Higher bat and bird γ-diversity in structurally complex forests is driven by distinct α- and β-diversity responses

(max. 10): BETA -FOR, biodiversity restoration, structural heterogeneity, multi -48 scale analysis, managed forests 49 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 4

50 Biodiversity is changing globally, often resulting in declines in ecosystem functioning and 51 stability at the local scale , with serious consequences for both wildlife and human well -being 52 (Sala et al. 2000, Cardinale et al. 2012, Gossner et al. 2016) . Human pressures, especially 53 intensive land use, have traditionally been associated with reduced local species diversity, shifts 54 in community composition, and biotic homogenization (McKinney and Lockwood 1999, Olden 55 and Rooney 2006, Hooper et al. 2012, Jaureguiberry et al. 2022). Despite decades of research, 56 the impacts of anthropogenic drivers can be complex, making predictions of biodiversity 57 patterns difficult. Recent studies have challenged long -held assumptions that human impacts 58 lead to universal patterns of homogenization and landscape fragmentation effects, reinforcing 59 that biodiversity responses are highly taxon-, habitat- , and scale-dependent (Gonçalves-Souza 60 et al. 2025, Keck et al. 2025) . Moreover, biodiversity is multi -faceted, with patterns varying 61 among rare, common, and dominant species within species assemblages, captured by orders of 62 q in Hill numbers (Chao et al. 2014, McGill et al. 2015) . Understanding the different 63 mechanisms driving biodiversity change across different dimensions of diversity and spatial 64 scales is essential for developing effective restoration strategies, especially those aimed at 65 increasing habitat heterogeneity. 66 Throughout large parts of Europe, centuries of economically oriented silviculture have 67 produced structurally simplified, even -aged forest landscapes that lack both early and late 68 successional stages. Yet, these two stages provide sunny conditions and deadwood build -up, 69 respectively, and both are critical for supporting biodiversity (Brunet et al. 2010, Hilmers et al. 70 2018, Aszalós et al. 2022) . The generality of a positive relationship between environmental 71 heterogeneity, such as structural complexity, and diversity has been widely debated (MacArthur 72 and MacArthur 1961, Lundholm 2009, Stein et al. 2014). According to the habitat heterogeneity 73 hypothesis, structurally complex habitats support more species , up to a certain area threshold , 74 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 5 by offering a greater variety of niches (MacArthur and MacArthur 1961, Allouche et al. 2012, 75 Stein et al. 2014, Heidrich et al. 2020). While previous studies support this relationship for bats 76 and birds, with strong evidence for vertical and particularly for horizontal structures (Renner et 77 al. 2018, Heidrich et al. 2020, Rigo et al. 2024), empirical evidence that spans multiple scales, 78 regions, and controlled experimental settings remains scarce. Addressing this challenge 79 requires large-scale, nested study designs and analytical approaches, such as standardized meta-80 analyses, that ensure consistency across regions and account for differences in sample 81 completeness, both of which we present in this study. Moreover, habitat heterogeneity can 82 affect within -patch α - and between -patch β -diversity in contrasting ways, diverging across 83 dissimilar habitats or aligning in more similar or connected ones, ultimately shaping landscape-84 level γ-diversity (Pastro et al. 2011) . Dissimilarities between species assemblages can arise 85 from different processes and are most commonly attributed to species turnover, that is, the 86 replacement of species (Legendre 2014, Soininen et al. 2018) . However, little is known about 87 the relationship dynamics between α-, β-, and γ-diversity, and the role of α- and β-diversity in 88 shaping γ in a controlled experiment. 89 Beyond species richness, understanding functional diversity is crucial for biodiversity 90 conservation. Functional dissimilarities and similarities among species can be quantified by 91 calculating pairwise functional distances between species, derived from differences in species’ 92 traits (Violle et al. 2017) . Functional dissimilarity promotes resource partitioning and 93 coexistence, enhancing functional diversity and contributing to stronger, more stable ecosystem 94 functioning over time . In contrast , functional similarity reflects the overlap in species’ 95 functions, indicating potential ecological interchangeability (Eisenhauer et al. 2023) . While 96 often described as functional ‘redundancy’, such overlap can serve as an ecological insurance 97 mechanism under environmental change , buffering the loss of species and ecosystem 98 functioning (the insurance hypothesis; Yachi and Loreau 1999) , and thus contributing to 99 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 6 ecosystem resilience (Eisenhauer et al. 2023) . Therefore, linking biodiversity to ecosystem 100 functioning requires a multifaceted approach that goes beyond taxonomic diversity (Cadotte et 101 al. 2009, Bae et al. 2018, Bełcik et al. 2020) . Functional diversity, measuring differences in 102 species traits, and phylogenetic diversity, measuring distances in evolutionary relatedness, both 103 reflect ecosystem resilience and stability (Cadotte et al. 2011, 2012, Mori et al. 2013) . 104 Phylogenetically distant species can enhance ecosystem functioning through niche 105 complementarity, which is often reflected in their traits (Cadotte 2013, 2017). Since many traits 106 exhibit phylogenetic signals, meaning closely related species share similar traits (Cadotte et al. 107 2019), phylogenetic diversity can be used as a surrogate for unmeasured but conserved traits 108 (Rothacher et al. 2025). Thus, functional and phylogenetic diversity often respond similarly to 109 environmental change, but not necessarily similarly to taxonomic diversity (Bae et al. 2018). 110 Bats and birds are ideal model taxa to study biodiversity responses to habitat change. 111 As mobile, higher trophic-level organisms, they respond sensitively to both abrupt and gradual 112 environmental disturbances, including bottom -up processes, such as declines in prey 113 availability (Russo et al. 2021) . Their ecological role as predators is crucial for maintaining 114 ecosystem stability (Terborgh et al. 2006) , e.g., by controlling herbivorous insects and plant 115 damage (Mooney et al. 2010, Böhm et al. 2011), and their protection under European law makes 116 them highly conservation -relevant taxa (European Union 1992, 2009) . Bats and birds exhibit 117 diverse morphological, behavioral, and ecological traits that enable them to colonize various 118 forest habitats and to navigate through the complex three-dimensional structure of forest 119 environments (Jung et al. 2012, Froidevaux et al. 2016). With their diversity strongly linked to 120 forest structure (Bradbury et al. 2005, Heidrich et al. 2020) , they particularly benefit from 121 heterogeneous forests, whether at early or late successional stages, which typically include 122 canopy gaps and deadwood structures (Hilmers et al. 2018, Braunisch et al. 2019, Hendel et al. 123 2023, Jung et al. 2025) . Forest bats and birds are associated with deadwood for roosting, 124 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 7 nesting, and foraging (Bouvet et al. 2016, Kortmann et al. 2018, Wild et al. 2025) , as well as 125 with general stand characteristics, such as canopy openness, basal area, or vegetation height 126 (Renner et al. 2018, Rigo et al. 2024) . These responses are further modulated by foraging 127 strategy and spatial scale, for example, open -space foraging bats cover larger areas than 128 gleaning species (Tews et al. 2004, Hendel et al. 2023) . Thus, forest structure can shape 129 diversity patterns differently across spatial scales. Schall et al. (2018), for instance, found higher 130 α-diversity of birds in fine -grained, uneven -aged forests in line with the early work by 131 MacArthur and MacArthur (1961), whereas γ -diversity increased in coarse -grained forests of 132 even-aged stands of different stages. 133 We experimentally enhanced forest structure in homogenized temperate forests by 134 manipulating canopy cover and deadwood structures (Müller et al. 2023). Using a paired design 135 across 11 sites and 234 forest patches, we compared structurally heterogeneous to homogeneous 136 mixed beech ( Fagus sylvatica) production forests, the dominant forest type in Germany. We 137 assessed how enhanced structural heterogeneity affects multiple diversity facets of bats and 138 birds, with emphasis on the responses of infrequent, frequent, and highly frequent species across 139 different spatial scales. 140 Specifically, we tested three hypotheses: 141 1. γ-diversity of bats and birds is higher in structurally heterogeneous than in 142 homogeneous forests, due to a greater variety of available niches. 143 2. Structural heterogeneity at the landscape level (γ -diversity) is mediated by varying 144 contributions of α - and β -diversity, reflecting mechanisms of local species gain or 145 dissimilarity in assemblages among sites. 146 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 8 3. Taxonomic diversity responds differently from phylogenetic and functional diversity in 147 bats and birds, reflecting whether gained species are functionally and evolutionarily 148 similar or dissimilar to the control assemblages. 149 We present a large -scale, spatially explicit (site-specific), and well -replicated field 150 experiment to investigate the effects of forest heterogenization on biodiversity. Using 151 autonomous data collection and automated species identification, a novel meta -analytic 152 framework for paired experimental designs, and multifaceted diversity metrics, we show 153 that γ -diversity increased in structurally enhanced forests for both bats and birds, but 154 through distinct mechanisms. These results add new experimental evidence to the scientific 155 debate on the effects of anthropogenic homogenization on diversity across scales, with key 156 implications for forest restoration. 157 158 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 9

