Vertical Stratiﬁcation of Beetles in Deciduous Forest Communities in the Centre of European Russia

: Studies on the vertical distribution of arthropods in temperate forests have revealed the uneven vertical distribution of communities. Many factors inﬂuence these patterns simultaneously. However, there are still many questions related to the vertical distribution of Coleoptera in deciduous forests of the temperate zone. The research was carried out within the territory of the Republic of Mordovia (the center of the European part of Russia). Fermental traps with a bait made of fermenting beer with sugar were used to collect Coleoptera. The collections were carried out from May to September 2020 at ﬁve sites in a deciduous forest. We set traps at a height of 1.5, 3.5, 7.5 and 12 m above the ground) on the branches of trees. Ninety-two species were identiﬁed at the end of studies at different heights. The families Nitidulidae (15 species), Cerambycidae (14 species), Elateridae (7 species), Curculionidae (7 species) and Scarabaeidae (7 species) had the greatest species diversity. The greatest species diversity was recorded at a height of 1.5 m, while the smallest one was recorded at a height of 7.5 m. The minimum number of specimens was recorded at a height of 12 m. The largest differences in the Jaccard similarity index were obtained between samples from a height of 1.5 and 12 m. The Shannon’s diversity index was higher near the ground than in the tree crowns (at heights of 7.5 and 12 m), and the Simpson index had the opposite tendency. Glischrochilus hortensis and to a lesser extent Cychramus luteus preferred to live in the lowest layers of deciduous forest (1.5 m). Cryptarcha strigata was mainly found with relatively high numbers at heights of 3.5 m and 7.5 m. The abundance and occurrence of Protaetia marmorata and Quedius dilatatus were higher in the uppermost layers of the crowns. The number of saproxylic beetle species at heights of 3.5–12 m was almost the same, while in the surface layer it decreased. The number of anthophilic beetle species was also lower at a low altitude. Our data conﬁrm the relevance of sampling in forest ecosystems at different altitudes while studying arthropod biodiversity.


Sampling Procedures
Coleoptera was collected from May to September 2020 in the spring-autumn period, when the activity of insects was the highest. All the places where the material was collected were solid forests. There were five such parts of deciduous forest in total. Each fragment of the forest was located more than 1.5 km from each other and was independent of the other fragment under study (replicate). This distance between points is an attempt to ensure non-dependence between samples; i.e., these areas were separate forest parts. Therefore, the sampling points were considered as replicates (n = 5).
Each sampling point had a set of four traps installed at different heights (1.5, 3.5, 7.5 and 12 m above the ground) on the branches of trees. We chose this height difference in order to facilitate and optimize the manual installation of traps without the need to climb  Figure 1).

Sampling Procedures
Coleoptera was collected from May to September 2020 in the spring-autumn period, when the activity of insects was the highest. All the places where the material was collected were solid forests. There were five such parts of deciduous forest in total. Each fragment of the forest was located more than 1.5 km from each other and was independent of the other fragment under study (replicate). This distance between points is an attempt to ensure non-dependence between samples; i.e., these areas were separate forest parts. Therefore, the sampling points were considered as replicates (n = 5).
Each sampling point had a set of four traps installed at different heights (1.5, 3.5, 7.5 and 12 m above the ground) on the branches of trees. We chose this height difference in order to facilitate and optimize the manual installation of traps without the need to climb trees or use special techniques and/or equipment. To avoid a possible edge effect, traps were installed inside forest areas. The total sampling effort was 172 trap exposures. There were nine replicates at each height (there were seven repetitions in one locality). 144 expositions were made in four collection localities (nine replicates at four heights), 28 expositions were made in one locality (seven replicates at four heights).
All collections were carried out using traps of our own design. A five-liter plastic container with a window cut out on one side at a distance of 10 cm from the bottom was used as a trap [42]. Beer was used as bait. Sugar was added to it for fermentation.
The collected samples were delivered in plastic bags containing 70% alcohol from the forest to the laboratory, then sorted and conserved in alcohol.

