Vous êtes sur la page 1sur 7



Institut J~r Bodenbiologie
Forschungsanstalt f i r Landwirtschaft
D 33 Braunschweig
Bundesallee 50

The disappearance of the herbicide, Avadex (40% diallate), from five agricultural soils
(differing in either pH, carbon content, or nitrogen content), incubated under sterile and
non-sterile conditions, was followed for a period of 20 weeks. Avadex was rapidly lost from
microbiologically active soils, with over 50% of the applied (2.5 ppm) dosage
disappearing within four weeks; losses from sterile soils were much slower with
recoveries of over 50% after 20 weeks. Incubation of soil with Avadex to which
*~C-labeled diallate had been added resulted in rapid formation of t4CO2 from
microbiologically active samples and only very slow 14CO2 formation from sterile
samples. Substantial quantities of radioactivity were retained as unextractable residues
in both sterile and non-sterile soils after seven days incubation. From these data it was
concluded that the disappearance of the herbicide from non-sterile soils was mainly due
to microbial degradation and to binding of diallate or its metabolites as residues to
undefined soil components. Losses from sterile soils were attributed to both binding of
residues and to a slow chemical degradation. Avadex degradation by pure cultures was
studied using representative fungi isolated from the five soils. Of the fungi tested,
Phoma eupyrena, Penicillium j a n t h i n e l l u m , and Trichoderma harzianium could de-
grade at least 20% of the applied (2.5 ppm) herbicide after ten days incubation.
Degradation of Avadex in soil cultures of T. harzianum was found to be slower than
degradation in liquid nutrient cultures.

Carbamate herbicides are of importance in agriculture because of their effectiveness at

low levels of application, their low mammalian toxicity, and their general short-term
persistence in soils. The metabolites of these compounds appear to be comparatively
innocuous (Kaufman 1967), and are themselves degraded in soil.

In the present study, the fate of the thiolcarbamate herbicide, diallate (S-2,3-
dichloroallyl N,N-diisopropylthiolcarbamate), in soil, was investigated. Of particular in-
terest was the disappearance of diallate in relation to microbial activity and soil characteris-

It has been previously reported that diallate rapidly disappears from soils, and it has been

1Presented in the Third International Congress of Pesticide Chemistry (IUPAC), Helsinki, 1974.

Archives of Environmental Contamination

and Toxicology Vol. 4, I-7 (1976)
9 1976 by Springer-Verlag New York Inc.
J. P. E. Anderson and K, H. Domsch

suggested that disappearance of the compound is due to its degradation (Banting 1967;
Smith 1970) and its partial mineralization to COu (Kaufman 1967).

Materials and methods

Avadex (BASF, Ludwigshafen, FRG), a commercial formulation containing 40% diallate,
was used in this work. Also used was z~C-carbonyl-labeled diallate, purchased from the
Radiochemical Centre, Amersham, United Kingdom.
Solvents were redistilled benzene, isopropanol, and acetone.

Five agricultural soils, having the following characteristics, were used: soil A, a brown
earth, pH = 5.2, total carbon (Ct) = 0.61%, total nitrogen (Nt) = 0.49%, maximum water
holding capacity (MWC) = 36 g water/100 dry soil, moisture content during the ex-
periments (% MWC) = 39; soil B, pH = 6.7, Ct = 1.25, Nt = 0.10, MWC = 33.6,
% MWC = 41.5; soil C, a brown earth, pH = 5.5, Ct = 3.7, Nt = 0.40, MWC = 51.3,
% MWC = 46; soil D, abrown podzol, pH = 5.4, Ct = 1.26, Nt = O. 11, MWC = 36.2,
% MWC = 30.5; soil E, a chernozem, pH = 7.5, Ct = 2.33, Nt = 0.20, MWC = 54.8,
% MWC = 38.0.
Both sterile and microbiologically active aliquots of sieved (2 mm) soil (2000 g dry
weight) were treated with herbicide by tumbling, in sterile containers, for 30 min, with
enough acetone-sterilized Avadex, adsorbed on sterile talcum (0.5 talcum/100 g dry weight
of soil), to give a final concentration of one ppm of diallate. After treatment, cultures of the
soils were prepared which consisted of40-g aliquots held in sterile, cotton-plugged 125 ml

