Metagenomic Influential Insights in the Formation of Biogenic Iron Hydroxysulfate Precipitates by Ferrous Oxidative Microbial Consortia

Abstract The formation of iron hydroxysulfate precipitated minerals by microbial bioleaching is an undesired dominant process because it decreases the solubilization efficiency of targeted metals. To date, the microbial role in forming these precipitated minerals is unclear. Herein, we compared two microbial community diversity (consortia M1-18 and M8-15) with the distinct formation of secondary iron minerals. In the case of M1-18, a 16S rRNA metagenomic analysis revealed a higher microbial diversity and a prevalence of Ferrimicrobium and Acidimicrobiacea. This consortium also showed a slightly higher oxidation capacity and a remarkably higher particulate precipitate capacity of secondary minerals. In contrast, the consortium M8-15 showed lower diversity and poor precipitate capacity. XRD, EDS, and FE-SEM analysis of secondary iron minerals showed that the microbial consortia promote the mineralogical formation of the secondary iron precipitates in the form of schwertmannite, an early metastable phase that can be easily transformed to jarosite.

In recent years, the formation of iron hydroxysulfates minerals gained further importance by their ability to act as a vehicle for the transport of heavy metals and toxic elements (Gan et al. 2017;S anchez-España et al. 2012;Zhu Jianyu et al. 2013). This ability is accomplished by adsorption, co-precipitation or structural incorporation/substitution mechanisms, playing a key role in the remediation of environments contaminated by AMD (Bai et al. 2012;Gan et al. 2017;Hou et al. 2015;Liao et al. 2009;Song et al. 2018). Researches show that schwertmannite and jarosite can remove heavy metals such as As, Zn, Cu, Hg, Pb, Cd, and Cr from the aqueous solution (Bao et al. 2018;Houng Aloune et al. 2014;Zhou et al. 2007;Zhu Jianyu et al. 2013). This high scavenger capability is conferred by their surface reactivity, crystal structure, and solubility (Bai et al. 2012;Gan et al. 2017;Zhu Jianyu et al. 2013). Iron hydroxysulfates minerals are also applied in the recovery of contaminated environments (Bai et al. 2012;Gan et al. 2017;Hou et al. 2015;Song et al. 2018;Wang H et al. 2006;Zhu Jianyu et al. 2013), and hydrometallurgical and industrial processes such as the production of pigments, magnetic materials, and catalyst enhancer (Hou et al. 2015;Jim enez et al. 2019;Zhu Jianyu et al. 2013). The efficient and inexpensive production of iron hydroxysulfates minerals has gained significant attention, being the synthesis by Fe-oxidizing microorganisms the most attractive method (Bai et al. 2012;Bao et al. 2018;Hou et al. 2015).
Iron hydroxysulfates minerals formation begins with the oxidation of Fe 2þ to ferric iron (Fe þ3 ), which is catalyzed by iron-oxidizing bacteria (FeOB) (Mori et al. 2016;Song et al. 2018). After the Fe 2þ oxidation process, the Fe þ3 faces a subsequent hydrolysis reaction and precipitates as a mineral phase of secondary iron (Daoud and Karamanev 2006;Houng Aloune et al. 2014;Mori et al. 2016;Song et al. 2018). Acidithiobacillus ferrooxidans, one of the most common acidophilic FeOB (Bai et al. 2012;Gan et al. 2017;Liao et al. 2009;Wang X et al. 2012), is reported as a mediator in the formation of iron hydroxysulfates in acid conditions with high concentrations of Fe þ3 and SO 2À 4 (Daoud and Karamanev 2006;Hou et al. 2015;Song et al. 2018;Zhu Jianyu et al. 2013). However, this is not the only acidophilic FeOB involved in the formation of iron hydroxysulfates. Previous reports indicate that the iron hydroxysulfates minerals formation is promoted by to Acidithrix ferrooxidans and Ferrovum myxofacies (Hedrich et al. 2011;Mori et al. 2016) as well as by indigenous bacterial community conformed by FeOB and iron reducing bacteria (FeRB) performing activities related to the cycling of iron (Lu et al. 2013).
The microbial populations linked to these activities are usually mesophilic and moderate thermophilic acidophiles, the predominance of chemoautotrophic FeOB, such as Acidithiobacillus and Leptospirillum, and Ferroplama, and heterotrophic Fe þ3 reducing bacteria, such as Acidiphilium (Johnson et al. 2014;Mori et al. 2016). Moreover, heterotrophic bacteria such as Ferrimicrobium and Acidimicrobiacea have also been related to AMD and the iron hydroxysulfates formation (Gupta et al. 2019;Johnson et al. 2009). However, the involvement of this bacteria is controversial. Our previous reports in microbial genomics showed to unveil important physiological traits in FeOB (Ccorahua-Santo et al. 2017, 2021. In this work, we continue with our efforts to understand the FeOB microbial communities and conduct a metagenome approach correlated to physical-chemistry materials analysis to study the consortia influence on precipitates of secondary iron minerals.
Herein, to better understand the effect of a microbial community in the formation and transformation of secondary iron minerals, the 16S rRNA metagenomic analysis of two microbial consortia, with different oxidation abilities and different formation of secondary iron minerals were compared. Additionally, the morphology and composition of precipitates were studied with FE-SEM/EDS and XRD.

