Pest Management Science
Pest Manag Sci 61:936–950 (2005)
DOI: 10.1002/ps.1074
Differential response of Amaranthus tuberculatus (Moq ex DC) JD Sauer to glyphosate†
Ian A Zelaya∗ and Micheal DK Owen
Department of Agronomy, Iowa State University, Ames IA 50011-1011, USA
Abstract: Midwest USA farmers have reported inconsistent control of Amaranthus tuberculatus (= rudis)
(Moq ex DC) JD Sauer by glyphosate in glyphosate-resistant crops. The potential of selection for decreased
A tuberculatus sensitivity to glyphosate was therefore investigated in a reportedly resistant Everly, IA
population (P0 –EV). Evaluation of six A tuberculatus populations from the Midwest USA estimated a
seedling baseline sensitivity of 2.15 mM glyphosate. Based on these results, three generations of divergent
recurrent selection were implemented on P0 –EV to isolate resistant and susceptible populations. A seedling
assay was developed to screen large amounts of seeds and thus expedite the selection process. Whole-plant
and seedling rate responses of P0 –EV and a known pristine A tuberculatus population from Paint Creek,
OH (P0 –WT) identified no significant difference in response to glyphosate; however, greater phenotypic
variance was ostensibly evident in P0 –EV. The first recurrent generation selected for resistance at 3.2 mM
glyphosate (RS1 –R) had a 5.9- and 1.7-fold resistance increase at the seedling and whole-plant levels,
respectively, compared with the susceptible generation selected at 32 µM glyphosate. After three cycles of
recurrent selection, 14.6-fold difference in resistance at the seedling level and 3.1-fold difference at the
whole-plant level were observed when comparing the populations selected for resistance (RS3 –R) and
susceptibility (RS3 –S). Overall, recurrent selection increased the frequency of resistant individuals and
decreased the variability to glyphosate at the population level. Nevertheless, variability for glyphosate
resistance was still evident in RS3 –R. Results herein suggested that A tuberculatus is inherently variable to
glyphosate and that selection decreased the sensitivity to glyphosate. We purport that evolved glyphosate
resistance in A tuberculatus may require multiple cycles of selection under field conditions. Historic
estimated use of glyphosate alludes to the evolution of tolerant weed populations.
 2005 Society of Chemical Industry
Keywords: asexual propagation; baseline sensitivity monitoring; non-destructive assay; recurrent selection;
shikimic acid; herbicide resistance
1 INTRODUCTION
Since their commercialization in 1996, N-(phosphonomethyl)glycine (glyphosate)-resistant crops have
provided an excellent control strategy for broadspectrum weed suppression, thus possibly simplifying
farm management, presumably lowering production
costs, and permitting the management of weed biotypes resistant to herbicides other than glyphosate.1,2
These elements have propelled the broad adoption of
this technology in such crops as cotton (Gossypium hirsutum L), soybean (Glycine max L Merr), and canola
(Brassica napus L). In 2001, glyphosate-resistant soybean comprised more than 60% of the cultivated soybean area in the USA, and nearly 100% in Argentina.3
Producers can now spray high glyphosate rates without injury concerns to the crop, and, given the lack of
residual activity, to meet with weed control expectations, crops typically receive multiple glyphosate
applications per growing season. This represented an
increase from the average total of 340 Mg acid equivalents (AE) of glyphosate sprayed per state within the
Midwest USA in 1995 (prior to glyphosate-resistant
crops) to 2060 Mg AE glyphosate per state by 2002.3,4
In the past few years, much debate has centered
on the potential implications of such unparalleled
change of farming operations on the environment,
the world economy, and human health. Evidence
put forward suggests that the large-scale adoption
of herbicide-resistant crops may be accompanied by
significant biological implications; chiefly reduction in
biodiversity,5,6 weed population shift,7,8 and vertical
and horizontal transgenic flow to adjacent plants.9 – 12
Enough evidence exists to assert that the herbicide
selection pressure in the production of herbicideresistant crops increases the fitness of weeds in that
agroecosystem. In the case of glyphosate-resistant
∗
Correspondence to: Ian A Zelaya, 2104 Agronomy Hall, Ames, IA 50011-1011, USA
E-mail: iazelaya@iastate.edu
† Presented at the 13th Annual Australian Weeds Conference in Perth, Australia on 8–13 Sept 2002
Contract/grant sponsor: Monsanto Company
Contract/grant sponsor: Syngenta Crop Protection
(Received 28 January 2005; revised version received 22 March 2005; accepted 12 April 2005)
Published online 24 May 2005
 2005 Society of Chemical Industry. Pest Manag Sci 1526–498X/2005/$30.00
936
Differential response of Amaranthus tuberculatus to glyphosate
crops, cited weed responses include variable or
differential response, tolerance and resistance. The
first response encompasses the variable efficacy
within a weed species to a single, otherwise lethal
glyphosate rate; examples include Convolvulus arvensis
L, Abutilon theophrasti Medicus, and Amaranthus
tuberculatus (= rudis) (Moq ex DC) JD Sauer.13 – 15
Tolerance is the inherent ability of some weed
species to survive glyphosate rates that effectively
control other species in the weed community; species
in this category include Commelina communis L,
Chenopodium album L and Dicliptera chinensis (L) Juss.3
Mechanisms by which tolerant species compensate
for the damaging glyphosate effects include reduced
herbicide absorption and translocation, and higher
activity of the target enzyme, 3-phosphoshikimate
1-carboxyvinyltransferase (EPSPS; EC 2.5.1.19).16,17
Presumably, similar mechanisms exist in species that
demonstrate a differential response to glyphosate.
Resistance is the evolved ability of some species,
thorough selection pressure or genetic engineering,
to survive glyphosate rates that have effectively
controlled the species in the past and that are
lethal to the wild-type species. Glyphosate resistance
has been confirmed in seven species worldwide;
resistance mechanisms include target site modification
and differential glyphosate cellular transport.18 – 21 All
three, resistance, tolerance and differential response,
have been implicated in weed succession changes
within glyphosate-resistant agroecosystems.3,8,22
The cosmopolite pigweed complex is represented
by at least eight weedy species in the Great Plains
of the USA.23 Control with post-emergence herbicides is variable, depending on the pigweed species
and the application timing.15,24 Furthermore, interspecific hybridization and multiple resistance to triazine,
acetolactate synthase (ALS; EC 2.2.1.6) and protoporphyrinogen oxidase (PPO; EC 1.3.3.4) inhibiting
herbicides has complicated A tuberculatus management, even with alternating herbicide programs.25 – 27
In recent years, however, Iowa farmers have reported
inconsistent A tuberculatus control with glyphosate
in glyphosate-resistant crops. Most field reports of
inconsistent glyphosate control of A tuberculatus are
attributable to an ineffective herbicide application,
sub-lethal herbicide doses, or delayed glyphosate
application with respect to the weed’s phenological stage. Nevertheless, some reports of inconsistent
glyphosate control do not fit this profile.
