NITROGEN-REMOVING WASTEWATER TREATMENT SYSTEMS: MICROBIAL COMMUNITIES AND GREENHOUSE GAS EMISSIONS

Wastewater is a major source of anthropogenic nitrogen (N) pollution that causes groundwater contamination and eutrophication in coastal ecosystems. The negative effects of excess N from wastewater on human and environmental health have led the United States Environmental Protection Agency (USEPA) and many state and local agencies to set maximum N concentrations for treated wastewater before it can be discharged to ground and surface water bodies. Wastewater treatment systems that include biological nitrogen removal (BNR) can help meet these standards by promoting microbial N removal in centralized wastewater treatment plants (WTP) as well as decentralized, onsite wastewater treatment systems (OWTS; i.e., septic systems). Nitrogen removal in BNR wastewater treatment is accomplished by sequential nitrification in oxic zones and denitrification in hypoxic/anoxic zones. Wastewater treatment, including BNR, can produce the greenhouse gases (GHGs) CO2, N2O, and CH4 as by-products, potentially threatening air quality. In Manuscripts 1 and 2 of this dissertation, I investigated the dynamics of GHGs and microbial communities in OWTS that have a lignocellulose-amended, N-removing layered soil treatment area (STA). These systems are designed to promote sequential nitrification and denitrification by stratifying the two processes into layers that promote microbial N removal. Layered, nonproprietary STAs are a cost-effective alternative to proprietary, advanced N-removal OWTS. In Manuscript 3, I compared the microbial community structure and composition of nine proprietary, advanced N-removal OWTS and a WTP with BNR. Although BNR in OWTS and WTPs is designed to promote the same microbial processes – and their microbial communities assumed to be the same for OWTS management purposes – their nitrifying and denitrifying microbial communities have not been compared. In Manuscript 1, I describe CO2, N2O, and CH4 emissions from the septic tanks, Layered STAs, and Control STAs from three OWTS serving homes in southeastern MA, USA. Emissions did not differ significantly between Layered and Control STAs at any of the sites, and were controlled chiefly by temperature, soil moisture, and subsurface GHG concentrations. Per capita average emissions for these systems were higher than those reported by others, with mean values ranging from 0 to 1835 gCO2e capita-1 day-1 and from 48 to 1400 gCOe2 capita-1 day-1 in septic tanks and STAs, respectively. These results suggest that Layered STAs are unlikely to have a negative impact on air quality compared to conventional STAs. In Manuscript 2, I investigated the diversity, structure, and composition of ammonium oxidizing and nitrous oxide reducing bacterial communities in three Layered and Control STAs by targeting the functional genes amoA (ammonium monooxygenase) and nosZ (nitrous oxide reductase). The amoA community composition was similar throughout the profile of both STAs, but there was a somewhat different denitrifying communities between the sand and lignocllulose-amended layers within the Layered STA. The most common classified amoA taxon was Nitrosospira and the most common nosZ taxon was Afipia, with the most common taxa made up of unclassified bacterial strains for both amoA and nosZ. The septic tank pH and concentration of TN were correlated with differences in composition for both amoA and nosZ communities. A number of physical and chemical properties of the STA fill material were also correlated with community differences, some of which may be adjusted to support and promote specific amoA and nosZ populations and increase N-removal efficiency in Layered STAs. In Manuscript 3, I compared the bacterial communities responsible for N removal (ammonia oxidizers and denitrifiers) in a BNR WTP and nine advanced N-removal OWTS, all within the Greater Narragansett Bay watershed in RI, USA, targeting the bacterial genes amoA and nosZ. Diversity metrics were similar between oxic and anoxic zones within a type of treatment, suggesting the capacity for nitrification is present in hypoxic/anoxic zones, and the capacity for denitrification is present in oxic areas in both centralized and decentralized BNR wastewater treatment systems. The most widespread ammonia oxidizing genera at both the WTP and OWTS were Nitrosomonas and Nitrosospira, and the latter had higher relative abundance in OWTS than the WTP. The most widespread nitrous oxide reducers were in the genus Pseudomonas and Aeromonas. Thauera, Alicycliphilus, Oligotropha, Sinorhizobium, and Rhodopsueudomonas were also common nosZ genera in both types of treatment. Bradyrhizobium, Burkholderia, Massilia, and Paracoccus were more frequently associated with OWTS, whereas Azospirillum, Pseudogulbenkiania, Rhodoferax, Shinella, and Thiobacillus were more frequently associated with the WTP. These results point to major differences in the amoA and nosZ microbial communities between BNR treatment as a function of scale, even though they are designed to promote the same microbial processes. These differences need to be considered in the design and management of advanced N-removal OWTS to maximize N removal