Received: 31 August 2022 Accepted: 10 July 2023 DOI: 10.1111/afe.12589 ORIGINAL ARTICLE Understanding the Coleoptera community at the tree-line using taxonomic and functional guild approaches Jamie Dinkins Bookwalter 1,2 | Berta Caballero-Lo pez 3 | Roberto Molowny-Horas 2 |  pez 1,2 Bernat Claramunt-Lo 1 Unitat d’Ecologia, BABVE, Edifici Ciències, Abstract Universitat Autònoma de Barcelona, Bellaterra, Spain 1. Mountain species are at the forefront of climate change disruption, and montane 2 CREAF, Centre de Recerca Ecològica i saproxylic Coleoptera are facing large- and small-scale changes in their surroundings. Aplicacions Forestals, Bellaterra (Cerdanyola de Vallès), Spain Saproxylic Coleoptera are both functionally and taxonomically diverse and are the 3 Museu de Ciències Naturals de Barcelona, representatives of an imperilled fauna confronted with the realities of a changing Passeig Picasso, Castell Tres Dragons, landscape. Understanding the effects of elevation and other forest characteristics on Barcelona, Spain saproxylic and non-saproxylic Coleoptera is a step towards predicting the future of Correspondence taxonomic and functional group biodiversity at the tree-line and on mountains. Jamie Dinkins Bookwalter, Unitat d’Ecologia, BABVE, Edifici Ciències, Universitat Autònoma 2. The objective of this study was to examine the effect of elevation and other forest de Barcelona, 08193 Bellaterra, Catalonia, characteristics on the biodiversity of montane Coleoptera at the tree-line using Spain. Email: jdink123@gmail.com both taxonomic and functional feeding guild classifications. 3. Our results suggest that abundance of saprotrophs is closely linked to the density Funding information of large trees rather than the volume of wood. Edge effects and elevation seem to Collections Section of the Museu de Ciències Naturals de Barcelona; Earthwatch Institute drive abundance patterns of some species and also influence taxonomic and func- tional guild community patterns differently. Finally, we discuss the implications of climate change and land abandonment to future Coleoptera community structure. KEYWORDS altitude, Andorra, beetle, climate change, elevation, feeding guilds, new records, Pinus mugo, Pyrenees, saproxylic I N T R O D U CT I O N there may be no habitable terrain above where they currently exist that is available for colonization (Dieker et al., 2011; Wilson et al., 2005). The decline of insects in the last 50 years is well documented Montane saproxylic Coleoptera, or beetles that depend in some part of (Hallmann et al., 2017; Kotze & O’Hara, 2003; Wendorff & their life cycle on dead or dying wood (sensu Speight, 1989), are of par- Schmitt, 2019), driven at least in part by climate change and loss of ticular interest because they are ecologically important and taxonomi- habitat (Wagner et al., 2021). Montane insect species in particular are cally and functionally diverse (Nieto & Alexander, 2010). Saproxylic in peril (Dirnbock et al., 2011; Sánchez-Bayo & Wyckhuys, 2019) as Coleoptera not only play important roles in nutrient recycling but also temperature variation at higher elevations due to climate change can include many feeding guilds, including predators, parasites, fungivores, result in a loss of habitat. For example, studies have shown that con- detritivores, myxomycophages (slime mould feeders), wood-consumers tractions of lower elevation ranges may not correspond to an upward and omnivores (Gimmel & Ferro, 2018). Saproxylic Coleoptera are often shift of higher elevational ranges (Dahlhoff et al., 2019; Moret used as biodiversity indicators for wider forest ecosystem functioning et al., 2016). Other montane insects may ‘run out of mountain’, that is, ski et al., 2021). Therefore, understanding the (Burns et al., 2014; Karpin This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. © 2023 The Authors. Agricultural and Forest Entomology published by John Wiley & Sons Ltd on behalf of Royal Entomological Society. Agr Forest Entomol. 2023;25:669–680. wileyonlinelibrary.com/journal/afe 669 14619563, 2023, 4, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/afe.12589 by Readcube (Labtiva Inc.), Wiley Online Library on [19/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 670 BOOKWALTER ET AL. spatial dynamics of montane saproxylic Coleoptera communities at the richness of saprotroph Coleoptera (beetles that feed upon dead or tree-line is fundamental to forecasting the change in biodiversity pat- decaying organic matter) and saproxylic Coleoptera (beetles that terns in a shifting landscape. depend on dead or decaying wood at some point in their life cycle) to Following the trend of many montane flora and fauna be closely linked to the volume of dead wood and large trees and (Rahbek, 2005), Coleopteran biodiversity generally decreases with (2) forest characteristics that are related to higher levels of sunlight, increasing elevation (Corcos et al., 2018; Franc et al., 2007) or volume of dead wood and density of larger trees to predict higher tax- displays a hump-shaped distribution along an elevational gradient onomic abundance, taxonomic richness, functional feeding guild abun- (Tykarski, 2006). Coleoptera biodiversity can increase with increasing dance and functional feeding guild richness. This study was conducted elevation, but this is rare (Dolson et al., 2021). These trends vary as part of a larger research programme monitoring climate change in among taxonomic and functional groups and across geographical areas high elevation Andorra (Bookwalter et al., 2023). and depend on spatial scale and elevational gradient range (Chamberlain et al., 2016; Colwell et al., 2004; McCain, 2009). Numer- ous variables are known to drive community structure along an eleva- METHODS tional gradient, especially temperature, which can delay timing of flight and elongate life cycles of bark beetles and other herbivorous Study area and sample collection insects (Bale et al., 2002; Reymond et al., 2013). Rising temperatures are expected to shorten generation times of some pest bark beetles, The study spanned 1 year (2017) in a 20 km2 section of Vall d’Ordino, such as Ips typographus (Linnaeus, 1758), at higher elevations (Jakoby Ordino Parish, Principality of Andorra, and included Vall d’Ordino and et al., 2019). However, little is known about saproxylic Coleoptera Vall de Sorteny Natural Park. We selected paired 0.1 km2 plots at five community structure at the tree-line, an important ecotone. locations (N = 10 plots) in P. mugo forest and installed seven tradi- In most mountains, the delineation between a forest margin and tional insect traps (three attraction, three flight intercept and one mal- shrub-only terrain is a matter of scale, as canopies can open gradually aise) in each plot (Figure 1). or with a sharp transition depending on slope and other environmen- Each attraction trap consisted of a 1 L plastic soda bottle with a sin- tal factors (Holtmeier & Broll, 2007). The tree-line is generally defined gle 3 cm diameter hole in the side, hung upright from a tree branch and as the point in which the dominant stem of a tree no longer grows positioned 30 cm from the trunk. The soda bottle was filled to just above 3 m (Körner, 2012), and in the last 100 years, the tree-line below the hole with 250 g of a bulk bait mixture containing the fol- delineation in some mountains has migrated upwards (Harsch lowing ratio: 7 L sangria (Don Simon):2 L peach juice (Spar):1 kg et al., 2009). The relationship between the tree-line and climate salt:1 kg sugar (Viñolas et al., 2009). Flight intercept traps consisted of change is difficult to untangle from other biotic and abiotic variables; two transparent laminate plastic PVC panes (15 cm height by 12.5 cm soil temperature, local, current