To hold the moisture content of the soit cultures constant during the 20-week incubation
period of the experiment, sterile water was added as needed. Batchds of sterile soil were
prepared by autoclaving 2000-g samples (dry weight), in 2000-ml flasks, three times for
two hr over a period of four to six days. All operations with these soils were carried out in a
sterilizable room. Subsamples (0.5 g) of cultures were taken for sterility tests imme-
diately prior to extraction.

Soils were extracted as described by Smith (1969), and organic solvent extracts were
purified and dried for analysis by shaking with O. 15 g of active charcoal and one g of
sodium sulfate/20 ml of extract. Pure cultures of fungi in liquid nutrient medium were
prepared, treated with five ppm of Avadex (two ppm of diallate), and extracted as described
by Anderson et al. (1970), except that benzene was used in place of hexane. All cultures
were incubated in the dark at 22~

Benzene extracts of soils or of nutrient medium cultures were analyzed for diallate
by gas liquid chromatography on a Hewlett Packard Model 5700 gas chromatograph
which was equipped with an electron capture detector. A 6-foot by l/4-inch Pyrex glass
column, filled with 5% SE 30 on acid washed 80/100 Chromosorb W DMCS, was used
for analyses. Column, injector port, and detector temperatures were 19&C, 220~ and
198~ respectively. The argon/methane (9/1) and helium gas flow rates were each
adjusted to 40 ml/minute. Range, attenuation and pulse interval controls were set at 10,
32, and 150, respectively.
Microbial Degradation of Diallate

Soils treated with Avadex to which 14C-diallate had been added were incubated on a
WSsthoff "Ultragas 3" CO., analyzer which was modified to allow simultaneous analysis
of CO._, and t4CO., (Domsch et al. 1973). During incubation, gases evolving from the soils
were first passed through toluene and paraffin oil traps, which removed volatilized
diallate, and then passed through methanol/ethanolamine solution (7/3) which trapped
x4CO.,. The radioactivity in these solvents was determined by liquid scintillation

Results and discussion

Disappearance of diallate from five agricultural soils. The results of these experi-
ments (Table I) indicate that diallate disappeared most rapidly from soils containing living
microbial populations. In microbiologically active samples of soil D, ca. 50% of the
initially applied diallate had disappeared after 1.5 weeks of incubation. In soils A and E, and
in soils B and C, 50% of the applied dosage had disappeared after 2.5 and 4 weeks,
respectively. Disappearance of diallate from samples of the soils which were sterile was
relatively slow. After 20 weeks of incubation, between 54 and 69% if the applied herbicide
could be recovered. Correlations between pH, carbon content or nitrogen content of the
soils and the persistence of diallate could not be detected.

It was concluded from these studies that the major cause of diallate disappearance from
microbiologically active soil is its degradation by the soil microflora. As is shown in the
following experiments, the major losses of diallate from the sterile samples can be attributed
to both non-biochemical degradation and incomplete extraction of residues from the soil.

Confirmation of diallate degradation in soil. To confirm that diallate was degraded in

soil, and not merely lost by volatilization, adsorption, or retention by microflora, experi-

Table I. Disappearance of diallate from microbiologically active (Act.) and sterile (Ster.)
samples of five agricultural soils. Results, which are averages from triplicate experiments,
show diallate recoveries in per cent of those obtained immediately after soil treatment.


Weeks of
incubation Act. Ster. Act. Ster. Act. Ster. Act. Ster. Act. Ster.

