Increasing temperatures are driving rapid upward range shifts of species in mountains. An altitudinal range retreat of 10 m is predicted to translate into a $10-km latitudinal retreat based on the rate at which temperatures decline with increasing altitude and latitude, yet reports of latitudinal range retractions are sparse. Here, we examine potential climatic, biological, anthropogenic and methodological explanations for this disparity. We argue that the lack of reported latitudinal range retractions stems more from a lack of research effort, compounded by methodological difficulties, rather than from their absence. Given the predicted negative impacts of increasing temperatures on wide areas of the latitudinal distributions of species, the investigation of range retractions should become a priority in biogeographical research.Climate change and shifting plant distribution The geographical distribution of plants is strongly influenced by climate [1]. Minimum temperatures are particularly important in limiting the poleward (see Glossary) expansion of plant species, whereas limited water availability interacts with high temperatures to exert a direct climatic limitation on their expansion in the opposite, or equatorial, direction in many regions [1][2][3][4][5][6][7]. Changes in climate are, therefore, predicted to alter the geographic distribution of plant species at global to local scales.Contemporary plant range shifts are most frequently reported from mountain regions, with elevational shifts of the mountain treeline being the most commonly documented response to increasing temperatures [8][9][10][11][12]. The distributions of species in mountain regions are typically restricted to relatively narrow and well-delineated altitudinal bands, in comparison with often broad and poorly defined latitudinal distributions in the lowlands. This strong altitudinal zonation of mountain vegetation, and its climatic sensitivity [13], makes mountains ideal 'natural laboratories' and favoured locations for the study of climate change impacts worldwide [13][14][15].The altitudinal compression of montane vegetation zones is due to a rapid decrease in temperature with increasing elevation ($5-6.5 8C per 1000 m) [3,14], which contrasts with a similar temperature decline occurring over $1000 km of latitude (6.9 8C at 458 N or S) [3]. An implication of this altitude-for-latitude temperature model is that general biogeographical patterns resulting from the climatic limitations imposed on plants in mountains should also be observed in neighbouring lowland regions with similar temperature and moisture regimes. Distributional shifts already observed in mountain plants [8,[16][17][18] should, therefore, be mirrored by lowland range changes occurring over distances that are several orders of magnitude larger (Figure 1).Here, we discuss the implications of the altitude-forlatitude model for shifts in plant distributions and examine why lowland range shifts, particularly range retractions, are rarely reported. We discuss potential climati...