Consortium and culture condition
The samples were collected from the in-situ leaching (ISL) and leach/Solvent Extraction/Electrowinning (L-SX-EW) zone, located in the Toquepala copper mine (Tacna, Peru). They were placed in sterile polythene bags with a proper label, stored in an ice box, and transported to the laboratory. The samples were stored at 4 C for further analysis.
Subsequently, samples were transferred on a 9 K medium to enrich the iron-oxidizing consortia. The composition of 9 K medium was prepared as it follows: 3.33% w/v FeSO 4 .7H 2 O, 0.04% w/v MgSO 4 .7H 2 O, 0.01% w/v (NH 4 ) 2 SO 4 , and 0.004% w/v KH 2 PO 4 (Ram ırez et al. 2004). The pH value of these media was adjusted to 2 with a solution of 1.0 mol/l H 2 SO 4 . The cultures were incubated aerobically at 30 C and 180 rpm on a rotary platform. The bacterial consortia were adapted and maintained in freshly prepared 9 K media for posterior analysis.

Kinetic study and precipitation experimental design
Leaching kinetic experiments were performed in a 500 ml flask with a 200 ml reaction volume of 9 K medium at initial pH 2. Each flask was inoculated with cell suspension containing bacterial consortium at 10% (v/v) from active culture. The cultures were stirred at 200 rpm and incubated aerobically at 30 C. All experiments were conducted in triplicate. The experiments were monitored respectively at a scheduled interval of 24 h by 5 days. The number of bacteria was verified by a Petroff Hauser counting chamber. The concentration of Fe 2þ and total iron was quantified by spectrophotometry.

Consortium identification by 16S rRNA gene sequencing
Preliminary bacterial consortium cells were grown on 9 K medium in two 500 ml flasks to complete a total of 1 l of culture volume. When bacterial consortium reached the exponential growth period (72 h), the cells were harvested by centrifugation at 13,000 rpm for 10 min at 5 C. Then, the cell pellets went through a washing process, according to Ccorahua-Santo et al. (2017). Briefly, the cells were centrifuged at 13,000 rpm, and the pellet was washed with acid water (pH 2.0) and centrifuged at 13,000 rpm. The process was repeated four times until the removal of the jarosite precipitate. Subsequently, the pellet was washed once with sodium citrate (10 mM, pH 7.0).
Afterward, the cell pellets were used for gDNA isolations using Soil DNA Isolation Plus Kit (Norgen,Cat No. 64000) according to the specifications of the manufacturer, where 100 ml elution buffer was used lastly to elute gDNA. The quality of gDNA was checked on 1% agarose gel (loaded 5 ml) for the single intact band, while the agarose gel was run at 70 V for 60 min. DNA quantification and A260/280 ratio were determined using Epoch TM Microplate Spectrophotometer (Biotek Instruments, Inc., USA), where 2 ml of each sample was loaded in the Take 3 plate. The extracted DNA was sent to Macrogen, Inc., Seoul, South Korea, for amplicon sequencing. In detail, the amplicon libraries were prepared using the Nextera XT Index Kit (Illumina Inc.) as per the 16S Metagenomics Sequencing Library preparation protocol (Illumina 2018).

