Faunal Remains: Results by Species

PA RT 1 0 Animals ‘Like it or not, this rich site had been stressed and re-assessed in some form or other almost annually; the archaeological pack has chased ater the only juicy bone apparently available, gnawing at it and extracting the marrow of its information base.’ (Schadla-Hall 1988, 27) C HAPTER 23 Faunal Remains: Results by Species Becky Knight, Nicky Milner, Terry O’Connor, Ben Elliott, Harry K. Robson, Mike Buckley, Piotr Witkowski, Sophy Charlton, Oliver Craig and Matthew Collins Introduction he faunal remains from the original excavations were studied by Frederic Fraser and Judith King of the then Department of Zoology, British Museum (Natural History), and the bird bones were identiied by Marjorie Platt of the Royal Scottish Museum, Edinburgh. he collection was noted as being important because it contained several animals now absent from the fauna of this country and there was enough material to give an impression of the composition of the fauna at the start of the postglacial period in Yorkshire (Clark 1954, 70). Clark, in his synthesis chapter, highlights the importance of red deer, roe deer, elk, aurochs and pig [sic] in terms of their relative abundance and suggests that it is likely that the ‘total bag’ of these species would have been twice as much. In doubling these igures and then converting to dead weight, clean carcass weight and calories, he suggested that a group of four families (comprising an active man, a moderately active woman and three children) could live of this food supply (approximately 50,000 kg) for 6 1/4 years (Clark 1954, 16). However, it was noted that it cannot be assumed that all the meat was consumed on site and some could have been dried for use elsewhere. We now know that the nature of occupation was highly complex: the site is much larger than previously thought, it spans c. 800 years and the faunal remains that survive are only likely to be a small percentage of what was used and deposited. In addition, we know that Clark did not retain everything: bone, antler and lint appear to have been purposefully deposited in several parts of the backill which will have skewed previous analyses. What had been collected was then dispersed across a number of museums mainly around England, but also farther aield, which makes it harder to re-examine. he faunal remains found in the recent excavations at Star Carr are generally in fairly poor condition (Chapter 22). Nevertheless, even some of the really badly preserved bones have revealed important data using both traditional zooarchaeological techniques, as well as biomolecular approaches such as Zooarchaeology by Mass Spectrometry (ZooMS) and stable isotope analysis. Our recent excavations have resulted in the discovery of some new species to add to Clark’s list. In addition, the open area excavation, 3D plotting, application of GIS and the dating programme allowed us to take a new approach and consider the variability of the data through time and across space (see Chapter 7). Figure 23 (page 193): Fracturing long bones for marrow extraction (Copyright Aimée Little, CC BY-NC 4.0). How to cite this book chapter: Knight, B., Milner, N., O’Connor, T., Elliott, B., Robson, H. K., Buckley, M., Witkowski, P., Charlton, S., Craig, O. and Collins, M. 2018. Faunal Remains: Results by Species. In: Milner, N., Conneller, C. and Taylor, B. (eds.) Star Carr Volume 2: Studies in Technology, Subsistence and Environment, pp. 195–254. York: White Rose University Press. DOI: https://doi. org/10.22599/book2.i. Licence: CC BY-NC 4.0 196 Star Carr Volume 2 he faunal remains from 2004–2010 were initially assessed by Sarah Viner, Rachael Parks and Cluny Johnson (reports in the ADS) but the whole assemblage has been reanalysed by BK who has been the faunal remains specialist since 2012 and was onsite throughout the 2013–2015 seasons. A large quantity of sediment was also sampled for lotation (see Chapter 15) and a signiicant quantity of bone fragments were retrieved from this, though much remains unidentiied to species or element. his chapter begins by outlining the original assemblage and subsequent reinterpretations. It then presents the methodologies used in examining the faunal assemblage. An overview of taphonomic factors is given, followed by an overview of the quantiication of species across the site. he diferent species are then detailed in turn, in terms of spatial distribution, elements, age and sex where possible, seasonality where possible, modiications by carnivores and humans, palaeopathology, and the MNI. he faunal remains data have been collated on a spreadsheet which is available via the ADS and spatial data is available allowing all specimens which have been 3D recorded to be plotted (https://doi.org/10.5284/1041580). he original assemblage and subsequent reinterpretations In the faunal remains chapter in the original monograph (Clark 1954), the total number of identiied specimens (hereater NISP) were not provided but the minimum number of individuals (hereater MNI) per species were listed (Table 23.1) with the note that no human or ish remains were found. A description of the remains for each species was given and the discussion drew attention to the size of some species, notably red deer which were much larger than modern Scottish counterparts. In addition, Clark requested that the analysts should consider the season of occupation (Clark 1972, 22). his was determined, based on red deer, elk and roe deer antlers and when these are shed, with the conclusion that people must have inhabited the site during the winter until spring, notably April. Common name Scientiic name Common crane Grus grus (Linnaeus, 1758) MNI 1 ? White Stork Ciconia ciconia (Linnaeus, 1758) 1 Red-breasted merganser Mergus serrator (Linnaeus, 1758) 1 Red-throated diver Colymbus stellatus (Pontoppidan, 1763) 1 Great crested grebe Podiceps cristatus (Linnaeus, 1758) 1 Little grebe Podiceps ruicollis (Pallas, 1764) 1 Lapwing Vanellus vanellus (Linnaeus, 1758) 1 Buzzard Buteo buteo (Linnaeus, 1758) 1 Duck (size of pintail) Anas acuta (Linnaeus, 1758) 1 Wolf Canis lupus (Linnaeus, 1758) 2 Fox Vulpes vulpes (Linnaeus, 1758) 2 Pine marten Martes martes (Linnaeus, 1758) 2 Badger Meles meles (Linnaeus, 1758) 1 Hedgehog Erinaceus europaeus (Linnaeus, 1758) 1 Pig [sic] Sus scrofa (Linnaeus, 1758) 5 Elk Alces alces (Linnaeus, 1758) 11 Red deer Cervus elaphus (Linnaeus, 1758) 80 Roe deer Capreolus capreolus (Linnaeus, 1758) 33 Aurochs Bos primigenius (Bojanus, 1827) 9 Hare Lepus cf. europaeus (Pallas, 1778) 1 Beaver Castor iber (Linnaeus, 1758) 7 Table 23.1: Mammals and birds identiied by Fraser and King (Clark 1954). Faunal Remains: Results by Species 197 During Clark’s original excavations, the inds were hand collected and there was no sieving protocol implemented. his methodology will have inluenced the recovery rate of smaller species and more delicate elements, as acknowledged by Legge and Rowley-Conwy (1988, 12) in their reanalysis. Also, it is clear that Clark had a selection process for the inds although this is never outlined in his publications. hrough the rediscovery of inds in the backill during the most recent excavations (see also Chapter 7), it appears that Clark mainly kept elements that were complete or large fragments of bone and antler that were easily identiied to species. Butchered fragments were kept, but only if they could be identiied to species and element. Smaller fragments of bone and antler, or elements such as ribs that were diicult to identify to species, were not retained. In the 1970s, a number of re-evaluations of the data were carried out. Because red deer antlers had been used for making artefacts such as barbed points, it was suggested that red deer antler should be discounted from analyses (Jacobi 1978; Caulield 1978; Pitts 1979; Grigson 1981), which radically reduced the MNI of red deer. Pitts (1979) noted that the roe deer shed antler had not been used for making tools and so could be used for calculations of MNI. As noted in Chapter 2, this led to a study undertaken by Legge and Rowley-Conwy (1988) in which all of the faunal material from Clark’s excavations, including material housed in various museums around the country, was re-assessed. his primarily set out to re-examine the bones from the large mammals: aurochs, elk, red deer, roe deer and wild pig [sic] (Legge and Rowley-Conwy 1988). heir re-quantiication of the bones changed the MNI for all of these species (Figure 23.1). heir work excluded red deer antler (shed and unshed), but also roe deer antler (38 let and 39 right) on the basis that virtually all have been broken out of the skull and if they are taken into account a very extreme and unlikely sex ratio is arrived at (Legge and Rowley-Conwy 1988, 10). It is suggested that the roe deer antler may have been collected for an unknown purpose. In addition, the representation of elk increased due to some bones having previously been identiied as aurochs. here is no explanation as to why the MNI for aurochs increased, but this is perhaps a result of the corpus of Danish aurochs being published since the original faunal assessment, as noted by Legge and Rowley-Conwy (1988, 10). he wild pig [sic] decreased by an MNI of one because a mandible had been re-identiied by Sebastian Payne as bear (Legge and Rowley-Conwy 1988, 10). In addition, other important analyses were undertaken which changed the species list to some degree. Harrison (1987) re-examined the Star Carr birds. Nine bird species were identiied in the original analysis by Platt, but Harrison states that of these nine, four ‘appear to be invalid’ (Harrison 1987, 141) (Table 23.2). Harrison disagreed with the identiication of white stork (Ciconia ciconia) as this was based on a fragment of long Figure 23.1: Diferences in the MNI between the analyses presented in Clark (1954, 15) and Legge and Rowley-Conwy (1988, 9) (Copyright Star Carr Project, CC BY-NC 4.0). 198 Star Carr Volume 2 Common name Scientiic name Element Side Red-throated diver Colymbus stellatus Distal humerus Right Ulna Let Great crested grebe Podiceps cristatus Tibiotarsus Let Little grebe Tachybaptus ruicollis Ulna Right Brent goose Branta bernicla Humerus shat Let Red-breasted merganser Mergus serrator Tibiotarsus Right Tibiotarsus Let Carpometacarpus Let Common scoter Melanitta nigra Ulna Let Common crane Grus grus Humerus Right Humerus Let Ulna Right Ulna Let Table 23.2: Birds reclassiied by Harrison (1987). bone midshat only, and he revised this identiication to ‘indeterminate’ (Harrison 1987, 141). here were several other identiications which Harrison changed: the common buzzard (Buteo buteo) humerus midshat was re-identiied as Brent goose (Branta bernicla); the pintail (Anas acuta) carpometacarpus was reassigned to red-breasted merganser (Mergus serrator) and the lapwing (Vanellus vanellus) ulna was re-assigned to common scoter (Melanitta nigra). here has been much discussion about whether wolf or domesticated dog was found at Star Carr. Fraser and King (1954, 72) reported wolf; however, Degerbøl’s (1961) and Benecke’s (1987) analyses make it clear that they are domestic dogs, and a further dog right femur was found in the collection by Rick Schulting (Schulting and Richards 2009, 499). A further notable addition to the record is brown bear (Ursus arctos). An upper let canine was found from this species by Tot Lord, who visited the excavations before Clark’s trenches were backilled (Dark et al. 2006, 191). In particular, he searched the deposits next to the birch trees in cutting II and found items by pushing his hands into the sot sediment at the base of the sections. In doing this he felt a skull but could not retrieve it; however, he was able to extract a tooth from it (Edwards et al. 2014). During our excavations we expected to ind this skull; however, despite excavating all around cutting II there was no sign of it and the conclusion is that it has either demineralised (if it was on the east side—SC24 only contained two pieces of jellybone) or that it was retrieved by someone else. Two other brown bear bones have since been identiied: as already noted, a broken mandible previously identiied as wild boar (Legge and Rowley-Conwy 1988, 10) and an axis vertebra (Noe-Nygaard 1983). It has been suggested that the brown bear skull, mandible and axis vertebra possibly belong to the same animal (Dark et al. 2006). Methods Introduction A number of traditional zooarchaeological methods were used to assess taphonomy, taxa, quantiication, ageing, sexing, palaeopathology and seasonality. In addition, a range of biomolecular techniques were used: ZooMS to aid in the identiication of taxa, and stable isotope analysis to assess diet for deer. In addition, aDNA was attempted on the bones of dog and wolf (by Greger Larson and team, University of Oxford) and beaver (by Melissa Marr, Danielle Schreve and Ian Barnes, Royal Holloway and the Natural History Museum); however, no positive results were produced, thought to be due to the poor preservation environment for aDNA at the site. Faunal Remains: Results by Species 199 Taphonomy he faunal remains excavated from the dryland areas of the site were found to be in a particularly poor condition: specimens were very desiccated, fragmentary and fragile and exhibited a high frequency of root damage. Due to these factors, almost half (49%) of the specimens could not be identiied to species or element, although this is fairly typical for material found within these types of sediments and of this age. he bones from the wetland part of the site exhibited a range of problems including demineralisation, lamination and concretion of minerals as set out in Chapter 22 which also provides the methodology used in analysing the varying states of preservation. All bones were analysed for evidence of natural taphonomic changes such as weathering and exposure, root or bioturbation damage and trampling, but also for evidence of modiication: (1) anthropogenic modiications such as charring, cut marks or butchery evidence; (2) modiications made by animals such as uneven (ragged-edged) breakage, gnaw marks, crushing or cracking and tooth impressions (Lee Lyman 1994). he anatomical location of these modiications on each of the elements was described in detail and added to the faunal database. Elements with particularly clear or unusual evidence were then selected for photography. Identiication of the causes of the modiications was also noted; for example, classic percussion damage for marrow extraction (Noe-Nygaard 1977) and longitudinal splitting for bone tool production (David 2005). In terms of the animal modiication evidence, attempts were made, where possible, to identify the species responsible by examining any clear tooth impressions or gnawing patterns. Some species, such as canids and felines, tend to favour particular elements and the pattern of gnawing can be very distinctive (Haynes 1983): for example, dogs have a tendency to drag their teeth over the surface of a bone creating tooth scores and gouges and rodents leave parallel tooth scores (O’Connor 2000). Taxa identiication he species or genus was assessed using the University of York zoological reference collection and the atlas of Quaternary mammals (Pales and Lambert 1971a; 1971b; Pales and Garcia 1981a; 1981b) by BK and TOC. In addition, measurements were taken where possible in order to try to diferentiate between red deer and elk using von den Driesch (1976). However, this was sometimes problematic due to the large size of the red deer and the post-depositional changes that have occurred to a large quantity of the remains. It was not possible to identify the species or genus of a large number of specimens, particularly those on the dryland because they were severely degraded. When species could not be determined, specimens were categorised to family