Occurrence of Pesticides in Selected Water Sources in Israel

The concentrations of 12 organic pesticides were deter- mined in selected water samples in Israel during 1972-73. Where pesticides were identified, th...
0 downloads 0 Views 443KB Size
NOTES

Occurrence of Pesticides in Selected Water Sources in Israel Yair Kahanovitch Tahal, Water Planning for Israel, Tel-Aviv, Israel

Noam Lahav" Department of Soil and Water, The Faculty of Agriculture, Rehovot, The Hebrew University of Jerusalem

The concentrations of 12 organic pesticides were determined in selected water samples in Israel during 1972-73. Where pesticides were identified, their concentrations were much lower than the permitted levels; in the great majority of the water samples studied, the pesticidal concentrations were in the order of a few nanograms per liter. The only samples that did not contain pesticides were from Dan River and drainage water from agricultural fields. The most widekpread compounds in the water samples under study were r-BHC and a-BHC. The great concentrations of T-BHC in municipal sewage indicate that this is an important source of BHC in the water bodies under study. The relative low and constant pesticide levels in Lake Kinneret and the other water reservoirs probably result from the buffering action of the bottom sediment. At the present time more than 90% of the water resources of Israel are being used for agricultural, urban, and industrial purposes, pumped in substantial amounts from Lake Kinneret. This lake is the main water reservoir in Israel and its waters are pumped all the way down to the southern arid area by a water carrier system (Figure 1). Preliminary studies on the content of several chlorinated hydrocarbon pesticides in water samples from the northern part of this country (1, 2 ) have shown that most of the water samples did not contain detectable concentrations of the pesticides under study. Lindane, however, was present in several water samples. The increasing use of pesticides. on the one hand, and the limited water resources of this state, on the other hand, make the purity of the water a problem of utmost importance, and it has been decided to start a more detailed and sensitive investigation on pesticide residues in Lake Kinneret as well as in other water bodies elsewhere in Israel. In the present work, we report on the study concerned with surface water.

Experimental Sampling sites. The sampling sites are shown in Figure 1. At the following six places, water was sampled at least three times during the survey period: The entrance of the Jordan River into Lake Kinneret (No. 4) The exit of the Jordan River from Lake Kinneret (No. 12) Drainage waters from agricultural areas which flow into the lake (No. 11) Kishon reservoir which receives its water from both Lake Kinneret and winter floods of the Kishon watershed with sewage from two neighboring small towns ( N o . 231 A small water reservoir which collects drainage and runoff water from the nearby agricultural areas (No. 24) 762

Environmental Science & Technology

Drainage and runoff water before entering into site No. 24 (No. 2 5 ) All the water bodies under study are located in areas of heavily textured soils. In addition to the above, water was sampled from the lower Jordan River (No. 13) and from the Bet-Netofa (No. 22) and Zohar (No. 26) reservoirs. The latter two are part of the water carrier system in which the water from Lake Kinneret is pumped into the southern part of Israel. Water samples were usually taken from depths of about 10-20 cm. Several samples from Lake Kinneret were taken from the upper surface. Extraction Procedure. Water samples of 9-10 liters were usually extracted a t the sampling sites in 2-liter separating funnels with about 100 ml of hexane according to Hindin (3). Special precautions were taken to prevent contamination during the extraction; these included three washings of the funnels with acetone followed by three washings with hexane prior to the extraction. The reliability of these precautions can be seen from the fact that certain water samples were pesticide-free even though they were extracted after the extraction of contaminated samples. Preliminary studies have shown that it made little difference whether the extractions were carried out for 1 min or for 10 min. N o pretreatment was applied prior to extraction. Analytical Procedure. After drying over anhydrous sodium sulfate, the extract was evaporated in a Buchi rotavapor concentration apparatus; 2-3 ml of hexane were then added to redissolve the pesticides. The cleanup procedure was carried out according to Law and Goerlitz ( 4 ) , the only modification being the use of chromatographic columns of 8 cm instead of 4 cm. Ef-

l

7:

