Cyanobacterial blooms are increasingly reported worldwide, presenting a challenge to water treatment plants and concerning risks to human health and aquatic ecosystems. Advanced oxidative processes comprise efficient and safe methods for water treatment. Hydrogen peroxide (H2O2) has been proposed as a sustainable solution to mitigate bloom-forming cyanobacteria since this group presents a higher sensitivity compared to other phytoplankton, with no major risks to the environment at low concentrations. Here, we evaluated the effects of a single H2O2 addition (10 mg L-1) over 120 h in mesocosms introduced in a reservoir located in a semi-arid region presenting a Planktothrix-dominated cyanobacterial bloom. We followed changes in physical and chemical parameters and in the bacterioplankton composition. H2O2 efficiently suppressed cyanobacteria, green algae, and diatoms over 72 h, leading to an increase in transparency and dissolved organic carbon, and a decrease in dissolved oxygen and pH, while nutrient concentrations were not affected. After 120 h, cyanobacterial abundance remained low and green algae became dominant. 16S rRNA sequencing revealed that the original cyanobacterial bloom was composed by Planktothrix, Cyanobium and Microcystis. Only Cyanobium increased in relative abundance at 120 h, suggesting regrowth. A prominent change in the composition of heterotrophic bacteria was observed with Exiguobacterium, Paracoccus and Deinococcus becoming the most abundant genera after the H2O2 treatment. Our results indicate that this approach is efficient in suppressing cyanobacterial blooms and improving water quality in tropical environments. Monitoring changes in abiotic parameters and the relative abundance of specific bacterial taxa could be used to anticipate the regrowth of cyanobacteria after H2O2 degradation and to indicate where in the reservoir H2O2 should be applied so the effects are still felt in the water treatment plant intake.

Since conventional drinking water treatments applied in different countries are inefficient at eliminating potentially toxic cyanobacterial peptides, a number of bacteria have been studied as an alternative to biological filters for the removal of microcystins (MCs). Here, we evaluated the degradation of not only MCs variants (-LR/DM-LR/-RR/-LF/-YR), but also non-MCs peptides (anabaenopeptins A/B, aerucyclamides A/D) by Paucibactertoxinivorans over 7 days. We also evaluated the degradation rate of MC-LR in a peptide mix, with all peptides tested, and in the presence of M. aeruginosa crude extract. Furthermore, biodegradation was assessed for non-cyanobacterial peptides with different chemical structures, such as cyclosporin A, (Glu1)-fibrinopeptide-B, leucine-enkephalin, and oxytocin. When cyanopeptides were individually added, P. toxinivorans degraded them (99%) over 7 days, except for MC-LR and -RR, which decreased by about 85 and 90%, respectively. The degradation rate of MC-LR decreased in the peptide mix compared to an individual compound, however, in the presence of the Microcystis extract, it was degraded considerably faster (3 days). It was noted that biodegradation rates decreased in the mix for all MCs while non-MCs peptides were immediately degraded. UPLC–QTOF–MS/MS allowed us to identify two linear biodegradation products for MC-LR and MC-YR, and one for MC-LF. Furthermore, P. toxinivorans demonstrated complete degradation of non-cyanobacterial peptides, with the exception of oxytocin, where around 50% remained after 7 days. Thus, although P. toxinivorans was previously identified as a MC-degrader, it also degrades a wide range of peptides under a range of conditions, which could be optimized as a potential biological tool for water treatment.

Cyanobacteria and their toxic secondary metabolites present challenges for water treatment globally. In this study we have assessed TiO₂ immobilized onto recycled foamed glass beads by a facile calcination method, combined in treatment units with 365 nm UV-LEDs. The treatment system was deployed in mesocosms within a eutrophic Brazilian drinking water reservoir. The treatment units were deployed for 7 days and suppressed cyanobacterial abundance by 85% while at the same time enhancing other water quality parameters; turbidity and transparency improved by 40 and 81% respectively. Genomic analysis of the microbiota in the treated mesocosms revealed that the composition of the cyanobacterial community was affected and the abundance of Bacteroidetes and Proteobacteria increased during cyanobacterial suppression. The effect of the treatment on zooplankton and other eukaryotes was also monitored. The abundance of zooplankton decreased while Chrysophyte and Alveolata loadings increased. The results of this proof-of-concept study demonstrate the potential for full-scale, in-reservoir application of advanced oxidation processes as complementary water treatment processes.