In 1998, a producer from Everly, IA asserted
that application of glyphosate-isopropylammonium
at 0.63 kg AE ha−1 on 10-cm tall plants, followed
by 0.83 and 1.25 kg AE ha−1 one and two weeks
after the initial application, failed to effectively
control A tuberculatus. This incident occurred during
the producer’s first year adoption of glyphosateresistant soybean, yet different control levels were
observed between adjacent A tuberculatus plants
within soybean rows. Field assessment ruled out
delayed or other application problems as the probable
Pest Manag Sci 61:936–950 (2005)
cause for the inconsistent A tuberculatus control.
In 1999, another similar incident was reported in
Badger, IA where 12-cm tall A tuberculatus plants
reportedly survived two sequential applications of
1.25 kg glyphosate AE ha−1 . Since plants were treated
at the labeled height and glyphosate dose, and no
application problems were detected, the inconsistent
control at Badger, IA was attributed to a differential
response in the A tuberculatus population. Endogenous
shikimic acid accumulation estimates from putative
susceptible and resistant plants from Badger, IA, and
from untreated plants in adjacent areas, confirmed
the observations of variable whole-plant response
chronicled in the field.28 Similar cases of inconsistent
A tuberculatus control with glyphosate have been
reported in other Midwest states.29 Preliminary
greenhouse assessments suggested that, compared
with unselected A tuberculatus populations, plants
that originated from the Everly, IA population
demonstrated an enhanced fitness to glyphosate.
Subsequently, an investigation was undertaken to
estimate the baseline sensitivity of A tuberculatus
populations to glyphosate, and to assess the potential
for selection of glyphosate resistance in the Everly, IA
population.
2 MATERIALS AND METHODS
2.1 Plant materials
Ninety-seven plants (53 :44 ) with a resistant phenotype were collected from the Everly, IA field where
glyphosate reportedly failed to control A tuberculatus
and the rescued plants were grown in controlled greenhouse conditions (Section 2.2). At anthesis, plants
were allowed to cross-pollinate without restrictions.
The female inflorescences were then covered with
a PQ218 DelNet bag (Speciality Nets & Profiles.
601 Industrial Drive, Middletown, DE 19709-1083)
and the seed permitted to develop on the mother
plant. Seeds from individual female plants (henceforth maternal lines) were then collected, cleaned
in an air-column separator (Seedburo Equipment
Company. 1022 W Jackson Boulevard, Chicago, IL
60607–2914), and stored at 5 ◦ C and darkness until
utilized; the combined (53) maternal lines comprised
the parental Everly, IA population (P0 –EV). Collected
at Paint Creek Dam, OH, the wild-type A tuberculatus population (P0 –WT) was acquired through
the USDA–ARS National Plant Germplasm System (NPGS; PI 603888). The P0 –WT population
evolved in a wild environment without exposure to
glyphosate (D Pratt, pers comm). After collection,
P0 –WT was maintained at the North Central Regional
NPGS station in Ames, IA and propagated without
selection to any herbicide (D Brenner, pers comm).
Two A tuberculatus populations from Iowa agroecosystems were extracted from the Iowa State University
weed seed collection; records indicated that the seed
was collected at Curtiss Farm, Ames, IA in 1996
(P0 –AM) and Ogden, IA in 1997 (P0 –OG). The
937
IA Zelaya, MDK Owen
homogenous cross–resistant A tuberculatus population
(P0 –CR) was isolated in the greenhouse through two
cycles of recurrent selection at 2.8 kg imazethapyr
AE ha−1 +1.60 kg primisulfuron-methyl ha−1 from a
Davis City, IA population previously described.30 The
multiple-resistant A tuberculatus was selected from a
triazine-resistant population from Fillmore County,
NE.31 Two cycles of recurrent selection at 2.8 kg
imazethapyr AE ha−1 + 4.8 kg atrazine ha−1 isolated
the homogenous multiple-resistant A tuberculatus population (P0 –MR). The grain amaranth cultivar ‘Plainsman’ (A hypochondriacus L × A hybridus L cv K343)
(P0 –GA) was acquired from a local dealer (Calvin
Oliverius, PO Box 25, Albin, WY, 82050–9776).
2.2 Seed conditioning and growth environment
Dormancy was overcome by preconditioning seeds in
filter paper moistened (20 ml) with a 1 mM GA3 [90%
(+)-gibberellic acid] solution in plastic petri dishes
and stored for one month in darkness at 5 ◦ C. Seeds
were then planted in a 30 × 90 cm flat containing a
peat:perlite:loam (1:2:1) soil mix and seedlings were
grown in a greenhouse set at 25–30 ◦ C and 40–60%
relative humidity (RH) diurnal and 20–25 ◦ C and 50%
RH nocturnal conditions. One week post-emergence,
seedlings were transplanted to 12-cm diameter
pots, irrigated as needed, and fertilized (Miracle
Gro Excel, Scott–Sierra Co. 14111 Scottslawn
Road. Marysville, OH 43041–0001) twice during
the reproductive growth cycle. Natural light was
supplemented with artificial illumination delivering
600–1000 µmol m−2 s−1 photosynthetic photon flux
density (PPFD) to create a 16-h photoperiod.
2.3 Glyphosate efficacy assessment
2.3.1 Seedling dose response
All Amaranthus populations were evaluated for
response to glyphosate utilizing a novel seedling assay
that permitted large-scale screening of individuals.
