Size-biased mortality due to predation in a nesting freshwater turtle, Trachemys scripta.
|Abstract:||Differential survivorship within or between stages is an important component of most explanations of turtle reproductive patterns. We tested the null hypothesis that a sample of adult female Trachemys scripta found killed by predators was a random sample of adult females found nesting at a site in W-central Illinois by comparing plastron lengths of the two samples. Mean plastron length of 19 dead female T. scripta was significantly smaller than mean plastron length of 79 females found alive. Apparently smaller females were at greater risk of mortality than were larger ones. Cubic spline analysis indicated that mortality was strongly concentrated among the smallest and perhaps youngest nesting females and was distinctly nonlinear. This finding was consistent with the suggestion that the minimum threshold of maturation size was influenced by the size at which the probability of predation decreases.|
Predation (Biology) (Research)
Reproduction (Physiological aspects)
Tucker, John K.
Janzen, Fredric J.
|Publication:||Name: The American Midland Naturalist Publisher: University of Notre Dame, Department of Biological Sciences Audience: Academic Format: Magazine/Journal Subject: Biological sciences; Earth sciences Copyright: COPYRIGHT 1999 University of Notre Dame, Department of Biological Sciences ISSN: 0003-0031|
|Issue:||Date: Jan, 1999 Source Volume: 141 Source Issue: 1|
Life-history traits of turtles such as longevity and delayed sexual maturity fit a bet-hedging model of reproduction (Stearns, 1976; Galbraith and Brooks, 1987). Bet-hedging theory predicts that when juvenile mortality is high but adult mortality is low (i.e., the risk of mortality is size-dependent), lifetime reproductive success is enhanced by maximizing growth rates and delaying sexual maturity. Because the onset of sexual maturity coincides with a reduction in growth rate in turtles (Gibbons and Lovich, 1990), delaying sexual maturity permits attainment of larger body sizes more quickly and reduces the time spent at smaller and possibly more vulnerable sizes (Gibbons and Lovich, 1990; Kennett, 1996). Besides possible survival advantages, larger body size may allow turtles to produce more and larger eggs (see Tucker and Moll, 1997, for a review). If, however, survivorship of juveniles and small nesting females is not lower than survivorship of adult or larger nesting turtles, then reproduction at the smallest size at which viable eggs can be produced is favored (Kennett, 1996).
Differential survivorship between-stages (e.g., juvenile vs. adult) and/or within-stages (e.g., smaller vs. larger nesting females) is thus an important assumption of most explanations of turtle reproductive patterns (Galbraith and Brooks, 1987; Gibbons and Lovich, 1990; Kennett, 1996). For example, Moll (1994) attributed large size of adult females in a sea beach nesting population of a slider turtle (Trachemys scripta venusta) to the hazards of the long nesting excursions of this turtle. This hypothesis assumes that larger females are less susceptible to predation than smaller females. Similarly, Kennett (1996), in part, attributed differences in growth patterns between two Australian turtles to higher risks faced by Chelodina rugosa in an ephemeral habitat compared to those of Elseya dentata in more stable lotic environments. He suggested that rapid growth of the former was under strong selection due to suspected higher risk of size-dependent mortality in C. rugosa compared to E. dentata.
Higher mortality of juveniles relative to adults is common in natural populations of freshwater turtles. Examples include Kinosternon subrubrum, Trachemys scripta scripta, and Deirochelys reticularia (Gibbons, 1987; Frazer et al., 1990, 1991a); Chelydra serpentina (Galbraith and Brooks, 1987; Congdon et al., 1994); Chrysemys picta (Frazer et al., 1991b); Kinosternon flavescens (Iverson, 1991); and Emydoidea blandingii (Congdon et al., 1993; Herman et al., 1994).
Size-related differences in survivorship may also occur within-stages (e.g., Janzen, 1993), but comparisons of mortality between smaller and larger adult turtles have not been reported. Because they make periodic overland nesting forays, adult female aquatic turtles may be particularly at risk. We collected dead adult females, apparently killed by one or more predators, at a nesting area in W-central Illinois to determine whether or not they were a representative sample of the turtles found nesting at this site. Our study is important because selection for size at maturity should be a trade-off between the minimum size at which viable eggs can be produced and the size at which the risk of mortality is offset by reproductive fitness (Gibbons and Lovich, 1990; Kennett, 1996). This trade-off requires size-dependent mortality such that smaller turtles must be at greater risk of death than larger ones. Specifically, we tested the hypothesis that smaller nesting turtles are at greater risk of mortality than are larger ones.
