In many mature tropical forests, above ground biomass (AGB) has increased over the last few decades,. Such sustained biomass accumulation requires large quantities of N and other nutrients. Symbiotic biological Nfixation (BNF) can bring huge quantities of N into ecosystems, and legumes capable of BNF are abundant in many Neotropical forests suggesting that BNF might be critical for maintaining high rates of biomass growth. Unfortunately, there exist only a handful of sites in which BNF has been measured directly and most of these sites are not coupled with long-term biomass inventories. We asked: how are long-term patterns of forest AGB dynamics affected by legume demographics and BNF and, in turn, how is BNF affected by AGB change? We analyzed patterns of tree demographics and AGB dynamics in permanent forest inventory plots in lowland rainforest of Trinidad in which trees have been identified to species and measured for diameter change, recruitment and mortality for 30 years. We coupled these analyses with ground measures of nodulation, BNF and soil chemistry. We evaluate observed AGB-BNF dynamics using a simple ecosystem model that tracks fixer and non-fixer N accumulation and turnover.
Results/Conclusions
Trinidad forests are biomass rich (~150-380 Mg/ha) and span the range of AGB observed across mature Neotropical forests. While AGB change is heterogeneous and temporally variable across the landscape, it has been net positive over the last 30 years. These forests contain a diverse group of legume species capable of BNF but the most abundant is Pentaclethra macroloba. Sixty-six% of individual trees were nodulated and 80% of surveyed subplots contained nodulated individuals. Upscaling tree-specific measures suggests that BNF varies from ~4-24 kg N ha-1 yr-1 which, to our knowledge, are among the highest rates recorded for mature tropical forest and stand in contrast to the expectation that BNF is down-regulated as succession proceeds. Intriguingly, AGB change was most positively associated with BNF for non-fixing trees, suggesting a powerful indirect influence of legume N supply on forest dynamics. Our measures at the plot scale also suggest an influence of soil P on AGB change. However, at the scale of individual N-fixing trees, our measures suggest that BNF declines with soil N but does not vary with P. We present modeling results which evaluate the extent to which observed AGB growth could be fueled by leguminous BNF.