This approach was relevant as A tuberculatus produced
>80 000 seeds under the aforementioned greenhouse
conditions. Twenty seeds per well in a 24-well cell
culture cluster plate (Corning Incorporated. One
Riverfront Plaza Corning, NY 14831–0001) were
placed on four 3-mm diameter 3 MM chromatography
paper discs (Whatman Inc 100 Ames Pond Drive,
Tewksbury, MA 01876-1240), previously moistened
(500 µl) with distilled deionized water (dd H2 O;
control) or doses of 0.032, 0.1, 0.32, 1, 3.2,
10 or 32 mM glyphosate. Doses were prepared by
solubilizing the salt of glyphosate (Sigma-Aldrich
Corporation. 3050 Spruce Street, St Louis, MO
63103-2530) in dd H2 O by adjusting the solution to pH
7.0 with 0.5 M potassium hydroxide. Therefore, each
24–cell culture plate allocated for three replications
of the control and the seven glyphosate doses;
evaporation of the treatment solution was prevented
by enveloping plates with plastic wrap film. Plates
were also established with potassium hydroxide alone
to assess the effect of the base on the germination
938
and development of A tuberculatus seedlings. Finally,
plates were placed in a germinator (Model SG30,
Hoffman Manufacturing, Inc 353 29th Avenue SW,
Albany, OR 97322-3899) set at 30 ◦ C and 14 h
diurnal and 20 ◦ C and 10 h nocturnal conditions
at 500 µmol m−2 s−1 PPFD. Seed germination, and
seedling hypocotyl and radicle length were recorded
two weeks after plate establishment. Germination was
defined by radicle protrusion from the seed, while
dimensional measurements were recorded from the
hypocotyl base to the distal apex or root end.
2.3.2 Whole-plant rate response
Three seedlings per 12-cm diameter pot were equidistantly transplanted and permitted to grow in the
aforementioned greenhouse conditions (Section 2.2).
The manufacturer’s recommended glyphosate rate for
15-cm tall A tuberculatus is 0.85 kg AE ha−1 (Roundup
UltraMAX , Monsanto Co, 700 Chesterfield Parkway North, St Louis, MO 63198–0001), thus rate
responses were conducted at 0.0 (dd H2 O; control),
0.21, 0.54, 0.85, 1.27, 1.69 and 2.5 kg glyphosate AE
ha−1 .32 Treatments were sprayed at 30 cm from the
canopy of the plants through an even flat flow nozzle delivering 187 liter ha−1 at 2.8 kg cm−2 (275 kPa)
in a CO2 -powered spray chamber (Model SB5–66,
DeVries Manufacturing. Route 1 Box 184, Hollandale, MN 56045). Visual glyphosate injury was
assessed 14 days after application by estimating the
percentage foliage damage of treated plants compared
with the dd H2 O-treated controls; dead plants received
a 100% score while asymptomatic plants received 0%.
Typical A tuberculatus injury symptoms to glyphosate
included plant stunting, leaf chlorosis and meristem
necrosis. At evaluation, plant height was also measured by recording the distance from the soil surface
to the base of the plant apex. Furthermore, biomass
was assessed by cutting plants at the soil surface,
drying the tissue at 70 ◦ C for 48 h, and determining
the weight of individual plant samples. Endogenous
shikimic acid quantification was performed on the dry
plant apices (Section 2.3.3).
2.3.3 Endogenous shikimic acid determination
Dry apex samples (0.2 g) were ground in liquid
nitrogen with a mortar and pestle and were then
moistened with 0.25 M hydrochloric acid at a
1:20 (w:v) proportion. Cells were disrupted by
sonication (2210 Ultrasonicator. Branson Ultrasonic
Corporation. 9999 Markham Road, PO Box 150,
Markham, ONT. Canada L3P 3J6) at 25 ◦ C for
30 min and the solution and plant debris were
separated by centrifugation at 20 000 × g for 15 min.
The supernatant (5–10 µl) was then used directly
for shikimic acid determination, by mixing with
dd H2 O (240 µl) and 250 µl of an aqueous solution of
periodic acid (5 g liter−1 ) and sodium metaperiodate
(5 g liter−1 ). The mixture was then incubated at 37 ◦ C
for 30–45 min, quenched with a mixture of aqueous
sodium hydroxide (1.0 M; 300 µl) and aqueous sodium
Pest Manag Sci 61:936–950 (2005)
Differential response of Amaranthus tuberculatus to glyphosate
sulfite (0.056 M; 200 µl), and absorbance detected
at 382 nm in a Lambda 18 UV/Vis Spectrometer
(Perkin–Elmer, 710 Bridgeport Avenue, Shelton, CT
06484-4750). Apex samples were assayed in duplicate
and standard curves constructed with shikimic acid
(Sigma-Aldrich, 3050 Spruce Street, Saint Louis,
MO 63103-2530) at a range of 1–60 µmol ml−1 . The
shikimic acid chromophore had a molar absorptivity
(ε382 ) of 4.76 × 104 liter mol−1 cm−1 at pH 11–12.
2.4 Asexual Amaranthus tuberculatus
propagation
An asexual propagation method (cloning) was developed in an attempt to perpetuate specific A tuberculatus genotypes, permit non-destructive assessment
of glyphosate efficacy (Section 2.5), and facilitate seed
increase for recurrent selection purposes (Section 2.6).
The method comprised defoliating A tuberculatus
branches in active vegetative growth, cutting branches
at the base insertion with the stem, and treating the
abaxial end of branches with Rootone (TechPac,
LLC. PO Box 24830. Lexington, KY 40504). The
branch propagules were then planted in vermiculite
and allowed to develop in a growth cabinet (Model
CMP–3244, Conviron, 222 South 5th Street, Pembina, ND 58271) set at 16 h of 1100 µmol m−2 s−1
PPFD and approximately 95% RH. Adventitious roots
developed from propagules within one to four weeks
after planting. Propagules were then transplanted to
an autoclaved soil-mix and fertilized weekly for one
month. Note that the long photoperiod and frequent
fertilization instigated the re-differentiation of floral A
tuberculatus meristems back to vegetative reproduction.
