| Summary: | Mining for the recovery of minerals and coal can result in acid mine drainage (AMD) which presents an environmental risk. Acid mine drainage, as the name suggests, is acidic run-off water from mostly mine waste dumps. It affects water quality by lowering its pH and increasing its metal and sulphate loading, thus making it unsuitable for use by many forms of life. AMD must therefore be treated before entering nearby water systems and soils. An effective treatment technology is considered as the one that can result in water neutralisation and removal of metals and sulphate. Biological sulphate reduction (BSR) technologies, mediated by sulphate reducing bacteria (SRB), have attracted attention as a sulphate remediation strategy as they offer a cheap alternative to other sulphate removal technologies such as chemical approaches. In addition, the concomitant generation of alkalinity and soluble sulphide assist in neutralisation and heavy metal removal. One of the challenges associated with BSR is the supply of a cost-effective carbon source which also acts as an electron donor for the anaerobic reduction of sulphate. Studies have reported that both the choice of carbon source and electron donor and the microbial communities present influence the sulphate reduction process, the former frequently defining technoeconomic feasibility. The feed sulphate concentration and residence time, together defining the volumetric sulphate loading rate, have also been reported to influence the efficacy of the sulphate reduction process and needs to be optimised for the microbial community present and the chosen electron donor. The identification and characterisation of the microbial communities involved and investigating how these change with changes in operating conditions is crucial in the optimisation of BSR processes. Currently, there are no commonly used molecular tools which can be used for routine analysis of SRB communities in real time and on a regular basis and cost effectively. This makes it difficult to understand the link between changes in the mixed BSR microbial community structure and process performance. The study presented in this thesis had three main objectives. Firstly, to evaluate the use of an anaerobic digestate, obtained from a partially anaerobically digested Cyanobacteria species (Arthrospira platensis, commonly known as Spirulina), as a carbon source and electron donor for BSR. Secondly, to validate, optimise and apply the molecular tools for analysis of the relative abundances of species within the mixed BSR microbial community in this study. Thirdly, to compare the microbial community dynamics and performance of BSR using the complex anaerobic digestate as carbon source and electron donor to BSR using a single electron donor source, lactate. Chemostat studies using a mixed SRB consortium were carried out using anaerobic digestate, characterised as containing a mixture of acetate, propionate and butyrate, as a carbon source and electron donor for BSR. Upon reaching steady-state, the concentrations of sulphate, bicarbonate, acetate, propionate and butyrate were measured and used to estimate the BSR kinetics and reaction stoichiometry. A 16S rRNA gene survey of the BSR inoculum used for this thesis was performed by constructing a 16S rRNA gene clone library and analysis of the diversity of clones was performed using amplified ribosomal DNA restriction analysis (ARDRA). These 16S rRNA sequences were used to provide insight into the diversity and phylogenetic relatedness of the bacterial community and key species within the mixed BSR inoculum. In silico analysis of the 16S rRNA sequences captured from the clone library was performed to design novel genus specific quantitative real-time PCR (qPCR) primers and to validate the specificity of previously published primers. Fluorescence in situ hybridisation (FISH) techniques were optimised for the visual characterisation of this microbial community. FISH and qPCR were then applied to assess how the mixed microbial community structure was affected by the changes in the volumetric sulphate loading rate (VSLR), mediated through dilution rate and feed sulphate concentration, when anaerobic digestate (mixed carbon source) and lactate (simple carbon source) were used as an electron donor for BSR. The results obtained were used to examine and compare the link between microbial community dynamics and performance of sulphate reducers between the mixed and the simple carbon source. The results obtained from this thesis suggested the simultaneous utilisation of all the three volatile fatty acids (acetate, propionate and butyrate) present in anaerobic digestate which contributed