Abstract
The phenomenon of plants migrating more rapidly than predicted from simple observations of actual seed dispersal from the parent plant (e.g., Reid's Paradox of Rapid Plant Migration) has generally been discounted for seagrasses. Previous knowledge of the general seed ecology of the clonal seagrass Zostera marina L. (eelgrass) suggests that sexual reproduction is not very important to the population dynamics of seagrass populations; however, researchers have hypothesized long-distance dispersal for nearly a century. In Chesapeake Bay, the distribution of eelgrass is radically different than 70 years ago because of the wasting disease of the 1930's and estuarine eutrophication and high sediment input into the Bay in the 1960's and 1970's. Although some recovery has occurred, many areas remain devoid of eelgrass or are only sparsely vegetated. In this study, I present a combination of observational, experimental, empirical, and theoretical studies, conducted at different scales, to study the reproductive ecology and ecological dispersal mechanisms of eelgrass.

From a bay-wide sampling effort, viable eelgrass seeds in the seed bank were found throughout most of the lower and middle Chesapeake Bay, but abundance of seeds was highly variable. Lower seed-bank densities were found in middle Chesapeake Bay, the region with slow recovery of eelgrass populations. From natural and artificially created eelgrass populations, regional environmental conditions were found to have a greater impact on reproductive shoot (reproductive effort) and seed (reproductive output) production than small-scale influences of location and patch structure.

Detached reproductive shoots of eelgrass (containing viable seeds) held in greenhouse tanks remained buoyant for several weeks before they degraded, sank, and lost all their seeds. In offshore shoal areas suitable for eelgrass growth and survival, seventy percent of tube caps of the polychaete Diopatra cuprea (found throughout the shallow regions of Chesapeake Bay) had fragmented reproductive shoots built into its walls, suggesting a mechanism for seeding these shallow areas. Viable eelgrass seeds were found throughout the shoreline of south Chesapeake Bay, up to 34 km away from the nearest bed. Additionally, a GIS exercise identified new eelgrass patches up to 108 km from the nearest source population.

The use of burlap bags for protecting seeds from predation, burial, or lateral transport maximized germination success over unprotected seeds in the field and provides a new mechanism for restoration efforts. An ecological model of eelgrass reproduction highlighted the potentially significant contribution of seeds to the long-term productivity of eelgrass at different water depths. Exploring theoretical scenarios, the model can be used to predict the total number of seeds produced for one to germinate and successfully establish as a seedling, as well as determine the size of patches, newly created from seeds, based on the number of viable seeds in the seed bank and the vigor of the seedlings that develop.

The results of this research suggest that, when combined with earlier work on seed dispersal ecology of this species, eelgrass has adaptive qualities that make it an excellent colonizer of new habitat and that propagule supply, in general, may not be a limiting factor.