Science of the Total Environment 587–588 (2017) 75–86

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

High frequency monitoring of pesticides in runoff water to improve

understanding of their transport and environmental impacts
Marie Lefrancq a,⁎, Alain Jadas-Hécart a, Isabelle La Jeunesse a,b, David Landry a, Sylvain Payraudeau c
a
University of Angers, LETG-Angers UMR CNRS 6554, 2 bd Lavoisier, 49045 Angers, France
b
University François Rabelais of Tours, Citeres UMR CNRS 7324, 33, allée Ferdinand de Lesseps, B.P. 60449, 37204 Tours cedex 3, France
c
University of Strasbourg, CNRS, ENGEES, LHyGeS UMR 7517, 1 rue Blessig, F-67084 Strasbourg, France

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• High frequency sampling enabled to es-

timate what the range of TU values is.
• The exported loads were homoge-
neously distributed throughout runoff
events.
• The C-Q hysteresis patterns were highly
dynamic and dependent on storm and
chemical characteristics.
• Intense rainfall events induced stronger
hysteresis patterns.

a r t i c l e i n f o a b s t r a c t

Article history: Rainfall-induced peaks in pesticide concentrations can occur rapidly. Low frequency sampling may therefore largely
Received 17 December 2016 underestimate maximum pesticide concentrations and fluxes. Detailed storm-based sampling of pesticide concen-
Received in revised form 31 January 2017 trations in runoff water to better predict pesticide sources, transport pathways and toxicity within the headwater
Accepted 3 February 2017
catchments is lacking. High frequency monitoring (2 min) of seven pesticides (Dimetomorph, Fluopicolide, Glyph-
Available online 24 February 2017
osate, Iprovalicarb, Tebuconazole, Tetraconazole and Triadimenol) and one degradation product (AMPA) were
Editor: J Jay Gan assessed for 20 runoff events from 2009 to 2012 at the outlet of a vineyard catchment in the Layon catchment in
France. The maximum pesticide concentrations were 387 μg L−1. Samples from all of the runoff events exceeded
Keywords: the legal limit of 0.1 μg L−1 for at least one pesticide (European directive 2013/39/EC). High resolution sampling
Concentration peak used to detect the peak pesticide levels revealed that Toxic Units (TU) for algae, invertebrates and fish often
Runoff exceeded the European Uniform principles (25%). The point and average (time or discharge-weighted) concentra-
Toxicity tions indicated up to a 30- or 4-fold underestimation of the TU obtained when measuring the maximum concentra-
Hysteresis tions, respectively. This highlights the important role of sampling methods for assessing peak exposure. High
Suspended matter
resolution sampling combined with concentration-discharge hysteresis analyses revealed that clockwise responses

⁎ Corresponding author at: University of Angers, LETG-Angers - UMR CNRS 6554, 2 bd Lavoisier, 49045 Angers, France.
E-mail address: marie.lefrancq@univ-angers.fr (M. Lefrancq).

http://dx.doi.org/10.1016/j.scitotenv.2017.02.022
0048-9697/© 2017 Elsevier B.V. All rights reserved.
76 M. Lefrancq et al. / Science of the Total Environment 587–588 (2017) 75–86
were predominant (52%), indicating that Hortonian runoff is the prevailing surface runoff trigger mechanism in the
study catchment. The hysteresis patterns for suspended solids and pesticides were highly dynamic and storm- and
chemical-dependent. Intense rainfall events induced stronger C-Q hysteresis (magnitude). This study provides new
insights into the complexity of pesticide dynamics in runoff water and highlights the ability of hysteresis analysis to
improve understanding of pesticide supply and transport.
1. Introduction
Headwater streams significantly impact downstream water quality
and contribute to the biodiversity of the overall stream network
(Rasmussen et al., 2013; Salmon-Monviola et al., 2013). Headwaters in
agricultural areas are subject to particularly high pesticide inputs as a
result of their location near agricultural plots and their low discharge
levels (Lorenz et al., 2016). However, headwaters are often not consid-
ered significant surface water bodies and are generally disregarded (e.g.,
by the European Water Framework Directive (WFD)) (Bundschuh et al.,
2014). In headwater catchments, pesticide export dynamics is generally
discontinuous and complex dependent upon rainfall occurrence and
seasonal use (Stehle et al., 2012). Rainfall-induced peaks in pesticide
concentrations can occur rapidly. However, even very short pulses
(from a few minutes to several hours) of high pesticide concentrations
may negatively impact aquatic organisms (Boxall et al., 2013; Cold
and Forbes, 2004). Low frequency sampling largely underestimates pes-
ticide concentrations and fluxes (Assoumani et al., 2013; Bieroza and
Heathwaite, 2015), and fails to evaluate acute pesticide exposure
(Lorenz et al., 2016; Stehle et al., 2012). The environmental effect of pes-
ticides can be evaluated by observing the number of exceedances of the
legal limit (0.1 μg L−1 for a single pesticide and 0.5 μg L−1 for the sum of
all pesticides in a given sample, European directive 2013/39/EC). Anoth-
er option consists in comparing pesticide concentrations with results
from ecotoxicological tests for key species (e.g., EC50, LC50, etc.)
(Papadakis et al., 2015). Aquatic ecosystems are typically exposed to a
mixture of pesticides, leading to possible combined action of co-occur-
ring pesticides, potentiation or synergistic interactions that may result
in higher toxicity than predicted (Abdo et al., 2015; Cedergreen,
2014). The combined toxicity of mixed pesticides is increasingly debat-
ed in the literature, stressing the need for a new assessment tool
(Carvalho et al., 2014). The toxic unit (TU) is an easy and common
tool that assumes a cumulative effect of pesticide toxicity. TU is the
sum of the ratios between the concentrations of different pesticides
and their respective toxicities for a given species (Allinson et al., 2015;
Bundschuh et al., 2014). This indicator evaluates the toxicological im-
pact of a given pesticide mixture. In the present study, TUs were esti-
mated using point, average and maximum concentrations, as well as a
discharge-weighted concentration of a runoff event, to assess the bene-
fit of high frequency sampling in evaluating ecological risk.
High frequency sampling allows: (i) an assessment of much more
reliable estimates of pesticide fluxes (Bieroza and Heathwaite, 2015),
(ii) an estimation of the contribution of each hydrograph component
(e.g., rising limb, falling limb, first flush, etc.) to pesticide transport
(Granger et al., 2009) and (iii) an in-depth investigation of pesticide re-
sponse patterns to rain events (Bende-Michl et al., 2013). Concentra-
tion-discharge (C-Q) patterns have already been largely documented
for turbidity, temperature, suspended solids and nutrients such as ni-
trates and phosphorus. These patterns were applied to better predict
and understand their sources and pathways within a catchment during
rainfall events (Bende-Michl et al., 2013; Bieroza and Heathwaite, 2015;
Donn et al., 2012; Dupas et al., 2015; Granger et al., 2009; Paul et al.,
2015; Pietroń et al., 2015). A C-Q pattern may be hysteretic and follow
either a clockwise or anti-clockwise hysteresis loop. A clockwise pattern
may indicate transport limitations, while an anti-clockwise pattern may
indicate supply limitations (Bieroza and Heathwaite, 2015). A clockwise
pattern is generally observed when the solute peak occurs before the
© 2017 Elsevier B.V. All rights reserved.
outflow peak. This is often related to a rapid flushing of solutes that
are readily available due to their ability to dissolve, and the exhaustibil-
ity or the proximity of the solute source. Correspondingly, anticlockwise
patterns can be the result of a delayed solute delivery due to the up-
stream source area or to its availability (sorption, complexation, etc.)
(Pietroń et al., 2015; Zuecco et al., 2016). Studying hysteretic relation-
ships for different events (Outram et al., 2016), different sites within a
catchment (Paul et al., 2015) or different solutes (Bieroza and
Heathwaite, 2015; Dupas et al., 2015) can thus reveal important infor-
mation about the spatial origin of solutes and the underlying hydrolog-
ical processes involved in solute transport (Fovet et al., 2015). However,
little is known about the C-Q patterns of organic pesticides (Taghavi et
al., 2011). Detailed storm-based sampling is thus required to better
evaluate pesticide sources and transport in runoff water, as well as
their ecological impact on aquatic ecosystems.
This study aims to explore the use of high frequency sampling to bet-
ter understand the sources, transport pathways and ecological impacts
of pesticides in runoff events. The objectives of this study were to eval-
uate the advantages of high frequency sampling in runoff water in the
assessment of the ecotoxicological impacts of pesticides and to provide
new insights into pesticide supply and transport pathways by evaluat-
ing first flush and concentration-discharge patterns. Runoff events
from 2009 to 2012 were selected arbitrarily to cover the growing season
and diverse rainfall and runoff occurrences during pesticide application
periods. Due to the high cost of analyses and difficulties associated with
the selected sampling procedure, only 20 runoff events were studied.
The experimental setup was designed to obtain a wide range of runoff
events and not to compare the yearly mass balance budget.
2. Material and methods
2.1. Description of the vineyard catchment
The 2.2-ha catchment is located in the commune of Rochefort sur
Loire, (West of France, 47°19″19.47″N; 0°38″21.39″W) (Fig. 1). Its alti-
tude ranges between 39.4 and 83.7 m along an ESE facing slope. Soils
overlay an impermeable Armorican substratum (Namurian Shale, Sand-
stone and Psammites). The catchment boundary was determined using
ArcGIS 10.1 (ESRI, Redlands, United States) and data derived from an
electronic theodolite (1 pt. per 10 m2) and was validated by field obser-
vations. The outlet of the catchment is located 37 m from an influent of
the Layon River that is itself 500 m downstream. The catchment is
characterised by three different gradients: (i) The upper catchment
has 0–5% slopes (51% of the total catchment area); (ii) The middle
catchment has 5–15% slopes (40%); and (iii) The lower catchment has
N15% slopes (9%), including agricultural terraces. Soil depths vary
from 30 cm in the lower zone to 120 cm in the upper zone. Spatial var-
iability of the soil was characterised using 50 surface soil samples (0–
20 cm) taken from across the three areas. Soil characteristics for the
catchment are as follows (mean ± SE): sand: 42.3 ± 5.1%; silt:
36.1 ± 3.0%; clay: 19.5 ± 2.3%; OM: 2.1 ± 0.4%; pH: 7.1 ± 0.4; CEC:
10.4 ± 0.8 meq 100 g −1; CaCO3: 0.1%. The structural stability of the
soils was measured by immersing soil aggregates in water followed by
the separation of the soil fraction using mechanical sieving (Le
Bissonnais et al., 2007). Fractions N250 μm were measured and consti-
tuted an index of soil stability. Grassed rows were comprised of 38 ±
12% stable aggregates while weeded rows were comprised of 18 ± 6%
 M. Lefrancq et al. / Science of the Total Environment 587–588 (2017) 75–86 77

subscripts A and B, Fig. 1). These fields are separated by a grassed strip
that has been reinforced with stones near the catchment outlet to limit
erosion. Vines are planted perpendicular to the slope. Grass covers the
vineyard plots along every two vine rows (1.4 m separated).

