Forest Restoration to Attract a Putative Umbrella Species, the Whitebacked

http://www.diva-portal.org

This is the published version of a paper published in Ecosphere.

Citation for the original published paper (version of record):

Bell, D., Hjältén, J., Nilsson, C., Jørgensen, D., Johansson, T. (2015) Forest restoration to attract a putative umbrella species, the white-backed woodpecker, benefited saproxylic beetles. Ecosphere, 6(12): 278 http://dx.doi.org/10.1890/ES14-00551.1

Access to the published version may require subscription.

N.B. When citing this work, cite the original published paper.

Permanent link to this version: http://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-114589 Forest restoration to attract a putative umbrella species, the white-backed woodpecker, benefited saproxylic beetles 1 1, 2 3 1 DAVID BELL, JOAKIM HJA¨ LTE´ N, CHRISTER NILSSON, DOLLY JøRGENSEN, AND THERESE JOHANSSON

1Department of Wildlife, Fish, and Environmental Studies, Swedish University of Agricultural Sciences, SE-901 83 Umea˚, Sweden 2Landscape Ecology Group, Department of Ecology and Environmental Science, Umea˚ University, SE-901 87 Umea˚, Sweden 3Department of Business Administration, Technology, and Social Sciences, Lulea˚ University of Technology, SE-971 87 Lulea˚, Sweden

Citation: Bell, D., J. Hja¨lte´n, C. Nilsson, D. Jørgensen, and T. Johansson. 2015. Forest restoration to attract a putative umbrella species, the white-backed woodpecker, benefited saproxylic beetles. Ecosphere 6(12):278. http://dx.doi.org/10. 1890/ES14-00551.1

Abstract. Umbrella species are often spatially demanding and have limited ability to adapt to environmental changes induced by human land-use. This makes them vulnerable to human encroachment. In Sweden, broadleaved trees are disadvantaged by forestry, and commercially managed forests are often deprived of dead wood. This has led to a situation where previously widespread top predators in saproxylic food webs, such as the white-backed woodpecker (Dendrocopos leucotos), have become species of conservation concern. The white-backed woodpecker is generally considered an umbrella species, and it has been linked to forests with large volumes of dead wood from broadleaved trees. In recent years, forest stands have been restored for the white-backed woodpecker, but post-treatment evaluations have rarely included other species that also occur in broad-leaved forests (co-occurring species). Many co-occurring species are saproxylic beetles. In this study, we collected saproxylic beetles and environmental data in restored and commercially managed forests to evaluate if habitat restoration for the white-backed woodpecker also benefited other species with similar habitat associations. We found that volumes of coarse woody debris were higher in restored than in commercially managed forests, and that a majority of man-made snags and downed logs were created from birch trees (Betula spp.). Most spruce trees (Picea abies) were extracted during forest restoration, and this opened up the forest canopy, and created stands dominated by broadleaved trees. Many saproxylic beetles were more common in restored forests, and there were significant differences in species composition between treatments. These differences were largely explained by species traits. Effects of sun- exposure were particularly important, but many beneficiary species were also linked to dead wood from broadleaved trees. Red-listed saproxylic beetles showed a similar pattern with more species and individuals in restored sites. The white-backed woodpecker is still critically endangered in Sweden, but important prey species are already responding to forest restoration at the stand level. We recognize that landscape-level improvements will be required to bring the white-backed woodpecker back, but also that the umbrella species concept can provide a useful framework for successful forest restoration as many co-occurring saproxylic beetle species seemingly benefitted from restoration for the white-backed woodpecker.

Key words: beetle; boreal; broadleaved; dead wood; deciduous; forest; natural disturbance; restoration; saproxylic; Sweden; umbrella species; woodpecker.

Received 29 December 2014; revised 9 June 2015; accepted 22 June 2015; published 21 December 2015. Corresponding Editor: T. Roulston. Copyright: Ó 2015 Bell et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. http://creativecommons.org/licenses/by/3.0/ E-mail: [email protected]

v www.esajournals.org 1 December 2015 v Volume 6(12) v Article 278 BELL ET AL.

INTRODUCTION 2005 as part of the Swedish Environmental Protection Agency’s species action plan for the Human land-use has changed forest ecosys- bird (Mild and Stigha¨ll 2005). To restore habitats tems world-wide. Forest management often for the white-backed woodpecker, forest manag- favors production of commercially important ers in Sweden have created dead wood from tree species at the expense of less profitable but broadleaved trees and selectively harvested in many cases ecologically important tree species. spruce trees to open up forests, and to make In Sweden, stand-level volumes of timber have deciduous trees more competitive (Mild and increased with 40–80% since the 1950s (SLU Stigha¨ll 2005, Blicharska et al. 2014). So far, more 2012). The explanation for this is an increased than 10000 ha of forested stands have been production of conifers at the expense of broad- restored to aid white-backed woodpecker popu- leaved trees that are disfavored by modern lations. In spite of these efforts, the white-backed forestry, e.g., during thinning. Commercial for- woodpecker has failed to recover—only nine ests are therefore denser and less permeable to birds were identified in 2012 as living perma- sunlight. This has led to an impoverished fauna nently in Sweden—which could lead to the of species associated with sun-exposed condi- conclusion that the restoration has been unsuc- tions and broad-leaved trees (Ga¨rdenfors 2010). cessful. Broadleaved trees are particularly important in However, the white-backed woodpecker is not the context of biodiversity management since the only species that is disfavored by modern most planted trees are coniferous and species forestry. Species of conservation concern are associated with broadleaves are particularly often limited by substrates or structures that disfavored (Bernes 2011). Broadleaved trees are develop slowly, e.g., dead wood. On the Swedish also disadvantaged when natural disturbance red-list, 25% (1126 species) of all red-listed regimes, such as recurrent wildfires in upland species are saproxylic (Dahlberg and Stokland forests and seasonal floods in riparian environ- 2004). Many wood-inhabiting beetles, including ments, are suppressed or altered (Linder et al. those that are woodpecker prey, are associated 1997, Johansson and Nilsson 2002, Hellberg with sun-exposed substrates (Kaila et al. 1997, 2004). Martikainen 2000, Sverdrup-Thygeson and Ims One species harmed by practices favoring 2002). Thus, although the white-backed wood- conifers over broadleaves is the white-backed pecker is yet to recover in Sweden, we could woodpecker, which is critically endangered in expect other species with similar habitat require- Sweden with less than a handful of breeding ments to benefit from restoration for this species, pairs/year recorded during the last decade and in fact the white backed woodpecker has (Stigha¨ll et al. 2011). The main reason for the been suggested as a putative umbrella species dramatic decline of this species is forest manage- (Angelstam et al. 2003, Roberge et al. 2008, ment and changes in important disturbance Stigha¨ll et al. 2011). regimes, such as flooding and fire. The habitat Umbrella species are wide-ranging species preferred by the white-backed woodpecker is whose requirements include those of many other relatively open forest with broad-leaved trees species (Groom et al. 2006). Results from some and large volumes of dead wood. The white- studies suggest that the conservation of putative backed woodpecker is known for being associat- umbrella species might confer protection of co- ed with edge habitats (Stigha¨ll et al. 2011), post- occurring species (Fleishman et al. 2000, 2001, fire sites (Mild and Stigha¨ll 2005), and areas of Suter et al. 2002, Caro 2003, Kerley et al. 2003) intermediate forest cover (Mikusin´ski and Angel- whereas other do not (Andelman and Fagan stam 2004).The white-backed woodpecker feeds 2000, Caro 2001, Rubinoff 2001). It can be on invertebrates, mostly saproxylic (wood living) difficult to quantify restoration needs at the beetles in dead wood of broad-leaved trees landscape level, but habitat requirements of (Aule´n 1988). umbrella species might provide a mechanism In an effort to save the white-backed wood- for determining tangible targets. In fact, recent pecker in Sweden the largest forest restoration findings suggest that umbrella species can guide project in the country to date was launched in management efforts (Branton and Richardson