159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 Figure 1 Results from three meta-analyses evaluating γ-diversity (A & B), β-diversity (C & D), 176 and α-diversity (E & F) across 11 experimental forest landscapes. Symbols represent mean 177 predicted changes in standardized diversity: blue triangles for taxonomic diversity (TD), orange 178 circles for phylogenetic diversity (PD), and brown diamonds for functional diversity (FD). Error 179 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 10 bars indicate 95% confidence intervals at fixed sample coverage levels. For bats, the level was 180 set to the 25th quantile of sample coverage across all samples extrapolated to double size: 0.819 181 (TD and PD) and 0.762 (FD) . For birds, it was set to 0.991, the minimum sample coverage 182 across all samples when each is extrapolated to double size. The standardized diversity 183 differences were calculated between each pair of control and treatment forest districts, across 184 all three diversity facets (TD, PD, FD) , and diversity orders referred to as q0, q1 , and q2 . 185 Positive values indicate increased diversity in structurally heterogeneous forests relative to 186 homogeneous controls. Differences are considered statistically significant when the confidence 187 interval does not include zero , as indicated by dar k-colored error bars. TD, PD, and FD are 188 expressed in the same units of species, lineage, or functional group equivalents and can be 189 directly compared within each diversity level (α, β, γ). To account for variation in patch 190 numbers between sites, we applied a Jaccard -type turnover transformation (1 -S) of 191 multiplicative β-diversity, which quantifies dissimilarity between assemblages relative to γ (see 192 methods). Across levels, only α and γ can be directly and meaningfully compared (e.g., for q0: 193 TD α < TD γ). For detailed results, see forest plots in the supplements (Document S1. Fig. S3-194 29 and Fig. S30-56). 195 We identified a total of 17 bat species from 936 bat call recordings and 72 bird species from 196 16,147 bird sound recordings across all sites. As autonomous recorders do not provide true 197 abundance data, we generally used an incidence-based approach for the analyses (Kortmann et 198 al. 2025). To equalize sample completeness and avoid potential biases within both taxonomic 199 groups, bat assemblages were standardized to the 25th quantile of sample coverage across all 200 samples, extrapolated to double size: 0.819 (TD and PD) and 0.762 (FD), and for birds to 0.991, 201 the minimum sample coverage across all samples when each is extrapolated to double size . 202 Taxonomic (TD), phylogenetic (PD), and functional diversity (FD) are expressed in the same 203 units of species, lineage, or functional group equivalents and can be directly compared within 204 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 11 each diversity level (α, β, γ). Across levels, comparisons are only meaningful between α and γ, 205 as multiplicative β -diversity is a ratio and represents how much of γ -diversity comes from 206 differences between patches. Across all forest landscapes, bat γ-diversity consistently increased 207 from homogeneous control to heterogeneous treatment forests across all diversity facets (TD, 208 PD, FD) and diversity orders, emphasizing infrequent (q0), frequent (q1), and highly frequent 209 (q2) species, lineages, and functional groups (Fig. 1, A). For birds, significant increases in γ -210 diversity were observed for TD and FD across all diversity facets and orders (Fig. 1, B). In bats, 211 gains in γ-diversity were primarily driven by increased β-diversity (Fig. 1, C & E), whereas in 212 birds, α-diversity was the only driver for γ (Fig. 1, D & F). Gains of TD for bats were strongest 213 in γ- and β-diversity across all diversity orders, exceeding gains in PD and FD (Fig. 1, A & C). 214 In contrast, birds showed the largest increases in γ - and α -diversity for FD at q1 and q2, 215 exceeding the increases of TD (Fig.1, B & F). 216 217 γ-diversity increased across heterogeneous forest landscapes for bats and birds 218 For bats, taxonomic γ-diversity increased significantly in structurally enhanced landscapes with 219 gains of approximately two effective numbers of species across all diversity orders (q0 = 1. 8, 220 q1 = 2.2, q2 = 2.1) (Fig. 1, A). These were the strongest gains among the three diversity facets, 221 exceeding those observed for PD (q0 = 1.4, q1 = 1.1, q2 = 0.90) and FD (q0 = 0.51, q1 = 0.37, 222 q2 = 0.30). 223 Birds also showed taxonomic gains of about two effective species (q0 = 2.0, q1 = 1.4, 224 q2 = 1. 8) (Fig. 1, B) . FD showed the strongest gains in γ-diversity for increasingly frequent 225 groups (q0 = 2.0, q1 = 2.2, q2 = 2.1). 226 β-diversity increased for bats but decreased for birds across treatment forest districts 227 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 12 Bats showed increases in taxonomic (q0 = 0.21, q1 = 0.17, q2 = 0.089) and phylogenetic (q0 = 228 0.11, q1 = 0.08 0, q2 = 0.0 49) β-diversity, for all orders of q, with effect sizes declining from 229 infrequent to highly frequent species (Fig. 1, C). 230 In contrast, birds showed overall significant decreases in β-diversity for taxonomic (q0 231 = -0.036, q1 = -0.031, q2 = -0.035) (Fig. 1, D), phylogenetic, except at q0 (q1 = -0.024, q2 = -232 0.040), and functional diversity (q0 = -0.018, q1 = -0.019, q2 = -0.021). 233 α-diversity showed stronger gains for birds than for bats in treatment patches 234 Bats showed significant increases in taxonomic α-diversity among highly frequent species (q2 235 = 0.37) (Fig. 1, E). Functional diversity increased significantly for frequent and highly frequent 236 groups (q1 = 0.30, q2 = 0.28). 237 Birds showed consistent significant gains in taxonomic α-diversity, with gains of about 238 two effective species (q0 = 2. 2, q1 = 2.0, q2 = 2.0) (Fig. 1, F) . Phylogenetic α -diversity 239 increased significantly across all orders (q0 = 0.55, q1 = 0.51, q2 = 0.58), but with smaller effect 240 sizes than for FD, which showed the strongest gains for increasingly frequent functional groups 241 (q0 = 2.1, q1 = 2.6, q2 = 2.3). 242

243 Increased γ-diversity in heterogeneous forest landscapes is driven by distinct α- and β-diversity 244 mechanisms for bats and birds 245 Overall, γ-diversity of bats and birds increased in structurally heterogeneous forests, except for 246 birds’ PD, supporting our first hypothesis and reinforcing the habitat-heterogeneity hypothesis: 247 small-scale interventions that enhance habitat complexity can promote biodiversity by 248 expanding ecological niche availability. However, consistent with our second hypothesis, the 249 underlying mechanisms driving γ-diversity differed, reflecting distinct contributions of α - and 250 β-diversity for bats and birds. 251 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 13 In bats, increases in landscape -scale diversity are primarily driven by increased β -252 diversity, indicating greater dissimilarity of assemblages between patches with different light 253 conditions and deadwood amounts. Local α-diversity gains of bats were small and only became 254 apparent at the landscape scale, particularly for species with high occurrence frequency (q1, 255 q2). This is consistent with the dominance of β -diversity hypothesis, which states that γ -256 diversity is driven by β- rather than α-diversity (Tscharntke et al. 2012). This pattern has been 257 supported by other studies (Farnsworth et al. 2014, Morante -Filho et al. 2016, Costa and 258 Schmidt 2022) and aligns with bats’ high mobility and wide -ranging foraging behavior. Their 259 habitat use is strongly shaped by forest structure, with different foraging guilds specializing in 260 distinct forest structures. This niche differentiation allows them to locate and exploit spatially 261 distributed resources while reducing interspecific competition (Schnitzler and Kalko 2001, 262 Müller et al. 2012, Charbonnier et al. 2014). Closed-space foragers (e.g., Myotis bechsteinii, M. 263 nattereri, Plecotus auritus ) prefer dense forest interiors or canopies. In contrast, edge -space 264 foragers (e.g., Barbastella barbastellus, Myotis myotis, Pipistrellus pipistrellus) prefer more 265 open forest stands with sparse understory and leaf-littered ground, and can exploit forest edges 266 and gap margins. Open-space foragers (e.g., Eptesicus and Nyctalus spp.) hunt in open canopies 267 or above gaps. In this study, homogeneous control forests with closed canopies mainly 268 supported closed- and edge-space foragers. Aggregated treatments created canopy gaps and 269 edge structures, likely attracting open-space foragers (Müller et al. 2013, Kortmann et al. 2018, 270 Jung et al. 2025) (also see Document S1, Fig. S1, and Tab. S1) . A study comparing forest 271 patches with high and low structural complexity showed that TD, PD, and FD of aerial-hawking 272 bats increased at habitat edges, highlighting the importance of heterogeneous landscape 273 elements (López-Baucells et al. 2022). Distributed treatments with deadwood structures under 274 closed-canopy likely favored closed -space specialists, adapted to navigate in dense habitats, 275 which harbor high prey abundance (Müller et al. 2012, Hochrein et al. 2025, Jung et al. 2025). 276 Additionally, all forest bats depend on structures associated with old or damaged trees for 277 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 14 roosting, with species-specific preferences for cavities or crevices (Kortmann et al. 2018). Our 278 treatments replicated these features across forest patches by varying deadwood volume and 279 type, such as snags and habitat trees. Even lying deadwood volume and understory vegetation 280 can enhance bat foraging activity, suggesting the influence of bottom-up processes (Jung et al. 281 2025). Social information transfer about new roosts (Kerth and Reckardt 2003) or food 282 resources may have facilitated resource exploitation in heterogeneous forest landscapes (Safi 283 and Kerth 2007), promoting differentiation in local bat assemblages across patches (increased 284 β-diversity), especially among rarer, specialized species (β: TD & PD at q0). 285 In contrast, bird γ -diversity increased exclusively through temporal gains in within -286 patch α-diversity, despite declining β-diversity, underlining the effect of small-scale and often 287 randomly distributed structural features. Although it has been suggested that opposing α - and 288 β-trends might cancel each other out (Farnsworth et al. 2014) , our results suggest that 289 structurally heterogeneous forests can simultaneously enhance local and regional bird diversity 290 while reducing assemblage differentiation among sites. Unlike bats, most forest bird species 291 operate during the breeding season on smaller spatial scales within fixed territories that are 292 actively and aggressively defended (Krebs 1982, Poesel and Dabelsteen 2005) , limiting their 293 spatial mobility and thus also potential changes in assemblages across patches. Deadwood and 294 canopy gaps can locally and temporally increase insect availability (Lettenmaier et al. 2022, 295 Rothacher et al. 2023, 2025) , leading to a localized diversification of resources , as already 296 shown at the scale of a few trees in the seminal work by MacArthur and MacArthur (1961) . 