UC Berkeley

Sunlight inactivation of microorganisms is a natural process that can occur in any sunlit water and has implications for microbial ecology, the fate of microbial contaminants in the environment, and natural wastewater treatment systems. This work focuses on sunlight inactivation of waterborne viruses, which present a special challenge in health-related water microbiology given that waterborne human viruses are important etiologies of disease, and are difficult to remove and inactivate in water. The research presented in this dissertation builds upon previous research by determining the mechanisms and rates of sunlight inactivation of human viruses in natural waters; these data were used to build predictive inactivation rate models that account for virus type, sunlight irradiance and water quality. Additionally, two applications of sunlight inactivation - the differential stability of indicator organisms and human viruses in the environment, and disinfection of irrigation water in wastewater-irrigated agriculture - were investigated.

There are three proposed sunlight inactivation mechanisms for viruses: the direct- and indirect-endogenous mechanisms, which require absorption of photons by virus components, and the exogenous mechanism, which involves reaction between the virus and exogenously produced reactive intermediates formed by photochemical reactions. All three mechanisms are affected by water quality. Natural organic matter, for example, is found in most aquatic environments and is capable of both attenuating sunlight - which would decrease sunlight exposure and therefore inactivation rates - and photosensitizing production of reactive intermediates - which could increase inactivation rates of viruses susceptible to the exogenous mechanism.

The first goal of this dissertation was to better understand the mechanisms of sunlight inactivation of select bacteriophage and human viruses in surface waters containing natural organic matter. Given the two main modes by which damage is delivered to viruses in sunlit surface waters - through direct absorption of photons (the direct- and indirect-endogenous mechanisms) and contact with reactive molecules formed by sensitizers in the water column (the exogenous mechanism) - a better understanding of sunlight inactivation mechanisms can help us predict how environmental conditions (e.g., sunlight irradiance, light attenuation, water quality, depth, mixing) can affect observed inactivation rates.

Through laboratory experiments using simulated sunlight and natural organic matter-containing waters collected from the environment, we determined that the sunlight inactivation rates of human poliovirus type 3 and bacteriophage PRD1 mainly involved endogenous inactivation mechanisms, while MS2 and human adenovirus type 2 were also affected by the exogenous mechanism. Different virus types were found to have different rates of inactivation, and inactivation rates differed among water types depending on light attenuation and natural organic matter in the water source; this finding made it clear that water quality conditions and virus type must be taken into account when predicting inactivation rates. Additionally, MS2 was inactivated at the slowest rate in all waters, and PV3 the fastest, at the water depth that was studied (i.e., 5 cm).

The data obtained in the initial virus inactivation study, along with findings from other researchers, were used to develop and test models that predict the sunlight inactivation rates of viruses in surface waters containing light-attenuating photosensitizers. Models were developed for poliovirus type 3 and MS2. Inactivation rates were modeled and measured in reactors containing different, well-mixed depths of water collected from an open-water wastewater treatment system; the models performed well in predicting inactivation rates in laboratory experiments. Given the tradeoff between decreased endogenous inactivation due to light attenuation and increased exogenous inactivation due to the presence of photosensitizers in natural organic matter-containing water, we compared the measured and modeled inactivation rates of poliovirus type 3 (which is reliant on endogenous mechanisms) and MS2 (which is susceptible to the exogenous mechanism) at different well-mixed depths. The inactivation rate of poliovirus type 3 was found to be lower than that of MS2 at deeper well-mixed depths, indicating that MS2 cannot be considered a conservative indicator of poliovirus sunlight inactivation under all conditions. Some research gaps - including the development of models that take into account the solar zenith angle, annual and diurnal variation in irradiance, and water body hydraulics - must be filled to successfully translate the models to environmental waters and natural sunlight.

Many viruses are not culturable, making them difficult to study in general, let alone their inactivation rates. To better understand factors that dictate differences in sunlight inactivation rates between viruses, which could help in predicting inactivation rates of non-culturable viruses, we investigated whether sunlight inactivation of poliovirus type 3 is caused by damage to the protein capsid of the virus. More specifically, this research focused on the inhibition of viral `life' processes that depend on an intact capsid. Comparing data from assays that quantify host cell attachment and infectivity, we found that although sunlight exposure leads to an inhibition of poliovirus type 3 attachment to host cells (which is the first step in the infection process), this damage mechanism plays a minor role in total inactivation.

The second part of this dissertation focuses on a case study of wastewater irrigation practiced in Accra, Ghana, with goals to better understand the health risks associated with wastewater irrigation in Accra, and to determine whether small, farmer-dug ponds can contribute to disinfection of irrigation water. To provide data that can be used to refine quantitative microbial risk assessment models of wastewater-fed agriculture in Accra, irrigation water samples were analyzed for concentrations of fecal indicator microorganisms (human-specific Bacteroidales, E. coli, enterococci, thermotolerant coliform, and F+ and somatic coliphages) and two human viruses (adenovirus and norovirus genogroup II). E. coli concentrations in all samples exceeded recommended limits set by the World Health Organization, human viruses were detected in 75% of samples analyzed, and virus concentrations were quantified in 60% of samples. Indicator organism and virus concentrations were compared as part of an analysis of differential stability of fecal indicator organisms and pathogens in the environment, and the appropriateness of assumptions used in quantitative microbial risk assessment to relate indicator organism concentrations to those of pathogens.

After determining indicator organism and pathogen concentrations in Accra irrigation water, we investigated the ability of a farmer-developed intervention (small, on-farm ponds) to disinfect wastewater before use in vegetable irrigation. Results indicated that sunlight inactivation dominated the removal of two bacteria (E. coli and enterococci) and two bacteriophages (F+ and somatic coliphages) in these ponds, and that the ponds can contribute to the multi-barrier approach to reducing health risks related to wastewater irrigation. On-farm pond design and management recommendations, as well as challenges, are also discussed.