v www.esajournals.org 2 December 2015 v Volume 6(12) v Article 278 BELL ET AL.

2014, Sheehan et al. 2014). 2002). This suggests that many saproxylic beetles Only a handful of studies have looked specif- could co-exist with the white-backed woodpeck- ically at spillover effects of ecological restoration er in restored forests that resemble naturally foranumbrellaspecies.IntheUSA,pine- disturbed environments. grassland communities have been restored for For the white-backed woodpecker to be useful the red-cockaded woodpecker (Picoides borealis) as an umbrella species, we would expect other with positive outcomes for other bird species organisms associated with similar forest types to associated with fire-induced environments (Wil- benefit from forest restoration guided by habitat son et al. 1995). Forest management for the requirements of the white-backed woodpecker. cerulean warbler (Setophaga cerulean) has also To be more specific, we hypothesized that (1) been shown to benefit other disturbance-depen- forest restoration for the white-backed wood- dent bird species (Sheehan et al. 2014). Cross- pecker would create open stands with more dead taxon studies are uncommon, and to our knowl- wood from broadleaved trees than commercially edge we are among the first to evaluate the managed forests; (2) saproxylic beetles would effects of ecological restoration for an umbrella respond positively, in terms of species richness species from the perspective of another taxonom- and abundance, to an increase in sun-exposed ic group. In fact, we are only aware of one dead wood from broadleaved trees; (3) forest previous study, i.e., Branton and Richardson restoration would cause changes in the species (2014), and it presented results for an aquatic composition of saproxylic beetles towards com- scenario. Branton and Richardson (2014) showed munities favored by sun-exposure and dead that efforts to restore floodplain ponds for coho wood from broadleaved trees; and (4) restored salmon (Oncorhynchus kisutch) provided benefits sites would support more saproxylic species and for other fish species: abundance of coho salmon individuals of threatened beetles than commer- were, for instance, positively associated with cially managed forests. biomass and abundance of other fish species and vertebrates of conservation concern. METHODS Earlier studies have revealed positive relationships between the occurrence of white-backed Study area woodpecker and species richness of other bird Nine restored and nine commercially managed species (To¨rnblom et al. 2007, Roberge et al. reference sites were included in this study. All 2008), beetles, and cryptogams; many of which study sites were located between the latitudes are threatened or red-listed (Martikainen et al. 59.58 and 59.98 N, and the longitudes 12.08 and 1998, Roberge et al. 2008). Several studies show 13.78 E, in Va¨rmland, southwestern Sweden, in that avian top predators can be efficient umbrella the center of the historical distribution range of species (Martikainen et al. 1998, Sergio et al. the white-backed woodpecker and one of the few 2003, 2004, 2005, 2006, Roberge et al. 2008), but areas in Sweden where the species still occasion- the evidence is not conclusive (Roth and Weber ally occurs (Stigha¨ll et al. 2011). In elevation, the 2008). However, the usefulness of the white- sites ranged from 67 to 246 m.a.s.l. Before backed woodpecker’s habitat requirements for treatment, restored sites resembled commercially guiding restoration efforts is yet to be properly managed sites. This was concluded from data evaluated. We therefore collected saproxylic provided by the Swedish Forest Agency and two beetles and environmental data in restored and forestry companies (Stora Enso and Bergvik commercially managed forests to evaluate spill- Skog), but also from field measurements. Site over effects of stand restoration for the white- selection was largely based on similarities in backed woodpecker. Many wood-inhabiting bee- forest age and stand basal area (Table 1), but tles dwell in coarse woody debris from broad- many more variables were considered (see leaved trees. In fact, birch trees alone support Appendices A and B). Restored sites were more than 400 species (Ehnstro¨m 2011). It is also selectively harvested to remove Norway spruce, clear that many saproxylic beetles are associated between 2 and 12 years ago (mean ¼ 6.7 years). with sun-exposed substrates (Kaila et al. 1997, Dead wood was also created from broadleaved Martikainen 2000, Sverdrup-Thygeson and Ims trees like birch and, rarely, European aspen

v www.esajournals.org 3 December 2015 v Volume 6(12) v Article 278 BELL ET AL.

Table 1. All tree species and their stand basal area (m2/ha) in restored and commercially managed reference forests.

