Biology_of_Turtles

(4) repulsion of egg predators through taste or smell, (5) compaction of dirt after nesting to ward off predators, and (6) camouflage” (Ehrenfeld, 1979, based on Patterson, 1971).

The inner surfaces of the ancillary bladders (and the cloaca) may be vascularized to assist in regulation of electrolyte balance by absorbing sodium (Chrysemys picta: Dunson, 1967; Minnich, 1982) and contribute to respiration as aquatic respiratory organs (Apalone spinifera asper: Dunson, 1960; Trionyx triunguis: Girgis, 1961). Other species may also use aquatic respiration, at least to some degree (Podocnemis spp.: Steen, 1971, cited by Seymour, 1982). Respiratory cloacal bursae are found in Rheodytes leukops, where they augment respiration (Legler, 1979). The inner lining is modified into vascularized villiform projections that facilitate the uptake of oxygen from the water and the release of carbon dioxide to the water; water is circulated by muscular contractions of the bursae and cloaca (Legler, 1979; Legler & Cann, 1980). The augmentation of respiration allows this species to remain submerged and active for extended periods to time (Gordos et al., 2003a, 2003b); typically, it remains in highly oxygenated water located downstream of riffles (Cann, 1998).

Male chelonians have a single penis situated in the ventral midline of the cloaca. The penis is comprised of erectile tissue (Andersson & Wagner, 1995). In cryptodires, the penis is divided into two unequal portions: the long proximal shaft and the shorter distal glans (Zug, 1966). In pleurodires, the shaft portion is virtually nonextant and the entire structure appears to be the glans (McDowell, 1983).

Two types of erectile tissue occur in the turtle penis: corpus spongiosum and corpus fibrosum (Zug, 1966). McDowell (1983) uses the term corpus cavernosum instead of corpus spongiosum based on the lack of clear homology. In cryptodires, the corpus spongiosum contains a thick layer of vascularized sinusoidal tissue situated on top of the corpus fibrosum, which is denser, less erectile connective tissue. The penis of pleurodires is morphologically quite different from that of cryptodires. The corpus cavernosum exists as a veneer with just enough “to form and close the lips of the seminal groove” (McDowell, 1983). The corpus fibrosum is not cavernous or erectile; instead, it is comprised of dense, cartilage-like material (McDowell, 1983).

The glans of both cryptodires and pleurodires range in structure from simple to elaborate. The glans of sea turtles and snapping turtles (Cryptodira: Cheloniidae, Chelydridae) are relatively simple (Zug, 1966). In contrast, the glans of Emydura australis (Pleurodira: Chelidae) is relatively elaborate, looking more like an opening flower (McDowell, 1983). Zug (1966) used the morphology of the glans to reconstruct the basic relationships among families within the cryptodires. For example, Williams (1950) placed the kinosternid genera into Chelydridae; however, Zug (1966) found the similarity of the “plica media suggests the kinosternids and the dermatemydids arose from a common ancestor,” as proposed by McDowell (1961). In the context of modern genetically based classifications extending Zug’s work to include more species of both cryptodires and pleurodires, their sexual dimorphism and their ecological characteristics (Stephens & Wiens, 2003, 2004) will aid analysis and understanding of the evolution of reproductive structures and strategies found in chelonians.

Although in both groups the glans becomes engorged with blood, the supportive components of the erection are not the same. During erection in cryptodires, the engorged glans and the shaft extrude from the cloaca of the male into the cloaca of the female (Zug, 1966). The length of the shaft ensures the proper placement of the glans in the anterior part of the cloaca of the female for sperm transfer. In contrast, the copulatory organ of pleurodires (e.g., Emydura, Phrynops, Platemys) includes both the glans and the partial eversion of the cloaca (McDowell, 1983), which functionally acts like the shaft of cryptodires and extends the total length of the organ, allowing insertion into the female cloaca.

Although the penis must have some rigidity to be inserted into the female cloaca, it is not fully erect before insertion. Once the distal portion has been inserted, the expansion of the glans, and supporting region (in conjunction with the partial relaxation of the female cloaca), moves the penis further into the female cloaca. The engorged erectile tissue of the seminal ridges presses them together, thereby ensuring the lumen of the seminal groove (now formed into a tube) remains open

for the passage of sperm. In pleurodires, because the seminal groove does not extend to the end of the glans the erect penis appears to be “doubled over” (McDowell, 1983), which facilitates the placement of sperm from the seminal groove into the cloacal vestibule of the female. During intromission, the penis fills the cloaca and is held in place by the engorged erectile tissue in the glans (McDowell, 1983). The enlarged glans orients the tip of the seminal groove in the vestibule at the opening of the oviducts. The enlarged glans helps to seal the boundary of the distal penis against the epithelium of the female’s cloaca.

