The herb layer restoration potential of the soil seed bank in an overgrazed oak forest

We investigated the potential contribution of the persistent soil seed bank in post-disturbance restoration of the herb layer in a long-term overgrazed, mixed oak forest (NW Greece). We examined the impacts of grazing on plant richness and density in the soil seed bank in regard to the different dispersal and life strategy types of the herb layer taxa. Soil seed bank was qualitatively and quantitatively analyzed and contrasting plant guilds were defined according to life strategy type and dispersal mode. Soil seed bank differences between a) the upper and lower soil layers and b) plant functional guild pairs (ruderals vs. non ruderals, including typical forest taxa, and physically-vs. animal-dispersed plants) were statistically tested in overgrazed and sporadically grazed plots. Moreover, correlations in soil seed bank species dominance between overgrazed and sporadically grazed plots were examined by Spearman's Rank correlation. The majority of seeds were found in the upper (0-5 cm) soil layer. Seed density in the deeper (5-10 cm) soil layer was rather poor and did not differ significantly between overgrazed and sporadically grazed plots. In the upper soil layer, both seed density and plant species richness were significantly lower in the overgrazed plots. Overgrazing reduced both species richness and seed density of non-ruderal species in general and typical forest herbs in particular, while it did not affect ruderal species richness and density. Plant species richness and seed density of animal-dispersed taxa were reduced by overgrazing while physically-dispersed species were not affected; it is therefore concluded that large herds of grazers fenced in relatively small areas cannot act as efficient dispersal vectors of the former species. Our findings suggest that, upon cessation of grazing, the soil seed bank is rather inadequate to restore the herb layer of overgrazed forest sites.