Assembly of heterotrophic communities during spontaneous succession in quarries: invertebrates model groups and macromycetes

Quarrying has a crucial impact on the environment, but it could enhance species diversity. Mining sites represent important refuges for countless species disappearing from homogenous landscapes. Our study focused on assemblages of heterotrophic communities such as moths (Lepidoptera), carabid beetles (Coleoptera: Carabidae), spiders (Araneae), and macromycetes (fungi: Basidiomycota, Ascomycota) in an active part of kaolin quarries and their immediate surroundings in the Pilsen region, Czech Republic. We compared differences between mined and unmined sites, sites with spontaneous succession and sites with replanted pine trees. In total, we recorded 178 moth, 63 spider, 27 carabid beetle, and 81 macromycetes species, including 21 Red‐listed species. The moths, carabid beetles, and macromycetes tended to inhabit unmined sites; on the contrary, open habitat spiders preferred open sites with replanted pine trees. Based on the life history traits analyses, moth species feeding on forbs and grasses prevail at the active part of kaolin quarries, where higher plant diversity was detected. Large body carabid beetles such as Carabus spp. favored unmined sites, as well as macromycetes with long‐lived fruit bodies. Open habitat and xerophilous spiders inhabited the replanted sites by pine trees where the sparse vegetation was obvious. Our results indicated that groups with radically different life histories such as moths, carabids, and macromycetes may react to mining remarkably similarly, although spiders, despite sharing predatory habits with the majority of carabids, reacted differently.


Introduction
Rapid economic development, climate change, industrial agriculture, and general land use change are collectively impacting species diversity (Cardoso et al. 2020;Raven & Wagner 2021).Therefore, ecologists strive to predict the consequences of biodiversity loss, or to understand factors that structure biotic communities (Lepš et al. 2018).An increasingly used tool for relating biotic assemblages' compositions to their environments are the functional traits of the constituent species.
Functional traits are well-defined, measurable properties of organisms that influence organismal performance (McGill et al. 2006).The application of functional traits in ecology is not new (Raunkiaer 1934;Hrubý 1964) and many faunistic and ecological syntheses since the mid-20th century onwards considered some traits of the targeted taxa (Juliano 1983;Hodgson 1993;Beneš & Kuras 1998).Contrary to a nomenclatural view, which tends to produce highly contingent rules and special cases (Lawton 1999), conclusions based on studying functional traits tend to be independent of the taxonomic composition of communities up to the global scale (Statzner et al. 2001).Rapid progression of functional traits' analyses can be illustrated in studies of plants, and vertebrates such as birds (Meffert & Dziock 2013;Saunders & Cuthbert 2014) or fishes (Villéger et al. 2009), and invertebrates including hoverflies (Schweiger et al. 2007), bees (McCravy et al. 2019), ground beetles (Kędzior et al. 2020), and spiders (Wolz et al. 2020).Examples of themes include searching for predictors of extinctions' propensity (Matilla et al. 2006), analyses of colonization (Summerville 2015), and responses of species to fragmentation and homogenization of landscapes (Prugh et al. 2008;Franzén et al. 2012;Bartoňov a et al. 2016).Traits are used to answer questions related to species composition changes during succession (Kröber et al. 2012), along environmental gradients (Pekin et al. 2011;Duivenvoorden & Cuello 2012;Fric et al. 2020), or for understanding declines of pollinators (McCravy et al. 2019).
Localities and habitats affected by mineral excavations are of high interest, both from the general scientific and the conservation-minded points of view.They are undergoing primary succession, that is, colonization by species from their surroundings, and de novo assembly of biotic communities (Pek ar 1997; Gobbi et al. 2017;Marténez-S anchez et al. 2020), rendering them ideal for studying community assemblage rules.The exposed bedrocks, extreme microclimates, rugged relief, and rudimentary soil layer at such sites often radically differ from surrounding landscapes, forming islets of regionally rare conditions offering colonization and establishment potential for regionally declining or rare organisms depending on such conditions.In early stages of post-excavation succession, such substrates are colonized by early successional specialists, which have declined in general rural landscapes across the world (Bobrek 2020;Kettermann & Poniatowski 2022;Münsch & Fartmann 2022), but even later succession stages can have conservation value and aesthetic appeal (Baczy nska et al. 2018).Based on these observations, many authors warned against technical approaches to postexcavation sites rehabilitation, which usually consist of leveling-off terrain and planting a few woody species, and advocating more balanced approaches to restoration (Prach et al. 2011;Moradi et al. 2018;Kettermann & Fartmann 2023).
In this article, we present a multi-taxa study of four groups of heterotrophic organisms (moths, spiders, carabid beetles, and macromycetes) recorded in active kaolin quarries and their immediate surroundings.Contrary to many of the earlier studies in quarry environments, which often targeted nutrient-rich mineral conditions within biotically rich areas (Hendrychov a & Bogusch 2016;Baranov a et al. 2021), kaolin is an extremely nutrient-poor substrate, and the region targeted is notable for its rather simplified landscape structure and low species richness, that is, poor regional species pool (Beneš et al. 2002;Zahradnický et al. 2004).Still, we previously showed that even sites affected by kaolin extraction may represent regionally important replacements for grasslands or wetlands (Walter et al. 2023).
The targeted groups represent herbivores (moths), predators (carabids and spiders), and saprotrophs, the latter little represented in postindustrial habitats community studies.We aimed to answer two questions.The first, more applied, is how the four groups responded to active mining and two different post-mining successional regimes: spontaneous succession and replanting the mined sites by trees.The second, more general, is which functional traits of the constituent species responded to mining/succession, and whether the four ecologically disparate groups show some detectable similarities.

