AFN vs. ASM or AFR

Some researchers may consider AFN identical to ASM or AFR, but it is not. ASM, as applied to females, is the age at which an individual female first becomes anatomically and endocrinologically capable of producing eggs and copulating. It precedes AFN by an undetermined and variable interval; i.e., ASM < AFN. According to Rostal (2005, 2007), sexual maturity in female Kemp's ridleys may occur in the same calendar year as first nesting. However, actual intervals between ASM and AFN in individual females are unknown and could be influenced by many factors, especially variation in distances traveled from foraging sites to nesting beaches (see discussions by Rostal 2005; Morreale et al. 2007; Rostal 2007). Although most Kemp's ridley nesting still occurs on the primary nesting beach near Rancho Nuevo, Tamaulipas, Mexico, the geographic range of nesting is expanding to the north and south of Rancho Nuevo (Heppell et al. 2007), which produces greater variation in routes and distances traveled to nesting sites. AFN precedes AFR (i.e., AFN < AFR), the age at which a first-time nester produces offspring (i.e., hatchlings). The interval between AFN and AFR is the incubation period, which also varies.

Previous Demographic Modeling

Previous demographic models were fitted to time series of annual numbers of Kemp's ridley nests (i.e., clutches laid) and hatchlings (females and males) released years earlier on nesting beaches in Tamaulipas (TEWG 1998, 2000; Heppell et al. 2005, 2007). Estimated or assumed mean ASM was among the inputs that represent parameters in these models. The number of years lapsed between the release of a cohort of hatchlings and nesting of females of the same cohort was set equal to estimated or assumed mean ASM. Hypothesized proportions of female hatchlings were included among model inputs, to convert annual numbers of hatchlings to annual numbers of female hatchlings. Model outputs included estimated annual numbers of nests as well as estimated survival rates for hatchlings and prebenthic juveniles, large benthic juveniles, and adults. However, the survival rate inputs for small benthic juveniles were not gender specific, having been derived from catch-curve analysis based on females and males combined. Model outputs were sensitive to estimated or assumed mean ASM inputs, and no doubt would also be sensitive to substituted inputs that represent AFN.

Availability of Males

Rostal (2005, 2007) showed that a Kemp's ridley female's first ovulation within a given nesting season follows courtship and mating, which indicates that a female cannot produce clutches within that season without having copulated, whether a first-time or repeat nester. Therefore, AFN as well as AFR (but not ASM) could be affected by the availability of males. However, for nesters that have copulated in a given season, there is no guarantee that all eggs that they lay in a nest or during that season will be fertilized. Coyne and Landry (2007) expressed concern that abundance of males could possibly become limiting if manipulative conservation methods altered natural sex ratios. The level of fertilization of clutches laid by first-time nesters that have copulated affects hatch rates of their clutches but not AFN of such nesters.

Previous Accounts of AFN

To our knowledge, there exist only 5 published direct observations of Kemp's ridley AFN. However, all involved individuals were captive-reared at Cayman Turtle Farm, Ltd (Wood and Wood 1984, 1988, 1989). Two 5-year-old turtles produced eggs for the first time in 1984. One produced 7 eggs in a seawater tank, and none developed; this was not nesting under our definition of AFN. The other 5-year-old turtle deposited 62 eggs in a nest cavity, but only 3 hatchlings emerged, all of which died. Four 7-year-old turtles nested for the first time in 1986, laying a total of 535 eggs that averaged 67 per nest (range, 11–103 eggs) and producing 78 viable hatchlings. Such young ages of first-time nesters are interesting, but they do not represent free-living nesters. It is possible that free-living Kemp's ridley females may also release eggs at sea if nesting is interrupted or otherwise prevented. If they do release eggs at sea, then this could be a complicating factor in determining AFN for such individuals. In other words, if a would-be first-time nester released eggs at sea and was later observed nesting for the first time, then the turtle might be misidentified via laparoscopy as a repeat nester.

