Faunal turnover in Indo-Australia across Wallace’s Line is one of the most recognizable patterns in biogeography and has catalyzed debate about the role of evolutionary and geoclimatic history in biotic interchanges. Here, analysis of over 20,000 vertebrate species with a model of geoclimate and biological diversification shows that broad precipitation tolerance and dispersal ability were key for exchange across the deep-time precipitation gradient spanning the region. Sundanian (Southeast Asian) lineages evolved in a climate similar to the humid ‘stepping stones’ of Wallacea, facilitating colonization of the Sahulian (Australian) continental shelf. In contrast, Sahulian lineages predominantly evolved in drier conditions, hampering establishment in Sunda, and shaping faunal distinctiveness. We demonstrate how the history of adaptation to past environmental conditions shapes asymmetrical colonization and global biogeographic structure. Faunal turnover in Indo-Australia across Wallace’s Line is one of the most recognizable patterns in biogeography and has catalyzed debate about the role of evolutionary and geoclimatic history in biotic interchanges. Here, analysis of over 20,000 vertebrate species with a model of geoclimate and biological diversification shows that broad precipitation tolerance and dispersal ability were key for exchange across the deep-time precipitation gradient spanning the region. Sundanian (Southeast Asian) lineages evolved in a climate similar to the humid ‘stepping stones’ of Wallacea, facilitating colonization of the Sahulian (Australian) continental shelf. In contrast, Sahulian lineages predominantly evolved in drier conditions, hampering establishment in Sunda, and shaping faunal distinctiveness. We demonstrate how the history of adaptation to past environmental conditions shapes asymmetrical colonization and global biogeographic structure.

All data was collated from published sources.

Spatial and phylogenetic data for terrestrial vertebrates

We obtained spatial and phylogenetic information for all 488 taxonomic families of birds, mammals, amphibians, and squamate reptiles, composed of 32,645 species. We obtained matching data on the geographic distribution and phylogenetic position of extant species of terrestrial vertebrates collected through the VertLife project (vertlife.org) in association with Map of Life (mol.org). We subset this data to include only families found in Sunda or Sahul. This data set contained matching phylogenetic and spatial data for 227 families composed of 20,433 species, with 7216 species of birds, 3551 amphibians, 6106 reptiles, and 3560 mammals.

Paleoenvironmental reconstruction

Our reconstruction of paleogeographic evolution is based on maps from Hall 2013. From that study, we use the maps from 30 Ma to the present-day, which are presented at 5 Ma intervals. From these maps, we developed a continuous reconstruction of paleogeography in 0.1 Ma timesteps, incorporating plate tectonic motion and the spatial distribution of ocean, land, and highlands. We reconstructed paleoenvironmental temperature from the reconstructed distribution of Köppen climate bands (Scotese et al 2021), corrected for elevation. We obtained reconstructed Köppen climate bands at 5 Ma intervals from 35 Ma to 0 Ma. These broad climate bands are estimated from different fossil and lithological indicators and give a broad categorization of spatial variation in climate type (tropical, sub-tropical arid, warm temperate, cool temperate, and polar) and approximate the steepness of the latitudinal gradient in temperature through time. We assigned each Köppen climate band a temperature value based on present-day averages of those climate bands (tropical=26, sub-tropical arid=22, warm temperate = 16, cool temperate = 5, polar=-20). To account for global fluctuations in temperature, we then modify these average values by a global temperature change curve based on oxygen isotope data. Finally, to account for the decrease in temperature with increasing elevation, we use a lapse rate of 9.8 °C/km. We model variation in the intensity of temperature oscillations during the Pleistocene. We derived mean daily precipitation data from the global circulation model of paleoclimate from. This paleoclimatic data is based on a HadCM3L model, which couples ocean, atmosphere and vegetation and was simulated at 5 Ma intervals.

Data_S9.rds requires the R software to open: data <- readRDS("Data_S9.rds")