Microbial diversity of the remote Trindade Island, Brazil: a systematic review

Data collection

This systematic review was based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Page et al., 2021). Records were searched on 2 August 2024 in Scopus, Web of Science, PubMed, PubMed Central, Dimensions, and Google Scholar databases using the Publish or Perish v8 tool (Harzing, 2007). The keywords used to search for records were: “trindade island” AND (metagenomics OR shotgun OR (amplicons AND (16S OR 18S OR ITS OR “internal transcribed spacer”))). There were no restrictions on document type, language or publication date to avoid excluding relevant records.

The results of the database searches were exported in CSV format. The scripts format_input.py (Jasper, 2023a) and remove_duplicates.py (Jasper, 2023b) were used to select unique documents. These scripts read the CSV files and filter out DOI-less and duplicate documents from the set of records (Dutra et al., 2023a, 2023b; Gomes et al., 2023; Sierra et al., 2023). The researchers García, G.J.Y. and Dutra, J.d.C.F. reviewed both the unique and DOI-less documents to select the scientific articles that met the selection criteria (Table 1). Disagreements in the selection of documents were resolved by the researcher Gomes, R.F.

After the initial review of titles and abstracts, only studies that addressed Trindade Island using shotgun metagenomics or amplicon sequencing were selected. These studies underwent a thorough evaluation, applying well-defined inclusion and exclusion criteria (Table 1), considering the collection area, sample type, methodologies used and the relevance of the results.

The following data were extracted from the selected documents: taxonomic abundance of microorganisms (Archaea, Bacteria, and Fungi) obtained by shotgun metagenomics and amplicon sequencing, type of substrate, samples or treatments, metagenomic approach, marker genes and sequencing platform used.

Extraction of taxonomic data

From the selected articles, taxonomic information on Archaea, Bacteria, and Fungi was extracted, both from the main text and Supplemental Materials. Considering that the selected articles used different databases and, when using the same databases, employed different versions, it was necessary to standardize the taxonomic nomenclature, identifying possible synonyms. For this, a local script developed in Python v3.10.8 accessed the Global Biodiversity Information Facility (GBIF) database (GBIF Secretariat, 2023) via its API. For taxa absent in the GBIF, manual searches were conducted using the NCBI Taxonomy database (Schoch et al., 2020).

Grouping samples

The samples analyzed in the selected articles were collected from various sources, including soil, water, and coral tissue. These studies employed different sequencing approaches, including shotgun metagenomics and amplicon sequencing targeting the 16S rRNA and ITS markers. Additionally, two studies employing amplicon sequencing also conducted microcosm experiments prior to sequencing (Morais et al., 2016; Rodrigues et al., 2018).

To streamline the explanation of analyses in this review, samples were grouped based on similar characteristics. Samples subjected to sequencing without prior treatment were categorized with the suffix “_Env” (Environment), while those treated with microcosms were grouped under the suffix “_MC” (Microcosms). Nevertheless, in most of the statistical analyses, the sample itself was considered the unit of analysis.

The groups were defined as follows:

• Soil_Env: Independent-cultivation soil samples (PD5, PD6, PF7; see Table 2). Samples PD5 and PD6 were collected from the Pico do Desejado region, and sample PF7 was from the Fazendinha region, both located in the highest areas of Trindade Island.

• Water_Env: Independent-cultivation water samples (NOR_Island_W, PRI_Island_W, SAN_Island_W; see Table 2). The NOR_Island, PRI_Island, and SAN_Island samples were collected from Ponta Noroeste, Enseada do Príncipe (south), and Trindade Shelf (east), respectively.

• Coral_Env: Independent-cultivation coral tissue samples (FAR_Island_C, NOR_Island_C; see Table 2). The NOR_Island and FAR_Island samples were collected from Ponta Noroeste and Ponta dos Cinco Farilhões (southwest), respectively.

• Soil_MC: Soil sample treated with microcosms (Crude Oil; see Table 2). This soil sample was collected from the eastern region of the island.

• Water_MC: Twelve microcosm treatments using seawater samples (OIL, FLU, HEX, PHE, PHE+FLU, PHE+HEX, PHE+OIL, PHE+PYR, PYR; see Table 2). These water samples were collected from the southeastern part of the island, near Praia das Tartarugas.

Rodrigues, Kalks & Tótola (2015) and da Silva et al. (2015) identified a group of bacterial isolates through partial sequencing of the 16S rRNA gene using the MegaBACE 1,000 DNA Analysis System (SANGER) platform (Table 2). However, these studies did not provide abundance data, which is why this information was not included in the comparative analyses. It is worth noting that taxonomic classification at the genus and species levels was only achieved in articles using independent-cultivation sequencing techniques (Env). To facilitate spatial understanding of the samples, a static map and an interactive map using the OSM (OpenStreetMap) platform and the JavaScript Leaflet library were created.

Analyses of the communities of Archaea, Bacteria, and Fungi

A series of sample analyses were performed using the programming language R v4.2.1. Relative abundances of all taxonomic categories were analyzed through bar charts generated with the ggplot2 v3.4.3 package (Wickham, 2011). Core (shared) and satellite (unique) taxa were examined using UpSet diagrams, generated with the UpSetR v1.4.0 package (Conway, Lex & Gehlenborg, 2017).

