The extremely high TBT content (as high as 2500 ng/g, dry weight) found in sediments from the dumpsites of dredged materials from Keelung Harbor and the second highest value (840 ng/g) from the Keelung Harbor near the China Ship Building Company clearly indicate the sources of TBT pollution. Among the Machu coastal sediments, a high TBT value (624 ng/g) was obtained at Fu-Au Harbor. For the sediments collected from the areas of copper recycling operations (Erhjin estuary) and the outlet from nuclear power plants (Chinshan), the TBT content is below the detection limit of the method. This suggests that the existence of green oysters (Crassostrea gigas) along the Erhjin Chi estuary and skeletal deformities in fish (Terapon jarbua and Liza macrolepis) along the Chinshan coastal areas are not caused by TBT pollution.
 

In recent decades, many investigators have analyzed TBT in water, sediments and organisms using various methods, such as graphite furnace-atomic absorption spectrophotometry (McKie, 1987), liquid chromatography (Yu and Arakawa, 1983), supercritical fluid extraction (Cai et al., 1994), and gas chromatography (Siu et al., 1989). However, there have been few studies analyzing the Machu and Taiwan coastal sediments. The purposes of this paper are to describe the QA/QC of a simple, sensitive determination of TBT in sediments and to evaluate the TBT in sediments collected from the Machu and Taiwan areas, particularly some suspected hot spots. For instance, the Erhjin Chi estuary has had cases of green oysters that were caused by the recycling of copper from waste cables (Hung & Han, 1992), and the Chinshan coast has been found to have skeletally-deformed fish (Terapon jarbua and Liza macrolepis) in the outlet area of nuclear power plants (Hung et al., 1998).
 

All organic solvents were HPLC grade; two kinds of tributyltin chloride (TBT-Cl) with purities of 96% (from E. Merck, Germany) and 95% (from Aldrich-Chemie/Steinhein, Germany) were used without further purification for calibration curves and confirmation checks, and analysis did not reveal any detectable impurities. Trimethyltin chloride (TMT-Cl, purity of 95%), used as an internal standard, was obtained from Tokyo Chemical Industry Company, Japan. The other chemicals were analytical reagent grade. Water used throughout all experiments was made extra-pure (18.3 MW ) by passing it through a Millipore Water Purification system (Milli-Ro-60).

Sediment Samples

Surface sediments were collected by the dredging technique aboard fishing boats along the coasts of Machu (Fig. 1A) and Taiwan (Keelung near Keelung Harbor and Fishing Harbor;
 

Fig. 1A Sampling stations along the Machu coastal area.

Chinshan near nuclear power plants; and Erhjin Chi estuary near copper recycling operations, Fig. 1B). The sediments were stored in plastic bags and kept in an icebox (about 4 ^oC) until analysis. The Machu, Keelung, Chinshan and Erhjin Chi estuary samples were collected during the periods of January 17-19, April 27-28, December 11-12, and on January 15, 1995, respectively.
 

Fig. 1B Sampling stations along the coast of Taiwan: China Ship Building Company of Keelung Harbor [1], Keelung Harbor [2], Fishing Harbor [3], near Power Plant [4] and dumpsites of harbor dredged materials [5 and 9]; Chinshan near nuclear power plants [K1 and K2]; and Erhjin Chi estuary near copper recycling operations [R1 and R2].
 

Extraction and determination of tributyltin in sediments were modified slightly from the method of Siu et al. (1989). Tributyltin in sediments (0.250 g) was released in a mixture of 2mL hydrochloric acid and 1mL methanol using an ultrasonics bath (AC, 110 volts; output, 400 watts for 14 min). After adding 400 m g/L trimethyltin (TMT, as an internal standard), the solution was extracted with 1.5% tropolone in toluene - isobutyl acetate solution (v/v, 80:20) using a shaker (Kasin, Model SB-301) for 2 hrs, and then water was removed by centrifuge at 3000 rpm. The contents of TBT and TMT in the final extract were determined by gas chromatography.

A Hewlett-Packard 5890 series II gas chromatograph (GC) equipped with a split/splitless injector, a capillary column (HP-1, 100% dimethylpolysiloxane, 30 m in length, 0.25 mm i.d., 1 m m film thickness) and a flame photometric detector (FPD) was used. The detector was operated with a 610 nm cut-off interference filter at a temperature of 210^oC. For analysis of TBT and TMT, the column oven temperature was programmed to run from an initial temperature of 80^oC to 250^oC at a rate of 14^oC/min. Nitrogen was used as a carrier gas at a flow rate of 3 mL/min. Hydrogen and air were introduced to the FPD detector at flow rates of 75 mL/min and 100 mL/min, respectively. The injection port of the gas chromatograph and the FPD detector were kept at 200^oC and 275^oC, respectively. Data on the chromatogram from the GC/FPD were processed with a Hewlett-Packard 3390A integrator and calculated on a dry basis.
 

Table 2 shows that the contents of TBT (< 50.5 ng/g to 624 ng/g, dry weight) in the Machu sediments varied with location. The highest value was observed at station M7, in the harbor, and the second highest value (104 ng/g) was at station M4. Fortunately, the content of TBT in sediments from station M6, the oyster mariculture site, was under the method detection limit (< 50.5 ng/g). In general, the pollution of TBT in the Machu coastal area is not serious. Higher values of TBT (140 - 1400 ng/g, dry weight) were found in sediments from Chesapeake Bay in the U. S. A. (Mathias et al., 1989), and also in the Masnou marine sediments (994 ng/g) of Spain (Cai et al., 1994) and the Main river sediments (953 ng/g) off the coast of Germany (Waldock and Thain, 1983).

