Questions: Have forest dynamics changed significantly in intact Amazon rainforests since the early 1980s? If so, what environmental drivers might potentially be responsible?
Location: Central Amazonia, north of Manaus, Brazil.
Methods: Within 20 1-ha plots scattered over ∼300 km2, all trees (≥10 cm diameter at breast height) were marked, identified, and measured five times between 1981 and 2003. We estimated stand-level dynamics (mortality, recruitment, and growth) for each census interval and evaluated weather parameters over the study period.
Results: We observed a widespread, significant increase in tree mortality across our plots. Tree recruitment also rose significantly over time but lagged behind mortality. Tree growth generally accelerated but varied considerably among census intervals, and was lowest when mortality was highest. Tree basal area rose 4% overall, but stem number exhibited no clear trend. In terms of climate variation, annual maximum and minimum temperatures increased significantly during our study. Rainfall anomalies were strongly and positively associated with ENSO events.
Conclusions: The increasing forest dynamics, growth, and basal area observed are broadly consistent with the CO2 fertilization hypothesis. However, pronounced shorter-term variability in stand dynamics might be associated with climatic vicissitudes. Tree mortality peaked, and tree recruitment and growth declined during atypically wet periods. Tree growth was fastest during dry periods, when reduced cloudiness might have increased available solar radiation. Inferences about causality are tenuous because tree data were collected only at multi-year intervals. Mean temperatures and rainfall seasonality have both increased over time in central Amazonia, and these could potentially have long-term effects on forest dynamics and carbon storage.