Resilience and alternative successional trajectories in temperate forests exposed to two large herbivores

Ungulate browsing is a key disturbance process influencing forest structure and composition, with the potential to alter forest resilience following canopy disturbance. We examined the effects of l5 years of white-tailed deer ( Odocoileus virginianus ) and moose ( Alces alces americana ) browsing in regenerating temperate forests of the northeastern U.S. We tested the alternative hypotheses that woody vegetation is (1) resilient to browsing, with structure and composition similar after stems have grown above the browser trap or (2) altered by browsing, resulting in long-term legacy effects. Three browser treatments—ungulate exclusion, deer, and deer + moose—were established in seven clear-cut stands. Basal area and species diversity did not differ significantly among treatments after 15 years, indicating resilience, whereas the abundance of two tree species was altered: pin cherry ( Prunus pensylvanica ) was lower by 3-fold in deer + moose plots, while white pine ( Pinus strobus ) was greater by over 5-fold. Forb cover was half as abundant in deer + moose plots compared to ungulate exclusion, while forb richness remained unaffected. Our results suggest that ungulate browsing during stand initiation can alter successional trajectories while still being compatible with long-term forest resilience; however, there are potential trade-offs between woody plant resilience and the vigor of the herbaceous layer when large herbivores are present. With relatively small long-term effects on structure and diversity and by accelerating succession towards longer-lived tree species, deer + moose browsing in northern temperate forests may be less of a management concern than is generally assumed. • Woody plant structure and diversity were resilient to browsing by large herbivores. • Browsing by two large herbivores resulted in alternative successional trajectories. • Prunus pensylvanica was reduced by 3-fold in areas browsed by deer + moose. • Pinus strobus was greater by 5-fold in areas browsed by deer + moose. • Forb cover was half as abundant in deer + moose plots compared to ungulate exclusion.