The Science

In tropical forests, knowing leaf age is a key component of understanding seasonal dynamics in carbon assimilation. However, a robust method for efficiently estimating leaf age across multiple species and environments did not exist. Here, scientists measured leaf age and leaf reflectance spectra and showed that their statistical model was able to predict leaf age across two contrasting forests in Peru and Brazil, and through diverse vertical gradients within the canopy.

The Impact

This study has three important implications for the broader plant science, remote sensing, and modeling communities: (1) It shows that it is possible to monitor and map leaf age of tropical forest canopies and landscape using an imaging spectroscopy approach. (2) In combination with previous spectroscopy work that demonstrated the possibility to retrieve plant functional traits from leaf spectral signatures, this work highlights the possibility to use a spectroscopy approach to reconstruct temporal dynamics of leaf traits (i.e., morphological, physiological, and biochemical). (3) This work enables the retrieval of age-dependent plant functional traits that can be used to parameterize new model structures in future terrestrial biosphere models.

Summary

Leaf age was estimated by tagging developing leaves at budburst and following their development with repeated in situ photo documentations. Scientists assembled 759 leaves from 11 tree species covering four canopy environments in an Amazonian evergreen forest in Brazil (August 2013–August 2014), including canopy sunlit leaves (red, n=4 trees), canopy shade leaves (yellow, n=4), mid-canopy leaves (green, n=3), and understory leaves (blue, n=4). Project results showed that a previously developed spectra-age model for Peruvian sunlit leaves also performed well for independent Brazilian sunlit and shade canopy leaves (R2 = 0.75–0.78), suggesting that canopy leaves (and their  associated  spectra) follow constrained developmental  trajectories even in contrasting forests. The Peruvian model did not perform as well for Brazilian mid-canopy and understory leaves (R2 = 0.27–0.29), because leaves in different environments have distinct traits and trait developmental trajectories. When the team accounted for distinct environment-trait linkages by re-parameterizing the spectra-only model to implicitly capture distinct trait trajectories in different environments, the resulting, more general, model was able to predict leaf age across diverse forests and canopy environments.

Principal Investigator

Jin Wu
Brookhaven National Laboratory
jinwu@bnl.gov

Program Manager

Daniel Stover
U.S. Department of Energy, Biological and Environmental Research (SC-33)
Environmental System Science
daniel.stover@science.doe.gov

Funding

J. Wu and S.P. Serbin were supported in part by the Next-Generation Ecosystem Experiments (NGEE)–Tropics project. The NGEE-Tropics project is supported by the Office of Biological and Environmental Research within the U.S. Department of Energy Office of Science.

References