Introduction

Soil micromycetes critically affect ecosystems by participating in the degradation of soil organic matter and nutrient cycling. They also play an important role in soil trophic chains as a basic food source for fungivorous organisms as mites, springtails, and earthworms (^{Dighton, 2003}). Micromycetes are mainly distributed in the superficial soil horizons and the rhizosphere zone, representing a high percentage of the total microbial biomass in soil (^{Swift et al., 1994}). Because microfungi synthesize different kind of enzymes, along with macromycetes, bacteria, actinomycetes, and soil mesoand macrofauna, they participate to the decomposition of plant and animal remains, thereby contributing toward the elimination of organic wastes generated by ecosystems and nutrient recycling. During decomposition, micromycetes also liberate several minerals in soluble form, such as inorganic ions that can be more easily absorbed by plant roots, therefore supporting the fertility of ecosystems (^{Dighton, 2003}).

Despite the ecological importance of soil saprobic micromycetes, there are very few studies addressing their diversity in tropical and semitropical forests. Tropical montane cloud forests (TMCF) exhibit a high diversity of plant species and communities, as a result of high humidity and mild temperatures that create favorable conditions for the coexistence of temperate and Neotropical flora (^{Rzedowski, 1996}). Impressively, even though TMCF only covers 0.8 % of Mexico, it contains 10 to 12 % of all plant species recorded in Mexico, including 2,500 endemic plant species. Therefore, TMCF is considered the most diverse type of vegetation per unit area (^{Rzedowski, 1996}). TMCF has an enormous value not only because of its high biodiversity and associated uses, but also because of the important ecosystem services that it provides, such as carbon capture, water catchment, water purification, erosion control, and sediment control in rivers.

These forests also provide regulating services, preventing or mitigating natural disasters such as landslides and flooding that affect large areas in Mexican lowlands (^{Bruijnzeel and Hamilton, 2001}). Unfortunately, TMCF is the most threatened ecosystem in the world, with more than half of its area that has been transformed to other land uses. In Veracruz, Mexico, TMCF has mostly been modified or disappeared because of human settlements and the establishment of agricultural crops. Human activities and disturbances, such as habitat destruction, urban development, industrialization, agriculture, contamination, and pesticide use, can potentially affect the fungal diversity of these forests (^{Tsui et al., 1998}). Given this context, the aim of the present study was to analyze the diversity and community structure of cultured saprobic filamentous soil fungi in four TMCF fragments with distinct landscape quality.

Materials and methods

Results

Species richness

In the four TMCF sites, 233 species from 51 genera were recorded, 95 sterile strains and 11 Xilarial type strains. The largest and the fewest number of species were found at the MASCOTA site (MS) and SANTUARIO site (SS), representing 45 and 25 % of the total number of species recorded respectively (Table 2). According to the Jackknife 1 and bootstrap estimators, the species inventory at the landscape level represented 52-64 % of the expected morphospecies. The estimators confirmed that 57 species (Jackknife 1) to 133 species (bootstrap) are missing from the inventory (Table 2). The accumulation curves for the morphospecies did not stabilize for any of the TMCF fragments (Figure 1). With respect to species richness per site, the MS and SS had the highest (107 spp.) and the lowest (60 spp.) species richness respectively. Statistical analyses showed significant differences (F= 4.58, p= 0.008) in species richness among sites (Table 3).

Table 2. Observed and estimate species of saprobic filamentous microfungi in four tropical montane cloud forest (TMCF) fragments of central Veracruz, Mexico 

TMCF	Observed species	Estimate species	Estimate species	Inventory
Study sites		Jack 1	Boostrap	%
MIRADOR	99	173	125	79%
SANTUARIO	60	105	78	77%
MASCOTA	107	179	137	78%
CAÑADAS	93	160	120	77%
LANDSCAPE	233	370	294	80%

Figure 1. Accumulation curves of observed species from the assembly of saprobic fungi in four fragments of tropical montane cloud forest in Veracruz state, Mexico. 

