Mapping local and global variability in plant trait distributions.

Area of Science:

• Ecology
• Global Change Biology
• Biogeochemistry

Background:

• Understanding ecosystem responses to environmental change requires quantifying vegetation functional diversity.
• Current Earth system models simplify plant diversity using global plant functional types (PFTs), collapsing trait variation into single mean values.
• This simplification limits the accurate representation of global vegetation functional diversity.

Purpose of the Study:

• To develop fine-grained global maps of plant trait distributions for improved Earth system modeling.
• To quantify trait variation within and among global grid cells, moving beyond simplified PFTs.
• To assess the impact of different PFT definitions on trait distribution modeling.

Main Methods:

• Utilized the largest global plant trait database combined with advanced Bayesian spatial statistical modeling.
• Mapped distributions of key plant traits (specific leaf area, leaf nitrogen, and phosphorus) across over 50,000 grid cells.
• Compared model predictions using no PFTs versus 4 or 14 PFTs, validating against out-of-sample data.

Main Results:

• Generated high-resolution global maps of plant trait distributions, preserving variability across scales.
• Revealed that the most diverse grid cells exhibit trait variability comparable to the range of global PFT means.
• Demonstrated the effectiveness of Bayesian spatial modeling for capturing fine-grained trait variation.

Conclusions:

• The new trait maps provide a more nuanced representation of global vegetation functional diversity for Earth system models.
• Modern Bayesian modeling and extensive databases significantly advance plant trait mapping capabilities.
• Accurate quantification of trait variability is essential for predicting ecosystem functions in a changing world.