How reserve selection is affected by preferences in Swedish boreal forests

It is important to consider the preferences of the various stakeholders involved when evaluating effective reserve selection, since it is largely their preferences that determine which of a given set of potential reserve networks that actually is “the best”. We interviewed eight conservation planners working at the county administrative boards in each of the eight administrative counties covering boreal Sweden to establish weightings for different structural biodiversity indicators by using the Analytic Hierarchy Process (AHP). The subjective weightings were applied in a reserve selection model based on a goal programming (GP) approach. The structural indicators were derived from the Swedish National Forest Inventory (NFI) and used as proxy for biodiversity potential. A biodiversity indicator score, based on the values of those indicators, was maximized. The model adjusted this score ensuring that all indicators were represented in the selection, and further also adjusted the influence of the indicators based on the subjective weightings. We evaluated the GP approach by comparing it to a simple linear programming (LP) formulation, only maximizing the indicator richness. In all cases the model was limited either by a budget or an area. The biodiversity potential in young forests are often neglected within present conservation policies, however, the proportion of selected forest under 15 years was relatively high in all our cost-effective cases, varying between 32% and 60% using the individual planners subjective weightings, compared to 80% when using a simple LP model. The proportion of selected forest over 100 years varied between 69% and 85% in the area-effective cases using the subjective weightings, compared to 80% when using a simple LP model. Middle-aged forest was not favored in any of the selections, although they make up a substantial part of the total area. We conclude that there are differences in how conservation planners prioritize the indicators, and depending on how specific biodiversity indicators are weighted the age distribution of the selected reserves differs. This demonstrates the importance of considering how to establish appropriate weightings. It is also important to consider the, at least in our case, substantial difference in how common the different indicators are to ensure that the weightings get their intended impact on the selections.