2.2. Pesticide properties and application

In the studied vineyard, 31 commercial products with 21 different ac-

tive ingredients were applied in the following amounts: 3.1, 4.8, 2.1 and
3.0 kg in 2009, 2010, 2011 and 2012, respectively. Of those products
used, 53, 28, 54 and 32% were fungicides, respectively. The study focussed
on 7 pesticides (Dimetomorph (DIM), Fluopicolide (FLU), Glyphosate
(GLY), Iprovalicarbe (IPR), Tebuconazole (TEB), Tetraconazole (TET)
and Triadimenol (TRI)) and one degradation product (AMPA) because
of their detection frequency and their yearly applied mass within the
study catchment (Tables 1 and 2). The physical and chemical characteris-
tics of these 8 compounds are provided in Table 1. The 7 pesticides were
mostly applied between March and July. TEB and TRI were generally ap-
plied to the upstream section of plot A, while FLU, IPR and TET were only
applied to plot B (Fig. 1). The ability of high resolution sampling to im-
prove our knowledge of the pesticide sources, transport pathways and
ecological impacts in runoff was assessed using these 8 compounds.

2.3. Hydrological and sampling procedures

Rainfall intensity was recorded using a tipping bucket rain gauge

Fig. 1. The study catchment with the experimental setup (Rochefort sur Loire, 47°19″
19.47″N;0°38″21.39″W).
stable aggregates, which indicate a limited risk of soil sealing (Goulet et
al., 2006).
Mean annual rainfall is 623 mm (±124 mm) (1985–2014, Beaulieu
sur Layon, 3 km from the study site). No permanent flows were observed
and a drainage system installed in the 1980s conducts water at 60–80 cm
depth to a central grassed strip in the middle of the catchment during
rainfall-runoff events (Fig. 1). Vineyards cover 89% of the catchment
and are divided into two fields managed by two farmers (noted with
(Fig. 1). Water discharge was measured at the catchment outlet every
30 s using a bubbler flow module with a 5 mm precision (OTT CBS,
Kempten, Germany) combined with a Venturi channel. As soon as the
water level increased above 2 cm, 500 mL of water were sampled
every 2 min using an automatic sampler (Buhler 1000, Hach-Lange,
GB) containing 24 polyethylene flasks of 500 mL. Automatic phone
calls notified people on duty if the rainfall event exceeded the capacity
of the sampler, enabling manual sampling of the rest of the event
when necessary. Water samples were then collected and placed on ice
for transportation to the laboratory.
2.4. Chemical analysis
Samples were filtered through a 0.45 μm regenerated cellulose filter
to measure the total suspended solid concentration (TSS). Raw and fil-
tered samples were kept at −18 °C in the dark prior to chemical analy-
sis. GLY and AMPA samples were analysed after filtering (0.45 μm), this

Table 1
Family, type, commercial formulations, physicochemical properties, toxicity and detection frequency of the 7 pesticides (DIM, FLU, GLY, IPR, TEB, TET, TRI) and degradation product
(AMPA).

Compound Abbrv. Commercial Chemical family Formula Physico-chemical characteristics of active substances
formulation
Log Log Henry DT50 DT5O EC50a,c Detection
Kowa Kocd constanta (aerobic-anaerobic fieldd frequency
(25 °C) soil)b (20 events)
[−] [−] [Pa m3 mol−1] [day] [day] [ppm] [%]

Dimethomorph DIM Panthéos/ Morpholine C21H22ClNO4 2.68 2.61 2.04 × 10−05 75–26 10–61 N 10.6 90.9
Arco DTI/
Fastime/ Tulsa
Fluopicolide FLU Profiler Benzamide C14H8Cl3F3N2O 2.90 2.50 4.15 × 10−05 415–561 132 N1.8 100
Glyphosate GLY Roundup/ Vival/ Phosphonoglycine C3H8NO5P −3.20 3.85 2.10 × 10−07 96–22 5–21 40 100
Sirlene
AMPA AMPA – Unclassified CH6NO3P −1.63 – 0.16 n.a. 76–240 n.a. 100
Iprovalicarbe IPR Ocarina Carbamate C18H28N2O3 3.20 – 1.40 × 10−06 15.5–n.a n.a. n.a. 40.9
Tebuconazole TEB Abilis Triazole C16H22ClN3O 3.7 3.00 1.00 × 10−05 597–1260 20–92 2.79 22.7
Tetraconazole TET Greman Triazole C13H11Cl2F4N3O 3.56 – 3.60 × 10−04 364–180 136–1688 3.0 86.3
Triadimenol TRI Abilis Triazole C14H18ClN3O2 3.18 2.40 3.5 × 10−06 250–n.a 110–375 51 31.8
a
Obtained from the PPDB (2007, 2008, 2009), The Pesticide Properties DataBase (PPDB) developed by the Agriculture & Environment Research Unit (AERU) at the University of
Hertfordshire.
b
Obtained from the PAN (PAN, 2006) pesticide data base.
c
Daphnia magna test, 48 h obtained from PPDB data base.
d
Obtained from AGRITOX.
78 M. Lefrancq et al. / Science of the Total Environment 587–588 (2017) 75–86

Table 2 MS). The remainder of the ethyl acetate extract was analysed by gas
Application amount [g] and number of applications [−] (in brackets) of the 7 pesticides chromatography/ion trap tandem mass spectrometry GC/IT-MS-MS.
(DIM, FLU, GLY, IPR, TEB, TET, TRI) in 2009, 2010, 2011 and 2012.
The pesticide quantification limit within the water samples was
Applied total mass [g] (number of applications [−]) 0.05 μg L−1. Recovery rates ranged between 86 and 96%.
Compound 2009 2010 2011 2012 Total
2.5. Data analysis and calculation
DIM 373.8 (2) 177.8 (1) 177.8 (1) 231.2 (2) 960.5 (6)
FLU 105.2 (1) 105.2 (1)
GLY 1087.2 (3) 2642.4 (3) 933.3 (2) 1999.8 (2) 6662.7 (10) 2.5.1. Climatic and hydrological data
IPR 213.3 (1) 89.6 (1) 302.9 (2) To compare the amount, intensity and duration of rainfall events,
TEB 124.2 (3) 54.0 (1) 54.0 (1) 232.2 (5) an event index (EVI) was calculated using the following ratio
TET 40.3 (2) 39.5 (2) 39.5 (2) 39.5 (2) 158.8 (8)
TRI 41.4 (3) 18.0 (1) 18.0 (1) 77.4 (5)
(Baartman et al., 2013):

step being required for the derivatization of both compounds prior to
their analysis by HPLC with fluorescence detection. Other compounds
were analysed in raw water in order to not underestimate the runoff ex-
port via the particulate phase (N 0.45 μm). However, for 22 arbitrarily se-
lected samples from the measured runoff events, fungicides were
analysed in both filtered and raw forms to investigate the partitioning
of those fungicides in the “dissolved phase” (b0.45 μm) and in the par-
ticulate phase (N0.45 μm). DIM, FLU, IPR and TET concentrations in
raw and filtered water (b 0.45 μm) did not differ significantly
(p N 0.05), indicating that fungicides were predominantly transported
in the dissolved phase (b 0.45 μm), which is supported by previous stud-
ies (Maillard and Imfeld, 2014). TEB and TRI were not quantifiable in the
22 samples but are hypothesized to behave similarly to TET because they
belong to the same triazole family and have a similar log Kow (Table 1).
The GLY and AMPA extraction and quantification procedures are de-
scribed in detail in supplementary material. Briefly, GLY and AMPA were
analysed using HPLC separation with spectrofluorimetric detection after
decomplexation of both analytes, followed by a derivatization using 9-
fluorenylmethyl chloroformate (FMOC-Cl). The average recovery rates
were 100% and 105% for glyphosate and AMPA, respectively. The detec-
tion and quantification limits were 0.03 μg L− 1 and 0.09 μg L−1 for
glyphosate and 0.04 μg L−1 and 0.1 μg L−1 for AMPA, respectively.
Other pesticide analyses were performed at the GIRPA laboratory in
Angers (France), which is accredited and meets the ISO 17025 standard
(ISO/IEC 17025:2005) for pesticide analysis. Briefly, after spiking with
surrogate standards chlorpyriphos-d10 and diuron-d6, water samples
(500 mL) were successively liquid–liquid extracted at 3 pHs (b 2, 7
and N12) using a mixture of dichloromethane: ethyl acetate 80:20.
The extracts were combined, dehydrated and evaporated to b 1 mL
under vacuum. The concentrated extract was transferred into a vial
and adjusted accurately to 1 mL with ethyl acetate. An aliquot of this ex-
tract was solvent exchanged with a mixture of water:methanol (50:50
with 0.1% acetic acid). Analysis was performed by liquid chromatogra-
phy/electrospray ionisation tandem mass spectrometry (LC/ESI-MS-
Imax  Rtot
EVI ¼ ð1Þ
D
where Imax is the maximum rainfall intensity [mm h−1], Rtot is the rain-
fall amount [mm] and D is the rainfall duration [min]. A high EVI repre-
sents a short but intense rainfall event, whereas a low EVI indicates an
event with a low intensity but long duration. The catchment response
time is defined as the time between the gravity centre of the rain
event and the peak outflow.
2.5.2. Pesticide export
The maximum pesticide concentration, which was measured, was
supposed to be the maximum of the event and was labelled the pesti-
cide peak. To calculate pesticide loads, linearity of the values between
two successive concentration data points or flow measurements was as-
sumed. When pesticides were not detected (336 analyses, i.e. 16%), con-
centrations were set to zero to calculate the mean concentrations,
occurrences and loadings. When pesticide concentrations were detect-
ed but lower than the quantification limit (186 analyses, i.e. 9%), the
sample concentration was set to half of the quantification limit. Pesti-
cide export coefficients were estimated as the ratio of the exported
loads from a runoff event compared to the cumulative application of
the year preceding the studied runoff event. Hydrological characteristics
and pesticide concentrations were compared using the paired nonpara-
metric Wilcoxon signed rank test and the Spearman rank correlation
test. Statistical tests were performed using the R software (R develop-
ment Core Team, 2008; version 2.6.2).
2.5.3. Ecotoxicological data
The impact of the pesticide mixture toxicity on the aquatic ecosys-
tem was evaluated using the TU approach (Bundschuh et al., 2014).
TU was estimated as follows for each event:
n
Ci
TU ¼ ∑ ð2Þ
i¼1 IC 50i

