The Science of the Total Environment 248 Ž2000. 115]122
Finding minimal herbicide concentrations in ground
water? Try looking for their degradates
D.W. Kolpina,U , E.M. Thurmanb, S.M. Linhart a
a
US Geological Sur¨ ey, 400 S. Clinton St., Box 1230, Iowa City, IA 52244, USA
b
US Geological Sur¨ ey, 4821 Quail Crest Place, Lawrence, KS 66049, USA
Abstract
Extensive research has been conducted regarding the occurrence of herbicides in the hydrologic system, their fate,
and their effects on human health and the environment. Few studies, however, have considered herbicide transformation products Ždegradates.. In this study of Iowa ground water, herbicide degradates were frequently detected. In
fact, herbicide degradates were eight of the 10 most frequently detected compounds. Furthermore, a majority of a
herbicide’s measured concentration was in the form of its degradates } ranging from 55 to over 99%. The herbicide
detection frequencies and concentrations varied significantly among the major aquifer types sampled. These
differences, however, were much more pronounced when herbicide degradates were included. Aquifer types
presumed to have the most rapid recharge rates Žalluvial and bedrockrkarst region aquifers. were those most likely
to contain detectable concentrations of herbicide compounds. Two indirect estimates of ground-water age Ždepth of
well completion and dissolved-oxygen concentration . were used to separate the sampled wells into general vulnerability classes Žlow, intermediate, and high.. The results show that the herbicide detection frequencies and concentrations varied significantly among the vulnerability classes regardless of whether or not herbicide degradates were
considered. Nevertheless, when herbicide degradates were included, the frequency of herbicide compound detection
within the highest vulnerability class approached 90%, and the median total herbicide residue concentration
increased over an order of magnitude, relative to the parent compounds alone, to 2 mgrl. The results from this study
demonstrate that obtaining data on herbicide degradates is critical for understanding the fate of herbicides in the
hydrologic system. Furthermore, the prevalence of herbicide degradates documented in this study suggests that to
accurately determine the overall effect on human health and the environment of a specific herbicide its degradates
should also be considered. Q 2000 Elsevier Science B.V. All rights reserved.
Keywords: Herbicides; Degradation products; Ground water
U
Corresponding author. Tel.: q1-319-358-3614; fax: q1-319-358-3606.
E-mail address: [email protected] ŽD.W. Kolpin.
0048-9697r00r$ - see front matter Q 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 8 - 9 6 9 7 Ž 9 9 . 0 0 5 3 5 - 5
116
D.W. Kolpin et al. r The Science of the Total En¨ ironment 248 (2000) 115]122
1. Introduction
Extensive research has been conducted regarding the occurrence of herbicides in the hydrologic
system ŽHolden et al., 1992; Walls et al., 1996.,
their fate ŽBintein and Devillers, 1996; Kruger et
al., 1997., and their effects on human health
ŽCarbonell et al., 1995; Bain and LeBlanc, 1996;
Ribas et al., 1997., and the environment ŽLongley
and Sotherton, 1997; Pratt et al., 1997; Carder
and Hoagland, 1998; Howe et al., 1998.. However,
current understanding of the biological impacts of
herbicide use is limited by the fact that most
investigations have focused on the active ingredients Žparent compounds. without considering their
transformation products Ždegradates..
Complete mineralization of most parent herbicide compounds rarely occurs in the environment
ŽStamper et al., 1997.. However, relatively stable
and persistent degradates can be formed during
the transformation of many herbicides ŽCoats,
1993.. These degradates can, in some cases, be
Table 1
Herbicides and herbicide degradation products analyzed
Common name
Chemical name
Use or origin
Method of
analysis
Acetochlor
Acetochlor ethanesulfonic
acid Žacetochlor ESA.
Acetochlor oxanilic acid
Žacetochlor OA.
Alachlor
Alachlor ethanesulfonic
acid Žalachlor ESA.
Alachlor oxanilic acid
Žalachlor OA.
Ametryn
Atrazine
Deethylatrazine ŽDEA.
