Re-Inventories at Two Study Sites in the Oberpfalz (Bavaria, Germany): Increase in RWA Nests, Herb Biodiversity and Dead Wood in Sustainably Managed Forests

: A large-scale re-inventory of red wood ant (RWA; Formica rufa -group) nests and a comparative analysis of their presence/absence data was carried out for the ﬁ rst time in 2023 at two study sites in the Oberpfalz, NE Bavaria, Germany, to investigate the suspected decline of nests, their relationship with forestry aspects, and possible changes in spatial distribution due to tectonic activity (GeoBio-Interactions). We inventoried, in two sustainably managed forests, outstanding nest occurrences (5393 nests including 5276 active nests) and an increase of ≈ 8% over 4–7 years. Our results do not support claims that the number of nests in the Oberpfalz is in sharp decline. RWA preferred mature and medium–mature pine-dominated forests. Standing and downed dead wood did not negatively a ﬀ ect nests. The number of nests with woodpecker cavities and the diversity of the herb layer increased. Nests clusters in NW–SE direction suggest a strong interaction between nest distribution and active tectonics. Our (re-)inventory approach combined with the extensive photo database is a valuable tool for (a) monitoring the entire forest habitat including natural re-stocking, herb biodiversity, woodpecker cavities, and dead wood in, at, and around nests, and (b) identifying GeoBio-Interactions. Such an approach will lead to more realistic counts of RWA nests and provide scienti ﬁ c evidence of the current situation of nests occurrence. We suggest seven to eight years as an appropriate time interval for re-inventories.


Introduction
Global tectonic, magmatic, and geologic processes are essential drivers of biodiversity patterns in all ecosystems (e.g., [1,2]).The resulting type, structure, composition, thickness, and physicochemical properties of the bedrock influence the nutrient supply, water retention, plant growth, composition, and productivity of forests, one of the prevalent ecosystems on land.Understanding the tectonic and lithologic controls on ecology and biology is therefore of fundamental importance [3,4].
Forests are long-lived ecosystems with multiple functions: (a) highly important economic factor and raw material supplier for forest-based industries, (b) climate regulator, (c) protector against and moderator of natural hazards, (d) conservator of biodiversity, natural cycles, and provider of habitats for flora and fauna, and (e) provider of natural experiences and recreational areas for humans, e.g., [5,6].However, forest ecosystems are threatened.Management (e.g., intensive use), as well as natural influences (e.g., climate change, loss of biodiversity and habitats), affect the ability of forests to recover after disturbances [7].Habitat loss, invasive species or climate change have also been suggested to negatively affect insect diversity and promote species decline in forests [8,9]).
Sustainable forest management (SFM) aims to support the forest and the multifunctional forest sector to (a) secure and improve the growth and stability of forest stands, (b) develop sustainable, economically, and ecologically balanced management practices, and (c) ensure the long-term integrity and biodiversity of forest ecosystems [5].Two forestry aspects, dead wood (DW) and red wood ants (Formica rufa-group; hereafter RWA), are considered as indicators of sustainable forest management.Dead wood is an important habitat for a wide range of organisms and plays a key role in carbon, nutrient, and hydrological cycles, and influences ecosystem processes.In many managed forests, DW such as decaying standing or downed trees, roots, and branches are scarce due to their removal [10].RWA are important forest species (e.g., [11]), related to, but not limited to, habitat biodiversity (e.g., [12]) and pest regulation (e.g., [13]).Furthermore, RWA are bioindicators of hidden active tectonic systems [14][15][16][17][18][19] and geogenic gases ("GeoBio-Interactions"), a prerequisite for their settlement [20][21][22][23].Declines (e.g., [24,25]), but also population increases (e.g., [26,27]), have been reported for RWA.RWA are also considered endangered in Germany, although statistically sound long-term monitoring was discontinued in the 1980s [28,29].Consequently, it is currently not possible to make any statements on the urgently needed conservation status of this species.To fill this knowledge gap, the two densely forested study sites (a) Münchsgrün (MG) and (b) Falkenberg (FB) in the tectonically active Oberpfalz (NE Bavaria, Germany) were re-investigated for the first time with respect to different forestry aspects and possible GeoBio-Interactions [14,18,30,31].
In this comparative analysis, not only was the number of RWA nests counted but the ecosystem around a RWA nest was also monitored and documented in a photo database [14,18,30].We asked six interrelated questions: (1) Are presence/absence data of RWA nests comparable to the two previous inventories?(2) What is the influence of changing forest composition (related to, e.g., tree species, tree age, natural restocking, herb layer) on RWA nests?(3) Does the enrichment of dead wood (standing and/or downed trees, lying branches) in, on, and around a nest during SFM influence RWA nest settlements and nest growth?(4) Was the number of woodpecker cavities (WpC) created by predation on the entire nest surface different from previous inventories?(5) What influence do tectonic structures have on RWA nest distribution (GeoBio-Interactions)?(6) What are the best time intervals for re-inventories?A novel and key aspect of our approach is to evaluate the presence/absence data and the spatial distribution of RWA nests with previous inventories in order to test the postulated decline of RWA in the Oberpfalz region, taking into account the impact of sustainable forest management and increased seismic activity in this region.

