Designing nature reserves in the face of uncertainty

Michael Mccarthy,Colin Thompson,Hugh Possingham

doi:10.1038/npre.2009.3387.1

Michael Mccarthy, Colin Thompson + Show 1 more

Open Access

https://doi.org/10.1038/npre.2009.3387.1

Copy DOI

Abstract

AbstractConservation reserves are one of the most important tools for managing biodiversity. Ever since Diamond, based on theory of island biogeography, proposed that a single large reserve was preferable to several small reserves of the same total area, there has been an enduring debate about the veracity of his assertion. The so-called SLOSS debate - should we have a Single Large reserve Or Several Small reserves - features in every conservation text book and is central to conservation theory. Population dynamic models suggest that the design that minimizes the risk of extinction of species is case-specific, with the optimal number of reserves ranging between one and very many. Uncertainty is pervasive in ecology, but, the previous analyses of the SLOSS debate have not considered how uncertainty in the model of extinction risk might influence the optimal design. Here we show that when uncertainty is considered, the SLOSS problem is simplified and driven more by the aspirations of the manager than the population dynamics of the species. For a given budget of land area to be reserved in a region, the optimal solution is to have on the order of twenty or fewer reserves for any species. This result shows counter-intuitively that considering uncertainty actually simplifies rather than complicates decisions about designing nature reserves.

Highlights

Conservation reserves are one of the most important tools for managing biodiversity[1]
Aspects of metapopulation dynamics such as extinction risk of patches are difficult to estimate with precision; bounds on estimated probabilities of extinction can sometimes encompass almost all possible values between zero and one for even well-studied species[18, 19]
We use models of metapopulation dynamics with and without dispersal to show that, when a given budget of land is available for the reserve system, a large number of reserves is never optimal in the presence of uncertainty

Summary

Designing nature reserves in the face of uncertainty

The model without dispersal has two parameters, the probability of extinction of the species within the timeframe of management concern when all the available area for the reserve system is in a single reserve (x1), and the scaling parameter describing how the mean time to local extinction changes with the size of an individual reserve (b, typically in the range 0.5 – 2.522). If we choose a land area budget of 15,000 ha of mountain ash forest to conserve the greater glider (Petauroides volans) in the Central Highlands of Victoria, b = 0.87 and the probability of extinction within 100 years for a single 15,000 ha patch is predicted to be x1 = 0.0027717 In this case, xn is minimized when n = 735, with each reserve being approximately 20 ha in size. In the presence of uncertainty about extinction risk, the optimal number of reserves for single species is on the order of twenty or fewer, and the number is driven primarily by the aspiration of the manager rather than the dynamics of the species

Methods summary

Findings

Methods