Low functional but large soil-related variations in growth trajectories characterise the widespread Neotropical tree Cecropia obtusa Trecul (Urticaceae)

Environment-driven interspecific functional variability has been convincingly described across Amazon lowland tree species. However, functional variability has rarely been addressed at the intraspecific level, especially through one key environmental driver such as soil composition. In this study, we assess whether patterns of soil-dependent interspecific variability are retained at the intraspecific level. We examined phenotypic variability for 16 functional leaf (dimensions, nutrient, chlorophyll) and wood traits (density) across two soil types, Ferralitic Soil (FS) vs. White Sands (WS), and on two sites for 70 adult trees of Cecropia obtusa Trecul (Urticaceae) in French Guiana. Cecropia is a widespread pioneer Neotropical genus that generally dominates forest early successional stages. We also examined the architectural development through a temporal-scaled retrospective analysis and lifespan-level growth trajectories, in order to evaluate the role of soil phenotypic variability. Cecropia's unique features, such as simple architecture and growth markers allowed us detailed description of temporal-scaled retrospective analysis of development. Functional trait responses to soil types were weak, as only two traits, namely petiole length and leaf area, exhibited significant differences between the two soil types. Soil effects were stronger on growth trajectories and tree architectural development, with WS trees having the slowest growth trajectories -underlined by smaller internode length-, smallest trunk heights and diameters for a given age, and less numerous branches. Soil-related functional traits did not mirror the divergence usually found at the interspecific level. By integrating a dynamic approach based on retrospective analysis of architectural development to the functional approach to understand tree ecology, here, we demonstrate how an improved understanding of environmental effects on tree phenotypic variance can be captured.