297 This increase in resources may have attracted infrequent species, such as rare cavity- or 298 deadwood-associated specialists (α: TD at q0), and increased local species richness (Whittaker 299 1972). Yet, as the same rare species may respond similarly across sites, this can lead to increases 300 in α - and γ -diversity without a corresponding rise in β -diversity. Similarly, increases also 301 occurred among more common species (α: TD at q1 and q2), likely leading to the same species 302 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 15 quickly occupying niche spaces across all patches , making assemblages more similar, and 303 further reducing β-diversity. However, unlike our results, a study in heterogeneous mixed 304 forests in Germany found no significant effects of structural features on bird α -diversity, but 305 did find effects on β-diversity, particularly related to light availability, which likely increased 306 insect resources and foraging space (Schauer et al. 2023) . At early successional stages, forest 307 gaps increase sunlight exposure, supporting herbaceous understory plants that produce seeds 308 and fruits, providing resources for both omnivorous and herbivorous bird species. β-diversity 309 effects may have decreased over time (Freitag Kramer et al. 2025) , as positive effects of 310 structural heterogeneity on diversity are typically strongest during the early and late 311 successional stages (Hilmers et al. 2018) . In our study, sampling occurred 4 to 7 years post -312 intervention, when forest succession had already reached an intermediate stage , with some 313 patches showing intensive regeneration (University Forest). 314 Diversity facet responses: bats show patterns of functional similarity, birds gain dissimilar 315 traits 316 Bats showed greater increases in TD than in PD or FD at both β - and γ-scales (TD > PD, FD), 317 supporting our third hypothesis, and suggesting that functionally similar species contributed the 318 most to the detected diversity gains (Flynn et al. 2009). An increase of two effective species at 319 standardized sample coverage is substantial for bats, given that only 25 species occur in 320 Germany (Bundesamt für Naturschutz 2025) . Unlike birds, bats are not highly territorial over 321 feeding areas and can coexist despite exploiting similar resources by differing in foraging 322 strategies and operating at large spatial scales. However, because bats can be classified into 323 only a few broad guilds, this lead s to trait overlap within guilds (Denzinger and Schnitzler 324 2013). Moreover, all observed bat species are closely related, belonging to the same family of 325 Vespertilionidae, which is divided into a few subfamilies that broadly correlate with their 326 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 16 foraging guilds, e.g., Myotinae (Myotis) occupy diverse niches, such as edge and closed spaces, 327 and Vespertilioninae (e.g., Nyctalus, Eptesicus, Vespertilio, Pipistrellus) dominate open spaces. 328 Phylogenetic diversity patterns in birds, contrary to our third hypothesis, did not align 329 closely with those of FD. Both diversity facets can capture complementary aspects about 330 community patterns (Cadotte et al. 2013), e.g., functional and phylogenetic bird diversity can 331 respond similarly to vegetation or landscape structure, but differently from taxonomic diversity 332 (Klingbeil and Willig 2016, Bae et al. 2018) . However, in our study, FD and TD responded 333 more strongly to heterogeneous forest structure than PD (FD, TD > PD), suggesting that the 334 gained species at the α -scale contributed only a few new phylogenetic lineages, but these did 335 not translate into an overall increase in PD at the γ-scale. Over 60% of the observed bird species 336 belonged to the species-rich order of Passeriformes, most of which were widely distributed and 337 occurred in both control and enhanced forests (see Document S1, Fig. S2). Only a few species 338 were unique to the enhanced forests, add ing a new lineage : the rare ring ouzel ( Turdus 339 torquatus), and the green sandpiper (Tringa ochropus), a small wader, breeding in wet forests, 340 introducing an entirely new order to our phylogenetic tree. Although PD and FD are often 341 correlated, they are not interchangeable. Phylogenies and traits may capture different ecological 342 dimensions: phylogeny reflects deeper, ancestral similarities, while traits may highlight more 343 recent evolutionary changes (Cadotte et al. 2019) . For example, this is highly pronounced in 344 the functional dissimilarity of bat species in the genus Myotis (Document S1, Fig. S1, Tab. S1). 345 The closer alignment of FD with TD, relative to PD, points to short -term ecological processes 346 driving species gains at the patch level (Aguirre et al. 2016) , e.g., temporal increase in food 347 resources or nesting sites to which birds quickly respond. Among increasingly frequent species 348 (at q1 and q2) of α- and γ-diversity, gains in FD were higher than in TD (FD > TD), suggesting 349 that the added species contributed dissimilar or complementary ecological functions (Moreno 350 et al. 2024). As birds are territorial, increases in effective species richness were modest at the 351 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 17 local scale (~2 species). However, these gains added 2–2.5 new functional groups among more 352 frequent species, suggesting greater functional diversification without increasing competition 353 (Moreno et al. 2024). 354 Diversity order patterns: infrequent species exhibit variable responses, increasingly frequent 355 species respond uniformly 356 Within each diversity level (α -, β -, and γ), response patterns remained relatively consistent 357 across diversity orders (q0–q2) for both bats and birds. However, diversity differences between 358 control and treatment forests were more variable at q0, while effect sizes were often stronger 359 and more stable at q1 and q2, suggesting a more heterogeneous response among infrequent 360 species (q0), while increasingly frequent species (q1, q2) responded more uniformly to 361 structural changes. These trends mirror findings from a bird sound study in Ecuador, which also 362 found more variance in diversity values for q0 (Kortmann et al. 2025). However, dissimilarity-363 based methods have been criticized for producing misleading results, when species are 364 undersampled and dissimilarity values become saturated when many sites share no species, 365 which is often the case for sampling rare species, emphasized at q0 (Tuomisto et al. 2012). Our 366 standardization of sample coverage within each taxa mitigates this sampling bias to some 367 extent. Furthermore, as we used acoustic data, which provides incidence-based data, we lacked 368 information on species abundances reflecting their dominance within assemblages (Tucker et 369 al. 2016). Yet, this was also addressed by our method, by using repeated measurements at each 370 patch and analyzing our data on a nightly basis for bats and a daily basis for birds, we increased 371 the number of sampling units. 372 Conservation and restoration implications 373 Environmental heterogeneity, particularly structural complexity, is strongly associated with 374 increased species diversity across spatial scales (Stein et al. 2014). Our study supports this link, 375 showing that restoring structural heterogeneity in production forests can promote bat and bird 376 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 18 diversity, however, effects vary across spatial scales and taxa (Schall et al. 2018, Heidrich et al. 377 2020, Schauer et al. 2023). 378 For bats, enhancing the spatial variation of structural heterogeneity at broader scales is 379 essential. Treatments that create contrasting canopy conditions (open vs. closed) increase 380 horizontal structural diversity and microclimate variation , which was confirmed on the same 381 patches by other studies (Thom et al. 2020, Pierick et al. 2025), altering understory vegetation 382 and prey availability (Hendel et al. 2025). For example, moths are often associated with closed 383 canopies, while beetles and other arthropods, both saproxylic and non-saproxylic, depend on 384 deadwood (Seibold et al. 2016, Hendel et al. 2025) . Diverse deadwood structures, e.g., lying, 385 standing, and different decay stages, provide diverse microhabitats for both arthropods and 386 roosting bats (Tillon et al. 2016, Seibold et al. 2023, Rothacher et al. 2023) , promoting 387 assemblage differentiation between stands and increasing bat diversity at the landscape scale 388 (Barnes et al. 2016, Jung et al. 2025). 389 In contrast, birds responded positively to increased heterogeneity regardless of spatial 390 distribution. Opening closed forest canopies can promote forest specialists and generalists, 391 enhancing overall γ-diversity (Vanderwel et al. 2007, Schall et al. 2018). In forests near cultural 392 landscapes, forest gaps can attract some opportunistic birds of open habitats (Żmihorski et al. 393 2016) such as the Yellowhammer ( Emberiza citrinella ), the Common linnet ( Linaria 394 cannabina), or the European turtle dove (Streptopelia turtur), as observed on our experimental 395 patches located in some regions (e.g. , Bavarian Forest, University Forest, and Lübeck). 396 Furthermore, creating gaps initiates forest succession, enhancing vertical structural diversity 397 and supporting overall bird diversity (Lesak et al. 2011, Vogeler et al. 2014, Hanle et al. 2020), 398 particularly shrub-nesting birds and short-distance migrants (Graser et al. 2025). Additionally, 399 deadwood structures can serve as nesting sites, foraging substrates, and singing perches (Bouvet 400 et al. 2016) . For example, snags within small forest gaps can expand territories and promote 401 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 19 bird diversity in deciduous forests , especially for cav ity-nesting species that arrive relatively 402 late from their wintering grounds (Müller 2005, Lewandowski et al. 2021). 403 Effective management practices should promote structural heterogeneity at the local and 404 landscape scales. This can be achieved through a benign neglect strategy in large , protected 405 areas, which deliberately allows natural disturbances to occur, thereby creating heterogeneous 406 landscapes with suitable structures for bats and birds (Müller et al. 2010, Kortmann et al. 2018, 407 Cours and Duflot 2025) . Based on our results, we suggest to locally promot e gaps and 408 deadwood structures whenever possible to enhance bird diversity in production forests . 409 Moreover, combining different silvicultural approaches, e.g., gap felling, shelterwood, or 410 single-tree selection, applied with varying intensity in a spatially explicit design (Schall et al. 411 2018) will foster habitat heterogeneity and benefit both birds and bats. In turn, supporting bats 412 and birds enhances natural pest control, contributing to healthy and resilient forest ecosystems 413 (Mäntylä et al. 2011, Blažek et al. 2021). 414 Study limitations 415 Some site -specific deviations from general patterns were observed, but our meta -analytic 416 approach minimized their influence through appropriate weighting. For example, the unrealistic 417 high estimate for taxonomic α -diversity of bats at a single study site in Hunsrück (TD q0 in 418 control = 35.8) had minimal influence on the overall outcomes (weight = 0.05%). Furthermore, 419 diversity estimates were based on incidence data from acoustic recordings, reflecting presence 420 or activity instead of absolute abundances, while this limits direct interpretation of species 421 counts, it provides robust, standardized insight into acoustic community structure (Kortmann et 422 al. 2025). However, this approach also requires a sufficient number of sampling units per patch 423 to ensure analytical robustness. Thus, we recommend using at least eight sampling units per 424 patch. To test the robustness of observed patterns, meta-analyses can be repeated with varying 425 sample coverage levels. Consistent results strengthen the confidence in inferences , e.g., using 426 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 20 the median of sample coverage across all samples extrapolated to double size, yielded the same 427 patterns (see bat analysis with SC threshold = C50% in Supplementary Document S1, Fig. S57). 428 Our study focused on assessing spatial differences in species assemblages, based on a one-year 429 sampling campaign at an intermediate stage of forest succession. This approach allowed species 430 enough time to redistribute among patches after the interventions, however, we could not 431 analyze temporal changes in assemblage, which can fluctuate over time (Rolls et al. 2023). This 432 should be considered in future studies. 433

