The role of bracket fungi in creating alpha diversity of invertebrates in the Białowieża National Park, Poland

Abstract Bracket fungi are seen mainly as the cause of economic losses in forestry, and their role as creators of biodiversity is relatively poorly understood. The aim of the study was defining the manner in which the degree of decay (DD) of the fruiting bodies determines the character of the invertebrate assemblages colonising them. The effect of this group of fungi on the modification of biodiversity of invertebrates (Aranae, Opiliones, Pseudoscorpionida, two groups of mites—Mesostigmata and Oribatida, and Collembola and Insecta) was investigated by analyzing 100 fruiting bodies of 10 species of bracket fungi divided into four DD classes. The material was collected at Białowieża National Park, which is considered to be the largest area of natural forests in the North European Plain. 16 068 invertebrate individuals classified into 224 species were obtained. Oribatid mites (12 543 individuals) constituted the largest group of individuals, which were classified into 115 species with the most numerous Carabodes femoralis (8,811 individuals). Representatives of this group of mites have been reported previously in the publications on bracket fungi; however, the contributions of Oribatida and other groups of invertebrates were not broadly compared. Moreover, the species such as Hoploseius mariae and H. oblongus, which were predominantly found in fruiting bodies of bracket fungi, have also been discerned. The invertebrate fauna differs depending on DD of the samples: In the more decayed samples, a higher number of both individuals and species were recorded compared to the samples with lower DDs; however, this trend proved to be nonlinear. The DCA and cluster analysis revealed a similarity of the invertebrate assemblages from the 2 DD and 4 DD samples. They also indicated that the group 3 DD differed the most from all the other samples. The indicator species analysis identified species characteristic to individual DDs: For group 1 DD, it was, for example, Hoploseius oblongus; for 2 DD—Orchesella bifasciata; and for 3 DD—Chernes cimicoides, while for 4 DD—Dinychus perforatus.

Unfortunately, the knowledge of the subject is still insufficient and there is a lack of comprehensive studies covering all invertebrate groups and the character of their assemblages.
Furthermore, there are almost no publications identifying factors, which determine the species composition and dependencies within those groups. In contrast to the studies investigating the fauna colonizing decaying wood, the publications concerning invertebrate assemblages which colonize fruiting bodies of bracket fungi neglect to analyze the degree of decay in bracket fungi. Studies have particularly been focusing on the fauna of invertebrates found in various fungal species (e.g., Gwiazdowicz & Łakomy, 2002;Makarova, 2002).
This study aimed to investigate the invertebrate fauna colonizing such a unique microhabitat as fruiting bodies of bracket fungi, as well as to identify how degree of decay (DD) of fruiting bodies determines the character of these assemblages. The research derives from previous studies conducted by Gwiazdowicz and Łakomy (2002) stating that the fungal species is not a key factor in determining the character of the invertebrate assemblage.
Other research also showed that it was not the species of the fungus, but other factors (whether the fruiting bodies were alive or dead) that influenced the invertebrate communities inhabiting it (Thunes & Willasen, 1997). Also in the O' Connell and Bolger (1997a) research in which the invertebrate fauna inhabiting the fruiting bodies of 40 species of fungi was examined, characteristic faunas were not detected for any species or higher taxon of fungus. For this reason, a hypothesis was formulated that the invertebrate assemblages (both the number and the species' richness) colonizing fruiting bodies of bracket fungi vary depending on their degree of decay (DD).

