This is the first study focusing on methanogenic archaea in samples collected from lakes nearby Czech polar Station JG Mendel on James Ross Island. Moreover, the lake Long Island from the Long Island was sampled just for the first time, and limnological data available for some other lakes like Lagoons Mesa 2 a 3 or Solorina 42 and 60 are those obtained during this research in the 2019 expedition.
Low temperature and ice cover during the sampling, thus reduced light availability could probably affect the final DNA yield from sediments of lakes BIB, CYA, MUD, and PHO. Finally, samples from all these lakes were not used for subsequent 16S rRNA analysis. Both the low intensity of irradiation and low temperature hinders the development of primary production which is supported by cyanobacteria and microscopic algae here. Restriction of the primary producers also limits the presence of other microorganisms depending on them as has been reported for some antarctic lakes (Laybourn-Parry and Pearce 2007; Nedbalová et al. 2013).
Bacteria were a dominant fraction of prokaryotic microorganisms in the samples, archaea rarely exceeded 1% of the total prokaryotic community. Exceptions were lakes ESM and LM1 with values to be 8.8% and 1.2%, respectively. The higher proportion of archaeal, namely methanogenic sequences in the lake Esmeralda (ESM) has been caused probably due to the lower proportion of Deltaproteobacteria which usually compete with methanogens for H2 (Karr et al. 2006). Generally, the low contribution of archaeal sequences has also been observed in many other studies of antarctic freshwater lakes (Bowman et al. 2000; Mulyukin et al. 2014; Gugliandolo et al. 2016; Chaya et al. 2019).
On the other hand, due to the development of next-generation sequencing, along with using more effective primers and a growing dataset of newly described sequences we were able to detect the prevalence of two relatively newly proposed archaeal classes Woesearchaeia (Woesearchaeota) and Nitrososphaeria (Thaumarchaeota) representatives in some lakes. Currently, Woesearchaeota is widely found in diverse environments (Liu et al. 2018) and belong usually among the most frequent and abundant archaea found in lakes (Ortiz-Alvarez and Casamayor 2016; Juottonen et al. 2020; Tóth et al. 2020). Woesearchaeota can enable carbon and hydrogen metabolism under anoxic conditions, which might associate with symbiotic and/or fermentation-based lifestyles (Castelle et al. 2015; Gründger et al. 2019).
Most members of the Thaumarchaeota are chemolithoautotrophic ammonia-oxidizers and may play important roles in nitrogen and carbon biogeochemical cycles (Pester, Schleper and Wagner 2011; Stieglmeier et al. 2014). Nitrification by ammonia-oxidizing archaea (AOA) contributes to N2O production which might have implications for climate change (Santoro et al. 2011). Many Thaumarchaeota are marine and live in the open ocean (Schleper and Nicol 2010), however, AOA representatives were also found in freshwater aquaculture ponds (Lu et al. 2021), rice paddy soils, reservoir sediments (Wang et al. 2014), rivers (Li et al. 2018) and lakes (Ortiz-Alvarez and Casamayor 2016; Juottonen et al. 2020).
Although the qPCR approach was not used in Antarctic lake sediment analyses until now, our results are comparable with those from shallow freshwater arctic lakes in Alaska where the mcrA gene occurred in orders of 102–104 the gene copies per one gram of the sediment (Matheus Carnevali et al. 2015). Slightly higher densities of the gene copies 105–106 were found in freshwater lakes in plateau Yunnan in China (Yang et al. 2020).
Results of 16S rRNA sequencing also show that methanogenic archaea form not only a small part of the total prokaryotic diversity but also within the archaeal community in the most of studied lakes. This trend has been reported already before in some papers on prokaryotic communities in the sediments of the antarctic lakes (Sjöling & Cowman 2003; Tang et al. 2013; Gugliandolo et al. 2016). The smaller proportion of methanogens within the prokaryote/archaeal community may probably indicate the prevalence of terminal acceptors used by other microorganisms (Purdy, Nedwell and Embley 2003).