While flaccid in the male’s cloaca, the more expansible corpus spungiosum is on the dorsal surface and the less expansible corpus fibrosum is on the ventral surface. As both become engorged with blood, the greater expansion of the corpus spongiosum causes the penis to extend from the cloaca and curve forward. As a result, the glans—which is posteriorly directed in the cloaca—becomes anteriorly directed and the dorsal surface of the flaccid penis becomes ventrally directed during the erection. Because the male is mounted above and behind the female, he must bring his cloaca into contact with the opening of her cloaca. This means that the penis enters the female cloaca upsidedown (or at least partially on its side) relative to the flaccid position in his cloaca. Penetration may be aided, at least to some degree, in some groups (e.g., Testudo, Pyxis: Obst, 1986) by the hardened tip of the male’s tail; the cap scale on the tail tip may be used to locate the opening leading into the female’s cloaca or simply to guide the engorging penis to the area preceding intromission. The role of the tail in copulation needs to be better defined by observations in the field and in captivity.

The duration of intromission is difficult to determine. Harrel et al. (1996) observed alligator snapping turtles (Macrochelys temminckii) in coitus for between 5 and 25 minutes. Wood and Wood (1980) reported that the average time spent mounted for green turtles (Chelonia mydas) in captivity was 25.5 hours. Their observations on captive sea turtles indicated higher fertility was positively correlated with longer periods of intromission (Wood & Wood, 1980); similar information is not available for other chelonians. The stimulus for ejaculation is unknown for any turtle. Ejaculation does not appear to be spasmodic, but rather sperm transfer seems to occur as a flow over time. However, this may be an illusion associated with the tendency of the male to remain mounted on the female for an extended period; this behavior may facilitate uptake of sperm through the ostia of the oviducts. The time involved may reflect the process of sperm transfer not in an ejaculatory rush but rather a more pulsed flow, the distance the sperm must travel, and the need for sperm to enter the ostia of the oviducts once the seminal fluid has been placed into the anterior of the cloaca. Intromission is an understudied part of chelonian reproduction.

The clitoris arises from the medial-ventral wall of the cloaca as a ridge in the proximal region of the cloaca. Unfortunately the size, shape, and histological structure of the clitoris have not been described for any species of turtle.

10.3.4 Hormonal Changes during Reproduction

The general pattern of endocrine control of the reproductive cycle in male and female sea turtles (species that produce multiple clutches in a single reproductive season) shows increases and decreases in four critical hormones and changes in associated behaviors (Licht et al., 1979; Owens & Morris, 1985; Owens, 1997; Lance & Rostal, 2002; Rostal, 2005; Hamann et al., 2003). For males, sperm formation occurs prior to migration and in conjunction with an increase in testosterone and a slight rise in gonadotropin. Levels of both hormones decrease during migration and mating periods, and by the time the male begins his remigration to his foraging area, the levels are greatly reduced. In females, estrogen reaches its maximum level during preparation of the follicles in the foraging area, then reduces during the migration to the mating and nesting areas, and is lower still during mating receptivity. Levels of testosterone are low in the foraging area but increase preceding and during the migration and mating periods. Mating ceases about two weeks prior to ovulation, which precedes nesting by about two weeks. In this time, the testosterone level dips but recovers prior to ovulation; during each subsequent inter-nesting period, testosterone decreases and recovers but each recovery

is less than the previous recovery. Estrogen level recovers just prior to ovulation from a low during the mating period. The dip and recovery occurs during each successive inter-nesting period; however, the recovery and dip levels stay about the same each time.

The level of gonadotropin is steady during follicle preparation and migration; it reduces slightly during mating and prior to the first ovulation. At this time, blood concentration of gonadotropin surges and crashes within a few days. The pattern is followed closely by a surge in progesterone such that the peak concentrations are offset slightly. All hormone levels reduce dramatically following the last oviposition of the nesting season. Sequential changes in the levels of hormones (estradiol, progesterone, AVT, neurophysin, PGE2, PGF) circulating during the nesting process correlate well with changes in observed behaviors (Guillette et al., 1991; Owens, 1997). It is likely that in species that produce two or three clutches of eggs, the pattern of cyclic fluctuations in hormone concentration levels is similar to that described for sea turtles, except for the number of sequential ovipositional cycles (e.g., Geochelone nigra; Schramm et al., 1999). In species that ovulate a single clutch, the cyclic pattern of hormonal concentration increase and decrease is truncated. Variation is to be expected in the actual levels of hormone concentrations when the general cycle is applied to a specific species.

