Exploring the Relationships between Macrofungi Diversity and Major Environmental Factors in Wunvfeng National Forest Park in Northeast China

In this paper, we analyze the macrofungi communities of five forest types in Wunvfeng National Forest Park (Jilin, China) by collecting fruiting bodies from 2019–2021. Each forest type had three repeats and covered the main habitats of macrofungi. In addition, we evaluate selected environmental variables and macrofungi communities to relate species composition to potential environmental factors. We collected 1235 specimens belonging to 283 species, 116 genera, and 62 families. We found that Amanitaceae, Boletaceae, Russulaceae, and Tricholomataceae were the most diverse family; further, Amanita, Cortinarius, Lactarius, Russula, and Tricholoma were the dominant genera in the area. The macrofungi diversity showed increasing trends from Pinus koraiensis Siebold et Zuccarini forests to Quercus mongolica Fischer ex Ledebour forests. The cumulative species richness was as follows: Q. mongolica forest A > broadleaf mixed forest B > Q. mongolica, P. koraiensis mix forest D (Q. mongolica was the dominant species) > Q. mongolica and P. koraiensis mix forest C (P. koraiensis was the dominant species) > P. koraiensis forest (E). Ectomycorrhizal fungi were the dominant functional group; they were mainly in forest type A and were influenced by soil moisture content and Q. mongolica content (p < 0.05). The wood-rotting fungus showed richer species diversity than other forest types in broadleaf forests A and B. Overall, we concluded that most fungal communities preferred forest types with a relatively high Q. mongolica content. Therefore, the deliberate protection of Q. mongolica forests proves to be a better strategy for maintaining fungal diversity in Wunvfeng National Forest Park.


Introduction
Fungal communities are essential for forest ecosystems and have many functions [1,2]. Ectomycorrhizal fungi (EM) participate in the soil nutrient cycle of forest ecosystems and promote the host plant's absorption of nutrients, such as nitrogen, phosphorus, and water, thereby maintaining the above-ground primary productivity of the forest ecosystem [3]. Saprotrophic fungi can degrade wood components (i.e., lignin, cellulose, and hemicellulose [4]), and they are considered essential wood-decay-promoting organisms. These functions indicate a crucial role in maintaining the forest ecosystem's stability [5].
Many biotic and abiotic factors can affect the diversity and composition of fungal communities [6,7]. The composition of EM is strongly influenced by the soil's nitrogen content [8,9], pH [10], temperature and moisture [11,12], the species composition of the host trees [13,14], and by the seasons [15]. Fungal communities living on the wood are closely dependent upon environmental factors, such as the amount, diameter, and stage of wood decomposition [16], wood chemistry [17], age [18], and tree species [19]. Factors influencing terricolous saprotrophic communities include litter quantity and pH [20], soil  Pinus. koraiensis sites (E): Single species; 40-year-old P. koraiensis forests planted in 1980, close to nature managed, no fallen trees removed. There are zero Q. mongolica with a coverage rate of 0%.
We collected samples 20-25 times per month from June to October 2019-2021. We randomly acquired macrofungi from each plot. We photographed the specimens in the field using a Canon EOS 800D digital camera and recorded fresh morphological characteristics and ecological characteristics ( Figure 2). We selected the context or stipe tissue (1-2 g) of the same specimen when it was fresh and stored it in a sealed bag with silica gel for DNA extraction; we dried them in an oven (45-50 • C) and placed them in specimen boxes. We then took a morphotype of each specimen to the laboratory and used it for morphological species identification.

Soil Sampling, Analysis and Environmental Data Collection
We collected soil samples four times per month during July-September 2020. After cleaning and removing plant material and debris from the surface, we collected individual soil samples from the center and 4 corners in 15 plots using an auger (5 cm radius, 5 cm depth). We mixed the soil samples from the same plots well and placed them in sealed bags. After removing impurities, we enclosed the fresh samples (20 g) from each clod in an aluminum box. We dried the samples to a constant weight in an oven at 105 • C to measure water content (SWC); we used natural air-dried composite samples (200 g) for each plot to analyze for pH, organic matter (SOM), available phosphorus (P), effective nitrogen (N), and available potassium (K) using the method described by Xing et al. [39]. Finally, we averaged the data from three plots of the five fore for subsequent analysis.

