Spatial modelling is increasingly being embraced by conservation practitioners and community groups to guide natural-asset management. Despite this boom in popularity, little guidance is often available on choosing the most applicable technique, and the relative merits and drawbacks of each approach. We present a case study in the biodiversity hotspot of south-western Australia where five forms of spatial modelling have previously been used by community conservation managers in this region: realistic GIS modelling, focal species, proximity analysis, Marxan and MaxEnt. We developed a set of criteria for effective spatial modelling and then evaluated each of these spatial-modelling techniques using these criteria. The criteria used in evaluating spatial-modelling tools were as follows: (1) capacity to quantify barriers to movement; (2) capacity to quantify habitat; (3) capacity to accurately predict target responses; (4) capacity to demonstrate how change will happen over differing time frames; (5) effectiveness of the tool within data limitations of a project; (6) ease to which outputs are understood; (7) utility of the tool within skill limitations of a project; and (8) efficacy within resource limitations of a project. None of the applications used met all criteria. When selecting spatial-modelling applications, a hybrid approach is advocated by using criteria that will reflect both the landscape-level conservation requirements and the capacity and objectives of conservation managers. To do this, we advocate using multiple modelling techniques to identify, quantify and ameliorate threats to meet the needs of regional biota.
Mapping intra-urban malaria risk using high resolution satellite imagery: a case study of Dar es Salaam
