Maxent and Soil Phosphorus Predictions in a Mixed-Use Montane Watershed.

Our study is centered on the Wallsburg Watershed, a mixed-use, montane watershed in north central Utah.  Throughout the lower elevations, one finds low intensity agriculture, reduced grazing livestock populations, and increased stream bank protections, and yet the surface waters still exhibit undesirably high phosphorus loads  in late-winter and early spring, loads that impact water quality in a major reservoir downstream.  Through our own exploratory sampling and soil testing, we found soils with high levels of extractable phosphorus throughout the watershed, most notably in non-agricultural areas.  Our objective is to use Maxent (Phillips, et al 2004), a modeling method and tool common to ecological studies, in a soil study. We aim to determine its capability of intelligently locating all areas of elevated soil phosphorus in our study watershed. Preliminary results of the Maxent model revealed interesting patterns based on various predictor variables (e.g., topographic, hydrologic, vegetative, pedologic, etc).  We also explore the potential of the model on two nearby watersheds similar in landscape. Our hope is that a proven model might aid in land management decisions for areas likely high in soil phosphorus that are near sensitive surface waters.