Leveraging remote sensing and crowd-sourced biodiversity data for enhanced plant functional trait mapping

Area of Science:

• Ecology
• Remote Sensing
• Biodiversity Science

Background:

• High-resolution maps of plant functional traits are vital for ecosystem process understanding.
• Integration into ecosystem models is limited by data scarcity and uncertainty.

Purpose of the Study:

• To map global community trait distributions and their moments at 1-km resolution.
• To assess the predictive skill of trait maps against plot-level data.
• To provide spatially explicit trait data for ecological modeling.

Main Methods:

• Combined optical remote sensing, global crowd-sourced biodiversity records, and plant trait databases.
• Estimated community-weighted means (CWMs) and higher-order moments for specific leaf area (SLA), leaf nitrogen (LNC), and leaf phosphorus (LPC).
• Benchmarked model outputs against sPlotOpen plot-level CWMs and canopy-weighted comparators (TWM).

Main Results:

• Limited explained variance (R² = 0.10–0.27) when benchmarking against plot-level CWMs, indicating scale mismatches.
• Increased agreement (R² = 0.22–0.38) with canopy-weighted comparators, aligning with optical sensor sensitivity.
• Provided spatially explicit distributions of trait moments, enhancing landscape-scale ecological insights.

Conclusions:

• The study delivers improved detail for understanding biodiversity patterns and ecosystem functioning.
• Findings offer landscape-scale insights into trait-mediated coexistence and enhance ecological modeling.
• This work provides a foundation for assessing global environmental change impacts on plant functional diversity and ecosystem resilience.