The Impacts of Climate on the Past, Present, and Future Distributions of Species in the American Southwest and Its Madrean Sky Island Archipelago

Abstract

The southwest United States (US) has a complex geologic and climatic history that have affected the diversity and distribution of species throughout the region. The ground beetle (Coleoptera: Carabidae) Synuchus dubius (LeConte, 1854) is a flightless species restricted to high-elevation (>1800 meters) conifer forests and has a widespread, yet fragmented distribution across the southwest US. Phylogeographic analyses revealed multiple diverging lineages that resulted largely from climate oscillations during the Pleistocene Epoch. S. dubius is not the only species in the southwest US with unique genetic lineages. Many others also have cryptic diversity, especially in the Madrean Sky Island Archipelago (Madrean Archipelago). This region is a biodiversity hotspot located in southeastern Arizona, southwestern New Mexico, and northern Sonora, Mexico. With its 65 isolated mountain ranges rising from low elevation desert to well over 2500 meters, the Madrean Archipelago supports up to eight biological communities that are ordered along elevational gradients—a phenomenon known as biome-stacking. As a result, the region contains an incredibly rich assortment of species. However, higher temperatures and aridity predicted under future climate change are expected to affect the region, with species in the uppermost biomes (i.e., conifer forests) being especially vulnerable. Species distribution models for multiple high-elevation taxa suggest that under 21st-century climate change, the Madrean Sky Islands will become increasingly isolated due to conifer forest habitat loss which may have negative consequences for the region’s cryptic diversity. MAXENT is a popular species distribution modelling program that uses presence-only data to estimate habitat suitability. Previous works have shown that MAXENT’s likelihood is identical to that of an inhomogeneous Poisson process (IPP) conditioned on the number of occurrence records. Consequently, MAXENT’s raw output may, under specific assumptions, be considered proportional to relative abundance. Using Arizona occurrence records for the mosquito (Diptera: Culicidae) Aedes aegypti (Linnaeus, 1762), a method is proposed that uses abundance data to rescale MAXENT’s raw output into an estimate of the probability of presence for a species. When the area of interest is environmentally heterogeneous and occurrence records are not equally distributed, as was the case for Arizona, it can be beneficial to divide the area into regions and then mosaic the resulting probability of presence maps.