| Summary: | 博士 === 國立成功大學 === 環境工程學系碩博士班 === 97 === Cyanobacteria are present in many drinking water reservoirs in Taiwan and the world, and some of them may produce cyanotoxins and release them to natural water bodies. However, the information relevant to the presence of toxic cyanobacteria and cyanotoxins in Taiwan’s drinking water reservoirs are very limited. Therefore, a systematic investigation of their occurrence is urgently needed. The objectives of this dissertation is to develop and apply different analytical approaches, including chemical and bio-molecular methods, for the determination of toxigenic cyanobacteria and cyanotoxins in Taiwan’s drinking water reservoirs and waterworks. In this dissertation, a solid phase extraction (SPE) coupled with liquid chromatography (LC)-mass spectrometry (MS) method was first developed to concentrate and detect nine commonly observed cyanobacterial toxins simultaneously, including six microcystins (MCs) congeners, nodularin (NOD), anatoxin-a (ATX) and cylindrospermopsin (CYN), in water samples. A surrogate standard (SS) and internal standard (IS) were applied in the analytical method for better quality control. The method detection limit (MDL) was 2-10 ng/L for MCs and NOD in pure water, and was 46 ng/L for ATX and 100 ng/L for CYN, respectively. In more complicated water matrix, reservoir water with high concentration of Microcystis spp., the MDL for the cyanotoxins increased by a factor of 3 to 10, with CYN = 500 ng/L as the highest.
The analytical method developed was then applied to monitor two groups of cyanotoxins (MCs and ATX) in nine major drinking water reservoirs and seven associated waterworks. Monitoring results suggested that microcystins were present in all the drinking water reservoirs studied, and some of them had concentration higher than the WHO guideline of MC-LR (1 μg/L). In addition, ATX was also found in four reservoirs, in Kinmen Island. In order to correlate the two groups of cyanobacterial metabolites (cyanotoxins and off-flavour compounds) and other environmental parameters, 22 water quality and meteorological parameters were monitored for two source waters (Moo Tan Reservoir, MTR, and Tseng Wen Reservoir, TWR) in south Taiwan from August 2003 to April 2005. Monitoring results showed that the cyanotoxins and off-flavour compounds (2-MIB and Geosmin) were present in the source waters. Concentrations of 2–30 ng/L of 2-MIB was observed for the two reservoirs, while that of the summation of five microcystin congeners measured were between 30 and 340 ng/L. The concentration of both 2-MIB and microcystins showed higher concentrations in warmer seasons. A stepwise regression technique was employed to correlate 2-MIB and MCs concentrations with all the corresponding water quality and meteorological parameters. Good correlations among 2-MIB concentration, MC concentration, water temperature and air temperature were found in the water samples collected from both reservoirs. The correlations may provide a simple means for the water utility to anticipate the two groups of cyanobacterial metabolites in the two source waters.
In addition to the two reservoirs monitored, the cyanobacterial metabolites were also commonly observed in reservoirs and their associated waterworks in Kinmen Island. To have a better water treatment efficiency in Kinmen’s waterworks, a more precise understanding of the algal metabolites at different time and depths in the water sources as well as the change of metabolites in the treatment processes are needed. Therefore, the diurnal concentration change of the two major cyanobacterial metabolites were monitored in a major source water (Tai Lake Reservoir, TLR and major waterworks (Tai Lake Waterworks, TLW) of the island. The samples for the reservoir water were collected at/near the water intake, and one of them was sampled at 4 different depths. Most of the parameters measured varied significantly at different depths and different time, and only 2-MIB concentration remained almost constant through out the 24 hour period and at different depths. This may imply that 2-MIB was likely to uniformly distribute in the reservoir water. For most of the cyanobacteria and cyanobacterial metabolites measured, no strong correlations were observed. However, a good correlation between Microcystis spp. and MCs concentrations was found, indicating that the probable relationship between the toxins and their producers. This simple correlation may also be used in the estimation of the cell-bound and dissolved concentration of MCs in the reservoir water. For the samples collected for the waterworks, more than 98% of cyanobacteria were removed in the treatment processes, and most were removed at the dissolved air floatation (DAF) unit. Although the overall removal efficiency of microcystins and 2-MIB in TLW is >75%, unlike that for the cyanobacertia cells, only 20-30% were removed before DAF. This may be attributed to that DAF cannot effectively remove dissolved microcystins that was already present in the raw water or was released into water from the breakage of mcirocystis cells by pre-chlorination. Compared with other conventional waterworks, the slow sand filters may provide an extra 20-30% of 2-MIB removal for TLW.
Finally, in order to identify the potential MC producers in MTR and its associated waterworks, two molecular methods were developed and employed to determine the DNA sequences and characteristics of cyanobacteria community, and to quantify the functional gene concentrations in water samples. Four toxigenic Microcystis spp. strains (TWNCKU01 - TWNCKU04) were first isolated from different locations in MTR. After laboratory cultivation, two of the strains, TWNCKU01 and TWNCKU02, were found to mainly produce MC-RR, and another two may produce MC-LR, -RR and -YR at different ratios. The bio-molecular results based on mcyA and mcyB sequencing showed that all the strains are toxic Microcystis spp. and may produce MCs. The two higher diversified regions, PC-IGS (cpcB) and 16S-23S rDNA (ITS), are used to further identify the four strains. In addition, the ITS region was also used in DGGE for the construction of a clone library and bio-makers for 11 strains observed in MTR. These ITS-DGGE biomarkers were successfully applied in monitoring the community changes of potential microcystin producers over a period of 5 years. To develop a rapid method for quantifying microcystin-producing genes, two highly specific primers were designed based on UPL probes to measure mcyB and cpcB concentrations in water samples, where the former one represents gene concentration for MC producers and the latter one is for gene concentrations of cyanobacteria. In the long term monitoring results of MTR, 39 of the 41 DGGE samples contained Microcystis spp., with 36 samples being TWNCKU01 or -02, the MC-RR producers. In addition, 3 of the samples contain Planktothrix spp. After analyzing the data from UPL-based real time PCR and other reservoir water quality parameters, including Microcystis cell counts, MC concentrations, and others, the gene concentrations based on UPL-mcyB correlates well with MC-RR concentrations, the major toxin type in the reservoir, and water temperature. In addition, the gene concentrations based on UPL-cpcB correlate with cyanobacteria as well as Microcystis cell concentrations in the water samples. Both DGGE and UPL-probe methods were further successfully applied in the water samples from MTW. Although toxin concentrations were very low, the DGGE bands clearly demonstrated the presence of MC-RR producers in both process water and finished water samples. The results of UPL-real time PCR also showed that mcyB concentrations were detected to be around 200 copies/mL in the finished samples, proving that Microcystis cells may penetrate through the treatment processes and pose a potential risk in drinking water systems.
|
|---|
