Seed banks and seedling establishment on an experimental productivity gradient

A major cause of the lower species richness of nitrogen enriched grasslands with high productivity is the lower colonization rate of such areas by new species. To determine what role the soil seed bank might play, we performed greenhouse germination tests of soil sampled in experimental plots that had received various rates of annual nitrogen addition for II yr. Species composition of the soil seed bank depended on the rate of N-addition, even though the species richness and the overall density of viable seeds in soil were independent of N-addition. The proportion of the species shared between the seed bank and the above-ground vegetation decreased with N-addition. The density of forb seedlings emerging in field plots was negatively correlated with the nitrogen input, productivity, and litter depth, but the probability of seedling survival from May through August was not different among treatments. Thus, suppression of seed germination, rather than limitation of the availability of seeds in soil or seedling survivorship, was the primary mechanism of lower colonization rates in the high-N plots. Burning in spring decreased litter depth and promoted germination of forbs, especially in no-N control plots. However, germination was still very low in the high-N plots even after fire decreased litter depth to a level similar to that of controls.