PS 32-23 - Rapid shifts in soil hydraulic properties in response to simulated rainfall

Wednesday, August 9, 2017
Exhibit Hall, Oregon Convention Center
Joshua S. Caplan1, Daniel Giménez1, Daniel R. Hirmas2, Nathaniel A. Brunsell2, John M. Blair3 and Alan K. Knapp4, (1)Department of Environmental Sciences, Rutgers University, New Brunswick, NJ, (2)Department of Geography & Atmospheric Science, University of Kansas, Lawrence, KS, (3)Division of Biology, Kansas State University, Manhattan, KS, (4)Department of Biology and Graduate Degree Program in Ecology, Colorado State University, Fort Collins, CO

Changes in precipitation regimes are known to alter soil hydraulic properties over relatively long timescales (i.e., centuries to millennia), with these changes responsible for numerous alterations to ecosystem composition and function. However, it is unclear if the relatively rapid shifts in precipitation expected in the coming century could also induce shifts in soil hydraulic properties. Given that soil pore structure is mediated by plant and microbial processes and that these biological processes can respond rapidly to precipitation change, we speculated that ecologically important hydraulic properties can likewise change over short (i.e., decadal) timescales. We evaluated this possibility in the context of a rainfall simulation experiment at the Konza Prairie Biological Station in eastern Kansas, USA. The experiment consisted of four elevation transects in a native grassland overlying a siltly loam soil, of which two transects experienced ~30% increases in annual rainfall for 22 years prior to this study. We measured infiltration rates and near-surface water retention (<5 cm depth) over typical ranges of pressure potentials. We then quantified effect sizes for shifts in the hydraulic properties as a function of precipitation regime and hillslope position (summit vs. footslope). We also quantified particle-size distributions, carbon content, and aggregate stability to gain greater insight into the mechanisms underlying any hydraulic changes we observed.


We found that infiltration rates were reduced by 21-33% in soils that experienced simulated rainfall. This effect was independent of hillslope position and spanned most of the pressure potentials evaluated (-1.5 to -5.5 cm). In addition, water retention was greater in irrigated vs. control plots at pressure potentials between -10 and -1000 cm, but reduced at potentials below -1000 cm. Water retention was also greater in soils from the summit than from the footslope; this appeared to be due to shifts in particle-size distributions associated with normal hillslope geomorphic processes. Our results suggest that regions in which growing season precipitation is expected to increase over the coming decades could experience rapid shifts in soil hydraulic properties. We recommend that rainfall simulation studies assess the effects of such changes on soil water balance, as these could help modeling efforts predict ecological responses to changing precipitation regimes.