Benthic macrofauna in Tasmanian estuaries: scales of distribution and relationships with environmental variables

Information on the distribution of species richness, faunal density, biomass and estimated productivity of benthic invertebrates in Tasmanian estuaries was quantified at a variety of spatial and temporal scales to assess general hypothesis relating community metrics to such environmental variables as salinity, seagrass biomass and sediment particle size. An associated aim was to assess appropriate scales of investigation for soft-sediment biota distributed in estuaries, including whether patterns identified at individual sites, estuaries, tidal levels or times are likely to have more general relevance. Faunal biomass and productivity varied principally at between-estuary (10 to 1000 km) and replicate-sample (1 m) scales, indicating that these two community metrics were largely responding to estuary-wide effects, such as nutrient loading, and to microhabitat features, rather than to locality characteristics at intermediate scales such as salinity, anoxia or sediment particle size. By contrast, faunal density showed greater response to tidal height (1 to 100 m) and to factors distributed at the locality scale within estuary (10 km) than to factors between estuary. Both faunal density and species richness in estuaries declined over three- and fivefold ranges down the shore from high water mark to the shallow sublittoral, while estimated productivity and biomass showed highest overall levels at low water mark. The greatest component of variance in species richness was associated with tidal height, with variance then distributed approximately evenly between other spatial scales examined. At the low-tide and shallow subtidal levels, species richness, faunal biomass and estimated productivity were all highly correlated with salinity and biomass of macrophytes, whereas faunal density was highly correlated with biomass of macrophytes only. Relationships between environmental and biological variables examined were poorly defined at high tidal levels. Seasonal plus interannual variance was much lower than spatial variance - a clear indication that sampling effort in studies would generally be better directed across a range of localities than for a single locality to be repeatedly investigated over time. © 2002 Elsevier Science B.V. All rights reserved.