OUP user menu

Yeast communities associated with the bulk-soil, rhizosphere and ectomycorrhizosphere of a Nothofagus pumilio forest in northwestern Patagonia, Argentina

M. Cecilia Mestre, Carlos A. Rosa, Silvana V.B. Safar, Diego Libkind, Sonia B. Fontenla
DOI: http://dx.doi.org/10.1111/j.1574-6941.2011.01183.x 531-541 First published online: 1 December 2011


Soil microorganisms play an important role in soil quality and they interact closely with vegetation. Little is known about yeast diversity and function in forest soil ecosystems and their interactions with other biotic soil components, particularly in the mycorrhizosphere. We studied the diversity of yeasts inhabiting the bulk-soil, rhizosphere and ectomycorrhizosphere of a Nothofagus pumilio forest in Nahuel Huapi National Park (Bariloche, Argentina). Ectomycorrhizal infection was observed in all N. pumilio trees studied. A total of 126 yeast isolates were obtained, including 18 known and three possibly new species. Basidiomycetous yeasts were predominant in all soil fractions, and the most frequently isolated species was Cryptococcus podzolicus. Diversity indices and multivariate analyses were used to study and compare yeast communities in the bulk-soil, rhizosphere and ectomycorrhizosphere. Yeasts able to ferment glucose were found associated with the rhizosphere. Many of the recovered yeast species were associated with lignocelluloses compound degradation, which suggest that yeast plays an important role as a decomposer in these forest soils. Each soil fraction has a distinct yeast assemblage related to their physiologic capacities and soil nutrient availability.

  • yeast ecology
  • ectomycorrhizae
  • Cryptococcus podzolicus


Nothofagus (Nothofagaceae) genus comprises c. 35 tree species that inhabit New Zealand, Australia, Tasmania, and cold-temperate woods in South America (Manos, ). In South America, its geographic distribution ranges from 35 to 56°S on both sides of the Andes in Chile and Argentina. These forests are very important as they contain a great number of endemic species; in Argentina, large areas of Nothofagus forest are included in protected areas (Laclau, ). The dominant tree species in the Andean-Patagonian forests are principally Nothofagus spp. (southern beech) and conifers such as Austrocedrus chilensis (southern cedar or ‘ciprés de la cordillera’), Araucaria araucana (commonly named pehuén), Fitzroya cupressoides (commonly named alerce or Patagonian cypress), and Pilgerodendrum uviferum (commonly named ‘ciprés de las Guaitecas’, ten or lahuán), which are among the most important species in Argentina and Chile (Donoso Zegers, ). These native forests are characterized by low anthropogenic impact and minimal atmospheric pollution (Satti et al., ). Nothofagus pumilio (Poep. Et Endl) Krasser, an endemic deciduous species, is the main species of the high altitude tree-line forests of the Andes (Souza et al., ). This species inhabits a wide altitude and latitude range, its appearance varying from tall trees (about 30 m) to a bush-like form on the highest (altitude) limit of distribution. Nothofagus pumilio forests account for 25% and 4% of native forest surface in Chile and Argentina, respectively, constituting one of the most important types of forest in both countries (Gonzalez et al., ). In the north Patagonian region of Argentina, about 70% of plant species in the forest present mycorrhizal colonization, principally arbuscular mycorrhizae, and Nothofagus species present high rates of ectomycorrhizal colonization (73–79% of infected tips in summer) (Diehl, ; Diehl et al., ). Extensive soil characterization and the nutrient dynamics of these soils can be found in Satti et al. () and Diehl et al. (, ).