Table 3. Abundance* (number of colonies isolated to 250 soil particle), Richness **, diversity ** (Shannon Index H´) and equitativity** of saprobic filamentous microfungi in four tropical montane cloud forest fragments of central Veracruz, México. s.d (Standar deviation) 

Forest	Abundance		Richness		Diversity (h’)		Equitativity
fragment	Total	Average±s.d	Total	Average±s.d	Total	Average±s.d	Total	Average±s.d
MIRADOR	184	18.43 ± 8.87 ab	99	13.2 ± 5.86 a	4.2	2.37 ± 0.49 a	0.93	0.96± 0.02 a
SANTUARIO	108	10.8 ± 3.93 b	60	7.3 ± 2.3 b	3.6	1.78 ± 0.44 b	0.90	0.91± 0.13 a
MASCOTA	217	21.9 ± 7.26 a	107	14.8 ± 4.51 a	4.3	2.52 ± 0.27 a	0.95	0.95± 0.03 a
CAÑADAS	155	15.5 ± 9.66 ab	93	11.5 ± 5.52 b	4.2	2.20 ± 0.48 b	0.94	0.96± 0.04 a

*Average of ten replicates in two sampling.

** Average of ten replicates in two sampling.

† Shared superscripts (a, b) by study sites indicate no signifiicant differences between them (p≤ 0.05).

Abundance and dominance patterns

The slopes of the abundance-dominance curves of the MIRADOR (MIS) and CAÑADAS sites (CS) are similar and less pronounced than those of the SS and MS (Figure 2). No species maintained the same position in the hierarchical order of dominance across the four forest fragments. Talaromyces flavus was abundant at the SS, whereas Myrmecridium schulzeri was abundant at the MS. Trichoderma atroviride and a sterile strain (53) were abundant at the SS, MS, and CS. Pestalotiopsis maculans was abundant at the MIS and moderately abundant at the CS. The majority of species were rare (68 %). Forty-seven percent were singletons (single individuals), and 20 % were doubletons (two individuals) (Figure 2). The total abundance of colonies varied from 108 to 217. Significant differences in total abundance were found among sites (F= 3.67, p= 0.02), with the highest abundance corresponding with the MS and the lowest with the SS (t= 0.014) (Table 3).

Figure 2. Whittaker plots (Rank/abundance plots) illustrating the abundance of colonies of the saprobic fiilamentous microfungi detected in the study sites. 

Relationship of abundance, species richness, and species diversity with landscape quality and soil physico-chemical characteristics

In the simple regression analyses, a significant relationship was only found between the diversity of saprotrophic filamentous fungi and the landscape quality index, indicating that a greater diversity of morphospecies was found as the landscape quality increased (p= 0.018) (Figure 3). Some interesting tendencies were observed between soil pH and morphospecies diversity (p= 0.14), abundance (p= 0.16), and richness (p= 0.14), even though the results were not significant (Figure 3).

Figure 3. Relationship between diversity index. A: With index landscape quality. B: With species richness. C: With abundance. D. With the soil pH of forest fragments. 

Discussion

Fungi are important components of the biodiversity of tropical forests. The colonization and decomposition of soil organic matter are performed by a wide range of microorganisms, yet fungi and bacteria play a particularly important role in nutrient recycling. The overview of studies on soil fungal diversity largely depends on the techniques used to detect fungal communities. The soil particle filtration technique is widely used because it favors the isolation of fungi whose hyphae are associated with soil particles and, therefore, are active in soil (^{Bills et al., 2004}). Soil washing generally eliminates the spores of rapid-growth fungi and allows slow growing species to be expressed. However, fungi are nutritionally diverse, so there is not a single cultivation medium appropriate for all fungal species. In addition, many isolated fungi may require natural substrates to form fruiting bodies, complicating taxonomic identification by traditional methods (^{Bills and Polishook, 1994}). Considering these limitations, molecular techniques would be of great assistance in identifying isolated fungal strains. Thus, a detailed morphological description was made of each of the 104 sterile strains, and their abundance was quantified in each survey. In this way, all the isolates were accounted for enabling calculation of the evaluated parameters.