Identification and Taxonomic Position of Samples
The classification of the family-group taxa used in this checklist follows predominantly Bouchard et al. [43]) with subsequent additions [44]. Changes for Coleoptera have been taken into account from the Catalog of Palaearctic [45][46][47][48][49][50][51], as well as for Cucujoidea from the publication of Robertson et al. [52] and for Curculionoidea from the publication of Alonso-Zarazaga et al. [53]. To clarify the nomenclature, the cited works were used, as well as the Catalog of Palaearctic Coleoptera [54,55]. The years of description of some species are specified according to Bousquet [56]. The species identification was carried out by L.V. Egorov. The samples are kept in the collection of the Mordovia State Nature Reserve (Pushta, Russia).

Data Analyses
When analyzing the results, we used only data on the quantitative parameter (number) of all Coleoptera individuals in traps for exposure time. Exposure time is the period between hanging a trap and taking samples for analysis (expressed in days). Mean number (M, expressed in %) was calculated based on the exposures of all traps at a given height. Occurrence is the ratio of the number of samples in which a species (taxonomic group) is present to the total number of samples (expressed in %). Saproxylic species were determined taking into account the approaches adopted by a number of authors [57][58][59]. The anthophilic species were classified according to our own long-term observations.
To compare species similarity between study plots we used Jaccard index. We did not take into account insects, which were not identified to species level. Based on the collected data, we calculated widely used biodiversity indices, namely the Shannon's Diversity Index and the Simpson's diversity index [60,61].
Statistical analyses were carried out using PAST 4.07. The ordination techniques, using the principal component analysis (PCA), defined the major gradients in the spatial arrangement of the studied species selected for the analysis. For ecological interpretation of the ordination axes, groups of the height of bait trap positions were plotted onto the PCA ordination diagram as supplementary environmental data. We analyzed the species, which were represented at least 30 individuals during the sampling period.

Results
As a result of processing the material, 92 species from 26 Coleoptera families (Table A1) were identified. A total of 7882 individuals have been studied. Some specimens from the families Staphylinidae, Nitidulidae, Ptinidae, and Buprestidae could not be identified to species. Such families as Nitidulidae (15 species), Cerambycidae (14 species), Elateridae (7 species), Curculionidae (7 species) and Scarabaeidae (7 species) had the greatest species diversity (Figures 3 and 4). Representatives of these five families and the family Staphylinidae made up a total of 91.1% of all studied specimens.
The greatest species diversity (58 species) was recorded at a height of 1.5 m, the smallest one (40 species)-at of 7.5 m (Table A1). According to the average number of specimens, the highest numbers were obtained at heights of 1.5 and 3.5 m (on average of sampling point, 427 and 428 specimens, respectively). The minimum number of specimens was caught at a height of 12 m. The relative number of saproxylic beetle species was lower at low altitude, whereas at other altitudes it increased slightly. At heights of 3.5-12 m, the number of saproxylic species was almost the same (Table A1). The relative number of anthophilic species was also lower at low altitude. However, at other heights it sharply increased (Table A1).     (Table 1). The calculation of the Jaccard similarity index revealed that there were certain differences among the heights at which Coleoptera were recorded ( Figure 5). The greatest differences were obtained between samples from a height of 1.5 and 12 m. At the same time, the differences between the heights of 3.5 and 7.5 m were minimal. The calculation of the Jaccard similarity index revealed that there were certain differences among the heights at which Coleoptera were recorded ( Figure 5). The greatest differences were obtained between samples from a height of 1.5 and 12 m. At the same time, the differences between the heights of 3.5 and 7.5 m were minimal. In Figure 6, the spatial arrangement of the selected beetle species demonstrated that the majority of these taxa have no specific preferences to the height in the forest ecosystem. However, five species were exceptions. So, Cryptarcha strigata mostly occurred at heights of 3.5 m and 7.5 m with a relatively high abundance. Protaetia marmorata had clear preferences to the highest layer of the forest community (12 m). Quedius dilatatus also preferred the highest layers of forest crowns, also occurring at an altitude of 7.5 m. In In Figure 6, the spatial arrangement of the selected beetle species demonstrated that the majority of these taxa have no specific preferences to the height in the forest ecosystem. However, five species were exceptions. So, Cryptarcha strigata mostly occurred at heights of 3.5 m and 7.5 m with a relatively high abundance. Protaetia marmorata had clear preferences to the highest layer of the forest community (12 m). Quedius dilatatus also preferred the highest layers of forest crowns, also occurring at an altitude of 7.5 m. In contrary to the previous species, Glischrochilus hortensis and in the lesser degree Cychramus luteus had preferences to inhabit the lowest layers of the forest ecosystem (1.5 m).  The total occurrence of Coleoptera was higher at a height of 1.5 m and gradually decreased as the height increased, i.e., the higher the traps were located, the lower the occurrence of beetles. Cryptarcha strigata, Protaetia marmorata, Glischrochilus hortensis and Soronia grisea had the highest occurrence rates. Thus, those species whose abundance in traps was high had the highest occurrence rates. Figure 7 shows the analysis of the main components based on the occurrence of the species selected for analysis. Cryptarcha strigata occupies a separate position on the chart due to the highest occurrence value with slightly higher numbers at an altitude of 3.5 m. Soronia grisea was more often found at an altitude of 3.5 m. The occurrence of Protaetia marmorata, Quedius dilatatus and Cryptarcha undata was higher at altitudes of 7.5 m and 12.0 m, whereas Glischrochilus hortensis, on the contrary, was more common at low altitudes-1.5 and 3.5 m. Differences are insignificant in the occurrence of other beetle species among heights. The total occurrence of Coleoptera was higher at a height of 1.5 m and gradually decreased as the height increased, i.e., the higher the traps were located, the lower the occurrence of beetles. Cryptarcha strigata, Protaetia marmorata, Glischrochilus hortensis and Soronia grisea had the highest occurrence rates. Thus, those species whose abundance in traps was high had the highest occurrence rates. Figure 7 shows the analysis of the main components based on the occurrence of the species selected for analysis. Cryptarcha strigata occupies a separate position on the chart due to the highest occurrence value with slightly higher numbers at an altitude of 3.5 m. Soronia grisea was more often found at an altitude of 3.5 m. The occurrence of Protaetia marmorata, Quedius dilatatus and Cryptarcha undata was higher at altitudes of 7.5 m and 12.0 m, whereas Glischrochilus hortensis, on the contrary, was more common at low altitudes-1.5 and 3.5 m. Differences are insignificant in the occurrence of other beetle species among heights.