Anecdotal estimates of AFN in free-living Kemp's ridleys also can be found in Marquez (1994) and Francis (1978), but they are only approximations. These accounts were based on releases of unmarked hatchlings on beaches where hatchling production was assumed to have been low or nonexistent for years, followed by nesting by putative neophyte nesters on the same beaches ≥ 10 years thereafter.

Identifying First-time Nesters

The absence of tags (external or internal) and the presence of other characteristics that suggest a female might be a first-time nester are not sufficient evidence to identify first-time nesters. However, all characteristics that have been used to identify “young” (Marquez 1994) or putative neophyte nesters (Pritchard 1990; Leo Peredo et al. 1999) should be applied in searching for first-time nesters. Laparoscopies and carapace length measurements must be conducted on subadult females and previous nesters as well as on first-time nesters, all of which have a wide range in carapace lengths. Results for subadults and previous nesters will serve as controls for comparison with those of first-time nesters.

Age Determination

Skeletochronology cannot be used to determine ages of live Kemp's ridleys, because sacrificing living specimens of endangered species is prohibited by the U.S. Endangered Species Act of 1973 as amended. Laparoscopy can identify first-time nesters but cannot determine their ages. One direct but long-term approach to determining age of nesters requires tagging large numbers of newly emerged hatchlings with permanent tags unique to year-classes, then later recapturing nesters from the same cohorts. At Rancho Nuevo, mass tagging of several year-classes of Kemp's ridley hatchlings with coded wire tags was conducted (Fontaine et al. 1993; Caillouet et al. 1997b; Higgins et al. 1997), as recommended by Eckert et al. (1994). However, relatively few recaptures from these releases were reported (Snover et al. 2005), none of which were nesters. Living tags have been successfully applied to hatchling loggerheads (Caretta caretta) and green turtles (Chelonia mydas) (Hendrickson and Hendrickson 1981). They also were experimentally applied to Kemp's ridley hatchlings, but mortalities occurred, and further testing was halted. Further development and testing of techniques for tagging hatchlings with permanent tags is recommended. However, age determination by mass tagging of hatchlings requires a substantial, long-term commitment to searching for the tagged turtles, especially on nesting beaches.

Age can be estimated from carapace length, with appropriate somatic growth curves. Mark–recapture such as that applied to postpelagic life stages of green turtles by Limpus and Chaloupka (1997) could be useful in developing improved somatic growth curves for Kemp's ridley. Such growth curves could then be applied to estimate AFN from carapace length at first nesting (CLFN) for samples of first-time nesters. CLFN is our generalization for any standard measure of carapace length (Bolten 1999); one standard measure can be transformed to another with an appropriate regression equation. Kemp's ridley ASM has been estimated from fitted somatic growth curves (Marquez 1994; Chaloupka and Zug 1997; Schmid and Witzell 1997; TEWG 1998, 2000; Snover et al. 2007, 2008). Chaloupka and Musick (1997) reviewed studies of sea turtle age, somatic growth, and population dynamics in the context of methodological issues relevant to population modeling, underlying statistical properties, and assumptions. They discussed numerous problems associated with application of somatic growth curves in estimating ASM, and their discussion would apply likewise to the use of somatic growth curves to estimate AFN from CLFN.

Chaloupka and Musick (1997) noted that there are no conclusive growth criteria to indicate onset of sexual maturity. Minimum or mean carapace length of nesters or an arbitrary carapace length set slightly below mean carapace length of nesters have commonly been used as indicators of size at sexual maturity. Limpus (1992) detected considerable variation in carapace length of first time nesting hawksbill turtles (Eretmochelys imbricata). Limpus and Chaloupka (1997) questioned the use of either minimum or mean carapace lengths of nesting green turtles in estimating ASM from somatic growth functions. The same issues apply to estimation of AFN in Kemp's ridley from somatic growth curves. The probability-based approach of Heino et al. (2002), in which age and size are treated as continuous, interdependent variables, might also be applicable to determining the bivariate distribution of age and carapace length of Kemp's ridley subadult females and nesters. However, it probably would require much larger sample sizes than when using somatic growth curves of females to estimate AFN from CLFN.