and historic land use and abiotic site width) perpendicularly crossed below a 14 cm diameter white hard plas- conditions can play a role in limiting tree growth at a specific elevation tic disk attached to a 13 cm diameter white plastic funnel. Each white (Hofgaard, 1997; Holtmeier & Broll, 2005; Körner, 2012). Research malaise trap measured 120  100  150 cm (Entosphinx S.R.O). A col- shows that the rise in the tree-line in the Pyrenees mountains is likely lection bottle containing 70% propylene glycol (VWR Chemicals) and a influenced more strongly by land abandonment rather than climate few drops of dish detergent (Fairy) (to lower surface tension) was change (Ameztegui et al., 2016; Batllori & Gutiérrez, 2008). There is attached to each flight intercept and malaise trap. Hymenoptera and little debate, however, that the eastern Pyrenean tree-line is migrating Diptera are traditional targets of malaise traps (Karlsson et al., 2020), upwards and that the population of the dominant tree at the Pyre- but Coleoptera can be captured as well (Skvarla & Dowling, 2017). nean tree-line, the Pinus mugo complex (hereafter Pinus mugo), has Traps were suspended 1.8–1.9 m above the ground and were spaced at become denser and less patchy over the last 50 years (Batllori least 30 m apart within each plot. Representative photos of the three et al., 2010; Batllori & Gutiérrez, 2008), although these two spatial types of traps are shown in Supplementary Material 1. Traps were phenomena are driven by different factors (Feuillet et al., 2020). installed 23-28 May 2017, and their contents were removed and baits In this work, we examined taxonomic and functional saproxylic were refilled every 13–15 days until 30–31 September 2017. All speci- and non-saproxylic Coleoptera community responses to stand and mens captured in the traps were kept in 70% ethanol until processed landscape characteristics at the tree-line and 200–300 meters below and deposited in the Museu de Ciències Naturals de Barcelona. the tree-line in a forest in the eastern Pyrenees. Other studies have For each of the five pairs of plots, one plot was established at the examined saproxylic Coleoptera community responses to stand level elevation that generally is associated with the tree-line, and the other characteristics in Mediterranean mountains (Parisi et al., 2020), Scan- plot was established well below the tree-line. To the fullest extent dinavian forests (Brunet & Isacsson, 2009a, 2009b; Gibb et al., 2006), possible, each tree-line plot was directly upgradient from its paired boreal Canadian forests (Saint-Germain et al., 2006) and the Swiss below tree-line plot. Distances between the two plots that formed a Alps (Schiegg, 2000, 2003); but to our knowledge, this is the first pair (each pair consisted of one higher elevation plot at the tree-line study to examine saproxylic Coleoptera community responses to land- and one lower elevation plot below the tree-line) to other paired plots scape and stand characteristics specifically at the tree-line. In this ranged from 430 to 1000 m. Plots at the tree-line ranged from 2055 study, we expected the following outcomes: (1) abundance and to 2217 meters above sea level (hereafter m.a.s.l.) and plots below the 14619563, 2023, 4, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/afe.12589 by Readcube (Labtiva Inc.), Wiley Online Library on [19/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License THE TREE-LINE COLEOPTERA COMMUNITY 671 FIGURE 1 Map of study site. tree-line from 1719 to 1998 m.a.s.l. In Andorra, the tree-line is consid- Morphotypes of Staphylinidae were sent to a Staphylinidae expert for ered to be positioned from 2200 to 2400 m.a.s.l., with local isolated species identification. Coleoptera species were assigned to functional boundaries from 2100 to 2500 m.a.s.l. (Carreras et al., 1996). larval and adult feeding guilds based on the literature currently avail- At each trap location, percentage of open space (1 km radius), able regarding each species’ life cycle as well as the FRISBEE database aspect and slope were calculated using QGIS3.4 and the MCSA 2012 (Bouget et al., 2008). Species were assigned to one of the three func- landcover map downloaded from the Institute of Andorran Studies tional feeding guild groups: phytophage, saprotroph (including wood- (Centre de Biodiversitat de l’Institut d’Estudis Andorrans 2012). Basal feeding, detritivore and fungivore) or predator (including parasitoid). area, volume of dead wood above 7.5 cm diameter and density of live In addition to functional feeding guild classification, each species was P. mugo at or above 30 cm diameter breast height (dbh) (medium tree also classified as saproxylic or non-saproxylic. As mentioned previ- density) and 50 cm dbh (large tree density) were measured at each trap ously, saprotroph Coleoptera are beetles that feed upon dead or location (Table 1). Volume of dead standing stems above 30 cm diame- decaying organic matter, and saproxylic Coleoptera are beetles that ter within a 20 m radius of trap was initially calculated, but we chose to depend on dead or decaying wood at some point in their life cycle. discard this variable due to extremely low inventories. Percentage of Taxa collected in this study, including status as saproxylic versus non- open space was considered a landscape-scale forest characteristic, and saproxylic and feeding guild, are listed in Appendix 1. all the other forest characteristics were considered stand-scale. Statistical analyses Species identification Species abundance (total number of individual specimens caught per All Coleoptera specimens were morphologically identified to the spe- collection per trap) and species richness (total number of species cies level (list of experts involved in identification is given in Supple- caught per collection per trap) among plots at the tree-line and below mentary Material 2) with the exception of Scydmaeninae. the tree-line were calculated for two groups of Coleoptera: saproxylic 14619563, 2023, 4, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/afe.12589 by Readcube (Labtiva Inc.), Wiley Online Library on [19/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 672 BOOKWALTER ET AL. TABLE 1 Description of forest characteristic variables, month and was tested using the both cor() function of the R package corrplot trap type. (Wei & Simko, 2021) and the collin.diag() function of the R package Variable Description Scale misty (Yanagida, 2022). The results of these analyses examining species richness and abundance of combined Coleoptera and saproxylic Cole- Elevation At the tree-line (2055–2217 m.a.s.l.) Stand (i.e., high) or below the tree-line optera are referred to as ‘taxonomic classification’ results. (1719–1998 m.a.s.l.) (i.e., low) The above analyses were also performed on Coleoptera species par- Aspect Flat, South, North, East, West Stand titioned into functional feeding guild groups, using species richness or Basal area Combined basal area (area of stem) Stand species abundance within each guild as a response variable, and forest of all Pinus mugo at or over characteristics, month and trap type as explanatory variables. Functional 7.5 cm diameter at breast height feeding guilds, as mentioned previously, were comprised of the following within 10 m radius of trap categories: phytophage, saprotroph (including wood-feeding, detritivore (m2 ha 1) and fungivore) or predator (including parasitoid). In the models examining Dead wood volume Volume of dead wood with Stand circumference at or over 7.5 cm species richness and abundance of secondary consumers (parasitoids within 10 m radius of trap combined with predators, hereafter referred to simply as ‘predators’), the (m3 ha 1) species richness or species abundance