Modelling the spatial heterogeneity of ecological processes in an intertidal estuarine bay: dynamic interactions between bivalves and phytoplankton

摘要

Spatial patterns in ecological communities result from a combination of physical and biological factors. In an estuarine intertidal bay, spatial differences have been found in the structure of phytoplanktonic communities and in the biological performance of cultivated oysters. It has been hypothesised that trophic heterogeneity exists, although the mechanisms controlling it remain undefined. Spatial and temporal interactions in the structure of phytoplanktonic biomass and in the biological performance of cultivated oysters were highlighted in this estuarine intertidal bay using a 2-dimensional hydrodynamic model coupled to a nutrient-phytoplankton-zooplankton bivalve food web model. The coupled models allowed a reproduction in space and time of variations in the main variables (i.e. nutrients, chlorophyll a (chl a) and bivalve growth and reproduction). Spatial patterns of chl a at the bay scale showed a dichotomy between the eastern and western parts of the bay, with a sharp drop in concentrations above the oyster area. At the smallest scale, significant spatial heterogeneity was obtained in terms of oyster dry weight (DW), with a difference of around 3.0 g between the lowest and the highest oyster DW. Influences of physical and biological factors were discriminated for spatial patterns of phytoplankton at a large scale and for spatial patterns of bivalves at a small scale. Bivalve density, immersion time (i.e. feeding time) and current velocity were identified as the main factors controlling the spatial patterns of phytoplankton and bivalve growth. The results of the model indicate that the effects of spatial scales are much larger than those of temporal scales; this conclusion differs from that expected through observations only. Top-down effects of oysters on phytoplankton biomass at local scales were revealed, whereas bottom-up effects drove primary productivity at the whole bay scale. In general, we conclude that spatial modelling is particularly appropriate to reveal spatial properties which would be difficult to observe directly. Knowledge of ecosystem functioning would be enhanced accordingly.