| Study area
To gain an insight into the colonization process of fruiting bodies of bracket fungi by invertebrates, it was decided to select Białowieża National Park (BNP) in eastern Poland as the study area. This selection was not an arbitrary one, since the BNP is considered to be the largest area of natural forest in the North European Plain (Gutowski & Jaroszewicz, 2004). Therefore, the BNP is widely recognized as a model forest, to which various observations and research of natural deciduous and mixed forests in this climate zone can be referred (Gutowski, 2004). Białowieża National Park is the oldest Polish national park (established in 1947). It is located in the northeastern part of Poland (52°69′89′′ N-52°81′89′′ N, 23°71′76′′ E-23°93′95′′ E) (Figure 1). The maximum altitude in the BNP is 176.3 m above sea level. According to Sokołowski (2004) (Jaroszewicz, 2010;Kwiatkowski & Gajko, 2009). Both bracket fungi (Domański, 1964(Domański, , 1967Niemelä, 2013) and numerous invertebrate groups (e.g., Gutowski & Krzysztofiak, 1995;Jaworski et al., 2014;Stańska et al., 2016) have already been researched at Białowieża primeval forest, which facilitates a comparison of the obtained results with the current knowledge on the subject matter. It is worth noting that this research is the first one on the subject of invertebrate assemblages inhabiting fruiting bodies of bracket fungi in the BNP area.

| Collecting material
In the course of a survey of bracket fungi conducted in the years 2008-2012, a total of 142 species were recorded at BNP (Niemelä, 2013). However, during the fieldworks for this manuscript only the most numerously represented species were collected. In the months from June to August in 2014, a total of 100 specimens (fruiting bodies) belonging to 10 species of bracket fungi were harvested. For that purpose, an axe was used to cut the bracket fungi off of tree trunks.

| Laboratory procedures
The fruiting bodies were identified at the laboratory, and their DD was determined (Table 1). The classification of the DDs of fruiting bodies had not been used so far-it is proposed by the authors, and like the wood decay scales (e.g., Maser et al., 1979), it is based on the differences in the occurrence of various features in the substrate with a different degree of decay. Next, in order to extract mesofauna, the fruiting bodies were placed in Tullgren funnels for 72 hr.
The collected invertebrates were preserved in 75% ethanol and then classified into seven taxonomic groups: spiders (Aranae), Opiliones, pseudoscorpions (Pseudoscorpionida), two groups of mites (Mesostigmata, Oribatida), springtails (Collembola), and insects (Insecta). Due to the fact that there was only one species of Opiliones in the research, and its trends were very similar to that of spiders, Aranae and Opiliones were connected into one group; therefore, in the further part of the article, six groups are mentioned, and spiders and Opiliones are analyzed and discussed together.
The collected Araneae and Opiliones were identified and counted. The taxonomic keys of valid spiders and Opiliones (Nentwig et al., 2020;Roberts, 1985;Rozwałka, 2017) were used to identify the species. F I G U R E 1 Locality of the study area (in the context of Poland and Białowieża Forest) The collected individuals of Pseudoscorpionida were examined using a stereomicroscope. For some species, it was necessary to prepare temporary microscope slides so that they could be studied under the compound microscope. The taxonomic keys of Beier (1963) and Christophoryová et al., (2011), Christophoryová Šťáhlavský, andFedor (2011) were used to identify the species.
The Oribatida were identified using a light microscope, preferably with a phase contrast and a differential interference contrast.
Prior to the examination, the cuticles were rendered transparent using concentrated lactic acid, 60% lactic acid, or lactophenol.
The clearing process was performed at room temperature over a course of several days, and sometimes weeks. Oribatid mites were determined at a species level by identifying their key features and original species descriptions (Niedbała, 2008;Olszanowski, 1996;Weigmann, 2006). The names of the species were confirmed according to Weigmann (2006). The Collembola were identified using a light microscope. The extracted specimens of springtails were mounted with Hoyer's medium (Coleman et al., 2004) to prepare the permanent microscopic slides necessary for the taxonomic analysis. The taxonomic identification of the Collembola was carried out by following the manuals of Fjellberg (1998Fjellberg ( , 2007, Bretfeld (1999), Potapov (2001, Thibaud et al., (2004), Dunger andSchlitt (2011), andJordana (2012).
The identification of immature stages of collected Insecta was carried out by following Stehr's manuals (Stehr, 2005;Stehr, 2008).
Here, invertebrates analyzed in this study were identified as higher taxonomic units, for example, order or family, and were in-

| Statistics
In order to assess the differences between the samples in terms of the invertebrate composition to avoid interpretation problems related to data compression, the detrended correspondence analysis (DCA) developed by Hill and Gauch (1980)

| General information
The number of bracket fungi varied within each species, which was a result of the biology and ecology of these fungi and was also due to their more numerous or sporadic occurrence. The following 10 species were collected: Fomitopsis pinicola (47 specimens),