Soil biota is essential for ecosystem function and significantly influences the diversity and structure of aboveground communities (Kardol & Wardle, ). Bacteria, filamentous fungi, and yeasts are found in soils worldwide (di Menna, ; Botha, ; Vishniac, ). The relevance of yeasts to soil function is not yet fully understood, although it is known that they influence soil aggregation, contribute to nutrient cycles, and interact with the vegetation (Cloete et al., ; Botha, ). Soil yeasts are found in the bulk-soil, the rhizosphere (Zacchi et al., ), and inside roots (Isaeva et al., 2009). As the rhizosphere is largely influenced by root exudates and symbiotic activities, large microbial populations inhabit the zone surrounding roots and upper soil layers, among which mycorrhizal fungi are one of the predominant microorganisms (Barea et al., ; Bertin et al., ). Mycorrhizal fungi provide key services in forest ecosystems as they facilitate plant growth and nutrient uptake, and increase the absorptive surface area of their host plant root systems (Aerts, ). Some authors maintain that for many ectomycorrhizal trees in the world (Pinaceae, Fagaceae, Betulaceae), survival and normal growth is directly related to ectomycorrhizal colonization (Aerts, ). While the interaction of yeasts with soil animals has received some attention (Yurkov et al., ), interactions with other biotic components, particularly in the mycorrhizosphere, remain unknown. A few studies showed that yeasts have a probable role as colonization helper for ectomycorrhizal fungi (Garbaye & Bowen, ). In addition, several yeasts associated with roots and spores of arbuscular mycorrhiza (AM) have been reported (Fracchia et al., ; Renker et al., ; Sampedro et al., ).

Yeast diversity in the Andean Patagonia region has been surveyed in many natural habitats such as high altitude lakes, extreme environments, stromata of Nothofagus parasitic fungus, and the flowers of some herbs (Brizzio & van Broock, ; Libkind et al., 2004, ; Russo et al., ; de García et al., ; Ulloa et al., ). Soil-related substrates have been little studied (Mestre et al., , 2010, 2011). Knowledge of yeast diversity in other forest soils in the Southern Hemisphere is also scarce and is limited to a few studies performed in New Zealand (di Menna, , ; Spaak, ). In this study, the diversity, distribution, and physiologic properties of yeast inhabiting the bulk-soil, rhizosphere, and ectomycorrhizosphere of N. pumilio were investigated in Patagonia.

Materials and methods

The sampling site was located on the southeast slope of Cerro Otto in San Carlos de Bariloche (1350 m altitude), in a pure N. pumilio forest which has three vegetation strata and a high degree of plant cover and root density. Annual precipitation in this area reaches 800 mm (Souza et al., ), and soils are Andisols with a low degree of development, characterized by a high capacity for stabilizing organic matter, storing water, buffering pH, and retaining phosphorus (Satti et al., ). Five adult N. pumilio trees were randomly selected at a distance of at least 15 m from each other. Three soil cores (10 cm in diameter) from each tree were aseptically collected at a distance of 60 cm from a trunk. These cores represented the upper 30 cm of the soil profile. The litter layer had previously been removed. The soil cores, considered as sub-samples for each tree, were individually transported to the laboratory in plastic bags and maintained at 4 °C until processed (no longer than 1 week).

Soil fractions recovery

The three soil sub-samples from each tree were pooled together aseptically in the laboratory. Each composite sample was sieved using a 2-mm mesh to separate soil from the roots. Soil without roots was considered the bulk-soil (BS). Root fragments, with strongly adhered soil, were then shaken gently to recover the rhizospheric soil (R). Following this, roots were washed twice with sterile tap water. The clean roots were inspected using an Olympus SZX9 stereomicroscope for selection of predominant and/or the most abundant ectomycorrhizal morphotypes (ECM-Mt). These ectomycorrhizal morphotypes were grouped by morphologic features (e.g. color, surface appearance, presence of emanating hyphae, and rhizomorphs), according to Agerer (). Specimens of the same morphologic ECM-Mt were aseptically pooled together into plastic micro-tubes for the recovery of yeast from the ectomycorrhizosphere (E).

Medium for count and isolation

Bulk-soil and R soils were suspended 1 : 25 (w/v) in 0.9% NaCl solution (final dilution was 40 g of wet soil per liter) and shaken at 250 r.p.m. for 30 min. An aliquot of 100 μL was distributed on the surface of solid MYP medium (% w/w, malt extract 0.7, yeast extract 0.05, peptone-soytone 0.25, agar 1.5) supplemented with rose bengal 25 μg mL−1 and chloramphenicol 200 μg mL−1.