level (e.g. Cervid sp., Bovid sp.), category (e.g. bird sp., ungulate sp.) and size (large, medium or small mammal). he composition of bird bones is very diferent to that of mammals as they are usually very gracile, lightweight and have a very diferent internal structure, making them much easier to identify as bird, though not necessarily to species. It was possible in some cases to get to family level using size, robusticity and internal matrix detail of the bones. However, the majority of the material categorised to family level came from the application of ZooMS. A total of 280 bones were sampled and analysed by MB from the dryland excavations carried out in 2007 and 2008. In addition, a further 60 bones were analysed by PW and MC from the dryland excavations carried out in 2013 and 2014. Only specimens from the dryland were sampled for ZooMS because this is where identiication using traditional methods was most challenging due to the deterioration of the bone. Collagen, the dominant protein in mineralised tissues, is known to persist with extraordinary longevity and can be examined using ZooMS (Buckley et al. 2009), whereby peptides are fractionated using C18 ZipTip® pipette tips, and analysed by Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry (MALDI-ToF MS). his method introduces the sample into the mass spectrometer as an air-dried co-crystallised acidic mixture of peptides and a coloured matrix. Under vacuum, the multiple samples spotted onto the plate are each in turn volatilized by a laser shot at the crystals. he energy is absorbed by the matrix within the crystals, causing it to partially decompose and volatilise. his pulse of energy also lits the co-crystallised peptides of the plate and adds a proton, giving them a positive charge. hese charged peptides are then accelerated by an electric ield and guided down a light tube, which assesses the mass of each peptide by its time of light to reach a detector, peptides with lower mass arriving earlier than those with higher mass. A small amount of each sample was removed and placed in an eppendorf micro-centrifuge tube. hen 200 μl of 0.1M sodium hydroxide (NaOH) was added to the sample to remove tannins and other chromophoric 200 Star Carr Volume 2 compounds. he sample was then vortexed and centrifuged. he NaOH was removed, and 200 μl of 50mM ammonium bicarbonate solution (NH₄HCO₃) pH 8.0 (AmBic) was added in order to ‘rinse’ the sample. he eppendorfs were then vortexed briely before being centrifuged again for one minute. his rinsing process was repeated twice more. 75 μl of AmBic was then added to the samples. Next, 1 μl of trypsin solution was added to each of the ‘second’ eppendorfs. hese were then incubated for four hours at 37 °C. Following incubation, the samples were centrifuged, and 1 μl of 5% TFA solution was added to each to terminate trypsin activity. Peptides were then extracted from the sample solution using C18 ZipTip® pipette tips and eluted with 50 μl of conditioning solution. hen 1 μl of sample was spotted onto a Bruker ground steel target plate, following which 1 μl of matrix was added on top. Each sample was spotted in triplicate and the plate was then run on the Bruker Ultralex. Quantiication NISP and MNI were used in quantifying this material with the aim of calculating relative abundance of animals, or parts of animals, deposited at the site. Weights were not recorded because the deterioration of the deposits have afected the faunal remains so much that weight calculations would not be comparable across site, or with other sites, and would therefore be meaningless. NISP is deined as the number of identiied specimens. Usually the term ‘specimens’ refers to either a single bone, or a group of fragments originally derived from a single bone that could be reitted together. hese calculations are used to illustrate the relative numbers of each element in order to assess which elements are abundant and which elements appear to be missing from the assemblage. For this analysis it was decided to include individual fragments in the NISP values; however, the potential bias of high fragmentation through working bone into tools was taken into account when discussing the dominance of diferent taxa within the assemblage. MNI is deined as the minimum number of individuals. his was calculated using the most abundant skeletal element for each species within the assemblage, sorting into let and right specimens, and identifying repeating areas of the element. Age and size of the element were also taken into account in the calculations where possible. As set out above, there has been much debate about whether to include antler in the quantiication of deer from the site. In this study, the calculations of both NISP and MNI include unshed antler which still retained some attached crania, but do not include shed antler, as this material may have been collected from elsewhere and brought to the site for making artefacts. Similarly, bone and antler tools such as barbed points, bodkins and bone chisels are not included in these calculations because these artefacts might have been mobile within the landscape and may have been curated over signiicant periods of time. For convenience here, even though antler is of course a specialised form of bone, the terms ‘bone’ and ‘antler’ are used as separate categories. Ageing, sexing and palaeopathology In terms of age assessment, the two main methods used were epiphyseal fusion and tooth development. In general, as an animal ages the bone surfaces become more sculptured, muscle attachments become more pronounced and epiphyseal fusion occurs which obliterates the sutures. he presence of these characteristics can therefore add additional support for diferentiation between adult and juvenile remains. Studies by Lesbre (1897–8), Grigson (1982), Purdue (1983) and Fanden (2005) investigated epiphyseal fusion in a number of different taxa, and these studies have been used to assess age for the Star Carr faunal remains. For tooth development, studies by Severinghaus (1949), Habermehl (1961), Briederman (1965), Matschke (1967), Silver (1969) and Hillson (2005) have been used for the diferent taxa. For some species it is also possible to use tooth wear data to aid with estimations of age (Severinghaus 1949; Aitken 1975; Brown and Chapman 1991a; Brown and Chapman 1991b; Hewison et al. 1999); where possible, these methods of analysis have been applied in this study. Sex assessment is generally much more diicult to establish, especially in assemblages of archaeological animal bone of this age, due to deterioration and fragmentation. Sexing of the pelves could not be carried out because although some fairly complete specimens were found, the detail was obscured by concretions on the bone surfaces or exploded bone caused by salt crystals, or they had been compressed, warped or highly Faunal Remains: Results by Species 201 demineralised (Chapter 22). he canine teeth of pigs and wild boars were assessed as these show marked sexual dimorphism, allowing the sex identiication of mandibles and maxillae. Metric analysis of the bones for size comparisons to published datasets of comparable species is another way to aid sex identiication; however, this is highly dependent on good preservation levels (Klein and Cruz-Uribe 1984). Legge and Rowley-Conwy (1988) made measurements of the astragalus, metacarpal, scapula and humerus, measuring the width of the distal end (Bd) and the thickness (Dd) ater von den Driesch (1976). However, due to issues of compression, demineralisation and bloating this was not possible for the majority of bones recovered during the recent excavations. Only the astragalus of aurochs could be assessed in this way. In addition, calculations for age and sex were not possible for bones from the dryland due to poor preservation and high fragmentation of the material. Palaeopathological changes can be assessed by looking for bone modiication, either in terms of the bone structure or appearance of an element. here are many diferent ways in which pathologies can exhibit themselves on a skeleton. Arthropathies are pathologies of the joint, and this can encompass arthritic changes such as lesions on the articular surfaces or osteophyte formation (bony lipping or spurs occurring at the joints) and changes caused by rapid weight gain or uneven weight distribution across the limbs such as osteochondrosis (localised articular lesions caused by malformation of subchondral bone), for example. Various diseases can also exhibit themselves within the skeleton, such as periodontal disease which can afect the mandible and maxilla, causing a widening of the tooth sockets, and can result in the eventual loss of teeth. Trauma to the skeleton can also be exhibited and recorded, for example healed breaks, blunt force traumas or evidence of human damage such as projectile perforations. Very few specimens exhibit evidence of pathological alterations (n=7) and the evidence is oten very subtle. he majority of instances are healed fractures on ribs (n=4) which are a result of wounding, possibly from red deer rutting, or possibly from a hunting incident. One of the large mammal ribs with a healed break just below the proximal head also exhibits an area of eburnation (polish) to the articular surface of the rib head. his is generally associated with bone rubbing on bone and is a diagnostic feature of osteoarthritis. Seasonality assessments he interpretation of seasonal occupation of the site is, for the most part, reliant on the ability to assess the age and development stage of the animal remains found there. By using information such as tooth development and epiphyseal fusion to age the faunal remains, it is oten possible to estimate the age at death, and from this the time of year these animals died through an understanding of breeding and birthing patterns. For example, red deer in Britain tend to begin mating in September/October each year and their gestation period is typically around eight months. he young are therefore born in May/June (Dobronika 1988). Combining this information with the data gathered from age assessment at death, it should be possible to approximate the season of death. Other methods used to interpret seasonality information include the development and shedding of deer antler (which today in Britain occurs between March and April), and the combination of age data and seasonal migration patterns of certain species such as birds. Stable isotope analysis Carbon (δ13C) and nitrogen (δ15N) stable isotope analysis was undertaken on ive faunal remains from the assemblage. One dog (Canis familiaris) and four red deer specimens were sampled; one of the latter was taken from a worked antler frontlet. Analysis followed a modiied Longin collagen extraction protocol using ultrailtration (30kDa MWCO) on c. 100–200 mg of bone (Brown et al. 1988; Richards and Hedges 1999; Colonese et al. 2015). he analysis was undertaken in order to contribute to the existing isotopic dataset for the site. Interest was particularly focused on comparison of the isotope values generated here with those previously obtained from the site (e.g. Schulting and Richards 2000; 2002a; 2002b; 2009), which have resulted in much debate surrounding potential movement of human populations between the coast and inland during the Early Mesolithic (Clutton-Brock and Noe-Nygaard 1990; Day 1996b; Schulting and Richards 2002a; Dark 2003; Schulting and Richards 2009). We also aimed to investigate the potential of any isotopic diferences between the individual elements of red deer and the antler frontlet. 202 Star Carr Volume 2 Samples were initially cleaned manually using a scalpel, and then were demineralised in 0.6 M aq. HCl solution at 4°C, and the resulting insoluble fraction was gelatinised in pH3 HCl for 48h at 80°C. he supernatant solution was then ultrailtered (30kDa MWCO, Amicon) to isolate the high molecular weight fraction, which was then lyophilised. Puriied collagen samples (1 mg) were analysed in duplicate by Elemental Analysis Isotope Ratio Mass Spectrometry (EA-IRMS) on a Sercon GSL analyser coupled to a Sercon 20-22 Mass Spectrometer at the University of York. Accuracy was determined by measurements of international standard reference materials within each analytical run. hese were IAEA 600 δ13Craw = -27.9 ± 0.1, δ13Ctrue = -27.8 ± 0.1, δ15Nraw = 0.6 ± 0.1, δ15Ntrue = 1.0 ± 0.2; IAEA N2 δ15Nraw = 20.5 ± 0.1, δ15Ntrue = 20.3 ± 0.2; IA Cane, δ13Craw = -11.8 ± 0.1; δ13Ctrue = -11.6 ± 0.1. he overall uncertainties on the measurements of each sample were calculated based on the method of Kragten (1994) by combining uncertainties in the values of the international reference materials and those determined from repeated measurements of samples and reference materials. hese are expressed as one standard deviation (1σ). In addition, a homogenised bovine bone extracted and analysed within the same batch as the samples produced the following values; δ13C = -22.9 ± 0.1; δ15N = 7.0 ± 0.2. he overall mean value among 50 separate extracts of this bone sample produced values of δ13C = -23.0 ± 0.7 and δ15N = 6.7 ± 0.4. Stable isotope values are presented here relative to the internationally deined standards of VPDB for δ13C and AIR for δ15N. Collagen quality fell within reported ranges (DeNiro 1985; van Klinken 1999). Collagen yields were calculated from retentate samples only, following ultrailtration. Variability was seen in the yields obtained from the samples, from 22.5% in the one dog sample, but ranging from 2–12% for the deer samples; however, all samples exhibited acceptable atomic C:N ratios of between 3.3–3.4. Taphonomic analysis Introduction Chapter 22 sets out the results of the bone degradation at Star Carr. In sum, the bone found on the dryland is very friable which has resulted in serious diiculties for the zooarchaeological assessment. It was possible to identify some specimens to taxa using ZooMS but other standard methods such as identiication of elements, anatomical distributions, sexing and ageing proved problematic. Some of the bone that does survive appears to have been quickly buried within the sediment following deposition. Any bone which had not been immediately buried is likely to have been subject to a variety of taphonomic factors such as clearing, dog-gnawing and trampling. It is impossible to quantify how much bone has completely deteriorated from this part of the site but it is highly likely that we are only seeing a very small percentage of what was originally there. As seen in Chapter 22, the bone from the wetland has been subjected to high levels of acidity though this varies in strength across the site. he bone found in Clark’s area was better preserved than elsewhere (the Clark backill assemblage and the material obtained during the most recent excavations), perhaps because the bone was deposited as a dump, which might have helped bufer the acidity, and it was also surrounded by less acidic/ near neutral backill. In addition, these bones were probably deposited in shallow water, meaning they were less prone to other destructive factors such as dogs and trampling. Preservation in other areas of the wetland varies signiicantly and the presence of ‘jellybone’ (see Chapter 22) suggests much bone has probably disappeared completely due to the high levels of acidity. Even in the small pocket of less-acidic sediment to the south of the detrital wood scatter it is important to note that severely deteriorated bone has been found (Chapter 22). he only evidence of natural taphonomic occurrences is the rounding and smoothing of the edges of bones. For the most part this is found within the waterlogged deposits of Clark’s area (n=18) and Clark’s backill (n=8) and is likely the result of water lowing over the bones. here are also four specimens from the dryland which exhibit similar rounding, and this will occur when sediment moves over exposed bone, or bone is let lying exposed on the ground surface for a while (Klein and Cruz-Uribe 1984). Anthropogenic factors he diferent types of taphonomy related to anthropogenic factors were examined by area (Table 23.3). In Clark’s area, most signs of modiication are present, bar heating, and here percussion breaks are present on 39% Faunal Remains: Results by Species 203 of the bone found. he prevalence of anthropogenic modiication in this area is most likely due to the large numbers of bone found as opposed to a real pattern: loss of bone elsewhere on site means the areas are not directly comparable. Spiral fractures and percussion breaks are found in all areas of the site and are the result of people using heavy objects, such as stones or rocks, to break into the central cavity of the bones in order to retrieve the marrow (Figures 23.2 and 23.3). Although it should be possible to achieve this goal by producing Figure 23.2: Photograph of an example of a percussion breakage, and clear percussion point, on the faunal remains from Star Carr (red deer radius <109242>) (Photograph taken by Paul Shields. Copyright University of York, CC BY-NC 4.0). Figure 23.3: An example of longitudinal splitting on a red deer tibia <110290> (Photograph taken by Paul Shields. Copyright University of York, CC BY-NC 4.0). 