Alpha and beta diversity analyses were conducted based on the abundance at the order level, as it was the category where data was available for all samples. Shannon indices were calculated using the vegan v2.6.4 package (Dixon, 2003) and plotted with ggplot2 v3.4.3, based on relative abundances. Principal coordinate analyses (PCoA) were generated using the vegan v2.6.4, ggrepel v0.9.3 (Slowikowski et al., 2018) and ggplot2 v3.4.3 packages, based on Hellinger transformations.

Strong association networks between groups and taxa were calculated using Hellinger transformations with the indicspecies v1.7.14 package (De Caceres, Jansen & De Caceres, 2016). The multipatt function with the point biserial correlation coefficient (r.g) was used to calculate association strength, considered statistically significant at p < 0.05. Networks were constructed using Cytoscape v3.10.1 software (Shannon et al., 2003) with the edge-weighted spring-embedded layout. In the networks, circles represent taxa, the size of the circles indicates taxa abundance, diamonds represent sample groups, and the length of edges represents the association strength, with shorter edges indicating stronger sample-taxon associations.

Non-parametric multivariate analyses were conducted using the PERMANOVA test and the Bray–Curtis dissimilarity index with 9,999 permutations in PAST v4.11 software (Hammer & Harper, 2001). The aim was to compare community structure among sample groups based on the Hellinger transformation. Additionally, pairwise PERMANOVA models were used to evaluate statistically significant differences between pairs, with a significance level of p < 0.05.

Study selection

Search protocols across databases identified a total of 274 studies (Scopus (n = 24), Web of Science (n = 1), PubMed (n = 2), PubMed Central (n = 13), Dimensions (n = 98), and Google Scholar (n = 136)) (Data S1, Data S2). Among these, 67 duplicate records were automatically removed prior to the initial screening, resulting in 207 articles being evaluated. During screening, six documents were excluded for lacking a DOI and having ambiguous or incomplete titles, leaving 201 unique articles for the eligibility analysis.

After a manual review of the methodologies and application of inclusion and exclusion criteria, 193 articles were discarded by researchers García, G.J.Y. and Dutra, J.d.C.F., who reached agreement on the selected documents, with only two cases of disagreement. These were resolved by researcher Gomes, R.F., who accepted one document and rejected the other for not meeting the selection criteria (Table 1). Ultimately, eight documents met the established criteria (Fig. 1, Table 2).

It is worth noting that the studies by Câmara et al. (2022, 2023) and used the same samples and presented equivalent results. As such, the Câmara et al. (2022) study was excluded from the discussion for being redundant, but it remains part of the table of the eight selected articles.

Profile of selected articles

Meirelles et al. (2015) collected seawater and coral tissue samples, analyzing them through shotgun metagenomics. Câmara et al. (2023) collected soil samples and performed analyses using amplicon sequencing of 16S rRNA and ITS markers. Rodrigues et al. (2018) collected seawater samples and established nine microcosms (closed systems created from these samples) with the addition of various hydrocarbons. These systems were monitored for temperature, time, and hydrocarbon degradation and analyzed through 16S rRNA amplicon sequencing. Morais et al. (2016) collected soil samples, set up microcosms with the addition of aged crude oil, and subsequently performed analyses using 16S rRNA and ITS amplicon sequencing.

The majority of the selected studies focused on the bioremediation of hydrocarbon contamination (da Silva et al., 2015; Morais et al., 2016; Rodrigues, Kalks & Tótola, 2015; Rodrigues et al., 2018) using water and soil samples (Table 2, Fig. 2).

For more detailed visualization of the collected sample locations, an interactive map is available: https://glenjasper.github.io/leaflet-trindade-island-review-map.

Recovered communities of Archaea, Bacteria, and Fungi

This review focused on analyzing Archaea, Bacteria, and Fungi communities. Taxa from other kingdoms were also extracted but not analyzed; they can be consulted in Data S3. The selected articles reported results at varying taxonomic levels. Meirelles et al. (2015) and Câmara et al. (2023) presented results up to the genus level, while Rodrigues et al. (2018) reported data at the family level, and Morais et al. (2016) at the order level. Rodrigues, Kalks & Tótola (2015) and da Silva et al. (2015) used the SANGER method to sequence the 16S rRNA gene, identifying bacterial communities restricted to a few species. As a result, the findings from these two studies were included only in the discussion.

Because of differences in the taxonomic data presented in the selected studies, comparative analyses were primarily conducted at the phylum, order, family, and genus levels. Although Pylro et al. (2014) met the selection criteria, the authors did not report taxonomic data, limiting their study to pipeline comparisons for amplicon analysis. Consequently, no relevant information was extracted, and this study was excluded from further analyses.

Bacterial dominance and high rates of unclassified sequences in soil, with abundant low-frequency taxa in water and coral sample

In the analysis of Archaea and Bacteria, the majority of microbial reads annotated in the selected articles were identified as Bacteria, while Archaea were almost absent from the samples. No Archaea reads were detected in the microcosm samples, appearing only in the environmental samples. The highest relative abundances of Archaea were found in the coral samples NOR_Island_C (2.47%) and FAR_Island_C (5%) (Data S4).

A notable amount of unclassified reads was observed, particularly in environmental soil samples (PD5, PD6, PF7). At the family level, 36% of reads were unclassified, increasing to 68.9% at the genus level and 99.75% at the species level. However, one bacterial species, Dinghuibacter silviterrae, was identified in the Fazendinha region (PF7) (Data S4).

