Ethics Approval
Chelonia mydas sample collection was authorized and coordinated by the Direction of the Environment of French Polynesia. Non-lethal skin and muscle biopsies were performed, which do not require any other specific permits. Samples were exported to France for processing with CITES permits n° FR1298700118-E and n° FR2098700187-E. All samples remain the DIREN’s property.
Abstract
The reproductive systems of natural populations can greatly impact their genetic diversity by preventing or encouraging inbreeding. It is therefore crucial to have a comprehensive understanding of the mating system to evaluate a population’s ability to maintain genetic diversity over time. In this study, we examine the mating system of an endangered population of green sea turtles in Tetiaroa, French Polynesia. We determine if different mating behaviours serve as strategies to avoid inbreeding. We genotyped 107 nesting females and 1483 hatchlings from 549 nests and used 23 microsatellite markers to reconstruct the genotypes of the fathers. We assessed the level of inbreeding and relatedness of the parent pairs and explored the correlation between relatedness and fitness parameters in the offspring. We determined the mating behaviours of both males and females and investigated if specific behaviours were linked to different levels of relatedness. Our results showed that 27 fathers and 31 mothers were responsible for the genotypes of 445 hatchlings from 105 nests. Global Fis was significant, and levels of relatedness were higher than expected through random mating, indicating inbreeding and non-random partner selection. However, we did not find any mating behaviours that were associated with lower relatedness levels than the general population, suggesting that they are not part of an inbreeding avoidance strategy. Ultimately, this study illuminates the reproductive system of green turtles and shows that this population is susceptible to inbreeding. Additionally, our research demonstrates the effectiveness of parentage analysis in understanding the reproductive behaviour of elusive species.
Keywords: genotype reconstruction, mating behaviour, fitness, relatedness, multipaternity, inbreeding
Introduction
Reproductive systems strongly influence genetic diversity in natural populations (Clegg et al., 1992). Loss of genetic diversity may lead to inbreeding depression, loss of adaptive potential, and accumulation of deleterious alleles, and ultimately to extinction (Charlesworth, 2009). The number of breeders, their reproductive success, and the mating strategies they use are key elements present in the mating system of a population (Sugg and Chesser, 1994). A good understanding of these factors is essential to assess the intrinsic capacity of natural populations to maintain themselves through generations and to maintain their genetic diversity (Anthony and Blumstein, 2000). Several components of the mating system may either mitigate or favour inbreeding. They may be precopulatory such as the operational sex-ratio (OSR, i.e. relative number of breeders from both sexes, (Emlen and Oring, 1977)), reproduction frequency, and partner choice and number (Blouin and Blouin, 1988; Taylor et al., 2014). Postcopulatory factors include sperm storage and sperm competition (Michalczyk et al., 2011).
OSR is a key feature of the mating system, especially for species with a temperature-dependent sex determination (TSD) like the marine turtles (Standora and Spotila, 1985). The primary sex-ratio of TSD species is solely dependent on environmental conditions and can be heavily biased under climate warming situations (Janzen, 1994). The primary sex-ratio of several marine turtle populations is female-biased (Casale et al., 2000; Jensen et al., 2018; Santidrián Tomillo et al., 2014; Zbinden et al., 2007), but the OSR is usually much more balanced (Hays et al., 2022), mitigating concerns about the vulnerability of a population. However, as highlighted by Wright et al. (2012a), if the OSR reflects a small number of males that breed more frequently than females, it will still lead to a loss of genetic diversity and inbreeding. Due to the high energetic cost of reproduction, which may include long migrations between foraging and nesting grounds, female marine turtles typically reproduce at intervals of several years (Hays et al., 2014). For the green turtle, Chelonia mydas , reproduction frequency in females is estimated between 2 and 5 years (Seminoff et al., 2015). In contrast, males can mate more frequently, and annual migration to breeding grounds has been observed in loggerhead (Caretta caretta ), leatherback (Dermochelys coriacea ) and green turtles (James et al., 2005; Limpus, 1993; Wibbels et al., 1990). Knowing the reproduction frequency of both sexes is thus important to correctly interpret the OSR concerning inbreeding risk.