Soil Sampling, Analysis and Environmental Data Collection
We collected soil samples four times per month during July-September 2020. A cleaning and removing plant material and debris from the surface, we collected individ soil samples from the center and 4 corners in 15 plots using an auger (5 cm radius, 5 depth). We mixed the soil samples from the same plots well and placed them in sea bags. After removing impurities, we enclosed the fresh samples (20 g) from each clod an aluminum box. We dried the samples to a constant weight in an oven at 105 °C measure water content (SWC); we used natural air-dried composite samples (200 g) each plot to analyze for pH, organic matter (SOM), available phosphorus (P), effec nitrogen (N), and available potassium (K) using the method described by Xing et al. [ Finally, we averaged the data from three plots of the five fore for subsequent analysis We obtained the temperature and relative humidity of the air and soil temperat from July-September 2020 from meteorological monitoring sites in the forest park. T tested results for the soil are included in Table 1.  We obtained the temperature and relative humidity of the air and soil temperature from July-September 2020 from meteorological monitoring sites in the forest park. The tested results for the soil are included in Table 1.

Species Identification
We identified the macrofungi using morphological observations methods. We used molecular methods for the species that were morphologically difficult to identify. We measured different microscopic structures of taxonomic importance (e.g., spores, basidia, cystidia) [40]. We examined the morphological features of the fruiting bodies using appropriate monographs, including by Li et al. [41] and Liu et al. [42], to identify each macrofungi specimens. The specimens are currently housed in the Herbarium of Mycology of Jilin Agricultural University (HMJAU), Changchun, China.
Molecular identification involved sequencing the internal transcribed spacer (ITS). For this, we extracted the DNA of the macrofungi using a NuClean Plant Genomic DNA Kit (Cowin Biosciences, Taizhou, China), following the manufacturer's instructions. We conducted final elutions in a total volume of 50 µL. We showed a polymerase chain reaction (PCR) with the primer pairs ITS-1F and ITS-4 [43]; finally, we sequenced the PCR products using the Sanger method. We conducted the PCR in 25 µL reactions consisting of 2 µL genomic DNA, 0.5 µL Taq, and one µL upstream and downstream primers, respectively. We used 14.5 µL ddH 2 O, five µL 5 × PCR buffer, and 1 µL dNTP in the PCR reactions that we ran under the following conditions: 95 • C for 3 min, followed by 35 cycles of 94 • C for 40 s, 55 • C for 45 s, 72 • C for 1.5 min, and a final extension step at 72 • C for 6 min before storage at 4 • C We purified the PCR products and sequenced them at Sangon Biotech Co., Ltd. (Shanghai, China). We performed molecular identification via BLAST comparisons. Species with >98% sequence similarity were also identified with morphological characteristics. GenBank accession numbers obtained are provided in Appendix B.

Statistical Analysis
We used three alpha diversity indices to analyze the community composition of the macrofungi. The Menhinick richness index (R) reflected the species richness of the community. The Shannon index (D) reflected the diversity of the community species. Pielou's evenness index (E) reflected the distribution of the number of individuals in each species. The diversity index formulae were as follows: where P i is the proportion of species i to the total number of individuals of all species in the plot; ln is the natural logarithm; S is the total number of species in the plot; and N is the total number of individuals observed in the plot. We analyzed the relationships between ectomycorrhizal fungi communities and selected variables using the canonical correlation analysis (CCA) from Canoco 5.0 [45]. We first used detrended correspondence analysis (DCA) to determine the appropriate model for direct gradient analysis. The results indicated that a unimodal model (gradient lengths > 3 standard units) would best fit our study data; we utilized CCA. Furthermore, we tested explanatory variables using the Monte Carlo permutation test provided by Canoco 5.0 software (with 999 randomizations). The species data matrix for the CCA analysis was based on the presence-absence data of ectomycorrhizal fungi species in each forest type (three-year accumulation of the five forest types).
We used Origin 9.0 software to construct species stacked histograms at the genera level [46] to compare community compositions of the macrofungi species in the five forests and provide the relative proportion of macrofungi species richness (data include the number of species at the genera level in each forest type). Additionally, we generated pie charts, Venn diagrams, and species accumulation curves using Hiplot (available online: https://hiplot.com.cn/basic/venn (accessed on 20 October 2021)). The pie chart data were derived from the number of macrofungi types. The Venn diagram data included the species in each forest type. The accumulation curve data consisted of the cumulative number of species per collection.

