Effects of seagrass landscape structure, structural complexity and hydrodynamic regime on macro- faunal densities in North Carolina seagrass beds

Seagrass habitats exhibit high structural variability at local (<1 to 10s of m) and landscape (100 to 1000s of m) scales, which is closely linked with physical setting. In this study, we conducted 2 spring and 2 fall field surveys in 1991 and 1992 over a 25 km long portion of Core and Back Sounds, North Carolina, USA, to relate macrofaunal abundance to measures of seagrass landscape structure and associated ecological variables. Independent variables included seagrass bed structure (percent cover and total linear edge), local-scale ecological attributes (shoot density, shoot biomass, percent sediment organic matter) and elements of physical setting (water depth and energy regime as estimated by a relative wave exposure index [REI]). Seagrass beds were composed of eelgrass (Zostera marina: fall/spring-dominant), shoalgrass (Halodule wrightii: summer-dominant), some widgeongrass (Ruppia maritima) and minor amounts of macroalgae. Seagrass coverage ranged from highly patchy (13% cover) dune-like beds to continuous (100% cover) low-relief beds within 18 replicate 50 × 50 m plots. Twelve species (8 decapods and 4 fishes) made up 95% or more of the catch, and densities of nearly all varied significantly between seasons and years. Multiple regression analysis indicated that relationships between faunal densities and environmental variables varied greatly between species and between collection periods. In addition, species-specific correlations between faunal density and environmental variables generally were not consistent among the 4 collection periods. REI and seagrass shoot biomass appeared to have the greatest influence on species’ densities, with REI having more influence on densities in 1991 and shoot biomass having more influence on densities in 1992. Seagrass percent cover and total linear edge explained little of the variation in species’ densities. Only blue crab Callinectes sapidus density was (positively) correlated with seagrass percent cover in the spring of 1992. In principal components analyses, species groupings were inconsistent between collection periods, though grouping by relative abundance was evident in some collection periods. There was little separation between crustaceans and fishes in principal component space, but sites of high faunal abundance were distinct from sites of low faunal abundance. Sites with consistently high faunal abundance generally were found in western Core and Back Sounds, whereas sites with consistently low faunal abundance were found in eastern Core Sound, suggesting that processes operating at larger than landscape spatial scales (e.g. larval delivery by currents) may influence faunal community patterns in these seagrass landscapes. The influence of a variety of covarying factors on fauna operating over a range of spatial scales highlights fundamental differences in the relationship between landscape structure and animal abundance in seagrass versus terrestrial habitats.