Living turtles were collected on 28 and 29 May and between 11 June and 25 June 1996. Collections were not made between 30 May and 10 June due to flooding of roads to the site. Dead turtles were collected between 11 June and 16 June but were also searched for each day that the site was visited to collect nesting turtles.
The site is a grass airstrip maintained by occasional mowing and managed by the U.S. Fish and Wildlife Service, Mark Twain National Wildlife Refuge. Turtles nesting at this site occupy Pohlman Slough, a backwater of the Mississippi River, in Calhoun County, Illinois. Turtles were collected on the airstrip and along the road separating the airstrip from Pohlman Slough. The plastron length of living (n = 79) and dead turtles (n = 19) was measured to the nearest millimeter with Haglof[R] 40 cm calipers. Living turtles were released at the collecting site following measurement. All dead turtles found were females as judged by their large size and domed carapaces (Cagle, 1950; Gibbons and Lovich, 1990).
We used SAS for statistical analyses (SAS Institute, 1988). We compared plastron lengths of living and dead turtles using the Wilcoxon rank sums test (= z) because variables were not normally distributed in each classification. We used Spearman's rank correlation coefficient to examine the association between female plastron length and date of collection during the nesting period. Means in the text are accompanied by [+ or -] 1 SD and the range.
We evaluated selection on plastron length of nesting turtles in three complementary ways. First, to estimate the standardized strength of selection ([Beta][prime]), we employed linear regression analysis, with relative fitness as the dependent variable (0 for dead turtles and 1/(79/98) = 1.24 for living turtles) and plastron length (divided by its standard deviation) as the independent variable (Lande and Arnold, 1983; Janzen, 1993; Brodie et al., 1995). With this method, the slope of the regression is equivalent to the strength of selection acting on plastron length in terms of standard deviations. However, because this approach may not provide accurate statistical testing due to nonnormality of the dependent variable (e.g., Mitchell-Olds and Shaw, 1987), we used logistic regression to determine how well plastron length predicted survivorship (0 for dead turtles and 1 for living turtles). Finally, we used a cubic spline algorithm to aid in visualizing the form of selection on plastron length (Schluter, 1988; Brodie et al., 1995). This technique of nonlinear regression permits visualization of the potentially complex relationships between a continuously distributed trait (i.e., plastron length) and a discrete fitness estimate (i.e., survivorship), which typical linear and even logistic regression approaches may fail to do adequately (Schluter, 1988). Standard errors for the best-fit spline function were calculated by bootstrapping the original data 100 times.
Overall, 19 dead, female, red-eared sliders (Trachemys scripta elegans) were collected. The heads and all four legs of all dead turtles were missing at discovery. Mean plastron length of dead female Trachemys scripta elegans (205.3 [+ or -] 11.4 mm, 179-220 mm, N = 19) was significantly less than that of females found alive (213.5 [+ or -] 14.0 mm, 187-258 mm, N = 79) at Pohlman Slough (Z = -1.99, P = 0.05, 1 df). Our results did not support the null hypothesis that dead turtles were a representative sample of turtles found nesting at the site, suggesting that the risk of mortality was size-dependent [ILLUSTRATION FOR FIGURE 1 OMITTED].
The standardized strength of selection on plastron length was moderate in magnitude ([Beta][prime] = 0.115, SE = 0.049). In other words, this level of selection produced a "within-generation" increase in plastron length of nearly one-eighth of a standard deviation. Logistic regression analysis indicated that this relationship between plastron length and survivorship was statistically significant ([[Chi].sup.2] = 5.37, P = 0.02, 1 df). Cubic spline analysis showed that mortality was distinctly nonlinear with respect to plastron length, however [ILLUSTRATION FOR FIGURE 2 OMITTED]. Survivorship was greatly reduced among females smaller than 200 mm in plastron length, high and stable among females 200-220 mm in plastron length, and then sharply increased again among females with plastron lengths exceeding 220 mm [ILLUSTRATION FOR FIGURE 2 OMITTED].
The cubic spline analysis of the form of selection on plastron length suggests that mortality is distinctly nonlinear and is concentrated among the smallest and, perhaps, youngest nesting turtles (Tucker and Moll, 1997) [ILLUSTRATION FOR FIGURE 2 OMITTED]. Survivorship levels are high and relatively constant for turtles with plastron lengths of 200-220 mm and then improve even more for the very largest individuals. Although linear regression analysis documented the largely directional pattern of selection on plastron length and estimated its strength, it could not capture the complexity of the form of selection elucidated by the nonparametric cubic spline approach (see discussion in Schluter, 1988). In fact, the shape of the spline function is crucial to interpretation of the results of this study, because it suggests that the minimum threshold of maturation size in female turtles may be influenced by the size at which the probability of predation decreases (Gibbons and Lovich, 1990). Thus, our study provides evidence that supports critical assumptions of life-history theory.