10.3.5 Eggs

All turtles lay amniotic eggs (Ewert, 1979, 1985). By definition, the structure of the cleidoic egg allows the exchange of only gases between the external environment and the embryonic tissue. This condition is met by the hard-shelled eggs of terrestrial chelonians, but more pliable eggs may allow liquid water to penetrate the shell matrix and enter the egg in a fluid state as well as a vapor (Thompson, 1983, 1987; Ackerman et al., 1985). The egg shape is defined by the shell. Inside the shell matrix is a fibrous membrane that surrounds the albumen (Packard & DeMacro, 1991). Within the albumen is the yolk membrane that contains the yolk material and on which the embryonic disc is situated. Regardless of the actual size or shape of the egg, or the type of egg shell, the proportions of egg components are relatively constant among the species. In terms of overall egg mass, the shell typically contributes about 11 to 12% (range: 4.3 to 24.7%), the albumen contributes about 46 to 47% (range: 34.4 to 61.9%), and the yolk contributes about 41 to 42% (range: 32.1 to 55.0%) (Gibbons, 1990; see Table 7 in Ewert, 1979).

The shell is composed of two general layers: an outer crystalline structure and the inner, supporting proteinaceous, fibrous membrane (Packard & DeMacro, 1991; Schleich & Kästle, 1988; Gad, 1994). The flexibility of the egg shell can be categorized into two groups: inflexible (hard-shelled) and flexible (soft-shelled). The difference in the rigidity of the shell is the result of differences in the matrix of the calcified layer (Packard & DeMarco, 1991). In hard-shelled eggs, the calcified matrix is densely packed calcium columns (Packard & DeMarco, 1991). In contrast, in soft-shelled eggs the matrix is comprised of aragonite crystals that form rosettes on the fibrous inner shell membrane (Packard & DeMarco, 1991). In both types, the shell matrix must allow the movement of oxygen and carbon dioxide; more hydric flux occurs through the shells of soft-shelled eggs (Chelydra serpentina: Packard et al., 1987, Chrysemys picta: Paukstis et al., 1984). In general, the structure of the egg shell correlates with the environmental conditions in which the eggs incubate (Ewert, 1979). Typically, soft-shelled eggs are laid by marine turtles and some river turtles; these species incubate their eggs in a moist substrate and the eggs gain weight during their relatively short incubation period. Some pliable eggs also follow this pattern. Hard-shelled eggs are laid by terrestrial turtles nesting in arid and semi-arid environments and by species nesting in areas that become dry during their longer incubation periods (see Table 5 in Ewert, 1979). Equally as important in considering the size and shape of turtle eggs are the thickness, functional pore area, and structure of the shell because the egg shell provides the critical boundary between the external and internal environments affecting the developing embryo by regulating the movement of gases, including water vapor (Packard & Packard, 1988a; Thompson & Speake, 2004). The shell structure is not continuous; the pores

facilitate gas and vapor exchange. The shell also provides calcium for embryonic bone development (Apalone spinifera: Packard & Packard, 1991; Chrysemys picta: Packard & Packard, 1986; Chelydra serpentina: Packard & Packard, 1989; Dermochelys coriacea: Simkiss, 1962).

Within the protective boundary of the egg shell, the albumen serves the embryo in three important ways (Palmer & Guillette, 1991; Thompson & Speake, 2004). First, the albumen provides antimicrobial protection through its fibrous structure and its colloid form. Second, its viscosity provides support for the expanding vitelline membrane in early development and to the growing embryo later in development. Third, albumen also provides for at least some water storage and may assist in accumulating water from the substrate, depending on the egg shell structure (Palmer & Guillette, 1991; Thompson & Speake, 2004).