204 Star Carr Volume 2 only one break, there are several examples of fractures occurring to both ends of a long bone, removing both the proximal and distal articular ends. For these specimens it appears likely that ater the marrow was extracted, the long bone was prepared for tool production. Marks created by cutting or scoring are rarely found, with the majority of examples coming from Clark’s area, probably due to better conditions in this area. A large number were found on ribs (n=54): there was no clear patterning, with cut marks occurring along the full length of the rib, and they occur on the ventral and dorsal aspects, suggesting they represent a mixture of both skinning and dismembering activities. Cut marks have also been observed around the joints and along the midshat of long bones, providing evidence of muscle and ligament removal during dismemberment, and on the cortical bone surfaces of crania and antler frontlets, providing evidence of skinning. Bones which have been longitudinally split are found from most areas of the site, with the exception of the western platform and to the north of Clark’s area, though due to small numbers of bone in these areas this is unlikely to be a signiicant pattern. he percentage of split bone is always under 10%. his type of modiication was predominantly found on long bones (n=65; 90.3%), and is part of the process of bone tool manufacture. In terms of human modiication, Clark’s area exhibits all of the diferent types of human interaction with the most common forms being for marrow extraction (percussion breaks and spiral fractures) (Table 23.3). Both of these techniques can also been seen in all areas across the site. Interestingly, both Clark’s backill and the detrital wood scatter assemblages exhibit a similar range of activities to Clark’s area, but in smaller numbers. Evidence of heating (in the form of blackening, charring or calcination) is lacking from large areas of the site (Figure 23.4). Due to the conditions of the bone and the staining from the peat it is very diicult to identify signs of heating; however, a total of 500 fragments of bone (43 hand collected bones and 447 from lotation) exhibit evidence of heat exposure. Of these remains, the majority were found on the dryland (n=486), 11 were found in Clark’s area and three in Clark’s backill. Of the dryland heat afected specimens, the main concentration is from within and around the eastern structure (n=431). he afected specimens were found to exhibit evidence of having been heated to varying degrees (Table 23.4): burnt/blackened (n=84), charred (n=170) and calcined (n=246), suggesting ires of varying diferent heat intensities and durations. he material from the dryland mainly consists of bone that was intensely heat exposed (calcined), but there is evidence of bone from the full range of heat exposures. In terms of the material from Clark’s area, 11 fragments exhibit evidence of heating, and they also illustrate the full range of colour change. Clark’s Detrital wood scatter (n=160) Spiral fractures Percussion breaks 13 (8.1) Clark’s area (n=560) Wetland Clark’s backill (n=331) Central platform (n=24) Eastern platform (n=34) Western platform (n=20) Wood peat (n=157) Marl Bead Test Dryland (n=11) area pits (n=601) (n=13) (n=11) 99 (17.7) 36 (10.9) 3 (12.5) 6 (17.6) 3 (15) 11 (7) 3 (27.3) 3 (23.1) 6 (54.5) 33 (5.5) 18 (11.3) 219 (39.1) 66 (19.9) 6 (25) 9 (26.5) 3 (15) 14 (8.9) 2 (18.2) 5 (38.5) 2 (18.2) 12 (2) Cut marks 1 (0.6) 93 (16.6) 18 (5.4) Longitudinally split 2 (1.3) 23 (4.1) 15 (4.5) 1 (4.2) 3 (8.8) 2 (0.3) Groove-andsplinter 5 (3.1) 3 (0.5) 3 (0.9) 4 (16.7) 3 (8.8) 3 (15) 4 (2.6) Worked (antler) 5 (3.1) 2 (0.4) 5 (1.5) 1 (4.2) 2 (5.9) 1 (5) 1 (0.6) 8 (5.1) 1 (9.1) 17 (2.8) 2 (0.3) Table 23.3: Total number (and percentage) of the types of taphonomy exhibited for the excavated faunal remains within each area of the site. Faunal Remains: Results by Species § Legend Burnt bone 0 5 10 20 Unburnt bone 205 30 Metres Figure 23.4: Plot showing the number and distribution of the hand-collected heat-afected bone compared to unburnt bone (Copyright Star Carr Project, CC BY-NC 4.0). Area Feature Eastern structure Central hollow 49 148 175 Posthole [169] 1 15 28 Charred (black/grey) Calcined (white) Posthole [185] 3 Pit [177] 3 Surrounding area Central structure Burnt (blackened) Pit [336] 3 2 19 4 6 Posthole [382] Surrounding area Northern structure 1 Posthole [358] Posthole [459] Surrounding area 1 1 1 West of northern structure Posthole [462] Western structure 9 Central hollow 2 Surrounding area 1 Clark’s area Clark’s baulk 7 3 14 Clark’s backill Backill 1 2 1 Table 23.4: Number of fragments of bone exhibiting evidence of heat exposure (excavated and bone recovered from lotation). 206 Star Carr Volume 2 Clark’s Detrital wood scatter (n=160) Clark’s area (n=560) Wetland Clark’s backill (n=331) Tooth impressions 1 (0.6) 14 (2.5) 7 (2.1) Tooth scores 3 (1.9) 12 (2.1) 4 (1.2) Uneven breakage 4 (2.5) 20 (3.6) 2 (0.6) Central platform (n=24) Eastern platform (n=34) Western platform (n=20) 1 (4.2) Wood Marl peat (n=11) (n=157) Bead Test pits Dryland area (n=11) (n=601) (n=13) 1 (7.7) 1 (9.1) 1 (7.7) 1 (9.1) 1 (7.7) 1 (9.1) Table 23.5: Total number (and percentage) of the animal modiication exhibited for the excavated faunal remains within each area of the site. he focus of these heat afected fragments of bone around the diferent structures on the dryland is very interesting. he eastern structure contains the most evidence, particularly within the central hollow, suggesting the presence of a hearth within the structure or possibly distribution of ire ash across the structure loor during hearth clearance. he other possibility is that the structure may have burnt down; however, this probably would have resulted in greater quantities of burnt material in this area. he heat-afected bone may also be the result of cooking, but gentle heating for marrow extraction purposes, bone waste disposal and accidental heating cannot be discounted. Modiications by animals hree types of modiication by animals have been observed: tooth impressions, tooth scores and uneven breakage (which is oten associated with one or both of the other two characteristics). A total of 50 specimens exhibit these modiications and they are found across the entire site (Table 23.5). Long bones are the most common elements afected (n=19), along with ribs (n=9) and podial elements (n=8). he remainder are all represented by elements that could be considered to be waste products of butchery and are not meat-bearing (vertebrae, pelvis, scapula), so may be easily scavenged from a processed carcass. Interestingly, one red deer antler frontlet from Clark’s area (<116888>; Chapter 26) also exhibits tooth impressions and score marks consistent with a small amount of carnivore activity. he majority of modiications were consistent with wolf or dog gnawing: chewing concentrated on the ends of long bones (Figure 23.5) or the edges of fragments, and on elements at the joints such as the podial elements; and tooth impressions and tooth scores at the edges of elements (Figure 23.6) (Shipman 1981, in Lyman 1994). here is one exception: a wild cat humerus <116175> has tooth impressions around the edge of the proximal articular surface consistent with feline teeth, possibly another wild cat, or mustelid teeth. he detrital wood scatter, Clark’s area, north of cutting III (the bead manufacturing area), test pits and Clark’s backill all provide evidence for all three types of modiication, with additional evidence of uneven breakage from the central platform. Given that dog bones have been found on site, it is not surprising that such evidence exists. he lack of evidence from other areas is more likely to be a factor of taphonomy due to deterioration than a real pattern: the numbers of bone found in the wetland is relatively small and the bone from the dryland is badly degraded and unlikely to show such modiications. he fact that this evidence does survive in some places suggests that bone was subject to animal gnawing, and there is a strong likelihood that much bone was destroyed during the occupation of the site through this process. Discussion A number of diferent factors have inluenced the preservation of bone from Star Carr. he damage caused by the acidiication of the deposits is impossible to quantify; however, because a number of ‘jellybones’ have been found and some areas have produced very small quantities of faunal remains (such as the western platform), the likelihood is that signiicant quantities may have been lost in some areas. Faunal Remains: Results by Species 207 Figure 23.5: Photograph of the uneven (ragged-edge) breakage associated with carnivore gnawing on the proximal head of an ulna <109243> found in the area of the detrital wood scatter (Photograph taken by Paul Shields. Copyright University of York, CC BY-NC 4.0). Figure 23.6: Photograph of carnivore gnawing on a red deer navicular-cuboid <108662>, from the detrital wood scatter, with clear tooth impressions (on the articular surface) (Photograph taken by Paul Shields. Copyright University of York, CC BY-NC 4.0). 208 Star Carr Volume 2 On the dryland there is also evidence of severe deterioration of the bone because here the site is not waterlogged; this is fairly normal for dryland sites of this period, and the level of degradation is likely to be more or less consistent across the area. However, other factors caused by both humans and animals, such as fragmentation through trampling, working bone for tools and processing it for food, and gnawing by dogs, all have further negative impacts on preservation. From the data it is not possible to assess how signiicant these actions were, but they are likely to have had an impact across the site. In sum, as with most sites, we have to interpret the faunal remains data knowing that we are looking at a very partial record. Not only have we excavated less than 10% of the site (potentially missing material deposited in other areas), but the assemblage has been partially destroyed by acidity in the wetland, oxidation on the dryland, carnivores and a multitude of human actions in the Mesolithic period related to consumption practices, tool making and deposition of remains in various parts of the landscape. It is with these caveats in mind that the faunal remains analysis was undertaken. Results Overall quantiication NISP by taxa A total of 2414 specimens of bone and antler were found from the site (Figure 23.7). Of these, 1925 archaeological bone and antler specimens were recovered through excavation on site (Figure 23.8), and 489 small fragments were recovered from lotation of soil samples. A further 12 have not been included in the analysis since they are undoubtedly intrusive (mole, rabbit, modern cattle, modern dog), or were found within re-deposited upcast from the nearby River Hertford. Figure 23.7: NISP of the taxa for the whole site (hand excavated and lotation) (Copyright Star Carr Project, CC BY-NC 4.0). Faunal Remains: Results by Species 209 Figure 23.8: NISP for taxa from the hand-excavated faunal remains (not including lotation remains) (Copyright Star Carr Project, CC BY-NC 4.0). A total of 16 species were identiied, 12 of which had been found previously, and four of which were new: wild cat (Felis silvestris), ield vole (Microtus agrestis), northern pike (Esox lucius) and European perch (Perca luviatilis). he ish species are particularly signiicant because of the debate concerning the apparent lack of ish remains at Star Carr (Wheeler 1978). It is also noteworthy that microfaunal remains such as ield vole, which have never been recovered before from Star Carr, have now been discovered through lotation. he NISP values illustrate the dominance of red deer remains within this assemblage, followed by aurochs, and then roe deer (Figure 23.8). Dog is also dominant; however, this is represented by one almost complete skeleton which has the efect of skewing the results. In contrast, several species were represented by only a small number of remains: wolf, wild cat, pine marten, beaver, common crane, red- or black-throated diver and white-fronted or bean goose. he NISP data also shows large numbers of cervid species, large mammals and unidentiied bones (Figures 23.7 and 23.8). here were a total of 510 specimens that could only be categorised as medium or large mammals due to fragmentation and poor preservation: 506 excavated and four found during lotation. Of the 510 specimens, 375 were identiied as belonging to large mammals (374 excavated and one from lotation) and 135 as medium mammals (132 excavated and three from lotation). here were also 162 specimens that could only be identiied as cervid species (Figure 23.8). A further 731 could not be assigned to either species or general category and so were labelled ‘unidentiied’ (290 from excavated deposits and 441 fragments from lotation) (Figure 23.7). Of the 489 fragments of bone found through lotation, the majority of specimens were found from the eastern structure, the central structure, the possible northern structure and Clark’s area. Seven fragments can be identiied to species: perch, northern pike, pine marten, red deer and ield vole (Figure 23.9). here are also a number of fragments that can only be identiied to family, genus or medium/large mammal (n=41) and the majority (n=441) were unidentiied, either due to the small size of the fragments or due to taphonomic processes afecting the preservation. he category of ‘cervid species’ includes specimens that could be elk, red deer or roe deer. Of the elements identiied only as cervid species, a large number were identiied by ZooMS, mainly due to the bone fragments being either too poorly preserved or too small to identify macroscopically. his is relected by the large number of specimens that were identiied as ‘cervid species’ by ZooMS but could not be identiied to element (n=78) (Figure 23.10). 