Correspondence: Luiz H. Rosa, Departamento de Microbiologia, Instituto de Ciências Biológicas, Universidade Federal de Minas Gerais, PO Box 486, Belo Horizonte, MG CEP 31270-901, Brazil. Tel.: +55 31 3409 2751; fax: +55 31 3409 2730; e-mail: firstname.lastname@example.org
This study assessed the diversity and distribution of filamentous fungi obtained from water sampled from six lakes in the Antarctic Peninsula. One hundred and twenty-eight fungal isolates were purified and identified by analysis of nuclear rDNA ITS region sequences as belonging to 31 fungal different operational taxonomic units (OTUs). The most frequently isolated fungi were Geomyces pannorum and Mortierella sp.; these species occurred in six and three of the lakes sampled, respectively, and displayed the highest total colony-forming unit per L. Different species that have not been found to these lakes and/or had adapted to cold conditions were found. In general, the fungal community displayed low richness and high dominance indices. The species Cadophora cf. luteo-olivacea, Cadophora malorum, Davidiella tassiana, G. pannorum, Mortierella cf. alpina and Thelebolus cf. microsporus that were found in the lakes in question were also previously found in other cold ecosystems, such as Arctic, temperate and Alpine regions. The results of this study suggest the presence of an interesting aquatic fungal web, including symbionts, weak and strong saprophytes and parasite/pathogen fungal species. This aquatic web fungal may be a useful community model for further ecological and evolutionary studies of extreme habitats.
Antarctica is one of the most pristine regions of the planet, and its environments include habitats that combine cold, dry and oligotrophic extremes and offer unique opportunities to discover extremophile microorganisms. In Antarctica, the microbial communities represent the largest reservoir yet to be described of biodiversity adapted to withstand extreme conditions. According to Vincent (2000), some Antarctic habitats have remained sealed for hundreds of thousands of years or longer and represent an opportunity for exploring microbial evolution.
The majority of lakes of the Antarctic Peninsula are shallow systems (< 10 m deep), are highly transparent and allow high levels of light, including UV radiation, which may be a stress factor on microbial diversity (Izaguirre et al., 1993). In addition, the water of Antarctic lakes is cold and contains few available nutrients. However, the complex geology of the different islands of the Antarctica Peninsula results in variable lake chemistry. In this environment, microbial life is the predominant and plays a major role in the transfer of inorganic, organic material and energy (Ellis-Evans, 1996).
Although present in different ecosystems and substrata throughout the world, Antarctic fungi still represent an unknown proportion of fungal diversity, which occur in different substrates, such as soil and ice, and are associated with plants. Although scarce, studies of Antarctic terrestrial fungi have been carried out by a number of researchers (Ruisi et al., 2007). The mycota in the lakes of Antarctica is subject to a number of potential limiting factors, such as extreme low temperatures (including frequent freeze–thaw cycles), high salinity, different pH values, high UV radiation and low nutrient availability. However, few mycological studies were performed in the freshwater lakes until now, and the diversity and geographical distribution of fungal communities across the Antarctic lakes has yet to be thoroughly examined. In this study, we investigated the presence, distribution and diversity of the fungi in Antarctic lakes from different locations in the Antarctica Peninsula, including Admiralty Bay, King George Island and Deception Island.
Materials and methods
Sampling sites and physico-chemical parameters
The samples were collected from five lakes in Admiralty Bay (King George Island) and one lake in Port Foster (Deception Island), Antarctica (Fig. 1). Five samples of 500 mL from three different points (each point was approximately 50 m apart) of the littoral region of each lake were collected during the austral summer season between December 2008 and January 2009. The samples were collected using sterile bottles and were transported to the laboratory on ice within 24 h for processing. The physico-chemical parameters (temperature, conductivity, salinity, pH, chlorophyll a) of the water of each lake were obtained using the YSI 650 multi-parameter display system (YSI Environmental).
A total of 90 samples (15 samples of 500 mL per lake) were filtered through 0.45-μm membranes with a 47 mm diameter (Millipore). The membranes were placed on YM agar (0.3% yeast extract, 0.3% malt extract, 0.5% peptone, 2% glucose, 2% agar) containing 100 mg mL−1 chloramphenicol (Sigma) and incubated at 15 °C for until 30 days. Fungal colony-forming units (CFUs) were counted and subcultures made of all morphologically distinct colonies from each sample. The subculture were grouped into different morphotype according to their cultural (colony colour and texture, border type and radial growth rate) and micro-morphology characteristics on YM agar. All fungal isolates obtained were deposited in the Collection of Microorganisms and Cells of the Universidade Federal de Minas Gerais, Brazil, under the code UFMGCB.
