Cyanobacterial blooms lead to deterioration of the water quality of drinking water reservoirs and recreational waters, and can cause severe ecological and economic damage. In particular, toxins produced by cyanobacteria, known as cyanotoxins, can be directly harmful to many organisms. Cyanobacterial blooms are fueled by anthropogenic eutrophication and climate change, and have become increasingly problematic over the past decades. An effective way to reduce cyanobacterial blooms is through the long-term reduction of nutrient input. However, efforts to reduce nutrient loads from catchment areas and control the internal nutrient release from sediments have proven to be very challenging. Therefore, the demand is growing for short-term emergency measures that rapidly suppress toxic cyanobacterial blooms in recreational lakes and drinking water reservoirs, preferably with the least possible impact on the ecosystem. Hydrogen peroxide (H₂O₂) is a promising candidate for an emergency method as cyanobacteria are very sensitive to this oxidative stressor, while other organisms are less affected. Furthermore, H₂O₂ is degraded biotically and abiotically to water and oxygen, leaving no chemical traces behind. However, knowledge about the efficacy of H₂O₂ treatments and their effects on other organisms in the ecosystem is still in its infancy. This thesis investigates these knowledge gaps by addressing the following questions:

1. How do variations in light and nutrient availability influence the effectiveness of H₂O₂ in treating cyanobacteria? (Chapter 2 and 3)
2. How do natural populations of cyanobacterial species respond to H₂O₂ treatments of entire lakes? (Chapter 4 and 5)
3. What is the fate of cyanotoxins during H₂O₂ treatments? (Chapter 2, 3, and 5)
4. What are the impacts of H₂O₂ treatments of entire lakes on non-target organisms such as the microbial community, eukaryotic phytoplankton, and zooplankton? (Chapter 4 and 5)

The first question was addressed through a series of laboratory batch experiments, where the Microcystis strain PCC 7806 was exposed to various light intensities and treated with different H₂O₂ concentrations. Light intensity was found to strongly enhance the effectiveness of an H₂O₂ treatment. The photosynthetic vitality of Microcystis PCC 7806 decreased faster when exposed to higher light intensities, requiring lower H₂O₂ concentrations to achieve similar suppression. In parallel, cell size measurements as well as intra- and extracellular toxin concentrations indicated that cellular damage occurred at lower H₂O₂ treatment concentrations when exposed to higher light intensities. H₂O₂ degradation rates were not affected by the light intensity. Exposure to orange light, which leads to more energy reaching PS II than exposure to blue or green light, resulted in a more effective treatment. Our results provide evidence that stronger light exposure, i.e. more light energy reaching PS II, during a H₂O₂ treatment results in increased photoinhibition and therefore induces faster cell damage. Consequently, a H₂O₂ treatment during sunny days or a treatment of a surface bloom will be more effective than a treatment during cloudy days or a treatment of a bloom at deeper layers of the waterbody.

The same experimental setup was used to compare the H₂O₂ treatment effectiveness under nutrient-replete conditions with the effectiveness under nitrogen (N) limitation and phosphorus (P) limitation. Microcystis PCC 7806 was first acclimatized to these specific growth conditions and subsequently treated in batch cultures with a range of H₂O₂ concentrations under both high light and low light conditions. The N-limited and the P-limited cultures were less sensitive to the H₂O₂ treatment than the nutrient-replete cultures under low light conditions, whereas the differences were much less pronounced under high light exposure. Gene expression analyses revealed that genes involved in anti-oxidative stress response were more highly expressed in nutrient-limited cultures than in nutrient-replete cultures, providing a head start in oxidative stress response. While nutrient limitation at high light intensities (i.e., on sunny days or at the surface) does not impact the treatment effectiveness, it might play a role in treatments at much lower light intensities.