Data analysis
Sequencing raw data generated by Illumina MiSeq System were analyzed by CosmosID metagenomic software (CosmosID Inc., Rockville, MD), which describes and reveals microbial community composition and quantification of relative abundance (Yan et al. 2019). The raw sequence files were uploaded to the CosmosID cloud application, with no parameters being set or modified for data upload. Briefly, the CosmosID platform utilizes highly curated dynamic comparator databases (GenBook V R ) and high-performance data-mining algorithms (k-mer-based) to rapidly disambiguate millions of metagenomic sequence reads into discrete microbial taxa. OTUs were identified against the CosmosID curated 16S database using a closed-reference OTU picker and 97% of sequence similarity through the QIIME framework (Roy et al. 2018;Yan et al. 2019).

Chemical analysis
The samples were periodically taken throughout the experiment in order to analyze the ferrous iron. This was determined by the colorimetric o-phenanthroline method. The Fe 2þ was measured by the 1,10-phenanthroline analytical method at 510 nm using a UV-Vis spectrophotometer (Biotek Instruments, Inc., USA) (Zhu Jian et al. 2018). Total iron concentrations were measured through the reduction of Fe 3þ to Fe 2þ by hydroxylamine. The concentration of Fe 3þ was calculated as the difference in concentrations between total iron and Fe 2þ . In addition, sodium fluoride was added to the Fe 3þ complex to eliminate interference.

Utilization of organic substrate
The consortium cells were grown for 5 days in a liquid modified 9 K medium containing 10 mM FeSO 4 .7H 2 O and 0.2% (w/v) yeast extract at pH 2.0, according to Johnson et al. (2009). Samples were collected after inoculation (0 h) and after 120 h in order to analyze the total organic carbon (TOC). The TOC was determined by the elemental carbon combustion method using a LECO CHN628 analyzer (Michigan, USA). Briefly, the samples were combusted at 950-1050 C under a constant O 2 stream, and the generated carbon gas was detected by infrared absorption.

Morphological and structural characterization
The precipitates were collected with a 0.45 mm filter paper through filtration, washed twice with the acidic-distilled water (pH 2.0), and lyophilized. Field Emission Scanning Electron Microscope (FE-SEM) and X-ray diffraction (XRD) analysis were performed to characterize the precipitates.

SEM characterization
Ultra-high-Resolution Field Emission Scanning Electron Microscope (FESEM) was used to assess the morphology of samples and verification of nanostructure and surface. SEM tests were carried out using a Thermo Scientific QUATTRO-S FE-SEM at 0.8, 3 and 6 kV, depending on magnification. Specimens were attached to a carbon tape and analyzed without using sputtering. SEM instrumentation was equipped with an EDS (Energy dispersive X-ray Spectroscopy) analyzer (Ultra Dry EDS, Thermo Fisher Scientific, Inc., UK).
XRD characterization XRD characterization was conducted with an X-ray diffractometer PANalitycal, model AERIS (Malvern Panalytical Ltda., Almedo, The Netherlands). The spectra were recorded using a Benchtop X-ray diffractometer with a copper laser and a PIXcel3D detector. Ni-filtered CuKa radiation (wavelength of 0.1542 nm) was produced at 40 kV and 15 mA. Scattered radiation was detected in the angular range of 10-80 (2h) at a rate of 5 /min. The data were analyzed using specialized software (Match and Origin). The diffraction patterns were compared to the reference patterns from Powder Diffraction File (PDF-2) database of the International Center for Diffraction Data (ICDD) with the search-match software.