0 100 100 100 100 100 100 100 I00 100 100
l 66 I00 72 94 70 85 63 90 66 94
2 62 98 57 87 68 -- 41 88 53 93
3 43 85 53 81 52 88 32 86 42 91
4 38 82 46 83 54 83 16 80 39 95
6 28 66 30 72 43 83 13 82 31 96
8 17 62 25 75 36 72 -- -- 22 90
10 18 62 28 81 31 74 9 53 14 88
20 6 54 6 59 12 69 1 57 3 62
J. P. E. Anderson and K. H. Domsch

ments were conducted in which 14C-labeled diallate (one ppm) was added to soil and the
evolution of radio-active carbon dioxide was followed. Soil D was used in these experi-
ments. Soil samples were prepared and treated with the herbicide as described above.

The results from this experiment are given in Figure I and in Table II. Figure 1 shows
that radioactive diallate was degraded to radioactive carbon dioxide in both mic-
robiologically active and in sterile soil. The greatest quantity of diallate was degraded
in the microbiologically active soil. During the first two days of incubation, over 25%
of the applied radioactivity was evolved as carbon dioxide. In contrast to this, evolution
of labeled carbon dioxide from sterile soil samples totaled only 3% of the applied
dosage. Only trace quantities of diallate were volatilized from the soils during the seven
day course o f this experiment. A series of solvent (toluene and paraffin oil) traps built
into the system to retain diallate escaping from the soils in the vapor phase contained
only trace quantities of either diallate or radioactivity.

The distribution of radioactivity in the benzene and aqueous extracts of the soils, and in
the soils after they had been extracted, are shown in Table II. Included in the table is the

20 I

o Active soil
Sterile soi I
o 15

.~ 10


0 2 4 6 8
Days of incubation
Fig. 1. Evolution of 14CO~ from microbio|ogically active and sterile soils treated with one ppm of
l~carbonyl-labeled diallate.
Microbial Degradation of Diallate

percent of the radioactivity recovered as labeled carbon dioxide. The benzene extract of the
samples incubated for zero days contained 100% of the initially recoverable radioactivity.
No radioactivity was recovered in the aqueous extracts, and only trace quantities of activity
were retained as unextractable residues in the soils. After seven days o f incubation,
microbiologically active soils contained only 56% of the initial radioactivity. Forty six
percent was recovered in the benzene extract of the soil, and 10% was detectable as an
unextractable residue retained by the soil. Recovery of radioactivity in the extracts of the
sterile soil was 92% of the applied dosage: 85% was recovered in the benzene extract, and
7% was retained in the solvent extracted soil. Water-soluble radioactive compounds were

Table I I . Distribution of radioactivity in soil extracts and in the gas evolved

from soil after its treatment with one ppm 14C-diallate. Results are in % of the
radioactivity recovered a

Days of Soil
Incubation Material analyzed Microb. active Sterile

Extract of soil
Benzene 100 100
Aqueous 0 0
Soil Residue b Tr. e Tr.
Gas Evolved ('4CO~) 0 0

Extract of soil
Benzene 45.9 84.8
Aqueous 0 0
Soil residue 10.4 7.5
Gas evolved (14CO~.) 43.7 7.7

aTotal radioactivity recovered was 95-100% of applied dosage.

bSoil residue = soil after extraction with organic solvents.
eTr. = trace quantity.

Table I I I . Comparison of diallate recoveries from pure cultures of Trichoderma

harzianum grown in liquid nutrient medium and in sterile soil a

Recovery (% of applied dosage)

Days of Liquid medium Soil
Incubation Live Autoclaved Live Autoclaved
0 100- 13 100+__4 100---3 100--.3
5 46-4-2 89---6 84---2 89-+2
10 34---0 65+2 79--.3 90---6

aTriplicate experiments at two ppm applied dosage.

J. P. E. Anderson and K. H. Domsch

not detected in these studies. The only degradation product of diallate detected in this
experiment was radioactive carbon dioxide. The results presented here show that diallateis
broken down in soil, and support the contention that soil microorganisms are the major
agents responsible for degradation.

Preliminary investigations on diallate degradation by pure cultures of soil fungi.

Preliminary investigations on diallate degradation by pure cultures of microorganisms were
begun with soil fungi. Fungi were isolated as described by Gams and Domsch (1967).
Those species which appeared in high frequency and were present in all five soils were used
in the investigations.

