Growth, body size, and body proportions are important parameters of the life-history patterns of animals often related to ecological and ethological aspects of the individuals (Peters 1983; Schmidt-Nielsen 1984; Roff 1992; Stearns 1992). Differences between the sexes in body size, morphology, and coloration (sexual dimorphism) are common among reptiles (Dunham et al. 1988; Shine 1994; Randriahamazo 2000). Particularly, morphometry, growth, and sexual dimorphism have been extensively described in turtles (Berry and Shine 1980; Iverson 1985; Gibbons and Lovich 1990; Forsman and Shine 1995; Lambert 1995; Zuffi and Gariboldi 1995; Yasukawa et al. 1996; Ernst et al. 1998; Graham and Cobb 1998; Willemsen and Hailey 1999; Zuffi et al. 1999; Ayres and Cordero 2001; Bonnet et al. 2001; Boone and Holt 2001). However, most of the descriptive studies have not focused on the evolutionary causes implicated in the biometric results or the possible consequences on ecology, behavior, or reproductive success of individuals (see, e.g., Berry and Shine 1980).

Growth patterns are often related to sexual selection processes, so that the differences between sexes in reproductive strategies can involve a differential investment of resources and energy in the developmental processes, thus resulting in sexual dimorphism in size and general body proportions (Berry and Shine 1980). For example, Bonnet et al. (2001) found that sexual dimorphism in some morphological characters of Testudo horsfieldii could be related to mobility of males. They documented that shell morphology in males would enhance their locomotor performance due to greater leg mobility. They showed that legs of males were longer than those of females, and argued that the greater locomotor performance of males would enhance mate searching, thus increasing their reproductive success (Berry and Shine 1980; Bonnet et al. 2001). When major sexual differences exist in diet, habitat, or predator-escape tactics, the ecological side of natural selection can also be responsible for sexual dimorphism, promoting niche divergence (Hedrick and Temeles 1989). For example, Lindeman and Sharkey (2001) have found that sexual differences in head proportions of map turtles could be due to their diet divergence. On other hand, fecundity selection may result in increased turtle female size or variation in body proportions to enhance egg-carrying capacities (Ernst et al. 1998; Zuffi et al. 1999; Bonnet et al. 2001). Although the hypothetical selective forces that promote differences between sexes are well discussed, it is often difficult to elucidate the relative importance of each one (sexual selection, niche divergence, and/or fecundity selection) on sexual dimorphism (Braña 1996).

The stripe-necked terrapin (Mauremys leprosa) is a riverine species that mainly inhabits the center and south of the Iberian Peninsula, but also inhabits northwest Africa and southwest Europe (Salvador and Pleguezuelos 2002). Morphometric studies of this species are few, and none have explored the adaptative significance of its morphology (Pérez et al. 1979; Meek 1987; Keller 1997a). In this study we measured individuals of M. leprosa originating from a wide range of central Spain, and we discuss the possible evolutionary causes of the observed morphometric patterns.

METHODS

We measured 140 individuals of M. leprosa from central Spain, 30 of which were preserved specimens from the collection of the National Museum of Natural Sciences (Spain), and the rest were captured during spring 2001 and 2002. Turtles were hand- or creel-captured using sardines as bait (under a licence of the Consejería de Medio Ambiente de Toledo). Measurements were taken immediately after capture and individuals were liberated at the capture site. Turtles were classified into 3 categories: 1) juvenile (sexually immature), 2) adult female, and 3) adult male. Turtles were sexed based on plastron shape. Those with a concave plastron were considered as adult males, whereas those with a flat or convex plastron were identified as adult females (Pérez et al. 1979; Meek 1987). The terminology used in this classification refers to morphological characters (the appearance of the secondary sexual characters in the males), even though sexual maturity in females occurs at larger sizes than considered as an “adult female” in our study. Thus the term “adult” was used for turtles that could be sexed accurately, and not only for turtles showing physiological and behavioral features of adults.