2.5 Non-destructive plant assay
Branches from asexually propagated A tuberculatus
plants (clones) were cut at the base insertion with
the stems and placed in 13 × 100 mm culture tubes
with Hydro-grow hydroponic solution (Hummert
International. 4500 Earth City Expressway, Earth
City, MO 63045–1329). The branch cuttings were
then sprayed with 0.85 kg glyphosate AE ha−1 and
placed under a 73% PAK shade fabric (Hummert
International) with natural light supplemented to 16 h
of 600 µmol m−2 s−1 PPFD for one week; the Hydrogrow solution was replenish as needed. Glyphosate
efficacy was estimated visually (Section 2.3.2) and by
measuring the accumulation of shikimic acid in the
leaves of treated branches (Section 2.3.3). Susceptible
phenotypes accumulated at least twice the amount
of shikimic acid of resistant phenotypes. The nondestructive plant assay therefore permitted assessment
of glyphosate efficacy in selected phenotypes, without
compromising the survival of cloned plants.
2.6 Selection of P0 –EV
2.6.1 Baseline sensitivity monitoring
The natural variation to glyphosate between samples
of the same population was estimated by assessing
glyphosate efficacy in the 53 P0 –EV maternal lines
with the seedling dose response (Section 2.3.1).
Glyphosate resistance was defined as the reduction
in susceptibility beyond that estimated in the
natural variation of the baseline sensitivity for
the population.33 Maternal lines were therefore
considered resistant if the glyphosate dose required to
reduce A tuberculatus biomass accumulation by 50%
(GR50 ) lay outside the 95% confidence intervals (CI)
for the baseline sensitivity across all six A tuberculatus
populations evaluated (Sections 2.1; 2.8).34
2.6.2 Divergent recurrent selection
Using the P0 –EV baseline sensitivity data as reference,
maternal lines were selected for three generations of
recurrent selection in a divergent strategy, thus isolating resistant (R) and susceptible (S) populations
(Table 1). Recurrent selection was performed with
a modified seedling assay (Section 2.3.1). Six well
cell culture clusters were used and approximately 500
seeds per well on two 3.5-cm diameter chromatography paper disks were moistened (1 ml) with the
screening glyphosate dose (Table 1). The amount of
seed used was estimated by a previous determination that 100 P0 –EV seeds weighed 20.8 mg. Plates
Table 1. Summary of the divergent recurrent selection program to isolate glyphosate resistant (R) and susceptible (S) plants within the Everly, IA
Amaranthus tuberculatus population (P0 –EV). Identification of R and S plants was performed at the seedling stage (Section 2.6.2)
Source
materiala
P0 –EV
P0 –EV
RS1 –S
RS1 –R
BCR
RS2 –S
RS2 –R
Selected
lines
Estimated seeds
selectedb
Screening
dose
Rescued
seedlings
Relative
frequencyc
24
29
24
7
7
16
7
3.54 × 105
3.84 × 105
2.51 × 105
1.38 × 105
2.01 × 105
1.73 × 105
1.37 × 105
32 µM
3.2 mM
10 µM
10 mM
8 mM
10 µM
8 mM
376
159
320
113
121
242
209
1.06 × 10−3
4.14 × 10−4
1.27 × 10−3
8.19 × 10−4
6.02 × 10−4
1.40 × 10−3
1.53 × 10−3
: ratio at
maturity
24: 29
7: 6
16: 21
0: 12
7: 11
12: 15
5:18
Relative
frequencyc
z (P > z)d
1.50 × 10−4
3.38 × 10−5
1.47 × 10−4
8.70 × 10−5
8.96 × 10−5
1.56 × 10−4
1.68 × 10−4
0.69 (0.49)
0.28 (0.78)
0.82 (0.41)
3.46 (<0.01)
0.94 (0.35)
0.58 (0.56)
2.71 (<0.01)
a RS –S, first susceptible recurrent generation; RS –R, first resistant recurrent generation; BC , back cross of RS –R males to RS –R females;
1
1
2
1
R
RS2 –S, second susceptible recurrent generation; RS2 –R, second resistant recurrent generation.
b Based on a mean A tuberculatus seed weight of 20.8 mg per 100 seeds.
c
Proportion of confirmed R or S individuals relative to the estimated seed selected.
d
The one-tailed sign test assessed whether the observed sex rations in selected lines were equally proportioned (1 : 1 ).
Pest Manag Sci 61:936–950 (2005)
939
IA Zelaya, MDK Owen
were treated as described in Section 2.3.1 and evaluated two weeks after establishment. The R phenotype
was classified as A tuberculatus seeds that germinated
and the resulting seedlings grew above the baseline
glyphosate dose; conversely, the S phenotype comprised those seeds that germinated in the medium,
but seedlings developed glyphosate toxicity symptoms
below the baseline dose (Table 1). Individuals that met
the R or S classifications were rescued from the selection medium, rinsed of superfluous glyphosate with

dd H2 O, and roots treated with Rootone . Seedlings
were then transplanted into an autoclaved soil mix in
12-cm diameter pots and grown in a growth cabinet
(Section 2.4) at 500 µmol m−2 s−1 and approximately
80–90% RH; pots were covered with plastic bags
to maintain the high RH. Plants that survived these
conditions were transferred to the aforementioned
greenhouse environment (Section 2.2). At anthesis,
R or S plants were intercrossed in clear plastic tents
where pollen contamination was less than 0.01%.35
Finally, seeds from maternal lines were collected,
cleaned and conditioned as previously indicated (Sections 2.1; 2.2). At least 12 A tuberculatus plants per
recurrent generation were crossed to mitigate the
potential for genetic drift and inbreeding depression
and sustain enough genetic variability for recurrent
selection.36
2.7 Estimation of the number of expressed
EPSPS homologs in plants
The National Center for Biotechnology Information
(NCBI) database was used to identify the ten plant
species with the largest number of available expressed
sequence tags (dbEST).37 The keyword ‘EPSP synthase’ was used to acquire annotated tentative consensus (TC) sequences from these species from the
Eukaryotic Gene Orthologs (EGO) database38 available at the Institute for Genomic Research (TIGR).39
To verify that the TC sequences encoded different
genes, pairwise nucleotide–nucleotide comparisons of
EPSPS orthologous genes within species were conducted with the BLASTN algorithm in the Basic
Local Alignment Search Tool (BLAST).40 Moreover,
the BLASTX algorithm was utilized to compare TC
sequences with previously characterized EPSPS genes
in the GenBank non-redundant database and to identify the highest homologous species.41