Figure 1 . Sampling sites for pesticide residues from Israel

fluent volumes of 0.5-2 ml were collected and their amounts determined gravimetrically. The total effluent volume was usually 10 ml. In several cases an additional volume of 10 ml of 1:1 benzene:hexane solution was passed through the column to elute malathion. Pesticide mixtures of known concentrations were used to identify the eluent fraction in which each compound was eluted. The pesticidal determination was made with a Packard Model 7400 gas chromatograph equipped with an electron capture detector. The glass columns used were of 4 mm i.d. and 180 cm length and packed with ( a ) 5% OV-1 + 2% QF-1, and ( b ) 5% QF-1.

Nitrogen was used as the carrier gas. The temperatures were: Injection point, 210°C; column oven, 190°C; detector, 200°C. Usually aliquots of 0.5-8 pl. were injected into the two columns and only if the presence of a pesticide could be confirmed on both columns was it reported. Because some organic compounds might possibly have the same retention time on one column as the pesticide of interest, the lowest value is reported. Remembering that the cleanup procedure is also a chromatographic separation, the identification of the pesticides is based on the retention time of three columns.

Table I. Concentration of Five Pesticides in Water Samples from Lake Kinneret (Figure 1) Tests were carried out for heptachlor, heptachlor epoxide, aldrin, endrin, a n d o,p'-DDD but none was detected in a n y of the samples except No. 16. The presence of Thiodan, Treflan, a n d Diazinon was also tested for, b u t none was detected in a n y of the samples. The presence of Malathion was tested in all the samples except No. 6 (50.cm depth) a n d No. 16. Only sample 17 was f o u n d to contain this compound (302 ng/l.) Sampling time and date

Concentration, ng/l. ?-BHC

p.p'-DDT

Dieldrin

0.9

22.1

ND

ND

0.5

1.0

-

-

8/72

2.7 4.6 2.4 2.7 1.8 1.1

1.6 6.3 1.6 7.3 1.9 1.2

ND 15.8 ND ND 2.5 ND

ND 0.3 ND ND 0.6 ND

12 12 12 12 12

14. 8/72 9. 10172 21. 11/72 1 2 . 12/72 9. 1/73

0.9 1.8 4.4 0.9 4.4

0.9 1.1 3.4 1.4 9.2

ND ND ND ND

ND

2.6

0.5 ND ND ND

12 14 15 16

9 . 5/73 9. 1/73 12. 12/72 9. 1/73

19.5 0.9 3.1 1.3

3.4 0.8 1.9 9.9

ND ND ND

ND ND ND

-

-

1.0

14.

8/72

7.6 22.6 0.6

1.1

ND ND ND

ND ND 0.2

9.

1/73

3.7

ND

ND

Site no.

Description

Lake Kinneret

6

21.

9/72

6 7 7 8 9 10 12

9. 1/73 9. 5/73 21. 11/72 9. 1 / 7 3

Ein-Gev Ha h o n

Maagan Deganya. Exit from the lake

4.

Tiberias

17 18 19

Ginnosar T a b h a pumping station

20

a-BHC

8.0

1.8

Remarks

Sample taken from uppermost layer Sample taken from a depth of about 50 cm

Sample taken from uppermost layer

Taken near the entrance of sewage Water taken from depth of several meters

N D = n o n e detectable.

Table II. Concentrations of Seven Pesticides in Water Samples from Watershed of Lake Kinneret Heptachlor epoxide was not detected in a n y of the samples. Testswere carried out for Malathion, T h i o d a n , Treflan, a n d Diazinon, but none was detected in samples 2, 3, 4, a n d 11. E n d r i n was found only in sample 4 (14. 8/72) at the rate of 1.0 ng/l. Concentration, ng/l.

No.

