Environmental correlates of tree species distributions vary among age classes in a northern temperate forest

Processes involved in the structuring of forest communities include: (1) ecological sorting, where species poorly suited to local conditions are subject to environmental filtering and competitive displacement; (2) disturbance, resulting in stochastic removal of individuals and reinitiating successional regimes and (3) dispersal limitation, inhibiting the infiltration of species into preferred sites. Temporal dynamics in these processes lead to difficulty inferring causal landscape–biota correlations. Complicating factors include potential for ontogenetic variation in habitat preferences among age classes, and inherent ambiguity regarding severity and coverage of historical disturbance events. Sorting species into age groups can provide relevant temporal information. Fundy National Park is a northern, mixed-temperate forest in Atlantic Canada (Acadian forest type), which was pervasively altered upon European settlement. Species frequency data for three tree age classes (saplings, juveniles and adults) in permanent sample plots (400 m2, n = 33) were compared to environmental data, including soil chemistry, understory light conditions, physiography and disturbance history using ordination and randomization techniques. Abiotic and disturbance-related predictors of species distributions differed among life stages. Specifically, in the adult stage, stand age was a critical predictor of distribution, whereas in younger age classes environmental variables such as nutrient availability and soil moisture and drainage were key drivers of distribution. It is concluded that older populations were increasingly less constrained by environmental conditions, suggesting that adult populations bear the legacy of stochastic landscape alteration, thus appearing randomly distributed along environmental gradients. As younger populations gradually expand in distribution, they are filtered into preferred conditions by ecological sorting. These findings indicate the importance of considering age-class effects and site history in further assessments of interactions between landscapes and flora.