Result and discussion
Consortium identification and diversity DNA of iron-oxidizing consortia was extracted and sequenced, giving detailed information about the bacterial diversity of each sample. In order to evaluate the metagenomic profile of the consortia, two conditions were strictly accomplished: Before sequencing, (i) the bacteria were collected from the exponential growth phase. And after sequencing (ii) the reads were obtained from sequences representing the hypervariable V3-V4 region of the 16S rRNA gene. This region provided high fidelity (in terms of readlength, throughput capacity, error rates, and classification efficiencies) and the longest read length obtained from 16S rRNA (Onywera and Meiring 2020). The sequencing quality reached a Phred quality Q20 with a base call accuracy of 91.5% and 92.2% in consortia M1-18 and M8-15, respectively which means that the precision of the base call (that is, the probability of a successful base call) is greater than 90%. 374,712 and 359,160 reads were obtained from consortia M1-18 and M8-15, respectively, as raw data from Illumina. We discarded more than 200,000 reads per sample based on the Q Phred values of quality by using default parameters of Cosmos ID metagenomic software, getting a total of 158,263 and 164,620 reads for consortia M8-15 and M1-18, respectively. One hundred five bacterial taxa at the genera level were detected for the consortium M8-15, from which only 21 represented !0.1% of relative abundance. On the other hand, 130 were detected for the consortium M1-18, where 23 represented !0.1% of relative abundance. The iron-oxidizing microbial community assigned at the genus level is displayed in Figure 1(a). Overall, the microbial communities in all two samples were dominated by Nitrospirae (91.9%, 88.76% for M8-15 and M1-18, respectively) at phyla levels, including Actinobacteria, Acidobacteria, Bacteroidetes, Chloroflexi, Euryarchaeota, Firmicutes, Gemmatimonadetes, Proteobacteria, Tenericutes, and Verrucomicrobia, as the most representative. At the genus level in both M8-15 and M1-18 samples, identification of bacterial groups revealed the presence of 49 genera in common, being dominated by Leptospirillum (91.72% and 88.63% for M8-15 and M1-18, respectively), and other with less abundant genera, varying relative abundance percentage across samples: Syntrophaceae (1.14 and 0.83%), Ferroplasma (0.97 and 0.72%) and Desulfurivibrio (0.71 and 0.51%). Mycobacterium and Thermogymnomonas were only detected in the M8-15 sample, accounting for 0.82-0.27% of the population. Whereas in the M1-18 sample it was only detected Ferrimicrobium and Acidimicrobiaceae, representing 4.29-0.54% of the population. Genera with relative abundance below 0.1% in the two samples were classified as 'others'. Overall, isolated consortia revealed different microbial communities, with nonshared populations representing 1.09 and 4.83% in M8-15 and M1-18 samples, respectively.
Based on the Chao1 estimator and Shannon index (Figure 1(b)), the structure of the bacterial population of M8-15 and M1-18 samples were significantly different. The total numbers of OTUs estimated by the Chao1 estimator were 121 (M8-15) and 146 (M1-18), indicating that M1-18 had the greatest richness while M8-15 had the lowest. The Shannon diversity index suggests how the abundance of each species is distributed (i.e., a higher value represents more diversity) among all the species in the community. M1-18 had the highest diversity (Shannon ¼ 1.4) among the M8-15 community (Shannon ¼ 1.3). Therefore, the variation in OUTs richness and Shannon diversity within each sample type was larger in M1-18 compared to M8-15, based principally on a diversity of Fe/S oxidizers bacteria and Fe/S reducers but with a low abundance percentage.
Iron oxidizing bacteria such as Leptospirillum and Ferroplasma were present in both consortia because they are bacteria commonly isolated from areas of mining processes where sulfur/iron oxidation usually occurs (Johnson et al. 2014;Mori et al. 2016). However, there was substantial variation in richness and diversity because OTUs capable of Fe or S reduction were more abundant in the M1-18 sample. The consortium M1-18 was isolated from a sample of mineral sediment, where OTUs capable of Fe/S reduction due to low oxygen levels are usually detected (Niu et al. 2016). Consequently, this generated a greater abundance of anaerobic or facultative anaerobic bacteria, such as bacteria belonging to the genus Ferrimicrobium and Acidimicrobiumrelated group, such as those found in the consortium M1-18. Indeed, the genus Acidimicrobium contains a single species, Acidimicrobium ferrooxidans, which is a moderate thermophile, and it can be found in distinct environments where the conditions are not optimal. For instance, Watkin et al. (2009) reported the isolation of the strain N39-30-03, corresponding to a sub-strain Acidimicrobium ferrooxidans DSM 10331, growing at 30 C in the bioleaching heap.
Genus Desulfurivibrio is one of the less abundant genera found in both consortia. Despite being a neutrophilic and strictly anaerobic sulfate-reducing bacteria, it is highlighted because its ability to use a variety of electron donors (e.g., lactate, acetate, pyruvate) and acceptors (e.g., Fe þ3 , NO À 3 ) (Lentini et al. 2012). Given the low S to Fe ratio in the culture medium during this study, the sulfate reduction could happen before or simultaneously with Fe þ3 reduction and continue with Fe-S cycling. Koschorreck (2008) showed that the pH might not be a significant factor in controlling sulfate-reducing bacteria (SRB), which are typically present and active in acidic mine tailings. In this sense, some authors propose a syntrophic relationship between the SRBs in mixed cultures because sulfate-reducing at a low pH is easy but very difficult in pure culture. However, the mechanism of interactions of SRB with other bacteria at a low pH remains unsolved (Kimura et al. 2006;Koschorreck 2008). In addition, Desulfurivibrio alkaliphilus is the only species validated by SRB genus. Its presence at low pH might be due to the acidic-tolerant decarboxylases enzymes in its genome (Melton et al. 2016). Moreover, Kjeldsen et al. (2004) reported that SRB tolerates oxygenic environments for up to 120 h with the same viability and maintained sulfate-reducing activity as the initial culture. Likewise, Desulfurivibrio alkaliphilus is present under oxygenic conditions. Therefore the presence of oxidoreductases and peroxidases in its genome might allow its adaptability and tolerance to oxygenic environments (Melton et al. 2016).

