If climate change results in wholesale replacement of one forest community with another, effects on ecosystem functions ranging from carbon storage to wildlife habitat could be dramatic (Tilman, 1998). Possible wholesale changes in forest structure have been a great source of uncertainty and concern. The North America assessment in the IPCC's Special Report on Regional Impacts of Climate Change concluded that there were equal probabilities of considerable forest dieback and enhanced forest growth, given state-of-the-art models and GCM predictions (Shriner and Street, 1998).
Many types of models (biogeographic individual-based forest growth, gap, dynamic global vegetation, regression tree analysis, response surface, richness, and rare and endangered species) have been used with numerous climate scenarios to examine broad-scale climate change-induced changes in vegetation (VEMAP Members, 1995; Shugart and Smith, 1996; Aber et al., 2001). There is great uncertainty about the precision and accuracy of each of these models and the assumptions that underlie them (Loehle and LeBlanc, 1996; Repo et al., 1996; Beuker et al., 1998). Many studies suggest that all major forest types in North America will expand northward and most will increase in extent in the next 50-100 years. The increase in forests is predicted to be driven by slight warming coupled with increases in water-use efficiency (WUE) associated with increased atmospheric CO2 (Saxe et al., 1998). However, with continued warming, increased water use associated with higher temperatures overwhelms the CO2 effect, resulting in potentially important decreases in forest area (Aber et al., 2001).
Wholesale changes in forest ecosystem structure over time are likely to be
mediated by changes in disturbance regimes and/or catastrophic events that provide
opportunities for forest regeneration over large areas (Suffling, 1995; Loehle
and LeBlanc, 1996). Of particular interest in North America are changes in fire
and insect outbreaks.
Stocks et al. (1998) used outputs from four GCMs to project forest fire danger levels in Canada under a warmer climate. Their analysis shows an earlier start to the fire season and significant increases in the area experiencing high to extreme fire danger. Increased lightning frequency associated with global warming also may increase fire frequency (Goldammer and Price, 1998). Changes in fire frequency have a wide range of effects, from production of tropospheric aerosols that influence climate (Clark et al., 1996) to changes in ecosystem carbon storage and trace gas fluxes. The long-term effects of fire will depend heavily on changes in human fire management activities, which are uncertainespecially in remote boreal forests, where fire is a critical issue (e.g., see Section 126.96.36.199). In mid-latitudes, climate effects on fire frequency are much less important than human management factors (Veblen et al., 2000).
Insects represent dominating disturbance factors (Hall and Moody, 1994) in
North America's forests; during outbreaks, trees often are killed over
vast areas (Hardy et al., 1986; Candau et al., 1998). Because the potential
for wildfire often increases in stands after insect attack (Stocks, 1987; Wein,
1990), uncertainties in future insect damage patterns also lead to uncertainties
in fire regimes. Insect outbreaks also lead to changes in ecosystem carbon and
nutrient cycling, biomass decomposition, energy flow (Mattson and Addy, 1975;
Schowalter et al., 1986; Szujecki, 1987; Haukioja et al., 1988; Chapin, 1993;
Haack and Byler, 1993), and competitive relationships between plants (Morris,
1963; Holling, 1992)hence successional pathways, species composition,
and forest distribution.
Climate change already appears to be accelerating the seasonal development
of some insects (Fleming and Tatchell, 1994). Forecasts based on historical
relationships between outbreak patterns and climate in specific areas are likely
to predict change as the climate in those areas changes. Such analyses suggest
more frequent [mountain pine beetle (Thomson and Shrimpton, 1984), spruce budworm
(Mattson and Haack, 1987), eastern hemlock looper (Carroll et al., 1995), jack
pine budworm (Volney and McCullough, 1994), western spruce budworm (Thomson
et al., 1984)] or longer [forest tent caterpillar (Roland et al., 1998), spruce
budworm (Cerezke and Volney, 1995)] outbreaks or range shifts northward and
to higher elevations [spruce budworm (Williams and Liebhold, 1997)] as climate
Vegetation in human settlements plays two potentially important roles that
are relevant to climate change: modification of local climate and sequestration
Trees reduce demands for seasonal heating and cooling of the interiors of buildings
and absorb air pollutants (Heisler, 1986; Nowak et al., 1994). In the city of
Chicago, calculations suggest that increasing tree cover by 10% could reduce
building energy use by 5-10% (Nowak et al., 1994). Thus, urban tree planting
represents an adaptive strategy that can reduce energy use and associated CO2
emissions and counteract temperature increases in urban areas, which are predicted
to be extreme in some cases.
Carbon density in residential areas can be significant, amenable to management, and often overlooked in evaluation of landscape, regional, and national carbon budgets. Freedman et al. (1996) found that aboveground tree biomass of an old residential neighborhood in Halifax, Nova Scotia, was only slightly smaller than that of a natural forest in a nearby reserve. Nowak et al. (1994) estimated that carbon storage by urban forests in the United States was 440-990 Mt. There is a strong need for better estimates of "natural" carbon fluxes in human-dominated environments.
Other reports in this collection