The ECM-Mts were also placed in 0.9% NaCl solution (1 : 25, w/v) and vortexed (three times for 2 min) to release tightly adhered yeast cells from the ectomycorrhizosphere (E). An aliquot of 100 μL was distributed on the surface of solid MYP medium supplemented with rose bengal 25 μg mL−1 and chloramphenicol 200 μg mL−1. Each sample was plated in triplicates. Incubation of all plates was performed at 20 °C, and colonies were counted after 72 h of incubation, to avoid the interference of mold growth in yeast count procedure (Mestre et al., ). Incubation temperature was selected based on environmental condition (mean temperature 10 °C; Gonzalez et al., ) and data from other yeast diversity studies in the region (Libkind et al., ; Ulloa et al., ). Yeast quantity was expressed as colony forming units per gram of fraction (CFU g−1 of each fraction). After 120 h of incubation, all yeast colonies grown were differentiated into macro-morphologic types (color, aspect, margin) using a dissection microscope. Three representative colonies of each morphotype from each count plate were purified. Pure cultures were maintained on MYP medium.

Yeast identification

Physiology profile

All isolated yeasts were grouped according to macro and micromorphology and physiologic characterization, which were performed according to Yarrow (). At least two representative isolates of each group were identified using rDNA sequencing. Fifty isolates were sequenced and all isolates within a group belonged to the same species. All yeast isolates were cryopreserved (−20 °C) using MYP liquid medium with 12% glycerol.

DNA extraction

DNA extraction was performed by suspending one colony loop of 72 h-old yeast culture into 100 μL of lysis buffer (Tris 50 mM, NaCl 1 M, EDTA 0.1 mM, SDS 0.5%) and incubated for 30 min at 50 °C. After incubation, 100 μL of phenol : chloroform : isoamyl alcohol (25 : 24 : 1) was added, and samples were vortexed for 2 min at a maximum velocity and then centrifuged for 15 min at 21 400 g. The upper phase was transferred to a clean tube and DNA precipitation was performed twice with 100 μL ethanol and subsequent centrifugation at 21 400 g for 15 min. Finally, the DNA was re-suspended in 100 μL TE buffer (10 mM Tris, 1 mM EDTA).

Sequence analysis

The D1/D2 domain at the 5′ end of large subunit (LSU) rRNA gene was amplified using primers NL-1 (5′-GCATATCAATAAGCGGAGGAAAAG) and NL-4 (5′-GGTCCGTGTTTCAAGACGG). Amplification was performed for 35 PCR cycles with annealing at 54 °C, extension at 72 °C for 20 s, and denaturation at 94 °C for 15 s. Fragment amplification and size were conferred by 1.5% agarose gel electrophoresis and stained with Ethidium Bromide (5 μg mL−1). Sequencing of amplicons was carried out using an ET Dynamic Terminator Kit in a MegaBACE 1000/Automated 96 Capillary DNA sequencer (GE Healthcare) using primer NL-1 (5′-GCATATCAATAAGCGGAGGAAAAG). Comparisons with sequences from GenBank database (http://www.ncbi.nlm.nih.gov/) were carried out using blastn search algorithm. Sequence alignment and neighbor-joining tree construction (based on 1000 bootstrap iteration) was performed using Mega version4 (Tamura et al., ).

Yeast community

The yeast community was described for each soil fraction (BS, R, and E), using ecologic indices for species diversity, species similarity, and the physiologic features of the isolated yeasts.

Yeast diversity in each soil fraction was studied using Shannon-Weaver (H) index with Hutcheson's t-test (α = 0.05) as described in Moreno (). Similarities between communities were studied with Jaccard index (J) according to Chao et al. (), using the occurrence frequency of each isolated species.

Twelve physiologic tests were selected from the previously performed phenotypic tests to study the functional aspects of each yeast community: assimilation of known root exudate carbon compounds (cellobiose, d-xylose, l-arabinose, trehalose, lactose, and M-inositol); assimilation of different forms of nitrogen (nitrate and nitrite in inorganic form and lysine in organic form), and glucose fermentation and related tests (ethanol assimilation, acid production, and solubilization of CaCO3). The functional analysis was performed using three different approaches: (1) soil fraction level; (2) species level; and (3) linking both soil fraction and species levels:

  • Percentage of feature occurrence per soil fraction (calculated as percentage of isolates within each soil fraction showing each feature) was used to describe the physiologic properties of each yeast community.

  • Percentage of feature occurrence per species (calculated as percentage of isolates within each species showing each feature) was used for principal component analysis (PCA) and hierarchical clustering (following Ward's aggregation criterion), with spad 5.6 software, to obtain species clusters with similar physiologic responses.

