Skip to main content
SLU publication database (SLUpub)

Research article2009Peer reviewed

Quantifying Habitat Requirements of Tree-Living Species in Fragmented Boreal Forests with Bayesian Methods

Berglund, Hakan; O'Hara, Robert B.; Jonsson, Bengt Gunnar


Quantitative conservation objectives require detailed consideration of the habitat requirements of target species. Tree-living bryophytes, lichens, and fungi are a critical and declining biodiversity component of boreal forests. To understand their requirements, Bayesian methods were used to analyze the relationships between the occurrence of individual species and habitat factors at the tree and the stand scale in a naturally fragmented boreal forest landscape. The importance of unexplained between-stand variation in occurrence of species was estimated, and the ability of derived models to predict species' occurrence was tested. The occurrence of species was affected by quality of individual trees. Furthermore, the relationships between occurrence of species at the tree level and size and shape of stands indicated edge effects, implying that some species were restricted to interior habitats of large, regular stands. Yet for the habitat factors studied, requirements of many species appeared similar. Species occurrence also varied between stands; most of the seemingly suitable trees in some stands were unoccupied. The models captured most variation in species occurrence at tree level. They also successfully accounted for between-stand variation in species occurrence, thus providing realistic simulations of stand-level occupancy of species. Important unexplained between-stand variation in species occurrence warns against a simplified view that only local habitat factors influence species' occurrence. Apparently, similar stands will host populations of different sizes due to historical, spatial, and stochastic factors. Thus, habitat suitability cannot be assessed simply by population sizes, and stands lacking a species may still provide suitable habitat and merit protection.


conservation planning; hierarchical model; logistic regression; random effect; predictability

Published in

Conservation Biology
2009, Volume: 23, number: 5, pages: 1127-1137

    UKÄ Subject classification

    Environmental Sciences

    Publication identifier


    Permanent link to this page (URI)