Location and Geologic Setting
Bavaria is one of the most densely forested states in Germany, with 2.56 million hectares accounting for one-third of its total area.The Bavarian State Forest (BSF) is the main forest owner in the two re-inventoried study sites MG and FB, both under the responsibility of the Waldsassen forest division; smaller shares belong to private owners (private forests; PF) or municipalities (municipal forests; MF; [14]).In addition, more than one third of the state forest area in Bavaria is currently classified as at least of the following protection categories: Natura 2000, nature reserves, or natural forests [32][33][34].
MG and FB are located in the Oberpfälzer Lake district between the towns of Mitterteich, Tirschenreuth, and Falkenberg (NE Bavaria, Germany; Figure 1a).The gently NE-SW sloping (MG) and partially rugged (FB) terrains are located between 470-550 m a.s.l. and are characterized by low mean annual temperature, short vegetation period, and high precipitation rate [14].
Since the Paleozoic, intricate tectonic, magmatic, and geologic processes have shaped both sites.Two different basement units, the Saxo-Thuringian (in the north) and the Moldanubian (in the south), are separated by a large, tens-of-kilometers-long fault system, the "Erbendorfer Line", that mainly strikes NW-SE and cuts the MG study area in the SW into two sections (Figure 1a).Permo-Carboniferous granite intrusions, the (a) Mitterteicher/Steinwald Granite Complex and (b) Falkenberger Granite Complex, were formed after the collapse of the Variscan orogen.During the Alpine orogeny, rifting (e.g., Eger rift system), volcanism, subsidence, and uplift were dominant.Today, a NW-SE to NNW-SSE-oriented compressional stress field and weak to moderate shallow crustal earthquakes (<20 km; ML ≤ 3.5 Richter scale) characterize this region [14].

Mapping Approach and Data Collection
The re-inventory followed the approach developed by Berberich et al. [14,18,30,31] and already applied in the 2016 (MG) and 2019 (FB) inventories [14].To provide a scientific comparison between the 2016/2019 and 2023 inventories and data collection, the same field methodology was applied: all 352 ha of the mapped area (Figure 1b,c) were walked and mapped by the same observers.To ensure an area-wide inventory, tracks were recorded during mapping with transect spacing <10 m.Mapping was conducted in April, May, and October at both sites and across all inventories to compare forest composition and herbaceous layer.
The fieldwork was divided into two main steps to investigate the presence/absence data of RWA nests: (a) Re-inventory of all RWA nests previously mapped in 2016 and 2019.The GPS coordinates of all these RWA nests were transferred to a GPS receiver (Garmin 60CSx).Previously mapped nests were revisited and mapped with a second GPS receiver (Garmin 62S) together with surrounding forest parameters, and (b) Inventory of new nests including their surrounding forest parameters with the second GPS receiver.
Finally, the total number of nests (ntot) in all inventories and the spatial distribution of RWA nests were examined.
During this 2023 re-inventory, ≈131 ha (MGBSF) and ≈170 ha (FBBSF) were mapped in the BSF.The same adjacent areas of the PF and MF, previously inventoried in 2016/2019, were also included in this re-inventory due to the high abundance of RWA nests.The total mapped areas are 149 ha (MGtot) and 203 ha (FBtot; Figure 1b,c) in 2023.

Data Analysis
Octave 8.4.0,MATLAB 2023b (online), and QGis 3.34 were used for the analyses.Uniform or random distributions or clusters were analyzed using nearest neighbor distribution statistics in MATLAB 2023b (online).For MGBSF and FBBSF, a Multivariate ANalysis Of Variance (MANOVA), one-way ANOVA, Kruskal-Wallis-Test, and multiple comparison with a Bonferroni-adjusted post hoc analysis were applied to physical nest parameters (NH, ND) of active nests, medium tree age (mTA), and woodpecker cavities (WpC) to examine patterns of multiple dependent variables.Information on tree species and medium tree age (mTA) was only available for BSF [44].Density plots of RWA nests were generated to provide insight into their underlying spatial distribution and possible changes over time using the code developed by He [45].This code generated a bivariate plot of scattered point data (RWA nests) by setting the number of equally spaced bins to 20.Color-coded graphs show that multiple RWA nests in the same location result in the highest concentrations of the numerical variable (hotspots), colored in red.Flora Incognita, a freely available mobile image-based application, was used to automatically identify wild plants in the field [46].

