I investigated the role of a flood in particulate organic matter (POM) dynamics of the Little Tennessee river/floodplain ecosystem in western North Carolina, USA. I measured litter inputs, leaf breakdown, floodplain litter, and aboveground herbaceous net primary production at 12 sites. Annual litter inputs (274-625 g m-2 y-1) were typical of a temperate deciduous forest, but lower than other floodplain forests in the eastern United States. Mean aboveground herbaceous net primary production ranged from 61-439 g m-2 y-1 and leaf breakdown rates of 4 tree species ranged from 0.001-0.010 d-1. Following a flood on 8 January 1998, sites were separated into three inundation classes: inundated, partially-inundated, and non-inundated. Sites inundated by a January flood had significantly less forest floor leaf litter and coarse woody debris after the flood. There was no significant change at the partially-inundated or non-inundated sites. In addition, there was no significant difference in herbaceous material between pre- and post-flood collections for any inundation classes. Litter input, leaf breakdown, and floodplain litter standing crop data suggest that flood entrainment of POM from the floodplain of Little Tennessee River is a source of POM to the active channel. The impact of floods on the floodplain POM dynamics, however, is highly dependent on the time of year and magnitude of the flood, and on the structure of floodplain tree assemblages, which is strongly influenced by anthropogenic land use. Compared to direct litterfall and instream primary production, the floodplain may be a small annual source of POM for the river. / Master of Science

Surface waters receive large amounts of dissolved organic matter (DOM) via runoff from land. The DOM is rich in organic carbon that serves as an energy source for the aquatic biota. During uptake of this energy, aquatic organisms mineralize organic carbon. The resulting inorganic carbon is partially released to the atmosphere as carbon dioxide and methane that are greenhouse gases, and which are of concern for the ongoing global warming. The rate at which organic carbon is mineralized depends strongly on DOM quantity and quality that vary with respect to both time and space. In this thesis, DOM quantity and quality were addressed using spectroscopic methods that build on the absorptive and fluorescent properties of chromophoric DOM (CDOM). New techniques to measure CDOM absorption and fluorescence were applied and further developed that allowed us to present novel CDOM variability patterns. Addressing the lake-rich Scandinavian landscape, strong focus was placed on water retention by lakes that tightly links to lake DOM quantity and quality. An analysis of 24,742 lakes from seven large Swedish river systems indicated that the majority of lakes in Sweden exchange their water within one year. From headwaters to the Sea, summed lake volumes in the catchments of lakes were found to increase at rates comparable to discharge, which indicated effective water renewal along flow. A strong relationship between lake water retention and CDOM was apparent and further investigated based on samples from a lake district to a regional scale. Results from in situ high-frequency monitoring of CDOM absorption in a eutrophic humic lake showed intra-annual variability patterns known from oligotrophic lake systems. The patterns for CDOM absorption contrasted results obtained for synchronously measured partial pressures of carbon dioxide that showed diurnal signals. Measurements of CDOM fluorescence and DOC concentrations indicated lake-internal DOM production. A comparison of these results with results from addressing 560 lakes distributed across Sweden, showed that a well-calibrated CDOM fluorescence measurement captures signals from lake-internal DOM production. I conclude that improved CDOM fluorescence measurements are promising to address lake-internally produced DOM.

No description available.

No description available.

No description available.

