Model of Transient Changes in Arctic and Boreal Vegetation in Response to Climate and Land Use Change

Abstract

One of the greatest challenges in global‐change research is to predict the future distribution of vegetation. Most models of vegetation change predict either the response of a patch of present vegetation to climatic change or the future equilibrium distribution of vegetation based on the present relationship between climate and vegetation. Here we present a model that is, to our knowledge, the first model of ecosystem change in response to transient changes in climate, disturbance regime, and recruitment over the next 50‐500 yr. The frame‐based model uses quantitative and qualitative variables to develop scenarios of vegetation change from arctic tundra to boreal forest in response to global changes in climate (as predicted by general circulation models [GCMs]), fire, and land use. Seed availability, tree growth rate, and probability of fire were the model parameters that most strongly influenced the balance between tundra and boreal forest in transitional climates. The rate of climatic warming strongly affected the time lag between the onset of climate change and the simulated ecosystem response but had relatively little effect on the rate or pattern of ecosystem change. The model calculated that, with a gradual ramped change of 3°C in the next century (corresponding to average rate of warming predicted by GCMs), any change from tundra to forest would take 150 yr, consistent with pollen records. The model suggested that tundra would first be invaded by conifer forests, but that the proportion of broad‐leaved deciduous forest would increase, reflecting increased fire frequency, as climatic warming continued. The change in fire frequency was determined more strongly by climatically driven changes in vegetation than by direct climatic effects on fire probability. The pattern of climatic warming was more important than the rate of warming or change in precipitation in determining the rate of conversion from tundra to forest. Increased climatic variability promoted ecosystem change, particularly when oscillations were long relative to the time required for tree maturation. Management policies related to logging and moose‐predator control affected vegetation as much or more than did changes in climate and must be included in future scenarios of global changes in ecosystem distribution. We suggest that frame‐based models provide a critical link between patch and equilibrium models in predicting ecosystem change in response to transient changes in climate over the coming decades to centuries.