Tree species Mean (m2/ha) SD t df P Direction Alder (Alnus spp.) Reference 2.09 4.92 Restored 1.39 4.00 0.64 15.55 0.53 NS Birch (Betula spp.) Reference 6.80 1.48 Restored 8.64 1.33 2.77 16.00 0.014 Ref , R Bird cherry (Prunus padus) Reference 0.03 0.08 Restored 0.02 0.02 Na Common juniper (Juniperus communis) Reference 0.01 0.03 Restored 0.00 0.00 Na European aspen (Populus tremula) Reference 1.36 1.48 Restored 1.21 1.66 0.04 15.88 0.97 NS Goat willow (Salix caprea) Reference 0.09 0.14 Restored 0.04 0.03 0.95 15.45 0.36 NS Norway maple (Acer platanoides) Reference 0.04 0.07 Restored 0.00 0.00 Na Norway spruce (Picea abies) Reference 13.02 5.14 Restored 1.29 0.59 6.80 8.21 0.0001 Ref . R* Oak (Quercus spp.) Reference 0.06 0.17 Restored 0.03 0.06 Na Rowan (Sorbus aucuparia) Reference 0.05 0.08 Restored 0.22 0.32 1.51 13.63 0.15 NS Scots pine (Pinus sylvestris) Reference 0.69 0.90 Restored 0.45 0.48 0.31 15.18 0.76 NS Notes: NS ¼ non-significant, Na ¼ no statistical test conducted due to low sample size. Ref ¼ reference, R ¼ restored, * denotes P-values significant after sequential Bonferroni adjustment due to repeated independent t-tests.

(Populus tremula). Ring-barking or girdling killed facing triangles (height ¼ 1m,base¼ 1 m). many broadleaved trees, but harvesters were also Collected invertebrates were conserved in bottles used to create high-stumps or snags, i.e., remove filled with propylene glycol and detergent, tree tops approximately 4 m from the forest floor. attached to the bottom of each trap. Sampling All tree tops were deposited at restored sites to bottles were also equipped with vents to divert provide downed logs. On average, restored sites rain water. All flight-intercept traps were strung covered 5.6 ha (SE ¼ 0.77). This can be compared between two trees at breast height, and posi- to an average stand size of 7.4 ha (SE ¼ 1.67) for tioned at least 30 m and 120 degrees apart. The commercially managed sites. Trees were, on first trap at every site was positioned towards the average, 43.4 years old (SE ¼ 5.45) in restored north, and the remaining traps faced south-east sites and 41.8 years old (SE ¼ 4.85) in commer- and south-west, respectively. All saproxylic cially managed sites. beetles were identified by expert taxonomists to the species-level except for specimens of the Invertebrate sampling genus Acrotrichis (family: Ptiliidae). All species Three flight-intercept traps (Polish IBL2-traps, were also divided into categories based on their CHEMIPAN, Warsaw, Poland) were positioned habitat association. Substrate (broadleaved or in the center of every study site, from early June coniferous trees) and microclimatic (sun or until late October in 2012, to catch airborne shade) association were given special attention. saproxylic beetles. Flight-intercept traps were Classifications were made from empirical data semi-transparent and shaped like downward- provided by other researchers (Lindhe et al. 2005;

v www.esajournals.org 4 December 2015 v Volume 6(12) v Article 278 BELL ET AL.

J. Hja¨lte´n et al., unpublished data), but claims were Understory plants of different species require also cross-referenced in scientific databases and varying amounts of nutrients. The forest produc- in books on species traits (Koch 1989, Ehnstro¨m tivity can be estimated by dividing forests into and Axelsson 2002; www.beetlebase.com). When categories based on such differences. In this the scientific literature failed to provide answers, study, we used a method developed by the preferences were listed as unknown. Many Swedish University of Agricultural Sciences and species (39%) were also linked to dead wood the Swedish Forest Agency (Ha¨gglund and from both broadleaved and coniferous trees. Lundmark 2004). The data were collected in Others (12%) thrived in both open and closed three random survey plots per study site. Each forests. circular plot (radius ¼ 10 m) covered an area of 315 m2, and all plant species typical for this forest Stand-level characteristics type were identified within every plot. All Dead wood was surveyed in four random surveys were conducted by the same field transects, 50 m long and 5 m wide, per study site. personnel, and within four days in August 2012. All coarse woody debris objects (dbh 0.1 m and length 1.3 m) were measured within 2.5 m on Statistical analyses both sides of each transect. Objects partly outside Stand-level characteristics were assessed inde- of the surveyed area were measured to the outer pendently in pair-wise comparisons of restored limit, but not beyond the transect. If, by chance, and commercially managed forests. If require- transects continued outside the boundaries of the ments for normality and homoscedasticity were study site the outside area was withdrawn from fulfilled, a Student’s t-test was employed. Other- the total area. Dead wood was divided into lying wise, an unequal variance t-test was used and standing coarse woody debris for each tree (Ruxton 2006). Alpha levels were sequentially species. The volume of lying coarse woody debris Bonferroni adjusted due to repeated independent was calculated with a formula for a truncated t-tests. We examined differences in forest age, circular cone: stand basal area, coarse woody debris volume, and canopy cover. Equivalent analyses were also V ¼ðph=3Þ 3ðr2 þ r r þ r2Þ 1 1 2 2 performed to examine differences in species where h ¼ height or length, r1 ¼ maximum radius, richness and abundance for red-listed species, and r2 ¼ minimum radius. For standing coarse and saproxylic beetles of different substrate woody debris, r1 and r2 were estimated from (broadleaved or coniferous trees) and microcli- average changes in radius/m for lying coarse matic (sun or shade) preferences. All analyses woody debris. were carried out in IBM SPSS Statistics (Version We counted pixels covered by vegetation in 21). digital images to measure the canopy cover. All To explore differences in community structure images were captured with a compact camera between treatments, we performed multivariate (Panasonic DMC-TZ20, resolution 14 Megapixel) analyses in PRIMER 6 (version 6.1.12) and placed 0.1 m above the forest floor. Forest stands PERMANOVAþ (version 1.0.2) by PRIMER-E were photographed between four and 10 times Ltd. Prior to analysis, a Bray-Curtis similarity depending on their size to give an average matrix was calculated for raw data and trans- estimate per study site. The percentages of visible formed (presence-absence) data. Non-metric sky on all the photos were calculated using the multidimensional scaling (nMDS) was used to magic-wand and the pixel-count functions in illustrate differences in dominance structure (raw Photoshop Elements (Ver. 8.0, 2009). Angle count data) and species composition (presence-absence sampling was performed to determine the stand data) between restored and commercially man- basal area (m2/ha) of all tree species in every aged sites. For conclusive statistical tests, how- study site. We used a relascope to assess the ever, we used PERMANOVA in PRIMER. Species situation from fixed positions, randomly allocat- contributions to the observed dissimilarity be- ed across every study site, in scopes of 360 tween treatments were calculated with SIMPER degrees. Study sites were sampled between four in PRIMER for all saproxylic beetles in trans- and 10 times depending on their size. formed and untransformed datasets.

v www.esajournals.org 5 December 2015 v Volume 6(12) v Article 278 BELL ET AL.