Species Richness
We collected 1235 specimens from 5 forest types, 940 (76.11%) of which we identified at the species level, and we classified these into 283 fungal species. We identified 244 species based on morphology and 39 species using morphology and molecular methods (Appendix B). The unidentified sporocarps were not part of our further analysis. We classified the macrofungi species into 116 genera, 62 families, 18 orders, and 2 phyla. Basidiomycota was the dominant phylum, divided into 12 orders, 50 families, 102 genera, and 265 species. Ascomycota was divided into 6 orders, 12 families, 14 genera, and 18 species. The Russulaceae was the most diverse family with 36 different species, followed by Tricholomataceae (21 species), Boletaceae (19 species), and Amanitaceae (16 species). Together, these accounted for 32.51% of the total collected species. The most abundant genera were Amanita, Cortinarius, Lactarius, and Russula. The Agaricales were the most prevalent order in the five forest types (59.36%). In terms of the trophic groups, most of the species were ectomycorrhizal fungi (47%), followed by wood-decaying fungi (20.14%) and soil saprotrophs (18.37%).

Macrofungal Types
The most significant genera of Agarics accounted for 69.26% of the identified species, followed by boletes, larger ascomytetes, and polyporoid fungi, accounting for 9.89%, 6.36%, and 6.01% of the identified species, respectively. In contrast, hydnaceous and cantharelloid fungi were less abundant, accounting for 1.06% and 0.71%, respectively (see Figure 3). For more detailed information, see Appendix B. For images of some species, see Appendix A.

Analysis of Dominant Families and Genera
Among the identified species, there were 9 dominant families (number of specie 10 species) of macrofungi ( Table 2). The Russulaceae was the most diverse family. In add tion, 53 families contained less than 10 species, accounting for 85.48% of the families an 48.06% of the identified species (Appendix B).

Analysis of Dominant Families and Genera
Among the identified species, there were 9 dominant families (number of species ≥10 species) of macrofungi ( Table 2). The Russulaceae was the most diverse family. In addition, 53 families contained less than 10 species, accounting for 85.48% of the families and 48.06% of the identified species (Appendix B). Among the identified species, there were 15 dominant genera (number of species ≥ 5 species) of macrofungi ( Table 3). The Amanita, Lactarius, and Russula were the most diverse genera. In addition, 34 genera contained 2-4 species, accounting for 29.31% of the genera and 29.68% of the identified species; 67 of the genera contained only 1 species, accounting for 57.76% of the genera and 23.67% of the identified species (Appendix B).

Cumulative Abundance of Macrofungi in Five Forest Types
The accumulation curves for the species identified in the five forests show a steady increase with more samplings ( Figure 5). We reached saturation of macrofungi richness after 150 surveys. The species accumulation curves of A (Q. mongolica forest) and B (broad-leaved forest) showed relatively steep upward slopes and produced higher macrofungi abundance values than the other forests. Nevertheless, forest type A (Q. mongolica forest) obtained the highest macrofungi diversity values.

Cumulative Abundance of Macrofungi in Five Forest Types
The accumulation curves for the species identified in the five forests show a increase with more samplings ( Figure 5). We reached saturation of macrofungi ri after 150 surveys. The species accumulation curves of A (Q. mongolica forest) and B ( leaved forest) showed relatively steep upward slopes and produced higher macr Two genera (Clitocybe and Tricholoma) were shared in five forest types, but no species were shared. The unique species (found only in 1 forest) increased from E < C < D < B < A and consisted of 13, 22, 31, 50, and 104 fungal species, respectively ( Figure 6). Forest type A shared 27 species with B, 7 species with C, 11 species with D, and 2 species with E. Forest type B shared six species with C, eleven species with D, and three species with E. Forest type C shared three species with D and three species with E. The Gymnopus densilamellatus The species richness increased from E < C < D < B < A (Table 4). Broad-leaved forests A and B, with the highest richness indices of 7.4023 and 5.4832, respectively, accounted for 80.57% of the total species. Among them, 142 species were found in forest type A, accounting for 50.18% of the total species. This indicates that the broadleaf forest was the main habitat of macrofungi in the area, especially regarding the Q. mongolica forest. Mixed forests C and D, with richness indices of 2.8296 and 4.6509, contained 84 species, accounting for 29.68% of the total species. However, we found that the species abundance was higher in forest type D (49 species) than in forest type C (35 species), indicating that macrofungal species are associated with Q. mongolica. In P. koraiensis forest E, with the smallest species richness index of 2.286, we only found 18 species, accounting for 6.36% of the total species. This indicates that P. koraiensis forests can only provide habitats for a few fungal species.  Two genera (Clitocybe and Tricholoma) were shared in five forest types, but no species were shared. The unique species (found only in 1 forest) increased from E < C < D < B < A and consisted of 13, 22, 31, 50, and 104 fungal species, respectively ( Figure 6). Forest type A shared 27 species with B, 7 species with C, 11 species with D, and 2 species with E. Forest type B shared six species with C, eleven species with D, and three species with E. Forest type C shared three species with D and three species with E. The  The species richness increased from E < C < D < B < A (Table 4). Broad-leaved forests A and B, with the highest richness indices of 7.4023 and 5.4832, respectively, accounted for 80.57% of the total species. Among them, 142 species were found in forest type A, accounting for 50.18% of the total species. This indicates that the broadleaf forest was the

