The investigation, published in Environmental Geochemistry and Health, monitored bottom sediments at four strategic sites between March 2019 and July 2024: the Aimorés and Mascarenhas reservoirs upstream, and two estuarine stations near the river mouth, one of which is directly influenced by saltwater wedge intrusion. Researchers collected twenty-four mud-fraction samples covering both wet and dry seasonal cycles, using a sterile Van Veen grab sampler to retrieve the uppermost centimeters of sediment. This spatiotemporal design allowed the team to separate the slow mineralogical fingerprint inherited from the tailings from the faster changes driven by present-day environmental forcing.

To dissect the iron chemistry, the team combined selective dissolution procedures with X-ray diffraction, diffuse reflectance spectroscopy, magnetic susceptibility measurements, and multivariate statistics. Acid ammonium oxalate extraction targeted poorly crystalline iron forms such as ferrihydrite and lepidocrocite, while citrate–bicarbonate–dithionite extraction captured more crystalline phases like goethite and hematite. Alkali fusion provided total elemental inventories for iron, manganese, and the trace-element suite of arsenic, cadmium, cobalt, chromium, molybdenum, nickel, lead, vanadium, and zinc, quantified by ICP-MS with certified reference materials confirming recoveries between 93 and 109 percent.

The results reveal a sediment system dominated by crystalline iron oxyhydroxides. Dithionite-extractable iron ranged from roughly 15 to 70 grams per kilogram, while oxalate-extractable iron stayed below 20 grams per kilogram. Yet the ratio between the two fractions varied dramatically, from 0.14 to 0.82, with the strongest fluctuations at the estuarine stations. This temporal decoupling, the authors argue, is evidence of ongoing post-depositional transformation: crystalline phases are periodically destabilized by reducing conditions, and subsequent oxidation events generate short-range-order phases with high sorptive capacity, resetting the geochemical clock again and again.

X-ray diffraction confirmed goethite as the dominant iron oxyhydroxide across every environment and campaign, with hematite appearing as a subordinate but highly variable phase, particularly in estuarine sediments and at the Mascarenhas reservoir. Diffuse reflectance spectroscopy, interpreted through the Kubelka–Munk function, tracked the relative abundance of the two minerals and showed goethite-dominated assemblages giving way episodically to hematite-rich intervals, especially at the saltwater-influenced station during 2023. Magnetic susceptibility remained consistently low, indicating that antiferromagnetic oxyhydroxides, not ferrimagnetic magnetite, control the magnetic signal of these sediments.

Perhaps most striking are the crystallographic details. Goethite mean crystal dimensions ranged from about 20 to 40 nanometers along the 110 direction, with crystals preferentially elongated along that axis, a morphology consistent with assemblages inherited from the Fundão tailings. Specific surface areas of goethite spanned 57 to nearly 120 square meters per gram, and aluminum substitution in its structure varied from 0.097 to 0.199 mol per mol. Because aluminum substitution inhibits crystal growth, shrinks crystallite size, and boosts surface area, these variations translate directly into differences in the density and accessibility of sorption sites where trace elements bind through inner-sphere complexation, adsorption, and co-precipitation.

The geochemical data show that estuarine stations consistently carry the highest contaminant loads. Nickel reached 448 milligrams per kilogram at the saltwater-influenced station in March 2024, chromium peaked at 321 milligrams per kilogram, and arsenic concentrations climbed steadily between 2019 and 2024 at the neighboring estuarine site. Dithionite-extractable arsenic at that site surged nearly sixfold over the monitoring period, and wet-period campaigns brought marked increases in cobalt, chromium, molybdenum, nickel, and zinc. Reservoir sediments, by contrast, showed lower concentrations and less temporal volatility, behaving as relatively stable retention zones.

Principal component analysis tied the story together. The first two components explained 56 percent of total variance and cleanly separated reservoir from estuarine samples. Most potentially toxic elements loaded strongly alongside dithionite-extractable iron, total iron, and hematite crystallographic parameters, while the upstream Aimorés reservoir clustered instead with goethite surface area and aluminum substitution. Hierarchical clustering confirmed four distinct sample groups, demonstrating that spatial differences in contaminant partitioning are governed by iron mineralogy and crystallography rather than by total elemental abundance alone.

Why does this matter beyond the Doce River? Estuaries are often treated as terminal sinks where contaminated sediments come to rest. This study shows they are better understood as reactive biogeochemical reactors. Salinity fluctuations alter colloidal stability and surface reactivity, organic matter degradation drives suboxic conditions that partially dissolve iron phases, and resuspension during floods exposes fresh mineral surfaces. Each cycle can release a pulse of trace elements and then re-sequester them into newly formed, highly reactive phases. The contaminated pool is therefore dynamic, capable of responding to environmental perturbations rather than remaining permanently immobilized.

