Disturbance and Tree Life History on the Shifting Mosaic Landscape

An analytical model of disturbance and plant population dynamics is developed to explore the optimal life history for a plant within a "shifting mosaic" meta—population. The population dynamics consist of short—lived recruitment events followed by longer intervals of thinning. Plants balance costs and benefits of delayed maturation time that result from cohort thinning, a correlation between maturation time and longevity, and the distribution of recruitment events in space and time. Two fundamentally different responses to disturbance are explored: (1) the plant is killed by the disturbance that allows for new recruitment (type A response), and (2) the plant may survive many such disturbances (type B response). Species maximize either the probability of being reproductively mature at the time of the next recruitment opportunity (type A) or the total number of recruitment opportunities to occur during the period of reproductive maturity (type B). Predictions of the theory are compared with the actual life histories of trees that occur in different disturbance regimes. The costs and benefits associated with delayed maturation from an energy standpoint must be weighed against the probability that a recruitment opportunity (disturbance) will occur at a particular age. Trees subject to low thinning rates should reach reproductive maturity t 1 at t 1 ° 0.4 x (expected disturbance interval in years). At high thinning rates, this optimum is t 1 ° 0.4/(mortality rate per year). Disturbance probabilities that increase with time since the last disturbance select for maturation times that are greater than these values. Species that are not killed by disturbances have optimal maturation times that are independent of disturbance frequency. However, when such species are susceptible as juveniles, optimal maturation time does depend on disturbance frequency. This optimum maturation time is still greater than it is for the case of a mortality response (type A) to disturbance, but less than the case of no susceptibility period (type B). Application of the theory to real—world disturbance regimes results in predictions that closely match the life histories of species that actually occur there. The optimal maturation time for a gap species in temperate North American forests is 30—60 yr, a value that agrees with observed maturation time. A second test involved fire regimes where two species having very different responses to fire and life histories co—occur, Pinus resinosa and P. banksiana. The maturation times of these species both match the predicted optima for a species that survives fire (P. resinosa) vs. one that is killed by fire (P. banksiana) subjected to identical fire regimes. Different modes of dispersal are predicted to have important effects on reproductive potential, but little effect on the optimal maturation time. Application of the models to these actual cases is consistent with that prediction. The "intermediate" disturbance is predicted to be that which implies the optimum life history that coincides with the life histories of the greatest number of species.