. W. Protozoa von marinen Fischen des Kanals und von Nord-westafrika. Wissenschaft. Zeitschr., Padag. Hochsch. (1976) 143-147. 12. Thelohan P. Recherches sur les myxosporidies. Bull. Sci, Fr. Beig. (1895) 26: 347. Appendix II – Published paper on myxosporean parasitofauna of horse mackerel Acta Parasitologica, 2005, 50(2), 000–000; ISSN 1230-2821 Copyright © 2005 W. Stefañski Institute of Parasitology, PAS The myxosporean parasitofauna of the Atlantic horse mackerel, Trachurus trachurus (L.) in the North-East Atlantic Ocean and Mediterranean Sea Stefański Neil Campbell School of Biological Sciences (Zoology), Zoology Building, University of Aberdeen, Aberdeen, Scotland, AB24 2TZ, U.K. Abstract As part of a multidisciplinary stock identification study 1002 horse mackerel [Trachurus trachurus (L.)] from 12 sites in the North-East Atlantic and Mediterranean Sea were examined for the presence of myxosporean parasites, with the aim of identifying species that could be used as biological tags. Five species of myxosporean parasites were found. Examination of gall bladders revealed infections with Alataspora serenum, Alataspora solomoni and an unidentified Kudoa sp., Kudoa nova was found in the red muscles. The liver of a single fish was found to be infected with Myxobolus spinacurvatura. Infection with A. solomoni and M. spinacurvatura represents new host records. Kudoa sp. is likely to be a previously undescribed species. A. serenum, A. solomoni and K. nova are shown to be potentially useful tags for stock identification. Key words Myxosporean parasites, fish, Trachurus trachurus, Atlantic Ocean, Mediterranean Sea Introduction Skóra The horse mackerel (Trachurus trachurus) is a small pelagic fish, common in the North Atlantic and Mediterranean Sea. It is the most northerly representative of the genus Trachurus and is distributed from West Africa to the Norwegian Sea, including Iceland, and throughout the North Sea, the Mediterranean and Black Seas (Smith-Vaniz 1986). There has been uncertainty for several years about the validity of stock definitions for this species in the North-East Atlantic (Murta 2000). Currently, ICES considers there to be three stocks in this area, the North Sea Stock, the Western Stock, in waters to the west of the British Isles, extending into the Bay of Biscay, and the Southern Stock, in waters around the Iberian Peninsula (ICES 1998). There are no stocks defined in the African coastal area, or the Mediterranean Sea. To support the possible use of parasites as biological tags for this species, a study was made of the parasite fauna of the horse mackerel throughout European waters. Sixty-eight different parasite taxa have been recorded by this study (Mac Kenzie et al. 2004). Three myxosporeans have been described from the horse mackerel in the North Atlantic; Alataspora serenum Gaevskaya et Kovaleva, 1979, Kudoa nova Naidenova, 1975 and Kudoa quadratum (Thélohan, 1892) (Gaevskaya and Kovaleva 1979a, Gaevskaya and Kovaleva 1980). A number of species of myxosporeans have been described from other members of the genus Trachurus. Alataspora solomoni Yurakhno, 1988 was described from the Mediterranean horse mackerel, T. mediterraneus, from the Black Sea, near Sevastopol (Yurakhno 1988). Davisia donecae Gaevskaya et Kovaleva, 1979 and Ceratomyxa australis Gaevskaya et Kovaleva, 1979 have been described from the Cape horse mackerel, T. capensis from Namibian waters (Gaevskaya and Kovaleva 1979b). Materials and methods Samples of horse mackerel were collected by a combination of commercial and research vessels at 19 sites in the North Atlantic and Mediterranean Sea during both 2000 and 2001, Corresponding address: Neil.Campbell@abdn.ac.uk Śląski mon. Other sites where the parasite was recorded are shown on Figure 1. Results Five myxosporean parasites from three families were recorded infecting the horse mackerel in the Atlantic Ocean and Mediterranean Sea. Alataspora serenum, Alataspora solomoni and Kudoa sp. were found to be coelozoic in the gall bladder. Kudoa nova was found in the red musculature. Myxobolus spinacurvatura was found in the liver of one fish from the Aegean Sea. Alataspora serenum Gaevskaya et Kovaleva, 1979 (Fig. 2) Order: Bivalvulida Shulman, 1959 Family: Alatasporidae Schulman et al., 1979 Genus: Alataspora Schulman et al., 1979 Fig. 1. Locations of sampling sites in European waters. Sites where myxosporeans were recorded are signified by filled circles and denoted as follows: 1. Alataspora serenum. 2. Alataspora solomoni. 3. Kudoa sp. 4. Kudoa nova. 5. Myxobolus spinacurvatura. Sites where myxosporeans were not present are shown as empty circles as part of the multidisciplinary stock identification project “HOMSIR”. For a detailed description of sampling procedures, see Abaunza et al. (in press). One thousand and two fish from 12 sites over both years were deep-frozen and transported to the University of Aberdeen for complete parasitological examination. These sites are shown in Figure 1. Wet smears of gall bladder and liver tissue were examined under phase contrast microscopy at a magnification of × 325, using a Zeiss Photomicroscope II, for the presence of myxosporean parasites. Fish were then filleted and candled and the musculature examined for the presence of cysts. For descriptive purposes, spores were examined at magnifications of up to × 2000, using an agar monolayer and oil immersion microscopy. Air-dried smears were fixed in methyl alcohol and stained with Giemsa to enhance spore morphology. For electron microscopy, gall bladders containing spores of A. solomoni were fixed in glutaraldehyde, opened longitudinally, attached to a stub, sputter-coated with gold and examined using a Jeol 35CF scanning electron microscope. As far as is possible when dealing with frozen material, the guidelines of Lom and Arthur (1989) were adhered too. All measurements of dimensions are given in micrometres (µm), as ranges followed, in parentheses, by means plus or minus one standard deviation. Date and location of capture and prevalence refer to the site where the species was most com- Description: No pathological changes in host noted. Trophozoite stage not observed. Spores transparent, crescent shaped, anterior end convex, posterior end concave. Spore valves equally sized. Suture straight and distinct. Triangular body containing polar capsules, with alate projections extending laterally from the anterior of the body. Polar capsules spherical. Spore dimensions, based on observations of 50 spores: length 3.8–7.7 (5.80 ± 0.83), width 11.5–19.2 (14.40 ± 1.56). Spore length:width ratio, 1:1.67–3.50. Polar capsules equally sized, opening to anterior of spore. Polar capsule dimensions, based on 40 measurements, diameter 1.3–2.6 (1.87 ± 0.39). Host: Trachurus trachurus (L.). Site of infection: Gall bladder. Location of capture: West of Ireland, 52°52.8´N 12°03.6´W. Date: March 25, 2001. Prevalence: 6/25 (24%). Host length range: 19.6–43.1 cm. Alataspora solomoni Yurakhno, 1988 (Fig. 3) Order: Bivalvulida Shulman, 1959 Family: Alatasporidae Schulman et al., 1979 Genus: Alataspora Schulman et al., 1979 Description: No pathological changes noted. Vegetative stages not observed. Spores transparent, crescent shaped, ante- Fig. 2. Alataspora serenum. Scale bar = 10 µm Zdzisław Stanisła rior end convex, posterior end concave. Spore valves equally sized. Suture straight and distinct. Triangular body, with slight thickening around suture, containing polar capsules. Polar capsules opening onto different sides of anterior side of the spore. Spore dimensions (µm), based on observations of 50 spores: length 5.1–9.0 (6.7 ± 0.8), width 17.9–33.3 (24.3 ± 3.2). Spore length:width ratio, 1:2.67–1:5.75. Polar capsules spherical, inequal in size. Dimensions of polar capsules, based on 40 observations, diameter of larger polar capsule, 1.6–2.88 (2.24 ± 0.39), diameter of smaller polar capsule 0.96–2.56 (1.94 ± 0.4). Host: Trachurus trachurus (L.). Site of infection: Gall bladder. Location of capture: Ionian Sea 40°28´N 24°55´E. Date: June 7, 2001. Prevalence: 11/50 (22.0%). Host length range: 12.4–30.3 cm. Fig. 3. Alataspora solomoni. Scale bar = 5 µm Host: Trachurus trachurus (L.). Site of infection: Coelozoic in the gall bladder. Location of capture: South of Ireland, 48°45´N 09°29´W. Date: April 2001. Prevalence: 3/46 (6.5%). Host length range: 26.4–31.8 cm. Kudoa nova Naidenova, 1975 (Fig. 5) Order: Multivalvulida Shulman, 1959 Family: Kudoidae Meglitsch, 1960 Genus: Kudoa Meglitsch, 1947 Description: Pseudocysts macroscopic, white, spherical, up to 3 mm in diameter, found in clusters in the dorsal, ventral and lateral red muscles. Vegetative state of spore not observed. Spore subquadrate in apical view, with rounded valves. Sutural line thin and indistinct. Slightly elongate, ventrally flattened, in lateral view. Polar capsules large, pyriform. Spore dimensions based on 30 spores: length 5.1–7.7 (6.2 ± 0.9), width 5.1–7.7 (6.2 ± 0.9), polar capsule length 1.3–2.6 (1.8 ± 0.4). Host: Trachurus trachurus (L.). Site of infection: Dorsal, ventral and lateral red muscles. Location of capture: North African Coast, 19°58´N 17°28´W. Date: January 2001. Prevalence: 3/46 (6.5%). Host length range: 21.3–24.6 cm. Kudoa sp. (Fig. 4) Order: Multivalvulida Shulman, 1959 Family: Kudoidae Meglitsch, 1960 Genus: Kudoa Meglitsch, 1947 Description: No pathological changes noted. Vegetative stages not observed. Spore transparent, subquadrate in apical view, with rounded valves. Sutural line thin and indistinct. Anterior side of spore convex, posterior side “bell” shaped. Polar capsules large, pyriform. Spore dimensions, based on 17 measurements: length 5.1–7.7 (5.97 ± 0.79), spore width 5.3–7.3 (6.11 ± 0.53), polar capsule length 2.6–3.8 (3.10 ± 0.45). Fig. 5. Kudoa nova in lateral (left) and apical (right) views. Scale bar = 5 µm Myxobolus spinacurvatura Maeno, Sorimachi, Ogawa et Egusa, 1990 (Fig. 6) Order: Bivalvulida Shulman, 1959 Family: Myxobolidae Thélohan, 1892 Genus: Myxobolus Butschli, 1882 Fig. 4. Unidentified Kudoa sp. in apical (left) and lateral (right) views. Scale bar = 2 µm Description: No pathological changes noted in host. Spore oviform in lateral view, flattened perpendicular to sutural line. Two polar capsules, opening separately on the apical side of Roborzyński rosbśźćv Fig. 6. Myxobolus spinacurvatura viewed in lateral (left) and sutural (right) planes. Scale bar = 5 µm spore. Polar filaments coiled in three or four turns. Spore dimensions, based on measurements of 40 spores: length 8.9– 11.5 (9.97 ± 0.90), thickness 3.8–6.4 (4.92 ± 0.82), width 7.1–9.2 (8.56 ± 0.61). Length of polar capsule 3.8–5.1 (4.24 ± 0.60), width of polar capsule 2.3–2.9 (2.78 ± 0.26). This species has been comprehensively described by Maeno et al. (1990) and Bahri and Marques (1996). Host: Trachurus trachurus (L.). Site of infection: Liver. Location of capture: 40°28´N 24°55´E. Date: June 7, 2001. Prevalence: 1/43 (2.3%). Host length range: 23.8 cm. Discussion This is the first comprehensive study of the myxosporean parasites of T. trachurus throughout its geographic range, and reveals a diverse community. Several of the myxosporean species that infect horse mackerel show potential for use as biological tags. Alataspora serenum was originally described from T. trachurus in the Celtic Sea, infecting around 15% of fish (Gaevskaya and Kovaleva 1979a). It was recorded in the present study in three samples, one site to the north-west of Ireland, one in the Celtic Sea and one in the western English Channel, in both 2000 and 2001. Prevalence in all samples was around 15–20%. It was not recorded in samples from the adjacent waters of the North Sea and Bay of Biscay. This finding supports the current management system by suggesting that movements between the Western stock and the Southern and North Sea stocks are limited. A. serenum was also recorded from a single fish (1/46) from the Atlantic coast of North Africa (19°58´N 17°28´W), suggesting that some horse mackerel may be highly migratory. This species was not recorded fjad kadsććżć in the Mediterranean Sea. It could be useful as a tag, indicating mixing of stocks in the north east Atlantic. Alataspora solomoni was originally described infecting the Mediterranean horse mackerel, T. mediterraneus, from Quarantine Bay, near Sevastopol, in the Black Sea (Yurakhno 1988). This is the first record of A. solomoni infecting the Atlantic horse mackerel (T. trachurus), and the first record of infection outside the area of the Black Sea. A. solomoni was only recorded in fish from the samples taken in the eastern part of the Mediterranean Sea. It was not recorded in any of the samples taken in the western part of the Mediterranean, and suggests the existence of discrete stocks in this body of water. Over 450 species of Myxobolus have been described from fish hosts. Of these, the majority have been described from freshwater fishes (Landsberg and Lom 1991). This is the first record of Myxobolus spinacurvatura infecting T. trachurus, and the first record of this species from the European coast of the Mediterranean Sea. The rarity of this finding suggests that this was an accidental infection. This species does not provide any useful information on stock identity or distribution. An unidentified Myxobolus sp. has been previously recorded in the Adriatic Sea, infecting the livers of mullet, Mugil cephalus, in aquaculture facilities in Italy (Fioravanti et al. 2001). It is unknown whether this was the same species which has been recorded here, but the geographical proximity of this record to our finding and the fact that this species was first described from an infection of M. cephalus (Maeno et al. 1990) would suggest that this is possible. The genus Kudoa (Myxozoa, Myxosporea) comprises of species which are typically histozoic parasites (Moran et al. 1999). This is only the third report of a Kudoa species being found free in the gall bladder, the other two being Kudoa tachysurae Sarkar et Mazumder, 1983, reported from the three-spined catfish, Arius tenuispinis Day, 1877, and Kudoa haridasae Sarkar et Ghosh, 1991, from the gold-spot mullet, Liza parisa (Hamilton, 1822), both from the Bay of Bengal (Sarkar and Mazumder 1983, Sarkar and Ghosh 1991). This species was found infecting horse mackerel from one site to the south-west of Ireland only. Where present, infections were of a very low intensity, which made a detailed description or electron microscopy impossible. It is likely that this finding represents the discovery of a previously undescribed species. This species needs further study and description before any useful information can be extracted from its distribution. Despite examining samples from the Bay of Biscay, where it had previously been recorded (Gaevskaya and Kovaleva 1980), no fish infected with Kudoa quadratum were encountered. Kudoa nova has been reported from T. trachurus, T. mediterraneus and T. capensis, from the Atlantic Ocean, Black Sea and Mediterranean (Kovaleva et al. 1979). In this study, it was most frequently recorded in the sample collected from the African coast, at a prevalence of 6.5%. This agrees with the southerly distribution proposed by Kovaleva et al. (1979) who found infected T. trachurus only along the Morocco- falfęś addjfkęś Saharan coast. This species could prove a useful biological tag to study migration of fish from African waters into European seas. A single fish from a sample collected off the Galician coast was found to by infected with K. nova. Again, this supports the idea that some horse mackerel are highly migratory. Cruz et al. (2003) reported a number of horse mackerel from the coast of Portugal as being infected with a Kudoa sp. Further sampling should be carried out in this area to determine the identity of this species. If found to be K. nova, it would support the findings of Murta (2000), who proposed a degree of mixing between North African and European horse mackerel stocks. Alataspora serenum, A. solomoni and K. nova all show potential as useful biological tags for stock identification in different areas. A. serenum is most commonly found in fish from the Western stock, A. solomoni has only been found in fish from the eastern Mediterranean, and K. nova shows a more southerly distribution. It is a simple procedure to examine a fish for infections with these species, and examination of greater numbers or of further samples from areas that have not been covered in this study would be an simple and useful contribution to understanding the stock distribution of this species. Acknowledgements. I would like to thank Kevin MacKenzie of the Electron Microscopy Unit at the University of Aberdeen for preparing specimens. I am deeply grateful to Dr Ken MacKenzie for careful supervision of my work and helpful comments on an earlier draft of this work. I also am grateful to all partners in the HOMSIR project for the provision of samples and helpful discussions. I would like to offer my thanks to two anonymous reviewers for providing helpful comments on this paper. This work was funded by the EU under the 5th Framework programme (QLK5-CT1999-01438). References Abaunza P., Murta A., Campbell N., Cimmaruta R., ComesaZa S., Dahle G., Gallo E., García Santamaría M.T., Gordo L., Iversen S., MacKenzie K., Magoulas A., Mattiucci S., Molloy J., Nascetti G., Pinto A.L., Quinta R., Ramos P., Ruggi A., Sanjuan A., Santamaría M.T., Santos A.T., Stransky C., Zimmermman C. Considerations on sampling strategies for an holistic approach to stock identification: the example of the HOMSIR project. Fisheries Research, in press. Bahri S., Marques A. 1996. Myxosporean parasites of the genus Myxobolus from Mugil cephalus in Ichkeul lagoon, Tunisia: Description of two new species. Diseases of Aquatic Organisms, 27, 115–122. Cruz C., Vas A., Saraiva A. 2003. Occurrence of Kudoa sp. (Myxozoa) in Trachurus trachurus L. (Osteichthyes) in Portugal. Parasite, 10, 69–79. (Accepted April 20, 2005) Fioravanti M.L., Caffara M., Delgado M.L., Florio D., Marcer F., Quaglio F., Restani R. 2001. Survey on the diffusion of myxosporeans (Myxozoa) in marine fish farmed in Italy. Bollettino Societa Italiana di Patologia Ittica, 13, 12–25. Gaevskaya A.V., Kovaleva A.A. 1979a. New and rarely encountered forms of myxosporidia from fishes of the Celtic Sea. Parazitologiya, 13, 159–165 (In Russian). Gaevskaya A.V., Kovaleva A.A. 1979b. Two new species of myxosporidia from horse mackerel in the south eastern Atlantic. Biologiya Morya, 3, 80–83 (In Russian). Gaevskaya A.V., Kovaleva A.A. 1980. Eco-geographical characteristics of the parasite fauna of Atlantic Ocean scad. In: Investigations on the biological resources of the Atlantic Ocean. IBSS, Sevastopol (In Russian). ICES 1998. Working group on the assessment of mackerel, horse mackerel, sardine and anchovy. ICES CM 1998/ACFM, 6. Kovaleva A.A., Shulman S.S., Yakovlev V.N. 1979. Myxosporidia of the genus Kudoa (Myxosporidia: Multivalvulida) of the Atlantic Ocean basin. Trudy Zoologicheskogo Instituta AN SSSR, 87, 42–64 (In Russian). Landsberg J.H., Lom J. 1991. Taxonomy of the genera of the Myxobolus/Myxosoma group (Myxobolidae: Myxosporea), current listing of species and revision of synonyms. Systematic Parasitology, 18, 165–186. Lom J., Arthur J.R. 1989. A guideline for the preparation of species descriptions in the Myxosporea. Journal of Fish Diseases, 12, 151–156. MacKenzie K., Campbell N., Mattiucci S., Ramos P., Pereira A.L., Abaunza P. 2004. A checklist of the protozoan and metazoan parasites reported from the Atlantic horse mackerel, Trachurus trachurus L. Bulletin of the European Association of Fish Pathologists, 24, 180–184. Maeno Y., Sorimachi M., Ogawa K., Egusa S. 1990. Myxobolus spinacurvatura sp. n. (Myxosporea: Bilvalvulida) parasitic in deformed mullet, Mugil cephalus. Fish Pathology, 25, 37–41. Moran J.D.W., Whitaker D.J., Kent M.L. 1999. A review of the myxosporean genus Kudoa Meglitsch, 1947, and its impact on the international aquaculture industry and commercial fisheries. Aquaculture, 172, 163–196. Murta A. 2000. Morphological variation of horse mackerel (Trachurus trachurus) in the Iberian and North African Atlantic: implications for stock identification. ICES Journal of Marine Science, 57, 1240–1248. Sarkar N., Ghosh S. 1991. Two new coelozoic Myxosporidia (Myxosporea: Myxosporea) from estuarine teleost fishes (Mugilidae) of West Bengal, India. Proceedings of the Zoological Society of Calcutta, 44, 131–135. Sarkar N., Mazumder S.K. 1983. Studies on myxosporean parasites (Myxozoa: Myxosporea) from marine fishes in West Bengal, India: I. Description of three new species from Tachysurus spp. Archiv für Protistenkunde, 127, 59–63. Smith-Vaniz W.F. 1986. The Carangidae. In: Fishes of the NorthEastern Atlantic and the Mediterranean. Vol. II (Eds. P.J.P. Whitehead, M.