Northern Prairie Wildlife Research Center
Factors Associated with Duck Nest Success
in the Prairie Pothole Region of Canada
Data Analysis
Nest success for an area with >1 habitat class is a function of the number and
success of nests in each habitat in that area. For many study areas, our samples
of nests probably did not reflect their true distribution among habitat classes;
we were denied access to some land, and we could not search all habitats in equal
proportion (e.g. Woodland and Wetland were more difficult to search than was Grass;
thus we searched smaller proportions of these habitats than of Grass). We attempted
to minimize the effects of unequal search effort on our estimates of nest distribution
for each of the 5 common species in all habitat classes on each study area by
deriving an index to the number of nests initiated in each habitat class on each
half-area-year; this exercise, however, did not overcome other potential biases
with nest drags (e.g., unequal effectiveness in different habitats). The index
was the product of the total amount of habitat in each class in the half-area-year,
the number of breeding pairs in the half-area-year, and species preference for
nesting in the particular habitat class. By scaling the index values to sum to
one, we obtained estimates of the proportions of nest initiations in each habitat
class.
We measured the amount of habitat available in each class (Table
2) and number of breeding pairs (Tables 1-5 of Appendix
A), but had to derive preference values (see Habitat Preference).
Preference of a species for a habitat was defined by Klett et al. (1988) as
the estimated probability that a female will select a particular habitat class
for nesting, given that all habitat classes are equally available.
Previous Section -- Data Collection
Return to Contents
Next Section -- Habitat Preference
U.S. Department of the Interior |
U.S. Geological Survey