ABSTRACT

In France, among the numerous active substances in legal use on winter oilseed rape, herbicides - most applied on bare soils before emergence - are first concerned by potential environmental risks. As a first step, we tried to acquire a better understanding of transfer mechanisms from rape fields to water. Numerous experiments have been performed, in the field, in semi-field conditions and in the laboratory. Particular attention has been given to trifluralin - Different research teams demonstrated this substance was not so critical as expected. It is strongly adsorbed on the soil and poorly restituted to water. In practice, risks seem to be limited to associated transports on soil particles in case of rather erosive runoff. Metazachlor and tebutam have also been studied - Metazachlor stands in an intermediate situation between trifluralin and tebutam. Our attention was more especially focused on tebutam because, in all the experiments it was taken in account, results were alarming. Without considering toxicity-ecoxicity aspects of the problem, tebutam appears easily transferable to water and highly susceptible to lead to large overpasses of the legal 0.1 µg/l (water for human consumption). In practice, this results are confirmed by lots of routine surface water or ground water analysis ordered by different authorities. Trifluralin is rarely detected. From time to time, metazachlor is pointed out, but at relatively low concentrations. At first not integrated in the detection lists, tebutam is now included, often detected and associated to relatively high concentration levels in water.

INTRODUCTON

According to European agricultural policy and possibility to affect set-aside to industrial crops (1993), french authorities asked farmers to evaluate the impact on the environment of such cultivation compared to typical set-aside put out of production and necessitating low pesticide input. At first CETIOM (Metropolitan oil-producing networks) elaborated an ?Environmental Charter for the cultivation of rape? intended for the production of bio-fuels. After a theoretical evaluation of the topic* confronted to actual knowledge, experts concluded in a first step it was pertinent to focus the investigations to water for environment and to herbicides - most applied on bare soils before emergence - for pesticides. There, trifluralin appeared especially concerned as it was pointed out in the water monitoring lists of CORPEN (Steering Committee for the Reduction of Pollution of Water by Nitrates, Phosphates and Crop Protection Substances arising from Agricultural Activities).

* This development does not consider the second aspect of the question : fertilization.

MATERIALS AND METHODS

The main action was managed by a competent national steering committee and realized in a context of a large partnership based upon two regional groups of participants (Lorraine, Centre Region). Public authorities and organizations responsible for water quality were associated in this project : ``A contribution to the understanding of the Transfer of Herbicides from Oilseed Rape to Water''? (CETIOM 1996, Pierre 1996).

Investigations were conducted for a better knowledge of the actual practices and to quantify the substances used by farmers in France and more especially in the two concerned regions.

Numerous experiments have been performed under natural conditions - in the field or in semi-field conditions - and in the laboratory. Observations have concerned soil and water.

- Dissipations of substances have been measured in different types of soils. Quality was checked upon water coming from drainage, runoff and in ditches in close relation with rape fields. The same investigations were made from bore-holes sited at the top of water tables (1993-96).

- A fundamental work have been devoted to trifluralin and its metabolites at the Soil and Environment Laboratory of ENSAIA (Malterre, 1997). Adsoption of trifluralin on soil was determined by batch experiments, its degradation followed under controlled conditions in incubators and under outdoor conditions in microlysimeters (Malterre et al., 1997) and soil columns. A field scale monitoring was performed to validate the results of the previous experiments and to observe trifluralin behavior compared to tebutam and metazachlor (Pierre et al., 1997 a). Observations on tebutam are now under way (Grébil, 1997).

- Field observations have been performed on a disposal consisting in four cultivated bands (50 x 5 m) connected to basal grassed strips, 0, 6, 12 or 18 m wide (Anonyme 1998).

RESULTS

1 - Investigations.

In France, farmers dispose of around 20 different herbicide active substances. In practice, the control of weeds is based on only few substances applied pre-plant or pre-emergence, on bare soils. They can be used alone or in most of the cases in combination of two sprays. Herbicide applications on developed covers are restricted to specific treatments or previously unsolved difficulties.

Table 1 - France : Application timing of herbicides (fields %)

	1993	1994	1995	1996	1997
Fields without herbicides	2	2	2	1	1
Pre-sowing	50	50	60	55	58
Pre-emergence	55	61	68	74	77
postemergence	21	23	16	17	25

(From Reau and Wagner, 1998).