Iron concentration in precipitation process
In order to evaluate oxidation ability and the secondary iron precipitate formation of the two microbial communities, essays on Fe 2þ oxidation were carried out. Figure 2(a) illustrates the influence of the different microbial communities on Fe 2þ consumption. The consortia showed a similar decreasing trend, where Fe 2þ concentration decreased linearly in the first 48 h of the incubation period and then remained constant. However, a clear difference was observed at 24 hours of incubation. The Fe 2þ oxidation efficiency percentage of the M8-15 consortium was 46.4% within 24 h, while the M1-18 consortium was 78.1%, representing more than 30% of oxidation ability for M8-15. Moreover, the Fe 2þ oxidation rate of M1-18 consortium resulted faster ( k Fe 2þ ¼ 0.07 ± 0.004, r 2 ¼ 0.99) than the one inoculated with M8-15 consortium ( k Fe 2þ ¼ 0.04 ± 0 .009, r 2 ¼ 0.94) which is consistent with the oxidation ability of M1-18. Besides, it was noted that Fe 2þ in the solution had been oxidized at around 99% by both consortia at 48 h, which was similar to previous reports for A. ferrooxidans containing acidic medium (pH 2-3) (Gan et al. 2017;Zhou et al. 2007). Although O 2 could oxidize Fe 2þ to Fe 3þ , the pH of the culture was 2 during the incubation period. Hence the two consortia can use the Fe 2þ and O 2 at the same time to obtain energy for its growth (Hou et al. 2015;Wang X et al. 2012), being M1-18 the dominant consortium. After a continuous decrease in Fe 2þ concentration, Fe 3þ concentration in all two consortia reaction systems gradually increased, but the increase dynamics at two consortia showed a significant difference at 24 h. The Fe 3þ concentration produced by M8-15 and M1-18 were 305.5 mg/l and 452.3 mg/l, respectively. Despite the presence of FeOB and FeRB in both consortia, according to Figure 2(a) microbial oxidation was the most dominant. As the amount of Fe 2þ depleted in 48 h for both consortia, Fe 3þ gradually became the predominant Fe species in the culture medium. This was directly reflected in the dynamic changes in the total iron concentration (Figure 2(b)). The total Fe concentration decreased 10% for M1-18 and 14% for M8-15, remaining at constant values between 72-120 h. The decrease in total iron concentration is associated with the hydrolysis reaction and formation of Fe-precipitates (Hou et al. 2015;Huang and Zhou 2012;Song et al. 2018;Zhu Jianyu et al. 2013).
On the other hand, the microbial community growth during the Fe 2þ consumption is displayed in Figure 2(c), which shows the increase of the cell concentration as a function of time. During the lag phase (0-24 h), the Fe 2þ oxidation dominated the first stage of the reaction, and Fe þ3 concentration increased. After the lag phase, the growth curves reached an exponential growth phase between 24 and 96 h, where Fe þ3 concentration reached the highest concentration value. Subsequently, cell concentration decreased between 96 and 120 h. The microbial growth status showed a significant difference during this time frame because the cell concentration for the M1-18 consortium was 10.9% greater than the M8-15 consortium at 96 h and 23.8% at 120 h. In addition, after 450 h, the cellular concentration of the M1-18 consortium kept increasing to 3.3 Â 10 8 cells/ml, while the M8-15 consortium fell at values of 3.13 Â 10 7 cells/ml. The microbial community of the M1-18 consortium comprises not only autotrophic but also heterotrophs microorganisms with the ability to oxidize and reduce iron, such as those belonging to the genera Ferrimicrobium and Acidimicrobiacea. These heterotrophic bacteria could have reached high population densities due to the presence of organic substrates that derived from death autotrophic primary producers when grown as mixed cultures from iron oxidization/reducing (Gupta et al. 2019;Johnson et al. 2009).
On the other hand, Fe 2þ was consumed entirely during the exponential growth phase, and the Fe þ3 and the dissolved total Fe concentration decreased slightly during this phase, which shows that both consortia significantly promoted secondary iron precipitate formation through ferrous iron oxidation. Some researchers report that the secondary iron minerals formed during the Fe 2þ oxidation are associated with the microbial growth status and the oxidation ability. Different growth state corresponds to different consumption of necessary elements like Fe 2þ and monovalent cations, leading to the formation of different precipitates (Huang and Zhou 2012;Wang X et al. 2012;Zhu Jianyu et al. 2013). For instance, Huang and Zhou (2012) found that minerals such as schwertmannite, ferrihydrite, and other impurities of Fe(OH) 3 were formed by rapid oxidation of Fe 2þ , while jarosite minerals are formed by slow oxidation of Fe 2þ , evidencing that Fe 2þ oxidation rate influenced the formation of secondary iron minerals and their crystallization.
Therefore, the consortia could affect the formation of precipitates because they have different microbial communities with different Fe 2þ oxidation abilities, which play an essential role in the precipitate nucleation, grading, and the crystal growth process. As shown in insets in Figure 2(b), there is a distinctive feature in the precipitates formed by two consortia with different Fe 2þ oxidation abilities. Secondary iron mineral precipitate produced by the consortium M8-15 was uniform, loose, and slowly formed at the bottom of the flask, whereas that precipitate formed by the consortium M1-18 was dense, which resulted in the formation of a quick precipitate at the flask bottom.

