Canopy closure determines arthropod assemblages in microhabitats created by windstorms and salvage logging

Abstract Conifer-dominated forests of the Northern Hemisphere are prone to natural disturbances, such as windstorms, which create canopyopenings by downing single trees to complete stands. The size of a windthrowdetermines how strongly microclimatic conditions are changed. After windstorms, damaged trees are commonly logged to ‘salvage’ economic returns, but effects on biodiversity are often negative. One cause of these negative effects on biodiversity might be the alteration of storm-created microhabitats, for example, as branchesare cutoff and root plates are flipped back into their pits when trunks are removed. The effect of these alterations of microhabitats on biodiversity under open and closed canopies, which represent extremes of microclimatic conditions, remains unclear. To investigate the relative importance of canopyclosure and microhabitats for arthropodassemblages, we created six artificial windthrowslocated under an open canopyand six under a closed canopyby uprooting three spruce trees per plot. We used permutational analysis of variance to test whether cutting branches off uprooted trees and allowing them to rot on the forest flooraffects branch-dwelling saproxylic beetles and whether repositioning upright root plates in their pits affects assemblages of carabids, epigeal spidersand harvestmen. Our final data set comprised 7657 arthropodindividuals belonging to 28 species of saproxylic beetles, 28 species of carabids, 85 species of epigeal spiders, and 8 species of harvestmen. Carabid abundance was higher under a closed canopyand the number of epigeal spiderspecies was higher under an open canopy. Canopyclosure also significantly affected the assemblage composition of saproxylic beetles, carabids and epigeal spiders, with a higher mean moisture affinity of these assemblages on plots under a closed canopy. Arthropodassemblages of root plates and root plate pits did not differ, but the composition of branch-dwelling saproxylic beetle assemblages on cut and uncut branches significantly differed. Overall, assemblages of saproxylic beetles, carabids, epigeal spidersand harvestmen were pre-dominantly determined by canopyclosure and not by alteration of storm-created microhabitats. Nevertheless, entire crowns of downed spruce trees should be retained in salvage-loggedforests to provide this microhabitat for saproxylic beetle assemblages typical for naturally disturbed forests.