Intensification of the Amazon hydrological cycle inferred from tree-ring stable isotopes

Understanding how the hydrological cycle of tropical regions has responded to recent climate change is critical for assessing ecosystem resilience, carbon cycling, and the risk of large-scale forest transitions. However, long-term observational records of precipitation remain sparse across much of the tropics, and existing datasets often disagree on both the magnitude and direction of rainfall trends. In particular, whether recent changes reflect general drying, wetting, or an amplification of rainfall seasonality remains unresolved.

This talk examines how stable oxygen isotope ratios (δ¹⁸O) preserved in annually resolved tree rings can provide large-scale, seasonally resolved insights into hydroclimate change. The analysis draws on an Amazon study based on oxygen isotope chronologies from two tree species with contrasting growth phenologies and hydrological settings: Cedrela odorata from terra firme forests, which forms annual rings during the wet season, and Macrolobium acaciifolium from seasonally flooded forests, which grows during the terrestrial phase coinciding with the Amazon dry season. Although sampled from sites separated by approximately 1000 km, large-scale atmospheric moisture transport and Rayleigh distillation processes impart a coherent basin-scale climatic signal to both records, allowing wet- and dry-season trends to be evaluated independently.

The two δ¹⁸O chronologies exhibit opposing long-term trends since around 1980, with increasing δ¹⁸O values in the dry-season record and decreasing values in the wet-season record, consistent with an intensification of rainfall seasonality. The talk highlights the specific ecohydrological and phenological conditions that make this type of inference possible, and discusses the distinct sources of uncertainty that can affect interpretation across different records. Key conditions relating to growth seasonality, moisture sourcing, and signal integration must be met in order to draw comparable conclusions from other tree-ring isotope datasets. The talk therefore outlines the potential and common pitfalls associated with applying tree-ring isotope approaches to assess large-scale changes in climate seasonality.