Presence/Absence Data of RWA Nests
A total of 5393 RWA nests (ntot) were mapped across the sites.This re-inventory clearly shows an overall increase in total nest numbers (ntot) of ≈10% in MG and stable situation in FB (≈+0.3%).In the 2023 inventory, ≈58% of RWA nests (ntot) mapped in 2016 and 2019 were re-identified; ≈42% were newly mapped nests.About one third of the nests (ntot) mapped in 2016/2019 could not be re-mapped for various reasons (Figure 3; Table 1; Table 2): (a) natural processes: some RWA nests disappeared because the colony died due to natural causes, including nest destruction by, e.g., wild boars (≈2.5%) and other unidentified processes after the 2016/2019 inventories; (b) thickets: In 2023, ≈4% of the nests at both sites could not be re-mapped due to access constraints.These nests were now located in very dense thickets, so their status-quo is uncertain; (c) Pond management (FB): In 2019, 22 nests were mapped on the banks of a fish pond.Pond management caused the water level to rise, submerging these nests.They could not be re-mapped in 2023.Only nests that were active in 2023 will be discussed below.Active RWA nests (nact) were 2513 (MG; ≈10% increase) and 2763 (FB; ≈6% increase).An increase in the numbers of nests (nact) was also observed in the differently owned forests: BSF (≈5%), MF (≈36%), and PF (≈93%; Table 1; Table 2).In particular, small active nests (ND <0.5 m) showed a high percentage in all inventories (MG ≈ 72%; FB ≈ 64%), and their sum increased by ≈9% in both study areas (Table 1).Most active nests were located in forest stands and natural regeneration areas (MG: ≈67%; FB: ≈54%), and up to a quarter (MG: ≈16%; FB: ≈26%) along forest roads and skid trails.Open areas such as meadows, fields, or pond banks were not preferred by RWA nests (MG: ≈3%; FB: ≈14%).
Nearly two thirds of active nests mapped in 2016/2019, and their forest habitat, were re-identified (nactR) at both study sites using the GPS records and the 2016/2019 photo database.More than ≈50% (MG: ≈57%; FB: ≈53%) of start-ups and short nests survived since the last inventories and were re-identified in 2023.Overall, nest heights developed positively: In 2023, half of the re-identified nests showed the same NH class (MG: ≈46%; FB: ≈54%) or even developed into larger nests (MG: ≈28%; FB: ≈23%).Approximately a quarter of the nactRs were smaller than those mapped in 2016/2019 (Figure 4).We observed an ≈5% increase in the number of active nests between the two inventories in MGBSF and FBBSF (Table 2).The statistically significant results of the MANOVA for MGBSF and FBBSF for the 2016/2019 to 2023 inventories confirm that the data are dependent and non-random (d = 1; p <0.05; Wilk's λ: 0.976 (MGBSF) and 0.983 (FBBSF); Table 3).The Wilk's lambda results indicate a small group separation in 2023.This result is supported by the grouped plot matrices and grouped scatterplots of the first two canonical variables of the MANOVA (Figure 5), which show that changes over time of the inventories were associated with (a) small separations between groups, (b) a shift to larger NH and ND, (c) an increase in WpC, and (d) a shift of mTA to more early mature forests for MGBSF due to timber harvesting and a recent assessment of tree age classes and to more mature forests in FBBSF for 2023.The Kruskal-Wallis zero p-value confirms that all data come from the same distribution at the 1% level of significance (Table 4).Bonferroni-adjusted posthoc test showed significant differences in all tested parameters (Table 5).
Forest composition observed in the field differed from records, with a slight shift from pure spruce stands (2016: ≈62%; 2023: ≈52%) to more mixed spruce-pine stands (2016: ≈12%; 2023: ≈18%) in MG due to naturally grown trees.In FB, the proportion of pure spruce (≈53%) and mixed spruce-pine stands (≈37%) remained the same in both inventories.In both study areas, the proportion of naturally grown deciduous trees such as birch (Betula pendula), beech (Fagus silvatica), oak (Quercus robur), red elder (Sambucus racemosa), alder (Alnus glutinosa), or a combination of two deciduous species increased in 2023 (Figure 6), as well as a combination of coniferous and deciduous trees in FB (Figure 6b).Nest numbers of all NH classes multiplied in areas of natural restocking in MG, and short nests especially increased sevenfold; in FB, especially short nests and medium-sized nests predominated.Spruce-and pine-dominated restocking or combinations with, e.g., larch (Larix) characterized one third of the flora around nests in MG and FB.Pure deciduous species or combinations of these, e.g., beech (Fagus silvatica), oak (Quercus robur), red elder (Sambucus racemosa), and rowan (Sorbus aucuparia), remained robust with low proportions of ≈4% in both inventories.
Overall, there was an increase in partial logging activity at each site between the two inventories, e.g., due to bark beetle infestation, windthrow, and snow break.The number of active nests on clearcuts and partially cleared plots tripled by 2023 (Table 2).In MG, the numbers of short nests almost quadrupled (from 73 to 260) and the number of mediumsized nests almost tripled (from 21 to 57).In FB, start-ups (from 10 to 39) and short nests (from 32 to 122) quadrupled.
The nactDW/ha ratio for both inventories was the same for active and re-identified nests in MGBSF and FBBSF.This ratio doubled in MGMF and MGPF and remained robust in FBPF.Re-identified nests in MGBSF and FBBSF showed twice the rate of MGMF and MGPF (Table 5).

Woodpecker Cavities in RWA Nests
In all inventories in MG and FB, woodpecker cavities (WpC) due to predation were observed throughout the nest surface.In 2023, the number of nests with WpC and the number of WpC in the nest increased in BSF, MF, and PF compared to the 2016/2019 inventories.In MGBSF, nest numbers with WpC increased by ≈22% (FBBSF: ≈16%); in FBPF, nest numbers with WpC and WpC counts doubled.In MG, nest numbers doubled in 2023 for nests with 6-10 woodpecker cavities; in FB, nest numbers also increased 1.5-fold (Table 6, Figure 8).Short, medium, and tall nests had the highest proportions of 1-5 cavities at both sites.In MGBSF, one third of the nests were observed in early mature and mature spruce forests and two-thirds in medium mature and mature pine forests in both inventories.In 2019, newly planted and mature spruce forests (≈40%) and mature pine forests (two-thirds) were the preferred locations in FBBSF.In 2023, mTA changed to newly planted, young, and mature spruce forest (≈40%) and medium mature and mature pine forests (≈60%; Figure 8).Deciduous trees played a minor role.
Medium-sized nests with large diameters had an average of 2.7 WpC (MGBSF) in mature and 2.8 WpC in medium mature forests (FBBSF).In both inventories, larger nests (tallextra tall) had fewer cavities than smaller nests.Tall nests with large diameters had 1.9 WpC in MGMF.Medium-sized nests with large diameters had an average of 1.4 WpC (MGPF) and 3.5 WpC (FBPF; Table 6).

GeoBio-Interactions
A predominantly NW-SE spatial distribution pattern of RWA nests (Figure 9), parallel to the current main stress direction in the area, is evident from the density plots.In addition, there appears to be a shift of nest hotspots from the NE section towards the "Erbendorfer Line" (fault line) in the SW section of MG (Figure 9a).

Time Intervals for Re-Inventories
The time intervals for re-inventories were different at the two study sites: MG at seven years; FB at four years.

