Electron Microscopy Evidence of Zn Bioauthigenic Sulfides Formation in Polluted Organic Matter-Rich Sediments from the Chicamocha River (Boyacá-Colombia)

Electron microscopy and sediment geochemical data from a river basin (the upper Chicamocha river basin, UCRB, Boyacá province, Colombia) affected by anthropogenic activities (wastewater discharges, smelting and agricultural activities) showed the existence of heterogeneously distributed Zn particles in the sediments and sediments with Zn contents above the regional background (42 mg/kg). The objective of this study was to evidence the ZnS sedimentary neoformation in organic matter rich sediments deposited in anthropogenic reservoirs to reveal the processes involved in the sedimentary uptake of Zn from potential pollution sources. The highest Zn concentrations are found in clay minerals and organic matter-rich sediments (up to 427 mg/kg) deposited in slow-flowing reaches of the river associated to La Playa dam. Quartz-rich sediments poor in organic matter deposited in fast flowing segments of the river show very low Zn contents (1–12 mg/kg). Electron microscopy images showed ZnS nanoparticles forming cell-shaped aggregates suggesting that sulfate-reducing microorganisms acted as templates for the partial binding of Zn and for the nucleation and growth of zinc sulfide minerals. A good correlation of Zn with total organic carbon (r = 0.936) and the low potential redox of these sediments (−233 mV) suggest that organic matter was able to maintain oxygen depleted conditions appropriate to the Zn accumulation in the sediments. Our results demonstrate that potentially toxic Zn, originating from anthropic activities, was partially immobilized in organic matter-rich sediments through the precipitation of sulfides.


Introduction
Pollution by heavy metals is of particular concern for humans because of their detrimental health effects on people in excessive quantities. Contaminated geomaterials in areas heavily utilized by people, including extensively used recreational zones in towns [1] or agricultural fields [2,3], can create potential risk of metal exposure to humans. Alluvial sediments in river basins can accumulate or release heavy metals depending on flow regime and flooding [4]. The presence of dams regulate the surface water outflow in river basins and causes slow-flowing sites and flooding. The accumulation of clays and organic matter is favored behind these dams, promoting in many cases binding of metals.
This study was especially focused on the use of high resolution microscopy techniques (field emission scanning electron microscopy, FESEM, and high-resolution transmission electron microscopy HRTEM), coupled to energy dispersive X-ray spectrometry (EDX) to evidence the ZnS sedimentary neoformation in organic matter rich sediments deposited in anthropogenic reservoirs to reveal the processes involved in the sedimentary uptake of Zn from potential pollution sources in the UCRB area.

Background Context
The UCRB is located in the high mountains of the equatorial Andes of Colombia in the Boyacá department ( Figure 1). The landscape is characterized by plains of lacustrine and fluvio-lacustrine origin, with average heights over 2500 m, surrounded by elevated massifs and plateaus (2650-2800 m), including the Tunja Plateau [23]. The area of the UCRB is around 214,000 ha, with a length of 62.46 km and an average slope of 0.12% [24]. The Chicamocha river is formed by the confluence of the Tuta and Jordán rivers with a length of 227 km flowing to the Caribbean Sea. Regarding the climate of the UCRB, annual rainfall ranges from 650-800 mm in the southwestern area to 900-1200 mm in the northwestern area. The highest precipitations occur during April-May and October-November. The average annual temperature is around 13.1 • C. March and November are the warmest months. Prevailing wind directions are from east and north.
The Chicamocha River Basin is characterized by the presence of Paleozoic to Quaternary sedimentary and volcanic rocks. Sedimentary rocks are predominantly siliceous (conglomerates, sands, shales and clays). The Paipa sector is characterized by the presence of one of the main geothermal areas in the Colombian territory. At the northwestern sector of the UCRB the products of the Paipa volcano crops out, which is an eroded explosive volcanic building formed by acid pyroclastic deposits (alkaline rhyolites and trachyandesites) of the Pliocene-Pleistocene age [24].
