3.3. Structural diversity of microbial communities
3.3.1. Environmental microbiome
The taxonomic composition of sequences from environmental samples identified 10 phyla (>1%). The most abundant representatives of the community were Proteobacteria (55.84%), Firmicutes (10.95%), Acidobacteria (9.77%), Verrucomicrobiota (5.64%), Actinobacteria (5.78%), Desulfobacterota (4.86%), Bacteroidetes (2.0%), Chloroflexi and Myxococcota (both representing 1.87%), and Cyanobacteria (1.38%) (Figure 3C and Supplementary Table 8).
The high abundance of the Proteobacteria phylum (55.84%) (Figure 3C) corroborates other studies carried out in sediments associated with water bodies (Pascault et al. 2014, Zarraonaindia et al. 2015). Members of the phylum play an essential role in biogeochemical cycles, mainly in the nitrogen cycle (Spain et al. 2009). Members of the phylum Firmicutes (10.95%) are linked to the nitrogen cycle and actively participate in the replacement of this compound in the environment (Anderson et al. 2018; Jung et al. 2013). Acidobacteria (9.77%) are physiologically diverse, and the cultivated members of this phylum are heterotrophic and metabolize a wide variety of complex carbohydrates and nitrates (Kielak et al. 2016). The presence of other phyla such as Actinobacteria, Verrucomicrobiota, Bacteroidetes, Chloroflexi, and Cyanobacteria are important bacterioplankton participants that make up freshwater ecosystems (Crump and Hobbie 2005; Allgaier and Grossart 2006; He et al. 2017).
The Cyanobacteria phylum was identified among the ten most abundant groups in environmental samples; however, at a more specific taxonomic level, such as genus, representatives of the cyanobacteria phylum showed low diversity and abundance in relation to other groups. Through the taxonomic composition, only 17 genera were classified, which reveals the difficulty of identifying representatives of this phylum in analysis that encompass the new generation environmental sequencing, even in a freshwater environment and a photic zone, which, theoretically, is an environment that favors the adaptation of photosynthetic species (Richardson et al. 2018).
Cyanobacteria are one of the most diverse groups in the bacterial domain (Gaysina et al. 2018) and, even with advanced next-generation DNA sequencing techniques, there is no high-level sensitivity to the point of recognizing or approaching the total diversity of low-abundance groups in environmental samples (Tromas et al. 2017). In addition, the databases still have a large sequence gap when it comes to cyanobacteria, and Alvarenga et al. (2017) demonstrated that among all genomic sequences deposited in the NCBI, only 1% is related to the phylum Cyanobacteria.
In general, between the sampling points (Pedra Caída and Farinha River), the richness of ASVs measured by the Chao1 estimator, the diversity estimated by Shannon, and the number of observed ASVs (Figure 1A) did not demonstrate an effect on the estimated indices (Mann–Whitney, W = 8, p = 0.1; t-test, t = −1.4434, df = 6.3, p = 0.1; Mann–Whitney, W = 8, p = 0.1; respectively).
The non-metric multidimensional scale (nMDS) based on dissimilarities calculated using the Bray–Curtis metric showed that there is dissimilarity between the microbial communities found in the regions of Rio flour and Pedra Caída (Figure 1B). Based on the multivariate permutation analysis of variance (PERMANOVA), the variation in the structure of the microbial community is, in part, explained by the sample points analyzed (ANOSIM, R = 0.356, p = 0.01).
The analysis carried out through the envifit revealed that part of the microbial community variation can be explained by environmental factors characteristic of each sampled region. Parameters such as temperature (envifit; R2 = 0.65; p = 0.01), NO−2 (envifit; R2 = 0.58; p = 0.02), and NO-3 (envifit; R2 = 0.68; p = 0.02) had a significant effect (Supplementary Table 10) on microbial communities, as variations in environmental temperature and nutrient availability can modulate the community and delimit microbial niches (Samad et al. 2020).
3.3.2. Enriched microbiome
Studies that characterized the cyanobacterial consortia aimed to analyze their ability to degrade compounds such as crude oils (Hamouda et al. 2016) and hydrocarbons from petroleum (Ichor et al. 2016), or selection of cyanobacteria that can be used biotechnologically for the treatment of effluents (Arias et al. 2017). Unlike the aforementioned approaches, these consortia aimed to favor the growth of photosynthetic microorganisms, selecting them from culture in culture medium (BG-11) for further taxonomic analysis. Thus, it is possible to enrich and study under sampled photosynthetic organisms in sequencing environmental samples.
