No abstract

Experimental restoration of Halodule wrightii (shoalgrass) to its former range on Galveston Island, Texas, began in April 1994. We tested the effects of site, planting density, water depth, and fertilizer addition on survival and growth through June 1996. Temperature, salinity, light transmittance, turbidity, and sediment properties at two restoration sites, Redfish Cove and Snake Island Cove, were similar to those in naturally occurring grassbeds in nearby Christmas Bay. Halodule survival, coverage, and new shoot densities were affected by site (significantly higher at Redfish Cove than at Snake Island Cove, which eventually failed), by planting density (significantly higher when planted on 0.25-m or 0.5-m centers rather than on 1.0-m centers), and by water depth (significantly higher when planted in relatively shallow water). Propagation (spreading from transplant units) was significantly greater from 0.25-m or 0.5-m center plantings but was not consistently affected by site or water depth. Fertilizer enhanced propagation but not survival. After two years, Redfish Cove produced belowground biomass similar to that observed in Christmas Bay, but aboveground biomass remained significantly less. Snake Island Cove plant mortality in September 1995 may have been presaged by low root-rhizome carbohydrate levels observed in October 1994, but causes remain unknown. Further restoration of Halodule to Galveston Bay is possible at selected sites, but structural equivalency will take longer than two growing seasons to achieve.

We compared nekton and benthos densities and community compositions in a natural mixed seagrass bed dominated by Halodule wrightii (shoalgrass) with those found in three shoalgrass transplant sites and adjoining sand habitats in western Galveston Bay, Texas, U.S.A. Quantitative drop traps and cores were used to compare communities up to seven times over 36 months post‐transplant where transplant beds survived. Total densities of fishes, decapods, annelids, benthic crustaceans, and most dominant species were significantly higher in natural seagrass than in transplanted shoalgrass or sand habitats during most sampling periods. On occasion, fish and decapod densities were significantly higher in transplanted shoalgrass than in adjoining sand habitats. No consistent faunal differences were found among transplant sites before two of three sites failed. Taxonomic comparison of community compositions indicated that nekton and benthos communities in natural seagrass beds were usually distinct from those in transplanted beds or sand habitats, which were similar. We conclude that reestablishing a shoalgrass bed that resembles a natural seagrass bed and its faunal communities in the Galveston Bay system will take longer than 3 years, provided that transplants persist.