Mechanistic analytical models for long-distance seed dispersal by wind

Area of Science:

• Ecology
• Mathematical Biology
• Plant Sciences

Background:

• Estimating seed dispersal is crucial for understanding plant population dynamics.
• Existing models often lack mechanistic underpinnings or require extensive dispersal data for parameterization.

Purpose of the Study:

• Introduce the Wald analytical long-distance dispersal (WALD) model.
• Provide a mechanistic framework for estimating wind-dispersed seed kernels and canopy escape probability.
• Enable parameter estimation from independent factors like wind statistics and seed traits.

Main Methods:

• Developed a two-parameter Wald (inverse Gaussian) distribution by simplifying 3D Lagrangian stochastic approaches.
• Derived model parameters from wind statistics, seed release height, and terminal velocity.
• Validated the model using three empirical dispersal datasets (forest trees, heathland shrubs, grassland forbs).
• Compared WALD with other analytical mechanistic models (tilted Gaussian Plume, advection-diffusion equation).

Main Results:

• WALD's parameters can be estimated independently of dispersal data.
• The model exhibits an asymptotic power-law tail with an exponent of -3/2, consistent with meta-analyses.
• WALD showed the fairest agreement with empirical measurements compared to other tested analytical models.

Conclusions:

• The WALD model offers a simple yet mechanistically informed approach to seed dispersal.
• Analytical mechanistic models like WALD are valuable for large-scale, long-term plant population modeling.
• WALD advances the understanding of wind-driven seed dispersal and canopy escape.