Fig. 2. Representation of the four different TUmax, TUmean,TUpond, TUinst calculations for an artificial runoff event with n samplings for two different compounds, called 1 (purple) and 2
(orange).
 M. Lefrancq et al. / Science of the Total Environment 587–588 (2017) 75–86 79

where Ci is the concentration of pesticide i within a mixture of n pesti-
cides [μg L−1] and IC50i is the concentration of pesticide i which induces
a response halfway between the baseline and maximum after a speci-
fied exposure time [μg L−1]. In our study, IC50 values were taken from
the PPDB database (Lewis et al., 2016) and were related to three trophic
levels: algal growth inhibition (acute 72 h), invertebrate immobility
(acute 48 h) and fish mortality (acute 96 h). Although no fish are likely
to live in intermittent streams within a dry catchment, as was the case
in the present study, TU values enable a comparison and evaluation of
the toxicity of runoff waters. Although other species may be more sen-
sitive, for comparison purposes, Daphnia magna and Oncorhynchus
mykiss were used to study the effects of all pesticides, except for TET,
on invertebrate and fish species, respectively. TET effects on fish species
were measured using Lepomis macrochirus. Pseudokirchneriella
subcapitata and Scenedemus subspicatus were primarily used to study
the effects of the 8 target compounds on green algae. If no data for
these green algae species were available (i.e., for GLY, FLU, IPR and
TET), IC50-data for any other green algae species were used
(Bundschuh et al., 2014). Four different methods were used to estimate
Ci to test the loss of ecotoxicological information associated with the dif-
ferent sampling methods. The formulas used are illustrated in Fig. 2.
Within these formulas, Ci represents: (i) the point concentration within
each sample (Cinst,t used to estimate TUinst,t), (ii) the maximum concen-
tration during the runoff event (Cmax used to estimate TUmax), (iii) the
average concentration during runoff event (Cmean used to estimate
TUmean) and (iv) the discharge-weighted average concentration (Cpond
used to estimate TUpond). Cmean represents the concentration of a pool
of samples obtained at regular time intervals, whereas Cpond represents

Fig. 3. Rainfall, outflow, total suspended solids (TSS) and pesticide and degradation product loads (GLY, AMPA, DIM, FLU, IPR, TEB, TET, TRI) for 20 runoff events in a vineyard catchment
(Rochefort sur Loire, France).
80 M. Lefrancq et al. / Science of the Total Environment 587–588 (2017) 75–86

the concentration of a pool of samples obtained for a constant outflow direction (1 for clockwise, −1 for anticlockwise and 0 for no or an un-
volume. Cinst,t represents the potential concentration that may occur clear hysteresis pattern). Therefore, ΔR varied between − 1 to 1. The
for a random sample. Estimated TU values were compared to the Euro- magnitude parameter, ΔC, represents the relative change in pesticide
pean Union TU threshold of 0.1 for algae and 0.01 for invertebrates and concentrations during the runoff event and is measured as follows:
fish, which are known as the European Uniform Principles (European (Bieroza and Heathwaite, 2015).
Commission, 2011). TUs were estimated for all studied runoff events ex-
cept October, 20 2009 and October, 14 2012. Data for these two events C max −C min
ΔC ¼ ð5Þ
were omitted because not all compounds were analysed. C max

2.5.4. First flush calculation
A first-flush effect is defined to occur when a disproportionately
greater pesticide load is transported by a relatively small proportion of
the runoff volume during the beginning of a runoff event. The first
flush (FF [%]) is defined as follows (Bertrand-Krajewski et al., 1998):
t
∫ 0X C ðt ÞQ ðt Þdt
FF X ¼ T
 100
∫ 0 C ðt ÞQ ðt Þdt
where X is the defined runoff volume of a sample as a percent of the
total runoff [%], here, 10, 25, 50 and 75%; C(t) [μg L− 1] and Q(t)
[L s−1] are the pesticide concentration and the runoff outflow at time
t, respectively; T is the duration of the runoff event [min]; and tx is the
time at which X% of runoff has been delivered [min]. A FFX value signif-
icantly larger than X indicates a disproportionate phenomenon.
Bertrand-Krajewski et al. (1998) assumed that a significant first flush
occurred if at least 80% of the total pollutant mass was transported in
the first 30% of runoff discharged during a rainfall event.
2.5.5. Hysteresis pattern
Runoff events for which at least 2 sample points were quantifiable
have been taken along both the rising and falling limb were used in
the present study to investigate the hysteresis patterns. To compare
the hysteretic loops of different runoff events and solutes, two quantita-
tive indices were used. First, the rotational parameter ΔR, which inte-
grates information on the hysteresis area and direction, was estimated
as follows (Bieroza and Heathwaite, 2015; Butturini et al., 2006):
ΔR ¼ R  Ah
where Ah is the normalized hysteresis area, calculated as the polygon
area of the convex-hull of the C-Q hysteresis curve after standardizing
discharges and concentrations to a unity scale; and R is the hysteresis
where Cmax and Cmin are the maximum and minimum pesticide concen-
trations, respectively.
3. Experimental results
3.1. Characteristics of selected rainfall-runoff events
ð3Þ
Twenty rainfall-runoff events occurring between 2009 and 2012
were studied. Each event yielded N1 m3 total discharge at the outlet of
the catchment. These events represented a wide range of rainfall inten-
sities and durations, runoff percentages and volumes (Fig. 3 and Table
3). Selected rainfall events exhibited return periods ranging from 0 to
10 years (Data from MeteoFrance, Table 3). The runoff coefficient
ranged from 0.3 to 47%. The catchment response time ranged between
5 and 482 min, with an average of 88.5 min. For events with a high
EVI, i.e., intense and short rainfall, discharge occurred rapidly (short re-
sponse time) (p b 0.001) (Table 3). The concentration of TSS ranged be-
tween 11 and 6454 mg L−1 (Table 4) and was positively correlated with
outflow (p b 0.001), suggesting that rill erosion occurred.
Maximum pesticide concentrations reached 13, 8, 386.9, 47, 3, 81, 68
and 4.2 μg L−1 for DIM, FLU, GLY, AMPA, IPR, TEB, TET and TRI, respec-
tively (Table 4). These concentrations are more than one order of mag-
nitude higher than reported in French vineyards with grab or integrated
sampling (Gregoire et al., 2010; Lefrancq et al., 2014; Rabiet et al., 2010)
and a similar order of magnitude for short time interval sampling
(Louchart et al., 2001). For each runoff event, maximum pesticide con-
centrations decreased with increasing time following the last applica-
tion of the analysed pesticide (p b 0.016) (Table 4). This indicates the
ð4Þ occurrence of a dissipation effect, as previously mentioned in literature
(Arias-Estévez et al., 2008). The exported pesticide and AMPA loads for
each of the events are represented in Fig. 3. Maximum exported loads
for a single event reached 154, 142, 2229, 660, 39, 185, 255 and
39 mg, for DIM, FLU, GLY, AMPA, IPR, TEB, TET and TRI respectively.

Table 3
Climatic and hydrology characteristics of the 20 studied runoff events (Rochefort sur Loire, France). Values in bold are extremes. Grey cases represent hysteresis analysis events with at
least two measured points along the rising and falling limbs.

2009 2011 2012

15 16 28 13 17 20 28 9 13 27 30 4 19 20 9 11 13 14 16 17
APR APR APR MAY SEP OCT JUN DEC DEC APR APR MAY MAY MAY OCT OCT OCT OCT OCT OCT
Rainfall
Rainfall amount [mm] 8.4 4.4 6.6 12.4 16.4 27.0 15.8 12.0 12.6 40.0 8.4 6.6 15.8 10.6 7.6 31.5 8.8 20.2 5.0 19.6
Rainfall duration [h] 2.6 1.6 1.6 1.6 15.8 4.9 1.5 14.9 7.7 22.1 1.9 1.8 7.5 9.7 1.4 5.0 2.9 12.1 1.8 24.0
6 min–peak rainfall
[mm h–1] 14.0 26.0 28.0 68.0 8.0 38.0 78.0 14.0 14.0 6.0 14.0 6.0 10.0 8.0 36.0 64.0 16.0 12.0 22.0 14.0
intensity
Seven day
[mm] 20.0 31.5 6.5 2.6 9.9 0.4 2.0 11.7 19.0 38.0 61.6 61.8 3.2 17.6 30.1 29.7 51.7 53.5 64.6 64.0
antecedent rainfall
EVI [] 0.8 1.2 2.0 9.1 0.1 3.5 13.8 0.2 0.4 0.2 1.0 0.4 0.4 0.1 3.3 6.8 0.8 0.3 1.0 0.2
Return period [y] <2 <2 <2 <2 <2 2–5 2 <2 <2 2–5 <2 <2 <2 <2 <2 5–10 <2 <2 <2 <2
Hydrology
Total discharge [m3] 8.0 8.2 4.0 10.2 2.9 30.1 12.6 1.5 21.9 413.7 24.9 6.9 0.9 5.9 2.3 44.6 3.5 50.7 3.3 79.8
Runoff event
[h] 3.4 3.5 1.8 7.8 8.6 5.6 1.6 12.2 11.0 89.7 24.7 23.1 5.3 22.3 1.2 7.9 4.7 17.9 7.6 30.0
duration
Runoff coefficient [%] 4.3 8.5 2.8 3.7 0.8 5.1 3.6 0.6 7.9 47.0 13.5 4.7 0.3 2.5 1.4 6.4 1.8 11.4 3.0 18.5
Maximum outflow [L s–1] 3.1 3.9 1.6 5.3 0.5 11.2 11.1 0.3 3.3 10.5 4.1 0.8 0.3 0.7 2.2 25.2 1.1 6.0 0.7 3.3
Peak time
[min] 68.0 16.0 53.0 13.5 923.0 141.5 27.5 721.5 428.0 1095.0 64.5 119.0 367.5 135.0 58.0 21.5 111.5 301.5 87.0 300.0
discharge
–1
Mean outflow [L s ] 0.6 0.6 0.6 0.3 0.0 1.4 2.1 0.0 0.6 1.3 0.3 0.1 0.0 0.1 0.5 1.2 0.2 0.8 0.1 0.8
Response time [min] 33 6 33 5 16 27 11 341 116 482 24 69 181 58 14 8 19 42 17 286
 M. Lefrancq et al. / Science of the Total Environment 587–588 (2017) 75–86 81

