Purple phototrophic bacteria (PPB) show an underexplored potential for resource recovery from wastewater. Raceway reactors offer a more affordable full-scale solution on wastewater and enable useful additional aerobic processes. Current mathematical models of PPB systems provide useful mechanistic insights, but do not represent the full metabolic versatility of PPB and thus require further advancement to simulate the process for technology development and control. In this study, a new modelling approach for PPB that integrates the photoheterotrophic, and both anaerobic and aerobic chemoheterotrophic metabolic pathways through an empirical parallel metabolic growth constant was proposed. It aimed the modelling of microbial selection dynamics in competition with aerobic and anaerobic microbial community under different operational scenarios. A sensitivity analysis was carried out to identify the most influential parameters within the model and calibrate them based on experimental data. Process perturbation scenarios were simulated, which showed a good performance of the model.

Purple non-sulphur bacteria (PNSB) are an emerging group of microbes attractive for applied microbiology applications such as wastewater treatment, plant biostimulants, microbial protein, polyhydroxyalkanoates and H₂ production. These photoorganoheterotrophic microbes have the unique ability to grow selectively on organic carbon in anaerobic photobioreactors. This so-called selectivity implies that the microbial community will have a low diversity and a high abundance of a particular PNSB species. Recently, it has been shown that certain PNSB strains can produce antimicrobials, yet it remains unclear whether these contribute to competitive inhibition. This research aimed to understand which type of antimicrobial PNSB produce and identify whether these compounds contribute to their selective growth. Mining 166 publicly-available PNSB genomes using the computational tool BAGEL showed that 59% contained antimicrobial encoding regions, more specifically biosynthetic clusters of bacteriocins and non-ribosomal peptide synthetases. Inter- and intra-species inhibition was observed in agar spot assays for Rhodobacter blasticus EBR2 and Rhodopseudomonas palustris EBE1 with inhibition zones of, respectively, 5.1 and 1.5–5.7 mm. Peptidomic analysis detected a peptide fragment in the supernatant (SVLQLLR) that had a 100% percentage identity match with a known non-ribosomal peptide synthetase with antimicrobial activity.

Purple non-sulfur bacteria (PNSB) show potential for microbial protein production on wastewater as animal feed. They offer good selectivity (i.e., low microbial diversity and high abundance of one species) when grown anaerobically in the light. However, the cost of closed anaerobic photobioreactors is prohibitive for protein production. Although open raceway reactors are cheaper, their feasibility to selectively grow PNSB is thus far unexplored. This study developed operational strategies to boost PNSB abundance in the biomass of a raceway reactor fed with volatile fatty acids. For a flask reactor run at a 2 day sludge retention time (SRT), matching the chemical oxygen demand (COD) loading rate to the removal rate in the light period prevented substrate availability during the dark period and increased the PNSB abundance from 50-67 to 88-94%. A raceway reactor run at a 2 day SRT showed an increased PNSB abundance from 14 to 56% when oxygen supply was reduced (no stirring at night). The best performance was achieved at the highest surface-to-volume ratio (10 m2 m-3 increased light availability) showing productivities up to 0.2 g protein L-1 day-1 and a PNSB abundance of 78%. This study pioneered in PNSB-based microbial protein production in raceway reactors, yielding high selectivity while avoiding the combined availability of oxygen, COD, and darkness.

Mixed-culture biotechnologies are widely used to capture nutrients from wastewater. Purple non-sulfur bacteria (PNSB), a guild of anoxygenic photomixotrophic organisms, rise interest for their ability to directly assimilate nutrients in the biomass. One challenge targets the aggregation and accumulation of PNSB biomass to separate it from the treated water. Our aim was to enrich and produce a concentrated, fast-settling PNSB biomass with high nutrient removal capacity in a 1.5-L, stirred-tank, anaerobic sequencing-batch photobioreactor (SBR). PNSB were rapidly enriched after inoculation with activated sludge at 0.1 gVSS L^–1 in a first batch of 24 h under continuous irradiance of infrared (IR) light (>700 nm) at 375 W m^–2, with Rhodobacter reaching 54% of amplicon sequencing read counts. SBR operations with decreasing hydraulic retention times (48 to 16 h, i.e., 1–3 cycles d^–1) and increasing volumetric organic loading rates (0.2–1.3 kg COD d^–1 m^–3) stimulated biomass aggregation, settling, and accumulation in the system, reaching as high as 3.8 g VSS L^–1. The sludge retention time (SRT) increased freely from 2.5 to 11 days. Acetate, ammonium, and orthophosphate were removed up to 96% at a rate of 1.1 kg COD d^–1 m^–3, 77% at 113 g N d^–1 m^–3, and 73% at 15 g P d^–1 m^–3, respectively, with COD:N:P assimilation ratio of 100:6.7:0.9 m/m/m. SBR regime shifts sequentially selected for Rhodobacter (90%) under shorter SRT and non-limiting concentration of acetate during reaction phases, for Rhodopseudomonas (70%) under longer SRT and acetate limitation during reaction, and Blastochloris (10%) under higher biomass concentrations, underlying competition for substrate and photons in the PNSB guild. With SBR operations we produced a fast-settling biomass, highly (>90%) enriched in PNSB. A high nutrient removal was achieved by biomass assimilation, reaching the European nutrient discharge limits. We opened further insights on the microbial ecology of PNSB-based processes for water resource recovery.