Distance-based linear modeling (DistLM) in debris (Table 2). In total, there were no significant PRIMER showed to what extent structural (can- differences between treatments in terms of opy cover, elevation, coarse woody debris vol- standing coarse woody debris (Table 3). ume, productivity, aspect, tilt, and tree assemblage) and temporal (forest age and time Saproxylic beetles—species richness and since restoration) variables associated with the abundance community composition of all saproxylic beetles. Flight-intercept traps caught 4380 saproxylic Variables with less than six observations were beetles, and 292 species were identified. Altogeth- omitted from the analysis. To explore relation- er, 91 species were unique to restored sites, but 59 ships for individual variables, marginal tests were additional species were unique to commercially performed. All variables were subsequently sub- managed sites, and 142 species were found in jected to a step-wise selection procedure (selection both treatments. In total, there were no significant criterion: AICc) in order to develop models. Prior differences between restored and commercially to analysis, environmental variables were plotted managed sites in terms of species richness (F ¼ against each other in a Draftsman plot to control 0.39, t ¼1.54, df ¼ 16, P ¼ 0.14) and abundance (F for collinearity. Pearson correlation analysis was ¼ 0.77, t ¼0.88, df ¼ 16, P ¼ 0.39), but results on used to quantify the relationships. If the correla- species traits showed that microclimatic and tion coefficient was higher than 0.6 one variable substrate preferences mattered. Taxonomic groups was excluded from the analysis. Variables exclud- favored by sun-exposure were more species-rich (t ed were the basal area of Picea abies, Sorbus ¼5.0, df ¼ 9.92, P ¼ 0.001, Fig. 1A) and abundant acuparia and Salix caprea, the volume of lying (t ¼5.03, df¼ 16.0, P ¼ 0.0001, Fig. 1B) in restored Betula coarse woody debris and the productivity forests. There was also a significant difference in class ‘‘Vaccinium myrtillus type’’.Inallstep-wise abundance for species favored by shade; with procedures and marginal tests, P-values were more individuals in commercially managed sites obtained with 999 permutations. Models with a (t ¼ 2.14, df ¼ 13.5, P ¼ 0.05, Fig. 1B). Further DAICc , 2 were considered equal. results showed that taxonomic groups associated with coarse woody debris from broadleaved trees RESULTS were more species-rich in restored forests (F ¼ 0.85, t ¼2.16, df ¼ 16, P , 0.05, Fig. 1C), but Forest structure observed differences in abundance were not Commercially managed sites were dominated significant for such species (F ¼ 2.84, t ¼1.86, by coniferous trees, and had a significantly df ¼ 16, P ¼ 0.08, Fig. 1D). higher stand basal area of Norway spruce than restored sites where spruce had been selectively Dominance structure and species composition harvested (Table 1). But after sequential Bonfer- Species assemblages in restored forests were roni correction of P-values due to multiple significantly different from species assemblages repeated independent tests, the basal area of all in commercially managed forests (Fig. 2), and other tree species did not differ between restored results based on presence-absence data were in and commercially managed sites (Table 1). agreement with results based on counts. More Additionally, there were no significant differenc- than 30% of the variation in the raw data analysis es in forest age between forest types (F ¼ 0.57, t ¼ was explained by 10 species (Table 4). Among 0.23, df ¼ 16, P ¼ 0.82), but restored sites had them, only one, Tomoxia bucephala, was unique to significantly less canopy cover than commercial- restored forests. T. bucephala was also influential ly managed sites (t ¼ 5.0, df ¼ 9.10, P ¼ 0.001). in the analysis based on presence-absence data Dead wood was also created during forest (Table 4). Influential species were often linked, in restoration, and restored sites contained larger terms of their ecology, to stand conditions total volumes of lying coarse woody debris than associated with each treatment (Table 4). Sun- commercially managed sites (Table 2). Most of exposure was particularly important. Abundant the coarse woody debris in restored sites was and widely distributed species in restored forests derived from birch trees. Birch volumes were were generally associated with sun-exposure, in higher in restored sites for lying coarse woody contrast to common species in commercially

v www.esajournals.org 6 December 2015 v Volume 6(12) v Article 278 BELL ET AL.

Table 2. Volume of lying coarse woody debris (m3/ha) per tree species in restored and commercially managed reference forests.

Tree species Mean (m3/ha) SD t df P Direction Alnus spp. Reference 1.23 2.25 Restored 0.39 0.70 Na Betula spp. Reference 1.71 2.09 Restored 14.14 10.97 3.55 13.22 0.003 Ref , R* Juniperus communis Reference 0.02 0.06 Restored 0.00 0.00 Na Picea abies Reference 0.59 0.85 Restored 2.97 3.20 1.76 15.01 0.098 NS Pinus sylvestris Reference 0.19 0.58 Restored 0.18 0.33 Na Populus tremula Reference 0.04 0.11 Restored 0.96 1.63 1.30 12.08 0.22 NS Prunus padus Reference 0.03 0.08 Restored 0.00 0.00 Na Salix caprea Reference 0.18 0.39 Restored 0.13 0.40 Na Sorbus aucuparia Reference 0.00 0.00 Restored 0.49 1.10 Na Total Reference 3.97 3.35 Restored 19.27 12.19 3.63 9.20 0.005 Ref , R* Notes: NS ¼ non-significant, Na ¼ no statistical test conducted due to low sample size, Ref ¼ Reference, R ¼ restored, * denotes P-values significant after sequential Bonferroni adjustment due to repeated independent t-tests. managed forests that were associated with Red-listed species sheltered conditions (shade). Nineteen species on the Swedish red-list (Ga¨rdenfors 2010) were caught in the flight- Environmental drivers of community shifts intercept traps (Table 6). Analyses revealed Distance-based linear modeling showed that significant differences in species richness (F ¼ several independent variables were significantly 0.25, t ¼2.84, df ¼ 16, P ¼ 0.012; Fig. 3A) and correlated with the species-derived, multivariate, abundance between treatments (t ¼2.53, df ¼ data cloud (Appendices A and B), i.e., canopy 16.0, P ¼ 0.022; Fig. 3B); with more species and cover, time since restoration, elevation and broad- individuals in restored sites. Hylis procerulus was leaf grass vegetation type. Each of the mentioned particularly common (with 46 individuals unique variables single-handedly explained more than to restored sites) and explained most of the 10% of the variation. In subsequent, step-wise difference in total abundance (Table 6). Altogeth- procedures to differentiate between models, can- er 10 species were unique to restored forests, opy cover occurred in three out of four models for compared to three species in commercially raw data and in four out of the eight best models managed forests (Table 6). for presence–absence data (models with DAICc , 2; Table 5). Other important variables were time DISCUSSION since restoration, elevation and vegetation type (presence–absence data only). The canopy cover Restoration efforts guided by the white-backed was, however, negatively correlated with the woodpecker created stands significantly different volume of lying coarse woody debris from birch from those impacted by commercial forestry, as and with spruce basal area. predicted in our first hypothesis. Stand-level

v www.esajournals.org 7 December 2015 v Volume 6(12) v Article 278 BELL ET AL.

