Resistance to quinclorac and ALS-inhibitor herbicides
in Galium spurium is conferred by two distinct genes
L L VAN EERD, M D M C LEAN, G R STEPHENSON & J C HALL
Department of Environmental Biology, University of Guelph, Guelph, Ontario, Canada
Received 10 July 2003
Revised version accepted 14 May 2004
Summary
Classical Mendelian experiments were conducted to
determine the genetics and inheritance of quinclorac
and acetolactate synthase (ALS)-inhibitor resistance in
a biotype of Galium spurium. Plants were screened with
the formulated product of either quinclorac or the
ALS-inhibitor, thifensulfuron, at the field dose of 125
or 6 g active ingredient (a.i.) ha)1 respectively. Segregation in the F2 generation indicated that quinclorac
resistance was a single, recessive nuclear trait, based on
a 1 : 3 segregation ratio [resistant : susceptible (R : S)].
Resistance to ALS inhibitors was due to a single,
dominant nuclear trait, segregating in the F2 generation
in a 3 : 1 ratio (R : S). The genetic models were
confirmed by herbicide screens of F1 and backcrosses
between the F1 and the S parent. F2 plants that
Introduction
Following lack of control from treatment with triasulfuron, an acetolactate synthase (ALS)-inhibitor herbicide, Galium spurium L. (false cleavers) seeds were
collected from a field in central Alberta, Canada (Hall
et al., 1998). Greenhouse experiments confirmed that
this G. spurium biotype was resistant to several ALSinhibitor herbicides and coincidently to the auxinic
herbicide quinclorac (Hall et al., 1998). The field from
which resistant (R) seed was harvested had been sprayed
3 of 6 years with ALS inhibitors, but quinclorac had
never been used (Hall et al., 1998). By comparing the R
and susceptible (S) biotypes, the resistance was determined to be >6.7- and >14-fold for quinclorac and
ALS inhibitors respectively (Hall et al., 1998). The
mechanism of ALS-inhibitor resistance in the R biotype
was due to altered herbicide inhibition of the ALS
survived quinclorac treatment set seed and the resulting
F3 progeny were screened with either herbicide. Quinclorac-treated F3 plants segregated in a 1 : 0 ratio
(R : S), hence F2 progenitors were homozygous for
quinclorac resistance. In contrast, F3 progeny segregated into three ratios: 1 : 0, 3 : 1 and 0 : 1 (R : S) in
response to ALS-inhibitor treatment. This segregation
pattern indicates that their F2 parents were either
homozygous or heterozygous for ALS-inhibitor resistance. Therefore, there were clearly two distinct resistance mechanisms encoded by two genes that were not
tightly linked as demonstrated by segregation patterns
of the F3.
Keywords: weed, genetics, inheritance, auxinic, thifensulfuron, multiple-herbicide, recessive trait, quinclorac,
Galium spurium, resistant.
enzyme (EC 4.1.3.18) (Hall et al., 1998), which resulted
from a point mutation in the ALS gene that conferred an
amino acid alteration in the ALS enzyme (Horsman &
Devine, 2000). However, the mechanism of quinclorac
resistance in G. spurium is not known.
Quinclorac, a quinolinecarboxylic acid, is classified as
an auxinic herbicide with monocot activity. At low
concentrations, auxinic herbicides produce effects similar to the natural plant hormone auxin [indole-3-acetic
acid (IAA)]. At high concentrations, auxinic herbicides
are phytotoxic and produce a number of phenotypic
characteristics such as cell and internode elongation, as
well as leaf and stem epinasty/hyponasty, ultimately
leading to plant necrosis and death. These symptoms are
characteristic in susceptible dicots, such as G. spurium,
treated with a quinolinecarboxylic acid. Although
auxinic herbicides have been used for c. 60 years, their
mechanism of action is not fully understood. In some,
Correspondence: J C Hall, Department of Environmental Biology, University of Guelph, Guelph, Ontario, Canada, N1G 2W1. Tel: (+011)-519-8244120 ext. 52740; Fax: (+011)-519-837-0442; E-mail: [email protected]
Present address: L L Van Eerd, Ridgetown College, University of Guelph, Ridgetown, Ontario, Canada N0P 2C0
European Weed Research Society Weed Research 2004 44, 355–365
356 L L Van Eerd et al.
but not all (Valenzuela-Valenzuela et al., 2002), susceptible dicot species, auxinic herbicides induce
1-aminocyclopropane-1-carboxylic acid synthase activity, resulting in increased biosynthesis of ethylene,
abscisic acid and ultimately hydrogen peroxide, which
leads to cell necrosis (Hansen & Grossmann, 2000;
Grossmann et al., 2001). The actual target site of
quinclorac and other auxinic herbicides is not known.
Sulfonylurea herbicides, such as thifensulfuron, inhibit ALS, also known as acetohydroxyacid synthase
(AHAS), the first dedicated enzyme in the branchedchain amino acid biosynthetic pathway. The inhibition
of ALS by sulfonylurea herbicides results in plant death,
either by depletion of branched-chain amino acids or by
accumulation of toxic intermediates such as a-ketobutyrate and a-amino-butyrate.
Soon after its registration in the 1990s, resistance to
quinclorac was observed in monocot species, Digitaria
ischaemum L. (Schreb) Muhl. (Koo et al., 1994; Abdallah et al., 2004) and Echinochloa spp. (Lopez-Martinez
et al., 1997; Menezes et al., 2000). Despite the widespread use of the auxinic herbicides for almost six
decades, the incidences of auxinic herbicide resistance
have been rather minor: only 24 resistant weed species
reported worldwide (Heap, 2004). The mechanism of
resistance in most of the auxinic herbicide-resistant
species is not known. Absorption, translocation, root
exudation or metabolism were not responsible for auxinic herbicide resistance in several weed biotypes including Centaurea solstitialis L. (Valenzuela-Valenzuela
et al., 2001), Kochia scoparia L. (Cranston et al., 2001)
and Sinapis arvensis L. (Penuik et al., 1993). In the
aforementioned auxinic herbicide-resistant biotypes, the
mechanism of resistances is hypothesized to be due to
altered auxin binding to the target site or altered signal
transduction along the auxin pathway (Penuik et al.,
1993; Valenzuela-Valenzuela et al., 2001, 2002; Goss &
Dyer, 2003). The inheritances of only two auxinic
herbicide-resistant biotypes have been investigated.