to the robustness of the chemostat reactors as indicated by higher sulphate, propionate and butyrate conversion efficiencies. The kinetic profiles of the volumetric sulphate reduction rate (VSRR) obtained with anaerobic digestate were well matched with the kinetics observed in previous studies when single carbon sources and electron donors were used for BSR. At a feed sulphate concentration of 1.0 g l-1 , the oxidation of acetate, propionate and butyrate and concomitant sulphate reduction were observed across the dilution rates of 0.0083 to 0.083 h -1 . The stoichiometry of BSR utilising propionate and butyrate as carbon and electron donor suggested that by increasing feed sulphate concentrations from 1.0 to 2.5 and 5.0 g l-1 acetogenic reactions were favoured at the higher dilution rates of 0.042 and 0.083 h-1 . However, increasing the feed sulphate concentration at the lower dilution rates of 0.0083 to 0.021 h-1 did not alter the oxidation of volatile fatty acids (VFAs) and concomitant sulphate reduction, suggesting that the sensitivity of the propionate and butyrate oxidisers was related to specific growth rate rather than the sulphate loading. A previous mathematical model developed by Moosa et al. (2002) was used to determine microbial growth constants (μmax and Ks) and energetic coefficients (Yx/s) for SRB at each feed sulphate concentration to describe the microbial growth kinetics obtained with anaerobic digestate. A 16S rRNA gene survey, performed by 16S rRNA library construction and 16S rRNA gene amplicon sequencing, revealed a more diverse microbial community in the inoculum obtained from a lactate operated BSR reactor than previously reported. qPCR was used to confirm the presence and relative abundance of these species within the reactors receiving anaerobic digestate or lactate as carbon source and electron donor. The 16S rRNA sequences captured were found to have high similarity to well- known SRB species belonging to the Desulfomicrobium, Desulfovibrio, Desulfuromonas, Desulfobulbus and Desulfocurvus genera. Other “non-traditional SRB” species belonging to the Firmicutes and Citrobacter genera containing a specific molecular target for the detection of SRBs, the dissimilatory sulphite reductase gene (dsrAB), within their genomes were also detected. DsrAB is the key enzyme catalysing the last and main energy-generating step during sulphate reduction. Non-SRB species present were identified as members of the Sphaerochaeta, Synergistetes, Chloroflexi, Mesotoga, Acholeplasma, Bacteriodetes, Petrimonas and Bacteriodes genera. A 16S rRNA gene survey by 16S rRNA variable region amplification from metagenomic DNA extracted from microbial biomass associated with continuous stirred tank reactors (CSTRs) operated on anaerobic digestate or lactate was performed to validate the qPCR results and assist with the identification of the “other SRB” and nonSRB species. The 16S rRNA gene survey suggested the presence of 13 known SRB species Desulfomicrobium groups (Desulfomicrobium hypogeium and Desulfomicrobium aestuarii), Desulfovibrio species (D. aminophilus, D. vulgaris, D. desulfuricans, D. intestinalis, D. oxamicus, and D. sulfodismutans), Desulfobulbus oligotrophicus, Desulfocurvus vexinensis, Desulfococcus biacutus, Desulfarculus baarsii, Desulfomonile tiedjei and Desulfobacca acetoxidans in CSTRs operated on anaerobic digestate. Only up to 10 SRB species, Desulfomicrobium hypogeium, Desulfomicrobium aestuarii, Desulfovibrio groups (Desulfovibrio aminophilus, Desulfovibrio vulgaris, Desulfovibrio desulfuricans, Desulfovibrio intestinalis, Desulfovibrio sulfodismutans, Desulfovibrio mexicanus, Desulfobulbus oligotrophicus and Desulfocurvus vexinensis were observed in reactors with lactate, suggesting that the multiple VFAs present in the anaerobic digestate (acetate, propionate and butyrate) were able to support a more diverse SRB community than a single electron donor (lactate). Various non-SRB bacterial genera as well as known elemental sulphur reducing bacteria Desulfuromonas acetexigens and Dethiosulfovibrio acidaminovorans were also found to be present, with the latter being associated only with the lactate operated reactor. qPCR results indicated that despite being present in high proportions at the lowest VSLRs, the Desulfomicrobium species were washed out of the reactors at higher VSLRs regardless of carbon source and electron donor was provided. Species from the Desulfovibrio genera, which were present at lower abundances than the Desulfomicrobium species, were more resistant to changes in dilution rates and remained present within the reactors at the higher VSLRs, 0.104 and 0.208 g l-1 h -1 . In the reactors operated on anaerobic digestate, the decline in the abundance of Desulfovibrio species at VSLRs of 0.052 and 0.104 g l-1 h -1 , correlated with a noticeable decline in sulphate conversion from 60.4 to 49.4% at feed sulphate of 2.5 g l-1 , and from 66.9 to 22.6% at feed sulphate of 5.0 g l-1 . These findings suggest that Desulfovibrio species may play a critical role in sustained sulphate reduction at lower VSLRs. 