The microbial composition of the consortia presented seven phyla (>1%) of the bacterial domain, according to the classified sequences. The most abundant representatives were Proteobacteria (45.2%), Cyanobacteria (26.5%), Bacteroidetes (18%), Planctomycetes (3.3%), Verrucomicrobiota (2.5%), Acidobacteria (2.2%), and Armatimonadota (1.8%) (Supplementary Table 9).
Photosynthetic microbial communities have a strong relationship with heterotrophic organisms because of the different ecological roles played by these groups. The carbon secreted by the phytoplankton community promoted the growth of the associated microbial communities. In natural environments, such as Chapada das Mesas, the relationships between cyanobacteria and the heterotrophic community tend to be balanced, which guarantees local ecological stability (Xu et al. 2018). In artificial communities, such as enriched cultures, the presence of these photosynthetic microorganisms makes it possible for the heterotrophic community to grow in minimal culture medium, as they act as primary producers in the microcosm.
Microbial consortia come from the enrichment of environmental samples with the objective of favoring certain target groups through the use of selective culture media for a determined objective. Mixed populations in bacterial cultures may seem undesirable objectives; however, with this approach, it is possible to analyze functions that are difficult to perform or even impossible for individual strains or species (Brenner et al. 2008). Cyanobacteria often have heterotrophic microbes in symbiotic association with their cells, and it is assumed that these associations exist even among the oldest known life forms, which represent important ecological interactions during the evolutionary history of this group (Alvarenga et al. 2017).
By analyzing the taxonomic level of the genera of photosynthetic microorganisms, it was possible to identify a total of 34 different microbial groups, including organisms of the phylum Cyanobacteria and organisms classified as “Chloroplast,” which includes the eukaryotic representatives of microalgae, in the samples of the consortia of both the Complex of Pedra Caída and in the Farinha River region. The most abundant genera were “Chloroplast” (45.60%), Synechocystis_CCALA_700 (10.10%), Geminocystis_PCC-6308 (9.38%), YB-42 (7.38%), Ancylothrix_8PC (4.86%), JSC-12 (3.99%), Cyanobacteria (3.62%), Phormidium_IAM_M-71 (2.91%), Cyanobacterium_PCC-10605 (2.01%), Pleurocapsa_PCC-7327 (1.71%), SU2_symbiont_group (1.12%), and HAVOmat113 (1.06%) (Figure 2, Supplementary Table 11 and Supplementary Table 12).
Some groups of photosynthetic microorganisms were identified in the environmental microbiome, but at low abundance (Figure 2). The park’s photosynthetic diversity has microbial genera that would not be identified using only the cultivation-independent approach, because these groups may be in low abundance in these water bodies, limitations of the technique itself, among others. When the culture-dependent approach was used, cyanobacterial diversity increased by 65%. Some genera such as Acaryoclhoris, Aetokthonos, Calotrhrix, Chalicogloea, Cyanobium, Gastranaerophilales, Gleobacter, Microseria, Myxosarcina, Scytonema, and SepB.3 were identified only in the environmental analysis and probably failed to compete under controlled cultivation conditions with other microorganisms that were enriched in the consortia.
As the most abundant of the two treatments, the group classified as “Chloroplast” was taxonomically assigned as representative of the phylum Cyanobacteria. The high abundance of this less specifically classified group may represent a need to update the SILVA 138 database for this group (Lesack and Birol 2018), because many phytoplankton representatives have not yet been identified, and they may also be related to microalgae that were probably enriched in the consortia and are present in the natural environment, as there is a strong association between the presence of these other phytoplankton components and the enrichment of cyanobacteria (Perera et al. 2019).
