A key climatic feature of the Shortgrass Steppe is the summertime pattern of moderate to large precipitation events, separated by weeks of hot, dry conditions. These precipitation events have potential to revive plants, microbes and soil fauna that became less active in the intervening dry period, and thus structure ecosystem-scale patterns including net carbon exchange. Previous results from the SGS LTER indicate that the system typically becomes a net carbon source immediately following a precipitation event, and that net carbon uptake occurs only if the event is of sufficient size to elevate soil moisture for more than a few days. To better understand the responses (both timing and magnitude) of plants, microbes and soil fauna to precipitation events of different size, we conducted a rainfall addition experiment on the SGS LTER during July 2009. Our treatments included simulated +1cm and +2cm rainfall events to plots that had received no rainfall for 3 weeks; replication, including controls, was n=5 per treatment. This synthetic experiment brought together the expertise of 13 SGS scientists and their labs to better understand the individual and integrated responses of plants, microbes and soil fauna to precipitation pulses.
Although samples from this experiment are still being processed, some preliminary results are already clear. Within the first 24 hours, ecosystem respiration in both +1cm and +2cm plots peaked at rates that were 5x higher than background. The +1cm plots returned to baseline soil moisture and baseline ecosystem respiration more quickly than the +2cm plots (3 days vs. 5 days). Despite the rapid rise in ecosystem respiration, soil microbial enzyme dynamics were slower, peaking after about 2-3 days, depending on the enzyme and treatment. Soil fauna, many of whom derive their C from grazing on microbes, showed dynamics on a similar scale to the enzyme kinetics, suggesting that interactions between fauna and microbes may be important for soil C and nutrient fluxes. Soil ammonium levels peaked 1 day post-pulse and returned to baseline by day 2, while soil nitrate peaked at day 2 and returned to baseline more quickly in the +1cm plots than in +2cm (4 days vs. 7 days). Photosynthetic activity of the dominant grass (Bouteloua gracilis) peaked on day 2, and when corrected for pre-pulse rates, the +1cm treatments returned to baseline more quickly than the +2cm plots (3 days vs. 5 days). Additional samples that are still being processed will (1) use high-throughput sequencing to quantify changes in the microbial community composition, and (2) follow gross rates and fates of 15N-labeled ammonium and nitrate at days 1-2 and days 6-7. When these samples are analyzed, we will use a soil food web model (Moore et al.) and an ecosystem biogeochemistry model (DAYCENT) to identify key ecological interactions and constraints on ecosystem function. These models, coupled with data from an eddy flux tower in the vicinity of the experimental area, will allow us to place our findings in a broader spatial and temporal context.