DNA extraction, amplification and sequencing for fungal identification
The fungal pure cultures were submitted to DNA extraction according to protocols previously described by Rosa et al. (2009). The primers ITS1 and ITS4 were used to amplify ITS regions between the SSU and LSU regions (ITS1-5.8S-ITS2) (White et al., 1990). Sequencing was performed using the methods described by Rosa et al. (2009). The consensus sequence was obtained using the bioedit v. 18.104.22.168 software. To identify species by rRNA gene sequencing, the consensus sequence was aligned with all of the sequences of related species retrieved from the GenBank database using the fasta 2.0 program (Altschul et al., 1997) via blastn searches to find the most likely taxonomic designation. The consensus sequences were deposited in GenBank, and the accession numbers are shown in Table 1. According to Gazis et al. (2011) and Ko et al. (2011), sequencing of the ITS region may fail to recognise some Ascomycota taxa. For this reason, the following criteria were used to interpret the sequences of the GenBank database: for sequence identities ≥ 99%, the genus and species were accepted; for the sequence with 98% of identities, the genus and species were accepted, but term ‘cf.’ (Latin for confer = compares with) used to indicate that the specimen resembles, but has certain minor features not found on the comparison with the reference species, were used; for sequence identities between 95% and 97%, only the genus was accepted; and for sequence identities ≤ 95%, isolates were labelled at the order or family name or as ‘unknown’ fungi. However, the taxa that displayed identities ≤ 97% or inconclusive taxonomic position were submitted to phylogenetic inferences, which were estimated using mega Version 4.0 (Tamura et al., 2007). The maximum composite likelihood model was used to estimate evolutionary distance with bootstrap values calculated from 1000 replicate runs. In addition, sequences of reference species deposited in GenBank were added to phylogenetic analysis accuracy. Information about the fungal taxonomic hierarchical levels follows the databases MycoBank (http://www.mycobank.org/) and Index Fungorum (http://www.indexfungorum.org/).
Table 1. Fungal taxa obtained from Antarctic lakes and identified by sequence comparison with the best blastn match with the NCBI GenBank database
Species abundance, diversity, richness, similarity and correlation
The percentage abundance of each taxon was calculated according to the following formula: percentage abundance of taxon A = occurrence of taxon A × 100/occurrence of all taxa. These values were used to determine the prevalence of each taxon in comparison with the total fungal taxa obtained from the Antarctic lakes. Sample-based data were used for the calculation of diversity and similarity indices, species accumulation curves and estimators of species richness. For measurement of the species diversity, we used the richness and dominance categories provided by two indices: (1) Shannon H = ∑ni/n ln (ni/n) and (2) Simpson's index = ∑(ni/n)2. The similarities among fungal taxa from different areas were estimated using a Bray–Curtis (B) measure and Sorensen coefficient (QS). The diversity, similarity and the correlation between water physico-chemical parameters, fungal density (CFU L−1), Shannon H and Simpson's indices, and the principal components analyses (PCA) calculations were carried out using the past computer program version 1.90 (Ryan et al., 1995). Species accumulation curves and the estimators of species richness were inferred using estimates 8.0 (Colwell, 2006). Further information for all of these indices can be found in the study by Magurran (1988). All results were illustrated by plotting the number of species against the CFU L−1, which provide a quantitative analysis (density) of fungal viable cell occurrence in a determinate habitat.
A total of 90 water samples from six different lakes were analysed, and the average counts of each fungal morphospecies ranged from 0.02 to 12.09 CFU L−1. Previous studies of Antarctic fungi identification (Rosa et al., 2009, 2010; Santiago et al., 2012) have been shown that these extremophile fungal groups require specific conditions of temperature, humidity, pH and photoperiod to produce asexual structures in conventional mycological media, which are used to morphological identification; however, most of these fungi did not produce conidia or spores and the morphological identification is not possible. To solve this problem, different studies have been carried out based on rDNA blastn results to characterise the Antarctic fungi (Connell et al., 2006; Arenz & Blanchette, 2010; Blanchette et al., 2010; Rosa et al., 2010; Arenz et al., 2011; Farrell et al., 2011; Santiago et al., 2012). For this reason, all 128 fungal isolates obtained were identified based on sequencing of the ITS region and blastn comparisons with the GenBank database. A total of 31 distinct operational taxonomic units (OTUs) were identified in this study based on their unique sequences. The taxa identified in this study belong to the phylum Ascomycota, subphylum Mortierellomycotina (traditional Zygomycota phylum), and one zoosporic fungus (Oomycota, Straminopila). In addition, three taxa did not have match with any species deposited in GenBank. These taxa were identified only by order or family level and require additional taxonomy analyses to suggest their appropriate taxonomic placement (Table 1). In addition, phylogenetic trees (Fig. 2) were constructed to illustrate the relationship among unidentified individual sequences to their closest relatives retrieved from GenBank database.
