Sensitivity of grasslands to rainfall variation: the role of resource colimitation

Abstract

Identifying the factors shaping ecosystem sensitivity to global change is an important unresolved question in ecology. Grasslands and savannas account for one-third of Earth's terrestrial primary production, and variation in that production is strongly driven by rainfall. The sensitivity of ecosystems to rainfall should depend on the availability of other co-limiting resources, such as nutrients. However, it is unclear how the strength of that co-limitation varies among sites - across gradients of aridity, soil fertility and community composition. I leverage long term experiments within the Nutrient Network research cooperative to analyse how fertilization affects the sensitivity of grassland plant communities to rainfall. I expected that sensitivity to rainfall would decline from arid to mesic ecosystems, while the effects of fertilization on sensitivity would increase along that same gradient. I found the sensitivity of grasslands to extreme drought in 13 sites did not change with aridity, but it was increased by fertilization. For 37 grassland sites around the world, fertilization generally increased the sensitivity of biomass to rainfall variation irrespective of aridity. This showed that grasslands are co-limited by nutrients and water all across the aridity gradient present among our sites. Finally, I found that community shifts due to chronic nutrient addition at more than half the sites studied will still retain species that perform well in both high and low rainfall years. Thus composition change is unlikely to amplify the effects of nutrient addition on biomass sensitivity to precipitation.