The most important anthropogenic modification of the UCRB is La Playa dam. The La Playa dam is located at the Tuta area in the upper sector of the URCB (Figure 1). This reservoir supplies water for livestock and agricultural activities and receives wastewaters (urban sewage) from Tunja, Oicatá, Combita and Tuta, together with runoff waters of the surrounding farming activities, generating high organic and nutrient load [25]. All of these inputs produce an intense eutrophication with a strong proliferation of Eichhornia crassipes and colonies of small algae, such as Trachelomonas and C. microporum [19]. Nearby La Playa dam, in the Tunja-Paipa road (km 27), is a manufacturing facility belonging to one of the world's largest steel producers. This facility produces recycled carbon steels causing the emission of slags with high Zn contents [21]. The presence of other heavy metals has not been investigated.

Materials
Sediments of the river were sampled from a 20 points network through the UCRB, including proper alluvial deposits and sediments from La Playa dam ( Figure 1D-F). A standard stainless Shelby tube was used for sampling, and cores down to a depth of 50 cm were obtained. The physico-chemical parameters of sediments (redox potential, pH and electrical conductivity) were determined in situ with a Hanna Instruments multiparameter for soils and sediments (HI98168, Hanna Instruments (S) Pte Ltd., Singapore). The sediment samples were dried in oven at 40 • C and prepared for further treatments. Sediment fragments were picked every 10 cm in the cores for the electron microscopy studies. For the X-ray diffraction (XRD), X-ray fluorescence (XRF) and Inductively Coupled Plasma-Mass spectrometry (ICP-MS) studies, each core was homogenized to yield a single sample. Water samples were also taken from the same 20 sampling points (Figure 1). At each site, water samples were collected in several (15, 50 and 150 mL) clean high-density polyethylene bottles (HDPE), rinsed several times with groundwater to be sampled. One of the bottles (15 mL) containing sample was acidified to pH < 2.0 with nitric acid (HNO 3 ), ACS reagent and a purity of 70 %, to stabilize trace metals. The filling of the bottles was performed using sterile syringes and filtered by a 0.45-µm filter pore size then acidified. The samples were stored in ice boxes, brought to laboratory and stored at 4 • C for ICP-MS analysis.

Mineralogical and Geochemical Analysis
The XRD data were obtained of oriented aggregates from the homogenized whole sample and from the <2µm fraction of the same sample separated by centrifugation, after washing with ultrapure water to remove salts. Oriented aggregates were prepared by settling a dispersion in a glass holder. Furthermore, ethylene glycol treatment was carried out to assist in the identification of expandable minerals (smectite, interstratified layers, etc.). The diffraction patterns were obtained in a Panalytical X'Pert Pro diffractometer (CuKα radiation, 45 kV, 40 mA, Panalytical, Eindhoven, The Netherlands) equipped with an X'Celerator solid-state linear detector, using a step increment of 0.008 • 2θ and a counting time of 10 s/step (Centro de Instrumentación Científico-Técnica of the Universidad de Jaén, Spain). A sweep between 3 • and 62 • 2θ was done on the dry samples, while for the glycolated samples the sweep was done between 2 • and 30 • 2θ in order to confirm the identification of expandable minerals. Deconvolution routines included in the HighScore software were applied.
Textural and chemical observations in a field emission scanning electron microscope (FESEM, Merlin Carl Zeiss, Carl Zeiss, Oberkochen, Germany) were made on polished sections using back-scattered electrons (BSE) in atomic number contrast mode, and on sediment fragments in secondary electron mode (SE). Chemical composition of minerals was obtained using an energy dispersive X-ray spectrometry (EDX) coupled to the electron microscope. The FESEM study was carried out with a Merlin Carl Zeiss (Carl Zeiss, Oberkochen, Germany) equipment with an Oxford Inca EDX system in the Centro de Instrumentación Científico-Técnica of the Universidad de Jaén.