This export corresponded to 0.04, 0.13, 0.21, 0.02, 0.23, 1.29 and
0.22% of the application loads during the preceding year of the runoff
event, respectively (Table 4). Such high export for single event may lead
to significant ecotoxicological impact on the surrounding ecosystem.
3.2. Toxicity impact
All runoff water samples contained at least one pesticide with a con-
centration exceeding 0.1 μg L−1 (Table 4). Thus, pesticide levels in the
studied catchment continuously exceed mandated acceptable concentra-
tions (European directive 2013/39/EC). Toxic units based on maximum
concentrations (TUmax) reached 0.29, 0.05 and 0.04 for algae, invertebrate
and fish, respectively (Table S1). The percentage of runoff events that
exceeded the European Uniform Principles threshold for these species
was 15, 5 and 25%, respectively. Several researchers questioned the rele-
vance of the TU threshold set by the EU for invertebrates (red line in
Fig. 4) (Bundschuh et al., 2014; Liess and Von Der Ohe, 2005; Peters
et al., 2013). Instead, these researchers preferred to use a TU value of
0.001 for invertebrates (blue line in Fig. 4). Based on this threshold, 55%
of events analysed in the present study may represent a risk to the integ-
rity of the aquatic ecosystem. Dilution occurs when these flows reach the
Layon River 500 m downstream. However, approximately 182 km2 of

Table 4
Number of days after treatment (DAT) [d], export coefficient (EC) [%] and total suspended solids (TSS) [mg L−1] and pesticide concentrations [μg L−1] (C; provided as min – max and
mean ± standard deviation) (DIM, FLU, GLY, AMPA, IPR, TEB, TET, TRI) in the study catchment (Rochefort sur Loire, France). Values in bold signify extremes for each lines. n is the number
of samples for each event. Grey cases represent hysteresis analysis events with at least two measured points along the rising and falling limbs. EC are expressed in four classes: “≥0.1” for
EC ≥ 0.1, “≥0.01” for 0.1 N EC ≥ 0.01, “≥10−3” for 0.01 N EC ≥ 0.001 and “≥10−6” for 0.001 N EC ≥ 10−6. * showed cases when positive exported loads of pesticide occurred while this pesticide
was not applied during the previous year.

2009
15 APR 16 APR 28 APR 13 MAY 17 SEPT 20 OCT
n 16 6 9 36 12 34
TSS 59.2 – 353.6 74.5 –187.3 180.2 –530.6 NA NA 13.8 –1547.5
[mg L–1]
149.4 ± 80.9 139.7 ± 42.7 330.9 ± 105.2 NA NA 489.8 ± 385.2
C DAT EC C DAT EC C DAT EC C DAT EC C DAT EC C DAT EC
[µg L –1] [d] [%] [µg L –1] [d] [%] [µg L –1] [d] [%] [µg L –1] [d] [%] [µg L –1] [d] [%] [µg L –1] [d] [%]
0.1 –0.3 0.3 –0.3 0.3 –0.4 0.2 –2.3 4.7 –8 0.1 –4.5
DIM 320 >10–6 321 >10–6 333 >10–6 348 >10–6 97 >10–3 130 >0.01
0.2 ± 0.1 0.3 ± 0.02 0.4 ± 0.1 0.9 ± 0.6 6.2 ± 0.9 3.4 ± 0.9
NA NA NA NA NA NA
FLU NA NA NA NA NA NA NA NA NA NA NA NA
NA NA NA NA NA NA
17.6 –59.6 160.7 –386.9 24.2 –51.9 1 –62 0.5 –1.8 –6 NA
GLY 9 >0.01 0.1 >0.1 12 >0.01 27 >0.01 154 >10 187 NA
43.2 ± 13.9 285.1 ± 83.2 41.3 ± 9.2 20.8 ± 15.4 1.0 ± 0.5 NA
7.7 –20.8 11.9 –23.2 14.6 –28 1.2 –47 1.5 –3.6 NA
AMPA – – – – – – – – – – – –
15.3 ± 4.3 17.3 ± 4.3 22.5 ± 4.2 18.0 ± 13.5 2.1 ± 0.8 NA
0 –0.03 0 –0.03 0 –0.03 0 –0.1 2 –3 0.06 –1.7
IPR 232 >10–6 233 >10–6 245 >10–6 260 >10–6 69 >10–3 102 >0.01
0.01 ± 0.01 0.02 ± 0.01 0.02 ± 0.01 0.03 ± 0.03 2.5 ± 0.3 1.2 ± 0.3
0 –0.03 0.03 –0.03 0 –0.03 0 –81 0.08 –0.19 0 –0.1
TEB 288 >10–6 289 >10–6 301 >10–6 0 >0.1 63 >10 –6
96 >10–6
0.01 ± 0.01 0.03 ± 0 0.01 ± 0.01 16.0 ± 19.6 0.1 ± 0.03 0.03 ± 0.02
0 –0.03 0.03 –0.03 0.03 –0.03 3.5 –68 0.17 –0.86 0.07 –0.4 130
TET 341 >10–6 342 >10–6 354 >10–6 1 >0.1 97 >10–3 >0.01
0.02 ± 0.01 0.03 ± 0 0.03 ± 0 19.0 ± 16.6 0.3 ± 0.2 0.3 ± 0.1
0 –0.03 0–0 0–0 0.03 – 4.2 0 – 0.14 0 – 0.1
TRI 288 >10–6 289 0 301 0 0 >0.01 63 >10–6 96 >10–3
0.01 ± 0.01 0±0 0±0 1.6 ± 1.2 0.1 ± 0.1 0.04 ± 0.04

2011 2012
28 JUN 9 DEC 13 DEC 27 APR 30 APR 4 MAY

n 21 15 25 68 12 16

TSS 29.1 –6454.5 95.2 –330 81.3 –587.1 28.9 –1071.1 96 –369.6 40 –314.5
[mg L–1]
2436.9 ± 1729.4 128.0 ± 58.1 269.0 ± 186.0 201.9 ± 136.9 271.8 ± 90.5 208.6 ± 87.1
C DAT EC C DAT EC C DAT EC C DAT EC C DAT EC C DAT EC
[µg L –1] [d] [%] [µg L –1] [d] [%] [µg L –1] [d] [%] [µg L –1] [d] [%] [µg L –1] [d] [%] [µg L –1] [d] [%]
0.1 –13 0.7 –1.2 0.4 –1.3 0 –0.4 0.1 –0.3 0.1 –0.2
DIM 21 >0.01 185 >10–6 189 >10–3 325 >0.01 328 >10–3 332 >10–6
11.0 ± 3.9 1.0 ± 0.2 0.9 ± 0.3 0.2 ± 0.1 0.2 ± 0.08 0.2 ± 0.07
0.1 –8.2 3.4 –6 1.2 –5 0.3 –1.5 0.5 –1.1 0.2 –0.6
FLU 48 >0.01 212 >10–3 216 >0.01 352 >0.1 355 >0.01 359 >10–3
5.5 ± 2.2 5.3 ± 0.9 3.8 ± 1.1 0.7 ± 0.4 0.8 ± 0.3 0.4 ± 0.2
0 –13.7 0.3 –0.9 0.1 –4.5 0.1 –2 0.2 –3.7 0.1 –0.5
GLY 85 >10–3 249 >10–6 253 >10–3 389 * 392 * 396 *
2.5 ± 2.7 0.4 ± 0.2 0.8 ± 1.1 0.9 ± 0.5 1.1 ± 0.9 0.4 ± 0.1
0 –7.8 1 –2.2 0.7 –2 0.1 –3.7 0.5 –7.6 0.5 –3.1
AMPA – – – – – – – – – – – –
4.2 ± 2.1 1.4 ± 0.3 1.1 ± 0.3 1.8 ± 0.9 3.0 ± 1.9 2.3 ± 0.7
0 –0 0 –0 0 –0 0 –0 0 –0 0 –0
IPR 718 0 882 0 886 0 1022 0 1025 0 1029 0
0±0 0±0 0±0 0±0 0±0 0±0
0.03 –0.07 0 –0 0 –0 0 –0.03 0 –0 0 –0
TEB 53 >10–6 217 0 221 0 357 >10–3 360 0 364 0
0.03 ± 0.02 0±0 0±0 0.01 ± 0.01 0±0 0±0
0.03 –1.4 0.1 –0.3 0.1 –0.2 0.03 –0.14 0.05 –0.1 0.2 –3.3
TET 21 >0.01 185 >10–6 189 >10–3 325 >0.01 328 >10–3 1 >0.01
0.9 ± 0.4 0.2 ± 0.05 0.2 ± 0.04 0.06 ± 0.04 0.08 ± 0.02 1.8 ± 1.3
0.07 –0.1 0 –0 0 –0.1 0 –0.12 0 –0.03 0 –0
TRI 53 >10–3 217 0 221 >10–3 357 >0.1 360 >10–3 364 0
0.08 ± 0.02 0±0 0.02 ± 0.03 0.05 ± 0.05 0.01 ± 0.01 0±0
(continued on next page)
82 M. Lefrancq et al. / Science of the Total Environment 587–588 (2017) 75–86

