Abstract: I investigated river otter (Lutra canadensis) latrines
as a means to determine habitat selection along forested lake shore on
Huntington Wildlife Forest (HWF), Newcomb, NY. Habitat characteristics,
including structure (e.g., ground cover, canopy closure), shoreline features
(e.g., morphology, distance to beaver activity), and forest-cover composition
(e.g., conifer, deciduous), were used to discriminate latrine from non-latrine
sites. The variables most effective for distinguishing latrines form
non-latrines were forest-cover (concordance=97%), followed by structure
(concordance=85%), and shoreline characteristics (concordance=68%).
Several major characteristics distinguished latrines from non-latrine
sites. Latrines had more bare ground, greater coniferous forest cover,
and occurred on areas of topographic prominence (e.g., point of land).
In addition, latrines were characterized by a boulder dominated shoreline
and were in close proximity to beaver dams. Latrines were monitored
for a 6 week period and visitation rates were tested for independence (X2=38.503).
Rates of visitation were regressed against structure and forest-cover composition.
Again, the model for forest-cover type described the most variability (R2=0.671),
with the structural model describing much less (R2=0.284).
Results suggest that differences in otter visitation among latrine is strongly
influenced by forest-cover type, similar to the differentiation of latrine
from non-latrine sites. I compared visitation to an increasing shoreline
distance from a preselected latrine in an effort to estimate otter abundance.
By isolating otter visits at one end of the spatial gradient, I found that
a distance of 3500 m between latrine visits could be used as a criterion
for distinguishing otter groups (> 1 otter). The results indicate
at least 2 groups of otters are present on HWF.