1990). Using a gear type selective for small resident macroepibenthic fauna (1 m2 dropnets), we found that in 1.8 yr H. wrightii planted on 0.5 m centers developed an animal density, number of taxa, and species composition equivalent to that found in natural beds. Animals tended to be larger in planted beds over the course of the study. Comparison of planted S. filiforme and mixed H. wrightii and S. fjliforme with natural beds was impaired due to failure of several planting areas but exhibited a pattern of development similar in some ways to that of planted H. wrightii. Macroepibenthic animal density in planted beds displayed an asymptotic relationship with areal shoot density, where animal densities became equal to natural beds at shoot densities only a third of the average density for natural beds. This pattern corroborates the existence of threshold values of habitat structure in seagrass beds influencing numerical abundance of some associated animal communities. Macroepibenthic fauna1 abundance and composition in planted beds could be inferred from the amount of areal coverage of seagrass and its persistence over time, while measurement of areal shoot density may provide an important first check point on the road to functional restoration of seagrass habitat.