Three novel methods were examined to assess the effects of nutrient enrichment in three sub-tropical east Australian estuaries with differing levels of catchment disturbance. The Manning and Wallamba River catchments are 8927 and 500 km2 respectively and support intensive livestock agriculture and some residential development, whereas the smaller Wallingat River catchment is 185 km2 and is mostly forested. The Wallamba and Wallingat Rivers flow into Wallis Lake, which was the site of an environmental crisis in 1997 when the consumption of sewage-contaminated locally grown oysters caused an outbreak of Hepatitis A. This study contributes valuable and extensive water quality data to assess rehabilitative works in Wallis Lake, and in other estuaries, to reduce the input of nutrients such as sewage. The importance of various sources, loads and sinks and the dispersion and assimilation of nutrients were quantified and compared for each estuary. An ecological model coupled with a one-dimensional hydrodynamic box model revealed that nearly all of the nitrogen inputs to the modified Wallamba River entered as runoff from the intensively livestock farmed upper catchment. In contrast, in the forested Wallingat River nitrogen was generated internally from anoxic sediments and was also longitudinally dispersed upstream from the junction with the modified Coolongolook River. Bloom conditions of up to 59 mg m-3 were observed in the Wallamba River between 7 and 11 days after the storms, but in spite of high concentrations of biologically available nitrogen in the forested Wallingat River, phytoplankton biomass was low due to phosphate limitation. Longitudinal dispersion could not counter phytoplankton growth in the modified Wallamba River, in spite of large increases in river flow. Estimates of grazing pressure by the pygmy mussel, Xenostrobus securis, demonstrate that it could contribute up to half of the phytoplankton loss. The inter-tidal mangrove pneumatophore habitat of X. securis allows filtering of the upper water column from the lateral boundaries in a vertically stratified water column, exerting top-down control on phytoplankton biomass. The optical plankton counter (OPC) can provide an in situ, rapid assessment of zooplankton productivity over large temporal and spatial scales from the size distribution of zooplankton. High concentrations of sub-resolved particles, including suspended detritus, have hampered the use of the in situ OPC in estuaries. Up to 58 counts L-1 due to the coincidence of sub-resolved particles passing through 100 ??m mesh were detected by the laboratory OPC in Manning, Wallamba and Wallingat River water samples. The influence of these erroneous counts on in situ OPC measurements was assessed by comparison with measurements of simultaneously collected net zooplankton measured using the laboratory OPC. In situ OPC measurements of total zooplankton abundance could be corrected for erroneous counts of sub-resolved particles using laboratory OPC measurements of 100 ??m mesh filtered water samples from the same site, but estimates had large associated error and information on the size structure of the zooplankton community is sacrificed. In contrast to expected relationships, no meaningful or significant correlations were found between the number or biomass of sub-resolved particles and in situ light attenuance. Laboratory OPC measurements of net zooplankton in each estuary revealed that the modified Manning and Wallamba Rivers supported a greater biomass of zooplankton compared to the forested Wallingat River. The normalised biomass size spectra of net zooplankton responded to both production of small particles and predation and loss of large particles. The effect of catchment disturbance on the nitrogen and carbon stable isotope ratios (??15N and ??13C) of X. securis was investigated in the three estuaries. Manning and Wallamba River mussels were enriched in ??15N by an average of 3.2??? and 1.5??? respectively compared to mussels from the forested Wallingat River. The isotope values of particulate organic matter showed a similar pattern to mussels, indicating a direct link between them within each estuary. A multiple regression model of mussel ??15N using the fractions of land used for livestock agriculture and residential development within 5 km zones from river networks to a distance equivalent to a tidal ellipse from sites explained 67% of the variation in mussel ??15N with 95% of the differences lying within 1.6??? of observed values. Increasing fractions of land used for livestock agriculture in the regression equation depleted estimated mussel ??15N indicating the use of cow manure as a nutrient source with a value of 2.0???. Increasing fractions of land used for residential development enriched estimated mussel ??15N, indicating the use of human-derived waste with a value of 20.8???.
| Identifer | oai:union.ndltd.org:ADTP/258638 |
| Date | January 2005 |
| Creators | Moore, Stephanie Kay, School of Biological, Earth & Environmental Sciences, UNSW |
| Publisher | Awarded by:University of New South Wales. School of Biological, Earth and Environmental Sciences |
| Source Sets | Australiasian Digital Theses Program |
| Language | English |
| Detected Language | English |
| Rights | Copyright Stephanie Kay Moore, http://unsworks.unsw.edu.au/copyright |
Page generated in 0.0017 seconds