Yeasts of the soil – obscure but precious

Abstract Pioneering studies performed in the nineteenth century demonstrated that yeasts are present in below‐ground sources. Soils were regarded more as a reservoir for yeasts that reside in habitats above it. Later studies showed that yeast communities in soils are taxonomically diverse and different from those above‐ground. Soil yeasts possess extraordinary adaptations that allow them to survive in a wide range of environmental conditions. A few species are promising sources of yeast oils and have been used in agriculture as potential antagonists of soil‐borne plant pathogens or as plant growth promoters. Yeasts have been studied mainly in managed soils such as vineyards, orchards and agricultural fields, and to a lesser extent under forests and grasslands. Our knowledge of soil yeasts is further biased towards temperate and boreal forests, whereas data from Africa, the Americas and Asia are scarce. Although soil yeast communities are often species‐poor in a single sample, they are more diverse on the biotope level. Soil yeasts display pronounced endemism along with a surprisingly high proportion of currently unidentified species. However, like other soil inhabitants, yeasts are threatened by habitat alterations owing to anthropogenic activities such as agriculture, deforestation and urbanization. In view of the rapid decline of many natural habitats, the study of soil yeasts in undisturbed or low‐managed biotopes is extremely valuable. The purpose of this review is to encourage researchers, both biologists and soil scientists, to include soil yeasts in future studies.


| HISTORY
In the years following the first observation of yeasts in 1680 by Antonie van Leeuwenhoek, these small eukaryotic organisms were considered to be associated mainly with alcoholic fermentation of beer and wine.
However, the question of the origin of yeasts found in fermented products soon became the starting point for research on yeasts outside man-made environments. Louis Pasteur was one of the first to attempt to answer this question. In 1875, he began a series of investigations to find out whether yeasts could be isolated from the skin of the grapes used in making wine and whether they were present only at one time of the year (reviewed in Guilliermond, 1920). His experiments indicated that, during autumn, yeasts existed on practically all parts of the vine and disappeared during the winter. Emil Hansen investigated the life cycle of the yeast Saccharomyces apiculatus (Hanseniaspora uvarum) that was widespread on fruits (Hansen, 1880). He thought that yeasts were distributed by air currents, insects and rainfall to other fruits as well as to the soil on which fruit trees grow (reviewed in Guilliermond, 1920;Starkey & Henrici, 1927;Bouthilet, 1951). Hansen was also able to observe living yeasts in soil under fruit trees. Using both cultivation and artificial inoculation experiments, he demonstrated that yeasts can survive in soils throughout the year. Yeasts have been frequently observed in the surface layers but rarely in the deeper soil layers ( Figure 1a). In the following years yeasts were found in soils of vineyards and orchards down to a depth of 12-13 and 20-30 cm by the pioneering microbiologists Amedeo Berlese and Hermann Müller-Thurgau, respectively (reviewed in Starkey & Henrici, 1927). Hansen believed that yeasts hibernating in soil during the winter were carried by the wind on dust particles and inoculated fruits above-ground (reviewed in Guilliermond, 1920;Starkey & Henrici, 1927), while Berlese suggested that insects served as vectors of yeast cells (Figure 1b) (reviewed by Brysch-Herzberg, 2004). Later, Hansen investigated the presence of yeasts in soils in the Copenhagen area and also found them outside orchards and gardens under beech, fir, pine and oak trees, although only in about 30% of samples (reviewed in Guilliermond, 1920).
Other pioneering studies demonstrated that yeasts are present in soils (reviewed in Guilliermond, 1920;Starkey & Henrici, 1927;Bouthilet, 1951). However, they were not recognized as indigenous soil organisms and the ability of yeasts to propagate in soils was repeatedly questioned (discussed by Danielson & Jurgensen, 1973;Phaff, Miller, & Mrak, 1966;Phaff & Starmer, 1987). Yeasts were often equated with fermenting ascomycetes that colonize above-ground sugar-rich substrates, such as ripe fruits and flowers. Starkey and Henrici (1927) and Ciferri (1928) analysed yeast numbers and noticed that bacteria and filamentous fungi outnumber yeasts in most soils.