and outlook 434 Global biodiversity is undergoing significant changes, with α - and β-diversity playing distinct 435 roles in shaping γ-diversity (Gonçalves-Souza et al. 2025, Keck et al. 2025) . In this study, we 436 experimentally enhanced structural between -patch complexity in homogenized temperate 437 production forests by modifying canopy cover and deadwood availability. Our results 438 confirmed that (1) γ -diversity of bats and birds is higher in structurally heterogeneous than 439 homogeneous forest landscapes, except for birds’ PD, and (2) γ-diversity is driven by distinct 440 contributions of α - and β-diversity in each group. Bats primarily respond to spatial variation 441 across the landscape (β-diversity), consistent with their high mobility and foraging range, while 442 birds respond mainly through local species gains (α -diversity), influenced by opportunistic 443 behavior and competition. In line with our third hypothesis, (3) TD increased more strongly 444 than PD and FD in heterogeneous forests for β - and γ-diversity of bats, suggesting functional 445 similarity in gained species. In contrast, for birds, FD aligned more closely with TD than PD, 446 indicating that newly gained bird species were functionally dissimilar to the control 447 assemblages. These contrasting responses underscore the importance of taxon -specific 448 approaches in biodiversity-oriented forest management, even within the same trophic level. By 449 contributing to a broader understanding of how α - and β -diversity respond to habitat 450 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 21 heterogeneity, our study gives critical insights for developing effective, scale -sensitive 451 restoration strategies. 452 Resource availability 453 Lead contact 454 Requests for further information and resources should be directed to and will be fulfilled by the 455 lead contact, Clara Wild ([email protected]). 456

availability 457 This study did not generate new unique reagents. 458 Data and code availability 459 ● Bat species, trait, and phylogenetic data have been deposited at Zenodo and are publicly 460 available as of the date of publication at [DOI]. 461 ● Bird species, trait, and phylogenetic data have been deposited at Zenodo and are 462 publicly available as of the date of publication at [DOI]. 463 ● All original code has been deposited at Zenodo and is publicly available at [DOI] as of 464 the date of publication. 465 ● Any additional information required to reanalyze the data reported in this paper is 466 available from the lead contact upon request. 467 Acknowledgments 468 We thank all local managers for their support in our research. In particular, we thank K. Kraus, 469 A. Kieffer, G. Bach, J. Torres, K. Kallnik, M. Mauermann, T. Becker, T. Rosch, W. Weisser, 470 the interns of the Bavarian Forest National Park, and all other assistants. We also thank J. Thein 471 and C. Franz for providing bat recorders, and R. Martin for validating our bird sound data. The 472 experimental sites were established within the project 'Beta -Diversität experimentell für 473 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 22 nachhaltige Waldbewirtschaftung in Mitteleuropa' of the German Federal Environmental 474 Foundation (DBU) (project no. 34488/01) and the BioHolz project ( grant no. 01LC1323A ). 475 CW, OM, JR, ST, SK, and JM received funding from the German Research Foundation (DFG) 476 within the research unit ’BETA -FOR’ (project no. 459717468). JR was further supported by 477 funding from the Bavarian State Ministry for Food, Agriculture, and Forestry (StMELF) (grant 478 no. L062). MK acknowledges funding by the Bavarian Research Institute for Digital 479 Transformation (bidt), an institute of the Bavarian Academy of Sciences and Humanities 480 (ROOT: Real-time earth Observation of fOrest dynamics and biodiversiTy (KON-22-024)). 481 Author contributions 482 Conceptualization: JM, CW; Data curation: CW, SK, OM; Formal analysis: CW, JM, OM; 483 Funding acquisition: JM, ST; Investigation: CW, JR, RP, RH, KS , SH , SM , MJ, ND, JS ; 484 Methodology: AC, OM, SK, MK; Project administration: CW, JR, RP , OD; Resources: JM; 485 Software: AC, PYC, SK, OM; Supervision: JM, ST , MC ; Validation: JM, CW, OM; 486 Visualization: CW, MK; Writing – original draft: all authors; Writing – review & editing: all 487 authors. 488 Declaration of interests 489 The authors declare no competing interests. 490 Declaration of generative AI and AI-assisted technologies 491 During the preparation of this work, the authors used ChatGPT and Grammarly in order to 492 improve the orthography, grammar, and writing style. Additionally, the authors used the AI -493 based tools BirdNET and batIdent to assist in the identification of bird and bat species. After 494 using these tools, the authors reviewed and edited the content as needed and take full 495 responsibility for the content of the published article. 496 Supplemental information 497 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 23 Document S1. Figures S1–S57 and Table S1–S4 498 Table S2. Excel file containing additional data too large to fit in a PDF, related to Figure 2 499 500 501 502 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 24

503 Study area and experimental design 504 505 506 507 508 509 510 511 512 513 Figure 2 Experimental design of the BETA-FOR research unit. (A) Map of Germany showing 514 six study regions with 11 experimental forest landscapes: Lübeck (L11), Saarland (S10), 515 University Forest (U01 –U03), Passau (P08), Hunsrück -Hochwald National Park (H09), and 516 Bavarian Forest National Park (B04–B07). Each landscape includes two paired forest districts: 517 a treatment district with enhanced structural heterogeneity (orange), and a control district 518 representing homogeneous production forest (green). (B) Schematic illustration of an 519 experimental landscape, illustrating spatial scales and diversity levels assessed: each district 520 contains 9 patches (15 in the University Forest), used to evaluate within -patch α -diversity, 521 between-patch β-diversity, and overall γ -diversity. (C) Overview of 14 treatment types and a 522 control patch. The depicted control patch serves as local control within each treatment district 523 and further represents the stand structure of all patches in the control district. Treatment patches, 524 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 25 established only in treatment districts, vary in spatial arrangement, either aggregated, within a 525 forest gap, or distributed, under a closed canopy. The control patch and the first eight treatments 526 from left to right (Stumps remain, Logs, Snags, Logs + Snags) are replicated across all 527 experimental landscapes, while six additional treatments (Total tree removed, Crowns remain, 528 Habitat trees) are exclusive to the University Forest. Adapted from a graph provided by Rike 529 Schwarz. 530 531 This study is part of the collaborative research unit BETA-FOR (Müller et al. 2023). The study 532 design comprises 234 experimental patches distributed across 11 forest landscapes in six 533 regions of Germany, spanning a gradient of mean elevation above sea level (a.s.l.): Lübeck (41 534 m), Saarland (296 m), the University Forest of the Julius -Maximilians-Universität Würzburg 535 (336 m), Passau (496 m), the Hunsrück -Hochwald National Park (654 m), and the Bavarian 536 Forest National Park (931 m) (Fig. 2 , A; for detailed site information, see supporting 537 information, Table S2. Excel file). All forests are managed and reflect the typical tree 538 composition of Central European mixed forests. These are primarily dominated by deciduous 539 tree species, with conifers comprising 1% to 30% of the total basal area per forest. European 540 beech (Fagus sylvatica) is the dominant species in five of the regions, while the University 541 Forest features a mixed composition of beech, maple ( Acer spp.), ash ( Fraxinus excelsior), 542 hornbeam (Carpinus betulus), and oak (Quercus spp.). 543 Each experimental landscape consists of two paired forest districts: the treatment district 544 is characterized by enhanced structural heterogeneity (heterogeneous forest), and the control 545 district represents an untreated production forest (homogeneous forest). This design allows for 546 contrasting findings to a reference control and to study biodiversity at different scales (Keck et 547 al. 2025) . Within each forest district, nine patches, and 15 in the University Forest, were 548 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 26 established, each measuring 50 m × 50 m. This design enables assessment of within -patch α-549 diversity, between-patch β-diversity, and overall γ-diversity within each district (Fig.2, B). 550 Structural enhancements in heterogeneous forests increased between-patch complexity 551 by assigning each patch within a treatment district a distinct combination of canopy openings 552 and deadwood structures. These treatments were not repeated within districts but were 553 replicated across the 11 experimental landscapes, except for six treatments implemented only 554 in the University Forest. Aggregated treatments created canopy gaps with diameters of 25 –30 555 m, whereas distributed treatments maintained a closed canopy. These interventions resulted in 556 microclimatic contrasts, with forest gaps exhibiting higher UV radiation and greater 557 temperature fluctuations, warmer during the day and cooler at night, compared to the more 558 thermally buffered conditions under closed canopies (Thom et al. 2020). Deadwood structures, 559 such as stumps, snags, logs, and combinations thereof (with additional elements in the 560 University Forest), were established both in canopy gaps and under closed canopies (Fig. 2, C). 561 Patches and treatments were established in the winters of 2015/16 (Bavarian Forest and 562 Passau), 2016/17 (Saarland, Lübeck, and Hunsrück), and 2018/19 (University Forest), allowing 563 sufficient time for colonization by forest fauna. 