Field experiments in ecology have long been touted as a means of providing mechanistic insight into contemporary ecological patterns and dynamics. Perhaps more importantly, they are one of few ways by which we can gain insight into how ecological systems may function in a future predicted to have climatic conditions with no analog today. Results from field experiments that manipulate important ecological drivers, such as temperature, precipitation, or nutrients, are often compared with observed responses to natural variations in these drivers. When these are consistent, this gives us confidence in the insights gained from experiments. But in many cases, experiments and observations yield conflicting results. This inconsistency suggests that a more rigorous “reality check” of experiments should be conducted to both improve confidence and potentially constrain inferences derived from climate change experiments. We conducted such as assessment by comparing results from precipitation manipulation experiments with long-term observations of aboveground net primary production (ANPP) and precipitation from the Konza Prairie LTER site.
Results/Conclusions
Based on the compiled results from nine independently conducted precipitation experiments, that either increased or decreased growing season precipitation (GSP), we determined that estimates of mean ANPP for the site, as well as rain-use-efficiency (ANPP/mm precipitation) did not differ from values based on long-term monitoring. This was despite the fact that observational data spanned 34 years, but experiments were conducted in only 15 of those years. We also compared the GSP-ANPP relationship estimated from experimental data versus long-term observations. We determined that within similar ranges of GSP, experimental and observational data estimated the same functional relationship between ANPP and GSP. However, because some experiments altered GSP to higher or lower levels than represented in the observational data, experiments also provided insight into how ecosystem function may change as precipitation regimes become more extreme in the future. In sum, precipitation experiments conducted in this relatively mesic, but water-limited grassland passed this reality check. Future assessments should include other ecological functions and additional climate change drivers. This will increase our confidence that climate change experiments can provide the mechanistic insight needed to improve forecasts of our ecological future.