We characterized the general shape of turtles with external features of the shell and the tail. Details of shell and tail measurements are illustrated in Fig. 1. Measurements recorded from each individual included 1) carapace length (CL), as the straight distance from the anterior edge of the nuchal scute to the midline posterior notch between the supracaudal scutes; 2) plastron length (PL), as the straight midline distance between the gular scutes and the anal notch; 3) anterior carapace width (ACW), as the straight distance between the midline point of the external edges of the fifth marginal scutes; 4) posterior carapace width (PCW), as the straight distance between the midline point of the external edges of the eighth marginal scutes; 5) curved carapace length (CCL), taken from the same points as for CL, but following the dorsal curvature of the shell; 6) curved carapace width (CCW), taken from the same points as for ACW, but following the dorsal curvature of the shell (these two curvilinear measurements may reveal large variations in body shape [i.e., flat or domed shells] when included in ANCOVAs); 7) tail length (TL), from the posterior midline anal notch of the plastron to the top of the tail; and 8) preanal tail length (PTL), measured from the midline anal notch of the plastron to the anterior border of the cloaca (in both tail measurements the tail was stretched firmly). Straight measurements were obtained with calipers and curved measurements with a tape measure (to the nearest mm in both cases). For each turtle, all measurements were taken two times by the same person (A.M.) in non-consecutive order, and were shown to be highly repeatable (r > 0.99, p < 0.0001 in all cases).

View Figure

• Download as Powerpoint
• Download Figure

As in other morphometric studies on turtles (e.g., Zuffi and Gariboldi 1995; Bonnet et al. 2001), we included data on both live and preserved turtles. We only measured tail and preanal tail length in the preserved turtles that had a straight tail which could be stretched as in live individuals. To analyze growth patterns and sexual dimorphism, we used regression models, analysis of variance (ANOVA), and covariance (ANCOVA) (Packard and Boardman 1987). All measurements are presented as mean ± standard error, unless noted, and all data were analysed using STATISTICA '99.

RESULTS

Turtles considered as juvenile ranged from 24 to 94 mm CL (mean = 61 ± 3 mm, n = 52). Size (CL) was significantly different between adult males and adult females (males = 144 ± 3 mm, range = 9.9–175 mm, n = 45; females = 165 ± 4 mm, range = 96–205 mm, n = 43; ANOVA, F_1, 86 = 13.3, p < 0.001). Table 1 shows the regression lines from each measurement against the CL. Adult females showed a higher relative growth of PL than juveniles, whereas in the case of adult males the relative growth of PL was lower than in juveniles (Table 1). As a consequence, adult females have significantly longer plastrons in proportion to CL than adult males (Table 2). Adult females also showed a significant increment of the relative growth of ACW, as compared with adult males and juveniles, whereas no differences exist between adult males and juveniles (Tables 1 and 3). Females develop wider shells than males, both for ACW and PCW (Table 2). The value of the difference between PCW and ACW increases quickly in juveniles. In males this value continues increasing, though less markedly than in juveniles, and in females the value remains practically constant (Fig. 2). Thus, the relative growth of PCW with respect to ACW is much higher in males than in females, despite that females have wider carapaces both in the anterior and the posterior measurements. The slope of the regression line for CCW is higher in adult females than in juveniles or males (Tables 1 and 3). Thus, as an adult the female carapace becomes increasingly domed. Moreover, ANCOVA to compare CCW between the sexes (with ACW as covariate) also showed that CCW in females is significantly greater than in males (F_1, 86 = 5.2, p = 0.025). We found no sexual differences in tail length of M. leprosa (Table 2), but preanal tail length is a sexually dimorphic trait, much longer in males (Table 2). In juveniles PTL increases quickly in relation to the CL, and then stabilizes in males and females (Table 1, Fig. 3).

Table 1. Regression statistics to compare the variation of the measurements with the carapace length (CL) in males (M), females (F), and juveniles (J).^a

View Fullscreen

Table 2. ANCOVA (with the carapace length as covariate) of the comparisons of the measurements (in mm) between males (M) and females (F).^a

View Fullscreen

Table 3. Comparisons of the growth rates of each measurement—relative to carapace length (CL)—between the three categories (males, females, juveniles).^a

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

DISCUSSION

Most explanations of sexual dimorphism have been related to ecological, sexual selection, or fecundity aspects of individuals (Hedrick and Temeles 1989). Intrasexual selection or male competition for mates could be the cause of greater body size in males than in females in many chelonian species, favoring their fight capability (Berry and Shine 1980; Niblick et al. 1994; Lovich et al. 1998; Lindeman 1999; Rovero et al. 1999).Berry and Shine (1980) suggested that the smaller body size of males in most freshwater turtles could be the result of low importance of intrasexual selection processes in these species. Sexual selection should enhance mate searching and acquisition abilities in males favoring their mobility (Berry and Shine 1980; Bonnet et al. 2001). We have observed that activity and mobility of males of M. leprosa are higher than that of females in the mating season, and that males can detect females by using chemical cues during this season (Muñoz 2004). A greater energy investment of males for locomotion would be reflected in a lower investment in growth. Moreover, the smaller size would facilitate the mobility of males and would reduce their transportation costs (Bonnet et al. 2001). Furthermore, being smaller, males hide more easily in refuges, since by their greater mobility they may be more exposed to predators (Magnhagen 1991). It has been suggested for other species of turtles and tortoises that the small size of one of the sexes can also be related to age at maturity (Gibbons and Lovich 1990). In M. leprosa, males attain sexual maturity at smaller sizes than females (Keller 1997b), and this could be one of the causes of the sexual size dimorphism in this species.