The low quantity of yeasts was interpreted as evidence for the minor importance of this group of fungi in soil functioning (e.g. Starkey & Henrici, 1927 and in later reviews by Phaff & Starmer, 1987;Starmer & Lachance, 2011). Starkey and Henrici (1927) did not find any correlation between the occurrence of yeasts and the type of soil, vegetation or season of the year. In contrast, Pumpyanskaya (1938) showed that yeast quantities depend on physical and chemical soil properties (reviewed in Babjeva & Golovleva, 1963). It is important to mention that approaches that were used at that time to study yeasts strongly favoured the isolation of fast-growing fermenting ascomycetes. Margret di Menna (1957) revised isolation techniques used for soil yeasts. She showed that suitable culture media, cultivation conditions and sample pre-treatments resulted in higher yeast colony counts. Culture media, supplements and incubation techniques have been changing with the evolving knowledge of the taxonomic composition of soil yeast communities and the ecology of the dominant species (e.g. Babjeva, 1969;Boundy-Mills, 2006;di Menna, 1959;Miller & Webb, 1954). The use of nitrogen-free media (e.g. More thorough and systematic sampling of soil worldwide is needed outside of temperate and boreal zones, generally from Asia, Africa and the American continent, and from unmanaged soils, particularly in the tropics and the subtropics.

Future research on soil yeasts
FIGURE 1 Schematic representation of ecology and dispersal routes of yeasts in soils. Yeasts from ripe fruits are carried to the soil (a), hibernate during the winter and inoculate fruits above-ground (b). Indigenous soil yeasts multiply in the topsoil, and their number decreases in deeper soil layers (c). The presence of some yeasts is related to the deposition of plant and animal residues on the soil; these transient species are quickly eliminated (outcompeted or preyed on) in the soil (d). Some yeasts are associated with soil plant toots (e) and invertebrates (f) Brown's Azotobacter agar) facilitated the isolation of slow-growing Lipomyces species from soils. Reliable isolation techniques and a growing number of studies showed that yeast numbers in soils may exceed those on decaying plant material and convinced researchers that yeasts do live, and not only reside, in soils. A number of soil-borne yeasts were isolated and described during the beginning of the twentieth century, e.g. Apiotrichum dulcitum, A. porosum, Cyberlindnera saturnus (originally Willia saturnus), Lipomyces starkeyi, Nadsonia starkeyi-henricii (originally Schizoblastosporion starkeyi-henricii), Schwanniomyces polymorphus (originally Pichia polymorpha) and Vanrija humicola (originally Torula humicola). As outlined by Phaff and Starmer (1987), the repeated isolation of the same yeasts from soils and their absence in other sources above-ground were employed as another argument to demonstrate the soil origin of several yeast species.
The diversity of soils and vegetation types encouraged scientists to study yeasts across different climates, biotopes and soil types. Di Menna (1960Menna ( , 1965a was probably the first scientist to study soil yeasts through a series of biotopes characterized by different types of vegetation, land management and soil properties. Jensen (1963) analysed yeasts in Danish beech forests in different seasons. Capriotti (1967) sampled soils across a wide geographical range in the USA.
Starting in 1956, Johannes van der Walt described more than 30 yeast species from soils, many of which were isolated in South Africa (Smith & Groenewald, 2012). Later, in cooperation with Maudy Smith, he intensified studies of the yeast family Lipomycetaceae from an evolutionary perspective. Inna Babjeva and co-workers systematically studied the distribution of yeasts in major soil types in the USSR. Babjeva and Golovleva (1963) provided the first comprehensive review of soil yeasts in zonal, intrazonal and azonal soils (see also Babjeva & Chernov, 1995). Helen Vishniac analysed soils collected over a period of nearly 30 years along a latitudinal gradient in western North America covering polar to tropical climates (Vishniac, 2006a). At about the same time, Ivan Chernov (2005) performed a similar study. He collated data derived from a total of 114 localities and ca. 7000 samples previously analysed by Babjeva and co-workers in order to study the influence of geographic latitude and natural zones on yeast community parameters.