Table 3. Volume of standing coarse woody debris (m3/ha) per tree species in restored and commercially managed reference forests.

Tree species Mean (m3/ha) SD t df P Direction Alnus spp. Reference 1.01 1.76 Restored 0.70 1.69 0.38 15.98 0.71 NS Betula spp. Reference 2.50 3.62 Restored 9.59 10.21 2.34 16.00 0.033 Ref , R Picea abies Reference 0.64 0.67 Restored 1.65 3.42 0.87 8.62 0.41 NS Pinus sylvestris Reference 0.15 0.10 Restored 0.00 0.00 Na Populus tremula Reference 0.15 0.46 Restored 0.34 0.70 Na Salix caprea Reference 0.80 2.40 Restored 0.00 0.00 Na Sorbus aucuparia Reference 0.00 0.00 Restored 0.02 0.07 Na Total Reference 5.26 5.04 Restored 12.30 13.07 1.11 15.68 0.28 NS Notes: NS ¼ non-significant, Na ¼ no statistical test conducted due to low sample size, Ref ¼ Reference, R ¼ restored, * denotes P-values significant after sequential Bonferroni adjustment due to repeated independent t-tests. characteristics showed that restored sites con- important to remember that 59 species were tained larger volumes of coarse woody debris unique to commercially managed sites. than commercially managed forests. Restored There were no differences in total species forests were also more open, and permeable to richness and abundance between restored and sunlight. Saproxylic beetles with explicit sub- commercially managed forests, but in agreement strate and microclimatic preferences were, as with our third hypothesis there were significant predicted in our second hypothesis, more com- differences in species composition, largely ex- mon in restored sites (Appendix C). Taxonomic plained by canopy cover. A reason for this could groups favored by dead wood from broadleaved be that many saproxylic beetles have developed trees were, for instance, more species-rich in adaptations to natural disturbances, such as fire restored sites. This is not surprising since large (Wikars 2002, Boulanger and Sirois 2007, Hja¨lte´n quantities of dead wood were created from birch et al. 2007, Hekkala et al. 2014). Wildfires, storms, trees. Selectively harvested forests had a higher floods, and pest outbreaks can create open species richness and abundance of saproxylic habitats with large volumes of dead wood (Esseen beetles favored by sun-exposure. In this study, et al. 1997, Kuuluvainen 2002), much like those we were able to anticipate the effects of sun- restored for the white-backed woodpecker (Tables exposure thanks to earlier assessments of forestry 2 and 3). Forest restoration for the white-backed impacts. Several studies show that clear-felled woodpecker created favorable environmental con- forests, with retained dead wood from broad- ditions for warmth-demanding species leaved trees, are utilized by different saproxylic currently disadvantaged by forestry practices. In beetles than mature forests with similar sub- addition, 91 species (31% of all species caught in strates (Kaila et al. 1997, Martikainen 2000, the traps) were unique to restored sites. Sverdrup-Thygeson and Ims 2002). In this study, Red-listed species were positively affected by pair-wise comparisons showed that forest resto- forest restoration, both in terms of species ration can benefit saproxylic beetles favored by richness and abundance. This provides support sun-exposure without affecting the species rich- for our final hypothesis, but unfortunately the ness of shade-tolerant species. However, it is sample size was too small for us to analyze

v www.esajournals.org 8 December 2015 v Volume 6(12) v Article 278 BELL ET AL.

Fig. 1. Species richness (A and C) and abundance (B and D) of saproxylic beetles (mean þ/ SE) of different microclimatic (sun, shade, both, unknown) and substrate (broadleaved trees, coniferous trees, both) preferences. P-values denote significant (P 0.05) differences between restored and commercially managed reference forests.

Fig. 2. The community structure of saproxylic beetles in restored sites (triangles) and commercially managed reference sites (squares) illustrated in two nMDS-plots. The left-hand plot (dominance structure) was composed from untransformed (raw) data. The right-hand plot was composed from transformed (presence-absence) data. Each symbol represents one study site. P-values denote significant (P 0.05) differences in dominance structure and species composition.

v www.esajournals.org 9 December 2015 v Volume 6(12) v Article 278 BELL ET AL.

Table 4. SIMPER results for raw data and presence-absence (P-A) data. Only the ten most influential species were included in this table. All listed species have been categorized in terms of their microclimatic (sun or shade) and substrate (broadleaved trees or coniferous trees) preferences.

Average abundance Substrate/tree preference Microclimate Species, by data type Reference RestoredContrib% Cum% Broadleaf Conifers Sun Shade Raw Triplax russica 2.44 13.56 4.14 4.14 x x x Trypodendron domesticum 11.78 8.44 4.07 8.2 x x Tomoxia bucephala 0 13.33 3.91 12.11 x x x Bibloporus bicolor 15.22 10.89 3.79 15.9 x x Dryocoetes autographus 14.44 5 3.51 19.41 x x Cerylon ferrugineum 12.56 7.67 3.39 22.81 x x x Euplectus punctatus 7 14.78 3.06 25.87 x x x Cerylon histeroides 11.78 8.56 2.62 28.49 x x x x Enicmus rugosus 0.56 9.11 2.44 30.93 x x Haploglossa villosula 2.56 9.11 2.3 33.24 x x x x P-A Tomoxia bucephala 0 0.78 0.98 0.98 x x x Sepedophilus littoreus 1 0.33 0.87 1.85 x x x Leptura quadrifasciata 0 0.67 0.82 2.67 x x Hylis cariniceps 0 0.67 0.81 3.47 x x (?) Dacne bipustulata 0 0.67 0.79 4.27 x x x Pityogenes chalcographus 0.22 0.78 0.78 5.05 x x Cychramus luteus 0.89 0.33 0.77 5.82 x ? ? Endomychus coccineus 0.11 0.67 0.77 6.59 x x Ennearthron cornutum 0.67 0.11 0.75 7.34 x x x* Mycetochara flavipes 0 0.67 0.75 8.1 x (?) Notes: Contrib% ¼ contribution in percent, Cum% ¼ cumulative percentage, (?) ¼ no empirical evidence, but unique to restored forests, ? ¼ no empirical evidence, x* ¼ weak empirical evidence.