Resistance to the auxinic herbicide picloram in
C. solstitialis is based on a recessive allele at a single,
nuclear locus (Sabba et al., 2003). In contrast, resistance
to the auxinic herbicide dicamba in an S. arvensis
biotype was due to a completely dominant trait at a
single, nuclear gene (Jasieniuk et al., 1995). The inheritance of quinclorac resistance has not been previously
characterized.
In contrast, weed resistance to ALS-inhibitor herbicides developed shortly after their introduction in 1982.
Of the 286 herbicide-resistant weed biotypes, 83 have
ALS-inhibitor resistance (Heap, 2004). In most ALSinhibitor-resistant species, resistance is a result of an
altered ALS enzyme because of a point mutation in the
ALS gene (Tranel & Wright, 2002; Heap, 2004), similar
to that characterized in this R G. spurium biotype (Hall
et al., 1998; Horsman & Devine, 2000).
Understanding the inheritance of herbicide resistance
provides insight into the likely speed at which resistance
may spread. Furthermore, if quinclorac and ALSinhibitor resistance genes are linked, molecular analyses
may provide insight into the mechanism of quinclorac
resistance and perhaps the mode of action of quinclorac
and auxinic herbicides in dicots. Consequently, conventional breeding and genetic techniques were conducted
to characterize the inheritance of both quinclorac and
ALS-inhibitor resistance in this G. spurium biotype.
Thifensulfuron was used as the representative ALSinhibitor herbicide because only one ALS-inhibitor
herbicide is necessary to demonstrate inheritance of
target site-based ALS-inhibitor resistance (MallorySmith et al., 1990; Tranel & Wright, 2002). Classical
Mendelian inheritance was studied to determine (i) the
number of genes controlling resistance to both herbicides, (ii) gene linkage, (iii) whether resistance alleles
were dominant or recessive, and (iv) whether the herbicide resistance locus/loci were nuclear or organellar.
Materials and methods
Growth of plants
Seeds were sown in Premier Promix, a peat-mossbased potting medium (Premier Horticulture, Red Hill,
PA, USA). After emergence, plants were thinned to one
plant per 450-mL pot. The plants were irrigated with
water as required and fertilized three to four times a
week with 20 : 20 : 20 (N : P : K) fertilizer (20 g L)1)
containing micronutrients. Plants were grown in a
controlled environment growth room maintained at
24/18 ± 1C day/night temperature with a 16-h photoperiod and an average relative humidity of 65%. The
irradiance level was constant at c. 400 lEinstein m)2 s)1.
Development of parental lines
In 1996, R G. spurium seeds from plants which survived
triasulfuron treatment were collected from a field in
central Alberta, Canada and known S seed were
collected from a field outside of Edmonton, Alberta,
Canada. Based on flower and seed morphology, both R
and S biotypes were identified by Dorothy Fabian of the
Vascular Plant Herbarium, University of Alberta,
Edmonton, Canada as G. spurium L. and not G. aparine
L. (common cleavers) (LM Hall & D Fabian, pers.
comm.). G. spurium is a diploid (2n ¼ 20) and no
polyploidy has been observed (Malik & Vanden Born,
1988). In contrast, G. aparine is typically polyploidy with
the hexaploid (2n ¼ 66) most often observed (Malik &
European Weed Research Society Weed Research 2004 44, 355–365
Herbicide resistance in Galium spurium 357
Vanden Born, 1988). Although, very little is known
about the reproduction biology of G. spurium, generally,
the Galium spp. are considered to be predominately selfpollinating (Ehrendorfer, 1971; Faegri & Van den Pijl,
1971; Moore, 1975; Malik & Vanden Born, 1988), but
the level of out-crossing in G. spurium is not known.
Plants were grown for seed production in the
greenhouse in Alberta. At the University of Guelph,
Canada, four to five breeding lines were established
from the Alberta seed stock by subjecting R and S plants
to two more generations of self-pollination. To ensure
self-pollination, the entire plant was placed in a glassine
pollination bag (Cryovac, Mississauga, ON, Canada)
approximately 6 weeks after planting and prior to
flowering. Plants readily self pollinated, with >1000
seeds produced per plant. Plants, which developed from
seed produced after the second self-pollination, were
used for reciprocal crosses to produce F1 progeny, or
sprayed with either quinclorac or thifensulfuron as
described below.
Genetic crosses and production of F1, F2, F3,
and backcross generations
Reciprocal crosses were performed with individual R
and S plants from the parental lines. G. spurium flowers
were 1–1.5 mm in diameter, with four anthers and two
pistils, with two seeds per flower. For detailed descriptions of flower morphology see Malik and Vanden Born
(1988). To produce F1 progeny, individual axillary
flower peduncles were selected. All open flowers and
immature flower buds were physically removed. Under
microscopic examination, unopened flowers were dissected with ultra fine forceps. All four anthers were
removed without damaging the two pistils and ovary.
The emasculated flowers on individual axillary peduncles were wrapped and sealed in a 5 cm · 12 cm glassine
bag to prevent unwanted pollination, and the entire
plant was enclosed in another glassine bag. The second
day after emasculation, pollen from the male parent
plant was placed on the receptive stigmas. Any nonpollinated flowers and immature flower buds were
removed prior to resealing each individual axillary
flower peduncle in a small pollination bag and enclosing
the entire plant in another pollination bag. Seeds were
allowed to develop to maturity on the plant inside the
pollination bags. Terminal flower stems were not used
for crosses because of the high density of flowers at
various stages of development, including dormant flower
buds. To ensure that pollination was controlled (i.e.
selfing was zero), flowers were emasculated but not
pollinated.
F1 seed was germinated and grown to maturity to
produce plants for F2 seed production, for backcrosses
European Weed Research Society Weed Research 2004 44, 355–365
with S parents (designated F1BC1), or for screening with
either herbicide. Ten of 149 germinating F1 plants were
grown to flower and self-pollinated to produce F2 seed.
Six pairs of backcrosses between F1 hybrids and S
parents were performed as previously described.
F2 plants at approximately the three-whorl stage of
development were screened with either herbicide as
described below. Fourteen days after herbicide treatment, plant survival was determined and plants were
classified as R or S by comparing their responses with
those of herbicide-treated R and S parental plants.
Twelve F2 plants that survived quinclorac treatment
were grown to maturity and self-pollinated to produce
F2-derived F3 progeny. F3 seeds were sown and grown
to approximately the three-whorl stage of development,
treated with either herbicide and classified as R or S as
described elsewhere.