Discussion
Information on abundance and distribution of species is essential for ecology and conservation [48].The protection of RWA as keystone species that shapes ecosystems and controls pests in an integrally and sustainably managed forest (SFM) is only possible if sufficient scientifically based information on their occurrence and spatial distribution is available.This requires not only censuses, e.g., of individual RWA nests, but also multidisciplinary integration of knowledge across forest ecosystem sciences, biological, spatial, temporal, behavioral, and geo-tectonic scales [14][15][16][17][18][19][20][21][22][23]49].In most European countries, including Germany, the data situation for the different sessile RWA and their occurrence in forests is poor [50].The re-inventories in this study are based on our area-wide and integrated approach [14,18,30,31], as suggested by [49].This approach provides scientifically reliable and very extensive records and documentation of not only the presence/absence data of RWA nests but also of the forest ecosystem and the underlying geo-tectonic processes at short intervals at and around the nests.In addition, each RWA nest and its habitat is documented in a photo database with at least two photos to facilitate comparisons of nests and forest ecosystems of different inventories.

Increase in Number of Active Nests
The outstanding occurrence of RWA nest numbers (F.polyctena) in the MG and FB study areas was confirmed by the 2023 re-inventory after seven (MG) and four years (FB).This study clearly showed that the total number of active RWA nests increased (≈8%) at both sites, but also for BSF (≈5%), MF (≈36%), and PF (≈93%; Table 1; Table 2).This strong increase in RWA nests in PF was observed on large clearing plots, which are thought to promote nest settlements.These results are confirmed by a comparable 12-year re-inventory in the West Eifel Volcanic Field (WEVF), which also showed an increase in RWA nests (ntot ≈ 10%; nact ≈ 6%) and a quadrupling of nest numbers on clearing plots [30].
Our results do not support the claim that the number of RWA nests in the Oberpfalz is in sharp decline [51].These differences may be due to the data collection methods (e.g., analogue instead of digital mapping, simple counting of nests without comparing whether nests were recorded in the previous mapping or not, vague data handling from memory), small sample size, and the conclusions based on this [52,53].In addition, a study on the detection probability of RWA showed that beginners and laypersons had a very low detection probability of RWA nests, in contrast to our high detection probability when mapping RWA nests.High detection of RWA nests is essential for more realistic RWA nest counts [31].
Furthermore, the IUCN Red List is inconsistent with regard to F. polyctena, which is listed as 'not threatened' by the Red List Centre [54] but as 'near threatened' by the IUCN Red List.This IUCN category based on data collected almost 30 years ago and has not been updated or corrected since [55].Such outdated and contradictory database is neither helpful nor applicable to RWA conservation.
The Sustainability and Regional Conservation Concept of the Waldsassen Forestry Division from 2010, updated in 2019, describes a wide range of conservation measures as an integral part of near-natural forest management.The primary objectives are to maintain and promote biodiversity, to effectively apply conservation measures for RWA species, and to create site-appropriate, near-natural, climate-resilient, productive, and species-rich mixed forests [33].To protect RWA nests in the BSF, no logging or processing is allowed near neon-colored logs in the vicinity of a nest.The GPS-based (re-)inventory database of RWA nests from the 2016/2019 and 2023 studies will enable the BSF to achieve its overall goal of further protecting RWA nests by prioritizing conservation over timber harvesting [14].