Both male and female marine turtles can reproduce with several partners (i.e. polygynandry), which tends to equilibrate the OSR (Jensen et al., 2013). Polyandry is observed through multiple paternity in clutches, which has been found at various degrees in all marine turtle species. For green turtles, it ranges from 15% to 92% of the clutches depending on the populations (reviewed in Lee et al., 2018). Multiple paternity is a direct consequence of seasonal sperm storage in females, a capacity well established in Testudines (Pearse and Avise, 2001) and observed in six of the marine turtle species (Crim et al., 2002; FitzSimmons, 1998; Kichler et al., 1999; Phillips et al., 2013; Sakaoka et al., 2011; Theissinger et al., 2009). Satellite tracking and direct observations showed that breeding usually occurs before the beginning of the nesting season and all the clutches laid within a season are usually sired by the same fathers (Hays et al., 2022). However, sperm storage across multiple seasons has not been formally demonstrated in marine turtles, as opposed to terrestrial and freshwater turtles (Owens, 1980; Whitaker, 2006), but several studies have suggested that it is likely (Howe et al., 2017; Theissinger et al., 2009; Wright et al., 2013).
It has been proposed that females would benefit from polyandry and sperm storage. The ‘good genes’ hypothesis relies on the assumption that by mating with different partners, better quality sperm would outcompete lower quality sperm, which would lead to an increased fitness of embryos sired by the dominant father (Kokko et al., 2002). However, while many studies have attempted to prove this hypothesis, no correlation has yet been found between multipaternity and fitness parameters in the clutches (Jensen et al., 2006; Lee et al., 2018; Lee and Hays, 2004; Wright et al., 2013, but see Howe et al., 2017). On the other hand, the ‘genetic compatibility’ hypothesis predicts that paternity would be biased towards genetically dissimilar males to avoid inbreeding (Bretman et al., 2009; Zeh and Zeh, 1997). In an attempt to test this latter hypothesis in a hawksbill turtle population in the Republic of Seychelles, Philips et al. (2013) did not find any correlation between relatedness and paternity contribution in the clutches. Ultimately, all of these studies tend to conclude that polyandry is likely to occur as an energy cost trade-off between mating several times and avoiding mating harassment (i.e., convenience polyandry, Lee and Hays, 2004). Male sea turtles are known to actively and aggressively attempt mating, and females may try to avoid it (Booth and Peters, 1972). Thus, levels of multipaternity may simply reflect the density of breeders on reproductive grounds and the number of encounters between the two sexes (Lee et al., 2018).
Understanding the mating system of sea turtles is challenging, notably due to the lack of observations of males and breeding grounds. For species as elusive as these, molecular analyses can provide important insights. The use of microsatellite markers allowed for the genotypes of unsampled fathers to be reconstructed and to thus assess the mating system parameters in several marine turtle populations (e.g. Bernatchez and Duchesne, 2000; Figgener et al., 2016; Horne et al., 2022; Phillips et al., 2013; Wright et al., 2012). Together, these studies reveal that regional variations in these parameters are the rule rather than the exception, and many regions are still lacking this key information for their populations.
French Polynesia, in particular, is an archipelago composed of 118 islands distributed over an exclusive economic zone of 5 million km2, a surface as wide as Europe. The country is thought to host approximately 1,000 female green turtle breeders, although no recent assessment exists (Groombridge and Luxmoore, 1989; Seminoff et al., 2015). In the centre of French Polynesia, Tetiaroa Atoll (Society archipelago) is one of the major nesting grounds. It is estimated that around 100 females nest annually on the atoll (Seminoff et al., 2015; Touron et al., 2018), and mating behaviours are observed around the island (Gaspar, pers. comm.). This location provides a unique opportunity to closely investigate the mating system of green turtles with molecular marker analyses. Here, we genotyped nesting females and hatchlings from Tetiaroa with 23 microsatellite markers, reconstructed male genotypes, and conducted parentage analysis to study the mating behaviours of both sexes.
The aim of this study is threefold. First, we characterize the level of inbreeding and relatedness in the population of green turtles nesting in Tetiaroa and explore their effect on fitness, by measuring three fitness parameters in the clutches: the number of eggs, successful hatchlings, and dead embryos. Second, using reconstructed pedigrees, we determined the features of the mating system and nesting parameters of this population. These parameters, which are key to inbreeding depression dynamics, include the OSR, the reproductive frequency, the level of multipaternity, the female’s nesting intervals, and the number of partners for the males. Third, we test whether the mating strategies deployed reduce the overall relatedness and contribute to avoiding inbreeding by comparing the relatedness of couples involved in these strategies with the overall relatedness of the population. To our knowledge, this work is the first to explore the relatedness bias of mating behaviours in green turtles. It improves our general understanding of the drivers of mating strategies in sea turtles.
Material and Methods