The local populations of the fungus Schizophyllum commune Fr. as drivers of its biodiversity

Understanding the regularity of genetic diversity in fungi is a fundamental problem that is associated with the development of strategies for their preservation (Urbanelli et al. 2003). Longest genetic distances correspond to a subdivision of the Eastern and Western Hemisphere, within which the population’s structure exists. For fungi species with the sex cycle and a big population size, genetic variation is very high, and most of the variations are observed within the population. In the last decades, using molecular markers has allowed to establish geographical limitation of gene flow for some fungi (Garbelotto et al. 1993; Hibbett et al. 1995; Kauserud and Schumacher 2003). Density and genetic diversity significantly vary for some fungi. It suggests that their dissemination is not limited only by the ability to spread the spore mass. The idea of global fungi distribution only by spores is not correct, and human intervention plays a significant role in fungi spreading (Fry et al. 1992; Maurice et al. 2014).

Schizophyllum commune is a xylotrophic cosmopolitan fungus. In most cases, it settles on dead wood and can be found on weakened live wood (Raper 1988). Its genetic variability allows discriminating substructures in common, widespread populations (James et al. 1999). The widespread growth of S. commune enables researchers to collect the required number of samples on the local population level. The use of S. commune as a model object in population studies will make it possible to find rare genotypes and identify factors that contribute to the isolation process with possible further interpolation of the results to other species. Very frequently, changes that occur on the local level can explain the processes on the level of population.

According to literature data, the stands of Holosiivskyi National Nature Park (HNNP), Lysa Hora Regional Landscape Park (LHRLP) and Feofaniya forest parcel (Ffp) were continuous forests in the 17th century (Netsvetov and Prokopuk 2016). The settlement of surrounding areas started in the second half of the 19th century and continued till the beginning of the 20th century. In the 1960s–1970s, boundaries of HNNP and Ffp were separated by a part of Great Ring Road. LHRLP was separated from HNNP by the overpass and developed living area. Based on this, at least for the last 50–60 years, the described region has not been a continuous forest because of anthropogenic factors, and therefore is an important object for population studies. The unique conditions of partial geographical isolation over a long period of time make it possible to research the possible genetic differentiation of the model fungus S. commune.

The study’s goal was to establish genetic differentiation of local populations of the fungus S. commune in partially isolated areas with complex terrain.

The locations of S. commune basidiocarp collection are separated from each other. The distance between HNNP (1) and Ffp (3), that is, between the nearest samples from both stands, was 0.7 ± 0.05 km, and the distance between HNNP (1) and LHRLP (2) was 1.2 ± 0.05 km. The genetic diversity of fungi is illustrated by indices that are calculated individually for each artificial local population and the general population (Tab. 1–3). For the HNNP population, locus Eg2 was represented only by allele Eg2¹⁰⁰, and in LHRLP appeared allele Eg2⁹³ with the frequency 0.056 (Tab. 1).

For Ffp, in addition to allele Eg2¹⁰⁰, allele Eg2¹⁰² was observed with the frequency 0.385. For locus Amy2, four alleles were found; however, allele Amy2¹⁰⁷ was not detected in the HNNP population and allele Amy2¹¹⁰ in the Ffp population. Despite being rare, allele Amy2⁹⁵ was found in all researched locations. For locus Cat, we observed striking differences in allele frequencies depending on the location. Specifically, allele Cat¹⁰⁷ was found only in the fungi of Ffp. In addition, the frequency of Cat⁹⁷ was relatively high in the HNNP and LHRLP populations (0.188 and 0.222, respectively), but significantly decreased in Ffp (0.038). Locus Got showed a relatively similar profile of allele frequencies in different populations, except that allele Got^null was observed only in the HNNP. Taking into account only dominant alleles, we observed a significant decrease in the frequencies of locus Eg2 in the Ffp population (0.615), locus Amy2 in the LHRLP population (0.667) and locus Cat in the HNNP population (0.656). In general, the rarest alleles were Eg2⁹³ and Got^null with a frequency of 0.013.

The average number of alleles per locus was from 2.5 (HNNP) to 2.75 (LHRLP and Ffp) (Table 2). The effective number of alleles, Shannon’s diversity index and expected heterozygosity were not significantly oscillating in the studied populations. The difference was bigger for observed heterozygosity. For all local populations, a lack of heterozygotes was observed; it was lowest in the Ffp, which is supported by H_o and F indexes. Of note, in the LHRLP population of S. commune, the lack of heterozygotes is insignificant and by all indexes of genetic variability has the highest rankings (Tab. 2).

Determination of the genetic distance between studied populations was conducted using F-statistic (Tab. 3). We found that a lack of heterozygotes was observed on the population level for some loci, and it can be a result of the inbreeding process (F_is = 0.323). It is highly influenced by Eg2 and Cat loci. We previously reported similar results for Eg2 locus on the territory of Ukraine (Boiko 2018b). Also, the uneven distribution of alleles for this locus is a characteristic feature of not only large territories, but also of local populations.

Importantly, the same S. commune cultures showed a balanced level of heterozygotes at Amy2 and Got loci on the population level (F_is was −0.016 and 0.007, respectively), which excluded the significant influence of the inbreeding process. Therefore, the lack of heterozygotes at the mentioned loci can be a result of new and rare allozymes or a small population size.

In the aggregate sample of populations, a single basidiocarp revealed a lack of heterozygotes on 35.5%. The degree of differentiation of genes between experimental populations relative to the frequencies of alleles (F_st) was average and indicated that 91.4% of the entire genetic diversity can be found within each population. This is the evidence for a certain genetic separation among the studied populations. Among the polymorphic loci, the greatest contribution to the interpopulation component of variability was made by the locus Eg2 (F_st = 0.244). The data we obtained indicates a relatively high gene flow among local populations (N_m = 6.78), and a significant contribution to this index was made by the Got locus. Low genetic diversity and high gene flow are observed in populations that are supported by the spread of spores.