of primary consumers was also Large tree density Density of Pinus mugo at or over Stand tested as an explanatory variable, as primary consumers can serve as prey 50 cm dbh within 20 m radius of  pez et al., 2016). When species exhibited different or host (Caballero-Lo trap (ha) guild behaviours in larval and adult stages, abundances of that species Medium tree density Density of Pinus mugo at or over Stand  pez et al. (2016). were counted into both guilds, following Caballero-Lo 30 cm dbh within 20 m radius of trap (ha) The results of these analyses are referred to as ‘functional feeding guild classification’ results. Month June, July, August, September NA Differences in forest characteristic variables consisting of contin- Percent open Percent of open space (land without Landscape tree cover) within 1 km radius of uous values were examined using Wilcoxon–Mann–Whitney tests fol- trap (1 = completely without lowing methods performed in Parisi et al. (2020). To examine tree cover, 0 = completely similarity among these continuous variables, a principal component closed) analysis (PCA) was also performed on the data using the R package Slope Gradient of land Stand ggfortify (Tang et al., 2016). Trap type Malaise, Flight intercept, Attraction NA RE SU LT S and non-saproxylic species combined (hereafter referred to as ‘com- bined Coleoptera’) and saproxylic only species (hereafter referred to Forest characteristic variables as ‘saproxylic Coleoptera’). For our purposes, both functional and obligate saproxylic species were considered to be within the Forest characteristic variables varied between the two elevations, saproxylic classification. To understand which forest characteristics with measurements of slope, percent of open space, and medium tree drive species richness and abundance, models were fitted to a gener- density comprising the largest differences (see r values; Table 2). Basal alized linear mixed-effects model using the glmmTMB package area and medium tree density were among the variables that were (Brooks et al., 2017) in R (R Core Team, 2021) using zero-truncated larger in plots below the tree-line. Slope was steeper and plots were Poisson, zero-truncated generalized Poisson or zero-truncated non- more open at the tree-line. There was no difference between the vol- binomial depending on residual plots created by the DHARMa package umes of dead wood between the two elevations (Table 2). These (Hartig, 2022). Model performance was tested by calculating the results are reinforced by a PCA displaying: (1) basal area and open Pearson, Kendall and Spearman correlation coefficients. Model accuracy space were strongly negatively correlated and (2) higher elevation was evaluated by measuring the root mean square error and the mean- plots tended to be more open (Figure 2). No multicollinearities absolute deviation of each model. Eight landscape and stand explana- between the forest characteristic variables were found. tory variables listed and described in Table 1 were used as explanatory factors. Moreover, a visual inspection of the datasets revealed that month had a parabolic-shaped response for abundance and richness, Combined Coleoptera which led us to also include month-squared as a predictor (in R syntax, I (month2)) in those analyses. These landscape and stand variables were A total of 8995 specimens were collected, representing 237 species chosen as they are characteristics often found to be influential in (146 saproxylic) and 41 families (Appendix 1). A total of 171 of these saproxylic species distributional patterns (Müller & Bütler, 2010; Oto species found in this study were new records for Andorra (112 of these et al., 2022; Schiegg, 2000; Thorn et al., 2016). Plot was inserted as a records were published in Bookwalter et al. (2023), as the data in these random factor. Potential multicollinearity between explanatory variables studies partly overlapped). The three most abundant species were 14619563, 2023, 4, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/afe.12589 by Readcube (Labtiva Inc.), Wiley Online Library on [19/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License THE TREE-LINE COLEOPTERA COMMUNITY 673 T A B L E 2 Differences in forest characteristics between plots below and at the tree-line analysed using Wilcoxon–Mann–Whitney z scores and associated p values. Low elevation High elevation Independent variable Mean value SD Mean value SD Wilcoxon–Mann–Whitney z p Effect size (r) Basal area 38.7 19.8 23 11.2 11 <0.001 0.41 Dead wood volume 7.28 11.3 13.3 45.5 0.61 0.54 0.02 Medium tree density 107 58.8 47.7 32.5 13.57 <0.001 0.51 Large tree density 10.5 9.65 9.93 16.3 4.43 <0.001 0.16 Percent open 51.2 7.81 68.5 16 14.1 <0.001 0.53 Slope 12.9 10.3 30.4 18.3 15 <0.001 0.56 Note: Extent of effect size is described using r scores (Parisi et al., 2020). Elevaon High Low F I G U R E 2 Principal component analysis of continuous forest characteristic values. found in higher elevation plots, although error bars overlapped (Figure 3). Basal area and percent open were found to negatively affect abundance (est.: 0.61, p: 0.03; est.: 0.47, p: 0.01, respectively), F I G U R E 3 Mean abundance per trap of the five most common whereas elevation positively affected abundance (est.: 0.57, p: 0.00; species found in high and low elevations. Species: (1) Eusphalerum Table 3 and Figure 4). Elevation, basal area, dead wood volume, tree umbellatarum, (2) Dasytes gonocerus, (3) Dasytes virens, (4) Pityogenes density and percent open were not associated with combined Coleop- trepanatus and (5) Anastrangalia sanguinolenta. tera richness. Aspect was also found to influence abundance; however, a boxplot did not show clear affiliation between abundance and a par- showed a moderately positive dependence on elevation (est.: ticular aspect level (Supplementary Material 3). Levels of month and 0.34, p: 0.07) and (2) saproxylic abundance and richness trap type were found to influence abundance and richness relative to depended negatively on medium tree density (est.: 0.48, p: 0.02; their reference levels (see Table 3 for estimates and p values). est.: 0.22, p: 0.04, respectively), unlike combined Coleoptera results (Table 3). Saproxylic Coleoptera Functional feeding guild classification results Model results describing abundance and richness patterns of saproxylic Coleoptera were similar to combined Coleoptera; basal Similar to taxonomic abundance and richness results (i.e., species area and percent open were found to negatively affect abundance abundance and richness results of combined Coleoptera and (est.: 0.9, p: 0.00; est.: 1.13, p < 0.0001, respectively), and saproxylic coleoptera), month and trap type significantly explained aspect was found to influence abundance (Table 3). Two prominent variability in species abundance and richness within phytophage, divergent results were found: (1) saproxylic Coleoptera abundance saprotroph and predator Coleoptera guilds (Table 4). 14619563, 2023, 4, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/afe.12589 by Readcube (Labtiva Inc.), Wiley Online Library on [19/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 674 BOOKWALTER ET AL. T A B L E 3 Results of generalized linear mixed-effects modelling: Effects of independent variables on abundance and richness of combined Coleoptera and saproxylic Coleoptera. Abundance Richness Saproxylic abundance Saproxylic richness Independent variable Est. p value Est. p value Est. p value Est. p value Elevation 0.57 0.00** 0.29 0.53 0.34 0.07. 0.11 0.44 Basal area 0.61 0.03* 0.13 0.13 0.90 0.00** 0.09 0.36 Dead wood volume 0.06 0.49 0.03 0.59 0.08 0.38 0.01 0.86 Medium tree density 0.29 0.16 0.18 0.08. 