Contrary to the sequences retrieved by 16S rRNA analysis, many more reads belonging to methanogens were found with mcrA gene sequencing. This finding is, however, not surprising because of targets the specific function gene which is common for methanogenic and methane-oxidizing archaea (Steinberg and Regan 2009). Nevertheless, the presence of taxons was almost the same in both of the used approaches except for the detection of Methanothermobacter sp. and Methanoculleus sp. sequences, and the absence of the Methanolobus sp. in mcrA gene sequencing.
On the other side, the mcrA gene sequencing confirmed a high proportion of sequences belonging to Methanoperedens-like archaea ANME-2d the anaerobic methanotroph mostly in BLA and SOL42 lake sediments. "Candidatus Methanoperedens nitroreducens" is a candidate species of methanotrophic archaea that couples anaerobic methane oxidation to denitrification (DAMO archaea) (Raghoebarsing et al. 2006; Haroon et al. 2013). Ca. Methanoperedens nitroreducens" carries out reverse methanogenesis and harbors the key enzyme methyl-coenzyme M reductase (MCR), which catalyzes either the last step of methane production in methanogens or the first step of methane oxidation in anaerobic methanotrophic archaea (Guerrero-Cruz et al. 2018). This microorganism is present at oxic-anoxic interfaces in a wide range of aquatic environments and man-made ecosystems, such as lakes, rivers, paddy fields, and wastewater treatment systems (Ding et al. 2015; Guerrero-Cruz et al. 2018; Ding and Zeng 2021) but no presence of this archaeal organism has been reported for antarctic lakes until now. To our current knowledge, the only ANME-3 sequences typical for sulfate-dependent anaerobic alkane-oxidizing archaea (Wang et al. 2021) were detected in cold seepages in the Weddel sea surrounding James Ross Island (Niemann et al. 2009). Both the ANME-2d and ANME-3 clades harbor mcrA gene (Cui et al. 2015) and belong to the same phylum Euryarchaeota and order Methanosarcinale, however, most environmental sequences of the ANME-2d group are derived mostly from freshwater sediments while members of the ANME-3 clade dominate methane-rich arctic mud volcanoes (Wang et al. 2021). The occurrence of this genus might indicate higher concentrations of nitrates in the water of the particular lakes, however, our analyses have shown almost zero concentrations of nitrates in all studied lakes (see Table 2).
Methanothrix and Methanosarcina, as the most dominating methanogenic genera in our samples, have also been found to be dominant in other antarctic lakes (Purdy, Nedwell and Embley 2003; Karr et al. 2006; Chaya et al. 2019). Genus Methanobacterium occurs mostly in an active layer of permafrost (Rivkina et al. 2007; Barbier et al. 2012; Wang et al. 2020), however, in Antarctica, this taxon has been captured for the first time to such a large extent.
Similarly, sequences of potentially thermophilic representatives belonging to Methanothermobacter and Methanomassiliicoccales (Thermoplasmata) were recorded for the first time in antarctic lakes. The species within the genus Methanothermobacter (Euryarchaeota, Methanobacteriales) are thermophilic, hydrogenotrophic methanogens and grow best at temperatures between 55°C and 65°C. However, Mickol et al. (Mickol, Laird, and Kral 2018) have suggested that thermophilic representatives of the Methanothermobacter genus can survive long-term extreme conditions in permafrost, reducing their metabolism while enduring the more convenient conditions. Members of both orders Methanobacteriales and Methanomassiliicoccales were already found in peat bogs in the Siberian tundra (Grodnitskaya et al. 2018).
It is too difficult to say which kind of methanogenesis could be prevalent in the studied lakes because many sequences were not aligned with any genus level. Generally, sequences of the methanogens which might be involved in all three types of methane production were found in our samples. Genera Methanobacterium, Methanothermobacter, Methanoculleus, and partially Methanosarcina are involved in hydrogenotrophic methanogenesis, while acetoclastic methanogenesis is mainly dominated by species of Methanosarcina and Methanothrix genera, respectively. Hence, we can assume the prevalence of hydrogenotrophic methanogenesis in lakes LA1, KAT1, and ESM. On the other hand, the acetoclastic pathway should dominate in the lakes LM2 and LM3 where the dominance of the genera Methanohrix a Methanosarcina was found.