The second question was addressed by conducting entire lake treatments with low concentrations of H₂O₂. In total, five lake treatments were performed in three lakes. The dominant cyanobacterial species included Dolichospermum spp., Planktothrix rubescens, Aphanizomenon spp., and Planktothrix agardhii. In all treatments, the cyanobacteria were much more sensitive to H₂O₂ than eukaryotic phytoplankton, and the cyanobacterial biovolume strongly declined between 84 and 100% after the treatments. Dolichospermum spp. and Aphanizomenon spp. were very sensitive and rapidly declined, while Planktothrix rubescens declined more slowly and Planktothrix agardhii even recovered several days after the treatment. The treatment was likely less effective for those cyanobacterial species that also thrive in deeper layers of the water column, where light intensities are much lower.

The third question was addressed through both laboratory experiments and whole-lake treatments. Laboratory results indicated that cyanotoxins, specifically microcystins, leak from cells upon cellular damage. This means that a more efficient treatment will result in a faster decrease of intracellular toxins, while the release of toxins during a less efficient treatment will occur more gradually. The total concentration of detectable toxins decreased within 24 h after the treatment compared to the pre-treatment concentrations. In the lake treatments, the cyanotoxins were first released by the lysing blooms and then declined by degradation, dilution or temporary binding to proteins. Overall, cyanotoxins disappeared within a few days after the removal of the cyanotoxin-producing cyanobacterial species.

The fourth question was addressed through research conducted during field experiments. Non-target organisms were monitored before, during, and after the H₂O₂ treatment of entire lakes. A drastic decrease of the cyanobacterial bloom in a lake will create new environmental conditions to which non-target organisms have to adapt. Eukaryotic phytoplankton experienced minimal negative effects and, for the majority of the H₂O₂ treatments, either remained unaffected or even benefited from the new conditions. Some species of zooplankton, however, were more sensitive to H₂O₂. Rotifers were frequently impacted and appeared to be the most sensitive zooplankton group. Cladocerans only experienced significant declines during treatments with slightly higher H₂O₂ exposure, while copepods were the most resistant group that did not suffer significant losses during any of the treatments.

The microbial community was monitored during two lake treatments of Lake Oosterduinsemeer, while a parallel incubation experiment offered control measurements and exposure of the same lake water to a similar H₂O₂ concentration as in the lake treatments (2.5 mg/L) and to a higher H₂O₂ concentration (10 mg/L). H₂O₂ exposure was clearly noticeable as short-term pulse disturbances of the alpha and beta diversity of the microbial community 24 h after the two lake treatments. However, the microbial community showed resilience and recovered within days after the lake treatments. At the higher H₂O₂ concentration applied in the incubation experiment (10 mg/L H₂O₂), the resilience of the microbial community was reduced, although the treatment-induced changes in microbial community structure were still relatively small compared to the naturally occurring seasonal variation observed between the two lake treatments.

This thesis contributes to a better understanding of the use of H₂O₂ to treat cyanobacterial blooms. The results demonstrate that, under the right circumstances, H₂O₂ is a highly effective tool to rapidly suppress cyanobacterial blooms. Cyanobacterial abundances stayed low for several weeks after the treatments. Thereafter, however, new cyanobacterial blooms appeared, which illustrates that H₂O₂ is an emergency method but not a long-term solution to control blooms. Based on the obtained results, this thesis makes a number of recommendations. First, lake treatments with H₂O₂ will be most effective in combination with high light exposure (i.e., during sunny days), while slightly higher H₂O₂ concentrations would be required to achieve the same effect during cloudy days or for treatments of blooms at deeper layers of the waterbody. Second, although effects of the added H₂O₂ on most non-target organisms are minimal, some zooplankton taxa (especially rotifers) are sensitive to the H₂O₂ concentrations applied in lake treatments. Therefore, high H₂O₂ concentrations during the treatments should be avoided, and we advise to run incubation tests with lake water prior to the treatment to assess the sensitivity of the cyanobacteria and zooplankton in the lake to the applied H₂O₂ concentration. Third, if the bloom is toxic, the lysing cells will release their cyanotoxins after H₂O₂ addition. Although our lake results showed that cyanotoxin concentrations rapidly declined within days after the treated cyanobacteria had disappeared, careful monitoring of the toxins following treatments of toxic blooms is therefore recommended. In conclusion, with these recommendations in mind, H₂O₂ serves as a highly successful, selective, and sustainable emergency method to temporarily free lakes from cyanobacterial blooms.

More information