2.8 Statistical analysis
Analyses were performed with the Statistical Analysis
Software.42 Experiments replicated in time were tested
for patterns of covariance matrices that satisfied
the Huynh–Feldt condition (option PRINTE), if
accepted, F –statistics tested the within time effects
and related interactions.43 Dose and rate response
experiments repeated once in time were arranged
in a randomized complete block (RCB) design
with four replications and analyzed by ANOVA
(PROC GLM); Fisher’s least significant difference
(LSD; α = 0.05) test was used to separate means
940
when ANOVA identified significant population or
maternal line effects. In addition, biomass data were
used to estimate gamma (γ̂ ) and test whether sublethal glyphosate concentrations had a stimulatory
effect on A tuberculatus development (hormesis).44 A
parameterized Brain–Cousens model was adopted to
estimate GR50 values if a positive hormetic response
was obtained; otherwise, data were analyzed by PROC
NLIN following the log–logistic model.45,46 P –values
for the difference between two estimated GR50
values (|λ50 |) were calculated by reparameterization
of the Brain–Cousens model.44 Visual injury data
were transformed to a binomial format utilizing
the classifications for R (30%) and S (>30%)
phenotypes and the glyphosate dose that inflicted
50% mortality in the A tuberculatus population (LD50 )
estimated by a modified Newton–Raphson algorithm
(PROC PROBIT).47 The overall data fit to the
models was assessed graphically by the distribution
of residuals, and statistically by a lack-of-fit (LOF)
tests and by pseudo coefficients of determination
[R2 (pseudo) ].44,46 Confidence intervals (CI) for means
and GR50 estimates were calculated with PROC
SURVEYMEANS, while the magnitude in variability
to glyphosate was calculated by the sensitivity index
(SI):
GR50 of most tolerant maternal line
[1]
SI =
GR50 of most sensitive maternal line
Total phenotypic variances (σ 2 ) for the components of
the ANOVA were estimated with PROC VARCOMP
using the restricted maximum likelihood (REML)
method and a one-tailed sign test was used to assess
whether the observed sex ratios in selected populations
were equally proportioned 1:1 ( : ). The relationship
strength between endogenous shikimic acid and the
measured plant response parameters was estimated by
Spearman’s correlation analysis.
3 RESULTS AND DISCUSSION
3.1 Differential Amaranthus tuberculatus
response to glyphosate
3.1.1 Natural variation of Amaranthus species to
glyphosate
The grain amaranth (P0 –GA) was the most susceptible
and least variable population to glyphosate (Fig 1).
Both A tuberculatus populations from agroecosystems
(P0 –AM and P0 –OG) had a similar median GR50
to glyphosate compared with the wild-type A
tuberculatus (P0 –WT); however, the agroecosystem
populations had a greater magnitude (at least 18fold) in variance to glyphosate as indicated by the
SI index (Fig 1). In addition, the herbicide-resistant
A tuberculatus populations (P0 –CR and P0 –MR)
had a higher SI index to glyphosate than P0 –WT.
P0 –CR particularly demonstrated a broad native
variability to glyphosate (Fig 1). This is relevant as
most A tuberculatus populations in the Midwest are
Pest Manag Sci 61:936–950 (2005)
Differential response of Amaranthus tuberculatus to glyphosate
30
(10)
24
(53)
6
(10)
4
18
(10)
(10)
(10)
12
(10)
RS2-EV
P0-EV
RS1-EV
P0-CR
P0-MR
P0-AM
P0-OG
P0-GA
P0-WT
RS2-EV
P0-EV
RS1-EV
16
P0-CR
0
P0-MR
6
0
P0-AM
2 (10)
P0-OG
20
8
P0-GA
GR50 (mM glyphosate)
24
(10)
SI (GR50)
28
36
10
P0-WT
32
GR50 (mM glyphosate)
36
12
8
3× baseline
4
baseline
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53
parental P0-EV lines
Figure 1. Main plot: response of Amaranthus tuberculatus from Everly, IA (P0 –EV) to glyphosate estimated by the dose that reduced plant growth
by 50% (GR50 ). The 1× and 3 × P0 –EV baseline sensitivity to glyphosate is indicated by the dashed line; 95% upper and lower confidence intervals
(CI) of the 1× baseline are depicted by the dotted lines. Insets: natural variability (box plot) and variability magnitude (bar plot) of Amaranthus
species to glyphosate as estimated by GR50 and sensitivity index (SI; eqn (1)) values. Boxes denote 5% and 95% percentiles; the line represents
the median. A hypochondriacus × A hybridus: P0 –GA (grain amaranth). A tuberculatus: P0 –WT (wild-type Paint Creek, OH), P0 –AM (Ames, IA),
P0 –OG (Ogden, IA), P0 –CR (cross-resistant Davis City, IA), P0 –MR (multiple-resistant Fillmore County, NE), P0 –EV (Everly, IA), RS1 –EV (P0 –EV
first recurrent), RS2 –EV (P0 –EV second recurrent). Numbers in parentheses designate the numbers of observations (n) in each box estimate. For all
graphs, extensions on boxes or bars designate the 95% upper and lower CI associated with the median or estimated GR50 value, respectively.
suspected to be resistant to ALS-inhibiting herbicides3
and further selection pressure by glyphosate may
select for multiple-resistant populations. Combined,
these data suggest that A tuberculatus populations
from agroecosystems are more variable to glyphosate
than pristine populations. Notably, the A tuberculatus
population collected from Everly, IA (P0 –EV)
demonstrated the most variability to glyphosate
and had a higher SI index (32-fold) than the
other evaluated A tuberculatus populations (Fig 1).
The median GR50 value for P0 –EV was 1.85 mM
glyphosate, compared with 1.11 mM glyphosate of the
P0 –WT population. The data therefore confirmed that
P0 –EV was inherently variable to glyphosate.
3.1.2 Baseline sensitivity monitoring of P0 –EV
Baseline sensitivity monitoring provided an estimate
of the natural variation of P0 –EV to glyphosate and
permitted identification of resistant maternal lines for
recurrent selection purposes. The range of sensitivity
across P0 –EV maternal lines was 0.33–10.66 mM
glyphosate, with a mean (n = 103) baseline GR50 of
2.15 mM glyphosate (95% CI = 1.75–2.54) across all
six A tuberculatus populations evaluated. Considering
our definition of resistance (Section 2.6.1), eleven
maternal lines 1, 3, 11, 19, 20, 23, 26, 27, 28, 44 and
47 were identified as resistant to glyphosate (Fig 1).