For the Water_Env and Coral_Env groups, about 8% of the reads were unclassified at the family and genus levels, decreasing to 1% at the species level (Data S4). This pattern may be attributed to the presence of bacterial species labeled as “uncultured.” However, a significant percentage of low-abundance taxa (<1%, <1.5%, or <3%) grouped as “Others” was observed in the Water_Env and Coral_Env samples at species, genus, family, and order levels (Data S4, Fig. 3B).

Taxon sharing among samples

UpSet diagram analyses revealed shared and unique taxa among Archaea, Bacteria, and Fungi across environmental samples (Soil_Env, Water_Env, and Coral_Env) and microcosm-treated samples (Soil_MC and Water_MC). The focus was on taxonomic categories at the phylum and genus levels (Fig. 4, Data S5).

Based on the data analyzed, 26 phyla of Bacteria and Archaea were present in both the environmental water and coral tissue samples. Among those samples, 14 were unique to the respective environments, namely Aquificota, Campylobacterota, Chrysiogenota, Deferribacterota, Deinococcota, Desulfobacterota, Fusobacteriota, Halobacteriota, Methanobacteriota, Poribacteria, Spirochaetota, Synergistota, Thermoproteota and Thermotogota. A total of 12 bacterial phyla were identified in the environmental soil samples, with two (GAL15 and WPS-2) being exclusive. In the samples treated with microcosms, only two bacterial phyla (Bacteroidota and Pseudomonadota) were identified, with the Pseudomonadota phylum being the sole presence across all the groups of samples analyzed. It is noteworthy that the highest diversity of phyla was observed in the environmental water and coral tissue samples, while the soil and water samples treated with microcosms exhibited the lowest microbial richness (Fig. 4A, Data S5).

Fungal phyla were more prevalent in soil environmental samples, followed by water and coral tissue samples. Only two fungal phyla (Ascomycota and Mucoromycota) were found in microcosm-treated samples. Ascomycota was the only phylum identified across all the fungal sample groups (Fig. 4C, Data S5).

At the genus level, 293 genera of Archaea and Bacteria were shared between water and coral tissue environments, with 279 unique to these environments. Seven genera (Baumannia, Buchnera, Carnobacterium, Chlamydia, Mobiluncus, Riemerella and Simonsiella) were exclusively shared by seawater samples, while eight (Brachybacterium, Brevibacterium, Halorubrum, Haloterrigena, Methanocorpusculum, Oribacterium, Pyrobaculum and Tolypothrix) were unique to coral tissue. Soil samples shared 12 bacterial genera, with five unique to this environment (Bryobacter, Edaphobacter, Gemmatimonas, Pedomicrobium, and Roseiarcus). The Fazendinha (PF7) region had seven unique genera (Chthonomonas, Dinghuibacter, Edaphobaculum, Inquilinus, Occallatibacter, Roseisolibacter, and Thermosporothrix). Only six genera (Conexibacter, Haliangium, Mucilaginibacter, Mycobacterium, Pedosphaera, and Rhodomicrobium) were shared across all sample groups. Bacterial genus diversity was significantly higher in marine regions compared to soil areas (Fig. 4B, Data S5).

No fungal genus was shared across all samples. Coral tissue samples had two unique genera (Grosmannia e Rhodonia). Soil samples shared 22 unique genera, including Alternaria, Antarctomyces, Apiotrichum, Archaeorhizomyces, Ciliophora, Cladosporium, Diaporthe, Hyphodiscus, Juncaceicola, Knufia, Lipomyces, Mortierella, Neoascochyta, Papiliotrema, Penicillium, Pestalotiopsis, Pseudogymnoascus, Purpureocillium, Recurvomyces, Rhexocercosporidium, Saitozyma, and Trichoderma. The Fazendinha (PF7) region displayed 18 unique genera (Acaulospora, Arachnomyces, Botrytis, Callistosporium, Chaetosphaeria, Circinella, Clavaria, Cryptodiscus, Gloeocystidiellum, Gloioxanthomyces, Hyphozyma, Mycena, Phaeotremella, Podospora, Solicoccozyma, Spencermartinsiella, Trechispora, and Triparticalcar), while the Pico do Desejado (PD6) region also had 18 unique genera (Acremonium, Arthrinium, Debaryomyces, Gibberella, Glarea, Gremmenia, Hyaloscypha, Ilyonectria, Leucosporidiella, Macrophomina, Pezicula, Pezoloma, Phenoliferia, Rhodotorula, Thelebolus, Trichosporon, Veronaeopsis, and Volutella). Unlike bacterial taxa, fungal genera exhibited greater diversity in soil samples compared to marine samples (Fig. 4D, Data S5).

Analysis of alpha and beta diversity and sample correlation

Alpha diversity analysis revealed distinct results across sample groups. Water_MC group showed a mean Shannon index of 1.109, indicating relatively low diversity in this specific environment. Soil_MC group exhibited a mean Shannon index of 1.683, reflecting moderate diversity. Soil_Env and Water_Env groups displayed significantly higher Shannon indices, 2.639 and 2.7, respectively, indicating substantial diversity compared to microcosm treated samples. Within Soil_Env, samples from Pico do Desejado (2.68 and 3.31) demonstrated greater diversity than those from Fazendinha (1.93). Coral_Env group had the highest mean Shannon index, 4.094, highlighting a notably rich biodiversity in independent-cultivation coral tissue samples (Fig. 5A).

