A model of the post-fire recruitment of Picea mariana and Pinus banksiana as a function of salvage timing and intensity

In this paper, we model the post-fire recruitment dynamics of two aerial seedbank species, Picea mariana and Pinus banksiana, in response to salvage logging. The model incorporates: (1) initial seed availability as a function of source tree basal area and proportion of stand salvaged; (2) seed abscission as a function of time; (3) seedling survivorship as a function of seed mass, seedbed proportion, and granivory; and (4) seedling and seed mortality as a function of salvage operations. We also elaborate a simulation of the effect of direct seeding via cone-bearing branches fed into a moving chipper. The model performed adequately when tested against data sets from two fires in Quebec and one in Saskatchewan. In particular, it showed that P. mariana was more adversely affected by early salvage than P. banksiana because of its far slower seed abscission rate. The model predicted that a delay in salvage or a decrease in salvage proportion would enhance tree regeneration densities, especially for P. mariana. Finally, model projections indicate that the use of a chipper to disseminate seeds during the harvesting would permit either species to be adequately regenerated cheaply even with low pre-fire basal area per area or very early salvage.