| 英文摘要: | Climate change will alter many aspects of the ecology of organisms, including dispersal patterns and population connectivity1. Understanding these changes is essential to predict future species distributions, estimate potential for adaptation, and design effective networks of protected areas2. In marine environments, dispersal is often accomplished by larvae. At higher temperatures, larvae develop faster3, 4, 5, but suffer higher mortality4, 5, 6, making the effect of temperature on dispersal difficult to predict. Here, we experimentally calibrate the effect of temperature on larval survival and settlement in a dynamic model of coral dispersal. Our findings imply that most reefs globally will experience several-fold increases in local retention of larvae due to ocean warming. This increase will be particularly pronounced for reefs with mean water residence times comparable to the time required for species to become competent to settle. Higher local retention rates strengthen the link between abundance and recruitment at the reef scale, suggesting that populations will be more responsive to local conservation actions. Higher rates of local retention and mortality will weaken connectivity between populations, and thus potentially retard recovery following severe disturbances that substantially deplete local populations. Conversely, on isolated reefs that are dependent on replenishment from local broodstock, increases in local retention may hasten recovery. Organisms experience natural and anthropogenic disturbances, such as storms, overexploitation and pollution, that cause destruction of habitat and fragment metapopulations7. Recovery from disturbance relies heavily on the success of subsequent recruitment events to the disturbed site8. Recruits can originate locally through reproduction of the surviving individuals, or disperse from other locations to the affected areas9. For most marine species whose adults are sessile or relatively sedentary, the degree of local retention and dispersal depends on the hydrodynamic regime and the survival and development rates of propagules, such as larvae10, 11. Larval survival and development depend on temperature12. Temperature increases the activity of enzymes, accelerating fundamental biochemical processes and consequently metabolic rates13. Increased metabolic rates hasten larval development, potentially reducing the time larvae spend in the plankton and thereby increasing the likelihood that larvae will settle before being flushed from their natal reef3, 4, 5. However, increased metabolic rates are also likely to increase mortality during larval development, reducing overall recruitment success, including recruitment back to the natal reef as well as successful dispersal elsewhere4, 5, 6. Thus, whether rising temperatures will strengthen or weaken local retention is not obvious. To determine how ocean warming will influence such patterns, the temperature dependence of the larval mortality rate–development rate trade-off must be quantified. Here, we extended a dynamic model of coral larval dispersal14 to characterize how temperature affects the survival and acquisition of competence to settle, and calibrated our models by conducting a series of experiments testing the effects of temperature on larval survival and competence. We then integrated our calibrated survival and competence models with a model of larval retention15 to project temperature-induced changes in the proportion of larvae produced on a reef that successfully become competent to settle while retained on that natal reef (that is, the number of larvae settling on a reef as a proportion of the total number produced on that reef, hereafter termed local retention). We selected three scleractinian coral species (Cyphastrea japonica, Favites stylifera and Acropora millepora) to evaluate the effect of temperature (27–31 °C) on survival and competence dynamics. These species have lecithotrophic (aposymbiotic) larvae and span an almost twofold range in egg diameter (296.8 ± 2.6, 453.6 ± 4.8 and 541 ± 5.9 μm, respectively). Recent work suggests that broadcast-spawning coral species with smaller eggs acquire the capacity to settle more quickly than species with larger eggs11. Consequently, we expect these differences in egg size to systematically affect the influence of temperature on competence dynamics. We find that, across all temperatures, our models characterize competence and survival dynamics of broadcasting coral larvae extremely well (Fig. 1); in particular, model selection by AIC (Akaike’s information criterion) revealed the consistently earlier rate of acquisition of competence at warmer temperatures, and, for C. japonica and A. millepora, the higher mortality rates at warmer temperatures (Fig. 1 and Supplementary Tables 1–3). Together, the temperature dependence of survival and competence dynamics implies that ocean warming will boost local retention of broadcasting coral larvae on reefs with mean residence times of about 5 days or less (Fig. 2 and Supplementary Fig. 2). This encompasses most fringing and patch reefs worldwide, whose estimated mean residence times typically range from a few hours to about 5 days (Supplementary Information).
|