A liquid chromatographic technique using ultrastable-Y (US-Y) molecular sieve as the stationary phase and n-pentane as the mobile phase has been developed to fractionate and enrich pentacyclic triterpanes from petroleum. The sieve provides a shape-selective window which distinguishes between the various pentacyclic components, thus fractionating them on the basis of molecular shape differences. This sieving technique has been applied to isolate various pentacyclic triterpanes from sedimentary organic matter to enable better analysis of these biomarkers to be carried out.Biodegraded crude oils from three Australian basins were analysed to assess the geochemistry of their rearranged hopanes. Enhanced abundances of 25-norhopanes, 18(alpha)-30-norneohopane and diahopanes relative to the regular hopanes were observed in the most severely biodegraded samples. Geochemical interpretation of these results suggests that the enhanced abundances are due to the greater resistance of rearranged hopanes to biodegradation compared to regular hopanes. These studies also indicate that enhanced relative abundances of 25-norhopanes in these samples is most likely due to selective bacterial demethylation of (alpha beta)-hopane precursors.A branched and cyclic alkane fraction from a higher plant-derived crude oil was subjected to the US-Y chromatography procedure and the fractions eluted from the column were analysed using GC-MS. The compositions of the first two eluted fractions were markedly different from the initial branched and cyclic alkane mixture in that they were enriched in higher plant-derived triterpanes, such as bicadinanes, spirotriterpane and the oleananes and other, previously unreported, C(subscript)29 and C(subscript)30 triterpanes. A comparison of mass spectral data, GC retention and molecular sieve sorption characteristics of these compounds with those of known ++ / triterpanes of known molecular structure was used to suggest structures for the unknown compounds.Isolation of crude oil fractions enriched in pentacyclic alkanes using the sieving procedure enabled lower concentrations of bicadinanes to be detected than was previously possible by applying selective ion detection GC-MS to branched and cyclic alkane fractions. Application of this technique to a higher-plant derived Jurassic crude oil and two Jurassic sediments from the Eromanga Basin, Australia has revealed the presence of bicadinanes. The occurrence of the cis-cis-trans and trans-trans-trans bicadinane biomarkers that have previously only been reported from angiosperms may indicate an early evolution of flowering plant like species in this basin.The molecular sieving technique has also been used to isolate three pentacyclic triterpanes from low rank coals in order to obtain unambiguous structural identification and to determine their geochemical significance. A major hopanoid component isolated from a Victorian brown coal was characterised by single crystal X-ray diffraction and (subscript)13C NMR spectroscopy as 22R 17(alpha),21(beta)(H)-homohopane. This compound was shown to correspond to the later eluting 17(alpha),21(beta)(H)-homohopane and hence, for the first time, confirmed the common practice of assigning the higher retention time peak in gas chromatograms of (alpha beta) homohopanes as the 22R diastereomer. Heating of the isolated 22R (alpha beta)-homohopane on anthracite produced a mixture of the 22S and 22R diastereomers which implied a product-reactant relationship between the two epimers. Furthermore, a C(subscript)29 and a C(subscript)30 triterpane present in the hydrous pyrolysate of a Bremer Basin coal were also isolated using the molecular sieving procedure. 28 Nor-18(alpha)-oleanane was characterised by single crystal X-ray analysis while lupane was ++ / characterised by (subscript)13C NMR spectroscopy and by co-chromatography with an authentic standard on four different GC phase columns. The unusual occurrence of these triterpanes was attributed to the high sulphur content of the coal.Finally, laboratory isomerisation and reduction of an isomeric mixture of oleanenes was carried out to investigate the origin of oleanane (18(beta)-oleanane) and 18(alpha)-oleanane. Laboratory results indicated that oleanane was mainly derived from olean-18-ene, while 18(alpha)-oleanane was derived from 18(alpha)-olean-12-ene. Analysis of oleanene/oleanane abundances in a sedimentary sequence from Indonesia provided results consistent with laboratory evidence showing that 18(alpha)- olean-12-ene, rather than oleanane, is the main sedimentary precursor of 18(alpha)- oleanane.