The identification of the genus Pantanalinema in the consortia is interesting, as this group was first recorded in saline-alkaline lakes and was characterized by its ability to survive and produce biomass at a pH ranging from 4 to 11 (Vaz et al. 2015). It was mainly described based on the phylogenetic information of its 16S rRNA sequences (Genuário et al. 2018). Microorganisms from this group have already been identified in Amazonian rivers and in the Cerrado Maranhense, are part of a group of bacteria that cannot be cultivated in isolation, and later recovered from an association with cnidarians (Genuário et al. 2018; Ferreira et al. 2021). The first genome of the genus was recently characterized (Ferreira et al. 2021), and the analysis revealed a still unknown environmental and biotechnological potential. Brazil has increased the number of newly described cyanobacterial taxa, highlighting regions that have fewer studies, such as the North and Northeast regions of the country (Menezes et al. 2015).
Environmental sequencing is an important tool for the study of microbial communities, as most microorganisms cannot be cultivated in vitro (Locey and Lennon 2016). However, combining this tool with classical microbiology techniques, through the use of culture media to enrich target organisms, has become an alternative for the study of bacterial groups that would not be identified only through the sequencing of the environmental 16S rRNA gene. In this study, 17 genera of photosynthetic microorganisms were identified in the environmental taxonomic annotation, while after enrichment with selective medium, it was possible to identify 34 different photosynthetic groups. Photosynthetic microorganisms, when associated with microbial consortia, can in many cases act as main primary producers, helping the growth of other microorganisms and creating a community that would not be cultivated independently (Rappé and Giovannoni 2003; Alvarenga et al. 2017).
The diversity indices (alpha and beta) were also calculated for the consortium samples to analyze the structure and composition of the microbial community enriched in the culture medium. The alpha diversity, analyzed between the sampling points (Pedra Caída and Farinha River), revealed that the richness of ASVs measured by the Chao1 estimator, the diversity estimated by Shannon, and the number of observed ASVs (Supplementary Figure 2) did not demonstrate an effect on the estimated indexes (t-test, t = −0.3274, df = 7.2985, p = 0.7732; t-test, t = −0.23153, df = 6.848, p =0.7526; t-test, t = −0.29907, df = 7.2977, p = 0.7732, respectively).
The non-metric multidimensional scale (nMDS) based on dissimilarities calculated using the Bray-Curtis metric demonstrated dissimilarity (Supplementary Figure 2) between the microbial communities found in the regions of Rio flour and Pedra Caída (Figure 1B). Based on the multivariate permutation analysis of variance (PERMANOVA), the variation in the structure of the microbial community is, in part, explained by the sample points analyzed (ANOSIM, R = 0.3667, p = 0.005). This dissimilarity corroborates the diversity directly analyzed in the samples of origin of the enrichments, since in the environmental microbiome, the composition of the communities also varied according to the origin of the samples, which explains part of the variation in the microbial communities in the enrichments.
3.3.3. Comparison of environmental and enriched microbiomes
Sequencing was carried out for both the environmental samples collected in the Chapada das Mesas National Park and the samples cultivated and enriched for photosynthetic microorganisms, in order to analyze the possible subsampling of this group of photosynthetic microorganisms in these environments when analyzed only by independent cultivation techniques. A total of 4,792 AVSs were identified (Figure 3A), with only 2.65% sharing both in the environmental samples and in the microbial consortia.
In samples obtained from environmental sediment, 2,385 ASVs were found exclusively in this treatment. Only 2,280 ASVs were identified in the cultured samples after enrichment. This indicates that 47.57% of the microbiota could be cultivated when we used a minimal and selective culture medium for photosynthesizing microorganisms.
The large number of ASVs that were identified exclusively in the cultures (Figure 3A) led us to suppose that the microbes cultivated in our experiment may be representative of the rare portion of the sediment microbiota, that is, likely microorganisms that would not be identified only with sequencing of environmental samples owing to their low relative abundance, and therefore, require a very high sequencing coverage and, in practice, often unfeasible to be represented.
It is also important to note that there was a significant increase (p < 0.05) in the abundance and diversity of organisms of the Cyanobacteria phylum in samples that were cultivated in BG-11 culture medium, a minimal medium that allows the growth of photosynthetic microorganisms (Figure 3B). These are also part of this portion, which is not very abundant in the environment and would not be identified only with the sequencing of environmental samples. Therefore, enrichment for specific groups and/or in low abundance can be a good alternative for identifying a better view of the microbiota in a broader way.
