Survival and growth of balsam fir seedlings and saplings under multiple controlled ungulate densities

Tree species composition in forests can be strongly modulated by high densities of cervid herbivores ultimately leading to local extirpation of species. To establish which cervid densities are compatible with the recruitment of a browse sensitive tree species, seedlings and saplings should be surveyed under variable cervid densities rather than in their presence or absence alone. We studied the growth and survival of different demographic stages of balsam fir (Abies balsamea) on Anticosti island (Québec, Canada) under controlled densities of white-tailed deer (Odocoileus virginianus). In a seven-year experiment using deer enclosures, we followed the life stage and fate of individually tagged balsam fir seedlings/saplings under forest cover and forest logged at the start of the experiment. Almost no regeneration into the sapling stage (>30 cm) was observed under ambient deer densities after 7 years and decreased survival and growth were observed under an experimental deer density of 15 km−2. However, mortality at ⩽15 deer km−2 decreased over time and with age and stem height, converging towards mortality observed at 0 deer km−2. Given the relatively high stem density of saplings at 15 deer km−2 7 years after the start of the treatment, our data indicate that at this density considerable balsam fir regeneration may occur, although the ultimate contribution of balsam fir to the canopy remains uncertain. The notion that small seedlings are most vulnerable to deer browsing and that balsam fir recruitment rapidly decreases after logging suggests that maintaining low deer densities is most crucial immediately after a stand-initiating disturbance (e.g. logging).

► Forest regeneration exposed to multiple controlled deer densities.
► Following individual seedlings and saplings of balsam fir over 7 years.
► Virtual fir eradication at ambient deer density, dense regeneration at low density.
► Mortality decreased over time at high (15 deer km−2), but not at ambient deer density.
► Regeneration most sensitive to deer during early regeneration.