However, extremely high TBT values of 2500 ng/g and 840 ng/g, as well as 496 ng/g (Table 2) were observed in the Keelung coastal sediments respectively at stations 5, 1 and 9. Stations 5 and 9 are the dumpsites for materials dredged from Keelung Harbor, and station 1 is inside Keelung Harbor where the China Ship Building Company is located. Thus the reason for those high TBT contents and the sources of the TBT can be easily explained.

Table 2 also shows the TBT contents in sediments collected from the Erhjin Chi estuary (R1 and R2) and the Chinshan coastal area (K1 and K2). The Erhjin Chi estuary is the place where, since January 1986, copper pollution from copper recycling operations gave rise to green oysters. The copper content in oysters (Crassostrea gigas) was 2200 m g/g (dry weight) in January 1986 and 4400 m g/g in February 1989 (Hung and Han, 1992; Han and Hung, 1990). Stations K1 and K2 were located along the Chinshan coastal area near the outlet of nuclear power plants where the skeletal deformities of fish (Terapon jarbua and Liza macrolepis) were observed in the summer of 1993 (Hung et al., 1998). Many environmentalists suspected that the deformities were caused by TBT pollution. As illustrated in Table 2, the sediments collected from the areas of copper recycling operations (Erhjin Chi estuary) and outflow from nuclear power plants (Chinshan) have TBT contents that are less than the method detection limit (< 50.5 ng/g), thus providing evidence that the occurrence of green oysters and fish-skeleton deformations were not caused by TBT pollution.
 

Beaumont, A.R. and P. B. Newman (1986) Lower levels of tributyltin reduce growth of marine micro-algae. Mar. Pollut. Bull., 17, 457-161.

Bryan, G. W., P.E. Gibbs, L.G. Hummerstone and G.R. Burt (1986) The decline of the gastropod Nucella lapillus around southwest England: Evidence for the effect of tributyltin from antifouling paints. J. Mar. Biol. Ass. U.K., 66, 611- 640.

Cai, Y., R. Alzaga and J.M. Bayoma (1994) In situ derivatization and supercritical liquid extraction for the simultaneous determination of butyltin and phenyltin compounds in sediments. Anal. Chem., 66, 116-167.

Chagot, D., C. Alzieu, J. Sanjuan and H. Grizen (1990) Sublethal and histopathological effect of trace level of tributyltin fluoride on adult oyster Crassostrea gigas. Aquac. Liv. Resour. 3, 121-130.

Han, B.C. and T.C. Hung (1990) Green oysters caused by copper pollution on the Taiwan coast. Environ. Pollut., 65, 347-362.

Hung, T.C., C.C. Huang and K.T. Shao (1998) Some events of ecological survey of coastal water adjacent to nuclear power plants in Taiwan. Chem. & Ecol. (Accepted).

Hung, T.C. and B.C. Han (1992) Relationships among the species of copper, organic compounds and bioaccumulation along the mariculture area in Taiwan. Sci. Total. Environ., 125, 359-372.

Kannan, K., S. Tanabe, H. Iwata and R. Tatsukawa (1995) Butyltins in muscle and liver of fish collected from certain asian and oceanian countries. Environ. Pollut., 90, 279- 290.

Mathias, C.L., J.M. Bellama, G.J. Olson, F.E. Brinckman (1989) Determination of dibutyltin and tributyltin in sediments and microbial biofilms using acidified methanol extraction, sodium-borohydride derivatization and gas chromatography with flame photometric detector. Intern. J. Environ. Anal. Chem., 35, 61-68.

Mckie, J.C. (1987) Determination of total tin and tributyltin in marine biological materials by electrothermal atomic absorption spectrophotometry. Anal. Chim. Acta., 197, 303.

Salazer, M.H. and S.M. Salazer (1991) Assessing site-specific effects of TBT contamination with mussel growth rates. Mar. Environ. Res., 32, 131-150.

Siu, K.W.M., P.S. Maxwell and S.S. Berman (1989) Extraction of butyltin species and their chromatographic determination as chlorides in a sediments certified reference material for trace metals. PACS-1, J. Chromatogr., 475, 373-379.

Spooner, N., P.E. Gibbs, G.W. Bryan and L.J. Goad (1991) The effect of tributyltin upon steroid titres in the female dogwhelk, Nucella lapillus and the development of imposex. Mar. Environ. Res., 32, 37-49.

Wade, T.L., B. Garcle-Romero and J.M. Brook (1988) Tributyltin contamination in bivalves from united States coastal estuaries. Environ. Sci. Technol., 22, 1488-1493.

Waldock, M.J. and J.E. Thain (1983) Shell thickness in Crassostrea gigas: Organotin antifouling of sediment induced. Mar. Pollut., Bull., 14, 411-415.

Wilken, R.D., J. Kuballa and E. Jantzen (1994) Organotin their analysis and assessment in the Elbe river system, northern Germany. J. Anal. Chem, 350., 77-84.

Yu, T.H. and Y. Arakawa (1983) High-performance liquid chromatographic determination of dialkyltin homologues using fluorescence detection. J. Chromatogr., 258, 189-197.