Historical Contribution of Different Sources to Environmental Dioxin Pollution Estimated from the Lake Shinji Sediment Core

(PDFファイルはここをクリック。PDFfile is here.)

Shigeki Masunaga^{1, 2}, Yuan Yao^{1, 2}, Isamu Ogura¹, Satoshi Nakai¹, Yutaka Kanai³,
Masumi Yamamuro³, and Junko Nakanishi^{1, 2}

^{1 Institute of Environmental Science and Technology, Yokohama National University
79-7 Tokiwadai, Hodogaya-ku, Yokohama, 240-8501, Japan
2 CREST, Japan Science and Technology Corporation
Kawaguchi, Kawaguchi 332-0012, Japan
3 Geological Survey of Japan, 1-1-3 Higashi, Tsukuba 305-8567, Japan}

^Introduction

^{A significant portion of dioxins accumulated in surface aquatic sediment in Japan was indicated to have originated from agrochemicals, especially pentachlorophenol (PCP) and chloronitrofen (CNP)(1). Since these chemicals were used extensively as paddy field herbicides in the past, their present contribution to pollution may be less than that in the past. Thus, it is of interest to estimate their historical contribution to environmental dioxin pollution. In this study, we analyzed dioxins in a dated sediment core taken from Lake Shinji which receives effluent from agricultural land and some local towns. A total of more than 80 gas chromatographic peaks corresponding to individual congeners or groups of congeners were quantified in order to perform a detailed statistical analysis.}

^{Materials and Methods}

^{Sediment core: A sediment core sample was taken from the western part of Lake Shinji, Shimane Prefecture, in 1994. The core was sliced into 1-cm-thick disks, and the average sedimentation rates were estimated to be 0.26 g/cm2/year by the Pb-210 method and 0.25 g/cm2/year by the Cs-137 method(2). Both estimates were similar and the value obtained using the Pb method was used in this study.}

^{Dioxin analysis: After the addition of 13C-labeled internal standards, dried sediment disks (about 4 g) were Soxhlet-extracted with toluene for 20 hours. They then were treated by alkaline hydrolysis and concentrated sulfuric acid. They were further cleaned using a series of silica gel, aluminum and carbon columns. The final PCDD/F and coplanar PCB fractions were concentrated to 25 m l and spiked with 13C-labeled recovery standards for HRGC/HRMS analysis. Both DB-5 and DB-17 columns (J&W Scientific) were used for quantification.}
　

Results

Dioxins in the sediment core: More than 80 gas chromatographic peaks corresponding to the individual tetra- through octa-chlorinated PCDD/F congeners or groups of congeners were quantified using the DB-5 column. All the 2378-chlorine-substituted congeners were quantified using both DB-5 and DB-17 columns. Some of the results are shown in Table 1.

Table 1. Dioxin concentrations in Lake Shinji sediment core
(pg/g dry sediment or pg TEQ/g dry sediment)

* Calculated using the I-TEFs (WHO/ICPS, 1988). ** Calculated using the TEFs for human (WHO, 1998)

A drastic increase in the total PCDD/F concentration in sediment occurred during 1945-1970 followed by a small decrease during 1972-1994 (Figure 1). The major components that increased during 1945-1970 were OCDD and HpCDD congeners, which are known impurities of PCP⁽³⁾. They decreased during 1972-1976 but have remained at the same level since 1980. The period between 1972 and 1976 corresponds well to the period during which PCP use declined rapidly in Japan (1970-1972). In contrast to the highly chlorinated dioxins, TCDDs, PeCDDs and TCDFs (especially 1368-TCDD, 1379-TCDD, 12368-PeCDD, 12379-PeCDD and 2468-TCDF) increased during 1964-1977 but have since remained at the same level. These congeners are reported to be the major impurities of CNP⁽³⁾. The period from 1964 to 1977 corresponds well with the period from 1966 to 1972 during which the use of CNP in Japan increased rapidly.
 
 

（クリックすると大きくなります）

Figure 1. Dioxin concentration in dated Lake Shinji sediment core.

Principal component analysis: To identify the possible sources of dioxin in the sediment, principal component analysis was performed using a correlation matrix calculated from congener-specific data (83 GC peaks as variables and 12 slices of sediment core as cases). Analysis after the varimax rotation yielded three major principal components (PCs) (Table 2). Based on the characteristic congeners in each PC, PC-1 and PC-2 were judged to be impurities of PCP and CNP, respectively. It was not possible to attribute PC-3 to any known sources confidently; however, PC-3 might correspond to another major dioxin generator, incineration. The principal component scores of all PCs are shown in Figure 2. The component score of PC-1 increased during the 1960s and reached its maximum in around 1970. The component score of PC-2 follows the same trend as that of PC-1 but with a delay of several years. The behaviors of PC-1 and PC-2 were in accordance with the amounts of PCP and CNP used in Japan, respectively.

　Table 2. Results of principal component analysis with varimax rotation

 
　
Estimated trends of different dioxin source contributions: Based on the result of PC analysis, we assumed that PCP, CNP and incineration (atmospheric deposition) are the three major sources of dioxin in Lake Shinji. Their contributions to pollution in sediment were estimated by multiple regression analysis using congener profiles of dioxin impurities in PCP⁽³⁾ and CNP⁽³⁾ and of

（クリックすると大きくなります）

Figure 2. Component scores of three principal components (PCs)

atmospheric dioxin deposition. For atmospheric deposition, data obtained in the Kanto area⁽⁴⁾ were used because of the lack of congener-specific data in this area. The result indicated that PCP had been the greatest contributor to aquatic sediment pollution since the 1950s (Figure 3). The contribution from CNP began in the 1970s. Atmospheric deposition increased during the 1950s and 1960s and subsequently leveled off.

（クリックすると大きくなります）

Figure 3. Contributions of different sources to dioxin pollution in sediment core

　

Discussion and Conclusion

Detailed analysis of a dated sediment core showed that dioxin input to aquatic sediment increased in accordance with PCP and CNP use. The input did not significantly decrease even after the decline of their use, indicating that dioxins remaining in agricultural land continued to run off and pollute the aquatic environment. A discrepancy between the contributions of different sources presented here and those estimated from the dioxin source inventory⁽⁵⁾ was noted. This may be due partly to the limitation of the present statistical analysis based on data consisting of a very wide range of concentrations (very high concentration of OCDD and low concentration of many other congeners).
　

Acknowledgements: This work has been supported by Core Research for Evolutional Science and Technology (CREST) of the Japan Science and Technology Corporation (JST).
 

Reference

1) Masunaga, S., Sakurai, T., Ogura, I., Nakanishi, J.: Organohalogen Compounds 1998, 39, 81-84.
2) Kanai, Y., Inouchi, Y., Yamamuro, M., Tokuoka, T.: Chikyukagaku (Geochemistry) 1998, 32, 71-85.
3) Masunaga, S., Nakanishi, J.: to be presented at Dioxin'99 in Venice, Italy, Sept. 1999.
4) Ogura, I., Masunaga, S., Nakanishi, J.: to be presented at Dioxin'99 in Venice, Italy, Sept. 1999.
5) Masunaga, S.: Proc. of 2nd International Workshop on Risk Evaluation and Management of Chemicals, in Yokohama, Japan, Jan. 1999.