Sampling time and date

~~~

a-BHC

?-BHC

Heptachlor

o,p'-DDD

ND ND

ND ND

1.3 ND

0.3 ND

1 2 3 4

9/72 21. 1 2 . 12/72 12. 12/72 7/72 4.

ND 0.7 1.6 2.0

ND 1.5 5.2 1.7

0.6 ND

4 4 4 5 11

14. 8/72 1/73 21. 5/73 9. 21. 9/72 9 . 10172

1.2 1.5 2.4 0.6 ND

1.7 0.3 3.0 0.1 ND

ND ND ND ND ND

Traces

11

12. 9.

0.1 ND

0.1 ND

ND ND

11

12/72 5/73

p,p'-DDT

ND ND

-

Dieldrin

ND ND ND ND

D a n River a t the foundainhead Jordan River, Northern H u l a Valley Jordan River, Southern Hula Valley Jordan River, before entering

Lake Kinneret

-

Traces

ND ND

4.0 ND 2.7 ND ND

ND ND

ND ND

ND ND

ND

Description of sampling site

ND ND ND

Neshushim River Drainage water from agricultural areas

N D = n o n e detectabie.

Volume 8 , Number 8 , A u g u s t 1 9 7 4

763

In several determinations, only one column of the gas chromatograph apparatus was used. In these cases the identification is based on two chromatographic columns. Concentrations were calculated from the peak height, since a linear relationship was found between the two. Due to the large water volumes used in the samples under study. pesticide concentrations lower than 1 ng/l. could be detected without difficulty and considered reliable.

Results and Discussion The results (Tables 1-111) show that in most of the water samples studied the pesticidal concentrations were in the order of several nanograms per liter and that the most widespread compounds were y-BHC (lindane) and a-BHC. The only samples free of BHC were those from the Dan River (No. 1) and the drainage water (No. 11). The Meshushim River (No. 5 ) contained only small amounts of y-BHC and a-BHC, presumably because the water sample was taken several kilometers from the fountainhead. No significant differences were found between the water reservoirs under study. A comparison of these results. using water criteria proposed by Ettinger and Mount ( 5 ) or by Nicholson (6) reveals that contamination of the water samples under study was very low. It is of interest, however. to discuss the possible mechanisms of pesticide pollution of the waters under study. Lake Kinneret, I t s Watershed and Water Carrier. It seems that the upper layer of the lake’s water has greater concentrations of y-BHC than the layers below. This may be due to the higher adsorption of y-BHC by algae (7. 8). It should be noted that the pumping station in site N o . 19 (Figure l), which pumps the water from the lake to the water carrier. takes the water from a depth of several meters. The water sample of site No. 16 was taken a t about 100 meters from the entrance of the sewage stream into the lake. This sewage flows regularly in a special pipeline and is emptied into the Jordan south of Lake Kinneret. During the sampling period the pipeline was broken and its content flowed into Lake Kinneret. The content of y-BHC and e-BHC in the sewage on the same day was 125 ppt