Other researchers have suggested that herbicideresistant populations comprise those with a 3×
difference in sensitivity to the baseline.48 Under these
Pest Manag Sci 61:936–950 (2005)
parameters, only three maternal lines, 3, 23 and 47,
were considered resistant to glyphosate (Fig 1).
Characterization of insect natural variation via
baseline monitoring has been proposed as an
important tool for insect resistance management49 and
similar strategies could be applied to plant systems.
Resistance risk assessment through controlled studies,
including baseline sensitivity monitoring, is pivotal
in developing effective pest resistance management
strategies.50
3.2 Amaranthus tuberculatus from Everly, IA is
inherently variable to glyphosate
To minimize possible environmental, rate and phenological factors that affect herbicide efficacy in the field,
the response of P0 –EV to glyphosate was assessed at
the seedling and whole-plant levels under controlled
environments. The pristine P0 –WT was used as a control since records indicated that the population evolved
in a non–agricultural environment without previous
exposure to glyphosate (Section 2.1). Most studies
that evaluate plant performance to herbicides model
the rate response to the log–logistic equation, but
this model is insensitive to discontinuous increasing or
decreasing trends that may result at sub-lethal herbicide rates.44 The Brain–Cousens model was proposed
as an alternative to the accepted log–logistic model,
specifically when discontinuous response trends are
observed that may distort proper estimation of the
response parameters.45 Therefore, the response of
941
IA Zelaya, MDK Owen
Table 2. Dose (hypocotyl; radicle) or rate (biomass; height) response summary of the pristine Paint Creek, OH (P0 –WT), the parental Everly, IA
(P0 –EV), and first (RS1 ) and third (RS3 ) recurrent Amaranthus tuberculatus populations selected for susceptibility (S) or resistance (R) to glyphosate
Parameter
Materiala
Model
GR50 b
CIc
LOFd
R2 (pseudo) d
|λ50 |e
Hypocotyl
P0 –EV
P0 –WT
RS1 –S
RS1 –R
RS3 –S
RS3 –R
P0 –EV
P0 –WT
RS1 –S
RS1 –R
RS3 –S
RS3 –R
P0 –EV
P0 –WT
RS1 –S
RS1 –R
RS3 –S
RS3 –R
P0 –EV
P0 –WT
RS1 –S
RS1 –R
RS3 –S
RS3 –R
Brain–Cousens
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Brain–Cousens
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
Log–logistic
1.56
1.57
0.79
4.71
0.59
8.65
0.07
0.27
0.39
2.31
0.11
1.78
0.62
0.24
0.53
0.89
0.29
0.91
0.47
0.37
0.39
0.87
0.27
0.59
1.12–2.00
1.14–1.99
0.57–1.01
3.19–6.22
0.35–0.84
6.62–10.68
0.05–0.10
0.19–0.34
0.25–0.54
1.81–2.82
0.08–0.15
1.19–2.37
0.52–0.72
0.15–0.32
0.48–0.59
0.83–0.94
0.14–0.43
0.71–1.11
0.39–0.55
0.29–0.44
0.33–0.45
0.80–0.94
0.22–0.32
0.50–0.69
0.78
0.78
0.004
0.32
0.34
0.84
3.918
0.33
0.21
0.84
8.061
0.02
0.16
0.81
0.193
0.16
0.07
0.89
1.919
0.07
0.52
0.88
1.664
0.28
0.44
0.69
0.381
0.25
<0.01
0.89
0.354
<0.01
0.75
0.44
0.624
<0.01
0.75
0.73
0.104
0.37
0.41
0.86
0.483
0.13
0.82
0.76
0.321
0.54
Radicle
Biomass
Height
P valuee
a
RS1 –S, first susceptible recurrent generation; RS1 –R, first resistant recurrent generation; RS3 –S, third susceptible recurrent generation; RS3 –R,
third resistant recurrent generation.
b Glyphosate dose in mM (hypocotyl; radicle) or rate in kg AE ha−1 (biomass; height) that reduced by 50% the evaluated plant parameter.
c
95% lower and upper confidence interval (CI) of the GR50 value estimated by PROC NLIN.
d
Lack-of-fit (LOF) tests and pseudo coefficients of determination [R2 (pseudo) ] estimated the proportion of the total variation explained by the model.
e
|λ50 | quantifies the difference between two estimated GR50 parameters by reparameterization of the Brain–Cousens model; the P value indicates
the statistical significance of the difference.
A tuberculatus at the seedling and whole-plant levels was tested for stimulatory glyphosate effects at
sub-lethal doses (hormesis). The response of P0 –EV
maternal lines to glyphosate was variable (Fig 1).
Resistant lines had seedlings survive the 3.2 mM and
10 mM glyphosate doses, while seedlings in susceptible lines demonstrated pronounced injury symptoms
at the 0.1 mM glyphosate concentration (data not
shown).
Modeling P0 –EV and P0 –WT radicle length,
plant height and biomass data to the parameterized
Brain–Cousens equation estimated γ̂ values that
disregarded a putative hormetic response. Therefore,
response parameters were calculated based on the
log–logistic model (Table 2). The rate response of
P0 –EV hypocotyl length data, however, resulted in
a γ̂ value (38.47) that did not overlap zero at the
95% CI and thus confirmed hormesis. Estimates
of the model fit to the data resulted in LOF tests
and R2 (pseudo) coefficients that confirmed adequacy
of the model in describing the response of P0 –EV
and P0 –W to glyphosate. Overall, comparisons at
the seedling and whole-plant levels provided nonsignificant |λ50 | values suggesting that the P0 –EV and
P0 –W populations responded similarly to glyphosate
(Table 2). While the P0 –EV population tended to
germinate and survive higher glyphosate rates than
942
P0 –W, the differences were statistically not significant
(Fig 2). Nevertheless, P0 –EV demonstrated a greater
overall σ 2 for hypocotyl (5.88) and radicle length
(2.17) in response to glyphosate compared with either
parameter in P0 –W (σ 2 < 0.02). Greater σ 2 in the
response of P0 –EV to glyphosate was also observed for
biomass and plant height measurements. These results
indicated that, while P0 –EV and P0 –WT responded
similarly to glyphosate, individuals within the P0 –EV
population were more variable to the effects of the
herbicide. In addition, results reaffirmed seedling assay
(Section 3.1.1) and visual estimates of variability to
glyphosate assessed in the Everly, IA field.