0.48 0.02* 0.22 0.04* Large tree density 0.27 0.24 0.05 0.65 0.09 0.63 0.16 0.20 Percent open 0.47 0.01* 0.31 0.44 1.13 <0.0001*** 0.02 0.90 Slope 0.17 0.18 0.03 0.74 0.37 0.00** 0.01 0.94 Month 1.36 <0.0001*** 0.37 <0.0001*** 1.23 <0.0001*** 0.36 <0.0001*** I (month2) 1.41 <0.0001*** 0.62 <0.0001*** 1.52 <0.0001*** 0.71 <0.0001*** Aspect N 0.27 0.74 0.72 0.05* 0.50 0.55 0.59 0.11 Aspect NE 3.21 0.00*** 0.05 0.90 3.85 <0.0001*** 0.32 0.48 Aspect E 0.54 0.50 0.73 0.05. 0.37 0.66 0.64 0.09. Aspect SE 0.49 0.55 0.53 0.19 0.32 0.72 0.51 0.20 Aspect S 0.95 0.23 0.56 0.12 1.56 0.06. 0.30 0.39 Aspect SW 0.29 0.78 0.66 0.10. 0.92 0.34 0.60 0.17 Aspect W 0.19 0.81 0.72 0.02* 0.62 0.42 0.47 0.16 Aspect NW 0.42 0.66 0.86 0.00** 0.19 0.84 0.53 0.11 Trap Type: FI 7.16 0.01* 2.56 <0.0001*** 5.82 <0.0001*** 2.69 <0.0001*** Trap Type: M 5.31 0.07. 2.30 <0.0001*** 4.55 <0.0001*** 2.27 <0.0001*** Abbreviations: Est., Beta-estimates; FI, Flight intercept; M, Malaise. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.000. abundance depended positively on abundance of primary consumers (est.: 0.26, p < 0.001; Table 4). Higher basal area and dead wood volume and lower medium tree density and elevation drove higher phytophage Coleoptera richness (est.: 0.06, p: 0.04; est.: 0.05, p: 0.01; est.: 0.14, p: 0.00; 0.88, Elevaon p < 0.0001, respectively). Aspect significantly affected richness within High phytophage and saprotroph Coleoptera. No forest characteristic Low besides aspect was significant in the saprotroph richness model. Finally, richness within predator families moderately and negatively depended upon dead wood volume (est.: 0.08, p: 0.07) and large tree density (est.: 0.13, p: 0.05), and depended positively upon rich- ness within primary consumers (est.: 0.14, p < 0.0001; Table 4). DI SCU SSION FIGURE 4 Abundance of combined Coleoptera at high and low Saproxylic Coleoptera are integral to a healthy ecosystem, and elevations. research in the last 20 years has focused upon describing habitat con- nectivity and relationships with forest variables at large and small scales (see Brin et al., 2011, 2016; Brunet & Isacsson, 2009a; Gibb Higher abundance of phytophage Coleoptera was associated with et al., 2006; Saint-Germain et al., 2006). The relationship between higher levels of dead wood volume and large-tree density (est.: 0.31, saproxylic Coleoptera functional and taxonomic biodiversity and for- p < 0.0001; est.: 0.78, p: 0.01). Higher saprotroph abundance was also est variables at the tree-line, however, is a topic both unexamined and related to higher large-tree density (est.: 0.14, p: 0.03). Predator pertinent, as land use shifts and climate change are driving abiotic T A B L E 4 Results of generalized linear mixed-effects modelling: Effects of forest characteristics variables and PCRich (Primary consumer richness) on species richness within functional groups as well as effects of forest characteristics variables and PCAbun (Primary consumer abundance) on species abundance within functional groups. Predator abun Saprotroph abun Phytophage abun Predator rich Saprotroph rich Phytophage rich Independent variable Est. p value Est. p value Est. p value Est. p value Est. p value Est. p value Elevation 0.37 0.32 0.07 0.59 2.66 0.03* 0.04 0.73 0.06 0.38 0.88 <0.0001*** THE TREE-LINE COLEOPTERA COMMUNITY Basal area 0.11 0.61 0.1 0.21 0.6 0.01** 0.02 0.77 0.04 0.51 0.06 0.04* Dead wood volume 0.01 0.84 0.05 0.36 0.31 <0.0001*** 0.08 0.07. 0.04 0.28 0.05 0.01* Medium tree density 0.08 0.73 0.17 0.07. 0.17 0.68 0.01 0.95 0.07 0.31 0.14 0.00*** Large tree density 0.19 0.38 0.14 0.03* 0.78 0.01** 0.13 0.05. 0.04 0.48 0.04 0.25 Percent open 0.07 0.86 0.06 0.6 0.18 0.86 0.03 0.79 0.05 0.38 0.92 <0.0001*** Slope 0.09 0.52 0.07 0.36 1.5 <0.0001*** 0.06 0.34 0.07 0.17 0.17 <0.0001*** Month 1.31 <0.0001*** 0.09 0.10. 3.7 <0.0001*** 0.22 <0.0001*** 0.08 0.07. 0.33 <0.0001*** I (month2) 1.34 <0.0001*** 0.18 <0.0001*** 2.9 0.00** 0.26 <0.0001*** 0.13 <0.0001*** 0.52 <0.0001*** PCAbun 0.27 <0.0001*** NA NA NA NA NA PCRich NA NA NA 0.14 <0.0001*** NA NA Aspect N 0.37 0.75 0.34 0.21 2.5 0.01 0.17 0.49 0.35 0.06. 0.44 <0.0001*** Aspect NE 2.01 0.10. 0.32 0.31 3.99 <0.0001*** 0.07 0.8 0.26 0.25 0.44 <0.0001*** Aspect E 0.08 0.95 0.46 0.12 1.43 0.15 0.3 0.22 0.63 0.00** 0.33 0.01** Aspect SE 0.57 0.65 0.01 0.97 0.68 0.56 0.17 0.5 0.39 0.04* 0.44 0.00** Aspect S 0.81 0.49 0.06 0.84 1.67 0.10. 0.12 0.63 0.21 0.25 0.12 0.36 Aspect SW 0.34 0.79 0.34 0.34 1.25 0.23 0.28 0.35 0.27 0.32 0.11 0.42 Aspect W 0.49 0.65 0.66 0.02* 0.73 0.31 0.4 0.12 0.38 0.05 0.23 0.02* Aspect NW 0.79 0.57 0.5 0.08. 4.01 0.000*** 0.22 0.41 0.27 0.16 0.4 <0.0001*** Trap Type: FI 3.81 0.00*** 0.5 <0.0001*** 5.87 <0.0001*** 0.52 <0.0001*** 0.56 <0.0001*** 1.71 <0.0001*** Trap Type: M 2.4 0.02* 0.22 0.2 5.09 <0.0001*** 0.62 <0.0001*** 0.32 0.01** 1.71 <0.0001*** Abbreviations: Abun, Abundance; Est., Beta-estimates; FI, Flight intercept; M, Malaise; Rich, Richness. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. 675 14619563, 2023, 4, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/afe.12589 by Readcube (Labtiva Inc.), Wiley Online Library on [19/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 14619563, 2023, 4, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/afe.12589 by Readcube (Labtiva Inc.), Wiley Online Library on [19/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 676 BOOKWALTER ET AL. and biotic transformations in mountains across the world. We found Hanula, 2009). Mechanisms that drive positive responses to dead (1) saprotroph abundance and density of large trees are closely linked, wood by non-saproxylic Coleoptera and other arthropods include the and dead wood correlations are not easily untangled, (2) edge effects addition of structural components, especially large logs. Large logs and elevation drive abundances of some species and (3) edge and other dead wood add structural and chemical complexities, sur- effects and elevation influence taxonomic and functional guild com- face area and moisture refugia (Castro & Wise, 2010; Marra & munity patterns differently. Edmonds, 1998). Dead wood can also increase leaf litter, which can insulate animals including non-saproxylic arthropods from extreme temperatures (Langlands et al., 2011). Abundance of saprotrophs was more closely linked to density of large trees rather than volume of wood Edge effects and elevation seem to drive abundance Large trees are a keystone feature in many habitats and can play an inte- of some species gral role in supporting forest biodiversity because they provide unique microhabitats (e.g., tree hollows, areas of dead wood within living trunks Higher levels of basal area and percent openness drove lower abun- and epiphytic lichens and mosses) (Hall & Bunce, 2011; Lindenmayer dance of both combined Coleoptera and saproxylic Coleoptera. Taken et al., 2014). Numerous studies have linked large, veteran trees to Coleop- together, these data could indicate association of some species of tera abundance (Horak, 2017; Müller et al., 2014; Ranius & Coleoptera with edge effects. In other words, some species of Cole- Jansson, 2000; Wetherbee et al., 2021); therefore, it was unsurprising optera are more successful in areas with higher landscape-level large- that saprotroph abundance was positively linked to large trees. scale tree coverage (i.e., lower percentage of openness) and lower The volume of dead wood is also known to be a predictor of stand-level smaller scale basal area. Forest edges have been found to ski et al., 2021), especially in saproxylic Coleoptera biodiversity (Karpin harbour greater diversity and richness of saproxylic Coleoptera com- cooler sites (Lachat et al., 2012); however, it was not associated in our pared with closed, interior forest habitats (Vodka et al., 2009; study with the abundance or richness