Table 5. Best models from distance based linear Table 6. Saproxylic beetles on the 2010 Swedish red-list

modelling showing all models with DAICc , 2 for (Ga¨rdenfors 2010) in this study, and their ranked raw data and presence–absence data respectively. df abundances. ¼ 16. Time ¼ time since disturbance, BRGR ¼ Species Status Reference Restored Sum broadleaved grasses vegetation type. Hylis procerulus NT 0 46 46 Model AIC DAIC R2 Xylophilus corticalis NT 3 12 15 c c Dolichocis laricinus NT 4 3 7 Raw data Agathidium mandibulare NT 4 2 6 Canopy cover 139.53 0.22 Hallomenus axillaris NT 0 4 4 Canopy cover þ Elevation 140.69 1.16 0.29 Cis quadridens NT 0 3 3 Canopy cover þ Time 141.08 1.55 0.28 Cis rugulosus NT 1 2 3 Time 141.35 1.82 0.14 Denticollis borealis NT 0 3 3 Elevation 142.34 2.81 0.09 Euryusa castanoptera NT 1 2 3 Canopy cover þ Time þ Elevation 142.78 3.25 0.34 Cis dentatus NT 1 1 2 Time þ Elevation 142.89 3.36 0.2 Peltis grossa VU 0 2 2 P-A data Amphicyllus globiformis NT 1 0 1 Canopy cover 134.33 0.18 Atomaria affinis NT 1 0 1 Time 135.13 0.8 0.14 Cerylon deplanatum NT 1 0 1 Canopy cover þ Elevation 135.33 1 0.26 Eucnemis capucina VU 0 1 1 Canopy cover þ Time 135.8 1.47 0.24 Harminius undulatus NT 0 1 1 Elevation 135.81 1.48 0.1 Orchesia fasciata NT 0 1 1 Canopy cover þ BRGR 135.81 1.48 0.24 Triplax rufipes NT 0 1 1 BRGR 136.03 1.7 0.09 Zilora ferruginea NT 0 1 1 Time þ BRGR 136.31 1.98 0.22 Total species richness 9 16 19 Time þ Elevation 136.45 2.12 0.21 Total abundance 17 85 102 Elevation þ BRGR 136.89 2.56 0.19 Canopy cover þ Time þ Elevation 137.35 3.02 0.31 Notes: Scientific names follow Catalogus Palearctic Coleop- Elevation þ Time þ BRGR 137.88 3.55 0.29 tera. Status ¼ 2010 Red List Category (NT ¼ Near Threatened, VU ¼ Vulnerable).

v www.esajournals.org 10 December 2015 v Volume 6(12) v Article 278 BELL ET AL.

Fig. 3. In A: number of species (saproxylic beetles, mean 6 SE) on the 2010 Swedish red-list in restored and commercially managed reference forests. In B: abundance of saproxylic beetles (mean 6 SE) on the 2010 Swedish red-list in restored and commercially managed reference forests. P-values denote significant (P 0.05) differences. differences in community composition. Two red- ciata are conspicuous and attracted to forbs like listed species, Hylis procerulus and Xylophilus Filipendula ulmaria and Angelica sylvestris in early corticalis, were particularly abundant in restored summer, and larval stages of S. ratzeburgi leave sites. In Sweden, H. procerulus has been associat- distinctive signs of foraging in dead wood ed with dead wood from Norway spruce (Ehnstro¨m and Axelsson 2002). This might make (Baranowski et al. 1999), but observations in them good sub-targets in landscape-level at- other countries (e.g., Ukraine and Russia) link it tempts to restore habitats for the white-backed to both broadleaved and coniferous trees (Bu¨ che woodpecker, as well as examples of restoration et al. 2010). X. corticalis will also use both success at the stand-level. substrates, but it is limited by other factors, such Umbrella species have mainly been discussed as the initial level of decomposition (Ehnstro¨m as management shortcuts in environmental pro- 1999). It is also possible that both species benefit tection (Caro and O’Doherty 1999), but ecological from sun-exposure, but we were unable to find restoration presents a different context. Co- support for this in the literature. occurring species will overlap at different spatial Canopy cover was identified as one of the most and temporal scales, and umbrella species with important environmental variables determining extensive habitat requirements will probably be beetle community composition. However, cano- among the last to recover. Under such circum- py cover was correlated with spruce basal area stances, umbrella species become testimonies of and lying birch CWD making it difficult to ecosystem recovery rather than management distinguish between the effects of dead wood shortcuts (Fig. 4). The white-backed woodpecker enrichment and those of spruce removal. Some has not yet recovered in Sweden, but at this point wood-inhabiting beetles are likely to benefit from in time habitat requirements have only been both measures, e.g., prey species targeted by the fulfilled at the stand-level. Forest restoration has white-backed woodpecker in edge habitats, post- created stands with coarse woody debris vol- fire sites, and areas of intermediate forest cover. umes similar to those found in typical white- Two examples of important prey species for the backed woodpecker habitats (10–20 m3/ha; An- white-backed woodpecker, unique to restored gelstam et al. 2003, Auleń et al. 2010), but white- sites in this study, were Scolytus ratzeburgii and backed woodpecker territories cover much larger Leptura quadrifasciata (Auleń 1988). Both species areas (150–650 ha; Auleń et al. 2010). have been linked to sun-exposed dead wood In conclusion, we argue that habitat require- from broadleaved trees (Ehnstro¨m and Axelsson ments of umbrella species can guide successful 2002, Dahlberg and Stokland 2004). They are also restoration efforts, if there is congruence in the easily surveyed. Adult specimens of L. quadrifas- response of co-occurring species to manipulated

v www.esajournals.org 11 December 2015 v Volume 6(12) v Article 278 BELL ET AL.