Lorraine as Centre Region apply herbicides on bare soils and/or on reduced cover of young plants. But as Centre Region goes on with a large use of metazachlor (tebutam is marginal), Lorraine turns itself toward a larger use of tebutam (Pierre et al., 1997 b).

Table 2 - France : Herbicide applications (fields %)

	1993	1994	1995	1996	1997
trifluralin	32	42	56	51	54
metazachlor	24	28	32	28	29
metazachlor + quinmerac	0	0	5	7	9
tebutam + clomazon	32	34	33	42	43
napropamid	16	12	11	9	8

Each all other products always < 5%. (From Reau and Wagner, 1998).

2 - Special attention has been given to trifluralin.

Different research teams have contributed to demonstrate this substance was not so critical as expected. It is strongly adsorbed on the soil and poorly restituted to water. In practice, risks seem to be restricted to associated transports on soil particles in case of runoff.

Trifluralin adsoption on soil is very rapide, strong (Koc = 19500) and hardly reversible. Trifluralin degradation is slow and leed to 8 products. Degradation products are in small amounts and migrate badly in deeper soil horizons. At all, it is estimated that their migration under tillage level is less than 1% of the applied amount (Malterre, 1997).

Table 3 - Lorraine (La Bouzule) : Herbicides in drainage water (µg/l)

				1994-95					1996-97
		30/10 ->	trif.	metaz.	tebut.		5/12 ->	trif.	metaz.	tebut.
	->	14/11	nd	nd	0.366	->	12/12	nd	nd	14.9
	->	22/11	nd	nd	1.164	->	19/12	nd	nd	18.4
	->	26/11	nd	nd	0.654	->	2/01	t	nd	35.6
	->	5/12	nd	nd	0.485	->	16/01	nd	0.3	26.8
	->	13/12	nd	nd	nd	->	23/01	nd	0.6	t
	->	19/12	nd	nd	nd	->	3/02	nd	nd	nd
	->	31/12	nd	nd	nd	->	13/02	nd	nd	nd
	->	11/02	nd	nd	0.599	->	22/02	nd	nd	nd
	->	04/03	nd	nd	nd	->	28/02	nd	t	nd
	->	29/03	nd	nd	nd	->	3/04*	nd	nd	nd

trif. : trifluralin, metaz. : metazachlor, tebut. : tebutam. nd : undetected -

t : traces - * 9/03+21/03+27/03+3/04 - (From G. Grébil, 1997).

3 - Metazachlor and tebutam have also been studied.

3.1 - Metazachlor - Metazachlor stands in an intermediate situation between trifluralin and tebutam. It was taken in account in field experiments in the same manner than trifluralin but without fundamental works.

3.2 - Tebutam - Attention was more especially focused on tebutam. At first, tebutam was not considered as int was not so in use as the others products. It is now still increasing.

In all the experiments tebutam was taken in account, results are alarming. Without considering toxicity-ecoxicity aspects of the problem, tebutam appears easily transferable to water and highly susceptible to lead to large overpasses of the legal ``0.1 µg/l'' (water for human consumption).

Table 4 - Lorraine and Centre Region : Dissipation time in the soil (50 and 90 days).

			DT	50			DT	90
		1993	-1994	1994	-1995	1993	-1994	1994	-1995
		L	C	C	c	L	C	C	c
	trifluralin	54	111	254	230	367	410	830	752
	metazachlor	15	34	45	91	185	257	150	297
	data	8	8	8	8	7	7	9	9

L : Lorraine (field), C : Centre region (f.), c : Centre r., field lysimeter. (From CETIOM, 1996).

4 - Grassed-buffer zones

After lots of experiments, buffer grassed zones are considered to be efficient to purify runoff waters and to limit the quantities of pesticides deversed in surface waters bording cultivated areas (CORPEN, 1997).

In the case of rape - after a 2-years experiment and run-off around 20 mm/ha/year - calibrated grassed strips appear able to reduce run-off by : 85-86 % for water quantities, 64-98 % for pesticides concentrations, 94-100 % for pesticides quantites (Réal, 1998).