210 Star Carr Volume 2 Figure 23.9: Number of fragments recovered by lotation and sieving of soil samples and identiied to taxa (Copyright Star Carr Project, CC BY-NC 4.0). Figure 23.10: NISP of the elements for the cervid species category (Copyright Star Carr Project, CC BY-NC 4.0). Faunal Remains: Results by Species 211 he category of ‘large mammal’ includes specimens where it was not possible to distinguish between elk, red deer and aurochs. his category (Figure 23.11) was represented by a range of elements, with the majority identiied as ribs (n=173) and long bone fragments (n=65). A similar pattern can be seen for the elements that can only be identiied as ‘medium mammal species’ (Figure 23.12). his category includes roe deer, wild boar, dog, wolf and wild cat. A range of elements can be identiied for this category, the most dominant being ribs (n=83) followed by long bone fragments (n=18). Although not much can be said about these three broad categories due to the large number of taxa they include, it is important to note the abundance of ribs and long bones, suggesting some of these elements will be missing from the taxa they represent: ignoring these elements when discussing the individual species creates a biased view of the assemblage. Unfortunately this is likely to have been the case in past analyses for this dataset (e.g. Clark 1954; Legge and Rowley-Conwy 1988), due to the large proportion of these types of elements that were found in Clark’s backill. In terms of the lotation specimens, a range of diferent elements were represented with the majority being identiied as fragments of loose teeth (Figure 23.13) from species of rodent (ield vole: n=3, rodent species: n=5, vole species: n=3), ish species (cyprinid species: n=2, pike/salmon species: n=3), carnivore species (n=1) and red deer (n=1). In terms of the vertebrae fragments, most belong to cyprinid species (n=4) and ish species (n=3), with one identiied as perch. he majority or ribs are identiied as rodent species (n=3) but also ish species (n=2), cyprinid species (n=1), mustelid species (n=1) and medium mammal (n=1). Figure 23.11: NISP of the elements for the large mammal category (excluding lotation data) (Copyright Star Carr Project, CC BY-NC 4.0). 212 Star Carr Volume 2 Figure 23.12: NISP of the elements for the medium mammal category (excluding lotation data) (Copyright Star Carr Project, CC BY-NC 4.0). Figure 23.13: Number of fragments of bone identiied to element from the lotation samples (Copyright Star Carr Project, CC BY-NC 4.0). Faunal Remains: Results by Species 213 Results by species Fish In total, 21 ish remains have been recovered from the site; however, it is likely that these represent only a fraction of the overall assemblage. In Clark’s area, there was better preservation than elsewhere allowing for the recovery of six ish remains. Two of these were found by hand during the excavation of the reed peat (312) and were located approximately 5 m away from one another (Figure 23.14; Figure 23.15). he remainder were recovered through bucket lotation in the laboratory from basal deposits in Clark’s area and from the ill of the hollow of the eastern dryland structure (context 149). Of the 21 remains, 14 could be identiied to family or species (Table 23.6). he assemblage was dominated by Cyprinidae (the majority of British freshwater ish species including carps and minnows), followed by northern pike/Salmonidae (Esox lucius L., 1758/salmons, trouts, chars and whiteishes), northern pike and European perch (Perca luviatilis L., 1758). here were 12 postcranial elements and nine cranial elements. Although identiication was attempted to genus and species, 11 specimens from two families could not be further identiied. Of these two families, two Salmonidae species and nine Cyprinidae species have been previously identiied in contemporaneous faunal assemblages throughout Northern Europe (Aaris-Sørensen 1976; Richter 1982; Zabilska-Kunek 2014; Zabilska-Kunek et al. 2015): brown trout (Salmo trutta L., 1758), Atlantic salmon (Salmo salar L., 1758), white bream (Blicca bjoerkna L., 1758), common bream (Abramis brama L., 1758), Crucian carp (Carassius carassius L., 1758), common carp (Cyprinus carpio L., 1758), asp (Aspius aspius L., 1758), Trench Context Skeletal element Taxon SC23 149 Unknown vertebra cf. Cyprinidae SC23 149 Pharyngeal tooth Cyprinidae SC23 149 Caudal vertebra Cyprinidae SC23 149 Rib Cyprinidae SC23 149 Caudal vertebra Cyprinidae SC23 149 Premaxilla Esox lucius SC23 149 Tooth Esox lucius/Salmonidae SC23 149 Tooth Esox lucius/Salmonidae SC23 149 Tooth Esox lucius/Salmonidae SC23 149 Vertebral fragment Perca luviatilis SC23 149 Vertebral fragment Unidentiiable SC23 149 Vertebral fragment Unidentiiable SC23 149 Vertebral fragment Unidentiiable SC23 149 Rib/spine Unidentiiable SC23 149 Rib/spine Unidentiiable SC34 312 Pharyngeal bone with tooth Cyprinidae SC34 312 Pharyngeal tooth Cyprinidae SC34 312 Posterior abdominal vertebra Esox lucius SC34 Basal deposits cf. posterior abdominal vertebra Cyprinidae SC34 Basal deposits Unknown Unidentiiable SC34 Basal deposits Unknown Unidentiiable Table 23.6: Contextual information and skeletal elements of the various ish identiied in the assemblage. 214 Star Carr Volume 2 0 2 4 8 Isolated Fish Bone Finds No. of Fish Bone Present: 0 1 5 No. of Fish Bone Present: 0 15 Figure 23.14: Location of the ish remains on site (Copyright Star Carr Project, CC BY-NC 4.0). 12 16 20 Metres § Faunal Remains: Results by Species 215 Figure 23.15: Photograph of the northern pike posterior abdominal vertebra in situ. Scale: 9.1 mm across the greatest cranio-caudal length of the centrum (Copyright Harry Robson, CC BY-NC 4.0). ide (Leuciscus idus L., 1758), roach (Rutilus rutilus L., 1758), common rudd (Scardinius erythrophthalmus L., 1758) and tench (Tinca tinca L., 1758). he identiied species spectrum consisted of freshwater ish. Northern pike and European perch are oten termed stationary freshwater ish although they can also reside in weakly brackish waters. In general, they commonly occur in stagnant or gently lowing reaches of a river (Brinkhuizen 2006). An estimation of total ish length (TL) was attempted for all specimens that were identiied to the family or species. By estimating TL it is possible to determine whether or not the assemblage was anthropogenically or naturally derived. For example, if numerous species ranging in size are represented in a given assemblage, an argument in favour of a natural death assemblage can be put forward. However, if the assemblage is dominated by one species that are generally similar in size, the assemblage can be interpreted as anthropogenic, and the data probably represents the ‘selective killing of ish of a certain size’ (Noe-Nygaard 1995, 170). In addition, since TL is inter- and intra-species speciic, size estimates can add weight to seasonality. For instance, the European eel (Anguilla anguilla) is sexually dimorphic; thus if a size frequency diagram demonstrates that the majority of eels were over 0.55 m in total length, then it can be argued that females were probably targeted during their autumnal migrational run (since males do not exceed 0.5 m) (Tersch 2003). Given the incompleteness of the specimens it was only possible to estimate the TL for seven of the specimens. Based on comparison with modern skeletons of known taxa, one northern pike specimen was estimated to be <0.2 m in TL, whilst the posterior abdominal vertebra was estimated to have derived from an individual with a TL of c. 0.7 m (Robson et al. 2016). In addition, it was possible to estimate the TL for four of the Cyprinidae specimens. Based on comparison with modern skeletons of known taxa, these are estimated to have derived from specimens that were <0.2 m in TL. Furthermore, it was estimated that the one European perch specimen in the assemblage was derived from an individual that was <0.1 m in TL (Robson et al. 2016). Whilst preservation on site was variable, it is unlikely to have impacted TL. he one pike posterior abdominal vertebra was in a very good state of preservation and had not been subjected to compression or warping. 216 Star Carr Volume 2 On the other hand, the majority of the ish remains recovered from within the structure were calcined, which is likely to have afected their structure and size. For these reasons broad estimates (i.e. 0.1 m increments) were employed. Overall, the majority of the ish remains derived from small individuals, <0.2 m, with the exception of the pike posterior abdominal vertebra. Although the sample size is very small, the ish sizes coupled with the microwear analysis on the lint (Chapter 8; Robson et al. 2016) demonstrate that the assemblage was anthropogenically derived. Moreover, the data are comparable with the assemblage from the broadly contemporary Early Mesolithic site of Friesack IV (Robson 2016). Birds Overview Fraser and King (1954) did not state the NISP values for birds present in the assemblage, though it was noted that not more than one individual of each species was represented. he current analysis found 20 bird specimens (2 from lotation): common crane (n=4), red-/black-throated diver (n=2), white-fronted/bean goose (n=1) and 13 specimens that can only be identiied as belonging to bird species (n=7), large bird species (n=4), medium bird species (n=1) and small bird species (n=1). Of the 20 bird specimens, the majority were found within Clark’s area (n=12), with the remainder being found within Clark’s backill (n=4), the dryland (n=3) and a 2005 test pit SC10 (n=1) (Figure 23.16). It should be noted that during Clark’s excavations, a section cut through a bird bone was found, thought to probably represent a bead (Clark 1954, 164). Of the 20 bird bones found here, seven demonstrate signs of human modiication: six demonstrate evidence of percussion breaks or spiral fractures; the seventh was found on the dry land and was charred. One of the large bird bones has a cut mark. he breakage of these bones is ! ( ! ( ! ! ( ! ( ( ! (! ! (! ! ( ! ! Legend ! Common crane ! ( Red or black throated diver ! ( White fronted or bean goose ! ( Tetraonidae sp. ! ( Large bird sp. ! ( Medium bird sp. ! ( Birds sp. 0 5 10 20 Metres § Figure 23.16: Spatial plot of the birds found at Star Carr within the main area of excavation (Copyright Star Carr Project, CC BY-NC 4.0). Faunal Remains: Results by Species 217 unusual because there is no marrow in bird bones, so there is the possibility that they are being prepared for bead manufacture. here is also evidence of what appears to be a healed break to the midshat of a large bird humerus, which could have occurred in a number of ways; for example related to human action, carnivore action or as a natural accident. White-fronted or bean goose he one specimen of goose was found in Clark’s area (Figure 23.16). It is identiied as the midshat of an ulna, but due to the partial nature of this element it is not possible to identify the species and it could derive from either a white-fronted or bean goose. he specimen was humanly modiied: both the proximal and distal articular ends of the specimen are lacking and the breakage is very uneven and ragged, possibly suggesting a percussion and snapping action. It is not possible to age and sex this specimen. In terms of seasonality, both of these species of geese are migratory birds: today, these geese are only present in this country in the winter but Early Holocene distributions may have been markedly diferent given the climate data for this site (Chapter 18). Red- or black-throated diver Red-throated diver was found in the original assemblage (Clark 1954, 70). he current analysis produced two specimens of diver, consisting of a partial humerus and radius, from within Clark’s area (Figure 23.16), but due to the partial nature of the elements it is not possible to identify to species. Both elements exhibit human modiication: the humerus represents the proximal half of the element and the distal end has been removed by a percussion break (Figure 23.17); the radius represents the distal half and the proximal end has been removed by a spiral fracture. It is not possible to age or sex these remains. In terms of seasonality, both species of diver are migratory birds and today both species can be found in Scotland and Northern Ireland during the summer months for breeding and around the UK coastline in winter; however, again it is diicult to assess the Early Holocene distribution. Common crane Common crane was found in the original assemblage (Clark 1954, 70). A further four specimens were found within Clark’s area (Figure 23.18): a partial radius, one complete and one partial carpometacarpus, and a frag- Figure 23.17: Removal of the distal end of diver humerus <116486> by a percussion break (Photograph taken by Paul Shields. Copyright University of York, CC BY-NC 4.0). 218 Star Carr Volume 2 Figure 23.18: Spiral fracture to one end of a common crane tibiotarsus midshat <115281> (Photograph taken by Paul Shields. Copyright University of York, CC BY-NC 4.0). ment of tibiotarsus. he MNI is one. Two of the specimens appear to be humanly modiied; one carpometacarpus is missing both its proximal and distal articular ends and the breakage appears to have been caused by percussion, and the fragment of tibiotarsus midshat has spiral fracture breakage to one end. Unfortunately, due to the fragmentary nature of the elements represented, it is not possible to age or sex these remains. In terms of seasonality, common crane became extinct in the UK in the 1600s, but in recent decades European birds have begun to repopulate East Anglia (Brand, pers. comm. 2016). hese populations are not migratory; however, cranes on the continent migrate and spend summer in the north and winter in the south (Svensson et al. 2008). Carnivora Wild cat Given the chronological period in which these remains were associated, it was possible to discount domestic cat and therefore six specimens of wild cat were found. Four of these were recovered from Clark’s area and two from the dryland (Figure 23.19). he two elements from the dryland are distal phalanges, both of which are burnt and calcined, and were found next to a spread of burnt lint on the peripheries of the western structure. he wild cat assemblage found in Clark’s area consists of a right humerus and radius, a sacrum and a second metacarpal (Figure 23.19), which might even represent one leg, particularly as these remains were found no more than 1.5 m apart at the northern end of the baulk. It is not possible to ascertain whether the phalanges on the dryland are part of the same animal but given that the western structure may date to the same period as the deposition in Clark’s area the MNI is one (Figure 23.20). he wild cat remains from Clark’s area are very well preserved. On the proximal head of the humerus, there are some very subtle tooth marks around the edge of the articular surface (Figure 23.21). he size and shape of these tooth marks seem to suggest that they may have been made by another wild cat or a mustelid. he partial radius appears to have sufered from a possible percussion break; however, modern damage has unfortunately obscured the clarity of this modiication. Faunal Remains: Results by Species 0 5 10 20 30 Metres 219 § Figure 23.19: Spatial plot of wild cat (Copyright Star Carr Project, CC BY-NC 4.0). Figure 23.20: Element representation for wild cat (Copyright Archeozoo.org/M. Coutureau 1996. Adapted by Becky Knight). 