Global climate change is resulting in regional increases in the frequency of extreme climatic events, including multi-year wet or dry periods. Ecological effects of these infrequently occurring multi-year events are difficult to examine such that observations of single events are often used to predict future dynamics. Most of the focus has been on drought periods, yet multi-year wet periods are also expected. Recent observations from the Chihuahuan Desert show strikingly different patterns in grass production responses in shrublands between two naturally-occurring sequences of wet years (1984-1988, 2004-2008). To explain these patterns, we tested two hypotheses, either that different responses in the two wet periods were: (1) related to differences in climate, such that wet periods are similar, but not the same (climatic modoki), or (2) contingent on pattern-process relationships that affected fine-scale plant recruitment and growth. To test our hypotheses, we analyzed long-term datasets from the Jornada USDA-LTER site on primary production (response variable), patch-scale properties (horizontal redistribution of soil, bare gap size distribution, soil texture by depth), and broad-scale drivers (precipitation, temperature, wind speed, livestock grazing). We compared grass recruitment in the two periods using a simulation model that integrated these datasets to generate patch-scale soil water dynamics.
Results/Conclusions
Results show that differences in production responses between the two wet periods can not be explained by differences in broad-scale climate (precipitation and wind speed at multiple scales [daily, seasonal, annual] or grazing. Differences in grass recruitment and growth between time periods were contingent on patch-scale pattern-process relationships associated with the effects of wind and water. In the first wet period, large bare soil gaps and low grass cover resulted in high soil flux by wind and low plant available water (PAW) that led to low recruitment of grass seedlings and a linear relationship between grass production and precipitation. By contrast in the second wet period, small bare soil gaps and high grass cover led to negligible soil flux and higher PAW. Positive feedbacks were initiated that increased grass recruitment and resulted in a nonlinear increase in grass production through time. Accounting for land surface properties and vegetation-soil-water-wind interactions can be used to distinguish impacts of wet period modokis. These results demonstrate the importance of patch-scale biotic contingencies to grass recovery in desertified shrublands, and to a potential state change reversal in wet periods.