Herbicides used
Quinclorac [Accord 75 DF, 75% (wt/wt) a.i., DF; BASF
Canada] was applied with Merge (BASF Canada), an
adjuvant containing a 1 : 1 surfactant and petroleum
hydrocarbon solvents blend. Thifensulfuron [Pinnacle
75 DF, 75% (wt/wt) a.i., DF; Dupont Canada] was used
with non-ionic surfactant Agral 90 (Dupont Canada),
which contains 90% nonylphenoxy polyethyoxy
ethanol.
Dose–response experiments
Dose–response experiments were conducted to determine
the appropriate dose for use in inheritance studies.
G. spurium were sprayed at approximately the threewhorl stage of foliar development with the formulated
product and recommended adjuvant of quinclorac or
thifensulfuron at doses ranging from 10.4 to 1500 g a.i.
1
ha)1 (12
· to 12· the recommended field dose of
1
125 g a.i. ha)1) or 0.38 to 96 g a.i. ha)1 (16
· to 16· the
)1
recommended field dose of 6 g a.i. ha ) respectively.
Spray applications were conducted with a motorized
hood sprayer equipped with a flat-fan nozzle (TeeJet,
80015E, Spraying Systems, Guelph, Canada) calibrated
to deliver 110 L ha)1 of spray solution at 250 kPa.
Quinclorac- and thifensulfuron-treated plants were harvested 14 or 21 days after treatment (DAT), respectively,
by severing the shoot at the soil level. Percentage survival
and shoot dry weights were determined.
Herbicide treatment of F1, F2, F3, and F1BC1
generations
Plants were screened with only one herbicide, either
quinclorac or thifensulfuron, at the recommended field
358 L L Van Eerd et al.
dose of 125 or 6 g a.i. ha)1 respectively. Plants were
sprayed at approximately the three-whorl stage of foliar
development. Spray applications were conducted as
previously described. Fourteen DAT, plant survival
was determined and plants were classified as either R or
S by comparing their phytotoxic response with that of
herbicide-treated R and S parental seedlings.
Statistical analysis
Dose–response experiments were conducted as completely randomized designs with at least five replications
per treatment and each experiment repeated once. Shoot
dry-weight data were expressed as a percentage of the
mean of the non-treated control prior to statistical
analysis with the general linear model (PROC GLM)
using SAS 8.02 software (SAS Institute, Cary, NC,
USA). Shoot dry-weight data were expressed as a
nonlinear model and the GR50 [dose (g a.i. ha)1)
required to reduce shoot growth by 50% compared
with the non-treated control] values were calculated with
the nonlinear model (PROC NLIN) using SAS 8.02
software (SAS Institute, Cary, NC, USA).
For each herbicide, plant mortality data was expressed as a generalized linear model (GLM) with a
binomial response distribution and a probit link function (Bailer & Oris, 1997). Lethal concentration
estimates (LD50; the lethal dose (g a.i. ha)1) to 50% of
the plants tested) were obtained by solving the generated
model for herbicide dose.
For each herbicide, R and S phenotype frequencies
were tabulated for F1, F2, F3, and F1BC1 generations.
Goodness-of-fit to specific genetic ratios was determined
using the chi-square (v2) test. The Yates correction
term was used when there was only one degree of
freedom. Homogeneity tests, using (v2) values without
the Yates correction term, were conducted and data
were pooled, where possible. For genetic models, twogene loci were considered: one locus for the quinclorac
response (Qq) and the other for the thifensulfuron
response (Tt), with Q and T being dominant over q and t
respectively. The type I error rate (a) was 0.05 for all
statistical tests.
These results indicated that the breeding lines were
homozygous for both quinclorac and ALS-inhibitor
response.
Dose–response
The R biotype was resistant to quinclorac; LD50 values
could not be calculated because there was no mortality
at any dose tested (10.4–1500 g a.i. ha)1) (Fig. 1A). The
calculated GR50 value for the R biotype treated with
quinclorac was 266.25 g a.i. ha)1. The calculated LD50
and GR50 values for the quinclorac-treated S biotype
were 48.75 and 6.25 g a.i. ha)1 respectively. At the
lowest dose tested (10.4 g a.i. ha)1), there was a significant reduction in the shoot biomass of the S biotype
compared with that of the non-treated control (Fig. 1C).
Symptoms of quinclorac phytotoxicity in the S biotype
included leaf hyponasty, reduced leaf area, chlorosis,
necrosis, and plant death. Furthermore, despite reduced
shoot growth in R plants treated with 1500 g a.i. ha)1 of
quinclorac, there were no phytotoxic symptoms other
than minor chlorosis of some leaf tips.
The R G. spurium biotype was also resistant to
thifensulfuron (Fig. 1B,D); LD50 and GR50 values could
not be calculated for the R biotype because there was no
mortality or significant dry weight reduction at any dose
tested (0.38–96 g a.i. ha)1) (Fig. 1B,D). The calculated
LD50 and GR50 values for the S biotype were 15.48 and
0.60 g a.i. ha)1 of thifensulfuron respectively. Symptoms of thifensulfuron phytotoxicity in the S biotype
included inhibition of shoot growth, chlorosis, necrosis,
and plant death. Despite up to a 60% reduction in shoot
growth in R plants treated with very high doses of
thifensulfuron, there were no observed phytotoxic
symptoms 21 DAT.
The recommended field doses of 125 or 6 g a.i. ha)1 of
quinclorac or thifensulfuron, respectively, were selected
for screening of F1, F2, F3, and F1BC1 plants. These doses
were used because they represent the LD95 and GR95
for the S biotype treated with quinclorac and thifensulfuron respectively. The LD95 and GR95 dose clearly
differentiates between R and S individuals (Fig. 1).
F1
Results
After each generation of controlled self-pollination, all
tested R progeny (n ¼ 178) were resistant and all tested
S progeny (n ¼ 175) were susceptible to both herbicides
applied separately at the field dose (data not shown).
Furthermore, no S plant was observed in the parental R
seed stock and vice versa, which was used in our dose–
response experiments (n ¼ 270) and in dose–response
experiments conducted by Hall and colleagues (1998).
The generation time, from seed to seed, was c. 4 months.
