Mutation Research 412 Ž1998.

245–250

In situ assessment of pesticide genotoxicity in an integrated pest

management program:
II. Maize waxy mutation assay
a,)
Geraldo Stachetti Rodrigues , David Pimentel b, Leonard H. Weinstein b,c

a
˜ de Impacto Ambiental, Caixa Postal 069, Jaguariuna,
EMBRAPAr CNPMA-Centro Nacional de Pesquisa de Monitoramento e AÕaliaçao ´
˜ Paulo, CEP 13820, Brazil
Sao
b
Cornell UniÕersity, Ithaca, NY 14853, USA
c
Boyce Thompson Institute for Plant Research, Ithaca, NY 14853, USA
Received 21 May 1997; revised 27 October 1997; accepted 30 October 1997

Abstract

The mutagenicity induced by pesticides applied in an integrated pest management ŽIPM. program was evaluated in situ
with the maize forward waxy mutation bioassay. Three pesticide application rates were prescribed as follows: Ž1. Low—no
field pesticide spray; Ž2. Medium—IPM test rate: banded cyanazine plus metolachlor Ž2.7 kg a.i. and 2.3 l a.i.rha of
herbicides, respectively.; and Ž3. High—a preventative pesticide application program: broadcast cyanazine plus metolachlor
Žsame application rates as above. plus chlorpyrifos Ž1 kg a.i.rha of insecticide.. In general, there was no significant
reduction in the genotoxic effects from the high to the medium treatment levels of the IPM program. This suggests that the
reduction in pesticide application rates attained with the implementation of the proposed IPM program was not sufficient to
abate the genotoxicity of the pesticides. The results indicate that replacing genotoxic compounds may be the only effective
remediation measure if concern about environmental mutagenesis were to result in changes in agricultural management.
q 1998 Elsevier Science B.V.

Keywords: Maize waxy pollen; Bioassay; In situ monitoring; Pesticide

1. Introduction Rodrigues et al. w1x. It was shown that it is possible

Studies on the genotoxicity of the pesticides ap-
plied to an integrated pest management ŽIPM. pro-
gram for corn and soybean utilizing the Tradescantia
micronucleus ŽTrad-MCN. assay were described by
)
Corresponding author. Tel.: q55-19-867-8735; fax: q55-19-
867-8740; E-mail: stacheti@cnpma.embrapa.br
to detect genotoxicity by exposing the plants in situ
to recently-sprayed fields, to extracts of soil samples
collected shortly after spraying, and to solutions of
the commercial pesticides.
However, due to the extreme sensitivity of the
Trad-MCN assay, a question remains regarding
whether less sensitive subjects, or more specifically,
the crop plants actually being exposed to those pesti-
cides on a regular basis, could respond in a compara-

1383-5718r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.

PII S 1 3 8 3 - 5 7 1 8 Ž 9 7 . 0 0 1 9 4 - 0
246 G.S. Rodrigues et al.r Mutation Research 412 (1998) 245–250

ble fashion under the same conditions. This informa-

water disp. granulesr2.7 kgrha; emul. conc.r2.34 lrha

water disp. granulesr2.7 kgrha; emul. conc.r2.34 lrha
tion would be valuable for evaluating the possibility
that an increased mutation rate could occur in pesti-
cide-treated fields, and whether the reduction in pes-
ticide application rate attained with an IPM program
could effectively lessen such genotoxicity.
In order to address these questions, the maize

carboxinqdiazinonrlindane 14q15q25%rapprox. 1.4 grkg

Ž Zea mays L.. waxy forward mutation assay in the

37q25%rapprox. 1.5 grkg

microgametophytes w2x was chosen for the in situ
assessment of mutagenicity in the IPM demonstra-

granularr1.12 kgrha
tion field. This assay was chosen as a surrogate for

FormulationrA.i.
one of the crops being tested in the IPM program.

application rate
Pesticide input levels for the plots assayed for mutagenicity with the bioassays, and details of the commercial formulations applied
2. Materials and methods

Details of the experimental conditions in the field,

the pesticides applied and the treatment levels, and

Bladex w 90DFrShelleqDual w rCibae cyanazineqmetolachlor

Bladex w 90DFrShelleqDual w rCibae cyanazineqmetolachlor
the design of the IPM program were given in w1x.
The pesticide treatments were applied as presented in

captanqdiazinon
Table 1.