and 4.1 ppt, respectively. Thus the sample taken a t site 16 reflects the fact that the lake was contaminated by yBHC. This contamination mechanism is much more evident in the Jordan water (site No. 13,Table 11). The relative constancy of the BHC concentrations along the beaches may be attributed to several factors such as the mixing rate of the water and the adsorption of this compound by sediment a t the bottom of the lake. In a preliminary study we have found BHC and p,p’-DDT at the bottom of the lake with concentrations in the order of several hundreds ng/kg. Lindane (y-BHC) is a common insecticide and its domestic uses are quite widespread in Israel. This is probably the reason for the high lindane concentrations in sewage. When we take into account the lack of BHC in the Jordan Fountainhead and that lindane is little used in agriculture, sewage would seem to be an important source of this compound in the Jordan River and Lake Kinneret. The BHC concentrations found a t the Jordan River’s entrance to the lake (site No. 4) are comparable to those found in the lake and presumably originate from inhabited areas in upper Galilee and from cultivated fields. The lake itself is probably supplied with lindane from Tiberias and several villages on the beaches. Recently, the sewage of Tiberias has been diverted into the lower part of the Jordan. This undoubtedly decreased the lindane concentrations in Lake Kinneret but deteriorated the situation in the Jordan. It is noted, however, that more determinations should be carried out to establish the pesticide level and fluctuations in the Jordan River. The reason for the great Malathion concentrations (No. 17, Table I) is not known. On the same date as this sample was taken, six additional samples were also taken from the lake and were found to be free of this insecticide. The artificial water reservoirs at Bet Netofa (No. 2 2 ) and Zohar (No. 26) are part of the water carrier system and they store the water of Lake Kinneret. Except for a t mospheric pollution, no other known mechanism can cause their contamination with the pesticides under study. Therefore it seems that the pesticides detected in their waters come from Lake Kinneret. Kishon and Gevat Reservoirs. There was no significant difference in the pesticide content of these two water

Table 111. Concentrations of Six Pesticides in Water Samples from Different Sources Heptachlor, heptachlcr epoxide a n d aldrin were not detected in a n y of the samples. Malathion, T h i o d a n , Treflan, a n d Diazinon were tested for b u t not detected in water samples 13, 21, 23 (10. 1/73) a n d 26 Concentration, ng/l.

Sampling time and date

a-BHC

yBHC

0.p’-DDD

p,p’-DDT

Dieldrin

Endrin

23.6 9.3 4.7 1.0 1.8 3.5 2.8 42.0 4.8 9.2 3.6 1.7

119.0 14.2 2.7 2.1 1.7 3.0 3.4 21.0 6.3 4.1 1.9 0.4

ND ND ND ND 1.1 ND 0.4 ND

ND ND 0.9 ND 6.0

ND 0.3

25

9 . 5/73 21. 11/72 14. 8/72 1 3 . 8/72 3. 8/72 13. 8/72 22. 9/72 10. 1/73 3. 8/72 1 3 . 8/72 22. 9/72 3 . 8/72

ND ND 1.5 0.8 0.6 8.0

0.2 ND 1.0 ND ND 0.7 5.0 ND

ND ND ND ND ND ND ND ND 0.7 ND

26

1 3 . 8/72 22. 9/72 9. 10172 2 . 1/73

1.3 1.3 2.4 4.3

0.4 0.7 4.4 2.6

ND 4.3 2.0 ND

4.2 1.3 0.6 0.1

0.2 ND 1.1 0.6

2.4 ND 9.0 ND

No.

13 21 22 23 24

25