CCA Analysis of Macrofungal Communities and Selected Environmental Factors
We performed a canonical correspondence analysis (CCA) for the 130 ectomycorrhizal fungi (EM) species recorded in the 5 forest types. The variables included Q. mongolica content, effective soil nitrogen, soil available phosphorus, soil available potassium, soil organic matter, soil pH, soil temperature, air temperature, soil water content, effective soil nitrogen, and air relative humidity. The CCA results show that all samples were roughly separated into five groups according to their corresponding locations. Eigenvalue axis 1 (0.8963) is higher than axis 2 (0.7955), with a cumulative contribution of 28.8% and 25.57%, respectively. Of all the variables, the Q. mongolica content and soil moisture content were the most significant factors influencing the EM fungi. Many EM fungi (e.g., Amanita, Cortinarius, Lactarius) positively correlate with Q. mongolica and soil water (Figure 7).

Discussion
This study is the first systematic survey of macrofungal diversity in Wunvfeng National Forest Park, Ji'an, China. We divided the forests into five main types: Q. mongolica forests (A), mixed broad-leaved forests (B), artificial P. koraiensis forests (E), and mixed forests (C, D). This enabled us to analyze the composition of macrofungi according to the relative content change of Q. mongolica in different forest types. The results show differences in species richness and diversity among forest types with different relative contents of Q. mongolica. The species richness increases with the relative content of Q. mongolica. Forest types with a high cover of Q. mongolica may provide a stable environment for the growth of macrofungi [47]. More importantly, Quercus is the main host plant of EM fungi [48], such as Lactarius [49,50], Amanita [51], Russula [52], and Cortinarius [53]. Our results reveal that the EM fungi are mainly distributed in the Q. mongolica forest. Most EM fungi had a significant positive correlation with Q. mongolica content (Figure 7), especially Amanita, Cortinarius, and Lactarius. In addition, we found that 11 species are shared in forest types A and D (e.g., Amanita ibotengutake T. Oda, C. Tanaka

