Uitl Kl OiT fHII

■ ttfTIOOTfl

■ ttfTIOOTfl

Figure 6.5b. Changes in vegetation composition.

types rather than a single dominant category.The major conclusions of the study are: (1) there are likely to be significant changes in the composition of U.S. vegetation under all four assumed scenarios of climate change; (2) warmer and drier climates are likely to cause reduction in the prevalence of many coniferous species (e.g., spruce, fir, larch, etc.) and deciduous species (maple, birch, etc.); (3)

Figure 6.6. Changes in vegetation prevalence under GISS and GFDL climate change scenarios.

warmer and drier climates may substantially increase prevalence of shrubs and grass, particularly in the Western United States; and (4) wetter climates without significant warming could increase prevalence of pine species in the Southeast.

This probabilistic modeling approach to biogeography has a number of weaknesses. It does not consider factors such as elevated CO2, nitrogen deposition, and changes in the disturbance regimes (e.g., fires). Although it captures the nonlinear relationship between vegetation prevalence and climate, it does not employ a process-based, mechanistic explanation of this relationship. It uses information from climate models, which operate on a scale much larger than the scale of a plant resource allocation or the scale of population or community dynamics. Several different strategies have been proposed to address the problem of scaling. Root and Schneider (1995) proposed a "strategic cyclical scaling" to model climate-ecosystem interactions: Given a model parameterization being scale dependent, one must cycle back and forth across the scales testing mechanistic constructions at a large scale and then cycle back to see if the processes are adequately represented. Hurtt et al. (1998) advocate development of biological models that are formulated at an intermediate scale of biological detail and are tested at different spatial and temporal scales.

Most important, as in many other biogeography models, the probabilistic model relies on the assumption that the current distribution of vegetation is in equilibrium with the current climate. The dynamics and end points of the changes in vegetation composition will depend on the rate of climate change and the degree of fragmentation in land cover as well as on nonlinear responses of individual components of ecosystems. For example, different taxa— insects, birds, and animals—can have differential transient responses in response to changes in climate and/or disturbance regime. Communities of species may become disaggregated during such transitions and may have responded in unexpected ways to their environment and climate (Root and Schneider 1993). Changes in land practices due to the growing population will further exacerbate impoverishment of the biosphere. While the assumption of equilibrium is clearly not true in an absolute sense, the errors due to current disequilibrium are presumed to be small enough that the predictions can still serve as a useful starting point for assessing where and how vegetation composition can change.

0 0

Post a comment