Freshwater cyanobacterial blooms are recognized as problems, particularly in lentic waterbodies. These problems can include toxin production, the occurrence of taste- and odour-causing compounds, and low dissolved oxygen events causing fish kills. These global problems have prompted research into effective in-situ methods of bloom treatment and prevention. The phytotoxicity of terrestrial dissolved organic matter (DOM) has been explored as a novel and promising form of cyanobacterial bloom control. Direct application of whole plant material to lentic waterbodies has been found to be an effective mitigation against blooms, however, these results are not consistent with the existing large body of literature classing DOM as recalcitrant and of little ecological consequence. The compounds that make up a DOM pool are chemically complex and dynamic in nature. In freshwater ecosystems, plant-derived DOM is an important source of carbon (C) and nutrients to aquatic microorganisms. A major barrier to understanding the ecological effect of DOM rests is a chemical description of the labile versus recalcitrant DOM pools, its degradation rates, and in the case of cyanobacterial control, its toxicity. Furthermore, these chemical descriptors would also allow its sources to be described. Methodological challenges have meant that little progress has been made in identifying toxic DOM characteristics or specific compounds, and their potential persistence in the environment. While some optical methods now allow the main constituent functional groups of DOM to be characterized, it is apparent these measures have limited usefulness for predicting DOM sources or toxicity. Research is needed to align results from current DOM analytical methods (thus clarifying the DOM chemical complexity) with its ecological impact (i.e. toxicity towards cyanobacteria). Fundamental information on DOM toxicity is needed to establish the suitability of vegetation derived DOM as a targeted natural cyanobacterial bloom control in inland waterways. This thesis has combined ecological and chemical approaches to the study of terrestrial DOM, by investigating how vegetation source and environmental conditions affect the impact of terrestrial DOM on cyanobacteria. Accordingly, this study had two key areas of focus:

1. Ecological effects of DOM a) Determine the potential for DOM from a range of terrestrial plants to differentially affect a cyanobacterial species compared with a eukaryotic alga. b) Investigate terrestrial sources (e.g. plant source at an order or family level) and transformation processes (i.e. as a result of exposure to bio- or photodegradation) to determine whether they affect DOM toxicity towards cyanobacterial species.
2. Chemistry of DOM a) Assess the effectiveness of spectroscopic and elemental chemical characterisation methods to determine functional indicators to predict leached DOM phytotoxicity. b) Fractionate DOM using Preparative HPLC, and identify molecular components of DOM responsible for toxic activity using nuclear magnetic resonance (NMR) 1H NMR and 2D NMR characterisation techniques. This study developed a unique approach in connecting the chemistry and ecology of DOM. Specific pools of DOM were characterized, their associated toxicity towards a cyanobacterial monoculture was investigated, and measured chemical parameters were assessed to establish if they were predictors of its source or fate (toxicity). The specific pools of DOM included; DOM leached from different plant sources, before and after environmental degradation, and isolated groups of constituent compounds (i.e. HPLC fractionated DOM). DOM was spectroscopically characterized using methods ranging from relatively inexpensive and simple ultraviolet–visible absorption spectroscopy (UV-Vis spectroscopy) to more sophisticated 1H and 2-Dimensional NMR techniques, with preparative HPLC, that provided more specific chemical identification. The first data chapter (Chapter 2) showed DOM chemical characteristics depended upon its plant source phylogeny, as measured using UV-Vis spectroscopy and elemental methods. Eight DOM absorbance and concentration parameters were used in combination to be indicative of the DOM source it was leached from viz. non-woody angiosperm, woody-angiosperm and gymnosperm trees. Although plant source affected the chemical characteristics of DOM, bioassays completed on these same DOM pools showed its toxicity towards the cyanobacterial species Cylindrospermopsis raciborskii (C. raciborskii) was less affected by its source. However, unique DOM spectral characteristics estimated toxic DOM samples to be dominated by C-rich, ligneous compounds. In separate dose-response experiments it was determined that cyanobacteria differed in sensitivity to eukaryotic algae, when exposed to angiosperm derived-DOM. Specifically, cultures of the cyanobacterium C. raciborskii had a greater reduction in photosynthetic yield in response to DOM leached from four of the five different terrestrial plant sources compared to a eukaryotic alga (Monoraphidium spp.) isolated from the same waterbody. While plant species had limited effect on sample toxicity, plant species may be important when considering biomass production in terms of the quantity of leaf material available for leaching, as determined by dose response curves. Leached