Self- and cross-pollinated seeds readily germinated and
no seed dormancy was observed. Based on emasculation
without pollination, the pollination procedure was
deemed successful because no seeds were produced as
a result of uncontrolled pollination. F1 seed was sown
and grown to generate the F2 population or used for
backcrossing with S parental plants to produce F1BC1
progeny. The remaining F1 seeds were used in herbicide European Weed Research Society Weed Research 2004 44, 355–365
Herbicide resistance in Galium spurium 359
100
Resistant
Resistant
Survival (%)
80
60
40
20
Susceptible
0
Susceptible
a
c
Shoot dry weight
(% of untreated control)
100
80
Resistant
60
Resistant
40
Susceptible
20
Susceptible
c
d
0
0
7.8
15.6 31.25 62.5 125
250
500 1000 1500 0
Quinclorac Dose (g a.i. ha–1)
0.38 0.75
2
3
6
12
24
48
96
192
Thifensulfuron Dose (g a.i. ha–1)
Fig. 1 Response characterized by percent survival (A, B) and reduction in shoot dry weight (C, D) of herbicide-resistant and -susceptible
Galium spurium treated with the formulated products of either quinclorac (left) or thifensulfuron (right) and harvested 14 or 21 days
after treatment respectively. Symbols represent sample means with standard error bars. Where no standard error bar is shown the
standard error is smaller than the symbol.
response experiments. F1 plants treated with 125 g a.i.
ha)1 of quinclorac segregated in a 0 : 1 ratio (R : S;
v2 ¼ 2.05 and P ¼ 0.15; data not shown) suggesting
that quinclorac resistance may be a recessive trait. Of
the 45 F1 plants sprayed with quinclorac, there was only
one plant with an intermediate phenotype 21 DAT.
There were no observed thifensulfuron-treated F1
plants (n ¼ 35) with an intermediate phenotype. At
the recommended field dose, thifensulfuron-treated F1
plants segregated in a 1 : 0 ratio (R : S; v2 ¼ 0.86 and
P ¼ 0.35; data not shown) indicating that thifensulfuron
and thus ALS-inhibitor resistance may be a dominant
trait.
Dominance versus co-dominance of herbicide resistance based on herbicide dose–response experiments on
the F1 could not be investigated in G. spurium plants
because: (i) seed production was severely limited because
of the laborious nature of crossing, i.e. only two seeds
per 1.5 mm flower, (ii) there were no phenotypical
markers for resistance (i.e. there were no phenotypic
differences between the R and S biotypes); therefore, it
European Weed Research Society Weed Research 2004 44, 355–365
was not possible to distinguish between successful F1
crosses and self-pollinated individuals without spraying
and thereby sacrificing S plants and (iii) in cases where
herbicide resistance is a recessive trait, treating F1 plants
would be ineffective because the F1 heterozygote would
have a susceptible phenotype and all F1 plants would die
as a result of herbicide treatment. Therefore, non-treated
F1 G. spurium plants were self-pollinated to produce
F2 progeny.
F2
There were no inheritance differences between reciprocal
crosses for either herbicide, suggesting that nuclear
gene(s) were involved (Table 1). With respect to quinclorac, segregation in the F2 generation did not differ
from a 1 : 3 ratio (R : S; pooled v2 ¼ 0.10 and P ¼
0.75; Table 1) indicating that quinclorac resistance at
125 g a.i. ha)1 in G. spurium was controlled by a single,
recessive nuclear gene system. In the F2 generation
sprayed with the field dose of thifensulfuron, segregation
360 L L Van Eerd et al.
Table 1 R and S segregation in Galium spurium F2 generation in response to treatment with quinclorac or thifensulfuron*
Quinclorac
Cross $ · #
F2 designation
R2 · S2
R3 · S3
S1 · R1
S1 · R1
S1 · R1
S3 · R3
S3 · R3
R3 · S3
R2 · S2
R3 · S3
1
2
3
4
5
6
7
8
9
10
Thifensulfuron
2 No. of plants
v
R:S§
1:3 expected
6:21
3:30
11:23
5:31
10:24
22:58
10:23
10:27
13:61
25:59
P
No. of plants
v2 R:S§
3:1 expected
P 0.11
4.45
0.98
2.37
0.35
0.27
0.49
0.08
2.18
1.02
0.74
0.03
0.32
0.12
0.55
0.60
0.48
0.78
0.14
0.31
21:6
21:15
25:9
23:10
23:11
63:19
26:14
30:7
48:20
60:15
0.11
5.33
0.04
0.49
0.98
0.11
2.13
0.73
0.70
1.00
0.74
0.02
0.84
0.48
0.32
0.74
0.14
0.39
0.40
0.32
12.30
0.10
12.20
0.26
0.75
0.20
340:126
11.62
1.03
10.59
0.31
0.31
0.30
Deduced
F1 genotypeà
Qq
Qq
Qq
Qq
Qq
Qq
Qq
Qq
Qq
Qq
Tt
Tt
Tt
Tt
Tt
Tt
Tt
Tt
Tt
Tt
Homogeneity Chi Square: All populations
Total
Pooled
Homogeneity
d.f. ¼ 10
d.f. ¼ 1
d.f. ¼ 9
115:357
*Fourteen days after application of the formulated product of quinclorac or thifensulfuron at the recommended field dose of 125 or
6 g a.i. ha)1 respectively.
Chi square values (v2) with associated probabilities (P) were the result of tests for goodness of fit to a 1:3 and 3:1 (R:S) segregation ratio for
quinclorac and thifensulfuron respectively. The type I error rate (a) was 0.05 for all statistical tests.
à
For genetic models, two gene loci were considered; one locus for the quinclorac response (Qq) and one for the thifensulfuron response (Tt),
with Q and T being dominant over q and t respectively.
§
Resistant (R) and susceptible (S) classification was based on comparison with herbicide-treated R and S parental plants.
did not differ from a 3 : 1 ratio (R : S; pooled v2 ¼ 1.03
and P ¼ 0.31; Table 1). These results indicate that
thifensulfuron resistance at the field dose in G. spurium
was conferred by a single, dominant nuclear allele.
orac did not differ from the expected 0 : 1 (R : S) ratio
(Fig. 2; Table 2). These results confirm that quinclorac
resistance at the field dose in G. spurium was due to a
recessive trait at a single, nuclear locus. Thifensulfurontreated F1BC1 plants segregated in the expected 1 : 1
(R : S) ratio (Fig. 2; Table 2), providing further evidence that ALS-inhibitor resistance was a single, dominant nuclear trait when treated at the field dose.
