Feeding Habits of Stinkpot Turtles

The common musk turtle, Sternotherus odoratus, is the northernmost ranging species of the widespread freshwater turtle family Kinosternidae. It inhabits slow moving waters of rivers, ponds, lakes and streams over the eastern portion of the North American continent and in 72 Texas counties. It is small with a narrow, domed carapace and two stripes running down the sides of the head towards the neck. Common musk turtles inhabit heavily vegetated areas with slow currents (Berry, 1975).
Common musk turtles inhabit both lentic and lotic areas of Spring Lake, Hays County, Texas. Differences have previously been documented in the feeding and reproductive habits of common musk turtles in lentic and lotic systems (McPherson and Marion, 1983). This may be a result of differences in the availability of food resources between the systems or competition with other species.
Common musk turtles are chiefly nocturnal (Lagler, 1943; Vermersch, 1992) and omnivorous, feeding on a variety of foods ranging from algae (Cladophora sp., Spirogyra sp., Lyngbya sp.), to snails (Marisa cornuarletis and Hydrobia sp.), and small vertebrates (tadpoles, frogs and adult fish) (Ernst, 1986). Ford and Moll (2004) described seasonal and sexual differences in common musk turtle diets in Missouri, however detailed and
quantified dietary studies on common musk turtles and how they differ by gender and season in Texas are lacking. Food habit studies need to be done on a local scale because of differences in availability of foods in different geographic areas. Results from one area may not be applicable to another (Rosas-Rosas et al., 2003).
Dietary studies provide basic information to enhance understanding of the ecology of a species and are important prerequisites to developing successful conservation, management and recovery plans (Huygens et al., 2003; Hernandez et al., 2006). Species management and recovery plans depend on knowledge of suitable habitat and distribution of nutritional factors in order to predict the potential distribution of a species (Jones et al., 1998). The quality of nutritional factors within the habitat determine the stability of the breeding population (Mindell et al., 1987), sustainability and the amount of seasonal stress on populations (Madison et al., 2002), and the productivity of an organism or population (Bjorndal, 1985).
Food habits studies lend predictive value to the distribution of the animals in question but also to the floral structure of their habitat (Clark et al., 2001). Birds (Compton et al., 1996; Wutherich et al., 2001), bats and other mammals (Clark et al., 2005), and even turtles (Braun and Brooks, 1987) are important agents of seed dispersal. Understanding the feeding and movement patterns of these seed dispersers enables biogeographers to analyze structural vegetation regimes from a habitat management and conservation standpoint, and allows for the identification of habitat parameters vital to the continued occupation of the species (Ford and Moll, 2004).
A variety of methods are employed to study animal diets. The removal of stomach material by dissection is one of the more reliable ways of obtaining data because foods in
the stomach are often undigested and readily identifiable. This method has been successfully used to study the diets of birds (Day and Byrd, 1989; Falk et al., 1992; Jamieson et al., 2001), deer (Nixon et al., 1970; Weckerly and Nelson 1990), bears (Sato et al., 2000), fish (Hall et al., 1995; Beaudoin et al., 1999), amphisbaenians (Lopez et al., 1991), snakes (Diller and Johnson, 1988) and turtles (Hulse, 1974; Bjorndal, 1985). Stomach flushing does not require sacrificing animals and has been used to study the diets of birds (Gionfriddo et al., 1995; Hess, 1997), small mammals (Wrazen and Svenson, 1979), fish (Culp et al., 1988), lizards (Legler and Sullivan, 1979; Pietruszka, 1981), and turtles (Legler, 1977; Parmenter, 1980; Fields et al., 2003; Seminoff et al., 2002; Caputo and Vogt, 2008).
Less invasive methods of studying animal diets include observational records and analysis of fecal material. Analysis of fecal material is a convenient method to obtain dietary data without causing undue stress to the organism. Fecal studies have been done on bats (Hurst and Lacki, 1997) wildcats including leopards, ocelots and pumas (Johnson et al., 1993; De Villa Meza 2002; Rosas-Rosas et al., 2003), deer and antelope (Koerth et al., 1984; Hodgman et al., 1996), rodents (Karels et al., 2004) and reptiles, including turtles and tortoises (Braun and Brooks, 1987; MacDonald and Mushinsky, 1988; Seminoff et al., 2002; Caputo and Vogt, 2008). In some studies, more items were identified from fecal samples than from samples obtained from stomach contents (Wrazen and Svendson, 1979).
Many individual resident turtles in Spring Lake, including common musk turtles, are marked and currently being used in population studies (Francis Rose, pers. comm.), so fecal analysis is the preferred method of studying food habits. Stomach flushing was attempted with common musk turtles, but resulted in unacceptable rates of mortality (Thomas Simpson, pers. comm.).
This thesis presents the results of a comparative study of the diets of common musk turtles in the lentic and lotic habitats of Spring Lake. My objectives were to quantify the diet of the common musk turtle in Spring Lake, to determine whether differences exist in the dietary habits of the common musk turtle in lentic and lotic areas of the lake, and to determine whether dietary habits of common musk turtles in Spring Lake vary by gender or season. I also assess whether or not common musk turtle feeding habits in Spring Lake are similar to those previously documented in other portions of their range.

