Recent years have seen a broadening in the consideration of climate change impacts on ecosystems. There has been upsurge in studies of ecosystem structure and behavior and in the application of the results of such studies to considerations of the response of ecosystems to a climate change [ Field et al., 1992; Peters and Lovejoy, 1992; Smith et al., 1992]. Bonen [1993] and Malanson [1993] have listed and described the variety of approaches to ecosystems modeling. In the application of advances in understanding ecosystems there have been a number of specialized studies; some have considered climate change variables other than temperature and precipitation. O'Brien et al. [1992], for example, considered the impact on tropical forests of changes in hurricane frequency or intensity.
Forests represent the largest component of terrestrial biomass, and as such are the most frequent subject of climate change/ecosystem studies. The relation of forests to a changing climate is complex, in the first place it is a two-way street. A changing climate can damage forests in several ways; meanwhile forest changes can modify the climate by placing the carbon formerly stored in the forest biomass and soil into the atmosphere increasing climate forcing, by changing the surface albedo and so modifying the local or regional heat balances, by changing surface roughness and hence momentum exchanges with the atmosphere, or by modifying the local or regional latent heat exchange or hydrological cycle and through it the climate [ Botkin et al., 1992; Bonen et al., 1992]. An improved understanding of forest ecology or physiology, achieved either for improving estimates of the influence of forests on climate or the other way around is a contribution to climate impact studies [ Neilson, 1993].
Much of the progress in understanding possible forest response to climate change has come from pollen studies of the species composition of paleoforests in times of changing climate, especially during the retreat of the last glaciation in North America [ Webb, 1992]. The inferred migration rates of individual species differ, giving rise to concerns that a more rapid climate change could tear apart ecosystems and thereby reduce their productivity or viability. The inferred migration rates are also slower, for most species, than that required to keep up with the projected movement of climate zones in the future.
A number of forest models have been developed to apply to the problem of climate change impacts [ McGuire et al., 1993; Shugart et al., 1992]. Recent improvements in forest models include the addition of needed complexity to the representation of tree growth and the examination of the sensitivity of this improved model to parameter errors [ Botkin and Nisbet, 1992], and the use of tree ring evidence of response of individual species to current climate as corrections to model assumptions [ Cook and Cole, 1991].
As is the case with agricultural production, changes in the pest prevalence caused by climate change must be considered in addition to the direct impacts [ Franklin et al, 1991]. Tree ring studies have found correlations of pest outbreaks with past shifts in climate [ Swetnam and Lynch, 1993].
Freshwater ecosystems may be stressed by changes in water temperature or streamflow [ Covich, 1993], and the complex behavior of freshwater taxa must be considered in the consideration of impacts [ Irons et al., 1993; Chang et al., 1992].