Using sediment cores to reconstruct historical pollution records: digging up the Yarra's dirty past

Despite the increasing threat of flooding due to climate change, there is limited understanding of the level of contamination of flood deposits, and their risk to humans and the environment. It is expected that sediment cores can provide a better understanding of the contaminants deposited by floods, because sediment cores can preserve both the pollution and flood histories of aquatic systems. This PhD thesis aims to use sediment cores to identify pollutant levels (heavy metal concentrations) in sediments deposited by past fluvial floods. The Yarra River, which flows through a metropolitan area (Melbourne) in South-East Australia is used as a case study, and contaminant levels within fluvial flood deposits during the 20th century are identified in sediment cores from two floodplain lakes (billabongs). First, overall pollution trends within the Yarra River billabongs are explored. It is found that current sediment quality trigger values used in Australia do not reflect background heavy metal concentrations in the Yarra River billabongs. This highlights the need to use sediment cores to identify the background conditions of aquatic systems, when developing environmental management targets, instead of relying on generic trigger values. Also, urban stormwater from purely residential catchments appears to result in heavy metal pollution of aquatic systems. Although the installation of a stormwater treatment wetland has coincided with a slight decrease in heavy metal levels within one of the billabongs, background levels have not been restored. Second, two methods for identifying discrete flood deposits within sediment cores are presented. One method utilises the elemental composition of sediments, a flood proxy not previously sufficiently explored, to identify flood layers. The historical flood records reconstructed using billabong sediment cores can be used to infer that flooding frequency of the Yarra River has decreased through the 20th century. These reconstructed records were checked using measured flow data An uncertainty framework for using sediment cores to obtain historical pollution records is also presented. The greatest source of uncertainty is the assumption that all metals entering the aquatic system are deposited on the sediment bed. Observations during a 12-month field monitoring period indicate that whilst there may be a discrepancy between the total mass of metal inputs and total mass of metals deposited on the bed sediment, they vary throughout the year in a similar manner. These results suggest that sediment core heavy metal profiles are indicative of historical pollution trends in the aquatic system. The thesis also demonstrates how high resolution historical pollution and hydrologic trends, both reconstructed using sediment cores, can be used together to determine the pollution sources of aquatic systems. This shows that the main source of pollution in billabongs can vary greatly even if these billabongs are close to each other spatially. Whilst one billabong had higher heavy metal concentrations in flood-deposited sediments, the other had higher concentrations in sediments that were not deposited by floods. Cumulative distribution functions of heavy metal concentrations (lead and zinc) in sediments deposited by Yarra River floods at the two billabongs over the 20th century are also presented. These functions could be used to help predict the contaminant deposition by future floods of the Yarra River at the two billabongs. The methods presented in this thesis can be applied to other river catchments to better understand their sources of contamination; in particular, the importance of flooding in the deposition of contaminants in floodplains and billabongs. Furthermore, having these data of heavy metal concentrations in flood deposits will better equip us for the future, enabling us to better understand the risks associated with fluvial floods, and the management strategies that are required.