Medium-Mature and Mature Pine-Dominated Forests Preferred
Forests in Germany are predominantly commercial forests [56].At both study areas, human intervention in the forest communities dates back to the Middle Ages and continued until after the Second World War [33].Where previously mixed forests of pine, spruce, fir, beech, oak, and lime were dominant, the forest composition was changed by wood-(over)consuming industries and livestock farming with intensive use of litter to coniferdominated forest stands that characterize the landscape.
In the study areas, the tree age classes preferred by RWA nests are medium mature (61-80 years) and mature (≥81-140 years) pine-dominated forests, corresponding to the average tree age of 77 years of German forests ( [57]; Figure 5), confirming our findings from previous inventories [14] in both the Oberpfalz and the WEVF [30] that F. polyctena and all nest heights are more abundant in mature forests.Mature spruce-dominated forests are habitats for start-ups, short nests and medium-sized nests, but with lower proportions compared to pine forests.This confirms findings in the WEVF [30], but contrasts with findings that newly planted forests (≤20 years) with an open canopy promote nest settlements [58].The fourfold increase in nest numbers in young (21-40 years; FBBSF) and newly planted spruce-dominated areas (MGBSF) also contradicts the findings of [59], who did not witness nests in 20-year-old Scots pine stands.
The best methods to manage forests to mitigate climate change and provide multiple ecosystem services for present and future generations is vehemently debated [60].As a natural change in tree composition is likely to occur very slowly due to low speed of seed dispersal, an active conversion to deciduous and structurally rich mixed forests is suggested to have stabilizing effects on tree composition, species, and habitat diversity [56,60,61].Therefore, one of the main objectives of the Bavarian Forest Policy and the Sustainability and Regional Conservation Concept of the Waldsassen Forestry Division is to create site-appropriate, near-natural, climate-tolerant, productive, and species-rich mixed forests and to protect valuable forest structures and important habitats through SFM.To support this goal, site-appropriate tree species are planted according to soil variability and conditions [33]: (a) fir (Abies alba), oak (Quercus robur), and beech (Fagus sylvatica) are planted on sandy, silty, or clayey parent material that results in periodically wet soils (MG, [62]), and (b) beech (Fagus silvatica), birch (Betula pendula), Douglas fir (Pseudotsuga menzisii), fir (Abies alba), Russian elm (Ulmus laevis), sycamore (Acer pseudoplatanus), and walnut (Juglans regia) are the preferred trees species on bedrock (granite, granodiorite), resulting in more acidic soil conditions [62][63][64].The forest composition observed in the field at and around RWA nest sites suggests (a) a shift from pure coniferous forests to a combination of coniferous and naturally grown deciduous trees and (b) naturally regenerated mixed stands, e.g., spruce-beech-oak, spruce-birch, pine-oak at nest sites.Furthermore, the proportion of deciduous trees such as birch (Betula pendula), beech (Fagus silvatica), oak (Quercus robur), red elder (Sambucus racemosa), alder (Alnus glutinosa), or a combination of two deciduous species increased in 2023 (Figure 6).In addition, areas of natural restocking showed pure deciduous species or combinations of these in both inventories, but at lower levels.This also confirms findings that ≈86% of young stocking with trees up to 4 m in height are naturally regenerated [56].
The Fauna-Flora-Habitat (FFH) Directive [65] serves to conserve, maintain, and restore and enable the long-term recovery of protected species and habitat types of common European interest.Natural disturbances can increase tree species diversity, as well as the diversity of biotope trees and deadwood, but can also decrease it, often exacerbated by pest infestations and the removal of damaged timber.The response of value-determining animal species to this particular conservation area has been little studied [66].The Natura 2000-Fauna-Flora-Habitat area in MG is characterized as a large wetland complex with traditional ponds and pond farming, wet meadows, bog and open habitats, siltation zones, and open water areas [67].The increased number of nests (especially short nests and medium-sized nests) of RWA, as a key ecological group of forest species, confirm favorable nest locations in these protected areas.In addition, the threefold increase in WpC in nests indirectly indicates a suitable habitat for various woodpecker species in this particular protected area [68].
Finding suggests that climate change-related natural disturbances in forests create early successional forest stages that provide high deadwood availability, light, and habitat heterogeneity.Such stages increase the diversity of many forest species, in particular promoting threatened red-listed species [60,66].Although the number of cleared plots and partly cleared plots increased between the inventories due to bark beetle (Ips typographus, Pitogenes chalcographus) infestations, windthrow, and snow break, the number of active nests on these plots tripled by 2023 (Table 2).This result confirms the results of the recent WEVF study [30] and contrasts with the results of [69], suggesting that bark beetle-induced degradation of lignin or holocellulose may negatively affect plant-sucking insects (Aphidae, Coccidae, Psyllidae) living in trophobiosis with RWA [70,71].To control bark beetle infestations, rapid salvage logging of the infested tree is performed, which has been suggested to reduce the survival of RWA nests [63].The fourfold increase in the number of new nests, especially of start-ups to medium-sized nests, contradicts the general statement of nest decline [72] on such plots.The spatial distribution of xerothermophilic species, such as the black-backed meadow ant (F.pratensis), has been suggested to be promoted by such dry habitats [71,73].In addition, it has been suggested that the reduction in wood quality in bark beetle-infested spruce stands alters nutrient cycling [70], and therefore smaller colonies in smaller RWA nests may be more likely to survive in cleared or partially cleared areas [30].

Increasing Herbaceous Biodiversity at and around RWA Nests
Important information on forest site characteristics can be obtained from spatial and temporal disturbances in forests that influence and change the diversity of species in the herb layer [74].In general, the observed composition of the herb layer at, on, and around a nest can be addressed in relation to the specific tree species such as cranberry-sprucefir-pine forests (Vaccinio vitis-idaeae-Abietetum) in MG and FB, and spruce, rowan, and birch (Betula pubescens-Sorbus aucuparia) on skeleton-rich soils (silicate blocks) with an acidic humus layer in sections of FB [33].This study confirms the findings of the previous inventories at and around RWA nests at both sites [14].In addition to the known main herbs that characterize the forest communities, an increase in herb biodiversity was observed in 2023, although all surveys were conducted in the same months to allow for comparison (Figure 7).This contrasts findings of the WEVF survey, where the herb layer composition decreased dramatically [30].Reasons for this increase in diversity may reflect inertia due to past environmental changes [75].Results from Great Britain showed that in addition to bird and butterfly communities, plant communities increased overall in species richness over short (20 years) time scales, associated with increasing temperatures and precipitation.These results cannot be fully confirmed because (a) the time scale of the re-inventory might be too short (four and seven years) and (b) little difference was found in the averaged weather data (e.g., temperature, humidity) for the inventory months in 2016, 2019, and 2023 [76].A more plausible explanation is a shift in forest composition due to logging, which created open spaces and allowed pioneer plants such as foxglove or mullein to colonize.