LITERATURE CITED

Andelman, S. J., and W. F. Fagan. 2000. Umbrellas and flagships: efficient conservation surrogates, or expensive mistakes? Proceedings of the National Academy of Sciences USA 97:5954–5959. Angelstam, P., and L. Andersson. 2001. Estimates of the needs for forest reserves in Sweden. Scandina- vian Journal of Forest Research 3:38–51. Angelstam, P. K., R. Bu¨ tler, M. Lazdinis, G. Mikusin´ski, and J.-M. Roberge. 2003. Habitat thresholds for focal species at multiple scales and forest biodiversity conservation: dead wood as an example. Annales Zoologici Fennici 40:473–482. Auleń, G. 1988. Ecology and distribution history of the white-backed woodpecker Dendrocopos leucotos in Sweden. Dissertation. Report 14. Department of Fig. 4. Many species are resource- and process- Wildlife Ecology, Swedish University of Agricul- limited, but at different spatial and temporal scales. tural Sciences, Uppsala, Sweden. Local improvements will not necessarily fulfill habitat Auleń, G., A. Carlsson, and K. Stigha¨ll. 2010. Artfakta- requirements of top predators like the white-backed blad: Vitryggig hackspett (Dendrocopos leucotos). woodpecker, but less demanding species, like many Swedish Species Information Centre, Swedish saproxylic beetles, will probably respond more rapidly. University of Agricultural Sciences, Uppsala, Swe- In fact, forest restoration will most probably have den. bottom-up effects on top predators in saproxylic food Baranowski, R., B. Ehnstro¨m, and S. Lundberg. 1999. webs. Under such circumstances, umbrella species Artfaktablad (Hylis procerulus). Swedish Species Information Centre, Swedish University of Agri- become testimonies of ecosystem recovery rather than cultural Sciences, Uppsala, Sweden. management shortcuts. Bernes, C. 2011. Biodiversity in Sweden. Pages 36–92 in Monitor 22. Chapter 1: Biodiversity in the forest (environmental) attributes. Resource- and pro- landscape. Swedish Environmental Protection Agency, Stockholm, Sweden. cess-limited species with extensive area require- Blicharska, M., P. W. J. Baxter, and G. Mikusin´ski. 2014. ments, like the white-backed woodpecker, are Practical implementation of species’ recovery plans: particularly useful in this context (Lambeck 1997, lessons from the white-backed woodpecker action Angelstam and Andersson 2001). It is not certain plan in Sweden. Ornis Fennica 91:108–128. that white-backed woodpecker populations in Boulanger, Y., and L. Sirois. 2007. Postfire succession of Sweden will recover, validating successful resto- saproxylic arthropods, with emphasis on Coleop- tera, in the north boreal forest of Quebec. Environ- ration efforts at the landscape-level, but in this mental Entomology 36:128–141. study there were many beneficiary species at the Branton, M. A., and J. S. Richardson. 2014. A test of the stand-level. Positively affected species were often umbrella species approach in restored floodplain linked to environments much like those inhabit- ponds. Journal of Applied Ecology 51:776–785. ed by the white-backed woodpecker, and we Bu¨ che, B., E. Hyvarinen, and K. Alexander. 2010. Hylis argue that this shows that there is merit to the procerulus. The IUCN list of threatened species. umbrella species concept in restoration design. Version 2014.2. www.iucnredlist.org Caro, T. M. 2001. Species richness and abundance of small mammals inside and outside an African ACKNOWLEDGMENTS national park. Biological Conservation 98:251–257. Caro, T. M. 2003. Umbrella species: critique and Funds for this project were provided by the Swedish lessons from East Africa. Animal Conservation Research Council FORMAS (via C. Nilsson), and the 6:171–181. Swedish University of Agricultural Sciences (via J. Caro, T. M., and G. O’Doherty. 1999. On the use of Hja¨lteń). Stig Lundberg and Jacek Hilszczan´ski iden- surrogate species in conservation biology. Conser- tified all saproxylic beetles. We thank A˚ ke Nordstro¨m vation Biology 13:805–814. and Adriaan de Jong for field assistance, and Roger Dahlberg, A., and J. Stokland. 2004. Vedlevande arters Pettersson for expert advice. krav pa˚ substrat: sammansta¨llning och analys av

v www.esajournals.org 12 December 2015 v Volume 6(12) v Article 278 BELL ET AL.