1.
Fruiting body with fresh hymenophore, without visible signs of decay.

2.
Fruiting body with dry hymenophore, without visible signs of decay.

3.
Fruiting body with few traces of decomposition, for example, single (up to 10) insect galleries.

4.
Fruiting body with numerous traces of decay, such as insect galleries, easily crumbled.

| Diversity of invertebrate assemblages depending on DD
The study assumed that both the number and the species' richness of invertebrates depend on the DD of bracket fungi. In order to verify this thesis, the data analyses were conducted in order to indicate a trend for the occurrence of a higher mean number of individuals and species per sample in the higher DD classes (Figure 2a While analyzing the obtained results, it may also be observed that certain invertebrate species, such as Eniochthonius minutissimus, were present in all DD classes, except 3 DD. There were also some others, which were detected both in fresh fruiting bodies In order to determine whether DD has effect on the number of individuals and species of invertebrates colonizing bracket fungi, the DCA was conducted (using the decorana function in the R platform).
The results of this analysis are presented in Figure 3. The first two axes indicate 54.86% species variation, with the first axis accounting for 30.35% and the second axis for 24.52%. It is evident that most points corresponding to the individual samples form one cluster. The application of DCA made it possible to "expand" this cluster, thanks to which a division of samples into groups in terms of their DD may be observed. Thus, the analysis provided a partial grouping of bracket fungi in terms of their DDs and indicated similarities between these groups. In the DCA, the groups are considered to be different if they are separated linearly. Figure 3 shows no marked divisions between groups; however, certain similarities may be observed.
The DCA and cluster analysis reveal a degree of differences between DDs. Thus, the permutation analysis of variance (adonis function) was conducted. The homogeneity criterion of multivariate dispersions is fulfilled (p = .731, F = 0.468; df: 3, 94, function permutest). It was shown that the effects of DDs differ significantly (p = .001, see Tab. 5), but since R 2 = 0.061 this suggests an even distribution of high and low ranks, both within and between groups.
F I G U R E 2 Minimum, maximum, mean number ± SE of individuals (a) and species (b) of invertebrates depending on DD of samples (1-4degrees of decay of fungi) In addition, only 6% of the variation in distance between the groups is explained by grouping testing. As a result, the groups differ significantly, but these differences are not distinctly evident.
The result obtained in the discussed analysis is consistent with the results of cluster analysis (Figure 4). This shows that samples belonging to 1 DD and 4 DD form one cluster (in Figure 3 mixed points), continuing on 2 DD can also be added to this cluster, and lastly 3 DD was also added, but it may also be considered a separate cluster.
Moreover, the pairwise comparisons were conducted using a permutation analysis of variances method (adonis function) with the Bonferroni correction of p-values (Table 2) The indicator species analysis (using the multipatt function from the indicspecies package), performed in order to show which invertebrates contributed to the group differentiation, indicates 11 taxa to distinguish the groups. There are 10 taxa associated with 1 group and 1 species associated with 2 groups (