  • Occurrence frequency of each species in clusters in BS, R, and E (calculated as number of isolates within each cluster) was used in a correspondence factorial analysis (CFA) with spad 5.6 software, to assess the association between species clusters (and their physiologic features) and soil fraction.


All N. pumilio trees in this work presented ectomycorrhizal infection, and five ECM-Mts were selected as the most abundant ones (Table 1). Yeasts were recovered from all five ECM-Mts and were grouped together to constitute the E fraction. Yeasts were isolated from every soil sample; 126 isolates were obtained: 38 from the bulk-soil, 43 from the rhizosphere, and 45 from the ectomycorrhizosphere fractions. Average yeast count was 3.75 × 103 CFU g−1 for BS and 3.42 × 103 CFU g−1 for R. The yeast count was different for each ECM-Mt, with an average value of 7.93 × 103 CFU g−1 for the ectomycorrhizosphere. Lower count values were associated with ECM-Mt 4 with a smooth surface (9.72 × 102 CFU g−1), and higher values were observed on ECM-Mt with a rough and woolly surface (ECM-Mt5: 1.60 × 104 CFU g−1 and ECM-Mt3: 1.26 × 104 CFU g−1, respectively; Table 1).

View this table:

Morphologic features of ectomycorrhizal morphotypes (according to Agerer, ) and yeast counts in Nothofagus pumilio roots

Ectomycorrhizal morphotypesSurfaceRamificationRhizomorphEmananting hyphaeYeast count
ShapeColorShape–frequencyColor–frequencyColor–frequency(CFU g−1)
ECM-Mt 1SmoothWhiteIrregularly pinnate – fewWhite fan like – abundantAbsent5.83 × 103
ECM-Mt 2SmoothWhiteSimpleAbsentAbsent4.25 × 103
ECM-Mt 3WoollyBrownSimple or monopodial-pinnateAbsentGolden abundant1.26 × 104
ECM-Mt 4SmoothLight brownMonopodial-pinnate – abundantAbsentAbsent9.72 × 102
ECM-Mt 5RoughBlackSimpleAbsentBlack – frequent1.60 × 104

Basidiomycetous yeasts prevailed in all soil fractions, the highest value being observed in the E fractions. Ascomycetous yeasts were higher in the R fraction than in the other two fractions. Eighteen known yeast species (12 Basidiomycetes and six Ascomycetes), two possibly new species and one recently described species were recovered in this work. Species isolated in this study and their distribution in the three soil fractions is shown in Table 2.

View this table:

Yeast taxa isolated in this study and their distribution and number of isolates in the three soil fractions. Occurrence frequencies ≥ 0.1 are shown in parenthese

Yeast taxaTotal 126Soil fraction
BS 38R 43E 45
Basidiomycetes 96 (0.76)30 (0.79)27 (0.63)39 (0.87)
Cryptococcus aerius 12 (0.10)246 (0.13)
Cryptococcus sp.5023
Cruptococcus phenolicus 17 (0.13)5 (0.13)39 (0.20)
Cryptococcus terreus 65 (0.13)10
Cryptococcus terricola 5122
Cryptococcus podzolicus 29 (0.23)8 (0.21)11 (0.26)10 (0.22)
Tremella sp.1001
Holtermanniella wattica 4004 (~0.10)
Asterotremella albida 4130
Trichosporon dulcitum 1100
Trichosporon porosum 65 (0.13)01
Rhodotorula colostri 2200
Rhodotorula fujisanensis 2002
Guehomyces pullulans 2011
Ascomycetes 25 (0.20)7 (0.18)14 (0.33)4 (~0.10)
Candida railenensis 2110
Candida maritima 725 (0.12)0
Candida sp.7133
Pichia delftensis 2101
Lindnera rhizosphaerae 2020
Hanseniaspora valbyensis 3120
Saccharomyces sp.1010
Dothiora cannabinae 1100
Underminated strain
Mycorrhizal fungal5122
  • Species isolated from a single substrate.

  • Possible new species.

  • New species recently described.

  • Uncertain affinity with Saccharomyces uvarum/bayanus group.

  • BS, bulk-soil; R, rhizosphere; E, ecto-mycorrhizosphere.