220 Star Carr Volume 2 Figure 23.21: Photograph of a tooth mark to the edge of the proximal head of the humerus <116175> (Photograph taken by Paul Shields. Copyright University of York, CC BY-NC 4.0). Wolf One wolf specimen was found in Clark’s area (Figure 23.22), and this was identiied as a right second metatarsal from an adult animal. his specimen is attributed to wolf on the basis of its size and length-breadth proportions, considering the size and morphology of other canid remains from the site. he distal head is missing, and the breakage appears to have been caused by percussion damage; however, there is also a small amount of excavation damage to the anterior break edge. Not much can be inferred from this one element, though it is of interest because no other wolf remains have been found from the site: the remains identiied as wolf from Clark’s excavations were later reassigned to dog. Dog Overview An almost complete skeleton of a dog was found within the peat above the marl (Figure 23.23). Although it was found at the base of the wood peat (310), the date obtained on the let canine (OxA-33678; Figure 17.16) dates it to the 90th century cal. BC (Chapter 17). Unfortunately the preservation of the remains is poor with demineralisation and compression of the bones and teeth; in particular, the teeth enamel was splitting and peeling away from the root. Elements he majority of the elements were represented (Figure 23.24; Figure 23.25); however, some of the elements from the extremities of the animal were missing, including the majority of the podial elements. Given that these are the smaller, more delicate elements, this may be as a result of the process of degradation. In addition, some elements may have been moved by water at the time of decomposition. he skeleton was positioned in two separate groupings about 0.5 m apart. Some of the elements of the two groupings were still found to be semi-articulated: from group 1 the humerus, radius and ulna, and the vertebrae were semi-articulated; within group 2, the humerus, radius and ulna, and the femur and tibia were semi-articulated (Figure 23.26; Table 23.7). 0 5 10 20 30 Metres § Figure 23.22: Spatial plot of wolf (Copyright Star Carr Project, CC BY-NC 4.0). 0 5 10 20 30 Metres Figure 23.23: Spatial plot of dog (Copyright Star Carr Project, CC BY-NC 4.0). § 222 Star Carr Volume 2 Figure 23.24: NISP of elements represented in the dog skeleton (Copyright Star Carr Project, CC BY-NC 4.0). Figure 23.25: Element representation for dog (Copyright Archeozoo.org/M. Coutureau and V. Forest 1996. Adapted by Becky Knight). Faunal Remains: Results by Species 223 Figure 23.26: Photograph of the in situ dog skeleton (group 1= let, group 2= right) (Copyright Star Carr Project, CC BY-NC 4.0). Group 1 Mandible 2 Cranium 1 Premolar 1 1st molar 1 Humerus 1 Scapula 1 Radius 1 Humerus 1 Ulna 1 Radius 1 Atlas 1 Ulna 1 Axis 1 Pelvis 1 Cervical vertebrae 9 Femur 2 Tibia 2 Calcaneus 1 horacic vertebrae 4 Group 2 Ribs 22 Pelvis 1 Talus 1 Sacrum 1 Metatarsal 5 Phalanges 3 Metapodials 4 Table 23.7: NISP and element distribution of the dog skeleton between the two separate groupings of bone. 224 Star Carr Volume 2 his assemblage does not look as if the skeleton has split in half because each group of elements contains a mix of front and back parts of the dog. Given that group 1 is more tightly packed and in basic anatomical position, it seems likely that the bones in group 2 have moved to some degree: here the cranium and some forelimb bones are found with some hind limb bones. his is most likely to be water action: the water would have been lowing in that direction, out of the lake and towards the west. Given the context (shallow water reed swamp environment) and examination of the sediment on site, it is unlikely to have been a formal burial. he dog may have died a natural death in the lake or was placed within the water once it had died. Age Only the mandible is complete enough to aid with the age estimation: although the maxilla is present, a large number of teeth are either missing or are too degraded to be assigned. he let side of the mandible contains complete adult dentition. However, the right side contains only partial dentition: the incisors are missing and the third molar has not erupted. It is not possible to make any observations on the tooth wear due to the demineralisation and delamination of the enamel. Although the third molar from the let mandible is fully erupted, the third molar from the right side is still in the crypt. herefore, based on the dentition, this animal would have been between six and seven months old at death (Silver (1969) and Habermehl (1961) in Hillson (2005)). Due to a combination of the individual being immature and the poor preservation of the bones, it was not possible to assess the size or robustness of the animal. Isotope values Due to the lack of human remains on British Mesolithic sites, dogs have previously been used as an analogue for both human diet and movement (e.g. Noe-Nygaard 1988; Fischer et al. 2007; Guiry 2012). his has extended to dog specimens from the Vale of Pickering both at Star Carr and Seamer Carr (e.g. Clutton-Brock and Noe-Nygaard 1990; Day 1996b; Schulting and Richards 2002a; Dark 2003; Schulting and Richards 2009). Both the initial δ13C and subsequent δ13C and δ15N isotopic analysis of dog remains from Seamer Carr yielded values suggested to indicate some degree of marine resource consumption, and therefore potentially relective of seasonal movements from the coast to the Vale of Pickering (Clutton-Brock and Noe-Nygaard 1990; Schulting and Richards 2002a), although this interpretation has resulted in signiicant debate (Day 1996b; Dark 2003; Schulting and Richards 2009). Conversely, isotopic analysis of dog remains from Star Carr have thus far indicated no evidence of the consumption of marine resources, and instead have been suggested to relect a diet based upon terrestrial resources, with some possible freshwater protein input (Schulting and Richards 2002a; 2009). One let rib sampled from the skeleton found during this excavation yielded suicient amounts of collagen of suitable quality for δ13C and δ15N stable isotope analysis (Table 23.8) and exhibited isotopic values comparable to those previously reported for dog remains from Star Carr (Schulting and Richards 2002a; 2009). hese results also indicate that this dog was unlikely to have consumed signiicant quantities of marine protein (Table 23.8; Figure 23.27). he elevated δ15N values of the dog remains from Star Carr have previously been interpreted as being indicative of a degree of freshwater resource and/or aquatic bird consumption, or possibly of low levels of marine protein (Schulting and Richards 2002a). he δ15N value of 10.5‰ for the newly excavated Star Carr dog falls in line with these previous interpretations. In particular, when we consider the available isotope values for terrestrial herbivores from Star Carr (Table 23.11; Schulting and Richards 2009), it can be seen that the dog values fall more than a trophic level (3–5‰) above these, therefore indicating that there must be additional protein source(s) in the diets of the dogs. Given the isotopic data already available for the site, it seems most probable Sample No. Element 108261 Rib δ13C (‰) −20.3 ± 0.1 δ15N (‰) 10.6 ± 0.2 Table 23.8: Carbon and nitrogen stable isotope data of the dog. C:N Collagen yield 3.4 22.5% Faunal Remains: Results by Species 225 Figure 23.27: Dog isotope data from several British Mesolithic sites (data compiled from Schulting and Richards 2000; 2002a; 2002b; 2009; this study) (Copyright Star Carr Project, CC BY-NC 4.0). that the Star Carr dog analysed here consumed a non-marine diet which comprised a degree of freshwater resource consumption. his therefore also lends weight to the hypothesis put forward by Schulting and Richards (2009) that movements to and from the Vale of Pickering and the coast were possibly not a particularly regular occurrence. Alternatively, perhaps, dogs did not oten accompany people to the coast. Pine marten Pine marten was found in the original excavations (Fraser and King 1954, 71) and included cranial elements, long bones and ribs, with an estimation of at least two animals. A total of seven specimens of pine marten were recovered from the recent excavations: six from Clark’s area and one from the detrital wood scatter (Figure 23.28). he element from the detrital wood scatter is the right side of a mandible, and from Clark’s area there is a let radius and ulna, one lumbar vertebra and one caudal vertebra, the let half of a pelvis and a let tibia (Figure 23.29). Although all of the remains are well preserved, there is no clear evidence of human or animal modiication. It is not possible to combine the data from the original excavations with those from the recent excavations because not all the elements were sided in Fraser and King’s report. In terms of the recent excavations, the MNI is one, as all the remains represent fully developed elements; however, as one specimen derives from the detrital wood scatter which is dated to much earlier than Clark’s area (Figure 17.20), the MNI can be adjusted to two. Artiodactyl Wild boar Overview Wild boar was identiied in the original assemblage (Clark 1954, 74) and re-examined by Legge and Rowley-Conwy (1988) who counted 22 fragments and estimated an MNI of four. In the recent excavations, a 226 Star Carr Volume 2 0 5 10 20 30 Metres § Figure 23.28: Spatial plot of pine marten (Copyright Star Carr Project, CC BY-NC 4.0). Figure 23.29: Element representation of pine marten (Copyright Archeozoo.org/M. Coutureau 2015. Adapted by Becky Knight). Faunal Remains: Results by Species 0 5 10 20 30 Metres 227 § Figure 23.30: Spatial plot of wild boar (Copyright Star Carr Project, CC BY-NC 4.0). further 22 specimens have been found across the site (Figure 23.30): three on the southern edge of the eastern platform, three in the detrital wood scatter, one above the western platform, four in Clark’s area, a concentration of eight around the eastern dryland structure and two on the peripheries of the western dryland structure. Six specimens originating from the dryland have been identiied to species through ZooMS but cannot be identiied to element. Elements Mandibles and loose teeth, scapulae and hind leg bones have been identiied but the small number of specimens make it diicult to identify any signiicance in their spatial patterning (Figures 23.31 and 23.32). In terms of missing elements, the cranium, torso and front limbs were almost completely absent, apart from two scapulae (Figure 23.31). his is very similar to the results of Legge and Rowley-Conwy (1988, table 1E). However, it should be noted that there are a large number of rib elements that can only be identiied as belonging to medium mammals, some of which may be wild boar, and therefore we cannot necessarily assume that torso elements are missing. Age and sex Two specimens were used to aid in ageing the wild boar: an unerupted irst incisor and an unfused femur (Table 23.9). he irst incisor was a loose tooth ind, it had no wear on the occlusal surface and it appears 228 Star Carr Volume 2 Figure 23.31: Element representation of wild boar (Copyright Archeozoo.org/M. Coutureau 2003. Adapted by Becky Knight). Figure 23.32: NISP values of wild boar (Copyright Star Carr Project, CC BY-NC 4.0). Element NISP Age range First incisor 1 14–16 months Femur (prox.) / (dist) epiphyses 1 36–42 months Table 23.9: Wild boar maximum age ranges when eruption and fusion are complete. Tooth development based on Briederman (1965) and Matschke (1967), and bone fusion based on Lesbre (1897–8). Faunal Remains: Results by Species 229 to be only partially developed and is likely to have been an unerupted tooth. he femur is only represented by the midshat, which is gracile and small and missing both the proximal and distal epiphyses. he lack of any evidence for fusion having begun on the femur suggests that this element belonged to an animal less than 3.5 years of age: the small size cannot be explained by sexual dimorphism as bone development is incomplete. However, it should be noted that due to the small amount of information available about the timings of epiphyseal fusion in pigs and wild boar, estimations of age from this method should be treated with caution. he other wild boar remains all appear to be fully developed and the scapulae and metapodials fully fused (with the fusion lines obliterated). his corresponds to the two specimens which Legge and Rowley-Conwy (1988, 44) noted as being from dentally mature animals. Overall, there appears to be both young adult and adult wild boars in the assemblage. It was not possible to sex the majority of the remains; however, the presence of a partial mandible with large canines (which was found amongst the timbers of the western platform; Chapter 7) would suggest that the remains of at least one adult male are present at the site. Modiication here is no evidence for modiication by carnivores on these specimens. In Clark’s area, of the four specimens found, three show evidence of human modiication. he mandible has been broken using a percussion break, beneath the tooth row, for marrow extraction or to remove the canine. he distal end of the second phalanx has been removed by a percussion break for marrow extraction. he scapula exhibits ephemeral cut marks around the posterior aspect of the glenoid, which is likely to be the result of cutting through ligaments to separate the forelimb (a major meat-bearing limb) from the carcass. From around the eastern platform there are two third and one fourth metatarsal which exhibit percussion breaks at the distal ends; however, these are small elements with very little marrow. MNI he most dominant elements within the assemblage are metatarsals and loose teeth. Based on metatarsals an MNI of one is calculated; however, taking into account the age proiles provided by the incisor and femur, an MNI of two can be posited: a juvenile piglet of under a year and a fully developed adult. As the juvenile incisor was found on the dryland around the eastern structure, this is less likely to correspond with the juvenile femur from the detrital wood scatter, which could increase the MNI to three. However, as the dating of the dryland has a degree of uncertainty to it, the MNI should remain at two. If the MNI results are added to Legge and Rowley-Conwy’s (1988, 9) MNI of four (from the let mandibles), the presence of a juvenile increases the overall MNI for the site to ive. Elk Overview Within the waterlogged areas, the elk remains were generally spread across most of the trench, with the majority located in Clark’s area (Figure 23.33). Elk was identiied in the original assemblage (Clark 1954, 76–79) and re-assigned by Legge and Rowley-Conwy (1988) who noted 247 fragments and an MNI of 12. In comparison, only 34 specimens have been identiied from across the site during the recent excavations (including six from Clark’s backill). 