Most of the studies on saprobic soil micromycetes in primary ecosystems have been carried out in temperate regions, as coniferous forests in Canada (^{De Bellis et al., 2007}), the United States (^{Gochenaur, 1978}), and China (^{Cheng et al., 2015}). Fewer studies have been carried out in Africa (^{Persiani et al., 1998}), Malaysia (^{Varghese, 1972}; ^{Bettuci and Roquebert, 1995}), Brazil (^{Pfenning, 1997}), and India (^{Panda et al., 2010}; ^{Saikia and Joshi, 2012}; ^{Chandini and Rajeshwari, 2017}).

The present study demonstrated the high richness of saprotrophic soil fungi in TMCF in Mexico. A total of 233 species were identified from 666 isolates. Most studies on saprobic soil micromycetes in tropical forests have reported a smaller number of taxa, except for ^{Persiani et al. (1998)}, who reported 519 species in a forest along the Ivory Coast. Previous studies recorded 38 species in Malaysia (^{Varghese, 1972}); in India 63 species (^{Saikia and Joshi, 2012}), 81 species (^{Chandini and Rajeshwari, 2017}) and 118 species (^{Panda et al., 2010}) and 80 species in a tropical forest of Brazil (Pfenning, 1996). In recent years, DNA sequence-based methods have been used to survey soil fungal diversity of various terrestrial ecosystems, including forests; ^{He et al. (2017)} recorded 7501 OTU, 43 % were saprobic fungi. ^{Shi et al (2014)} detected 7630 OTUs, saprobic fungi representing 80 % of these. ^{Bueé et al. (2009)} recorded an average of 830 fungal operational taxonomic units (OTUs) per site in temperate forest.

Soil is an important habitat because it shelters a wide diversity of species and provides a wide variety of important environmental services. At the landscape level, the representativeness of the inventory was 52 to 64 % according to the richness estimators (Jackknife 1 and bootstrap). For each site, the representativeness of the inventories ranged from 42 to 79 %. Overall, the inventories of all forested sites and each forested site included more than 50 % of the estimated species. We consider that the sampling effort allowed to detect an overview of the composition of the soil fungal community. Molecular methods offer the prospect of a more thorough documentation of diversity. However, while numerous recent studies (^{Buée et al., 2009}, ^{Jumpponen and Jones 2009}) have utilized high‐throughput sequencing to examine various aspects of fungal biodiversity, these studies have failed to saturate a sampling (rarefaction) curve.

In this study the community of soil fungal species in the TMCF fragments presented a Raukier pattern: few species were abundant, and many species were uncommon or had low abundances. A group of species common to cloud forests were detected in all fragments: M. schulzeri, P. maculans, T. atroviride, V. nigra, T. cremeum, F. solani, C. terrestris, A. fischeri, and P. olsoni. However, these species were not always the most abundant. The most abundant species were T. flavus at the SS, M. schulzeri at the MS, and P. maculans at the MIS. Talaromyces flavus has been frequently observed in North America, Europe, and Australia (^{Beuchat, 1986}). Its ascospores can reportedly survive high temperatures. Pestalotiopsis maculans is a common and abundant habitant in the leaf litter of some TMCF trees in Mexico (^{Heredia et al., 2000}) and Costa Rica (^{Bills and Polishook, 1994}). The presence of this latter microfungi at all the studied sites shows that it can successfully grow from the leaf litter through the first soil horizons. Myrmecridium schulzeri, a lignicolous ascomycete that represents the imperfect state of the genus Ophiostoma, is associated with the decomposing wood of different TMCF tree species.