DOM from Casuarina cunninghamiana and Eucalyptus tereticornis had similar chemistry and toxicity towards the cyanobacteria, and thus were selected for further exploration in Chapters 3 and 4. In Chapter 3, dose-response bioassays were used to test the relative importance of DOM source versus environmental breakdown on toxicity towards a cyanobacterial species (C. raciborskii). DOM was leached from Casuarina and Eucalyptus leaves, and exposed to photochemical and microbial degradation for periods of 24 and 120 h. Interestingly, DOM phytotoxicity was most affected by photochemical degradation processes, with the source of DOM (i.e. plant species) and microbial breakdown processes playing a lesser role. Combining these findings with spectroscopic and elemental chemical parameters, a statistical model was developed which indicated a number of parameters that were useful in estimating toxicity of a DOM sample (measured as DOC in units of carbon). Findings from this study contrast with previous paradigms that terrestrial DOM is recalcitrant and inert to bacteria and algae, or assumptions that active compounds were relatively short-lived after exposure to sunlight and bacterial degradation processes. A novel characterisation method, 1H and two-dimensional NMR (2D-NMR), was successfully applied to determine the dominant structures of leaf leached DOM. The similarities and differences of these structures were compared to elucidate the effects of plant source, or environmental breakdown. Casuarina and Eucalyptus DOM consisted mainly of α- and 𝛽-glucose as major components, and gallic acid was detected from both sources. The chemical differences in these sources were that proline was only in the Casuarina DOM and 4-hydroxybenzoic acid in the Eucalyptus DOM. The resolution of this approach allowed changes to the dominant chemical profiles of the leachates after various environmental exposures to be identified. In both samples, microorganisms rapidly assimilated glucose. Two products of the bacterial metabolism were formed as a result (viz. myo-inositol and 2,3-butanediol). However as this work was completed on bulk DOM samples (i.e. unfractionated), it was not possible to determine if these compounds were responsible for toxic action. The reduction in glucose reduced the overall DOC concentration, while toxicity remained the same irrespective of the incubation time. These findings indicate DOC concentrations in water samples may not be useful as a proxy for predicting the fate of DOM. In isolation, DOC concentrations measured by a catalytic oxidation method do not have any meaning with respect to DOM concentrations or presence of toxic compounds and may overlook any ecological impact of DOM, i.e. microbial growth. However, in this study preliminary mixed effect models indicated DOC could be useful in predicting source and toxicity when combined with spectroscopic parameters. The application of a more sophisticated coupled analytical technique, preparative HPLC coupled with 1H and 2 D NMR spectroscopy was used to examine Casuarina and Eucalyptus leached DOM (Chapter 4). This allowed a more comprehensive characterisation of DOM than previous published approaches. These techniques were applied to determine the structures of compounds responsible for DOM toxicity, and their susceptibility to degradation. Specific compounds were eluted separately using HPLC, and added to cultures of C. raciborskii to determine which of the fractions were responsible for the toxic response. This study found that the toxic compounds were the amino acid proline in the case of Casuarina, and polyphenols and phenols including gallic acid in the case of Eucalyptus. Therefore, although Casuarina and Eucalyptus leached DOM had similar broad chemistry and toxicity towards the cyanobacteria (Chapter 3), the specific compounds responsible for this toxicity differed. This study also found that these chemical and biological transformation processes investigated did not transform or break down the compounds identified as toxic viz. proline, polyphenols and gallic acid, and no other toxic products were created. Therefore, it was concluded that sunlight and microbial activity increased the toxicity of DOM in leaf leachate through quite different mechanisms i.e. likely generation of reactive oxygen species in sunlight, and degradation of non-toxic compounds by microbes. Microbial degradation processes resulted in a rapid loss in sugars (carbon), while toxic compounds did not necessarily decrease in an analogous manner with degradation time. By identifying the structures and sources of the phytotoxic compounds present in DOM, this thesis showed that phytotoxicity of DOM towards the cyanobacterium (C. raciborskii) may occur irrespective of the plant source selected, and may persist irrespective of phototransformation and/or microbial degradation. The fate of DOM is dependent upon its chemical composition, and this thesis presented novel characterisation methods, whereby DOM toxicity was assessed against the samples optical properties and elemental composition, and clarified after the identification of specific compounds. Future control of cyanobacteria is most likely to be successful if catchment vegetation composition is modified towards plant species identified as a source of phytotoxic DOM (or toxic DOM precursors).