The yolk constitutes the primary source of energy provided by the female (Ar et al., 2004). The actual energy content varies among the species (see Table 7.1 in Ar et al., 2004); dry mass energy content of turtle eggs ranged between 24.18 kJ/g for Chelydra serpentina and 28.46 kJ/g for Caretta caretta. Congdon et al. (1983c) reported differences in the percentage of yolk lipid used during incubation in seven species of freshwater turtles (Chrysemys picta, Emydoidea blandingii, Graptemys ouachitensis, Trachemys scripta, Sternotherus odoratus, Graptemys geographica, Chelydra serpentina) incubated under common hydric conditions. Although their data were not unequivocal, they speculated the emergence tactics of hatchlings were linked in some part to lipid levels in turtle eggs (Congdon et al., 1983c). Certainly, the hydric conditions at the nest site influence yolk utilization and the hatchling size in Chrysemys picta (Packard et al., 1991). Chelonians seem to be using a different strategy than lizards because their hatchlings contain a relatively larger energy reserve at hatching (Ar et al., 2004); this reserve supports the hatchling as it moves away from the nesting area to a reasonably safe area before beginning to feed (Packard & Packard, 1988a; Ar et al., 2004). If the embryo remains unhatched in the nest (e.g., Deirochelys reticularia, Kinosternon baurii: Ewert & Wilson, 1996), or if the hatchling remains in the nest over winter, its yolk reserves must sustain it through changes in temperature (and hence changes in metabolism) and moisture with enough to provide for it until it can reach food in the spring (e.g., Chrysemys picta: Costanzo et al., 1995).

10.4Reproductive Strategies

Reproductive strategies in reptiles tend to be conservative (Shine, 1985); yet within and among groups, there is variation. The general reproductive cycle of reptiles is under environmental control (Duvall et al., 1982). The reproductive ecology of chelonians has received more emphasis than morphology in recent years (Fox, 1977), with the primary focus on sexual dimorphism, mating system/pattern, fecundity, seasonal timing of reproduction, ovipositional site, incubation period, mode of sex determination, embryology, and hatchling size and survivorship (Shine, 1985; Ewert, 1979; Moll, 1979; Bury, 1979; Gibbons, 1990; Lutz & Musick, 1997; Lutz et al., 2003). Reproductive cycles have been described in varying detail for several turtle species (e.g., Sternotherus odoratus: Risley, 1938; Terrapene carolina: Atland, 1951; Terrapene carolina, T. ornata, Sternotherus: Legler, 1960a; Trachemys scripta elegans: Burger, 1937; Moll & Legler, 1971; Gibbons, 1990).

10.4.1 Sexual Dimorphism

Chelonians exhibit a variety of sizes and shapes. Although some fossil forms are much larger (e.g., Archelon ischyros: Carr, 1964; Obst, 1986), the largest living chelonian is the leatherback sea turtle, Dermochelys coriacea—the largest individual on record was 256 cm curved carapace length and 915 kg (Eckert & Luginbuhl, 1988). The smallest chelonian is the speckled cape or padloper tortoise, Homopus signatus, at less than 10 cm, less than 140 g (Boycott & Bourquin, 1988). Between these extremes are large Galapagos tortoises and smaller freshwater turtles such as the Kinosternids; some river turtles are quite big (more than 100 cm carapace length) and some terrestrial turtles are

rather small (less than 15 cm) (for dimensions, see Carr, 1952; Ernst & Barbour, 1972, 1989; Ewert, 1979; Pritchard, 1979; Pritchard & Trebbau, 1984; Cann, 1998; Bonin et al., 2006).

Often, the adult females are larger than adult males but the generalization is not universally true, even within a single family (Carr, 1952; Ernst & Barbour, 1972, 1989; Pritchard, 1967, 1979; Ewert, 1979; Bonin et al., 2006). For example, in freshwater turtles (Emydidae), females are larger than males in some genera (e.g., Chrysemys, Graptemys, Kachuga), whereas in other genera, the difference in size is slight (e.g., Emys, Clemmys, Mauremys). In sea turtles (Cheloniidae, Dermochelyidae), the adult females are larger than the males (Wibbels et al., 1991); in contrast, the sexes are about equal in size in several species of mud turtles (Kinosternidae) and land tortoises (Testudinidae). In a few chelonians, the males are noticeably larger than females (e.g., Geochelone,

Melanochelys, Macrochelys).

Regardless of their size, mature males can be easily identified by their longer tail, which houses the penis. Other distinguishing sexual characteristics found in turtles include eye color (e.g., orangered in Terrapene carolina: Ernst & Barbour, 1972; Emys orbicularis: Pritchard, 1966), color of head skin marking (e.g., orange spots in Clemmys guttata: Ernst & Barbour, 1972), extended gular spurs in Gopher tortoises, plastral concavity in some sea turtles (Wibbels et al., 1991), and extralong claws on the front feet of male slider turtles (Trachemys scripta: Viosca, 1933). Sexual dichromatism is widespread in both pleurodires and cryptodires (Cooper & Greenberg, 1992); seasonal dichromatism, related to reproduction, is less well represented but definitely occurs (Cooper & Greenberg, 1992). More fieldwork must be conducted to elucidate the roles of these secondary sexual characteristics in the reproductive behavior of chelonians.