For the textural and chemical characterization at the nanometer scale, selected samples according to XRD analysis and FESEM observations were prepared for high resolution transmission electron microscopy (HRTEM) study. Samples were prepared using Au and Cu grid surface coated in a perforated formvar resin from a dispersion of finely ground sample particles, in alcohol or distilled water. The monomineralic character of each grain is proven by its electron diffraction pattern, checking the existence of a single network and, therefore, a single crystalline phase. The sections were then fixed on an Au or Cu grid for TEM analysis. The TEM data were obtained using two electron microscopes at the Centro de Instrumentación Científica (C.I.C.) of the Universidad de Granada (Granada, Spain): the HAADF FEI TITAN G2 microscope, operated at 300 kV, and the Philips CM20 microscope, operated at 200 kV. Qualitative analyses (AEM) of particles were obtained with an EDX microanalysis system in the Philips CM20.
Whole-sediment analyses of the major and trace elements of selected samples were carried out using XRF in a Philips Magix Pro (PW-2440) spectrometer. Every sample was finely powdered in a mill and then pressed into a pellet using wax, producing a dense and homogenous sample to enhance the analytical performance of trace elements [26]. Glass beads with lithium tetraborate were also employed for the major elements analyses to minimize the preferential orientation of phyllosilicates. The detection limit was 0.01 wt.% for major elements and from 1 to 5 mg/kg for trace elements. An empirical calibration was performed with an array of more than 25 international geostandards, including a wide range of silica-containing common rocks (Pro-trace reference materials, Panalytical). Residual matrix effects are adjusted taking into account a rough analysis of the main elements and using variable influence coefficients (alpha) obtained by fundamental parameters. Loss on ignition (LOI) was determined using 0.5 g of powdered sample, first dried at 110 • C and then heated at 1000 • C for one hour. Trace elements were also analyzed using a NexION 300D inductively coupled plasma-mass spectrometer (ICP-MS) (PerkinElmer Inc., Shelton, CT, USA). ICP-MS analyses were performed following sample digestion with HNO 3 + HF. Sample powders (0.1 g) were attacked in a teflon-lined vessel at high temperature and pressure, and then evaporated to dryness. Subsequently, residues were dissolved in 100 mL of 4% vol. HNO 3 . ICP-MS analyses quantification was made with certified standards (MAG-1, UB-N, AGV-N, DR-N, GS-N, GA, GH and BR-N). Waters were also analyzed by ICP-MS. Detection limit for all the trace elements analyzed was below 1 µg/kg. Three replicates were performed for each sample. Standard deviation was always below 5%. Both techniques, XRF and ICP-MS, were carried out at the Centro de Instrumentación Científica, CIC, Universidad de Granada, Granada, Spain.
The total organic carbon (TOC) was analyzed using a Shimadzu Total Organic Carbon Analyzer (TOC-V sch) from the Instituto de Recursos Naturales Agrobiología (IRNAS) from CSIC-Sevilla (Sevilla, Spain).

XRD Results
XRD diffraction patterns reveal that the mineral assemblage of the alluvial sediments from the Chicamocha River is dominated by the presence of quartz and kaolinite. Some sediments from the southwestern and northeastern parts of the UCRB are enriched in quartz ( Figure 2a). Kaolinite is the dominant phyllosilicate in all the studied samples with intense and sharp 7.2 and 3.6 Å peaks ( Figure 2a). Mineral assemblage is contrastingly different in sediments from La Playa reservoir at the central part of the UCRB, where oriented aggregates of the <2 µm fraction have a broad peak of 10 to 13 Å (Figure 2b) due to the presence of smectite and illite-dioctahedral vermiculite mixed layers (I-DV) [27]. The presence of these minerals and the higher intensity of the kaolinite peaks indicate that the sediments from La Playa are enriched in clay minerals.