Beta diversity, represented by Principal Coordinate Analysis (PCoA) (Fig. 5B), revealed clustering of samples based on similarity: Water_Env and Coral_Env groups were closely positioned, showing no statistically significant differences according to PERMANOVA analyses (Data S6). Only Water_MC samples exhibited significant differences when compared to Soil_Env (p = 0.0044), Water_Env (p = 0.0058), and Coral_Env (p = 0.019).

Strong association networks at the phylum level showed that Coral_Env had the highest number of associated phyla, with no significant associations in only five phyla (Fig. 6A). Coral samples shared more associations with Water_Env samples than with soil samples (Soil_Env and Soil_MC), consistent with PERMANOVA results. Unexpectedly, Soil_Env and Soil_MC groups exhibited limited overlap in associations, with Soil_Env showing more connections with Coral_Env samples.

Water_MC and Soil_MC groups displayed fewer associations: soil-associated phyla included Ascomycota, Actinobacteriota, and Acidobacteriota, while water-associated phyla included only Pseudomonadota. Across all substrates, bacterial phyla dominated associations, while fungal phyla (Ascomycota and Basidiomycota) were associated only with Soil_Env and Soil_MC, respectively. Methanobacteriota and Halobacteriota, representing Archaea, were associated exclusively with Coral_Env. Despite their high abundance (Fig. 6A), Ascomycota and Pseudomonadota were not significantly associated with all substrates. Actinobacteriota, Chlamydiota, and Pseudomonadota showed the highest number of associations, linked to three out of five substrate types.

At the order level, the association patterns mirrored those at the phylum level, with Coral_Env displaying the majority of the associations. Microcosm-treated samples exhibited fewer associations across orders, while Coral_Env and Water_Env showed a substantial number of shared associations. Soil samples demonstrated no overlap in their associations (Fig. S1).

At the genus level, only environmental samples (Soil_Env, Water_Env, and Coral_Env) identified communities of Archaea, Bacteria, and Fungi. Coral tissue samples exhibited the highest number of significantly associated genera, including most archaeal genera, followed by water samples. Shared associations between water and coral tissue samples were predominantly bacterial, with fewer archaeal and fungal genera represented. In independent-culturing seawater samples, the most abundant bacterial genera were Prochlorococcus, Pelagibacter, Pseudoalteromonas, and Alteromonas, while in coral tissue samples, the most abundant were Cyanothece, Streptomyces, Mycobacterium, and Geobacter (Fig. 6B).

Soil samples showed fewer strong associations, with bacterial genera Acidothermus and Mycobacterium as those of the most abundant. Fungal genera Mortierella and Lipomyces exhibited high abundance and strong associations with samples from Pico do Desejado and Fazendinha (Fig. 6B).

Bacterial and fungal communities in the ecological niches of Trindade Island marine environments

Meirelles et al. (2015) provided a comprehensive overview of microbial communities in the marine areas of Trindade Island, revealing notable differences among habitats. The Trindade Shelf region (SAN_Island)—which includes the Jaseur Seamount, Columbia Bank, Almirante Saldanha Seamount, Vitória Seamount, and Eclaireur Seamount, all part of the Vitória-Trindade Chain (VTC)—displayed unique microbiota, especially in terms of bacterial diversity. These sites, which featured rhodolith beds overgrown by fleshy algae, exhibited both high abundance and diversity of microbial communities in the water column.

The study reported the highest concentrations of chlorophyll a (0.18–0.27 μg L⁻¹) and orthophosphate (0.55 μM) in these areas, suggesting higher primary productivity and potential benthic–pelagic coupling. These conditions presumably favor more diverse bacterial communities because of the consistent input of nutrients from fleshy algae and the accumulation of dissolved organic carbon (DOC) released by these algae.

In the Trindade Shelf region, the phylum Pseudomonadota was dominant, playing an especially significant role in heterotrophic processes and nutrient recycling. The phyla Bacteroidota, Actinobacteriota, and Verrucomicrobiota were also prominent, performing key functions in organic compound degradation and carbon cycling—processes essential for sustaining high productivity on the Trindade Shelf.

Meirelles et al. (2015) also investigated three other sites on Trindade Island: Ponta dos Cinco Farilhões (FAR_Island), Ponta Noroeste (NOR_Island), and Enseada do Príncipe (PRI_Island). These sites featured rocky reefs predominantly covered by turf algae, with higher DOC concentrations (up to 219.17 μM at Ponta dos Cinco Farilhões). The associated microbiota appeared adapted to these environmental conditions. Turf algae, which extensively cover these reefs, release exudates rich in organic compounds, potentially promoting bacterial growth while also creating local hypoxia that can stress corals. Cyanobacteria, such as Lyngbya, are often integral to turf algae in these environments, releasing DOC and exacerbating unfavorable conditions for corals.

Bacteroidota and Pseudomonadota were the dominant bacterial phyla in these reefs. Bacteroidota, known for their ability to decompose organic matter, and Pseudomonadota, typically fast-growing heterotrophic bacteria found in DOC-rich environments, both reflect the substantial influence of turf algae and cyanobacteria. Elevated irradiance and organic input may further drive microbial activity, potentially worsening coral health.