10.4.2 Mating Behavior

Although turtle reproductive behavior is complex, descriptions of pre-copulatory and mating behavior of turtles are relatively few (e.g., Pseudemys concinna, P. floridana: Cagle, 1944; Gopherus: Auffenberg, 1978; Chelydra serpentina, Apalone: Legler, 1955; Chelonia mydas: Bustard, 1972;

Macrochelys: Harrel et al., 1996; Macrochelys, Chelydra: Ewert, 1976; Caretta caretta: LeBuff, 1990; Terrapene: Dodd, 2001; Table 10.6) and often repeated (see illustrations in Oliver, 1955; Bellairs, 1970; Obst, 1986). Other reproductive behaviors include nest construction and concealment, as well as other actions (e.g., thermoregulation) that increase the fitness of an individual adult and, potentially, its eggs. Turtles exhibit stereotyped nesting behaviors that can be compared among groups because they contain several easily described phases (Ehrenfeld, 1979; Carpenter & Ferguson, 1977). The sequential behaviors include finding the site, selecting the specific location for the nest, digging the nest cavity, oviposition, covering the nest, and leaving the nest site (Ehrenfeld, 1979; Carpenter & Ferguson, 1977).

Chelonian courtship involves a variety of behaviors including following, head bobbing or swaying, nudging or biting, cloacal touching or sniffing, titillation via long foreclaws (Carpenter, 1980; Carpenter & Ferguson, 1977; Harless, 1979; Bels & Crama, 1994; Mason, 1992). Three general types of courtship patterns have been proposed based primarily on male behavior: premounting, mounting, and intermediate (Bels & Crama, 1994). Premounting courtship behaviors are found among aquatic species (e.g., Cheloniidae, Dermochelyidae, Emydidae, Chelidae) swimming in the three dimensional environment; mounting courtship behaviors occur while the male is mounted and preparing for intromission. The intermediate classification allows for variation in some precopulatory behaviors but does not contain species-specific behaviors (Bels & Crama, 1994). All three types are preceded by approaching, following/chasing or sniffing behaviors (Bels & Crama, 1994). Premounting and mounting courtship occur in both pleurodires and cryptodires. Premounting courtship seems to be widespread in cryptodires (Bels & Crama, 1994). The courtship type seems to be correlated with habitat type (i.e., open water, bottom dwelling, terrestrial), which may limit the range of movement of the turtle. The most elaborate courtship rituals occur in the three-dimensional environment of water.

Table 10.6

Examples of Sensory Reproductive Communication in Turtles*

* Based on Carpenter (1980); see also Carpenter & Ferguson (1977) and Harless (1979).

Courtship involves both male-female interactions and male-male interactions (Carpenter & Ferguson, 1982; see illustrations in Obst, 1986). The courtship includes olfactory recognition, gaping, and biting (Table 10.6). Multiple males may be attracted to a female in estrus and compete among themselves for the privilege to court her. Interactions among male terrestrial turtles (e.g., Gopherus) may be combative. The rival males may circle each other, head butt, bite, and ram each other in the flanks; this “jousting” may result in the loser being turned over, possibly to die. Male-female courtship is reminiscent of male-male interactions without attempting to turn her over, and involves olfactory recognition, ramming, and biting before mounting.

Among aquatic turtles, males interact by gaping at each other, biting, and shoving with their claws. The sequence of premating behaviors includes circling, sniffing, biting, or tickling. Painted turtles (Chrysemys, Trachemys) have evolved elaborate courtship rituals (Cagle, 1944). For

example, both Pseudemys concinna and P. floridana position themselves above their chosen female and extend their front legs and long claws down to tickle the sides of her neck. In both Trachemys scripta and T. terrapen, the male approaches the front of the female, almost touching nose to nose, and reaches forward to tickle the female along the sides of her neck before attempting to mount. Among some groups of freshwater turtles (e.g., Kinosternidae, Trionychidae), courtship rituals are simple and almost nonexistent; often, receptive females are mounted following olfactory inspection and a minimum of other interaction.

Among sea turtles, two or more males may attempt to court the same female; they bite and push each other while trying to attract her attention (Booth & Peters, 1972). As the male quickly moves to attempt to secure his position attached to her carapace, the female attempts to twist away. If successful in mounting, the male holds on to the lateral front and lateral rear parts of the female’s carapace with his longer, curved (first digit) claws. In this position, he is carried by the female; if she surfaces to breathe, he can breathe. He is also vulnerable to biting by attendant males along the exposed edges of his flippers and the dorsal part of his tail. It is not uncommon to find males with the posterior margins of their flippers raw and swollen—sometimes bleeding—during the mating season (Limpus, 1993). Flippers and tails may be gnawed to expose bone. These interactions may not dislodge the mounted male but serve to reduce his ability (fitness) to court and mount a female in subsequent mating activity, at least during the current breeding season.

