Humans have widely extirpated large carnivores and simultaneously promoted overabundance of deer. The intense pressure imposed by these herbivores in forests has led to extremely low rates of natural forest regeneration. In natural old-growth forests, deadwood functions as a key driver of biodiversity and promotes ecosystem functioning, such as water retention and nutrient recycling. An as yet unappreciated function of deadwood is its ability to act as a physical barrier, excluding large herbivores from the obstructed patches and thereby reducing browsing pressure. However, this benefit may be minimized by an increase in rodent herbivory in the sheltered patches. In this study, a field experiment was conducted in a total of 384 plots in which tree crowns (0–4) from logging residuals were used as increasingly dense physical barriers to shelter five newly planted saplings of silver fir (Abies alba Mill.). Generalized linear mixed-effects models were applied to determine whether sapling browsing by roe deer and rodents was differentially affected by these barriers. The probability of roe deer browsing decreased from 26% (no crowns) to 2% (4 crowns) while that of rodent browsing increased from 1% to 17%, respectively, as the number of deadwood crowns used in barrier construction increased. In broadleaf stands, browsing by roe deer and rodents was generally higher than in coniferous stands. In forests with high numbers of visitors, browsing by roe deer was reduced, but browsing by rodents was not influenced. The retention of large amounts of deadwood or active deadwood increments may thus provide an effective barrier to roe deer browsing but promote the browsing activity of rodents. The landscape-level heterogeneity of browsing patterns associated with the presence of deadwood suggests that deadwood shelters in homogenized forests may encourage both natural forest regeneration and forest biodiversity, despite an overabundance of roe deer.
Read full abstract