Variation in Near-Surface Soil Water Content and Matric Potential Across a Semiarid Shrub-Grass Ecotone: Sevilleta Wildlife Refuge, New Mexico

The 20th century shrub invasion of semiarid grasslands in New Mexico is one example of the transition of a semiarid herbaceous to woody plant ecosystem. A key factor controlling plant productivity and reproduction in semiarid environments is water availability in the soil. We characterize the hydrologic changes resulting from the shrub invasion, specifically the hydrologic advantage that shrubs hold over grasses, to pinpoint the primary factor(s) that contribute to the invasion. Observations are based on measurements of near surface soil moisture and matric potential across the transition. We focus on the differences between the grass and shrub environments. Our previous research has shown that in both environments, maximum spatial soil moisture variations exist at the plant-to-interspace scale, on the order of meters. Therefore, we now focus on individual plant canopy and interspace patches. We hypothesize that the critical hydrologic difference(s) contributing to the shrub invasion exist at this scale. During the 2001 summer monsoon season we used the TDR (time domain reflectometry) method to measure soil water content. In both environments, TDR probe arrays are installed laterally covering the surface of nine plant canopies in total and their adjacent interspaces, using 128 probes. One array in each environment contains TDR probes inserted both at the surface and at depths. In addition, soil matric potential is measured adjacent to selected TDR probes both laterally and vertically. Matric potential is measured using a total of 64 heat dissipation sensors. The probe arrays are designed to capture the effects of topographic relief and plant shading. Measurements are made and recorded hourly. We focus on initial dry soil conditions, wetting events, and subsequent dry down sequences following natural rainfall events. Results suggest that during any rain event, shrub canopies accumulate a greater percentage of precipitation applied to an area than grass canopies, most critically during low precipitation events on the order of millimeters. Both shrub canopies and interspaces additionally lose soil moisture via evaporation and transpiration more quickly than in the grassland, on the order of days. This induces a greater water stress than in the grasslands. Currently, we are analyzing more rainfall events in order to reliably quantify the differences in water content, matric potential gradients, and time-scales of soil moisture consumption between the two environments.