No Negative Impact from Dead Wood
Every forest ecosystem has dead wood, a dynamic resource that is produced and controlled by factors such as tree productivity, forest type and age, disturbances such as harvesting, windthrow, and snow break, and decomposition rates [10,77].Dead wood plays a key role in forest ecosystems by providing a habitat for a variety of species, contributing to the complexity of forest structure, and influencing fire behavior [78].In commercially managed forests, standing and coarse woody debris are severely depleted and need to be enriched to combat depletion biodiversity loss [79].This is the case in the Bavarian state forests, where dead wood is enriched in all stands, but especially in nearnatural-forests (≥140 and ≥100 years) with natural tree species composition during harvesting.The enrichment targets are 40 m 2 /ha and 20 m 2 /ha, respectively [80].In addition, special forest structures of particular ecological importance are created: (a) coarse DW, (b) biotope trees, and c) tree torsos up to 6 m high, which provide valuable habitats for insects, birds, bats, and small mammals as standing dead wood until their natural decay ( [34,80], Figure 2).
The qualitative DW analyses of this study showed that there is a high proportion of DW around RWA nests and identical nactDW/ha ratios at both sites (Table 5).The average amount of ≈44% (MGBSF) and ≈55% (FBBSF) for both inventories showed that BSF foresters at both study sites contribute to the overall objective of the Bavarian state forests and to the "2020 Nature Conservation Concept for the Waldsassen Forestry Division" of enriching DW in forests [33].
Private forest owners hold average forest sizes just under 2 ha.The large number of small forests results in a wide range of management options for these owners: from intensively managed "tidy" forests without any DW to extensively managed forests with whole islands of deadwood or irregularly managed forest in intermittent operation with DW there for decades.Most forest owners are open to the issue of dead wood, as evidenced by their willingness to have their forest certified according to the PEFC criteria (74% in Bavaria;).This means that forest owners voluntarily commit to leave an appropriate amount of dead wood and hollow trees [81].This could be confirmed by the results of this study, which show almost similar proportions of DW at and around RWA nests for both inventories (MGPF: ≈51%; ≈46% FBPF), which are well comparable to those of the BSF.This development of DW at both sites confirms findings of the federal forest inventories of 2002 and 2012, which showed an increase in DW from 11.6 to 20.6 m 3 per ha [79].
Overall, visual inspection and comparison of more than 12,000 photographs of the newly inventoried nests taken during the fieldwork for this study with those from the previous inventories showed that RWA nests are not negatively affected by downed DW (class 2 and 3), as confirmed by robust nest numbers in all differently owned forests and increased nest sizes (Table 5).The photo comparison showed that RWA simply enlarge their nests and incorporate fallen trees into their nest structure.This is corroborated by the findings of [82] that an artificial "highway" of dead logs in the vicinity of ant nests can promote nest growth, increase survival rates, and increase speed of movement on branches and logs [83].This contrasts to the findings of Zakharov [84], who observed that dead wood may contribute to the extinction of RWA nests in Ukraine.
Also, the nactDW-ha ratio for the different forest owners remained robust in both inventories, with the DW proportion decreasing by ≈12% at and around nests in FBBSF in 2023.Reasons for this may be that (a) DW observed in previous inventories is obscured by growth in nest height (NH) and diameter (ND) between inventories.Comparison and analysis of photographs of re-identified nests taken during the 2016/2019 and 2023 inventories confirmed that RWA incorporated branches and tree stems into the nest structure, eventually obscuring them with nest building material.In addition, standing DW that became downed trees is incorporated by RWA nests.(b) More logs were produced prior to the 1st inventory in 2019, the accumulated DW of which has already decomposed during the current inventory.(c) There has been no scheduled logging since 2018.As a result, less timber and processing volume was produced.

Increase in WpC in Nests and Cavities
Structured forests with several developmental stages, biotope trees, and a large supply of DW with a variety of (micro)habitats, e.g., bark pockets or woodpecker holes, are important for biodiversity.DW provides food, shelter, breeding, and drumming grounds for various specialists such as woodpeckers [57,68,85].
An indicator of forest biodiversity is the year-round active woodpecker, which uses large breeding territories and is adapted to forest habitat structures of old forests [86].The observed increase in WpCBSF (≈22%) at both sites, especially in short nests and mediumsized nests (Table 6; Figure 8), confirms the findings of [30] for the WEVF, although the increase is smaller compared to the WEVF.The monitoring of abandoned food cavities, which can accumulate over time, can be excluded because woodpecker cavities in nests are immediately maintained by ants within days and nights.Abandoned cavities negatively affect the temperature and humidity inside the nest and provide access to the nest and food for other predators, such as the European robin (Erithacus rubecula), common blackbird (Turdus merula), Eurasian jay (Garrulus glandarius), and common redstart (Phoenicurus phoenicurus), as shown by our year-long 24/7 on-site monitoring of various RWA species (>50.000h of video; AntCam data by Berberich and Berberich, unpublished).The increase in WpC in nests is also consistent with population increases in several woodpecker species (up to 20%) in Bavaria, such as the Three-toed (Picoides trydactylus), Green (Picus viridis), and Middle Spotted (Dendrocopos medius) Woodpecker [68].A large selection of DW is a prerequisite for their habitat and was observed at both sites in this study.
Our findings could also indicate a significant increase in the number of woodpeckers at both sites, which has also been observed by the responsible foresters.In particular, Black (Dryocopus martius), Green (Picus viridis), and Gray (Picus canus) woodpeckers forage not only on tree trunks but also in RWA nests and feed on ants [87].It would be more effective and advantageous for woodpeckers to forage in smaller nests, where prey can be captured more quickly.Foraging in large nests, as seen in the Green Woodpecker (Picus viridis), requires digging small tunnels into the nest and a longer stay [30].The higher cavity-to-NH class ratio in larger nests suggest that such nests contain a large number of queens and their brood.It might be more nutritious to forage in a larger supply of prey compared to smaller nests.
Furthermore, our results suggest that woodpeckers were more abundant in the BSF, suggesting a different forest managing than in the MF and PF (Table 6), e.g., in terms of preservation, enrichment, and replenishment of biotope trees and dead wood.
It is also possible that at the time of the two inventories at both sites, the number of habitat trees was decreasing and a stable woodpecker population was therefore more concentrated in RWA nests, as already suggested for the WEVF [30].Woodpecker cavities in RWA nests have been suggested to be a valuable, albeit indirect, indicator for assessing woodpecker populations and their forest habitats [30,86].