Although several fungal taxa were isolated from each lake, a few species dominated the assemblage in each case (Table 2). The most frequently isolated taxa were Geomyces pannorum, Thelebolus sp. and Mortierella sp., which were found in six, four and three of the sampled lakes, respectively, and displayed the highest total CFU L−1. In contrast, 12 taxa (37.5%) were found in only one lake with a density of ≤ 0.05 CFU L−1, thereby representing the minor assemblage of the fungal taxa of the lakes. Comparison of the fungal assemblages showed that Cadophora cf. luteo-olivacea, Cladosporium sp., Cosmospora cf. vilior, Fontanospora sp., Gibberella moniliformis, Heydenia sp., Microdochium nivale, Penicillium cf. verrucosum, Pleosporales sp., Phoma fimeti, Saprolegniaceae sp., and Trichoderma cf. longibrachiatum occurred in only one lake each. In addition, the taxon Antarctomyces psychrotrophicus, found in two lakes, is considered endemic to Antarctica.
Table 2. Fungal species and density (CFU L−1) collected at five lakes of Antarctica
The diversity indices, water physico-chemical parameters and fungal CFU L−1 are shown in Table 3. The lakes displayed average values of total fungal counts ranging from 6.5 (in Wanda point) to 62 CFU L−1 (in Crater Lake). In general, the fungal compositions of Antarctic lakes displayed low richness (Shannon H =1.83) and high dominance (Simpson's = 0.73) indices. The Lakes Stain House and Wanda point displayed the highest Shannon values. In contrast, Crater Lake and Brazilian Refuge II had low Shannon values. The Simpson's values were high for Stain House and Wanda point lakes and low for Brazilian Refuge II and Crater Lake. Additionally, the fungal species accumulation curves and bootstrap richness estimators are shown for each lake (Fig. 3), which fell within the 95% confidence intervals, suggesting that our samples were representative of the species richness of these fungal assemblages. Figure 4 shows the Sorensen and Bray–Curtis coefficient comparisons among the fungal assemblages of the six lakes. In the Sorensen index, the most similar assemblages were found at Macchu Picchu and Wanda point lakes (QS =0.44); in contrast, Agat Point Lake displayed the lowest similarity in comparison with the other five sites. The Bray–Curtis index demonstrated that the lakes that had the most similar species assemblages were the Macchu Picchu Lake and Crater Lake (B =0.4), and the assemblages of Wanda point and Agat point displayed the most differences in relation to the other assemblages. However, the bootstrap values among some lakes fell below than 50%. The PCA among physico-chemical parameters of water and fungal diversity revealed that chlorophyll a had a positive correlation with Shannon index (Fig. 5).
Table 3. Physico-chemical parameters of water, total counts of CFU L−1 and diversity indices of fungal communities in six Antarctic lakes
Within the Antarctic environment, lakes represent unexplored habitats that may be used to study microbial diversity and ecology in extreme conditions. According to Ellis-Evans (1996), the knowledge of fungal communities in Antarctic lakes is scarce and, until now, only a few genera have been described, such as Penicillium, Cladosporium, Arthrobotrys, Geomyces, Mortierella and several mycelia sterilia, most of which have also been isolated from Antarctic soil. Based on the molecular characterisation, our results demonstrated that a significant number of fungal genera and some unidentified taxa exist in the lakes of the Antarctica Peninsula.