Moreover, the segregation pattern of reciprocal
F1BC1
Segregation among the progeny of backcrosses between
F1 hybrids and the S parental line treated with quincl-
S parent (QQ tt)
Alleles
Q
F1 (Qq Tt)
t
Q T
QQ Tt
S-quinclorac, R-ALS
q T
Qq Tt
S-quinclorac, R-ALS
Q t
QQ tt
S-quinclorac, S-ALS
q t
Qq tt
S-quinclorac, S-ALS
Genotype
Expected phenotypic ratios:
Phenotype
Quinclorac 0:1 (R:S)
ALS 1:1 (R:S)
Fig. 2 Segregation pattern of a two-gene model for backcrosses between F1 hybrids and S parental Galium spurium plants (F1BC1)
treated with the formulated product of either quinclorac or thifensulfuron at 125 or 6 g a.i. ha)1 respectively. For the genetic model,
there was one locus for the quinclorac response (Qq) and one for the thifensulfuron response (Tt), with Q and T being dominant over q
and t respectively. ÔSÕ indicates susceptible response and ÔRÕ indicates resistance response to the herbicide.
European Weed Research Society Weed Research 2004 44, 355–365
Herbicide resistance in Galium spurium 361
Table 2 R and S segregation in Galium spurium progeny among backcrosses between F1 hybrids and S parents (F1BC1) in response to
treatment with quinclorac or thifensulfuron F1 backcrosses $ · #
Genotype
of parents
S · (R · S)
S · (S · R)
(R · S) · S
(S · R) · S
Total
Pooled
Homogeneity
QQtt · QqTt
QQtt · QqTt
QqTt · QQtt
QqTt · QQtt
d.f. ¼ 3
d.f. ¼ 1
d.f. ¼ 2
Quinclorac
Thifensulfuron
No. of plants
No. of plants
v2à
R:S§
R:S§
1:1 expected
0:12–
0:5
0:15
0:16
3:7
4:2
8:6
7:7
18:20
1.6
*
0.28
0
1.88
0.10
1.78
Pà
0.20
*
0.59
1
0.60
0.74
0.41
*Chi-squared test (v2) requires a minimum of 5 in the smallest class; the test could not be performed.
Fourteen days after application of the formulated product of quinclorac or thifensulfuron at the recommended field dose of 125 or
6 g a.i. ha)1 respectively.
à
Chi-square values (v2) with associated probabilities (P) were the result of tests for goodness-of-fit to a 1:1 (R:S) segregation ratio for
thifensulfuron. The type I error rate (a) was 0.05 for all statistical tests.
§
Resistant (R) and susceptible (S) classification was based on comparison to herbicide-treated R and S parental plants.
–
A 0:1 ratio was expected. Chi-squared tests were not possible because there is only one class.
backcrosses [i.e. S · (R · S), S · (S · R), (R · S) · S,
(S · R) · S] provide further evidence that the locus/loci
for resistance were located in the nucleus for both
herbicides (Table 2).
thifensulfuron, further supporting the genetic model that
ALS-inhibitor resistance in G. spurium was conferred by
a single, dominant nuclear allele.
Discussion
F3
To test the two-gene model, F2 plants that survived
quinclorac treatment were self-pollinated and allowed to
set seed. The resulting F2-derived F3 progeny were
screened for herbicide resistance. As expected, F3 plants
sprayed with 125 g a.i. ha)1 of quinclorac segregated in
a 1 : 0 ratio (R : S; v2 ¼ 0.03 and P ¼ 0.86; Table 3).
Thus, the quinclorac-resistant F2 progenitors were
homozygous at the quinclorac-resistance locus and
had a genotype of qq. This provides further support
for the genetic model that at the field dose quinclorac
resistance was mediated by a single, recessive nuclear
allele in G. spurium. In the 12 F2-derived F3 populations
tested, thifensulfuron response segregated into three
R : S ratios, 1 : 0, 3 : 1, and 0 : 1, indicating three F2
parental genotypes of TT, Tt, and tt, respectively,
providing further evidence that ALS inhibitor resistance
was mediated by a single, dominant nuclear allele
(Table 3).
Because the F2 progenitors were homozygous for
quinclorac resistance (qq) and either homozygous
(TT or tt) or heterozygous (Tt) for ALS-inhibitor
resistance, resistance to each herbicide was controlled
by two separate loci, which were not tightly linked.
Six F3 families (designated 1c, 2e, 2f, 3g, 3h, and 4j)
segregated in a 3 : 1 ratio (R : S; pooled v2 ¼ 0.65 and
P ¼ 0.42; Table 3) when sprayed with the field dose of
European Weed Research Society Weed Research 2004 44, 355–365
The R G. spurium biotype had one gene system
controlling quinclorac resistance and another one for
ALS-inhibitor resistance. The inheritance of ALS-inhibitor resistance in G. spurium was similar to all other
characterized target-site based ALS-inhibitor-resistant
plant species, which is due to a dominant allele at a
single, nuclear locus. However, based on other researchersÕ results, the degree of dominance can vary from
incomplete to complete (Mallory-Smith et al., 1990;
Tranel & Wright, 2002). In contrast, quinclorac
resistance at the field dose in G. spurium was mediated
by a recessive allele at a single, nuclear gene. The
inheritance of quinclorac resistance in various monocot
species has not been characterized.
The use of the F3 to identify two separate herbicide
resistance mechanisms in one plant was a novel
approach. Many other options were explored including foliar application of two herbicides at once on
one plant, but this approach would only identify
individuals that were resistant to both herbicides.
Likewise, the method of treating numerous plants with
one herbicide followed by treatment of survivors with
the other herbicide 2 weeks later would only provide
information on plants that were resistant to the initial
herbicide. With sequence treatments there would be
the potential for synergism of the two herbicides or
age-related effects of the plants. Plant clones via
362 L L Van Eerd et al.
Table 3 R and S segregation in Galium spurium F2-derived Fy3 generation in response to treatment with quinclorac or thifensulfuronà
Quinclorac
Thifensulfuron
–
F1 Cross $ · #
F3 designation§
Expected
ratio
R:S
(No. of plants)
Expected
ratio
R2 · S2
1a
1b
1c
2d
2e
2f
3g
3h
3i
4j
4k
4l
1:0àà
1:0
1:0
1:0
1:0
1:0
1:0
1:0
1:0
1:0
1:0
1:0
19:2
30:0
13:0
28:1
25:0
20:0
24:0
11:0
37:0
20:0
35:0
26:0
0:1àà
0:1
3:1**
1:0
3:1
3:1
3:1
3:1
0:1
3:1
0:1
0:1
R3 · S3
S1 · R1
S1 · R1
Homogeneity chi-square
Total
Pooled
Homogeneity
R:S–
(No. of plants)
0:39
0:56
10:3
43:0
38:9
36:10
31:14
20:10
0:33
23:13
0:50
0:35
v2**
P**
*
*
0.86
0.26
0.90
1.11
0.36
0.61
0.34
0.29
2.37
0.12
5.50
0.65
4.85
0.36
0.42
0.3
Deduced F2
genotype qq
qq
qq
qq
qq
qq
qq
qq
qq
qq
qq
qq
tt
tt
Tt
TT
Tt
Tt
Tt
Tt
tt
Tt
tt
tt
Segregating families: 2e, 2f, 3g, 3h, 4j
d.f. ¼ 5
d.f. ¼ 1
d.f. ¼ 4
148:56
*Chi-squared test (v2) requires a minimum of 5 in the smallest class; the test could not be performed.