GeoBio-Interactions: Shift of Nest Hotspots from NE to SW
Dense forest, agriculture, and sediment cover limit the knowledge of the overall tectonic regime in both study areas.Recent studies in European countries, e.g., Denmark, Romania, and Germany, as well as the 2016/2019 inventories in the Oberpfalz with a focus on GeoBio-Interactions, have shown that tectonic-volcanic processes and geochemical composition of the bedrock, e.g., high soil radon concentrations, promote nest settlement, their high nest numbers, and the spatial distribution of RWA nests [14][15][16][17][18][19][20][21][22][23].
During the period between the inventories (four to seven years), seismic activity has increased, creating faults and providing degassing pathways in this region [88,89].Such structures are the most important prerequisite for nest settlements [14][15][16][17][18][19][20][21][22][23]30,31].The ≈9% increase in the total number of small nests (ND < 0.5 m; Table 1), the survival of more than ≈50% of start-ups and short nests since the last inventories, and the positive growth of nest size from start-ups and short nests to larger nests (nactR) are considered to indicate favorable settlement conditions due to active tectonics and confirm that RWA nests are bioindicators of such tectonic structures [14][15][16][17][18][19][20][21][22][23]30,31].In addition, the long persistence of large nests is suggested to indicate that the underlying fault structure and degassing over time are stable, providing a suitable location for RWA, e.g., the continuous, moderately warm granites [90] support the survival of small nests during winter in the study areas [14,91] and their growth into larger nests [14].This is also consistent with evidence from the tectonically active WEVF [30].
The highly clustered nests, especially in NW-SE direction, suggests a strong interaction between RWA nests distributions and tectonic structures in all inventories, namely (1) the NW-SE present-day stress field, (2) the fault line "Erbendorfer Linie" [36,47], and (3) the NW-SE to NNW-SSE-trending intrusion direction of the Falkenberger Granite Complex [90], confirming the GeoBio-Interaction findings in this area (Figure 1a; 9; [14]).This is in contrast with entomological studies, which suggest that the spatial distribution pattern of RWA nests and their local occurrence are controlled by, e.g., foraging and food supply, e.g., [92].
Furthermore, a shift of the nest hot spots towards the NW-SE-trending fault line "Erbendorfer Linie" in MG (Figure 9) is suggested to be caused by stronger degassing of the tectonically and structurally controlled "Erbendorfer Linie", which promotes RWA nest settlements [16,19] due to an increase in regional seismicity as an important driving force causing micro-fracturing (Figure 1a).These new fractures provide pathways for juvenile fluids with mantle-derived CO2 and Helium to migrate over long distances, indicating hidden volcanic structures in the vicinity of these degassing sites [42,43,88,89].In addition, seismic events lead to the formation of new emanation surfaces and pathways and a significant increase in the radon signal, even though the applied stress remains constant over time [93].At the same time, RWA nest hot spots are decreasing in the NE section of the MG study area.These areas are hypothesized to have a reduced degassing output or even closure of degassing structures due to the regional seismicity [19,42,43].The spatial distribution of RWA nests complements and clarifies the tectonic regime, as information on small-scale fault systems in MG and FB study area is obscured by vegetation cover.This hypothesis, already established in previous studies, has now been further confirmed [14,19].

Seven to Eight Year Re-Inventory Interval
A suitable and adequate time interval for re-inventories is a critical aspect for longterm monitoring of RWA nests.Re-inventories after different decades show different results: (a) constant nest numbers after two decades (England and Romania), (b) RWA being "under intense pressure" after three decades (Belgium), and (c) decreases and increases in nest counts after six decades (The Netherlands; [94][95][96][97]).Re-inventories with a time interval of ≥20 years are not suitable for documenting the dynamics of RWA populations and the associated forests compositions because (a) the dynamic systems of forests change rapidly, e.g., human intervention (e.g., logging) or natural effects (e.g., windthrow, snow break; [14,30]) and (b) the vitality of German forests is affected by climate change (hot summers, lack of precipitation) much faster than expected and might affect RWA population dynamics [57].
Our multiple experiences and findings from re-inventories of RWA nests carried out in the WEVF after 50, 25, 18, and 12 years [18,30] have already shown an increase in in the number of RWA nests, with maximum increases up to 85-fold (2021) compared to the 1984 inventory [30].An increase in active RWA nests (nact) of ≈10% for MG and ≈6% for FB was also confirmed in this study, which was conducted after seven years (MG) and four years (FB).Due to the longer timeframe of the re-inventory, growth and changes in RWA nest distribution are more apparent in MG than in FB.Based on our extensive experience with re-inventories of RWA nests and because the spatial distribution patterns of RWA nests can change rapidly over time, a time interval for re-inventories of seven to eight years seems appropriate for quantifying changes.This interval would also allow changes in forest ecology to be documented prior to updating the 10-year forest inventory and management plan at both sites.In order to compare the inventory methods, it is recommended that the inventory approach we have described be used and supplemented with the extensive photo database.This approach can be used to document not only the effects of climate change on forest vitality but also the potential effects on RWA population dynamics and to identify interactions between species and the BioGeo-Ecosystem [18,30,75].

