Incorporating ontogenetic dispersal, ecological processes and conservation zoning into reserve design

Our article in Biological Conservation has become available online today. As promised I shall attempt to write a readable summary.

The aim of our paper is to extend existing methods for placing marine reserves to protect habitats by making the methods more realistic and thus hopefully more likely to protect species of importance and be acceptable to fishermen.

Marine reserves are basically areas of the ocean that are closed to extractive activities like fishing and the removal of stone or pharmaceutical products. Selecting locations for reserves is a tricky business as it's difficult to predict how species and habitats will respond to reserves and even more difficult to convince local stakeholders that going another few miles every day to a potentially less-stocked fishing area is a good thing. For example, closing areas to fishing groupers can benefit parrotfish - an ecologically-important family of fish because parrotfish are stupid enough to go and sit in the grouper fish traps (even though they're not targeted by the fishermen). However, closing areas to fishing groupers allows groupers to grow to a larger size before they die and since groupers eat parrotfish that means that they're able to eat the larger parrotfish (the ones more likely to reproduce).

Rather than choosing sites arbitrarily, computer algorithms (lists of instructions followed by the computer) exist to help select locations. These are required because once you start selecting locations from hundreds or thousands of sites it becomes impossible to do it by hand. Basically you divide the seascape up into lots and lots of potential reserve sites - ours were square and made by overlaying a 1km by 1km grid over an aerial map of Belize, and specify roughly what proportion of sites you want to be in your final network of marine reserves (say 20% of the total area of the seascape). The computer selects a random group of sites as your first reserve network and then repeatedly tries making changes to the reserve network by adding or removing a site and seeing if the resulting reserves are 'better'. Better is usually defined as being more likely to protect the species or habitats of interest with minimal cost to fishermen and reserve managers.

Ian and I wrote a new computer algorithm to make this process more realistic by

1) taking into account that some species migrate between habitats (that's the ontogenetic dispersal part),
2) evaluating how important each fish species is ecologically and economically and then using predicted numbers of each species in each habitat to determine how much of each habitat you aim to have in reserves (that's the ecological processes part) and
3) taking into account that areas outside reserves are also important - previous algorithms assessed if a reserve network was 'better' only by looking at what was in the reserve network. We also evaluate what is outside the reserves at the same time (that's the conservation zoning part).

So now you know!!