-L. Bauchot, J.-C. Hureau, J. Nielsen and E. Tortonese). UNESCO, Paris, 815–844. Yurakhno V.M. 1988. On new myxosporidia of Black Sea fishes. Parazitologiya, 22, 521–524 (In Russian). Appendix III – Horse Mackerel Discriminant analysis “R” Code # # # # # # # # # Discriminant Analysis sum(homsir$Area==2|homsir$Area==3) 133 sum(homsir$Area==5) 50 sum(homsir$Area==11) 50 Others = 252 homsir<-read.table("C:/northwest.txt", header=T) homsir.2001<-read.table("C:/northwest01.txt", header=T) ## reads in data files ANIS HYST PROD TERG ECTE DERO PSEUD GASTRO <- c(homsir$Anis, homsir.2001$Anis) <- c(homsir$Hyst_L, homsir.2001$Hyst_L) <- c(homsir$P_pol, homsir.2001$P_pol) <- c(homsir$T_lat, homsir.2001$T_lat) <- c(homsir$E_lep, homsir.2001$E_lep) <- c(homsir$D_var, homsir.2001$D_var) <- c(homsir$P_trach, homsir.2001$P_trach) <- c(homsir$G_trach, homsir.2001$G_trach) ## extracts the columns of specific parasites from both years rANIS<-log(ANIS+(abs(rnorm(336,0,0.1)))) rHYST<-log(HYST+(abs(rnorm(336,0,0.1)))) rPROD<-log(PROD+(abs(rnorm(336,0,0.1)))) rTERG<-log(TERG+(abs(rnorm(336,0,0.1)))) rECTE<-log(ECTE+(abs(rnorm(336,0,0.1)))) rDERO<-log(DERO+(abs(rnorm(336,0,0.1)))) rPSEUD<-log(PSEUD+(abs(rnorm(336,0,0.1)))) rGASTRO<-log(GASTRO+(abs(rnorm(336,0,0.1)))) ## log transforms and adds small random varaible to parasite nos stocks <- data.frame(rANIS, rHYST, rPROD, rTERG, rECTE, rDERO, rPSEUD, rGASTRO) ## combines these into a new matrix comb.areas <- c(homsir$Area, homsir.2001$Area) comb.areas[comb.areas==2|comb.areas==3]<-"W" comb.areas[comb.areas==5]<-"N" areas<-as.factor(comb.areas) areas[areas=="W-00"|areas=="W-01"] <- "W" areas[areas=="NS-00"|areas=="NS-01"] <- "N" ## specifies if a fish is from a western or north sea sample class.matrix <- matrix(nrow=1000, ncol=2) for( a in (1:1000)){ rows <- sample(c(1:336), 168) temp.stocks test.stocks temp.areas test.areas <<<<- stocks[rows,] stocks[-rows,] areas[rows] areas[-rows] z <- lda(temp.stocks, temp.areas) results <- predict(z, test.stocks) classification.success <- vector(length=2) no.classified <- vector(length=2) for (i in (1:2)){ classification.success[i] <sum(results$class[test.areas==levels(results$class)[i]]==test.areas[test.are as==levels(results$class)[i]]) no.classified[i] <- length(test.areas[test.areas==levels(test.areas)[i]]) } class.matrix[a,] <- classification.success/no.classified } ## carries out discriminant analysis 1000 times and stores the results of ## successful classification in a vector for further analysis Appendix IV – Horse mackerel neural network analysis “R” code. library(nnet) # Loads the neural network library HOMSIR<-read.table("C://HOMSIR/alt01.txt", header=T) # reads data file with sample area, biological information # and parasite abundance for each fish attach(HOMSIR) plot(Anisakis~Length) # plots the relationship between no. of Anisakis sp. # infecting each fish vs. length lg.an<-log(Anisakis+1) # performs a ln(n+1) transformation on the Anisakis abundance # value to produce a linear relationship plot(lg.an~Length) # plots Log(Anisakis+1) against length tmp<-lm(lg.an~Length) abline(tmp) rs.lg.an<-residuals(tmp) # extracts residuals of Log(anisakis+1)~length for later use par(mfrow=c(2,2), mar=c(4.2,4.2,1,0.5)+.1,mex=0.8,cex.axis=0.5, cex.main=0.8) plot(Length~Stock, ylab="Length (mm)") mtext("A", side=3, adj=0.02, cex=0.8, padj=1.5) plot(Anisakis~Length, xlab="Length(mm)", ylab="Number of Anisakis") mtext("B", side=3, adj=0.02, padj=1.5, cex=0.8) plot(lg.an~Length, xlab="Length (mm)", ylab="Log(Anisakis +1)") mtext("C", side=3, adj=0.02, padj=1.5, cex=0.8) abline(tmp) plot(rs.lg.an~Stock, ylab="Residual") mtext("D", side=3, adj=0.02, padj=1.5, cex=0.8) parasites<data.frame(as.factor(A.serenum),as.factor(Goussia),rs.lg.an,Hystero.L,H ystero.A, Tergestia,Derogenes,Ectenurus,Pseudaxine,Gastrocotyle,Heteraxinoides) # assembles data frame of parasite data samp <-(c(sample(1:34,25), sample(35:131,25), sample(132:181,25),sample(182:281,25),sample(282:333,25),sample(334:433 ,25),sample(434:483,25))) # creates vector of numbers relating to half of the fish from each area sampled area<-class.ind(HOMSIR$Stock) # creates a class.ind object containing the stock to which each fish belongs para1 <- nnet(parasites[samp,], area[samp,], size = 8, rang = 0.1, decay =5e-4, maxit = 1000) # creates a neural networh with 8 hidden layer objects, trained on the fish # specified in the samp vector test.cl <- function(true, pred){ Sample <- max.col(true) Assigned <- max.col(pred) table(Sample, Assigned) } test.cl(area[samp,], predict(para1, parasites[samp,])) # a function which uses the trained neural network to classify the remaining # of fish to the relevant stock detach(HOMSIR) # we don't need this any longer horse.mack<-function(repetitions, layers){ # creates a carries out a neural network analysis over as many # repetitions as specified one.as.one<-vector(length=repetitions) one.as.two<-vector(length=repetitions) one.as.three<-vector(length=repetitions) one.as.four<-vector(length=repetitions) two.as.one<-vector(length=repetitions) two.as.two<-vector(length=repetitions) two.as.three<-vector(length=repetitions) two.as.four<-vector(length=repetitions) three.as.one<-vector(length=repetitions) three.as.two<-vector(length=repetitions) three.as.three<-vector(length=repetitions) three.as.four<-vector(length=repetitions) four.as.one<-vector(length=repetitions) four.as.two<-vector(length=repetitions) four.as.three<-vector(length=repetitions) four.as.four<-vector(length=repetitions) # creates vectors into which percentage classifications are stored for (i in (1:repetitions)){ # loops the operation for the specified number of repeats samp <-(c(sample(1:131,66), sample(132:181,25), sample(182:433,125),sample(434:483,25))) area<-class.ind(HOMSIR$Stock) para1 <- nnet(parasites[samp,], area[samp,], size = layers, rang = 0.1, decay = 5e-4, maxit = 2000) test.cl <- function(true, pred){ Sample <- max.col(true) Assigned <- max.col(pred) table(Sample, Assigned) } temp.table<-test.cl(area[-samp,], predict(para1, parasites[samp,])) # carries out neural network classifications on random half of horse mackerel data # using the specified number of layers temp.matrix<-matrix(ncol=4, nrow=4, 0) ncols<-as.numeric(colnames(temp.table)) nrows<-as.numeric(rownames(temp.table)) for (j in 1:length(nrows)){ for (k in 1:length(ncols)){ temp.matrix[ncols[j],nrows[k]]<-temp.table[j,k] } } # Sometimes the classification fails to return four categories. This works out # when that occurs and deals with the problem... one.as.one[i]<-temp.matrix[1,1] one.as.two[i]<-temp.matrix[1,2] one.as.three[i]<-temp.matrix[1,3] one.as.four[i]<-temp.matrix[1,4] two.as.one[i]<-temp.matrix[2,1] two.as.two[i]<-temp.matrix[2,2] two.as.three[i]<-temp.matrix[2,3] two.as.four[i]<-temp.matrix[2,4] three.as.one[i]<-temp.matrix[3,1] three.as.two[i]<-temp.matrix[3,2] three.as.three[i]<-temp.matrix[3,3] three.as.four[i]<-temp.matrix[3,4] four.as.one[i]<-temp.matrix[4,1] four.as.two[i]<-temp.matrix[4,2] four.as.three[i]<-temp.matrix[4,3] four.as.four[i]<-temp.matrix[4,4] } sort.1.1<-sort(one.as.one) sort.1.2<-sort(one.as.two) sort.1.3<-sort(one.as.three) sort.1.4<-sort(one.as.four) sort.2.1<-sort(two.as.one) sort.2.2<-sort(two.as.two) sort.2.3<-sort(two.as.three) sort.2.4<-sort(two.as.four) sort.3.1<-sort(three.as.one) sort.3.2<-sort(three.as.two) sort.3.3<-sort(three.as.three) sort.3.4<-sort(three.as.four) sort.4.1<-sort(four.as.one) sort.4.2<-sort(four.as.two) sort.4.3<-sort(four.as.three) sort.4.4<-sort(four.as.four) # sorts the classifications scores perc.1.1<-((sort.1.1/25)*100) perc.1.2<-((sort.1.2/25)*100) perc.1.3<-((sort.1.3/25)*100) perc.1.4<-((sort.1.4/25)*100) perc.2.1<-((sort.2.1/25)*100) perc.2.2<-((sort.2.2/25)*100) perc.2.3<-((sort.2.3/25)*100) perc.2.4<-((sort.2.4/25)*100) perc.3.1<-((sort.3.1/127)*100) perc.3.2<-((sort.3.2/127)*100) perc.3.3<-((sort.3.3/127)*100) perc.3.4<-((sort.3.4/127)*100) perc.4.1<-((sort.4.1/65)*100) perc.4.2<-((sort.4.2/65)*100) perc.4.3<-((sort.4.3/65)*100) perc.4.4<-((sort.4.4/65)*100) # transforms the classification score into a percentage cols<-c("African", "North Sea", "Southern", "Western") rnames<-c("5%", "50%", "95%") nsea<-matrix(nrow=3, ncol=4) colnames(nsea)<-cols rownames(nsea)<-rnames west<-matrix(nrow=3, ncol=4) colnames(west)<-cols rownames(west)<-rnames south<-matrix(nrow=3, ncol=4) colnames(south)<-cols rownames(south)<-rnames masah<-matrix(nrow=3, ncol=4) colnames(masah)<-cols rownames(masah)<-rnames masah[,1]<-c(perc.1.1[round(0.05*repetitions)], perc.1.1[round(0.5*repetitions)], perc.1.1[round(0.95*repetitions)]) masah[,2]<-c(perc.2.1[round(0.05*repetitions)], perc.2.1[round(0.5*repetitions)], perc.2.1[round(0.95*repetitions)]) masah[,3]<-c(perc.3.1[round(0.05*repetitions)], perc.3.1[round(0.5*repetitions)], perc.3.1[round(0.95*repetitions)]) masah[,4]<-c(perc.4.1[round(0.05*repetitions)], perc.4.1[round(0.5*repetitions)], perc.4.1[round(0.95*repetitions)]) nsea[,1]<-c(perc.1.2[round(0.05*repetitions)], perc.1.2[round(0.5*repetitions)], perc.1.2[round(0.95*repetitions)]) nsea[,2]<-c(perc.2.2[round(0.05*repetitions)], perc.2.2[round(0.5*repetitions)], perc.2.2[round(0.95*repetitions)]) nsea[,3]<-c(perc.3.2[round(0.05*repetitions)], perc.3.2[round(0.5*repetitions)], perc.3.2[round(0.95*repetitions)]) nsea[,4]<-c(perc.4.2[round(0.05*repetitions)], perc.4.2[round(0.5*repetitions)], perc.4.2[round(0.95*repetitions)]) south[,1]<-c(perc.1.3[round(0.05*repetitions)], perc.1.3[round(0.5*repetitions)], perc.1.3[round(0.95*repetitions)]) south[,2]<-c(perc.2.3[round(0.05*repetitions)], perc.2.3[round(0.5*repetitions)], perc.2.3[round(0.95*repetitions)]) south[,3]<-c(perc.3.3[round(0.05*repetitions)], perc.3.3[round(0.5*repetitions)], perc.3.3[round(0.95*repetitions)]) south[,4]<-c(perc.4.3[round(0.05*repetitions)], perc.4.3[round(0.5*repetitions)], perc.4.3[round(0.95*repetitions)]) west[,1]<-c(perc.1.4[round(0.05*repetitions)], perc.1.4[round(0.5*repetitions)], perc.1.4[round(0.95*repetitions)]) west[,2]<-c(perc.2.4[round(0.05*repetitions)], perc.2.4[round(0.5*repetitions)], perc.2.4[round(0.95*repetitions)]) west[,3]<-c(perc.3.4[round(0.05*repetitions)], perc.3.4[round(0.5*repetitions)], perc.3.4[round(0.95*repetitions)]) west[,4]<-c(perc.4.4[round(0.05*repetitions)], perc.4.4[round(0.5*repetitions)], perc.4.4[round(0.95*repetitions)]) # sticks the percentages into a matrix labs<-c("African","North Sea","Southern","Western") par(mfrow=c(2,2), mar=c(4.2,4.2,1,0.5)+.1,mex=0.8,cex.axis=0.5, cex.main=0.8) boxplot(nsea[,1],nsea[,2],nsea[,3],nsea[,4], range=0, boxlty=0, medpch=0, whisklty=1,medlty="blank", names=labs, ylab="%", staplewex=0.1, ylim=c(0,100)) mtext("A", side=3, adj=0.02, cex=0.8, padj=1.5) boxplot(masah[,1],masah[,2],masah[,3],masah[,4], range=0, boxlty=0, medpch=0,whisklty=1,medlty="blank", names=labs, ylab="%",staplewex=0.1, ylim=c(0,100)) mtext("B", side=3, adj=0.02, padj=1.5, cex=0.8) boxplot(south[,1],south[,2],south[,3],south[,4], range=0, boxlty=0, medpch=0,whisklty=1,medlty="blank", names=labs, ylab="%",staplewex=0.1, ylim=c(0,100)) mtext("C", side=3, adj=0.02, padj=1.5, cex=0.8) boxplot(west[,1],west[,2],west[,3],west[,4], range=0, boxlty=0, medpch=0,whisklty=1,medlty="blank", names=labs, ylab="%",staplewex=0.1, ylim=c(0,100)) mtext("D", side=3, adj=0.02, padj=1.5, cex=0.8) # plots the results product<-list("North_Sea"=nsea, "Southern"=south, "Western"=west, "African"=masah) return(product) } temp<-horse.mack(100,8) temp Appendix V - Spatial and temporal variations in parasite prevalence and infracommunity structure in herring (Clupea harengus L.) caught to the west of the British Isles and in the North and Baltic Seas: implications for fisheries science 1 Journal of Helminthology (2007) 81, 137–146 DOI: 10.1017/S0022149X07747454 Spatial and temporal variations in parasite prevalence and infracommunity structure in herring (Clupea harengus L.) caught to the west of the British Isles and in the North and Baltic Seas: implications for fisheries science Neil Campbell1,3*, Marcus A. Cross2†, James C. Chubb2, Carey O. Cunningham3, Emma M. C. Hatfield3 and Ken MacKenzie1 1 School of Biological Sciences (Zoology), The University of Aberdeen, Aberdeen AB24 2TZ, UK: 2School of Biological Sciences, The University of Liverpool, Crown Street, Liverpool, L69 7BZ, UK: 3FRS Marine Laboratory, PO Box 101, 375 Victoria Road, Aberdeen AB11 9DB, UK Abstract Herring Clupea harengus L. viscera were examined for endoparasitic infections as part of a multidisciplinary stock identification project (WESTHER, EU Contract no. Q5RS-2002-01 056) which applied a range of stock discrimination techniques to the same individual fishes to obtain comparable results for multivariate analysis. Spawning and non-spawning adults, and juvenile herring were caught, over 3 years, by commercial and research vessels from numerous locations to the west of the UK and Ireland, along with control samples of spawning fish from the eastern Baltic Sea, and juveniles from sites in the eastern and western North Sea, and the north of Norway. The metacercariae of two renicolid digeneans (Cercaria pythionike and Cercaria doricha), one larval nematode (Anisakis simplex s.s.) and one larval cestode (Lacistorhynchus tenuis) were selected as tag species. Results were compared with those from herring collected between 1973 and 1982, which suggested remarkable stability in the parasite fauna of herring in the study area. These species were used to compare the parasite infracommunities of spawning herring. A significant variation in infracommunity structure was observed between different spawning grounds. These results suggest that the parasite fauna of herring are spatially variable but remain temporally stable in both the short and long term. Significant differences in prevalence and abundance of infections and comparisons of parasite infracommunity enabled the separation of putative herring stocks west of the British Isles. Distinctive patterns of parasite infection in two different spawning groups off the north coast of Scotland suggest that this area is occupied by two spawning populations, one recruiting from the west of Scotland, the other from outside this area, and most likely from the eastern North Sea. The distribution patterns of L. tenuis, C. doricha and C. pythionike suggest the potential for fish that spawn in three distinct International Council for the Exploration of the Seas *E-mail: N.Campbell@marlab.ac.uk †Present address: School of Life Sciences, University of Dundee, Dow Street, Dundee DD1 5EH, UK 138 N. Campbell et al. (ICES) management units to be present in mixed aggregations found over the Malin Shelf, with significant implications for management in this area. Introduction The identification of discrete fish stocks for the development and appropriate management of commercial fisheries is vitally important (e.g. Waldman, 2005). The application of management measures that assume erroneous population boundaries can lead to depletion and extinction of local sub-stocks (Butterworth & Penney, 2004). The present investigation was part of the WESTHER project that had four major aims: (1) to identify discrete spawning populations; (2) to investigate the composition of non-spawning aggregations; (3) to link recruits on nursery grounds to spawning adult populations; and (4) to deliver advice on these findings to fisheries managers. The null hypotheses of these investigations were that herring to the west of the British Isles live as a single, well-mixed