PS 30-97
An experimental approach to understanding browsing by moose and deer in recently logged forests

Tuesday, August 11, 2015
Exhibit Hall, Baltimore Convention Center
Justin A. Compton, Biology, Springfield College, Springfield, MA
Stephen DeStefano, Massachusetts Cooperative Fish and Wildlife Research Unit, USGS, Amherst, MA
Edward Faison, Highstead, Redding, CT
David Foster, Harvard Forest, Petersham, MA

With the recent recolonization of moose (Alces alces) in northern temperate forests it has been challenging to determine the role of herbivory by white-tailed deer (Odocoileus virginianus) in shaping regeneration and recruitment of hardwoods. We have come across no studies that have experimentally tested if, and how moose + deer influence tree regeneration and recruitment in recent clear cuts. One of the challenges of studying these two animals is that their molar zones overlap considerably; thus determining the source of herbivory between moose and deer based on height or stem diameter can be very difficult and lead to numerous sampling and recording errors. We created an experimental design that manipulated three levels of large herbivore diversity (moose + deer, deer only, and neither). The experimental design was replicated in eight recent patch clear cut blocks in central New England. At each of these eight sites, we constructed a full exclosure (fence to the ground to keep out moose and deer), a partial exclosure (fence two-feet off the ground to keep out moose but allow deer in), and a control plot (no fencing). We monitored effectiveness of the experimental design with animal-activated cameras and pellet count surveys.


Browsing intensity differed by treatment (full, partial, and control). Data collected from animal-activated cameras between 2008-2011 from 6 sites indicate moose and deer visits in control plots were similar (paired t test; t = 0.19; DF = 4; P = 0.86). Deer visitation rates within partial exclosure (0.27 visits wk-1; SE = 0.11) and control plots (0.52 visits wk-1; SE = 0.19; paired t test; t = -1.15; DF = 4; P =0.31) did not differ. One site had a higher visitation rate in the control, causing the higher control mean. Pellet count surveys are 2 year mean values from 2012-2013 and values are standardized to hectare from 7 harvested sites. Pellet counts indicate no difference between moose and deer within the control plots (t = -0.26; DF = 6; P = 0.81). There was no difference for pellet counts for deer with within the control (210.7; SE = 63.5) or partial exclosure (192.8; SE = 83.3; t = -0.15; DF = 6; P = 0.88). No moose pellets were found in partial exclosure plots. Our results indicate the effectiveness of our experimental design in excluding one large herbivore while simultaneously not altering another large herbivores movements within the treatments.