Seasonal Variation in Basking in Two Syntopic Species of Map Turtles (Emydidae: Graptemys)
Abstract
The purpose of this study was to quantify seasonal variation in basking habits within 2 syntopic species of map turtles (Graptemys pseudogeographica kohnii and Graptemys ouachitensis sabinensis) in eastern Texas. Basking surveys were conducted twice per week from January 2005 to January 2006 along a 1.5-km transect of the Sabine River in Smith County. During this period, 312 observations of male and 178 observations of female G. pseudogeographica kohnii were recorded, while 2761 observations of male and 1334 observations of female G. ouachitensis sabinensis were made. Basking activity peaked in March (G. pseudogeographica kohnii) and May (G. ouachitensis sabinensis) with a peak of lesser magnitude for both species occurring in November. Within G. ouachitensis sabinensis, significantly more males than females were recorded basking throughout the year; however, in G. pseudogeographica kohnii significantly more males were recorded only during spring and fall.
In many freshwater turtle genera, such as Graptemys, Pseudemys, Trachemys, and Chrysemys, basking is a highly visible component of daily activity (Boyer 1965). Turtles may bask for several reasons, including 1) to promote digestion, 2) to remove parasites and algae, 3) to promote synthesis of vitamin D, 4) to increase metabolic rates, and 5) to accelerate the rate of egg development in adult females (Auth 1975; Avery 1982; Janzen et al. 1992).
Previous researchers have suggested that seasonal differences in basking between males and females of Trachemys and Graptemys result from their different reproductive strategies. Males require energy to search for a mate; whereas, females expend energy to produce and lay eggs, suggesting that during the annual cycle males will become active before females and remain active later in the year (Trivers 1972; Morreale et al. 1984; Thomas et al. 1999). Morreale et al. (1984) found that male Trachemys scripta increased activity prior to the breeding season, usually in early spring and late fall; whereas, numbers of basking females increased to approximate those of males in April, May, and June. The inequality in parental investment leads to an assumption that males will maximize fitness by basking for longer periods between April and October (Thomas et al. 1999). In studies of Graptemys oculifera (Jones and Hartfield 1995) and Graptemys geographica (Pluto and Bellis 1986), basking males were found to outnumber basking females. Because basking is highly developed in map turtles and most map turtle species are found in the southern United States, they provide a good opportunity to examine sexual differences in basking throughout the seasons.
The objective of this study was to investigate whether there are seasonal and/or sexual differences in basking frequencies within and between 2 syntopic species of map turtles, Graptemys pseudogeographica kohnii and Graptemys ouachitensis sabinensis. This study provides a quantitative comparison of basking habits in these 2 species throughout 1 year.
Methods. — Texas Parks and Wildlife Department's Old Sabine Bottom Wildlife Management Area (WMA) is located 8.1 km north of Lindale, Texas, and is situated between Smith and Wood counties. The WMA is a 2318-ha bottomland hardwood forest, with the Sabine River marking the boundary between the 2 counties. The study site is a 1.5-km transect along the Sabine River that consists of steep banks, sandy shores, and bends in the river. Hardwood trees consisting of oak, elm, ash, and a mixture of diverse understory species line both sides of the river. Trees that have fallen into the river here create snags, which are ideal for basking turtles.
Basking turtles were observed with binoculars (Bausch & Lomb 10 × 42) and a spotting scope (Eagle Optics ×20–60 straight angle) along a 1.5-km shoreline of the Sabine River, from 1 January 2005 through 1 January 2006. Observations were made twice per week, at approximately the same time of day (between 1000 and 1500 hours). Every turtle observed was aged (adult vs. juvenile), sexed, and identified to species. If the carapace appeared to be less than 8 cm the turtle was classified as a juvenile. Sex was determined by looking at 3 dimorphic traits: size of tail, overall size, and foreclaw length (Ernst et al. 1994). The size of the tail provided the best indication of sex. Overall size and foreclaw length were used only when the tail was not visible or for supporting evidence. These dimorphic traits were not visible (e.g., obscured by other turtles, basking angle, or location on basking site) on 303 individual turtles, thus those individuals were excluded from the study. Identification of species was based on head coloration (Lindeman 2003; Vogt 1993). In G. pseudogeographica kohnii the postorbital mark is a narrow downward crescent stripe; whereas, in G. ouachitensis sabinensis, it is an oval. Determination of species was not possible (e.g., head obscured from view or turtle retreated prior to identification) for 203 individuals; therefore, these individuals were excluded from the study. Air and water temperatures were recorded at the beginning and end of each observation period. Water temperatures were obtained by placing a mercury thermometer approximately 2.5 cm below the surface of water (shaded areas); air temperatures were taken in the shade next to the river.