population which does not show significant regional variations in its parasitic faunal composition, prevalence and mean intensity; that herring from particular spawning grounds remain in discrete groups on their feeding grounds; and that parasites do not link juveniles on nursery grounds to spawning adult populations. Our investigations of the spatial and temporal variations in parasite prevalence and infracommunity of spawning herring Clupea harengus L. caught to the west of the British Isles and in the Baltic Sea are reported here. Parasites are widely used as indicators of stock boundaries, recruitment, migration and mixing of commercially important species of marine, anadromous and freshwater fish. Since their first practical application (Herrington et al., 1939), numerous reviews and guidelines for their use in this field have been published (see, for instance, MacKenzie, 2002) Although infection statistics of a single parasite species can be enough to determine the origin of fish and to identify stocks, to build up a more detailed picture of host population structure it is preferable to use either combined infection data from several parasite species (MacKenzie, 1985; MacKenzie & Longshaw, 1995; Larsen et al., 1997) or to analyse entire parasite assemblages (e.g. Lester et al., 2001; Moore et al., 2003). A considerable amount of research has been carried out on the complex of commercially important Atlantic herring Clupea harengus L. stocks to the west of Great Britain and around Ireland. Despite this intensive research, levels of mixing, recruitment patterns and genetic interactions within the complex are still poorly understood (Molloy, 2006). Parrish & Saville (1965) considered the herring stocks to the west of the British Isles as consisting of two main components, the ‘oceanic’ and ‘shelf’ populations. Since then, two opposing theories of Atlantic herring population structure have been developed: the discrete population concept (Iles & Sinclair, 1982) and the dynamic balance concept (Smith & Jamieson, 1986). Current opinion is that neither of these concepts adequately explains herring population structure and dynamics, and that they are better described using the metapopulation concept (McQuinn, 1997), where population structure in a given area can be considered as a complex of local subpopulations linked by variable degrees of gene flow. Herring to the west of the British Isles are currently managed on the basis of five different management areas; the west of Scotland [ICES Division VIa (North)]; the Firth of Clyde; the area to the north and west of Ireland [ICES Divisions VIa (South) and VIIb,c]; the Irish Sea to the north of 52830’N [ICES Division VIIa (North)], and finally, the Celtic Sea and Division VIIj, to the south of Ireland (fig. 1). Methods To study the stock identity of herring, a sampling scheme appropriate to herring life-history needs to be devised. First, samples of herring at different life stages were collected: juveniles, spawning adults on their respective spawning grounds, and putative mixtures of non-spawning adults from known feeding grounds. In order to determine the scale of differences within and between the potential metapopulation being sampled, it was necessary to collect samples from populations that are known not to mix with populations to the west of the British Isles. Consequently, samples of spawning herring were collected in the Baltic Sea, near Rügen, and juveniles from the northern coast of Norway. To investigate possible mixing of recruits from the North Sea with west-coast spawners, juveniles were collected from the east and west coasts of the North Sea. Samples were collected and processed at sea onboard research vessels, or were caught by commercial vessels and processed that day after the catch was landed. Approximately 100 spawning, juvenile and mixed stock aggregations of fish were subjected to parasitological examination. Samples were collected at a variety of times and locations, between late 2002 and early 2005, the details and realized sample sizes of which are presented in table 1. The viscera were removed and preserved individually in vials of 80 – 96% ethanol for parasitological study. Due to the constraints of sample collection, ectoparasitic species were not recovered. Visceral examinations were carried out at Aberdeen and Liverpool universities, which helped increase the number of samples processed and standardize the methodology across institutes. In the laboratory, the visceral organs of herring were removed from their vials and placed in tap water for several minutes to allow the tissue to soften, aiding examination. Viscera were then examined for helminth macroparasites under a dissecting microscope at magnifications of from £ 60 to £ 120. Numbers of pyloric caeca were counted and recorded, and smears of liver, spleen and gonad were examined for microparasites (myxozoans, protozoans, fungi) under a research microscope at Helminth tags informing fisheries science 139 Fig. 1. Location of sampling sites. Spawning samples: A, Cape Wrath; B, Skye; C, Clyde; D, Irish Sea; E, Celtic Sea; F, Dingle; G, Rosamhil; H, Donegal. Juvenile samples: 1, Hebrides; 2, Scottish sea lochs; 3, Stanton Bank; 4, Western Irish Sea; 5, Eastern Irish Sea; 6, Celtic Sea. Mixed stock aggregations: diamonds, Management Unit VIa (North); circles, Management Unit VIa (South), VIIb,c; triangles, Irish Sea [Management Unit VIIa (North)]; squares, Celtic Sea and ICES subdivision VIIj. £ 300 to £ 400. All parasites present were identified to species level, recorded, and all macroparasites counted. Samples of larval Anisakis sp. nematodes and renicolid metacercariae were preserved in 96% ethanol for development of molecular markers. Morphological identification of parasites was carried out by referring to the scientific literature, including descriptions of parasites previously reported from herring. Semi-permanent mounts of representative specimens of each helminth parasite species found, with the exception of Anisakis sp. nematode larvae, were stained with acetocarmine and mounted in DePeX. All specimens of Anisakis spp. larvae were preserved in ethanol and retained for investigation of molecular genetics, together with several hundred specimens of each species of renicolid metacercariae (see Cross et al., 2007). Prevalences were compared with chisquared tests, parasite infracommunities were compared using Mann– Whitney U-tests. Results A total of 4073 herring viscera samples were examined for parasitic infections. This comprised 1649 juvenile herring, 1315 spawning herring, 715 mixed adults and 394 outliers (table 1). In total, 14 species of parasite were recorded in this study. Details of species and infection sites are presented in table 2. To confirm that renicolid infections remain constant for the life of the fish, abundances of Cercaria pythionike and C. doricha were log transformed, to normalize the data. Linear regression revealed no significant relationship between age and abundance of either C. pythionike (r 2 ¼ 0.0021, P ¼ 0.55) or C. doricha (r 2 ¼ 0.016, P ¼ 0.09). This indicates that further infection with these parasites does not occur in adult herring, that heavily infected individuals are not subject to increased mortality, and that infections are therefore strongly indicative of the nursery ground of origin. Long-term temporal variation Two of the areas sampled in MacKenzie’s (1985) study, Stanton Bank, to the south of the Outer Hebrides, and a number of Scottish west coast sea-lochs, were sampled again during the present study, so it was possible to compare prevalences of infection of C. doricha and 140 N. Campbell et al. Table 1. Location of capture and sample size of spawning, nonspawning and juvenile herring samples examined. the Scottish sea lochs sampled (P ¼ 0.002 for C. doricha, P ¼ 0.039 for C. pythionike), but the prevalences were still remarkably similar (fig. 3). Number of herring examined Area sampled Spawners Celtic Sea (spring spawners) Celtic Sea (autumn spawners) Dingle Rosamhil Donegal (spring spawners) Donegal (autumn spawners) Clyde Irish Sea Skye Cape Wrath Juveniles Celtic Sea Eastern Irish Sea (spring) Eastern Irish Sea (autumn) Western Irish Sea (spring) Western Irish Sea (autumn) Scottish sea lochs (spring) Scottish sea lochs (autumn) Minch Stanton Bank Mixed Stock Aggregations IVa (South)/ VIIb–c IVa (North) Irish Sea Celtic Sea Outliers Baltic (spawners) Eastern North Sea (juveniles) Moray Firth (juveniles) North Norway (juveniles) Grand total 2002 2003 2004 2005 Total – – – – – – – – – – – 104 – – – 103 43 104 – 100 95 – 92 97 94 – – 138 124 99 – – – – 85 – – – 37 – 95 104 92 97 179 103 43 242 161 199 – – – – – – 100 – – – – 70 – 87 100 99 83 108 97 – 86 121 120 103 101 49 112 – 104 – 109 – – – – – 97 104 156 230 207 203 300 132 220 – – – – – 110 99 – 116 98 125 69 – – 98 – 116 208 322 69 – – – – – 104 – – 24 – 102 50 49 21 – – – 44 – – 206 50 93 45 4073 C. pythionike in samples taken from the same areas about 30 years apart (fig. 2). There were no significant differences between years in prevalence of either C. doricha (P ¼ 0.059) or C. pythionike (P ¼ 0.282) at Stanton Bank. There were significant differences between years in Short-term temporal variation There were no significant differences in prevalence or abundance of any of the four tag species between autumn spawners collected off Donegal in October 2003, and spring spawners collected in the same area in February 2004, therefore we are confident that these samples were taken from a single population spawning in this area over an extended period and can be combined into a single sample. When combined with spring spawning herring collected in 2005 from a similar location, no significant differences were found between the mean intensities of any of the four species at any site across the two periods (table 3). The samples collected at Donegal showed very similar prevalences between years, only varying in the prevalence of C. pythionike. There were no significant differences in prevalence between samples collected in the two different sampling periods in the Irish Sea or at Cape Wrath, but there were significant differences in prevalence of the three tag species present in the samples collected off Skye (table 3). The similarity between the samples collected at Cape Wrath and those in the Baltic Sea is interesting, but without samples from intermediate locations, should not be seen to imply that these are samples taken from the same stock. Spatial variability of infracommunities Spawning herring were pooled over different sampling periods and, where appropriate, all samples collected within an ICES management area were combined. A comparison of herring parasite infracommunities, grouped by management area, was performed (table 4). Visual inspection of the parasite community data highlights a number of obvious groupings. Herring from north-western Ireland [ICES Divisions VIa (South) and VIIb,c] and spring spawning fish from Skye [ICES Division VIa (North)] appear to have very similar parasite Table 2. Parasite species recorded from herring in this study. New host records are highlighted with an asterisk. Parasite group Parasite species Site of infection Protozoa Myxosporea Goussia clupearum Ceratomyxa auerbachi Myxobolus sp.* Cercaria pythionike Cercaria doricha Hemiurus luehei Derogenes varicus Brachyphallus crenatus Lecithaster confusus Pronoprymna ventricosa* Anisakis simplex sensu stricto Hysterothylacium aduncum Lacistorhynchus tenuis Echinorhynchus gadi Liver Gall bladder Gall bladder Visceral cavity Visceral cavity Stomach Stomach Stomach Intestine Pyloric caeca Visceral cavity Visceral cavity Visceral cavity, stomach and intestine Intestine Digenea Nematoda Cestoda Acanthocephala Helminth tags informing fisheries science 141 Fig. 2. Comparison of prevalences of Cercaria doricha (black) and Cercaria pythionike (white) between herring collected in Scottish sea lochs (A) and at Stanton Bank (B) between 1973 and 1982 (MacKenzie, 1985) and those collected in the present study. infracommunities. The two samples collected at different times of year in VIa (North), i.e. Skye, and Cape Wrath, have very different parasite infracommunities. Samples of Celtic Sea and Division VIIj herring are not selfconsistent, suggesting this is home to some degree of stock mixing. While spawning herring from the northwest of Ireland [VIa (South) and VIIb,c] and Skye [VIa (North)] appear to be similar, with the other areas appearing to be more different from each other, Mann– Whitney U-test results between the different areas dilute these differences somewhat. The sample taken at Cape Wrath remained significantly different to all other samples. Other differences are less significant or produce confusing patterns, suggesting that the rather low number of species recorded may make this a rather blunt tool for identifying these populations. Links between juveniles and adults Anisakis simplex infections are not a useful tag species to link adults to juveniles on nursery grounds, due to the cumulative nature of their mode of infection. The prevalence of this species varied greatly between inshore and offshore samples collected to the west of Scotland, Fig. 3. Prevalences of Anisakis simplex in juvenile herring samples at different sites in ICES Division VIa (North): (A) Scottish sea lochs; (B) Hebrides; (C) Stanton Bank. 142 N. Campbell et al. Table 3. Significance of temporal variability in prevalences (Pr) of selected tag parasites between spawning herring collected at the same site in multiple years. A. simplex C. pythionike C. doricha L. tenuis Period Sample code Pr (%) P Pr (%) P Pr (%) P Pr (%) P 0 Donegal 69.0 71.8 94.0 98.0 64.5 91.9 81.7 60.2 100 96.2 .0.05 51.3 36.5 0 0 58.1 21.6 0 0 0 0 ,0.05 20.3 14.1 0 0 25.0 5.4 0 0 0 0 .0.05 0 0 0 0 0 0 8.7 4.3 0 0 NA 0 03– 04 2005 2003 2004 2004 2005 2003 2004 2003 2004 Cape Wrath Skye Irish Sea Baltic .0.05 ,0.001 .0.05 .0.05 with prevalences of over 40% in juveniles collected from Stanton Bank and offshore of the Hebrides, but between 0% and 7% in Scottish sea loch juvenile samples (fig. 3). This illustrates the differences in parasite faunas between juveniles in adjacent nursery areas. The prevalences of the three remaining tag species were not significantly different between spawning samples collected to the north-west of Ireland [ICES Divisions VIa (South) and VIIb,c] and juveniles collected at Stanton Bank (table 5). Samples collected in other management areas show significant differences in prevalence between juveniles and adults, although some of these are rather surprising. Links between spawning and non-spawning populations Comparisons were made between the prevalences of tag parasites in spawning samples and non-spawning mixed aggregations collected in the same management area in the same year (table 6). Interestingly, fish infected with C. doricha and C. pythionike were found in mixed NA ,0.001 NA NA NA ,0.01 NA NA NA NA .0.05 NA stock aggregations in the Irish Sea, but were absent from the Irish Sea spawning population, showing that herring from different spawning populations do not remain in distinct groups on feeding grounds. Discussion Pelagic marine fish species have been the subjects of many biological tag studies (Williams et al., 1992; MacKenzie, 2002), a number of which have focused on the clupeids – the Atlantic herring Clupea harengus, and the Pacific herring C. pallasi (see MacKenzie, 1985; Moser & Hsieh, 1992). Anisakid nematode larvae have been used most frequently, but other tag groups include larval and adult digeneans, cestode plerocercoids, monogeneans and myxosporeans. MacKenzie (1985) used two species of renicolid metacercariae, Cercaria doricha and C. pythionike, and a cestode plerocercoid, Lacistorhynchus tenuis, in his study of herring recruitment around the Scottish coasts. The value of the metacercariae as tags derived from three features of the host– parasite Table 4. Parasite infracommunity percentages observed in spawning samples. Area sampled Infracommunity Uninfected C. pythionike C. doricha A. simplex L. tenuis C. pythionike and C. doricha C. pythionike and A. simplex C. pythionike and L. tenuis C. pythionike and C. doricha and A. simplex C. pythionike and C. doricha and L. tenuis C. pythionike and C. doricha and A. simplex and L. tenuis C. pythionike and A. simplex and L. tenuis C. doricha and A. simplex C. doricha and L. tenuis C. doricha and A. simplex and L. tenuis A. simplex and L. tenuis Celtic Sea (south-west Irish Sea) Dingle 13.93 8.20 North-west Ireland Irish Sea Skye 11.04 7.36 27.78 13.70 8.66 10.24 1.29 76.71 0.68 41.73 98.71 35.25 81.82 46.82 66.67 2.46 36.07 9.09 3.01 19.06 5.56 3.28 4.55 11.71 14.17 4.55 1.00 0.79 0.82 Cape Wrath Clyde 5.51 18.90 8.90 143 Helminth tags informing fisheries science Table 5. Differences between prevalences (Pr) of tag parasites in spawning (S) and juvenile (J) samples from each ICES functional unit. C. pythionike ICES management area VIa (South) VIIb,c VIa (North) Irish Sea Celtic Sea and Division VIIj C. doricha L. tenuis Sample Pr (%) P Pr (%) P Pr (%) P Stanton Bank (J) NW Ireland (S) Sea lochs/Hebrides (J) Skye (S) W & E Irish Sea (J) Irish Sea (S) Celtic Sea (J) Celtic Sea (spring and autumn) (S) 41.8 42.7 76.8 25.8 22.4 0 1.0 34.0 .0.05 19.5 16.9 54.3 9.4 5.9 0 0 6.9 .0.05 0 0 0 0 0.6 6.2 0 0.3 NA relationship: (1) herring were only susceptible to infection in their first year of life and no further infection occurs thereafter; (2) the parasites had life spans in herring extending to several years and possibly as long as that of the host itself; and (3) levels of infection varied significantly between different nursery areas. This meant that herring were effectively tagged for life as juveniles and these parasites could be used to trace samples of adult herring to their nursery grounds. The plerocercoid of Lacistorhynchus tenuis also infected herring as juveniles, but its value as a tag was constrained by the fact that further infection of herring as adults was possible. The study area of MacKenzie (1985) overlapped that of the present study, and