Electron Microscopy Results
SEM images of the sediments from the La Playa dam are characterized by alternating bands of microlaminated organic matter-rich layers and very fine-grain-sized clay-rich layers (<2 µm) ( Figure 3a). TEM-AEM study of the sediments from the La Playa dam confirmed the existence of three groups of clays: kaolinite, I-DV and smectite ( Figure 3b).  Quevedo et al. [27] indicated that the mineral assemblage in the sediments trapped in La Playa dam is characterized by the presence of Fe 2+ -bearing smectite (up to 0.4 atoms per formula unit), pyrite microframboids and vivianite (iron phosphate), suggesting the existence of reduced conditions in the sediments. Small smelting particles containing Fe and Zn near pyrite framboids can also be observed (Figure 3c).
SEM and TEM imaging revealed the presence of ZnS aggregates associated to plant fragments (Figures 3a and 4a). Clumps of fine ZnS nanoparticles (50-100 nm) were observed forming aggregates with oval and spherical shape and size between 1 to 2 µm, closely resembling bacterial cell morphology (Figure 4b). Surfaces of the cell-shaped aggregates are heavily encrusted. Zn nanoparticles apparently not associated with cell-like morphologies were also observed in electron microscope images. HRTEM images (Figure 4c,d) shows that ZnS nanoparticles can occur scattered or attached to mineral surfaces. Several cell-shaped aggregates are apparently flattened probably as a consequence of the electron microscopy vacuum conditions or breaking down of the cells (Figure 4b,h). Grape-like clusters made up of multiple fragments of aggregated spheres (i.e., botryoidal habit) formed by coalescence or aggregation of microspheroids can also be observed (Figure 3a). SEM-EDX elemental mapping revealed that cell-shaped aggregates and extracellular precipitates contained Zn and S (Figure 4e-g); however, it was not possible to assign mineralogy at this scale because we could not obtain electron diffraction patterns even in the HRTEM. Although EDX analysis in FESEM confirmed that precipitates were mainly composed of zinc and sulfur, some of the ZnS particles have small contents of copper (Figure 4h

Chemical Composition and Physicochemical Properties of the Sediments
The SiO 2 and Al 2 O 3 contents of the sediments define two main group of samples. Clay-rich sediments from la Playa reservoir have lower SiO 2 contents (mean 61.24%) and higher Al 2 O 3 contents (mean 17.17%) than quartz-rich sediments from the rest of segments of the UCRB (Table 1 and Table S1). The La Playa sediments are also characterized by the highest content in organic matter (TOC up to 13.84%), and LOI values (up to 15.43%). The quartz-rich sediments from the SW and NE part of the UCRB show the highest content in SiO 2 (up to 95.4%) and the lowest content in organic matter (TOC from 0.15% to 0.52%) and LOI values (0.72-1.55%). A significant variation on the P 2 O 5 content of the sediments can also be observed, ranging from a mean value of 0.58% in the sediments from La Playa to 0.14% in the sediments from the northeastern segment of the UCRB. Heavy metal concentrations in sediments from the UCRB vary depending on the sediment location. Sediments from La Playa reservoir contain comparatively high levels of heavy metals (Table 2 and  Table S2) following a decreasing order as: Zn > Cu > Cr > Ni > Pb. On the other hand, lower mean values of these elements in the sediments from the NE and SW segments can be observed. Water composition of the sediment sampling sites (Table 3 and Table S3) is also characterized by higher heavy metals concentrations in samples from La Playa reservoir, especially for Zn (mean value 268 µg/L).  The concentrations of Zn in the sediments from the UCRB exhibited large variability, ranged from 5 to 427 mg/kg ( Table 2). The lowest concentrations (5-12 mg/kg) were found in the quartz-rich sands from the northern part of the basin. In the sediments rich in clay minerals and organic matter (TOC up to 13.84%) from the La Playa reservoir, Zn can be up to 427 mg/kg, i.e., more than 400 times the content in the quartz-rich sediments poor in organic matter. Zn is well correlated with TOC and LOI ( Figure 5, Table 4) and with the other heavy metals and P 2 O 5 .   Table 5 and Table S4 show important differences in the physicochemical properties of the UCRB sediments and waters. La Playa sediments are characterized by lower redox potential (around −230 mV) and higher electrical conductivity (2625 µS/cm) than the rest of sediments.