Natural organic matter (NOM), a complex mixture of organic compounds, influences drinking water quality and water treatment processes. The presence of NOM is unaesthetic in terms of colour, taste and odour, and may lead to the production of potentially carcinogenic disinfection by-products (DBPs), as well as biofilm formation in drinking water distribution systems. Some NOM removal processes such as coagulation, magnetic ion exchange resin (MIEXTM) and membrane filtration produce sludge and residuals. These concentrated NOM-containing sludges from alum precipitation, membrane treatment plants and MIEX regeneration must therefore be treated prior to disposal. The white-rot fungi possess a non-specific extracellular oxidative enzyme system composed of lignin peroxidase (LiP), manganese-dependent peroxidase (MnP) and laccase (Lac) that allows these organisms to mineralise lignin and a broad range of intractable aromatic xenobiotics. Rojek (2003) has shown the capabi lity of Phanerochaete chrysosporium ATCC 34541 to remove 40-50% NOM from solution, however, this was found to be mainly due to adsorption and to be a partially metabolically linked activity. Consequently, the bioremediation of NOM wastes by selected white-rot fungi was further investigated in the present study. The P. chrysosporium seemed to preferentially remove the very hydrophobic acid (VHA) fraction, and so was most effective for a NOM preparation with a high proportion of hydrophobic content (and so high in colour and specific UV absorbance (SUVA)). The extent of NOM decolourisation by P. chrysosporium in three growth media with different C:N ratios followed the trends: Waksman (C:N = 6) > Fahy (C:N = 76) > Fujita medium (C:N = 114), such that the lower the C:N ratio, the greater NOM removal. This was consistent with the findings of Rojek (2003), who used a different NOM preparation and demonstrated that the removal of NOM increased with decreased C:N ratio (1.58-15.81). As removals of NOM with P. c hrysosporium ATCC 34541 were low, and little biodegradation occurred, this organism was compared with P. chrysosporium strain ATCC 24725, Trametes versicolor ATCC 7731, and three strains of yeast (Saccharomyces species arbitrarily denoted 1, 2 and 3). T. versicolor gave the greatest removal (59%) which was attributed largely to degradation, whereas the NOM removal by the two strains of P. chrysosporium (37%) and the yeast was predominantly due to adsorption as indicated by the deep brown colouration of the biomass. Saccharomyces sp. 1, 2 and 3 removed 12%, 61% and 23% of the colour, respectively. Although Saccharomyces sp. 2 had similar high colour reduction to T. versicolor, the specific removal values differed markedly: 0.055 compared to 0.089 mg NOM/mg biomass, respectively. The low level of the ligninolytic enzymes secreted by both strains of P. chrysosporium corresponded with the low degree of NOM removal by biodegradation as shown by high performance size exclusion chromatography (HPSEC). The high NOM removal attained by T. versicolor was attributed to the activities of the ligninolytic enzymes, especially laccase. The NOM removal was attributed to the breakdown of the high molecular weight compounds to form a pool of low molecular weight materials, which were then most likely utilised by the T. versicolor. Growth of T. versicolor cultures at 36oC caused inhibition or denaturation of the activity of the phenoloxidase enzymes compared to those grown at 30oC. The low activity of LiP in both cultures suggested that this enzyme may not play much of a role in NOM removal. The higher levels of MnP and Lac activities at 30oC were responsible for the greater NOM removal (73% vs. 59%) and thus the cleavage of aromatic rings, conjugated and C-Cβ αbonds in phenolic moieties, as well as catalysing alkyl-aryl cleavage in the NOM structures. T. versicolor cultured in Waksman medium with higher initial glucose (5 g/L cf. 2 g/L) led to lower ligninolytic enzyme activities and a lower degree of NOM removal (25% less colour reduction), probably due to preferential use of glucose over NOM as carbon source. NOM removal (mg removed) increased linearly with NOM concentration up to 600 mg C/L (62 mg (A446); 31 mg (A254)), above which removal decreased markedly. This trend coincided with increasing total ligninolytic enzyme activity, where the level of Lac increased up to 600 mg C/L NOM although MnP decreased gradually across the range while LiP was only detected for 100 and 300 mg C/L NOM. Hence, the removal of NOM from solution by T. versicolor was associated with high oxidative enzyme activity, particularly of laccase. Laccase was the major extracellular enzyme secreted by T. versicolor and by deduction, played a major role in NOM removal. The optimum temperature for Lac activity secreted by T. versicolor cultured in Waksman medium supplemented with 4.5 g/L wheat bran plus 0.5% Tween 80 was determined to be 50oC. The optimum pH for the Lac activity for guaiacol and NOM was identified as pH 4.0-4.5. Although the optimum enzyme activity occurred at 50oC, 30oC was recommended for enzymatic removal of NOM as the phenoloxidase enzyme activity may be denatured if the NOM removal process were considered to run for long period at high temperature. Although agitation led to apparent enzyme denaturation, fermentations with continuous agitation promoted enzyme activity faster than those with occasional agitation (agitated every 6 hours for 30 minutes at 130 rpm and 30oC) as it provides better mass transfer. However, it seemed that continuous agitation had an adverse effect on the fungal growth and enzyme production over extended fermentation periods. Addition of 4.5 g/L wheat bran to modified Waksman medium in the absence of NOM led to high production of Lac activity compared with LiP and MnP activities, showing its great potential as a laccase inducer. Addition of Tween 80 alone to the cultures led to a small improvement in Lac activity; however, with the presence of wheat bran it caused marked increases in LiP, MnP and Lac activit ies. When NOM was added to cultures of T. versicolor with the two supplements, it led to markedly reduced Lac activity, but increased LiP and MnP activities, and no improvement in NOM removal compared with the cultures in the absence of supplements (12 mg (or 61%) cf. 15 mg (or 73%) for 100 mg C/L after corrected for colour from and adsorption by wheat bran).