of saproxylic Coleoptera or Wermelinger et al., 2007). saprotroph functional group Coleoptera (Tables 3 and 4). The lack of Higher elevations also drove higher abundance of combined Cole- correlation between saproxylic Coleoptera and volume of dead wood optera, and the abundance of saproxylic Coleoptera was moderately agrees with studies that posit that this connection could be more affected by higher elevations. Plots at higher elevations had lower muted and complex than previously hypothesized (Franc et al., 2007; levels of basal area and were more open (Table 2). While the mecha- Lassauce et al., 2011; Vodka et al., 2009). Furthermore, the volume of nisms driving higher abundance at higher elevations do not necessarily dead wood measured in our sites (10.1 ± 32.2 m3 ha 1) is lower include the higher levels of open space at higher elevations, the affilia- compared to other European alpine coniferous forests (26.0 ± tion of saproxylic Coleoptera to both open space and edge habitat is 5.7 m3 ha 1) (Puletti et al., 2019) as well as to other Spanish alpine supported by other studies (Oto et al., 2022; Seibold et al., 2016; 3 1 coniferous forests (21.04 ± 30.50 m ha ) (Alberdi et al., 2020). A Wermelinger et al., 2007). Open space and edge habitat can be a volume of 10.1 ± 32.2 m3 ha 1 is probably lower than the dead wood proxy for sunlight, and higher amounts of sunlight-exposed substrate thresholds needed to sustain many rare species. For example, Müller are known to be important predictors of saproxylic biodiversity and Bütler (2010) found that an average threshold of 24 to >70m3 of (Jonsell et al., 1998; Lindhe & Lindelöw, 2004; Thorn et al., 2016; dead wood was necessary to sustain a variety of saproxylic Coleop- Vogel et al., 2020). It is hypothesized that higher amounts of sunlight tera in a boreal Scandinavian coniferous forest. Some studies have could play a role in warming the substrate and the organisms within, suggested that because early-successional species like bark beetles leading to direct and indirect effects on saproxylic biodiversity. Indi- can be highly mobile, larger volumes of dead wood within a 100 m rectly, warmer temperatures could provoke changes in the type of local might be less important than temperature when predicting abun- wood-eating fungi available to saproxylic Coleoptera, for example, dance of saproxylic Coleoptera (Gibb et al., 2006). Finally, it is possible higher abundance of certain wood-inhabiting fungi on sun-exposed that spatial arrangement of the dead wood, or connectivity, is more logs (Bässler et al., 2010; Vogel et al., 2020) and directly through fas- important to saproxylic Coleoptera abundance than the total volume ter reproductive and growth rates of saproxylic Coleoptera (Brown of dead wood, as suggested by Schiegg (2000). et al., 2004). Open space can promote a complex architecture of for- Phytophage functional group abundance and richness, however, est with more sublevels, a positive influence upon abundance of the were driven by the volume of dead wood. Other studies have found forest-associated taxa (Ampoorter et al., 2020). positive correlations between non-saproxylic Coleoptera groups and higher amounts of dead wood (Seibold et al., 2016). For example, an experiment in North American loblolly pine (Pinus taeda) forests found Edge effects and elevation influence taxonomic and only positive correlations between the volume of dead wood and functional guild community patterns differently ground beetles (Carabidae), a family with many non-saproxylic mem- bers. Similar to our study, no correlations between the volume of dead While elevation positively affected abundance in combined Coleop- wood and saproxylic Coleoptera as a group were found (Ulyshen & tera and moderately and positively affected abundance of saproxylic 14619563, 2023, 4, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/afe.12589 by Readcube (Labtiva Inc.), Wiley Online Library on [19/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License THE TREE-LINE COLEOPTERA COMMUNITY 677 Coleoptera, elevation also significantly positively affected abundance processes are difficult to untangle (Galván et al., 2015). Poikilothermal ani- of Colephytophage functional group Coleoptera. Phytophage richness, mals such as arthropods are particularly sensitive to temperature however, was negatively affected by elevation. It is likely that a few (Roitberg & Mangel, 2016), and climate change and land use shifts could very successful species could be driving the greater abundance found elicit phenological desynchronization of species interactions, further com- at higher elevations (Figure 3). Furthermore, the lower richness of plicating community functioning (Konvicka et al., 2016). phytophage Coleoptera at higher elevations could be linked to colder temperatures more likely to be found at the higher elevations. Tem- perature is most often the dominant abiotic factor affecting herbivo- CONC LU SIONS rous insects (Reymond et al., 2013). Temperature fluctuation is also wider at higher elevations in the Pyrenees (Navarro-Serrano Our results are generally consistent with those of Wermelinger et al. et al., 2020). Although Steven’s extension to Rapaport’s rule (i.e., the (2007), Seibold et al. (2016), Vogel et al. (2020) and Oto et al. (2022). elevational range of a species is wider at higher elevations) is contro- These studies showed high affiliation between Coleoptera and open versial (Kim et al., 2019; McCain & Bracy Knight, 2013), Rasmann space, forest edges and/or sunlight. Saproxylic Coleoptera or et al. (2014) found evidence for less host specialization of Buprestidae saprotroph-group Coleoptera did not show marked differences in species (a saproxylic Coleoptera family) and Apiformes (a bee group) richness as a function of elevation in our study. However, some Cole- at higher elevations in the Swiss Alps. Thus, the greater abundance of optera groups, especially phytophage Coleoptera, were richer and herbivorous insects in the higher elevation plots may be strongly influ- more abundant at plots closer to the tree-line, which were more open. enced by families that are less specialized and able to adapt to more Ecological processes initiated and mediated by climate change and variable conditions. land use shifts are driving changes in biodiversity in mountains, espe- Phytophage species richness was also positively affected by both cially within the Pyrenees range (OPCC-CTP, 2018). Some of these landscape-level open space and stand-level basal area and negatively changes include densification of the tree-line. The strong biodiversity affected by medium tree density, indicating that, as a group, phytoph- correlations with edge effects and open space at the tree-line indicate age Coleoptera are also more successful in edge habitat, albeit in edge landscape and climate change effects on Pyrenean tree-line could habitat with more open space and smaller copses of trees. This is in greatly affect spatial patterns of montane and saproxylic montane contrast to the edge affects found in our study linked to taxonomic Coleoptera in the future. abundance, that is, larger areas of higher landscape-level tree cover (low percent openness) interspersed with interior gaps in the stand- AUTHOR CONTRIBU TIONS level tree cover (low basal area). Our results show the importance of Jamie Dinkins Bookwalter: Conceptualization; data curation; formal testing Coleoptera data using both taxonomic and functional feeding analysis; investigation; methodology; visualization; writing – original guild focused approaches as well as landscape- and stand-level vari- draft; writing – review and editing. Berta Caballero-Lopez: Conceptu- ables for a fuller understanding of saproxylic community data. alization; funding