Improving the estimation of landscape scale seed dispersal by integrating seedling recruitment

Abstract

Dispersal and recruitment play a critical role in determining the abundance and the spatial structure of regeneration. In this study, we estimated landscape scale seed dispersal using the inverse modeling method by fitting seedling counts to seed tree location and fertility in the European silver fir Abies alba Miller. Seed trees and seedlings from three age cohorts were counted in 30 plots on Mont Ventoux (southeastern France) at elevations from 995 to 1,605 m, where the following ecological variables were measured: elevation, slope, aspect, light availability, and tree density. We developed and fitted a dispersal model, including a new parameter, recruitment rate, which depended both on the ecological characteristics of the plots and on seed production. Elevation was the factor that affected seedling recruitment the most, with an optimum value at around 1,200 m. Estimated A. alba seed dispersal distances were short (median values for the three cohorts, respectively, 16.4–13.2–19.6 m). We demonstrated that including the ecological characteristics of plots and post-dispersal recruitment processes realistically reduced estimates of seed dispersal distances which are otherwise grossly over-estimated using inverse modeling procedures.