Background/Question/Methods Despite the fact that the number of seeds produced in most plant communities is typically huge, many communities are seed limited, such that experimentally adding seeds increases adult plant abundance or diversity. Seed limitation may occur because either not enough seeds are available (“pure”); enough seeds are hypothetically available but not all seeds are able to situate in microsites (“apparent”); or not enough seeds are even hypothetically available and, of those that are, not all are able to situate in microsites (“aggravated”). In a seed limited, Minnesota, USA old field, we quantified the number of undispersed late-successional forb and C4 grass seeds m^-2 and, using ambient seed rain, experimentally modified field conditions for establishment with four factors (rainfall amendment, adult plant thinning, litter reduction, and small mammal fencing). We tracked the fates of individual seedlings. After germination slowed in the field, we collected soil seed bank cores.

Results/Conclusions Of the nearly 2600 undispersed seeds m^-2, only 164 seeds m^-2 were present and living after overwintering, as measured by field and seed bank germination. Moreover, only 9 of those 164 seeds m^-2 germinated in the field, even under our four experimental treatments. Together, these suggest that seed limitation was aggravated. Of the seeds that did germinate in the field, all of our four experimental treatments significantly improved at least one stage of establishment, and rainfall amendment had the greatest overall effect across species. Although their adults dominate the field, surprisingly few C4 grass seedlings germinated. We conducted a second experiment the following year by adding seed of the three dominant C4 grasses within the same plots. The ability of the C4 grasses to recruit into control plots compared to treated plots was relatively lower than for the ambiently recruited forbs from our first experiment, suggesting that C4 grasses have greater difficulty establishing in the extant community. This effectively lowers their colonization rate and should theoretically permit more species to coexist with them via the competition-colonization tradeoff. Our results also suggest a mechanism by which C4 grasses come to be locally abundant over the course of succession despite their low viable seed numbers and establishment success in late-successional communities.