Meirelles et al. (2015) observed that Mussismilia hispida colonies in these locations showed impaired health, including tissue necrosis and bleaching in over 90% of colonies in shallow reefs. The increased presence of Bacteroidota and Ascomycota in these corals reinforces their possible links to stress and disease, as both phyla have been reported in diseased corals in other reef systems (Schmieder & Edwards, 2011). Although the authors did not examine fungal communities in detail, the broad taxonomic composition supports differences in bacterial and fungal assemblages between deeper areas (SAN_Island) and shallow reefs (FAR_Island, NOR_Island, PRI_Island). The deeper Trindade Shelf region displays higher bacterial diversity, correlating with elevated nutrient levels and diminished light stress, whereas the shallow, turf algae–dominated reefs show bacterial communities adapted to high DOC levels and low-oxygen conditions, adversely affecting coral health.

Terrestrial environments

Câmara et al. (2023) investigated three soil samples from two high-elevation sites on the island: Pico do Desejado (PD5 and PD6) and Fazendinha (PF7). These soils were characterized by low base saturation (PBS < 50%), high cation exchange capacity (CEC), and high organic matter content, mainly composed of plant fibers. The presence of iron from weathered ferromagnesian minerals fosters microbial communities adapted to acidic, metal-rich conditions. Such soil properties significantly influence both the structure and function of bacterial and fungal communities.

In this study, the dominant bacterial phyla were Actinobacteriota, Pseudomonadota, Acidobacteriota, Chloroflexota, and Verrucomicrobiota, all typical of acidic soils with high organic matter. Actinobacteriota were most abundant at Pico do Desejado (PD5), reflecting their crucial role in decomposing complex organic matter (e.g., lignin and cellulose) and driving carbon cycling. Their abundance may be linked to the active breakdown of plant residues, such as fibers from giant ferns (Cyathea gigantea), which predominate in the Fazendinha region. By contrast, PD6 was dominated by Pseudomonadota and Verrucomicrobiota, which are pivotal in the nitrogen cycle, including nitrification and nitrogen fixation. Their prevalence could be explained by the buildup of nutrients due to the weathering of the island’s rocky substrate.

At Fazendinha (PF7), Acidobacteriota and Bacteroidota were the most prominent phyla. Acidobacteriota, commonly found in acidic environments, drive essential decomposition and mineralization processes in nutrient-poor soils. The high potential acidity (H+Al) and low pH in Fazendinha soils account for the predominance of this phylum, critical for maintaining fertility under challenging conditions. Bacteroidota further support nutrient cycling by breaking down organic matter.

Fungal communities in these soils were also highly diverse, with Ascomycota, Basidiomycota, and Mucoromycota as the main phyla (Câmara et al., 2023). Notably, Antarctomyces psychrotrophicus and Pseudogymnoascus sp. exhibited adaptations to cold conditions, assisting in carbon recycling at low temperatures. Mortierella humilis was particularly abundant at Fazendinha (PF7), highlighting its capacity to decompose complex organic compounds and form mycorrhizal associations that enhance phosphorus uptake by plants. These mycorrhizal interactions are instrumental in soil structuring and improve both stability and water retention.

Overall, these microbe–soil interactions emphasize the resilience of Trindade Island’s ecosystems, even under adverse environmental conditions, and underscore the importance of microbial communities in maintaining the ecological balance of island habitats.

Bacterial and fungal communities from microcosm-treated samples

Rodrigues et al. (2018) conducted microcosm experiments by enriching seawater with various hydrocarbons to evaluate their effects on microbial communities. The nine experimental treatments—PHE (phenanthrene), OIL (weathered oil), PHE+OIL (phenanthrene + weathered oil), PYR (pyrene), PHE+PYR (phenanthrene + pyrene), FLU (fluoranthene), PHE+FLU (phenanthrene + fluoranthene), HEX (hexadecane), and PHE+HEX (phenanthrene + hexadecane)—had distinct impacts on both the structure and function of the microbial communities. These differences reflect how microorganisms adapt to the presence of aliphatic versus aromatic hydrocarbons and their mixtures.

According to Rodrigues et al. (2018), alpha and beta diversity differed markedly among treatments. Taxonomic shifts underscored how the chemical complexity of hydrocarbons influences the adaptation and metabolic activity of marine microorganisms. Treatments containing only aromatic hydrocarbons (PHE, PYR, and FLU) significantly decreased microbial diversity, likely due to the inherent toxicity of polycyclic aromatic hydrocarbons (PAHs). Phenanthrene (PHE) promoted the dominance of Alteromonas, indicating the genus’s specialized ability to degrade recalcitrant compounds. Pyrene (PYR), being highly hydrophobic, resulted in a more selective community dominated by taxa such as Oceanospirillaceae. Fluoranthene (FLU) had a subtler effect on beta diversity, with Alteromonas maintaining a strong presence but Flavobacteriaceae showing sensitivity, suggesting difficulties in metabolizing isolated PAHs.

Treatments combining phenanthrene with other PAHs (PHE+PYR and PHE+FLU) exacerbated the toxic effects, further reducing microbial diversity. In PHE+PYR, the family Sphingomonadaceae was abundant, suggesting a competitive advantage under PAH-rich conditions. In PHE+FLU, Alteromonas again emerged as dominant, highlighting its importance in degrading recalcitrant compounds. However, the reduced abundance of Cryomorphaceae implies that certain families struggle to adapt to multiple toxic compounds.

