%0 Journal Article %J Biology Letters %D 2009 %T Scale effects in species distribution models: implications for conservation planning under climate change %A Seo, Changwan %A Thorne, James H. %A Hannah, L %A Thuiller, Wilfried %K global climate models %K grid size sensitivity analysis %K sensitivity analysis %K species range %X

Predictions of future species' ranges under climate change are needed for conservation planning, for which species distribution models (SDMs) are widely used. However, global climate model-based (GCM) output grids can bias the area identified as suitable when these are used as SDM predictor variables, because GCM outputs, typically at least 50×50 km, are biologically coarse. We tested the assumption that species ranges can be equally well portrayed in SDMs operating on base data of different grid sizes by comparing SDM performance statistics and area selected by four SDMs run at seven grid sizes, for nine species of contrasting range size. Area selected was disproportionately larger for SDMs run on larger grid sizes, indicating a cut-off point above which model results were less reliable. Up to 2.89 times more species range area was selected by SDMs operating on grids above 50×50 km, compared to SDMs operating at 1 km. Spatial congruence between areas selected as range also diverged as grid size increased, particularly for species with ranges between 20 000 and 90 000 km. These results indicate the need for caution when using such data to plan future protected areas, because an overly large predicted range could lead to inappropriate reserve location selection.

%B Biology Letters %V 5 %P 39-43 %8 2009 %G eng %U http://rsbl.royalsocietypublishing.org/content/5/1/39.abstract %! Biology Letters %0 Journal Article %J Journal of Conservation Planning %D 2005 %T Choosing surrogates for biodiversity conservation in complex planning environments %A Stoms, D. M. %A Comer, P. J. %A Crist, P. J. %A Grossman, D. H. %K coarse-filter %K conservation planning %K fine-filter %K Napa County %K reserve selection %K sensitivity analysis %K Sites %K surrogates %X The coarse filter/fine filter hypothesis suggests that by conserving high-quality examples of all ecological systems along with imperiled species and communities, we could protect the majority of native biodiversity. Given the cost of data collection, conservation planners might wonder how large this set of elements must be. We conducted an analysis of the sensitivity of selecting a set of reserves to the choice of surrogates in Napa County, California, USA. The study evaluated the extent to which conservation goals for the coarse/fine-filter elements were met by surrogates and whether the same general locations were being selected. Napa County represents a data-rich setting, whereas the test surrogates portrayed a range of circumstances where less data are available. A worst (data-poor) case, based only on landscape condition with no biological data, was tested to identify the value of improved information. Our results suggest that in complex planning environments, there are no simple shortcuts in collecting data. None of the surrogate sets was particularly effective at meeting all the goals for the full set of baseline elements. There was also relatively low spatial congruence between the test solutions and the baseline. However, we did find that all combinations of surrogates provided some degree of protection in notional reserves, suggesting that in less complex planning problems, simpler surrogates can serve a useful function. Studies like this will help planners gauge how much effort it is prudent to spend in compiling spatial data relative to the risks and irreplaceability to native biodiversity. %B Journal of Conservation Planning %V 1 %P 44-63 %8 2005 %G eng %U http://www.journalconsplanning.org/2005/volume1/issue1/stoms/manuscript.pdf %0 Journal Article %J Conservation Biology %D 1993 %T Geographic analysis of California condor sighting data %A Stoms, D. M. %A Davis, F. W. %A Cogan, C. B. %A Painho, M. O. %A Duncan, B. W. %A Scepan, J. %A Scott, J. M. %K habitat suitability %K sensitivity analysis %X Observation and habitat data were compiled and analyzed in conjunction with recovery planning for the endangered California Condor (Gymnogyps californianus). A geographic information system (GIS) was used to provide a quantitative inventory of recent historical Condor habitats, to measure the association of Condor activity patterns and mapped habitat variables, and to examine spatio-temporal changes in the range of the species during its decline. Only five percent of the study area within the historic range is now used for urban or cultivated agricultural purposes. Observations of Condor feeding perching, and nesting were nonrandomly associated with mapped land cover, in agreement with life history information for the species. The precipitous decline in numbers of Condors in this century produced only a small reduction in the limits of the observed species 'range, as individual birds continued to forage over most of the range. Some critical risk factors such as shooting and lead poisoning are difficult to map and bave not been included in the database Besides the applications demonstrated in this case study, GIS can be a valuable tool for recovery planning, in the design of stratified sampling schemes, or for extrapolation of habitat models over unsurveyed regions. We conclude with recommendations from this case study regarding when to consider using GIS and the importance of pilot studies and sensitivity analysis. %B Conservation Biology %V 7 %P 148-159 %8 1993 %G eng %U ://A1993KR98600021 %0 Journal Article %J Photogrammetric Engineering and Remote Sensing %D 1992 %T Effects of habitat map generalization in biodiversity assessment %A Stoms, D. M. %K habitat suitability %K scale %K sensitivity analysis %K species richness %X Species richness is being mapped as part of an inventory of biological diversity in California (i.e., gap analysis). Species distributions are modeled with a GIS on the basis of maps of each species' preferred habitats. Species richness is then tallied in equal-area sampling units. A GIS sensitivity analysis examined the effects of the level of generalization of the habitat map on the predicted distribution of species richness in the southern Sierra Nevada. As the habitat map was generalized, the number of habitat types mapped within grid cells tended to decrease with a corresponding decline in numbers of species predicted. Further, the ranking of grid cells in order of predicted numbers of species changed dramatically between levels of generalization. Areas predicted to be of greatest conservation value on the basis of species richness may therefore be sensitive to GIS data resolution. %B Photogrammetric Engineering and Remote Sensing %V 58 %P 1587-1591 %8 1992 %G eng %U ://A1992JV67200007 %0 Journal Article %J Photogrammetric Engineering and Remote Sensing %D 1992 %T Sensitivity of wildlife habitat models to uncertainties in GIS data %A Stoms, D. M. %A Davis, F. W. %A Cogan, C. B. %K habitat suitability %K sensitivity analysis %X Decision makers need to know the reliability of output products from GIS analysis. For many GIS applications, it is not possible to compare these products to an independent measure of "truth." Sensitivity analysis offers an alternative means of estimating reliability. In this paper, we present a GIS-based statistical procedure for estimating the sensitivity of wildlife habitat models to uncertainties in input data and model assumptions. The approach is demonstrated in an analysis of habitat associations derived from a GIS database for the endangered California condor. Alternative data sets were generated to compare results over a reasonable range of assumptions about several sources of uncertainty. Sensitivity analysis indicated that condor habitat associations are relatively robust, and the results have increased our confidence in our initial findings. Uncertainties and methods described in the paper have general relevance for many GIS applications. %B Photogrammetric Engineering and Remote Sensing %V 58 %P 843-850 %8 1992 %G eng %U ://A1992HX38700006