Climatic and environmental drivers of extinction in Mediterranean island reptiles since the height of the last Ice Age

Abstract

As the earth experiences an accelerating wave of climate change-driven species extinctions, it is increasingly important to understand how natural species communities respond to the dual stressors of climate and landscape modification. Island species represent an excellent system in which to study the effects of shifting climate and concomitant landscape change on wildlife populations. Reptiles in particular are ideal study organisms because they are, like most species worldwide, 1) ectotherms, making them directly susceptible to changes in the thermal environment and 2) poor dispersers, attenuating their ability to respond to environmental changes by movement or migration. In this study we take advantage of a natural long-term fragmentation process that occurred when, following the end of the last ice age, rising sea levels led to the separation of numerous Mediterranean land-bridge islands from the nearby mainland. Reptile species occurring on these islands became isolated and subject to numerous stressors such as small population sizes and an increasingly inhospitable thermal environment, and many taxa went extinct over these millennia. Because these stressors closely reflect the problems that present-day wildlife species face, they can help us understand the process of extinction and serve as a model to understand the long-term implications of anthropogenic landscape change on wildlife. In this study, we assess how reptile population persistence on Mediterranean islands has been influenced by the interacting effects of island area, timing of fragmentation, changing climate, and changing topography, in the period since the height of the last ice age (the Last Glacial Maximum, LGM). We use bathymetry, topography, and geophysical models of sea-level rise to reconstruct the sequence and timing by which 83 islands in the Aegean and Ionian seas progressively separated from paleo-landmasses, and to characterize the topography of paleoislands from which they fragmented. We also reconstruct the progressive sequence of local reptile population extinctions that occurred on these islands going back to the LGM. Combined with publicly available paleo-climate reconstructions, we characterize a set of landscape and climate variables for each present-day and paleo-island for which we have reconstructed extinction events. Our analysis revealed that extinction increased directly with diminishing island size and tended to accumulate linearly with increased duration of isolation. We also find that extinctions are significantly positively associated with higher warmest-season temperatures, and with higher topographic roughness, which may be a measure of diminishing resource availability. We also find that extinctions are positively associated with higher pre-fragmentation precipitation, an indication that reptile population persistence may be negatively affected by large differences between pre- and post-fragmentation vegetation lushness. These conclusions point forward to predicting, anticipating, and eventually preventing future species extinctions due to environmental change.