Study Site
Spring Lake was created in 1847 by construction of a dam downstream from the San Marcos Springs system, which is comprised of 200 springs that flow from the Edwards Aquifer in San Marcos, Texas (Stovall et al., 1986; Brune, 2002). This 7.9 ha reservoir (Fields et al., 2003) is the headwaters of the San Marcos River. Sink Creek, an intermittent waterway, empties into Spring Lake through an eastern extension of the lake called the slough, causing periodic localized flooding. Spring Lake, the slough area and the San Marcos River are home to numerous turtle species, including the Texas river cooter (Pseudemys texana), red eared slider (Trachemys scripta elegans), the spiny softshell (Apalone spinifera), the common snapping turtle (Chelydra serpentina) and the common musk turtle. Common musk turtles were collected from lentic and lotic areas of the lake.

Sample Collection
I collected 41 common musk turtles during spring (March 2008 - May 2008), 40 during the winter (December 2008 - February 2009), and 33 each during the fall (September 2008 - November 2008), and summer (June 2008 - August 2008). Sixty-eight turtles were collected from the lotic habitat and 79 from the lentic habitat of Spring Lake. I used notching codes on the marginal scutes of individuals to determine that no turtles used in my study traveled between habitats.

I collected common musk turtles from shore with a dip net during 45 minutes before and after dusk. I baited the capture area with a small amount of commercial canned cat food about 10 minutes prior to engaging in capture attempts. After capture, each turtle was placed in a separate 3.8 l bucket with enough water to cover the limbs and base of the shell. Immediately after each evening of collection, I transferred the buckets of turtles to a laboratory at Texas State University-San Marcos. I rinsed each turtle, then weighed, measured, and determined their gender. I replaced the water in each bucket with fresh, clean water and held the turtles in separate buckets in the laboratory for 48 hours to allow voiding of intestinal contents.
After 48 hours, fecal material was removed from each bucket and immediately stored in 70% ethanol for identification. After sample collection, turtles were released at the site of capture.

Sample Identification
Sample material was sorted and identified in four steps. First, samples were separated into categories including mollusk, crayfish, vegetation (including algae), insects, and "other." The "other" category included everything that could not be placed into any of the preceding four categories, such as non-insect invertebrates, what appeared to be eggs, and unidentifiable material. The five categories from each sample were stored in separate vials and labeled with a sample number and composition for later analysis.
For identification and quantification, each vial was placed in a 100 x 20 mm petri dish with four lines drawn onto the bottom forming a 30 x 30 mm square. Items closest to each corner and the mid-point of the square were identified to the lowest taxonomic level possible under a dissecting microscope. After the components of the samples were identified, the contents of each vial was dried and weighed to the nearest 0.0001 gram. To standardize size differences between samples, the percent weight of each component was used in the analysis.