Hexadecane (HEX), an aliphatic hydrocarbon, supported the highest microbial diversity among the treatments. This less recalcitrant compound enabled a balance of Flavobacteria, Gammaproteobacteria, and Alphaproteobacteria, with Flavobacteriaceae being the most prevalent family. The combination of phenanthrene and hexadecane (PHE+HEX) displayed significant synergy, as hexadecane acted as a non-aqueous phase liquid (NAPL), enhancing the bioavailability and degradation of phenanthrene. This treatment stimulated the growth of Alcanivorax and Alteromonas, highlighting a complementary metabolic interaction between these genera.

In treatments involving weathered oil (OIL and PHE+OIL), Alcanivorax predominated, reflecting its specialization in degrading complex hydrocarbons and long-chain alkanes. Weathered oil, containing a diverse array of organic compounds, promoted increased abundance of taxa such as Alteromonas and Marinomonas, which are adapted to polluted environments. In the PHE+OIL treatment, Oceanospirillaceae and Flavobacteriaceae likewise increased, suggesting metabolic interactions that facilitate PAH degradation in the presence of alkanes.

Across these treatments, Alteromonas, Alcanivorax, and the Flavobacteriaceae family proved central to hydrocarbon degradation, each responding differently to aliphatic and aromatic compounds. While Alteromonas was particularly prevalent in PAH treatments, Alcanivorax was favored by alkanes and weathered oil. Overall, Rodrigues et al. (2018) concluded that interactions among various hydrocarbons modulate microbial metabolism and community composition, enhancing degradation efficiency under mixed contamination scenarios.

Morais et al. (2016) performed microcosm experiments with soil to assess how crude oil influences microbial and fungal communities on Trindade Island. Oil addition altered community structure, diversity, and composition, emphasizing an adaptive response to complex hydrocarbons. Microbial alpha diversity decreased in the presence of oil, with a reduction of up to 40% in fungal taxa—likely due to PAH toxicity, which eliminates less-adapted organisms. Taxonomic analysis revealed Actinobacteriota, especially Nocardia and Streptomyces, as dominant in contaminated soil. These genera are known to produce biosurfactants and extracellular enzymes that enhance hydrocarbon bioavailability and promote degradation (Balachandran et al., 2012; Rodrigues, Kalks & Tótola, 2015).

Among Pseudomonadota, Betaproteobacteria and Gammaproteobacteria showed relative increases; this included the family Comamonadaceae, noted for its heterotrophic denitrification capabilities. In contrast, Alphaproteobacteria and Deltaproteobacteria, such as Rhodospirillales and Syntrophobacteraceae, decreased, indicating lower resilience to hydrocarbon exposure.

In the fungal community, Mortierella accounted for 70% of relative abundance under oil treatment. This oleaginous fungus metabolizes hydrocarbons (Hughes, Bridge & Clark, 2007) and accumulates lipids, making it central to soil adaptation in contaminated environments. Genera such as Hypocreales declined, reflecting limited capacity to endure adverse conditions.

Bacterial communities from cultivated samples

Rodrigues, Kalks & Tótola (2015) isolated and characterized hydrocarbon-degrading microorganisms by collecting coastal sediment samples from Trindade Island. Fifteen bacterial isolates were identified, spanning the phyla Actinobacteriota, Firmicutes, and Pseudomonadota. Of these, the genera Rhodococcus and Nocardia showed remarkable catabolic versatility, capable of utilizing a variety of hydrocarbons—toluene, octane, xylene, naphthalene, phenanthrene, pyrene, hexadecane, anthracene, eicosane, tetracosane, triacontane, and pentacosane—as carbon and energy sources. In particular, Rhodococcus rhodochrous, which harbors the alkB and C23DO genes, exhibited rapid growth on both long-chain hydrocarbons and polycyclic aromatic hydrocarbons (PAHs). Nocardia farcinica, containing the C23DO gene, proved specialized in degrading specific aromatic hydrocarbons.

Although Cellulosimicrobium cellulans and Microbacterium lacticum were capable of degrading hexadecane and naphthalene, respectively, the authors recommend further studies to clarify the specific roles of these relatively understudied isolates. They conclude that the hydrocarbon-degrading capacity of these strains—isolated from an environment without a documented history of anthropogenic pollution—underscores the potential of indigenous microbiota for bioremediation, particularly given the island’s proximity to Brazil’s oil-producing regions.

da Silva et al. (2015) focused on bacteria producing biosurfactants tolerant to high salt concentrations, with potential applications in biotechnology. Soil and sediment samples collected from Trindade Island yielded 14 bacterial strains, mostly within the genus Bacillus (including Bacillus subtilis and its subspecies Bacillus subtilis subsp. subtilis and Bacillus subtilis subsp. spizizenii). Some strains remained unidentified, suggesting the possibility of novel species. These bacteria produced effective biosurfactants even at salt concentrations as high as 175.0 g L⁻¹; strains TR13 and TR14 were particularly successful in reducing interfacial and surface tension. However, a genetic analysis by Dunlap, Bowman & Zeigler (2020) proposed that the four Bacillus subtilis subspecies (subsp. subtilis, subsp. spizizenii, subsp. inaquosorum, and subsp. estercoris) may warrant reclassification as separate species due to differences in their production of bioactive secondary metabolites. These findings underscore the complexity and wide-ranging biotechnological potential of the genus Bacillus.

