Abstract
Habitat trees, which provide roosting, foraging and nesting for multiple taxa, are retained in managed forests to support biodiversity conservation. To what extent their spatial distribution influences provisioning of habitats has rarely been addressed. In this study, we investigated whether abundance and richness of tree-related microhabitats (TreMs) differ between habitat trees in clumped and dispersed distributions and whether the abundance of fifteen groups of TreMs is related to tree distribution patterns. To identify habitat trees, we quantified TreMs in temperate mountain forests of Germany. We determined clumping (the Clark–Evans index), size of the convex hull, diameter at breast height, as well as altitude, slope and aspect of sites for their possible influence on TreMs. We additionally determined the difference in TreM abundance and richness among four options of selecting five habitat trees per ha from 15 candidates: (a) the most clumped trees, (b) five randomly selected and dispersed trees, (c) the single tree with highest abundance or richness of TreMs and its four closest neighbors and (d) a “reference selection” of five trees with known highest abundance or richness of TreMs irrespective of their distribution. The degree of clumping and the size of the convex hull influenced neither the abundance nor richness of TreMs. The reference selection, option (d), contained more than twice the number of TreMs compared to the most clumped, (a), or random distributions, (b), of five habitat trees, while option (c) assumed an intermediate position. If the goal of habitat tree retention is to maximize stand-level abundance and richness of TreMs, then it is clearly more important to select habitat trees irrespective of their spatial pattern.