Are compound leaves an adaptation to seasonal drought or to rapid growth? Evidence from the Amazon rain forest

Aim: To assess the hypotheses that compound leaves of trees in the Amazon forest are an adaptation to drought and/or rapid growth. Location: Amazon rain forest, South America. Methods: Genera from 137 permanent forest plots spread across Amazonia were classified into those with compound leaves and those with simple leaves. Metrics of compound leaf prevalence were then calculated for each plot and regression models that accounted for spatial autocorrelation were used to identify associations between climate variables and compound leaf structure. We also tested for associations between compound leaf structure and a variety of ecological variables related to life history and growth strategies, including wood density, annual increase in diameter and maximum height. Results: One plant family, Fabaceae, accounts for 53% of compound-leaved individuals in the dataset, and has a geographical distribution strongly centred on north-east Amazonia. On exclusion of Fabaceae from the analysis we found no significant support for the seasonal drought hypothesis. However, we found evidence supporting the rapid growth hypothesis, with possession of compound leaves being associated with faster diameter growth rates and lower wood densities. Main conclusion: This study provides evidence that possession of compound leaves constitutes one of a suite of traits and life-history strategies that promote rapid growth in rain forest trees. Our findings highlight the importance of carefully considering the geographical distribution of dominant taxa and spatial clustering of data points when inferring ecological causation from environment–trait associations