acquisition; methodology; project administration; resources; supervision; writing – review and editing. Roberto Molowny-Horas: Formal analysis; writing – review and editing. Climate change and land use shift prediction  pez: Conceptualization; funding acquisition; Bernat Claramunt-Lo methodology; project administration; resources; supervision; Land use shifts and climate change are driving the densification and writing – review and editing. upward migration of the Pyrenean tree-line (Batllori et al., 2010; Batllori & Gutiérrez, 2008). In the Alps, it is likely that Coleoptera AC KNOW LEDG EME NT S endemic to habitats above the tree-line will face disproportionate spe- Jorge Mederos was indispensable regarding trap design and installa- cies loss in even the most conservative climate change and abandoned tion support. Jana Marco, Gerald and Barbara Dinkins, Georgia Haw- pasture scenarios (Dirnbock et al., 2011). These dynamic processes kins, Brent Bookwalter and volunteers with the Earthwatch Institute will play different roles in future saproxylic Coleoptera assemblages, graciously donated hours of effort cleaning, collecting and organizing as these communities are dependent on wood. In the Iberian eastern samples. Gerald Dinkins also contributed by attentively proofreading. range of Pyrenees, Batllori and Gutiérrez (2008) found only 50% of Sorting the samples would not have been possible without the com- forest densification at the tree-line co-occurred with the tree-line bined assistance of Toni Carrasco, Marc Vilella, Xavi Mendez Camps, upward migration. Our study indicates a closed forest without exterior Pablo Fernandez, Jasmine Leather, Claudia Pla-Narbona and especially or interior edges negatively affects the abundance of montane Cole- David Hernández. Saproxylic Coleoptera experts Josep Muñoz Batet optera, montane saproxylic Coleoptera and phytophage Coleoptera, and Amador Viñolas were integral to the success of this project by suggesting that further forest densification without an upward migra- dispensing trap advice and sharing their vast knowledge of saproxylic tion of the tree-line could be detrimental for many tree-associated Coleoptera morphology and classification. Gianfranco Liberti, Miguel Coleoptera populations. While it is possible that the future forest den- Alonso-Zarazaga and Eduard Petitpierre Vall were especially generous sification and the tree-line upward migration shift could be mitigated with their Coleoptera group expertise. The site map was graciously by forecasted climate change-triggered drought-stress, these and skillfully created by Darren Green. This work was supported by 14619563, 2023, 4, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/afe.12589 by Readcube (Labtiva Inc.), Wiley Online Library on [19/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 678 BOOKWALTER ET AL. the Earthwatch Institute and the Collections Section of the Museu de Brunet, J. & Isacsson, G. (2009a) Restoration of beech forest for saproxylic Ciències Naturals de Barcelona. beetles-effects of habitat fragmentation and substrate density on species diversity and distribution. Biodiversity and Conservation, 18, 2387–2404. CONF LICT OF IN TE RE ST ST AT E MENT Brunet, J. & Isacsson, G. (2009b) Influence of snag characteristics on The authors have no relevant financial or non-financial interests to saproxylic beetle assemblages in a south Swedish beech forest. Jour- disclose. nal of Insect Conservation, 13, 515–528. Burns, M., Smith, M., Slade, E. & Ennos, R. (2014) The saproxylic activity index: a new tool for the rapid assessment of deadwood species dur- DATA AVAI LAB ILITY S TATEMENT ing forest restoration. Open Journal of Forestry, 4, 144–150. Data available from author upon reasonable request. Caballero-Lo pez, B., Blanco-Moreno, J.M., Pujade-Villar, J., Ventura, D., Sánchez-Espigares, J.A. & Sans, F.X. (2016) Herbivores, saprovores ORCID and natural enemies respond differently to within-field plant charac- teristics of wheat fields. Journal of Insect Conservation, 20, 467–476. Jamie Dinkins Bookwalter https://orcid.org/0000-0001-5722-1750 Carreras, J., Carrillo, E., Masalles, R., Ninot, J., Soriano, I. & Vigo, J. (1996) Delimitation of the supra-forest zone in the Catalan Pyrenees. Botan- RE FE R ENC E S ical Journal of the Linnean Society, 47, 7–36. Alberdi, I., Moreno-Fernández, D., Cañellas, I., Adame, P. & Hernández, L. Castro, A. & Wise, D.H. (2010) Influence of fallen coarse woody debris (2020) Deadwood stocks in south-western European forests: ecolog- on the diversity and community structuree of forest-floor spiders ical patterns and large scale assessments. Science of the Total Environ- (Arachnida: Araneae). Forest Ecology and Management, 260, 2088– ment, 747, 141237. 2101. Available from: https://doi.org/10.1016/j.foreco.2010. Ameztegui, A., Coll, L., Brotons, L. & Ninot, J.M. (2016) Land-use legacies 08.051 rather than climate change are driving the recent upward shift of the Chamberlain, D., Brambilla, M., Caprio, E., Pedrini, P. & Rolando, A. (2016) mountain tree line in the Pyrenees. Global Ecology and Biogeography, Alpine bird distributions along elevation gradients: the consistency of 25, 263–273. climate and habitat effects across geographic regions. Oecologia, Ampoorter, E., Barbaro, L., Jactel, H., Baeten, L., Boberg, J., Carnol, M. 181, 1139–1150. et al. (2020) Tree diversity is key for promoting the diversity and Colwell, R.K., Rahbek, C. & Gotelli, N.J. (2004) The mid-domain effect and abundance of forest-associated taxa in Europe. Oikos, 129, 133–146. species richness patterns: what have we learned so far? The American Bale, J.S., Masters, G.J., Hodkinson, I.D., Awmack, C., Bezemer, T.M., Naturalist, 163, E1–E23. Brown, V.K. et al. (2002) Herbivory in global climate change research: Corcos, D., Cerretti, P., Mei, M., Vigna Taglianti, A., Paniccia, D., direct effects of rising temperature on insect herbivores. Global Santoiemma, G. et al. (2018) Predator and parasitoid insects along Change Biology, 8, 1–16. elevational gradients: role of temperature and habitat diversity. Bässler, C., Müller, J., Dziock, F. & Brandl, R. (2010) Effects of resource Oecologia, 188, 193–202. availability and climate on the diversity of wood-decaying fungi. The Dahlhoff, E.P., Dahlhoff, V.C., Grainger, C.A., Zavala, N.A., Otepola- Journal of Ecology, 98, 822–832. Bello, D., Sargent, B.A. et al. (2019) Getting chased up the mountain: Batllori, E., Camarero, J.J. & Gutiérrez, E. (2010) Current regeneration pat- high elevation may limit performance and fitness characters in a terns at the tree line in the Pyrenees indicate similar recruitment pro- montane insect. Functional Ecology, 33, 809–818. cesses irrespective of the past disturbance regime. Journal of Dieker, P., Drees, C. & Assmann, T. (2011) Two high-mountain burnet Biogeography, 37, 1938–1950. moth species (Lepidoptera, Zygaenidae) react differently to the Batllori, E. & Gutiérrez, E. (2008) Regional tree line dynamics in response global change drivers climate and land-use. Biological Conservation, to global change in the Pyrenees. Journal of Ecology, 96, 1275–1288. 