The χ2 goodness-of-fit test was used to assess deviations from a 1:1 ratio of basking males to basking females in each species and was also used to test whether the number of G. pseudogeographica kohnii observed differed significantly from the number of G. ouachitensis sabinensis observed. A χ2 contingency table was used to evaluate sexual and seasonal differences in basking patterns between the 2 species. A second-order polynomial regression was used to investigate the relationships of air and water temperature to the number of basking turtles. Statistical analyses were conducted with the computer software, SYSTAT 12 (Systat Software, Inc., Chicago, IL).
Results. — Significantly more G. ouachitensis sabinensis (n = 4095) than G. pseudogeographica kohnii (n = 490) were observed basking during the year (χ2 = 0.283E+04; p < 0.001). Eighty-four observations of basking G. pseudogeographica kohnii were made in the fall (18 September–20 December), 82 in the winter (21 December–19 March), 220 in the spring (20 March–20 June), and 104 in the summer (21 June–17 September). For the same seasons for G. ouachitensis sabinensis, 866, 539, 1849, and 841 observations were made, respectively (Table 1). Within G. ouachitensis sabinensis, in all seasons, the number of males observed basking significantly exceeded the number of females (χ2 = 107.6; p < 0.05); however in G. pseudogeographica kohnii, numbers of males and females differed significantly only in spring and fall (Table 1; χ2 = 18.8; p < 0.05).
Seasonal basking patterns were not significantly different for G. pseudogeographica kohnii and G. ouachitensis sabinensis (χ2 = 7.6; p > 0.05). Both species were seen in relatively low numbers from December to February, with a marked increase in basking observations in March (Fig. 1). The relative abundance of both species remained constant throughout the year, except during May and June. Numbers of G. ouachitensis sabinensis peaked in May; whereas, numbers of G. pseudogeographica kohnii peaked in March; additional peaks in basking numbers were observed in June for G. pseudogeographica kohnii and in November for both species (Fig. 1).
Citation: Chelonian Conservation and Biology 7, 2; 10.2744/CCB-0708.1
Results showed a polynomial correlation between air temperature and number of turtles basking (Fig. 2). The coefficients of determination relating basking numbers to air temperature were low but the relationships were significant for both sexes in both species (G. pseudogeographica kohnii males R2 = 0.11, p = 0.01; females, R2 = 0.06, p = 0.05; G. ouachitensis sabinensis males R2 = 0.16, p = 0.003; females, R2 = 0.09, p = 0.015).
Citation: Chelonian Conservation and Biology 7, 2; 10.2744/CCB-0708.1
The number of turtles basking was greatest at water temperatures of 15°C −25°C in both G. pseudogeographica kohnii and G. ouachitensis sabinensis (Fig. 3). On observation days when water temperatures were above 25°C or below 15°C, basking numbers were low (Fig. 3). The coefficients of determination relating basking numbers to air temperature were low, but the relationships were significant for males of both species (G. pseudogeographica kohnii: R2 = 0.12, p = 0.04; G. ouachitensis sabinensis: R2 = 0.18, p = 0.04). The relationships were not significant for females (G. pseudogeographica kohnii: R2 = 0.09, p = 0.20; G. ouachitensis sabinensis: R2 = 0.12, p = 0.10).
Citation: Chelonian Conservation and Biology 7, 2; 10.2744/CCB-0708.1
Discussion. — The results suggest that G. pseudogeographica kohnii has a much smaller population than G. ouachitensis sabinensis along this portion of the Sabine River (Table 1). Lindeman (1999) also found the smaller species, G. ouachitensis ouachitensis, to be more abundant than G. pseudogeographica kohnii in the Tennessee River; although, the interspecific ratio (1.7:1) was much more similar than that observed in our study on the Sabine River (8.4:1). Microcephalic species seem to predominate in basking counts, which may reflect a difference in basking prevalence between mollusk and nonmollusk consumers (Lindeman 1999). Shively and Jackson (1985) reported that upstream distributional termination of G. ouachitensis sabinensis results from increasingly difficult dispersal and progressively reduced dispersal pools upstream, along with changing environmental factors that limit population numbers. However, ideal habitat characteristics (e.g., river width, open canopy, and prey availability; Shively and Jackson 1985) within the study site may provide an explanation for the higher numbers of G. ouachitensis sabinensis observed in our study.