Cryptococcus podzolicus was the most frequently occurring species with similar occurrence frequency values in BS (0.21), R (0.26), and E (0.22). A few species were recovered with frequency values higher than 0.1, and they were different in each soil fraction: Cryptococcus phenolicus, Cryptococcus terreus, and Trichosporon porosum in the BS fraction; Cr. aerius and Candida maritima in the R fraction; Cr. aerius, Cr. phenolicus, and Holtermanniella wattica in the E fraction (Table 2). The genus Holtermanniella was recently described to include the anamorphic Cryptococcus species within Holtermannia clade (Wuczkowski et al., ). Eight species, five basidiomycetous and three ascomycetous yeasts, occurred exclusively in one fraction. Holtermanniella wattica and Rhodotorula fujisanensis seemed to be linked to the E fraction.

One possible new species with affinity to Cr. aerius was recovered from the R (two isolates) and E (three isolates) soil fractions, which showed 16–18 nucleotide differences with the type strain (Cr. aerius CBS 155T, GenBank sequence AF075486). One possibly new ascomycetous yeast with affinity to Candida ralunensis was recovered from all three soil fractions and it showed 11 nucleotide differences with the type strain (Candida boleticola CBS 8179T, designated type strain of C. ralunensis, Genbank sequence U45786). The formal description of Lindnera rhizosphaerae, a new ascomycetous species isolated only from the R fraction, has recently been published (Mestre et al., ).

Rhodotorula colostri, the only pigmented species isolated in this study, was isolated from the BS fraction. An additional group of isolates (treated separately throughout this work) showed sequence homology with unidentified mycorrhizal fungi (GenBank sequence AY394920) from phylum Ascomycota, and they represented 4% of total isolates.

Species diversity values, measured with Shannon–Weaver index, were higher in the BS and R communities (H = 1.07 and H = 1.06 respectively) than in E (H = 0.98). There were significant differences for Shannon–Weaver diversity indices between BS and E communities (P = 0.03), but no significant differences were found between BS and R (P = 0.1) or R and E communities (P = 0.31). The Jaccard index for community similarity analysis showed that BS and R yeast communities were the most similar (J = 0.66). The E species community was more similar to R (J = 0.56) than to BS (J = 0.52).

The functional aspect of the three yeast fraction communities, studied by the percentage of occurrences of 12 selected physiologic features (Fig. 1), showed that glucose fermentation and organic acid production were rare features that occurred more frequently in the R community than in the other two communities. A higher percentage of isolates from BS and E communities assimilated l-arabinose, trehalose, lactose, and M-inositol than in R community. d-xylose was assimilated by a high percentage of isolates in the three communities. Cellobiose was assimilated by a higher percentage of isolates in R than in BS or E communities. Less than 50% of isolates from R community assimilated NO2, whereas over 60% of isolates in the other two communities assimilated it. NO3 and lysine were assimilated by over 50% of isolates in each of the three communities.

Percentage of occurrence of physiologic feature per soil fraction. C source assimilation test: Cel, cellobiose; d-Xyl, d-Xylose; l-Ara, l-Arabinose; Tre, trehalose; Lac, lactose; M-ino, M-inositol; Eth, ethanol; Acid, acid production test as solubilization of CaCO3; Ferm, glucose fermentation. N source assimilation test: NO3, nitrate; NO2, nitrite; Lys, lysine (organic form). All tests were performed at 20 °C.

Results for the 12 selected physiologic tests for yeast species isolated in this study are presented in Table 3. Species of the orders Filobasidiales, Tremellales, Holtermanniales, Cystofilobasidiales and Trichosporonales, and the mycorrhizal fungal group showed a capacity for sugar assimilation. Cryptococcus podzolicus and Dothiora cannabinae were the species with the widest range of sugar assimilation. Species within the orders Sporidiobolales and Cystofilobasiales, species within the Microbotryomycetes, the mycorrhizal fungal group, and ascomycetous yeasts were able to assimilate ethanol. Isolates within the order Saccharomycetales were the only ones capable of fermenting glucose and producing organic acid (with the exception of Asterotremella albida in Trichosporonales).