Previous estimates of Kemp's ridley ASM from somatic growth curves were based on various assumed carapace lengths at sexual maturity. However, the data to which these growth curves were fitted included males and females. Marquez (1994) assumed that female Kemp's ridleys reached maturity at 58.0-cm straight carapace length (SCL); his estimates of 6–7 years at maturity were based on mark–recapture and captive-rearing data. Snover et al. (2007) reviewed 7 estimates of Kemp's ridley age at maturity that ranged from 10 to 17 years. All were estimated from von Bertalanffy growth curves over a range of assumed SCL at maturity of 60–65 cm. Gregory and Schmid (2001) and Coyne and Landry (2007) suggested that 60-cm SCL should be considered the minimum size of adults. The U.S. Sea Turtle Stranding and Salvage Network uses 60-cm SCL as the minimum size that defines the adult life stage for both females and males (D. Shaver, pers. comm., September 2009). Zug et al. (1997) suggested the size at which females mature could be as high as 65-cm SCL. Caillouet et al. (1995) and Snover et al. (2008) chose 60-cm SCL as the hypothetical size at which head-started females matured after release into the wild.

Most somatic growth curves for free-living Kemp's ridleys have been based on samples that contained males and females in unknown proportions. It was assumed implicitly that the growth pattern was the same in both sexes. In many cases, the data sets used to fit somatic growth curves contained few if any nesters (e.g., Schmid and Witzell 1997). In contrast, growth curves for head-started Kemp's ridleys tagged in multiple ways and released into the Gulf of Mexico were based on samples dominated by females, but some males could have been included (Caillouet et al. 1995; Shaver and Wibbels 2007; Wibbels 2007; Snover et al. 2008). The overall sex ratio of the hatchlings received for head-starting was strongly female biased (Caillouet 1995, 2000; Shaver 2005; Shaver and Wibbels 2007; Wibbels 2007). The overall sex ratio of recaptures was also female biased (Caillouet 1995, 2000; Landry et al. 2005; Shaver 2005; Shaver and Wibbels 2007; Wibbels 2007). Obviously, samples that contain only females should be used for purposes of determining the relationship between age and size for purposes of estimating AFN from CLFN. Laparoscopy will assure that only females are included in the samples.

The estimates of free-living female Kemp's ridley ASM could have been disproportionately influenced by more numerous smaller turtles in the samples, i.e., adults were less numerous in these samples. We recognize that this was, in part, due to declining numbers with age, which makes it difficult to obtain comparable samples over the size and age ranges of females in the population. Nevertheless, there seems little if any reason to include small juveniles in the samples when the objective is to estimate AFN from CLFN. Excluding small juveniles will assure that they do not influence the shape of fitted somatic growth curves or bivariate distributions of age and size. Day and Taylor (1997), Chaloupka and Musick (1997), and Chaloupka and Zug (1997) advised against applying the von Bertalanffy growth model over the entire life span. Day and Taylor (1997) pointed out that, during the prematuration phase, surplus energy supports exponential growth; during the maturing phase, surplus energy supports growth and maturation; and, during the postmaturation phase, surplus energy supports reproduction. Furthermore, there is evidence that growth over the life span is biphasic for head-started Kemp's ridleys (Snover et al. 2008) and polyphasic (Chaloupka and Zug 1997) for free-living Kemp's ridleys. Eliminating small juveniles from the samples might avoid such problems. Growth models other than the von Bertalanffy might also be applicable to samples that exclude small juveniles.

Limiting samples to females ≥ 40-cm SCL would include subadult and adult life stages, and would capture the transition from initiation of ovarian development within the subadult stage (Gregory and Schmid 2001; Schmid and Barichivich 2005) through sexual maturity and beyond. Ovarian development in free-living Kemp's ridleys begins around 50 cm SCL, and it may take several years thereafter for females to fully mature (Schmid and Barichivich 2005). By excluding turtles < 40-cm SCL from the data set used to determine the relationship between age and carapace length, most of the prematuration phase of growth will be excluded, and application of the von Bertalanffy growth model might then be valid.