Table 4 (continued)

19 MAY 20 MAY 9 OCT 11 OCT 13 OCT 14 OCT

n 12 13 12 36 4 5
TSS 83.7 –155 17.5 –65.9 132.5 –702 100.6 –1542 144.7 –315.2 47.1 –130.6
[mg L–1] 125.8 ± 21.1 42.4 ± 12.6 287.2 ± 153.3 355.5 ± 314.9 229.5 ± 72.1 94.1 ± 31.9
C DAT EC C DAT EC C DAT EC C DAT EC C DAT EC C DAT EC
[µg L –1] [d] [%] [µg L –1] [d] [%] [µg L –1] [d] [%] [µg L –1] [d] [%] [µg L –1] [d] [%] [µg L –1] [d] [%]
0.03 –0.03 0.03 –0.03 0.8 –2.3 0 –1.9 0.8 –1.4 NA
DIM 347 >10–6 348 >10–6 76 >10–3 78 >0.01 80 >10–3 81 NA
0.03 ± 0 0.03 ± 0 1.7 ± 0.6 1.2 ± 0.5 1.0 ± 0.3 NA
1.3 –1.7 0.7 –1.4 0.2 –0.4 0 –0.4 0.2 –0.3 NA
FLU 374 * 375 * 517 * 519 * 521 * 522 NA
1.54 ± 0.15 1.1 ± 0.4 0.4 ± 0.1 0.2 ± 0.1 0.3 ± 0.1 NA
0.9 –3 0.5 –2.1 1 –3.8 1.4 –3.5 1.4 –2.4 0.5 –1.7
GLY 8 >10–6 9 >10 –6
112 >10–6 114 >10 –3
116 >10–6 117 >10–3
1.8 ± 0.7 1.1 ± 0.6 2.9 ± 0.8 2.4 ± 0.6 1.9 ± 0.5 1.0 ± 0.6
2.3 –4.4 1.4 –4.1 1.9 –5.6 2.5 –5.5 2.4 –3.4 0.9 –3.2
AMPA – – – – – – – – – – – –
3.6 ± 0.5 2.7 ± 0.9 4.3 ± 1.0 3.7 ± 0.9 3.0 ± 0.5 1.9 ± 1.0
0 –0 0 –0 0.2 –0.4 0 –0.4 0.1 –0.3 NA
IPR 1044 0 1045 0 76 >10–3 78 >10–3 80 >10–6 81 NA
0±0 0±0 0.3 ± 0.1 0.2 ± 0.1 0.2 ± 0.1 NA
0 –0 0 –0 0 –0 0 –0 0–0 NA
TEB 379 0 380 0 522 0 524 0 526 0 527 NA
0±0 0±0 0±0 0±0 0±0 NA
0.8 –1.5 0.5 –1.1 0.2 –0.4 0 –0.3 0.2 –0.2 NA
TET 16 >10–3 17 >0.01 123 >10–3 125 >0.01 127 >10–3 128 NA
1.2 ± 0.3 0.8 ± 0.3 0.3 ± 0.1 0.2 ± 0.1 0.2 ± 0.03 NA
0 –0 0 –0 0 –0 0 –0 0 –0 NA
TRI 379 0 380 0 522 0 524 0 526 0 527 NA
0±0 0±0 0±0 0±0 0±0 NA

2012
16 OCT 17 OCT Total on the 20 runoff events
n 3 28 313
TSS 65.9 –92 11 –145 11.0 –6454.6
[mg L–1] 78.07 ± 13.16 72.5 ± 27.4 374.2 ± 720.0
C DAT EC C DAT EC C EC
[µg L –1] [d] [%] [µg L –1] [d] [%] [µg L –1] [%]
0.4 –1.2 0.2 –1.1 0 – 13
DIM 83 >10–3 84 >0.01 0.05
0.9 ± 0.4 0.5 ± 0.2 1.8 ± 2.8
0.1 –0.3 0.1 –0.3 0 –8.2
FLU 524 * 525 * 0.03
0.3 ± 0.1 0.2 ± 0.04 1.6 ± 2.1
0.9 –1 0.2 –1.6 0 –386.9
GLY 119 >10–6 120 >10–3 0.33
0.9 ± 0.05 0.5 ± 0.3 12.1 ± 41.6
0.1 –1.9 0.4 –2 0 – 47
AMPA – – – – –
1.3 ± 1.0 0.8 ± 0.4 5.6 ± 8.1
0.11 –0.27 0 –0.18 0–3
IPR 83 >10–6 84 >10–3 0.01
0.2 ± 0.1 0.07 ± 0.06 0.4 ± 0.7
0–0 0 –0 0 –81
TEB 529 0 530 0 0.06
0±0 0±0 2.6 ± 9.7
0.17 –0.21 0.08 –0.2 0 –68
TET 130 >10–3 131 >0.01 0.19
0.2 ± 0.02 0.2 ± 0.03 3.3 ± 9.5
0 –0 0 –0 0 –4.2
TRI 529 0 530 0 0.06
0±0 0±0 0.3 ± 0.7

Fig. 4. Toxic unit for Daphnia magna immobility (acute 48 h) based on observed point concentrations (boxplot), maximum concentrations (red) and mean concentrations (green) for the
20 runoff events (log scale). Horizontal lines represent the toxic unit threshold of the European commission for invertebrates (0.01 in red), based on ecotoxicological studies (Bundschuh et
al., 2014) (0.001 in blue) and calculated using mandated acceptable concentrations of 0.1 μg L−1 for each pesticide (European directive 2013/39/EC) (0.00014 in green).
 M. Lefrancq et al. / Science of the Total Environment 587–588 (2017) 75–86 83

vineyards feed the Layon River, suggesting the potential combination of
contaminated flows from N8000 small headwater catchments with fea-
tures similar to our study site. TU (max and mean) for fish and inverte-
brates were negatively correlated with seven-day antecedent rainfall
(p b 0.001) (Table S1 in supplementary material), highlighting a dissipa-
tion effect with preceding rainfall (Olsson et al., 2013).
The variations between the different TU estimations are represented
for invertebrates in Fig. 4. Surprisingly, very little variation was ob-
served between TUmean and TUpond. There was, on average, 1.6 (and
up to 4) times greater TUmax than TUmean and 3.4 (up to 30) times great-
er than TUinst. The method used to estimate TU results in significant dif-
ferences in the values obtained, which is partly due to the variability of
pesticide concentrations patterns throughout the hydrograph.
3.3. First flush and concentration-discharge patterns
The mean and range of the first flushes FF10, FF25, FF50 and FF75 for
all chemicals across the 20 runoff events were 9.8 ± 5.2, 25.3 ± 10.2,
50.6 ± 13, 75.5 ± 11.6%, respectively. It suggests that no disproportion-
ate event occurred. In other words, no “first flush effect” as defined by
Bertrand-Krajewski et al. (1998) was observed, contrary to expecta-
tions. Nevertheless, this phenomenon may occur when pesticides are
rapidly mobilized at the beginning of a runoff event if those pesticides
are less-sorptive than those in the present study or if their source area
is very near the catchment outlet. 15 runoff peaks allowed us to study
the differences between the rising and falling limb of the hydrograph
(at least two sample points for each rising and falling limb were taken
during these events) (Fig. 5). Between 57 and 99% of the water dis-
charge volume occurred during the falling limb, which may be partly
due to natural and artificial drainage features that delayed the flow.
TSS and pesticide concentrations did not significantly differ between
the rising and falling limb (p N 0.05) (except for AMPA and the triazole
family: TEB, TET, TRI) and thus require a deeper analysis of concentra-
tion patterns.
Based on previously determined hysteresis classifications (Bieroza
and Heathwaite, 2015), 52% of the concentration versus discharge