Of the fungi thus far tested, only Phoma eupyrena, Penicillium janthinellum and
Trichoderma harzianum have shown the capacity to degrade substantial quantities ( > 20%
of the applied two ppm) of diallate. Of these fungi, only T. harzianum appeared to be able to
use the compound as a supplementary source of carbon for growth.

Diallate degradation by T. harzianum in liquid nutrient medium was rapid, with ca. 50%
more of the herbicide disappearing from actively metabolizing cultures of the fungus than
from control cultures containing autoclaved mycelia (Table III). Degradation by pure
cultures also appeared to occur in soil, but the rate of degradation was much slower than
in liquid cultures. The reasons for differences in degradation rates under these two
conditions of culture are under further investigation.

The data presented in this report support the results of B anting (1967), Kaufman (1967)
and Smith (1970), who found that diallate rapidly disappeared from microbiologically
active soils, and who suggested that disappearance of the compound is mainly due to its
degradation by the soil microflora.

In previous studies, either bioassay tests (Banting 1967, Kaufman 1967) or chemical
analyses for diallate residues (Smith 1970) were used to follow the disappearance of the
compound from soil. In the present study, degradation of 14C-diallate was followed by
chemical analyses and was confirmed by trapping a volatile metabolite, radioactive carbon

Degradation of diallate by isolated soil microorganisms has been recently reported by

Kaufman and Blake (1973), who found that pure cultures of the fungus Fusarium oxy-
sporum dehalogenated the compound within a 20-day incubation period. Other metabolites
were not detected in those studies. A recent publication by Cullimore and Smith (1972) has
indicated that the closely related comp.ound, triaUate (S-2,3,3-trichloroallyl N,
N-diisopropylthiolcarbamate) can be taken up by some species of fungi. Degradation of
triallate was not reported. The preliminary results reported in the present study indicate that
at least three fungi can degrade diallate. Of these organisms, one can possibly use the
compound as a supplementary source of carbon. The results of these studies, however, have
not yet been confirmed by isolation of diallate metabolites.
Microbial Degradation of Diallate 7


This work was sup.ported by the Deutsche Forschungsgemeinschaft. The authors wish to
thank Mrs. Ingrid Langelfiddecke and Mr. Kurt Steffens for their excellent technical


Anderson, J. P. E., E. P. Lichtenstein, and W. F. Whittingham: Effect ofMucor alternans

on the persistence of DDT and dieldrin in culture and in soil. J. Econ. Entomol. 63,
1595 (1970).
Banting, J. D.: Factors affecting the activity of di-allate and tri-allate. Weed Res. 7,302
Cullimore, D. R., and A. E. Smith: Initial studies on the microbial breakdown of triallate.
Bull. Environ. Contam. Toxicol. 7, 36 (1972).
Domsch, K. H., J. P. E. Anderson, and R. Ahlers: Method for simultaneous measurement
of radioactive and inactive CO~ evolved from soil samples during incubation with
labeled substrates. Appl. Microbiol. 25, 819 (1973).
Gams, W., and K. H. Domsch: Beitr~ige zur Anwendung der Bodenwaschtechnik ffir die
Isolierung von Bodenpilzen. Archiv. Mikrobiol. 58, 134 (1967).
Kaufman, D. D. : Degradation of carbamate herbicides in soil. J. Agr. Food Chem. 15,
582 (1967).
Kaufman, D. D., and J. Blake: Microbial degradation of several acetamide, acylanilide,
carbamate, toluidine and urea pesticides. Soil Biol. Biochem. 5, 297 (1973).
Smith, A. E. : Estimation ofS-2,3-dichloroallyl N,N-diisopropylthiolcarbamate (dial-
late) residues in soils by electron-capture gas chromatography. J. Agr. Food
Chem. 17, 1052 (1969).
Smith, A. E.: Degradation, adsorption and volatility of di-allate and tri-allate in prairie
soils. Weed Res. 10, 331 (1970).

Manuscript received September 14, 1974; accepted October 31, 1974.