it was therefore anticipated that these three parasite species would prove to be useful tags once again. Long-term stability of herring parasite fauna has not been observed previously (MacKenzie, 1985). Significant trends over time have, however, been reported. MacKenzie (1987) recorded a decline in the prevalence of L. tenuis in herring from the North Sea. Between 1973 and 1978 prevalences remained at a stable level of around 10%, declining to a stable level of less than 1% in the years after 1978. MacKenzie (1987) attributed this decline to the change in the distribution of an unknown copepod, ,0.001 ,0.001 ,0.001 ,0.001 ,0.001 ,0.01 NA ,0.001 .0.05 required as an intermediate host (Mudry & Dailey, 1971) associated with the end of the ‘great salinity anomaly’ (Dickson et al., 1988). The prevalence of this parasite in this study remains at about the level that MacKenzie (1987) reported from 1978 onwards; however, this could now be associated with the long-term decline in abundance of the spurdog (Squalus acanthias), the final host of this species (ICES, 2006). The apparent stability of a parasite community over a period of about 30 years is an interesting feature that has not been reported previously for herring. Admittedly, there are no data from the intervening period, and it could be the case that large fluctuations have taken place in prevalence and community composition that have not been observed. However, the lack of short-term fluctuations observed in this study would suggest that this is not the case. The finding of significant differences in the parasite fauna of herring sampled at different sites to the west of the British Isles allows us to reject the null hypothesis that herring in this area exist as a single, fully mixed population. We can also reject the null hypothesis that herring remain in discrete groups on their feeding grounds. This has implications for management, particularly in assigning catches of non-spawning fish to particular populations. Table 6. Comparison of prevalences of tag parasites (As, Anisakis simplex; Cp, Cercaria pythionike; Cs, C. doricha; Lt, Lacistorhynchus tenuis) from spawning and non-spawning samples collected in each functional unit. Parasite prevalences (%) 2003–04 Management area VIa (South) and VIIb,c VIa (North) Irish Sea Celtic Sea and Division VIIj 2004– 05 Sample As Cp Cd Lt Mixed ’03–’04 spawners ’04–’05 spawners 03–’04 Mixed ’03–’04 spawners ’04–’05 Mixed ’04–’05 spawners 03–’04 Mixed ’03–’04 spawners ’04–’05 Mixed ’04–’05 spawners Mixed ’03–’04 spawners ’04–’05 spawners 80.2 70.7 33.6 44.6 15.5 17.7 0.9 0 90.9 76.8 9.1 28.1 4.6 11.6 1.8 0 80.0 81.7 68.3 5.0 0 44.7 1.0 0 5.5 As Cp Cd Lt 71.8 36.5 14.1 0 95.9 96.3 1.0 21.3 1.0 5.1 1.0 0 48.9 60.2 75.4 12.1 0 10.2 4.9 0 5.8 4.0 4.3 0 89.1 10.9 9.8 1.1 2.0 8.7 0 144 N. Campbell et al. The analysis of parasite community composition shows a complex relationship between spawning herring populations to the west of the British Isles. Most apparent is the difference between the two spawning groups in area VIa (North). Autumn spawning herring taken from Cape Wrath have a very different parasite community to the spring spawning fish from Skye. The absence of either species of renicolid metacercariae in fish from Cape Wrath suggests that these fish have not recruited from the Scottish west coast, as these parasites are commonly found in west coast juveniles, and infection with these species is permanent. This pattern of infection was only observed in the juvenile herring collected in the eastern North Sea and from North Norwegian waters. While this might not be indicative of a recruitment of herring from the northern North Sea to the spawning population north of Scotland, we can say conclusively that the fish spawning at Cape Wrath have not been recruited from any of the nursery grounds to the west of the British Isles which have been sampled in this study. This is in agreement with the results of the Bløden herring tagging project (Anon., 1975), which showed conclusively that there was a migration of some adult herring from the North Sea to the west coast of Scotland. A number of studies have demonstrated a drift of herring larvae away from Cape Wrath, along the north coast of Scotland, around the Orkney Islands and into the North Sea (Heath & MacLachlan, 1987; Heath & Rankine, 1988). Heath et al. (1987) reported that larvae of herring spawning to the west of the Hebrides are entrained into the Scottish coastal current and carried towards the North Sea, while those spawning in other areas are inshore of the main current flow, and the transport route of these larvae has not been identified convincingly. Our results support this pattern of larval drift and adult migration in the opposite direction. The parasitological data allow us to make a number of inferences about stock identity of herring. The similarities in parasite prevalences in spawning fish collected to the north and west of Ireland, the mixed stock sample from management area VIa (South) and VIIb,c, and the juveniles collected at Stanton Bank and around the Hebrides suggest that these fish comprise a single stock. There is a relationship between juveniles in the Irish Sea and those that spawn in the Celtic Sea, as evinced by the presence of C. doricha and C. pythionike in these samples. The absence of these parasites in fish that spawn in the Irish Sea suggests that not all juveniles in the Irish Sea go on to recruit to the Irish Sea spawning stock. The intermediate prevalence of C. doricha and C. pythionike in the Irish Sea mixed stock sample would suggest that this comprises a mixture of herring spawning both in the Irish Sea and Celtic Sea. The sample of Skye spawners taken in 2004 recruited mainly from Scottish coastal nurseries, whereas the one taken in 2005 included a component recruited from the eastern North Sea. The parasite fauna of the mixed stock sample taken from ICES Division VIa (North) was similar in its composition to that of the spawning fish taken at Cape Wrath. However, the finding of L. tenuis in these fish suggests the presence of some Irish Sea spawners, the only area where this species was recorded, and the finding of C. doricha and C. pythionike is indicative of the presence of some fish from the Skye spawning sample. There is, therefore, the potential for fish from three ICES management units to be present in the mixed aggregations found over the Malin Shelf, with significant implications for management in this area. Interestingly, it is apparent that C. doricha was never observed as a single infection but occurred only in the presence of other parasites. The absence of L. tenuis in the mixed stock aggregation sample from the Celtic Sea indicates that no component of Irish Sea spawners could be detected in the Celtic Sea, suggesting that the fish spawning in the Irish Sea are more closely linked with populations on the Malin Shelf, rather than with those in the south-western approaches. There are obvious differences in parasite fauna between the fish spawning in the Irish Sea and the juveniles found on the nursery grounds there. Juveniles are infected with renicolid metacercariae, which are absent from the adults that spawn in the Irish Sea. The picture is reversed in fish from the Celtic Sea and Division VIIj management area, with juvenile fish having a significantly lower prevalence of C. doricha and C. pythionike than the spawning adults. The contiguity of these two management areas could mean that there is a drift of larvae from the Celtic Sea into the Irish Sea, which then migrate back as spawning adults. This would be supported by the finding of L. tenuis in a spawning sample to the south of Ireland – the only spawning fish infected with this parasite outside of the Irish Sea. There could also be recruitment of juveniles from the Celtic Sea, with their very low prevalences of C. pythionike and C. doricha, to the Irish Sea spawning population. However, it is clear that the southern Irish Sea is an area of stock mixing between populations with very different parasite prevalences. Anisakis simplex is one of the most heavily studied of all fish parasites, and has been used as a tag species in herring a number of times (Grabda, 1974; Horbowy & Poldoska, 2001). It reaches maturity in the intestines of cetaceans, uses a wide range of intermediate hosts and has a certain element of plasticity in its life cycle (Smith, 1983; Klimpel et al., 2004). The anisakid nematodes found in this study were identified as Anisakis simplex sensu stricto using restriction fragment length polymorphism (RFLP) (M. Cross, unpublished data). This is in agreement with the findings of Mattiucci et al. (2007), who found this species to be the dominant species infecting horse mackerel Trachurus trachurus in this region of the northeast Atlantic. Infection with anisakid nematodes has been shown to be cumulative with age (Gibson & Jones, 1993), which means that care must be taken when interpreting these data, particularly in relation to juvenile fishes, as differences in the mean age of otherwise similar samples of juvenile fishes could have a significant bearing on the levels of accumulated parasites observed. The finding of a Myxobolus sp. free living in the gall bladder is uncommon. Over 450 species of Myxobolus have been described from fish hosts. Of these, most have been described from freshwater fishes (Landsberg & Lom, 1991). This is the first record of a Myxobolus sp. infecting herring. The rarity and low intensity of this finding suggests that this was an accidental infection. This species does not provide any useful information on stock identity or distribution. The finding of Pronoprymna ventricosa is a new host record for herring, having previously been reported from a range of small pelagic, mainly clupeid, fishes, Helminth tags informing fisheries science and most frequently from shads (Alosa spp.). Bray & Gibson (1980) reported and described specimens from Alosa alosa and A. fallax caught in the Celtic Sea and the River Severn. The other adult digeneans found in the digestive tract are generalist species that have previously been reported from a wide range of species. In conclusion, the helminth fauna of the herring is a valuable tool for the identification of both discrete stocks and linkages between different life history stages in waters to the west of the British Isles. It is possible to build testable hypotheses from these results, which will hopefully lead to better understanding of stock dynamics and improved resource management in the future. Acknowledgements This project was funded by the EU Commission within the 5th framework program, Quality of Life and Management of Living Resources (Key Action 5: Sustainable agriculture, fisheries and forestry) Contract QLRT2001-01 056. The authors are deeply grateful to all involved in the collection of herring samples, and to our partners in the WESTHER project for helpful comments and thoughtful discussion. The authors also wish to thank the anonymous reviewers for their comments on this work. References Anon. (1975) Report of the ICES working group on the Bløden tagging experiment, 1969/70. ICES Cooperative Research Report 47. 136 pp. Bray, R. & Gibson, D. (1980) The Fellodistomidae (Digenea) of fishes from the northeast Atlantic. Bulletin of the British Museum Natural History (Zoology) 37, 1 –199. Butterworth, D.S. & Penney, A.J. (2004) Allocation in high seas fisheries: avoiding meltdown. pp. 165– 189 in Payne, A.I.L., O’Brien, C.M. & Rogers, S.I. (Eds) Management of Shared Fish Stocks. Oxford, Blackwell Publishing. Cross, M.A., Collins, C., Campbell, N., Watts, P.C., Chubb, J.C., Cunningham, C.O., Hatfield, E.M.C. & MacKenzie, K. (2007) Levels of intra-host and temporal sequence variation in a large CO1 sub-units from Anisakis simplex sensu stricto (Rudolphi 1809) (Nematoda: Anisakidae): implications for fisheries management. Marine Biology, 151(2), 695– 702. Dickson, R.D., Meincke, J., Malmberg, S-A. & Lee, A.J. (1988) The ‘great salinity anomaly’ in the Northern North Atlantic 1968– 1982. Progress in Oceanography 20, 103– 151. Gibson, D.I. & Jones, J.B. (1993) Fed up with parasites? A method for estimating asymptotic growth in fish populations. Marine Biology 117, 495– 500. Grabda, J. (1974) The dynamics of the nematode larvae, Anisakis simplex (Rud.) invasion in the South Western Baltic herring (Clupea harengus L.). Acta Ichthyologica et Piscatoria 4, 3– 21. Heath, M.R. & MacLachlan, P. (1987) Dispersion and mortality of yolk-sac herring (Clupea harengus L.) 145 larvae from a spawning ground to the west of the Outer Hebrides. Journal of Plankton Research 9, 613– 630. Heath, M. & Rankine, P. (1988) Growth and advection of larval herring (Clupea harengus L.) in the vicinity of the Orkney Isles. Estuarine, Coastal and Shelf Science 27, 547– 565. Heath, M.R., MacLachlan, P.M. & Martin, J.H.A. (1987) Inshore circulation and transport of herring larvae off the north coast of Scotland. Marine Ecology Progress Series 40, 11 – 23. Herrington, W.C., Bearse, H.M. & Firth, F.E. (1939) Observations on the life history, occurrence and distribution of the redfish parasite Sphyrion lumpi. United States Bureau of Fisheries Special Report 5, 1 – 18. Horbowy, J. & Poldoska, M. (2001) Modeling infection of Baltic herring (Clupea harengus membras) by larval Anisakis simplex. ICES Journal of Marine Science 58, 321– 330. ICES (2006) Report of the Working Group on Elasmobranch Fishes (WGEF). ICES CM 2006/ACFM:31. Iles, T.D. & Sinclair, M. (1982) Atlantic herring: stock discreteness and abundance. Science 215, 627– 633. Klimpel, S., Palm, H.W., Ruckert, S. & Piatkowski, U. (2004) The life cycle of Anisakis simplex in the Norwegian Deep (northern North Sea). Parasitology Research 94, 1 – 9. Landsberg, J.H. & Lom, J. (1991) Taxonomy of the genera of the Myxobolus/Myxosoma group (Myxobolidae: Myxosporea), current listing of species and revision of synonyms. Systematic Parasitology 18, 165–186. Larsen, G., Hemmingsen, W., MacKenzie, K. & Lysne, D.A. (1997) A population study of cod, Gadus morhua L., in northern Norway using otolith structure and parasite tags. Fisheries Research 32, 13 – 20. Lester, R.J.G., Thompson, C., Moss, H. & Barker, S.C. (2001) Movement and stock structure of narrowbarred Spanish mackerel as indicated by parasites. Journal of Fish Biology 59, 833– 842. MacKenzie, K. (1985) Relationships between the herring, Clupea harengus L., and its parasites. Advances in Marine Biology 24, 263–319. MacKenzie, K. (1987) Long-term changes in the prevalence of two helminth parasites (Cestoda: Trypanorhyncha) infecting marine fish. Journal of Fish Biology 31, 83– 87. MacKenzie, K. (2002) Parasites as biological tags in population studies of marine organisms: an update. Parasitology 124, S153– S163. MacKenzie, K. & Longshaw, M. (1995) Parasites of the hakes Merluccius australis and M. hubbsi in the waters around the Falkland Islands, southern Chile, and Argentina, with an assessment of their potential value as biological tags. Canadian Journal of Fisheries and Aquatic Science 52 (Suppl 1), 213– 224. Mattiucci, S., Farina, V., Campbell, N., Mackenzie, K., Ramos, P., Pinto, A.L., Abaunza, P. & Nascetti, G. (2007) Anisakis spp. larvae (Nematoda: Anisakidae) from Atlantic horse mackerel: their genetic identification and use as biological tags for host stock identification. Fisheries Research, in press. McQuinn, I.H. (1997) Year– class twinning in sympatric seasonal spawning populations of Atlantic herring, Clupea harengus. Fisheries Bulletin 95, 126– 136. 146 N. Campbell et al. Molloy, J. (2006) The herring fisheries of Ireland (1900 – 2005). Biology, research, development and assessment. 