Fig. 5. Hysteresis rotational parameter ΔR (clockwise = positive (blue), anti-clockwise = negative (red) and no or unclear hysteresis pattern = null) for the 15 runoff peaks and TSS, GLY,
AMPA, DIM, FLU, TRI, IPR, TET and TEB. Values represent the normalized area of the C-Q hysteresis. Grey cases represent undetected or unavailable data. Runoff peaks are named with the
runoff date and a subscript which represents the number of the peak within the runoff event.
84 M. Lefrancq et al. / Science of the Total Environment 587–588 (2017) 75–86
graph of quantified suspended solid or pesticide values exhibited clock-
wise patterns, 27% exhibited anticlockwise patterns and 21% had no or
an unclear hysteresis pattern (Fig. 5). ΔR ranged from − 0.47 to 0.48,
with an absolute average of 0.15 ± 0.11 (Fig. 5). ΔC ranged from 0.21
to 1, with an average of 0.64 ± 0.24. ΔR for TSS and EVI were correlated
(p b 0.01), indicating that the clockwise hysteresis of TSS occurred for
intense rainfall events, whereas an anti-clockwise hysteresis pattern
was observed for mild rainfall events (Fig. 5). The potential erosion
power of intense rainfall events may lead to rapid mobilisation of TSS
along the central grass strip, resulting in clockwise responses (Bieroza
and Heathwaite, 2015; Bowes et al., 2005).
The direction of the hysteresis loops for pesticides were not consistent
between substances within an event, nor for one substance across all
events (Fig. 5). AMPA, DIM, FLU, IPR, TEB and TET presented predominant-
ly clockwise hysteresis patterns (52%), while TRI exhibited anti-clockwise
pattern (67%) and GLY exhibited unclear pattern (54%). Fig. 5 shows that
for intense events (high EVI), stronger hysteresis patterns (clockwise or
anti-clockwise) were predominant, as indicated by a greater loop area.
However, this tendency was significant only for GLY, FLU and TET
(p b 0.05). ΔR for GLY, AMPA, DIM and FLU was significantly and positively
correlated with maximum outflow and the runoff coefficient (p b 0.05).
On the other hand, the number of days since the last application, air tem-
perature and antecedent rainfall did not correlate with ΔR. No hysteresis
trends were observed based on the pesticide affinity for water (Kow), as
might have been expected (Taghavi et al., 2011). That study found that
pesticide molecules with low to moderate solubility resulted in clockwise
hysteresis loops while soluble molecules resulted in anticlockwise loops
in an 1110 km2 groundwater-based catchment.
4. Discussion
High frequency sampling is costly but reveals important information
about the ecotoxicity and underlying hydrological and hydrochemical
processes governing pesticide transport in headwater catchments.
4.1. Hydrological functioning of catchments
Understanding surface runoff trigger mechanisms (Hortonian or sat-
uration excess) and the contribution of drainage characteristics to run-
off is crucial to the simulation of suspended solids and/or pesticide
transport (Tilahun et al., 2016). Often, saturation excess runoff is related
to low soil depth, good soil structure, high organic matter content, and
low erosion potential. On the other hand, Hortonian runoff is associated
with steep slopes, the absence of base flow, and crusted soils
characterised by low clay and organic matter contents and low structur-
al stability (Descroix et al., 2007; Tilahun et al., 2016). In this study, in-
dicators of both types of possible runoff scenarios are present such as
low soil depth (30–120 cm), 20% clay content and high structural stabil-
ity indicating saturation excess runoff but also steep slopes and the ab-
sence of a base flow indicating Hortonian runoff. For all runoff events,
pesticide concentrations correlated with flow rate (significant except
for FLU and TET, p b 0.05), indicating a concentration effect and not a di-
lution effect. The concentrations of the 8 compounds correlated with
TSS concentrations (p b 0.0025). This suggests that pesticide
mobilisation and transport occurred along similar pathways as TSS,
which was largely a function of Hortonian runoff associated with in-
tense rainfall. In addition, maximum concentrations are positively cor-
related with EVIs (only significant for GLY and AMPA, p b 0.01). This
may be explained by a minimum rainfall intensity required to mobilise
GLY and AMPA from the soil (Delpla et al., 2011; Ulrich et al., 2013),
which supports also the Hortonian runoff concept.
A clockwise hysteresis loop was the most observed pattern within
the study site, as was expected for a small catchment (Hudson, 2003;
Seeger et al., 2004). This suggests the direct and rapid mobilisation of
TSS and pesticides via runoff and indicates that drainage had a minimal
impact on pesticide and TSS export (Marttila and Kløve, 2010). Seeger et
al. (2004) hypothesized that saturation excess overland flow caused
clockwise or anticlockwise hysteresis while Hortonian runoff resulted
in a figure-eight shaped pattern and thus a small hysteresis area. In
the present study, intense rainfall events caused stronger hysteresis pat-
terns (whether clockwise or anti-clockwise) with greater loop areas
(Fig. 5). This may be explained by the combined result of rainfall inten-
sity, catchment hydrological status and the spatial distribution of pesti-
cides (Fauvel et al., 2016). This is partly due to two different,
chronological occurrences. For clockwise hysteresis patterns, intense
rainfall events rapidly induced surface runoff and a higher runoff coeffi-
cient. Rapid mobilisation of pesticides can thus occur as a flush of avail-
able pesticides prior to peak outflow. For anti-clockwise hysteresis
patterns, intense rainfall events can activate pesticide sources further
from the catchment outlet, less hydrologically connected or dryer
(Doppler et al., 2014). On the contrary, mild rainfall events did not pos-
sess enough energy or power to rapidly mobilise pesticides nearby or to
activate and transport pesticides from further away, resulting in diffuse
pollution with a small or non-existent hysteresis area. The shift from a
clockwise to anti-clockwise hysteresis pattern for different substances
within an event or for one pesticide across runoff events was highly dy-
namic and dependent on the storm and substance rather than only on
the catchment characteristics as previously suggested (Bieroza and
Heathwaite, 2015; Thompson et al., 2012) (Fig. 5).
4.2. Ecotoxicological impact of runoff events
Maximum TU values were observed for runoff events that occurred
very near the application date, for intense rainfall events and after a
dry period; In other words, for the first significant rainfall event after
application.
There are only small differences in the TUmean and TUpond values (Fig.
4) indicating that frequent sampling at short time intervals gives rela-
tively the same value for TU regardless of the method of computation.
The potential range of concentrations with random sampling are
reflected with the point concentration (or TUinst values). Random sam-
pling, e.g., every month, as often performed by national monitoring pro-
grams (Botta et al., 2012; Bundschuh et al., 2014), may underestimate
peak exposure. This highlights the importance of the sampling method
in assessing the ecotoxicological impact of contaminated runoff on
nearby ecosystems.
Where FLU represented 19% of the total pesticide load in all runoff in
2011 and 2012, this pesticide accounted for 59, 79 and 96% of the com-
posite TU value for invertebrates, fish and algae, respectively (data not
shown). FLU was extremely persistent and was always detected
(N525 days) after a single application of the pesticide on plot B. Quanti-
tative data on the behaviour of FLU in water is currently very limited in
the literature (Jeschke, 2016), although its negative effects on aquatic
ecosystems have been shown (Lewis et al., 2016). A limited number of
pesticides often account for N80% of the toxicity of a given pesticide
mixture (Vaj et al., 2011; Verro et al., 2009). The restriction of these sub-
stances would significantly reduce the ecotoxicological impact of pesti-
cides mixtures on aquatic ecosystems.
4.3. Supply limitation vs transport limitation
Pesticides primarily enter agricultural streams during rainfall events
via runoff; their movement is dependent on the presence of a sufficient
amount of the given pesticide and its availability (supply), as well as its
ability to be mobilized via runoff (transport).
No first flush effect was observed in the present study and the con-
tribution of pesticide exports was similar during almost all runoff
events. Pesticide transport rather than pesticide sources appeared thus
to be the limiting factor in pesticide exports from the catchment. The se-
quence of several runoff peaks, with the clockwise followed by anti-
clockwise runoff peaks was observed on both May 13, 2009 and October
20, 2009 (13/05/2009-1 followed by 13/05/2009-2 and 20/10/2009-1
 M. Lefrancq et al. / Science of the Total Environment 587–588 (2017) 75–86
followed by 20/10/2009-3 in Fig. 5). This sequence supports the hypoth-
esis that an exhaustion effect was present, i.e., the rapid mobilisation of
pesticides or suspended solids occurred during the first peak (transport
limitation), which limited the source during the second peak (supply
limitation) (Bieroza and Heathwaite, 2015; Bowes et al., 2009). Degra-
dation, and thus a supply limitation, can be evaluated for GLY in the
presence of AMPA. The relationship between AMPA and glyphosate
were evaluated by calculating %AMPA as a percentage of the molar
load of AMPA compared to the total molar loads of GLY and AMPA
(Imfeld et al., 2013). A gradual increase in %AMPA from the last applica-
tion was observed, indicating degradation of glyphosate (p b 0.05).
%AMPA generally exceeded 60%, except in April 2009, near the glypho-
sate application dates, and averaged 67.0 ± 19.3% across runoff events.
AMPA and GLY always followed the same hysteresis patterns; however,
%AMPA did not correlate significantly with ΔR or ΔC (p N 0.05).
TEB and TRI exhibited similar concentration patterns (p b 0.01), with
first flush calculations that differed significantly compared to the other
studied chemicals (p ≤ 0.05). These pesticides exhibited a predominant-