Overall, the ability of these microorganisms to reduce interfacial tension and withstand challenging conditions suggests considerable promise for bioremediation and enhanced oil recovery in industrial and contaminated environments.

Comparison between microbial communities of Trindade Island and the Galápagos Islands

In the study by Zhao (2019), 48 environmental samples collected in 2015 and 2016 at 23 stations throughout the Galápagos archipelago were analyzed. Filters of 3 and 0.22 μm pore size were used to collect bacterial cells, and environmental parameters included particulate organic carbon (POC), dissolved organic carbon (DOC), chlorophyll a, phosphate, nitrate, silicate, and cell abundance measured by flow cytometry. Samples were sequenced using the Illumina HiSeq platform, allowing both taxonomic composition and ecological functions to be assessed based on unassembled reads.

Taxonomic analyses revealed communities dominated by bacteria, cyanobacteria, and archaea. Among the primary bacterial groups, Candidatus pelagibacter (SAR11) was the most abundant taxon. This group plays a key role in the carbon cycle, recycling DOC in oligotrophic environments and thereby contributing to the sustainability of marine ecosystems. Other important genera included Formosa, Polaribacter, and Tenacibaculum (all within Flavobacteria), which were linked to upwelling events. These microorganisms are involved in organic matter degradation and nutrient cycling, particularly in areas of high primary productivity.

El Niño 2015 and La Niña 2016 climatic events exerted a substantial influence on microbial community composition (Zhao, 2019). During the La Niña event, stronger upwelling brought nutrient-rich deep water to the surface, leading to increases in taxa suited to eutrophic environments, such as Rhodobacterales HTCC2255 and Gammaproteobacteria HTCC2207, alongside archaea like Candidatus nitrosopelagicus. This archaeon, dominant in some high-productivity areas, plays a central role in the nitrogen cycle through ammonia oxidation, especially in deeper waters. By contrast, the El Niño 2015 event weakened upwelling, resulting in more homogeneous microbial communities and relatively stable DOC and nutrient levels.

Cyanobacteria were pivotal to primary productivity, notably Synechococcus and Prochlorococcus. Synechococcus was associated with high-productivity sites (e.g., Darwin Bay, where chlorophyll a increased), whereas Prochlorococcus was predominant in more oligotrophic eastern regions. Both genera directly contribute to carbon fixation and support the marine food chain (Zhao, 2019).

Whereas the microbial communities of Trindade Island were dominated by Bacteroidota, Actinobacteriota, and Pseudomonadota, Galápagos communities were marked by Candidatus pelagibacter (SAR11), Synechococcus, and Prochlorococcus. In the Galápagos, archaea such as Candidatus nitrosopelagicus significantly influence the nitrogen cycle, a feature less pronounced in Trindade Island communities. Additionally, climatic events like El Niño and La Niña notably shaped Galápagos microbial assemblages through direct impacts on nutrient availability via upwelling, whereas on Trindade Island, microbial patterns appear to be primarily driven by local processes such as benthic–pelagic coupling and exudates released by turf algae.

These disparities demonstrate how microbial communities are molded by unique environmental conditions in each region, underscoring their essential role in the ecological resilience and sustainability of marine ecosystems.

Comparison between microbial and fungal communities of Trindade Island and the Cuatro Ciénegas Basin

Souza et al. (2018) analyzed microbial communities from the Churince system in the Cuatro Ciénegas Basin (CCB), Mexico—a closed hydrological system with extremely high microbial diversity—by sampling water, sediment, and soil. Environmental DNA underwent 16S rRNA gene sequencing via 454 pyrosequencing, revealing 60 bacterial phyla and three archaeal phyla.

Aquatic environments in this system were dominated by Pseudomonadota, Actinobacteriota, and Bacteroidota, while sediments and soils showed higher overall diversity. The latter included Firmicutes and Cyanobacteria, which play key roles in nitrogen fixation and nutrient cycling. In sediments, cyanobacteria form microbial mats, whereas in soils, they participate in microbial crusts along with Acidobacteriota and Nitrospirota, reflecting localized adaptations.

Phylogenetic reconstructions of the genus Bacillus indicated endemic lineages in sediment and marine environments dating from the Ediacaran and Jurassic (Souza et al., 2018). These lineages perform crucial nutrient-recycling functions, exhibiting metabolic cooperation and local antagonism mechanisms that ensure survival under harsh conditions.

The CCB is often described as a microbiological “lost world,” not only because of its high biodiversity but also its ancient microbial lineages, limited horizontal gene transfer, and adaptations to extreme nutrient imbalance. These factors create ecological barriers that prevent even similarly adapted taxa from successfully colonizing the basin.

Both the CCB and Trindade Island harbor high microbial diversity, supported by eco-evolutionary adaptations to extreme environments. Yet, while the CCB’s microbial diversity is shaped by geologically stable, nutrient-poor, and often saline conditions—with notable phyla including Cyanobacteria, Pseudomonadota, Actinobacteriota, and Firmicutes—Trindade Island’s microbial communities thrive in organic-rich, acidic soils, featuring phyla such as Actinobacteriota, Pseudomonadota, Acidobacteriota, Bacteroidota, and Verrucomicrobiota, alongside diverse fungal groups (notably Ascomycota and Mortierella humilis).

