Predicting species' maximum dispersal distances from simple plant traits

Area of Science:

• Ecology
• Botany
• Evolutionary Biology

Background:

• Plant dispersal distances are linked to life-history traits.
• The predictive power of different traits for dispersal distances remains unclear.

Purpose of the Study:

• To assess how well various plant traits predict maximum dispersal distances.
• To develop models for estimating dispersal distances using measurable traits.

Main Methods:

• Utilized cross-validation and a global dataset of 576 plant species.
• Modeled maximum dispersal distances using traits like dispersal syndrome, growth form, seed mass, and terminal velocity.
• Compared predictive accuracy of models with varying trait combinations.

Main Results:

• Models incorporating dispersal syndrome, growth form, and terminal velocity achieved the highest predictive power (R2 = 0.60).
• Simpler models using dispersal syndrome, growth form, and taxonomy also provided reasonable predictions (R2 = 0.53).
• More specific traits, like terminal velocity, enhance prediction accuracy but may limit species applicability.

Conclusions:

• Plant traits, particularly dispersal syndrome and growth form, are valuable predictors of maximum dispersal distances.
• A function (dispeRsal) is available in R for researchers to estimate dispersal distances with confidence intervals.
• Accessible trait data allows for broad predictions across numerous plant species.