Geoscience Reference
In-Depth Information
map will be interpreted as the species is present only at that exact location;
most would assume that the detection is indicative that the species is pres-
ent in some larger area. The area over which the detection is believed to be
indicative of presence would define the sampling unit.
Consideration of the home range, or territory, size of an individual of the
target species has also been suggested as a means of defining a sampling
unit, particularly for studies in which occupancy is interpreted as a surrogate
for abundance. This might be a reasonable approach where individuals tend
to occupy relatively well-defined natural features (e.g., drainages or habitat
patches) but might not be that useful when sampling units (e.g., grid cells) are
of an arbitrary size in a more contiguous landscape. For example, just because
the grid cell has been defined at the approximate size of an individual's home
range, it is unlikely that the home range will fall fully within a single grid
cell because the species has no knowledge of the grid lines placed on a map;
hence, a single individual might occupy multiple cells. Therefore, attempting
to draw inferences about abundance from an occupancy metric might be mis-
leading in this situation. In fact, generally speaking, there are relatively few
cases for which studies can be suitably designed such that occupancy metrics
could be meaningfully interpreted as something close to abundance.
It should be remembered that different definitions of a sampling unit may
be suitable for different studies with different objectives even when working
on the same species in the same general area. There are no hard-and-fast
rules that will apply in most situations.
9.6.3 Defining a Sampling Season
Much like the definition of a sampling unit, sampling seasons have often been
defined implicitly and encouraging such assumptions to be explicitly stated
should improve the rigor of the study. Recall that the sampling season is the
time period when the occupancy state of each sampling unit in the popula-
tion is assumed to be static, or at least not changing in a nonrandom manner.
This assumption is required so that the true occupancy state is observable
for each survey within the season (i.e., that the units are closed to change in
the occupancy state). Whenever the results of a single survey for a species
are extrapolated beyond the actual survey period, then a similar assumption
is being made. For example, if the detection of a species in a 5-minute point
count is regarded as evidence the species is present at a site for the next 2
weeks (say), then that implies the occupancy status of the site is closed to
changes for the next 2 weeks (i.e., there is a 2-week sampling season).
The sampling season does not have to correspond to a naturally occurring
season such as a climatic or breeding season (although it may). It is the time
period for which the occupancy status of a unit is applicable for a particular
snapshot of the population. Therefore, the objective of the study or monitor-
ing program should be the primary determinant of the sampling season.
