Stochasticity Abundance Latitudinal and Elevational Range Shifts under Contemporary Climate Change Latitude/altitude March Abundance (a) Latitude/altitude Lean Abundance (b) Latitude/altitude Crash Abundance (c) (d) Latitude/altitude 601 locations within the peripheral area of a species’ range that were too cold and therefore less suitable might become more suitable in a warmer climate, thus turning sink populations into source populations with higher growth and colonization rates On the other hand, locations within the core area of a species’ range that were highly suitable might become too warm and therefore less suitable, turning source populations into sink populations with higher decline and extinction rates Across species ranges, the net result of these changes in growth, decline, colonization, and extinction processes will affect their geographic distribution in different ways However, an increase in temperature is likely to have an overall directional impact on species range shifts, because temperatures are autocorrelated in space, linking warmer conditions at lower latitudes and elevations with cooler conditions at higher latitudes and elevations Therefore, one expects to observe poleward and upward range shifts as climate warms, even after accounting for dispersal limitations (Engler et al., 2009) or biotic interactions (Arau´jo and Luoto, 2007) Accordingly, poleward and upward range shifts in the warming climates following the glacial recessions of the past interglacial periods have been widely reported for plants, birds, and mammals (Brown and Lomolino, 1998) For instance, Pleistocene fossils of several species of rodents have been found several thousand kilometers southward of the southern limit of their modern distribution in northern America (Graham, 1986), suggesting strong poleward shifts during the Holocene Similarly, Pleistocene macrofossils of Podocarpus have been found c 1000 m below the lower limit of their contemporary distribution on the Andean flank in western Amazonia (Ca´rdenas et al., 2011), thus suggesting a large upward shift during the current interglacial period Although the expectations given temperature increase alone are relatively straightforward, it is difficult to predict how concurrent changes in other climatic factors, especially precipitation, will affect species ranges Temperatures are negatively correlated along the latitudinal and elevational gradients and have risen globally over the past decades (IPCC, 2007b) In contrast, precipitation changes are more heterogeneous across space and time (IPCC, 2007b), leading to strong regional differences Such regional variation in climate change, mainly due to the effect of precipitation changes on the water balance equation (balance between evapotranspiration and precipitation), may affect species ranges as well, leading to unexpected regional range shifts as climate warms globally (Crimmins et al., 2011) Additionally, biotic interactions may also affect the magnitude of species range shifts as climate warms (Arau´jo and Luoto, 2007) with some suggestions that interactions could explain unexpected range shifts as well (Lenoir et al., 2010a) Therefore, at global to continental scales, poleward and upward range shifts are likely Figure Conceptual representations of latitudinal and elevational range shifts and their mechanics under a steady-state environment (a) and under climate warming (b)–(d) In each case, the figure shows the relative importance of growth, decline, colonization, and extinction processes across the species range For the march-, lean-, and crash-range shifts, gray shades represent conditions before a climate warming event whereas overlaying transparent colors represent shifting conditions under a climate warming event