On the other hand, the larger body size of females could be the consequence of fecundity selection to accommodate larger clutches and larger eggs (Ernst et al. 1998; Zuffi et al. 1999; Bonnet et al. 2001). In many turtle species (including M. leprosa) the largest females produce more and larger eggs (Congdon and Van Loben 1991, 1993; Da Silva 1995; Nieuwolt-Dacanay 1997; Keller 1997b). Moreover, a direct relationship has been found between the size of eggs and the size of hatchlings (Congdon et al. 1983; Packard and Packard 1988; Bobyn and Brooks 1994), and the largest hatchling turtles may have a greater survival probability (Miller et al. 1987; Janzen 1993; Bobyn and Brooks 1994; Janzen et al. 2000; but see Congdon et al. 1999; Kolbe and Janzen 2001). Therefore, fecundity selection would favor the largest females, permitting the ability to carry larger clutches, and thus favoring greater hatchling survival. In addition, Forsman and Shine (1995) found that, across freshwater turtle species of northern and Central America, sexual size dimorphism in favor of females was positively correlated with an increase in the annual frequency of reproduction. Thus, larger females might have fecundity advantages and, as a consequence, might be preferred as mates for males, as has been suggested by Pearse et al. (2002) in Chrysemys picta.

The greater width and height of the female carapace may also be an adaptation to carry more eggs (Ernst et al. 1998). In the European pond terrapin (Emys orbicularis) the number of eggs carried by females is closely correlated with the height of its shell (Zuffi et al. 1999). In many other turtle species, females also have higher and wider shells than males (Zuffi and Gariboldi 1995; Rowe 1997), and this has also been reported for populations of M. leprosa in southern Spain (Keller 1997a, 1997b) and northern Africa (Meek 1987). However, sexual differences in width and height of the shell may not be exclusively related to increased female fecundity. A narrower and lower carapace in males would permit greater efficiency in aquatic locomotion, and thus a greater efficiency in mate searching. Moreover, the higher relative growth of PCW with respect to ACW in males than in females may also be adaptative in permitting greater stability during aquatic locomotion. Thus, selective forces shape sexes differently, with male body proportions facing strong selection to enhance male mobility, and female fitness being influenced by the ability to carry eggs.

Although we have not directly measured the carapace height, this trait does not seem to be sexually dimorphic in M. leprosa, since no differences were found in CCL between males and females. Because CCW is a sexually dimorphic trait suggests that females of this species have increased their shell capacity through increased doming of the portions of the carapace lateral to the vertebral midline. This could be because doming of the posterior carapace might make it difficult for males to mount females, since they approach from behind and mount the back of the shell (Sidis and Gasith 1988). This possibility has not been considered in previous chelonian morphometric studies since most of them only consider carapace height measured at one point (e.g., Pérez et al. 1979; Zuffi and Gariboldi 1995; Keller 1997a; Ernst et al. 1998; Bonnet et al. 2001).

Perhaps the most common sexually dimorphic trait in chelonians is PTL (Dunham et al. 1988; Gibbons and Lovich 1990). The greater PTL of males could be a consequence of its shorter plastron in relation to carapace length, as noted by this and other studies (Pérez et al. 1979; Gibbons and Lovich 1990; Ernst et al. 1994; Rowe 1997; Litzgus and Brooks 1998; Bonnet et al. 2001). Moreover, it could also be an anatomical adaptation to the penis in the base of the tail. In addition, the greater PTL would facilitate mating in males by permitting greater freedom in tail movement and thus greater efficiency in cloacal contact during copulation.

Our knowledge of the ecology, behavior, social organization, and sexual selection processes in M. leprosa is very limited (but see Muñoz 2004). It is necessary to undertake numerous ecological and ethological studies to complement morphometric ones. Such studies might also provide important tests of the hypotheses arising from the morphometric studies.