Discussion
This study is the first systematic survey of macrofungal diversity in Wunvfeng National Forest Park, Ji'an, China. We divided the forests into five main types: Q. mongolica forests (A), mixed broad-leaved forests (B), artificial P. koraiensis forests (E), and mixed forests (C, D). This enabled us to analyze the composition of macrofungi according to the relative content change of Q. mongolica in different forest types. The results show differences in species richness and diversity among forest types with different relative contents of Q. mongolica. The species richness increases with the relative content of Q. mongolica. Forest types with a high cover of Q. mongolica may provide a stable environment for the growth of macrofungi [47]. More importantly, Quercus is the main host plant of EM fungi [48], such as Lactarius [49,50], Amanita [51], Russula [52], and Cortinarius [53]. Our results reveal that the EM fungi are mainly distributed in the Q. mongolica forest. Most EM fungi had a significant positive correlation with Q. mongolica content (Figure 7), especially Amanita, Cortinarius, and Lactarius. In addition, we found that 11 species are shared in forest types A and D (e.g., Amanita ibotengutake T. Oda, C. Tanaka  However, they are not found in the P. koraiensis forest. These species may be associated with Q. mongolica (Figure 7, Ama7, Ama8, Ama9, Ama16), because the macrofungal communities change accordingly with the forest's succession [54]. Therefore, these macrofungi shared in forest types D and A are likely to be in the early stages that develop from spore banks present in the soil [55].
The species richness of P. koraiensis forests is the lowest in our study. We only found 18 species. Theoretically, species richness may be similar between Q. mongolica and P. koraiensis forests because the EM fungi in temperate forests are significantly associated with Quercus and Pinus [56]. However, we only observed ten species of EM fungi in P. koraiensis forest (E) (e.g., Hydnellum aurantiacum (Batsch) P. Karst., Hydnellum peckii Banker, and Tricholoma matsutake (S. Ito & S. Imai) Singer). Our results differ from those of Gao [57], who found more EM fungi in P. koraiensis forests (aged < 150 years). On the one hand, EM fungi may need more time to form a stable symbiosis with the host plants [58]; on the other hand, the exotic trees have difficulty developing long-lasting symbiotic relationships with local EM fungi [59].
In our study, wood-dwelling fungi (57 species) are also a critical taxon that is mainly distributed in the Q. mongolica and mixed broad-leaved forests and grows on larger diameter Q. mongolica fallen wood (e.g., Armillaria gallica Marxm. & Romagn. and Neolentinus cyathiformis (Schaeff.) Della Magg. & Trassin). The wood-dwelling macrofungi may be related to forest type, as they tend to favor specific forest types under similar climatic conditions. In general, these combinations are determined by fungi closely related to the dominant tree, mainly because their enzymes have adapted to wood with different chemical and physical properties [60]. Another reason may be that large logs that provide a larger surface area have a greater chance of being colonized by fungal spores and mycelium than small logs. Species that produce large fruiting bodies also require more space [61]. Furthermore, we only found a few fallen trees in the Q. mongolica and mixed broadleaved forests; we found no fallen trees in the P. koraiensis forest, which is another factor that might affect fungal assemblage. The amount of deadwood also affects the macrofungal assemblage, which previous authors highlighted as the most crucial microhabitat in the forest [62,63]. The diversity of woody macrofungi strongly depends on the presence and amount of deadwood [64,65].
Saprophytic soil fungi (52 species) rely mainly on the decomposition of soil organic matter for nutrients, and they tend to prefer specific forest types under similar climatic conditions. Generally, the deciduous leaves of broad-leaved trees are more conducive to soil organic matter accumulation than coniferous forests [66]. The forest types with high soil organic matter have more opportunities to be colonized by fungal spores and mycelium [67]. Moreover, we only found thicker deciduous leaves in the Q. mongolica and mixed broad-leaved forests, affecting the grass rot fungal assemblage because litter saprotroph fungi strongly depend on deciduous leaves' presence and volume [68].
The composition of fungi is also influenced and constrained by soil environmental variables [69,70]. These include soil moisture [71], soil pH [72], soil nutrients [73] and soil total C [74]. We analyzed the correlation between the main functional groups (ectomycorrhizal fungi) and selected environmental factors. The results showed that most ectomycorrhizal fungi are closely related to soil water content, especially Amanita, Cortinarius, Lactarius, and Russula (Figure 7). This result suggests that specific fungal communities respond to soil parameters differently [75][76][77]. Previous studies have shown that soil moisture is one factor that regulates the composition of the ectomycorrhizal fungi community [78][79][80][81]. Hydraulic lift contributes to maintaining EM fungi roots' integrity and viability of extraradical hyphae [82]; further, EM fungi take up water and organic and inorganic nutrients from the soil via the extraradical hyphae and translocate these to colonized tree roots, receiving carbohydrates from the host in return [83]. This may be an important reason why most ectomycorrhizal fungi prefer forest types with relatively high soil water content.
This study with three years of species data is a small contribution that allows us to understand the distribution of fungal species in forest types with the different covers of Q. mongolica. The Wunvfeng National Forest Park has a strict protection policy for animals and plants, including soil protection. Thus, our soil data (with permission) are from July to September 2020 only. Nevertheless, our results illuminate the potential links between community composition and environmental factors because the July-September 2020 species data include almost all our species.

Conclusions
The Q. mongolica forests we analyzed are rich in macrofungal species. Although our data are based on only three years of sampling, we conclude that, as Q. mongolica increases in the forest, the abundance and diversity of macrofungal taxa also increase. We also observed that most EM species favored forest types with high Q. mongolica content (e.g., Amanita, Cortinarius, Lactarius, and Russula), indicating that some EM fungal communities are closely associated with Q. mongolica. We call for further studies to support this claim. In addition, we have only found Tricholoma matsutake (S. Ito & S. Imai) Singer in P. koraiensis forests, which is classified as an endangered species and considered an ectomycorrhizal fungus that is closely associated with Pinus trees. Therefore, according to our research, maintaining P. koraiensis forests is beneficial for conserving endangered species. However, deliberate conservation of Q. mongolica forests would be more useful for maintaining the diversity of macrofungal communities. Whether P. koraiensis affects other fungal species will need to be monitored over 3-5 years.