Discussion
This study shows the location of Coleoptera clusters selected using beer traps installed at different heights in temperate forests of European Russia. Different species of Coleoptera fall into such traps, but most of them are species that fly to the fermenting bait. Previously, it was determined that such traps attract a small number of species compared to the total species diversity that falls into these traps. However, the number of specimens actively flying to bait is extremely high and usually amounts to more than 90% of the total number of specimens [62]. In these studies, we have obtained similar results.
Our results show that the abundance and species diversity of Coleoptera is higher when the trap is set at a height of 1.5 m. The Shannon's diversity index was the highest near the ground than in the tree crowns (at heights of 7.5 and 12 m), but there were no differences in total abundance or species richness between the two layers. Ulyshen and Hanula [63] obtained similar results. Thus, there is a small species diversity and dominance of one or more species in the crowns of the tree.
According to our research, despite the fact that most beetle species are distributed more or less evenly within the vertical section of the forest, we still identified species that reliably preferred a certain height. For example, Glischrochilus hortensis and to a lesser extent Cychramus luteus preferred to live in the lowest layers of the forest ecosystem (1.

Discussion
This study shows the location of Coleoptera clusters selected using beer traps installed at different heights in temperate forests of European Russia. Different species of Coleoptera fall into such traps, but most of them are species that fly to the fermenting bait. Previously, it was determined that such traps attract a small number of species compared to the total species diversity that falls into these traps. However, the number of specimens actively flying to bait is extremely high and usually amounts to more than 90% of the total number of specimens [62]. In these studies, we have obtained similar results.
Our results show that the abundance and species diversity of Coleoptera is higher when the trap is set at a height of 1.5 m. The Shannon's diversity index was the highest near the ground than in the tree crowns (at heights of 7.5 and 12 m), but there were no differences in total abundance or species richness between the two layers. Ulyshen and Hanula [63] obtained similar results. Thus, there is a small species diversity and dominance of one or more species in the crowns of the tree.
According to our research, despite the fact that most beetle species are distributed more or less evenly within the vertical section of the forest, we still identified species that reliably preferred a certain height. For example, Glischrochilus hortensis and to a lesser extent Cychramus luteus preferred to live in the lowest layers of the forest ecosystem (1.5 m). Glischrochilus hortensis adults are found on the fermenting sap of Quercus robur and under the bark of fallen and dying trees Betula pendula, Populus tremula. Larvae develop under the bark of dying and damaged B. pendula, P. tremula, Q. robur leaves and in their fermented juice. They can also occur on fermented berries, mushrooms and vegetables [72][73][74]. Cychramus luteus imago are anthophiles and are found on flowers in summer. Later they switch to feeding on Armillaria mellea mushrooms, where their larvae develop [75]. Therefore, we assume that both species prefer the lower ground level of the forest. Cryptarcha strigata was mainly found at medium altitudes (undergrowth) with relatively high numbers. Usually imago of this species live near the leaking fermenting juice of Q. robur, where preimaginal phases develop. They occasionally fall on the leaking juice of P. tremula [73]. The greatest numbers are obtained in biotopes with the predominance of these species [71].
Protaetia marmorata had a clear preference for the highest layer of the forest. Larvae of this species develop in the hollow of dead deciduous trees for 3 years, most often in oaks [68,71,76]. Quedius dilatatus also preferred the highest layers of forest layers, occurring as well at an altitude of 7.5 m. It shows a connection with Vespa crabro nests, where its larvae feed on Diptera larvae in the nest debris [77]. It also occurred in wasp nests living in natural conditions. Such nests are located on old oak trees, apple trees, and other deciduous trees. The species was also found on fermenting sap on an oak trunk [78]. Both species of bark beetles Anisandrus dispar and Xyleborinus saxesenii (Ratzeburg, 1837) were caught in the largest number at an height of 1.5 m. The ambrosia beetles Scolytinae (Curculionidae) usually prefer to inhabit the lower parts of the tree crowns, so most of them are trapped at a height of up to 2 m [79][80][81][82].
