| Summary: | Biologists and wildlife managers recognise the need for systematic reserve selection
techniques to conserve habitat for species. Geographic information systems (GIS) provide a
tool that helps identify conservation areas using geographically referenced data. Mapping
continuous geographical phenomena with discreet boundaries affects the spatial organisation
of data. However, most published studies of systematic reserve selection techniques have
been completed at only one level of data organisation and usually for large regions at coarse
scales. In this thesis, I examined the effects of data organisation on reserve selection in the
South Okanagan, a small region in British Columbia. I used the software program "C-Plan"
with ArcView GIS to identify the minimum amount of area required to achieve explicit
conservation targets that maintain species within the region. I evaluated the reserve selection
technique using terrestrial ecosystem mapping (TEM) and species habitat models that predict
the suitability of T EM polygons for each of twenty-nine threatened vertebrate species. C-Plan
selected 37.2% of the region to represent habitat that maintains current population sizes of
these threatened vertebrate species. Although habitat area targets were achieved, these
priority sites were small and scattered throughout the region and were therefore not practical
for implementation or viable for many species. I examined the effects of data organisation on
priority site selection by altering three algorithm parameters: (1) size and shape of the unit
used to map data and select sites, (2) type of species included in selections, and (3) quantity of
the conservation target for each species. The spatial overlap of priority sets of sites that were
identified for different values of each parameter was low. Therefore, the spatial distribution
of priority conservation sites depends on values for these parameters. Data organisation also
influenced the evaluation of existing protected areas in the region for maintaining the
threatened vertebrate species. Both selection unit size and assignment of protection status to
selection units, based on area of overlap with actual protected areas, resulted in different
evaluations of reserve performance. I demonstrate that systematic reserve selection cannot be
performed with data at only one spatial organisation unless the consequences are recognised.
|
|---|