The findings also carry a clear message for environmental monitoring: total concentration measurements alone miss the story. Two sediments with identical arsenic totals may behave entirely differently depending on whether the element is locked inside well-crystallized goethite or loosely held on freshly precipitated ferrihydrite. The authors argue that mineralogical indicators and selective extraction procedures deserve a permanent place in long-term monitoring frameworks, not just in the aftermath of disasters. As climate variability reshapes hydrological regimes and saltwater intrusion pushes further inland, the mineral-scale processes documented here will increasingly determine whether legacy contaminants stay buried or re-enter the food web. A decade on, the Doce River’s iron minerals are still writing the disaster’s next chapter.

Subject of Research: Long-term evolution of iron oxyhydroxide mineralogy and potentially toxic element partitioning in mining-impacted sediments of the Lower Doce River fluvial–estuarine system, Brazil

Article Title: Long-term evolution of Fe oxyhydroxide mineralogy and potentially toxic element partitioning in mining-impacted sediments of the Lower Doce River fluvial–estuarine system

Article References: Long-term evolution of Fe oxyhydroxide mineralogy and potentially toxic element partitioning in mining-impacted sediments of the Lower Doce River fluvial–estuarine system. (n.d.). https://doi.org/10.1007/s10653-026-03495-z

Image Credits: AI Generated

DOI: 10.1007/s10653-026-03495-z

Keywords: Fundão dam collapse, Doce River, iron oxyhydroxides, goethite, hematite, potentially toxic elements, sediment geochemistry, selective dissolution, X-ray diffraction, estuarine biogeochemistry, trace-element partitioning, mining disaster legacy

Gypsum caves form when water dissolves calcium sulfate minerals, producing passages, chambers and sediment deposits that can look deceptively uniform to the naked eye. Chemically, however, these sediments may differ sharply over relatively short distances. Gypsum itself is composed primarily of calcium sulfate dihydrate, while anhydrite is its water-poor counterpart. As groundwater moves through the rock, it can transport dissolved ions, alter acidity and redistribute nutrients. Organic material entering from the surface may also accumulate unevenly, creating localized supplies of carbon, nitrogen and phosphorus. In darkness, where photosynthesis is impossible, microbial survival depends on exploiting these chemical gradients. Some organisms may use organic compounds as sources of energy, while others rely on oxidation-reduction reactions involving sulfur, nitrogen, iron or manganese. The new research indicates that the mineral and nutrient context helps determine which bacterial lineages can establish themselves in each sediment.

The research team examined sediments from Gypsum Cave, C3 Cave and Water Cave, three sites within the Sorbas gypsum karst. Their chemical analyses revealed pronounced heterogeneity among samples, particularly in concentrations of nitrogen, ammonium, nitrate, phosphorus, magnesium and calcium. These elements are not simply passive ingredients in the sediment. Nitrogen compounds can serve as nutrients or as electron acceptors in microbial metabolism; phosphorus is essential for nucleic acids, cell membranes and energy-transfer molecules; and calcium and magnesium influence mineral equilibria, cellular processes and the chemistry of pore water. Strontium, although not generally required in large quantities by bacteria, can act as a geochemical tracer because it follows pathways associated with calcium-bearing minerals and groundwater movement. Together, these variables provide a chemical fingerprint for each sediment microhabitat.

To connect chemistry with biology, the investigators compared the geochemical measurements with the composition of bacterial communities. Their statistical analyses used PERMANOVA, or permutational multivariate analysis of variance, a method commonly applied to ecological datasets containing many species or sequence-defined microbial groups. Rather than asking whether one chemical factor changes the abundance of one organism, PERMANOVA tests whether groups of samples with different environmental conditions also have systematically different community profiles. In this study, pH, carbonate content, magnesium, calcium and strontium each showed significant relationships with bacterial community structure. Individually, the variables explained between 15.8 and 21.3 percent of the observed variation, a substantial signal in the complex world of microbial ecology, where communities are also influenced by moisture, organic inputs, physical structure, seasonal changes and historical colonization. The statistical associations do not prove that each chemical variable directly causes the biological differences, but they identify geochemical conditions as important ecological filters.

Across the samples, the researchers detected bacteria affiliated with 38 phyla, although only 16 occurred at relative abundances greater than 1 percent. A phylum is a broad taxonomic category that groups organisms sharing deep evolutionary relationships, so the result reflects considerable diversity rather than a collection dominated by a handful of closely related species. The most prominent groups included Actinomycetota, Pseudomonadota, Bacillota, Planctomycetota, Acidobacteriota, Bacteroidota, Gemmatimonadota and Chloroflexota. Many of these lineages are familiar from soils, sediments and caves around the world, but their proportions shifted from cave to cave and from sample to sample. Such variation is consistent with the idea that underground microbial communities are assembled through a combination of environmental selection and the arrival of organisms from broader regional species pools.

