Understanding whether a species’ ecological niche is stable or not through time is important for interpreting the paleoecology and evolution of extinct organisms and especially for providing insights into how extant species will respond to ongoing global change. Here, I examine the ecological niches of 15 mammalian species (from Eulipotyphla, Lagomorpha, Rodentia, Carnivora and Artiodactyla) during the last glacial maximum (LGM) to their present ecological niches in order to test for niche stability through time. I focus on “environmental niche space” as opposed to traditional studies that are more concerned with modeling species distributions in geographic space. This approach allows differentiating between the environmental space, potential niche space, and realized niche space. The two key niche axes defining the niche space are mean annual temperature and mean annual precipitation. Occurrence data are from the FAUNMAP and GBIF databases. In order to make data comparable from the fossil record and the present record, I standardized sample sizes by subsampling the present realized niche space 100 times, setting sample size equal to that which defined the fossil realized niche space. Analyses performed to test for statistical power concluded that fossil sample sizes are adequate to test for niche stability and niche shifts through time.

Only two of the species (Sylvilagus audobonii, Thomomys bottae) I examined had LGM realized niche space within their present day realized niche space, suggesting niche stability in those cases. However, the majority of the species had LGM realized niche space either completely or partially outside their present day niche space suggesting niche shifts are common. These differences cannot be accounted for by the lack of available environmental space between the two time periods, further suggesting that the realized niche shifted for these species. Modeling that assumes niche similarity through time needs to better account for the reality that some species will occupy shifting or different environments through time, and comparisons of fossil and recent ranges is an effective way to identify those taxa.