Flow Regime and Flooding of the Chicamocha River and Sediment Conditions
The influence of flow regime and flooding of the Chicamocha river determined the formation of two main types of alluvial sediments in the basin. The river segment with slow-flowing conditions due to the construction of La Playa dam is characterized by the deposit of organic matter (TOC up to 16.22%, Table 1) and clay-rich sediments by periodic flooding. This reservoir receives input of untreated urban wastewaters and farming activities in the surrounding areas [27]. Domestic wastewater is considered one of the major sources of phosphate and nitrogen compounds in the natural environment, along with agricultural runoff [28]. Phosphate is a nutrient for algae that promotes eutrophication of water bodies. Eutrophication has many negative effects on aquatic ecosystems, such as proliferation of algae, degradation of water quality and oxygen depletion. The pollution of water by P in urban wastewater and agricultural runoff caused the intense eutrophication of the La Playa reservoir segment in the UCRB. The authigenic precipitation of Fe 2+ -bearing sulfides, phosphates and smectites in the La Playa reservoir [27] reveals the existence of anoxic conditions (with measured redox potential around −230mV) (Table 5) in the sediments rich in decaying organic matter with high microbiological activity caused by eutrophication.
In contrast, alluvial sediments deposited in fast-flowing sites, located both upstream and downstream of the La Playa dam, hardly contained organic matter (TOC frequently below 2.5%, Table 1), except for sample 14, situated in a local meander). In these cases, fast-flowing segments avoided the organic matter deposit and favored the water and sediment oxygenation (measured redox potential around 150mV).

Heavy Metal Distribution in Sediments and Pollution Sources
A decreasing order of Zn > Cu > Cr > Ni > Pb was observed in the mean values of the heavy metal contents in sediments from the La Playa slow-and fast-flowing segments located upstream and downstream of the La Playa. However, the heavy metal contents of the UCRB sediments vary depending on the location of the sediments, the mineral composition and the organic matter content (Tables 1 and 2).
There were significant differences between the Zn concentrations in different types of sediments from the UCRB. Zn contents in quartz-rich and organic matter-poor sediments (TOC contents below 2.5%) deposited in fast-flowing reaches of the Chicamocha river located upstream and downstream the La Playa dam are frequently below the corresponding values determined for some proposed unpolluted reference soils in Colombia (Table 2) [3] and clearly lower than crustal and average worldwide soils (around 70 mg/kg, [29]). Some sediments from these river segments richer in clay minerals and moderate TOC content (around 2.5%) have Zn concentrations similar to the crustal and average worldwide soils, although higher than average Zn contents in several regions (around 30-40 mg/kg [30,31]). According to these data, we considered the mean value of the quartz-rich with TOC contents below 2.5% deposited in fast-flowing reaches of the Chicamocha river located upstream and downstream the La Playa dam as a reference for the unpolluted sediments of the UCRB area (Tables 1 and 2).
On the other hand, the concentrations of Zn in the clay minerals and organic matter-rich sediments from the UCRB located in the slow-flowing reaches of the La Playa dam clearly exceeded the world normal averages (up to six times) and the mean level of Zn in these sediments was clearly higher than the corresponding value determined for the unpolluted reference sediment of the UCRB (Table 2). Although Colombia does not have a regulatory framework for concentrations of contaminants in sediments and soils [32,33], the Zn contents of the sediments from La Playa dam are above the threshold background values in European soils [34]. These Zn concentrations can be directly associated with an anthropogenic source.
