La Roche Percée, in New Caledonia, is the most important loggerhead turtle rookery outside of Australia for the eastern Pacific genetic stock. The females nesting on this beach are genetically similar to the females found at the Mon Repos rookery in Queensland, Australia. This study shows how nest temperature affects the phenotype of genetically similar populations. During the 2010–11 breeding season, mean nest temperatures were significantly higher at La Roche Percée (31.8°C) than at Mon Repos (29.5°C) and the mean for the three-days-in-a-row maximum nest temperatures was also significantly higher at La Roche Percée (34.6°C), than at Mon Repos (31.7°C). Differences were found in mean hatching success (La Roche Percée 83 ± 3%, Mon Repos 96 ± 2%) and emergence success (La Roche Percée 76 ± 3%, Mon Repos 93 ± 3%). Hatchlings from La Roche Percée also had significantly lower fitness characteristics, having smaller carapace size (La Roche Percée 1565 ± 7 mm2, Mon Repos 1634 ± 5 mm2), slower self-righting times (La Roche Percée 4.7 ± 0.1 s, Mon Repos 2.7 ± 0.1 s) and slower crawling speed in terms of both absolute speed and body lengths per second (La Roche Percée 2.5 ± 0.2 cm s–1 or 0.57 ± 0.05 body lengths s–1, Mon Repos 4.6 ± 0.1 cm s–1 or 1.04 ± 0.02 body lengths s–1). Nest temperatures at La Roche Percée approached the upper limit of embryo thermal tolerance towards the end of incubation (34°C) and this condition may contribute to the lower hatching and emergence success and lower fitness characteristics of hatchlings at the La Roche Percée rookery.

Context Migratory species are known to pose a challenge for conservation because it is essential to understand their complex life history in order to implement efficient conservation actions. Aims In New Caledonia, large seagrass habitats in the Grand Lagon Sud (GLS) are home to resident green turtles (Chelonia mydas) of unknown origins. To assess the stock composition in the GLS, 164 foraging turtles were sampled for genetic analysis of ~770 base pairs of the mitochondrial DNA (mtDNA) control region. Methods Foraging turtles ranging in size from 48.0 to 108.4 cm curved carapace length were captured at five different sites within the GLS between September 2012 and December 2013. To provide baseline data for mixed stock analysis, published data from rookeries were used in addition to 105 samples collected at rookeries in the d’Entrecasteaux Islands and Chesterfield Islands in New Caledonia and at Malekula Island in Vanuatu. Exact tests of population differentiation and pairwise FST estimates were used to test for differences in mtDNA haplotype frequencies. Key results These analyses indicated that rookeries in the d’Entrecasteaux Islands and Vanuatu form unique management units and that the Chesterfield Islands rookeries are linked to the Coral Sea management unit. Mixed stock analysis indicated the highest proportion (mean = 0.63) of foraging turtles originate from the d’Entrecasteaux stock. Conclusions The larger contribution is estimated to be from a large rookery from New Caledonia, but smaller contributions are suggested from other rookeries in the South Pacific. Implications Marine conservation policies in New Caledonia need to consider the links between the foraging and nesting populations of C. mydas in New Caledonia and other rookeries and foraging grounds in the Coral Sea.

No abstract