2-Chloro-N-Žethoxymethyl.-N-Ž2-ethyl-6-methylphenyl .acetamide
2-wŽ2-Ethyl-6-methylphenyl .Žethoxymethyl.aminox-2-oxoethanesulfonic acid
2-wŽ2-Ethyl-6-methylphenyl .Žethoxymethyl.aminox-2-oxoacetic
acid
2-Chloro-29-69-diethyl-N-Žmethoxymethyl.-acetanilide
2-wŽ2,6-Diethylphenyl.Žmethoxymethyl.aminox-2-oxoethanesulfonic acid
2-wŽ2,6-Diethylphenyl.Žmethoxymethyl.aminox-2-oxoacetic acid
Herbicide
Herbicide degradate
Žacetochlor.
Herbicide degradate
Žacetochlor.
Herbicide
Herbicide degradate
Žalachlor.
Herbicide degradate
Žalachlor.
Herbicide
Herbicide
Herbicide degradate
Žatrazine, propazine.
Herbicide degradate
Žatrazine, cyanazine,
simazine.
Herbicide degradate
Žatrazine.
GCrMS
HPLC
2-ŽEthylamino.-4-isopropylamino-6-methyl-thio-S-triazine
2-Chloro-4-ethylamino-6-isopropylamino-S-triazine
2-Amino-4-chloro-6-Žisopropylamino.-S-triazine
HPLC
GCrMS
HPLC
HPLC
GCrMS
GCrMS
GCrMS
Deisopropylatrazine
ŽDIA.
2-Amino-4-chloro-6-Žethylamino.-S-triazine
Hydroxyatrazine ŽHA.
2-Hydroxy-4-Žethylamino.-6-Žisopropylamino.-S-triazine
Cyanazine
2-ww4-Chloro-6-Žethylamino.-1,3,5-triazin-2-ylxaminox-2-methylpropionitrile
2-Chloro-4-Ž1-carbamoyl-1-methyl-ethylamino.-6-ethylamino-Striazine
2-Chloro-N-Ž2-ethyl-6-methylphenyl .-N-Ž2-methoxy-1-methylethyl.acetamide
2-wŽ2-Ethyl-6-methylphenyl .Ž2-methoxy-1-methylethyl.aminox-2oxoethanesulfonic acid
Herbicide
GCrMS
Herbicide degradate
Žcyanazine.
Herbicide
GCrMS
Herbicide degradate
Žmetolachlor.
HPLC
2-wŽ2-Ethyl-6-methylphenyl.Ž2-methoxy-1-methylethyl.aminox-2oxoacetic acid
4-Amino-6-Ž1,1-dimethylethyl.-3-Žmethylthio.-1,2,4-triazin5Ž4H.-one
2,4-BisŽisopropylamino.-6-methyoxy-S-triazine
2,4-BisŽisopropylamino.-6-Žmethylthio.-S-triazine
2-Chloro-N-isopropylacetanilide
2-Chloro-4,6-bisŽisopropylamino.-S-triazine
2-Chloro-4,6-bisŽethylamino.-S-triazine
2-Tert-butylamino-4-ethylamino-6-methylthio-S-triazine
Herbicide degradate
Žmetolachlor.
Herbicide
HPLC
GCrMS
Herbicide
Herbicide
Herbicide
Herbicide
Herbicide
Herbicide
GCrMS
GCrMS
GCrMS
GCrMS
GCrMS
GCrMS
Cyanazine amide
Metolachlor
Metolachlor ethanesulfonic acid
Žmetolachlor ESA.
Metolachlor oxanilic acid
Žmetolachlor OA.
Metribuzin
Prometon
Prometryn
Propachlor
Propazine
Simazine
Terbutryn
GCrMS
HPLC
GCrMS
D.W. Kolpin et al. r The Science of the Total En¨ ironment 248 (2000) 115]122
117
Fig. 1. Location of wells sampled in Iowa during 1995]1998.
more toxic than their parent compounds ŽBelfroid
et al., 1998; La Clair et al., 1998..