View this table:

Physiologic profiles determined for yeast species associated with the studied Nothofagus pumilio forest soil, and species cluster composition generated by a principal component analysis followed by hierarchical clustering

Yeast taxaClusterPhysiologic test
Celd-Xyll -AraTreLacM-inoEthAcidFermNO3NO2Lys
Cryptococcus aerius IIv++v+++
Cryptococcus sp.IIvvvv+v+v
Cryptococcus phenolicus II+++++++
Cryptococcus terreus II+v++v+++
Cryptococcus terricola IV++++++v++
Cryptococcus podzolicus IV++++++v+++
Tremella sp.III+++++
Holtermanniella wattica III++v+++v+
Asterotremella albida III++++++++
Trichosporum dulcitum IVvvvv+v+v++
Trichosporon porosum III++++++++
Rhodotorula colostri IIv++v++
Rhodotorula fujisanensis I++
Guehomyces pullulans IV+++++++++
Candida railenensis I+v+v++v+
Candida maritima I+++++v+
Candida sp.Iv+++v+
Pichia delftensis Ivv+++++
Lindnera rhizosphaerae Iv++vvv+
Hanseniaspora valbyensis Iv++++
Saccharomyces sp.I+++
Dothiora cannabinae IV++++++++++
Undermined strain
Mycorrhizal fungalIV++++++++++
  • C source assimilation test: Cel, cellobiose; d-Xyl, d-Xylose; l-Ara, l-Arabinose; Tre, trehalose; Lac, lactose; M-ino, M-inositol; Eth, ethanol. Acid: acid production test as solubilization of CaCO3. Ferm, glucose fermentation. N source assimilation test: NO3, nitrate; NO2, nitrite and Lys, lysine (organic form). All tests were performed at 20 °C.

A multivariate approach (PCA) was used to group species from this study into four clusters using percentage of feature occurrence per species for the selected tests presented in Table 3. Cluster composition is also shown in Table 3. Cluster I was characterized by the ability to ferment glucose and produce acid, and inability to assimilate trehalose, lactose, cellobiose, l-arabinose, M-inositol, and inorganic forms of nitrogen; this cluster includes all Saccharomycetales species recovered in this study. Cluster II was characterized by its ability to assimilate NO3, but not ethanol nor lysine. Cluster III has no characteristic feature and includes Tremella sp. and Trichosporonales species, except Trichosporon dulcitum and H. wattica. Cluster IV was characterized by the assimilation of lactose, l-arabinose, and NO2, and includes some basidiomycetes, some ascomycetes, and the mycorrhizal fungal group. When the species cluster association with soil fraction was studied using CFA (Fig. 2) Cluster I was associated with R fraction, cluster II with E fraction, and cluster III with BS fraction. Cluster IV had no clear association with either fraction, although it was closer to R and E than to BS fraction.

Correspondence factorial analysis on species clustering associated with the different soil fractions. BS, bulk-soil; R, rhizosphere; E, ecto-mycorrhizosphere.


Yeast numbers, in different soil types and locations, ranged from 1 to 106 yeasts per gram of soil (Botha, ). In our study, yeast count values for the three soil fractions (about 103CFU g−1 soil) are within the range previously reported for other forest soils (di Menna, ; Slaviková & Vadkertiová, ; Maksimova & Chernov, ).

Occurrence frequencies of many yeast species isolated in this study were lower than 0.1, which could indicate that such species might not be strictly pedobiont (soil associated), but transient or epibiont (associated with leaves and above ground substrate) species that fall onto soil (Maksimova & Chernov, ; Vishniac, ). Cryptococcus podzolicus, Cr. phenolicus, and Cr. aerius are pedobiont species in this study. Rhodotorula fujisanensis, a low frequency species, could be exogenous or transient, as it has been associated with leaf litter and leaves in temperate woods (Glushakova & Chernov, ; Maksimova & Chernov, ), and with Nothofagus seed in the region studied (N. Fernández, pers. commun.). Species with high occurrence frequencies in a given soil fraction (but not the total amount of yeast) could be associated with this particular substrate (di Menna, ; Zacchi et al., ; Maksimova & Chernov, ; Vishniac, ). In this work, certain soil fractions present some distinct yeast distributions and also different yeast communities with particular physiologic capacities. For example, Cr. terreus and Tr. porosum were isolated from the BS fraction, C. maritima from the R fraction, and H. wattica from the E fraction. This also happened with fermentative yeasts, usually associated with fruit and other sugar-rich substrates. The occurrence frequency of fermentative yeasts in the R fraction reached 0.33, indicating that they were associated with this fraction, which could be thought of as a rich soil substrate due to its high sugar level from root exudates.

