Diet composition of the collared lizard (Crotaphytus collaris) in west-central Texas.
Crotaphytus collaris has a broad distribution in the southwestern United States, ranging from eastern Utah and Colorado to southwestern Illinois, and south into Mexico. Its optimal habitat is rocky, limestone outcroppings and rocky areas (Smith 1946; Conant & Collins 1991). It has been the subject of numerous ecological studies, including its diet (see Best & Pfaffenberger 1987 for a complete listing). Diet has been studied quantitatively and anecdotally for this species in New Mexico (Best & Pfaffenberger 1987), Kansas (Fitch 1956), Oklahoma (Blair & Blair 1941), Arkansas and Missouri (McAllister 1985), and Utah (Knowlton 1938), but not in Texas, a state where the species is widespread and relatively common. Other quantitative studies have focussed on this species at the periphery of its range (e.g., McAllister 1985,) or in areas of habitat other than limestone outcroppings (e.g., Best & Pfaffenberger 1987). The general feeding ecology of this species has been thoroughly described, but this study focuses on sites that have optimal habitat for the species, so that a better understanding of the geographic variation in diet can be determined by comparison to other studies.
Crotaphytus collaris is significantly and variably sexually dimorphic in head width and length, with males having larger heads than females (McCoy et al. 1994; McCoy et al. 1997). One hypothesis for the evolution of sexual dimorphism proposes that the sexes differ in morphology so that they can utilize different resources and minimize niche overlap, reducing intraspecific competition (Selander 1966). Males and females could consume different types and/or sizes of prey items as part of the resource partitioning. The larger heads of males may be explained by this hypothesis if males were found to eat larger and/or harder prey items (e.g., beetles as opposed to grasshoppers) to account for the increased musculature. Best & Pfaffenberger (1987) did not find a difference in size of prey consumed by males and females, but they did find sexual differences in diet composition. They postulated that this difference may reduce intraspecific competition. A goal of this study was to determine if there is a difference in diet between sexes, and whether or not the data would lend support to the hypothesis that differential niche use is a selective force driving the evolution of sexual dimorphism in this population.
METHODS AND MATERIALS
Collared lizards (N=48) were captured by noosing in June-July 1997 and 1998 at six locations in west-central Texas: a private ranch in Irion County, Texas consisting of rocky limestone outcroppings on a redberry juniper (Juniperus pinchottii) and mesquite (Prosopis glandulosa) dominated hillside; a private ranch in Concho County, Texas consisting of a dry, limestone creekbed surrounded by mesquite brushland; and four sites in Tom Green, Runnels and Coke counties, consisting of rip-rap boulder dams surrounded by mesquite brushland.
Upon capture, lizards were returned to the lab where they were measured (SVL, total length, head width, and head length, all to the nearest 0.1 mm), euthanized and necropsied. The gastrointestinal tract was removed and the stomach was separated and preserved with its contents in 10% formalin. Stomach contents were later removed and prey items were counted and identified to the lowest possible taxon following Borror et al. (1989). Total weight of prey items was determined to the nearest 0.01 g and total volume of prey items was determined to the nearest 0.01 mL by volumetric displacement of water in a small, calibrated syringe. Volume and weight for individual prey items were not determined to avoid any discontinuity that might result by not including prey items in advanced stages of digestion.
Two-sample t-tests were performed to compare total volume, total weight, number of prey items per stomach and number of kinds of prey items per stomach between sexes. An artificial average volume and average weight for each prey item was calculated and tested between sexes with a two-sample t-test. Levins' (1968) value for niche breadth was calculated for each sex individually and for both sexes combined. Discriminant analysis was used to statistically compare composition of diet between sexes. Procedures for using discriminant analysis in dietary analysis follow Best & Gennaro (1984). There were no significant differences (two-sample t-tests) found in total volume, total weight, number of prey items per stomach, or number of kinds of prey items per stomach between the rip-rap dam and natural habitat, nor were there significant differences in the four categories between sexes in the two different habitat types (ANOVA) at the 0.05 significance level. Therefore, data were pooled from all of the sites for all comparisons between sexes.
A summary of the diet of C. collaris in west-central Texas is presented in Table 1. The mean number of prey items per stomach was 3.06. The mean number of kinds of prey items per stomach was 2.29. The mean total volume of prey items per stomach was 0.99 mL. The mean total weight of prey items per stomach was 1.11 g. The prey items with the highest frequency of occurrence (number of specimens containing a prey item / total number of specimens X 100) were: Orthoptera, 87.5%; Coleoptera, 31.3%; and Hymenoptera, 25.0%. Within the orthopterans, the family Acrididae had a frequency of occurrence of 85.4%. Coleopterans were evenly spread among the six families identified, but were dominated by Curculionidae with a frequency of occurrence of 10.4%. Hymenopterans were represented by four families, and were dominated by Formicidae with a frequency of occurrence of 18.8%. Orders that represented the highest percentages of the prey items (number of prey items in a category / total number of prey items X 100) were: Orthoptera, 39.2%; Hymenoptera, 23.6%; Coleoptera, 17.6%. Homopterans were less common, accounting for only 9.5% of the prey items and being present in only 20.8% of the lizards. Hemiptera and Odonata were even less common, both accounting for only 2.0% of the prey items and being present in only 6.3% and 4.2% of the lizards respectively. Plant material of various sorts accounted for 4.7% of the prey items and was found in 12.5% of the lizards. This is believed to be due to incidental ingestion as suggested by McAllister (1985). Small pebbles (2-3 mm) were common, being found in 29.2% of the lizards. These are probably used as digestive aids (Johnson 1966). All of the pebbles were of approximately uniform size and shape, suggesting that they are used for breaking up arthropod exoskeletons.
