Evaluation of shrub encroachment and brush control on water availability in the Upper Guadalupe River watershed

Abstract

Wooded plant encroachment has dramatically changed the composition of rangelands in the arid and semiarid rangelands of the southwestern United States and may have significantly affected hydrologic and biogeochemical process in these environments. In particular, suspicions that encroaching species waste an undue amount of water through evapotranspiration (ET) has prompted much discussion concerning the possibility of using brush control to enhance water supplies in Texas. This study focuses on two broad goals for evaluating the effects of wooded growth in rangelands. The first of these is the assessment of wooded cover with the use of remotely sensed imagery. A methodology for delineating differing land cover classes, including different levels of brush cover, is described, applied, and validated for the Upper Guadalupe River watershed, Texas. This portion of the research resulted in an 81.81% success rate for correctly matching land cover varieties and showed that 88.8% of the watershed was covered with various amounts of woody plant growth. The second portion of this study incorporated the previously developed land cover product along with a number of other highly detailed data sources to model the North Fork of the Upper Guadalupe River watershed using the Soil and Water Assessment Tool (SWAT). The role of topography, brush cover, and soil slope, which are hypothesized to contribute to successful implementation of brush removal for water yield, were tested in a scientifically conscious and practical experiment to determine their influence upon water availability at a watershed scale. The effects of brush removal were found to be comparable to the quantities documented in field experiments, but less than the levels presented in previous modeling studies. Brush density was found to be the most important factor in determining locations for successful brush removal in regards to reducing ET. Slope was also found to have significant effect in increasing lateral flow while shallow soil had lesser effects on hydrology than other criteria. Large quantities of deep recharge simulated by the model raise questions concerning measurement of ET in the Edwards Plateau region and the extent of deep water recharge to the Trinity Aquifer.