Lags in the response of ecosystems to directional changes in water availability

Background/Question/Methods

In most arid and semiarid ecosystems the major impact of climate change will occur through changes in soil water availability. Current understanding of arid-ecosystem responses to changes in water availability is mostly based on observations or short-term manipulations. These methods are insufficient to estimate the effects of long-term directional changes on the different components of populations, communities and ecosystems. The infrequent nature of exogenous phenomena, such as El Niño, and the cumulative effect on populations, communities and ecosystem components require long-term manipulative experiments. This paper is based on a conceptual framework that includes a hierarchical set of responses to directional changes in water availability, starting with plant-physiological responses, followed by changes in relative abundance and finally changes in composition all affecting ecosystem functioning. Thresholds are expected as a result of endogenous and exogenous mechanisms. This paper reports on an experiment of rainfall manipulations (80% reduced PPT, ambient, 80% increased PPT) located in the Jornada Basin LTER site in the Chihuahuan Desert grassland (NM) for the last seven  years. There were three types of response variables: (1) those that assessed the effectiveness of precipitation manipulation; (2) those used to test hypotheses; (3) those that assisted in unraveling the mechanisms behind hypotheses.

Results/Conclusions

Results from the seven-year manipulations show that ecosystem responses depend on water availability and on time of exposure to chronic resource alterations (water-treatment F_2,23=92.69, P<.0001; year F_6,26=107.87, P<.0001). Difference in total aboveground net primary production (ANPP) among precipitation treatments increased within the first three years of manipulation. Sustained drought or wet years had different effect on grass and shrub ANPP. Grass ANPP showed a fast and substantial response while shrub ANPP had little or no response to changes in water availability. The year 2013 was wetter than the previous 4 years however grasses that have been exposed to drought for 7 years were not able to respond to this increase in soil water, suggesting that grasses have crossed a threshold and lost their ability to respond to an increase in precipitation. The dominant grass, black grama, was more sensitive to both drought and wet conditions than other subdominant perennial grasses. The shrub mesquite was the least sensitive and, in some years, showed a positive response to drought perhaps benefiting from the demise of the dominant grass. Species richness was similar among treatments until 2011, when there was a significant decrease in number of species with drought compared to irrigated treatments.