Biodiversity Conservation Informatics Group

Introduction

Reserve selection algorithms aim at finding a reserve network where amount of biodiversity represented is maximized while minimizing the cost. In particular, the maximum coverage approach aims at maximizing biodiversity within the limits of a fixed budget constraint. In practice, maximum coverage studies have most often concentrated on maximizing the numbers of species represented in the network. Species representation is treated as a threshold – a species is either represented or not depending on whether or not a given target for representation is reached. This approach does not distinguish between total absence and a representation that is only slightly below the target, nor between a representation exactly on the target or ten times above the target. It seems reasonable to assume, that underrepresentation should have some value, and that overrepresentation should have more value since it benefits the species by improving its chances of long-term persistence.

Species are not of equal value in conservation planning. There are numerous criteria that are commonly used for prioritizing species for conservation action. For example, some species are more threatened than others, and to be able to persist in the long term, they are more in need of protection compared to common and non-threatened species. Another typical basis for prioritization is that species are not of equal value regarding biological diversity pattern: Phylogenetically distinct species possess more unique features and thus contribute more to biological diversity than species that have many similar, close relatives.

RSW2 introduces the use of continuous benefit functions for species representation, where the value of a species in a network increases with increasing representation: the more populations there are, the better. It further employs species specific weights, giving priority to endangered or phylogenetically distinct species. Use of weights and benefit functions improves the chances for these species of becoming protected, and helps in finding a solution where all the important species are adequately represented.


For vascular plant species in Finnish herb rich forests, using a continuous benefit function with species weights (black bars) resulted in much higher levels of protection for the high priority species as compared with using fixed targets and considering all species equal (grey bars). Note that the higher protection levels were achieved with exactly the same limited budget.

RSW2 allows for one to use a refinement for site size based on the species-area relationship. Usually reserve selection algorithms aiming at efficient resource use tend to select small sites due to their lower cost and collectively higher species richness. This is not the best strategy if the aim is to improve long-term persistence of species: Local extinctions are more likely at smaller sites.

Another available feature is the possibility to perform a "replacement cost" analysis. This feature uses a new definition of irreplaceability based on the benefit function approach: how much does including or excluding a specific site affect the value of the solution. The analysis produces replacement cost values for the selected sites.

The latest version also includes an option for sequential selection: This approach combines irreplaceability and vulnerability in a quantitative manner through the use of replacement cost and habitat loss rates within sequential site selection. Essentially, the analysis considers to what extent is there time to replace lost conservation value from elsewhere before replacement options become lost or degraded, and produces an ordered sequence of sites in which they should be protected.

Note that the current version of RSW2 is non-spatial optimisation, meaning that the spatial pattern of selected sites has no effect on perceived solution value.