Geoscience Reference
In-Depth Information
tion than they do. Peripheries expand and contract as home range cores con-
tain or fail to contain necessary resources. How are peripheries expanded?
How does an animal choose where and how to expand its peripheral home
range? Does the expansion depend on past knowledge and, if so, how did the
animal obtain that knowledge? How important are occasional sallies? And if
occasional sallies are important, should they be included in home ranges?
Conversely, what critical resources are contained in a home range core?
How can estimates of animals' home ranges be improved? No universally
best method can be developed for estimating home ranges of all animals be-
cause the best method for one research project may be different than that
for another. Optimal methods depend on the hypotheses being tested and
different hypotheses demand different analyses. Researchers must understand
the strengths and limitations of each home range estimator and choose the
one that provides the most accurate information related to the hypotheses
being tested or to research objectives. Nonetheless, some generalities exist for
improvement.
Home ranges must be quantified over biologically meaningful periods of
time. Data collected over too short a time period may not sample an animal
within all areas that it considers to be its home range. Data collected over too
long a time period may suffer from changes in an animal's home range during
the time period.
Kernels and brownian bridges incorporating time sequence information
(i.e., in what direction is an animal most likely to travel next?) promise to be
useful for many studies (see Bullard 1999). Future kernel estimators should
also use information on telemetry error or area of trap attraction and informa-
tion on an animal's radius of perception when calculating band width. Using
probabilities generated by kernel estimators to explore use of space at different
scales has much promise.
The fractal estimator, 1/
C
, may have promise. A future fractal estimator
could become an estimator of choice for some research if development allows
C
for each cell to include time and multiscale information from the entire data
set (not just from one cell), finds a way to determine optimal cell size for
research questions of known scale, and solves the critical problem of
C
depend-
ing on an unstable estimate of home range area.
All methods of estimating and quantifying animals' home ranges have
problems and no method is best for all research. Because of the myriad prob-
lems associated with quantifying home ranges, especially boundaries and areas
of home ranges, White and Garrott (1990) suggested that the home range con-
cept is obsolete and that our understanding of animal ecology and behavior