564 Bat sampling and data preparation 565 Due to the wide geographic distribution of study sites across Germany, sampling was conducted 566 in 2022 in the University Forest, Bavarian Forest National Park, and Passau regions and in 2023 567 in the Saarland, Hunsrück, and Lübeck regions. Bats were surveyed using autonomous bat 568 recorders (batcorder, generations 2 and 3; ecoObs GmbH, Nürnberg, Germany), which recorded 569 echolocation calls for at least one night per patch and month in May, June, and July. Additional 570 sampling nights in May, June, and August were included to compensate for adverse weather or 571 other suboptimal sampling conditions. The limited availability of recording devices restricted 572 sampling to one pair of forest districts per night. One batcorder was installed in each patch 573 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 27 within a 15-meter radius of the patch center, mounted at a height of 1.50 to 1.70 m on the trunk 574 of a thin tree, leafless at recorder height to minimize background noise. The antenna was angled 575 upward toward the patch center or, when available, toward structural features such as canopy 576 gaps or skid trails, which are often used by bats for navigation. Batcorders were configured 577 with the following settings: recording quality set to 20, post -trigger duration of 400 ms, 578 threshold level at −27 dB, and a critical frequency of 16 kHz. Recording was continuous and 579 set to start one hour before sunset and end one hour after sunrise. Acoustic data were processed 580 using bcAdmin4 (Version 1.2.6, ecoObs GmbH) to organize and analyze call sequences, 581 followed by species-level identification using batIdent (Version 1.5, ecoObs GmbH), following 582 the approach in Müller et al. (2013) . The minimum detection probability was set to 0.6. For 583 subsequent analyses, only species-level identifications were considered, with the exception of 584 Myotis brandtii/mystacinus and Plecotus spp. , which cannot be reliably distinguished 585 acoustically. However, for the assignment of trait data and phylogenetic information, these 586 groups were attributed to Myotis mystacinus and Plecotus auritus, respectively. Additionally, 587 species not known to occur in the respective study regions were excluded based on expert 588 knowledge and verification of individual audio file s. Raw data were initially expressed as 589 ‘minute counts’, defined as the number of minutes within a recording session during which at 590 least one bat call was detected (Hochrein et al. 2025) . Since individuals cannot be reliably 591 identified from acoustic recordings, we converted these data to incidence -based presence –592 absence format (0 = no detection, 1 = detection) for each sampling night (‘ sampling unit’ for 593 bats; Document S1, Tab. S3). 594 To construct a phylogenetic tree, species -specific information was extracted by 595 subsetting the mammalian megatree published by Upham et al. (2019). A customized 596 phylogeny was generated using the ‘phylo.maker’ function from the U.PhyloMaker package, 597 following the workflow provided in Jin and Qian (2023). The resulting tree was saved in .nwk 598 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 28 format. For visualization, we used the ‘ggtree’ function from the ggtree package (Yu 2022), 599 following the approach described by Kortmann et al. (2025) (Document S1. Fig. S1). 600 We selected seven traits that represent morphological, behavioral, and functional 601 adaptations of bats related to using forest structures (Document S1. , Tab. S1). Trait data for 602 each species were primarily obtained from the EuroBaTrait 1.0 dataset (Froidevaux et al. 2023) 603 and supplemented with information from Müller et al. (2012, 2013). All morphological traits 604 were loge-transformed. For Ear length and Tail length, we calculated residuals from a linear 605 model with the ‘lm’ function (R Core Team 2024) using loge-transformed Forearm length as 606 the predictor to account for body size scaling. In addition, loge(Body mass), Call duration, and 607 Maximum frequency were standardized (mean = 0, SD = 1) using the ‘decostand’ function 608 (method = standardize) from the vegan package (Oksanen et al. 2022). 609 Bird sampling and data preparation 610 Bird sampling, similar to bat surveys, was conducted in 2022 in the University Forest, Bavarian 611 Forest National Park, and Passau regions, and in 2023 in the Saarland, Hunsrück, and Lübeck 612 regions. Birds were sampled using autonomous sound recorders (BAR and BAR -LT; Frontier 613 Labs, Australia), which recorded ambient sound daily from March to Ju ne. One recorder was 614 installed within a 15 -meter radius of the patch center, mounted on the trunk of a thin tree, 615 leafless at recorder height to optimize recording quality. Recorders were placed at a height of 616 1.50 to 1.70 m with the microphone facing downward. Recorders were set to record for 2 617 minutes every 12 minutes, using a sampling rate of 44.1 kHz. Recordings were scheduled daily 618 to begin two hours before sunrise and continue until four hours after sunrise, and again from 619 three hours before sunset to three hours after sunset. For subsequent analyses, the recording 620 period was restricted to March 24 to June 15, coinciding with the peak breeding and singing 621 activity of most bird species. Sound data were analyzed using BirdNET version 2.4 . We 622 extracted all identifications with a confidence value above 0.5. In a second step , ~15,000 623 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 29 expert-validated species identifications from our study sites were used to calculate a species -624 specific threshold, following the method of Wood and Kahl (2024). For some rare species, we 625 applied a strict confidence threshold of 0.95 (Seibold et al. 2024). To achieve the same level as 626 for bats (one-night campaigns), we aggregated the bird data on a daily basis, with each species 627 counted only once per daily recording period . The resulting dataset was compiled in an 628 incidence-based presence–absence format (0 = no detection, 1 = detection) for each sampling 629 day (‘ sampling unit’ for birds ; Document S1, Tab. S4 ). The final data set was once more 630 checked by JM for plausibility of species occurrences in the study regions, and implausible 631 species were verified and, if erroneous, removed. 632 Phylogenetic information was obtained from the bird megatree (Jetz et al. 2012) and 633 adjusted to the observed species using the same procedure applied to bats, employing the 634 U.PhyloMaker package (Jin and Qian 2023). The resulting tree was saved in .nwk format and 635 visualized using the ‘ggtree’ function from the ggtree package (Yu 2022) (Document S1., Fig. 636 S2). 637 Similar to the approach used for bats, we selected seven bird traits from the AVONET 638 dataset (Tobias et al. 2022) that reflect adaptations to forest structure use (Document S1., Tab. 639 S2). All morphological traits were log e-transformed. For Beak length, Tarsus length, and Tail 640 length, we calculated residuals from linear models using the ‘lm’ function (R Core Team 2024), 641 controlling for log e-transformed Body mass. Additionally, log e-transformed Body mass was 642 standardized (mean = 0, SD = 1) using the ‘decostand’ function (method = "standardize") from 643 the vegan package (Oksanen et al. 2022). 644 Data analysis 645 All analyses were conducted using R version 4.4.2. To assess diversity differences between 646 treatment (structurally heterogeneous) and control (homogeneous) forest districts, we applied a 647 comparative meta-analytic approach. This approach combines two core concepts, embedded 648 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 30 into the iNEXT .3D framework (Chao et al. 2021) : sample coverage, to standardize sample 649 completeness, and Hill numbers, to unify diversity measurements across different orders of q, 650 including species richness (q = 0), Shannon diversity (q = 1), and the Simpson diversity (q =2). 651 These tools provide robust, comparable biodiversity estimates across sites (Roswell et al. 2021, 652 Kortmann et al. 2025). 653 As our data is based on incidences (presence-absence), and not abundances, we use Hill-654 Chao numbers (q = 0, 1, 2; hereafter ‘diversity orders’), which extend the original Hill number 655 approach, developed for taxonomic diversity (TD), to phylogenetic (PD) and functional 656 diversity (FD) (hereafter ‘diversity facets’) and incidence-based occurrence data (Chao et al. 657 2021). For bats, only four sampling units (nights) per patch were available, which can be 658 insufficient for reliable inference using incidence -based models, particularly for FD (Colwell 659 and Chao 2022) (see Document S1, Tab. S3). To address this for FD, we used species incidences 660 summed per patch as a “detection frequency”, serving as a proxy for abundance frequency. The 661 outputs for FD of bats were generated using abundance-based models, including the estimation 662 of sample coverage . TD quantifies the effective number of species and represents species 663 richness, whereas PD quantifies the effective number of lineages, and FD quantifies the 664 effective number of virtual functional groups, or functional species, both expressing ecosystem 665 functioning and resilience (Cadotte et al. 2012, Chao et al. 2021) . Diversity orders capture the 666 relative frequency of species incidence within the assemblage, emphasizing infrequent (q0), 667 frequent (q1), and highly frequent (q2) species (Kortmann et al. 2025). 668 To calculate the functional diversity, a species-pairwise distance matrix is needed. Thus, 669 we computed Gower’s distance (Gower 1971) between species for bats and birds based on their 670 traits, using the function ‘daisy’ from the cluster package (Maechler et al. 2024). Patches with 671 no observed species were excluded from the analysis, reducing the number of patches in some 672 districts. Due to malfunction of recorders or adverse weather conditions, recording had to be 673 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 31 repeated occasionally. This resulted in varying numbers of sampling units per patch, which are 674 defined as one recording night for bats and one recording day for birds (Document S1, Tab. S3 675 and S4). In some cases, the number of patches differed between treatment and paired control 676 districts (e.g., E = 8 vs. C = 9 patches), which can influence the range of β -diversity values. 677 Thus, a Jaccard-type turnover (1-S) transformation was applied to β-diversity (for details, see 678 Tab. 1 in Chao et al. 2019). This metric quantifies dissimilarity between assemblages relative 679 to γ -diversity, standardized by patch number, w ith a value of zero indicating identical 680 assemblages in the patches and one indicating no shared species . To account for sample 681 incompleteness within the meta-analytic framework, sample coverage (SC) was estimated using 682 the function ‘Coverage’ from the iNEXT.3D package (Chao et al. 2021), setting the minimum 683 quantile across all SC(2n) values to be greater than 0.5. At the fixed sampling coverage level, 684 rarefaction and extrapolation curves were generated for each of the eleven treatment -control 685 district pairs. 