Previous explorations of the environmental variables and landscape features that shape species distribution patterns have largely relied upon data collected at coarse grains, such as protected area, regional, and national biodiversity assessments. However, coarse-grain data carry analytic risks, including numeric bias and distortions of range porosity. In addition, recent research suggests that ecological processes acting at fine spatial scales, as well as perceptual and physical constraints imposed by life history traits, strongly influence ecological communities and potentially underlie species distribution patterns.
I propose to investigate environmental variables and landscape features that influence select East African bird species at multiple spatial scales, using species distribution models generated from a combination of coarse- and fine-grain data and analyses of fine-grain movement decisions throughout a heterogeneous landscape. Specifically, I propose to observe movement trajectories and behavioral states in target bird species across multiple land cover types and biophysical gradients at Mpala Research Centre (MRC) in Laikipia, Kenya.
I will use 2m-resolution Quickbird satellite images, Shuttle Radar Topography Mission (SRTM) elevation profiles, and other remotely sensed data to build a supervised land cover classification of the Laikipia region. The proposed field work at MRC will enable collection of ground-truth locations for refining and testing the ultimate accuracy of this classification. I will also derive fine-grain environmental variables and landscape features that reflect potential factors driving species occurrence patterns such as:
- distance to road and permanent water sources
- productivity (represented by NDVI)
- vegetation structure (evaluated by texture metrics)
- precipitation (as reported by rain gauges across MRC)
- temperature (interpolated from MRC records and Kenya Meteorological Department climate stations)
I will combine location and behavioral data with layers derived from remote-sensing analyses to generate fine-grain species distribution models. I will then compare these to models generated from coarse-grain data sources (such as natural history collections, online data portals, and scientific literature). Evaluating these sets of models will identify environmental and landscape factors associated with species distribution patterns at various spatial scales. I will also construct fine-grain step selection functions for target species in order to gain a more mechanistic understanding of species spatial patterns, connect fine-grain processes to regional patterns, and increase predictive capacity for species responses to environmental and anthropogenic influences.
By targeting species that represent (functionally or taxonomically) larger groups within East African avian taxa, and compiling fine-grain data from sites stratified across biophysical gradients and land cover types, the proposed research will offer novel perspectives on environmental associations and landscape connectivity for a substantial portion of regional avian diversity. This information will clarify conservation goal-setting and planning for multiple species, and may also aid in resolving competing management objectives among diverse East African taxa, facilitating pragmatic, multidimensional conservation decision-making.