Presence of heterotrophic microorganisms
The organic substrate was incorporated into the medium to promote the growth of heterotrophic microorganisms in both consortia and then characterized through TOC quantification. TOC contents were quantified by the combustion method from inoculated 9 K medium samples. Table 1 shows that TOC contents decreased after 5 days of incubation in both consortia. However, a significant difference was observed between the two consortia. The total assimilated carbon of the consortium M1-18 was 61% greater than the consortium M8-15. This could be due to the promotion of the genera Ferrimicrobium, which is present only in the consortium M1-18, capable of oxidation of ferrous iron, and reduction of ferric iron in the presence of organic carbon (Johnson et al. 2009).
Although the consortium M8-15 does not show heterotrophic microorganisms in the microbial community, the carbon consumption could be due to the presence of the genus Ferroplasma, a chemomixotrophic organism that usually uses inorganic and organic sources (Merino et al. 2016). Furthermore, both consortia include autotrophic Fe oxidizers bacteria (dominated by genus Leptospirillum), and its growth in media with carbon sources could have been favored by mixed cultures that generate secreted extracellular proteins (secretome) (Bobadilla Fazzini et al. 2011;Merino et al. 2016). These proteins enhance the interaction with ferrous iron and facilitate the transport of the substrate or the electron transport chain (Merino et al. 2016). Microorganisms such as A. thiooxidans (Bobadilla Fazzini et al. 2011) and A. ferrooxidans (Chi et al. 2007) have also shown this phenomenon.

