Soil seed banks are the product of disturbance events over time and provide a history of site colonization. They also enable us to predict the future site vegetation composition. Overbrowsing by deer and non-sustainable forest harvesting in some eastern forests have resulted in species-depauperate understories incapable of regeneration and with depeted seed banks. Although many studies show a lack of correspondence between the existing vegetation and the seed bank, many non-weedy (including non-recalcitrant, shade-tolerant species) should be present and have been documented in more pristine forests.
We compared the soil seed bank composition of sites that were (1) disturbed within the past 45 years and invaded by an exotic tree (Ailanthus altissima -- AA), (2) disturbed within the past 45 years and dominated by a native tree (Robinia pseudoacacia -- RP), (3) disturbed within the past 45 years and containing both AA and RP, and (4) not disturbed in the past 90 years and composed of a mixture of native trees. Two WV locations where these four forest types were found in close proximity to each other were selected. We sampled soil under 10 AA trees at each AA site, 10 RP trees at each RP site, 5 AA and 5 RP trees at the sites dominated by both species, and 10 mixed native trees at the control sites. Significant differences among the different seed bank species compositions were determined using nonmetric multidimensional scaling and one and two-way permutation-based non-parametric multivariate analysis of variance.
We found that the sites with both AA and RP had seed banks with the most species and the control seed banks had the fewest species. The sites with both AA and RP had seed banks with the highest density of individuals, while the control sites had the lowest density. The AA-dominated sites had the highest exotic:native species ratio, while the RP-dominated sites had the lowest. Rather disconcerting, however, was the fact that none of the seed banks differed significantly compositionally. Indeed, the mature forest seed banks contained about the same number of AA individuals as the AA-dominated sites. If ever cut, the mature forests appear likely to follow a successional trajectory similar to that of the AA sites. Seed banks can provide a positive or negative legacy that helps define different successional trajectories, the two extremes of which include those that lead to a healthy forest and those that lead to an invasional meltdown.