Chown 2003). However, CTmin declines to a greater extent with altitude than does CTmax (Fig. 5.25). Indeed, CTmax shows a slight, but non-significant, increase with altitude. Thus, even at the intraspe-cific level, and over a rather small spatial scale, there is some evidence that upper thermal limits are less variable than lower ones. However, this variation appears to be entirely a consequence of phenotypic plasticity. Investigations of acclimation that were undertaken as part of the same study indicated that treatment temperature has a pronounced effect on short term (seven days) acclimation of the CTmin (indeed, the acclimation effect accounts for all of the altitudinal variation), although there is no effect on CTmax.
Such decoupling between upper and lower lethal limits has been documented in several other studies, mostly investigating interspecific differences (e.g. Chen et al. 1990; Kingsolver and Huey 1998; Hercus et al. 2000, but see Hoffmann et al. 2002), and has also been documented in the responses of Drosophila species to selection (Gilchrist et al. 1997; Hoffmann et al. 1997). The apparent decoupling of upper and lower lethal limits, and the hemispheric asymmetry in both tolerance limits and cold hardiness strategies have several implications for macroecological and climate change investigations, and also have bearing on widely held ideas concerning correlated responses to stress.
1. Because of considerable inter-annual and seasonal variability in low temperatures, and a relatively shallow latitudinal gradient in low temperature extremes in the Southern Hemisphere, and because low temperatures seem more likely to be geographically limiting than high temperatures (Kukal et al. 1991; Jenkins and Hoffmann 1999; Chown 2001), physiological range limitation might be less stringent in the Southern than in the Northern Hemisphere. In consequence, geographic range size gradients of southern ectotherms should be much less steep than those of northern species. There should also be a less pronounced shift in range sizes near the Tropic of Capricorn than at the Tropic of Cancer because the interplay between the geographic and climatic determinants of range size should be less pronounced (Gaston and Chown 1999c). Moreover, if 'zonal bleeding' is important as a factor leading to the establishment of species richness gradients (Chown and Gaston 2000), then species richness gradients should be less steep in the Southern than in the Northern Hemisphere. Of course, these predictions might also apply to vertebrates, and indeed in these groups the evidence is suggestive (Gaston et al. 1998).
2. The information on tolerances and climatic variation supports the idea that towards higher latitudes annual climatic variation increases and that, at least in some species, this also true of the thermal tolerance range (Addo-Bediako et al. 2000). Thus, a major assumption of Rapoport's Rule appears to hold, although whether this translates into larger geographic ranges at high latitudes for these insect species is not well known because coupled data on tolerance limits and latitudinal ranges are rare (see Gaston and Chown 1999a for an altitudinal example).
3. If absolute temperatures are less important in constraining the ranges of Southern than Northern Hemisphere species, and opportunistic utilization of warm spells is the norm in the former hemisphere, then climate change might be expected to have much less of a direct effect on high latitude insects in the Southern than in the Northern Hemisphere. While there is evidence for pronounced recent changes in the geographic ranges of Northern Hemisphere species (Parmesan et al. 1999), information for the south is scanty.
4. Although a generalized stress response, involving cross-tolerance to several stresses may be present at the biochemical level in insects (Hoffmann and Parsons 1991), this general response does not appear to hold for higher levels of organization, as decoupling between upper and lower lethal temperatures (Huey and Kingsolver 1993) and differences in measures of heat resistance (Section 5.1.2) clearly demonstrate.
Clearly, there is much to be gained from understanding the relationship between upper and lower lethal limits at several scales. At the cellular level, these limits might either arise from problems associated with oxygen delivery and demand (Portner 2001), or with the effects of temperature on neuronal functioning and protein folding (Feder and Hofmann 1999; Hosler et al. 2000; Wu et al. 2002), and the effects might differ at either end of the temperature spectrum. In turn, these effects will not only influence whole-organismal tolerances, but potentially also the distribution of species. These interactions are explored further in Chapter 7.
Was this article helpful?