Analysis
Four samples were not included in the analysis because they were physically destroyed prior to being sorted and identified. The raw sample weights for the remaining 143 samples were converted to percent weights to standardize the sample comparisons. To prevent computational errors relating to zero (for example, cannot divide by zero, cannot take the log of zero), fecal weights of 0.0000 were converted to an extremely small nonzero value (0 .00000001).
Because strong correlations would violate the assumptions of a univariate model, Pearson's Product Moment Correlation was calculated to determine the magnitude and direction of association between food items for each of the independent variables (habitat, gender and season) to determine if ANOVA modeling was appropriate for the dependent variables. To determine if there were any differences across the five dependent variables for the three-way classification variable (season, habitat, gender) a three-way univariate analysis of variance was computed for each category, as appropriate.
A three-way univariate analysis of variance was computed for each measure of fecal composition. Following D'Agostino et al. (1990), normality was determined using Shapiro-Wilk, Fisher's Skewness, Fisher's Kurtosis and D'Agnostino-Peason omni-bus test for Kurtosis were used to determine which of the samples were appropriate for the general linear models (analysis of variance) (see D'Agostino et al., 1990 for a complete description of the selected tests). The Brown-Forsythe test was used to test for homogeneity of group variances.

I identified 36 taxa including 6 mollusks, 1 crayfish, 13 insects, 12 plants, and 4 taxa in the 'other' category. Mollusks were the most common component in the fecal samples, which occurred in 88% of samples. Crayfish parts were identified in 44% of the samples, insects or insect parts were identified in 46% of the samples and plants were identified in 41% of the samples. Every sample had a portion of material that was classified as 'other.' The most frequently occurring mollusks were snails in the genus Physa, and Gyralus, which were identified in 57% and 43% of the samples, respectively.

Normality Assumption
The results of the normality tests indicated that the percent weight of mollusks and 'other' were appropriate for modeling without transformation. The distribution of percent weight values in the crayfish, insect and vegetation categories were transformed according to Stevens (1996).

Homogeneity Assumption
The Brown-Forsythe test was used to test variances within fecal component categories. The hypothesis of equal variances was supported for all categories except 'other' (p = .3086 (mollusks), .3411 (crayfish), .9927 (insects), .2065 (plants), .0273 (others)). The log transformation was applied to the 'others' variable and achieved homogeneity (p = 0.3433).

Independence
Overall, there was a negative relationship between the relative amount (percent weight) of mollusks consumed and the amount of 'other' components consumed. There was little overall correlation between any of the other components.

Analysis of Variance
The strong correlation between mollusks and 'other' (r = -0.56) make an ANOVA for the variable 'other' unnecessary because no new information would be gained by adding it to the analysis. ANOVA tests for each of the remaining response variables are adequate because no strong correlations exist among the remaining food items.
For the percent weight of mollusks, a significant main effect was found for the season and habitat interaction. Subsequently, this interaction was also significant for the percent weight of plants. It should be noted that the interaction patterns for each fecal measure match and are inverted by habitat. Specifically, from spring to summer s increase mollusks in lotic habitats (and decrease in lentic) while plants decrease in lotic (and increase in lentic). From summer to fall, mollusks decrease in lotic habitats (and increase in lentic) while plants increase in lotic (and decrease in lentic). Finally, from fall to winter, both mollusks and plants increase in lotic and lentic habitats. Additionally, for the percent weight of insects, a significant main effect was found for the season and gender interaction.

Fecal Composition of Turtles in Lentic vs. Lotic Habitats
Habitat accounted for a significant portion of the variation in the amount of insects consumed between the two habitats. The frequency of mollusks, crayfish, insects and 'other' was higher in samples from the lentic habitat.

Fecal Composition Measures of Male and Female Turtles
Chi-square tests of independence revealed that differences in capture rates between genders for each habitat were not significantly different for the spring, summer and fall seasons, but that significantly more males were captured during the winter from the lentic habitat than any other season. Female turtle consumed slightly higher average percentages of mollusks and plants and male turtles consumed slightly higher average percentages of mollusks and other.