Variable responses to glyphosate within Amaranthaceae have been documented, and A tuberculatus
and A hybridus were found to be the most tolerant species within the family.22 Reports of incomplete A tuberculatus control with glyphosate in Illinois
and Missouri suggested that 10–15% of the individuals within these populations survived 0.84 kg
glyphosate AE ha−1 , compared with less than 2%
in a known glyphosate-susceptible population.29 Furthermore, evaluations at 0.21 kg glyphosate AE ha−1
identified broad injury ranges of 10–100% within
59 A tuberculatus populations in Illinois.15 Collectively these reports confirm that A tuberculatus in the
USA Midwest is inherently variable to glyphosate.
Pest Manag Sci 61:936–950 (2005)
Differential response of Amaranthus tuberculatus to glyphosate
16
P0-EV
P0-WT
LSD = 1.73
12
10
8
4
2
radicle length (mm)
6
14
P0-WT
P0-EV
12
10
100
80
8
60
6
40
4
20
2
0
0.0
0.032 0.1 0.32 1.0 3.2 10.0 32.0
germination (%)
hypocotyl length (mm)
14
0
0
0.0
0.032 0.1
0.32
1.0
3.2
10.0 32.0
glyphosate (mM)
conditions. Thus, a seedling assay was developed to
cope with the prolific nature of A tuberculatus and
segregate individuals that demonstrated relatively high
or low responses to glyphosate. The seedling assay
effectively identified individuals at 10−4 frequencies
within the population which would have otherwise
required significant greenhouse resources if selection
were conducted at the whole-plant level (Table 1).
Since no difference was observed between germinating
A tuberculatus seeds in dd H2 O or potassium hydroxide
solution (data not shown), survival of plants in the
seedling assay was attributed to the effect of glyphosate
and not to the potassium hydroxide used to solubilize
the herbicide (Section 2.3.1).
3.0
P0-EV
2.5
P0-WT
2.0
1.5
0.5
50
45
40
35
30
25
20
15
10
P0-WT
P0-EV
100
80
60
40
survival (%)
1.0
plant heigth (cm)
biomass (g)
LSD = 0.38
20
0
0.0
0.2
0.5 0.8 1.2 1.7 2.5
0.0
0.2
0.5 0.8 1.2 1.7 2.5
glyphosate (kg AE ha-1)
Figure 2. Upper: Amaranthus tuberculatus seedling dose responses
of (P0 –EV; gray bars; ) parental Everly, IA, and (P0 –WT; white bars;
) pristine A tuberculatus Paint Creek, OH populations. Inset: A
tuberculatus radicle length (circles) and germination (bars) as affected
by glyphosate. Individual points or bars in the graph represent the
mean of four separate experiments, each containing three replications
(n = 12). Lower: whole-plant rate response of (gray bars; ) P0 –EV
and (white bars; ) P0 –WT to glyphosate. Inset: A tuberculatus height
(circles) and mortality (bars) associated with glyphosate rates. Data
points represent the mean of two experiments with four replications
(n = 8). Least significant difference at the P 0.05 level (LSDα=0.05 )
was estimated by Fisher’s test. Extensions on circles or bars
designate the standard error associated with individual means (σM ).
ž
°
°
ž
More important is that survival of A tuberculatus
plants with a pre-existing fitness to glyphosate may
increase the proportion of resistant individuals within
the population and result in reduced herbicide efficacy in the field due to tolerance or evolved resistance.
3.3 P0 –EV seedling recurrent selection
3.3.1 The seedling assay permitted screening of a large
Amaranthus tuberculatus population
Considering that a variable response to glyphosate was
confirmed in P0 –EV, divergent recurrent selection was
implemented to ascertain whether isolation and mating
of A tuberculatus individuals with confirmed resistant or
susceptible phenotypes changed the overall response
to glyphosate at the population level. Greenhouse
estimates suggested that an individual A tuberculatus
plant produced more than 80 000 seeds under ideal
Pest Manag Sci 61:936–950 (2005)
3.3.2 Isolation of the first recurrent (RS1 ) generation
Dose responses were conducted on seeds of the 53
P0 –EV maternal lines to characterize the inherent
glyphosate response within each line (Fig 1). Based
on the baseline sensitivity data (Section 3.1.2), 24
maternal lines were identified susceptible to glyphosate
and the corresponding 3.54 × 105 seeds were selected
for susceptibility at 32 µM glyphosate. Conversely, the
remaining 29 maternal lines were classified resistant
to glyphosate and an approximate 3.84 × 105 seeds
were selected for resistance at 3.2 mM glyphosate
(Table 1). Seedlings selected at 32 µM glyphosate
demonstrated chlorosis; nonetheless, the development
of stems and roots was visually not affected by
glyphosate. In contrast, seedlings surviving 3.2 mM
glyphosate had green healthy tissues but with severely
atrophied roots. Treatment of these stunted roots with
Rootone promoted the development of adventitious
roots from stems and facilitated the rescue of seedlings
selected for resistance to glyphosate. The divergent
glyphosate selection of P0 –EV resulted in the isolation
of 376 susceptible and 159 resistant A tuberculatus
seedlings, which comprised the first recurrent selection
generation (RS1 ). RS1 seedling mortality post isolation
from the selection medium was 86% and 92% for
the susceptible and resistant populations, respectively.
The 53 susceptible (RS1 –S) and 13 resistant (RS1 –R)
plants eventually transferred to the greenhouse had sex
ratios at anthesis that did not differ statistically from
the 1:1 ( : ) proportion (Table 1).
3.3.3 Isolation of the second recurrent (RS2 ) generation
The second recurrent generation (RS2 ) was selected
from 2.51 × 105 seeds for susceptibility at 10 µM
glyphosate, while 1.38 × 105 seeds were selected for
resistance at 10 mM glyphosate (Table 1). The 113
asymptomatic A tuberculatus seedlings rescued after
10 days in 10 mM glyphosate developed injuries when
transplanted to the growth chamber and 89% eventually perished. Similarly, 88% mortality was estimated in the 320 seedlings selected for susceptibility
to glyphosate. At maturity the susceptible (RS2 –S)
and resistant (RS2 –R) populations were comprised
of 16:21 ( : ) and 0:12 ( : ) plants, respectively
943
IA Zelaya, MDK Owen
(Table 1). It is interesting that all 12 individuals
selected for resistance differentiated into male plants.
