Long-term CO2 fertilization increases vegetation productivity and has little effect on hydrological partitioning in tropical rainforests

Understanding how tropical rainforests respond to elevated atmospheric CO2 concentration (eCO2) is essential for predicting Earth's carbon, water, and energy budgets under future climate change. Here we use long-term (1982–2010) precipitation (P) and runoff (Q) measurements to infer runoff coefficient (Q/P) and evapotranspiration (E) trends across 18 unimpaired tropical rainforest catchments. We complement that analysis by using satellite observations coupled with ecosystem process modeling (using both “top-down” and “bottom-up” perspectives) to examine trends in carbon uptake and relate that to the observed changes in Q/P and E. Our results show there have been only minor changes in the satellite-observed canopy leaf area over 1982–2010, suggesting that eCO2 has not increased vegetation leaf area in tropical rainforests and therefore any plant response to eCO2 occurs at the leaf level. Meanwhile, observed Q/P and E also remained relatively constant in the 18 catchments, implying an unchanged hydrological partitioning and thus approximately conserved transpiration under eCO2. For the same period, using a top-down model based on gas exchange theory, we predict increases in plant assimilation (A) and light use efficiency (ε) at the leaf level under eCO2, the magnitude of which is essentially that of eCO2 (i.e., ~12% over 1982–2010). Simulations from 10 state-of-the-art bottom-up ecosystem models over the same catchments also show that the direct effect of eCO2 is to mostly increase A and ε with little impact on E. Our findings add to the current limited pool of knowledge regarding the long-term eCO2 impacts in tropical rainforests.