zyxwvutsrq zyxwvuts zy zyx J . A u ~ tent. . Soc., 1986, 25: 37-40 37 RELIABILITY OF USING HEAD CAPSULE WIDTH AND BODY LENGTH TO IDENTIFY LARVAL INSTARS OF AEDES VZGZLAX (DIPTERA :CULICIDAE) L. SHINKARENKO', K. HULSMAN', P. MOT TRAM^, P. DALE'and B. H. K A Y ~ 'School of Australian Environmental Studies, GriJith University, Nathan, Qld 411 1. 2QueenslandInstitute of Medical Research, Bramston Terrace, Herston, Qld 4006. Abstract Discriminant function analyses on head capsule width and body length of 940 larvae of the saltmarsh mosquito Aedes vigilax (Skuse) were used to determine their reliability in distinguishing between different instars. Of larvae reared under different temperature/salinity regimes, 99.4% were correctly graded on the basis of head capsule width whereas only 88.0% of determinations were correct when body length was used. Thus head capsule width was found to be a reliable character to grade instars regardless of rearing conditions. zyxwv Introduction Accurate grading of the 4 larval instars of the saltmarsh mosquito Aedes vigilax (Skuse) was needed for field studies of the species. In the field larvae develop under a variety of conditions that affect their size. If the overall size range associated with each instar can be established then larvae can be correctly graded. Peters (1943), Kettle (1948), Sen and Das Gupta (1958) and McDonald et al. (1977) successfully used head capsule width to separate different instars of mosquitoes. The reliability of head capsule width and also of body length was investigated using discriminant function analysis to determine their accuracy in separating larval instars of Aedes vigilax reared under different temperature and salinity regimes. zyxwvuts Materials and methods The accuracy of grading larvae reared under the different temperatureisalinity regimes which are normally encountered in the field was tested. Temperatures used were 25 and 35°C while salinities were 10, 17, 20, 35 and 45 O/,,,. Soil samples (approx. 30 x 30 x 5 cm) and their vegetation cover were collected from Coomera Island 2 7 9 1 'S, 153'23'E, southeast Queensland and stored dry in the laboratory for at least a week at room temperature (about 25°C). Each soil sample was then placed in a large jar and flooded with seawater. Hatching was stimulated by evacuation to lower the dissolved oxygen (Borg and Horsfall 1953). On hatching, 5 groups of 20 larvae were each transferred to 1 L plastic beakers in which they were reared in 400 ml of water at a specific salinity and temperature. Every second day, larvae were fed 75% W/Vyeast suspended in water of the appropriate salinity. Ten larvae from each treatment except 2 5 T , 17O/,,, were sampled daily. lnstars of larvae were known from the number of skinscast. Larvae were immobilised with 70% alcohol and maximum head capsule width and body length were measured immediately with an eyepiece micrometer. Body length was measured anteriorly from the head under the clypeal spines to the junction of the eighth and anal segments. In the treatment 25°C and 17°/,, larvae were reared individually in glass containers. They were measured 24 h after hatching in the case of first instar and 24 h after moulting in the case of the other instars. Statistically each larva (of specific instar reared under specific conditions) measured is a replicate. Ten to 63 larvae of each instar were measured in each treatment. Discriminant functions were calculated for each treatment and the whole data set. The programme was from the Statistical Package for the Social Sciences (SPSS, Nie er al. 1975). Information about the mathematics of discriminant function analysiscan be found in Cooley and Lohnes (1971). Two way analysis of variance was used to test for significant treatment effects. Results and discussion Within each treatment, discriminant function analysis separated instars perfectly on the basis of head capsule width but separation was variable on the basis of body length (Table 1) ranging from 85.51% at 35 "Cand 35 o/oo to 98.88% at 25 "Cand 45 o/oo. For practical purposes we were more interested in evaluating the usefulness of these characters to grade larvae reared under the range of conditions commonly occurring in the field. For this we calculated discriminant functions for the complete data (940 larvae) reared under all treatments. Based on head capsule width 99.4% of larvae were correctly graded (Table 2). This is less than the perfect discrimination zyxwvu L. SHINKARENKO ET AL. 38 within each treatment because the discriminant function for the whole data set is different from that for each treatment. The overall high level of discrimination indicates that head capsule widths are generally discrete between instars over a range of rearing conditions (Fig. 2, see also McDonald.et al. 1977). Only 87.8% of larvae were correctly graded on body length (Table 3). That body length was more variable than head capsule width can be seen from the coefficients of variation in Figs 1 and 2. zyxwvu zyxwvuts zyxwvutsrq Table 1 . Percentage of Aedes vigilax larvae in each treatment which were correctly graded by a discriminant function analysis on head capsule width and body length Temperature ( "C) Percentage of larvae correctly graded Salinity ("loo) Head width ~ 25 35 Body length zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ 10 17 20 35 45 100.00 