Methods
The study was conducted near the town of Horní Bříza, Pilsen Region, Czech Republic (Fig. 1).The town is an important center for mining kaolin, a clay mineral for porcelain, cement, and ceramics production, consisting predominately of Al 2 Si 2 O 5 (OH) 4 (Langhammer & Kaplick a 2005;Yahaya et al. 2017).The surroundings of the still operating quarry and adjacent factory consist mainly of plantations of Norway spruce (Picea abies [L.] H. Karst.) and Scots pine (Pinus sylvestris L.), with an admixture of deciduous trees such as Silver birch (Betula pendula Roth) and Sessile oak (Quercus petraea [Matt.]Liebl.).
We used a fully factorial design, with four sites within the quarry where the kaolin is still extracted, and four outside of the active part of the quarry, but impacted by the mining (typically, spoil heaps surrounding the quarry pit).The sites were 100 m apart (Fig. 1).Within these two treatments, two sites were revegetated by spontaneous vegetation succession, the other two were planted with pine (P.sylvestris) trees (≈1.5 m height, wide spacing).In August 2022, we recorded a standard phytosociological relevé, that is, a list of all vascular plant species present in a 10 Â 10 m area with their percentual covers, at each site (Braun-Blanquet 1964;Chytrý 2000;Kaplan et al. 2019).We also described the sites using the total % cover of vegetation layers E 3 (mature trees), E 2 (shrubs and tree saplings), E 1 (herbs), and E 0 (lichens and mosses) (Table S1).
We sampled the moths, carabid beetles, spiders, and macromycetes in 2022 using methods suited for the four groups.
The moths were sampled by using portable light traps powered by a 12 V dry battery and equipped with 8 W UV diodes (390-405 nm) placed on both sides of the aluminum bar.We used one trap per site, the distances from other traps were always greater than 100 m (Truxa & Fiedler 2012).The sampling started at dusk, the traps were emptied the next morning.We placed the light traps eight times from April 29 to October 24, 2023.All moths were identified to species level (Macek et al. 2007(Macek et al. , 2009(Macek et al. , 2012)), using genital preparation, if necessary.
The carabids and spiders were sampled using pitfall traps consisting of plastic cups (upper diameter 9 cm, depth 15 cm) filled to two-thirds by 8% acetic acid.At each site, two such traps were placed 3 m apart (Walter et al. 2023).We controlled the pitfall traps nine times, from April 29 to October 24, 2023.The material was identified to species level using standard monographs (Heimer & Nentwig 1991;Roberts 1995;Hu ˚rka 1996;Ku ˚rka et al. 2015).
The macromycetes' (Ascomycota and Basidiomycota) fruiting bodies were recorded within a 10 m radius surrounding the center pitfall traps (Hor ak et al. 2018).Only two visits were conducted (early spring and late autumn), because of the extreme aridity during the summer, typical for the sandy and sunny conditions.Each visit lasted 20 minutes, the fruiting bodies were identified during the visits, difficult species were processed in the laboratory (Bernicchia & Gorj on 2010;Knudsen & Vesterholt 2018).