Two-sample t-tests revealed no significant difference (p > 0.05) in total volume, total weight, number of prey items per stomach, number of kinds of prey items per stomach, average volume, or average weight between sexes. Despite the lack of significant difference between sexes, there were very evident trends in dietary composition that were not accounted for by the statistical analyses (Figure 1). Females appeared to consume more, smaller prey items than males, but the number of kinds of prey items per stomach was similar. Females also appeared to eat more hymenopterans (32.9%) and coleopterans (21.5%) and less orthopterans (29.1%) than males (13.0%, 13.0% and 50.7% respectively).
Levins' measures of niche breadth for the males (B=3.30) and females (B=3.94) were similar. For the sexes pooled (B=3.96), it was more similar to that of females. Discriminant analysis revealed no significant difference in diet between sexes. For both sexes, a total of 75% were classified as the correct sex based on diet, with 85% of the males being classified and 64% of the females being classified correctly.
[FIGURE 1 OMITTED]
The jackknifed matrix, however, only classified 38% total correctly, with 46% of the males and 27% of the females correctly classified. The jackknifing procedure, which drops each observation one at a time and repeatedly re-classifies groups, is not as influenced by outliers during the classification process when groups (e.g., sexes) are being predicted based on certain variables (e.g., diet components). This procedure represents a better way of predicting and classifying groups based on true separations of the variables being used to do the classifications.
The diet composition for this population of C. collaris differed from other studies in New Mexico, Kansas and Arkansas and Missouri as should be expected in populations that are geographically separated. Orthopterans made up a much greater portion of the diet in this study, followed by hymenopterans and coleopterans. It is, however, consistent with findings in northeastern Oklahoma (Blair & Blair 1941) and Utah (Knowlton 1938) where orthopterans were by far the most abundant food items taken.
The Levins' niche values are similar, suggesting that there is little variation in diet niche breadth between the sexes. They are also relatively low values (see Pianka 1987), suggesting that C. collaris in west-central Texas does not have as diverse a diet as might be expected for a lizard that is usually considered a strict opportunist (e.g., B=7.34 for Uta stansburiana, B=5.83 for Cnemidophorus tigris; Pianka 1987). This is evident by the extreme dominance of certain prey groups such as orthopterans and hymenopterans. Other studies have suggested that this species is an extreme generalist, such that availability was more important than preference (McAllister 1985). More specifically, it is worth noting that no dipterans, lepidopterans, blattids, scorpions or vertebrates were found, all of which are common in the area. It is not surprising how important orthopterans were found to be in the diet, but it is surprising how relatively unimportant abundant groups such as coleopterans and arachnids were found to be to the lizards' diet overall when compared to other studies. Blair & Blair (1941) found similarly high values for orthopterans in northeastern Oklahoma, but suggested that it was due to availability. Knowlton (1938) found similar dominance by orthopterans, but they constituted a smaller percentage of the diet and had a lower frequency of occurrence than found in this study. Hymenopterans were found to be more common and coleopterans were found to be about the same as found in this study. Best & Pfaffenberger (1987) hinted that preference may play a role in the diet of C. collaris based on the dominance of certain prey taxa. There was a dominance of certain taxa to an even higher degree in this study. It is unclear whether this population may be discriminating to any degree as is predicted by optimal foraging theory. They apparently have an abundant supply of diverse, palatable prey items, but they are consuming one prey item much more than any other taxa. Prey abundances were not determined in this study, but grasshoppers often appeared to be the most abundant prey item at the study sites. Quantitative studies of prey densities and diet are needed to confirm this.
