Paternity analysis of pollen-mediated gene flow for Fraxinus excelsior L. in a chronically fragmented landscape

Abstract

Paternity analysis based on microsatellite marker genotyping was used to infer contemporary genetic connectivity by pollen of three population remnants of the wind-pollinated, wind-dispersed tree Fraxinus excelsior, in a deforested Scottish landscape. By deterministically accounting for genotyping error and comparing a range of assignment methods, individual-based paternity assignments were used to derive population-level estimates of gene flow. Pollen immigration into a 300ha landscape represents between 43% and 68% of effective pollination, mostly depending on assignment method. Individual male reproductive success is unequal, with 31 of 48 trees fertilising one seed or more, but only three trees fertilising more than ten seeds. Spatial analysis suggests a fat-tailed pollen dispersal curve with 85% of detected pollination occurring within 100m, and 15% spreading between 300m and 1900m from the source. Identification of immigrating pollen sourced from two neighbouring remnants indicates further effective dispersal at 2900m. Pollen exchange among remnants is driven by population size rather than geographic distance, with larger remnants acting predominantly as pollen donors, and smaller remnants as pollen recipients. Enhanced wind dispersal of pollen in a barren landscape ensures that the seed produced within the catchment includes genetic material from a wide geographic area. However, gene flow estimates based on analysis of non-dispersed seeds were shown to underestimate realised gene immigration into the remnants by a factor of two suggesting that predictive landscape conservation requires integrated estimates of post-recruitment gene flow occurring via both pollen and seed.