Impact of marine heatwaves for sea turtle nest temperatures

Abstract

There are major concerns about the ecological impact of extreme weather events. In the oceans, marine heatwaves (MHWs) are an increasing threat causing, for example, recent devastation to coral reefs around the world. We show that these impacts extend to adjacent terrestrial systems and could negatively affect the breeding of endangered species. We demonstrate that during an MHW that resulted in major coral bleaching and mortality in a large, remote marine protected area, anomalously warm temperatures also occurred on sea turtle nesting beaches. Granger causality testing showed that variations in sea surface temperature strongly influenced sand temperatures on beaches. We estimate that the warm conditions on both coral reefs and sandy beaches during the MHW were unprecedented in the last 70 years. Model predictions suggest that the most extreme female-biased hatchling sex ratio and the lowest hatchling survival in nests in the last 70 years both occurred during the heatwave. Our work shows that predicted increases in the frequency and intensity of MHWs will likely have growing impacts on sea turtle nesting beaches as well as other terrestrial coastal environments.

Keywords: Chagos Archipelago; Granger causality testing; Hadley SST; climate change; temperature-dependent sex determination.

Figure 1.

Time series of mean monthly (a) sand temperature at nest depths on Diego Garcia, (b) coral reef water temperature at 15 m on Diego Garcia and (c) Hadley SST measured more broadly across the Indian Ocean. The warmest months in all time series occurred in early 2016, corresponding with an MHW and major coral-bleaching event. The relationships between mean monthly (d) water temperature at 15 m on Diego Garcia and Hadley SST, (e) sand temperature at nest depths on Diego Garcia and water temperature at 15 m on Diego Garcia and (f) sand temperature at nest depths on Diego Garcia and Hadley SST. In each case, these regression equations were highly significant (p < 0.01) with r² values of 0.81, 0.72 and 0.77, respectively. Due to logistical constraints of working at this remote nesting area, there was not a continuous rolling deployment of loggers, so some gaps when no loggers were deployed remain in our time series.

Figure 2.

Based on the mean monthly sand temperature (figure 1a), the predicted (a) hatchling sex ratio and (b) hatchling success on Diego Garcia. In both cases, extreme values were predicted for March 2016, which was the warmest month in the time series. (c) The time series of mean monthly Hadley SST around the Chagos Archipelago. The fitted line is the mean annual temperature which increased linearly over time at a mean rate of 0.15°C per decade (F_1,68 = 162, p < 0.001).