F3 generation was derived from F2 progenitors, which survived quinclorac treatment.
à
Fourteen days after application of the formulated product of quinclorac or thifensulfuron at the recommended field dose of 125 or
6 g a.i. ha)1 respectively.
§
The different letters in the F3 designation indicate different F2 progenitors. F3 populations designated with the same number were generated
from the same F2 population, which resulted from the F1 cross indicated in the table.
–
Resistant (R) and susceptible (S) classification was based on comparison with herbicide-treated R and S parental plants.
**Chi-square values (v2) with associated probabilities (P) were the result of tests for goodness-of-fit to a 3:1 (R:S) segregation ratio for
thifensulfuron. The type I error rate (a) was 0.05 for all statistical tests.
For genetic models, two gene loci were considered; one locus for the quinclorac response (Qq) and one for the thifensulfuron response (Tt),
with Q and T being dominant over q and t respectively.
àà
Chi-squared tests were not possible where 1:0 and 0:1 ratios were expected because there is only one class.
cuttings and in vitro multiplication was investigated
but G. spurium was not easily amenable to cloning and
the few acceptable clones that were produced quickly
flowered. Furthermore, there was a concern that the
response of clones to auxinic herbicide treatment
might be affected by the propagation techniques,
particularly because herbicide efficacy changes with
plant age, especially with auxinic herbicides. In retrospect, a non-destructive test would have been an
alternative approach to testing two herbicides in one
plant. For instance, to determine acetyl-CoA carboxylase inhibitor resistance or susceptibility 3 lL of
1.05 mM diclofop-methyl solution was spotted onto
Avena fatua L. leaves (Seefeldt et al., 1998). If necrosis
occurred, then the plant was classified as susceptible.
A similar test could be modified for G. spurium to
characterize response to thifensulfuron based on the
Ôspot testÕ. In the same plant, whole-plant response to
quinclorac could be determined by foliar application a
few days after the spot test. Regardless, the relatively
quick generation time of G. spurium facilitated the use
of the F3.
It has been generally accepted that herbicide resistance is mainly controlled by single, nuclear dominant
or co-dominant alleles (Jasieniuk et al., 1996). However, it is evident from the increasing discoveries of
recessive herbicide resistance, such as trifluralin resistant Setaria viridis (L.) Beauv. (Jasieniuk et al., 1994),
triallate resistant A. fatua (Kern et al., 2002), dinitroanaline resistant Eleusine indica (L.) Gaertn. (Zeng &
Baird, 1999), auxinic herbicide resistant C. solstitialis
(Sabba et al., 2003) and now quinclorac-resistant
G. spurium, that recessive alleles do confer herbicide
resistance. All the aforementioned weed species with
recessive herbicide resistance, with the exception of
C. solstitialis, are predominantly self-pollinating species.
A self-pollinating breeding system facilitates the accumulation of a recessive trait within these populations
(Jasieniuk et al., 1996). Although the rate of outcrossing of G. spurium is not known, G. spurium is a
predominately self-pollinating species (Malik & Vanden
Born, 1988), which provides support for the accumulation of R individuals without quinclorac selection
pressure.
European Weed Research Society Weed Research 2004 44, 355–365
Herbicide resistance in Galium spurium 363
Even when the resistance trait is a dominant allele at
a single nuclear locus, as in acetyl-CoA carboxylase
inhibitor resistant A. fatua, a predominantly selfpollinating species, the observed rate of outcrossing
when grown in a Triticum aestivum L. crop is only
0.05% (Murray et al., 2002). Likewise, the spread of
quinclorac resistance in G. spurium through pollenmediated gene flow is expected to be minimal. Therefore,
it is imperative that control of this R G. spurium biotype
focus on containment and eradication of its seed.
Provided that seed of the R biotype is contained within
the field where it is located, the likelihood for widespread quinclorac resistance in other G. spurium populations throughout the Canadian prairies is likely to be
low for the following reasons: (i) viability of seed is
typically 1–3 years (Malik & Vanden Born, 1988), (ii)
the resistance trait is recessive, and (iii) pollen-mediated
gene flow is expected to be minimal because of the
predominantly self-pollinating nature of G. spurium. In
fact, numerous G. spurium populations were screened for
herbicide resistance, but no others had quinclorac
resistance (LM Hall, pers. comm.). Moreover, resistance
in other auxinic herbicide-resistant species has not
become widespread, regardless of the inheritance mechanism (LM Hall, pers. comm.; Cranston et al., 2001;
Sabba et al., 2003; Heap, 2004). Consequently, there
is little evidence to suggest that quinclorac-resistant
G. spurium will be a problem in the Canadian prairies or
elsewhere. Nonetheless, good agronomic management
practices should be maintained to avoid resistance
development.
Previously, it was generally assumed that for expression of auxinic herbicide resistance, there would need to
be mutations at several loci. With the accumulation of
information on the inheritance of auxinic herbicide
resistance (Jasieniuk et al., 1995; Sabba et al., 2003)
including quinclorac-resistant G. spurium, it is evident
that a mutation at a single nuclear locus is adequate to
cause the expression of auxinic herbicide resistance.
Therefore, it is unlikely that auxinic herbicides have
multiple mechanisms of action. The single gene product
that controls quinclorac resistance in G. spurium and
auxinic herbicide resistance in S. arvensis (Jasieniuk
et al., 1995) and C. solstitialis (Sabba et al., 2003) is
hypothesized to be either an auxin target-site protein or
a protein along the auxin signal transduction pathway.
When herbicide resistance to two or more target-site
chemistries is caused by more than one genetic mutation,
it is considered multiple resistance, as is the case with
this R G. spurium biotype. There are a few weed biotypes
with multiple-herbicide resistance such as Lolium spp.
(Preston et al., 1996; Kuk et al., 2000), A. fatua (Nandula & Messersmith, 2003), K. scoparia (Foes et al.,
1999) and Amaranthus powellii S. Wats. (Diebold et al.,
European Weed Research Society Weed Research 2004 44, 355–365
2003). Moreover, there are at least four other additional
weed biotypes with resistance to the auxinic herbicides
and the ALS inhibitors, including Papaver rhoeas L. in
Spain, K. scoparia in Montana, USA, as well as
Limnophila erecta Benth. and Limnocharis flava (L.)
