Evaluating metapopulation source strength in a multispecies context for marine protected area planning

Background/Question/Methods

The idea that some patches within a marine metapopulation can act as sources is key to marine spatial management. Ecological theory dictates that sources should be protected from harvest in order to ensure population persistence and higher population densities. However, the somewhat nebulous "source" concept actually encompasses a number of distinct and potentially unrelated dynamic processes. I describe a suite of potential source definitions developed in order to guide the design of marine protected area networks in California. These metrics were calculated for each of six nearshore fish species using a spatially explicit metapopulation model that included information on habitat quality and larval connectivity patterns. Using that system as an example, I investigated how the alternative source metrics for each proposed MPAs were related, the factors affecting source status, and how those factors varied among species.

Results/Conclusions

I considered 3 source metrics: 1) self persistence (degree to which the MPA can persist over time in isolation), 2) deletion index (relative change in mean metapopulation biomass if an MPA is removed), and 3) larval production index (rate at which larvae are successfully exported to elsewhere in the metapopulation). For the modeled California MPAs, self-persistence and the deletion index were both correlated with total biomass in the MPA, and both increased with the size and amount of suitable habitat in the MPA. Larval production, by contrast, was not related to habitat quality or biomass within the MPA, apparently because ocean currents caused larvae produced in many otherwise successful  MPAs to simply be lost from the system. 

The source value of a particular MPA varied among species and depended on the spatial scale and directionality of larval dispersal, the scale of adult fish movement, and the level of harvest outside MPA boundaries. Together these factors make it difficult to assign an overall, multispecies source value to a particular MPA.

In terms of the entire metapopulation, networks of MPAs with high average deletion and larval production indices had higher total biomass and total fishery yield, whereas self-persistence was not a good predictor of overall metapopulation performance. These results emphasize the need to clarify the role a "source" is intended to play in a management context.