Our findings are important because differential mortality related to size is a central tenet of bethedging theory. Evidence from between-stage comparisons (i.e., juvenile vs. adult) from demographic studies cited above and a few other studies (e.g., Burbidge, 1981; Kennett, 1996) support the hypothesis that mortality is higher among juvenile turtles than adults. However, the within-stage hypothesis that smaller nesting females are at greater risk of death than are larger ones has never been tested. Our study comparing size of turtles found dead vs. turtles captured alive is the first empirical test of the within-stage fitness hypothesis for adult turtles. Importantly, the small but significant difference in size between dead and living turtles is consistent with the hypothesis that smaller nesting turtles experience greater risk of predation.
Because we did not witness any one predation event at the site, the dead turtles may not have been killed by predators. However, this seems unlikely to us. In one instance, we interrupted a predation event at 1900 h CDT on the evening of 12 June. We found an overturned gravid female (225 mm plastron length, 2025 g mass) that was alive but had most of her tail removed and had wounds on both hind feet. Her wounds were bleeding when she was collected.
However, we cannot identify the actual predator. We found scats and tracks of both coyotes (Canis latrans) and raccoons (Procyon lotor) at the study site near Pohlman Slough. Both are known to prey on adult turtles (reviewed by Gibbons, 1987; Ernst et al., 1994). We suspect that the predation that we observed was due to raccoons because raccoon tracks were associated with carapaces in areas where tracks could be detected.
Deviation of the size distribution of dead turtles from live turtles may occur for a number of reasons in addition to size-biased predation. For instance, predators may have preferentially carried larger animals away to locations outside the search area or to sites where they would be less likely to be seen compared to smaller turtles. However, these possibilities are unlikely to account for our observations at the Pohlman Slough nesting site. First, flooding isolated the elevated airstrip near Pohlman Slough resulting in an elongated island. One or more predators remained on this island and exploited turtles as food when they emerged to nest during the flood episode. Thus, a predator would have had to enter flood waters to remove turtles from the area. Moreover, the vegetation on the airstrip is mowed and carapaces of dead turtles were easily seen. The airstrip is bordered by agricultural fields that at the time of collection were newly planted and also were easily searched.
Another possible source of bias might be that small live turtles are less likely to be found than larger live turtles due to some unrecognized collecting bias. For instance, smaller turtles may be more likely to nest at times when investigators were not present. Smaller females might also be more easily overlooked than larger ones. If either were the case, then smaller turtles would have been underrepresented in the sample of living turtles examined. However, nearly all turtles in our study area nest between 0700-1300 h CDT (Tucker, 1997) and we concentrated our collecting efforts during these hours. Moreover, little cover exists at the Pohlman Slough site making it unlikely that we would miss even the smallest nesting females.
Because the predation event that we studied occurred early in the nesting season but living turtles were collected throughout the nesting season, it was important to examine the possibility that small turtles are more likely to nest earlier than larger ones. Importantly, we found no evidence that size of nesting females is correlated with Julian date of collection (Spearman's Rho = 0.09, P = 0.45, N = 79) among turtles collected alive at Pohlman Slough.
We are left with the conclusion that predators have caused size-dependent mortality in our sample of nesting red-eared slider turtles, consistent with important life-history theory. Although possibly rare, such events of mass mortality may be important nonetheless in influencing life-history evolution of long-lived vertebrates. Considering that [less than]2% of female Trachemys scripta hatchlings may reach sexual maturity (Frazer et al., 1990), any source of size-dependent mortality might be particularly significant in driving evolution of delayed sexual maturity in this (and other) long-lived species. Consequently, study of predation in adult turtles, even if it is rare, is important for understanding the evolution of turtle life-history strategies.
Acknowledgments. - Moynell M. Tucker helped search for dead turtles. Karen L. Drews (U.S. Fish and Wildlife Service, Mark Twain National Wildlife Refuge) allowed us to collect at Pohlman Slough. We thank E. D. Brodie III and A.M. Bronikowski for helpful comments on the manuscript. This study was partially supported by the Illinois Natural History Survey and the Long Term Resources Monitoring Program. Janzen received support from NSF grant DEB-96-29529. Journal Paper No. J-17831 of the Iowa Agriculture and Home Economics Experiment Station, Ames, Iowa, Project No. 3369, and supported by the Hatch Act and State of Iowa funds.
BRODIE, E. D. III, A.J. MOORE AND F. J. JANZEN. 1995. Visualizing and quantifying natural selection. Trends Ecol. Evol., 30:313-318.
