Explaining patterns of deformity in freshwater turtles using MacCulloch s hypothesis

Transcription

1 433 Explaining patterns of deformity in freshwater turtles using MacCulloch s hypothesis Christina M. Davy and Robert W. Murphy Abstract: A growing body of literature details the effects of teratogenic chemicals on embryonic development in freshwater turtles. However, other factors affecting developmental deformities have not been recently considered and evaluation of the significance of deformities in adults is lacking. We collected 193 wild Midland Painted Turtles (Chrysemys picta marginata Agassiz, 1857) and 39 Common Snapping Turtles (Chelydra serpentina (L., 1758)) from an uncontaminated site in Ontario and recorded incidence of deformity of the shell, limbs, face, and tail. We tested MacCulloch s hypothesis (that incidence of deformity increases along a latitudinal gradient) by comparing our data with previously published deformity records from both uncontaminated and heavily polluted sites at varying latitudes. Incidence of nonembryonic deformity varied among wild populations and was not correlated with pollution levels. Thus adult deformity cannot be used as an indicator of site quality. Frequency of deformity increased with latitude in C. picta, supporting MacCulloch s hypothesis, whereas deformities in C. serpentina did not. We refer to essential differences in the biology of the two species to explain this disparity and recommend that latitudinal variation be included as a covariate in the future when developmental trends are compared among distant sites. Résumé : Une littérature de plus en plus abondante décrit les effets des produits chimiques tératogènes sur le développement embryonnaire des tortues d eau douce. Cependant, on n a pas considéré récemment les autres facteurs qui affectent les déformations du développement, ni évalué la signification des déformations chez les adultes. Nous avons récolté 193 tortues peintes (Chrysemys picta marginata Agassiz, 1857) et 39 tortues serpentines communes (Chelydra serpentina (L., 1758)) sauvages dans un site non contaminé en Ontario et avons noté la fréquence des déformations de la carapace, des membres, de la face et de la queue. Nous avons testé l hypothèse de MacCullough (selon laquelle la fréquence des déformations augmente en fonction d un gradient latitudinal) en comparant nos données à celles de la littérature, tant de sites non contaminés que fortement pollués, à diverses latitudes. La fréquence des déformations non embryonnaires varie d une population sauvage à une autre et n est pas en corrélation avec les niveaux de pollution. Les déformations des adultes ne peuvent donc pas servir d indicateurs de la qualité d un site. La fréquence des déformations augmente avec la latitude chez C. picta, ce qui appuie l hypothèse de MacCulloch, mais non chez C. serpentina. Nous invoquons les différences fondamentales dans la biologie de ces deux espèces pour expliquer cette disparité; nous recommandons d inclure dorénavant la variation longitudinale comme covariable dans la comparaison des tendances du développement entre des sites éloignés. [Traduit par la Rédaction] Introduction Developmental deformities with varying severity are frequently observed in wild populations of freshwater turtles (e.g., Zangerl and Johnson 1957; Ernst 1971; Ewert 1979; MacCulloch 1981; Mosimann 2002; Alarcos et al. 2005; Bell et al. 2006). Deformities range from being lethal, e.g., brain protruding from head (Bell et al. 2006), to very minor, e.g., extra scutes (Pavaliko 1986). Lethal deformities and deformities that reduce survival probability or reproductive fitness affect population dynamics, especially if the frequency Received 8 November Accepted 27 March Published on the NRC Research Press Web site at cjz.nrc.ca on 2 May C.M. Davy. 1 Department of Ecology and Evolutionary Biology, University of Toronto, 25 Wilcocks Street, Toronto, ON M5S 3B2, Canada. R.W. Murphy. Department of Ecology and Evolutionary Biology, University of Toronto, 25 Wilcocks Street, Toronto, ON M5S 3B2, Canada; Department of Natural History, Royal Ontario Museum, 100 Queen s Park, Toronto, ON M5S 2C6, Canada. 