Conclusions
In a first systematic, large-scale, area-wide re-inventory in two sustainably managed forests, we recorded presence/absence data of red wood ant (RWA; Formica rufa-group) nests in two study areas in the Oberpfalz, NE Bavaria, Germany.Results are as follows: (1) RWA nest numbers: We inventoried outstanding nest occurrences (5393 nests including 5276 active nests) and an increase of ≈8% over 4-7 years.(2) Forest composition: RWA nests preferred mature (≥81 years) and medium mature (61-80 years) pine-dominated forests.The increased biodiversity of the herb layer at and around nests is thought to be related to a shift in forest composition due to logging, which created open spaces and allowed pioneer plants such as foxglove or mullein to colonize.(3) Dead wood: High proportions of standing and downed dead wood, which are enriched in the course of SFM at both study sites, do not negatively affect nests, as confirmed by robust nest numbers and increased nest sizes in all differently owned forests.(4) Woodpecker cavities: The observed increase in woodpecker cavities and the number of nests with woodpecker cavities is an indirect indicator for the assessment of woodpecker populations and their forest habitats.( 5) GeoBio-Interactions: The highly clustered nests, especially in NW-SE direction, suggests a strong interaction between the distribution of RWA nests and their tectonic environment at both sites and in all three inventories.(6) Time intervals: We suggest seven to eight years as an appropriate time interval for re-inventories to quantify changes in nest numbers, forest composition, dead wood, woodpecker cavities, and the herb layer.
In addition, we were able to show that (1) the near-natural sustainable forest management not only maintains and promotes biodiversity and species-rich mixed forests, but also effectively protects RWA species, and (2) in order to protect RWA, it is not enough to simply monitor their presence/absence, but it is also important to identify interactions between species and the BioGeo-Ecosystem.

Figure 1 .
Figure 1.Tectonic setting of both study sites with (a) major tectonic units, faults (black lines), and earthquake events (blue dots) taken from literature [35-43].The inset shows the location of both sites in Bavaria, Germany, close to the Czech border; (b) Münchsgrün (MG) and (c) Falkenberg (FB) study areas detailing mapped sites for the 2016/2019 inventories and the 2023 re-inventory [14].

Figure 2 .
Figure 2. Examples of dead wood classes 1-3 in (a) MG and (b) FB study area: DW-1: biotope trees (weathered red signature "A" in MG is a marker for the RWA nest) and dying standing tree, DW-2: branches of different size and diameter as downed DW, and DW-3: downed DW with additional tree trunks.Photo credit: M.B.Berberich.

Figure 3 .
Figure 3. Gain and loss of RWA nests (ntot) for the (a) MG and (b) FB study area for the 2016/2019 inventories and the 2023 re-inventory.

Figure 4 .
Figure 4. Growth of active, re-identified nests (nactR) between the 2016/2019 inventories and the 2023 re-inventory for (a) MG and (b) FB study area.

Figure 5 .
Figure 5. Results of MANOVA for active nests for nest height (NH), nest diameter (ND), medium tree age (mTA), and woodpecker cavities (WpC), showing a grouped plot matrix (top) and six grouped scatter plots of the first two canonical variables and their centers (bottom) for the 2016/2019 inventories and the 2023 re-inventory for (a) MGBSF and (b) FBBSF.Tree ages with a signature of −100 represent no trees or clearings around a nest.Blue: 2016/2019 data; red: 2023 data.

Figure 6 .
Figure 6.Visual representation of the forest composition (log-normed) observed in the field at active RWA nests (nact): deciduous only (DT), coniferous only (CT), combined deciduous and coniferous
Number of woodpecker cavities (WpC) on the entire nest surface created by predation, e.g., by the Green Woodpecker (Picus viridis; divided into three classes: class 1: 1-5 cavities; class 2: 6-10 cavities; class 3: >10 cavities); 6.All visible herbs at, on, and around each RWA nest (within 1.5 m of the nest); 7. Tree species at, on, and around each RWA nest (within 1.5 m of the nest); 8. Qualitative information on standing or lying dead wood (DW; classified into three classes: DW-1: standing DW (dying, decaying or biotope trees), DW-2: downed DW (small and large branches of varying diameter lying on or around a nest), and DW-3: downed DW of class 2 with additional decaying logs of at least 15 cm in diameter intentionally enriched by foresters as part of sustainable forest management (Figure2).There were too few stumps for an additional class 4 "dead tree stumps" because RWA nests on, around, and on top of tree stumps (partially) covered the stumps with nest material; 9. Cleared plots (CP) due to forest management; 10.Finally, at least two photographs (landscape and normal format) of each nest were taken and stored in a photo database.These photographs were used to compare and 3. Activity (nact) or inactivity of RWA nests (ninact; nest structures were still present, but no ant activity was observed during the inventory; therefore, the nest was considered dead); 4. Location of RWA nests (e.g., within the forest, forest roads, forest edges); 5. re-identify (a) active nests (nactR), (b) forest composition, (c) dead wood classes, and (d) herbaceous layers mapped during the 2016/2019 campaigns; 11.Random sampling of RWA species in the field (≈10% of mapped RWA nests) identified mainly F. polyctena, as F. rufa and F. polyctena are difficult to distinguish without a stereomicroscope.

Private Forest (PF) FFH Area * Number of Active Nests (nact) Number of nact in BSF, MF, and PF hatot
Area and nests related to the FFH area "Waldnaabtal between Tirschenreuth and Windisch-Eschenbach" (DE6139371) are integrated into those of BSF, MF and PF areas and nests.

Table 3 .
Results of the MANOVA for the 2016/2019 inventories and the 2023 re-inventory for (a) MGBSF and (b) FBBSF.

Table 5 .
Descriptive statistics of total mapped area (hatot;) for forest owners (BSF, MF, PF), numbers of active nests (nact) in BSF, MF, and PF, numbers of re-identified active nests (nactR) in BSF, MF, and PF, dead wood classes 1-3 (DW), numbers of active nests with DW (nactDW), and ratio of active nests with dead wood per ha for the 2016, 2019, and 2023 inventories for (a) MG and (b) FB. -= not present.

Table 6 .
Descriptive statistics of number of active nests (nact) with WpC and number (n) of woodpecker cavities (WpC) in active nests for the MG and FB study site for the 2016/2019 inventories and the 2023 re-inventory.-= not present.