10.4.3 Seasonal Timing of Reproduction

Although turtles lay their eggs on land, they do not show a uniform pattern of nesting. Some species show highly seasonal patterns (e.g., Chrysemys picta), whereas others may nest almost all year (e.g., Pseudemys floridana) (Figure 10.1). The pattern displayed by a species often correlates well with the temperatures required for adult activity, embryonic development, and hatchling survival (Moll, 1979; Georges et al., 1993). Males and females may share the same habitat but do not necessarily exhibit the same patterns of sexual selection or selection for fecundity. For example, male Chrysemys picta mature (become sexually active) at a smaller size than females (Wilbur, 1975b); this may result from the different responses to a reduced risk of predation with increasing size and the reproductive constraint imposed by egg production.

Chelonians breed when they have accumulated adequate stores of energy (Congdon & Tinkle, 1982; Congdon et al., 1983b; Gibbons, 1982). This depends on food quality and quantity, which are impacted—if not regulated—by environmental conditions. Some species residing at the same location may be affected more than others by environmental conditions, depending on their food requirements (Gibbons et al., 1983). The correlation between the El Nino-Southern Oscillation and the breeding biology of green turtles (Chelonia mydas) suggests that a two-year time lag is required between favorable foraging and oviposition (Limpus & Nichols, 1988, 2000); the proportion of females breeding each year varies as well (Limpus & Reed, 1985a, 1985b). Variation in environmental conditions experienced by Geochelone gigantea on Aldabra resulted in differences in the number of females breeding, clutch size, egg mass, and size at maturity in three subpopulations when comparisons were made over a long period of time (Swingland, 1977; Swingland & Coe, 1978, 1979).

The time of oviposition varies considerably among turtles and is often related to latitudinal distribution and the seasonality of rainfall. Licht (1984) described three patterns of reproduction for reptiles. Reproduction can either be continuous with comparable breeding levels all year, continuous with variable reproductive intensities, or discontinuous. Turtles fit into the second or third patterns. For instance, green turtles in the tropics nest all year round at some sites, although there is a substantial change in numbers between the summer and winter periods (Miller, 1997). An example of the discontinuous pattern occurs with the painted turtle in North America, where nesting occurs in the late spring or early summer regardless of latitude (Moll, 1973).

Deirochelys reticularia

FL

SC

VA

J F M A M J J A S O N D

Figure 10.1  Patterns of nesting seasons of North American freshwater turtles, showing changes in season associated with different latitudes. Emydoidea blandingii and Chrysemys picta show a northern nesting pattern with the nesting season restricted to the late spring and summer months. Deirochelys reticularia and Pseudemys floridana show a southern nesting pattern. FL = Florida, SC = South Carolina, VA = Virginia, Pen. FL= Peninsular Florida, NW FL = Northwestern Florida, MC = Michigan, WI = Wisconsin, IL = Illinois, LA = Louisiana, NE = North East, N. Scotia = Nova Scotia. Data: Aresco (2004); Gibbons & Greene (1978); Buhlmann (1995); Jackson (1988); Congdon et al. (2000); Iverson & Smith (1993); Moll (1973); Dinkelacker (unpublished).

Temperate species typically have discontinuous nesting seasons that are primarily influenced by seasonal temperatures (Legler, 1985; Kuchling, 1999). Kuchling (1999) describes the basal asynchronous reproductive cycles of temperate turtles. Briefly, most freshwater turtle species exhibit a seasonal nesting pattern in which eggs are typically oviposited in the late spring and early summer, development occurs during the summer, and hatching occurs in the early fall. There are no examples of continuous reproducing species in North America. However, several species approach a nearly continuous cycle in Florida (e.g., Deirochelys reticularia, Pseudemys floridana, and Kinosternon baurii), although in more northerly populations the nesting season becomes progressively more discontinuous and similar to other temperate species (Figure 10.1) (Gibbons & Greene, 1978; Jackson, 1988; Buhlmann, 1995; Ewert & Wilson, 1996; Aresco, 2004). Jackson (1988) provides an interesting discussion of “winter nesting” patterns in these species, as well as the effects of radiation into northern habitats. Legler (1985) described a similar temperate pattern for several species of Australian turtles in which females produced small flexible-shelled eggs with short incubation periods and the emergence of hatchlings before winter. The “temperate” pattern was exemplified by Emydura and Elseya, which nested during the early southern spring. All species in this study were widely distributed across the temperate and tropical zones, yet retained the cycle (i.e., temperate or tropical) of their origin. In other words, Emydura and Elseya retained a temperate nesting pattern even when found in tropical zones.