chlorpyrifos
ingredient
Remarks
2.1. Plant stock, propagation and maintenance

Active

captan
Maize Ž Z. mays L.. seeds of the Early–Early
Synthetic variety were obtained from the Maize Ge-
netics Stock Center, University of Illinois at Ur-
bana–Champaign, from Mr. E. B. Patterson. The

Lorsban w 15GrDowElancoe
lineage used in the experiments was homozygous
Wx, consequently amenable for the forward mutation Agrox DL-PlusrAgwaye
Agrox 2-wayrAgwaye

test. The seeds were germinated and the plants prop-

agated in a greenhouse Žsee w1x for settings., and the
Low Medium High namerProduser

harvested seeds were air-dried and kept in paper

bags in a well-ventilated facility.
Pesticide input level Trade

2.2. Exposure, sampling, and analysis

q

q
q

q
q

2.2.1. Maize waxy pollen grain forward mutation

assay
Early–Early Synthetic maize seeds were sown
1=
q
q

amid the commercial corn seeds planted for the IPM

2=

program experiments on the second day after pesti-

q

cide treatment. Each seed was marked with a small

Broadcast herbicide

flag, and upon germination each seedling was tagged

Banded herbicide
Soil insecticide
Seed treatment

Grower 2-way
Grower 3-way

in order to permit differentiation from the commer-

Weed control
Management

Commercial

Cultivation

cial corn plants. There were 50 seeds in each of the

Table 1

three pesticide treatment regimes in two blocks, or a

total of 300 seeds for three replicated treatments.
 G.S. Rodrigues et al.r Mutation Research 412 (1998) 245–250
Laboratory controls were established in addition
to the in situ control for the low pesticide level.
Maize seeds were soaked in deionized water Žnega-
tive control. or in a 100-ppm EMS aqueous solution
Žpositive control. for 24 h, thoroughly washed in
distilled water, and sown individually in 12.5-cm
recyclable cardboard pots containing a soil–peat
moss–vermiculite mix ŽCornell mix w3x. supple-
mented with Osmocote w Žslow-release. fertilizer and
micronutrient elements.
Upon tasselling, portions of the inflorescences
with unopened florets Žpreanthesis. were collected
and fixed in 70% Žvol.. ethanol. Scoring consisted of
randomly selecting five florets, washing them in
70% ethanol to remove any extraneous pollen grains,
and dissecting them to extract the anthers. The an-
thers were then minced in a petri dish containing
gelatin–iodine stain Ž500 mg potassium iodide q 95
mg iodine in 25 ml water, and 1.5 g of bacteriologi-
cal gelatin in 2.5 ml water, mixed immediately be-
fore use. under a stereomicroscope. Any fragments
were removed and the stained pollen grains were
mounted on a microslide w2x.
The total number of pollen grains on the slide was
estimated by counting Žunder 40 = magnification.
all full Žviable. pollen grains in 20 randomly chosen
1-mm2 grids of the microslide. The microslide was
then carefully scanned for mutant Žtan-colored pollen
grains among the black-colored nonmutated popula-
tion. and aborted pollen grains. Approximately
20,000 pollen grains were examined per tassel Žper
plant., and ten plants per treatment, totaling an aver-
age of 200,000 pollen grains per treatment. The
frequency of mutated pollen grains was calculated
for each plant, and the mean mutation frequency for
the ten plants compared between treatments.
Due to the presence of plants showing no mutated
pollen grains Žan occasional zero among the nor-
mally large numbers resulting from the numerous
pollen grains on a slide., a transformation was per-
formed in the data before statistical analysis in order
to normalize the variance. A transformation indicated
for counts of rare events was chosen from w4x as
follows:
Y s 'X q 'X q 1
where X is the observed and Y is the transformed
mutation frequency. The transformed mutation fre-
247
quencies and the rate of pollen abortion computed
per plant was compared statistically between treat-
ments by analysis of variance ŽANOVA, a s 0.05..
3. Results
A statistically significant Ž a s 0.05. increase in
forward mutation frequency occurred when maize
plants were exposed in situ to the pesticide-sprayed
field. If one considers the sequential pesticide treat-