686 Using the ‘iNEXTmeta_beta’ function (Chao 2025) , landscape-level γ -diversity was 687 decomposed into its components of within -patch α-diversity and between -patch β -diversity. 688 The three diversity facets (TD, PD, and FD) were calculated for each district pair and diversity 689 level across all three diversity orders , referred to as q0, q1 , and q2 (Chao et al. 2023a) . All 690 metrics are expressed in the same units of species, lineage, or functional group equivalents and 691 can be directly compared within each diversity level (α, β, γ) (Chao et al. 2021). Across levels, 692 comparisons are only meaningful between α and γ, as multiplicative β -diversity is a ratio, 693 representing the effective number of non -overlapping assemblages, calculated based on a 694 multiplicative diversity composition (β = γ/α). Thus, α -diversity captures temporal changes of 695 the within -patch assemblage, β -diversity reflects spatial differentiation among assemblages, 696 and γ-diversity integrates the spatio-temporal variation of communities across all study sites. 697 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 32 For each treatment -control district pair, we calculated diversity differences, with 698 positive values indicating higher diversity in heterogeneous forests compared to homogeneous 699 forests. The overall effect across all 11 experimental landscapes was estimated through a meta-700 analysis (see Data and code availability ). Confidence intervals (CI) and average differences 701 between all forest pairs were calculated using bootstrapping (nboot = 50 ). Significance is 702 inferred when the CI does not include zero. For reproducibility, all models were run with 703 ‘set.seed(1)’ or ‘ clusterSetRNGStream(cl, iseed = 1) ’ for parallel c omputation of PD with 704 ‘parLapply’ from the package parallel (R Core Team 2024) . Forest plots illustrating the 705 standardized diversity differences were generated using ‘ggiNEXTmeta’(Chao 2025) and are 706 presented in the supplements (Document S1., Fig. S3–29 and Fig. S30-56 ) (Chao et al. 2023a, 707 b). 708 709

710 711 Aguirre, L. F., Montaño-Centellas, F. A., Gavilanez, M. M. and Stevens, R. D. 2016. Taxonomic and 712 Phylogenetic Determinants of Functional Composition of Bolivian Bat Assemblages. - PLOS ONE 713 11: e0158170. 714 Allouche, O., Kalyuzhny, M., Moreno-Rueda, G., Pizarro, M. and Kadmon, R. 2012. Area–715 heterogeneity tradeoff and the diversity of ecological communities. - Proceedings of the 716 National Academy of Sciences 109: 17495–17500. 717 Aszalós, R., Thom, D., Aakala, T., Angelstam, P., Brūmelis, G., Gálhidy, L., Gratzer, G., Hlásny, T., 718 Katzensteiner, K., Kovács, B., Knoke, T., Larrieu, L., Motta, R., Müller, J., Ódor, P., Roženbergar, 719 D., Paillet, Y., Pitar, D., Standovár, T., Svoboda, M., Szwagrzyk, J., Toscani, P. and Keeton, W. S. 720 2022. Natural disturbance regimes as a guide for sustainable forest management in Europe. - 721 Ecological Applications 32: e2596. 722 Bae, S., Müller, J., Lee, D., Vierling, K. T., Vogeler, J. C., Vierling, L. A., Hudak, A. T., Latifi, H. and Thorn, 723 S. 2018. Taxonomic, functional, and phylogenetic diversity of bird assemblages are oppositely 724 associated to productivity and heterogeneity in temperate forests. - Remote Sensing of 725 Environment 215: 145–156. 726 Barnes, A. D., Weigelt, P., Jochum, M., Ott, D., Hodapp, D., Haneda, N. F. and Brose, U. 2016. Species 727 richness and biomass explain spatial turnover in ecosystem functioning across tropical and 728 temperate ecosystems. - Philosophical Transactions of the Royal Society B: Biological Sciences 729 371: 20150279. 730 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 33 Bełcik, M., Lenda, M., Amano, T. and Skórka, P. 2020. Different response of the taxonomic, 731 phylogenetic and functional diversity of birds to forest fragmentation. - Scientific Reports 10: 732 20320. 733 Blažek, J., Konečný, A. and Bartonička, T. 2021. Bat aggregational response to pest caterpillar 734 emergence. - Scientific Reports 11: 13634. 735 Böhm, S. M., Wells, K. and Kalko, E. K. V. 2011. Top-Down Control of Herbivory by Birds and Bats in 736 the Canopy of Temperate Broad-Leaved Oaks (Quercus robur). - PLOS ONE 6: e17857. 737 Bouvet, A., Paillet, Y., Archaux, F., Tillon, L., Denis, P., Gilg, O. and Gosselin, F. 2016. Effects of forest 738 structure, management and landscape on bird and bat communities. - Environmental 739 Conservation 43: 148–160. 740 Bradbury, R. B., Hill, R. A., Mason, D. C., Hinsley, S. A., Wilson, J. D., Balzter, H., Anderson, G. Q. A., 741 Whittingham, M. J., Davenport, I. J. and Bellamy, P. E. 2005. Modelling relationships between 742 birds and vegetation structure using airborne LiDAR data: a review with case studies from 743 agricultural and woodland environments. - Ibis 147: 443–452. 744 Braunisch, V., Roder, S., Coppes, J., Froidevaux, J. S. P., Arlettaz, R. and Bollmann, K. 2019. Structural 745 complexity in managed and strictly protected mountain forests: Effects on the habitat suitability 746 for indicator bird species. - Forest Ecology and Management 448: 139–149. 747 Brunet, J., Fritz, Ö. and Richnau, G. 2010. Biodiversity in European beech forests - a review with 748 recommendations for sustainable forest management. - Ecological Bulletins: 77–94. 749 Bundesamt für Naturschutz, (BfN) 2025. Artenportraits: Fledermäuse in Deutschland. 750 https://www.bfn.de/artenportraits. 751 Cadotte, M. W. 2013. Experimental evidence that evolutionarily diverse assemblages result in higher 752 productivity. - Proceedings of the National Academy of Sciences 110: 8996–9000. 753 Cadotte, M. W. 2017. Functional traits explain ecosystem function through opposing mechanisms. - 754 Ecology Letters 20: 989–996. 755 Cadotte, M. W., Cavender-Bares, J., Tilman, D. and Oakley, T. H. 2009. Using Phylogenetic, Functional 756 and Trait Diversity to Understand Patterns of Plant Community Productivity. - PLOS ONE 4: 757 e5695. 758 Cadotte, M. W., Carscadden, K. and Mirotchnick, N. 2011. Beyond species: functional diversity and 759 the maintenance of ecological processes and services. - Journal of Applied Ecology 48: 1079–760 1087. 761 Cadotte, M. W., Dinnage, R. and Tilman, D. 2012. Phylogenetic diversity promotes ecosystem 762 stability. - Ecology 93: S223–S233. 763 Cadotte, M., Albert, C. H. and Walker, S. C. 2013. The ecology of differences: assessing community 764 assembly with trait and evolutionary distances. - Ecology Letters 16: 1234–1244. 765 Cadotte, M. W., Carboni, M., Si, X. and Tatsumi, S. 2019. Do traits and phylogeny support congruent 766 community diversity patterns and assembly inferences? - Journal of Ecology 107: 2065–2077. 767 Cardinale, B. J., Duffy, J. E., Gonzalez, A., Hooper, D. U., Perrings, C., Venail, P., Narwani, A., Mace, G. 768 M., Tilman, D., Wardle, D. A., Kinzig, A. P., Daily, G. C., Loreau, M., Grace, J. B., Larigauderie, A., 769 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 34 Srivastava, D. S. and Naeem, S. 2012. Biodiversity loss and its impact on humanity. - Nature 486: 770 59–67. 771 Chao, A. 2025. AnneChao/iNEXT.meta. GitHub. https://github.com/AnneChao/iNEXT.meta. 772 Chao, A., Chiu, C.-H. and Jost, L. 2014. Unifying Species Diversity, Phylogenetic Diversity, Functional 773 Diversity, and Related Similarity and Differentiation Measures Through Hill Numbers. - Annual 774 Review of Ecology, Evolution, and Systematics 45: 297–324. 775 Chao, A., Chiu, C.-H., Wu, S.-H., Huang, C.-L. and Lin, Y.-C. 2019. Comparing two classes of alpha 776 diversities and their corresponding beta and (dis)similarity measures, with an application to the 777 Formosan sika deer Cervus nippon taiouanus reintroduction programme. - Methods in Ecology 778 and Evolution 10: 1286–1297. 779 Chao, A., Henderson, P. A., Chiu, C.-H., Moyes, F., Hu, K.-H., Dornelas, M. and Magurran, A. E. 2021. 780 Measuring temporal change in alpha diversity: A framework integrating taxonomic, phylogenetic 781 and functional diversity and the iNEXT.3D standardization. - Methods in Ecology and Evolution 782 12: 1926–1940. 783 Chao, A., Thorn, S., Chiu, C.-H., Moyes, F., Hu, K.-H., Chazdon, R. L., Wu, J., Magnago, L. F. S., 784 Dornelas, M., Zelený, D., Colwell, R. K. and Magurran, A. E. 2023a. Rarefaction and extrapolation 785 with beta diversity under a framework of Hill numbers: The iNEXT.beta3D standardization. - 786 Ecological Monographs 93: e1588. 787 Chao, A., Chiu, C.-H., Hu, K.-H. and Zelený, D. 2023b. Revisiting Alwyn H. Gentry’s forest transect 788 data: latitudinal beta diversity patterns are revealed using a statistical sampling-model-based 789 approach. - Japanese Journal of Statistics and Data Science 6: 861–884. 790 Charbonnier, Y., Barbaro, L., Theillout, A. and Jactel, H. 2014. Numerical and Functional Responses of 791 Forest Bats to a Major Insect Pest in Pine Plantations. - PLOS ONE 9: e109488. 792 Colwell, R. K. and Chao, A. 2022. Measuring and comparing class diversity in archaeological 793 assemblages: A brief guide to the history and state-of-the-art in diversity statistics. - In: Defining 794 and Measuring Diversity in Archaeology : Another Step Toward an Evolutionary Synthesis of 795 Culture. (Eren, M. I., and B. Buchanan, editors). Berghahn Books, pp. 263–294. 796 Costa, M. M. S. da and Schmidt, F. A. 2022. Gamma, alpha, and beta diversity of ant assemblages 797 response to a gradient of forest cover in human-modified landscape in Brazilian Amazon. - 798 Biotropica 54: 515–524. 799 Cours, J. and Duflot, R. 2025. Effects of landscape heterogeneity on bird communities in temperate, 800 boreal, and montane forests – a review. - Journal of Avian Biology 2025: e03458. 801 Denzinger, A. and Schnitzler, H.-U. 2013. Bat guilds, a concept to classify the highly diverse foraging 802 and echolocation behaviors of microchiropteran bats. - Frontiers in Physiology 4: 1–15. 803 Eisenhauer, N., Hines, J., Maestre, F. T. and Rillig, M. C. 2023. Reconsidering functional redundancy in 804 biodiversity research. - npj Biodiversity 2: 9. 805 European Union 1992. Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural 806 habitats and of wild fauna and flora. - OJ L 206: 7–50. 807 European Union 2009. Directive 2009/147/EC of the European Parliament and of the Council of 808 30 November 2009 on the conservation of wild birds (Codified version). - OJ L 020: 7–25. 809 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 35 Farnsworth, L. M., Nimmo, D. G., Kelly, L. T., Bennett, A. F. and Clarke, M. F. 2014. Does pyrodiversity 810 beget alpha, beta or gamma diversity? A case study using reptiles from semi-arid Australia. - 811 Diversity and Distributions 20: 663–673. 812 Flynn, D. F. B., Gogol-Prokurat, M., Nogeire, T., Molinari, N., Richers, B. T., Lin, B. B., Simpson, N., 813 Mayfield, M. M. and DeClerck, F. 2009. Loss of functional diversity under land use intensification 814 across multiple taxa. - Ecology Letters 12: 22–33. 815 Freitag Kramer, J. M., Bordin, K. M., Scarton Bergamin, R., Klipel, J., Caroline Picolotto, R., Zanini, K. J., 816 Zwiener, V. P. and Cristina Müller, S. 2025. Temporal changes in taxonomic and functional alpha 817 and beta diversity across tree communities in subtropical Atlantic forests. - Oikos 2025: e10961. 