Recently, Alfred Botha wrote two reviews solely dedicated to soil yeasts (Botha, 2006(Botha, , 2011. These sources and the two recent book chapters by Vadkertiová et al. (2017) and Yurkov (2017) are recommended.
Although soil communities are frequently regarded as species poor, low species richness in a single plot (alpha diversity) contrasts with the larger number of yeasts that can be isolated from a forest or a region.
The distribution of yeasts in soils is often fragmented with a few species only shared between sampling sites. For example, Vishniac (2006a) reported nearly 40% of yeasts to be restricted to a single locality.

Likewise, temperate forests in Germany (three regions) had only
Apiotrichum dulcitum in common (Yurkov et al., 2012). Three Mediterranean xerophyl forests sampled in a single locality had eight out of 57 species shared between all three plots (Yurkov, Röhl, et al., 2016). The dissimilarity in species composition between sites results in high diversity values on the regional level (e.g. Yurkov, Kemler, & Begerow, 2011;Yurkov, Röhl, et al., 2016). Recent studies showed that fairly well analysed soils yield a large number of as yet undescribed yeasts. The proportion of potentially novel taxa was estimated to exceed 30% in temperate beech and Mediterranean xerophyll forests (Yurkov, Röhl, et al., 2016;Yurkov, Wehde, et al., 2016). The same holds true for a few other temperate forests (Mašínová et al., 2017;Mestre et al., 2011;Takashima et al., 2012) and is likely to be true for tropical biotopes.
Not every yeast species isolated from soil is an indigenous soil inhabitant but may originate from other sources other than soils (e.g. Phaff et al., 1978;Phaff & Starmer, 1987 Yurkov, 2017). Observation of fermenting ascomycetous yeasts frequently found on fruit surfaces, such as Hanseniaspora species, suggests that they reside in soils (e.g. Phaff & Starmer, 1987). However, the ability to ferment sugars does not predict well the transient habit of a yeast species, since several autochthonous soil yeasts possess this trait, e.g.
Ascomycetous yeasts are generally more frequent and abundant in agricultural soils, orchards and grasslands (Vadkertiová et al., 2017;Yurkov et al., 2012). Ascomycetous yeasts of the genus Lipomyces are typical soil yeasts, some of which (L. starkeyi and L. tetrasporus) are distributed worldwide (Kurtzman & Smith, 2011). The genus Myxozyma represents asexual forms of Lipomyces. Interestingly, several Myxozyma and Lipomyces species have been isolated from insect-associated habitats such as frass, decaying cactus tissues and tree fluxes (Kurtzman & Smith, 2011).
Trichosporon is another prominent yeast genus reported from soils.
This genus has been recently reclassified (Liu et al., 2015) and common soil-related species have been transferred in the genera Apiotrichum, Cutaneotrichosporon and Tausonia (Table 1) Sniegowski, Dombrowski, & Fingerman, 2002;Sylvester et al., 2015), this yeast should be viewed as a transient soil species propagating on above-ground substrates (e.g. fruits, bark, leaves, tree fluxes) and residing in soils (Sampaio & Gonçalves, 2017). In most cases the isolation of these yeasts have been made from soils using a sugar-rich enrichment culturing medium, with or without 7-8% (v/v) ethanol (Kowallik & Greig, 2016;Sampaio & Gonçalves, 2008;Sniegowski et al., 2002;Sylvester et al., 2015). Such selective conditions allow the isolation of Saccharomyces but not most of indigenous soil yeasts.