BURBIDGE, A. A. 1981. The ecology of the western swamp tortoise Pseudemydura umbrina (Testudines: Chelidae). Aust. Wildl. Res., 8:203-223.
CAGLE, F. R. 1950. The life history of the slider turtle, Pseudemys scripta troostii (Holbrook). Ecol. Monogr., 20:31-54.
CONGDON, J. D., A. E. DUNHAM AND R. C. VAN LOBEN SELS. 1993. Delayed sexual maturity and demographics of Blanding's turtles (Emydoidea blandingii): implications for conservation of long-lived organisms. Conserv. Biol., 7:826-833.
-----, ----- AND ------. 1994. Demographics of common snapping turtles (Chelydra serpentina): Implications for conservation and management of long-lived organisms. Am. Zool., 34:397-408.
ERNST, C. H., J. E. LOVICH AND R. W. BARBOUR. 1994. Turtles of the United States and Canada. Smithsonian Institution Press, Washington, D.C. 578 p.
FRAZER, N. B., J. W. GIBBONS AND J. L. GREENE. 1990. Life tables of a slider turtle population, p. 183-200. In: J. W. Gibbons (ed.). Life history and ecology of the slider turtle. Smithsonian Institution Press, Washington, D.C.
-----, ----- AND ----- 1991a. Life history and demography of the common mud turtle, Kinosternon subrubrum in South Carolina, USA. Ecology, 72:2218-2231.
-----, ----- AND -----. 1991b. Growth, survivorship and longevity of painted turtles Chrysemys picta in a southwestern Michigan marsh. Am. Midl. Nat., 125:245-258.
GALBRAITH, D. A. AND R.J. BROOKS. 1987. Survivorship of adult females in a northern population of common snapping turtles, Chelydra serpentina. Can. J. Zool., 65:1581-1586.
GIBBONS, J. W. 1987. Why do turtles live so long? Bioscience, 37:262-269.
----- AND J. E. LOVICH. 1990. Sexual dimorphism in turtles with emphasis on the slider turtle (Trachemys scripta). Herpetol. Monogr., 4:1-29.
HERMAN, T B., T D. POWER AND B. R. EATON. 1994. Status of Blanding's turtles, Emydoidea blandingii, in Nova Scotia, Canada. Can. Field-Nat., 109:182-191.
IVERSON, J. B. 1991. Life history and demography of the yellow mud turtle, Kinosternon flavescens. Herpetologica, 47:373-395.
JANZEN, F.J. 1993. An experimental analysis of natural selection on body size of hatchling turtles. Ecology, 74:332-341.
KENNETT, R. 1996. Growth models for two species of freshwater turtle, Chelodina rugosa and Elseya dentata, from the wet-dry tropics of northern Australia. Herpetologica, 52:383-395.
LANDE, R. AND S.J. ARNOLD. 1983. The measurement of selection on correlated characters. Evolution, 37:1210-1226.
MITCHELL-OLDS, T. AND R. G. SHAW. 1987. Regression analysis of natural selection: statistical inference and biological interpretation. Evolution, 41:1149-1161.
MOLL, D. 1994. The ecology of sea beach nesting in slider turtles (Trachemys scripta venusta) from Caribbean Costa Rica. Chelonian Conserv. Biol., 1:107-116.
SAS INSTITUTE, INC. 1988. SAS/STAT user's guide. Statistical analysis systems institute, Inc., Cary, N.C. 1028 p.
SCHLUTER, D. 1988. Estimating the form of natural selection on a quantitative trait. Evolution, 42:849-861.
STEARNS, R. C. 1976. Life-history tactics: A review of the ideas. Q. Rev. Biol., 51:3-47.
TUCKER, J. K. 1997. Natural history notes on nesting, nests, and hatchling emergence in the red-eared slider turtle, Trachemys scripta elegans in west-central Illinois. Ill. Nat. Hist. Surv. Biol. Notes, 140:1-13.
----- AND D. MOLL. 1997. Growth, reproduction, and survivorship in the red-eared turtle, Trachemys scripta elegans, in Illinois, with conservation implications. Chelonian Conserv. Biol., 2:352-357.
JOHN K. TUCKER, Illinois Natural History Survey, Great Rivers Field Station, Long Term Resource Monitoring Program-Reach 26, 4134 Alby Street, Alton, Illinois 62002; NIRVANA I. FILORAMO(1) and FREDRIC J. JANZEN, Department of Zoology and Genetics, Iowa State University, Ames, 50011. Submitted 14 April 1997; Accepted 9 April 1998.
1 Present address: Department of Ecology and Evolutionary Biology, University of Connecticut, Storrs, 06269
|Gale Copyright:||Copyright 1999 Gale, Cengage Learning. All rights reserved.|