The nesting patterns of temperate species are closely related to their gonadal cycles (Kuchling, 1999). Most temperate freshwater turtles show post-nuptial gamete maturation (Terrapene carolina:

Atland, 1951; Terrapene ornata: Legler, 1960a; Chrysemys picta: Ernst, 1971a, 1971b; Chelydra serpentina: White & Murphy, 1973; Sternotherus odoratus: McPherson et al., 1982; Chelodina longicollis: Parmenter, 1985). In males, spermatogenesis routinely occurs during the hottest months of the year or immediately following the nesting season. The sperm are then stored in the epididymus during the fall and winter for use during spring mating, or are used during fall mating (Kuchling, 1999). In females, vitellogenesis may take up to a year or longer to complete (Ernst & Zug, 1994). Typically, vitellogenesis begins following oviposition in the summer or fall (Moll, 1979). At latitudes with sufficiently cold winters, there may be a pause in vitellogenesis during the coldest months. Vitellogenesis is completed either before or after hibernation, and in some cases just prior to ovulation (Moll, 1979). Thus, the reproductive cycles of males and females may be asynchronous, with the possible exception of turtles living in high latitudes (Kuchling, 1999). An explanation for the asynchrony may stem from the relative costs of the gonadal cycles. Spermatogenesis is relatively cheap and can be accomplished quickly at a time of year when more vigorous reproductive events are absent. Alternatively, vitellogenesis is energetically expensive and may require a prolonged period for the female to accumulate or convert a substantial quantity of her fat reserves into yolk. As such, vitellogenesis may begin when resources (i.e., food) are high and other reproductive activities are reduced (i.e., during the summer and fall). In addition, adaptations such as sperm storage and multiple paternity have permitted the female to mate at a time when vitellogenesis is not completed (Moll, 1979; Kuchling, 1999). Because the female can store the sperm from a single mating event in the fall, she may fertilize her eggs in the absence of a spring mating event. In addition, she may use stored sperm as well as “fresh” sperm to fertilize her eggs (Roques et al., 2006). The advantages and disadvantages of multiple paternity have received considerable attention in the literature (Galbraith, 1993; Valenzuela, 2000; Pearse et al., 2001, 2002; Lee & Hays, 2004; Roques et al., 2004, 2006; Jensen et al., 2006).

Tropically distributed species often have discontinuous nesting patterns as well. Pre-nuptial gamete maturation occurs in tropical and subtropical species (Lissemys punctata: Singh, 1977; Chelonia mydas: Licht et al., 1985; Caretta caretta: Wibbels et al., 1990; Geocholone nigra: Schramm et al., 1999). Rainfall seems to be the main factor dictating nesting in tropical species (Kuchling, 1999). Legler (1985) reported that in Australia, tropical species nested during the first six months of the year, which correlates to the dry season. Vogt (1997) described five patterns of nesting in South America: rainy season, end of rainy season, initiation of lowering rivers, lowest river depth, and spring nesters. Those species that nest during the dry season probably rely on exposed beaches, sand banks, or other nesting habitats that are not available during the rainy seasons. Data regarding gonadal cycles are scarce for freshwater turtles (Kuchling, 1999), and as a result the proximate and ultimate factors regulating reproduction and nesting in tropical turtles are still poorly understood.

10.4.4 Insemination and Sperm Storage

Fertilization is achieved internally via the penis being inserted into the cloaca of the female, sperm being transferred to the cloacal vestibule of the female, and sperm being accepted into the ostia of the oviducts. Once in the oviducts, sperm are carried via cilia to glands (Chrysemys picta: Parker, 1928, 1931; Chrysemys picta, Trachemys scripta elegans: Crowell, 1932; Mauremys japonica: Yamada, 1952) where, presumably, they mix with other sperm from previous copulations. The ciliated bands that move the sperm may be single or double. As an ovulated follicle enters the infundibulum, the folds are expanded; if any folds contain sperm, they are spread onto the surface of the follicle. As with birds, when one sperm penetrates the follicular layers, the coating of the follicle changes to prevent further penetrations. Polyspermy probably occurs at about the same rate as in birds and other animals, but there are no data available for polyspermy in turtles.