The timing and composition of organic matter (OM) inputs to rivers are important as carbon plays a major role in river functioning. Management of Australian rivers since European settlement has altered inputs of organic matter to these systems. Heterotrophic microbes play a critical role in the transformation of OM in rivers, allowing transfer of carbon to other biota. Alteration to the proportions of OM from different sources affects microbial functioning due to differences in OM composition. Macrophytes can represent important sources of carbon to rivers, however their inputs and in-stream processing are poorly understood. The aim of my study was to examine inputs and microbial processing of macrophyte OM in Australian lowland rivers under different flows. Distributions of dominant macrophytes (Typha orientalis, Phragmites australis, Vallisneria gigantea and Persicaria prostrata) were mapped in three lowland river reaches in south eastern Australia. Integration with flow data in a GIS allowed the determination of macrophyte inundation patterns under different flows. Resource allocation (biomass and nutrients), live and dead shoot densities and litter production were monitored in the field over 18 months. DOM release from different macrophyte tissues was examined in the laboratory and leachate composition was assessed using nutrient and spectral analyses. Responses of riverine microbial communities to different OM sources were assessed from substrate-induced respiration and enzyme activity experiments and field measurements of respiration and enzymatic responses to varied OM inputs. Finally, all data were integrated into a model of microbial responses to macrophyte OM inputs induced by different flows. Large populations of macrophytes occurred at all three sites, at bed level, on in-channel benches and on banks. Bank slope, channel heterogeneity and the vertical distribution of macrophyte beds all affected macrophyte inundation patterns. Substantial differences in biomass allocation, nutrient dynamics and litter composition were observed among different plant growth forms and over time. While leaves represented the major shoot component in litter for all species, stems and reproductive structures were also important in some species. Aside from the litter pool, translocation to rhizomes represented a major sink for annual production in emergent plants. Patterns of shoot density and litter production over time varied among species, providing a source of variation for particulate, and hence dissolved OM inputs upon inundation. The majority of DOM release from POM occurred within 24 hours of inundation. Growth form, tissue type (blade, stem, etc.) and status (live or dead) affected rates, quantities and composition of DOM release, with implications for microbial utilisation. Both overall activity and patterns of carbon utilisation in riverine microbes changed in response to altered OM inputs. Patterns of microbial carbon use were shown to be specific to the carbon source which induced them. Modelling showed that flow regulation had a major impact on OM inputs and microbial metabolism, through the effects of flow variability on macrophyte vertical distributions, macrophyte bed inundation and dilution. Positive relationships between discharge, DOM inputs and microbial metabolism were observed at the most highly regulated site (drought < current < historic < flood). While a similar pattern occurred at the less regulated site in terms of total loading, dilution effects resulted in a reversal of this trend on a reach volume basis. Microbial metabolism and DOM inputs were restricted to summer/autumn under regulated flows compared to a greater emphasis on winter/spring inputs and microbial activity under unregulated flows. Continual OM inputs during winter with pulsed inputs in spring under natural flows probably benefit larger, slow-growing macro-invertebrates. River regulation promotes pulsed macrophyte OM inputs during spring/summer, potentially favouring riverine microbial and zooplankton production, although at lower levels due to the overall reduction in OM inputs. The predictive model of macrophyte OM inputs and microbial responses developed throughout this thesis represents a major step forward in our understanding of macrophytemicrobe interactions and our ability to manage our river systems. This work has shown that flow manipulation can be used to influence macrophyte organic matter inputs to rivers and microbial responses, affecting whole stream metabolism and food web interactions.

