Impacts of forest fragmentation on the mating system and genetic diversity of white spruce (Picea glauca) at the landscape level.

We studied the mating system of white spruce (Picea glauca) in a landscape fragmented by agriculture in northern Ontario, Canada. We sampled 23 stands that ranged in size from 1 to >500 trees isolated by 250-3000 m from the nearest other stand. Six polymorphic allozyme loci from four enzyme systems were used to genotype approximately 10 000 embryos from 104 families. We detected no allele frequency heterogeneity in the pollen pool among stands or families (Phi(FT)=-0.025). Overall, estimates of outcrossing were high (t(m)=94% and mean t(s)=91%) but significantly different from unity. Bi-parental inbreeding (t(m)-t(s)=3.2%) was low but significantly different from zero. Allozyme-based outcrossing estimates did not differ significantly among three stand-size classes (SSCs): small (<10 trees), medium (10-100 trees) and large (> or =100 trees). The number of effective pollen donors was high in all SSCs, but was significantly lower in small stands (N(ep)=62.5) than in medium-sized and large stands (both N(ep)=143). The primary selfing rate was significantly higher in medium stands than in large stands. We found no significant difference in genetic diversity measures in the filial (seed) population among SSCs. Overall, these results indicate that white spruce stands in this fragmented landscape are resistant to genetic diversity losses, primarily through high pollen-mediated gene-flow and early selection against inbred embryos. We discuss the importance of using seed data, in conjunction with genetic data, to evaluate the impacts of fragmentation on natural populations.