Effects of Connectivity and Spatial Resolution of Analyses on Conservation Prioritization across Large Extents | Efectos de la Conectividad y la Resolución Espacial de Análisis sobre Priorización de la Conservación en Grandes Extensiones

The outcome of analyses that prioritize locations for conservation on the basis of distributions of species, land cover, or other elements is influenced by the spatial resolution of data used in the analyses. We explored the influence of data resolution on prioritization of Finnish forests with Zonation, a software program that ranks the priority of cells in a landscape for conservation. We used data on the distribution of different forest types that were aggregated to nine different resolutions ranging from 0.1 × 0.1 km to 25.6 × 25.6 km. We analyzed data at each resolution with two variants of Zonation that had different criteria for prioritization, with and without accounting for connectivity and with and without adjustment for the effect on the analysis of edges between areas at the project boundary and adjacent areas for which data do not exist. Spatial overlap of the 10% of cells ranked most highly when data were analyzed at different resolutions varied approximately from 15% to 60% and was greatest among analyses with similar resolutions. Inclusion of connectivity or edge adjustment changed the location of areas that were prioritized for conservation. Even though different locations received high priority for conservation in analyses with and without accounting for connectivity, accounting for connectivity did not reduce the representation of different forest types. Inclusion of connectivity influenced most the outcome of fine‐resolution analyses because the connectivity extents that we based on dispersal distances of typical forest species were small. When we kept the area set aside for conservation constant, representation of the forest types increased as resolution increased. We do not think it is necessary to avoid use of high‐resolution data in spatial conservation prioritization. Our results show that large extent, fine‐resolution analyses are computationally feasible, and we suggest they can give more flexibility to implementation of well‐connected reserve networks.