Reports of these yeasts outside vineyard and orchard soils are extremely rare and most of them have been made from broadleaf (oak, beech, southern beech) forest litter and the underlying topsoil (Kowallik & Greig, 2016;Mestre et al., 2011;Sampaio & Gonçalves, 2008;Sylvester et al., 2015).

| YEAST PHENOTYPES
The presence of fermenting yeasts below-ground was traditionally viewed as the result of contamination from above-ground sources.
However, species of the genera Barnettozyma (formerly Williopsis and Zygowilliopsis), Cyberlindnera (formerly Pichia and Williopsis), Kazachstania (formerly Arxula and Saccharomyces) and allied Candida species are prominent in grassland and agricultural soils (reviewed in Vadkertiová et al., 2017). These yeasts display a typical copiotrophic lifestyle; they grow fast, consuming simple sugars but not complex substrates, and are capable of anaerobic fermentation. The importance of fermentation in soil has not been investigated. However, the ability to utilize sugars in the absence of oxygen (e.g. when soil pores are filled with water) is potentially useful for soil yeasts.
The other adaptation frequently reported to be advantageous for soil microorganisms is the ability to produce extracellular polysaccharide capsules (EPS). The formation of capsules is a known mechanism that enables microbes to sequester and concentrate nutrients while growing in low-nutrient environments or sustain low water activity and desiccation (Aksenov, Babjeva, & Golubev, 1973;di Menna, 1959;Raspor & Zupan, 2006). Semi-arid soils, low in nutrients and moisture, are mostly populated by encapsulated anamorphic basidiomycetous yeasts (Spencer & Spencer, 1997;Vishniac, 2006a).
The ability of some of these soil yeasts to survive in sandy soils owing to the production of EPS has been demonstrated with the soil yeast Naganishia albida (formerly Cryptococcus albidus, Vishniac, 1995
Soil yeasts inhabit and interact with the plant rhizosphere ( Figure 1f; Botha, 2011;Mestre et al., 2011). They were also studied as potential antagonists of soil-borne plant pathogens (reviewed in El-Tarabily & Sivasithamparam, 2006;Botha, 2011). Several yeast cultures originating from the rhizosphere were reported to reduce rates of plant diseases (reviewed in Botha, 2011;Fu et al., 2016).
Different species of yeasts also showed different mechanisms of antagonism towards growth of fungal root pathogens (e.g. Botha, 2011;El-Tarabily & Sivasithamparam, 2006). However, only a few of the tested potential biocontrol species are true soil inhabitants.
Barnettozyma californica and Galactomyces candidum isolated from the rhizosphere of Drosera spatulata exhibited significant antagonistic effects against Glomerella cingulata in culture (Fu et al., 2016). Likewise, the soil yeast Vanrija albida showed the best negative effect on the growth of plant pathogens Verticillium dahliae and Pythium aphanidermatum (Mestre et al., 2016).
Pathogenic yeasts Cryptococcus neoformans, Coccidioides immitis, several clinically relevant Candida and species formerly classified in the genus Trichosporon can be found in soils (e.g. Miceli, Díaz, & Lee, 2011). However, the proportion of yeasts from rural soils growing at elevated temperatures (usually above 30 or at 37°C) is low (e.g. di Menna, 1955;Mok et al., 1984;Sylvester et al., 2015). Clinically relevant yeasts are not common or abundant in soils and they are probably introduced with animal feces and waste.