1 Corresponding author ( christina.davy@utoronto.ca). of such deformities is very high. However, minor deformities, such as missing digits or abnormally shaped scutes, are unlikely to have a significant impact on fitness. No evidence suggests that minor deformities are indicators of reduced reproductive fitness. Several factors have been suggested as causes of developmental deformities in turtles. Chemical teratogens are thought to cause severe deformity in turtle embryos, as was recently shown in wild populations of Painted Turtles (Chrysemys picta (Schneider, 1783)) and Common Snapping Turtles (Chelydra serpentina (L., 1758)) at a highly polluted site in Pennsylvania (Bell et al. 2006). Kyphosis, a dorsal ventral contortion of the spine, is thought to be a congenital anomaly (Wyneken et al. 2008). Lynn and Ullrich (1950) found that suboptimal moisture levels during incubation increase deformity frequency in C. picta; a lack of moisture during incubation can also cause a shell abnormality known as pyramiding (Wiesner and Iben 2003). Yntema (1960) also found that eggs of C. serpentina incubated at low temperatures produce abnormal hatchlings. Yntema s observations led MacCulloch (1981) to hypothesize that turtles in colder parts of their ranges might show higher levels of deformities, but this has not been tested. Can. J. Zool. 87: (2009) doi: /z09-028

2 434 Can. J. Zool. Vol. 87, 2009 Shell and spinal deformities in turtles range in severity. Severe deformities include missing plastra and extreme kyphosis (e.g., Bell et al. 2006; Wyneken et al. 2008). These are usually lethal and thus not observed in adults, and are probably rare under normal circumstances. However, mild deformities, such as abnormally shaped or extra scutes, are common in wild adult turtles (Ewert 1979; MacCulloch 1981; Ayres Fernández and Cordero Rivera 2004; Najbar and Szuszkiewicz 2006). Minor shell or spinal abnormalities are not known to affect fitness. Bell et al. (2006) provide a beautifully illustrated example of the devastating effects that pollutants can have on turtle embryonic development. Their data show that, on average, 55% of C. picta embryos taken from a site with high pollutant contamination developed lethal deformities, some involving the shell. But the frequency of minor deformities in adults (i.e., those deformities that probably do not decrease survivorship) was actually higher at their control site. The control site (the E.S. George Reserve, Michigan) was geographically distant, had no indication of contamination by... pollutants in the water, sediments or turtles, and no agricultural chemicals have been applied to the area since 1930 (E. Werner, personal communication). Furthermore, studies of the teratogenic effects of polychlorinated biphenyls (PCBs) and organochlorides on developing C. serpentina produce equivocal results. Bishop et al. (1998) indicate that teratogenic pollutants, especially polycyclic aromatic hydrocarbons, can significantly affect developing C. serpentina, although the impact may not be as severe as that documented for C. picta. However, exposure to other chemicals such as non-ortho PCBs and organochloride pesticides does not appear to increase rates of deformity. More recently, de Solla et al. (2008) found that although hatchling deformity levels in Ontario were higher at some polluted sites than at their control sites, other polluted sites did not show the same trend. They conclude that chemicals such as PCBs and organochlorides may not have a biologically significant effect and that other factors may also be acting as teratogenic agents. In adult turtles, they found deformity levels ranging from 0% (at a polluted site) to 21.1%, with a rate of 7% deformed adults at their clean reference site. Clearly, factors other than chemical pollutants are involved. Emphasis on embryonic deformities in the literature makes it difficult to assess what the natural background rate of deformity in wild populations might be. Data on normal levels of deformity in wild turtles from a variety of sites are necessary to provide context for more specialized studies, such as those mentioned above. This study compares published levels of deformities in juvenile and adult Midland Painted Turtles (Chrysemys picta marginata Agassiz, 1857) and C. serpentina with those observed at a natural site with good water quality. Our objectives were to describe deformities in wild adult turtles at a site not contaminated by teratogens and to determine (i) whether MacCulloch s hypothesis (i.e., levels of deformity increase along a latitudinal (temperature) gradient) is supported by the current data for C. picta or C. serpentina and (ii) whether frequencies of minor deformity are similar among populations from sites with varying levels of pollution, in which case they might provide a general indicator of site quality. Materials and methods We collected