Seasonal Variations in Fecal Composition Measures
Mollusks occurred in 54.4% of samples taken in the spring, 46.9% of samples taken in the summer, 60.6% of samples taken in the fall and 55.1% of samples taken in the winter. Crayfish occurred in 6.8 to 14.3% of samples in all seasons, insects occurred in 14.3 to 20.4% of samples in all seasons, and plants occurred in 11.6 to 30% of samples in all seasons. The amount of plants consumed varied significantly between habitats. There was a higher number of mollusks consumed during the summer and winter, and a higher number of crayfish consumed during the fall and spring.

It is not surprising that the most commonly counted food items in fecal matter samples from common musk turtles were mollusks, crayfish, and insects due to the persistence of hard parts such as shell and chitinous exoskeleton. Larger shells such as those of the invasive Giant Ramshorn Snail, Marisa cornuarietis, were always excreted in fragments, while many smaller shells were excreted intact. Thinner shells with a more fragile structure, such as those of snails in the genus Physa, were sometimes broken, but were typically recognizable. Mollusks found in the digestive tracts of common musk turtles in Michigan (Lagler, 1943), were generally crushed beyond recognition.
Both mollusks and insects occurred with a higher frequency in the lentic habitat, with a pattern strong enough to be indicative of environmental variation rather than variation in the sampling. Habitat alone accounted for 0.8% difference in the average percentage of mollusks consumed, but habitat combined with seasonal variation showed a difference that was significant, specifically that there was a large increase in the number of mollusks consumed in the summer in the lotic habitat, followed by a decrease in the fall and simultaneous increase in consumption in the lentic habitat. Plant consumption in the lentic habitat dropped significantly from summer to fall, while in the lotic habitat there was a slightly higher percent weight of plants present in fall and winter than in summer samples. Insects occurred at a much lower frequency than mollusks, but the difference in the amount of insects consumed between habitats proved significant.
The differential rates of digestion between chitinious, calcareous and cellulose parts probably accounts for some of the variation in the results. Although some studies have been done on differential digestion rates and on the interaction of diet items in turtles (Bjorndal, 1991), no information is available for digestion rates of individual food items in Sternotherus sp. If differential digestion rates for common musk turtles were available for this study then the results might have shown less variability between food categories containing hard parts, such as mollusks and crayfish, and those containing mostly soft parts, such as insects and plants. One way to reduce variation introduced by differential digestion rates without having to calculate a digestion coefficient is to obtain food items from the stomachs of the animals prior to digestion, either by dissection or stomach flushing.
Although this study employed only fecal analysis and did not attempt to correct for differential rates of digestion among food items, the results did not disagree with studies that employed other methods of dietary analysis for the common musk turtle (Lagler, 1943; Ford and Moll, 2004). Common musk turtles are typically described as omnivores and scavengers (Mahmoud, 1969), both designations were supported by my study. Common musk turtles actively hunt for food such as snails and insects (referred to by Ernst (1986) as the peer and probe method) and I observed them actively attacking crayfish, but they also scavenge on carcasses of dead animals (Ernst, 1986). One sample was made up almost entirely of mammal fur and skin, and another contained bird feathers.
Only two juveniles were captured (Mitchell, 1988 only captured 7% juveniles out of 107 turtles, so this is not uncommon), so an analysis of diet variation with age was not possible. It is possible that the dietary tendencies of common musk turtles change over the course of their lives, such as in Pseudemys and Graptemys species (Clark and Gibbons, 1969; Moll 1976), this possibility should be explored more in depth with future research.
Some Chelonians are agents of seed dispersal, which may explain patterns in riparian vegetation (Ford and Moll, 2004). Seeds have been identified in the fecal matter of snapping turtles, box turtles, and tortoises (Lagler, 1943; Braun and Brooks, 1987), however Lagler (1943) noted that seeds were crushed after passing through the digestive tract of the common musk turtles. Nuphar advena seeds identified in fecal samples from common musk turtles in Spring Lake were identifiable, but had also been crushed. It is possible that larger Chelonians may pass seeds without mechanical damage and are better suited to seed dispersal.
There was not a significant difference in the amount of plant matter measured in common musk turtle fecal material between the lentic and lotic areas of Spring Lake. This analysis, however, did not take into account possible differences in species composition of the samples or of the environment itself. The percent weight of mollusks, insects and 'other' was different between the two habitats, indicating perhaps that resources differ between the habitats. Because the differences only occurred in non-vegetative categories, it is possible that the vegetative composition is similar but the habitat differences affect the community composition of the rest of the food resources available to the turtles. Lindeman (1996) postulated that differences observed in two populations of Painted Turtle, Chrysemys picta, in different habitats were due to the differential availability of chironomid larvae in the two environments in which the populations were found.
Ford and Moll (2004) found that seasonality affected the sex ratio of their captures. There were more males than females captured during the winter portion of my study, but caution should be used when interpreting this because of the relatively low sample size (22 males, 18 females). Seasonal trends may be more apparent in northern latitudes where the water temperature changes throughout the year.
Season did, however, have a significant effect on the percent weight of mollusk, crayfish and plant categories. These seasonal feeding trends were also apparent for the insect category as a season by gender interaction, with females consuming significantly more insects than males in the spring season.
One reason for the shift in feeding trends of females in the spring might be that invertebrates provide more protein and energy than plants, algae and detritus (Bowen, 1995). Mollusks and crayfish have more hard, indigestible parts than insects. Perhaps turtles can gain more energy from eating insects because the chitinous exoskeleton of insects occupies less space in turtle digestive tracts. There would be more digestible, usable food there per unit volume with insects than for mollusks and crayfish.
The results of this study largely agreed with past dietary analyses of common musk turtles, most of which noted algae, seeds, plants, crayfish, mollusks, insects and other invertebrates and scavenged materials in the stomach, gut and fecal material (Lagler, 1943; Ford and Moll, 2004). Based on the results of this study, fecal analysis is an appropriate method of quantifying the diet of common musk turtles, and the diet of common musk turtles in Spring Lake is consistent with that in other portions of their range.