There is an interesting finding in our study. We have registered a relatively small species diversity of saproxyl beetles at a height of 1.5 m. Saproxylic beetles usually account for 30% of all Coleoptera species in forest ecosystems [83,84]. Their species diversity is usually higher in warmer forest areas with an abundance of dead wood, dead trees, stumps, coarse wood debris [85][86][87][88][89][90]. Some authors also associate a significant increase in the species diversity of saproxylic beetles with an increase in temperature in forest areas with dead wood, on illuminated edges [87,91,92]. Schroeder et al. [93] found differences in the composition of Coleoptera living in wood between the understory and canopy of deciduous forests in Canada. Bouget et al. [27] recorded an increase in abundance and species richness of the proximal species in the undergrowth of beech-fir and oak forests. In temperate forests, the proportion of saprophages is higher in the lower tiers of the forest, but in tropical forests, it increased with an increase in the height of traps. This is due to greater competition among individual groups of insects in the lower tiers of tropical forests [4]. However, Vodka and Cizek [24] noticed that the diversity of saproxylic species was higher in the undergrowth than in the canopy at the edge of the forest, while the opposite situation was observed in the depths of the forest. Preisser et al. [94] noticed that 86 out of 101 collected insect families were more numerous in traps at ground level than in traps under the canopy (the authors used two types of traps). Gossner et al. [95] noted an increase in species richness also at the lower levels of the forest. These differences can be caused by reactions at the level of species and families, which are caused by differences in behavior, ability to settle, ecological interactions, microclimate or spatial heterogeneity of the quality and quantity of food [96][97][98][99].
It is possible that a bait with a mixture of fermenting beer and sugar has a certain effect on the process. Some authors [100][101][102] point out that for traps like ours, alcoholic fermentation is a key process for attracting beetles, since in the wild fermented tree sap attracts them. It is possible that at high altitudes there are more such saproxylic beetle species that are attracted to our baits than in the near-surface layer of the forest.
There is no well-developed herbaceous cover in the studied forest, what can cause the decrease in the number of anthophilic beetle species at a height of 1.5 m. A closed grove of trees and a good undergrowth do not allow sunlight to reach the surface of the earth. That is why herbaceous plants do not develop well. On the other hand, many flowering shrubs grow in the undergrowth, where anthophiles find their food. It is also possible that anthophilic species are lured into traps at adjacent heights (3.5 and 7.5 m). The significant similarity of species diversity at these heights, calculated by the Jaccard similarity index, can also prove it.
The vertical profile of the air temperature in the forest canopy depends on the time of day, season, crown shape and species of the main tree species [103,104]. For example, the increased openness of the canopy noticeably changes the undergrowth and grassy layers. This gradient contributes to the vertical distribution of arthropod species in the forest canopy. Consequently, the availability of resources, the richness of microhabitat and abiotic conditions can be considered as critical factors affecting the number of arboreal arthropods. Thus, the insect species composition of the upper parts of the crowns should differ from the lower layers of the forest [4,11,[105][106][107]. In our study, it turned out that the main differences in biodiversity were obtained between heights of 1.5 and 12 m. At the same time, the differences between the heights of 3.5 and 7.5 m were minimal.
The microclimate of the upper tiers of the forest differs to a certain extent from the microclimate of the undergrowth. In temperate forests, it has special parameters of temperature, humidity, light, and interspecific interactions [108][109][110]. For example, the surface of the leaves of trees can be much warmer than the air temperature in the upper part of the crown, because they intercept a large amount of incoming radiation [111][112][113]. Among other things, the complex three-dimensional structure of tree crowns provides an ecological space for reducing insect predation [114][115][116].
On the other hand, the microclimate in the undergrowth is also different from the crown. For example, the temperature and humidity changes are not so significant. This buffer effect is present in all forests at different latitudes and is relatively independent of tree species [117][118][119]. The contribution of the understory to the functioning of temperate forests is significant but varies depending on the ecosystem function and ecological context, and, more importantly, the characteristics of the understory [120].
In the surface layer of temperate forests, the average values of temperature and humidity depend on the elements of the structure of the stand. The diameter of the tree, the base area and the variety of sizes affect the amount of scattered light [121]. The shrub layer, the species composition of tree species, dead wood, fallen trees, the remains of stumps are especially important. They can lead to an increase in the species diversity of insects of this tier under certain conditions, such as an increase in the temperature of the nearsurface layer [87,[122][123][124][125][126]. In addition, there are species that use microhabitats under the canopy to find partners or prey. Therefore, vertical migrations of individual species occur permanently or temporarily from one tier of the forest to another, depending on the season, the stage of the life cycle and even the time of day [114,127,128]. This means that we still have little information about the functioning of arthropod communities living in various tiers of deciduous forests of the temperate zone. To improve the understanding of such interactions, it is necessary to use different methods of data collection, increase the number of research areas, and expand the taxonomic composition of the studied communities.