C. p. marginata and C. serpentina from the Old Ausable Channel (OAC), Pinery Provincial Park, Ontario (43816 N). Water quality in the OAC is excellent (Killins 2008; K. Jean, personal communication (2008)), as it is spring-fed and does not collect drainage from agricultural lands. The OAC is home to a large, healthy population of C. p. marginata and a high number of adult C. serpentina. Turtles were caught by hand and with dip nets from 22 May to 7 August Each turtle was measured (curved carapace length) and marked by shell notching following a method modified from Cagle (1939) to identify recaptures. Each turtle was sexed if possible and any deformities of the face, tail, limbs, or shell were recorded and photographed. We considered any deviations from the normal body plan of the two species that did not appear to be caused by traumatic injury (e.g., extra scutes, fused or extra digits, missing digits with no evidence of scarring to explain their absence) to be deformities. The types of deformities we observed were consistent with abnormalities termed deformities in previous studies. All turtles were released at their capture site. All fieldwork was conducted using an animal use protocol approved by the Animal Care Committee of the Royal Ontario Museum. Data from other C. picta populations were taken from Bell et al. (2006) and MacCulloch (1981). These included the John Heinz National Wildlife Refuge (JHNWR) in Pennsylvania (a heavily polluted site; N), the E.S. George Reserve (ESGR), Michigan (a site with no evidence of significant pollution; N), and the Qu appelle River (50842 N; a site with some agricultural drainage). Data from C. serpentina for comparison were taken from Bell et al. (2006) and de Solla et al. (2008). Sites used for comparison were the JHNWR, the ESGR, Tiny Marsh (a clean control site from de Solla et al. 2008; N), and Dead Creek and Raisin River (contaminated sites from the same study; N and N, respectively). We chose to use the Dead Creek and Raisin River sites because they represented the contaminated sites with the highest and lowest recorded rates of deformity, and therefore provided a wide range for comparison. We tested the relationship between latitude and frequency of deformity by calculating a standard Pearson s correlation coefficient (r 2 ). However, we were not able to use this approach to assess the role of pollutants. Previously published literature did not always provide measurements of pollution levels at all sites and the chemicals evaluated varied between studies. Thus, although comparable quantitative assessment of contamination would be ideal, this was not available for all sites. Therefore, we used Spearman s rank correlation coefficient (r) to test for correlations between relative contamination levels and frequency of deformities at sites. Grouping of sites and rationale for this classification are presented in Table 1. Results Chrysemys picta marginata We collected data from 193 C. p. marginata in the OAC.

3 Davy and Murphy 435 Table 1. Evidence for site contamination and contamination score used to rank sites. Contamination score* Source Site Quality and evidence 1 Killins 2008; K. Jean (née Killins), personal communication, 2008 OAC, Ontario NM; spring-fed, no agricultural runoff, other tests (nutrient levels) show very high water quality ESGR, Michigan NM; not treated with chemicals since at least Bell et al. 2006; E. Werner, personal communication, 2008 Tiny Marsh, Ontario Low levels of PCBs detected in eggs from this site (<50ng/g wet mass) 1 de Solla et al Qu appelle River, Saskatchewan NM; river received some agricultural runoff 2 R. MacCulloch, personal communication, de Solla et al Raisin River, Ontario Low to moderate levels of PCBs, PBDEs, and chlordanes detected in eggs from this site (~200 ng/g wet mass) 3 de Solla et al Dead Creek, Ontario Moderate levels of PCBs, PBDEs, and chlordanes detected in eggs from this site (~375 ng/g wet mass) JHNWR, Pennsylvania High levels of PAH, alkanes, and heavy metals measured 3 Bell et al Note: OAC, Old Ausable Channel; ESGR, E.S. George Reserve; JHNWR, John Heinz National Wildlife Refuge; NM, not measured (no quantitative assessment available); PCB, polychlorinated biphenyl; PBDE, polybrominated diphenyl ether; PAH, polycyclic aromatic hydrocarbons. *1, unpolluted or showing low levels of contamination; 2, showing evidence of some contamination by teratogens or evidence suggests that some contamination is likely and cannot be discounted; 3, highly contaminated by known teratogens. Fifty-three (27.5%) individuals had some type of developmental abnormality or asymmetry. Of the 53 abnormal individuals, 35.8% were male, 58.5% were female, and 5.7% were juveniles too small to sex accurately. In total, 97 abnormalities were recorded (some turtles had more than one visible abnormality). All abnormal individuals appeared healthy and behaved normally when handled and released. The most frequently observed abnormalities (n = 28 individuals) involved the scutes. Some of these turtles had extra or merged scutes on the carapace and plastron. Ten of these 28 individuals had abnormally shaped scutes, indicating developmental stress that nevertheless did not cause extra scutes to form. In three cases, turtles with abnormally shaped scutes also had extra scutes, and in one of these three, spinal malformation was also present. We also found several individuals with a convoluted central seam on the plastron (n = 22; Fig. 1); these are included in the other category of Table 2 to maintain consistency among compared studies. If these 22 turtles showing asymmetrical development of the plastral seam were included in the total count for shell abnormalities, then 42 individuals (21.7%) would be included. One turtle had a spinal defect (scoliosis), while four had deformed tail tips, such that the tails ended in spiky, modified scales, similar to those described in Bell et al. (2006). One individual had an unfused mandibular symphysis and one had a deformed limb, which appeared to have an extra process growing from the carpal joints (Fig. 1). We did not observe any incidence of ankylosis or kyphosis. Frequency of shell deformities in C. p. marginata among four sites increased with latitude (Fig. 2). There was a significant correlation between latitude and incidence of minor shell deformity (r 2 = , P < 0.05). The frequency of extra scutes is probably underestimated in our OAC data, as the number of scutes in the plastral bridge was not always recorded; this was the area that most frequently showed extra scutes in the Saskatchewan population (MacCulloch 1981). Chelydra serpentina Of 39 C. serpentina captured in the OAC, 9 (23.1%) had visible deformities (Table 2). Two of these were males and 7 were females. The heavy algal loads on the turtles carapaces made it difficult to assess carapacial scute patterns, so only extra scutes on the plastron were recorded. The most noteworthy deformities we observed were a cleft palate (unfused maxillae) in an adult female and an unarticulated plastral bridge in another adult female. We were not able to isolate and compare shell-only deformities for C. serpentina, since deformity types for adults were not defined for all comparison sites. We therefore compared total frequency of observed deformity at the six sites. Latitude and frequency of deformity were not significantly correlated (Fig. 3), although P was approaching significance (r 2 = , P=0.056). As in C. p. marginata, there was no apparent correlation between site contamination and minor deformities. There was no correlation between contamination score (Table 1) and incidence of deformities in either C. picta (r = 0.105, P > 0.05) or C. serpentina (r = 0.639, P > 0.05); the lowest observed frequency of deformity in adults occurred at the most heavily polluted site.

4 436 Can. J. Zool. Vol. 87, 2009 Fig. 1. Examples of scute and limb deformities from turtles from Old Ausable Channel, Ontario. Midland Painted Turtle (Chrysemys picta marginata): (A) convoluted central plastral seam; (B) extra vertebral scutes; dovetail syndrome; mild scoliosis (spine curves toward the right; see lateral view, C); (D) growth on left hind leg at joint between carpals and tibia, which felt as though it contained several small bones; looked like a separate deformed (almost phocomelic) appendage. Common Snapping Turtle (Chelydra serpentina): (E) disarticulated plastral bridge in an adult female; (F, G) adult female with unfused maxillae. The left eye is also either undeveloped or missing owing to an infection. Discussion We found that site contamination does not explain the variation in frequency of deformity in either C. p. marginata or C. serpentina. Therefore, the incidence of deformity at a site cannot be used as an indicator of habitat quality or contamination. However, the strong correlation between latitude and incidence of minor deformities in wild C. p. marginata supports MacCulloch s hypothesis, and is most likely caused by 2004 climatic factors, including temperature. These factors may be involved