Functional Traits
We selected functional traits that are well-defined and potentially related to succession/colonization capacities, and hence, implicitly, to the effects of mining processes: For moths (Potocký et al. 2018), these were wing span as a proxy for mobility and resource demands for growth; overwintering stage, because species overwintering in later stages utilize season more efficiently; adult flight period length, known to be short in specialists and long in generalists; voltinism, defining the temporal niche of a species; trophic range, related to niche breadth; habitat range, related to habitat requirements; host plant form, related to food availability and plant defense modes; host plant part consumed, also related to food availability; and humidity preference, as a description of habitat association (Table S1).
For carabids, the traits were body size; presence of wings, directly related to mobility; type of breeders, related to seasonality of adults; openness preference, related to the light requirements of the species; and humidity preferences (Hu ˚rka 1996) (Table S1).
For spiders, we used the same traits as for carabid beetles, except for the presence of wings, which we replaced with hunting strategy, which describes not only hunting sensu stricto, but also mobility (Heimer & Nentwig 1991;Buchar & Ru ˚žička 2002) (Table S1).
We selected the following functional traits for macromycetes: permanence of fruiting body, distinguishing short-lived (days to months) and long-lived (years), describing the persistence of the fruiting body and availability of spores; life strategy (parasitic, mycorrhizal, and saprotrophic), reflecting the available resources and symbiotic partners (Zanne et al. 2020) and often covarying with phylogenetic position similarly to the next trait; wood affinity (lignicolous/non-lignicolous); and spore color (light/dark), associated with environmental conditions so that dark spores are associated with the r-strategy of saprotrophic agarics (Halbwachs et al. 2015) (Tables S1).

Statistical Analyses
We analyzed the assemblages separately for each targeted group using ordination analyses in CANOCO 5 ( Šmilauer & Lepš 2014).Firstly, to characterize vegetation conditions of the sites, we subjected data from the vegetation relevés to principal component analysis (PCA), extracting four major gradients of variation, Veg1-Veg4.The sample scores of the four gradients were subsequently used as predictors describing the impact of vegetation composition on the four target groups (Table S2).
Next, we used PCAs of the life history traits (herein LHT) of the targeted group, to visualize their mutual positions and to extract four major gradients of target traits, LHT1-LHT4.Inputs were sets of traits characterizing each of the recorded species.We used the case scores of thus extracted PCA axes in further analyses, to evaluate the effects of focal predictors on the traits' composition of the entire samples (Table S2).
To test the effects of focal predictors, we used canonical correspondence analysis (CCA), which allows direct testing of relationships between the composition of species assemblages and environmental variables.The presence of active mining (yes/no) and succession (spontaneous/replanted) were categorical predictors.Other predictors were vegetation (Veg1-Veg4) and the percentual representation of E0-E3 vegetation layers.We used square root transformation of species abundances and down-weighting of the rare species.CCA significances were evaluated using 999 Monte Carlo permutations.
Statistically significant ( p ≤ 0.05) CCAs were further interpreted by the constituent species LHT.This was done by redundancy analysis, a linear multivariate method, of species scores of the significant CCA axes from the previous analyses against the LHT.We used forward selection from all the traits, and also the compositional traits variables LHT1-LHT4.

Describing the Vegetation of the Sites
The results of PCA analyses of vascular plant composition (eigenvalues of ordination axes 1-4: 0.3451, 0.2597, 0.2129, and 0.0995) indicated that the main Veg1 gradient distinguished sites planted by pine from those of spontaneous succession (Fig. S1a).The Veg2 gradient distinguished nutrient-poor (and species-poor) vegetation from nutrient-rich sites (Fig. S1a).The Veg3 gradient distinguished sunny and xeric open-sward sites from those with close-sward (Fig. S1b).Finally, the Veg4 gradient did not clearly differentiate sites, but did indicate a connection between disturbed vegetation and an early stage of successional processes (Fig. S1c).

Traits
For moths, the main MothsLHT1 gradient ([eigenvalue] = 0.176) distinguished large-bodied species with larvae feeding on foliage of apparent plants, having a long flight period, and preferring humid conditions from smaller species, often developing on flowers or seeds of unapparent plants growing in xeric conditions (Fig. S2a).The MothsLHT2 gradient (0.137) distinguished large-bodied moths with broad trophic range and few generations per year, overwintering in early stages, from smaller moths with narrow trophic range and multiple generations per year (Fig. S2a).The MothsLHT3 gradient (0.107) distinguished species occurring in a broad range of habitats, but having a narrow trophic niche, from those occurring in a narrow range of habitats, but having a broad trophic niche (Fig. S2b).The MothsLHT4 gradient (0.099) did not convey to a clear pattern.
For spiders, the SpidersLHT1 gradient (0.526) distinguished web-building species from actively hunting species (Fig. S4a).The SpidersLHT2 gradient (0.225) distinguished species of shaded and humid habitats from those of sunny and xeric habitats (Fig. S4a).The SpidersLHT3 gradient (0.144) ordered the species according to body size, with small species in negative and large species in positive values (Fig. S4b).The SpidersLHT4 gradient (0.105) partly distinguished species of shaded and xeric habitats from those of open and humid habitats but did not convey to a clear pattern.