144, 2810–2818. Bookwalter, J.D., Niyas, A.M.M., Caballero-Lo  pez, B., Villari, C. & Dirnbock, T., Essl, F. & Rabitsch, W. (2011) Disproportional risk for habitat Claramunt-Lo  pez, B. (2023) Fecal matters: implementing classical loss of high-altitude endemic species under climate change. Global Coleoptera species lists with metabarcoding data from passerine bird Change Biology, 17, 990–996. feces. Journal of Insect Conservation, 27, 557–569. Available from: Dolson, S.J., Loewen, E., Jones, K., Jacobs, S.R., Solis, A., Hallwachs, W. https://doi.org/10.1007/s10841-023-00479-7 et al. (2021) Diversity and phylogenetic community structure across Bouget, C., Brustel, H. & Zagatti, P. (2008) The French information system elevation during climate change in a family of hyperdiverse neotropi- on saproxylic beetle ecology (FRISBEE): an ecological and taxonomi- cal beetles (Staphylinidae). Ecography, 44, 740–752. cal database to help with the assessment of forest conservation sta- Feuillet, T., Birre, D., Milian, J., Godard, V., Clauzel, C. & Serrano- tus. Revue d’Ecologie, 10, 33–36. Notivoli, R. (2020) Spatial dynamics of alpine tree lines under global Brin, A., Bouget, C., Brustel, H. & Jactel, H. (2011) Diameter of downed warming: what explains the mismatch between tree densification woody debris does matter for saproxylic beetle assemblages in tem- and elevational upward shifts at the tree line ecotone? Journal of Bio- perate oak and pine forests. Journal of Insect Conservation, 15, geography, 47, 1056–1068. 653–669. Franc, N., Gotmark, F., Okland, B., Norden, B. & Paltto, H. (2007) Factors Brin, A., Valladares, L., Ladet, S. & Bouget, C. (2016) Effects of forest conti- and scales potentially important for saproxylic beetles in temperate nuity on flying saproxylic beetle assemblages in small woodlots mixed oak forest. Biological Conservation, 135, 86–98. embedded in agricultural landscapes. Biodiversity and Conservation, Galván, J.D., Büntgen, U., Ginzler, C., Grudd, H., Gutiérrez, E., Labuhn, I. 25, 587–602. et al. (2015) Drought-induced weakening of growth–temperature Brooks, M.E., Kristensen, K., van Benthem, K.J., Magnusson, A., associations in high-elevation Iberian pines. Global and Planetary Berg, C.W., Nielsen, A. et al. (2017) glmmTMB balances speed and Change, 124, 95–106. flexibility among packages for zero-inflated generalized linear mixed Gibb, H., Hjalten, J., Ball, J.P., Atlegrim, O., Pettersson, R.B., Hilszczanski, J. modeling. The R Journal, 9, 378–400. et al. (2006) Effects of landscape composition and substrate availabil- Brown, J.H., Gillooly, J.F., Allen, A.P., Savage, V.M. & West, G.B. (2004) ity on saproxylic beetles in boreal forests: a study using experimental Toward a metabolic theory of ecology. Ecology, 85, 1771–1789. logs for monitoring assemblages. Ecography, 29, 191–204. 14619563, 2023, 4, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/afe.12589 by Readcube (Labtiva Inc.), Wiley Online Library on [19/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License THE TREE-LINE COLEOPTERA COMMUNITY 679 Gimmel, M. & Ferro, M. (2018) General overview of Saproxylic Coleoptera. Lindenmayer, D.B., Laurance, W.F., Franklin, J.F., Likens, G.E., Banks, S.C., In: Ulyshen, M. (Ed.) Saproxylic insects. Cham, Switzerland: Springer, Blanchard, W. et al. (2014) New policies for old trees: averting a global pp. 51–128. crisis in a keystone ecological structure. Conservation Letters, 7, 61–69. Hall, S.J.G. & Bunce, R.G.H. (2011) Mature trees as keystone structures in Lindhe, A. & Lindelöw, Å. (2004) Cut high stumps of spruce, birch, aspen Holarctic ecosystems—a quantitative species comparison in a north- and oak as breeding substrates for saproxylic beetles. Forest Ecology ern English park. Plant Ecology & Diversity, 4, 243–250. and Management, 203, 1–20. Hallmann, C.A., Sorg, M., Jongejans, E., Siepel, H., Hofland, N., Schwan, H. Marra, J.L. & Edmonds, R.L. (1998) Effects of coarse woody debris and soil et al. (2017) More than 75 percent decline over 27 years in total fly- depth on the density and diversity of soil invertebrates on clearcut ing insect biomass in protected areas. PLoS One, 12, e0185809. and forested sites on the Olympic peninsula, Washington. Environ- Harsch, M.A., Hulme, P.E., McGlone, M.S. & Duncan, R.P. (2009) Are tree- mental Entomology, 27, 1111–1124. lines advancing? A global meta-analysis of treeline response to cli- McCain, C.M. (2009) Global analysis of bird elevational diversity. Global mate warming. Ecology Letters, 12, 1040–1049. Ecology and Biogeography, 18, 346–360. Hartig, F. (2022) DHARMa: Residual diagnostics for hierarchical (multi- McCain, C.M. & Bracy Knight, K. (2013) Elevational Rapoport’s rule is not per- level / mixed) regression models. R package version 0.4.6. Moret, P., de los Ángeles Aráuz, M., Gobbi, M. & Barragán, Á. (2016) Cli- Hofgaard, A. (1997) Inter-relationships between treeline position, species mate warming effects in the tropical Andes: first evidence for diversity, land use and climate change in the central Scandes moun- upslope shifts of Carabidae (Coleoptera) in Ecuador. Insect Conserva- tains of Norway. Global Ecology and Biogeography Letters, 6, 419–429. tion and Diversity, 9, 342–350. Holtmeier, F.K. & Broll, G. (2005) Sensitivity and response of northern Müller, J. & Bütler, R. (2010) A review of habitat thresholds for dead wood: hemisphere altitudinal and polar treelines to environmental change a baseline for management recommendations in European forests. at landscape and local scales. Global Ecology and Biogeography, 14, European Journal of Forest Research, 129, 981–992. 395–410. Müller, J., Jarzabek-Müller, A., Bussler, H. & Gossner, M.M. (2014) Hollow Holtmeier, F.K. & Broll, G. (2007) Treeline advance—driving processes and beech trees identified as keystone structures for saproxylic beetles adverse factors. Landscape Online, 1, 1–33. by analyses of functional and phylogenetic diversity: hollow beech Horak, J. (2017) Insect ecology and veteran trees. Journal of Insect Conser- trees identified as keystone structures. Animal Conservation, 17, vation, 21, 1–5. 154–162. Jakoby, O., Lischke, H. & Wermelinger, B. (2019) Climate change alters ele- Navarro-Serrano, F., Lo  pez-Moreno, J., Azorin-Molina, C., Alonso- vational phenology patterns of the European spruce bark beetle (Ips González, E., Aznarez-Balta, M., Buisan, S. et al. (2020) Elevation typographus). Global Change Biology, 25, 4048–4063. effects on air temperature in a topographically complex mountain Jonsell, M., Weslien, J. & Ehnström, B. (1998) Substrate requirements of valley in the Spanish Pyrenees. Atmosphere, 11, 656. red-listed saproxylic invertebrates in Sweden. Biodiversity and Con- Nieto, A. & Alexander, K.N.A. (2010) European Red List of Saproxylic Bee- servation, 7, 749–764. tles Publications Office of the European Union. https://doi.org/10. Karlsson, D., Hartop, E., Forshage, M., Jaschhof, M. & Ronquist, F. (2020) 2779/84561 The Swedish malaise trap project: a 15 year retrospective on a coun- OPCC-CTP. (2018) Climate change in the Pyrenees: Impacts, vulnerabil- trywide insect inventory. Biodiversity Data Journal, 8, e47255. ities and adaptation, Jaca, Comunidad de Trabajo de los Pirineos. Karpinski, L., Elek, M. & Wegierek, P. (2021) The role of nature reserves in Oto, N., Václav, Z., Jiří, R., Markéta, M. & Vítězslava, P. (2022) Heritage preserving saproxylic biodiversity: using longhorn beetles trees as an important sanctuary for Saproxylic beetles in the central (Coleoptera: Cerambycidae) as bioindicators. The European Zoological European landscape: a case study from Litovelské Pomoraví, Journal, 88, 487–504. Czech Republic. Forests, 13, 1128. Kim, J.Y., Seo, C., Hong, S., Lee, S. & Eo, S.H. (2019) Altitudinal range-size Parisi, F., Frate, L., Lombardi, F., Tognetti, R., Campanaro, A., distribution of breeding birds and environmental factors for the Biscaccianti, A.B. et al. (2020) Diversity patterns of Coleoptera and determination of species richness: an empirical test of altitudinal saproxylic communities in unmanaged forests of Mediterranean Rapoport’s rule and non-directional rescue effect on a local scale. mountains. Ecological Indicators, 110, 105873. PLoS One, 14, e0203511. Puletti, N., Canullo, R., Mattioli, W., Gawrys, R., Corona, P. & Czerepko, J. Konvicka, M., Benes, J., Cizek, O., Kuras, T. & Kleckova, I. (2016) Has the (2019) A