| The effect of DD fruiting bodies on individual invertebrate groups
While analyzing the obtained results, it was also observed that the individual invertebrate groups react differently to the changes in the DD of the substrate. In the case of spiders and Opiliones in bracket fungi of the 1 DD group, three species were reported and were rep-  (Figures 5a and 6a, Appendix ).
The results concerning Pseudoscorpionida were slightly different. In the 1 DD group, three individuals from the Chernetidae family were identified. The same species was also found in the samples of all DD groups. In the 2 DD samples, two species were reported (5 individuals in total), while both 3 DD and 4 DD reported three species of pseudoscorpions each. In addition, the number of individuals was also similar, with 32 and 30 individuals, respectively (Figures 5b   and 6b).
As far as Mesostigmata mites in the 1 DD samples are concerned, a total of 23 species and 315 individuals were reported. In all of the samples, a total of 319 individuals and 19 species of Collembola were reported. There were 57 individuals belonging to 9 species found in the fruiting bodies with 1 DD. Two species (Pseudisotoma sensibilis and Sinella myrmecophila) were recorded in the samples from this DD group more frequently than in the fruiting bodies from the other DD groups. In the 2 DD samples, 24 individuals belonging to 8 species were found. Orchesella bifasciata was the only species recorded solely in the 2 DD samples, while Pseudachorutella assigilata and Pseudosinella immaculata were found in bracket fungi from these DD groups in greater numbers than in the other groups. In the 3 DD samples, 26 individuals representing 7 species were recorded. Entomobrya multifasciata was the only species found solely in the 3 DD samples. Springtails were most numerously represented in terms of both individuals (212) and species (14) in the 4 DD samples (Figures 5e and 6e). Six species were reported, among them Arrhopalites coecus and Cyphoderus albinus, and found only in this DD group, while Caprainea marginata, Entomobrya corticalis, and Tomocerus minor were recorded in the 4 DD samples more frequently than in the other groups.
The last group of invertebrates identified in this study comprised of insects. In the 1 DD samples, a total of 320 individuals were detected belonging to 8 species. Cyphon sp. was the only species of insects recorded solely in the 1 DD samples, while Coleoptera and Diptera were found in bracket fungi samples with this DD more numerously than in the other groups. In the 2 DD samples, more individuals (251) and a larger number of species (11) of insects were recorded. Nargus velox, Dermaptera and Psocoptera, were found only in the fruiting bodies of the DD group, while Thysanoptera were identified in the 2 DD samples more frequently than in the other groups. In the 3 DD samples, even greater numbers of individuals (407) and species (16) of insects were found. Only in the fruiting bodies of this DD group did the analyses show the presence of Attagenus sp., Bolitophagus reticulatus, Corticaria sp., or Dermestidae, whereas Ennearthron sp., Octotemnus sp., and Ciidae in the 3 DD samples were found more frequently than in the other groups. In the 4 DD samples, the presence of 717 individuals belonging to 13 insect species was reported (Figures 5f and 6f). Five of them, for example, Acrotrichis sp., Cerylon fagi, or Cerylon ferrugineum, were found solely in the samples from this DD group, whereas Cis spp. were markedly more numerous in the most decayed samples compared with the other DD groups.

