First deer densities are notoriously difficult to

First, deer densities are notoriously difficult to measure
(Burnham et al. 1980), certainly more difficult than the
coverboard and stem density counts we used to measure
vegetation density. Second, deer management by state
agencies operates at a county or district level (Knox
1997) that might contain multiple units of conservation
land with multiple conservation objectives. Third, bird
species respond to vegetation volume, not deer density.
Control sites located in close proximity to each other
maintained either high or low densities of birds because
of differences in vegetation volume, not densities of
deer. Soil and presumably moisture traits at each site de-
termined the rate at which the vegetation responded to
reduced deer densities. Productive sites can tolerate
higher deer densities, whereas sites with low soil poten-
tial and/or no canopy opening will respond slowly to
deer reduction. DeGraaf et al. (1991) showed that vege-
tation parameters, in their case forest thinning, took pre-
cedence over deer densities in predicting bird numbers.
Lowering deer densities is one means to increase vegeta-
tion density and diversity, but there can be no target
deer density; rather, vegetation measurements will de-
termine when the deer densities at that site are suffi-
cient to allow a vegetation response.
It is difficult to provide a specific vegetation index
that would gauge deer effect. A long-term index would
measure the density and diversity of understory shrubs,
but changes in these values will occur over 5–10 years,
not on the annual basis needed for deer management de-
cisions. An annual index should not be based on seed-
ling densities, because these values showed great annual
variation (personal observation). It is possible to use an
index based on the proportion of browsed twigs for fa-
vored tree species (Balgooyen & Waller 1995). There is
often variability in the relative abundance of preferred
trees, and this may dilute the sensitivity of the index.
Plants within the Liliaceae or Orchidaceae families are
common throughout the eastern United States, and both
the number of plants and the proportion flowering are
sensitive to changes in deer densities (Balgooyen &
Waller 1995; Augustine & Frelich 1998; Fletcher 1999).
Shifts in the abundance and diversity of the bird com-
munity at our sites were dynamic, with birds responding
to annual changes in site condition. Release from deer
browsing caused rapid successional changes in the for-
est understory as vegetation progressed from grasses to
forbs to Rubus spp. to woody saplings. These changes
corresponded to a shift in bird species composition
from Chipping Sparrows to Indigo Buntings to Hooded
Warblers to Ovenbirds. This successional process, in
combination with site differences, makes it difficult to
say whether or not a particular species will increase in
response to lower deer densities, because the answer
depends on the site characteristics and time span in-
volved. For example, Indigo Buntings responded imme-
diately to removal of deer but then declined at exclosure