16S rRNA gene amplicon data validated the qPCR data showing that increasing the VSLR, resulted in a change in the SRB community structure and a decrease in the proportion of total SRB within the microbial community. In agreement with the FISH and qPCR findings, Desulfomicrobium hypogeium was identified as the most abundant operational taxonomic unit (OTU) belonging to SRB present at the lowest dilution rate (D) tested (0.0083 h-1 , retention time (RT = 1/D) of 5 d) when anaerobic digestate was used as an electron donor for BSR. Washout of most SRB species was also observed when the dilution rate was increased from 0.0083 to 0.042 h-1 (RT of 5 to 1 d) in these reactors. Species such as Desulfovibrio sulfodismutans, Desulfomonile tiedjei, the acetate oxidiser Desulfococcus biacutus and the elemental sulphur reducing Desulfuromonas acetexigens were found to tolerate higher VSLRs of 0.104 and 0.208 g l-1 h -1 (dilution rate of 0.042 h-1 ), suggesting fast enough growth rates to remain in these reactors at the higher dilution rate of 0.042 h-1 . A decrease in the abundance of the incomplete propionate oxidiser Desulfobulbus oligotrophicus correlated to a decrease in propionate oxidation at a VSLR of 0.104 and 0.208 g l-1 h -1 suggesting that this SRB was responsible for the oxidation of propionate and concomitant sulphate reduction observed in these reactors. Similar to the reactors receiving anaerobic digestate, increasing the dilution rate from 0.0083 to 0.042_h -1 (RT of 5 to 1 d) resulted in washout of most SRB OTUs in CSTRs operated on lactate. At a feed sulphate concentration of 10.0 g l-1 , increasing the dilution rate from 0.0083 to 0.042 h-1 resulted in an increase in the proportion of the lactate oxidiser Desulfocurvus vexinensis from 25 to 98% of the total SRB proportion. At this dilution rate (0.042 h-1 ), other SRB species observed were the lactate oxidisers Desulfovibrio sulfodismutans and Desulfobulbus oligotrophicus which can oxidise lactate and the product of its incomplete oxidation, propionate. Although the abundance of these two SRB at the dilution rate of 0.042 h-1 was much lower than that of Desulfocurvus vexinensis, studies with anaerobic digestate suggested Desulfobulbus oligotrophicus which was abundant at only 0.004% and was identified as the only propionate degrader in the CSTR resulted in propionate conversion of 21.7%. This suggested that the less abundant Desulfovibrio sulfodismutans and Desulfobulbus oligotrophicus may have also played a role in sulphate reduction at the dilution rate of 0.042 h-1 in the CSTR with lactate. In addition, Desulfocurvus vexinensis and Desulfobulbus oligotrophicus were able to function at a VSLR of 0.42 g l-1 h -1 which suggests these two SRB species could be used effectively to reduce sulphate to hydrogen sulphide in wastewaters with higher VSLRs of up to 0.42 g l-1 h -1 when lactate was provided as an electron donor for BSR. The acetate specialist, Desulfobacca acetoxidans, the butyrate oxidiser Desulfarculus baarsii and the propionate oxidiser Desulfobulbus oligotrophicus, were able to function at a VSLR of 0.208 g l-1 h -1 suggesting that a combination of these three SRB species could be used in BSR treatment processes with VSLRs of up to 0.208 g l-1 h -1 where anaerobic digestate is provided as an electron donor. The ability for anaerobic digestate to support diverse SRB communities even at higher VSLRs may add to the robustness of the reactors to maintain sulphate reduction even at high VSLRs. This thesis showed that both the presence and diversity of SRB species are subject to the carbon source and VSLR. To the author's knowledge, this is the first study to indicate the relationship between the change in SRB community structure and sulphate reduction performance when anaerobic digestate (a complex carbon source) is used as a carbon source and electron donor for BSR. Results from this thesis suggest that the use of a mixed volatile fatty acid stream generated for the partial digestion of a suitably digestible biomass may be used as electron donor and carbon source to support a robust BSR process for the treatment of AMD. Using a mixed volatile fatty acid stream also has potential to result in the development of a more economically viable AMD treatment process.
|
|---|
