COS 32-5 - The role of climatic regime shifts, extreme events, and invasive species in desert grassland dynamics of Big Bend National Park, Texas, USA.

Abstract: We examined long-term dynamics of the Chihuahuan Desert grasslands in Big Bend National Park, Texas, USA. The area had been heavily grazed prior to Park establishment in 1944 and vegetation was further impacted by the 1950’s drought. In 1955, permanent study plots were established on lower and upper piedmont slope landforms. Transects were re-measured six times through 2014. Over this period, rainfall averaged 339 mm/yr, with 80% falling during the summer. Precipitation and temperature did not show the long-term trends from 1955 through 2014. Rather, there were distinct, wet-and-cool versus warm-and-dry regimes that aligned with regime shifts in the Atlantic Multi-Decadal Oscillation. We observed strong directional recovery from grazing and drought on both landforms between 1955 and 1981 in response to a wetter and cooler climate after the AMO shifted to its cool phase in the mid-1960s. On the lower piedmont slopes, which are now dominated by Larrea tridentata, the wetter and cooler conditions during the cool phase of the AMO led to large increases in grass cover, and competition between shrubs and grasses appeared to contribute to the decline in the cover of L. tridentata. Those wet conditions did not persist after the AMO shifted to its warm phase. Warmer weather, more erratic precipitation, and droughts led to the loss of perennial grasses and large increases in the density and cover of L. tridentata. Community dynamics were quite different on upper piedmont slopes. The shift to warmer and drier conditions during the warm phase of the AMO led to the loss of cover of many species by 1996, but this was short lived. Wetter conditions led to rapid recovery by 2007. However, a severe freeze followed by extreme drought in 2011 killed off many subshrubs and native perennial bunch grasses. After the drought, the invasive perennial grass, Eragrostis lehmanniana, replaced long-lived native perennial bunch grasses on upper piedmont slopes creating a novel shrub-steppe community with several species of co-dominant shrubs immersed in a non-native perennial grassland. Climate was a dominant driver of community dynamics, but to date has not shown simple long-term trends. Rather we find distinct regimes, correlated to the phases of the AMO, and punctuated by rare extreme events. Thus, while future changes in climate are likely to strongly influence plant communities, they are likely to do so in unpredictable ways.