Female turtles from at least six families possess tubules in the oviduct in which sperm can be stored (Hattan & Gist, 1975; Palmer & Guillette, 1988; Gist & Jones, 1989; Sarkar et al., 2003). Sperm storage tubules have been reported from the anterior half of the oviduct, including

albumen-secreting glands and in the “uterine region of the oviduct near the uterovaginal junction” (Palmer & Guillette, 1988; Gist & Congdon, 1998). Sperm were not found in the center of the glandular region (Gist & Congdon, 1998).

An ultrastructural study of sperm storage in T. carolina (Gist & Fischer, 1993) reported that the albumen glands containing sperm were cytologically the same as those that lacked sperm. Because sperm were observed in the lumen and not in contact with the epithelium, the sperm were considered to be resident in the tubules, rather than tubules being “specialized for the maintenance of stored sperm” (Gist & Congdon, 1998). The oviducts of Trachemys scripta and Sternotherus odoratus store sperm in “undifferentiated tubules and ducts of submucosal glands at the periphery of large glandular segments of the albumen and uterine regions of the oviduct” (Gist & Congdon, 1998). Sarkar et al. (2003) reported sperm storage tubules in the wall of the isthmus of the oviduct of the Indian freshwater soft-shelled turtle (Lissemys punctata punctata). The tubules were lined by both ciliated and nonciliated epithelial cells. The depth and activity of the tubule epithelia were highest during the breeding season and lowest in the non-breeding season. The tubules of the isthmus also regress in the non-breeding season. Scattered, short tubules appeared in the infundibulum only during the breeding phase. Sperm were retained in the tubules of the isthmus from the time of mating until just prior to ovulation. Sperm then moved into the tubules of the infundibulum via material secreted from the tubule epithelium. They speculated that the release of sperm “might have been stimulated by estrogen secreted from the ovarian follicles of pre-ovulatory turtles” (Sarkar et al., 2003).

Gist and Congdon (1998) found that sperm transfer began in October for Sternotherus odoratus and in January for Trachemys scripta. They speculated that stored sperm were most likely used for fertilization of eggs in the second and subsequent egg clutches. In Chrysemys picta, testis weight changes as a result of annual sperm production cycle but the epididymis does not, although the epididymis contains sperm throughout the entire year in Chrysemys picta and Trachemys scripta (Gist et al., 2002). Sperm of both in Chrysemys picta and Trachemys scripta exhibit relatively low motility and velocity (Gist et al., 2002).

Sperm may be stored in the oviduct for extended periods. Sperm were found in seminal receptacles in the oviducts of Terrapene carolina 14 months after mating (Hattan & Gist, 1975). Sperm were identified in tubules of female turtles isolated from males for as long as 423 days (Gist & Jones, 1989). Viable eggs were produced by isolated females after 2 and 4 years in Malaclemys (Hildebrand, 1929; Barney, 1922) and after 4 years in Terrapene carolina (Ewing, 1943). These may be extreme cases but they provide evidence that sperm can be stored beyond the current mating season. Storage of viable sperm in females is facilitated by low motility and longevity; this in turn allows time for sperm movement within the oviduct (Gist et al., 2002). Pearse and Avise (2001) reported that one female Chrysemys picta used the sperm from the same male to fertilize her eggs in three consecutive years; given the lack of pair bonding, sperm storage seems the mostly likely explanation. Actual fertilization may be offset from the time of insemination (e.g., Chrysemys picta: Gist et al., 1990). The storage of sperm in males ensures that sperm will be available for transfer to the female, and sperm storage in the female ensures that sperm will be available for fertilization, even though the seasonal transfer may be restricted.

Pearse and Avise (2001) reported multiple-paternity in nine species (Pleurodira: Podocnemis expansa; Cryptodira: Chelydra serpentina, Glyptemys insculpta, Chrysemys picta, Dermochelys coriacea, Caretta, Chelonia mydas, Lepidochelys kempi, Gopherus agassizii) with 33% of 321 clutches analyzed having multiple paternity (range per species studied: 0 to 100% in 14 studies) and concluded that multiple paternity is a “common phenomenon” (albeit not universal). Many temperate species produce two or more clutches per reproductive season, depending on endogenous and exogenous factors (Ewert, 1979; Kuchling, 1999). Short-term sperm storage alleviates the female’s need to re-mate between oviposition of sequential clutches, whereas long-term sperm storage “puts strong selective pressure on males to produce sperm that can survive long-term storage” (Pearse & Avise, 2001).