Heavy metal accumulation in sediments and soils can be the result of different types of anthropic activities. Smelting, urban wastewater and agricultural activities can be considered as potential pollution sources that provided heavy metals to sediments of the UCRB. Smelting is one of the most significant sources of Zn pollution [35]. During smelting, Zn is usually released into the surrounding environment, e.g., [36,37], discharged with smelting gases which particles can accumulate in nearby soils and sediments leading to significant pollution [38,39]. The presence in the study area of a major manufacturing facility in the region, suggests this smelting site as one of the potential sources of pollution. According to Parra and Sánchez [21], this facility produced smelting slags representing a significant environmental hazard due to its high Zn contents (1728 mg/kg), but the presence of other metals was not reported. Slags are processed by grinding and milling producing additional emission of dust particles. The presence of small Fe-Zn particles in the sediments (Figure 3c) suggest that part of Zn was deposited in the sediments by atmospheric deposition of particles from the smelting plant. Zn-transport in solution or as particulate by runoff from this source is less probable because the smelting plant of the UCRB is located downstream the site that contain sediments with highest Zn-pollution (La Playa reservoir).
On the other hand, good correlations among P 2 O 5 , heavy metals and organic matter contents (Table 4) suggest that the wastewater discharged by urban districts (e.g., Tunja, Figure 1) into the Chicamocha River, the use of wastewater irrigation and the frequent use of heavy metals-bearing products (fertilizers, pesticides, etc.) in the agricultural activities upstream the La Playa dam can also be considered as other potential sources for the concentration of heavy metals in the sediments deposited by periodic flooding in this area. In Colombia, the intensive use of fertilizers and pesticides was identified as an important source of pollutants in soils from river basin systems [3,32,40]. Flooding conditions could provide adequate conditions for the concentration of Zn and other heavy metals in the organic matter-rich sediments deposited by repeated inundation events [4].

The ZnS Formation
Good correlation of Zn with TOC (r = 0.942) and LOI (r = 0.936) ( Table 4) suggest that clay minerals and organic matter are related with the accumulation of this metal in the sediments of the La Playa reservoir in the UCRB. Organic matter-rich layers are able to maintain oxygen depleted conditions due to capability of decaying organic matter to provide electrons for reactions of reduction. The presence of Fe 2+ -bearing minerals, such as pyrite, vivianite and smectite in the La Playa sediments, as well as the low redox potential (around −230 mV, Table 5) of these sediments suggest that mineral transformation processes occurred in a reduced environment. In saline environments, such as that of La Playa dam (mean E.C. 2625 µS/cm, Table 5), following the availability of easily degradable organic matter, microorganisms most likely can create unevenly microenvironments with reduced metals and aqueous sulfides, in which some processes lead to the precipitation of insoluble, heavy metal-bearing minerals.
The striking morphological and dimensional features of spherical and oval-shaped ZnS nanoparticle aggregates suggest bacterially induced precipitation of ZnS in the organic matter-rich sediments of La Playa sediments. Some studies have shown that ZnS grain crystallinity can be improved by the presence of Desulfovibrio desulfuricans producing biogenic crystals (sized 4-12 nm) of sphalerite and wurtzite [44]. In a Pb-Zn mine in Tennyson, Wisconsin (USA), the predominance of nanocrystals of sphalerite over those of wurtzite was revealed based on TEM data [43]. However, the small size of the nanoparticles from La Playa sediments made it difficult to obtain electron diffraction patterns and did not allow to distinguish the type of ZnS polymorph that form the observed cell-shaped aggregates.
