B32D-04:
The Ups and Downs of Rhizosphere Resource Exchange

Wednesday, 17 December 2014: 11:05 AM
Zoe G Cardon, Marine Biological Laboratory, Woods Hole, MA, United States, Congsheng Fu, University of Connecticut, Groton, CT, United States, Guiling Wang, University of Connecticut, Storrs, CT, United States and John Stark, Utah State University, Logan, UT, United States
Abstract:
Hydraulic redistribution (HR) of soil water by plants occurs in seasonally dry ecosystems worldwide. During HR, soil water flows from wet soil into roots, through the root system, and out of roots into dry rhizosphere soil. Hydraulic redistribution affects plant physiology and landscape hydrology, and it has long been hypothesized that upward HR of deep water to dry, nutrient-rich surface soil may also stimulate soil nutrient cycling and thus enhance nutrient availability to plants in the field. We report results from a sagebrush-steppe field experiment in northern Utah, USA, showing that stimulation of sagebrush-mediated HL increased rates of nitrogen cycling in the surface soil layer around shrubs at summer’s end, and more than quadrupled uptake of nitrogen into developing sagebrush inflorescences.

We have built on these empirical data by folding Ryel et al.’s (2002) HR formulation into CLM4.5 and examining how well the combined model can simultaneously simulate measured evapotranspiration, the vertical profile of soil moisture, and the amplitude of HR-associated diel changes in water content, at multiple seasonally-dry Ameriflux sites: Wind River Crane (US-Wrc), Southern California Climate Gradient (US-SCs,g,f,w,d,&c), and Santa Rita Mesquite Savanna (US-SRM). The simulated hydraulic lift during the dry periods has an average value in the range from 0.09 (at US-SCc) to 0.64 (at US-SCf) mm H2O d-1. In many cases, the combined model reproduced seasonal and daily (diel) observations with reasonable accuracy. Among the many model parameters tested, the Clapp and Hornberger parameter “B” in CLM4.5 was critical for a realistic simulation of soil moisture. Modeled HR was also sensitive to the maximum radial soil-root conductance and the soil water potential where that conductance is reduced by 50%. Our next step is to explore how modeled carbon and nutrient cycling in soil layers are affected by redistributed water in the soil column caused by inclusion of HR in CLM4.5.