Percent cover for Bouteloua eriopoda and Eragrostis lehmanniana was collected over four years at Mt. Summerford (latitude 32.516, longitude -106.802), New Mexico State University's Chihuahuan Desert Rangeland Research Center (CDRRC), Doña Ana County, New Mexico. In June 2016, permanent 2m × 2m plots were installed to evaluate change in percent cover of B. eriopoda and E. lehmanniana in response to precipitation manipulation treatments of +80% (irrigated), -80% (drought), and ambient (control) precipitation. We used a randomized complete block design with six replicates per treatment, with blocks perpendicular to the slope, for a total of 18 plots. Plot locations were adjusted to exclude ant nests, and plant species other than annuals. Cover of other plant species was <1% at the beginning of the study. This cover included Aristida turnipes (0.13%), Panicum hallii (0.20%), and Sporobolus spp. (0.15%). At the upslope border of each plot, we installed galvanized flashing to a depth of approximately 11.4 cm to block surface flow of water onto plots. Metal rebar stakes were driven into the ground at each corner of each plot as permanent markers. A large nail embedded in the center of each plot was used to standardize placement of data collection frames. Rainout shelters, constructed according to the design by Yahdjian and Sala (2002), extended 0.5 m beyond the boundaries of the plots. Rainout shelters manipulated precipitation received, so 151 that irrigated plots received 180% ambient precipitation, drought plots received 20% of ambient precipitation, and control plots received 100% ambient precipitation. Due to repeated malfunctions including pipe leaks and electronic component failures, the maintenance of the irrigation systems became untenable after 2017. Therefore, irrigation plots were excluded and only control and drought data were included in 2018 and 2019 analyses.

Data were collected in August and September of each year from 2016 to 2019 because this is the peak of growth for these species in this monsoon-driven system. Live, green, foliar cover for each species was mapped for each plot using four one m² data collection frames each divided into 16 sections each measuring 0.0625 m². These frames were anchored on the nail in the center of the plot for standardized placement.