~~~~~~

N D = none detectable.

764

Environmental Science & Technology

~~~~

~

~

~~

-

-

Description of sampling site

Jordan River

Yasuor reservoir Bet.Netofa reservoir Drainage a n d runoff water Kishon reservoir, North Kishon reservoir, South Gevat reservoir Drainage a n d runoff. E n trance to Gevat reservoir

Zohar reservoir

bodies even though they are supplied by different sources. More than 60% of the Kishon reservoir waters are pumped from Lake Kinneret. The rest is contributed by winter floods from the watershed. In addition, some sewage from two neighboring small towns is also emptied into the Kishon reservoir. On the other hand, the Gevat reservoir gets its water only from the drainage of neighboring rural areas and from winter floods. A possible explanation for the observed similarity in the pesticide concentrations in the two reservoirs is the buffering capacity of the bottom sediment. Indeed preliminary study has shown that BHC, DDT, endrin, and dieldrin were found in the bottom sediment in concentrations on the order of several hundred nanograms per kg.

Literature Cited

(1) Hindin, B., “Pesticides, Plant Growth Regulators and Food

Ackno~cledgment

Additives,” Gunter Zweig, Ed., Academic Press, New York and London, 1967. (2) Berginsky, H., “Survey of Pesticides in Selected Water Sources in Israel,” Continuation of Hindin’s Survey, Mekorot Water Co. Ltd. (in Hebrew), 1969. (3) Hindin, E., “Chlorinated Hydrocarbon Pesticides in Selected Water Sources in Israel, a Preliminary Study,” Mekorot Water Co. Ltd., Water Quality Department, 1967. (4) Law, L. M . , Goerlitz, D. F., J . AOAC, 53, 1276-86 (1969). (5) Ettinger, M . B., Mount, D. I., Enciron. Sci. Technol., 1, 203-5 (1967). (6) Nicholson, H . P., Proc. Wash. Acad. Sci., 59, 77-89 (1969). ( 7 ) King, P. H.. Yeh, H . H . , Warren, P . S., Randall, C. W., J . Amer. Water Works Ass., 61,483-6 (1969). ( 8 ) Pionke, H. B.. Chesters, G., J . Enuiron. Qual., 2, 29-45 (1973).

We thank S. Milchen and H . Popisky for their help in the analyses of the water samples.

Received for recieu August 13, 1973. Accepted April 4, 1974. ,stud? u a s financed by Tohal, Water Pianningfor Israel Ltd.

This

Personal and High-Volume Air-Sampling Correlation Particulates Phillip M. Duvall and Richard C. Bourke” Detroit Diesel Allison Division, General Motors Corp., Indianapolis, Ind. 46206

Airborne particulate measurements are required both outside and inside an industrial plant to determine compliance with existing control regulations. The Environmental Protection Agency (EPA) specifies a high-volume sampling method for outside ambient air while Occupational Safety and Health Administration (OSHA) personnel generally use personal monitoring for determining particulate levels inside a plant. A personal monitoring method was developed by Detroit Diesel Allison which correlates with the EPA reference method in three types of ambient atmospheres: a “clean” laboratory area, outside ambient air in an industrial location, and inside an industrial plant in an area where oil mist is the predominant contaminant. Good agreement was obtained in all three cases after compensation was made for the effects of adsorbed moisture.

The need for reliable measurements of particulate concentrations in ambient air both outside and inside an industrial plant is well known. However, two different methods are used in these two areas where permissible limits are set by two different federal agencies-EPA and OSHA. In the industrial organization, however, both types of measurements generally are made by the same personnel and, for the sake of reliability in measurements and economy, a need for correlation is apparent. Ideally, the same instrument could be used in both cases or, at least, the monitors should be interchangeable so that the number required is reduced. Since both methods are gravimetric determinations of particulate concentrations, they should agree and correlation would mean a higher degree of confidence in both sets of data, Particulate concentrations in outside ambient air are generally measured by a high-volume reference technique

specified in detail by the EPA ( I ) . High-volume sampling has been used for some time and, if the recommended procedure is carefully followed, the obtained data have been found to be quite reliable. N o similar procedure has been established by OSHA, although samples are generally collected on membrane filters by means of personal monitors and general-area samplers ( 2 ) . The airflow of these monitors is approximately one thousandth of that of a high-volume sampler. Also, the filter material used iic the high-volume sampler is usually glass fiber purportedly capable of collecting particles greater than approximately 0.3 p in diameter, while the personal samplers use a variety of membrane filters of different pore-size diameters. In this study, filters with an average 0.8-p diameter pore size were used because of their general usage and availability. Differences also occur in the positioning of the impingement surfaces of the two different types of filters. The surface of the high-volume filter is facing upward toward an approximately vertical downward airflow, while the membrane filter surface is usually facing downward at approximately 45” to the horizontal, toward an upward airflow to simulate personnel breathing-zone conditions as closely as possible. In spite of the differences between the two methods, if any reliance is to be placed in the data obtained by either technique, a correlation between the two must be established. The purpose of this Kote is to describe the development of a membrane-filter, personal-monitoring technique that correlates with the high-volume air sampling reference method of the EPA in three different types of atmospheres-a clean laboratory area, outside ambient air in an industrial location. and inside an industrial plant where oil mist is the predominant contaminant. Selection of these three atmospheres permits comparison of the collection efficiencies of the two methods for particulates of widely varying size, chemical composition, and ambient air concentrations. For example, from optical and scanVolume 8 , Number 8, August 1974

765