Beaudoin, C. P., W. M. Tonn, E. E. Prepas and L. I. Wassenaar. 1999. Individual specialization and trophic adaptability of northern pike (Esox lucius): an isotope and dietary analysis. Oecologica 120: 386-396.

Berry, J. F. 1975. The population effects of ecological sympatry on musk turtles in northern Florida. Copeia 1975: 692-701.

Bjorndal, K. A. 1985. Nutritional ecology of sea turtles. American Society of Ichthyologists and Herpetologists 1985: 736-751.

Bjorndal, K. A. 1991. Diet mixing: nonadditive interactions of diet items in an omnivorouse freshwater turtle. Ecology 72: 1234-1241.

Bowen, S. H., E. V. Lutz, and M. O. Ahlgren. 1995. Dietary protein and energy as determinants of food quality: trophic strategies compared. Ecology 76: 899-907.

Braun, J. and G. Brooks, Jr. 1987. Box turtles (Terrapene carolina) as potential agents for seed dispersal. American Midland Naturalist 117: 312-318.

Brune, G. 2002. Springs of Texas. Volume 1. Texas A&M University Press, College Station, Texas, USA.

Caputo, F. P. and R. C. Vogt. 2008. Stomach flushing vs. fecal analysis: the example of Phrynops rufipes (Testudines: Chelidae). Copeia 2008: 301-305.

Clark, C. J., J. R. Poulsen and V. T. Parker. 2001. The role of arboreal seed dispersal groups on the seed rain of a lowland tropical forest. Biotropica 33: 606-620.

Clark, C. J., J. R. Poulsen, B. M. Bolder, E. F. Conner and V. T. Parker. 2005. Comparative seed shadows of bird-, monkey-, and wind-dispersed trees. Ecology 86: 2684-2694.

Clark, D. B. and J. W. Gibbons. 1969. Dietary shift in the turtle Pseudemys scripta (Schoepff) from youth to maturity. Copeia 1969: 704-706.

Compton, S. G., A. K. Craig and W. R. Waters. 1996. Seed dispersal in an African fig tree: birds as high quantity, low quality dispersers? Journal of Biogeography 23: 553-563.