| DISTRIBUTION OF SOIL YEASTS
The recent review by Botha (2011) describes the diversity of interactions of soil yeasts with the environment, including both abiotic and biotic factors. Soil yeasts respond to changes in abiotic factors, including soil organic matter content, pH, conductivity, temperature and availability of water and macronutrients, such as N, P, K, Na and Mg (e.g. Botha, 2006Botha, , 2011Chernov, 2005;Sláviková & Vadkertiová, 2003;Vishniac, 2006a). Similarly, changes in the yeast community of soils correlate with soil moisture (or rainfall) following seasonal changes in forest soils (Sláviková & Vadkertiová, 2000), microclimate (Yurkov, Röhl, et al., 2016;Yurkov, Wehde, et al., 2016) and latitudinal changes of physico-chemical environmental conditions (Chernov, 2005;Vishniac, 2006a). At the same time, abiotic soil parameters have little effect on soil yeast communities within the same type of habitat. It has been shown that yeast quantity, diversity and community structure reflect vegetation properties, such as age and management history, but not basic abiotic properties, including pH, nitrogen content and C/N ratio (Birkhofer et al., 2012;Yurkov et al., 2012). Likewise, yeast communities in Mediterranean forest soils reflected the properties of the forest cover, which in turn is shaped by the local precipitation regime (Yurkov, Wehde, et al., 2016).
The diversity and composition of soil yeast communities is influenced by vegetation, i.e. plant diversity and composition.
Ascomycetous yeasts are more prominent in grassland and agricultural soils and the proportion of these yeasts increases with the intensity of land management (Sláviková & Vadkertiová, 2003;Yurkov et al., 2012). Agricultural practice is often associated with monoculture cropping, which negatively affects soil yeasts (reviewed in Vadkertiová et al., 2017 (Glushakova, Kachalkin, & Chernov, 2015aGlushakova et al., 2015b). Compared with typical meadow vegetation, the abundances of Saitozyma podzolica, Schwanniomyces castelli and Torulaspora delbrueckii were negatively affected by the invasion of Impatiens parviflora, whereas the soil-borne species Apiotrichum dulcitum and Apiotrichum laibachii were more prominent as a result of the invasion (Glushakova et al., 2015a). Similarly, Candida vartiovaarae, Schwanniomyces castelli and Tausonia pullulans were less abundant in a ruderal and invasive Heracleum sosnowskyi and Aster salignus (Glushakova et al., 2015b regime. A common feature of all three studied floral invasions is an increased species richness trend and the proportion of ascomycetous yeasts, most of which are not typical for meadow soils. This observation is consistent with the earlier report of ascomycetous yeasts dominating soil yeast communities in managed grasslands (Yurkov et al., 2012).
Many yeast species are adapted to soil habitats. Some of them are widespread and others are found in a certain type of soil. Several studies that surveyed yeasts in a broad range of soils attempted to correlate soil properties with distribution of yeast taxa (e.g. Babjeva & Golovleva, 1969;Babjeva & Chernov, 1995;Chernov, 2005;di Menna, 1965a;Vishniac, 2006a). Chernov (2005) and Vishniac (2006a) performed the two largest studies of soil yeasts along a latitudinal gradient in the former USSR and western North America, respectively. They examined basic environmental parameters as factors that may influence the distribution of yeasts in these soils.
Both authors reported substantial dissimilarity between sampling regions. In samples collected on the East European Plain, the quantity of yeasts showed a unimodal distribution reaching the highest values in boreal and temperate climates and rapidly declined towards the North and the South (Chernov, 2005). Similarly, the diversity of yeast communities increased from subtropical deserts to the tundra but most of the increase was observed in boreal climate between forest biotopes and the tundra (Chernov, 2005(Chernov, , 2013. Among potential indicator species figured Saitozyma podzolica, associated with acid well-drained soils and Nadsonia (Schizoblastosporion) starkeyi-henricii, frequent in cold and temperate hydromorphic soils (see also Babjeva & Blagodatskaya, 1972;di Menna, 1965b;Yurkov et al., 2012).
Although soil yeasts respond to environmental parameters, mechanisms explaining their distribution patterns are not yet understood.
The observed spatial heterogeneity and endemism of soil yeasts may result from undersampling or reflect the distribution and availability of ecological niches yeasts occupy in soils. In contrast to the common view on yeasts as free-living soil organisms, their distribution may not depend on abiotic factors (e.g. Birkhofer et al., 2012) but is determined by plant, insect and fungal hosts and vectors.