Editor’s note: Jeffrey Leis is principal research scientist for ichthyology at the Australian Museum. Our March 2003 edition (MPA News 4:9) featured his remarks on larval dispersal and marine reserves, accompanied by an extended online interview.

By Jeffrey M. Leis

[Note: A full list of the literature cited in the following essay is available.]

MPA networks are a series of reserves that individually may be too small to be self-seeding, but that are close enough together so that one reserve can seed another (Palumbi 2002). A major unknown in planning MPA networks is how far apart the individual components of the network should be separated. Most connectivity between the populations of demersal organisms protected within the individual network components is by dispersal of larvae, so the geographic scale over which larval dispersal takes place in the ocean has become a major concern. It is increasingly obvious that it is inappropriate to assume larvae of fishes and decapod crustaceans (if not other taxa) are passive particles whose dispersal can be understood as a purely physical process applied over the pelagic larval duration (PLD). Dispersal is a much more biological process than was thought only a few years ago: it is now clear that larvae behave (Sponaugle et al. 2002; Leis 2006), and this can greatly influence dispersal outcomes.

However, hard data on how far larvae actually disperse are rare. In addition, dispersal of relevance for genetic connectivity (i.e., evolutionary connectivity) is likely to be over much greater distances than for demographic connectivity (i.e., ecological connectivity) (Leis 2002; Leis 2006). Further, dispersal distance will differ among:

  • Species because of species-specific behaviors and PLDs;
  • Locations because of site-specific differences in hydrography and the interaction of behavior of larvae with hydrography over both small and large scales; and
  • Seasons, years, and other time periods, due to (a) differences in hydrography among seasons and years as well as long-term oceanographic variations and trends, and (b) temperature-dependant rates of development and physiology in the cold-blooded organisms that dominate marine communities.

As a result, it is clear that no single inter-reserve spacing will be suitable for all MPA networks, and one that is suitable from an ecological point of view will probably differ from one suitable from an evolutionary point of view. A fixed inter-component spacing in MPA networks that is suitable for a sponge is unlikely to be suitable for a fish. This lack of clarity on ideal inter-component spacing in MPA networks causes headaches for planners and managers, and can delay the planning and consultation process.

A solution that provides a rule of thumb for inter-component spacing in MPA networks has been hinted at in the literature, but I have not seen it clearly and simply stated. It is this:

All distances between all components of an MPA network (not just adjacent components) should be scaled so that there is an approximately equal number in each distance category (e.g., very short, short, medium, large, very large), and there should not be large size gaps between distance categories.

Put more formally, the frequency distribution of all inter-component distances should be “flat”, or even. Smaller distances might be able to be accommodated within individual components of the MPA network, whereas moderate and large distances would be among the different components of the network. Extremely long dispersal distances, if they are thought to exist, might be accommodated by taking into account more distant MPA networks. The distances must be within water (e.g., not cutting across islands or peninsulas.)

This rule of thumb assumes that the frequency distribution of dispersal distances for all the individual species that the MPA network is supposed to protect, integrated over time and over the space occupied by the network, will be approximately flat, and that MPAs are designed to provide both ecological and evolutionary protection. We will not be able to adequately test the first part of the assumption for some time, but based on current knowledge, it seems to be the best-informed guess we will be able to make within the near future. The advantage of this rule of thumb is that it can be applied now.

This rule of thumb was hinted at by several authors in the seminal Ecological Applications volume on MPAs (13[1] supplement 2003). For example, Roberts et al. (2003) state: “it will be safest to have a range of distances among reserves”. But perhaps the clearest statement is in Steve Palumbi’s report for the Pew Oceans Commission (Palumbi 2002): “Because the models show that these different types of species require reserves with different spacing, the simple conclusion is that reserve networks should have a variety of spacing – from quite low to high – in order to accommodate the whole community.” The rule of thumb does not, however, seem to have reached a wide audience of MPA planners and managers. I have been raising it at conferences for a few years, yet it always seems to be greeted with surprise. This essay is an attempt to fully and clearly state the rule of thumb and introduce it to a wider audience.

My thanks to Bob Warner for pointing out important literature, Alan Jordan for comments, and Sue Bullock for editorial assistance.

For more information:

Jeffrey M. Leis, Ichthyology, Australian Museum, 6 College St, Sydney, NSW 2010, Australia. E-mail: jeff.leis@austmus.gov.au