Leveraging Remote Sensing and Crowd-Sourced Biodiversity Data for Enhanced Plant Functional Trait Mapping

Abstract

High-resolution maps of plant functional traits are essential for understanding terrestrial ecosystem processes, yet their integration into ecosystem models has been hindered by uncertainties and a lack of spatially detailed data. To address this, we developed an approach combining remotely sensed data, global crowd-sourced biodiversity records, and plant trait databases to estimate community-weighted mean values and additional statistical descriptors—standard deviation, skewness, and kurtosis—at a global 1 km resolution. Comparisons with plot-level trait estimates from thousands of sites revealed strong correlations (r  > 0.5) and low relative errors (rME <6% and rRMSE < $11%) for traits including Specific Leaf Area (SLA), Leaf Nitrogen Concentration (LNC), and Leaf Phosphorus Concentration (LPC). Notably, our results reveal a non-Gaussian structure in community trait distributions over large areas, suggesting potential biases in previous estimates. By providing spatially explicit distributions and their higher-order moments, our findings deliver unprecedented detail for understanding plant functional diversity, improving predictions of biodiversity patterns, species coexistence, and ecosystem functioning.