230 Star Carr Volume 2 0 5 10 20 30 Metres § Figure 23.33: Spatial plot of elk specimens (Copyright Star Carr Project, CC BY-NC 4.0). Figure 23.34: Element representation of elk (Copyright Archeozoo.org/M. Coutureau and J. Treuillot 2013. Adapted by Becky Knight). Faunal Remains: Results by Species 231 Figure 23.35: NISP for elk (Copyright Star Carr Project, CC BY-NC 4.0). Elements A range of elements are represented; however, the majority belong either to the skull or are leg bones (Figure 23.34; Figure 23.35). here is a distinct lack of long bones and also elements associated with the torso (such as vertebrae, ribs and pelvis), although as noted previously, most of the ribs found at the site have not been assigned to species. his pattern of element representation is diferent to the patterning observed by Legge and Rowley-Conwy (1988, table 1B) where most elements of the skeleton were found; however, their sample size was much larger. Age and sex In terms of the age proile for the elk remains, only one element possesses developmental information. he specimen, a calcaneal tuberosity epiphysis, is unfused and missing on the calcaneus of one elk specimen. According to Purdue (1983) this element fuses between the ages of 1.6–2.4 years, and so this gives a maximum age of 2.4 years for this animal. Legge and Rowley-Conwy (1988, 44) also noted that of the 10 elk jaws found, nine retained third molars, and in seven cases these were at a relatively early wear stage suggesting a high proportion of young adults in the cull. It was not possible to sex the remains. Legge and Rowley-Conwy (1988, 63) also found this because the animals, in early dental maturity, exhibit little dimorphism. However, the inclusion of elk antler within this assemblage identiies that male animals are present. Modiications here are no signs of carnivore modiication with these specimens. In total 16 specimens have evidence of human modiication. Within the detrital wood scatter an astragalus exhibits a small round hole suggestive of a possible projectile puncture wound. In addition, a palmate portion, one antler fragment and metacarpal have been humanly modiied but only the metacarpal is modiied by both a percussion break and spiral fracture, likely for marrow extraction. A piece of antler found by the western platform exhibits evidence of groove-and-splinter working. here was also one specimen of elk found above the eastern platform and this is a partial metatarsal; the distal half of the 232 Star Carr Volume 2 element has been removed by a percussion break and there are clear radiating fractures from the percussion point. his would have been carried out for marrow extraction. In Clark’s area, the proximal end of a radius and the proximal end of a metacarpal have been removed by a mixture of percussion breaks and spiral fractures for marrow extraction. here are also four irst phalanges, one second phalanx and a navicular-cuboid that have been broken open by spiral fractures and percussion breaks for marrow extraction: it is interesting to note that ‘although there is relatively little marrow within these, what is present is tasty’ (Speth 2010, 34). One of the irst phalanges also has cut marks just above the break edge on the posterior aspect of the midshat which probably occurred during skinning. In Clark’s backill, the proximal end of a irst phalanx and the proximal end of a second phalanx has been removed by percussion breaks and spiral fractures for marrow extraction. he irst phalanx also exhibits cut marks just below the break edge which probably occurred during skinning. Seasonality A neonatal let maxilla of an elk was noted by Noe-Nygaard (1975), coming from an animal no more than a few weeks old and using modern analogy thought to represent a summer death. Legge and Rowley-Conwy (1988, 31–32) found that one mandible with tooth wear gave an indication of season and was probably killed later in the year, in September or October, though if the animal had been late born it would have been killed in November or December. An elk skull with shed antlers was originally taken to indicate midwinter occupation (Fraser and King 1954), but this was reinterpreted by Legge and Rowley-Conwy (1988, 31) who stated that the elk could have been killed anytime between December to late April/May. No seasonality of death could be ascertained from the faunal remains recovered from the recent excavations. MNI Although the most dominant element according to NISP was the irst phalanx, it is very diicult to side these elements and so they cannot be used for calculations for MNI. Antler is also not used for calculations of MNI and so the calcanei have been used. hree specimens were found: two let and one right. Two specimens belong to adult animals, one let and one right element, and these were found in the detrital wood scatter. he inal calcaneus (young adult) was found within Clark’s area and is represented by a partial element with an unfused distal epiphysis. hese data suggest two individuals. Legge and Rowley-Conwy (1988, 9) provided an MNI of 12 (let astragali); however, as elk has been found in the detrital wood scatter, which is dated much earlier compared to Clark’s area (Figure 17.20), the MNI for the site can be raised to 13. Red deer Overview Red deer was identiied in the original assemblage (Clark 1954, 79-86) and re-examined by Legge and Rowley-Conwy (1988), who found 541 fragments and calculated an MNI of 26 (based on the let mandible). A total of 535 specimens of red deer were identiied in the recent assemblage, including 73 specimens that were found within Clark’s backill. Red deer specimens were found across the entirety of the site (Figure 23.36). he largest concentrations are within the detrital wood scatter and within Clark’s area. On the dryland there are several smaller concentrations of remains: around the eastern structure and around the western structure, and along the shore edge. Elements Red deer is the most dominant species from Star Carr and is represented by the most diverse range of elements (Figures 23.37 and 23.38), as was identiied by Legge and Rowley-Conwy (1988, table 1A). Antler is the most Faunal Remains: Results by Species 0 5 10 20 30 Metres 233 § Figure 23.36: Spatial plot for red deer (Copyright Star Carr Project, CC BY-NC 4.0). Figure 23.37: NISP values for red deer (including backill and lotation inds) (Copyright Star Carr Project, CC BY-NC 4.0). 234 Star Carr Volume 2 Figure 23.38: Element representation of red deer (Copyright Archeozoo.org/J.-G. Ferrié 2004. Adapted by Becky Knight). dominant element (NISP=147), with the most dominant skeletal element being metatarsals (NISP=34). Elements that are only represented by a small number of specimens (ive or less) are the maxilla, sternum, hyoid, rib, sacrum, patella, ibula and cuneiform. Ribs appear to be lacking but they are diicult to identify to species, even when having been analysed using ZooMS, and so they may have ended up in categories such as ‘cervid species’ or ‘large mammal species’. Faunal Remains: Results by Species 235 Age and sex Of the 22 mandibles that Legge and Rowley-Conwy (1988, 42–44) used for ageing, the age proile spanned 2–9 years with 15 (60%) falling into the three to ive year classes. In terms of attempting to calculate the age proile for the recently excavated red deer remains, nine bone specimens were used. One of the mandibles and the ulna come from the detrital wood scatter, and the rest are from Clark’s area. here are a range of ages represented suggesting there were at least ive diferent individuals represented across the two areas. In the detrital wood scatter, two specimens provide age data (Table 23.10). he irst individual is represented by a partial mandible with mixed dentition: deciduous third and fourth premolars and a permanent adult irst molar. Using data generated by Severinghaus (1949) for white-tailed deer, this mandible closely matches two mandibles that were aged between three to four months. In addition, a partial ulna with a missing and unfused proximal epiphysis was found 0.4 m from the mandible but has a diferent age: this is estimated to be no older than 26–42 months/2.2–3.5 years. In Clark’s area seven specimens exhibit age data (Table 23.10). Two mandibles show a similar age: a partial let mandible with a partially erupted permanent third molar which match mandibles of the age 11–12 months in Severinghaus’ (1949) study, and a partial right mandible with deciduous second, third and fourth premolars and permanent adult irst and second molars which match mandibles of the age 12–13 months. hese were found c. 5 m apart and may or may not be from the same animal. A further partial mandible containing a deciduous fourth premolar with the permanent fourth premolar partially erupted out of the crypt, a permanent irst and second molar and a partially erupted permanent third molar match mandibles of the age of 18–19 months (1.6–1.7 years of age). here are three specimens with an age maximum of 26–29 months/2.2–2.5 years: two metacarpals with unfused distal epiphyses and a partial calcaneus with an unfused proximal epiphysis. One further specimen, a distal tibia with a partially fused distal epiphysis, has a maximum age of 26–42 months/2.2–3.5 years. In terms of sex assessment, Legge and Rowley-Conwy (1988, 58) used measurements from animals of known age and sex from the Isle of Rhum to establish sex assessment for the Star Carr red deer. hey concluded that there was a roughly even sex ratio among the Star Carr red deer, contrary to previous interpretations that relied heavily on the antler data. hey also noted that the common age class was three year olds (n=10); however, most of the antler has come from animals of four years old and above (Fraser and King 1954, 80) and so much of the antler must have been brought to site (Legge and Rowley-Conwy 1988, 58). For the recently excavated assemblage it was not possible to measure the specimens with any accuracy due to the deterioration and warping and thus the datasets could not be compared. However, at least two females can be identiied within the assemblage from the cranial remains. Both appear to be adult individuals in terms of size, robustness and development, and so it is unlikely they represent young male animals. Element NISP Age range Seasonality Detrital wood scatter Mandible (dp3/4, M1) 1 3–4 months Ulna (proximal) 1 26–42 months (2.2–3.5yrs) Aug/Sept = mid-late summer Clark’s area Mandible (partial eruption M3) 1 11–12 months April/May = mid-summer Mandible (dp2,3,4, M1, M2) 1 12–13 months (1–1.1yrs) May/June/July = mid-summer Mandible (dp4, partial eruption P4, M1, M2, partial eruption M3) 1 18–19 months (1.6–1.7yrs) Nov/Dec = mid-winter Metacarpal (distal) 2 26–29 months (2.2–2.5yrs) - Calcaneus (proximal) 1 26–29 months (2.2–2.5yrs) - Tibia (proximal—partial fusion) 1 26–42 months (2.2–3.5yrs) - Table 23.10: Age estimation of the Star Carr red deer remains using epiphyseal fusion ater Purdue (1983) and tooth eruption ater Severinghaus (1949). 236 Star Carr Volume 2 Seasonality he seasonality of the red deer has been much debated. It was originally proposed by Clark (1954, 1962) that people were based at Star Carr in the winter, following the red deer as they migrated to the North York Moors in the summer. It has since been suggested that red deer would not have migrated in this area (Legge and Rowley-Conwy 1988, 38). here is evidence from the work of Legge and Rowley-Conwy (1988, 38) that the young deer were killed in the late spring and summer and they state ‘it is reasonable to assume that the adults were too’. More recently, Carter (1998) carried out analysis of the red deer mandibular ramus of red deer from the site and showed that one red deer younger sub-adult would have been killed in early winter, and another in the winter or spring. A further pairing of rami may be a late summer kill or an autumn/winter death. From the recently excavated remains, using the age estimations of the younger individuals, it is possible to provide further seasonality data. his can only be based on the data gathered from dental development as this tends to be a more reliable age indicator (Purdue 1983). Red deer rut and begin mating around September/ October each year, and the gestation period typically has a duration of around eight months with animals being born in May/June (Dobronika 1988). From this, three mandibles suggest a summer death (one in the detrital wood scatter and two in Clark’s area), but a fourth suggests that it was killed in the winter (from Clark’s area) (Table 23.10). Modiications here are four red deer elements (scapula, ulna, navicular-cuboid and metatarsal) which come from within the detrital wood scatter which exhibit evidence of carnivore modiication in the form of uneven breakage with associated tooth marks and tooth scores. From Clark’s area, there are seven specimens with evidence of modiication by carnivores (a scapula, axis vertebra, tibia, astragalus, cuneiform, metatarsal and an antler frontlet), with uneven breakage associated with tooth impressions and tooth scores. In Clark’s backill three specimens (femur, calcaneus, metatarsal) exhibit uneven breakage, tooth scores and tooth impressions. Within test pit SC20 one tibia specimen has been broken, and exhibits tooth scores and tooth impressions. Overall, 186 specimens (not including antler) from across the site have been modiied by humans, in some cases exhibiting a number of diferent types of evidence. In terms of percussion breaks and spiral fractures, for the extraction of marrow, there are 10 specimens from the detrital scatter, 21 in the wetlands, 93 in Clark’s area, eight on the dryland, 20 in the backill and four in test-pits. his processing has been carried out mainly on long bones (98) but also on phalanges (37), mandibles (13), pelves (3) and cuneiform (1). Although most of this has probably been carried out for marrow, it is unlikely that the percussion breaks on the pelves would have been for this purpose. In terms of longitudinal split bone (n=24) for the production of tools, there is one from the detrital wood scatter, seven from the wetland, 10 from Clark’s area; two from the dryland, and four from the backill. Of these, 23 are long bones, of which 18 are metapodials, and one is a phalanx. Finally, cut marks are also evident: two from the detrital wood scatter, 27 from Clark’s area, and three from the backill. However, it should be noted that most of the cut mark evidence is exhibited on ribs categorised as cervids or large and medium mammals. Palaeopathology An interesting pathology was noted on a red deer skull with heavy remodelling to the cortical bone surface of the frontal and parietal bones (Figures 23.39 and 23.40). he cortical bone surface on the frontal bone undulates and clearly still had active bone remodelling occurring. It appears that the cause of this may have been the tearing of muscles that run across the surface of the cranium which help to support the weight of the head and antlers. he likelihood is this will have occurred during the rut. he porous nature of the bone surface suggests that there may have also been some infection associated with this damage. here is also some remodelling of the skull around the parietals with deep grooves in the skull, again associated with the muscles supporting the antlers, suggesting that this individual may have been an older individual with particularly large and heavy antlers, producing the need for highly developed muscles and anchor points on the cranium for these muscles. Faunal Remains: Results by Species 237 Figure 23.39: Red deer cranium <116020> with active remodelling and infection of the cortical bone across the entirety of the frontal bone (Copyright Neil Gevaux, CC BY-NC 4.0). Figure 23.40: Red deer cranium <116020> with remodelling of the skull and pronounced muscle attachments on the parietals (Copyright Neil Gevaux, CC BY-NC 4.0). 