This paper describes the results from a study to
compare the occurrence of selected herbicide
degradates in ground water with that of their
parent compounds. Results are presented from
the sampling of 131 municipal wells during
1995]1998 ŽFig. 1. covering all the major aquifer
types in Iowa. The degradates analyzed ŽTable 1.
include acetochlor ethanesulfonic acid Žacetochlor ESA., acetochlor oxanilic acid Žacetochlor
OA., alachlor ethanesulfonic acid Žalachlor ESA.,
alachlor oxanilic acid Žalachlor OA., deethylatrazine ŽDEA., deisopropylatrazine ŽDIA., cyanazine
amide, hydroxyatrazine, metolachlor ethanesulfonic acid Žmetolachlor ESA., and metolachlor
oxanilic acid Žmetolachlor OA.. This research is
an extension of the multi-agency Iowa Ground
Water Monitoring Program ŽIGWM. ŽDetroy,
1985; Kolpin et al., 1997b..
2. Materials and methods
Wells screened in each major aquifer type Žalluvial, glacial drift, bedrockrkarst region, and bedrockrnon-karst region. were randomly selected
from total population of over 2000 Iowa municipal wells. The 328 water samples collected from
the final set of 131 wells ŽFig. 1. represent the
annual sampling carried out for the IGWM from
1995 to 1998. The sampling protocol for this study
Fig. 2. Frequency of detection for selected herbicide compounds Žadjusted to a common detection threshold of 0.20
mgrl.. See Table 1 for compound abbreviation definitions.
has been reported previously ŽKolpin et al., 1997a,
1998.. All wells were pumped for at least 30 min
before dissolved oxygen, pH, specific conductance, and water temperature were measured.
Once the values for these parameters stabilized,
the water samples were collected. Water samples
were filtered through a 0.7-mm glass-fiber filter
into amber baked-glass bottles and immediately
chilled.
All water samples were sent to the US Geological Survey, Organic Research Laboratory in
Lawrence, KS, USA to determine concentrations
of 13 herbicides and 10 herbicide degradates ŽTable 1.. The 13 parent compounds and three of the
triazine degradates ŽTable 1. were analyzed from
125-ml water samples by gas chromatographyr
mass spectrometry ŽGCrMS. following solidphase extraction onto C 18 cartridges ŽThurman et
al., 1990; Meyer et al., 1993.. The analytical reporting limit for this method was 0.05 mgrl for all
target compounds. The other seven herbicide
degradates ŽTable 1. were analyzed from 125-ml
water samples by high-performance liquid chromatography ŽHPLC. with diode-array detection
D.W. Kolpin et al. r The Science of the Total En¨ ironment 248 (2000) 115]122
118
following solid-phase extraction on C 18 cartridges
ŽHostetler and Thurman, 2000.. The analytical
reporting limit for this method was 0.2 mgrl for
all target compounds. Confirmation was achieved
for alachlor OA, acetochlor OA, metolachlor OA,
and metolachlor ESA by negative ion spray mass
spectrometry ŽFerrer et al., 1997..
The variation in the total number of samples
collected per well for this study Žfrom one to four
samples. would have created a spatial bias in the
dataset if left unmodified. Thus, mean concentrations for each of the 23 compounds were calculated from all water samples collected from each
well.
3. Results and discussion
The most important finding of this study was
the high frequency with which herbicide degradates were detected in ground water. In fact,
herbicide degradates were eight of the 10 most
frequently detected compounds ŽFig. 2.. Atrazine
was the only herbicide for which the parent compound was found more often than any of its
degradates examined, perhaps because of the
greater environmental persistence of atrazine
compared to the other parent compounds under
investigation ŽWauchope et al., 1992; Stamper et
al., 1997.. In addition, a majority of the measured
concentration for a given herbicide was from its
degradates: acetochlor Ž94.2%., alachlor Ž99.2%.,
atrazine Ž55.5%., cyanazine Ž85.0%., and metolachlor Ž86.3%.. This was true even for a relatively persistent compound such as atrazine.
No analyte concentrations exceeded their respective US Environmental Protection Agency
ŽUSEPA. maximum contaminant levels or health
advisory levels for drinking water ŽTable 2.. However, over half of the herbicide compounds under
investigation have had no such levels established.