100.00 100.00 100.00 100.00 90.07 89.00 94.32 93.40 98.88 10 100.00 100.00 86.6 94.83 85.51 87.43 20 35 45 100.00 100.00 Table 2. Number and percentage of Aedes vigilax larvae correctly graded on basis of head capsule width by discriminant function analysis. Percent correctly graded was 99.36 ~ Number examined 1 182 182 174 (100%) 1 2 3 226 4 358 ~~ Predicted instar membership Actual instar 1 2 I73 (99.4%) 3 (1.3%) (0.6%) 3 4 223 (98.2%) I (0.3%) 357 (99.7%) I 241 I IV zyxw zyx zyxw zyxwvut 0 1 2 3 4 5 6 7 8 Body Length (mm) FIG.1-Distribution of body lengths of each of the 4 instars of Aedes vigilax. The mean ( f 1 sd)body length and coefficient of variation (CV) for each instar were: I, 1.253 f 0.339 CV = 27.1%; II,2.490 rf: 0.351 CV = 14.1%; 111, 3.757 f 0.545 CV = 14.5%; IV, 5.763 f 0.821 CV = 14.2%. zyxwvu zyxwv zy GRADING INSTARS OF AEDES VIGILAX o 0.1 39 zyxwvu zyxwvu zyx 0.2 0.3 0.s 0.6 0.4 0.7 0.8 0.9 1.1 1.0 1.3 1.2 Width of Head Capsule (mm) FIG.2-Distribution of head capsule widths ofeach of the 4 instars of Aeda vigitax. The mean ( f 1 sd) head capsule width and coefficient ofvariation (CV) for each instar were: I, 0.294 f 0.015 CV = 5.1%; 11,0.498 f 0.03 CV = 6.0%; 111,0.775 f 0.053 CV = 6.8%; IV, 1.151 & 0.066 CV = 5.7%. Table 3. Number and percentage of Aedes vigihx larvae correctly graded on basis of body length by discriminant function analysis. Percent correctly graded was 87.77 Actual instar Number examined 1 182 2 174 3 226 4 358 Predicted instar membership __ 1 2 181 (99.5%) 10 (5.7%) 1 (0.4%) 1 (0.5%) 164 (94.3%) 32 (14.2%) 3 4 170 (75.2%) 48 (13.4%) 23 (10.2%) 310 (86.6%) Table 4. Summary of two-way anova on the effects of temperature and salinity on body length and head capsule width of third instar Aedes vigilax Source df F Body length F Head capsule width 45.81*** 5.32** 5.42*** 270.81*** 77.62*** 5.78*** ~ Temperature* Salinity+ Temperature x Salinity Error * 1 3 3 I58 temperature adjusted for salinity effects. t salinity adjusted for temperature effects. ** P < 0.01 *** P < 0.001. Results of the two way analyses of variance showed that variability in both body length and head capsule width was caused by different rearing conditions, in this case, temperature and salinity. Temperature had a much greater effect than salinity on body 40 zyxwv zyxwvuts zy L. SHINKARENKO ET AL. size of-all instars and this is well illustrated by the third instar (Table 4). There was a significant interaction between temperature and salinity on both body length and head capsule width (Table 4). That is, the effect of temperature on these characters varied with salinity. Despite the significant effects of temperature and salinity on body size of larvae, the discriminant function reliably identified each instar on the basis of head capsule width. This is because the variation within an instar was far less than the variation between instars and there was no overlap in the width of head capsules of any instar (Fig. 2). The few larvae that were incorrectly graded were early seconds, thirds and fourths categorized as first, seconds and thirds respectively (Table 2). One would expect the size of larvae reared under field conditions to be somewhat more variable than those in our experiments for two reasons. First, food supply and level of crowding vary in the field and these sources of variation were excluded in our experiments. Second, field temperatures and salinities sometimes exceeded the range used in our experiments (19-38 "C and 7->40°/,,). But our experiments covered the range that occurred most frequently in the field. Nevertheless we still expect each instar to differ distinctly in head capsule width and that it remains a reliable indicator of instar. However these expectations still require testing. In conclusion, head capsule width is an extremely useful character to use when grading Aedes vigilux larvae. This character accurately grades larvae irrespective of the temperature/salinity regimes in which they were reared. Acknowledgments We thank Drs Roger Kitching and Angela Arthington for their comments about a draft of this paper. We also gratefully acknowledge funds and support from Griffith University, Gold Coast City Council and Queensland Institute of Medical Research. We thank Mrs Janis Wedge for typing the manuscript. References zyx zyxwvu zyxw BORG,A. F. and HORSFALL, W. R. (1953)-Eggs of floodwater mosquitoes, 11 Hatching Stimulus. A n n . ent. SOC.Am. 46: 472-478. COOLEY, W. W. and LOHNES,P. R. (197l)-Multivariate data analysis: 243-261. Wiley: New York. KETTLE,D. S. (1948)-The growth of Anopheles sergenti Theobald (Diptera, Culicidae), with special reference to the growth of the anal papillae in varying salinities. Ann. trop. Med. Parasit. 4 2 5-29. MCDONALD, G., SMITH,1. R. and YHELDEN, G. P. (1977)-Identification of instars ot Cufex annulirostriA Skuse (Diptera:Culicidae). J. Aust. ent. SOC.1 6 359-360. NIE,N. H., HULL,C. H., JENKINS, J., STEINBRENNER, K. and BENT,D. H. (1975)-Statisticalpackagefor the Social Sciences. 2nd Edn. McGraw-Hill New York, 675 pp. PETERS, H. T. (1943)Studies on the biology of Anopheles walkeri Theobald (Diptera:Culicidae). J. Parasit. 2 9 117-122. SEN,P.and DASGUPTA,S.K. (1958)-"Dyar's Law" in the determination of larval instars in mosquitoes. Bull. Calcutta Sch. trop. Med. 6 69-10. [Manuscript received 11 January 1985. Revised 15 April 1985.1