Different studies suggest that fungal diversity is greater in tropical regions than temperate regions (^{Suryanarayanan et al., 2009}). In tropical forests, employing traditional survey methods there have been reported diversity index (H´) values of 1.8 (^{Grantina, 2011}), 2.4-2.5 (^{Saravanakumar and Kaviyarasan, 2010}), 3.4 (^{Devi et al., 2012}), 3.1-3.5 (^{Satish et al., 2007}), 3.7 (^{Banakar et al., 2012}), and 4.1 (^{Panda et al., 2007}). ^{Devi and Dkhar (2014)} reported H´ values of 2.7 for fungi in an Indian forest with similar vegetation and altitudinal characteristics as cloud forest. Comparatively, the fungal diversity index in the present study ranged from 3.6 to 4.3, confirming the high species diversity of the studied TMCF.

Of the 233 species detected, the MS had the highest number (107 spp.), followed by MIS (99 spp.), CS (93 spp.) and, finally, SS (60 spp.). With respect to beta diversity, the study sites presented a high richness, diversity, and equitativity of filamentous saprotrophic fungal morphospecies at the regional level. In addition, high morphospecies complementarity and high morphospecies turnover among the fragments was found. This may be related with the degree of alteration of the sites reflected by the amount of open areas and the landscape quality. In particular, the SS had the lowest diversity. ^{López-Barrera and Landgrave (2008)} assigned this latter site a low value on the landscape quality index, while the rest of the fragments were considered high quality. The landscape quality index considered a 2-km radius around each point and contemplated open areas and areas without vegetation cover in addition to incipient, intermediate, and dense shade, which may be interpreted to correspond with certain types of vegetation.

The qualitative and quantitative differences in fungal populations between fragments indicate that vegetation and proximity to urban areas influence the micro-fungal populations of the soil. Disturbances to soil and vegetation could be a determining factor of the lower fungal count at the disturbed forest site (SS). ^{Liu et al. (2016)} reported that soil microbial diversity decreased with increasing disturbance. ^{Carney and Matson (2006)} showed evidence of the human alteration of soil microbial communities via the modification of plant community composition and diversity. On the other hand, high plant diversity also has a strong impact on the soil microbial community, mainly because of increased plant production (^{Liu et al., 2016}).

Environmental factors that influence plant and animal distribution have been more widely studied than those that control the patterns of distribution and community composition of soil microorganisms (^{Rousk et al., 2010}). Regarding the microbial communities, ^{McCulley and Burke (2004)} found that plant species composition, soil type, seasonal variability, temperature, and availability of organic substances are primarily factors that controlled them. Meanwhile, saprophytic fungi are sensitive to soil conditions, pH, texture, moisture, and oxygen level (^{Setala and McLean, 2004}).

Contrary to what we expected, in this study we didn’t found any significant correlation (p>0.05) between the physico-chemical soil parameters and the diversity, richness, and abundance of saprobic fungi. However, some interesting tendencies were observed for soil pH, which influenced the abundance, richness, and diversity of saprobic fungal species (70 % related), although not significantly (p= 0.14, p= 0.16, p= 0.14 respectively). Specifically, abundance, diversity, and richness were higher in the sites with a pH of around 4-5 (CS, MIS, and MS) and lower in the SS, with a pH fluctuating around 6. The impact of pH on soil microbial communities has been widely described in previous studies. Generally, studies on fungal biomass have shown that lower pH tends to favor fungal development (^{Rousk and Nadkarni, 2009}). In a boreal forest, ^{Högberg et al. (2007)} observed that fungi strongly decreased as the range of soil pH increased. ^{Pennanen et al. (1998)} also found that fungal growth increased at pH values of lower than 4 in forest soil, and similar findings were reported by ^{Banakar et al. (2012)}. The results described herein estimate the composition and diversity of the soil fungal community in four TMCF sites. The findings clearly reveal that TMCF soil in the central region of Veracruz, Mexico, exhibits a high richness, diversity, and equitativity of saprobic filamentous fungal species, as well as a high turnover rate and high number of exclusive species. Greater knowledge of the ecology of soil fungal communities can enable the development of strategies to conserve native saprobic species or to use them for bioprospecting industrially important enzymes or biomolecules.