602 Latitudinal and Elevational Range Shifts under Contemporary Climate Change to be the main responses to the global increase in temperature conditions, albeit counterintuitive shifts should be expected as well at more regional scales due to regional variation in climate change and biotic interactions Types of Range Shifts There are different types of range shifts Many different terms have been used to refer to the many ways species may shift poleward or upward as climate warms, among others: leadingedge expansion, trailing-edge retraction, optimum shift, and abundance changes Recently, a catalog of the possible species range shifts observable along a given gradient has been laid out, suggesting that observed species distribution changes are rarely complete shifts, but intermediate states in an ongoing shifting process (Maggini et al., 2011) More generally, three main categories of observed range shifts can be distinguished as climate warms (Breshears et al., 2008): (1) ‘‘march’’; (2) ‘‘lean’’; and (3) ‘‘crash’’ (Figures 2(b)–(d)) • • • March-range shifts represent the most dynamic category of range shifts and involve dynamic edges with establishment from enhanced colonization and growth at the leading edge and mortality from enhanced decline and extinction at the trailing edge Such dynamics driven by the edges usually lead to expansion toward higher latitudes or elevations (leading-edge expansion) or retractions from lower latitudes or elevations (trailing-edge retraction) (Maggini et al., 2011) Therefore, out of this first category, one can develop several kinds of range shifts involving any combinations of expansion at the leading edge or retraction at the trailing edge, the most dynamic being a combination of both or march (Figure 2(b)) Short-lived, small-sized, and mobile species for which colonization and extinction processes occur on a relatively short time scale are likely to exhibit such dynamic range shifts For instance, short-lived and small-sized butterfly species not strongly limited by their dispersal abilities have been reported to shift northward at both their leading and trailing edges (Parmesan et al., 1999) Lean-range shifts involve stable edges with the optimum shifting within the existing range (Figure 2(c)) This represents the net result of enhanced growth at the highest latitudes or elevations (front edge) and decline at the lowest latitudes or elevations (rear edge) of the existing range, resulting in a shifting abundance pattern This kind of shift might be transient and might correspond to early stages of a full shifting process (Maggini et al., 2011) Longlived, large-sized, and immobile species for which colonization and extinction processes occur on a much longer time scale than the period of observation are likely to exhibit such transient stages (time lag) across their ranges For instance, long-lived and large-sized tree species, which tend to be strongly limited by their dispersal abilities, have been reported to shift their optimum elevation upward without shifting their range edges: persistence at the rear edge without establishment success beyond the front edge (Kelly and Goulden, 2008) Crash-range shifts involve stable edges and a stable optimum, but overall declines across the existing range (Figure 2(d)) Rare species having a restricted distribution for which the degree of climate warming far exceeds their ability to shift toward suitable conditions (because of dispersal limitation and natural or anthropogenic habitat fragmentation) are likely to exhibit such changes across their ranges Similarly, there is no possible escape for endemic species having a distribution restricted to a specific mountain summit and facing contemporary climate change In such situations, these threatened populations may simply crash For instance, in 1987, a multipopulation crash of the endemic golden toad of Costa Rica’s Cordillera de Tilara´n has been reported to have led to the disappearance of this species (Pounds et al., 1997) Additionally, species range shifts at the bottom of a food web may trigger trophic mismatch leading to multipopulation crash within the higher levels of the food web For instance, fluctuations in plankton have resulted in a general decline in cod biomass and recruitment throughout a bottom-up control in the North Sea (Beaugrand et al., 2003) Of course, these three categories are not mutually exclusive and can be combined to develop all possible patterns of poleward or upward shifts as climate warms (Breshears et al., 2008) Using the species response curve representation along one of the three geographic dimensions (Figure 2), the different combinations may affect the edges, optimum, level of abundance, or skewness of each species’ response curve, thus leading to many possible patterns of range shifts (Maggini et al., 2011) Contemporary Evidence for Latitudinal and Elevational Range Shifts An overwhelming number of reports of latitudinal and elevational range shifts under contemporary climate change have recently appeared in the scientific literature (IPCC, 2007a) The expected poleward and upward range shifts in warming regions have been reported for many taxonomic groups across the plant and animal kingdoms (Parmesan et al., 1999; Sturm et al., 2001; Lenoir et al., 2008; Moritz et al., 2008) A recent meta-analysis (Chen et al., 2011a) concluded that the distributions of species have recently shifted poleward at a median rate of 16.9 km per decade and upward at a median rate of 11.0 m per decade, which is two to three times faster than reported in a previous meta-analysis (Parmesan and Yohe, 2003) However, these global trends represent averages across considerable variation in range dynamics under climate change (e.g., Chen et al., 2011a) Indeed, most of the studies that have reported the expected range shifts toward higher latitudes or elevations as climate warms globally have also detected, at a smaller frequency and to a lesser extent, species moving toward lower latitudes or elevations (Parmesan et al., 1999; La Sorte and Thompson, 2007; Lenoir et al., 2008, 2010a; Moritz et al., 2008) In most cases, modern latitudinal and elevational range shifts have been documented by using and sometimes resurveying historical occurrence data such as museum collections and field notes (Tingley and Beissinger, 2009) These historically based studies can be divided into two main groups (Parmesan, 2006): (1) studies that directly assess