3600 arter. Rapport 7. Swedish Forest Agency, Biodiversity and Conservation 6:1–18. Jo¨nko¨ping, Sweden. Kerley, G. I. H., R. L. Pressey, R. M. Cowling, A. F. Ehnstro¨m, B. 1999. Artfaktablad (Xylophilus corticalis). Boshoff, and R. Sims-Castley. 2003. Options for the Swedish Species Information Centre, Swedish conservation of large and medium-size mammals University of Agricultural Sciences, Uppsala, Swe- in the Cape Floristic Region hotspot, South Africa. den. Biological Conservation 112:169–190. Ehnstro¨m, B. 2011. Insekter som missgynnas av Koch, K. C. 1989. Die Ka¨fer Mitteleuropas. O¨ kologie ta¨tnande skogar. Skog & mark 2011: om tillsta˚ndet Band 1-3. Goecke & Evers Verlag, Krefeld, Ger- i svensk landmiljo¨. Swedish Environmental Protec- many. tion Agency, Stockholm, Sweden. Kuuluvainen, T. 2002. Natural variability of forests as a Ehnstro¨m, B., and R. Axelsson. 2002. Insektsgnag i reference for restoring and managing biological bark och ved. Swedish Species Information Centre, diversity in boreal Fennoscandia. Silva Fennica Swedish University of Agricultural Sciences, Up- 36:97–125. psala, Sweden. Lambeck, R. J. 1997. Focal species: a multi-species Esseen, P.-A., B. Ehnstro¨m, L. Ericson, and K. Sjo¨berg. umbrella for nature conservation. Conservation 1997. Boreal forests. Ecological Bulletins 46:16–47. Biology 11:849–856. Fleishman, E., R. B. Blair, and D. D. Murphy. 2001. Linder, P., B. Elfving, and O. Zackrisson. 1997. Stand Empirical validation of a method for umbrella structure and successional trends in virgin boreal species selection. Ecological Applications 11:1489– forest reserves in Sweden. Forest Ecology and 1501. Management 98:17–33. Fleishman, E., D. D. Murphy, and P. F. Brussard. 2000. Lindhe, A., A˚ . Lindelo¨w, and N. A˚ senblad. 2005. A new method for selection of umbrella species for Saproxylic beetles in standing dead wood density conservation planning. Ecological Applications in relation to substrate sun-exposure and diameter. 10:569–579. Biodiversity and Conservation 14:3033–3053. Ga¨rdenfors, U. 2010. Ro¨dlistade arter i Sverige 2010. Martikainen, P. 2000. Conservation of threatened Swedish Species Information Centre, Swedish saproxylic beetles: significance of retained aspen University of Agricultural Sciences, Uppsala, Swe- Populus tremula on clearcut areas. Ecological Bulle- den. tins 49:205–218. Groom, M. J., G. K. Meffe, and C. R. Carroll. 2006. Martikainen, P., L. Kaila, and Y. Haila. 1998. Threat- Principles of conservation biology. Sinauer, Sunder- ened beetles in white-backed woodpecker habitats. land, Massachusetts, USA. Conservation Biology 12:293–301. Ha¨gglund, B., and J.-E. Lundmark. 2004. Handledning Mikusin´ski, G., and P. Angelstam. 2004. Occurrence of i bonitering med Skogsho¨gskolans boniteringssys- mammalsandbirdswithdifferentecological tem. Del 3 Markvegetationstyper—Skogsmarks- characteristics in relation to forest cover in Europe: flora. Swedish Forest Agency, Jo¨nko¨ping, Sweden. Do macroecological data make sense? Ecological Hekkala, A.-M., M.-L. Paä¨talo, O. Tarvainen, and A. Bulletins 51:265–275. Tolvanen. 2014. Restoration of young forests in Mild, K., and K. Stigha¨ll. 2005. A˚ tga¨rdsprogram fo¨r eastern Finland: benefits for saproxylic beetles bevarande av vitryggig hackspett (Dendrocopos (Coleoptera). Restoration Ecology 22:151–159. leucotos) och dess livsmiljoër. Rapport 5486. Swed- Hellberg, E. 2004. Historical variability of deciduous ish Environmental Protection Agency, Stockholm, trees and deciduous forests in northern Sweden. Sweden. Effectsofforestfires,land-useandclimate. Roberge, J.-M., G. Mikusin´ski, and S. Svensson. 2008. Dissertation. Swedish University of Agricultural The white-backed woodpecker: Umbrella species Sciences, Umea˚, Sweden. for forest conservation planning? Biodiversity and Hja¨lteń, J., T. Johansson, O. Alinvi, K. Danell, J. P. Ball, Conservation 17:2479–2494. R. Pettersson, H. Gibb, and J. Hilszczan´ski. 2007. Roth, T., and D. Weber. 2008. Top predators as The importance of substrate type, shading and indicators for species richness? Prey species are scorching for the attractiveness of dead wood to just as useful. Journal of Applied Ecology 45:987– saproxylic beetles. Basic Applied Ecology 8:364– 991. 376. Rubinoff, D. 2001. Evaluating the California gnatcatch- Johansson, M. E., and C. Nilsson. 2002. Responses of er as an umbrella species for conservation of riparian plants to flooding in free-flowing and southern California coastal sage scrub. Conserva- regulated boreal rivers: an experimental study. tion Biology 15:1374–1383. Journal of Applied Ecology 39:971–986. Ruxton, G. D. 2006. The unequal variance t-test is an Kaila, L., P. Martikainen, and P. Punttila. 1997. Dead underused alternative to Student’s t-test and the trees left in clear-cuts benefit saproxylic Coleoptera Mann-Whitney U test. Behavioral Ecology 17:688– adapted to natural disturbances in boreal forest. 690.

v www.esajournals.org 13 December 2015 v Volume 6(12) v Article 278 BELL ET AL.

Sergio, F., L. Marchesi, and P. Pedrini. 2004. Integrating species: the white-backed woodpecker Dendrocopos individual habitat choices and regional distribution leucotos. Scandinavian Journal of Forest Research of a biodiversity indicator and top predator. 26:576–585. Journal of Biogeography 31:619–628. Suter, W., R. F. Graf, and R. Hess. 2002. Capercaillie Sergio, F., I. Newton, and L. Marchesi. 2005. Top (Tetrao urogallus) and avian biodiversity: testing predators and biodiversity. Nature 436:192. the umbrella-species concept. Conservation Biolo- Sergio, F., I. Newton, L. Marchesi, and P. Pedrini. 2006. gy 16:778–788. Ecologically justified charisma: preservation of top Sverdrup-Thygeson, A., and R. A. Ims. 2002. The effect predators delivers biodiversity conservation. Jour- of forest clearcutting in Norway on the community nal of Applied Ecology 43:1049–1055. of saproxylic beetles on aspen. Biological Conser- Sergio, F., P. Pedrini, and L. Marchesi. 2003. Reconcil- vation 108:377–378. ing the dichotomy between single species and To¨rnblom, J., E. Degerman, J. -M. Roberge, P. Angel- ecosystem conservation: black kites (Milvus mi- stam, and T. Eriksson. 2007. Vitrygg och lax: tva grans) and eutrophication in pre-Alpine lakes. paraplyarter fo¨r funktionella land- och vattenmil- Biological Conservation 110:101–111. jo¨er? Fakta skog. Om forskning vid Sveriges Sheehan, J., et al. 2014. Avian response to timber lantbruksuniversitet. Swedish University of Agri- harvesting applied experimentally to manage cultural Sciences, Umea˚, Sweden. cerulean warbler breeding populations. Forest Wikars, L.-O. 2002. Dependence on fire in wood-living Ecology and Management 321:5–18. insects: an experiment with burned and unburned SLU. 2012. Forest statistics 2012. Official Statistics of spruce and birch logs. Journal of Insect Conserva- Sweden. Swedish University of Agricultural Sci- tion 6:1–12. ences, Umea˚, Sweden. Wilson, C. W., R. E. Masters, and G. A. Bukenhofer. Stigha¨ll, K., J.-M. Roberge, K. Andersson, and P. 1995. Breeding bird response to pine-grassland Angelstam. 2011. Usefulness of biophysical proxy community restoration for red-cockaded wood- data for modelling habitat of an endangered forest peckers. Journal of Wildlife Management 59:56–67.

SUPPLEMENTAL MATERIAL

ECOLOGICAL ARCHIVES

Appendices A–C are available online: http://dx.doi.org/10.1890/ES14-00551.1.sm

v www.esajournals.org 14 December 2015 v Volume 6(12) v Article 278