dataset of forest volume deadwood estimates for Europe. currently warming climate affected populations of the mountain ring- Annals of Forest Science, 76, 68. let butterfly, Erebia epiphron (Lepidoptera: Nymphalidae), in low- R Core Team. (2021) R: a language and environment for statistical comput- elevation mountains? European Journal of Entomology, 113, 295–301. ing. R Foundation for Statistical Computing, version 1.3.1056. Vienna, Körner, C. (2012) Alpine Treelines: functional ecology of the global high eleva- Austria: R Core Team. tion tree limits. Basel: Springer Basel. Rahbek, C. (2005) The role of spatial scale and the perception of large- Kotze, D.J. & O’Hara, R.B. (2003) Species decline: but why? Explanations scale species-richness patterns. Ecology Letters, 8, 224–239. of carabid beetle (Coleoptera, Carabidae) declines in Europe. Oecolo- Ranius, T. & Jansson, N. (2000) The influence of forest regrowth, original gia, 135, 138–148. canopy cover and tree size on saproxylic beetles associated with old Lachat, T., Wermelinger, B., Gossner, M.M., Bussler, H., Isacsson, G. & oaks. Biological Conservation, 95, 85–94. Müller, J. (2012) Saproxylic beetles as indicator species for dead- Rasmann, S., Alvarez, N. & Pellissier, L. (2014) The altitudinal niche-breadth wood amount and temperature in European beech forests. Ecological hypothesis in insect-plant interactions. In: Voelckel, C. & Jander, G. Indicators, 23, 323–331. (Eds.) Insect-plant interactions. Wiley-Blackwell, pp. 338–359. Langlands, P.R., Brennan, K.E.C., Framenau, V.W. & Main, B.Y. (2011) Pre- Reymond, A., Purcell, J., Cherix, D., Guisan, A. & Pellissier, L. (2013) Func- dicting the post-fire responses of animal assemblages: testing a trait- tional diversity decreases with temperature in high elevation ant based approach using spiders. Journal of Animal Ecology, 80, fauna. Ecological Entomology, 38, 364–373. 558–568. Roitberg, B.D. & Mangel, M. (2016) Cold snaps, heatwaves, and arthropod Lassauce, A., Paillet, Y., Jactel, H. & Bouget, C. (2011) Deadwood as a sur- growth. Ecological Entomology, 41, 653–659. rogate for forest biodiversity: meta-analysis of correlations between Saint-Germain, M., Buddle, C. & Drapeau, P. (2006) Sampling saproxylic deadwood volume and species richness of saproxylic organisms. Eco- Coleoptera: scale issues and the importance of behavior. Environmen- logical Indicators, 11, 1027–1039. tal Entomology, 35, 478–487. 14619563, 2023, 4, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/afe.12589 by Readcube (Labtiva Inc.), Wiley Online Library on [19/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 680 BOOKWALTER ET AL. Sánchez-Bayo, F. & Wyckhuys, K.A.G. (2019) Worldwide decline of the Wermelinger, B., Flückiger, P.F., Obrist, M.K. & Duelli, P. (2007) Horizontal entomofauna: a review of its drivers. Biological Conservation, 232, and vertical distribution of saproxylic beetles (Col., Buprestidae, Cer- 8–27. ambycidae, Scolytinae) across sections of forest edges. Journal of Schiegg, K. (2000) Effects of dead wood volume and connectivity on Applied Entomology, 131, 104–114. saproxylic insect species diversity. Ecoscience, 7, 290–298. Wetherbee, R., Birkemoe, T., Burner, R.C. & Sverdrup-Thygeson, A. (2021) Schiegg, K. (2003) Saproxylic insect diversity of beech: limbs are richer Veteran trees have divergent effects on beetle diversity and wood than trunks (vol 149, pg 295, 2001). Forest Ecology and Management, decomposition. PLoS One, 16, e0248756. 175, 589. Wilson, R.J., Gutierrez, D., Gutierrez, J., Martinez, D., Agudo, R. & Seibold, S., Bässler, C., Baldrian, P., Reinhard, L., Thorn, S., Ulyshen, M.D. Monserrat, V.J. (2005) Changes to the elevational limits and extent et al. (2016) Dead-wood addition promotes non-saproxylic epigeal of species ranges associated with climate change. Ecology Letters, 8, arthropods but effects are mediated by canopy openness. Biological 1138–1146. Conservation, 204, 181–188. Yanagida, T. (2022) R package ‘misty’: Miscellaneous Functions. (Version Skvarla, M.J. & Dowling, A.P.G. (2017) A comparison of trapping tech- 0.4.6). https://CRANR-projectorg/package=misty niques (Coleoptera: Carabidae, Buprestidae, Cerambycidae, and Cur- culionoidea excluding Scolytinae). Journal of Insect Science, 17, 7. SUPPORTING INF ORMATION Speight, M.C.D. (1989) Saproxylic Invertebrates and Their Conservation. Council of Europe. 92-871-1679-2. Additional supporting information can be found online in the Support- Tang, Y., Horikoshi, M. & Li, W. (2016) Ggfortify: unified interface to visu- ing Information section at the end of this article. alize statistical result of popular R packages. RJ, 8, 478–489. Data S1: Supplementary Material 1. Representative photos of the Thorn, S., Bußler, H., Fritze, M.A., Goeder, P., Müller, J., Weiß, I. et al. (2016) Canopy closure determines arthropod assemblages in micro- three types of traps used in this experiment. A: Attraction trap. B: habitats created by windstorms and salvage logging. Forest Ecology Flight intercept trap. C: Malaise trap. and Management, 381, 188–195. Supplementary Material 2. List of experts involved with identifying Tykarski, P. (2006) Beetles associated with Scolytids (Coleoptera, Scolyti- specimens. Column labeled “Level of Assistance Provided” indicates if dae) and the elevational gradient: diversity and dynamics of the com- munity in the Tatra National Park, Poland. Forest Ecology and the expert in the respective row assisted with author J. Bookwalter’s Management, 225, 146–159. specimen identification of the specified group (“assisted”), or the Ulyshen, M.D. & Hanula, J.L. (2009) Responses of arthropods to large- expert in the respective row performed all identifications of the speci- scale manipulations of dead wood in loblolly pine stands of the fied group (“identified”). southeastern United States. Environmental Entomology, 38, 1005– 1012. Supplementary Material 3. Species abundance (in log scale) found in Viñolas, A., Bentanachs, J. & Maso  , G. (2009) Biodiversitat de coleòpters various aspects. en el Parc Natural de Cadí-Moixero  . Informe. Museu de Ciències Data S2. Appendix 1. List of species captured. AR=Andorran Record. Naturals (Zoologia). Obligate=Obligate saproxylic, facultative=Facultative saproxylic, Vodka, S., Konvicka, M. & Cizek, L. (2009) Habitat preferences of oak- feeding xylophagous beetles in a temperate woodland: implications NS=Not saproxylic, or ND=No data. "Multiple" is a designation given for forest history and management. Journal of Insect Conservation, 13, to a grouping with both non-saproxylic and saproxylic members. AR=- 553–562. Andorran Record. Phyt=Phytophagous. Wfee=Wood-feeding. Para=- Vogel, S., Gossner, M.M., Mergner, U., Müller, J. & Thorn, S. (2020) Opti- Parasitic. Pred=Predator. Fung=Fungivore. Detr=Detritivore. mizing enrichment of deadwood for biodiversity by varying sun exposure and tree species: an experimental approach. Journal of Applied Ecology, 57, 2075–2085. pez, B., How to cite this article: Bookwalter, J.D., Caballero-Lo Wagner, D.L., Grames, E.M., Forister, M.L., Berenbaum, M.R. & Stopak, D. (2021) Insect decline in the Anthropocene: death by a thousand cuts.  pez, B. (2023) Molowny-Horas, R. & Claramunt-Lo Proceedings of the National Academy of Sciences, 118, e2023989118. Understanding the Coleoptera community at the tree-line Wei, T. & Simko, V. (2021) R package ‘corrplot’: Visualization of a Correla- using taxonomic and functional guild approaches. Agricultural tion Matrix (Version 0.92), Version. https://github.com/taiyun/ and Forest Entomology, 25(4), 669–680. Available from: corrplot Wendorff, A. & Schmitt, M. (2019) Leaf beetle decline in Central Europe https://doi.org/10.1111/afe.12589 (Coleoptera: Chrysomelidae s.l.)? ZooKeys, 856, 115–135.