Culp, J. M., I. Boyd and N. E. Glozier. 1988. An improved method for obtaining gut contents from small, live fishes by anal and stomach flushing. Copeia 1988: 1079-1082.Day. R.H. and G.V. Byrd. 1989. Food habits of the whiskered auklet and Buldir Island, Alaska. The Condor 91: 65-72.

D'Agostino, R. B., A. Belanger and R. B. D'Agostino, Jr. (1990). A suggestion for using powerful and informative tests for normality. The American Statistician, 44: 316-321.

de Villa Meza, A., E. M. Meyer, and C. A. Gonzalez. 2002. Ocelot (Leopardus pardalis) food habits in a tropical deciduous forest of Jalisco, Mexico. American Midland Naturalist 148: 146-154.

Diller, L. V. and D. R. Johnson. 1988. Food habits, consumption rates, and predation rates of western rattlesnakes and gopher snakes in southwestern Ohio. Herpetologica 44: 228-233.

Ernst, C. H. 1986. Ecology of the turtle, Sternotherus odoratus, in southeastern Pennsylvania. Journal of Herpetology 20: 341-352.

Falk, K., J. Jensen and K. Kampp. 1992. Winter diet of Atlantic puffins (Fratercula arctica) in the northeast Atlantic. Colonial Waterbirds 15: 230-235.

Fields, J. R., T. R. Simpson, R. W. Manning and F. L. Rose. 2003. Food habits and selective foraging by the Texas River Cooter (Pseudemys texana) in Spring Lake, Hays County, Texas. Journal of Herpetology 37: 726-729.

Ford, D. K. and D. Moll. 2004. Sexual and seasonal variation in foraging patterns in the stinkpot, Sternotherus odoratus, in southwestern Missouri. Journal of Herpetology 38: 296-301.

Gionfriddo, J. P., L. B. Best and B. J. Giesler. 1995. A saline-flushing technique for determining the diet of seed-eating birds. The Auk 112: 780-782.

Hall, S. J., W. S. Gurney, H. Dobby, D. J. Basford, S. D. Heaney and M. R. Robertson. 1995. Inferring feeding patterns from stomach contents data. The Journal of Animal Ecology 64: 39-62.

Hernandez, F., W. P. Kuvlesky, R. W. Deyoung, L. A. Brennan and S. A. Gall. 2006. Recovery of a rare species: case study of the masked bobwhite. The Journal of Wildlife Management 70: 617-631.

Hess, C. A. 1997. Stomach flushing: sampling the diet of red-cockaded woodpeckers. The Wilson Bulletin 109: 535-539.

Hodgman, T. P., B. B. Davitt and J. R. Nelson. 1996. Monitoring Mule Deer Diet Quality and Intake with Fecal Indices. Journal of Range Management 49: 215-222.

Hulse, A. C. 1974. Food habits and feeding behavior in Kinosternon sonoriense (Chelonia: Kinosternidae). Journal of Herpetology 8: 195-199.

Hurst, T. E. and M. J. Lacki. 1997. Food habits of Rafinesque's big-eared bat in southeastern Kentucky. Journal of Mammalogy 78: 525-528.

Huygens, O. C., T. Miyashita, B. Dahle, M. Carr, S. Izumiyama, T. Sugawara and H. Hayashi. 2003. Diet and feeding habits of Asiatic black bears in the Northern Japanese Alps. Ursus 14: 236-245.

Jamieson, S. E., G. T. Robertson and H. G. Gilchrist. 2001. Autumn and winter diet of long-tailed duck in the Belcher Islands, Nunavut, Canada. Waterbirds: The International Journal of Waterbird Biology 24: 129-132.

Johnson, K. G., W. Wei, D. G. Reid and H. Jinchu. 1993. Food habits of Asiatic leopards (Panthera pardus fusea) in Wolong Reserve, Sichuan, China. Journal of Mammalogy 74: 646-650.

Jones, M. D., G. S. Warburton and M. R. Pelton. 1998. Models for predicting occupied black bear habitat in costal North Carolina. Ursus 10: 203-207.

Karels, T. J., L. Koppel and D. S. Hik. 2004. Fecal pellet Counts as a Technique for Monitoring an Alpine-Dwelling Social Rodent, the Hoary Marmot (Marmota caligata). Arctic, Antarctic, and Alpine Research 36: 490-494.