| D ISCUSS I ON AND CON CLUS I ON S
Numerous studies have shown that bracket fungi inhabit specific habitats, described as patchy (often ephemeral), resembling an extremely heterogeneous environmental mosaic rather than "islands" in the true sense of the word (e.g., O'Connell & Bolger, 1997a, 1997b. Sporocarps of polypore fungi represent well-defined, patchy, and temporary habitats, which have long-fascinated ecologists and biodiversity researchers (Hågvar & Steen, 2013). They may be termed biological hotspots, supporting high numbers of species within small volumes (Komonen, 2003). The evidence for the unique character of this type of microhabitat may be provided, for instance, by the occurrence of organisms reported practically only in such habitats, as it is the case with such insect species as, for example, Megaselia lobatafurcae and M. parspallida (Disney & Pagola-Carte, 2009), or certain mite species, for example, from the Hoploseius genus (e.g., Gwiazdowicz, 2002;Mašán & Halliday, 2016;Walter, 1998). It also needs to be emphasized here that some invertebrate taxa, such as Hoploseius tenuis, even indicate certain morphological adaptations (an elongated and narrow body) to live within the pores of bracket fungi (Lindquist, 1965(Lindquist, , 1995. Some species reported in this microhabitat are also found in other similar habitats, such as feeding galleries of insects located under bark or in decaying wood (e.g., Salmane & Brumelis, 2010).
The invertebrate species found in the greatest numbers in this study were also reported in microhabitats other than fungal fruiting bodies. Carabodes femoralis, a mycophagous, decomposer oribatid mite (Manu & Honciuc, 2010;Schneider et al., 2005), is the most numerous species among all the invertebrates present in this study. It is found mainly in litter and soil (e.g., Błoszyk & Olszanowski, 1997;Manu & Honciuc, 2010;Seniczak et al., 2006), as well as in the nests of Formica rufa ants (Sell, 1990), cave mud, deadwood, leaves, and guano (Maślak & Barczyk, 2011). In addition, the invertebrate species ranking second in the terms of its numbers, Carabodes subarcticus, was also found in such microhabitats as soil (e.g., Hågvar et al., 2014;Kagainis, 2010Kagainis, , 2015 and bark of deadwood (Bluhm et al., 2015). In the case of the most numerously found invertebrate taxa from the other groups, it may also be observed that apart from fungal fruiting bodies they are also observed in other substrates. Spiders from the Linyphiidae family are also found in various habitats, with only a slight preference to forests (Wiśniewski et al., 2018), agricultural habitats (Downie et al., 2000;Schmidt & Tscharntke, 2005), and grass (Thomas & Jepson, 2003). Chernes cimicoides, the pseudoscorpion taxon most frequently recorded in this study, was also identified in tree hollows under tree bark (Krajčovičová & Christophoryová, 2014), under deadwood bark (Krajčovičová et al., 2018), in window traps at the bark and trunk eclectors (Muster & Blick, 2015), with an indigenous species found in bird nests (Christophoryová, Krumpálová, et al., 2011;Christophoryová, Šťáhlavský, et al., 2011).
The Mesostigmata mite recorded in greatest numbers in this study, Dendrolaelaps pini, was also found in decayed wood (Gwiazdowicz et al., 2011), Ips typographus galleries (Salavatulin et al., 2018), pine stumps and under wing covers of Hylurgus ligniperda and Hylastes sp. (Hirschmann & Wiśniewski, 1982). Entomobrya corticalis, the most numerous among the identified springtail species, is also known to live in such microhabitats as deadwood, epiphytic mosses, litter, and soil (Daghighi et al., 2013;Skarżyński et al., 2016;Yahyapour et al., 2019). Also, less numerous species, such as Pseudisotoma sensibilis and Orchesella bifasciata, are mentioned as corticolous and bryophilous species (Fjellberg, 2007;Weiner, 1981 A study conducted by Gwiazdowicz et al., (2011) also suggested that certain mite species are more associated with a specific but not necessarily high DD of the substrate, while in the other DD groups they are found in fewer numbers. Also, in the case of this study, despite the analysis of the slight substrate differences, similar dependencies have been observed-this was shown by means of indicator species analysis, used to identify the invertebrate species, which distinguished individual DD classes. This analysis showed not only which specific invertebrate species distinguish individual DD groups, but also the fact that the number of these species is small, since a con- The results of this study show that fruiting bodies of bracket fungi constitute a microhabitat colonized by unique invertebrate assemblages, among which certain taxa are found almost solely in this habitat type. It also appears that the DD of fruiting bodies can play role in terms of biodiversity modification of colonizing invertebrates.
This information may be useful both as supplementary data concerning the currently little-known problems of invertebrate ecology. The data also entail sustainable forestry based on ecological principles, in which bracket fungi are treated not only as an indicator of potential economic losses, but also as an element of the natural environment having an impact on biodiversity.

ACK N OWLED G M ENTS
The publication is co-financed within the framework of Ministry of Science and Higher Education program as "Regional Initiative

CO N FLI C T O F I NTE R E S T
We have no conflicts of interest to disclose.

AUTH O R CO NTR I B UTI O N S
Anna Gdula: Data curation (equal); Investigation (equal); Resources