during incubation. Correlations between experimental suboptimal moisture or temperature conditions and abnormal development of turtle embryos (Lynn and Ullrich 1950; Yntema 1960) are well documented. Temperature may also play an important role in the development of eggs within the female, since egg content (including water content) is correlated with climatic factors such as temperature and rainfall (Finkler et al. 1994), and can affect embryo development. Interestingly, MacCulloch s hypothesis is only weakly supported by the data for C. serpentina. The correlation between latitude and deformity in this species was only approaching statistical significance, but given the number of other factors that may influence deformity frequency, the correlation may be biologically significant. The incidence of deformities in adult C. serpentina in the OAC is higher than that reported for this species elsewhere. We may be underestimating deformity levels in this species at our site for the reasons previously stated. Adult C. serpentina in the OAC move in and out of the channel (C.M. Davy, unpublished data), thus our sampled adults may have been incubated elsewhere. Nevertheless, teratogenic compounds cannot be convincingly implicated in the deformities we report here. Genetic factors may be responsible for some deformities, although this possibility remains to be tested. Regardless, our data will contribute to future comparative studies. Where data are available, C. serpentina populations consistently show lower incidence of deformity compared with the sympatric C. p. marginata. Bell et al. (2006) suggest that the higher rate of deformities in C. picta might reflect a greater susceptibility to pollution in this species; their embryonic data support this hypothesis. However, their adult data do not support the pollution hypothesis, and yet

5 Davy and Murphy 437 Fig. 2. Frequency of shell (scute and spinal) deformities in juvenile and adult Midland Painted Turtles (Chrysemys picta marginata) from four sites, in order of increasing latitude. Latitude and deformity frequency are significantly correlated (r 2 = , P < 0.05). JHNWR, John Heinz National Wildlife Refuge; ESGR, E.S. George Reserve. Fig. 3. Frequency of observed external deformities in Common Snapping Turtles (Chelydra serpentina) from six sites, in order of increasing latitude. Latitude and frequency of deformity are positively correlated (r 2 = , P = 0.056). JHNWR, John Heinz National Wildlife Refuge; ESGR, E.S. George Reserve.

6 438 Can. J. Zool. Vol. 87, 2009 Table 2. Frequency of types of externally visible deformities in wild Painted Turtles (Chrysemys picta and Chrysemys picta marginata) and Common Snapping Turtles (Chelydra serpentina) from sites at varying latitudes. Type of abnormality Shell Spine alignment Limbs Other Source Jaw or facial Scutes Other No. of recorded abnormalities Tail No. of abnormal individuals Location n Chrysemys picta and Chrysemys picta marginata OAC, Ontario (27.5) 69 4.(2.1) 1.(0.5) 28.(14.5) ND 1.(0.5) 1.(0.5) 22.* (11.4) This study JHNWR, Pennsylvania (17.9) (3.5) 3.(1.0) 27.(8.5) 7.(2.2) 2.(0.6) 21.(6.6) ND Bell et al ESGR, Michigan (14.2) (14.2) ND 34.(0.8) Bell et al Qu appelle River, Saskatchewan { (25) 39 ND ND 28.(21.88) ND 2.(1.5) ND 2.* MacCulloch 1981 Chelydra serpentina OAC, Ontario 39 9 (23.1) (15.4) 0. 2.(5.1) 1.(2.5) 1. { (2.5) This study JHNWR, Pennsylvania (5.8) 11 2.(1.2) 1.(0.6) 3.(1.7) 2.(1.2) 2.(1.2) 0. 1.(0.6) Bell et al ESGR, Michigan (6.0) 72 ND ND ND ND ND ND ND Bell et al Note: Values in parentheses are percentages; OAC, Old Ausable Channel; JHNWR, John Heinz National Wildlife Refuge; ESGR, E.S. George Reserve; ND, no data available. Deformity data from de Solla et al. (2008) are not included, since sample size and type of deformity are not available for these sites. *Convoluted central seam on plastron. { Values from Table 1 of MacCulloch (1981) have been adjusted to include two individuals with kyphosis and two individuals with a convoluted central seam on the plastron, following the text. { Disconnected plastral bridge. still show higher total deformity in C. picta, as do our data (27% for C. p. marginata vs. 23% for C. serpentina). We hypothesize that the variance reflects different nesting strategies between the two species. Chelydra serpentina generally lays only one clutch per season (Congdon et al. 2008), so the hatchlings usually emerge before temperatures drop significantly. In contrast, C. picta may lay one or two clutches per season (Congdon et al. 2003; Samson 