Many factors lead to a disturbance of equilibrium in nature (Maurice 2014). In the S. commune populations, for each enzyme system, we determined the actual distribution of genotypes to the expected by Hardy–Weinberg principle. For the loci Eg2 (Ffp) and Cat (HNNP and Ffp), the Hardy–Weinberg equation was violated, which indicates that they are influenced by factors that cause deviation of allele frequencies. The possible reasons are aforementioned.

Calculation of genetic distance according to Nei and cluster analysis (UPGMA algorithm) allowed us to determine the isolation of Ffp (3) from HNNP (1) and LHRLP (2). The geographical distance between Ffp (3) and HNNP (1) is almost twofold less than between HNNP (1) and LHRLP (2). PCoA, which shows the interconnection of samples by genetic material, determined isolation of the Ffp cultures (Fig. 2). Of note, this exact population of Ffp makes the greatest contribution to the diversity of the first main coordinate.

Figure 2

Ewens–Watterson test on neutrality for each locus showed that the frequency of alleles at all loci was selectively neutral and the F values were within 95% of the expected value (Tab. 4). Spatial structure is an important characteristic for a particular population and is a result of the interaction among many components. To determine the geographical isolation of the studied populations, the Mantel test was conducted. As a result, it can be reasonably argued that the null hypothesis was true (R² = 0.0016); there was no interconnection between the genetic component and the geographical coordinates of the studied samples.

Considering that the distance between populations is relatively small, we conducted the spatial structure analysis, which establishes the correlation of the genotypes’ diffusion in space (Fig. 3).

Figure 3

Except for the locus Eg, no correlations were detected. A positive probabilistic correlation was observed within 0.049–0.056 interval and negative probabilistic correlation within 0.028–0.035 interval. The absence of a general correlogram trend may indicate an accidental peak. Considering that the positive correlation found mainly concerns cultures from HNNP and LHRLP, and that HNNP is represented only by the allele Eg2¹⁰⁰, we think this is exactly the reason for this maximum. A similar trend was observed with a probabilistic negative correlation for basidiocarps from HNNP and Ffp. Based on this, it can be argued that no genotype structure was observed at the studied locations.

In our research, we considered three artificial local populations. The critical question arising is: what will we observe if this is a single population? Applying UPGMA analysis to allele frequencies data, experimental cultures were divided into three clusters (Fig. 4). The first cluster, the most distant, was solely formed by fungi from the Ffp; the second cluster was formed by cultures from the HNNP (67%), LHRLP (22%) and Ffp (11%); and fungi forming the third cluster were from the HNNP (43%) and the other two locations (28.5% each). Of course, due to migration and drift of genes, there was no clear separation of genotypes by growth area, especially at small distances, but certain isolation of cultures was observed.

Figure 4

Cross-breeding system and gene flow are the main factors that determine the genetic structure of fungal populations. Considering that S. commune has a tetrapolar cross-breeding system that greatly prevents inbreeding (James et al. 1999), the main contributor to the population’s local genetic profile is the process of gene migration and drift. The genetic drift effect is greatest in small populations and lowest in largest populations.

For fungi, the basis for these processes is the dissemination of spores. Light spores of fungi can travel considerable distances and are at high altitudes, which indicates a high power of migration processes. The distance of spore dissemination is a controversial issue and depends on many components. According to some authors, up to 95% of spores are concentrated at a distance of up to 1 m from the fruit body (Reddi 1976; Galante et al. 2011). Other groups argue that the percentage of spores in the air as well as the covered distances are much higher (Norros et al. 2012; Hallenberg and Küffer 2001; Viljanen-Rollinson et al. 2007). Importantly, there is a common view that wind is the main, essential factor for disseminating spores (Kuparinen et al. 2007; Dam 2013). According to Andrew et al., temperatures and precipitation are positively correlated with fungal richness (Andrew et al. 2019).

In addition, in our opinion, relief is a factor that affects the direction and force of airflow and forms a hydration regime necessary for the normal growth and formation of basidiocarps. We tried to correlate these two factors to our local populations (Fig. 5).

Figure 5

Relief topography suggests that the HNNP (1) forms a hill between two other local populations and the wind direction from April to October (which is the optimal period for the formation of basidiocarps) is more favourable to disseminate spores in the direction LHRLP (2) → HNNP (1) → Ffp (3), rather than in the opposite direction. Taking into account that the altitude of the HNNP is the highest, effective ‘capturing’ of spores takes place from the LHRLP, and at the same time, the probability of their dissemination to the Ffp decreases (difference in altitude 28 m).

On comparing our data with similar results obtained for the steppe zone of Ukraine, it was noticed that the level of gene differentiation was smaller there and the gene flow conceded, although distances between populations were much higher (Boiko 2015). This may support our assumption of the effect of terrain topography on the genotype of the local population. It is possible that for such topography, ravine plays the role of ‘traps’ for rare alleles, which, due to drift, can lead to their disappearance or increase their fraction.

Our results show the similarity of genetic variability of S. commune in different local populations. The heterozygote deficiency of some loci that was discovered may have resulted from new forms of allozymes that have not become widespread or due to small population sizes. In the aggregate sample of populations, a single basidiocarp revealed a lack of heterozygotes on 35.5%. The degree of differentiation of genes between local studied populations was moderate due to the high flow of genes. The absence of spatial structuration of genotypes is established, and the Mantel test showed a lack of interconnection between the genetic component and the geographical coordinates of the samples. The direction of wind and terrain are the factors that influence the genetic structure of local populations.