The Regional Impacts of Climate Change

	Other reports in this collection

(continued...)

Thus, two contrasting scenarios of the North American forest future must be considered: one with considerable forest dieback, another with much enhanced forest growth. These contrasting scenarios represent endpoints on a spectrum of possible responses. In general, however, the enhanced-growth scenarios occur under the least amount of simulated global warming, whereas the severe decline or dieback scenarios occur under the greatest projected global warming. With the incorporation of direct CO₂ effects, small temperature increases can produce increased growth, but larger temperature increases still produce declines. Without a direct CO₂ effect, forest decline simulations are far more widespread, even under the least warming scenarios. These results suggest the possibility that early forest responses to global warming could exhibit enhanced growth; later stages could produce widespread decline or dieback. Most combinations of scenarios and CO₂ effects produce intermediate scenarios, with a regional mosaic of forest dieback and enhanced forest growth. When coupled with economic models, these internally consistent but potentially opposite regional responses provide the basis for regional, national, and globally integrated assessments. Also, it is not clear that greenhouse gases will stabilize at the equivalent of 2xCO₂ forcing; they could increase to 3xCO₂ or 4xCO₂ (IPCC 1996, WG I, Section 2.1.3).

Forests cannot move across the land surface as rapidly as the climate can. The faster the rate of climate change, the greater the probability of ecosystem disruption and species extinction.

Were temperature-induced drought dieback to occur, it likely would begin shortly after observable warming; if accompanied by short-term precipitation deficits, it could occur very rapidly (Solomon, 1986; King and Neilson, 1992; Martin, 1992; Smith and Shugart, 1993; Vose et al., 1993; Elliot and Swank, 1994; Auclair et al., 1996; Martin, 1996). That is, dieback could begin within a few decades from the present and might include potential increases in secondary impacts from pests and fire. Alternatively, forest growth might increase in the early stages of global warming, only to revert to widespread and rapid drought-induced forest dieback after higher temperatures have significantly increased evaporative demand. Vegetation change in areas of enhanced growth, especially previously unforested areas, would be more gradual (decades to hundreds of years), constrained by dispersal, establishment, and competition.

Under global warming the physical and biotic components of most animal habitats will likely change at different rates (Davis, 1986; Dobson et al., 1989; Malcolm and Markham, 1996; Markham, 1996). The faster the rate of change, the greater the disequilibrium between physical and biotic habitat components and the higher the probability of substantial ecosystem disruption and species extinctions (Malcolm and Markham, 1996; Markham, 1996). However, species will respond differently than biomes (Neilson, 1993a,b; Lenihan and Neilson, 1995). The relative mixtures of species in forest communities will change-and under either forest expansion or contraction, some important species could be at risk.

Forest ecosystems are expected to shift northward and upward in altitude, but expansion may be limited by dispersal and poor soils.

All three major forest types within North America expand north and forested areas, with a few exceptions, increase under all scenarios with or without a direct CO₂ effect (biogeography models of potential natural forests under equilibrium conditions). Total forest area increases by as much as 25-32% as projected under the FAR 2xCO₂ GCM scenarios (including a direct, physiological CO₂ effect)-much less than the 42-44% under the SAR scenarios with a direct physiological CO₂ effect. However, the projected forest-area increases under the SAR scenarios are reduced to 21-24% when a direct physiological CO₂ effect is not included. In the long term, more carbon would be sequestered by forests under these scenarios. Before equilibrium conditions are reached, however, the processes of forest redistribution could cause a temporary reduction in forest area and a carbon pulse to the atmosphere.

Boreal forests displace most of the taiga/tundra region and increase in area under the SAR scenarios but are projected to increase or decrease under the FAR scenarios (Table 8-2). It has been projected that the temperate evergreen forests shift northward into Canada and Alaska and expand under the climate projected by the FAR scenarios (Annex C, Figures C-2 to C-5; Table 8-2). Temperate evergreen forests may expand or contract in area-due in part to conversion from conifers to broadleaf deciduous forests and in part to severe forest dieback under some scenarios (Annex C, Figures C-2 to C-5; Table 8-2; VEMAP Members, 1995). The temperate mixed forest is projected to invade the boreal forest to the north and experience gains in area under all simulations. Smaller gains in area of both temperate forest types occur under more xeric scenarios; forest expansions to the north are balanced by forest dieback in the southern zones. Because dieback in the southern zones might occur more rapidly than northward advances, there could be a short-term reduction in the area of important temperate and boreal forests (King and Neilson, 1992; Smith and Shugart, 1993).

(continues on next page...)