ments Žfrom low to high, Table 1. as increasing
doses, a dose response-like rise in mutation fre-
quency occurred. The mutation frequency observed
for the negative control Ž2.66 = 10y5 ., as well as for
the low pesticide treatment level Ž6.45 = 10y5 ., were
well within the range of spontaneous mutation fre-
quency commonly described in the literature w5,6x.
The positive control ŽEMS. showed a significantly
higher mutation frequency than the negative control.
The medium pesticide treatment was intermediate
between the low and the high treatments, while these
two latter treatments were statistically different ŽFig.
1..
There were no physiological toxic effects induced
by the pesticides to the development of the microga-
Fig. 1. Means of forward waxy mutation frequencies Ž"SD. in
pollen grains of corn plants grown in pesticide-sprayed soil in situ.
The Control treatment consisted of plants grown in soil–peat
moss–vermiculite mix in a controlled-environment chamber. The
positive control ŽEMS. consisted of seeds soaked for 24 h in a
100-ppm aqueous solution of ethyl methanesulfonate. The Low
pesticide treatment received no pesticide, the Medium received
banded cyanazineqmetolachlor, and the High received broadcast
cyanazineqmetolachlor and chlorpyrifos. The pesticides were
sprayed 24 h before the seeds were planted in the field.
248 G.S. Rodrigues et al.r Mutation Research 412 (1998) 245–250
metophytes, at least as indicated in terms of changes
in pollen grain abortion frequency. The average fre-
quency of pollen abortion did not differ significantly
between treatments and remained below 2%, inferior
to the abortion frequencies normally cited w6,7x.
These results show that even after one single
application at recommended doses at the beginning
of the season, pesticides remain active and may
induce mutations in exposed plants, in this case in
the pollen grains formed last. Additionally, the re-
sponse pattern shown by the maize assay, with no
significant reduction in mutation frequencies from
high to medium pesticide treatment levels, corrob-
orates the results obtained with the highly sensitive
Trad-MCN assay w1x. The similarity between the
results obtained with these two essentially different
assays strongly suggests that the reduction attained
with the IPM program being tested Žfrom high to
medium application rates, Table 1. may be insuffi-
cient to abate the environmental mutagenesis of the
pesticides used.
4. Discussion
The in situ exposure of maize plants to pesticide-
treated soils resulted in detectable mutations. The
maize waxy pollen grain assay has been extensively
used for the in situ assessment of pesticide muta-
genicity, alone or incorporating microbial assays. All
three pesticides applied in the IPM program under
investigation had been previously evaluated for mu-
tagenicity with the maize waxy pollen assay. Cyana-
zine was shown to be mutagenic in the reversion test,
as well as in two microbial assays Ž Salmonella and
Saccharomyces cereÕisiae . after being activated with
maize microsomal homogenates or with extracts of
maize plants exposed to the compound w8x. In the
same series of experiments, metolachlor gave nega-
tive results. The mixture of cyanazineq metolachlor
was positive, but whether such an effect was caused
by the s-triazine alone or by any additive effect of
metolachlor could not be resolved.
In a later study w9x, metolachlor mixed with the
nonmutagenic herbicide dicamba produced negative
results. This time, however, extracts of maize ex-
posed to metolachlor, as well as microsomal ho-
mogenate-activated metolachlor, were mutagenic in
the Salmonella and S. cereÕisiae assays. Chlorpyri-
fos was shown not to induce mutation in the maize
waxy pollen reversion test, nor in the Salmonella or
in the S. cereÕisiae mutation assays w10x.
The findings of the present experiments corrobo-
rate the genotoxic effects detected in a variety of test
systems, including plant bioassays, of the three pesti-
cides under investigation. Chlorpyrifos induced an
increase in micronuclei frequency in erythroblasts of