Pigments, such as carotenoids and melanin, play an important role in UV and sunlight protection (Libkind et al., ; Sterflinger, ; Yurkov et al., ; Moliné et al., ). The so-called red and black yeasts therefore are associated with phylloplane, seed surface, and other environments with high irradiation rates (Fonseca & Inácio, ). Some authors mention the presence of red yeast in soil samples (Slaviková & Vadkertiová, ; Maksimova & Chernov, ; Wuczkowski & Prillinger, ). The scarce presence of pigmented yeast (both red and black yeast) in the soil fractions studied could reflect a different yeast community selection in high (phylloplane) and low (soil) irradiation environments. Different rates of carotenoid production have also recently been revealed for strains isolated from phylloplane and soil (Yurkov et al., ).

The polyphyletic genus Cryptococcus (Scorzetti et al., ) comprised 65% of isolates and soil-borne cryptococci from the lineages : Filobasidiales, Tremellales, Holtermanniales, and Trichosporonales, which were also found in our study (Table 2). These species possess some common adaptations, which facilitate their development in soils, such as the production of polysaccharide capsules that enable yeasts to sequester and concentrate nutrients (Botha, ; Fonseca & Inácio, ) or sustain low water activity (Fonseca & Inácio, ; Raspor & Zupan, ). Furthermore, these exogenous polysaccharide capsules may play an important role in soil aggregation (Lynch, ; Botha, ). Cryptococcus podzolicus was the most frequently isolated species among the three soil fractions. This species was originally isolated from soddy-podzolic soils (Podzols according to FAO classification) in a wide geographic range within the former USSR (Babjeva & Reshetova, ) and later in other regions of the world (Fell & Statzell-Tallman, ; Maksimova & Chernov, ). Interestingly, β-xylosidase and endoxylanase production as secretory enzymes by Cr. podzolicus has been reported by Shubakov (). These enzymes are important in wood degradation processes (Schmidt, ): hemicellulose xylan in wood is hydrolyzed, releasing xylose monomers. Middelhoven () reported that basidiomycetous yeasts isolated from decaying wood assimilated soluble starch, pullulan, dextran, xylan, polygalacturonate, galactomannan, and tannic acid, or at least some of these compounds; the most active species in this respect were Cr. podzolicus and T. porosum. Strains from these species isolated in our study might have a fundamental role in complex compound degradation leading to formation of soil organic matter.

In terrestrial ecosystems like ours, there is an important input of ligno-cellulosic material introduced into soil by leaf fall (Diehl et al., ; Bertiller et al., ). Cellobiose, d-xylose, and l-arabinose are derived from lignocellulosic material hydrolyzed by bacteria and mold in soils (reviewed by Botha, ). These and other carbohydrates with different chemical complexity (such as trehalose, lactose and myo-inositol) are also exuded by roots (Botha, ). Many basidiomycetous yeasts isolated in this study assimilated the above mentioned carbon compounds. Some species belonging to Trichosporon genus such as Tr. porosum were described previously as positive for phenolic compound assimilation that may be related to lignin degradation (Middelhoven, ; Middelhoven, ).

Glucose fermentation and acid production were associated with the Saccharomycetales (cluster I) and the R fraction. Diehl et al. () have determined that in the most abundant trees of the region studied, phosphorous (P) is not the limiting growth nutrient, in accordance with what was expected due to the soil chemistry (Andisoles which normally retain P). Phosphorous released from its insoluble organic forms could be attributed to ectomycorrhizal activity (Diehl et al., ), whereas weak organic acid production by yeast could contribute to inorganic P solubilization in R fraction, where it may be taken up by roots.

