Temporally-explicit models of fire and forest

Abstract

Pollen and charcoal data from varved sediments of Lake Ahvenainen, southern Finland, were examined through ordination analyses to determine temporal groupings of vegetation. The ordination `continually' changed throughout the pollen profile, indicating nonequilibrium system dynamics. However, temporal associations that were stationary existed during the history of vegetation change at Lake Ahvenainen, and these temporal associations were nested in a hierarchy. Weighted averaging regression and calibration were used to model the relationship between fire and vegetation within each temporal association. Fire was significantly related to the vegetation for nearly all temporal associations. Community dynamics within temporal associations were examined with cross-correlation analysis. Taxa were correlated with fire occurrence (charcoal fragments) within the temporal associations and classic successional recovery was confirmed. In addition, two temporal associations showed evidence of chaotic behavior. The long-term pollen-charcoal analysis supported three theories of system dynamics: (1) nonequilibrium long-term dynamics, (2) hierarchical structuring of temporally stationary associations, and (3) chaotic dynamics in stationary associations. These theories are not necessarily contradictory, but are scale dependent.