Buchenau in Malaysia (Heap, 2004). The inheritance of
auxinic herbicide resistance is not known in these
aforementioned biotypes. The widespread use of auxinics and ALS inhibitors has likely lead to the development of these biotypes (Heap, 2004). However, in
G. spurium, although ALS-inhibitor use was somewhat
intensive, quinclorac had never been used before.
If one considers mutation rates commonly observed in
plants at typical field densities of G. spurium, it is obvious
that even with no selection pressure from frequent
quinclorac use, the occurrence of at least one plant with
resistance to both ALS-inhibitors and quinclorac is very
likely (Table 4). For example, if one assumes a G. spurium
density of 5 plants m)2 (Leeson et al., 2002) in a 30-ha
field and a mutation rate of 1 · 10)6 gametes per locus
per generation for both quinclorac and ALS-inhibitor
resistance in a predominantly self-pollinating species
(Jasieniuk et al., 1996), the probability of occurrence of
one G. spurium plant with resistance to both herbicides
would be 58% (Table 4). However, these tabulated
probabilities do not take into account initial frequencies
of resistance alleles, selection pressure because of frequent applications of ALS-inhibitor herbicides, weed
fitness, and gene flow, which may increase the probability
of occurrence of a plant with resistance to both quinclorac and ALS-inhibitor herbicides. Thus, although
quinclorac was not previously used on the field where
the R biotype was located, it is conceivable, based on
mutation rates, weed densities and inheritance, that two
distinct resistance mechanisms may arise in one plant
without selection pressure to both herbicides. Furthermore, R individuals will rapidly accumulate in the field
because G. spurium is predominately self-pollinated, it is
usually found in the field in high abundance, i.e. >10
plants m)2 (Leeson et al., 2002), and one plant typically
produces 3500 seeds (Malik & Vanden Born, 1988).
With regard to the development of quinclorac
resistance without herbicide selection, it may be that
the gene controlling quinclorac resistance in G. spurium is
particularly prone to spontaneous mutations. Perhaps
Ônaturally occurringÕ resistant populations of G. spurium
existed prior to quinclorac registration. However, this is
unlikely because no other quinclorac-resistant G. spurium
biotype has been found in weed surveys conducted in
Alberta, Canada (LM Hall pers. comm.). Alternatively,
extensive auxinic herbicide use may have selected for the
resistance mutation. The R and S G. spurium biotypes
were tolerant to the auxinic herbicides 2,4-D and
clopyralid (Van Eerd, 2004). Only the R was highly
364 L L Van Eerd et al.
Mutation rate
(gametes/locus/generation)
Density
(per m2)
1 · 10)6
1
5
50
500
1
5
50
500
1
5
50
500
1 · 10)8
1 · 10)10
Probability of occurrence
of one resistant G. spurium plant*
Quinclorac ALSà
Both herbicides§
0.24
0.74
1.00
1.00
0.03
0.01
0.13
0.74
0.00003
0.0001
0.001
0.01
0.28
0.78
1.00
1.00
0.003
0.02
0.15
0.81
0.00003
0.0002
0.002
0.02
0.07
0.58
1.00
1.00
9 · 10)5
2 · 10)4
0.02
0.60
9 · 10)10
2 · 10)8
2 · 10)6
2 · 10)4
Table 4 Probabilities of occurrence of at
least one resistant mutant G. spurium plant
in a 30 hectare field with varying weed
densities. Adapted from Jasieniuk and
colleagues (1996)
*Assuming that G. spurium is a 95% self-fertilizing species.
Based on quinclorac resistance due to a single, recessive nuclear mutation.
à
Based on ALS-inhibitor resistance due to a single, completely dominant nuclear mutation.
§
Two gene mutations based on probability of quinclorac resistance · probability of
thifensulfuron resistance.
resistant to the quinolinecarboxylic acids, moderately
resistant to MCPA, and resistant to the auxinic herbicides
dicamba, picloram, fluroxypyr, and triclopyr (Van Eerd,
2004). Therefore, selection of the R biotype may have
been due to the previous use of auxinic herbicides other
than quinclorac. Ultimately, the elucidation and characterization of the target site of quinclorac is required to
provide insight into the possible link between quinclorac
and auxinic herbicide resistance in G. spurium.
This is the first characterization of quinclorac resistance inheritance. To our knowledge, this is the first time
an F3 population has been developed to illustrate that
resistance to two different modes of action was due to
two distinct genes. At the field dose, quinclorac resistance in G. spurium is mediated by a recessive allele at a
single, nuclear locus. Conversely, ALS-inhibitor resistance was a dominant trait at a single, nuclear locus. The
knowledge of the inheritance of quinclorac resistance
will likely facilitate a better understanding of the genetic
and biochemical mechanisms of auxinic herbicide
resistance.
Acknowledgements
JCH and LL Van Eerd thank the Natural Sciences and
Engineering Research Council (NSERC) of Canada for
providing an NSERC Research Grant and NSERC
graduate scholarship, respectively, and the Ontario
Ministry of Agriculture and Food (OMAF) for their
financial support of this research. The authors sincerely
acknowledge Dr LM Hall for providing the G. spurium
seeds and key information and Dr DJ Wolyn for
providing insight into the breeding and crossing procedures for G. spurium.
References
Abdallah I, Fischer AJ, Elmore CL & Saltveit ME (2004)
Mechanism of resistance to quinclorac in smooth crabgrass
(Digitaria ischaemum). In: Proceedings 2004 The Weed
Science Society of America, 44, Kansas City, MO, USA, 156.
Bailer AJ & Oris JT (1997) Estimating inhibition concentrations for different response scales using generalized linear
models. Environmental Toxicology and Chemistry 16, 1514–
1559.
Cranston HJ, Kern AJ, Hackett JL, Maxwell BD & Dyer WE
(2001) Dicamba resistance in kochia. Weed Science 49,
164–170.
Diebold RS, McNaughton KE, Lee EA & Tardif FJ (2003)
Multiple resistance to imazethapyr and atrazine in Powell
amaranth (Amaranthus powellii). Weed Science 51, 312–318.
Ehrendorfer F (1971) Evolution and ecogeographical differentiation in some southwest Asiatic Rubiaceae. In: Plant Life
in Southwest Asia (eds PH Davis, PC Harper & IC Hedge),
195–215. Botanical Society of Edinburgh University Press,
Aberdeen, UK.