238 Star Carr Volume 2 MNI he element with the highest NISP value is antler; however, this is not used for calculations of MNI (Legge and Rowley-Conwy 1988, 9) and so the second most abundant element is the metatarsal, NISP=31 (not including examples in the backill). Of these, nine were sided to the let and 13 to the right, whilst the other nine could not be sided. It should be noted that these are fragmentary remains and cannot be used directly for MNI counts: the specimens were examined to ind repeating ends in order to calculate the MNI. Nine specimens were found in the area of the detrital wood scatter, which has been dated as one of the earliest activity areas of the site and of these there are three repeating right distal ends of the metatarsal establishing an MNI of three. he inal 22 specimens were found across both the dryland and within the waterlogged deposits. As there are six repeating proximal ends of the metatarsals, the MNI here is six individuals. Taking into consideration the distribution of these elements across the site, the total MNI for red deer is nine individuals. Sample No. Element 108590 horacic vertebra 108594 Radius 108589 Second phalanx 103625 Skull (frontlet) δ13C (‰) −22.3 ± 0.1 −22.3 ± 0.1 −22.3 ± 0.1 −21.4 ± 0.1 δ15N (‰) 3.9 ± 0.2 C:N Collagen yield 3.3 4.0% 3.8 ± 0.2 3.3 11.1% 4.3 ± 0.2 3.3 5.5% 3.5 ± 0.2 3.3 2.5% Table 23.11: Carbon and nitrogen stable isotope data for the red deer samples analysed in this study. Figure 23.41: Red deer (Cervus elaphus), roe deer (Capreolus capreolus) and reindeer (Rangifer tarandus) isotope data from several British Mesolithic sites (data compiled from Bowen et al. 2000; Schulting 2005; Schulting and Richards 2002b; 2009; Schulting et al. 2013; this study) (Copyright Star Carr Project, CC BY-NC 4.0). Faunal Remains: Results by Species 239 During their reanalysis, Legge and Rowley-Conwy used the mandible to calculate MNI, and identiied 26 let specimens. During the most recent excavations within Clark’s area, a total of four let mandibles were identiied. By combining this data, the total MNI for red deer from this area of the site is 30 individuals. Isotope data he patterning of semi-articulated red deer bones in the detrital wood scatter, in close proximity to an antler frontlet (see Chapter 7), posed the question of whether these remains were derived from the same animal. As a way of addressing this question, isotopic analysis was undertaken on three of the bones and the frontlet. he four red deer samples yielded suicient amounts of collagen of suitable quality for δ13C and δ15N stable isotope analysis (Table 23.11), and exhibit isotope values comparable to those previously reported from the site (Schulting and Richards 2009; Figure 23.41). Interestingly, no signiicant diference was seen between the isotopic values of the disarticulated deer remains and antler frontlet sampled. his therefore suggests that the deer utilised for frontlets were not (isotopically) distinct from those utilised for other purposes (e.g. as a food or raw-material resource). Furthermore, the isotopic values obtained from the four skeletal elements fall within the error expected by replicate analysis of a single individual (Pestle et al. 2014). he isotopic values generated from the newly excavated red deer remains are also directly comparable to previous deer values obtained from Star Carr, and can also be seen to be broadly comparable to data obtained from deer at other British Mesolithic sites (Figure 23.41). However, whilst similar in terms of δ15N values, the Star Carr deer appear to be slightly more depleted in δ13C than deer at other British Mesolithic sites (Figure 23.41). Figure 23.42: Red deer (Cervus elaphus), roe deer (Capreolus capreolus), aurochs (Bos primigenius), elk (Alces alces) and wild horse (Equus ferus) isotope data from Star Carr and Flixton Island 2 (data compiled from Stevens and Hedges 2004; Schulting and Richards 2009; this study) (Copyright Star Carr Project, CC BY-NC 4.0). 240 Star Carr Volume 2 More depleted δ13C values in fauna have previously been suggested to represent a diet derived from dense woodlands, rather than from more open habitats, representing a ‘canopy efect’ (Van der Merwe and Medina 1991; Noe-Nygaard et al. 2005; Drucker et al. 2008). he hypothesis that deer species at Star Carr may have favoured more closed, forested environments has also previously been proposed by Schulting and Richards (2009); however, a study by Stevens et al. (2006) on red deer has shown that a δ13C canopy efect is not always present in fauna inhabiting diferent environments, and as such should be treated with caution. When we compare the Star Carr red deer with the other terrestrial herbivores from the site, and the horses sampled from the nearby Long Blade site at Flixton Island Site 2, there is somewhat of a division between the deer and the other terrestrial species, particularly elk and horses (Figure 23.42). he deer appear to be consistently more depleted in δ13C, which may indicate the consumption of diferent plant resources between the species, grazing in slightly diferent environments or occupying diferent ecological niches. his is not surprising given that site 2 on Flixton Island is earlier in date and would have had a diferent environment to that at Star Carr (Blockley et al. 2018). Roe deer Overview Roe deer were identiied in the original assemblage (Clark 1954, 79–86) and re-examined by Legge and Rowley-Conwy (1988, 9), who identiied 103 fragments and estimated an MNI of 17, based on right mandibles. A total of 88 specimens were identiied from the recent excavations, spread across the site, with a concentration 0 5 10 20 30 Metres Figure 23.43: Spatial plot of roe deer (Copyright Star Carr Project, CC BY-NC 4.0). § Faunal Remains: Results by Species 241 Figure 23.44: NISP for the roe deer in the assemblage (Copyright Star Carr Project, CC BY-NC 4.0). Figure 23.45: Element representation for roe deer (Copyright Archeozoo.org/J.-G. Ferrié 2005. Adapted by Becky Knight). 242 Star Carr Volume 2 of 52 in Clark’s area, eight in the detrital wood scatter, eight in the wetland, four in the dryland and 16 from Clark’s backill (Figure 23.43). A range of elements are represented, with the most common being mandibles. Elements Although the NISP of the roe deer from the site suggests there are no maxillae represented, one was still attached to a roe deer skull and represents the only example in the assemblage (Figures 23.44 and 23.45). Underrepresented areas of the body include elements from the torso (such as vertebrae and ribs) and also smaller elements such as carpals, tarsals, phalanges, patellae and ibulae. hese elements are smaller in size and could have easily moved by taphonomic processes such as water, sediment movement or bioturbation. hey are also less robust elements and so may degrade and disappear more quickly than some of the more robust elements. Equally, some of these elements such as the foot bones may have been let on hides. However, this patterning of element representation is similar to that found by Legge and Rowley-Conwy (1988, table 1D), particularly in terms of the dominance of mandibles and the lack of maxillae. Ribs may appear to be lacking, but they are diicult to identify to species and may have been assigned as medium mammals. Age and sex In this study, mandibles are used to calculate the age proiles of the roe deer. However, eruption is of very little use for aging older roe deer as all of the teeth tend to erupt within the irst year (Aitken 1975); therefore, this method can only identify individuals less than a year in age. Also, all of the mandibles containing permanent dentition appear to have little to no occlusal wear. Only one specimen has the mixed dentition of part deciduous, part permanent teeth. he surviving teeth are deciduous third and fourth premolars and permanent irst and second molars. his suggests that one individual was less than one year in age based on tooth development of roe deer by Severinghaus (1949). Legge and Rowley-Conwy (1988, 40) identiied a total of 21 roe deer mandibles that could be aged using tooth development criteria by Aitken (1975). hey estimated that 45% of the animals (10 mandibles) represented animals aged approximately one year of age, whilst 19% were aged to two years. Using these methods it was not possible to identify, with any accuracy, animals from older age categories. Carter (1997), used radiographs to further investigate this question and found that seven mandibular rami from Star Carr roe deer could be aged to approximately 10–11 months. It was noted by Legge and Rowley-Conwy (1988, 42) that roe deer females tend to give birth from two years of age, and oten have twins, meaning that the natural population has a high proportion of young. In terms of sex assessment, Legge and Rowley-Conwy (1988, 59) collected measurements from the scapula and distal humerus, the two most dominant elements, in order to examine the extent of sexual dimorphism using carcass weights from Prior (1968). However, there was little division between the measurements and any diferences emphasised by the measurements were coincidental and not related to sexual dimorphism. Due to the fragmentary or modiied nature of the majority of the roe deer remains from the recent excavations, there were few complete elements from which measurements could be obtained. However, as Legge and Rowley-Conwy (1988) noted, the diference between the sexes for roe deer is minimal and so the lack of metrics is inconsequential. Modiication hree specimens have evidence of carnivore modiications: a pelvis from the detrital wood scatter, a scapula from Clark’s area and a femur from Clark’s backill. hese show a mixture of uneven breakage associated with tooth impressions and tooth scores. A total of 44 bone specimens exhibit evidence of human modiication: two metatarsals and a mandible in the detrital wood scatter; 11 long bones, 10 mandibles, one scapula and one irst phalanx in Clark’s area; one humerus, one radius, one femur and one tibia in the wetland; and four mandibles, one humerus, one radius, one femur and one metatarsal in Clark’s backill. he evidence consists of a combination of percussion breaks and spiral fractures created during the process of marrow extraction. In terms of longitudinally split bones, thought to be created in the process of tool manufacture: one metacarpal and one metatarsal were found from Faunal Remains: Results by Species 243 Clark’s area, a tibia was found in the wetland and a metatarsal was recovered from Clark’s backill. Only three specimens, two crania and one mandible, exhibit cut marks from skinning, and they are all from Clark’s area. In addition, a radius from the dryland was calcined. his specimen was found in the same area outside the western structure as the calcined wild cat phalanges. Seasonality he seasonality data from Legge and Rowley-Conwy (1988, 22–30) points to a late spring and summer kill from dental development of 13 roe deer mandibles. In addition, 63 unshed antlers were used with caution to suggest kills between April and November. However, Carter (1997) examined 12 mandibular rami using radiography and concluded that these animals were being killed in the late winter and early spring. No further analysis for seasonality has been undertaken from the recently excavated assemblage. MNI Mandibles are used in the calculations of MNI for roe deer. Seven specimens can be sided as right, ive sided as let and one cannot be sided. Twelve of the mandibles contain teeth. he majority of the elements exhibit permanent adult dentition apart from one which has a mixture of deciduous and permanent teeth. Using these data, an MNI of six is established (ive adults and one young adult) as it is possible to reit two of the right mandibles to one specimen. Two of the mandibles were found above the detrital wood scatter and the remaining four were found in Clark’s area. Legge and Rowley-Conwy (1988, 9) had an MNI of 17, also based on right mandibles. By adding together our two groups of mandibles, the total roe deer MNI for the site is 23 individuals. 0 5 Figure 23.46: Spatial plot of aurochs. 10 20 30 Metres § 244 Star Carr Volume 2 Figure 23.47: NISP for aurochs (Copyright Star Carr Project, CC BY-NC 4.0). Figure 23.48: Element representation of aurochs (Copyright Archeozoo.org/M. Coutureau 2009. Adapted by Becky Knight). Faunal Remains: Results by Species 245 Aurochs Overview Aurochs bones were identiied in the original assemblage (Clark 1954, 79–86) and re-examined by Legge and Rowley-Conwy (1988, 9) who identiied 174 fragments and provided an MNI of 16. A total of 156 aurochs specimens have been identiied from the recent excavations (Figure 23.46) (Copyright Star Carr Project, CC BY-NC 4.0). Elements A wide range of elements are represented, with the most common being the humerus (Figures 23.47 and 23.48). Some elements are missing from the assemblage: cranial elements, hyoid, mandible, maxilla, ulna, sacrum and cuneiforms. However, these were found in the original excavations (Legge and Rowley-Conwy 1988, table 1C). here are also 25 fragments from the dryland that have been identiied as aurochs using ZooMS but cannot be identiied to element due to poor preservation or small size. Age and sex Legge and Rowley-Conwy (1988, 44) state that a single aurochs mandible with permanent teeth in wear suggest a relatively old individual. From the recently excavated assemblage, the majority of elements are too incomplete or too fragmentary to retain the developmental information required to gauge the age. However, age estimates have been made for a number of specimens using Grigson (1982), and although this is based on domesticated cattle, it is useful for a general guide to the development stages for this taxon. he majority of the elements appear to be from large and robust adult animals; however, there are also nine elements that have unfused epiphyses, and one element that exhibits partially fused epiphyses (axis inferior vertebral body epiphysis) (Table 23.12) from which ive separate individuals of diferent ages can be identiied. he irst is represented by an unfused distal end of a second phalanx. Fusion for this element occurs between 1.3–1.6 years of age, and therefore this specimen has a maximum age of 1.6 years. he second is represented by an unfused distal epiphysis of a metapodial. Fusion for this element occurs between 2–2.5 years of age, giving a maximum age of 2.5 years. he third individual has a maximum age of three years due to the unfused proximal calcaneal epiphysis. here are three specimens that are identiied as having fusion between 3.5–4 years of age (a humerus with an unfused proximal epiphysis, a femur with an unfused proximal epiphysis and a tibia with an unfused proximal epiphysis), and so it is possible that all of these elements represent one individual; the fourth individual with a maximum age of four years. he ith individual is Element NISP Age range Second phalanx (distal) 1 1.3–1.6 yrs Metapodial (distal) 1 2–2.5 yrs Calcaneus 1 3 yrs Humerus (proximal) 2 3.5–4 yrs Femur (proximal) 2 3.5–4 yrs Tibia (proximal) 1 3.5–4 yrs Axis vertebra (inferior vertebral body epiphysis) 1 4–5 yrs Table 23.12: Summary of the age ranges associated with the unfused aurochs remains from Star Carr, using Grigson (1982). 246 Star Carr Volume 2 Figure 23.49: Metric comparison of the right (square) and let (circle) aurochs astragali data illustrating sexual dimorphism (males within the blue ellipse, females within the red ellipse) (Copyright Star Carr Project, CC BY-NC 4.0). represented by an axis vertebra with a partially fused inferior vertebral body epiphysis, and a maximum age of ive years. Due to the fragmentary nature of the remains few specimens can be measured, except for four astragali, making it possible to do a comparative analysis with Legge and Rowley-Conwy’s (1988, 46) data in order to establish evidence for sexual dimorphism. hey found that from measurements on 13 astragali, seven were males, and among 15 metacarpals, 10 were males. In this study, plotting our astragali measurements against Legge and Rowley-Conwy’s data, there are two more male specimens and two more female specimens (Figure 23.49). Modiications A total of 13 specimens exhibit evidence of carnivore modiications, 10 from Clark’s area (six long bones, a cervical vertebra, a scapula, an astragalus and a irst phalanx), one from the wetland (a calcaneus) and two from Clark’s backill (a radius and a femur). hese all exhibit uneven breakage associated with tooth impressions and tooth scores. A total of 65 specimens exhibit evidence of human modiication. In terms of evidence of marrow extraction, there are 61 specimens, nine in the detrital wood scatter, 22 in Clark’s area, 11 in the wetland, 10 in the dryland, eight in Clark’s backill and one in a test-pit. here are 13 specimens that are longitudinally split, likely for tool production: one in the detrital wood scatter (metatarsal), ive in Clark’s area (humerus, two metatarsals, a one irst and one second phalange), two in the wetland (humerus and femur), four from the dryland (three humeri and one metatarsal) and one from Clark’s backill (metacarpal). Only ive specimens exhibit cut marks: three from Clark’s area (scapula, sternum and humerus), one from the dryland (thoracic vertebra) and one from the backill (metatarsal), and these are likely to have been produced through skinning. Faunal Remains: Results by Species 247 Palaeopathology here is also one aurochs metatarsal with a healed lesion to the midshat which appears to be spherical in shape. his is likely to be the result of a long-healed perforation or projectile wound. here is only a small amount of thickening to the outer surface of the afected area, and there is no associated deformity or excessive thickening, pitting or pock marking of the cortical bone surface to suggest disease or active healing. he internal surface of the medullary cavity also appears to be normal, further supporting that it is unlikely to be caused by disease. Due to the small size and shape of the afected area, it could be suggested that it is the result of a long-healed perforation, albeit caused by either human or natural causes such as trauma, or perhaps even the result of a possible projectile wound. MNI he humerus is the most dominant element. As there are four repeating fully developed distal articular ends, one right and three let, the MNI for aurochs is three. When looking at these remains spatially, two of the let distal humeri and the one right distal humerus were located above the central platform, and the third let distal humerus was found within Clark’s area. As the peat above the central platform may be of a similar date to Clark’s area, the MNI remains at three. Legge and Rowley-Conwy (1988, 9) provided an MNI of 16 based on metacarpals. 