To understand how landscapes
influence organisms, it is important to understand how habitat and
organisms are distributed in space and time. Because patterns may
emerge at different spatial and temporal scales, the ability to
detect spatial pattern is dependent on sampling procedures. Pattern
detection across multiple spatial scales can be described by the scope.
Scope is defined as the ratio of extent to grain size, where grain
size is the sample unit and extent is the spatial dimension of the study
(i.e., length, area, or volume). Extent provides the context within
which individual sample units are resolved. Another factor influencing
pattern detection is the uncertainty associated with areas that are not
sampled. This uncertainty is expressed mathematically as the ratio of
the number of possible sample units to the number of units actually sampled.
The uncertainty that results from incomplete sampling is termed the
magnification factor, and it provides a tool to evaluate the potential
tradeoffs between increasing extent or sampling density. When
determining spatial pattern is the ultimate research goal and little is
known about the scale at which patterns may emerge, a study design with
high scope and a low magnification factor is the most appropriate.
Research in lotic systems has focused on habitat associations in
which the most common response variable is organism abundance. Grain
size and extent have been variable but sampling has rarely been spatially
continuous because increases in extent typically require increases in
magnification factor. Consequently, scope is usually low and uncertainty
is high, and the capacity to detect spatial patterns in drainage networks
is diminished.
If the detection of spatial pattern is the goal, are highly accurate
and precise abundance estimates at the habitat-unit scale even necessary?
To address this question, patterns in abundance and size structure of
coastal cutthroat trout (Oncorhynchus clarki clarki) were evaluated
by comparing abundance and length-frequency distribution estimates
from single-pass electrofishing without blocknets to estimates from
multiple-pass electrofishing with blocknets.
For additional information about this research study see the
2003 CFER
Annual Report. (2.2 MB)
