PNNL

Abstract

Rising atmospheric carbon dioxide concentration ([CO2]) should stimulate biomass production directly via biochemical stimulation of carbon assimilation and indirectly via water savings caused by increased plant water use efficiency1. Because of these water savings2, the CO2 fertilisation effect should be stronger in drier sites3, yet large differences among experiments in grassland biomass response to elevated CO2 appear unrelated to annual precipitation2,4, preventing useful generalisations. Here we show that, as predicted, the impact of elevated CO2 on biomass production in 19 globally-distributed temperate grassland experiments reduces as mean precipitation in seasons other than spring increases but, unexpectedly, rises as mean spring precipitation increases. Moreover, because sites with high spring precipitation also tend to have high precipitation at other times, these effects of spring and non-spring precipitation on the CO2 response offset each other, constraining the response of ecosystem productivity to rising CO2. This explains why previous analyses were unable to discern a reliable trend between site dryness and the CO2 fertilisation effect2,4. Thus, the CO2 fertilisation effect in temperate grasslands worldwide will be constrained by their natural rainfall seasonality such that the stimulation of biomass by rising CO2 could be substantially less than anticipated.