235 pp. Galway, The Marine Institute. Moore, B.R., Buckworth, R.C., Moss, H. & Lester, R.J.G. (2003) Stock discrimination and movements of narrow-barred Spanish mackerel across northern Australia as indicated by parasites. Journal of Fish Biology 63, 765– 779. Moser, M. & Hsieh, J. (1992) Biological tags for stock separation in Pacific Herring Clupea harengus pallasi in California. Journal of Parasitology 78, 54 –60. Mudry, D.R. & Dailey, M.D. (1971) Postembryonic development of certain tetraphyllidean and trypanorhynchan cestodes with a possible alternative life cycle for the order Trypanorhyncha. Canadian Journal of Zoology 49, 1249– 1253. Parrish, B.B. & Saville, A. (1965) The biology of the northeast Atlantic herring populations. Oceanography and Marine Biology Annual Review 3, 323– 373. Smith, J.W. (1983) Anisakis simplex (Rudolphi, 1809, det. Krabbe, 1878) (Nematoda: Ascaridoidea): morphology and morphometry of larvae from euphausiids and fish, and a review of the life-history and ecology. Journal of Helminthology 57, 205– 224. Smith, W.G. & Jamieson, A. (1986) Stock discreteness in herring: a conceptual revolution. Fisheries Research 4, 223– 234. Waldman, J.R. (2005) Definition of stocks: An evolving concept. pp. 5 – 14 in Cadrin, S.X., Friedland, D.F. & Waldman, J.R. (Eds) Stock identification methods. Amsterdam, Elsevier. Williams, H.H., MacKenzie, K. & McCarthy, A.M. (1992) Parasites as biological indicators of the population biology, migrations, diet, and phylogenetics of fish. Reviews in Fish Biology and Fisheries 2, 144– 176. (Accepted 29 March 2007) q 2007 Cambridge University Press Appendix VI - Haplotypes of Anisakis simplex s.s. found in herring to the west of the British Isles and in the Baltic Sea. ID No. 3 4 16 18 19 20 25 31 34 40 49 52 53 54 58 64 73 79 94 Consensus 1 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G 3-S04B-086AF 1 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-089CF 2 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T C T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-064RF 3 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T C T T A C T T T A T G T G C C G C A T T T C A T G G C 3-S04B-064PF 4 T T G T T G T C C A T G T T C T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-064GGGF 5 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G T A T T T C A T A G C 3-S04B-082LF 6 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T T A T A G C 3-S04B-079GF 7 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C T G C A T T T C A T A G C 3-S06A-012BF 8 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T C T A T G T G C C G C A T T T C A T A G C 3-S04B-082UF 9 T T G T C G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A C T T C A T A G C 3-S04B-082JF 10 T T G T C G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-064CCCF 11 T T G T C G T C T A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-064FFFF 12 T T G T T G T C T A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-064NNF 13 T W G T T G T C T A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-064XXF 14 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C C T T A T G T G C C G C A T T T C A T A G C 3-S04B-082MF 15 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T T C C G C A T T T C A T A G C 3-S04B-064SSF 16 T W G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-082SF 17 T T A T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-064EEF 18 T A G T T G T C C A T G T G T T A T A G C G A T A G T G T C T T T T T T T A C T T T G T G T G C C G C A T T T C A T A G C 3-S04B-064OF 19 T A G T T G T C C A T G T G T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C G T A G C 3-S04B-064IIF 20 T A G T T G T C C A T G T G T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-064NF 21 T A G T C C T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C C T T A T G T G C C G C A T T T C A T A G C 3-S04B-064ZF 22 T A G T C C T C C A T G T G T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 103 104 106 109 112 113 116 118 121 124 130 131 136 137 148 157 160 163 172 175 178 181 188 190 193 199 211 213 220 224 226 239 241 247 255 256 262 263 266 268 271 280 C 3-S04B-064BBF 23 T A G T T G T C C A T G T T T T G T A G T G A T A G T G T C T T T T G T T A C T T T A T G T G C C G C G T T T C A T A G C 3-S04B-064HHF 24 T A G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T C T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-064WWF 25 T T G T T G T C C A T G T T T T A T A G T G A T A G C G T C T T T T T T T A C T T T A T G T G C T G C A T T T C A T A G C 3-S04B-082OF 26 T T G T T G T T C A T G T T T T A T A G T G A T A G T G T C C T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-064LLF 27 T T G T T G T C C A T G T T T T A T A G T G A T A G C G T C C T T T C? T T A C T T T A T G T G C C G C A T T T C A T A G C C 3-S06A-095EF 28 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G C? G C C G C A T T T C A T A G 3-S04B-064ZZF 29 T T G T T G T C C A T G T T T T A T A A T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-082BF 30 T T G T T G T C C A T G T T T T A T A G T G A T G G T G T C T T T T T T T A C T T T A T G T G C C G C A T T C C A T A G C 3-S04B-082FF 31 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T A T G C C G C A T T T C A T A G C 3-S04B-064HHHF 32 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T T T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-064FFF 33 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T T T T T T T T T A C T T T A T G T G C C G C A T T T T A T A G C 3-S04B-064TTF 34 T T G T T G T C C G T G T T T T A T G G T G G T A G T G T T T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S04B-082NF 35 T T A T T G T C C A T G T T T T A T A G T G A T A G T G T T T T T T T T T A C T T T G T G T G C C G C A T T T T A T A G C 3-S06A-057CF 36 T T A T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T A T G C C G C A T T T C A T A G C 3-S04B-064QQF 37 T W G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G T C G C A T T T C A T A G C 3-S04B-064AAF 38 T A G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G T C G C A T T T C A T A G C C 4-X01A-020AF 39 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G T C G C A T T T C A T A G 3-S04B-064DDF 40 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G T C G C A T T T T A T A G C 3-S04B-064EEEF 41 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T G T G T G T C G C A T T T C A T A G C 3-S04B-064CCF 42 T T G T T G T C C A T G T T T C A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G T C G T A T T T C A T A G C 4-X01A-007AF 43 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A T C 4-X01A-010BF 44 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T C? T A C T T T A T G T G C C G C A T T T C A T A G C 4-X01A-008AF 45 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T C A C T T T A T G T G C C G C A T T T C A C A G C 3-S04B-064AAAF 46 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C T C A T T T C A T A G C 4-X01A-013AF 46 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C T C A T T T C A T A G C 4-X01A-018BF 47 T T G T T G T C C A T A T T T T A T A G T G A T A G T G T T T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 4-X01A-020BF 48 T T G T T G T C C A T G T T T T A T A G C G A T A G T G T C T T T T T T T A C T T G A T G T G C C G C A T T T C A T A G C 3-X01A-044AF 49 T A G T T G T C C A T G T T T T A T A G T G A T A G T G T C T C T C T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-X01A-042BF 50 T A G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T C T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-X01A-038AF 51 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G A 3-S01B-012GF 52 T T G T T G T C C A T G T T T T A T A G T G A T A G T G C C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S01B-003FF 53 T T G T T G C? C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S01B-063BF 54 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T T T T T T T T T A C T T T A C G T G C C G C A T T T C A T A G C 3-S01B-063DF 55 T T G T T G T T C A T G T T T T A T A G T G A T A G T A T T T T T T T T T A C T C T A T G T G C C G C A T T T T A T A G C 3-S01B-012KF 56 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T C A C T T C A T G T G C C G C A T T T T A T A G C 3-S06A-002AF 57 T A G A T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G T A T T T C A T A G C 3-S06A-015AF 58 T A G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S06A-008BF 59 T A G A T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T C T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S06A-012AF 60 T A G A T T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T C T A T G T G C C G C A T T T C A T A G C 3-S06A-057BF 61 T T G T T T C C A T G T T T T A T A G T G A T A G C G T C T T T T T T T A C T T T A T G T G T T G C A T T T C A T A G C G G 3-S06A-057BF 61 T T G T T G T C C A T G T T T T A T A G T G A T A G C G T C T T T T T T T A C 3-S06A-057GF 62 T T G T T G T C C A T G T T T T A C A G T G A T A G C G T C T T T T T T T A 3-S06A-099BF 63 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A 3-S06A-095GF 64 T T G T T G T C C A T G T T T T A T A G C G A T A G T G T C T T T T T T T 3-S06A-099CF 65 T T G T T G T C C A T G T T T T A T A G C G A T A G T G T C T T T T T T 3-S06A-099DF 66 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T 3-S06A-095CF 67 T T G T T G T C C A T G T T T C A T A G T G A T A G T G T C T T T T T 3-S06A-109CF 68 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T T A T G T G T T G C A T T T C A T A G C C T T T A T G T G C C T T A T G T G C T G C A T T T C A T A G C T C G C A T T T C A T A G A C T T T A T G T C G C C G C A T T T C A T A G C T A C T T T A T G T A C T T T A T G T G C C G C A T T T C A T A G T Y G C C G C A N T T C A T A G C T T A C T T T A T T T A C T T T A T G T G C C G C A T T T T A T A G C G T G C C G C A T T T C A T G G C 3-S10B-061BF 69 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G C C G T A T T T C A T A G C 3-S10B-061CF 70 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T C? T A C T T T A T G T G C C G T A T T T C A T A G C 3-S10B-061EF 71 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T Y T T A C T T T A T G T G T? C G C A T T T Y A T A G C 3-S10B-061GF 72 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G T? C G C A T C? T C A T A G C 3-S10B-063AF 73 T T G T T G T C C A W G T T T T A T A G T G A T A G T G T C C T T T T T T A C T T T A T G T G C C G C A T T T C A T A G C 3-S10B-063BF 74 T W G T T G T C C A T G T T T T A T A G T G A T A T T G T C T T T T T T T A C T T T A T G T G C C G C A Y T T C A T A G C 3-S10B-098AF 75 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A T G T G A C G C A T T T C A T A G C 3-S10B-098BF 76 T T G T T G T C C A T G T T T T A T A G Y G A T A G T G T C T T T T T T T A C T T T A T G T G C C G T A T T T C A T A G C 3-S10B-098CF 77 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A Y G T G C C G C A T T T C M T A G C 3-S10B-098DF 78 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C C T T A T G T G C C G C A T C? T C A T A G C 3-S10B-098GF 79 T T G T T G T C C A T G T T T T A T A G T G A T A G T G T C T T T T T T T A C T T T A Y G T G C C G C A T N T C A T A G C Appendix VII R code to perform MDS Scaling of Tamura & Nei distances between Anisakis haplotypes. library(genetics) cat( "> 3-S04B-086AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-079FF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082PF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-037AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-061AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-089BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064YYF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064OOF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064PPF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064BBBF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064KKF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064MMF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064GGF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064IIIF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-030MF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-030NF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082KF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-061DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 4-X01A-014AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 4-X01A-018AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064UUF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-057AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064VVF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082IF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064JJF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082GF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-117GF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-117HF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-003AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-003CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-045CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-018EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-003BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-018AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-068UF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-068EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-068PF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-062AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-063GF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-057DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-057EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-X01A-050BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-035BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-030DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-037CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-030EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-039KF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-025F", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-095DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-030KF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-027F", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-001AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-099AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-018CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-015EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-063FF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-098EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-021AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-098FF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-030QF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-030SF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-117AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-067FF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-061FF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-X01A-038BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-030IF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-035GF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-030JF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-007HF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-029BR", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-063AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 4-X01A-007BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-025CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-025EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-095BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-012DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-002EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-015DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-008EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-007IF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-037CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-035IF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-003DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-003EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-013EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S05A-022CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-068BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S05A-034BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-062HF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-062IF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-015AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-021BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-018BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-019BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S05A-044BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S05A-044CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064SSF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-035EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-035FF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 4-X01A-015AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 4-X01A-008BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 4-X01A-015BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064SF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-030FF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTCTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064HHF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTCTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-X01A-029AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTCTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-089CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTCTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064RF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTCTTACTTTTATGGTGCTCGTCATTTTCATGGC", "> 3-S04B-064PF", "GTTGTCCATGTTCTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064GGGF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S04B-034DR", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S04B-034ER", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S06A-008FF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-021CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S06A-037EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S5B-061BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S5B-007GF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S06A-021EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S04B-082EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S04B-082VF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S5B-030CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTCATGGTGCTCGTTATTTTCATAGC", "> 3-S5B-030LF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTCTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S5B-061CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTCTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S04B-034BR", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTCYACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S06A-012EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTACTTTCATAGC", "> 3-S06A-012EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTACTTTCATAGC", "> 3-S5B-098BF", "GTTGTCCATGTTTTATAGYGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S04B-079GF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTTGTCATTTTCATAGC", "> 3-S06A-012BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTCTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-05AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTCTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-05EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTCTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064DDDF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTCTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082TF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTCTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082UF", "GTCGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCACTTTCATAGC", "> 3-S06A-015FF", "GTCGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-050AF", "GTCGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-037DF", "GTCGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-029CR", "GTCGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082JF", "GTCGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064CCCF", "GTCGTCTATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064RRF", "GTCGTCTATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-050CF", "GTTGTCTATGTTTTATAGTGATAGTGTCTTTCTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-015CF", "GTTGTCTATGTTTTATAGTGATAGTGTCTTTCTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-007DF", "GTTGTCTATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTYYCATAGC", "> 