A Distributed Hydrologic Model Application to Explore Ecohydrological Dynamics of Mesquite Encroachment during Winter and Summer Conditions

Woody plant encroachment is an important issue linked to water availability in semiarid ecosystems in the southwestern United States whose effect on the hydrologic cycle at the catchment scale is not well understood. In this study, we use observations from two small rangeland watersheds of the Santa Rita Experimental Range, Arizona. This Sonoran Desert landscape is representative of the vegetation shift from grasslands to a woody savanna due to the encroachment of velvet mesquite (Prosopis velutina). The paired watersheds have similar soil and meteorological conditions, but vary substantially in terms of vegetation cover, with one basin having undergone mesquite removal in 1974. Long-term observations from the watersheds exhibit changes in runoff production over time, indicating that the watershed with mesquite encroachment leads to less runoff for small rainfall events and more runoff for large rainfall events during the North American Monsoon (NAM). In this study, we explore the seasonality of precipitation and runoff in the two watersheds and investigate the link between winter and summer precipitation seasons. Ecohydrological observational datasets are obtained from a high-resolution environmental sensor network consisting of six rain gauges, twenty-one soil moisture and temperature profiles representing different vegetation types, two channel runoff flumes and an eddy covariance tower with a complete set of radiation, energy, carbon and water fluxes. High-resolution digital terrain models and image orthomosaics were obtained from a piloted aircraft with Light Detection and Ranging (LiDAR) measurements and an Unmanned Aerial Vehicle (UAV) with a digital camera. The field and remote sensing observations are used in simulations using the TIN-based Real-time Integrated Basin Simulator (tRIBS) at high spatiotemporal resolutions over three study years (June 2011 to June 2014). Model simulations are used to investigate the vegetation controls on soil moisture, evapotranspiration, and runoff generation for the bimodal precipitation regime. Through the spatiotemporal analysis of model outputs, we identify how and when mesquite trees affect the spatial patterns of energy and water fluxes and their linkage to runoff production. The model simulations are then evaluated to explore how antecedent conditions in winter affect soil moisture and runoff production in the NAM (and vice-versa). As a result, the distributed model application provides a more complete understanding of the impact of woody encroachment during the annual cycle and can be used to explain runoff threshold for different seasons. This information can provide useful interpretations for rangeland management in semiarid areas of the western United States.