Rainforests buffer against climate change by sequestering atmospheric CO2, but their climate sensitivity is uncertain. In prior work, we revisited the notion that tropical rainforest aboveground net primary productivity (ANPP) declines when rainfall exceeds 2.2 meters/yr. Using meta-analysis we showed that montane systems show this pattern, while lowland forest ANPP increases across the rainfall spectrum, as increasing precipitation accelerates nutrient turnover. Now, we’ve extended the database to include aboveground biomass (AGB) and demographics to address how forest dynamics shape the distribution and magnitude of primary forest carbon stocks. We also compare our estimates of ANPP and AGB to products derived from a coupled carbon – climate model (CLM-CN 4.0), Moderate-resolution Imaging Spectroradiometer, and Lidar approaches at both plot – to – pixel and regional scales.
Results/Conclusions
ANPP explains 65% of variation in AGB pantropically, and shows that highly productive forests also store high levels of carbon. The most C-rich forests are found in Southeast Asia, where lower stand mortality rates and higher allocation of photosynthate to wood over leaf growth enhance biomass accrual. This biogeographic pattern reflects a regional phylogenetic signal, and suggests Southeast Asian rainforests may be larger and longer-term C sinks than their New World counterparts. As a whole, pantropical AGB could total 343 ± 140 Pg C, after accounting for uncertainties in up scaling. This value is roughly 50% higher than estimates generated using remote approaches, but aligns well with recent extensive inventory-based assessments conducted in Amazonia and Africa. Finally, plot – to – pixel analyses reveal significant discrepancies between observed and modeled estimates of both ANPP and AGB in all state-of-the-art approaches, highlighting the need for increased communication between the global modeling community and field ecologists to improve estimates of tropical C dynamics.