3-S5B-007FF", "GTTGTCTATGTTTTATAGTGATAGTGTCCTTTCTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S05A-037CF", "GTTGTCTATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064FFFF", "GTTGTCTATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064NNF", "GTTGTCTATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082MF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTTCTCGTCATTTTCATAGC", "> 3-S04B-064EEF", "GTTGTCCATGTGTTATAGCGATAGTGTCTTTTTTTACTTTTGTGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064QF", "GTTGTCCATGTGTTATAGCGATAGTGTCTTTTTTTACTTTTGTGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064OF", "GTTGTCCATGTGTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCGTAGC", "> 3-S04B-064IIF", "GTTGTCCATGTGTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064ZF", "GTCCTCCATGTGTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064NF", "GTCCTCCATGTTTTATAGTGATAGTGTCTTTTTTTACCTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064XXF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACCTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S05A-044AF", "GTTGTCCATGTTTTGTAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCGTTTTCATAGC", "> 3-S04B-064BBF", "GTTGTCCATGTTTTGTAGTGATAGTGTCTTTTGTTACTTTTATGGTGCTCGTCGTTTTCATAGC", "> 3-S06A-057BF", "GTTGTCCATGTTTTATAGTGATAGCGTCTTTTTTTACTTTTATGGTGTTTGTCATTTTCATAGC", "> 3-S06A-057GF", "GTTGTCCATGTTTTACAGTGATAGCGTCTTTTTTTACTTTTATGGTGCTTGTCATTTTCATAGC", "> 3-S05A-041BF", "GTTGTCCATGTTTTATAGTGATAGCGTCTTTTTTTACTTTTATGGTGCTTGTCATTTTCATAGC", "> 3-S04B-034CR", "GTTGTCCATGTTTTATAGTGATAGCGTCTTTTTTTACTTTTATGGTGCTTGTCATTTTCATAGC", "> 3-S04B-006DR", "GTTGTCCATGTTTTATAGTGATAGCGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064LLF", "GTTGTCCATGTTTTATAGTGATAGCGTCCTTTCTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082OF", "GTTGTTCATGTTTTATAGTGATAGTGTCCTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-008DF", "GTTGTTCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-068JF", "GTTGTTCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-062EF", "GTTGTTCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGT", "> 3-S06A-013CF", "GTTGTTCATGTTTTATAGTGATAGTGTCTTTTTTTGCTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082BF", "GTTGTCCATGTTTTATAGTGATGGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTCCATAGC", "> 3-S5B-035HF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATAGTGCTCGTCATTTTCATAGC", "> 3-S04B-082FF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATAGTGCTCGTCATTTTCATAGC", "> 3-S06A-062BF", "ATTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATAGTGCTCGTCATTTTCATAGC", "> 3-S06A-057CF", "ATTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATAGTGCTCGTCATTTTCATAGC", "> 3-S04B-082SF", "ATTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082NF", "ATTGTCCATGTTTTATAGTGATAGTGTTTTTTTTTACTTTTGTGGTGCTCGTCATTTTTATAGC", "> 3-S04B-064FFF", "GTTGTCCATGTTTTATAGTGATAGTGTTTTTTTTTACTTTTATGGTGCTCGTCATTTTTATAGC", "> 3-S04B-064HHHF", "GTTGTCCATGTTTTATAGTGATAGTGTTTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-X01A-042AF", "GTTGTCCATGTTTTATAGTGATAGTGTTTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-05CF", "GTTGTCCATGTTTTATAGTGATAGTGTTTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-117EF", "GTTGTCCATGTTTTATAGTGATAGTGTTTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-063EF", "GTTGTCCATGTTTTATAGTGATAGTGTTTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-089AF", "GTTGTCCATGTTTTATAGTGATAGTGTTTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-063BF", "GTTGTCCATGTTTTATAGTGATAGTGTTTTTTTTTACTTTTACGGTGCTCGTCATTTTCATAGC", "> 3-S04B-064TTF", "GTTGTCCGTGTTTTATGGTGGTAGTGTTTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S05A-034AF", "GTTGGTCGTGTTTTATAGTAGTAGTGTTTTTTTTTGTTTTTATGTTGTTTGTTATTTTTACGGT", "> 3-S01B-063DF", "GTTGTTCATGTTTTATAGTGATAGTATTTTTTTTTACTCTTATGGTGCTCGTCATTTTTATAGC", "> 3-S01B-039LF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTTATAGC", "> 3-S01B-063CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTTATAGC", "> 3-S06A-095FF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTTATAGC", "> 3-S5B-007AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTTATAGC", "> 3-S01B-012IF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTTATAGC", "> 3-S06A-002FF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTTATAGC", "> 3-S01B-012HF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTTATAGC", "> 3-S04B-082LF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTTATAGC", "> 3-S01B-012KF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTCACTTTCATGGTGCTCGTCATTTTTATAGC", "> 3-S5B-045BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTTATAGC", "> 3-S5B-030GF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTTATAGC", "> 3-S04B-064DDF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTTATAGC", "> 3-S06A-068KF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTTATAGT", "> 3-S04B-064QQF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTCATAGC", "> 3-S04B-064AAF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTCATAGC", "> 3-S5B-030PF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTCATAGC", "> 4-X01A-020AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTCATAGC", "> 3-S06A-025DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTCATAGC", "> 3-S06A-062DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTCATAGC", "> 3-S5B-030TF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTCATAGC", "> 3-X01A-032BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTCATAGC", "> 3-S04B-082QF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTCATAGC", "> 3-S04B-064EEEF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTGTGGTGTTCGTCATTTTCATAGC", "> 3-S06A-099BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACCTTTATGGTGTTCGTCATTTTCATAGC", "> 3-S04B-037BF", "GTTGTCCATGTTTTATAGCGATAGTGTCTTTTCTTACTTTTATGGTGTTCGTCATTTTCATAGC", "> 3-S5B-061EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTYTTACTTTTATGGTGTTCGTCATTTTYATAGC", "> 3-S5B-061GF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTCTCATAGC", "> 3-S06A-062CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTYACTTTTATGGTGTTCGCCATTTTCATAGC", "> 3-S04B-064CCF", "GTTGTCCATGTTTCATAGTGATAGTGTCTTTTTTTACTTTTATGGTGTTCGTTATTTTCATAGC", "> 3-S06A-095CF", "GTTGTCCATGTTTCATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTTATAGC", "> 4-X01A-007AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATATC", "> 3-S06A-021DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTCTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 4-X01A-05BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTCTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 4-X01A-008AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTCACTTTTATGGTGCTCGTCATTTTCACAGC", "> 3-S06A-062GF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTCACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-035AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTCACTTTTATGGCGCTCGTCATTTTCATAGC", "> 3-S04B-064AAAF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCTTCATTTTCATAGC", "> 4-X01A-013AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCTTCATTTTCATAGC", "> 3-S06A-008BF", "GATGTCCATGTTTTATAGTGATAGTGTCTTTTCTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-008BF", "GATGTCCATGTTTTATAGTGATAGTGTCTTTTCTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-012AF", "GATGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTCTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-002AF", "GATGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S04B-082BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTGTGGTGCTCGTTATTTTCATAGC", "> 3-S06A-037AF", "GTTGTCCATGTTTTATAGTGATAGTGTCCTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S5B-063AF", "GTTGTCCAWGTTTTATAGTGATAGTGTCCTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-018DF", "GTTGTCCATGTTTTATAGTGATAGTGTCCTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-030BF", "GTTGTCCATGTTTTATAGTGATAGTGTCCTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-035CF", "GTTGTCCATGTTTTATAGTGATAGTGTCCTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-002CF", "GTTGTCCATGTTTTATAGTGATAGTGTCCTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S05A-022BF", "GTTGTCCATGTTTTATAGTGATAGTGTCCTTTTTTACTTTTATGGTGCTCGTCACCTTCATAGC", "> 3-S06A-013BF", "GTTGTCCATGTTTTATAGTGATAGCGTCCTTTTTTACTTTTATGGTGCTCGTCATTTTCATAAC", "> 3-S06A-117DF", "GTCGTCCATGTTTTATAGTGATAGTGTCCTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-062JF", "GTCGTCTATGTTTTATGGTGATAGTGTTTTTTTTTACTTTTATGGTGCTCGTTATTTTCATAGC", "> 3-S05A-037BF", "GTTGTCCATGTTTTATAGTGACAGTGTTTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-117FF", "GTTGTCCATGTTTTATAGTGATAGCGTTTTTTTTTACTTCTATGGTGCTTGTCATTTTCATAGC", "> 4-X01A-018BF", "GTTGTCCATATTTTATAGTGATAGTGTTTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S05A-037AF", "GTTGTCCATATTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-X01A-044AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTCTCTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-X01A-042BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTCTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-X01A-038AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGA", "> 3-S01B-012GF", "GTTGTCCATGTTTTATAGTGATAGTGCCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-012JF", "GTTGTCCATGTTTTATAGTGATAGTGCCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S01B-003FF", "GTTGCCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-030UF", "GTTGTCCATGTTTTATAGCGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-068HF", "GTTGTCCATGTTTTATAGCGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-095GF", "GTTGTCCATGTTTTATAGCGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-099CF", "GTTGTCCATGTTTTATAGCGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGT", "> 4-X01A-020BF", "GTTGTCCATGTTTTATAGCGATAGTGTCTTTTTTTACTTTGATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-025FF", "GTTGTCCATGTTTTATAGCGATAGTGTCTTTTCTTACTTTTATGGTGCTCGTCATTTTTATAGC", "> 3-S06A-099DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGYGCTCGTCANTTTCATAGC", "> 3-S06A-59CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATGGC", "> 3-S06A-068CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCAYTTTCATACC", "> 3-S5B-063BF", "GTTGTCCATGTTTTATAGTGATATTGTCTTTTTTTACTTTTATGGTGCTCGTCAYTTTCATAGC", "> 3-S5B-098AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGATCGTCATTTTCATAGC", "> 3-S5B-098DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACCTTTATGGTGCTCGTCATTCTCATAGC", "> 3-S5B-098CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTAYGGTGCTCGTCATTTTCMTAGC", "> 3-S5B-098GF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTAYGGTGCTCGTCATTNTCATAGC", "> 3-S06A-068GF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTGTGGAGCTCGTCATTTTCATAGC", "> 3-S06A-068IF", "GTTGTCCATGTTTTATAGTGATACTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-117BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGYCATTTTCATAGC", "> 3-S06A-002DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTACGGYGCTCGTCATTTTCATAGC", "> 3-S06A-05DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACCTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-012CF", "GTTGTCCATGTTTTATAGTGATGGTGTCTTTTTTTACTTTTATGGTGTTCGTCATTTTCATAGC", "> 3-S04B-064ZZF", "GTTGTCCATGTTTTATAATGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-008CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTCTTACTCTTATGGTGCTCGTCATTTTCATAGC", "> 3-S06A-037BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTCTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-030HF", "GTCGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTCTTATGGTGCTCGTCATTTTCACAGC", "> 3-S5B-030OF", "GTTGTCCATGTTCTATAGTGATAGTGTCTTTTTTTACTTTTATGGCGCTCGTCATTTTCATAGC", "> 3-S06A-095EF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGCGCTCGTCATTTTCATAGC", "> 3-S05A-041CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGCGCTCGTCATTTTCATAGC", "> 3-S04B-082AF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGCGCTCGTCATTTTCATAGC", "> 3-S5B-030WF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGCGCTCGTCATTTTCATGGC", "> 3-S5B-007BF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCATCATTTTCATAGC", "> 3-S5B-007CF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTATTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S5B-035DF", "GTTGTCCATGTTTTATAGCGATAGTGTTTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-006CR", "GTTGTYCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTAYGGTGCTTGTCATTTTCATAGC", "> 3-S04B-034AR", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTCTTTACTCTTATGGTGCTCGTCATTTTCATNNC", "> 3-S05A-041AF", "GTTGTCCATGTTTYACAGTGATAGTGTCTTTTTTTACTTTTATGGTGCTCGTCATTTTCATAGC", "> 3-S04B-082DF", "GTTGTCCATGTTTTATAGTGATAGTGTCTTTTTTTACTTTTATGGTGCCCGTCATTTTCATAGC", file = "exdna.txt", sep = "\n") ex.dna3 <- read.dna("exdna.txt", format = "fasta") dist.dna(ex.dna3, model="TN93", pairwise.deletion=TRUE)->distance numbers<c(2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,5,5,2,2,5,6,6,2,4,2,2,2,2,4,4,1,1,5,4, 1,4,1,1,1,4,1,4,4,6,5,5,2,5,1,4,4,5,4,4,4,4,4,1,5,4,5,5,5,4,4,5,6,5,5,5, 5,2,2,1,6,4,4,4,4,4,4,4,5,4,5,1,1,4,3,4,3,4,4,4,4,4,4,3,3,2,5,5,6,6,6,2, 5,2,6,2,2,2,2,2,2,4,4,4,5,5,4,2,2,5,5,5,2,4,4,5,2,4,4,4,2,2,2,4,2,4,2,2, 2,2,2,4,5,5,3,2,2,2,2,2,2,2,2,2,2,3,2,4,4,3,2,2,2,2,4,4,4,4,2,5,2,4,4,2, 2,2,2,6,4,4,1,2,1,2,3,1,1,1,4,5,1,4,1,2,1,5,5,2,4,2,2,5,6,4,4,5,6,2,2,4, 2,5,5,4,2,4,6,4,6,6,4,5,2,6,4,4,4,4,2,4,5,4,5,5,4,3,4,4,4,3,4,6,3,6,6,6, 1,1,1,5,4,4,4,6,4,4,4,4,5,5,5,5,5,4,4,4,4,4,4,2,4,4,5,5,4,3,2,5,5,5,5,2, 2,3,2) # # # # # # 1=Celtic Sea 2=Donegal 3=Clyde 4=Irish Sea 5=Cape Wrath 6=Baltic CMD<-cmdscale(distance) jc1<-jitter(CMD[,1], factor=1, amount=0) jc2<-jitter(CMD[,2], factor=1, amount=0) long<-c(jc1,jc2) new<-matrix(long, nrow=290, ncol=2) plot(new, main="Tamura & Nei Genetic Distance between Anisakis sp. from around the UK and Ireland", pch= 1, col=numbers, cex=2, xlab="1st Principle Component", ylab="2nd Principle Component") plot(CMD,type="n", main="Tamura & Nei Genetic Distance between Anisakis sp. from around the UK and Ireland") text(CMD[,1],CMD[,2], numbers, col=numbers, xlab="1st Principle Component", ylab="2nd Principle Component") Appendix VIII – Abstract of presentation to International Council for the Exploration of the Seas Annual Science Conference, Aberdeen, September 2004 CM 2005/K:24 Sequencing the Cytochrome Oxidase I (CO1) region of the parasitic nematode Anisakis simplex to identify spawning populations of herring, Clupea harengus, to the west of the British Isles. Campbell, N.1*, Cross. M.A.2, Collins. C.3, Watts, P.C.2, Chubb, J.C.2, Cunningham, C.O.3, MacKenzie, K.1, & Hatfield, E.M.C.3 1. School of Biological Sciences (Zoology), University of Aberdeen, Aberdeen, Scotland, AB24 2TZ. 2. School of Biological Sciences, Biosciences Building, University of Liverpool, Liverpool, L69 7ZB. 3. FRS Marine Laboratory, PO Box 101, 375 Victoria Road, Aberdeen, AB11 9DB. *Tel:+44(0)7976 832 417 email: Neil.Campbell@abdn.ac.uk Abstract The use of parasites as biological tags for investigating stock identity has been widely applied in fisheries science. Recently, the spatial pattern of intraspecific genetic variation in parasite populations has been investigated as a means of discerning their host population structure, This has had useful outcomes for stock identification research projects (Mattiucci et al., 2000), with the cytochrome-c-oxidase subunit 1 (COI) gene from the mitochondrial DNA identified as a good indicator of intraspecific population variance in many species of parasite. Anisakis simplex s.l. Rudolphi, 1809 was chosen as a candidate marker species because of its presence in all sample areas. A 276 bp region of the COI gene was sequenced from around 300 Anisakis spp., parasiting herring at spawning sites in the Baltic Sea, the Atlantic Ocean (Celtic Sea, Donegal, Cape Wrath) and the Irish Sea (Irish Sea and Clyde), as part of an ongoing multidisciplinary stock identification project, WESTHER. Analysis of spatial genetic structure was carried out to determine the utility of the COI fragment to discriminate between parasites from putative populations, and therefore their associated host, herring spawning stocks. The COI region is highly variable, with a high proportion (ca. 50 %) of haplotypes unique to each sampling site. Examination of the sequences revealed differences in presence and absence of haplotypes, and frequency of shared haplotypes, between the parasite populations. However, at present it is not possible to differentiate among distinct parasite populations (as delineated by spawning host populations), with the present data set. ARLEQUIN v. 2.000 software (Schneider et al., 2000) was used to calculate standard diversity indices and pairwise FST values between samples, as well as a hierarchical analysis of molecular variance (AMOVA); the AMOVA partitioned the level of genetic differentiation among various groups based on geographical location and possible migration routes of host populations, including Baltic vs Atlantic Ocean (Celtic Sea, Donegal, Cape Wrath) vs Irish Sea (Irish Sea and Clyde), and Baltic vs Celtic Sea vs (Irish Sea and Clyde) vs (Donegal and Cape Wrath). Nucleotide diversity (π) within populations was low (0.008-0.019), while haplotype diversity (h) was high (0.712-0.934), a pattern that is characteristic of recent population expansion. Pairwise FST values between sampling sites were not significant, indicating no genetically distinct populations. No significant genetic structure was found between groups of samples or among samples within groups indicating that most variation in the Anisakis CO1 was found within host populations rather than between host populations or groupings of host populations. These preliminary results may not reflect the true structure of the anisakid populations. A significant (r2=0.976 P<0.001), positive correlation was found between number of different haplotypes and sample size, suggesting that too few parasites per sample were analysed to attain a reliable estimation of genetic variation within, and therefore between, each population. Nevertheless, it is likely that the life-cycles of both parasite and host also effect a low level of genetic differences between anisakid populations, thereby limiting their usefulness for fine-scale differentiation of herring host populations. Although a number of parasites display genetic groupings that reflects the population biology of their hosts, such studies have focussed on parasite-host models with relatively simple life cycles, such as ticks and sea birds (McCoy et al., 2005), parasitoid wasps and butterflies (Kankare et al., 2005) and maternally inherited bacteria and fruit flies (Riegler et al., 2005). Anisakis spp., in contrast, have complex life cycles that involve free-living stages, various intermediate and paratenic hosts, and marine mammals (Klimpel et al., 2004). This diversity of potential hosts would suggest that although Atlantic herring may exist in semi-discrete populations at some stages in their life, the movements of all other hosts maintains mixing in the Anisakid population. In addition, the life history of the herring further complicates an evaluation of Anisakis population structure. Herring accumulate anisakids throughout their lives and there will be an overlap in exposure to anisakid populations between potentially different host populations, resulting from common feeding grounds or migratory routes. This overlap may mask differences due to anisakids acquired only on spawning grounds. The qualitative results of a population expansion may support our belief that Anisakis spp. are widely dispersed by many hosts. We are continuing our investigations, but these preliminary results suggest that an investigation of a herring parasite which has a simple life cycle, such as the monogenean, Gyrodactylus harengi, would be a more suitable candidate for revealing population structure of its host. It was not considered in this study due to limitations with the material available for examination. This study forms one part of the multidisciplinary WESTHER project, and was supported by funding from the EC Commission; Quality of Life and management of Living Resources (Contract QLRT-2001-01056). References Mattiucci S., Nascetti G., Tortini E., Ramadori L., Abaunza P. , Paggi L.