Partitioning Evapotranspiration to Illustrate the Effects of Shrub Competition and Soil Water on Ecosystem State Transitions

Abstract:

Woody-plant encroachment over the last 100 years has shaped the deserts in the southwestern United States. In this work, we examine the process of woody-plant encroachment in the Chihuahuan Desert at the Jornada Experimental Range in New Mexico. Our hypothesis is that differences in shrub phenology among mesquite, tarbush, and creosotebush relative to the seasonal timing of plant available water have governed the historic shifts in shrub-dominated states. This work tests that hypothesis using data obtained from a dense instrument network along with aerial images and phenological measurements. We installed sap flow sensors on 9 shrubs, 3 each of mesquite, tarbush, and creosotebush, to measure seasonal and interspecies variations in transpiration (T). These sensors are located within the footprint of an eddy covariance tower used to measure ecosystem-scale evapotranspiration (ET). We use a network of soil moisture sensors as well as a cosmic-ray sensing soil moisture sensor to measure soil water and relate this to overall ET and species-level transpiration. We partitioned ET into its components, bare soil evaporation (E) and T, using an empirical approach based on soil temperature and soil moisture. For the pre-monsoon and monsoon seasons of 2013 (2014), we found that T/ET was 0.26 (0.24) and 0.70 (0.71), respectively. We use a hydrologic model to partition the ET during this period to compare to the observations from the empirical approach. We further partitioned T into shrub components using sapflow data. We then analyze the activity of each species throughout the year and how they compete for water by examining time series of rainfall, soil moisture, soil temperature, ET, shrub phenological data and vegetation greenness. Our evaluations provide a first look into how plant phenology can impact shrub competition and the implications this has on short-term soil water availability and ET as well as long-term state dynamics in mixed shrublands of the Chihuahuan Desert.