Some phyla such as Myxococcota, Firmicutes, Actinobacteria, and Desulfobacterota were present in the environmental samples and were not identified as phyla with an abundance greater than 1% in the microbial consortia (Figure 3C). With the enrichment of the Cyanobacteria phylum in the intercropping samples, other phyla such as Bacteroidota, Armatimonadota, and Planctomycetota also showed an increase in their respective abundances compared to environmental samples (Figure 3C), suggesting a greater ecological interrelationship between these phyla.
3.3.4. Functional inference
The functional inference analysis performed in FAPROTAX aimed to explore the possible metabolic functions associated with environmental samples and samples grown in consortia.
Regarding the functions (Fig. 4) identified in the environmental samples, there were important ecosystem processes, such as highlighted methanotrophy (p < 0.05). Bacteria with this metabolic capacity use methane as a source of carbon and energy, thus mitigating net emissions of methane from natural sources (Crevecoeur et al. 2019), which may be associated with the presence of microorganisms belonging to the genera Methylomonas, Methylocaldum, Crenothrix, and Methylomonas, identified in environmental samples. Methylotrophic processes were also highlighted (p < 0.05) in samples collected in the park, associated with methanol oxidation, a process related to the identification of microorganisms from the Methylophilaceae, Methylophilus and Paracoccus groups. Furthermore, a function related to the degradation of hydrocarbons was also identified (p < 0.05), a process associated with the genus Halomonas, which is known for its metabolic capacity for the degradation of low- and high-molecular-weight PAHs (Govarthanan et al. 2020), naphthalene (Cheffi et al. 2020) and crude oil (Neifar et al. 2019).
The presence of the genera Prevotella, Faecalibacterium, Parabacteroides Eubacterium Ruminococcus, Bacteroides, and Clostridium in samples from the environmental microbiome highlighted the functions of the human gut and mammalian gut (p < 0.05) (Fig. 4). These groups are directly associated with the intestinal microbiota of mammals, which suggests a potential fecal contamination in these water bodies (Kiu et al. 2017; Suzuki et al. 2019; Vadde et al. 2019; Cheffi et al. 2020; Guo et al. 2020; Niestępsk et al. 2020). The presence of these species in the park’s samples may indicate that tourism in the area may lead to an anthropization of the place through tourism or agricultural activities, which, in the long term, may modify the local ecosystem.
Comparative functional inference analysis between the consortia where there is enrichment of photosynthetic groups compared to the environmental microbiome revealed a significant increase (p < 0.05) in functions related to photosynthetic activity, oxygen phototrophy, and nitrogen fixation. This was already an expected result because cyanobacteria and microalgae are microorganisms responsible for shaping the ecosystems in which they are inserted, as they make carbon sources available through photosynthesis (carbon fixation) (Durall and Lindblad 2015) and are directly involved in the cycle of nitrogen (Paerl 2017), carrying out the process of fixing it. Three genera of the Cyanobacteria phylum (Leptolyngbya, Phormidium, and Synechococcus) were identified only after enrichment in the microbial consortia. These groups can act as important participants in the production of biogenic methane in natural environments, as these genera have already been associated with the production of this hydrocarbon during the oxygen photosynthesis process (Bižić et al. 2020), an increased role in consortia (p <0.05). In addition to oxygenic photosynthesis, the SJA-28 group, belonging to the phylum Chlorobi, which is known as the green sulfur bacteria (GSB) group (Hiras et al. 2016) was also identified in the consortium samples. This strictly anaerobic, non-mobile, phototrophic microbial group is obligatory phototrophs that oxidize reduced sulfur compounds for CO2 fixation through the reverse tricarboxylic acid (rTCA) cycle, a process known as anoxygenic photosynthesis, and can also carry out N2 fixation (Paun et al. 2019).
Cyanobacteria are a group of microorganisms known for their broad metabolic capacity to produce secondary metabolites, including cyanotoxins (Harada 2004). The enrichment of these microorganisms in the consortia demonstrates the diversity present in the park, and serves as an alert to the risk of possible cyanobacteria and microalgae blooms, as the region is accessed by tourists for leisure in rivers and waterfalls. Alves et al. (2020) investigated the factors that favor blooms of photosynthetic microorganisms in artificial lakes and concluded that seasonality, including rainfall and water temperature, as well as hydrological characteristics such as pH and nutrient availability, are factors that favor blooms in water bodies.