Koerth, B. H., L. J. Krysl, B. F. Sowell and F. C. Bryant. 1984. Estimating Seasonal Diet Quality of Pronghorn Antelope from Fecal Analysis. Journal of Range Management 37: 560-564.

Lagler, K. F. 1943. Food habits and economic relations of the turtles of Michigan with special reference to fish management. American Midland Naturalist 29: 257-312.

Legler, J. M. 1977. Stomach flushing: a technique for chelonian dietary studies. Herpetologica 33: 281-284.

Legler, J. M. and L. J. Sullivan. 1979. The application of stomach-flushing to lizards and anurans. Herpetologica 35: 107-110.

Lindeman, P. V. 1996. Comparative life history of painted turtles (Chrysemys picta) in two habitats in the inland Pacific northwest. Copeia 1996: 114-130.

Lopez, P., J. Martin and A. Salvador. 1991. Diet selection by the ampbisbaenian Blanus cinereus. Herpetoloigca 47: 210-218.

MacDonald, L. A. and H. R. Mushinsky. 1988. Foraging Ecology of the Gopher Tortoise, Gopherus polyphemus, in a Sandhill Habitat. Herpetologica 44: 345-353.

McPherson, R. J. and K. R. Marion. 1983. Reproductive variation between two populations of Sternotherus odoratus in the same geographic area. Journal of Herpetology 17: 181-184.

Mindell, D. P., J. B. Albuquerque and C. M. White. 1987. Breeding population fluctuations in some raptors. Oecologica 72: 382-388.

Mitchell, J. C. 1988. Population ecology and life histories of the freshwater turtles Chrysemys picta and Sternotherus odoratus in an urban lake. Herpetological Monographs 2: 40-61.

Moll, D. 1976. Food and feeding strategies of the Ouachita map turtle (Graptemys pseudogeographica ouachitensis). American Midland Naturalist 96:478-482.

Nixon, C. M., M. W. McClain and K. R. Russell. 1970. Deer food habits and range characteristics in Ohio. The Journal of Wildlife Management 34: 870-886.

Parmenter, R. R. 1980. Effects of food availability and water temperature o the feeding ecology of pond slider (Chrysemys s. scripta). American Society of Ichthyologists and Herpetologists 1980: 503-514.

Pietruszka, R. D. 1981. An evaluation of stomach flushing for desert lizard diet analysis. The Southwestern Naturalist 26: 101-105.

Rosas-Rosas, O. C., R. Valdez, L. C. Bender and D. Daniel. 2003. Food habits of pumas in northwestern Sonora, Mexico. 2003. Wildlife Society Bulletin 31: 528-535.

Sato, Y., T. Mano and S. Takatsuki. 2000. Applicability of the point-frame method for quantitative evaluation of bear diet. Wildlife Society Bulletin 28: 311-316.

Seminoff, J. A., A. Resendiz and W. J. Nichols. 2002. Diet of east Pacific green turtles (Chelonia mydas) in the central gulf of California, Mexico. Journal of Herpetology 36: 447-453.

Stovall, F., M. Storm, L. Simon, G. Johnson, D. Schwarz and D.W. Kerbow. 1986. Clear Springs and Limestone Ledges, a History of San Marcos and Hays County for the Texas Sesquicentennial. Nortex Press, Austin, Texas, USA.

Vermersch, T. G. 1992. Lizards and turtles of south-central Texas. Eakin Press, Austin, TX.

Weckerly, F. W. and J. P. Nelson, Jr. 1990. Age and sex differences of white-tailed deer diet composition, quality and calcium. The Journal of Wildlife Management 54: 532-538.

Wrazen, J. A. and G. E. Svendsen. 1979. Sampling stomach contents of live small mammals using intubation. Journal of Mammalogy 60: 414-415.

Wutherich, D., A. Azocar, C. Garcia-Nunez and J. Silva. 2001. Seed dispersal in Palicourea ridida, a common treelet species from neotropical savannas. Journal of Tropical Ecology 17: 449-458.