An extreme precipitation event, abrupt grass recovery, and soil processes in the Chihuahuan Desert

Brandon T. Bestelmeyer, USDA ARS Jornada Experimental Range, Las Cruces, NM, Matthew Levi, Crop and Soil Sciences, University of Georgia, Athens, GA, Neeshia Macanowicz, USDA-ARS Jornada Experimental Range, Las Cruces, NM, Heather L. Throop, School of Life Sciences, Arizona State University, Tempe, AZ; School of Earth and Space Exploration, Arizona State University, Tempe, AZ and Nicholas Webb, USDA ARS Jornada Experimental Range/New Mexico State University, Las Cruces, NM

Background/Question/Methods
Landscapes in the northern Chihuahuan Desert have been undergoing a gradual shift from perennial grassland to shrubland over the past century, driven primarily by the legacies of historical disturbance events and ongoing aridification. Nonetheless, extreme high rainfall periods can result in instances of remarkable grassland recovery, although they are highly patchy. In one area on the USDA-ARS Jornada Experimental Range, abrupt grass recovery occurred in an eroding Prosopis glandulosa shrubland state in response to a high rainfall period (2006-2009)—an event that we did not imagine could happen. We are comparing ecosystem processes in this “recovered savanna” to 1) patches of grassland that are in a relatively unaltered “reference” state and 2) patches of shrubland state in which grass recovery did not occur. We asked a) is the recovered savanna similar to the reference grassland with respect to structure, composition, and vegetation-soil feedbacks and b) what led to spatial variation in grassland recovery? We compared vegetation and soil chemical and physical properties at twelve replicate study plots in recovered savanna, reference grassland, and shrubland patch types. In addition, we are monitoring soil volumetric water content and aeolian sediment flux within each state.
Results/Conclusions
We found that although the species composition of grasses differed strongly between reference grassland and recovered savanna, ecosystem functions measured were largely similar between these states but differed from shrubland states. Infiltration rate, soil aggregate stability, soil organic carbon, and soil C:N ratio were typically similar between reference grassland and recovered savanna but often far higher than in shrubland. Large bare gaps between plant canopies in the shrubland enabled far higher aeolian sediment fluxes than in the reference grassland and recovered savanna where bare gap connectivity was smaller. Static (edaphic) soil properties and volumetric water content in recovered savanna were intermediate between reference grassland and shrubland, suggesting that sites experiencing grass recovery may have been predisposed to grass establishment by virtue of their topoedaphic settings, yet were not resilient enough to avoid grass loss in the first place. The recovered savannas persist after a decade (including extreme dry years), provide ecosystem services comparable to the reference grasslands that originally occupied these sites, and are superior to the shrublands that they recently replaced. The question before us now is how to encourage more shrubland to savanna transitions.