Interpreting the Assemblages by Species Traits
Interpreting the significant results of constrained ordinations (Table S3) revealed that the differences in moths' assemblages between the mined and unmined parts of the kaolin quarry were due to differences in the host plant form.Moths associated with mined sites tended to develop on forbs and grasses, whereas moths associated with unmined sites preferred trees or nonvascular plants (Fig. 2A).
In the case of carabid beetles, large-bodied species of carabids prevailed at unmined sites (Fig. 2B).
Open habitat spiders prevailed at sunny sites with sparse vegetation, but also at sites replanted by pine trees (Fig. 2C).Moreover, the significant association with Veg3 (Fig. 2D) revealed a preference for large-bodied hunters of xeric habitats to sparsely vegetated sites (Table S3; Fig. 2E & 2F).
In the case of macromycetes, saprotrophic species prevailed at unmined sites (Fig. 2G), whereas parasitic and mycorrhizal species inclined toward active parts of kaolin quarries (Fig. 2H).Also, lignicolous species with long-lived fruiting bodies inhabited the sites with spontaneous succession (Table S3; Fig. 2I-K).

Discussion
Expectably, the active kaolin quarries and their surroundings hosted species-poorer assemblages than we recorded in closely situated disused kaolin quarries (295 moth, 147 spider, 55 carabid beetle, and 240 macromycetes species; Walter et al. 2022).Still, similarities with other post-industrial locations, including the relatively species-rich ones (limestone sites: Beneš et al. 2003;Tropek et al. 2010;black coal spoils: Tropek et al. 2012; garbage dumps: Twerd & Banaszak-Cibicka 2019) are evident.Perhaps most notable was the relatively high percentage of Red-listed species (6% of all species recorded), detected in all groups except carabids.The presence of these species agrees with other authors emphasizing the importance of post-industrial areas as secondary biodiversity refuges (Okrutniak et al. 2018;Ru ˚žičkov a & Hykel 2019;Macgregor et al. 2022).
The composition of moths', carabid beetles', and macromycetes' assemblages were affected by active mining in the quarries.On the other hand, spiders differed between sites with spontaneous succession and sites replanted by pine trees.

Moths
As herbivores in larval stage and nectar/sugary solutions feeders as adults, moths are obviously attracted to sites with well-developed vegetation cover.This was corroborated by the significant effect of the vascular plant gradient Veg2, which differentiated sites with species-poor vegetation from those with species-rich plant assemblages.This gradient tended to divide the sites with spontaneous vegetation succession from those replanted by pine trees.Other authors working in post-mining sites also reported more moth species from sites affected by spontaneous succession than from sites subject to technical reclamation (Tropek et al. 2010).Generally, higher vegetation diversity increases the probability of herbivores' presence (Huston 1979) and supports not only the herbivores, but also their predators (Miller et al. 2014).Analyses of LHT indicated that moths recorded from mined and unmined sites differed in their larval host plant forms.Species developing on forbs and grasses (e.g.Scotopteryx chenopodiata, Eupithecia simpliciata, Charanyca trigrammica, Mythimna impura, Xestia castanea) prevailed at the active part of the quarries, whereas unmined sites hosted species associated with woody plants (Calliteara pudibunda, Lymantria monacha, Eupithecia tantillaria) and even nonvascular plants diets (Eilema sororculum, Lithosia quadra, Laspeyria flexua).The host plant form is associated with development speed (voltinism, body size) and thus crucially affects herbivorous insects' life histories ( Čížek et al. 2006;WallisDeVries 2014;Potocký et al. 2018).Interestingly, Thorn et al. (2015) observed increased representation of lichens-developing moths in the decaying stage of mountain taiga forests.

Carabid Beetles
Mined and unmined sites differed in carabid beetles' assemblages so that mined sites hosted more species.These were mostly smallbodied predators (e.g.Olisthopus rotundatus, Poecilus cupreus, Pseudoophonus rufipes), known as pioneers in early succession (Okrutniak et al. 2018), able to colonize such habitats as short-cut lawns, or arable fields (Hu ˚rka 1996).The largebodied species, including those from the genus Carabus Linneaus, 1758, prevailed at unmined sites.These long-living species prefer shady microhabitats (Brooks et al. 2012), rich with feeding possibilities and shelters (Lee & Landis 2002).Here, they control the assemblages of many invertebrates and thus support nutrient cycling processes (Hunter et al. 2003).The most abundant species at such unmined sites was Calathus erratus, recorded in 307 individuals.This species inhabits sandy heaths dominated by Calluna vulgaris, abundantly present at unmined sites in the kaolin quarries' surroundings.