mouse bone marrow both by intraperitoneal and oral
treatments Ždermal treatment was negative. with
doses below the LD50 , and with no signs of clinical
effects w11x. By contrast, chlorpyrifos did not induce
increases in either sister chromatid exchange ŽSCE.
frequencies in the chick embryo assay, or in Chinese
hamster ovary cells. No chromosomal aberrations
were found in bovine blastocysts, but the reproduc-
tive performance of treated bulls became subnormal
after exposure to chlorpyrifos w12x.
The cytogenetic effects on human lymphocytes of
a mixture of 15 pesticides commonly found in foods
was investigated in Italy w13x. Chlorpyrifos com-
posed 2% of the mixture, which was genotoxic only
when benomyl was present. Chlorpyrifos was muta-
genic in the somatic as well as in the germ cells of
Drosophila, as determined by induction of mosaic
wing spots and sex-linked recessive lethals, at con-
centrations as low as 5 = 10y6 % w14x.
The mutagenicity of chlorpyrifos was evaluated
utilizing the microbial test systems Bacillus subtilis
Žrec-assay., Escherichia coli, and Salmonella ty-
phimurium Žreversion assays. w15x. None of the as-
says was positive. On the other hand, chlorpyrifos
was reported positive in the primary DNA damage
assay with these same organisms and with S. cere-
Õisiae in another study w16x. Other assays employed,
including reversion assay in S. typhimurium and E.
coli, recessive lethality in Drosophila melanogaster,
enhanced mitotic recombination in S. cereÕisiae, and
unscheduled DNA synthesis ŽUDS. in human lung
fibroblasts all produced negative results w16x. Chlor-
pyrifos also failed to induce mutations in S. cere-
Õisiae and S. typhimurium both before and after
activation with plant Ž1S. and animal ŽS9. microso-
mal homogenates, as well as in the Z. mays waxy
locus test w10x.
Significant increases in abnormal mitoses were
induced by saturated and 1r8-strength solutions of
 G.S. Rodrigues et al.r Mutation Research 412 (1998) 245–250
chlorpyrifos in Vicia faba root-tip meristematic cells
both after seed soaking and direct root treatment
w17x. Likewise, both mitotic and meiotic genotoxicity
were caused by chlorpyrifos in Hordeum Õulgare L.
root tip cells and pollen mother cells after seed
soaking and spray treatments w18,19x. Positive
dose–response relationships were obtained in both
tests, with chromosome stickiness being the main
effect. A review of the genetic toxicology of pesti-
cides w20x listed several chromosome effects caused
by chlorpyrifos, including fragmentation, bridges,
vagrancy, disturbed anaphase, and chlorophyll mu-
tant induction in Gossypium barbadence, V. faba,
and H. Õulgare. The genetic hazard potential of
chlorpyrifos was considered uncertain for develop-
mental and reproductive hazards, and positive for
hereditary genetic hazard.
Cyanazine has been shown to induce chromosome
damage, represented by increased break frequency,
in human lymphocytes in vitro at concentrations as
low as 1 ppm w21x. This result, however, is in
disagreement with other data reported for cyanazine
genotoxicity as tested in the UDS and SCE assays in
human peripheral lymphocytes Žconcentrations vary-
ing from 12 to 100 ppm. and a chromosome aberra-
tion analysis in rat bone marrow cells Ždoses varying
from 56 to 224 mgrkg.. Cyanazine was ineffective
in all cases w22x.
The rate of apparent dominant lethals in D.
melanogaster was shown to increase when flies
were fed a ration containing 0.01% cyanazine. The
absence of effects in chromosomal aberration tests,
however, led the authors to conclude that the de-
crease in egg hatchability was more likely due to
physiologic toxicity to sperm w23,24x.
Metolachlor induced chromosome damage in hu-
man lymphocytes in vitro at concentrations as low as
0.1 ppm w21x. The ability of animal microsomal
homogenates to activate metolachlor into mutagenic
metabolites was shown in S. typhimurium and in S.
cereÕisiae w9x. Metolachlor was also mutagenic in S.
typhimurium directly Žwithout activation..
The biological impact of these pesticides in the
environment is a function of the concentration at
