Relationships between soil seed banks and aboveground understory vegetation were examined in dry Douglas-fir (Pseudotsuga menziesii (Beissn.) Franco) forests near Kamloops, British Columbia, to compare the effects of different disturbance types (low and high severity fires and non-salvage logging), and to evaluate how seed banks contribute to post-disturbance vegetation establishment. Sites were selected to represent disturbances at 1, 5, and 10 years prior to sampling, and data were analyzed using nonparametric univariate and multivariate statistical techniques. Seed density and species composition of seed banks did not differ significantly among the 1 year disturbed sites; however, the rooted frequencies of seed-origin plants in the establishing vegetation decreased in the order: lightly burned > severely burned > logged. Seed dispersal within the first year following fire is believed to be an important recovery mechanism on burned sites. Low frequencies of seed-derived plants, particularly conifer seedlings, on logged sites pose concerns for the initial recovery of vegetation following logging; nonetheless, the lack of significant differences in understory species composition between sites burned 5 years prior to sampling and sites logged 10 years prior to sampling, and their corresponding undisturbed sites, suggests that vegetation recovery can occur relatively quickly. Rooted frequencies of plants establishing naturally in burned areas were similar to those in undisturbed areas, which suggests that post-fire erosion control seeding using agronomic species may not be necessary in these forests.

In order to assess the effects of disturbance on soil seed bank spatial structure, variation in seed bank species composition was analyzed at two spatial scales in mature (undisturbed) and recently disturbed stands of Interior Douglas-fir forest in south-central British Columbia. Coarse-scale analysis among stands 10-100 km apart using Multi-Response Permutation Procedures showed that individual mature stands were mainly characterized by unique seed bank species compositions. Disturbance had an homogenizing effect on seed bank species composition at the coarse scale--there was less variability among stands following disturbance by both low and high severity fires, and by non-salvage logging. In contrast, finer scale analysis among sampling units c. 10 m apart showed that heterogeneity in seed bank species composition was greater on severely burned and logged sites, commensurate with greater levels of soil disturbance, than on lightly burned and undisturbed sites. Despite the high intensity of seed bank sampling (sixty 25 cm 2 soil samples from each of 16 sites), species-area curves leveled off only when infrequent species were removed from the dataset. The number of seed bank samples required to account for the common species ranged from 84 on severely burned sites, to 196 on undisturbed sites; more than the 240 samples collected per disturbance class in this study would have been required to account for the remaining infrequent species. Overall, this study highlights the importance of sampling intensively within multiple stands to capture the variation in species composition inherent to these dry coniferous forest soil seed banks.

No abstract

Environmental heterogeneity can create differential opportunities for seedling recruitment among plant species. We collected soil seed banks from alvar habitats in southern Ontario and exposed them to three soil moisture treatments. Density and species richness of germinants were greatest in treatments kept moist compared with treatments where soil was either saturated with standing water at the soil surface or where drought was imposed. Contrary to previous studies, the drought treatment did not stimulate the germination of species that remained ungerminated in other treatments, but did increase the germination of five species that also germinated in wetter soils. Although 12 of the 40 species germinated in only one of the three treatments, overall community composition among watering treatments was relatively consistent; few species showed evidence of differential responses to soil moisture conditions. Variability in soil moisture in this system can alter population and community properties by rarefaction effects, as opposed to niche differences among species.