Maintaining structural and functional elements of ecosystems are essential in order to preserve biodiversity and ecosystem function. As a means of guiding conservation work, the umbrella species concept was developed. In Sweden, one putative umbrella species, the white-backed woodpecker, has guided conservation and restoration of deciduous forests for two decades.

Here, we evaluate the decadal effects of restoration aimed at the white-backed woodpecker on biodiversity of saproxylic beetles. We compare stands that were restored twelve to twenty-one years ago to non-restored stands and historical white-backed woodpecker habitats acting as restoration target stands.

Restored stands contained higher deciduous deadwood volumes than non-restored stands but lower volumes than restoration target stands. The deadwood in restored stands was concentrated in later decay stages whereas target stand deadwood was more evenly distributed across decay stages.

Restored stands had similar species richness and abundance of most groups of saproxylic beetles compared with non-restored stands while not reaching the levels of restoration target stands. Species assemblages differed among all stand types with restored stands supporting late decay stage and generalist species while target stands supported more deciduous associated and threatened species. 5. Synthesis and applications: We conclude that after one to two decades, restoration improve stand structure and benefit beetle diversity but that target levels are not yet reached. Thus, only partial restoration is achieved. Our results stress that for restoration to be successful both continuous and repeated restoration efforts are needed and that it is important to identify target levels of important habitat characteristics when assessing restoration outcome.

Three IBL2, flight-intercept traps were strung between trees at breast height in a North, Southeast, Southwest pattern based on the centre of each stand, approximately 30-70 m from the stand centre. IBL2 traps are large (base 1 m, height 1m, intercept area 0.3 m3); semi-transparent flight-intercept traps shaped like downward facing triangles (Bell et al., 2015). All flying invertebrates were contained in bottles, filled to one-third with 70% propylene glycol and some detergent, at the bottom of the traps. The traps were equipped with water-diverging modules that prevent rainwater from entering and flooding the bottles. The traps were set out in the first week of June 2021 and collected in mid-October 2021. An expert taxonomist identified all saproxylic beetles to species level. We then categorised beetles based on their affiliation to deadwood, decay stage, tree species, their conservation status and if they were recorded as prey species for the white-backed woodpecker. This resulted in the following categories: a) Saproxylics, b) Saproxylic species of conservation concern, meaning that they have had the conservation status of NT or higher during the last three Swedish red lists (Gärdenfors, 2010; Swedish Species Information Centre, 2015;2020), c) WBW prey species, species pointed out as especially attractive as food for the white-backed woodpecker according to Aulén, (1988), g) Coniferous preferring, h) Deciduous preferring, i) Generalist, with no tree species preference, j) Early, species occurring at early stages of deadwood decay, k) Middle, middle decay species, l) Late, late decay species and m) decay stage generalists, with no known preferences of decay stage. The same species can occur in several groups. Classifications of ecological preferences were based on available literature (Koch, 1992; Hagge et al., 2019; Seibold et al., 2015) and personal communication with taxonomic experts; nomenclature following the Swedish Dyntaxa system (Dyntaxa, n.d.).

Tree species and diameter at breast height (DBH, 1.3m) were recorded for all living trees higher than 1.3 m and > 5 cm in DBH within a 10 m circular sample plot at the centre of each stand. Coarse woody debris (DBH >10 cm, length/height >1.3 m) was measured within one 25 m circular sample plot, at the centre of each stand. We divided all deadwood by type (logs and snags), tree species and decay stage following Gibb et al., (2005) for logs and Jung et al., (1999) for snags. Top- and bottom diameter and length was measured for logs. For snags, we measured diameter in breast height and assessed height. Deadwood volumes for logs and snags up to 6 m height (as these snags were usually broken) were calculated as cylinders. For taller snags (>6 m) we used Brandel, (1990)’s southern Sweden volume functions for pine and spruce with birch being used for all deciduous tree species.