Head-started Kemp's Ridley Example

Hatchling emergence dates for head-started Kemp's ridleys were known, therefore, ages were known during captive rearing (Caillouet et al. 1997a) and at recapture after release into the Gulf of Mexico (Caillouet et al. 1995; Snover et al. 2008). The ages of Kemp's ridleys that spent their entire lives in captivity also were known (Wood and Wood 1984, 1988, 1989; Fontaine et al. 1989). We fitted a von Bertalanffy growth curve to SCL and known age, t (yr), of Kemp's ridleys that had been head-started, released into the Gulf of Mexico, and recaptured in the Gulf of Mexico. We not only limited the data set to sizes ≥ 40-cm SCL (Fig. 1; Tables 1 and 2) but also applied the same data selection criteria used by Caillouet et al. (1995). Data for recaptured prenesters with ≥ 40-cm SCL were provided by the National Marine Fisheries Service Laboratory, Galveston, Texas, and data for documented nesters were provided by D. Shaver (unpubl. data, 2009). The data subset contained 82 recaptures, including 49 known (documented) nesters and 33 other individuals (Tables 1 and 2). The “other” category may have contained some males, undocumented nesters, and some prenesters. One of the 49 nesters was only observed crawling on the beach, but we assumed that it nested. We included only Gulf of Mexico recaptures, because they were much more numerous, and we assumed they grew faster and matured sooner in the Gulf than in cooler waters of the US Atlantic coast (Fontaine et al. 1989; Caillouet et al. 1995); however, some might have spent time in the Atlantic before returning to the Gulf. Hatchlings within a year-class often had multiple emergence dates, because they came from different clutches and imprinting locations (most came from Padre Island National Seashore and Rancho Nuevo, and a few came from Cayman Turtle Farm Ltd). For each year-class, a weighted (by number of hatchlings) mean emergence date was used to calculate ages of recaptured individuals. Therefore, ages of recaptures were within ± 2 weeks of actual ages. Incidentally, no head-started Kemp's ridleys have been documented nesting outside the Gulf of Mexico.

View Figure

• Download as Powerpoint
• Download Figure

Table 1. Distribution of year-classes for 82 head-started Kemp's ridleys ≥ 40 straight carapace length (SCL) when recaptured in the Gulf of Mexico after release into the Gulf of Mexico (see Fig. 1).

View Fullscreen

Table 2. Straight carapace length (SCL) (cm) and known-age t (yr) statistics for 82 head-started Kemp's ridleys ≥ 40 SCL when recaptured in the Gulf of Mexico after release into the Gulf of Mexico (see Fig. 1).

View Fullscreen

Estimated asymptotic SCL, 64.08 cm (Fig. 1), was 1.81 cm higher than the 62.27 cm estimated by Caillouet et al. (1995) and 1.4 cm lower than the mean SCL (65.5 cm) of the 69 nesters measured at Rancho Nuevo during the 1990–2000 nesting seasons (Witzell et al. 2005). The sample used by Caillouet et al. (1995) included individuals < 40-cm SCL as well as ≥ 40-cm SCL, but it did not include known nesters or any individuals > 9 years old. The sample used by Snover et al. (2008) included individuals < 40-cm SCL and ≥ 40-cm SCL, but few nesters. The largest head-started nester in our sample (Fig. 1) had a 65.8-cm SCL, but it was only 12.9 years old; the oldest nester was 22.8 years old, with a 60.5-cm SCL. Interestingly, the 3 youngest head-started nesters were 9.7–9.9 years old, with 59.5–64.3-cm SCL. For the future, we recommend exclusion of turtles < 40 cm SCL from samples of free-living female Kemp's ridleys used to determine the relationship between age and carapace length for purposes of estimating AFN from CLFN.