Actinomycetota dominated the bacterial communities in Gypsum Cave, but this pattern weakened in Water Cave and C3 Cave. In four sediments collected from C3 and Water caves, Pseudomonadota became the dominant phylum instead. This contrast may reflect differences in nutrient availability, acidity, water movement or the amount of organic matter reaching each site. Actinomycetota includes many filamentous, soil-associated bacteria capable of producing resistant spores and breaking down complex organic compounds, traits that can be advantageous in dry or nutrient-poor environments. Pseudomonadota, a large and metabolically diverse group, contains bacteria adapted to rapidly changing conditions and a wide range of chemical resources. These broad ecological descriptions cannot identify the precise activities of the organisms in the caves, but they illustrate why shifts in major bacterial groups can signal changes in the constraints imposed by the environment.

The two most abundant genera reported in the study were Crossiella and wb1-P19. Crossiella belongs to the actinomycete community and has been detected in other terrestrial environments, including cave-associated settings. The label wb1-P19 refers to a bacterial lineage that is less well characterized and may represent organisms known primarily through genetic sequences rather than laboratory cultures. This is a common feature of microbiome research: DNA-based surveys can reveal the presence of organisms that scientists have not yet grown or studied in detail. Their sequences can show where a lineage occurs and how common it is, but they do not automatically reveal whether the cells are active, what they consume or how they interact with neighboring microbes. The study therefore provides a map of community composition and its environmental associations, while leaving many questions about physiology and ecosystem function open.

Water Cave contained the highest bacterial richness and diversity, whereas several samples from C3 and Gypsum caves exhibited comparatively low values. Richness refers to the number of detected taxa, while diversity incorporates both the number of taxa and how evenly their abundances are distributed. A sediment sample can therefore have many bacterial lineages but low diversity if one or two groups dominate. The greater diversity in Water Cave may be connected to more varied moisture conditions, chemical inputs or transport pathways, although the study’s results do not establish a single explanation. In subterranean systems, water can be both a resource and a vehicle, carrying dissolved nutrients, mineral particles and microorganisms through fractures and passages. Small differences in seepage, evaporation or sediment texture may create distinct niches, allowing more bacterial strategies to coexist.

One of the most striking implications is that gypsum caves appear to share a broad microbial ecological pattern with limestone and volcanic caves despite their different mineral foundations. The communities found in the Sorbas sediments included bacterial groups commonly reported from caves worldwide, suggesting that recurring environmental pressures—darkness, limited primary production, low nutrient supply and dependence on imported organic matter—may favor similar types of microbial life in geographically and geologically different settings. At the same time, the local chemistry of gypsum sediments added another layer of control, separating neighboring habitats within the same cave system. This two-level structure—general cave conditions shaping the overall microbial pool and sediment geochemistry refining the community—helps explain why underground ecosystems can be both globally recognizable and locally unique.

The study also highlights why evaporite karst deserves closer attention as climate, groundwater use and human activity alter subterranean environments. Gypsum dissolves more readily than many carbonate rocks, making gypsum cave systems particularly sensitive to changes in water flow and chemistry. Microbial communities may respond quickly to shifts in acidity, salinity, nutrient delivery or mineral dissolution, potentially making them useful indicators of environmental change. Yet the researchers’ results are a starting point rather than a complete ecological diagnosis. Future work will need to combine community DNA surveys with measurements of microbial activity, metagenomic analysis, microscopy and controlled laboratory experiments to determine which organisms perform specific chemical transformations. For now, the Sorbas caves reveal a hidden rule of underground life: in darkness, the composition of the sediment may be as important as the shape of the cave itself, and a few changes in dissolved minerals can help reorganize an entire microbial community.

Subject of Research: Geochemical controls on bacterial communities in gypsum cave sediments

Article Title: Deciphering the Geochemical Constrains Influencing Sediment Microbiomes in Gypsum Caves

Article References: Martin-Sanchez, P. M., Fernandez-Cortes, A., Calaforra, J. M. et al. “Deciphering the Geochemical Constrains Influencing Sediment Microbiomes in Gypsum Caves.” Microbial Ecology (2026). Original research article

Image Credits: AI Generated

DOI: 10.1007/s00248-026-02871-7

Keywords: gypsum caves, sediment microbiomes, cave microbiology, geochemistry, bacterial diversity, Actinomycetota, Pseudomonadota, Crossiella, PERMANOVA