Abstract:

We studied the persistent soil seed bank of an Uruguayan grassland under two grazing treatments: a continuously grazed area at moderate stocking rate, and an area excluded from domestic herbivores for 9 years. We also compared the composition of the soil seed bank with the aboveground vegetation using previous records of species frequency on the established vegetation. For the soil seed bank estimates, we used the seedling-emergence method. The seedlings were classified into four functional groups defined for the established vegetation: cool-season (C3) grasses, warm-season (C4) grasses, graminoids (sedges and rushes), and dicots. Under grazing, the soil seed bank density was significantly higher than in the exclosure and dominated by dicots (61% of the total seeds). Moreover, C3 and C4 grasses were scarcely registered (4 and 5% respectively). In the ungrazed area, the four groups were evenly represented. The level of similarity between the established vegetation and the soil seed bank in the grazed area was lower than in the exclosure. The differences between the soil seed banks may be explained by the differences in the aboveground species composition. The grazed area is dominated by C4 prostrate grasses which spread by means of rhizomes and stolons, and show a low seed output. A set of interstitial native forbs, that produce large quantities of seeds, constitute a subordinate group of the community. In the exclosure, the dominant species are replaced by erect C3 grasses that produce abundant inflorescences. The absence of herbivores also promoted canopy changes like the accumulation of a dense layer of litter that can act as a filter to the seeds, preventing their incorporation to the soil seed bank. This can explain the smaller size of the soil seed bank in the exclosure. Our results indicate that, in the soil seed bank, the different grazing conditions promote changes that are observable at the level of the functional groups defined for the established vegetation.