Distribution and dynamics of intertidal macrobenthos predicted from remote sensing: Response to microphytobenthos and environment

We investigated which variables, including environmental variables and food availability, could predict the spatial distribution and dynamics of benthic macrofauna on an intertidal flat. A time series of macrobenthos and sediment grain size samples was complemented by time series of microphytobenthos and saltmarsh vegetation biomass and sediment grain size from airborne hyperspectral remote sensing, and elevation from laser altimetry. Response models were constructed to predict biomass and species richness of macrobenthos as a function of the environmental variables. Total biomass and species richness was best predicted by a combination of microphytobenthos biomass and sediment characteristics as explanatory variables. Deep deposit feeders and surface deposit feeders also responded best to a combination of variables, with deep deposit feeders responding more strongly to sediment grain size and surface deposit feeders responding more strongly to microphytobenthos biomass. The environmental conditions to reach maximum biomass differed for each macrobenthos species. Application of the response models to the remote sensingderived maps of the environmental variables enabled significant predictions of the spatial distribution of macrobenthos biomass, demonstrating the differences in distribution of the macrobenthos species. The. models also revealed the sensitivity of the macrobenthic community to environmental. change. In situ and remote sensing data demonstrated a significant fining of the sediment and a (temporal) increase in average microphytobenthos biomass. Field observations also showed a significant increase in species richness and changes in the relative abundance of species, with a decrease in Bathyporeia pilosa, and an increase in Nereis diversicolor, Pygospio elegans and Heteromastus filiformis. Such changes in macrobenthos biomass and species richness were indeed predicted from the response models. The study demonstrates that the synoptic remote sensing techniques combined with field sampling allow efficient ecological mapping and monitoring.