Furthermore, these drinking-water criteria only
Table 2
Summary of herbicide results for the 131 wells sampled during this study, 1995]1998
Compound
Detection limit
Detection frequency
Ž%.a
Maximum concentration
Žmgrl.
MCL or HA
Acetochlor
Acetochlor ESA
Acetochlor OA
Alachlor
Alachlor ESA
Alachlor OA
Ametryn
Atrazine
DEA
DIA
HA
Cyanazine
Cyanazine amide
Metolachlor
Metolachlor ESA
Metolachlor OA
Metribuzin
Prometon
Prometryn
Propachlor
Propazine
Simazine
Terbutryn
0.05
0.2
0.2
0.05
0.2
0.2
0.05
0.05
0.05
0.05
0.2
0.05
0.05
0.05
0.2
0.2
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.8
11.4
3.8
6.9
48.1
19.1
0
37.4
32.1
21.4
11.4
6.1
19.8
18.3
51.9
25.2
1.5
18.3
0
0
0
0
0
0.77
11.5
0.8
0.63
8.5
33.4
]
2.1
0.59
1.1
1.3
0.51
0.64
11.3
8.6
15.3
0.27
1.0
]
]
]
]
]
]
]
]
a
Detected in at least one of the water samples collected from a well.
2
]
]
2000
3
]
]
]
1
]
100
]
]
200
100
]
90
10
4
]
D.W. Kolpin et al. r The Science of the Total En¨ ironment 248 (2000) 115]122
119
Fig. 3. Relation between total herbicide residue concentration
Žparent plus degradates. and number of herbicide compounds
detected per water sample.
apply to individual compounds and do not consider the effects of more than one herbicide compound. Studies have shown that some combinations of compounds may exhibit additive or synergistic toxic effects ŽMarinovich et al., 1996;
Thompson, 1996; Pape-Lindstrom and Lydy,
1997.. The detection of multiple compounds in
ground water was common during this study, particularly when herbicide degradates were included. While an average of two herbicides were
found in each ground-water sample in which a
herbicide parent compound was detected, an average of five herbicide compounds were found in
each sample with detections of either a parent
compound or one of its degradates. Similarly,
whereas up to six parent compounds were detected in a single water sample Žmaximum total
parent compound concentration exceeded 12
mgrl., as many as 15 herbicide compounds were
detected in a single sample when degradates were
included Žmaximum total herbicide residue concentration exceeded 78 mgrl. ŽFig. 3..
While herbicides were commonly detected in
ground water across Iowa during this study, their
frequencies of detection and concentrations varied significantly Ž P s- 0.001, Kruskal]Wallis
Fig. 4. Total herbicide concentration by aquifer type ŽALLUV
s alluvial, GDs glacial drift, BK s bedrockrkarst region,
BNKs bedrockrnon-karst region.. Numbers in brackets are
the frequency of herbicide detection for that aquifer type.
test. among the major aquifer types sampled ŽFig.
4.. These differences were much more pro-
120
D.W. Kolpin et al. r The Science of the Total En¨ ironment 248 (2000) 115]122
Fig. 5. The relation between total herbicide residue concentration Žparent plus degradates. and the two indirect estimates
of ground-water age. Numbers in brackets are the frequencies
of herbicide compound detection for each group. Dissolvedoxygen concentrations could not be measured in water from
nine of the municipal wells sampled. An explanation of a
boxplot is provided in Fig. 4.
nounced when herbicide degradates were included ŽFig. 4.. Aquifer types presumed to have
the most rapid recharge rates Žalluvial and bedrockrkarst region aquifers. were those most likely
to contain detectable concentrations of herbicide
compounds, indicating that ground-water age
could be an important factor in explaining these
variations in herbicide contamination.
Although no direct measures of ground-water
age were obtained for this study, two indirect
Fig. 6. The relation between total herbicide concentration
and vulnerability class Žlows well depth ) 50 m and dissolved-oxygen concentration - 0.5 mgrl; intermediate s well
depth ) 50 m and dissolved-oxygen concentration G 0.5 mgrl,
or well depth F 50 m and dissolved-oxygen concentration
- 0.5 mgrl; and high s well depth F 50 m and dissolvedoxygen concentration G 0.5 mgrl.. Numbers in brackets are
the frequencies of herbicide detection for each group. The
nine wells where dissolved-oxygen data were not available
were not used for this analysis. An explanation of a boxplot is
provided in Fig. 4.
estimates of age were available } dissolvedoxygen concentrations and well depth. These two
factors can be used as general indicators of water
age because oxygen tends to be consumed through
biotic and abiotic processes as water travels from
zones of recharge to greater depths, and well
depth provides a general indication of distance
from a recharge zone ŽKolpin et al., 1997a.. The
detection frequencies and concentrations of herbicides for this study were significantly related
Ž P- 0.001, Wilcoxon rank sum test. to these indirect estimates of ground-water age ŽFig. 5.. In
D.W. Kolpin et al. r The Science of the Total En¨ ironment 248 (2000) 115]122
general, these indicators of younger ground water
corresponded to more frequent detections and
higher concentrations of herbicide compounds.
The two indirect estimates of ground-water age
were used to separate the sampled wells into
general vulnerability classes Žlows well depth )
50 m and dissolved-oxygen concentration - 0.5
mgrl; intermediate s well depth ) 50 m and dissolved-oxygen concentration G 0.5 mgrl, or well
depth F 50 m and dissolved-oxygen concentration - 0.5 mgrl; and high s well depth F 50 m
and dissolved-oxygen concentration G 0.5 mgrl..
The results show that the frequencies of detection and concentrations varied significantly Ž P0.001, Kruskal]Wallis test. among the vulnerability classes regardless of whether or not herbicide
degradates were considered ŽFig. 6.. Nevertheless,
when herbicide degradates were included, the frequency of herbicide compound detection within
the highest vulnerability class approached 90%
and the median total herbicide residue concentration increased by over an order of magnitude,
relative to the parent compounds alone, to 2
mgrl ŽFig. 6..
This study demonstrates that obtaining data on
herbicide degradates is critical for understanding
the fate of herbicides in the hydrologic system.
Furthermore, the prevalence of herbicide degradates documented in this study suggests that to
accurately determine the overall effect on human
health and the environment from a specific herbicide its degradates should also be considered.
Acknowledgements
The authors thank the US Environmental Protection Agency, Office of Pesticide Programs for
funding much of the laboratory analysis of ground
water for this study; and the University of Iowa
Hygienic Laboratory and the Iowa Department of
Natural Resources, Geological Survey Bureau for
their support of the IGWM.
References
Bain LJ, LeBlanc GA. Interaction of structurally diverse pesti-
121
cides with the human MDRI gene product P-glycoprotein.
Toxicol Appl Pharm 1996;141:288]298.
Belfroid AC, van Drunen M, Beek MA, Schrap SM, van
Gestel CAM, van Hattum B. Relative risks of transformation products of pesticides for aquatic ecosystems. Sci Total
Environ 1998;22:167]183.
Bintein S, Devillers J. Evaluating the environmental fate of
atrazine in France. Chemosphere 1996;32:2441]2456.
Carbonell E, Xamena N, Creus A, Marcos R. Chromosomal
aberrations and sister-chromatid exchanges as biomarkers
of exposure to pesticides. Clin Chem 1995;41:1917]1919.
Carder JP, Hoagland KD. Combined effects of alachlor and
atrazine on benthic algal communities in artificial streams.
Environ Toxicol and Chem 1998;17Ž7.:1415]1420.
Coats JR. What happens to degradable pesticides? March
issue. Chemtech 1993; 25]29.
Detroy MG. Iowa ground-water-quality monitoring program.
US Geological Survey Open-File Rep 84-815, 1985.
Ferrer I, Thurman EM, Barcelo D. Identification of new ionic
chloroacetanilide-herbicide metabolites in surface and
ground water by HPLCrMS using negative ion detection.
Anal Chem 1997;69:4547]4553.
Holden LR, Graham JA, Whitmore RW et al. Results of the
National Alachlor Well Water Survey. Environ Sci Technol
1992;26:935]943.
Hostetler KA, Thurman EM. Determination of chloroacetanalide herbicide metabolites in water using high-performance liquid chromatography-diode array detection and
high-performance liquid chromatographyrmass spectrometry. Sci Total Environ 2000 Žthis issue..
Howe GE, Gillis R, Mowbray RC. Effect of chemical synergy
and larval stage on the toxicity of atrazine and alachlor to
amphibian larvae. Environ Toxicol Chem 1998;17Ž3.:
519]525.
Kolpin DW, Kalkhoff SJ, Goolsby DA, Sneck-Fahrer DA,
Thurman EM. Occurrence of selected herbicides and herbicide degradation products in Iowa’s ground water, 1995.
Ground Water 1997a;35:679]688.
Kolpin DW, Sneck-Fahrer DA, Hallberg GR, Libra RD. Temporal trends of selected agricultural chemicals in Iowa’s
groundwater, 1982]1995: are things getting better? J Environ Qual 1997b;26:1007]1017.
Kolpin DW, Thurman EM, Linhart SM. The environmental
occurrence of herbicides: The importance of degradates in
ground water. Arch Environ Contam Toxicol 1998;35:
385]390.
Kruger EL, Rice PJ, Anhalt JC, Anderson TA, Coats JR.
Comparative fates of atrazine and deethylatrazine in sterile
and nonsterile soils. J Environ Qual 1997;26:95]101.
La Clair JJ, Bantle JA, Dumont J. Photoproducts and
metabolites of a common insect growth regulator produce
developmental deformities in Xenopus. Environ Sci Technol 1998;32Ž10.:1453]1461.
Longley M, Sotherton NW. Factors determining the effects of
pesticides upon butterflies inhabiting arable farmland. Agric
Ecosyst Environ 1997;61:1]12.
122
D.W. Kolpin et al. r The Science of the Total En¨ ironment 248 (2000) 115]122
Marinovich M, Ghilardi F, Galli CL. Effect of pesticide mixtures on in vitro nervous cells: comparison with single
pesticides. Toxicology 1996;108:201]206.
Meyer MT, Mills MS, Thurman EM. Automated solid-phase
extraction of herbicides from water for gas chromatographyrmass spectrometry analysis. J Chromatogr 1993;629:
55]59.
Pape-Lindstrom PA, Lydy MJ. Synergistic toxicity of atrazine
and organophosphate insecticides contravenes the response
addition mixture model. Environ Toxicol Chem 1997;16:
2415]2420.
Pratt JR, Melendez AE, Barreiro R, Bowers NJ. Predicting
the ecological effects of herbicides. Ecol Appl 1997;7Ž4.:
1117]1124.
Ribas G, Carbonell E, Creus A, Xamena N, Marcos R. Genotoxicity of humic acid in cultured human lymphocytes and
its interaction with the herbicides alachlor and maleic
hydrazide. Environ Mol Mutag 1997;29:272]276.
Stamper DM, Traina SJ, Tuovinen OH. Anaerobic transformation of alachlor, propachlor, and metolachlor with sulfide. J Environ Qual 1997;26:488]494.
Thompson HM. Interactions between pesticides: a review of
reported effects and their implications for wildlife risk
assessment. Ecotoxicology 1996;5:59]81.
Thurman EM, Meyer MT, Pomes ML, Perry CE, Schwab AP.
Enzyme-linked immunosorbent assay compared with gas
chromatographyrmass spectrometry for the determination
of herbicides in water. Anal Chem 1990;62:2043]2048.
Walls D, Smith PG, Mansell MG. Pesticides in groundwater in
Britain. Int J Environ Health Res 1996;6:55]62.
Wauchope RD, Buttler TM, Hornsby AG, Augustijn-Beckers
PWM, Burt JP. The SCSrARSrCES pesticide properties
database for environmental decision-making. Rev Environ
Contam Toxicol 1992;123:1]156.