Cramer, W., Shugart, H. H., Noble, I. R., Woodward, F. I., Bugmann, H., Bondeau, A., Foley, J. A., Gardner, R. H., Lauenroth, B., Pitelka, L. F., Sala, O. & Sutherst, R. W., (1997). Ecosystem composition and structure. In: Walker, B. H., Steffen, W. L., Canadell, J. & Ingram, J.S.I. (eds.) The Terrestrial Biosphere and Global Change: Implications for Natural and Managed Ecosystems. Cambridge: Cambridge University Press.

Abstract:

• The structure of terrestrial ecosystems influences their responsiveness to most drivers of global change: for example, growth responses to enhanced CO2 are less at higher levels of organization and over longer periods of observation.

•The future structure and composition of terrestrial ecosystems will be affected by responses at the patch, landscape and global scales. Direct extrapolation from the patch to the globe is unlikely to yield realistic projections of ecosystem change; landscape-scale processes must be taken into account.

• A general finding from patch model studies is that many forests appear to be sensitive to global change on the time scale of centuries. On shorter time scales, e.g. for the next few decades, many forests will show little response due to the lag effects in demographic processes. However, in systems where intense disturbances are more common, or become more common under global change, there will be opportunities for mortality and replacement of existing trees, and changes in forest structure and composition may be more rapid.

•The interaction of global change and landscape phenomena can greatly modify both the magnitude and rate of change in community composition and structure. The importance of self-organization in landscape dynamics implies that change will not be incremental and smooth, but instead, punctuated and lumpy.

• Migration plays a critical role in the process of ecosystem adaptation to climate change; human modifications of landscapes affect the possible velocity of migration. Migration rates through the markedly non-random landscapes created by human activities are usually slower than those based on predictions derived from theoretical studies based on randomly fragmented landscapes. Many species may face a 'double bind' in which they need to migrate in response to climate change, but have few places to go and too much hostile territory to cross.

Assessment of the response of the terrestrial biosphere to global change are moving from an equilibrium towards a dynamic representation, triggered by the recognition that two major problems exist in the equilibrium assessments:

• By definition, equilibrium models simulate no transient changes in vegetation. Therefore, these simulations may at best be used to indicate the direction of possible change but not the time it might take to reach the new conditions.

• Evidence from the past shows that biomes are unlikely to be displaced as homogenous entities. Rather, differences in species' fundamental ecological niches, and their widely varying abilities to migrate, will result in quite different assemblages over a long period of time.

Future needs

There is a need to:

• gain a better understanding of physiological responses of plants to enhanced CO2 and changes in temperature and water availability, particularly since vegetation may show different adaptations over various time-scales;

• analyze further the influence of direct of indirect management on landscapes, e.g. through fragmentation or disturbance management; this is a prerequisite for more robust simulations of ecosystem sensitivities at the regional scale;

• improve global assessments of the terrestrial carbon, water and nitrogen cycles, which are necessary for improved predictions of the entire Earth system - such assessments are possible only with further improved Dynamic Global Vegetation Models, which include realistic disturbance regimes and lag times in changing ecosystem structure.