Date of this Version
Austral Ecology. 2023;00:1–16.
Species distribution models often suggest strong links between climate and species' distribution boundaries and project large distribution shifts in response to climate change. However, attributing distribution shifts to climate change requires more than correlative models. One idea is to examine correlates of the processes that cause distribution shifts, namely colonization and local extinction, by using dynamic occupancy models. The Cape Rock-jumper (Chaetops frenatus) has disappeared over most of its distribution where temperatures are the highest. We used dynamic occupancy models to analyse Cape Rock-jumper distribution with respect to climate (mean temperature and precipitation over the warmest annual quarter), vegetation (proportion of natural vegetation, fynbos) and land-use type (protected areas). Detection/non-detection data were collected over two phases of the Southern African Bird Atlas Project (SABAP): 1987–1991 (SABAP1) and 2008–2014 (SABAP2). The model described the contraction of the Cape Rock-jumper's distribution between SABAP1 and SABAP2 well. Occupancy probability during SABAP1 increased with the proportion of fynbos and protected area per grid cell, and decreased with increases in mean temperature and precipitation over the warmest annual quarter. Mean extinction probability increased with mean temperature and precipitation over the warmest annual quarter, although the associated confidence intervals were wide. Nonetheless, our results showed a clear correlation between climate and the distribution boundaries of the Cape Rock-jumper, and in particular, the species' aversion for higher temperatures. The data were less conclusive on whether the observed range contraction was linked to climate change or not. Examining the processes underlying distribution shifts requires large datasets and should lead to a better understanding of the drivers of these shifts.