The genus Geomyces includes only 11 known species. Although the genus has a wide distribution (http://www.mycobank.org/), G. pannorum (previously known as Chrysosporium pannorum) is one of the most frequently reported fungal species in Antarctica. Geomyces pannorum is a keratinophilic and psychrophilic fungus that has a ubiquitous distribution in the soils of Arctic, Alpine, temperate and Antarctic regions (Mercantini et al., 1989). In Antarctica, G. pannorum was isolated from thalli of seaweeds Adenocystis utricularis and Desmarestia anceps (Loque et al., 2010), as an endophyte of Colobanthus quitensis (Rosa et al., 2010), and is associated with mosses (Tosi et al., 2002). According to Vishniac (1996) and Arenz et al. (2011), this fungus occurs in many places of Antarctica. This species is halotolerant (Poole & Price, 1971) and moderately cellulolytic (Kuthubutheen & Pugh, 1979). Geomyces spp. have the ability to colonise and utilise different carbon sources such as leaf litter, and increase in abundance at lower temperatures (Arenz et al., 2006, 2011). According to Kirk et al. (2001), the Mortierella genus (subphylum Mortierellomycotina) occurs mainly in the soil of different ecosystems, including terrestrial habitats of Antarctica (Adams et al., 2006). In Antarctica, Mortierella species have been detected in soil (Bridge & Newsham, 2009) and associated with mosses (Tosi et al., 2002). The significance of fungi in Antarctic lakes is not fully clear, because data on the numbers of species in natural habitats are limited. According to Wong et al. (1998), in lakes of temperate and tropical areas, fungi are involved in the degradation of different organic matter sources. The presence of Geomyces and Mortierella species in soil and associated with plants in Antarctica may explain their abundance in the littoral region of Antarctic lakes. Fifteen taxa occur in only one lake at a low density, representing singleton taxa, which compose the minor portion of fungal communities. Earlier studies on fungi obtained in Antarctica (Azmiá & Seppelt, 1998; Adams et al., 2006; Rosa et al., 2009; Loque et al., 2010; Arenz et al., 2011) as well the present study show that there are few abundant species and a high occurrence of singleton taxa. All fungal species isolated from these lakes may have an important ecological role in Antarctica, such as the primary decomposition of organic material. According to Ellis-Evans (1996), with a few exceptions, most lakes in the Antarctic Peninsula are shallow systems (< 10 m deep) that are ice-free for only a few weeks each year and are notably nutrient poor. Increased levels of vegetation (Deschampsia antarctica, C. quitensis, mosses and lichens) and the proximity to penguin rookeries areas are probably also an important factor in higher fungal abundance in some Antarctica lakes. These substrates may be important sources of food for fungal mats in Antarctic lakes.
Few studies have been conducted to detect endemic or adapted fungi in Antarctica. According to Tosi et al. (2002), it is difficult to determine which fungi found in Antarctica are endemic to the region. According to Brunati et al. (2009), the majority of the fungi described in this study are cosmopolitan and cold tolerant, but some appear to be endemic species. The genus Thelebolus comprises 21 species (www.mycobank.org), and according to Kobayasi et al. (1967) and Montemartini et al. (1993), Thelebolus species, mainly Thelebolus microsporus, have often been isolated from Arctic and Antarctic climate zones. An investigation of the fungal diversity in mat material of some Antarctic lakes conducted by de Hoog et al. (2005) revealed that T. microsporus was one of the most abundant species found in association with skuas, petrels and other birds. In addition, Thelebolus was the predominant genus found by Brunati et al. (2009) in benthic mats of Antarctic lakes. The genus Antarctomyces is represented only by A. psychrotrophicus, which was isolated from Antarctic soil (Stchigel et al., 2001) and from the thalli of the seaweed A. utricularis (Loque et al., 2010). According to Arenz et al. (2006, 2011), A. psychrotrophicus is endemic to Antarctica; our results confirm that this species is able to survive in Antarctic lakes. In contrast, the species C. cf. luteo-olivacea, Cadophora malorum, Davidiella tassiana, G. pannorum, Mortierella cf. alpina and Thelebolus cf. microsporus have been isolated in cold ecosystems, such as Arctic, Alpine and temperate regions, and seem to have a wide distribution. In addition, the Saprolegniaceae sp. UFMGCB 3678 (Oomycota, kingdom Chromista) was isolated from water samples of Crater Lake. According to Petrisko et al. (2008), species of Saprolegniaceae often can be found on decaying animal and plant debris in freshwater habitats in worldwide, but few works have been published about the their diversity and ecology role. The presence of the Saprolegniaceae in Antarctic lake suggests that some species of this family can have interesting adaptation to survive in extreme conditions.