Basking habits exhibited an interaction between season and sex between the 2 map turtle species studied, with G. pseudogeographica kohnii exhibiting a significantly male-biased sex ratio for basking in fall and spring, and G. ouachitensis sabinensis having a significantly male-biased sex ratio for basking in all seasons (Table 1). The seasonal basking patterns observed in this study are consistent with those of previous studies of other freshwater emydids (Waters 1974; Vogt 1980; Pluto and Bellis 1986; Morreale et al. 1984; Craig 1992). Lindeman (2000) found that spring and autumn peaks observed in his study may have resulted from sampling bias because sampling occurred only when weather was ideal for basking (i.e., clear, sunny days). In our study, sampling continued throughout the year, regardless of weather conditions, thus eliminating that source of bias.
Thomas et al. (1999) suggested that male T. scripta bask more in September and October to facilitate the production of sperm in preparation for the following spring. Male predominance during the seasons may be associated with the change in temperatures during each season, as higher temperatures (> 20°C) have been reported to initiate testis growth, spermatogenesis, and testosterone production (Ganzhorn and Licht 1983). An explanation for predominance of basking males at cooler temperatures (< 15°C) might be an increase in their foraging activities in preparation for mating (Thomas et al. 1999), but this is not the only potential explanation and warrants further investigation.
Other possible explanations for predominance of basking males may be greater wariness of the larger females, as suggested for G. ouachitensis (Webb 1961), Graptemys ernsti (Shealy 1976), Graptemys spp. (Boyer 1965), and T. scripta (Auth 1975), or male-biased sex ratios in the Sabine River, as also reported in other map turtle species (Pluto and Bellis 1986; Fuselier and Edds 1994; Lindeman 1999). Temperature-dependent sex determination (TSD) is another factor that may influence the predominance of basking males, as documented among many species of reptiles (Janzen and Paukstis 1991; Burke 1993). Vogt and Bull (1982), working with 3 species of Graptemys, demonstrated that TSD strongly skews primary sex ratios in natural environments, with 75% of 236 Graptemys nests producing hatchlings of only one sex. A high frequency of male-biased nests could explain why a higher number of males were observed basking during this study. However, little is known about how hatchling sex ratios relate to functional sex ratios in turtle populations.
Air temperature appears to influence basking patterns in G. pseudogeographica kohnii and G. ouachitensis sabinensis. The number of male G. pseudogeographica kohnii and G. ouachitensis sabinensis basking can also be correlated to water temperature; however, a significant correlation between number of basking females of both species and water temperature was not found. Turtle body temperature has been noted to follow changing air temperature more closely than water temperature (Ernst 1971; Litzgus and Brooks 2000), which can explain to some extent why air temperatures correlated more strongly with the number of basking turtles (Figs. 2 and 3). The increased numbers observed basking in spring in both species may result from turtles increasing their metabolic rate so that physiological homeostasis can be reestablished from the cooler winter months. Temperature plays a significant role in many physiological processes, such as digestion, vitamin D synthesis, and avoidance of epizootic and epiphytic infestations (Edgren et al. 1953; Neill and Allen 1954; Boyer 1965; Belusz and Reed 1969). Therefore, the numbers of turtles observed basking should increase as air and water temperatures increase.
Basking surveys are useful for censusing aquatic turtle populations because large data sets can be compiled with relatively little time and effort (Lindeman 1997). Our study suggests that basking surveys should consider abiotic factors such as season and air temperature when evaluating relative abundance of species in a community.
Seasonal patterns of basking in Graptemys pseudogeographica kohnii and Graptemys ouachitensis sabinensis of the Sabine River from January 2005 to January 2006 at the Old Sabine Bottom Wildlife Management Area, Smith County, Texas. (A) Mean number of turtles seen per observation period within each month. (B) Proportion of total annual observations of each species by month.
Influence of air temperature on the number of basking (A) Graptemys pseudogeographica kohnii and (B) Graptemys ouachitensis sabinensis. The black diamonds and the solid lines (y =−0.005×2 + 0.28× and y =−0.05×2 + 2.59×, respectively) represent males; whereas, the open squares and dotted lines (y =−003×2 + 0.16× and y =−0.02×2 + 1.16×, respectively) represent females. Air temperature is an average of readings taken at the beginning and end of each sample period.
Influence of water temperature on the number of basking (A) Graptemys pseudogeographica kohnii and (B) Graptemys ouachitensis sabinensis. The black diamonds and solid lines (y =−0.01×2 + 0.41× and y =−0.10×2 + 3.99×, respectively) represent males; whereas, the open squares and dotted lines (y =−0.01×2 + 0.28× and y =−0.05×2 + 2.03×, respectively) represent females. Water temperature is an average of readings taken at the beginning and end of each sample period.