Discriminant analysis produced a high degree of correct classification, but the jackknifed matrix percentages were all very low. This gives a better understanding of the true separation, or lack thereof, because outliers do not play a large role in the jackknifed classification. This shows that there is no significant separation in prey composition between the sexes. Best & Pfaffenberger (1987) found a significant difference in diet composition between sexes in New Mexico and suggested that it might provide evidence that differential niche utilization by the sexes was a force driving sexual dimorphism in head size. The lack of significant difference between the sexes in this study does not support the theory that differential niche utilization by the sexes is a strong selective force driving sexual dimorphism in this population, at least in the sense of differential foraging niches based on diet. Although there were trends evident between the sexes, none are significant and many are due to a few outlying individuals. The trends were also contradictory to this hypothesis. Females were found to eat more, smaller prey, but they ate more hard prey items (more coleopterans and hymenopterans and less orthopterans) than males. Sexual dimorphism has been shown to vary geographically for this species (McCoy et al. 1994; McCoy et al. 1997), so it may be that other selective forces (e.g., different parasite loads, different predation rates, different densities of potential mates) are acting at higher degrees on this population, when compared to the populations examined in other studies. Differential niche utilization may still be a factor, but in this case, they may be partitioning resources other than prey.
Table 1. Stomach contents of 48 collared lizards (Crotaphytus collaris) from west-central Texas presented by sex. Sample sizes are given in parentheses and occurrence data are given as total items of that category in the stomachs. Category Male Female Total (26) (22) (48) Arachnida 2 0 2 Insecta Orthoptera 35 23 58 Hemiptera 2 1 3 Homoptera 5 9 14 Coleoptera 9 17 26 Hymenoptera 9 26 35 Odonata 3 0 3 Plant tissue 4 3 7 Total 69 79 148
Thanks to local ranchers, the U.S. Army Corps of Engineers, and the Colorado River Municipal Water District for access to the study sites. Thanks also to L. Holscher and T. King for help in the field and lab. Thanks to M. S. Husak, J. R. Arevalo and an anonymous reviewer for comments on earlier versions of the manuscript. Lizards were caught under Texas Parks and Wildlife Department Scientific Permit #SPR-0597-887. Funding was provided by an Angelo State University Research Enhancement Grant to J. Kelly McCoy.
Best, T. L. & A. L. Gennaro. 1984. Feeding ecology of the lizard, Uta stansburiana, in southeastern New Mexico. J. Herpetol., 18(3):291-301.
Best, T. L. & G. S. Pfaffenberger. 1987. Age and sexual variation in the diet of collared lizards (Crotaphytus collaris). Southw. Nat., 32(4):415-426.
Blair, W. F. & A. P. Blair. 1941. Food habits of the collared lizard in northeastern Oklahoma. Am. Midl. Nat., 26(2):230-232.
Borror, D. J., C. A. Triplehorn & N. F. Johnson. 1989. An Introduction to the Study of Insects. Saunders College Publishing, New York, xiv + 827 pp.
Conant, R. & J. Collins. 1991. A Field Guide to Reptiles and Amphibians of Eastern and Central North America. Houghton Mifflin Co., Boston., xiv + 450 pp.
Fitch, H. S. 1956. An ecological study of the collared lizard (Crotaphytus collaris). Univ. Kansas Publ., Mus. Nat. Hist., 8:213-274.
Johnson, D. R. 1966. Diet and estimated energy assimilation of three Colorado lizards. Am. Midl. Nat., 76(2):504-509.
Knowlton, G. F. 1938. Lizards in insect control. Ohio J. Sci., 38(5):235-238.
Levins, R. 1968. Evolution in changing environments: some theoretical explorations. Princeton University Press, Princeton, N.J., ix + 120 pp.
McAllister, C. T. 1985. Food habits and feeding behavior of Crotaphytus collaris collaris (Iguanidae) from Arkansas and Missouri. Southw. Nat., 30(4):597-619.
McCoy, J. K., S. F Fox & T. A. Baird. 1994. Geographic variation in sexual dimorphism of Crotaphytus collaris. Southw. Nat., 39(4):328-335.
McCoy, J. K., H. J. Harmon, T. A. Baird & S. F. Fox. 1997. Geographic variation in sexual dichromatism of Crotaphytus collaris. Copeia, 1997(3):565-571.
Pianka, E. R. 1987. Ecology and natural history of desert lizards: analyses of the ecological niche and community structure. Princeton University Press, Princeton, NJ., xx + 208 pp.
Selander, R. K. 1966. Sexual dimorphism and differential niche utilization in birds. Condor 68(2):113-151.
Smith, H. M. 1946. Handbook of Lizards. Lizards of the United States and Canada. Comstock Publishing Associates, Ithaca, N.Y., xxi + 557 pp.
Jerry F. Husak* and J. Kelly McCoy
Department of Biology, Angelo State University
San Angelo, Texas 76909
Department of Zoology, Oklahoma State University
Stillwater, Oklahoma 74078
JFH at: firstname.lastname@example.org
|Printer friendly Cite/link Email Feedback|
|Author:||Husak, Jerry F.; McCoy, J. Kelly|
|Publication:||The Texas Journal of Science|
|Date:||May 1, 2000|
|Previous Article:||Food habits of three species of striped whipsnakes, Masticophis (Serpentes: Colubridae).|
|Next Article:||Reproduction in the Mojave rattlesnake, Crotalus scutulatus (Serpentes: Viperidae).|