A total of 13 taxa showed low similarity with sequences of fungi deposited in GenBank database or high similarities with unidentified fungi. These taxa showed different levels of sequence similarity with taxa identified in genus (Cladosporium, Fontanospora, Helgardia, Heydenia, Microdochium, Mortierella, Penicillium, Phaeosphaeria, Pseudeurotium, Thelebolus), family (Saprolegniaceae) and order (Helotiales and Pleosporales) levels, which suggests that these fungi could represent new fungal species. However, more morphological and phylogenetic taxonomic studies will be necessary to determine the correct identification of these fungal species as well as describe whether or not they are new taxa.
Our diversity results, based on reported OTUs, indicated that aquatic mats are dominated by a relatively small number of fungal genera. This result was also observed by Brunati et al. (2009) in relation to benthic fungal communities in Antarctic lakes. According to Vazquez & Stevens (2004), environments with low species richness are thought to favour generalist ecological strategies over specialisation. Our results suggest that the dominant fungal genera, Mortierella and Geomyces, have the capability to resist extreme conditions as well to use the scarce available nutrients present in Antarctic lakes. According to Ellis-Evans (1985), the pattern of high numbers/low diversity of fungal communities in Antarctic lakes may reflect the permanently low temperature and extreme isolation factors. In addition, the water of the six lakes sampled displayed a broad range of temperature, conductivity, salinity and pH, demonstrating that fungi from these habitats have specific mechanisms to survive in extreme conditions. Chlorophyll a concentration represents a measure of the lakes trophic state; the low values of chlorophyll a demonstrated that all Antarctic lakes sampled are oligomesotrophic according to the definition of Lefranc et al. (2005). However, the Antarctic lakes with high values of chlorophyll a also displayed the highest Shannon index values. The same pattern was observed for fungal taxa that were most numerous when chlorophyll a concentration reached its maximum in an oligomesotrophic lacustrine temperate lake, as observed by Lepére et al. (2006). The extreme physico-chemical conditions of these Antarctic lakes have probably influenced the diversity of the fungal community. However, according to Hirano (1965), the low microbial diversity in Antarctic lakes might be more apparent than real because few lakes had been studied in detail. Our results may support this hypothesis because several fungal taxa were identified at genus, family or order levels, which may represent new and/or adapted species able to survive in the extreme environmental conditions of Antarctica. In addition, as observed by Ellis-Evans (1996) in lakes and Fell et al. (2006) in the soils of Antarctica, our results also show a presence of low fungal diversity that suggests the presence of a short food web in the Antarctic lakes, including possible symbionts (endophytic fungi), weak and strong saprophytes and parasite/pathogens genera, and unknown taxa, which may have interesting ecological roles in extreme conditions. Our results suggest that the fungal diversity in Antarctic lakes may be relatively low, but the complexity of interactions between the different taxa matches that of other aquatic ecosystems.
Studies describing the diversity of microbial communities in Antarctica have led to the discovery that different microorganisms have adapted to survive in extreme conditions. The results of this study indicate that the fungal species composition of lakes in the Antarctic Peninsula include taxa able to survive in an environment featured by low temperature, different salinities, different pH conditions and low availably of nutrients. The lakes of Antarctic Peninsula provide a unique environmental laboratory to study the origin and activities of fungal communities in contrast to the highly complex communities of tropical and temperate lakes. Together, the data obtained in this study demonstrate that Antarctic lakes of maritime Antarctica are an interesting reservoir of aquatic fungi, which may be a model for further ecological and evolutionary studies in extreme conditions. The origin, biology and ecological role of aquatic fungi web of Antarctic lakes deserves more research, which can be based on new collections and diversity characterisation, sequence analyses of different genes and microscopy studies and their relationship with other organisms, such as microalgae, bacteria and invertebrates.
This study had financial and logistic support from the Brazilian Antarctic Program, Marine of Brazil (in memoriam of Roberto L. dos Santos Carlos A. V. Figueiredo), and Luiz C. Consiglio and Beatriz Boucinhas from the Clube Alpino Paulista. This work is part of the API activity 403, contributes to Microbiological and Ecological Responses to Global Environmental Changes in Polar Regions and INCT Criosfera 704222/2009. We acknowledge the financial support from Fundação de Amparo a Pesquisa do Estado de Minas Gerais (FAPEMIG) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). We thank an anonymous reviewer for helpful comments to improve the quality of the manuscript.