In the CCB, adaptations center on coping with nutrient scarcity, with cyanobacteria facilitating nitrogen fixation and forming microbial crusts. On Trindade Island, microbial processes include decomposing organic matter—particularly plant fibers from giant ferns—highlighting the importance of Actinobacteriota in rich, acidic soils. Additionally, Pseudomonadota and Verrucomicrobiota play pivotal roles in nitrogen fixation and carbon cycling.

A striking contrast is the Bacillus genus in the CCB, which includes lineages dating to the Ediacaran and Jurassic, adapted to extremely oligotrophic habitats, compared to the predominance of Rhodococcus and Nocardia (both Actinobacteriota) on Trindade Island, noted for metabolizing complex hydrocarbons in an environment lacking historical anthropogenic contamination. The phylum Acidobacteriota, prevalent in Fazendinha soils, is vital for mineralization in nutrient-poor habitats, whereas Bacteroidota, found in reefs and soils, assists in organic-matter degradation.

Velez et al. (2016) collected water, sediment, and wood panel samples from three freshwater systems (Churince, Pozas Rojas, and Becerra) in the CCB, sequencing internal transcribed spacer (ITS) regions to investigate active fungal communities. The resulting fungal profiles showed moderate diversity, with lignocellulolytic and saprophytic anamorphic fungi dominating. Genera such as Alternaria, Cladosporium, Stachybotrys, Trichoderma, Phoma, and Acremonium played crucial roles in decomposing plant detritus and cycling nutrients under oligotrophic conditions. Species adapted to extreme environments (e.g., Acremonium persicinum and Emericellopsis pallida) thrived due to elevated salinity in the CCB.

The genus Stachybotrys—often associated with perennial grasses like Phragmites australis—exhibited strong cellulolytic capacities and acted as a primary saprobe in plant litter. Phoma, widespread in the environment, has both saprophytic and phytopathogenic phases, while Cladosporium (also broadly distributed) includes saprophytic and phytopathogenic taxa capable of tolerating various stressors. Local lignocellulose colonization was largely governed by fungi such as Volutella, Trichoderma, and Phoma, which exhibit strategies to compete intensely for available substrates (Velez et al., 2016).

By contrast, Trindade Island’s soil fungal communities are dominated by the phyla Ascomycota, Basidiomycota, and Mucoromycota. These fungi are adapted for carbon recycling under acidic, cooler conditions, as exemplified by Antarctomyces psychrotrophicus. Mortierella humilis—particularly prevalent in Fazendinha (PF7) soils—demonstrates adaptive capacity to organic-rich soils by decomposing complex compounds and forming mycorrhizal associations.

Despite differences in their respective habitats, the fungal communities of both the CCB and Trindade Island display specific adaptations to local environmental constraints. Whereas the CCB’s fungi are specialized for oligotrophic, often saline conditions, the fungi on Trindade Island thrive in organic-rich, acidic soils. In each setting, fungi are vital for nutrient cycling, sustaining key ecological processes.

The differences and similarities between the microbial and fungal communities of Trindade Island and the CCB illustrate how these organisms have evolved to occupy distinct but equally challenging ecological niches. In both ecosystems, they play a fundamental role in maintaining ecological resilience by supporting critical biogeochemical processes. Their diverse adaptations likewise provide insights into ancestral evolutionary pathways and interspecies interactions, pointing to the biotechnological potential of these communities, which holds considerable promise for biodiversity conservation and novel environmental applications.

Limitations and future perspectives

The reviewed studies highlight the richness and complexity of microbial and fungal communities on Trindade Island, illustrating how environmental factors affect the health of marine and terrestrial ecosystems. These findings have important implications for understanding natural mechanisms of habitat management and restoration on Trindade Island and in comparable regions.

Despite this progress, there are intrinsic limitations in molecular identification techniques, particularly for bacterial, archaeal, and fungal taxa. For instance, numerous taxa detected in soil samples collected at higher elevations such as Pico do Desejado and Fazendinha could not be assigned to specific genera or species. This limitation likely arises from the high number of undescribed microbial species and gaps in existing taxonomic databases (Konstantinidis, Rosselló-Móra & Amann, 2017; Lu et al., 2017; Marcelino et al., 2020; Seol et al., 2019). Furthermore, there were also potential biases due to distinct sequencing platforms, sequencing depths and coverage, and mainly, the lack of temporal (longitudinal) studies, which could indeed clarify aspects concerning how seasonality (summer × winter, or rainy × dry seasons) affects the microbial diversity in the different substrates and environments previously studied in the Trindade Island.

Another critical constraint was the lack of access to raw sequencing data (e.g., FASTQ files), which were unavailable in any of the reviewed articles. As a result, the analyses for this review relied solely on results reported in the main text and other supplemental materials. This was particularly problematic when examining studies of microcosm-based experiments, since many did not present detailed abundance tables at the genus level despite discussing genus-level findings. Such omissions limited the scope of inter-study comparisons and restricted our ability to profile the described microorganisms more comprehensively.

While DNA-based research on environmental samples from Trindade Island remains relatively limited, there is clear growth in what might be termed the “molecular era.” However, achieving a fuller understanding of the island’s diverse habitats, microbial communities, and their ecological functions requires expanded, dedicated efforts. This is especially vital given the uniqueness of Trindade Island, about which much remains to be discovered.