which they are found, their availability, and their
intrinsic biological activity. The three pesticides un-
der investigation must be considered potentially
genotoxic, since all of them have shown positive
249
results in test systems ranging from prokaryotes to in
vitro mammal assays, and two of them Žcyanazine
and metolachlor. were positive in in vitro human
assays w21x.
In addition to the potential genotoxic impacts
these compounds may have at the level of the agri-
cultural field assessed here, their distribution in the
environment may pose additional threats. Both
cyanazine and metolachlor were among the most
frequently detected pesticides in river water and in
paired drinking water samples in a 6-yr study con-
ducted in Canada w25x. These compounds were pre-
sent in almost 30% of the samples analyzed in both
river and drinking water at concentrations of up to
10 m grl for cyanazine and up to 28 m grl for
metolachlor. The maximum acceptable Žinterim. lev-
els of these herbicides in Canada are 10 and 50 m grl
for cyanazine and metolachlor, respectively, or equal
to the maximum level detected for cyanazine, and
only less than twice the levels found for metolachlor
w25x. These two herbicides are also among the most
common contaminants in rainwater w26x. Cyanazine
appeared in 25% of 325 rainwater samples collected
in Iowa, at concentrations as high as 28 m grl.
Metolachlor occurred in 20.7% of the samples, with
a maximum concentration of 2.70 m grl w27x. Cyana-
zine was also reported as a common contaminant in
groundwater w28x. Although lower than the concen-
trations normally referred to in evaluations of muta-
genicity in laboratory conditions, the widespread
presence of these compounds in the environment
may have adverse consequences.
The results of the present experiments corroborate
the adequacy of this assay for the in situ assessment
of pesticide mutagenicity. The short generation time
Žca. 40 d from seed to tasselling. of the Early–Early
synthetic variety allows for relatively expeditious
experimentation and differentiation between the ex-
perimental subjects and commercial varieties when
they are mixed in the field, as in the present study.
A clear and significant increment in mutation
frequency was observed for the high-pesticide treat-
ment level in relation to the nonsprayed Žlow-pesti-
cide treatment level. soil. The pesticide application
rate recommended in the IPM program under investi-
gation Žmedium treatment level. showed a mutation
frequency intermediate between the high and low
pesticide treatment levels. The decrease in relation to
250 G.S. Rodrigues et al.r Mutation Research 412 (1998) 245–250
the high treatment, however, was not statistically
significant, which agrees with the results described
for the Trad-MCN assay w1x. In other words, the
results yielded in situ by the maize waxy pollen
assay suggest that the reduction in pesticide dosages
attained with the implementation of the IPM pro-
gram proposed may not be sufficient to prevent
mutagenicity of the pesticides.
The conclusion that a partial reduction in pesti-
cide application rate may be ineffective in abating
mutagenicity indicates that a different strategy is
warranted if concern about environmental mutagene-
sis were to result in changes in agricultural manage-
ment. Such a strategy should rely on extensive test-
ing and selection for field use of alternative, nonmu-
tagenic compounds.
References
w1x G.S. Rodrigues, D. Pimentel, L.H. Weinstein, In situ assess-
ment of pesticide mutagenicity in an integrated pest manage-
ment program I— Tradescantia micronucleus assay, Muta-
tion Res., submitted.
w2x M.J. Plewa, Specific-locus mutation assays in Zea mays. A
report of the U.S. Environmental Protection Agency Gene-
Tox Program, Mutat. Res. 99 Ž1982. 317–337.
w3x J.W. Boodley, R. Sheldrake, Jr., Cornell Peat–lite mixes for
commercial plant growing, Brochure, Cornell Cooperative
Extension, Ithaca, NY, 1982.
w4x G.W. Snedecor, W.G. Cochran, Statistical Methods, Iowa
State Univ. Press, Ames, IA, USA, 1967, 593 pp.
w5x M.J. Plewa, J.M. Gentile, Plant activation of herbicides into
