Investigation of faecal pollution sources and bacterial transfer hydrodynamics in rural catchments

Abstract

Faecal bacteria from point and diffuse source pollution can impact water which poses a serious public health and environmental threat. Faecal pollution contains nutrients such as phosphorus and nitrogen and when in excess these nutrients can cause eutrophication. Observing and establishing the source of faecal pollution is imperative for the protection of water quality and human health. Conventional culture methods to detect such pollution via faecal indicator bacteria have been extensively utilised but do not determine the source of pollution. To combat this, microbial source tracking (MST), an important emerging molecular tool, can be applied to detect host-specific markers in faecally contaminated waters. The main aim of this project was to establish the source of faecal contamination and the pathways by which it was transferred in two rural catchments, Arable B (river catchment) and Grassland D (karst catchment containing turloughs), in Ireland using a multi-tiered approach including elements such as MST. The MST approach was to target ruminant and human-specific faecal Bacteroidales 16S rRNA genes within the two catchments. From a risk assessment perspective, the two catchments differ greatly due to the varying hydrology, agricultural land use, topography, geology, and human density located in each. These two catchments have combinations of potentially high source and transport pressures. Hydrological connectivity drives the transfer of faecal bacteria and nutrients, of human or agricultural origin, from land to water. In Chapter II, a novel pathway separation technique, Loadograph Recession Analysis (LRA), was applied to identify transfer pathways of faecal bacteria (E. coli). LRA separates the pathways into quick flow - which represents surface overland flow, preferential flow, and tile and ditch drainage; interflow; and delayed flow - which represents baseflow. The study illustrated that high loads of phosphorus and E. coli were transferred during the faster flow pathways in both catchments. Grassland D was found to be a transport limited system and Arable B a less transport limited system. Transfers of E. coli were shown to be dependent on flow and independent of season. Arable B had higher E. coli loads, even in low flow, despite Grassland D having a potentially higher faecal indicator organism FIO source. In Grassland D, higher loads of E. coli were moved during rising phases but this was more subtle in Arable B as the loads mostly remained unchanged during rising and falling phases. Potential chronic point sources in Arable B, were indicated by the asymptotic decline of sustained E. coli loads towards low flows.

MST has been used to differentiate between anthropogenic and agricultural faecal sources in faecally contaminated water but has not been used to show how the dominance of faecal sources changes between different phases of hydrological runoff. In Chapter III, it was hypothesised that agricultural ruminant faecal waste would be elevated increased and human faecal waste, where this was present, would be diluted by the quicker flow paths. The universal Bacteroidales qPCR assay, BacUni-UCD, and the host- specific qPCR assays BacBov-UCD and BacHum-UCD were applied to water samples taken during different flow phases. The water samples were also assessed for E. coli occurrence. Statistical analysis illustrated correlations between E. coli, total phosphorus (TP), and BacUni-UCD loads indicating the presence of faecal contamination. BacBov-UCD and BacHum-UCD were detected in Arable B with the agricultural ruminant contamination mostly dominating the quick flow pathway, and human faecal contamination mostly moving through delayed flow. Grassland D also showed low levels of host-specific faecal contamination with ruminant and human sources being highest in quick flows.

There are Bacteroidales bovine-specific assays currently available to detect cow faecal pollution but some of these assays have shown cross-amplification with sheep faecal pollution. As of yet, there has been no molecular Bacteroidales assay developed to identify and differentiate sheep-specific microbial pollution. The aim of Chapter IV was to use subtractive hybridisation to identify specific DNA sequences to develop such an assay, thus allowing further differentiation of sources of pollution, aiding water quality. Target sheep-specific faecal Bacteroidales rRNA gene fragments were differentiated from diverging, though closely related subtracter cow and human faecal Bacteroidales sources. The sequences specific to sheep were used to design five PCR assays which worked successfully as demonstrated by their ability to amplify the positive control. Probes were designed to develop qPCR assays to discriminate between sheep and cow faeces. The resultant five qPCR assays were tested against various faecal samples. One qPCR assay was able to successfully differentiate between sheep, cow, human, horse, goat and pig faecal samples. Using combining DNA based analysis of faecal bacteria to discriminate faecal matter sources in conjunction with high-resolution P analysis for hydrological pathway discrimination and E. coli analysis, during different events, will add to the understanding and mitigation of FIO transfers from land to water. This will result in a more targeted approach to best management practices which could limit the deterioration of water quality in the most cost effective way. This information can be used by agricultural policy makers or local county councils to help protect water quality.