2003), with the second clutch often overwintering in the nest (Breitenbach et al. 1984; Ernst et al. 1994). These later clutches can experience significant drops in temperature during their development, especially at higher latitudes. Thus, the higher frequency of deformity observed in wild C. picta may represent clutches laid late in the season; this hypothesis should be relatively simple to test. We recorded a slightly higher proportion of females with deformities in both species, as did MacCulloch (1981), and this trend conforms to our expectations. Both our study species exhibit temperature-dependent sex determination. Both produce females when incubated at high ( C) or very low ( C) temperatures and males when incubated at intermediate temperatures ( C; Yntema 1979; Gutzke and Paukstis 1984). Therefore, we would predict that if suboptimal temperatures are causing higher levels of deformity, then significantly more females (having developed at the lowest viable incubation temperatures) should develop deformities; the data support this prediction. The effects of latitude and environmental pollutants on embryonic development of turtles likely interact in complex ways. Perhaps some individuals are genetically predisposed to develop asymmetrically. Such individuals might develop asymmetries (asymmetrical scutation, slight scoliosis) if incubated at suboptimal temperatures but would not develop major deformities under reasonably good conditions. This would explain the higher levels of minor deformity at less polluted sites. However, these individuals might be more likely to develop severe deformities if exposed to teratogens, and therefore not survive. Thus, we would expect a slightly lower incidence of minor deformities at severely contaminated sites because individuals predisposed to asymmetrical development will develop more extreme, lethal deformities instead of surviving to adulthood. Indeed, Bell et al. (2006) found that although the overall percentage of deformed adult C. picta did not differ greatly between their polluted and nonpolluted sites (17.9% and 14.2%, respectively), the severity of deformities in adults at the polluted JHNWR site was much higher. This suggests that genetic or climatic predispositions to asymmetrical development were aggravated by the presence of teratogenic chemicals. We reiterate the findings of de Solla et al. (2008), keeping MacCulloch s hypothesis in mind. Their data show that the frequency of deformity in hatched individuals (i.e., those who have a chance to survive to become deformed adults) is not significantly correlated with chemical contamination. However, they did find a significant difference between their two reference sites (both of which are in the northern part of their sample range and one of which is their most northern site) and three contaminated sites (all of which are southerly and two of which are their most southern sites). Latitudinal variation is a potential explanation for this trend; we recommend that future evaluations of deformities among

7 Davy and Murphy 439 variably contaminated sites should consider latitudinal variation among sites as a covariable. Many factors clearly influence embryonic development; teasing these apart in wild populations may not always be possible. Clearly environmental pollutants can pose a threat to wild turtles and other wildlife; the recent studies showing this are valuable resources for population management and conservation. The maintenance or restoration of clean, highquality habitat is crucial to the protection of many species. However, without detracting from the importance of conserving a pristine environment, we nevertheless hope that the analysis presented here provides reason to think twice before attributing the bulk of observed deformities in wild freshwater turtles primarily to the effects of chemical pollutants. Acknowledgements This study was made possible by research permits from Ontario Parks and the Ontario Ministry of Natural Resources (permit # ). We thank Terry Green and Hugh Fraser for generously providing accommodations and canoe use at the field site. Suzanne Coombes, Eric Davy, and Leila Copes provided indispensable field assistance. Thanks go to Alistair Mackenzie (Pinery Provincial Park) for further access to canoes and Michael Dloogatch for access to a back issue of the Bulletin of the Chicago Herpetological Society. R. MacCulloch and an anonymous reviewer provided helpful comments on the manuscript. This research was funded by an Ontario Graduate Scholarship to C.D. and by a Natural Sciences and Engineering Research Council of Canada Discovery Grant to R.W.M. References Alarcos, G., Ortiz, M., Fernández-Benítez, M.J., and Lizana, M Preliminary data on the structure of an Emys orbicularis stream population in Los Arribes del Duero (Zamora, Spain). In Proceedings of the 4th International Symposium on Emys orbicularis, Valencia, Spain, 8 10 June pp Ayres Fernández, C., and Cordero Rivera, A Asymmetries and accessory scutes in Emys orbicularis from Northwest Spain. Biologia (Bratisl.), 59(Suppl. 