Multiple metal-gill modelling based on the toxic unit concept suggests that metals with the same toxic actions (e.g. Pb and Cd interacting at Ca-gill channels) will exhibit strictly additive binding at the gills of fish if the concentrations of the two metals sum to one toxic unit. Due to the non-linear nature in which metals bind to fish gills, the metal mixture will be more than additive below one toxic unit and less than additive above one toxic unit. This research tested the models by exposing rainbow trout to mixtures of Cd and Pb to investigate the metal-gill binding of these two metals when present in mixtures with fish. The relationship between Cd and Pb in mixtures was also investigated when natural organic matter (NOM) was added to the treatment solutions. Juvenile rainbow trout (5 g) were exposed to equal mixtures of Cd and Pb at 0.75, 1.5, 2.25 and 3.0 µM each. The high gill-Pb accumulation in relation to gill-Cd was unexpected and suggested an additional gill-binding site for Pb aside from the apical Ca channels on the gill. 96 h toxicity experiments revealed that Cd and Pb in mixtures were more toxic than either Cd or Pb alone. The bioaccumulation and partitioning of Cd and Pb within the body of 50 g rainbow trout was investigated when these fish were exposed to mixtures of Cd and Pb. The accumulation of metals within the trout body did not change between the single metal and mixture exposures. To investigate how Pb affects Cd binding to fish gills, and how these metals exert toxicity when present in mixtures, trout (2 g) were exposed to a range of Cd concentrations (0.75, 1.5, 2.25 and 3.0 µM) while maintaining constant Pb concentrations over the same concentration range. Gill binding results revealed that Pb reduced or inhibited gill-Cd binding. With the addition of NOM, this inhibition was lost. LT50 results for fish exposed to metal mixtures for 96 h revealed that without NOM, the mixture toxicity was dependent on Pb concentration. With the addition of NOM, mixture toxicity decreased with increasing Pb.

Multiple metal-gill modelling based on the toxic unit concept suggests that metals with the same toxic actions (e.g. Pb and Cd interacting at Ca-gill channels) will exhibit strictly additive binding at the gills of fish if the concentrations of the two metals sum to one toxic unit. Due to the non-linear nature in which metals bind to fish gills, the metal mixture will be more than additive below one toxic unit and less than additive above one toxic unit. This research tested the models by exposing rainbow trout to mixtures of Cd and Pb to investigate the metal-gill binding of these two metals when present in mixtures with fish. The relationship between Cd and Pb in mixtures was also investigated when natural organic matter (NOM) was added to the treatment solutions. Juvenile rainbow trout (5 g) were exposed to equal mixtures of Cd and Pb at 0.75, 1.5, 2.25 and 3.0 µM each. The high gill-Pb accumulation in relation to gill-Cd was unexpected and suggested an additional gill-binding site for Pb aside from the apical Ca channels on the gill. 96 h toxicity experiments revealed that Cd and Pb in mixtures were more toxic than either Cd or Pb alone. The bioaccumulation and partitioning of Cd and Pb within the body of 50 g rainbow trout was investigated when these fish were exposed to mixtures of Cd and Pb. The accumulation of metals within the trout body did not change between the single metal and mixture exposures. To investigate how Pb affects Cd binding to fish gills, and how these metals exert toxicity when present in mixtures, trout (2 g) were exposed to a range of Cd concentrations (0.75, 1.5, 2.25 and 3.0 µM) while maintaining constant Pb concentrations over the same concentration range. Gill binding results revealed that Pb reduced or inhibited gill-Cd binding. With the addition of NOM, this inhibition was lost. LT50 results for fish exposed to metal mixtures for 96 h revealed that without NOM, the mixture toxicity was dependent on Pb concentration. With the addition of NOM, mixture toxicity decreased with increasing Pb.