1196  2002 Estuarine Research Federation Estuaries Vol. 25, No. 6A, p. 1196–1204 December 2002 Seed Bank Patterns in Chesapeake Bay Eelgrass (Zostera marina L.): A Bay-wide Perspective MATTHEW C. HARWELL * and ROBERT J. ORTH Virginia Institute of Marine Science, College of William and Mary, Gloucester Point, Virginia 23062 ABSTRACT: The decline of eelgrass (Zostera marina) in Chesapeake Bay in the 1960s and 1970s has been studied in the context of changes in water quality and habitat suitability; little effort has focused on the importance of reproductive ecology in understanding current and potential recovery of these populations. The spatial variability of seed-bank char- acteristics of Z. marina in Chesapeake Bay was explored by a reproductive shoot and seed-bank sampling effort. Seed banks were sampled from 105 beds of submerged aquatic vegetation among 12 zones throughout the lower and middle Chesapeake Bay. Number of viable seeds was highly variable among and within zones, with seeds found in all but one zone and also found in cores not containing any Z. marina shoots. Number of reproductive shoots was also highly variable among and within zones, with differences probably driven by different local environmental conditions. Bay-wide, viable seeds were found in more monospecific Z. marina cores than in mixed species or monospecific Ruppia maritima cores suggesting local biological and environmental control on sexual reproduction. Lower densities of viable seeds in the middle Chesapeake Bay region reflect the lower abundance of Z. marina in these regions and provide context for discussion of historical changes in Z. marina in Chesapeake Bay. While this study focused on a snap shot of the seed bank immediately after establishment, we highlight critical questions for future study that may be important for their conservation and restoration. Introduction Seed banks, present in almost all areas where higher plants are found, are important for the pop- ulation dynamics of a particular species by offering a measure of insurance for plant persistence fol- lowing periods of disturbance (Fenner 1995). The literature is replete with studies on seed banks for many angiosperms, but little is known about seed banks in the seagrass group (McMillan 1991; Har- rison 1993; Inglis 2000b; Orth et al. 2000), which occur on every continent except Antarctica (den Hartog 1970). Because seagrasses, with reported seed bank densities similar to most plant commu- nities, are being threatened worldwide by anthro- pogenic influences, knowledge of seed banks will be vital for their conservation and restoration (In- glis 2000b; Orth et al. 2000). Even transient seed banks (those persistent for less than one year; Bas- kin and Baskin 1998) can be critical for recovery of populations after intra-annual disturbances (e.g., summer-time cownose ray [Rhinoptera bona- sus] foraging in Chesapeake Bay seagrass beds; Orth 1975). Zostera marina L. (eelgrass), the most common temperate seagrass throughout the Northern * Corresponding author; current address: A.R.M. Loxahatch- ee National Wildlife Refuge, 10216 Lee Road, Boynton Beach, Florida 33437; tele: 561/732-3684; e-mail: matthewharwell@ fws.gov. Hemisphere (den Hartog 1970), is found through- out the lower and middle Chesapeake Bay (Orth and Moore 1988) and is characterized by sexual reproduction ranging up to 11% to 19% (303–424 reproductive shoots m -2 , respectively) of total shoot density, and a seed production of 23 seeds per reproductive shoot (17–30 seeds per shoot; Sil- berhorn et al. 1983). Sexual reproduction in Ches- apeake Bay populations is initiated in late winter, culminating in seed release by early summer ( June; Silberhorn et al. 1983; Orth et al. 1994). Although seeds are negatively buoyant and do not disperse far when released at the sediment surface (Orth et al. 1994), seeds can also be dispersed by floating at the surface (when released from the parent plant) via gas bubbles (Churchill et al. 1985) or by rafting of fragmented reproductive shoots (Harwell and Orth in press and references within). The seed bank of Z. marina is transient, with seed viability demonstrated, under laboratory conditions, no longer than 11 mo after release (Harrison 1991; Moore et al. 1993; but see Chur- chill 1983). Other studies have demonstrated or suggested longer seed dormancy in Z. marina as well as in other seagrass species (e.g., McMillan 1988a,b, 1991; Harrison 1993; Conacher et al. 1994; Brenchley and Probert 1998); caution should be exercised in applying laboratory-deter- mined dormancy periods to field conditions when seeds are bleach cleaned and stored in sterilized