ly anticlockwise hysteresis pattern. Given that their sorption character-
istics fell within the same range as the other pesticides studied (Table
1), anti-clockwise patterns may be partly due to the application area,
which was mainly on the upstream section of plot A. Further location
of the application area may delay the pesticide arrival at the outlet of
the catchment (Hughes et al., 2012).
Hysteresis patterns for the different substances within an event or
for an individual substance across events were highly dynamic and
shifted between clockwise and anti-clockwise patterns. This may be
partly due to (i) the complexity of the studied outflow discharge,
which often had multiple peaks and indicated different flow pathways
within the catchment (transport limitation) and (ii) the complex inter-
play between the temporal and spatial evolution of the pesticide stocks
related to their application date, amount and mode (foliar or directly on
soil), as well as their degradation or their availability via sorption (sup-
ply limitation). Studied hysteresis patterns commonly found in the liter-
ature are based on groundwater-based catchments and nutrients
(Bieroza and Heathwaite, 2015; Dupas et al., 2015; Paul et al., 2015;
Roussiez et al., 2013; Taghavi et al., 2011). Few studies address surface
dominated catchments or organic pesticides such as in the present
study (Pietroń et al., 2015; Taghavi et al., 2011), which limited our abil-
ity to make direct comparisons.
5. Conclusion
High frequency sampling is certainly costly but enables the reliable
estimation of maximum pesticide concentration and fluxes. Further-
more, it reveals information about the underlying hydrological and
hydrochemical processes governing pesticide transport. Altogether,
the results highlight that (i) for all runoff events, the pesticide concen-
trations increased with outflow and significant pesticide export can
occur during a single event; (ii) when the TU of the pesticide mixture
was analysed, the European Uniform Principles for algae, invertebrates
and fish were regularly exceeded (15%) and FLU was responsible for
the majority of the toxicity (59–96%); (iii) random sampling may result
in an up to 30-fold underestimation of the TU for invertebrates obtained
using the maximum concentration, highlighting the important role of
the sampling methods for assessing peak exposure; (iv) no first flush
occurred, and the contribution of the pesticide loads from different sec-
tion along the hydrograph was mostly homogeneous; and (v) hysteresis
patterns were complex and highly dynamic. Individual events can be
interpreted in a particular way but not consistently given the complex
interactions between hydrology and reactive transport at the study
site. The primary limitation of the study was the lack of knowledge
about pesticide sources and availability in soils before each rainfall
event. This knowledge would help to better interpret hysteresis
patterns. Detailed off-site pesticide transport information may support
the design and adaptation of mitigation strategies and crop
85
management techniques. For example, here, the absence of an impor-
tant first flush phenomenon for all of the studied pesticides questions
the relevance of mitigation strategies based on the retention of the
first part of the runoff volume, such as small storm water wetland. Fur-
ther field studies that evaluate concentration-discharge patterns for
pesticides are needed to better understand the hysteresis behaviour of
pesticides and use it as a tool to predict the sources and pathways of
pesticides within agricultural catchments. Such an internal signature
for a catchment may help researchers to better understand pesticide
source availability, mobilisation and transport in runoff water.
Acknowledgements
This research was funded by the Pays de la Loire French region and
the Loire Bretagne Water Agency (contrat régional bassin versant:
CRBV1, CRBV2, CRBV3 et contrat territorial: Décision 2011D012, Déci-
sion 2012D005, Décision 2013D038), along with the wine growers of
France Agrimer, InterLoire and l'Institut Français de la Vigne et du Vin.
The authors wish to thank the farmers from the study site for their kind
contribution, Stephane Sourice for his invaluable help in the field,
Audrey Amiot for her contributions to our data collection efforts and
Etienne Neethling for his support related to the rainfall data. We would
like to thank Richard Coupe for proof-reading the manuscript.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.scitotenv.2017.02.022.
References
Abdo, N., Wetmore, B.A., Chappell, G.A., Shea, D., Wright, F.A., Rusyn, I., 2015. In vitro
screening for population variability in toxicity of pesticide-containing mixtures. Envi-
ron. Int. 85:147–155. http://dx.doi.org/10.1016/j.envint.2015.09.012.
Allinson, G., Allinson, M., Bui, A., Zhang, P., Croatto, G., Wightwick, A., Rose, G., Walters, R.,
2015. Pesticide and trace metals in surface waters and sediments of rivers entering
the Corner Inlet Marine National Park, Victoria, Australia. Environ. Sci. Pollut. Res.
23:5881–5891. http://dx.doi.org/10.1007/s11356-015-5795-6.
Arias-Estévez, M., López-Periago, E., Martínez-Carballo, E., Simal-Gándara, J., Mejuto, J.-C.,
García-Río, L., 2008. The mobility and degradation of pesticides in soils and the pollu-
tion of groundwater resources. Agric. Ecosyst. Environ. 123:247–260. http://dx.doi.
org/10.1016/j.agee.2007.07.011.
Assoumani, A., Lissalde, S., Margoum, C., Mazzella, N., Coquery, M., 2013. In situ applica-
tion of stir bar sorptive extraction as a passive sampling technique for the monitoring
of agricultural pesticides in surface waters. Sci. Total Environ. 463–464:829–835.
http://dx.doi.org/10.1016/j.scitotenv.2013.06.025.
Baartman, J.E.M., Temme, A.J.A.M., Veldkamp, T., Jetten, V.G., Schoorl, J.M., 2013. Exploring
the role of rainfall variability and extreme events in long-term landscape develop-
ment. Catena 109:25–38. http://dx.doi.org/10.1016/j.catena.2013.05.003.
Bende-Michl, U., Verburg, K., Cresswell, H.P., 2013. High-frequency nutrient monitoring to
infer seasonal patterns in catchment source availability, mobilisation and delivery.
Environ. Monit. Assess. 185:9191–9219. http://dx.doi.org/10.1007/s10661-013-
3246-8.
Bertrand-Krajewski, J.-L., Chebbo, G., Saget, A., 1998. Distribution of pollutant mass vs vol-
ume in stormwater discharges and the first flush phenomenon. Water Res. 32:
2341–2356. http://dx.doi.org/10.1016/S0043-1354(97)00420-X.
Bieroza, M.Z., Heathwaite, A.L., 2015. Seasonal variation in phosphorus concentration–dis-
charge hysteresis inferred from high-frequency in situ monitoring. J. Hydrol. 524:
333–347. http://dx.doi.org/10.1016/j.jhydrol.2015.02.036.
Botta, F., Fauchon, N., Blanchoud, H., Chevreuil, M., Guery, B., 2012. Phyt'Eaux Cités: appli-
cation and validation of a programme to reduce surface water contamination with
urban pesticides. Chemosphere 86:166–176. http://dx.doi.org/10.1016/j.
chemosphere.2011.10.005.
Bowes, M.J., House, W.A., Hodgkinson, R.A., Leach, D.V., 2005. Phosphorus–discharge hys-
teresis during storm events along a river catchment: the River Swale, UK. Water Res.
39:751–762. http://dx.doi.org/10.1016/j.watres.2004.11.027.
Bowes, M.J., Smith, J.T., Neal, C., 2009. The value of high-resolution nutrient monitoring: a
case study of the River Frome, Dorset, UK. J. Hydrol. 378:82–96. http://dx.doi.org/10.
1016/j.jhydrol.2009.09.015.
Boxall, A.B.A., Fogg, L.A., Ashauer, R., Bowles, T., Sinclair, C.J., Colyer, A., Brain, R.A., 2013.
Effects of repeated pulsed herbicide exposures on the growth of aquatic macro-
phytes. Environ. Toxicol. Chem. 32:193–200. http://dx.doi.org/10.1002/etc.2040.
Bundschuh, M., Goedkoop, W., Kreuger, J., 2014. Evaluation of pesticide monitoring strat-
egies in agricultural streams based on the toxic-unit concept — experiences from
long-term measurements. Sci. Total Environ. 484:84–91. http://dx.doi.org/10.1016/j.
scitotenv.2014.03.015.
86 M. Lefrancq et al. / Science of the Total Environment 587–588 (2017) 75–86
Butturini, A., Francesc, G., Jérôme, L., Eusebi, V., Francesc, S., 2006. Cross-site comparison
of variability of DOC and nitrate c–q hysteresis during the autumn–winter period in
three mediterranean headwater streams: a synthetic approach. Biogeochemistry
77:327–349. http://dx.doi.org/10.1007/s10533-005-0711-7.
Carvalho, R.N., Arukwe, A., Ait-Aissa, S., Bado-Nilles, A., Balzamo, S., Baun, A., Belkin, S.,
Blaha, L., Brion, F., Conti, D., Creusot, N., Essig, Y., Ferrero, V.E.V., Flander-Putrle, V.,
Furhacker, M., Grillari-Voglauer, R., Hogstrand, C., Jona, A., Kharlyngdoh, J.B., Loos,
R., Lundebye, A.-K., Modig, C., Olsson, P.-E., Pillai, S., Polak, N., Potalivo, M., Sanchez,
W., Schifferli, A., Schirmer, K., Sforzini, S., Sturzenbaum, S.R., Softeland, L., Turk, V.,
Viarengo, A., Werner, I., Yagur-Kroll, S., Zounkova, R., Lettieri, T., 2014. Mixtures of
chemical pollutants at European legislation safety concentrations: how safe are
they? Toxicol. Sci. 141:218–233. http://dx.doi.org/10.1093/toxsci/kfu118.
Cedergreen, N., 2014. Quantifying synergy: a systematic review of mixture toxicity stud-
ies within environmental toxicology. PLoS One 9, e96580. http://dx.doi.org/10.1371/
journal.pone.0096580.
Cold, A., Forbes, V.E., 2004. Consequences of a short pulse of pesticide exposure for surviv-
al and reproduction of Gammarus pulex. Aquat. Toxicol. 67:287–299. http://dx.doi.
org/10.1016/j.aquatox.2004.01.015.
Delpla, I., Baurès, E., Jung, A.-V., Thomas, O., 2011. Impacts of rainfall events on runoff
water quality in an agricultural environment in temperate areas. Sci. Total Environ.
409:1683–1688. http://dx.doi.org/10.1016/j.scitotenv.2011.01.033.
Descroix, L., Viramontes, D., Estrada, J., Gonzalez Barrios, J.-L., Asseline, J., 2007. Investigat-
ing the spatial and temporal boundaries of Hortonian and Hewlettian runoff in
Northern Mexico. J. Hydrol. 346:144–158. http://dx.doi.org/10.1016/j.jhydrol.2007.
09.009.
Donn, M.J., Barron, O.V., Barr, A.D., 2012. Identification of phosphorus export from low-
runoff yielding areas using combined application of high frequency water quality
data and MODHMS modelling. Sci. Total Environ. 426:264–271. http://dx.doi.org/
10.1016/j.scitotenv.2012.03.021.