14): Bell, B., Spotila, J.R., and Congdon, J High incidence of deformity in aquatic turtles in the John Heinz National Wildlife Refuge. Environ. Pollut. 142: doi: /j.envpol PMID: Bishop, C.A., Ng, P., Pettit, K.E., Kennedy, S.W., Stegeman, J.J., Norstrom, R.J., and Brooks, R.J Environmental contamination and developmental abnormalities in eggs and hatchlings of the common snapping turtle (Chelydra serpentina serpentina) from the Great Lakes St. Lawrence River basin ( ). Environ. Pollut. 101: doi: /s (98) PMID: Breitenbach, G.L., Congdon, J.D., and van Loben Sels, R.C Winter temperatures of Chrysemys picta nests in Michigan: effects on hatchling survival. Herpetologica, 40: Cagle, F.R A system of marking turtles for future identification. Copeia, 1939: Congdon, J.D., Nagle, R.D., Kinney, O.M., van Loben Sels, R.C., Quinter, T., and Tinkle, D.W Testing hypotheses of aging in long-lived Painted Turtles (Chrysemys picta). Exp. Gerontol. 38: doi: /s (03) PMID: Congdon, J.D., Greene, J.L., and Brooks, R.J Reproductive and nesting ecology of female snapping turtles. In Biology of the snapping turtle (Chelydra serpentina). Edited by A.C. Steyermark, M.S. Finkler, and R.J. Brooks. Johns Hopkins University Press, Baltimore, Md. pp de Solla, S.R., Fernie, K.J., and Ashpole, S Snapping turtles (Chelydra serpentina) as bioindicators in Canadian areas of concern in the Great Lakes Basin. II. Changes in hatching success and hatchling deformities in relation to persistent organic pollutants. Environ. Pollut. 153(3): doi: /j.envpol PMID: Ernst, C.H Observations of the painted turtle, Chrysemys picta. J. Herpetol. 5: doi: / Ernst, C.H., Barbour, R.W., and Lovich, J.E Turtles of the United States and Canada. Smithsonian Institution Press, Washington, D.C. Ewert, M.A The embryo and its egg: development and natural history. In Turtles: perspectives and research. Edited by M. Harless and H. Morlock. Wiley, New York. pp Finkler, M.S., Steyermark, A.C., and Jenks, K Geographic variation in snapping turtle (Chelydra serpentina) egg components across a longitudinal transect. Can. J. Zool. 82: doi: /z Gutzke, W.H.N., and Paukstis, G.L A low threshold temperature for sexual differentiation in the painted turtle, Chrysemys picta. Copeia, 1984: doi: / Killins, K A management plan for the Old Ausable Channel Watershed. Ausable Bayfield Conservation Authority, Port Franks, Ont. Available from news_events.html [accessed 7 November 2008]. Lynn, W.G., and Ullrich, M.C Experimental production of shell abnormalities in turtles. Copeia, 1950: doi: / MacCulloch, R Variation in the shell of Chrysemys picta belli from southern Saskatchewan. J. Herpetol. 15(2): doi: / Mosimann, D Situation einer Population von Europaischen Sumpfschildkrote, Emys orbicularis (Linnaeus 1758), 50 Jahre nach der ersten Ansiedlung in Moulin-de-Vert (Genf, Schwietz). Testudo, 11(4): Najbar, B., and Szuszkiewicz, E The morphometrics and colouration of the European pond turtle Emys orbicularis in Lubuskie province (West Poland). Biologia (Bratisl.), 61: doi: /s Pavaliko, P Shell and scute anomalies in some midwestern turtles. Bull. Chic. Herpetol. Soc. 21: Samson, J The life history strategy of a northern population of midland painted turtles, Chrysemys picta marginata. M.Sc. thesis, Department of Zoology, University of Guelph, Guelph, Ont. Wiesner, C.S., and Iben, C Influence of environmental humidity and dietary protein on pyramidal growth of carapaces in African spurred tortoises (Geochelone sulcata). J. Anim. Physiol. Anim. Nutr. (Berl.), 87: doi: /j x. PMID: Wyneken, J., Godfrey, M.H., and Bels, V Biology of turtles. CRC Press, Boca Raton, Fla. Yntema, C.L Effects of various temperatures on the embryonic development of Chelydra serpentina. Anat. Rec. 136: Yntema, C.L Temperature levels and periods of sex determination during incubation of eggs of Chelydra serpentina. J. Morphol. 159: doi: /jmor Zangerl, R., and Johnson, R.G The nature of shield abnormalities in the turtle shell. Fieldiana Geol. 10: