SC IE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m
w w w. e l s e v i e r. c o m / l o c a t e / s c i t o t e n v
Pesticides in fluvial wetlands catchments under intensive agricultural activities Laurier Poissant a,⁎, Conrad Beauvais a , Pierre Lafrance b , Christian Deblois c a
Direction des Sciences & Technologie, Environnement Canada, 105 McGill, 7e étage (Youville), Montréal, QC, Canada H2Y 2E7 INRS - Centre Eau, Terre et Environnement (ETE), Université du Québec, 490 de la Couronne, Québec, QC, Canada G1K 9A9 c Centre d'Expertise en Analyse Environnementale du Québec, 2700, rue Einstein, bureau E-2-220, Québec, QC, Canada G1P 3W8 b
AR TIC LE I N FO
ABS TR ACT
Article history:
A survey on pesticides (73 compounds) in the Bay St. François wetland and its catchment
Received 6 February 2008
(part of the wetlands of Lake St. Pierre area [St. Lawrence River, Québec]) was achieved in
Received in revised form 14 May 2008
2006. The metabolites as well as the active ingredients of pesticides (11 compounds) were
Accepted 17 May 2008
detected in the wetland and its catchment. This wetland ecosystem was active in the
Available online 14 July 2008
degradation of agricultural pesticides (e.g., atrazine). The measured pesticides were individually below the criteria for aquatic species in natural water, except chlorpyrifos.
Keywords:
Overall, the pesticides lost from agricultural field towards the streams were b1% of the
Pesticides
quantity applied. The environmental fates of the pesticides were found to vary according to
Agricultural
the size of the watershed. Over large catchments, half-life times were important in terms of
Losses from field
global loss from the agricultural lands to wetlands whereas over small catchments, soil
Wetlands impacts
organic carbon/water distribution coefficient (Koc) was an important term for pesticides losses to water system since half-life times were not limiting factors. Crown Copyright © 2008 Published by Elsevier B.V. All rights reserved.
1.
Introduction
Pesticides are used worldwide to control the various pests (UNEP, 1998). Substantial amounts (0.1–5%) of pesticides might be lost from the application fields to the surface waters via runoff and drainage (e.g., Flury, 1996; Leonard, 1990; Haarstad and Braskerud, 2003; Riise et al., 2004; Leu et al., 2004) as well as volatilization (e.g., Poissant and Koprivnjak, 1996; Garmouma and Poissant, 2004; Aulagnier and Poissant, 2005). Pesticides can move in ecosystems according to: their chemical properties such as soil organic carbon/water distribution coefficient, Koc, half-life time (t1/2), vapor pressure, etc. (Wania et al., 1998); the environmental conditions; and their application modes (e.g., Hansen et al., 2001; Lafrance et al., 1997). Hydrological regimes (Nash et al., 2002), soil properties (e.g., Lecomte et al., 2001), atmospheric transport and air–surface exchange of pesticides (Bidleman, 1999) contributed to local, regional as well as global distribution of
pesticides. Accordingly, pesticide residues and metabolites are omnipresent in the environment. Their occurrences have been found in rain in urban and agricultural areas (e.g., Coupe et al., 2000; Chevreuil, 2004). Traces of pesticides were also found in the Arctic (Barrie et al., 1992) and Antarctica (Tanabe et al., 1983). Recent studies have shown a significant quantity of currently-used pesticides detected in Canada (Yao et al., 2006). Pesticides used in Québec account for up to 3400 tons of active ingredients applied on about 1.6 millions of hectares (Fecteau and Poissant, 2001). This pesticide loading contributed to local or regional pesticides contamination in water (Lafrance and Banton, 2004; Giroux et al., 2006; Forrest and Caux, 1990), air (Poissant and Koprivnjak, 1996; Garmouma and Poissant, 2004; Aulagnier and Poissant, 2005; Aulagnier et al., 2008), soil (Ndongo et al., 2000) and biota (Boily et al., 2005). If exposure is above threshold concentrations, pesticide applications have the potential to cause toxic effects to
⁎ Corresponding author. E-mail address:
[email protected] (L. Poissant). 0048-9697/$ – see front matter. Crown Copyright © 2008 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2008.05.030
S CIE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
183
Fig. 1 – Yamaska watershed and Bay St. François catchments location.
nontarget species of plants and animals, including species found in wetlands (Donald et al., 2001). Wetlands are very rich in biodiversity and very important in the natural environment. They are considered to be major structural components of littoral habitats, acting as shelters, nesting and feeding grounds for fish and birds. However, wetlands are often perceived as sinks or storage areas for nutrients, metals and pesticides. Wetland ecosystems are composed of various biological, physical and hydrological components both from spatial and temporal considerations. Hence the response of the biota, water quality functions and others within a wetland to the loading of agricultural pollutants including pesticides remains unclear (Haycock et al., 1997). The objectives of this work are to assess the pesticides field losses in extensive agricultural catchments and to survey their potential interaction with downstream wetlands. A further investigation of the limiting physical–chemical factors involved in the pesticides losses from agricultural field is also within the scope of this work. This paper presents innovative approaches for pesticides fate and field losses studies in agricultural wetland catchments. Furthermore, this paper reports organochlorines, organophosphorus, aryloxyacid pesticides and other group of pesti-
cides, in 2006, in the Bay St. François (BSF) catchments. This allowing the determination of pesticide concentrations in wetlands water and their surroundings, pesticides losses from field crop lands along the growing season and related fates.
2.
Methods
2.1. rates
Site description, environmental factors and pesticides
The Yamaska watershed is located at the south shore of the St. Lawrence River in southern Québec, Canada (Fig. 1). The Yamaska watershed is a highly intensive agricultural region dominated by corn and soya cultivations where more than 400 tons of pesticides is applied annually (Poissant et al., 2007; Gorse et al., 2002). This watershed is extensively studied in respect to its pesticide environmental contamination and impacts (e.g., Boily et al., 2005; Giroux et al., 2006; Aulagnier et al., 2008). The Bay St. François (BSF) wetland, which is a fluvial wetland, covers 9.5 km2 and is part of the Lake St. Pierre (a fluvial lake of the St. Lawrence River) wetlands (112 km2) classified by the Ramsar Convention (United Nations Educational, Scientific, and
184
SC IE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
Fig. 2 – Bay St. François (BSF) catchments land uses, crops distribution and sampling locations: site 1 — BSF outflow; site 2 — Chemin de la castorerie; site 3 — Leblanc's Bridge; and site 4 — Chemin du bois de Maska (2006).
its surface; only ~ 4% is classified “very good” in term of drainage (Laplante and Choinière, 1953). Precipitation heights were measured at the BSF research station (Poissant et al., 2004) using a Fisher and Porter rain gauge. The precipitation height measured from May 26 to September 12, 2006 was 325.4 mm. Precipitation was more important in late May and early June (70 mm) and less in later June (21 mm) (Table 1). Precipitation volume was calculated for each sub-catchments using precipitation heights and their specific surface. The total precipitation volume for the BSF catchments was 12,963,936 m3 during the course of the field experiment. Assuming various runoff coefficients between 1% and 30% of the total precipitation in 2006 (May 26 to Sept. 12) lead to
Cultural Organization). Typically, the BSF's landscape is covered with mixed vegetation of approximately 1.5 m in height, including Scirpus fluviatilis (51–75%), Sagitaria latifolia (11–25%), and Butomus umbellatus (11–25%), soil and water during the summer period. The BSF catchments, which cover 39.8 km2, lid northwest of the Yamaska watershed, are extensively used for agricultural activities (Fig. 2). The top soils in the vicinity of the BSF are composed of about 45–50%, 24–32% and 15–20% of sand, silt and clay, respectively. The loss on ignition as a method for estimating organic and carbonate content in soil is about 3.1% to 10.8% (Laplante and Choinière, 1953). The overall drainage surface in the BSF catchments is flat (slope ~0.21%) and classified “imperfect” drainage over 87% of
Table 1 – Environmental parameters in 2006 in the Bay St. François location in 2006 Date 2006
Water Precipitation Site 1 (Bay St. François): 39.84 km2 Site 2 (Chemin du bois de Maska): 9.19 km2 temperature (°C)
May 26 to June 21 June 21 to June 29 June 29 to July 12 July 12 to July 24 July 24 to Sept. 12 Sum a
Assuming 100% runoff.
16 21 21 22 16
mm
70 21 31.8 19.6 183 325.4
Runoff potential volume (100%) a
Runoff potential volume (100%) a
m3
m3
2,788,800 836,640 1,266,912 780,864 7,290,720 12,963,936
643,300 192,990 292,242 180,124 1,681,770 2,990,426
S CIE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
Table 2 – Estimated pesticides loads on the Bay St. François in 2006 Active ingredient
Main uses
BSF Tonnage kg
Herbicides 2,4D
Common systemic herbicide, used on corn, oats and wheat field Atrazine Common systemic herbicide used on corn and soybeans field Bentazone Contact herbicide used on corn and soybeans field Bromoxynil Contact (with select systemic) herbicide used on cereals and corn field Clopyralide Contact (with select systemic) herbicide used on corn field Dicamba Selective systemic herbicide used on wheat and corn field Dimethenamid Selective, preemergence, chloroacetamide herbicides used on soybeans and corn field Flumetsulam A selective herbicide for the control of certain annual broadleaved weeds used on corn and soya field Glufosinate An herbicide often used with complimentary genetically modified plants used on canola and corn field Glyphosate A non-selective systemic herbicide used on corn, canola and hay MCPA Control most pasture broadleaf weeds used on oalt, wheat and corn field Mesotrione Herbicidal activity against broadleaf weeds used on corn field Metolachlor A preemergence herbicide commonly used on soybeans and corn field Nicosulfuron A selective systemic sulfonylurea herbicide, used on corn field. Primisulfuron A selective systemic sulfonylurea herbicide, used on corn field. Prosulfuron A selective systemic sulfonylurea herbicide, used on corn field. Rimsulfuron A selective systemic sulfonylurea herbicide, used on corn field. Diflufenzopyr A semicarbazone herbicide used for selective broadleaf weed control, used on corn field Mecoprop A selective hormone type phenoxy herbicide used for the selective control of surface creeping weeds. Insecticides Clothianidin Deltamethrin
Total
A neonicotinoid insecticide used on corn field A synthetic pyrethroid insecticeds used on corn field
33.95 397.14 172.85 47.84
185
About 15.5 km2 (39%) of the BSF catchments is dedicated to crop culture. Corn (50%) and soya (25%) are the main crops accounting for more than 75% of land use (Fig. 2) (Financière agricole du Québec, 2006). Pesticides application rates in the BSF catchments have been reconstructed using herbicides handbook for the Arable Crop Sector (CPVQ, 2000) and the crop land used in 2006 (Financière agricole du Québec, 2006). Accordingly, about 2.7 tons of pesticides was applied to the BSF catchments (Table 2) with a maximum application rate of b2.5 kg/ha (Fig. 3). Herbicides: glyphosate, metolachlor, atrazine and bentazone are the most extensively used pesticides on the BSF and Yamaska watersheds (Table 2).
2.2.
Sampling
11.97 85.17 162.55
6.79
80.06
915.26 53.66 23.08 695.10 6.89 0.42 0.40 3.46 10.73
1.05
11.06 0.17
2719.59
flow volumes at the BSF catchment's outlet estimated in between 129,639 m3 and 3,889,181 m3, respectively (Table 1). The water temperatures over the course of the experiment varied between 16 and 22 °C with typical warmer temperature in mid-summer (Table 1).
In 2006, a water quality survey was achieved at four sites surrounding the BSF catchments (Figs. 2 and 3) to assess the pesticides loss from field crop lands during the cultivation season [i.e., May to September (sampling periods: May 26; June 21; June 29; July 12; July 27 and; September 2006)]. Site 1 (BSF outflow) located downstream BSF catchments and is situated at the wetland outlet (to the St. Lawrence River — Lake St. Pierre) whereas sites 2 (Chemin de la castorerie), 3 (Leblanc's Bridge) and 4 (Chemin du bois de Maska) are located upstream and represented individual discharge of their respective sub-catchment. Site 1 integrated the whole BSF catchments including the wetland landscapes (Fig. 2). Site 2 is an enclosed sub-catchment (that is, a catchment entirely within the study area) whereas sites 3 and 4 are not completely enclosed (that is, catchments with runoff contributions or outflows from/to outside the study area), especially in early spring when flooding (i.e., April to early May). Although, the latter should be negligible along the experimental sampling period (i.e., late May to mid September), results and discussion will further focus on sites 1 and 2, hereafter. Twenty four samples (i.e., six times at four sites) were collected in 1 l bottles. Sampling details are presented in [http://www.ceaeq.gouv.qc.ca/methodes/chimie_org.htm] for each pesticides family (e.g., organochlorines and organophosphorus: glass bottles at 4 °C; aryloxyacid: glass bottle at 4 °C and at pH b2). A total of 73 pesticides and metabolites (Table 3) were analyzed for organochlorines, organophosphorus, aryloxyacid pesticides and others.
2.3.
Pesticides analysis
2.3.1.
Methods
Pesticide analyses (listed in Table 3, including their detection limits) were performed at the Centre d'expertise en analyse environnementale du Québec (CEAEQ) (Ste. Foy, Québec), a facility of the Government of Québec, which is an international standard organization/International Electrotechnical Commission 17025 (ISO/IEC 17025) accredited laboratory, using GC-MS in the full scan mode equipped with HP-5 MS capillary columns (30 m long, 0.25 mm i.d., 0.25 μm film thickness). The organochlorine pesticides were extracted with hexane by liquid–liquid extraction, concentrated with rotovap® and analyzed on a ThermoQuest Trace GC-MS operated with helium carrier gas at 1 ml/min. 2 μl samples were splitless
186
SC IE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
Fig. 3 – Pesticides loading to Bay St. François (BSF) catchments in 2006. Site 1 — BSF outflow; site 2 — Chemin de la castorerie; site 3 — Leblanc's Bridge; and site 4 — Chemin du bois de Maska (2006).
injected with an inlet temperature of 250 °C. The oven was first held at 100 °C for 1 min, ramped from 100 °C to 170 °C at 12 °C/ min, held at 170 °C for 3 min followed by a ramp from 170 °C to 250 °C at 15 °C/min, held at 250 °C for 5 min followed by a ramp from 250 °C to 280 °C at 20 °C/min. The final temperature (280 °C) was held for 5 min. The electron energy was 70 eV. The source and the transfer line temperatures were 230 °C and 280 °C respectively. The aryloxyacid herbicides were extracted on C-18 followed by esterification and analyzed on an Agilent 5973 GC/ MS operated with helium carrier gas at 1 ml/min. 2 μl samples were splitless injected with an inlet temperature of 280 °C. The oven was first held at 90 °C for 2 min, ramped from 90 °C to 200 °C at 15 °C/min, held at 200 °C for 3 min followed by a ramp from 200 °C to 280 °C at 15 °C/min. The final temperature (280 °C) was held for 2 min. The electron energy was 70 eV. The source and the transfer line temperatures were 230 °C and 300 °C respectively. The organophosphorus, organonitrogen and others pesticides were extracted with dichloromethane by liquid–liquid extraction, concentrated with rotovap® and analyzed on an Agilent 5973 GC/MS operated with helium carrier gas at 1 ml/ min. 3 μl samples were splitless injected with an inlet temperature of 280 °C. The oven was first held at 100 °C for 4 min, ramped from 100 °C to 150 °C at 10 °C/min followed by a ramp from 150 °C to 240 °C at 5 °C/min, held at 240 °C for 1.5 min followed by a ramp from 240 °C to 270 °C at 15 °C/min. The final temperature (270 °C) was held for 12 min. The electron energy was 70 eV. The source and the transfer line temperatures were 230 °C and 300 °C respectively. All the pesticides were identified by the retention time and the full scan spectra. The internal standard method was used
for quantification. Details on methods are available at http:// www.ceaeq.gouv.qc.ca/methodes/chimie_org.htm.
2.3.2.
Quality control
All the procedures were following ISO/CEI 17025 standards. Details on quality control are available at http://www.ceaeq. gouv.qc.ca/methodes/chimie_org.htm.
3.
Results and discussion
3.1.
Pesticide concentrations
Eleven pesticides (and metabolites) were detected of the 73 compounds listed in Table 3 in the 24 samples from the 4 sites at the BSF catchments (Fig. 2). Atrazine 0.53 ± 1.09 μg/l (83%) (respectively average, standard deviation and detection frequency) N metolachlor 0.17± 0.35 μg/l (71%) N dicamba 0.13 ± 0.15 μg/l (42%)N bentazone 0.12±0.12 μg/l (54%) =desisopropylatrazine 0.12±0.13 μg/l (13%)=deethylatrazine (DEA) 0.12±0.17 μg/l (63%)N chlorpyrifos 0.06±0.02 μg/l (13%)N dimethenamid 0.05± 0 μg/l (4%)N MCPA 0.04±0 μg/l (4%)=I-naphthol 0.04±0 μg/l (4%)N simazine 0.02±0 μg/l (4%). Atrazine and metolachlor showed larger coefficient of variations (i.e., standard deviation/average) (~200%) than dicamba, bentazone (~100%) and chlorpyrifos, dimethenamide, MCPA, Inaphthol, simazine (b100%). This might reflected the dose responses to environmental factors such as rainfall (Table 1), runoff, dilution factor, etc. Fig. 4a gives the concentrations of atrazine on the BSF catchment's sites. Interestingly, on the sample period of June 29th, sites 1 and 2 have different responses to environmental factors. Site 1 indicated atrazine concentration
187
S CIE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
Table 3 – List of pesticides, number of detected, frequency of detection, average concentrations and standard deviation and their detection limits (DL) in 2006 Pesticides
N
Pesticide type a
Organochloride Alpha-HCH Hexachlorobenzene Beta-HCH Lindane Heptachlor Aldrin Chlorthal Heptachlor epoxyde gamma Chlordane Endosulfan I Alpha chlordane p,p′-DDE Dieldrin Endrin Endosulfan-II p,p′-TDE p,p′-DDT Methoxychlor Mirex
I F I I I I H I I I I I I I I I I I I
Aryloxyacid Clopyralid Dicamba Mecoprop MCPA Dichlorprop 2,4-D Bromoxynil Triclopyr Fenoprop (Silvex) MCPB 2,4,5-T 2,4-DB Bentazone Piclorame Dinoseb Diclofop-methyl
H H H H H H H H H H H H H H H H
Organophosphorus Dichlorvos Diuron EPTC Butilate Mevinphos Tebuthiuron Desisopropylatrazine Deethylatrazine Trifluralin Phorate Dimethoate Simazine Carbofuran Atrazine Terbufos Fonofos Diazinon Disulfoton Chlorothalonil Metribuzin Parathion-methyl Carbaryl
I H H H I H H H H I I H I H I I I I F H I I
% of detection
Average (μg/l)
Standard deviation (μg/l)
DL (μg/l)
Guideline limit b (μg/l)
b0.02 b0.01 b0.03 b0.01 b0.01 b0.03 b0.01 b0.02 b0.01 b0.06 b0.02 b0.02 b0.05 b0.07 b0.05 b0.04 b0.04 b0.06 b0.04
√ √ √ √
10
42
0.13
0.15
1
4
0.04
0
13
54
0.12
0.12
√ √
√
√
4 15
13 63
0.12 0.12
0.13 0.17
1
4
0.02
0
20
83
0.53
1.09
b0.03 b0.03 b0.01 b0.01 b0.02 b0.02 b0.02 b0.02 b0.01 b0.01 b0.01 b0.02 b0.03 b0.02 b0.04 b0.02
b0.03 b0.24 b0.02 b0.02 b0.03 b0.29 b0.03 b0.03 b0.02 b0.03 b0.04 b0.02 b0.09 b0.02 b0.05 b0.01 b0.02 b0.03 b0.05 b0.02 b0.03 b0.07
10.0 2.6
–c
–c –c
10 1.8
1750270(continued (continued on next page)
188
SC IE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
Table 3 (continued) Pesticides Chloroxuron Fenitrothion Linuron Malathion Metolachlor Chlorpyrifos Parathion Cyanazine Chlorfenvinphos Myclobutanil Azinphos-methyl Phosalone Dimethenamid Methidathion I-naphthol Bendiocarbe
Pesticide type a H I H I H I I H I F I I H I I I
N
√
% of detection
Average (μg/l)
Standard deviation (μg/l)
17 3
71 13
0.17 0.06
0.35 0.02
1
4
0.05
0
1
4
0.04
0
DL (μg/l) b 0.13 b 0.03 b 0.07 b 0.02 b 0.01 b 0.03 b 0.02 b 0.05 b 0.05 b 0.05 b 0.2 b 0.04 b 0.03 b 0.02 b 0.03 b 0.05
Guideline limit b (μg/l)
7.8 0.0035
–c –c
Also presented the Canadian water quality guideline for protection of aquatic life. NB: Empty cell means below DL. √: Pesticides loads on the Bay St. François in 2006. a Pesticide type: I — insecticide; H — herbicide, F — fungicide. b Canadian water quality guideline for protection of aquatic life. c Guideline not established.
at 0.04 μg/l whereas site 2 measured 1.8 μg/l for the same period. It is suggested that site 1 was decoupled from its catchment during that sampling period and was more representative of the St. Lawrence River over flows. Moreover, Fig. 4b is supporting this assumption since DEA was not detected at site 1 during that period (see Section 3.2 for further details). Also, the average atrazine concentration in the St. Lawrence River is typically ~0.04 μg/l (B. Rondeau, personal communication). Among those above 11 pesticides, 7 (atrazine, metolachlor, dimethenamid, dicamba, MCPA, and bentazone) were listed in Table 2 as “estimated pesticides loads on the Bay St. François” in 2006. The other compounds are metabolites (deethylatrazine and desisopropylatrazine) or pesticides (simazine and chlorpyrifos) that might come from long range transport (atmosphere), catchment's residues from past applications or are missing in the estimated pesticide loads list on the Bay St. François in 2006. Atrazine was the most abundant pesticide measured in the BSF catchments followed by metolachlor, dicamba and bentazone. The latter were among the top 6 pesticides applied on the catchments (Table 2). Interestingly, atrazine showed larger average concentration than metolachlor even with a lower application rate than metolachlor in the BSF catchment (397 vs. 695 kg respectively) (Table 2). This might result from the shorter half-life time of metolachlor compared to atrazine (45 days vs. 80 days, respectively) (Gorse et al., 2002). The measured pesticides were individually below the Canadian water quality guideline for protection of aquatic life, except for chlorpyrifos (~17 times above the guideline) (Table 3). Chlorpyrifos is not listed on Table 2. However, chlorpyrifos was detected in air in 2004 in the Yamaska airshed at a level of 134–641 pg/m3 but not in precipitation (Aulagnier et al., 2008). Chlorpyrifos is one of the most widely used organophosphate insecticides. This pesticide could be transported over long distance and has been detected as far as the Arctic (Chernyak et al., 1996).
Tables 4 and 5 present the loads and the average measured pesticide concentrations at sites 1 and 2, respectively. These results also include the estimated maximum pesticides concentrations assuming various runoff coefficients (1%, 30% and 100%) and the respective pesticides losses to the field at 1% and 30% for reference. Atrazine + Deethylatrazine (DEA), bentazone, dicamba and metolachlor were measured at both sites whereas MCPA was measured at site 1. The average concentrations of atrazine (+DEA), bentazone, dicamba, MCPA and metolachor were 0.44 μg/l, 0.08 μg/l, 0.06 μg/l, 0.04 μg/l and 0.17 μg/l at site 1 and 0.83 μg/l, 0.10 μg/l, 0.29 μg/l, nd, 0.33 μg/l at site 2. On an average pesticide concentrations were generally higher at site 2 than site 1 since the former site was located upstream of the Bay St. François (BSF) catchments. The site 2 accumulated pesticides losses from the intensive agricultural field activities on its sub-catchment draining 9.19 km2 whereas site 1 is located downstream and collected the whole BSF catchments drainage area of 39.8 km2 including 9.5 km2 of wetlands as mentioned above.
3.2.
Pesticide degradations
Pesticides and their metabolites were lost from the agricultural field lands and were introduced in the BSF wetlands system through water discharge. Fig. 4(a and b) indicates the case of atrazine and its metabolite deethylatrazine “DEA” at the four sampling sites. Along its long transition (pathways) in the BSF catchments and wetlands, atrazine was degraded to DEA as indicated by the DEA/atrazine ratio (DAR) (Fig. 4(c)). The DAR might serve as an indicator of the transport mechanism of atrazine (soil residence times) and its degradation in the environment. Low DAR (b1) typically indicates fast surface runoff (as might be found when relatively little degradation occurs during transport) whereas a high DAR (N1) indicates an aging atrazine load with groundwater displacement and
S CIE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
interaction (Adams and Thurman, 1991). Sites 1 and 2 responded very well to atrazine catchments elution with a progressive increase of the DAR. The exception, as mentioned in Section 3.1, is on June 29th at site 1 due to the St. Lawrence River overflows. Atrazine is well known to be a pre- and post-emergent herbicide (Aulagnier et al., 2008). Pre-application typically occurred in midMay during plowing and seeding periods whereas post-emerging application occurred by mid-June during harrowing activities. The DAR varies between bDL and 4. The upper limit of the DAR appeared about 3 months after the last application of the parentcompound. This is in agreement with atrazine half-life range (60 days–100 days at 25 °C) (Gorse et al., 2002). The largest DAR is observed at site 1 situated at the downstream and an integration
189
of the whole BSF catchments including the wetlands (9.5 km2). This suggests that wetlands are effective for atrazine degradation possibly through groundwater advection.
3.3.
Losses from the fields to water system
Tables 4 and 5 present the estimated pesticides losses from the BSF catchments (site 1) and one of its sub-catchments (site 2) respectively, for various hydrological conditions and dilution scenarios. The maximum pesticide concentrations were estimated using a dilution scenario from the pesticide loadings for each active ingredient (Table 2) with the total volume of precipitation (100%
Fig. 4 – a) Concentrations of atrazine on the Bay St. François catchment's sites (2006), b) Concentrations of DEA on the Bay St. François catchment's sites (2006), c) DEA/atrazine ratio on the Bay St. François catchment's sites (2006).
190
SC IE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
Fig. 4 – (continued). runoff) during the 2006 agricultural season (i.e., 12,963,936 m3 from May to September). This scenario gives the maximum concentrations expected considering no atmospheric loading or any pesticide loads from previous years and 100% runoff. Since runoff water from agricultural field and wetlands are complex to assess and that the typical runoff fractions of the total precipitation amount in southern Québec are between 1 and 30% (Pesant et al., 1987), the estimated pesticide concentrations are further presented according to these “more realistic” dilution scenario intervals. Using the maximum estimated concentrations for both runoff water scenarios of 1% and 30% (Pesant et al., 1987) and the observed (measured) average concentrations at each sites,
it is permitting to assess the pesticides losses from the whole BSF catchments (site 1) (Table 4) and from its sub-catchments (site2) (Table 5). Hence, the losses from the fields estimated for the sampling sites are: atrazine (site 1: 0.43%; site 2: 0.55%), dicamba (site 1: 0.29%; site 2: 0.87%), bentazone (site 1: 0.18%; site 2: 0.15%), MCPA (site 1: 0.29%; site 2: ND %) and metolachlor (site 1: 0.09%; site 2: 0.13%). As a whole, less than 1% (0.09–0.87) of the pesticides applied are lost from the agricultural lands. These ranges are in agreement with other field studies dedicated to pesticides field losses (e.g., Leonard, 1990; Flury, 1996; Lafrance et al., 1997). Table 6 summarizes the field's losses from this study and others, such as Riise et al. (2004); Leu et al. (2004); and Lafrance et al. (1997). Lafrance et al.
S CIE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
191
Fig. 4 – (continued).
(1997) showed the evaluation of the masses of herbicides losses from runoff for various rainfall events in the Yamaska River upstream region (i.e., Noire River) from 0.02 to 8.36% for atrazine and from 0.02 to 5.0% for metolachlor.
3.4.
Fate of pesticides
Several factors affect pesticide fates, such as surface preparation, soil structure and chemistry, soil water content, type of irrigation, pesticide formulation, time of application and rainfall events.
In order to further investigate the pesticides losses from the fields some relationships with Koc and pesticide half-lives were overseen. The soil half-life is a measure of the persistence of a pesticide in soil. Pesticides can be categorized on the basis of their half-life as non-persistent, degrading to half the original concentration in less than 30 days; moderately persistent, degrading to half the original concentration in 30 to 100 days; or persistent, taking longer than 100 days to degrade to half the original concentration. A “typical soil half-life” value is an approximation and may vary
192
SC IE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
Table 4 – Estimated concentrations and losses to the field at Bay St. François catchments (site 1): precipitation 12,963,936 m3 in 2006 Pesticide
2,4D Atrazine Bentazone Bromoxynil Clothianidine Clopyralide Deltamethrine Dicamba Dimethenamide Flumetsulam Glufosinate Glyphosate MCPA Mesotrione Metolachlor Nicosulfuron Primisulfuron Prosulfuron Rimsulfuron Diflufenzopyr Mecoprop
Load
Conc. max. estimated
Conc. max. estimated
Conc. max. estimated
(100% runoff)
(1% runoff)
(30% runoff)
kg
µg/l
µg/l
µg/l
33.9 397.1 172.9 47.8 11.1 12.0 0.2 85.2 162.5 6.8 80.1 915.3 53.7 23.1 695.1 6.9 0.4 0.4 3.5 10.7 1.1
2.6 30.6 13.3 3.7 0.9 0.9 0.0 6.6 12.5 0.5 6.2 70.6 4.1 1.8 53.6 0.5 0.0 0.0 0.3 0.8 0.1
261.8 3063.4 1333.3 369.0 85.3 92.3 1.3 656.9 1253.9 52.4 617.6 7060.0 413.9 178.0 5361.8 53.1 3.2 3.1 26.7 82.8 8.1
8.7 102.1 44.4 12.3 2.8 3.1 0.0 21.9 41.8 1.7 20.6 235.3 13.8 5.9 178.7 1.8 0.1 0.1 0.9 2.8 0.3
Av. conc. measured (site 1)
Loss to the fields (estimated)
Loss to the fields (estimated)
(1% runoff)
(30% runoff)
µg/l
%
%
0.44 0.08
0.01 0.01
0.43 0.18
0.06
0.01
0.29
0.04
0.01
0.29
0.17
0.00
0.09
NB Atrazine + DEA
Table 5 – Estimated concentrations and losses at sub-watershed “Chemin du bois de Maska” (site 2): precipitation ~ 2,990,426 m3 in 2006 Pesticide
Load
kg 2,4D 10.9 Atrazine 134.9 Bentazone 59.2 Bromoxynil 26.4 Clothianidine 3.8 Clopyralide 4.6 Deltamethrine 0.1 Dicamba 29.8 Dimethenamide 46.9 Flumetsulam 1.9 Glufosinate 27.2 Glyphosate 249.4 MCPA 32.2 Mesotrione 7.8 Metolachlor 221.9 Nicosulfuron 2.3 Primisulfuron 0.1 Prosulfuron 0.1 Rimsulfuron 1.2 Diflufenzopyr 3.6 Mecoprop 0.8
Conc. max. estimated
Conc. max. estimated
Conc. max. estimated
(100% runoff)
(1% runoff)
(30% runoff)
µg/l
µg/l
µg/l
3.6 45.1 19.8 8.8 1.3 1.6 0.0 10.0 15.7 0.6 9.1 83.4 10.8 2.6 74.2 0.8 0.0 0.0 0.4 1.2 0.3
363.6 4509.6 1980.1 883.4 125.6 155.5 3.3 997.6 1567.6 62.3 909.1 8339.6 1076.8 262.0 7418.8 78.2 4.7 4.5 39.3 121.9 26.3
12.1 150.3 66.0 29.4 4.2 5.2 0.1 33.3 52.3 2.1 30.3 278.0 35.9 8.7 247.3 2.6 0.2 0.2 1.3 4.1 0.9
Av. conc. measured (site 2)
Loss to the fields (estimated)
Loss to the fields (estimated)
(1% runoff)
(30% runoff)
µg/l
%
%
0.83 0.10
0.02 0.00
0.55 0.15
0.29
0.03
0.87
0.33
0.00
0.13
NB Atrazine + DEA
193
S CIE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
Table 6 – Average half-lives (day) and Koc values and estimated losses to the field at sites 1 and 2 in 2006 Pesticides
Parameters Halflife (soil) a
Koc
a
yr Atrazine Bentazone Dicamba MCPA Metolachlor
80 11 18.5 120 45
122 37 10 629 194
Field loss (%) This study
This study
Riise et al. (2004)
Leu et al. (2004)
Lafrance et al. (1997)
Site 1
Site 2
SE Norway
Switzerland
Québec (Yamaska River upstream)
%
%
%
%
%
0.43 0.18 0.29 0.29 0.09
0.55 0.15 0.87
0.09–0.6
0.02–8.36
b0.5
0.04–0.3
0.02–5.0
0.13
Also presented field loss from other studies. a Gorse et al. (2002).
greatly because persistence is sensitive to variations in site, soil, and climate (e.g., Flury, 1996). On one hand, the shortest average soil half-life in Table 6 is for bentazone (~11 days). Its estimated losses to the field at sites 1 and 2 are 0.18% and 0.15%, respectively. Bentazone loads to about 173 kg on the BSF catchments (Table 2). Despites is short half-life bentazone is detected at an average level of 0.12±0.12 μg/l in the BSF catchments (frequency of detection of 54%). On the other hand, atrazine, second largest half-life (~80 days) in Table 6, accounted for field losses of 0.43% and 0.55% for sites 1 and 2, respectively. About 400 kg of atrazine is applied on the BSF catchments. Atrazine is detected at an average level of 0.53±
1.09 μg/l in the BSF catchments (frequency of detection of 54%). As mentioned above half-life time is important in term of residual concentration level in respect to pesticide loadings — see above in Section 3.1 for metolachlor. The sorption coefficient (Koc) describes the tendency of a pesticide to bind to soil particles. Sorption retards movement, and may also increase persistence because the pesticide is protected from degradation. The higher the Koc, the greater the sorption potential. Many soil and pesticide factors may influence the actual sorption of a pesticide to soil. The lowest Koc (Table 6) is for dicamba (~ 10) with estimated losses to the field at sites 1 and 2 of 0.29% and 0.87%, respectively. The latter
Fig. 5 – a–d. Relations between pesticides field loss, Koc and half-lives at BSF catchments and sub-catchments levels in 2006. a) Top left: site 1 field loss with Koc; b) Bottom left: site 1 field loss with half-life; c) Top right: site 2 field loss with Koc; d) Bottom right: site 2 field loss with half-life.
194
SC IE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
was the most important pesticide's fraction loss in the BSF catchments. Dicamba loads for ~85 kg on the BSF catchments and about 30 kg on site 2 sub-catchment (Tables 4 and 5). Fig. 5a–d presents the regression curves between the pesticides estimated field losses (30% runoff scenario), their average literature values of the Koc and half-lives, respectively (the average pesticides literature values of their half-lives and Koc are presented in Table 6) (Gorse et al., 2002). Interestingly, Fig. 5a and b shows that the most important factor (i.e., controlling factor) involved for the pesticides losses from the whole BSF catchments (site 1: surface 39.8 km2) is related to the half-life (R2 = 0.20, α b 0.01) of the pesticides whereas Koc is more important (i.e., controlling factor) (R2 = 0.28, α b 0.01) in the fate (loss) of pesticides in a small catchments (e.g., site 2: surface 9.19 km2) (Fig. 5c and d), since half-lives are not limiting factors in the latter case. This information is particularly important for pesticides management. According to the management scale, various physical properties are to be taken into account; on the watershed scale, half-lives of pesticide are controlling factor whereas on agricultural field scales, Koc is important factor retaining the pesticide as close as possible from the application site. Hence, the key factors to minimize the losses of pesticides to the catchment and the field are short half-lives and large Koc values. Although, other factors are important for pesticides management decision making (e.g., action mode, toxicity, etc.).
4.
Conclusions
About 2.7 tons of pesticides was applied to the BSF catchments in 2006. Herbicides: glyphosate, metolachlor, atrazine and bentazone are the most extensively used pesticides on the BSF catchments. Overall, pesticides losses from BSF agricultural field and catchments to the streams and wetlands were b1%. Although, considering the high variability of cultivation types, treatments in use, climatic conditions and catchment basin morphologies, this value (b1%) is similar to other studies in North America and Europe. BSF wetland is active in the degradation of pesticides (e.g., atrazine). Despite, pesticide metabolites as well as their pesticide active ingredients were present in BSF wetlands, which are rich in biomass and biodiversity, measured pesticides were individually below the criteria for aquatic species in natural water, except chlorpyrifos, which exceeded by 17 times the criteria. However, no information about the ecotoxicological impacts of their mixtures is known. Hence, b27 kg of pesticides was lost to the BSF wetlands. By scaling up the loss from field to the water system, we estimated that less than ~ 4 tons of pesticides was lost to the Yamaska river watershed (of 409 tons) and ~ 34 tons was lost to Québec stream systems (of 3400 tons), which mainly discharged to the St. Lawrence River. The principal pesticide property that minimizes pesticide losses to drain-flow is strong sorption, in most cases indicated by high Koc. Low application rates will also minimize concentrations (but not percentage losses) in drain-flow. More rapid degradation in soil will also tend to reduce residues in drain-flow.
Pesticides fate in the environment is variable according to the watershed scale and properties. Over large catchments, half-lives were found to be important in term of global loss from the agricultural lands to wetlands whereas over small catchments, Koc was important in governing the pesticides loss to water system since the half-lives were not limiting factors. This information should benefits the development of various pesticide management strategies and the decision makers.
Acknowledgments LP would like to thank The St. Lawrence Plan (Environ. Can.) and the Pesticides Science Fund (PSF) of Canada for support and funding. PL would like to thank Nat. Sci. Engin. Res. Council Can. (NSERC) for funding. They authors would like to thank Martin Pilote for technical support and the staff from CEAEQ for pesticide analyses. Special thanks are addressed to the “La financière agricole du Québec” for the agricultural land use data base.
REFERENCES Adams CD, Thurman EM. Formation and transport of deethylatrazine in the soil and vadose zone. J Environ Qual 1991;20:540–7. Aulagnier F, Poissant L. Some pesticides occurrence in air and precipitation in Québec, Canada. Environ Sci Technol 2005;39:2960–7. Aulagnier F, Poissant L, Brunet D, Beauvais C, Pilote M, Deblois C, Dassylva N. Pesticides measured in air and precipitation in the Yamaska Basin (Québec): occurrence and concentrations in 2004. Sci Total Envrion 2008;394:338–48. Barrie LA, Gregor D, Hargrave B, Laker R, Muir D, Shearer R, Tracey B, Bidleman T. Arctic contaminants—sources, occurrence and pathways. Sci Total Environ 1992;122:1–74. Bildeman TF. Atmospheric transport and air–surface exchange of pesticides. Water, Air and Soil Pollut 1999:115–60. Boily M, Bérubé V, DeBlois C, Dassylva N, Spear P. Hepatic retinoids in bullfrogs, Rana catesbeiana in relation to pesticide agricultural exposure. Environ Toxicol Chem 2005;24:1099–106. Chernyak SM, Rice CP, McConnell LL. Evidence of currently-used pesticides in air, ice, fog, seawater and surface microlayer in the Bering and Chukchi seas. Mar Pollut Bull 1996;32:410–9. Chevreuil M. L'atmosphère, vecteur de micropolluants organiques aux écosystèmes. L' Actualité Chimique 2004;277–78:40–8. Coupe RH, Manning MA, Foreman WT, Goolsby DA, Majewski MS. Occurrence of pesticides in rain in urban and agricultural areas of Mississippi, April–September 1995. Sci Total Environ 2000;248:227–40. CPVQ. Traitements herbicides, Grandes cultures 2000, Agdex 640, VT 049, Conseil des productions végétales du Québec Inc.; 2000. ISBN 2-89457-188-7, 358 p. Donald DB, Gurprasad NP, Quinnett-abbott L, Cash K. Diffuse geographic distribution of herbicides in Northern Prairie wetlands. Environ Toxicol Chem 2001;20(2):273–9. Fecteau M, Poissant L. Survol des impacts environnementaux potentiels liés à la production de maïs à des fins énergétiques au Québec. Environnement Canada, Service Météorologique du Canada; 2001. ISBN 0-662-84346-0, 78 p. Flury M. Experimental evidence of transport of pesticides through field soil — a review. J. Environ Qual 1996;25:25–45. Forrest S, Caux PY. Pesticides in tributaries of the St. Lawrence River (including certain CEPA priority list of substances).