0 5 10 20 30 Metres Figure 23.50: Spatial plot of beaver (Copyright Star Carr Project, CC BY-NC 4.0). § 248 Star Carr Volume 2 Figure 23.51: Element representation of beaver (Copyright Archeozoo.org/M. Coutureau 2003. Adapted by Becky Knight). Figure 23.52: NISP of beaver in the assemblage (Copyright Star Carr Project, CC BY-NC 4.0). Rodentia Beaver Overview Fraser and King (1954, 73–74) noted that beaver is fairly well represented in the collection with an estimate of at least seven animals represented. In the recent excavations, 15 specimens have been identiied, which mainly derive from Clark’s area and the dryland (Figure 23.50). On the dryland, these remains are represented by one fragment that has been identiied by ZooMS (but could not be identiied to element), and teeth and mandibles. Elements Beaver is represented by four loose teeth and ive mandibles, as well as a scapula, two humeri and two femora (Figures 23.51 and 23.52). his correlates with what was found in Clark’s excavation, which includes jaw bones, two humeri, four radii, four ulnae, six pelvic bones, eight femora, 10 tibia, one sacrum and six lumbar vertebrae (Fraser and King 1954, 73–74). It is important to note that this pattern of element representation in terms of the presence of large and distinctive incisors is unlikely to be a result of excavation bias or sampling strategy, as Faunal Remains: Results by Species 249 Figure 23.53: Two of the beaver mandibles (<116813> let and <115878> right) with similar modiication to remove the ascending ramus, and fragmentary incisors (let to right: <117547>, <116164> and <115881>) (Photograph taken by Paul Shields. Copyright University of York, CC BY-NC 4.0). small and more gracile remains have been recovered for other species (for example roe deer loose teeth, wild cat phalanges, pine marten caudal vertebrae and small fragments of ish remains). Age and sex In terms of age assessment, one humerus has a proximal epiphyses which is unfused and missing. According to Fandén (2005) this element remains unfused between the ages of three months and 6.4 years. A complete femur is fully fused; fusion is complete in this element by eight years of age, therefore this individual could have been 8+ years old. It was not possible to sex these remains. 250 Star Carr Volume 2 Figure 23.54: Polish just below the dental arcade on the mandible <115878> and on the buccal side of one of the molars, possibly suggesting the use of leather binding to create a tool from the mandible (Photograph taken by Paul Shields. Copyright University of York, CC BY-NC 4.0). Modiications here is only one specimen with evidence of animal modiication: a beaver femur from Clark’s backill exhibits uneven breaks and associated tooth impressions and tooth scores. In terms of human modiication evidence, six specimens exhibit percussion breaks, and they consist of four mandibles (two from Clark’s area, one from the wetland and one from the dryland), an incisor (from Clark’s area) and a humerus (from Clark’s area). he breakage pattern to the mandible is similar for all four specimens, with percussion breakage to remove the ramus, and two exhibit breakage around the incisor tooth socket, possibly for the removal of the incisor itself (Figure 23.53). Although the removal of the ascending ramus is associated with marrow extraction in other similar elements of red deer, as there is no breakage along the jawline below the tooth row for the beaver mandibles, this breakage is unlikely for this purpose. his modiication may therefore be related to the creation of beaver mandible tools, or for purposeful removal of the incisor from the mandible. Additionally, the presence of polish to the bone surface just below the dental arcade on mandible <115878> (Figure 23.54) on both the buccal and lingual sides suggests leather binding may have been applied to create a handle out of the mandible in order to use the incisor, possibly for woodworking. here are a number of archaeological examples, especially from Russia, of beaver incisors being utilised for decorative purposes or as a tool either whilst still within the mandible or loose and hated into antler (Zhilin 1997; Lozovskaya and Lozovski 2015). All of the loose beaver teeth are incisors and all are fragmentary (Figure 23.53). here is no clear sign of hating as seen in the Russian examples but dehating may have taken place, or leather may have been used for holding them. Faunal Remains: Results by Species 251 MNI During their original analyses, Fraser and King (1954) identiied 71 specimens of beaver; 15 loose teeth and 56 elements. he elements included one maxilla (with a complete set of teeth), 14 pieces of mandibles (eight right, six let rami of the lower jaw), 12 loose incisors, 3 loose molars, two humeri, four radii, four ulnae (two right, two let), six pelves (three right, three let), eight femora (one right, seven let), ten tibiae (three right, seven let), one complete sacrum and six lumbar vertebrae. Interestingly, the majority of the long bones (NISP=22) were found to derive from immature animals with one or more epiphyses unfused and missing. It is unclear which element was used to calculate the MNI, but they suggest that at least seven animals were represented, with the majority being immature individuals (Fraser and King 1954, 74). It is not clear why the rami were not used to give an MNI of 8, unless they were perhaps fragmentary. For the assemblage described here, the most common element is the mandible, which consists of ive specimens found in several locations: two on the dryland, one near the eastern platform and two within Clark’s area. Four mandibles are sided as right and one as let, all of which contain permanent teeth in wear; this provides an MNI of four. If the results from the most recent excavations are combined with Fraser and King (1954), based on the presence of mandibles, an MNI of 12 is estimated for the entire assemblage: four right mandibles from the recent excavations and eight right mandibles (rami) from those undertaken by Clark; however, without examining the original collection we cannot be sure that this does not double count some of Clark’s specimens, if they were fragmentary. Field vole A total of three ield vole specimens were found on the dryland, and they are all represented by fragments of loose teeth. he teeth are the easiest way to diferentiate between the diferent species of vole, and they were found during the lotation of soil samples taken from the central hollow of the eastern structure during the 2008 excavations. All three teeth appear to have been charred. It is unlikely that these teeth were intentionally brought into the structure, and given that they are charred, it is likely that they occur here by accident; possibly a vole was caught in a ire. here is very little that can be said about this species due to the small number of remains found; however, they are of interest as microfaunal remains have not been found at the site previously. Figure 23.55: MNI comparison between Legge and Rowley-Conwy analysis and the current research (Copyright Star Carr Project, CC BY-NC 4.0). 252 Star Carr Volume 2 Common name Clark 1954 Degerbøl Harrison 1961 1987 Legge and Tot Lord Rowley-Conwy 1988 Northern pike Present Overall study X European perch X X X X X Common crane X X White stork(?) X X Red-breasted merganser X X X Red-throated diver X X X Black-throated diver ? ? Great crested grebe X X X Little grebe X X X Lapwing X X Buzzard X X Duck (size of pintail) X X Brent goose X X Common scoter X X White-fronted or bean goose X Bear X Wild cat X X X X Pine marten X X Fox X X Wolf X Badger X Dog X X X X X X Hedgehog X X Roe deer X X Red deer X X Elk X X Pig X X Aurochs X X Hare X X Beaver X X Field vole X X Table 23.13: Revised list of taxa from Star Carr showing the presence (X) or absence of species. he names in red represent those species that are present in the amalgamated assemblages taking into account re-analyses of the material. × denotes taxa that have been re-analysed and classiied as something else. Faunal Remains: Results by Species 253 Discussion he data presented here provides some important new insights into the faunal assemblage at Star Carr. ZooMS as a complementary method has also provided additional information regarding species, allowing us to identify more of the dryland specimens, and although the cervids cannot be distinguished, the method identiied signiicant quantities of aurochs within the assemblage. he recent excavations at the site have yielded four previously unidentiied species: northern pike, European perch, wild cat and ield vole. here may also be a black-throated diver (though it may be a red-throated diver) and there is a new species of goose (white-fronted or bean). here appears to be both a dog skeleton and a wolf bone, which brings wolf back into the list following previous re-evaluations of dog to wolf. here are also a number of rodent specimens from lotation samples which cannot be identiied to the genus and species. Overall, with all the corrections to the data over the years, there now appears to be 26 species identiied from Star Carr (the potential black-throated diver has not been included in this total) (Table 23.13), as opposed to the 21 irst identiied (Fraser and King 1954). he ish bones are particularly noteworthy because there has been much debate as to why ish was not found previously at the site (Wheeler 1978): evidence has now been found both from the wetland (in Clark’s area) and on the dryland as calcined bones demonstrating ish were being caught, processed and discarded on site (see also Robson et al. 2016). It is interesting to note that of the MNI of the ive most represented species (red deer, roe deer, aurochs, elk and wild boar) there is very similar pattern to Legge and Rowley-Conwy’s (1988) data in terms of order of prevalence (Figure 23.55). his is probably due to the fact that a large proportion of the recent data also comes from Clark’s area, although it should be noted that red deer appears to predominate in most areas of the site. Legge and Rowley-Conwy (1988) hypothesised that the assemblage (as represented by Clark’s excavation) was strikingly similar to that of a Nunamiut hunting camp. his pattern was based on the presence of mandibles, upper forelimbs and limb extremities, with the assumption that the upper rear limbs are transported back to a diferent residential base (Legge and Rowley-Conwy 1988). However, from our dataset we can see that femora are present. A number of them are classed as ‘large mammal’ because most are fragments due to breakage for marrow extraction, and many were found in the backill. We now know that Clark did not collect many unidentiiable specimens and his assemblage had been handpicked, which skewed the dataset; this therefore throws doubt on the notion of this site being a hunting camp. he seasonality assessment has also changed since the analysis of Legge and Rowley-Conwy (1988). It should be noted that all seasonality assessments are reliant on the use of modern analogues for the season of birth, the migration patterns of birds and the shedding of antler (Milner 1999; 2005). here is oten some degree of variation in this data; for instance, some species can give birth over several months, which means the anchor point is moveable, and there is some variability in terms of the timing of tooth eruption and the development of wear patterns. However, what we do not know is how much further variability there might have been in the past, particularly for this site where we have clear evidence for signiicant luctuations in climate (Chapters 4, 9 and 18). he changes in temperature inevitably would have afected the behaviour of animals, and we should be mindful that, for instance, bird migration patterns might have been signiicantly diferent. What is clear is that there now appears to be evidence from Clark’s area for animals that have been killed in all four seasons. Legge and Rowley-Conwy (1988, 38) made the point that game was probably accessible at all times of the year: none of the major species show a marked tendency to migrate. However, because at that stage no winter kill stages were present in their dataset, they suggested that a year-round settlement was unlikely. Red deer skulls with antler shed are likely to be spring (Legge and Rowley-Conwy 1998, igure 7), with a concentration of roe deer mandibles and maxilla argued to be indicative of May/June deaths. Further summer evidence comes in the form of neonatal red deer, red deer maxilla, red deer mandibles, a neonatal elk and summer migratory birds (Grigson 1981). An elk mandible suggests an autumn kill, possibly alongside some red deer mandibles. However, Carter (1997; 1998) has since demonstrated that both roe deer and red deer exhibit evidence for winter kills and one of the red deer mandibles assessed as part of this study also suggests a winter kill. It is also important to consider the nature of the dataset. he seasonality data have always been based on the material from Clark’s assemblage. We now know that this is part of a larger picture and this deposition may have taken place over a short period of time (Chapter 17). he only other seasonality data from the rest of the site is from the detrital wood scatter, where the evidence from one red deer mandible suggests a kill in the summer. his simply suggests that a deer was killed in the summer during the earliest occupation of the site but does not necessarily mean that occupation did not happen in other seasons. 254 Star Carr Volume 2 Conclusions his new data provides some important changes to our understanding of the faunal assemblage at Star Carr, with new species, as well as a re-evaluation of both the site type, as assessed by faunal remains, and seasonality. As Legge and Rowley-Conwy (1988, 7) state: ‘no bone report is ever in a full sense a inal report. Conclusions will always be subject to modiication, and identiications to rechecking, as new methods and skills are developed.’ hey go on to say that having come to diferent conclusions to those of Fraser and King, they do so in the knowledge that their identiications and conclusions will also be subject to change, ‘possibly in much less than the 30 years that have elapsed since the appearance of the original bone report.’ It is now almost 30 years since their report, and in fact little has changed with the exception of Carter’s (1997; 1998) work on age and seasonality. However, the recent excavations have thrown more light on Clark’s area as well as new areas of the site: we also anticipate that this new dataset will generate further re-evaluations, new conclusions and new debate.