SEM images showing the concentration of ZnS nanoparticles on the surface of the cell-shaped aggregates forming thin encrustations in the UCRB sediments (Figure 4a,b,e,h) suggest that precipitation of ZnS was associated to the activity of cell surfaces. SEM and HRTEM images of these sediments also revealed ZnS nanoparticles apparently not associated with cell-like morphologies (Figure 4b,c) suggesting that extracellular substances can also be involved in the nucleation of ZnS nanoparticles in the UCRB sediments. Therefore, the SEM and HRTEM data from this study suggest that sulfate-reducing bacteria cells facilitated decisively the uptake of Zn particles from UCRB sediments through sulfide minerals formation. Some experimental studies [45,46] have confirmed the ability of microbial cells to take up metals (Fe) that concentrate in crusts on the surface of microorganisms as sulfide minerals. Microscopy analyses [46] revealed that iron sulfide minerals produced in the presence of sulphate reducing microorganisms exhibit cell morphology with crusts, suggesting that the cell surface of live and dead sulphate-reducing microorganisms, and the extracellular polymers produced by live cells, provide templates for the nucleation of sulfide minerals and favor mineral growth. In this process, the nucleation of nanoparticles that take up pollutant metals can be favored by the effect of organic substances (extracellular polymeric substances, biofilms) that decrease free energy and act as ligands [18,41,[47][48][49]. Particle attachment can produce larger nanocrystals and aggregates that can coarse to produce micrometric crusts [50]. Multiple stages of this process can promote botryoidal habit. Becker et al. [51] suggested that the remobilization of heavy metals from reduced sediments is hampered by the predominance of sulfidic bonding. Microbially driven framboidal pyrite formation can limit the mobility of metals from water and sediments and the abundance of plants can favor the incorporation of Zn within the roots [52]. Zn adsorption onto clay minerals cannot be discarded as an additional potential retention mechanism in the sediments [53][54][55].
Our results suggest that potentially toxic Zn, provided by different anthropogenic activities (wastewaters, agricultural and smelting activities), was concentrated in organic matter-rich sediments through precipitation of sulfides. The activity of sulfate-reducing microorganism controlled the formation of micron-scale cell-shaped aggregates made of ZnS nanocrystals.

Conclusions
Spatial distribution of Zn and other heavy metals in the sediments of the UCRB revealed that high concentrations of these metals are located in the slow flowing reaches associated to the clay minerals and organic matter-rich sediments from La Playa dam. Electron microscope evidence of Fe-Zn metallic particles in these sediments suggest that part of the supply of Zn into the UCRB was associated to the atmospheric deposit of particles from the activity of a local smelter located downstream the La Playa dam. Good correlations among organic matter, P 2 O 5 and heavy metals contents suggest that the origin of Zn and other heavy metals can be, partially at least, in the same source as organic matter (wastewater discharges) and the agricultural activities upstream the La Playa dam (application of fertilizers, pesticides and fungicides).
Electron microscope evidence has also shown that the supply of Zn into the UCRB has been partially immobilized by biogenic metal sulfide crystallization in microsites of clay minerals and organic matter-rich sediments. Electron microscope images revealed dispersed ZnS nanoparticles and nanoparticles forming encrusted microbial shapes suggesting that sulfate-reducing microorganisms play essential role in the precipitation of Zn sulfides. This study has shown HRTEM and SEM proofs of the micro-scale mechanisms involved in the precipitation of biogenic Zn sulfide in polluted sediments. Nanoparticle precipitation associated to microbial organic substances in layers rich in decaying organic matter was the main mechanism of sulfide crystallization.
Author Contributions: C.P.Q. and G.R.C. conducted field observations and sampling. J.J.-M., R.J.-E., C.P.Q. and G.R.C. performed microscopic observations SEM and TEM (mineralogical, textural, and geochemical analyses) and interpreted the X-ray diffractograms and chemical data (XRF and TOC). All the authors discussed the analytical results and prepared the manuscript. All authors have read and agreed to the published version of the manuscript.
Funding: This work has been financed by the Spanish research project PGC2018-094573-B-I00 from the MCIU-AEI-FEDER, and research group RNM-325 of the Junta de Andalucía (Spain). Our gratitude is also extended to Asociación Universitaria Iberoamericana de Postgrados (AUIP) and the Universidad de Boyacá. Additional thanks to Colombian Research groups Gestión Ambiental COL0005468 and Gestión de Recursos Hídricos COL0005477.
Acknowledgments: Technical and human support provided by CICT of Universidad de Jaén (UJA, MINECO, Junta de Andalucía, FEDER) is especially recognized for helping and for being able to provide us analyses during the COVID-19 lockdown. We acknowledge the suggestions and comments from two anonymous reviewers and from the editor.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.