Properties of iron precipitate
In order to determine why the two consortia form up different secondary iron precipitates morphology in terms of their removal ability and growth status, a comprehensive evaluation by FE-SEM, EDS, and XRD were conducted.

FE-SEM/EDS analysis
The morphology and elemental composition of the secondary iron minerals precipitates were studied with FE-SEM coupled with EDS. In Figure 3, the precipitates formed by the two consortia display an evident difference in surface morphology. As shown in SEM images in Figure 3(a), the crystals that are formed by the M1-18 consortium are pseudocubic with a smooth surface, and a dense coat of thin villous structure. In another scanning area, Figure 3(b) shows principally spheroids particles of uniform size with a diameter of approximately 1 mm covered by a villous structure. Figure 3(c) shows particles with a needle-shaped burrs with a length of approximately up to 220 nm and a width of approximately up to 40 nm. However, crystals formed by the consortium M8-15 were observed like a dense coat of thin villous structure on the surface of the mineral as can be seen in Figure 3(d-e), which would indicate a difference in the precipitates probably due to the different oxidizing ability of the consortia. Additionally, Figure 3(f) shows communities of curved rod-shaped bacteria corresponding to the consortium M1-18. Numerous studies have also reported these types of structures as those formed by the consortia in this study.
Regarding he micromorphological features on surface of the particles, in Figure 3, the particles are covered with needle-shaped burrs that correspond to schwertmannite species, and the pseudocubic shapes correspond to jarosite-like crystal structure, usually formed in acid mine environments (Hedrich et al. 2011;Wang X et al. 2012). The schwertmannite acquired by both consortia was consistent with the results obtained by Song et al. (2018), who found schwertmannite spheroids particles with sea urchin or chestnut shell morphology formed by A. ferrooxidans of an average size of 1.47 mm. Besides, S anchez-España et al. (2012) obtained schwertmannite spheroids with needles or whiskers of size 40-60 nm in diameter and 300-400 nm in length, which is similar to our results. Even the rounded spheroidal aggregates are observed not only in media synthetics, but also in natural environments (Lu et al. 2013). It was proposed that schwertmannite crystallite growth are generated by a radial aggregation of needles or whiskers, as well as by nucleation and crystallite growth around bacterial cells (Mori et al. 2016;S anchez-España et al. 2012). However, recent studies suggest that this arrangement can be entirely abiotic (Hedrich et al. 2011;S anchez-España et al. 2012). The pseudocubic crystals jarosite-like were in variable amounts in all the scanned areas, being apparently schwertmannite the dominant mineral phase, which might be due to that jarosite is primarily produced by transformation from metastable schwertmannite (Jim enez et al. 2019;S anchez-España et al. 2012;Wang H et al. 2006). Besides, rod shapes were observed, typical of the bacterial biomass in cultures containing soluble ferrous ions (Rojas-Chapana and Tributsch 2004). The visible presence of bacterial cells alongside the secondary iron minerals could also evidence their role as a template for mineral nucleation by adsorbing ions around the cellular surface membrane or cell wall (S anchez-Rom an et al. 2015).
EDX analysis detected elements Fe and S ( SO 2À 4 ) with a molar Fe/S ratio of 1.1 ± 0.02 and 3 ± 0.8 present in the minerals formed by consortia M1-18 and M8-15, respectively ( Table 2). The Fe/S molar ratio obtained might be related to the jarosite-like formation, which is consistent with the molar ratio range between 0.99 and 3.1 described by Wang H et al. (2006). Moreover, the element K was absent in precipitates formed by both consortia. This absence can be explained by its substitution by H 3 O þ . According to the thermodynamic data, the monovalent cations differed in terms of their jarosite-forming ability, specifically, K-jarosite (À3313.7 kJ/mol) <H 3 O þ -jarosite (3246.6 kJ/mol) <NH 4 -jarosite (À3095.0 kJ/mol) (Wang X et al. 2012). Furthermore, the oxygen excess in the composition of the mineral jarosite-like (58.5 ± 4.2 wt% and 55.1 ± 5.0 wt% for M1-18 and M8-15, respectively) could be compared to the theoretical value of 44 wt% (S anchez-España et al. 2012), which could corroborate the presence of the H 3 O þ cation. Likewise, H 3 O þ substitution could have been implicated because of K deficiency in the medium.
Although the morphological features of the precipitates present schwertmannite species and pseudocubic jarosite crystalline structures according to the SEM image analysis, the EDS analysisonly confirm the jarosite-like formation due to the Fe/S molar ratio obtained. Hence the formation of the jarosite precipitates might be generated by the transformation from schwertmannite to jarosite. It is well documented that low pH, over timescales from weeks to months, in the presence of an appropriate jarosite-directing cation (e.g., K þ , H 3 O þ , NH 4 þ or Na þ ) favors the transformation from schwertmannite, a metastable phase, to jarosite (Bai et al. 2012;Wang H et al. 2006). Therefore, we suggest that the consortia influence the secondary mineral precipitate  due to their direct metabolic activities and Fe sorption and nucleation reaction. The consortia M1-18 and M8-15 have different oxidation capacities of Fe 2þ and, consequently, a difference in catalyzing the transformation of Fe into precipitated secondary iron minerals.