818 Froidevaux, J. S. P., Zellweger, F., Bollmann, K., Jones, G. and Obrist, M. K. 2016. From field surveys to 819 LiDAR: Shining a light on how bats respond to forest structure. - Remote Sensing of Environment 820 175: 242–250. 821 Froidevaux, J. S. P., Toshkova, N., Barbaro, L., Benítez-López, A., Kerbiriou, C., Le Viol, I., Pacifici, M., 822 Santini, L., Stawski, C., Russo, D., Dekker, J., Alberdi, A., Amorim, F., Ancillotto, L., Barré, K., Bas, 823 Y., Cantú-Salazar, L., Dechmann, D. K. N., Devaux, T., Eldegard, K., Fereidouni, S., Furmankiewicz, 824 J., Hamidovic, D., Hill, D. L., Ibáñez, C., Julien, J.-F., Juste, J., Kaňuch, P., Korine, C., Laforge, A., 825 Legras, G., Leroux, C., Lesiński, G., Mariton, L., Marmet, J., Mata, V. A., Mifsud, C. M., Nistreanu, 826 V., Novella-Fernandez, R., Rebelo, H., Roche, N., Roemer, C., Ruczyński, I., Sørås, R., Uhrin, M., 827 Vella, A., Voigt, C. C. and Razgour, O. 2023. A species-level trait dataset of bats in Europe and 828 beyond. - Scientific Data 10: 253. 829 Gonçalves-Souza, T., Chase, J. M., Haddad, N. M., Vancine, M. H., Didham, R. K., Melo, F. L. P., Aizen, 830 M. A., Bernard, E., Chiarello, A. G., Faria, D., Gibb, H., de Lima, M. G., Magnago, L. F. S., Mariano-831 Neto, E., Nogueira, A. A., Nemésio, A., Passamani, M., Pinho, B. X., Rocha-Santos, L., Rodrigues, 832 R. C., Safar, N. V. H., Santos, B. A., Soto-Werschitz, A., Tabarelli, M., Uehara-Prado, M., 833 Vasconcelos, H. L., Vieira, S. and Sanders, N. J. 2025. Species turnover does not rescue 834 biodiversity in fragmented landscapes. - Nature 640: 702–706. 835 Gossner, M. M., Lewinsohn, T. M., Kahl, T., Grassein, F., Boch, S., Prati, D., Birkhofer, K., Renner, S. C., 836 Sikorski, J., Wubet, T., Arndt, H., Baumgartner, V., Blaser, S., Blüthgen, N., Börschig, C., Buscot, 837 F., Diekötter, T., Jorge, L. R., Jung, K., Keyel, A. C., Klein, A.-M., Klemmer, S., Krauss, J., Lange, M., 838 Müller, J., Overmann, J., Pašalić, E., Penone, C., Perović, D. J., Purschke, O., Schall, P., Socher, S. 839 A., Sonnemann, I., Tschapka, M., Tscharntke, T., Türke, M., Venter, P. C., Weiner, C. N., Werner, 840 M., Wolters, V., Wurst, S., Westphal, C., Fischer, M., Weisser, W. W. and Allan, E. 2016. Land-use 841 intensification causes multitrophic homogenization of grassland communities. - Nature 540: 842 266–269. 843 Gower, J. C. 1971. A General Coefficient of Similarity and Some of Its Properties. - Biometrics 27: 844 857–871. 845 Graser, A., Frank, C., Kunz, F., Schuldt, A., Senf, C., Sudfeldt, C., Trautmann, S. and Kamp, J. 2025. 846 Increase in disturbance-induced canopy gaps leads to reorganization of Central European bird 847 communities. - Basic and Applied Ecology 83: 88–97. 848 Hanle, J., Duguid, M. C. and Ashton, M. S. 2020. Legacy forest structure increases bird diversity and 849 abundance in aging young forests. - Ecology and Evolution 10: 1193–1208. 850 Heidrich, L., Bae, S., Levick, S., Seibold, S., Weisser, W., Krzystek, P., Magdon, P., Nauss, T., Schall, P., 851 Serebryanyk, A., Wöllauer, S., Ammer, C., Bässler, C., Doerfler, I., Fischer, M., Gossner, M. M., 852 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 36 Heurich, M., Hothorn, T., Jung, K., Kreft, H., Schulze, E.-D., Simons, N., Thorn, S. and Müller, J. 853 2020. Heterogeneity–diversity relationships differ between and within trophic levels in 854 temperate forests. - Nature Ecology & Evolution 4: 1204–1212. 855 Hendel, A.-L., Winiger, N., Jonker, M., Zielewska-Büttner, K., Ganz, S., Adler, P. and Braunisch, V. 856 2023. Bat habitat selection reveals positive effects of retention forestry. - Forest Ecology and 857 Management 531: 120783. 858 Hendel, A., Douma, J. C., Klingenfuß, S., Pereira, J. M. C., Ruppert, L., Spînu, A. P., Frey, J., Denter, M., 859 Liu, X., Storch, I., Klein, A. M. and Braunisch, V. 2025. Disentangling direct and indirect effects of 860 forest structure on biodiversity: Bottom‐up and top‐down effects between forestry, bats and 861 their insect prey. - Journal of Applied Ecology 62: 93–105. 862 Hilmers, T., Friess, N., Bässler, C., Heurich, M., Brandl, R., Pretzsch, H., Seidl, R. and Müller, J. 2018. 863 Biodiversity along temperate forest succession. - Journal of Applied Ecology 55: 2756–2766. 864 Hochrein, S., Liebhold, A. M., Bae, S., Wimmer, B., Leroy, B. M. L., Seibold, S., Weisser, W. W., Müller, 865 J. and Huszarik, M. 2025. Bat response to experimental insecticide application against a 866 defoliating moth in mixed oak forests. - Journal of Applied Ecology 00: 1–12. 867 Hooper, D. U., Adair, E. C., Cardinale, B. J., Byrnes, J. E. K., Hungate, B. A., Matulich, K. L., Gonzalez, A., 868 Duffy, J. E., Gamfeldt, L. and O’Connor, M. I. 2012. A global synthesis reveals biodiversity loss as 869 a major driver of ecosystem change. - Nature 486: 105–108. 870 Jaureguiberry, P., Titeux, N., Wiemers, M., Bowler, D. E., Coscieme, L., Golden, A. S., Guerra, C. A., 871 Jacob, U., Takahashi, Y., Settele, J., Díaz, S., Molnár, Z. and Purvis, A. 2022. The direct drivers of 872 recent global anthropogenic biodiversity loss. - Science Advances 8: eabm9982. 873 Jetz, W., Thomas, G. H., Joy, J. B., Hartmann, K. and Mooers, A. O. 2012. The global diversity of birds 874 in space and time. - Nature 491: 444–448. 875 Jin, Y. and Qian, H. 2023. U.PhyloMaker: An R package that can generate large phylogenetic trees for 876 plants and animals. - Plant Diversity 45: 347–352. 877 Jung, K., Kaiser, S., Böhm, S., Nieschulze, J. and Kalko, E. K. V. 2012. Moving in three dimensions: 878 effects of structural complexity on occurrence and activity of insectivorous bats in managed 879 forest stands. - Journal of Applied Ecology 49: 523–531. 880 Jung, K., Schall, P., Ammer, C. and Noelke, N. 2025. Green forests for bats: Response to stand-scale 881 management interventions – Results of an experiment. - Forest Ecology and Management 595: 882 122971. 883 Keck, F., Peller, T., Alther, R., Barouillet, C., Blackman, R., Capo, E., Chonova, T., Couton, M., 884 Fehlinger, L., Kirschner, D., Knüsel, M., Muneret, L., Oester, R., Tapolczai, K., Zhang, H. and 885 Altermatt, F. 2025. The global human impact on biodiversity. - Nature 641: 395–400. 886 Kerth, G. and Reckardt, K. 2003. Information transfer about roosts in female Bechstein’s bats: an 887 experimental field study. - Proceedings of the Royal Society of London. Series B: Biological 888 Sciences 270: 511–515. 889 Klingbeil, B. T. and Willig, M. R. 2016. Matrix composition and landscape heterogeneity structure 890 multiple dimensions of biodiversity in temperate forest birds. - Biodivers Conserv 25: 2687–891 2708. 892 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 37 Kortmann, M., Hurst, J., Brinkmann, R., Heurich, M., Silveyra González, R., Müller, J. and Thorn, S. 893 2018. Beauty and the beast: how a bat utilizes forests shaped by outbreaks of an insect pest. - 894 Animal Conservation 21: 21–30. 895 Kortmann, M., Chao, A., Schaefer, H. M., Blüthgen, N., Gelis, R., Tremlett, C. J., Busse, A., Püls, M., 896 Seibold, S., Kriegel, P., Rabl, D., de la Hoz, M., Şekercioğlu, Ç. H., Schleuning, M., Feldhaar, H., 897 Newell, F. L., Kümmet, S., Mitesser, O., Peters, M. K. and Müller, J. 2025. Sample coverage 898 affects diversity measures of bird communities along a natural recovery gradient of abandoned 899 agriculture in tropical lowland forests. - Journal of Applied Ecology 62: 480–491. 900 Krebs, J. R. 1982. Territorial defence in the great tit (Parus major): Do residents always win? - 901 Behavioral Ecology and Sociobiology 11: 185–194. 902 Legendre, P. 2014. Interpreting the replacement and richness difference components of beta 903 diversity. - Global Ecology and Biogeography 23: 1324–1334. 904 Lesak, A. A., Radeloff, V. C., Hawbaker, T. J., Pidgeon, A. M., Gobakken, T. and Contrucci, K. 2011. 905 Modeling forest songbird species richness using LiDAR-derived measures of forest structure. - 906 Remote Sensing of Environment 115: 2823–2835. 907 Lettenmaier, L., Seibold, S., Bässler, C., Brandl, R., Gruppe, A., Müller, J. and Hagge, J. 2022. Beetle 908 diversity is higher in sunny forests due to higher microclimatic heterogeneity in deadwood. - 909 Oecologia 198: 825–834. 910 Lewandowski, P., Przepióra, F. and Ciach, M. 2021. Single dead trees matter: Small-scale canopy gaps 911 increase the species richness, diversity and abundance of birds breeding in a temperate 912 deciduous forest. - Forest Ecology and Management 481: 118693. 913 López-Baucells, A., Rowley, S., Rocha, R., Bobrowiec, P. E. D., Palmeirim, J. M., Farneda, F. Z. and 914 Meyer, C. F. J. 2022. Interplay between local and landscape-scale effects on the taxonomic, 915 functional and phylogenetic diversity of aerial insectivorous neotropical bats. - Landsc Ecol 37: 916 2861–2875. 917 Lundholm, J. T. 2009. Plant species diversity and environmental heterogeneity: spatial scale and 918 competing hypotheses. - Journal of Vegetation Science 20: 377–391. 919 MacArthur, R. H. and MacArthur, J. W. 1961. On Bird Species Diversity. - Ecology 42: 594–598. 920 Maechler, M., Rousseeuw, P., Struyf, A., Hubert, M. and Hornik, K. 2024. cluster: Cluster Analysis 921 Basics and Extensions. https://CRAN.R-project.org/package=cluster4. 922 Mäntylä, E., Klemola, T. and Laaksonen, T. 2011. Birds help plants: a meta-analysis of top-down 923 trophic cascades caused by avian predators. - Oecologia 165: 143–151. 924 McGill, B. J., Dornelas, M., Gotelli, N. J. and Magurran, A. E. 2015. Fifteen forms of biodiversity trend 925 in the Anthropocene. - Trends in Ecology & Evolution 30: 104–113. 926 McKinney, M. L. and Lockwood, J. L. 1999. Biotic homogenization: a few winners replacing many 927 losers in the next mass extinction. - Trends in Ecology & Evolution 14: 450–453. 928 Mooney, K. A., Gruner, D. S., Barber, N. A., Van Bael, S. A., Philpott, S. M. and Greenberg, R. 2010. 929 Interactions among predators and the cascading effects of vertebrate insectivores on arthropod 930 communities and plants. - Proceedings of the National Academy of Sciences 107: 7335–7340. 931 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 38 Morante-Filho, J. C., Arroyo-Rodríguez, V. and Faria, D. 2016. Patterns and predictors of β-diversity in 932 the fragmented Brazilian Atlantic forest: a multiscale analysis of forest specialist and generalist 933 birds. - Journal of Animal Ecology 85: 240–250. 934 Moreno, D. J., Ribeiro, M. C. and Piratelli, A. J. 2024. Landscape heterogeneity increases bird 935 functional diversity within Neotropical vineyards. - Biotropica 56: e13328. 936 Mori, A. S., Furukawa, T. and Sasaki, T. 2013. Response diversity determines the resilience of 937 ecosystems to environmental change. - Biological Reviews 88: 349–364. 938 Müller, J. 2005. Waldstrukturen als Steuergröße für Artengemeinschaften in kollinen bis 939 submontanen Buchenwäldern. Dissertation. 940 Müller, J., Noss, R. F., Bussler, H. and Brandl, R. 2010. Learning from a “benign neglect strategy” in a 941 national park: Response of saproxylic beetles to dead wood accumulation. - Biological 942 Conservation 143: 2559–2569. 943 Müller, J., Mehr, M., Bässler, C., Fenton, M. B., Hothorn, T., Pretzsch, H., Klemmt, H.-J. and Brandl, R. 944 2012. Aggregative response in bats: prey abundance versus habitat. - Oecologia 169: 673–684. 