Miljøprojekt, 902 – The Influence of Sorption on the Degradation of Pesticides and other Chemicals in Soil – 10 Pesticides mainly bound to clay minerals: glyphosate, paraquat, and diquat

The Influence of Sorption on the Degradation of Pesticides and other Chemicals in Soil

10 Pesticides mainly bound to clay minerals: glyphosate, paraquat, and diquat

Both glyphosate and the metabolite of glyphosate, AMPA, are quickly and strongly adsorped in most soil types. Unlike most other pesticides glyphosate will not be very strongly bound to the soil organic matter, but on the other hand primarily to the soil minerals. The content of cations on the clay minerals is thought to be important for the binding. The binding is probably caused by metal cations (e.g. Fe³⁺ and Al³⁺) forming complexes with glyphosate and AMPA through the phosphonic acid part (Glass, 1987; Sprankle et al., 1975). The binding of glyphosate takes place in competition with free inorganic phosphate. The binding mechanisms have not been fully explained, however (Franz et al., 1997).

K_d-values for glyphosate in a number of surface soils are reported to be between 33 and 377 (Jacobsen et al., 1998). In a study of the sorption of ¹⁴C-marked glyphosate to a number of Nordic soils (Tiberg, 1998), K_d-values above 100 were similarly found in most soils. Measurements of the sorption in subsurface soils showed an even greater sorption than the corresponding surface soils. Studies in Sweden (Torstensson & Lindholm, 1988) have similarly shown K_d-values of 26 to 83 in soil sampled below railways in Sweden. Similar values have been found in Danish soils and in samples taken at a depth of about 1 metre (Petersen et al., 1998).

The degradation time for glyphosate can vary greatly from one soil to another. On measuring the degradation time on the basis of the liberation of ¹⁴CO₂- from ¹⁴C-labelled glyphosate, half-lives from a few days to 175 days were found. In soil sampled under the plough layer the degradation was considerably slower (Jacobsen et al., 1998).

In most cases the half-life seems to be relatively short in agricultural soils and has been shown to be similarly short in forest soils (Torstensson, 1985) even though a few half-lives of more than 10 years have been proved in 2 soils from Hawaii (Franz et al., 1998).

In a number of studies, agricultural soils have been analysed for residues of glyphosate and AMPA between 3 and 8 months after being treated with glyphosate. The results showed (Torstensson, 1985) that residues of glyphosate and AMPA can be proved in the soils after 3 to 8 months. In some cases mainly AMPA, in others mainly glyphosate. No results have been found in the refereed studies showing an accumulation of glyphosate or AMPA after incubation for a long time.

It has previously been shown that pesticides that are strongly bound to clay particles can be stabilised in the soil so that they get a very long half-life even though they are substances that in unbound state are relatively easily degradable. Especially long persistence after binding has been proved for the herbicide paraquat. Studies by Fryer et al. (1975) showed that almost the whole amount of paraquat, which was added in field experiments over 6 years, could be found again in the soil. Most remained in the upper 5 cm, but low contents were also found to a depth of 25-35 cm, which were attributed to the sampling method or to particle transport, cracks or earthworm activity. Soil milling did not influence the found residues of paraquat.

Similarly, Smith and Mayfield (1978) found that under a number of different incubation conditions paraquat was not degraded in 16 months, and Riley et al. (1976) stated that bound residues of paraquat were not available to living organisms such as plants, earthworms or microorganisms. Riley et al. (1976) distinguished between "loosely bound" and "tightly bound" paraquat as they defined "loosely bound" paraquat as the part that could be extracted from the soil with 5M NH₄Cl and "tightly bound" paraquat as the part that could be extracted by boiling with 18 N sulphuric acid. Riley et al. (1976) stated that it has been shown in greenhouses that "loosely bound" paraquat potentially might be available to plants again.

Eberbach (1998) studied the mineralisation of ¹⁴C-labelled glyphosate and demonstrated a biphasic course in which a relatively rapid mineralisation of available glyphosate and a slower mineralisation of bound glyphosate took place. The biphasic course of a degradation curve is thus not reserved for chemicals that are bound to the soil organic matter but occurs also for substances mainly bound to the clay particles of the soil. Nomura and Hilton (1977) and Rueppel et al. (1977) also mentioned the influence of sorption on the course of the degradation of glyphosate. Eberbach found that glyphosate is most strongly bound in soil with acid pH. Heinonen-Tanski (1989) found that liming increases the degradation rate for glyphosate and de Jonge et al. (2001) showed that the binding of glyphosate decreased with an increased presence of phosphate. Despite the biphasic course proved by Eberbach, there are many proofs that a higher biological activity in the soil increases the decomposition rate of glyphosate. Thus, it is to be expected that the decomposition will be quicker in a reduced tillage soil than in a conventionally tilled soil.

Fomsgaard et al. (2003a; 2003b) carried out lysimeter experiments with isoproturon and glyphosate. Experiments with ¹⁴C-labelled isoproturon and unlabelled glyphosate were carried out as repeat determination in soil that had been tilled conventionally while experiments with ¹⁴C-glyphosate and unlabelled isoproturon were carried out as repeat determination in soil that had been tilled with a reduced tillage technique. In the reduced tillage soil, glyphosate had been applied regularly until 3 years before sampling while glyphosate had not been applied in the last 5 years in the conventionally tilled soil. 16% and 14%, respectively, of the added radioactivity in the ¹⁴C-glyphosate lysimeters were found in the top 10 cm of the soil while 0.2% and 0.7%, respectively, were found in soil from 10-20 cm. Alkaline extractions were carried out in 0-10 cm depth and specific analysed on LC/MS of the residues of glyphosate and AMPA in the entire soil profile in both repeat determinations. The amounts of ¹⁴C found in the upper soil layers were on a level with findings of non-extractable radioactivity from ¹⁴C-mecoprop in Felding et al. (1996) and considerably lower than the findings of Burauel et al. (1998) of more than 80% of the added radioactivity in the upper 30 cm in 20 different outdoor lysimeter experiments carried out with ¹⁴C-labelled atrazine, terbuthylazine, chloridazon, dichlorprop-P, methabenzthiazuron, pyridate, and anilazine and considerably lower than the abovementioned results of ¹⁴C from ¹⁴C-isoproturon in lysimeter plough layers from the two lysimeters taken from conventionally tilled soil.

With the specific analysis significantly greater amounts of AMPA residues were found in the soil that had been cultivated with reduced tillage than in soil that had been conventionally tilled. Decomposition has either been slower in the soil that has been cultivated with reduced tillage than in the soil that has been tilled conventionally or the decomposition of glyphosate and AMPA has possibly not been sufficiently rapid for the residues of the previous years to disappear completely from the soil given the actual pattern of pesticide consumption. As glyphosate according to many studies is strongly bound, the second part of the two-compartment process can be expected to pass off exceedingly slowly. No direct comparisons can be made between the amount of ¹⁴C and the amounts of the substance, which were proved in the specific analysis, as the ¹⁴C-analyses do not measure the specific substances. It would be highly interesting to know how glyphosate was bound to the soil particles of different types and sizes. Gimsing et al. (2001) showed that the soil oxides are of great importance as regards the ability of the soil to adsorb glyphosate while the content and the type of clay are of small importance. It was shown by deJonge and deJonge (1998) that a larger dispersion of clay did not increase the binding of glyphosate to the clay minerals.