Modeling the trophic effects of marine protected area zoning policies: A case study

Marine protected areas (MPAs) are increasingly being recognized as an alternative management tool for conserving marine resources and ecosystems. By integrating organism dispersal rates, ecosystem interactions and fishing effort dynamics, ECOSPACE, a spatially explicit ecosystem-based modeling tool, allowed us to compare the ecological consequences of alternative MPA zoning policies within the proposed Gwaii Haanas National Marine Conservation Area, located off the west coast of British Columbia, Canada. The desired effects of MPAs include higher fishery yields, the conservation of biodiversity, and/or the preservation of intact ecosystems. However, ECOSPACE predicts that when MPAs are small, species interactions and movements may make these objectives difficult to achieve. ECOSPACE suggests that the effects of MPAs are reduced at their boundaries where fishing effort is predicted to concentrate. Furthermore, top predators may become more abundant within MPAs, which could lead to a depression of their prey species and a subsequent increase of species at even lower trophic levels. Trophic cascade patterns and density gradients across boundaries are nontrivial departures from our simple expectations of how MPAs protect areas and will force us to reconsider what constitutes effective conservation. Our ECOSPACE model indicates that the establishment of multi-use buffer zones may help alleviate these realistic but worrisome ecological predictions. When coupled with an overall reduction in harvest pressure, ECOSPACE suggests that a MPA with a large core `no-take' zone and large buffer will result in the greatest increase in organism biomass. The use of marine zoning may be an effective management tactic to reduce social conflict and conserve marine ecosystems.     

A model for simple cells as optimal edge detectors

Simple cells have often been characterised as edge detectors. This paper proposes a specific neural wiring for simple cells that operate as optimal edge detectors. The proposed simple cell is also selective for the edge contrast. Responses of this model simple cell to edges, bars, sinusoids, and flashed lights are simulated and are similar to real nondirectional simple cell responses.