Number of breeders and population density
The operational sex ratio (OSR) was slightly biased toward females (0.87:1 male: female), indicating that more females than males breed in this population. OSR in sea turtles has typically been reported as skewed towards males or even between both sexes (Hays et al., 2022). For example, Prakash et al. (2022) found an OSR of 1:1 in a population of hawksbill turtles in Fiji, and Wright et al. (2012b) found an OSR of 1.4:1 in a green turtle population in Cyprus. The primary sex-ratio of this population in Tetiaroa does not seem to suffer heavy female bias as observed elsewhere (Laloë et al., 2020), and so the discrepancy between the number of male and female breeders is likely small. However, a low level of multipaternity was observed (7% of the clutches, 16% of the females) and green turtles demonstrated multipaternity in as much as 92% of their clutches (Alfaro-Núñez et al., 2015). Our finding constitutes the lowest estimation of multipaternity level in this population, considering that 70 clutches were analysed over a total of 549 observed. Plus, the number of hatchlings genetically sampled was uneven in the clutches (most often < 10), thus the contribution of some fathers may have been missed. With no apparent fitness benefits, a low level of multipaternity may indicate low chances of encountering multiple males, directly linked to the density and number of male breeders (Lee et al., 2018). However, a high density of breeders is observed near the island (Gaspar, pers. comm.), and 11% of the males mated with multiple females within a season, indicating that males may encounter several available females. Coupled with the low level of multipaternity, it suggests that not all males are equally capable of mating with multiple females. In contrast, Phillips et al. (2013) found no male hawksbills that mated multiple times within a season and concluded that a low density and high turnover of males prevented this from occurring.
Another deduced behaviour of interest arose from the two reconstructed males *53 and *62. They both sired female CMY1384 in 2018/19 and a different female CMY2419 in the next season. This finding is surprising given the low number of breeders retained and the low level of multipaternity observed, though the probability of such an occurrence increases when the number of male breeders is low and the density in the mating area is high. This may indicate that some males are dominant on the breeding ground. Dominance hierarchies in males have been described in other species of freshwater turtles like the common snapping turtle (Chelydra serpentina ) and the wood turtle (Clemmys insculpta ), and in the gopher tortoises (Gopherus agassizii ) (Pearse and Avise, 2001). Such dominance could occur regularly on the breeding ground, or be year-specific in the case where CMY2419 was sired in 2018/19 and able to store sperm until the next season. Alternatively, non-random mating choices driven by a restrictive criterium may limit the number of suitable partners for a given female, increasing the probability of mating with the same pair of males. Nonetheless, these contrasting results are the first to look into the number of male breeders in French Polynesia and show that differences between the two sexes are likely to be small and that this number seems sufficient to maintain fitness in the population despite high levels of relatedness.