Doppler, T., Lück, A., Camenzuli, L., Krauss, M., Stamm, C., 2014. Critical source areas for
herbicides can change location depending on rain events. Agric. Ecosyst. Environ.
192:85–94. http://dx.doi.org/10.1016/j.agee.2014.04.003.
Dupas, R., Gascuel-Odoux, C., Gilliet, N., Grimaldi, C., Gruau, G., 2015. Distinct export dy-
namics for dissolved and particulate phosphorus reveal independent transport mech-
anisms in an arable headwater catchment: decoupling of dissolved and particulate
phosphorus. Hydrol. Process. 29:3162–3178. http://dx.doi.org/10.1002/hyp.10432.
European Commission, 2011. Implementing Regulation (EC) No 1107/2009 of the Euro-
pean Parliament and of the Council as regards uniform principles for evaluation
and authorization of plant protection products. Off. J. Eur. L155, 127–175.
Fauvel, B., Cauchie, H.-M., Gantzer, C., Ogorzaly, L., 2016. Contribution of hydrological data
to the understanding of the spatio-temporal dynamics of F-specific RNA bacterio-
phages in river water during rainfall-runoff events. Water Res. 94:328–340. http://
dx.doi.org/10.1016/j.watres.2016.02.057.
Fovet, O., Ruiz, L., Hrachowitz, M., Faucheux, M., Gascuel-Odoux, C., 2015. Hydrological
hysteresis and its value for assessing process consistency in catchment conceptual
models. Hydrol. Earth Syst. Sci. 19:105–123. http://dx.doi.org/10.5194/hess-19-105-
2015.
Goulet, E., Dousset, S., Chaussod, R., Bartoli, F., Doledec, A.F., Andreux, F., 2006. Water-sta-
ble aggregates and organic matter pools in a calcareous vineyard soil under four soil-
surface management systems. Soil Use Manag. 20:318–324. http://dx.doi.org/10.
1111/j.1475-2743.2004.tb00376.x.
Granger, S.J., Hawkins, J.M.B., Bol, R., White, S.M., Naden, P., Old, G., Bilotta, G.S., Brazier,
R.E., Macleod, C.J.A., Haygarth, P.M., 2009. High temporal resolution monitoring of
multiple pollutant responses in drainage from an intensively managed grassland
catchment caused by a summer storm. Water Air Soil Pollut. 205:377–393. http://
dx.doi.org/10.1007/s11270-009-0083-z.
Gregoire, C., Payraudeau, S., Domange, N., 2010. Use and fate of 17 pesticides applied on a
vineyard catchment. Int. J. Environ. Anal. Chem. 90:406–420. http://dx.doi.org/10.
1080/03067310903131230.
Hudson, P.F., 2003. Event sequence and sediment exhaustion in the lower Panuco Basin,
Mexico. Catena 52:57–76. http://dx.doi.org/10.1016/S0341-8162(02)00145-5.
Hughes, A.O., Quinn, J.M., McKergow, L.A., 2012. Land use influences on suspended sedi-
ment yields and event sediment dynamics within two headwater catchments, Wai-
kato, New Zealand. N. Z. J. Mar. Freshw. Res. 46:315–333. http://dx.doi.org/10.1080/
00288330.2012.661745.
Imfeld, G., Lefrancq, M., Maillard, E., Payraudeau, S., 2013. Transport and attenuation of
dissolved glyphosate and AMPA in a stormwater wetland. Chemosphere 90:
1333–1339. http://dx.doi.org/10.1016/j.chemosphere.2012.04.054.
Jeschke, P., 2016. Progress of modern agricultural chemistry and future prospects: prog-
ress of modern agricultural chemistry and future prospects. Pest Manag. Sci. 72:
433–455. http://dx.doi.org/10.1002/ps.4190.
Le Bissonnais, Y., Blavet, D., De Noni, G., Laurent, J.-Y., Asseline, J., Chenu, C., 2007. Erodibil-
ity of Mediterranean vineyard soils: relevant aggregate stability methods and signif-
icant soil variables. Eur. J. Soil Sci. 58:188–195. http://dx.doi.org/10.1111/j.1365-
2389.2006.00823.x.
Lefrancq, M., Payraudeau, S., García Verdú, A.J., Maillard, E., Millet, M., Imfeld, G., 2014.
Fungicides transport in runoff from vineyard plot and catchment: contribution of
non-target areas. Environ. Sci. Pollut. Res. 21:4871–4882. http://dx.doi.org/10.1007/
s11356-013-1866-8.
Lewis, K.A., Tzilivakis, J., Warner, D.J., Green, A., 2016. An international database for pesti-
cide risk assessments and management. Hum. Ecol. Risk Assess. Int. J. 22:1050–1064.
http://dx.doi.org/10.1080/10807039.2015.1133242.
Liess, M., Von Der Ohe, P.C., 2005. Analyzing effects of pesticides on invertebrate commu-
nities in streams. Environ. Toxicol. Chem. 24:954. http://dx.doi.org/10.1897/03-652.1.
Lorenz, S., Rasmussen, J.J., Süß, A., Kalettka, T., Golla, B., Horney, P., Stähler, M., Hommel, B.,
Schäfer, R.B., 2016. Specifics and challenges of assessing exposure and effects of pes-
ticides in small water bodies. Hydrobiologia http://dx.doi.org/10.1007/s10750-016-
2973-6.
Louchart, X., Voltz, M., Andrieux, P., Moussa, R., 2001. Herbicide transport to surface wa-
ters at field and watershed scales in a Mediterranean vineyard area. J. Environ. Qual.
30:982. http://dx.doi.org/10.2134/jeq2001.303982x.
Maillard, E., Imfeld, G., 2014. Pesticide mass budget in a stormwater wetland. Environ. Sci.
Technol. 48:8603–8611. http://dx.doi.org/10.1021/es500586x.
Marttila, H., Kløve, B., 2010. Dynamics of erosion and suspended sediment transport from
drained peatland forestry. J. Hydrol. 388:414–425. http://dx.doi.org/10.1016/j.
jhydrol.2010.05.026.
Olsson, O., Khodorkovsky, M., Gassmann, M., Friedler, E., Schneider, M., Dubowski, Y.,
2013. Fate of Pesticides and Their Transformation Products: First Flush Effects in a
Semi-Arid Catchment. Soil Air Water 41:134–142. http://dx.doi.org/10.1002/clen.
201100545.
Outram, F.N., Cooper, R.J., Sünnenberg, G., Hiscock, K.M., Lovett, A.A., 2016. Antecedent
conditions, hydrological connectivity and anthropogenic inputs: factors affecting ni-
trate and phosphorus transfers to agricultural headwater streams. Sci. Total Environ.
545–546:184–199. http://dx.doi.org/10.1016/j.scitotenv.2015.12.025.
Papadakis, E.N., Vryzas, Z., Kotopoulou, A., Kintzikoglou, K., Makris, K.C., Papadopoulou-
Mourkidou, E., 2015. A pesticide monitoring survey in rivers and lakes of northern
Greece and its human and ecotoxicological risk assessment. Ecotoxicol. Environ. Saf.
116:1–9. http://dx.doi.org/10.1016/j.ecoenv.2015.02.033.
Paul, A., Moussa, I., Payre, V., Probst, A., Probst, J.-L., 2015. Flood survey of nitrate behav-
iour using nitrogen isotope tracing in the critical zone of a French agricultural catch-
ment. Compt. Rendus Geosci. 347:328–337 (Geochemical and isotopic record of
anthropogenic activities (Part 2)Traditional and non-traditional stable isotopes). 10.
1016/j.crte.2015.06.002.
Peters, K., Bundschuh, M., Schäfer, R.B., 2013. Review on the effects of toxicants on fresh-
water ecosystem functions. Environ. Pollut. 180:324–329. http://dx.doi.org/10.1016/
j.envpol.2013.05.025.
Pietroń, J., Jarsjö, J., Romanchenko, A.O., Chalov, S.R., 2015. Model analyses of the contribu-
tion of in-channel processes to sediment concentration hysteresis loops. J. Hydrol.
527:576–589. http://dx.doi.org/10.1016/j.jhydrol.2015.05.009.
Rabiet, M., Margoum, C., Gouy, V., Carluer, N., Coquery, M., 2010. Assessing pesticide con-
centrations and fluxes in the stream of a small vineyard catchment – effect of sam-
pling frequency. Environ. Pollut. 158:737–748. http://dx.doi.org/10.1016/j.envpol.
2009.10.014.
Rasmussen, J.J., McKnight, U.S., Loinaz, M.C., Thomsen, N.I., Olsson, M.E., Bjerg, P.L.,
Binning, P.J., Kronvang, B., 2013. A catchment scale evaluation of multiple stressor ef-
fects in headwater streams. Sci. Total Environ. 442:420–431. http://dx.doi.org/10.
1016/j.scitotenv.2012.10.076.
Roussiez, V., Probst, A., Probst, J.-L., 2013. Significance of floods in metal dynamics and ex-
port in a small agricultural catchment. J. Hydrol. 499:71–81. http://dx.doi.org/10.
1016/j.jhydrol.2013.06.013.
Salmon-Monviola, J., Moreau, P., Benhamou, C., Durand, P., Merot, P., Oehler, F., Gascuel-
Odoux, C., 2013. Effect of climate change and increased atmospheric CO2 on hydro-
logical and nitrogen cycling in an intensive agricultural headwater catchment in
western France. Clim. Chang. 120:433–447. http://dx.doi.org/10.1007/s10584-013-
0828-y.
Seeger, M., Errea, M.-P., Beguerıia, S., Arnaez, J., Martıi, C., Garcıia-Ruiz, J., 2004. Catchment
soil moisture and rainfall characteristics as determinant factors for discharge/
suspended sediment hysteretic loops in a small headwater catchment in the Spanish
pyrenees. J. Hydrol. 288:299–311. http://dx.doi.org/10.1016/j.jhydrol.2003.10.012.
Stehle, S., Knäbel, A., Schulz, R., 2012. Probabilistic risk assessment of insecticide concen-
trations in agricultural surface waters: a critical appraisal. Environ. Monit. Assess.
185:6295–6310. http://dx.doi.org/10.1007/s10661-012-3026-x.
Taghavi, L., Merlina, G., Probst, J.-L., 2011. The role of storm flows in concentration of pes-
ticides associated with particulate and dissolved fractions as a threat to aquatic eco-
systems - case study: the agricultural watershed of save river (Southwest of France).
Knowl. Manag. Aquat. Ecosyst. 06. http://dx.doi.org/10.1051/kmae/2011002.
Thompson, J.J.D., Doody, D.G., Flynn, R., Watson, C.J., 2012. Dynamics of critical source
areas: does connectivity explain chemistry? Sci. Total Environ. 435–436:499–508.
http://dx.doi.org/10.1016/j.scitotenv.2012.06.104.
Tilahun, S.A., Ayana, E.K., Guzman, C.D., Dagnew, D.C., Zegeye, A.D., Tebebu, T.Y., Yitaferu,
B., Steenhuis, T.S., 2016. Revisiting storm runoff processes in the upper Blue Nile
basin: the Debre Mawi watershed. Catena 143:47–56. http://dx.doi.org/10.1016/j.
catena.2016.03.029.
Ulrich, U., Dietrich, A., Fohrer, N., 2013. Herbicide transport via surface runoff during in-
termittent artificial rainfall: a laboratory plot scale study. Catena 101:38–49. http://
dx.doi.org/10.1016/j.catena.2012.09.010.
Vaj, C., Barmaz, S., Sørensen, P.B., Spurgeon, D., Vighi, M., 2011. Assessing, mapping and
validating site-specific ecotoxicological risk for pesticide mixtures: a case study for
small scale hot spots in aquatic and terrestrial environments. Ecotoxicol. Environ.
Saf. 74:2156–2166. http://dx.doi.org/10.1016/j.ecoenv.2011.07.011.
Verro, R., Finizio, A., Otto, S., Vighi, M., 2009. Predicting pesticide environmental risk in in-
tensive agricultural areas. II: screening level risk assessment of complex mixtures in
surface waters. Environ. Sci. Technol. 43:530–537. http://dx.doi.org/10.1021/
es801858h.
Zuecco, G., Penna, D., Borga, M., van Meerveld, H.J., 2016. A versatile index to characterize
hysteresis between hydrological variables at the runoff event timescale. Hydrol. Pro-
cess. 30:1449–1466. http://dx.doi.org/10.1002/hyp.10681.