You are here

Harvard Forest >

Harvard Forest Symposium Abstract 2009

  • Title: Ungulate browsing in ambient vs. disturbed forest conditions at Harvard Forest
  • Primary Author: Audrey Barker Plotkin (Harvard Forest)
  • Additional Authors: Jhessye Moore-Thomas (University of Central Florida); Cassandra Rivas (University of Texas-Pan American)
  • Abstract:

    Herbivory by browsers can strongly influence forest stand development, especially during the regeneration phase following stand disturbances, such as timber harvesting. However, the advance regeneration that forms abundantly under many closed forest canopies dominates post-disturbance forest regeneration dynamics, so selective browsing in undisturbed forests may also shape long-term forest composition. In southern New England, a growing deer population has increasingly affected forest ecosystems over the last 20 years. Recently moose have re-colonized the region after a 200-year extirpation and represent a major new source of browsing pressure.



    At the Harvard Forest, deer are resident, although hunting has kept populations relatively low. In the last few years, moose sign is regularly observed and the presence of cow-calf pairs suggest a resident population. In summer 2008, we surveyed browsing intensity by both species across a suite of 30 plots in three large conifer removal experiments. In the hemlock experiment hemlock, forests were either logged, girdled or left as controls whereas in a plantation removal experiment, forests of spruce or red pine were harvested in patch clearcuts. Three of these plantations were harvested prior to the browsing sample, and five more of these plots were established in a plantation that was partially harvested about 10 years ago. By combining these different conifer plots, we sought to address the following questions:



    1) Is browsing intensity density dependent or independent?

    2) Is browsing intensity categorically different between disturbed and undisturbed stands?

    3) Are certain woody species preferred browse?



    We assessed browsing presence/absence by stem in a set of 4m2 plots arrayed across the Simes hemlock removal experiment plots (n=8 plots, with 56 subplots in each, covering 25% of the central 30x30m plot area) and plantation plots (n=22 plots, with 25 subplots in each, covering 25% of the 20x20m plot area). Woody tree and shrub stems (excluding Rubus) between 0.3-3.0m tall and <5cm diameter were included. While moose browsing can be distinguished from deer browsing when tall (>2m) or large (perhaps >0.5cm diameter) twigs are browsed, there were too few of these saplings in the sample to make clear distinctions. Pellet piles and anecdotal sightings of both ungulate species suggested that both deer and moose were active through much of the area covered by the sample plots.



    Across all plots, 30% of the stems were browsed. Browsing intensity did not vary significantly with sapling density (p=0.12, r2=0.08), but above-average browsing intensity occurred only in plots with sapling densities of >5000 stems per hectare (Figure 1), suggesting a possible threshold. Although more high-density plots were found in disturbed than control plots, both categories spanned the density gradient and disturbance itself did not appear to influence browsing intensity. Therefore, browsing can influence regeneration structure both prior to and after disturbance.



    Browsing is also likely to direct forest species composition. Overall, 39 woody species were observed across all plots, but browsing intensity values were not reliable (i.e., the small sample size for these generally showed 0% or 100% browsing due to chance) for samples less than 50 stems total. For the twelve species with n>50 stems, a series of exact binomial tests (with the critical p-value adjusted to 0.05/# of tests = 0.004) comparing proportion browsed to the overall proportion browsed browsing preference by species (Figure 2). Conifers including white pine (Pinus strobus), spruce (Picea abies/P. glabra) and hemlock (Tsuga canadensis) were avoided, whereas some hardwoods including red maple (Acer rubrum), birches (Betula lenta/alleghaniensis), beaked hazel (Corylus cornuta), pin cherry (Prunus pensylvanica) and blueberries (Vaccinium sp.) were preferentially browsed.



    We plan to re-survey browsing intensity in these plots in five years. In addition, a complementary exclosure experiment (see Faison et al., this volume) has been initiated in some of the harvested plantations to tease apart the effects of deer vs. moose browsing, and to better quantify their long-term influences on forest structure and composition.

  • Research Category: Conservation and Management, Large Experiments and Permanent Plot Studies, Physiological Ecology, Population Dynamics, and Species Interactions

  • Figures:
  • C:datarowse-analysisUngulate-browsing-ABP-symp09_fig1.pdf
    C:datarowse-analysisUngulate-browsing-ABP-symp09_fig2.pdf