Limited supply of clean water in urbanizing watersheds creates challenges for safely sustaining irrigated agriculture and global food security. On-farm interventions, such as riverbank filtration (RBF), are used in developing countries to treat irrigation water from rivers with extensive fecal contamination. Using a Bayesian approach incorporating ethnographic data and pathogen measurements, quantitative microbial risk assessment (QMRA) methods were employed to assess the impact of RBF on consumer health burdens for Giardia, Cryptosporidium, rotavirus, norovirus, and adenovirus infections resulting from indirect wastewater reuse, with lettuce irrigation in Bolivia as a model system. Concentrations of the microbial source tracking markers pepper mild mottle virus and HF183 Bacteroides were respectively 2.9 and 5.5 log10 units lower in RBF-treated water than in the river water. Consumption of lettuce irrigated with river water caused an estimated median health burden that represents 37% of Bolivia's overall diarrheal disease burden, but RBF resulted in an estimated health burden that is only 1.1% of this overall diarrheal disease burden. Variability and uncertainty associated with environmental and cultural factors affecting exposure correlated more with QMRA-predicted health outcomes than factors related to disease vulnerability. Policies governing simple on-farm interventions like RBF can be intermediary solutions for communities in urbanizing watersheds that currently lack wastewater treatment.

Wastewater use for irrigation is expanding globally, and information about the fate and transport of pathogens in wastewater systems is needed to complete microbial risk assessments and develop policies to protect public health. The lack of maintenance for wastewater treatment facilities in low-income areas and developing countries results in sludge accumulation and compromised performance over time, creating uncertainty about the contamination of soil and crops. The fate and transport of pathogens and fecal indicators was evaluated in waste stabilization ponds with direct reuse for irrigation, using two systems in Bolivia as case studies. Results were compared with models from the literature that have been recommended for design. The removal of Escherichia coli in both systems was adequately predicted by a previously-published dispersed flow model, despite more than 10years of sludge accumulation. However, a design equation for helminth egg removal overestimated the observed removal, suggesting that this equation may not be appropriate for systems with accumulated sludge. To assess the contamination of soil and crops, ratios were calculated of the pathogen and fecal indicator concentrations in soil or on crops to their respective concentrations in irrigation water (termed soil-water and crop-water ratios). Ratios were similar within each group of microorganisms but differed between microorganism groups, and were generally below 0.1mLg(-1) for coliphage, between 1 and 100mLg(-1) for Giardia and Cryptosporidium, and between 100 and 1000mLg(-1) for helminth eggs. This information can be used for microbial risk assessments to develop safe water reuse policies in support of the United Nations' 2030 Sustainable Development Agenda.

ABSTRACTViable ova ofAscaris lumbricoides, an indicator organism for pathogens, are frequently found in feces-derived compost produced from ecological toilets, demonstrating that threshold levels of time, temperature, pH, and moisture content for pathogen inactivation are not routinely met. Previous studies have determined that NH3has ovicidal properties for pathogens, includingAscarisova. This research attempted to achieveAscarisinactivation via NH3under environmental conditions commonly found in ecological toilets and using materials universally available in an ecological sanitation setting, including compost (feces and sawdust), urine, and ash. Compost mixed with stored urine and ash produced the most rapid inactivation, with significant inactivation observed after 2 weeks and with a time to 99% ovum inactivation (T99) of 8 weeks. Compost mixed with fresh urine and ash achieved aT99of 15 weeks, after a 4-week lag phase. Both matrices had relatively high total-ammonia concentrations and pH values of >9.24 (pKaof ammonia). In compost mixed with ash only, and in compost mixed with fresh urine only, inactivation was observed after an 11-week lag phase. These matrices contained NH3concentrations of 164 to 173 and 102 to 277 mg/liter, respectively, when inactivation occurred, which was below the previously hypothesized threshold for inactivation (280 mg/liter), suggesting that a lower threshold NH3concentration may be possible with a longer contact time. Other significant results include the hydrolysis of urea to ammonia between pH values of 10.4 and 11.6, above the literature threshold pH of 10.

Urine-diversion dehydration toilets (UDDT) are common throughout the developing world, and the toilet product is widely used as compost. There is no comprehensive research to date that characterizes the compost to determine its quality, extent of pathogen inactivation, and the effects of climate and bulking materials on the compost. Compost was collected from 45 UDDT in Bolivia and analyzed for physical, chemical, and biological parameters. Eighty percent and 56% of samples did not meet acceptable compost guidelines for moisture content and pH, respectively, indicating desiccation was the dominant process in UDDT. Bulking materials significantly impacted compost characteristics in terms of pH, carbon, carbon-to-nitrogen ratio, and carbon stability (P < 0.05). Composts with ash exhibited, on average, low carbon concentrations (4.9%) and high pH values (9.7), which can be harmful to plants and composting microorganisms. Composts with sawdust exhibited, on average, high carbon concentrations (40.0%) and carbon-to-nitrogen ratios (31.0). Climate had no significant impact on chemical characteristics, however composts from humid regions had significantly higher moisture contents (34.4%) than those from arid climates (24.8%) (P < 0.05). Viable Ascaris lumbricoides ova were identified in 31% of samples, including samples with high pH, low moisture contents, and long storage times.