Spiders
Spiders' assemblages differed between sites with spontaneous succession and replanted sites.Moreover, the vascular plant gradient Veg3 distinguished spiders of sunny and xeric sites from spiders of close-sward shady sites.As carabids, spiders can act as pioneer species (Bråten et al. 2012).They colonize newly exposed ground immediately (Hodkinson et al. 2001), easily dispersed by ballooning (Coulson et al. 2003;Tropek & Konvička 2008;Blandenier 2009).The spiders' assemblages were structured by the constituent species' preference for habitat openness, but also by the composed LHT gradient, Spi-dersLHT2s.Counterintuitively, open habitat species prevailed at sites replanted by pine trees, owing to sparse sward structures at the time of our survey.This suggests that rather than by species' composition of vegetation, spiders' assemblages are affected by levels of sward openness and moisture conditions.While studying restored grasslands, DiCarlo and DeBano (2019) also observed that sites with natural vegetation may not host higher spiders' diversity than degraded or restored sites (similar results ;Perner & Malt 2003;Borchard et al. 2014).Moreover, we observed several Red-listed species inhabited the replanted sites, namely Attulus distinguendus, Coriarachne depressa, Drassyllus pumilus, Micaria silesiaca, and Sittisax saxicola, all dwelling in such habitats as steppes, rocks, or sandy habitat.These records again illustrate the role of industrial habitats as a secondary refuge for invertebrates (Hendrychov a et al. 2008;Tropek et al. 2012).

Macromycetes
Not surprisingly, sites with higher cover of vascular plants hosted more macromycetes, especially saprotrophic species.In general terms, macromycetes' diversity increases in undisturbed conditions (Hofmeister et al. 2015;Hor ak et al. 2019).Saprotrophic macromycetes prevailed at sites with higher presence of E2 and E3 layers, or other plant biomass (cf.Xu et al. 2022), whereas parasitic and mycorrhizal species inclined toward actively mined parts of the quarries.In the initial stages of succession, there is a scarcity of suitable nutrients (still undeveloped litter layer), which limits the saprophytes.Interconnections among fungi and plants are not yet fully established, and the profit from mycorrhiza may be lower both in terms of nutrients and the protection of plants' roots.These conditions seem to favor the parasites compared to the later successional stages (Shi et al. 2019;Lee et al. 2021).The early successional plants, however, need their mycorrhizal partners, causing a prevalence of mycorrhizal macromycetes.Still, the extreme conditions are tolerated by only a few macromycetes, resulting in their low species richness (Bigg 2000).Young plant individuals will also not have such a rich representation of endophytes, which protect them against parasites (Wang et al. 2023).Examples of macromycetes adapted to extremely nutrient-poor sites were Pisolithus arhizus and Thelephora terrestris, dominant at bare surfaces of the mined sites.
The long-lived lignicolous species, in contrast, preferred unmined sites.Such sites offered a more variable plant material, including woody plants, and therefore supported woodassociated fungi.Perennial fungal species need some stabile substrate for several years, which was nonexistent in mined sites (Berglund et al. 2011;Stokland et al. 2012).
Our results reveal that moths, carabids, and macromycetes with radically different life histories react to mining remarkably similarly, whereas spiders, despite sharing predatory habits with the majority of carabids, react differently.Moths, as herbivores, track the development of vegetation (Hawkins & Porter 2003;Řehounkov a et al. 2016), reaching higher species richness at unmined sites, but specialists of disturbed conditions display opposite preferences (Habel et al. 2019).Carabids, in line with general theories of natural succession (Gobbi et al. 2010), comprise both abundant, small-bodied colonists of early successional conditions, and later-successional species, the former preferring mined and the latter unmined sites.The majority of macromycetes, especially so the saprotrophic and long-living lignicolous species, preferred unmined sites, but a few parasitic specialists preferred the opposite.Regarding spiders, our sampling method mainly targeted epigeic species, undersampling epiphytic species.The epigeic species, however, are rapid colonizers, which respond to the structure and composition of vegetation, and thus to temperature, moisture, and other factors, rather than to mining per se.

Figure 1 .
Figure 1.Location (A) of the study sites of kaolin quarries near the town of Horní Bříza (Czech Republic, western Bohemia, Pilsen Region).The sites replanted with pine trees (B) and spontaneous vegetation processes (C) are documented.

Figure 2 .
Figure 2. Redundancy analyses ordination biplots showing relations of life history traits (LHT) of moth (A), carabid beetle (B), spider (C-F), and macromycete (G-K) assemblages and significant focal predictors in ordination space.Only significant traits are depicted.