S CIE N CE OF T H E TOT AL E N V I RO N ME N T 4 0 4 ( 2 00 8 ) 1 8 2–1 95
1987–1988 Program Report, Environment Canada, Centre St. Laurent, Hydrology and Networks Quebec Region. Longueuil, Québec, Canada: Laboratoire du Capitaine Bernier; 1990. Financière agricole du Québec, 2006. Database available on http:// www.financiereagricole.qc.ca/. Garmouma M, Poissant L. Occurrence, temperature and seasonal trends of a- and g-HCH in air (Quebec, Canada). Atmos Environ 2004;38:369–82. Giroux I, Robert C, Dassylva N. Présence de pesticides dans l'eau au Québec: bilan dans les cours d'eau de zones en culture de maïs et de soya en 2002, 2003 et 2004, et dans les réseaux de distribution d'eau potable. Ministère du Développement Durable, de l'Environnement et des Parcs, Direction du suivi de l'état de l'environnement, Direction des politiques de l'eau et Centre d'expertise en analyse environnementale du Québec; 2006. ISBN 2-550-46504-0, Envirodoq n°ENV/2006/013, collection n°QE/00173, 57 p. et 5 annexes. Gorse I, Grégoire F, Laverdière C, Roussel T. Répertoire des principaux pesticides utilisés au Québec. Les Publications du Québec; 2002. ISBN 2-551-16754-X, 476 p. Haarstad K, Braskerud BC. Pesticide retention in the watershed and in a small constructed wetland treating diffuse pollution. Diffuse Pollution Conference, Dublin; 2003. p. 14–102. Poster Papers. Hansen NC, Moncrief JF, Gupta SC, Capel PD, Olness AE. Herbicide banding and tillage system interactions on runoff losses of alachlor and cyanazine. J Environ Qual 2001;30:2120–6. Haycock NE, Burt TP, Goulding KWT, Pinay G, editors. Buffer zones: their processes and potential in water protection. Harpenden, U.K.: Quest Environmental Publ.; 1997. 326 p. Lafrance P, Banton O. The impact of corn-cropping systems on the herbicides transport from agricultural fields to streams in the St. Lawrence lowlands (Québec). 11th Annual Int. Conf. on the St. Lawrence River Ecosystem: Managing Our Water: The Great Lakes/St. Lawrence River. Cornwall, Canada, St. Lawrence River Inst. Environ Sci; 2004. p. 1372. Lafrance P, Banton O, Gagné P. Exportation saisonnière d'herbicides vers les cours d'eau mesurée sur six champs agricoles sous quelques pratiques culturales du maïs (Basses-Terres du Saint-Laurent). Rev Sci Eau 1997;10 (04):439–59. Laplante L, Choinière L. Étude Pédologique des sols du comté d'Yamaska de agronomes-pédologues École Supérieure d'Agriculture. Division des sols Bulletin technique; 1953. No3: 153 p. Lecomte V, Barriuso E, Bresson LM, Koch C, Le Bissonnais Y. Soil surface structure effect on isoproturon and diflufenican loss in runoff. J Environ Qual 2001;30:2113–9.
195
Leonard RA. Movement of pesticides into surface waters. In: Cheng HH, editor. Pesticides in the soil environment — processes, impacts and modeling. . Soil Sci. Soc. Am.Madison, WI, USA: SSSA Publ.; 1990. p. 303–49. Book Series, No 2, chap. 9. Leu C, Singer H, Stamm C, Muller SR, Schwarzenbach RP. Variability of herbicide losses from 13 fields to surface water within a small catchment after a controlled herbicide application. Environ Sci. Technol 2004;38(14):3835–41. Nash D, Halliwell D, Cox J. Hydrological mobilization of pollutants at the field/slope scale. In: Haygarth PM, Jarvis SC, editors. Agriculture, hydrology and water quality, vol. 10. Oxon, UK: CABI Publ.; 2002. p. 225–42. Ndongo B, Leroux GD, Fortin J. Transport du linuron, de l'imidaclopride et du bromure au travers de colonnes de sol et de lysimètres drainants. Biotechnol Agron Soc Environ 2000;4 (1):33–40. Pesant AR, Dionne JL, Genest J. Soil and nutrient losses in surface runoff from conventional and no-till corn systems. Can J Soil Sci 1987;67:835–43. Poissant L, Koprivnjak JF. Fate and atmospheric concentration of α- and γ-hexachlorocyclohexane in Québec, Canada. Environ Sci Technol 1996;30:845–51. Poissant L, Pilote M, Xu X, Zhang H, Beauvais C. Atmospheric mercury speciation and deposition in the Bay St. François Wetlands. J Geophys Res 2004;109:D11301. _________________________ doi:10.1029/2003JD004364. Poissant L, Beauvais C, Copin N. Sommaire du milieu physique, modes de culture et estimation des taux d'application des pesticides sur le basin de la rivière Yamaska. Rapport technique, Environnement Canada; 2007. 39 p. Riise G, Lundekvam H, Wu QL, Haugen LE, Mulder J. Loss of pesticides from agricultural fields in SE Norway — runoff through surface and drainage water. Environ Geochem Health 2004;26(2):269–76. Tanabe S, Hidaka H, Tatsukawa R. PCBs and chlorinated hydrocarbon pesticides in Antarctic atmosphere and hydrosphere. Chemosphere 1983;12(2):277–88. UNEP. Report of the first Session of the INC for an International Legally Binding Instrument for Implementing International Action on Certain Persistent Organic Pollutants (POPs). UNEP Report; International Institute for Sustainable Development (IISD), vol. 15. ; 1998. Wania F, Haugen JE, Lei YD, Mackay D. Temperature dependence of atmospheric concentrations of semivolatile organic compounds. Environ Sci Technol 1998;32:1013–21. Yao Y, Tuduri L, Harner T, Blanchard P, Waite D, Poissant L, Murphy C, Belzer W, Aulagnier F, Li YF, Sverko E. Spatial and temporal distribution of pesticide air concentrations in Canadian agricultural regions. Atmos Environ 2006;40:4339–51.
