Predicting distributions of tortoise-beetles (Chrysomelidae, Coleoptera) using niche modeling and cost surfaces with remote sensing images

Despite the importance of tropical biodiversity, informative species distributional data are seldom available for biogeographical study, especially if the only data source is old locality data from museums. This makes decisions of potential collecting areas difficult, due to the uncertainty of the species distribution range. Modelling ecological niche distributions of species offers a potential solution. Ecological niche modelling (ENM) provides the realized niche based on the environmental space of species current distributions, therefore a better idea of their distribution range. However, not always the potential distribution model narrows down to the most accessible areas to collect, making optimization of field work decisions necessary. Here I present a new methodology that yield greater accuracy and permits taking better decisions for field work locality wise, by combining ENM techniques and spatial analysis tools, using R and ArcGIS software. From the ENM output an environmental centroid was calculated, and used to measure the distance of each occurrence point to it, followed by a density distribution, which yield how much of the species realized niche had been sampled. This step is followed by a Least Cost Path Analysis (LCPA), that uses land use layers to produce a cost surface. As an outcome, I obtained the most suitable places for field work, by combining ENM and LCPA. The likelihood of finding the specimen on field will still be tested.