Methylotrophic methanogens all belong to the family Methanosarcinaceae, except the genus Methanosphaera, which belongs to the order Methanobacteriales. The methylotrophic growth of Methanosphaera species is H2-dependent - they are obligate methylotrophic and hydrogenotrophic methanogens that are specialized to reduce methyl groups with H2. The metabolism of Methanosphaera sp. is restricted to methanol (Liu and Whitman 2008). More recently, methanogens that also reduce methanol with H2, belonging to the order Methanomassiliicoccales have been described (Dridi et al. 2012; Paul et al. 2012). These methanogens occur in animal gut systems but also have frequently been detected in other anoxic environments such as fens (Söllinger et al. 2016), mangrove sediments (Li et al. 2016), digesters (Wilkins et al. 2015), and anoxic paddy soils (Ji, Liu and Conrad 2018). Methanogens of the order Methanomassiliicoccales are probably even more widespread since they have only recently been added to sequence databases and probably were previously recorded as Thermoplasmatales-related Euryarchaeota. However, their role in carbon flux is presently unknown (Conrad, 2020). Due to the higher abundance of sequences belonging to Methanomassiliicoccales and Methanosarcina sp., methylotrophic methanogenesis is suggested to be prevalent in Lake SOL42. Nevertheless, as methanol is a product of pectin degradation, and pectin turnover in lake sediments is slow, methanol is a marginal methanogenic precursor in these sediments (Lovley and Klug 1983) and also in anoxic rice field soils (Conrad and Claus 2005). Moreover, Methanosarcina species can produce methane by using all three above-described pathways (Whitman, Bowen and Boone 2014).
Several factors like higher temperature, incomplete organic matter degradation, or the presence of syntrophic acetate oxidation may cause a higher proportion of hydrogenotrophic methanogens. In Yamal tundra lakes (Subarctic Eurasia), the rates of hydrogenotrophic methanogenesis appeared to be higher in the sediments of deep lakes than in those of the shallow ones (Savvichev et al. 2021). On the contrary, acetoclastic methanogenesis flourishes at a lower temperature, complete degradation of the organic matter, and the presence of acetogenic bacteria (Conrad, 2020). Prevalence of the hydrogenotrophic methanogenesis in freshwater lakes of maritime Antarctica has already been suggested in a study by Ellis-Evans (Ellis-Evans 1984). However, to better understand the physiological pathways, ecological interactions, and ecological roles of the present methanogens it would require culturing the microbes as suggested by Lascar et al. (Laskar et al. 2018).
Diversity analysis reveals a similarity between lakes located in two areas: Lagoons Mesa and Solorina but individual lakes from one area shared much higher similarity with the lakes from another area rather than with the lakes within the same area. This finding is supported further by a comparison of samples from Esmeralda and Katia 1 lakes which are several kilometers apart on different islands, suggesting that a lake location will be not the main factor influencing the diversity of the methanogens.
Results can differ even in samples taken from one lake as evident from prokaryote composition and mcrA gene copies number in samples taken from two sites in the White lake - WHI (J) and WHI (S). To get more complex information about the microbial diversity would require taking much more samples during various periods of the year from one lake. However, in Antarctica, this approach is hardly possible, at least due to financial reasons and impossible sampling during the rest of the year.
Nevertheless, for further investigation of the methanogenic diversity, we recommend sampling lakes from Lagoons Mesa and Solorina valley that showed both the highest diversity and number of mcrA gene copies. Many lakes located within these two areas allow for collecting more numerous samples. In addition, Esmeralda lake at Vega Island could be a further subject of the microbiological research due to distinct prokaryotic diversity and potentially high diversity of methanogens.