Faegri K & Van den Pijl L (1971) The Principles of Pollination
Ecology. Pergamon Press, Oxford, UK.
Foes M, Liu L, Vigue G, Stoller EW, Wax LM & Tranel PJ
(1999) A kochia (Kochia scoparia) biotype resistant to
triazine and ALS-inhibiting herbicides. Weed Science 47,
20–27.
Goss GA & Dyer WE (2003) Physiological characterization of
auxinic herbicide resistant biotypes of kochia (Kochia
scoparia). Weed Science 51, 839–844.
Grossmann K, Kwiatkowski J & Tresch S (2001) Auxin
herbicides induce H2O2 overproduction and tissue damage in
cleavers (Galium spurium L.). Journal of Experimental
Botany 52, 1811–1816.
Hall LM, Stromme KM, Horsman GP & Devine MD (1998)
Resistance to acetolactate synthase inhibitors and quinclorac
in a biotype of false cleavers (Galium spurium). Weed Science
46, 390–396.
European Weed Research Society Weed Research 2004 44, 355–365
Herbicide resistance in Galium spurium 365
Hansen H & Grossmann K (2000) Auxin-induced ethylene
triggers abscisic acid biosynthesis and growth inhibition.
Plant Physiology 124, 1437–1448.
Heap I (2004) International Survey of Herbicide Resistant Weeds
[WWW document]. URL http://www.weedscience.org
[accessed April 2004].
Horsman GP & Devine MD (2000) Molecular basis of ALS
resistance in a biotype of false cleavers (Galium spurium L.).
In: Proceedings 2000 Weed Science Society of America, 40,
Toronto, ON, Canada, 11.
Jasieniuk M, Brule-Babel A & Morrison IN (1994) Inheritance
of trifluralin resistance in green foxtail (Setaria viridis). Weed
Science 42, 123–127.
Jasieniuk M, Morrison IN & Brule-Babel AL (1995) Inheritance
of dicamba resistance in wild mustard (Brassica kaber).
Weed Science 43, 192–195.
Jasieniuk M, Brule-Babel A & Morrison IN (1996) The
evolution and genetics of herbicide resistance in weeds. Weed
Science 44, 176–193.
Kern AJ, Myers TM, Jasieniuk M, Murray BG, Maxwell BD
& Dyer WE (2002) Two recessive gene inheritance for
triallate resistance in Avena fatua L. Journal of Heredity
93, 48–50.
Koo SJ, Neal JC & DiTomaso JM (1994) Quinclorac-induced
electrolyte leakage in seedling grasses. Weed Science 42,
1–7.
Kuk YI, Burgos NR & Talbert RE (2000) Cross- and multipleresistance of diclofop-resistant Lolium spp. Weed Science 48,
412–419.
Leeson JY, Thomas AG & Hall LM (2002) Alberta Weed
Survey: Cereal, Oilseed and Pulse Crops 2001. Agriculture
and Agri-Food Canada, Saskatoon, Canada.
Lopez-Martinez N, Marshall G & De Prado R (1997) Resistance
of barnyardgrass (Echinochloa crus-galli) to atrazine and
quinclorac. Pesticide Science 51, 171–175.
Malik N & Vanden Born WH (1988) The Biology of Canadian
Weeds. 86. Galium aparine L. and Galium spurium L.
Canadian Journal of Plant Science 68, 481–499.
Mallory-Smith C, Thill DC, Dial MJ & Zemetra RS (1990)
Inheritance of sulfonylurea herbicide resistance in Lactuca
spp. Weed Technology 4, 787–790.
Menezes VG, Ramirez HV & Oliveira JC (2000) Resistance
Echinochloa crus-galli (L.) Beauv. to quinclorac in flooded
rice in southern Brazil. In: Proceedings 2000 International
Weed Science Congress, 3, Foz do Iguassu, Brazil, 297.
European Weed Research Society Weed Research 2004 44, 355–365
Moore RJ (1975) The Galium aparine complex in Canada.
Canadian Journal of Botany 53, 877–893.
Murray BG, Morrison IN & Friesen LF (2002) Pollen mediated
gene flow in wild oat. Weed Science 50, 321–325.
Nandula VK & Messersmith CG (2003) Imazamethabenzresistant wild oat (Avena fatua L.) is resistant to diclofopmethyl. Pesticide Biochemistry and Physiology 74, 53–61.
Penuik MG, Romano ML & Hall JC (1993) Physiological
investigations into the resistance of a wild mustard (Sinapis
arvensis L.) biotype to auxinic herbicide. Weed Research 33,
431–440.
Preston C, Tardif FJ & Powles SB (1996) Multiple mechanisms
and multiple herbicide resistance in Lolium rigidum. In:
Molecular Genetics and Evolution of Pesticide Resistance
(ed. TM Brown), 117–129. American Chemical Society
Symposium Series 645, Washington, DC, USA.
Sabba RP, Ray IM, Lownds N & Sterling TM (2003)
Inheritance of resistance to clopyralid and picloram in
yellow starthistle (Centaurea solstitialis L.) is controlled by a
single nuclear recessive gene. Journal of Heredity 94, 523–
527.
Seefeldt SS, Hoffman DL, Gealy DR & Fuerst EP (1998)
Inheritance of diclofop resistance in wild oat (Avena fatua
L.) biotypes from the Willamette Valley of Oregon. Weed
Science 46, 170–175.
Tranel PJ & Wright TR (2002) Resistance of weeds to ALSinhibiting herbicides: What have we learned? Weed Science
50, 700–712.
Valenzuela-Valenzuela JM, Lownds NK & Sterling TM (2001)
Clopyralid uptake, translocation, metabolism, and ethylene
induction in picloram-resistant yellow starthistle (Centaurea
solstitialis L.). Pesticide Biochemistry and Physiology 71,
11–19.
Valenzuela-Valenzuela JM, Lownds NK & Sterling TM (2002)
Ethylene is not involved in clopyralid action in yellow
starthistle (Centaurea solstitialis L.). Pesticide Biochemistry
and Physiology 72, 142–152.
Van Eerd LL (2004) Characterization and inheritance of
quinclorac resistance in a herbicide-resistant false cleavers
(Galium spurium L.) biotype. PhD Thesis, University of
Guelph, Guelph, ON, Canada.
Zeng L & Baird WV (1999) Inheritance of resistance to antimicrotubule dinitroaniline herbicides in an ÔintermediateÕ
resistant biotype of Eleusine indica (Poaceae). American
Journal of Botany 86, 940–947.