(2000) Composition and structure of metazoan parasitic communities of European hake (Merluccius merluccius) from Mediterranean and Atlantic waters: stock implications. Parassitologia 42 (Suppl. 1):176-186 Kankare, M; Van Nouhuys, S; Hanski, I. (2005) Genetic Divergence Among Host-Specific Cryptic Species in Cotesia Melitaearum Aggregate (Hymenoptera: Braconidae), Parasitoids of Checkerspot Butterflies. Annals of the Entomological Society of America 98(3):382-394. Klimpel, S; Palm, HW; Ruckert, S; Piatkowski, U. (2004) The life cycle of Anisakis simplex in the Norwegian Deep (northern North Sea). Parasitology Research 94(1):1-9. McCoy, KD; Boulinier, T; Tirard, C. (2005) Comparative host-parasite population structures: disentangling prospecting and dispersal in the blacklegged kittiwake Rissa tridactyla. Molecular Ecology. 14(9):2825-38 Riegler, M; Sidhu, M; Miller, WJ; O'Neill, SL. (2005) Evidence for a Global Wolbachia Replacement in Drosophila melanogaster. Current Biology 15(15): 1428-33 Schneider S, Roessli D, Exoffier C (2000) Arlequin version 2000, Genetics and Biometry. University of Geneva, Switzerland. Appendix IX – Identification of fish stocks through neural network analysis of parasitological data (published in Analysing Ecological Data, eds. A.F.Zuur, E.N. Ieno & G.M. Smith). 25 Fish stock identification through neural network analysis of parasite fauna Campbell, N., MacKenzie, K., Zuur, A.F., Ieno, E.N. and Smith, G.M. 25.1 Introduction The main aim of fisheries science is to interpret relevant information on the biology of the species in question, records of fishing effort and size of catches, in order to predict the future size of the population under different fishing regimes, allowing fisheries managers to make decisions on future fishing effort. The common approaches to evaluation, modelling and management of fish stocks assume discrete populations for which birth and death are the significant factors in determining population size, and immigration and emigration are not (Haddon 2001). Consequently, for successful management of fisheries, it is vital that populations which conform to these assumptions are identified. In areas where two stocks mix, it is useful to be able to quantify this so that catches can be assigned to spawning populations in the correct proportions. A number of techniques have been used to identify discrete fish stocks and quantify their mixing, such as physical tags, microchemistry of hard parts and a range of genetic markers. See Cadrin et al. (2005) for a comprehensive review. There is no single “correct” approach to stock identification, the trend being towards multidisciplinary studies which apply a range of methods to the same set of fish to allow cross-validation of findings. One of the more popular methods involves the use of parasites as biological tags, and has been used for over sixty years (Herrington et al. 1939). This technique has a number of advantages over other methods, such as low cost, suitability for delicate species and straightforward sampling procedures. Its main disadvantage is the limited knowledge available on the life-cycles and ecology of many marine parasites, but as research in these areas results in more and more information becoming available, the efficiency of the method increases accordingly (MacKenzie 1983; Lester 1990). The theory behind this technique is that geographical variations in the conditions which a parasite needs to successfully complete its life cycle occur between areas and so between fish stocks (factors such as distribution of obligatory hosts in the life cycle, environmental conditions or host feeding behaviour). This leads to differences in parasite prevalence (the proportion of a host population infected with a particular parasite species), abundance (the average number of a particular 452 25 Fish stock identification through neural network analysis of parasite fauna parasite species found per host) or intensity (the average number of parasites found in infected individuals) between areas. A “classical” parasites-as-tags study involves carrying out a preliminary study to identify parasite species which vary in prevalence, abundance or intensity within the study area, followed by the collection of data from a larger number of fish over a number of years to produce conclusive evidence of a lack of mixing between different parts within the study area. Note that the absence of a difference in parasite prevalence, abundance or intensity between samples is not necessarily indicative of homogeneous mixing. This approach allows migrations, recruitment of juveniles or mixing of different spawning populations to be observed and quantified. The practical application of this method was modelled and verified by Mosquera et al. (2000). It is particularly useful in areas where a small but significant degree of mixing between two populations occurs, obscuring genetic differences between populations. A number of more recent studies have taken a more complex approach to the statistical treatment of parasites as tags of their host populations. These studies have considered each fish as a habitat and treated the entire parasite fauna in that individual as a community. Discriminant analysis (DA) is applied to the parasite abundance data of the community, in order to identify groups of similar fishes (Lester et al. 1985). Moore et al. (2003) had some success with the application of DA to the parasite fauna of narrow-barred Spanish mackerel (Scomberomorus commersoni) around the coast of Australia, in order to quantify movement and mixing of stocks. Discriminant analysis, however, makes a number of assumptions which make it less suitable for this sort of analysis. Firstly, for testing of hypotheses, DA assumes normality and homogeneity within each group of observations per variable (in this case, parasite abundances). Secondly, DA works best with roughly equal sample sizes, and requires the number of variables to be less than the smallest sample size minus two. 25.2 Horse mackerel in the northeast Atlantic The Atlantic horse mackerel (Trachurus trachurus) is a small pelagic species of fish, with a maximum size of about 40cm. They are the most northerly distributed species of the jack-mackerels (family Carangidae) (FAO 2000), and support a sizeable fishery in the northeast Atlantic, both for human consumption and for industrial processing. Catches in the region have been over 500,000 tonnes per year in recent times. They feed at a slightly higher trophic level than many small pelagic fishes, their diet consisting of planktonic copepods, small fishes and benthic invertebrates. This diverse diet is reflected in a diverse parasite community, and 68 taxa have been reported to infect T. trachurus (MacKenzie et al. 2004). There has been uncertainty over the identity of stocks in the northeast Atlantic for over a decade. In this area, the International Council for the Exploration of the Seas (ICES) issues management advice for the horse mackerel which assumes the 25.7 Discussion 453 existence of three stocks. These are a (i) Western stock, (ii) a North Sea stock and (iii) a Southern stock (Figure 25.1). These stocks have been defined mainly from observations of the distribution of eggs in regular plankton surveys, and on historical records of the distribution of catches (Eltink 1992; ICES 1992). Until recently, publications dealing with the definition of stock structure in horse mackerel were rare and covered only a small part of the species distribution. In the southern stock there are some works dealing with differences in anisakid infestation levels (Abaunza et al. 1995; Murta et al. 1995), whilst using allozymes, some authors found differences between areas in the North-east Atlantic (Nefedov et al. 1978) whereas others did not (Borges et al. 1993). A recent, EU funded multidisciplinary study, HOMSIR, has resolved some of the problems with stock identity, but there are still unanswered questions. One of the problems with stock definition for T. trachurus is that it is a highly migratory species, spawning along the edge of the continental shelf, in water around 200 meters deep, then dispersing to feed over a wider area. It is thought that the Western and North Sea stocks overlap at certain seasons in the English Channel (Macer 1977), which may cause some degree of mixing between these stocks. Mixing between the Western and Southern stocks remains an unknown quantity, and there has been particular concern about the boundary between these stocks. Figure 25.1. A graphical representation of the distribution of horse mackerel stock in the north east Atlantic. 1. North Sea Stock. 2. Western Stock. 3. Southern Stock. 4. African Stocks, after ICES (1992, 2004). Locations of samples collected for verification of stock identity are marked with filled circles and those collected to investigate stock mixing with empty circles. 454 25 Fish stock identification through neural network analysis of parasite fauna 25.3 Neural networks The original “neural network” model was proposed in the 1940s (McCulloch & Pitts 1943), although it is only with the advent of cheap, powerful computers that this technique has begun to be applied widely. There has been a great deal of hype surrounding neural networks. They are, however, simply a non-linear statistical approach to classification. One of the attractions of neural networks to their users is that they promise to avoid the need to learn other more complex methods - in short, they promise to avoid the need for statistics! But this is a misconception: for example, extreme outliers should be removed, collinearity of variables should be investigated before training neural networks, and it would be foolish to ignore obvious features of the data distributions and summaries such as the mean or standard error. The neural network promise of "easy statistics" is, however, partly true. Neural networks do not have implicit assumptions of linearity, normality or homogeneity, as many statistical methods do, and the sigmoid functions which they contain appear to be much more resistant to the effects of extreme values than regression based methods. Many of the claims made about neural networks are exaggerated, but they are proving to be a useful tool and have solved many problems where other methods have failed. The name “neural network” derives from the fact that it was initially conceived of as a model of the functioning of neurons in the brain – the components of the network represent nerve cells and the connections between them, synapses, with the output of the nerve switching from 0 to 1 when the synapses linking to it reach a “threshold value”. For the purposes of this chapter, a neural network can be thought of as a classification model of a real world system, which is constructed from the processing units (“neurons”) and fitted by training a set of parameters, or weights, which describe a model that forms a mapping from a set of given values known as inputs to an associated set of values known as outputs (Saila 2005). The weights are trained by passing sets of input-output pairs through the model and adjusting the weights to minimize the error between the answer provided by the network and the true answer. A problem can occur if the number of training iterations, or “epochs” is too large. This reduction in classification success of the data not used in training is known as overfitting. Once the weights have been set by a suitable training procedure, the model is able to provide output predictions for inputs not included in the training set. The neural network takes all the input variables presented in the data and linearly combines them into a derived value, in a so-called “hidden layer object” or node (Smith 1993). It then performs a nonlinear transformation of this derived value (Figure 25.2). The use of multiple hidden layer objects in a neural network allows different non-linear transforms of data, with each neuron (node) having it’s own linear combination, increasing the classifying power of the network. Originally, the neuron was activated with a step function (represented as the dashed line in Figure 25.2), when the combined input values exceeded a certain value, however, this more flexible sigmoid function allows differentiation and 25.7 Discussion 455 0.0 0.2 0.4 Y 0.6 0.8 1.0 least squares fitting, leading to the back propagation algorithm, making it possible to tune the weights more finely. -10 -5 0 5 10 X Figure 25.2. An example of the transformation which the hidden layer applies to linear combinations of the input data: y=1/(1+exp(-x)). An example of the step function used in early neural network models is shown in the dashed line. A possible example is shown in Figure 25.3. In this case, we are interested in knowing the stock composition of a mixed sample of fish, and we have count data on six species of parasites from these fish. These counts are treated as the six input variables to our network. The network has four units in the hidden layer. The neural network in such an example would need to have been trained with data from fish which we knew belonged to stocks X and Y beforehand if it was to work successfully. This type of neural network is referred to in the literature by a number of names, such as feed-forward network, multilayer perceptron, or simply vanilla neural network, named for the generic ice-cream flavour. The number of units in a hidden layer is variable. Problems can arise if too few or too many units are used. If the network has too few units, it will not be flexible enough to correctly classify the data it is presented with. On the other hand, if it has too many units, a problem known as overparameterisation occurs, this reduces the chances of successful classification. Having many hidden layer objects also increases the computing power required to run the function. Often, a trial-and-error approach is used to determine the optimum number of hidden layer units for a particular data set, however there are other methods which take a more considered approach, such as cross-validation, bootstrapping, and early stop. Neural networks differ in philosophy from most statistical methods in several ways. A network typically has many more inputs than a typical regression model. Because these are so numerous, and because so many combinations of parameters result in similar predictions, the parameters can quickly become difficult to interpret and the network is most simply considered as a classifying “black box”. This means that areas where a neural-network approach can be applied in ecology are 456 25 Fish stock identification through neural network analysis of parasite fauna widespread. They are less useful when used to investigate or explain the physical process which generated the data in the first place. In general, the difficulty in interpreting what the functions contained within these networks mean has limited their usefulness in fields like medicine, where the interpretation of the model is vital. Figure 25.3. Graphical representation of the structure of a neural network. This one has six inputs (P1 to P6), one hidden layer with four neurons or units (A to D) and two output neurons into which it will classify the data (stocks X and Y) There have been a number of uses of neural networks in fisheries science (Huse and Grøjsæter 1999; Maravelias et al. 2003; Engelhaard and Heino 2004), however, these have mainly attempted to predict changes in fish abundance, recruitment or distribution, based on environmental and ecological inputs. The use of neural networks in recognising fish stocks is a relatively new development, and has been restricted to analyses of morphometry or of otolith microchemistry (Murta 2000; Hanson et al. 2004). These were reviewed and summarised by Saila (2005). A more specific introduction to neural network architectures can be found in Dayhoff (1990) and Smith (1993). These techniques produce continuously distributed values, such as the distance between two points on the body of a fish, or the quantity of a particular element in an otolith. Parasitological studies, on the other hand, are characterised by relatively large number of fish with low number of parasites, and a small percentage of observations with high numbers. This tends to give problems with classical statistical techniques that require normality assumptions, such as classical discriminant analysis. Furthermore, some forms of parasite, such as metazoan species, are too numerous to count, therefore a fish is either classed as infected or uninfected. Neural networks are able to cope with both forms of numeric data, as well as presenceabsence data, in any combination. 