ABSTRACT
Recent changes in forest management goals have led to a diversification of silvicultural practices to maintain complex, mixed-species stands. Managing such stands requires an understanding of competitive interactions among individual trees. In this project, we used an individual-based, spatially explicit "neighbourhood" analysis to develop new models of crown radius and crown length for subalpine fir (Abies lasiocarpa), lodgepole pine (Pinus contorta), and interior spruce (Picea glauca x engelmanii). We parameterized our models using spatially explicit crown measurement data from ~2400 trees collected in the subboreal spruce forest near Smithers, in north-central British Columbia.
Using maximum likelihood methods, we quantified crown radius and length for each species as a function of tree size and neighbourhood competition, estimated by the identity and spatial arrangement of neighbouring trees. In the absence of competition, spruce and fir showed similar patterns of increasing crown radius and length with tree dbh and height respectively. Compared to fir and spruce, pines growing in the absence of competition were associated with wider, shallower crowns.
The effect of competition on crown dimensions was strong. All species showed declines in predicted crown size with increasing neighbourhood competition, an index that varied with the abundance, size, distance, and species identity of neighbouring trees. The best models fit separate competition coefficients for each species of neighbouring trees, indicating that there are important differences among species in their competitive effects on tree crowns. The strongest competitors for fir and pine were conspecifics, while spruce was most sensitive to competition from fir neighbours. Although there was no single means to assess sensitivity to competition, pine was generally the most sensitive species to increasing levels of crowding in monospecific stands, while in mixed-species stands, spruce was most sensitive to crowding.
We used our results to parameterize a new, individual-based, distance-dependent crown model for the stand simulation model SORTIE-ND. Including a competition-dependent canopy model in SORTIE-ND is likely to improve the model's accuracy in predicting dynamics of mixed-species stands, understorey light availability, and sapling and sub-canopy tree dynamics.