Rhizospheric and BS communities were similar with respect to species diversity and species composition. The studied sites correspond to a mature forest with three vegetation strata, high plant cover, and high root density. Nothofagus, the dominant trees species, presents ectomycorrhizal infection, and over 70% of the remaining plant species present arbuscular mycorrhizal infection (Diehl, ; Diehl et al., ). The Bulk-soil is greatly influenced by the mycorrhizal mycelium that connects different plants by means of a common network of mycelia (Giovannetti et al., ), leading to more uniform conditions in both fractions (R and BS) than expected (Richter et al., ). The BS is expected to be more diverse, as it could include indigenous and exogenous (falling leaves, water runoff, etc) species, acting as a yeast repository in the forest. Rhizospheric community is directly influenced by root exudates, and although species are similar to BS, the proportion might change due to its higher sugar concentration, (e.g. in this case, fermentative species present in BS had a higher occurrence frequency in R).

Yeast from the Ectomycorrizosphere

Yeast counts in the E fraction were different for each ECM-Mt: ectomycorrhizal woolly morphotypes with large emanating hyphae would retain larger quantities of soil particles, and this could explain the higher number of yeasts recovered on the plates. Another factor affecting the number of yeasts obtained from each ECM-Mt could be a different, distinct exudation pattern for each ECM-Mt, which could influence the yeast (and other microorganism) communities associated with it (Garbaye, ). Our results suggest that the Ectomycorrhizosphere might be a common habitat for yeasts, in agreement with reports on isolation and detection of yeasts from mycorrhizal roots (Garbaye & Bowen, ; Zacchi et al., ; Renker et al., ). Ectomycorrhizosphere yeast populations could depend on the interaction between the mycorrhizal fungus and the plant species roots. Although it is known that Nothofagus spp. trees in Patagonia have a high rate of ectomycorrhizal infection, little is known about ectomycorrhizal fungal partner identity. Studies have been carried out in the region that reveal the presence of basidiomycetous fungi (such as Descolea antarctica, Thaxtherogaster albocanus, Cortinarius sp., Descomyces sp., Godoy & Palfner, ; Nouhra et al., ), which are able to form ectomycorrhizas with Nothofagus spp. and other tree species. Yeast association with these organisms has not yet been studied. Cryptococcus podzolicus, Cr. aerius, Cr. phenolicus, and H. wattica seem to be autochthonous species from E fraction (table 2), and H. wattica, in particular, is linked to E fraction, as it was not isolated from any of the other substrates. The BS and E yeast communities were significantly different with respect to species diversity (calculated with Shanon–Weaver index) and species composition similarity (calculated with modified Jaccard index). The E community was associated with NO3 assimilation. Yeast species from BS and E communities shared the ability to assimilate complex sugar compounds, such as trehalose, lactose and myo-inositol, and had a lower number of glucose fermenting yeasts, which separates them from R community. This suggests that BS and E fractions are more similar in their physiologic properties.

Final remarks

In this work, we presented data from an intensive survey on yeast diversity in Patagonian forest soil, which could lead to a better understanding of microbial–plant relationships. This study suggests that the participation of yeast in the C cycle is related to hemicelluloses degradation, and also to N and P cycles. Yeast might also be associated with soil aggregation processes, as abundant capsulated yeasts were recovered. The three soil fraction communities studied presented distinctive yeast assemblages with particular physiologic characteristics.

The ectomycorrhizosphere is a unique community, where three living organisms co-exist: plant + mycorrhizal fungi + yeast. In the ectomycorrhizosphere root exudates availability is limited by the symbiotic fungi that consume these compounds (Hakes et al., ; Bonfante & Genre, ). The interaction between these three organisms and nutrient availability may be an important factor in defining the yeast community (and its diversity). This yeast–ectomycorrhizal association might play an important role in the successful colonization and development of ectomycorrhizal symbiosis. Detailed mechanisms of interactions between ECM and yeast fungi remain uncertain, and our study highlights the need for additional research on this subject.


We thank the authorities of Administración de Parques Nacionales (Argentina) for their courtesy and cooperation. We thank Dr A.M. Yurkov for helpful comments on the manuscript and BSc (Hons) Audrey Urquhart for languages revision. This work was supported by project UNComahue (B143) and FONCyT projects PICT04-22200. Bilateral cooperation between Argentina and Brazil was supported by CAPES-MINCyT agreement (BR 06/011), and Conselho Nacional de Desenvolvimento Cientifico e Tecnológico (Prosul-CNPq-Brazil). M.C.M. was supported by an ANPCyT PhD grant and a CONICET PhD type II fellowship.


  • Editor: Philippe Lemanceau


View Abstract