X-ray diffraction (XRD) analysis
The XRD patterns of the secondary iron minerals precipitate formed by the consortia M1-18 and M8-15 are shown in Figure 4. In order to illustrate the mineral phase, XRD patterns were compared with a standard XRD pattern ( Figure 4). Phase identification reveals that the major crystalline phase formed by the M1-18 and M8-15 consortia was jarosite . Moreover, the three strongest characteristic peaks at 2h of 17.6 (d ¼ 5.103 Å), 28.84 (d ¼ 3.111 Å) and 29.15 (d ¼ 3.084 Å) corresponded to (012), (021) and (113) planes of jarosite, respectively. However, we observed a significant difference in the intensity of characteristic peaks between the patterns of the precipitates of the consortia M1-18 and M8-15. The maximum peak height of precipitate formed by the consortium M1-18 is approximately double that of the consortium M8-15. In addition, these are sharp and symmetrical, which indicates a high symmetry crystal structure and good crystallinity. In contrast, the maximum peak height for the precipitate formed by the consortium M8-15 demonstrates its low crystallinity. Although the secondary iron precipitates were uniformly identified as k-jarosite according to the PDF card, these results are not consistent with EDS results due to the absence of the K element in the mineral formed by both M1-18 and M8-15 consortia. It is assumed that the reason leading to this phenomenon is the substitution of monovalent cations, suggesting an early metastable phase. The mechanism of transformation from a metastable to a stable phase is unclear. However, some reports indicate that this occurs through the substitution of monovalent cations in the structure of the metastable phase. Monovalent cations would be incorporated into the structure by Fe þ3 substitution or absorbed onto the mineral surface (Jim enez et al. 2019).

Conclusions
The metagenomic analysis revealed that the found microorganism community includes autotrophic Fe oxidizers (dominated by Leptospirillum) in the consortium M8-15 and autotrophic/heterotrophic Fe oxidizers (those belonging to genera Ferrimicrobium and Acidimicrobiacea) in consortium M1-18. The oxidation capacity of M1-18 was higher than in M8-15, which indicates that the microbial composition intervened in the ferrous oxidation rate and, therefore, in the formation of secondary iron minerals, affecting their form, surface, percentage of elemental composition, and crystallinity of the precipitate. Moreover, FE-SEM, EDS, and XRD analysis revealed that the consortia diversity promoted the mineralogical formation of the secondary iron precipitates in the form of schwertmannite as an early metastable phase. Moreover, the symmetry crystal structure and crystallinity of the precipitate produced by the consortium M1-18 were higher than the precipitate produced by the consortium M8-15, indicating distinct precipitates. In addition, all the results obtained in this study revealed that bacterial community diversity affect and promote the differential mineralogical formation of schwertmannite/jarosite, exerting significant control on the geochemistry of AMDcontaminated systems. This comprehension of differential mineralogical formation offers new insights into the correlation between microbial diversity and control by precipitates of biotechnological performance during industrial mineral oxidation.