945 Müller, J., Brandl, R., Buchner, J., Pretzsch, H., Seifert, S., Strätz, C., Veith, M. and Fenton, B. 2013. 946 From ground to above canopy—Bat activity in mature forests is driven by vegetation density and 947 height. - Forest Ecology and Management 306: 179–184. 948 Müller, J., Mitesser, O., Cadotte, M. W., van der Plas, F., Mori, A. S., Ammer, C., Chao, A., Scherer-949 Lorenzen, M., Baldrian, P., Bässler, C., Biedermann, P., Cesarz, S., Claßen, A., Delory, B. M., 950 Feldhaar, H., Fichtner, A., Hothorn, T., Kuenzer, C., Peters, M. K., Pierick, K., Schmitt, T., Schuldt, 951 B., Seidel, D., Six, D., Steffan-Dewenter, I., Thorn, S., von Oheimb, G., Wegmann, M., Weisser, W. 952 W. and Eisenhauer, N. 2023. Enhancing the structural diversity between forest patches—A 953 concept and real-world experiment to study biodiversity, multifunctionality and forest resilience 954 across spatial scales. - Global Change Biology 29: 1437–1450. 955 Oksanen, J., Simpson, G. L., Blanchet, F. G., Kindt, R., Legendre, P., Minchin, P. R., O’Hara, R. B., 956 Solymos, P., Stevens, M. H. H., Szoecs, E., Wagner, H., Barbour, M., Bedward, M., Bolker, B., 957 Borcard, D., Carvalho, G., Chirico, M., De Caceres, M., Durand, S., Evangelista, H. B. A., FitzJohn, 958 R., Friendly, M., Furneaux, B., Hannigan, G., Hill, M. O., Lahti, L., McGlinn, D., Ouellette, M.-H., 959 Ribeiro Cunha, E., Smith, T., Stier, A., Ter Braak, C. J. F. and Weedon, J. 2022. vegan: Community 960 Ecology Package.: 2.6-6.1. 961 Olden, J. D. and Rooney, T. P. 2006. On defining and quantifying biotic homogenization. - Global 962 Ecology and Biogeography 15: 113–120. 963 Pastro, L. A., Dickman, C. R. and Letnic, M. 2011. Burning for biodiversity or burning biodiversity? 964 Prescribed burn vs. wildfire impacts on plants, lizards, and mammals. - Ecological Applications 965 21: 3238–3253. 966 Pierick, K., Link, R. M., Inderwies, H. L., Puggaard Koolen, S., Wu, B., Schuldt, B., Müller, J., Ammer, C. 967 and Seidel, D. 2025. Artificial Gap Creation But Not Thinning Interventions Persistently Enhance 968 Forest Structural Heterogeneity. SSRN. preprint. 969 Poesel, A. and Dabelsteen, T. 2005. Territorial responses of male blue tits to simulated dynamic 970 intrusions: effects of song overlap and intruder location. - Animal Behaviour 70: 1419–1427. 971 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 39 R Core Team 2024. R: A Language and Environment for Statistical Computing. R Foundation for 972 Statistical Computing, Vienna, Austria. 973 Renner, S. C., Suarez-Rubio, M., Kaiser, S., Nieschulze, J., Kalko, E. K. V., Tschapka, M. and Jung, K. 974 2018. Divergent response to forest structure of two mobile vertebrate groups. - Forest Ecology 975 and Management 415–416: 129–138. 976 Rigo, F., Paniccia, C., Anderle, M., Chianucci, F., Obojes, N., Tappeiner, U., Hilpold, A. and Mina, M. 977 2024. Relating forest structural characteristics to bat and bird diversity in the Italian Alps. - 978 Forest Ecology and Management 554: 121673. 979 Rolls, R. J., Deane, D. C., Johnson, S. E., Heino, J., Anderson, M. J. and Ellingsen, K. E. 2023. Biotic 980 homogenisation and differentiation as directional change in beta diversity: synthesising driver–981 response relationships to develop conceptual models across ecosystems. - Biological Reviews 982 98: 1388–1423. 983 Roswell, M., Dushoff, J. and Winfree, R. 2021. A conceptual guide to measuring species diversity. - 984 Oikos 130: 321–338. 985 Rothacher, J., Hagge, J., Bässler, C., Brandl, R., Gruppe, A. and Müller, J. 2023. Logging operations 986 creating snags, logs, and stumps under open and closed canopies promote stand-scale beetle 987 diversity. - Forest Ecology and Management 540: 121022. 988 Rothacher, J., Seidl, R., Thom, D., Kortmann, M., Chao, A., Chiu, C.-H., Heibl, C., Hothorn, T., Mitesser, 989 O., Mori, A. S., Morinière, J., Pierick, K., Wild, C., Wild, N. and Müller, J. 2025. The impact of tree 990 mortality and post-disturbance management on insect diversity in temperate forests: Insights 991 from a replicated experiment. - Journal of Applied Ecology 62: 1878–1888. 992 Russo, D., Salinas-Ramos, V. B., Cistrone, L., Smeraldo, S., Bosso, L. and Ancillotto, L. 2021. Do We 993 Need to Use Bats as Bioindicators? - Biology 10: 693. 994 Safi, K. and Kerth, G. 2007. Comparative Analyses Suggest That Information Transfer Promoted 995 Sociality in Male Bats in the Temperate Zone. - The American Naturalist 170: 465–472. 996 Sala, O. E., Stuart Chapin, F., III, Armesto, J. J., Berlow, E., Bloomfield, J., Dirzo, R., Huber-Sanwald, E., 997 Huenneke, L. F., Jackson, R. B., Kinzig, A., Leemans, R., Lodge, D. M., Mooney, H. A., Oesterheld, 998 M., Poff, N. L., Sykes, M. T., Walker, B. H., Walker, M. and Wall, D. H. 2000. Global Biodiversity 999 Scenarios for the Year 2100. - Science 287: 1770–1774. 1000 Schall, P., Gossner, M. M., Heinrichs, S., Fischer, M., Boch, S., Prati, D., Jung, K., Baumgartner, V., 1001 Blaser, S., Böhm, S., Buscot, F., Daniel, R., Goldmann, K., Kaiser, K., Kahl, T., Lange, M., Müller, J., 1002 Overmann, J., Renner, S. C., Schulze, E.-D., Sikorski, J., Tschapka, M., Türke, M., Weisser, W. W., 1003 Wemheuer, B., Wubet, T. and Ammer, C. 2018. The impact of even-aged and uneven-aged forest 1004 management on regional biodiversity of multiple taxa in European beech forests. - Journal of 1005 Applied Ecology 55: 267–278. 1006 Schauer, B., Thorn, S., Blaschke, M. and Kudernatsch, T. 2023. Conversion of pure spruce to mixed 1007 spruce beech stands: Effects on alpha and beta diversity of multiple taxonomic groups. - Forest 1008 Ecology and Management 545: 121297. 1009 Schnitzler, H.-U. and Kalko, E. K. V. 2001. Echolocation by Insect-Eating Bats. - BioScience 51: 557. 1010 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 40 Seibold, S., Bässler, C., Baldrian, P., Reinhard, L., Thorn, S., Ulyshen, M. D., Weiß, I. and Müller, J. 1011 2016. Dead-wood addition promotes non-saproxylic epigeal arthropods but effects are 1012 mediated by canopy openness. - Biological Conservation 204: 181–188. 1013 Seibold, S., Weisser, W. W., Ambarlı, D., Gossner, M. M., Mori, A. S., Cadotte, M. W., Hagge, J., 1014 Bässler, C. and Thorn, S. 2023. Drivers of community assembly change during succession in 1015 wood-decomposing beetle communities. - Journal of Animal Ecology 92: 965–978. 1016 Seibold, S., Richter, T., Geres, L., Seidl, R., Martin, R., Mitesser, O., Senf, C., Griem, L. and Müller, J. 1017 2024. Soundscapes and airborne laser scanning identify vegetation density and its interaction 1018 with elevation as main driver of bird diversity and community composition. - Diversity and 1019 Distributions 30: e13905. 1020 Soininen, J., Heino, J. and Wang, J. 2018. A meta-analysis of nestedness and turnover components of 1021 beta diversity across organisms and ecosystems. - Global Ecology and Biogeography 27: 96–109. 1022 Stein, A., Gerstner, K. and Kreft, H. 2014. Environmental heterogeneity as a universal driver of species 1023 richness across taxa, biomes and spatial scales. - Ecology Letters 17: 866–880. 1024 Terborgh, J., Feeley, K., Silman, M., Nuñez, P. and Balukjian, B. 2006. Vegetation dynamics of 1025 predator-free land-bridge islands. - Journal of Ecology 94: 253–263. 1026 Tews, J., Brose, U., Grimm, V., Tielbörger, K., Wichmann, M. C., Schwager, M. and Jeltsch, F. 2004. 1027 Animal species diversity driven by habitat heterogeneity/diversity: the importance of keystone 1028 structures. - Journal of Biogeography 31: 79–92. 1029 Thom, D., Sommerfeld, A., Sebald, J., Hagge, J., Müller, J. and Seidl, R. 2020. Effects of disturbance 1030 patterns and deadwood on the microclimate in European beech forests. - Agricultural and Forest 1031 Meteorology 291: 108066. 1032 Tillon, L., Bouget, C., Paillet, Y. and Aulagnier, S. 2016. How does deadwood structure temperate 1033 forest bat assemblages? - European Journal of Forest Research 135: 433–449. 1034 Tobias, J. A., Sheard, C., Pigot, A. L., Devenish, A. J. M., Schleuning, M., et al. 2022. AVONET: 1035 morphological, ecological and geographical data for all birds. - Ecology Letters 25: 581–597. 1036 Tscharntke, T., Tylianakis, J. M., Rand, T. A., Didham, R. K., Fahrig, L., Batáry, P., Bengtsson, J., Clough, 1037 Y., Crist, T. O., Dormann, C. F., Ewers, R. M., Fründ, J., Holt, R. D., Holzschuh, A., Klein, A. M., 1038 Kleijn, D., Kremen, C., Landis, D. A., Laurance, W., Lindenmayer, D., Scherber, C., Sodhi, N., 1039 Steffan-Dewenter, I., Thies, C., van der Putten, W. H. and Westphal, C. 2012. Landscape 1040 moderation of biodiversity patterns and processes - eight hypotheses. - Biological Reviews 87: 1041 661–685. 1042 Tucker, C. M., Shoemaker, L. G., Davies, K. F., Nemergut, D. R. and Melbourne, B. A. 2016. 1043 Differentiating between niche and neutral assembly in metacommunities using null models of β-1044 diversity. - Oikos 125: 778–789. 1045 Tuomisto, H., Ruokolainen, L. and Ruokolainen, K. 2012. Modelling niche and neutral dynamics: on 1046 the ecological interpretation of variation partitioning results. - Ecography 35: 961–971. 1047 Upham, N. S., Esselstyn, J. A. and Jetz, W. 2019. Inferring the mammal tree: Species-level sets of 1048 phylogenies for questions in ecology, evolution, and conservation. - PLOS Biology 17: e3000494. 1049 Vanderwel, M. C., Malcolm, J. R. and Mills, S. C. 2007. A Meta-Analysis of Bird Responses to Uniform 1050 Partial Harvesting across North America. - Conservation Biology 21: 1230–1240. 1051 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint 41 Violle, C., Thuiller, W., Mouquet, N., Munoz, F., Kraft, N. J. B., Cadotte, M. W., Livingstone, S. W. and 1052 Mouillot, D. 2017. Functional Rarity: The Ecology of Outliers. - Trends in Ecology & Evolution 32: 1053 356–367. 1054 Vogeler, J. C., Hudak, A. T., Vierling, L. A., Evans, J., Green, P. and Vierling, K. T. 2014. Terrain and 1055 vegetation structural influences on local avian species richness in two mixed-conifer forests. - 1056 Remote Sensing of Environment 147: 13–22. 1057 Whittaker, R. H. 1972. Evolution and Measurement of Species Diversity. - TAXON 21: 213–251. 1058 Wild, C., Brandl, R., Decker, O., Hochrein, S., Ingrosso, A., Mallick, S., Mitesser, O., Rothacher, J., 1059 Thorn, S. and Müller, J. 2025. Nest mass in forest tits increases with elevation and decreasing 1060 body mass, promoting reproductive success. - Journal of Avian Biology 2025: e03407. 1061 Wood, C. M. and Kahl, S. 2024. Guidelines for appropriate use of BirdNET scores and other detector 1062 outputs. - J Ornithol 165: 777–782. 1063 Yachi, S. and Loreau, M. 1999. Biodiversity and ecosystem productivity in a fluctuating environment: 1064 The insurance hypothesis. - Proceedings of the National Academy of Sciences 96: 1463–1468. 1065 Yu, G. 2022. Data Integration, Manipulation and Visualization of Phylogenetic Trees. - Chapman and 1066 Hall/CRC. 1067 Żmihorski, M., Berg, Å. and Pärt, T. 2016. Forest clear-cuts as additional habitat for breeding 1068 farmland birds in crisis. - Agriculture, Ecosystems & Environment 233: 291–297. 1069 1070 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 28, 2025. ; https://doi.org/10.1101/2025.08.25.671712doi: bioRxiv preprint