Conclusions
Ninety-two species were identified as a result of studies at different heights of deciduous forests in the temperate zone of the European part of Russia. The families Nitidulidae (15 species), Cerambycidae (14 species), Elateridae (seven species), Curculionidae (seven species) and Scarabaeidae (seven species) had the greatest species diversity. Seven species had the maximum number in the traps, and 15 Coleoptera species were common to all the studied heights. The greatest species diversity was obtained at an altitude of 1.5 m, the smallest-at an altitude of 7.5 m. The highest abundance values were obtained at altitudes of 1.5 and 3.5 m. The minimum number of specimens was caught at a height of 12 m. The largest differences in the Jaccard similarity index were obtained between samples from a height of 1.5 and 12 m. The Shannon's diversity index was higher near the ground than in the tree crowns (at heights of 7.5 and 12 m), and the Simpson's Diversity index had the opposite trend. Glischrochilus hortensis and to a lesser extent Cychramus luteus preferred to live in the lowest layers of deciduous forest (1.5 m). Cryptarcha strigata was mainly found at altitudes of 3.5 m and 7.5 m with relatively high numbers. The abundance and occurrence of Protaetia marmorata were higher in the uppermost layers of the crowns. Quedius dilatatus also preferred the highest parts of the crown, also occurring at an altitude of 7.5 m. The number of saproxylic beetle species was practically the same at altitudes of 3.5-12 m, while their relative number decreased in the surface layer. The relative number of anthophilic beetle species was also lower at low height. We think that in order to manage forests to increase species diversity, it is necessary to try not only to increase the amount of dead wood for saproxyl species. Of great importance is the improvement of the herbaceous cover for anthophilic insects.

Data Availability Statement:
The data presented in this study are available in Table A1.

Acknowledgments:
The authors are grateful to A.A. Khapugin (Saransk, Tyumen) for his help in conducting a statistical analysis of the obtained results.

Conflicts of Interest:
The authors declare no conflict of interest.