25.7 Discussion 457 Note that the question which is addressed by a neural network approach to parasite data is not “do these fish belong to different stocks”, but, “based on the parasitological data available, is it possible to successfully assign these fish to a stock”. The difference is subtle, but it should be apparent that low levels of successful classification between two sets of observations would not suggest that they belong to two different stocks, whereas high success would support a hypothesis that the samples were drawn from different populations. 25.4 Collection of data This work is based on samples of T. trachurus collected as part of the HOMSIR stock identification project (see Abaunza et al. In press) for a detailed explanation of the theory behind sample collection). Fish were collected with a pelagic trawl by a number of research vessels at eleven locations in the north-east Atlantic (see Figure 25.1) in 2001 and were immediately frozen and returned to the laboratory for examination. Between 34 and 100 fish were collected from each location (Table 25.1). To investigate spawning stock identity, three samples each from the Western and Southern stocks, and one each from the North Sea and African stocks, were examined. For estimation of stock mixing, one sample from a non-spawning seasonal fishery from the Norwegian coast, and one spawning sample from the boundary between the Western and Southern samples were examined. Fish were examined externally for parasites, before opening the visceral cavity. All organs were separated, irrigated with physiological saline and examined for the presence of parasites under a stereo-microscope (6-50x). The opercula (gill covers) were removed along with the individual gill arches, irrigated with physiological saline and examined for the presence of monogenean and copepod parasites under a stereo-microscope. Smears of liver and gall bladder were examined for protozoan and myxozoan infections at a magnification of 325x using phase contrast microscopy. Table 25.1. Location and size of samples collected for stock identification. Stock North Sea Western Southern African Mixed Norwegian Mixed Spanish Lat. 54.45N 52.53N 48.45N 51.35N 44.00N 41.00N 38.30N 37.00N 19.58N 57.41N 43.35N Long. 06.00E 12.03W 09.29W 11.06W 01.38W 08.50W 09.20W 08.30W 17.28W 05.10E 08.52W Sample Size 50 34 50 50 50 100 52 100 50 50 50 458 25 Fish stock identification through neural network analysis of parasite fauna 25.5 Data exploration A total of 636 fish from eleven sites were examined. Parasites which infected less than 2% of fish were deemed to represent either rare species or “accidental” infections and were discounted. Eleven species of parasites were found to be commonly present (Table 25.2). Exploration of any set of data is an essential first step in carrying out an appropriate analysis. One of the problems with this sort of study is that because fish vary in size, age or sex between samples, the examinations cannot be thought of strictly as replications of observations on a homogeneous population. It is therefore important to look for variation caused by these factors and remove it from further analysis. Plots were made of the abundance of different parasites against fish length, sex and age (Figure 25.4). No relationships were apparent, with the exception of the nematode, Anisakis spp.. This species encysts in the body cavity of the horse mackerel and therefore is cumulative with age. A ln(y+1) transformation was performed on the Anisakis abundance values, and a linear regression of the transformed value was carried out against fish length. The residual values of this regression were then taken forwards for use in the later analysis. This is one way of reducing the bias caused by differing lengths between samples. Table 25.2. Commonly encountered parasite species used for stock identification analysis. Class Myxosporea Species Location Data Type Alataspora serenum (Gaevskaya and Kovaleva 1979) Gall Bladder Presence/ Absence Goussia cruciata (Theolan 1892) Liver Anisakis spp. Hysterothylacium sp. (larval forms) Hysterothylacium sp. (adult forms) Body Cavity Abundance Body Cavity Abundance Intestine Abundance Tergestia laticollis (Rudolphi 1819) Derogenes varicus (Müller 1784) Ectenurus lepidus (Loos 1909) Intestine Stomach Stomach Abundance Abundance Abundance Pseudaxine trachuri (Parona and Perugia, 1889) Gastrocotyle trachuri (van Beneden and Hesse 1863) Heteraxinoides atlanticus (Gaevskaya and Kovaleva 1979) Gills Abundance Gills Abundance Gills Abundance Apicomplexa Presence/ Absence Nematoda Digenea Monogenea 25.7 Discussion B 400 300 0 150 100 200 Number of Anisakis 300 250 200 Length (mm) 350 500 A 459 AFRICAN NORT H SOUT H WEST ERN 150 200 250 300 350 3 Length(mm) C D 1 0 -1 Residual 4 3 -2 2 -4 0 -3 1 Log(Anisakis +1) 5 2 6 Stock 150 200 250 300 350 Length (mm) AFRICAN NORT H SOUT H WEST ERN Stock Figure 25.4. A: Fish lengths vary significantly between stock. B: There is a significant relationship between fish length and Anisakis spp. abundance. C: Natural logarithm transformation of Anisakis abundance produces a linear relationship. D: The residuals of this relationship are taken forward for use in the analysis after removal of length dependency. The values for the southern stock are generally lower as the parasite occurred less frequently in this area. 25.6 Neural network results The first problem in a neural network approach to a classification problem is to select an appropriate structure for the network. This is often done on an ad-hoc basis. A single hidden-layer, feed forward neural network was constructed using the nnet function (Venables and Ripley 2002) in the R statistical software environment v2.1.1 (R Development Core Team, 2005). This network was provided with half the fish from all samples on a random basis, for use as a training set, then used to reclassify the remaining fish to a stock. In the first instance, the hidden layer contained one object, and to remove chance results caused by selection of fish at random, the process was repeated 100 times with different random training sets. The mean percentage of successful classifications over all simulations is then taken and stored. We then increased the number of units in the hidden layer and repeated the process. Finally, mean successful classification is plotted against number of hidden layer units (Figure 25.5). This allows an educated guess to be made as to the most suitable number of hidden layer objects, which is sufficiently 460 25 Fish stock identification through neural network analysis of parasite fauna 80 60 40 20 0 Percentage Successful Classification 100 flexible to reclassify data successfully, but not so many that overfitting occurs and excessive processing time is required. All other settings remained at their default values. 0 10 20 30 40 50 Number of Units In Hidden Layer Figure 25.5. Estimation of the optimum number of units in the hidden layer. Mean successful classification (±1 standard deviation) reaches over 90% with 8 nodes. Further increases in the number of nodes cause a decrease in success, and an increase in variability, at the cost of increased computing time. From these data it would appear that the optimum number of hidden layer units is around eight. We decided to use eight hidden layer units for the later neural networks, as there was some decrease in performance of the network at higher values. To investigate stock identity, the same method of selecting half of the fish in a sample as a training set and reclassifying remaining fish (the “test set”) was used. The percentage of fish correctly classified was recorded, along with the numbers from each stock misclassified, and the stock to which they were assigned. This network had eleven inputs, eight objects in the single hidden layer, and four outputs. To obtain a measure of the error inherent in selecting a training sample at random, and allowing the starting weights of the network to be selected at random, the process was repeated 1000 times. Once outcomes were sorted, median successful classification was represented by the 500th value, and 95% confidence limits by the 50th and 950th values (Figure 25.6). From these results it is apparent that the neural network is able to correctly classify fish to a stock with a high degree of accuracy – median successful classification for the Southern stock is 95%. These findings support the ICES stock definitions as they are currently applied, and suggest that the application of this neural network to mixed stock analysis will give an accurate picture of stock composition. For investigation of the two mixed stock samples, the whole of the spawning data set was used to train the neural network. This was then used to re- 25.7 Discussion 461 100 B 60 % 40 20 0 0 20 40 % 60 80 A 80 100 classify the mixed data in question. The neural network was allowed to choose random starting weights, hence outputs are still variable, even when considering the same set of data. Consequently, this process was also repeated 1000 times in order to produce an estimate of inherent variability. Southern Western 100 North Sea African North Sea Southern Western African North Sea Southern Western D 60 % 40 20 0 0 20 40 % 60 80 C 80 100 African African North Sea Southern Western Figure 25.6. Percentage of fish from the test set assigned to each stock by the neural network; A: fish from the North Sea. B: fish from African waters. C: fish from the Southern stock. D: fish collected in the Western stock area. The Norwegian seasonal fishery shows a more mixed composition than any of the spawning samples, suggesting it may be made up of fish from more than one area. The neural network classifies around 65% of fish as belonging to the Western stock. The remaining 35% are a mixture of Southern and African stocks. Very few fish are assigned to the North Sea stock (Figure 25.7). The spawning sample from the area of stock uncertainty to the north of Spain is much less conclusive. The neural network assigns around 40% of the sample to the Western stock, 30% to the Southern stock and 20% to the African stock (Figure 25.8). 25 Fish stock identification through neural network analysis of parasite fauna 40 0 20 % 60 462 African North Sea Southern Western 0 10 20 % 30 40 50 Figure 25.7. Stock membership of horse mackerel from a mixed, non-spawning seasonal fishery which develops in the summer months off the Norwegian coast. African North Sea Southern Western Figure 25.8. Classification of horse mackerel in spawning sample taken from area of stock uncertainty on the north west coast of Spain. 25.7 Discussion Stock identification This study represents the first successful attempt to validate the existence of separate fish stocks by the application of a neural network to parasite abundance data. This technique successfully reclassifies over 90% of fish in the test set, from Western and Southern stocks. Success in reclassifying fish from the North Sea and African stocks is lower, although still over 80%. 25.7 Discussion 463 The neural network is good at distinguishing between members of the Western and Southern stocks. Median misclassification of fish from the Western stock to the Southern stock is 2%, and from the Southern stock to the Western is 0%. Using parasites as biological tags and multivariate analysis of morphometric distances, it is possible to distinguish between fish from the western and North Sea stocks, but it has previously been impossible to conclusively distinguish between fish from the western and southern areas using any method (ICES 2004). Although neural networks are robust enough to deal with differences in sample size, it is apparent that the areas with larger sample sizes (western and southern) have higher success rates than the two areas with only 50 fish (North Sea and African stocks). The confidence ranges for these two areas are also much wider than for the larger samples. It is interesting to note that where misclassification occurs, it tends to be towards stocks which are already believed to mix, rather than to those with which mixing is not regarded as possible. This might suggest that a small number of “alien” fish from adjacent stocks are present in our “discrete” spawning samples. These fish would produce such misclassifications, although the difficulties in investigating the processes that go on inside the neural network make this impossible to verify. It has been suggested that an element of stock mixing can take place between the Western and North Sea stocks while fish overwinter in the English Channel (Macer 1977). This has been supported by recent studies (MacKenzie et al. In press). If stock mixing is occurring in this area, it is likely that this effect is being reflected in the results obtained from the neural-network, and that there are a number of fish which belong to the Western stock in the North Sea, and viceversa. This will slightly confuse the picture which the neural network gives, and could explain the tendency of North Sea fish to be misclassified into the Western stock. A degree of mixing has also been proposed between the Southern and African stocks (Murta 2000; MacDonald 2005). This is supported by our findings. Although successful classification of fish from the Southern stock is over 95%, the neural network classifies around 20% of fish from the African sample as belonging to the Southern stock. Very few fish from the southern samples are classified as belonging to the African stock. This could suggest that mixing between these areas is a one way process. Norwegian non-spawning sample Having established the utility of a neural network approach to assign fish to spawning stocks based on host-parasite data, it is a straightforward matter to apply this to a non-spawning stock to investigate its composition. These findings suggest that fish in the Norwegian sample do not come from a single stock, but rather are drawn mainly from the Western stock, with a sizeable proportion from much more southerly stocks. This finding is in line with work by Abaunza (In press) who measured growth rates of horse mackerel, and found variability from stock to stock, with fishes from warmer waters growing more quickly than their more 464 25 Fish stock identification through neural network analysis of parasite fauna northerly conspecifics. When examining fish from the Norwegian area, growth rates were noticeably higher than that in neighbouring fisheries, to the west of Ireland and the southern North Sea. Abaunza proposed the existence of a highly migratory “infra-stock” which spawned and overwintered in the Southern stock area, then migrated northwards to feed in Norwegian waters. Evidence supporting this hypothesis came from the discovery of characteristically “southern” parasites in fish from the Norwegian area (MacKenzie et al. In press). The neural network classified few fish from this sample as belonging to the North Sea stock. This is interesting when considering how close it is to this stock. It suggests that very little mixing of North Sea and other fish in this area. This is an important finding for fisheries managers to consider when allocating catches of fish from the Norwegian fishery to particular spawning populations. La Coruña spawning sample The neural network was not able to classify fish from the north west coast of Spain to a particular stock with any great certainty. No one stock appears to dominate this area, and the 95% confidence limits are relatively small. The boundary between the Western and Southern stocks was recently moved from the north to the west coast of the Iberian peninsula (ICES 2004). These findings suggest that this change was appropriate, in that the sample is classed as more “Western” than “Southern”, but also suggest that a high degree of mixing takes place in this area, and that more intensive sampling in this area would be a worthwhile contribution to stock identification of the horse mackerel. Conclusions It is apparent that a neural network approach to classifying individuals into presupposed groups is a powerful tool for problems such as this. The ease with which it is possible to use neural networks, their lack of restrictive assumptions and their ability to cope with combinations of different types of data make them extremely useful for dealing with problems of classification in ecology. Acknowledgements We are grateful to all our partners in the HOMSIR stock identification project for their helpful advice and provision of samples. The work which this chapter was based on was funded by the European Commission, under the 5th Framework, contract no. QLRT-PL1999-01438. Visit www.homsir.com for more details on the project outcomes. We would like to thank Anatoly Saveliev for valuable comments on the statistical aspects of this chapter.

RELATED PAPERS

RELATED TOPICS

Log In

Population studies of horse mackerel (Trachurus trachurus (L.)) and herring (Clupea harengus L.) using parasites as biological tags.

Population studies of horse mackerel (Trachurus trachurus (L.)) and herring (Clupea harengus L.) using parasites as biological tags.

Related Papers

RELATED PAPERS

RELATED TOPICS