environmental mutagens: the waxy reversion bioassay, Maize
Genet. Coop. Newslett. 50 Ž1976. 44.
w6x P.K. Hopke, M.J. Plewa, J.B. Johnston, D. Weaver, S.G.
Wood, R.A. Larson, T. Hinesly, Multitechnique screening of
Chicago municipal sewage sludge for mutagenic activity,
Environ. Sci. Technol. 16 Ž1982. 140–147.
w7x M.W. Heartlein, D.M. DeMarini, A.J. Katz, J.C. Means, M.J.
Plewa, H.E. Brokman, Mutagenicity of municipal water ob-
tained from an agricultural area, Environ. Mutagen. 3 Ž1981.
519–530.
w8x M.J. Plewa, Activation of chemicals into mutagens by green
plants: a preliminary discussion, Environ. Health Perspect. 27
Ž1978. 45–50.
w9x M.J. Plewa, E.D. Wagner, G.J. Gentile, J.M. Gentile, An
evaluation of the genotoxic properties of herbicides follow-
ing plant and animal activation, Mutation Res. 136 Ž1984.
233–245.
w10x J.M. Gentile, G.J. Gentile, J. Bultman, R. Sechriest, E.D.
Wagner, M.J. Plewa, An evaluation of the genotoxic proper-
ties of insecticides following plant and animal activation,
Mutation Res. 101 Ž1982. 19–29.
w11x S.M. Amer, M.A. Fahmy, Cytogenetic effects of pesticides:
I. Induction of micronuclei in mouse bone marrow by the
insecticide Dursban, Mutat. Res. 101 Ž1982. 247–255.
w12x D.E. Muscarella, J.F. Keown, S.E. Bloom, Evaluation of the
genotoxic and embryotoxic potential of chlorpyrifos and its
metabolites in vivo and in vitro, Environ. Mutagen. 6 Ž1984.
13–23.
w13x P. Dolara, F. Torricelli, N. Antonelli, Cytogenetic effects on
human lymphocytes of a mixture of fifteen pesticides com-
monly used in Italy, Mutation Res. 325 Ž1994. 47–51.
w14x K.K. Patnaik, N.K. Tripathy, Farm-grade chlorpyrifos Durmet
is genotoxic in somatic and germ-line cells of Drosophila,
Mutat. Res. 279 Ž1992. 15–20.
w15x Y. Shirasu, M. Moriya, K. Kato, A. Furuhashi, T. Kada,
Mutagenicity screening of pesticides in the microbial system,
Mutation Res. 40 Ž1976. 19–30.
w16x M. Waters, S.S. Sandhu, V.F. Simmon, K.E. Mortelmans,
A.D. Mitchell, T.A. Jorgenson, D.C.L. Jones, R. Valencia,
N.E. Garrett, Study of pesticide genotoxicity, in: R.A. Fleck,
A. Hollaender ŽEds.., Genetic Toxicology: An Agricultural
Perspective, Plenum, New York, 1982, pp. 275–326.
w17x S.M. Amer, O.R. Farah, Cytological effects of pesticides:
XII. Effects of the phosphorothioate insecticide dursban on
the mitosis of Vicia faba, Cytologia 48 Ž1983. 27–33.
w18x P. Kaur, I.S. Grover, Cytological effects of some
organophosphorus pesticides: I. Mitotic effects, Cytologia 50
Ž1985. 187–197.
w19x P. Kaur, I.S. Grover, Cytological effects of some
organophosphorus pesticides: II. Meiotic effects, Cytologia
50 Ž1985. 199–211.
w20x C.B.S.R. Sharma, N. Panneerselvan, Genetic toxicology of
pesticides in higher plant systems, Crit. Rev. Plant Sci. 9
Ž1990. 409–442.
w21x B. Roloff, D. Belluck, L. Meisner, Cytogenetic effects of
cyanazine and metolachlor on human lymphocytes exposed
in vitro, Mutation Res. 281 Ž1992. 295–298.
w22x P. Hrelia, F. Vigagni, F. Maffei, M. Morotti, A. Colacci, P.
Perocco, S. Grilli, G. Cantelli-Forti, Genetic safety evalua-
tion of pesticides in different short-term tests, Mutation Res.
321 Ž1994. 219–228.
w23x M.R. Murnik, Mutagenicity of widely used herbicides, Ge-
netics 83 Ž1976. s54.
w24x M.R. Murnik, C.L. Nash, Mutagenicity of the triazine herbi-
cides atrazine, cyanazine, and simazine in Drosophila
melanogaster, J. Toxicol. Environ. Health 3 Ž1977. 691–697.
w25x R. Frank, B.S. Clegg, C. Sherman, N.D. Chapman, Triazine
and chloroacetamide herbicides in Sydenham river water and
municipal drinking water, Dresden, Ontario, Canada, 1981–
1987, Arch. Environ. Contam. Toxicol. 19 Ž1990. 319–324.
w26x R.P. Richards, J.W. Kramer, D.B. Baker, K.A. Krieger,
Pesticides in rainwater in the northeastern United States,
Nature 327 Ž1987. 129–131.
w27x B.K. Nations, G.R. Hallberg, Pesticides in Iowa precipita-
tion, J. Environ. Qual. 21 Ž1992. 456–492.
w28x A.R. Isensee, C.S. Helling, T.J. Gish, P.C. Kearney, C.B.
Coffman, W. Zhuang, Groundwater residues of atrazine
alachlor and cyanazine under no-tillage practices, Chemo-
sphere 17 Ž1988. 165–174.