Table 5 - Bretagne (La Jaillière) : Efficacy of grassed strips (% 0 m band)

Grass-band	6	m	12	m	18	m
	[c]	Q	[c]	Q	[c]	Q
trifluralin	75	94	75	93	98	96
metazachlor	64	94	79	98	96	97
tebutam	73	99	72	99	81	100

[c] : concentration - Q : quantities - (From Réal, 1998)

CONCLUSION

In practice, all these results are confirmed by lots of routine surface water or ground water analysis ordered by different authorities : (1) trifluralin is rarely detected ; from time to time, metazachlor is pointed out, but at relatively low concentrations ; at first not integrated in the detection lists, tebutam is now included, often detected and associated to relatively high concentration levels in water.

For example, if we consider the 1992-1997 period and the Centre region, official services report that for 655 routine surface water (rivers) analysis : 13 substances (12 herbicides) have been found in more than 10 % of the samples, with atrazin (number 1 : 99 %) ... lindan (number 4 : 58 %) tebutam (number 7 : 56 %) ... metazachlor (number 12 : 14 %) ... Trifluralin have been quite never observed (<1 %) (Chantrel et al., 1998).

All our investigations only consider risks directly associated to crop protection. Water pollution can come from many others causes (Ballanger et al., 1998) taken in account in monitoring programs.

We did not find trifluralin as so critical than expected. Tebutam is still in use in our country. This poses a real problem as this chemical degrades easily, but giving a major product that seems to have not so good environmental characteristics (Grébil, 1997).

For the future, we have to consider the risks encountered in a more global environmental vision, not restricted to water, not focused on herbicides and not limited to substances pointed out up on theoretical statements. We also would have to be more in accordance with actual farmers practices.

References

1. Anonyme, 1998. Etude de l'efficacité des dispositifs enherbés, ITCF - Agences de l'eau, campagnes 1993-94, 1994-95, 1995-96. Les Etudes des Agences de l'Eau (ISSN : 1161-0425), 63 : 30 pp

2. Ballanger Y., Cariolle M., Floquet A., Réal B., 1998. La chasse aux pollutions ponctuelles. Oléoscope, 48 : 6-8

3. CETIOM, 1996. Contribution à la connaissance du transfert des herbicides du colza vers l'eau. Synthèse globale. Résumé. Document CETIOM (ISNB : 2-908645-57-2), Paris : 17 pp.

4. Chantrel P., Charbois-Buffaut P., Kervevan C., 1998. Produits phytosanitaires, surveillance des eaux. Oléoscope, 47 : 23-26.

5. CORPEN, 1997. Produits phytosanitaires et dispositifs enherbés - Etat des connaissances et propositions de mise en oeuvre. Doc. CORPEN, Paris : 88 pp.

6. Grébil G., 1997. Comportement du tébutame dans le sol et l'eau. D.E.A. I.N.P.L. (Institut National Polytechnique de Lorraine) ENSAIA-Nancy : 30 pp.

7. Malterre F., 1997. Impact sur l'environnement d'un herbicide du colza d'hiver : la trifluraline. Thesis INPL (Institut National Polytechnique de Lorraine) ENSAIA-Nancy : 119 pp.

8. Malterre F., Grébil G., Pierre J.G., Schiavon M., 1997. Trifluralin behaviour in soil : a microlysimeter study. Chemosphere, 34 : 447-454.

9. Pierre J.G., 1996. Impact sur l'environnement des herbicides appliqués sur le colza d'hiver. Synthèse de l'action. Proc. Les Rencontres annuelles du CETIOM, 04 déc.1996 : 12-17.

10. Pierre J.G., Malterre F., Schiavon M., 1997. Transfert vers les eaux : des essais en conditions naturelles. Oléoscope, 37 : 15-18.

11. Pierre J.G., Malterre F., Schiavon M., 1997. Trois substances dominent le désherbage du colza. Oléoscope, 37 : 12-14.

12. Réal B., 1998. Bandes enherbées, une solution à la contamination des rivières par ruissellement. Oléoscope, 47 : 27-29.

13. Reau R., Wagner D., 1998. Charte environnement 1993-1997 : Evolution des pratiques agricoles pour la jachère énergétique. Doc. CETIOM : 21 pp.