Trace Determination of Pesticides and their Degradation Products in Water (BOOK REPRINT) (eBook)

1.3.3 Pesticides in river and estuarine waters

The current generation of pesticides is not as fat soluble as the organochlorinated pesticides, such as DDT. They are transported in aquatic systems primarily in the dissolved phase, generally exhibit much shorter half-lives than the organochlorinated pesticides, and have only a minor bioaccumulation potential. This is true for the agricultural use of herbicides, which are being used in increasingly larger amounts than insecticides. In the last few years, the monitoring programmes have started to consider the new generation of pesticides and their transport through rivers and estuarine areas.

Agrochemicals are applied annually to agricultural soils throughout the world. Many of these organic chemicals are transported to surface waters by a variety of mechanisms such as non-point source pollution, ground water discharge, or atmospheric deposition. Millions of pounds weight of relatively water-soluble pesticides, such as atrazine, simazine, alachlor and metolachlor, are applied each year in the USA and in Europe as pre- and post-emergence weed control agents on crops such as corn and soybean [2,100]. It is reasonable to assume that substantial amounts of these compounds may be present in the surface waters which drain agricultural areas of each country. Also, highways and railroads close to rivers may receive the impact of non-agricultural applications of pesticides, such as triazines, chlorinated acids, and phenylureas.

Following considerations based on usage information, physic-chemical properties, and persistency, a priority list of herbicides was established for the Mediterranean countries France, Italy, Greece and Spain. The list, shown in Table 1.14, considers selected herbicides and fungicides that can cause contamination of estuarine and coastal environments. The selection of pollutants has been based on the availability of usage data and the consideration of half-lives, so pesticides that do not exceed a total of 10 tonnes after 90 days of application have been excluded. Fungicides such as carbendazim, ethirimol, metalaxyl, captan, folpet, captafol, vinclozolin and chlorothalonil, which are each used in amounts varying from 20 to 300 tonnes per year in the cited Mediterranean countries are also included. Other fungicides such as mancozeb, maneb, ziram and thiram, which are also used in the same countries but in larger amounts, between 300 and 700 tonnes per year, were not included in a pilot monitoring study because of the need to develop specific methods for analysis at the low μg/l level in water samples [101].

1.3.3.1 Organonitrogen pesticides in river waters

Some of the most relevant information on the environmental behaviour of pesticides in aquatic systems is the extent to which, after application, they can reach surface and river waters and, at a later stage, estuarine waters, with the potential of contaminating coastal sea waters.

The fate of chlorotriazine herbicides in aquatic systems has occasioned several investigations in recent years. The modelling approach, under laboratory conditions, has established that around 92% of the herbicide will be in the dissolved phase whereas about 1% is in the particulate matter [68,100].

Field studies have agreed with the modelling approach, giving values of 99.5% and 0.5% of atrazine in the dissolved and particulate matter, respectively. Atrazine has been transported in a number of river basins with an estimated loss of the applied herbicide varying between 0.4 and 1.8% [48,100,102]. These values have been reported from monitoring in the Camargue region, on the Rhone estuary [103] where the concentration values from the different stations are the same order and agree with the concentration range expected from other rivers such as the Mississippi in Minnesota and others. The concentrations in the dissolved phase were ca. 95–97% whereas in the particulate matter they vary between 3 and 5%. This behaviour can be explained by the Koc (soil adsorption coefficients) which are low, thus giving the pesticide a tendency to leach instead of remaining adsorbed on the particulate matter. The losses of atrazine after application have been estimated to be comparable to other world rivers, with a value of ca. 0.4% of the load.

A number of approaches can be used to estimate fluvial loadings of contaminants to the sea. The approaches used to estimate fluvial loads of atrazine from the Ebro river to the Mediterranean sea [104] were averaging estimators, ratio estimators, and regression-based estimators.

Collective mode (method 1):

=∑i=112CiQi

where Ci is the pesticide concentration in the month i and Qi is the total water amount which outflows in the month i.

Averaging method (method 2):

=36512∑i=112CiQi

where Ci is the concentration on the ith day and Qi is the discharge on the ith day.

Ratio method (method 3):

=l¯q¯Q

where ¯ is the average daily load, ¯ is the average daily discharge and Q is the total flow.

Regression method (method 4):

=∑n=1365qnexpB0+B1lnqn

where qn is the daily discharge, B0 and B1 are the regression coefficients from

c=B0+B1lnq

where c is the concentration and q is the discharge.

In order to apply such expressions, it is necessary to know the discharge for the Ebro river during a period of at least 1 year. Figure 1.13 shows the discharge (m3 s− 1) for the Ebro river in the extended Julian day period, 1 June 1994 to 30 May 1995. Table 1.15 presents the annual usage and loads of various pesticides, applying the four methods mentioned above. The four methods can be seen to give very similar results.

The annual usages of the pesticides in the Ebro river basin, 130, 13, 58 and 37 tons/year, respectively, of atrazine, simazine, alachlor and metolachlor [93], make it possible to calculate ratios of loss/usage of these pesticides as 0.63, 3.85, 0.17 and 0.54%, respectively. However, it is necessary to keep in mind the fact that de-ethylatrazine is a direct metabolite from atrazine; therefore, if the two values are summed, the ratio of loss/usage of 0.97% is obtained (the recovery of de-ethylatrazine, 76%, was not applied). The atrazine ratio is within the expected range.

If we consider the usage of the herbicides studied above we can observe a parallelism between the pesticide detection, and the loads and usage for atrazine and metolachlor. In addition, the loads varied between 0.54 and 0.97% and are similar to values observed in other areas for alachlor, of which much more is applied than metolachlor, but only approximately half as much as atrazine, the pesticide loss is only 0.17%, which is much less than for all the other pesticides. This also occurred in Chesapeake Bay and can be attributed to the much larger Henry’s Law constant of this pesticide as compared to the other herbicides. In this respect, more air–soil volatilization of alachlor occurs compared to the other major organonitrogen herbicides. Along with its degradative process, this is responsible for the smaller relative percentage of the application...