These preliminary results suggested that glyphosate
may differentially affect the germination of male
and female A tuberculatus seeds. Alternatively, it is
possible that resistance to glyphosate is inherited
concomitantly with maleness in A tuberculatus. Limited
information exists regarding the inheritance of sex
in A tuberculatus. Crosses between dioecious and
monoecious Amaranthaceae provided evidence that
dioeciousness was epistatic over monoeciousness.
When a monoecious female parent was crossed to
a male dioecious plant, the progeny exhibited a 1:1
( : ) sex ratio; in the reciprocal cross however,
with a dioecious female as the parent, the progeny
resulted in 100% female hybrids.51 This suggested
that males may be heterogametic (XY) while females
are monogametic (XX), and that sex has a single locus
with two alleles, an X and Y loci, that determined
plant sex in A tuberculatus. Under this supposition,
the gene(s) associated with glyphosate resistance in
A tuberculatus may be located in the Y loci in the
chromosome.
3.3.4 Assessment of the association between maleness
and glyphosate resistance in RS2
Ascertaining whether glyphosate instigated differential
germination of A tuberculatus was difficult since seeds
cannot ordinarily be sexed. Therefore, the putative
linkage of maleness and glyphosate resistance was
tested by backcrossing RS2 –R individuals to RS1 –R
females, selecting the progeny at 8 mM glyphosate,
and inter-crossing the selected plants to generate
a new resistant lineage (BCR ). Differential sex
germination would have resulted in less germination
of female seeds in the 8 mM glyphosate media and
reduced the overall germination in BCR . Since RS2 –S
demonstrated a normal 1:1 ( : ) ratio (Table 1)
and the combined germination across rates (82%)
was not statistically different (LSDα,0.05 = 6%) from
BCR (81%), differential sex germination in the 8 mM
glyphosate media was disregarded. This selection
resulted in the isolation of 7:11 ( : ) BCR plants
which demonstrated statistically equal male and
female proportions (Table 1). Hence, the putative
sex–linkage to glyphosate resistance observed in the
RS2 –R was attributed to the small sample size in the
selected material (n = 12).
3.3.5 Isolation of the third recurrent (RS3 ) generation
Since the offspring of RS1 –R resulted in exclusively
male plants and thus no progeny was produced,
the third recurrent resistant population (RS3 –R)
was selected at 8 mM glyphosate from the 1.37 ×
105 seed produced by BCR (Table 1). The third
recurrent susceptible population was selected at
10 µM glyphosate from the 1.73 × 105 RS2 –S seeds.
Concurrently, 242 RS3 –S and 209 RS3 –R seedlings
were isolated. Mortality in these selected population
ranged from 85 to 90%, thus at maturity RS3 –S
944
and RS3 –R were comprised of 12:15 ( : ) and
5:18 ( : ) plants, respectively (Table 1). Noteworthy,
RS3 –R had sex ratios that statistically departed from
the 1:1 ( : ) ratio. Therefore, we concluded that
glyphosate resistance may be associated with maleness
in A tuberculatus. Further investigation is needed to
identify the genes and understand the process of sex
differentiation in A tuberculatus.
Few reports exist that cite departure from the
1:1 ( : ) distribution in natural A tuberculatus
populations. In isolated native Ohio A tuberculatus
populations, female plants can comprise up to 90%
of the individuals in a population.52 Conversely, a
tendency for more A tuberculatus males was reported
in response to composted swine manure.53 Spatial
segregation of sexes across environmental gradients
is well documented in dioecious plant species, such
that increased competition and stressful conditions
may favor the prevalence of males in a particular time
and space.54,55 In Amaranthus cannabinus (L) Sauer
no spatial segregation of sexes due to salt stress was
observed.55 However, temporally abnormal sex ratios
observed under controlled conditions were attributed
to flowering phenology differences and sex-specific
mortality. This suggested that differences in a gender
growth pattern may be related to the evolution
of dioecy in Amaranthaceae. While the distribution
of sex ratios in A tuberculatus populations has not
been investigated extensively, the report from Ohio
suggested that sex disparity in native populations is a
possible but rare event.52
3.4 The fitness of Amaranthus tuberculatus to
glyphosate increased with selection
Evaluation of RS1 –S seeds confirmed that the selected
material did not develop well above 1 mM glyphosate,
while radicles and hypocotyls of RS1 –R seedlings grew
in 3.2 mM glyphosate (Fig 3). Furthermore, less than
0.2% of RS1 –R seedlings demonstrated minor or no
toxicity symptoms at 10 mM glyphosate. Stimulatory
responses at sub-lethal glyphosate concentrations were
non-significant except for the radicle response of
RS1 –R (γ̂ = 20.59) (Table 2). Evaluations at the
seedling and whole-plant levels based on |λ50 | values
indicated that the response of RS1 –S and RS1 –R
differed only when comparing biomass as a function of
increasing glyphosate rates (Table 2). This difference
accounted for an approximately 1.7-fold increase in
resistance to glyphosate (Fig 3). At the population
level, 0.55 kg glyphosate AE ha−1 was required to
inflict 50% mortality in the RS1 –S compared with
0.78 kg glyphosate AE ha−1 in the RS1 –R (Table 3).
Concurrently, shikimic acid accumulation in the
apex tissue of treated RS1 plants correlated positively
with the visual herbicide ratings and negatively with
plant height or biomass measurements (Table 3). Phenotypic variance comparisons between the parental
and RS1 A tuberculatus populations suggested that the
recurrent selection process reduced the overall population variability to glyphosate. For example at the
Pest Manag Sci 61:936–950 (2005)
Differential response of Amaranthus tuberculatus to glyphosate
14
RS1-R
RS1-S
LSD = 1.40
10
8
4
radicle length (mm)
6
14
RS1-R
RS1-S
12
120
100
80
10
60
8
40
6
20
4
germination (%)
hypocotyl length (mm)
12
0
0.0
0.032 0.1 0.32 1.0 3.2 10.0 32.0
2
0.0
0.032 0.1 0.32 1.0
glyphosate (mM)
3.2 10.0 32.0
3.0
RS1-R
RS1-S
LSD = 0.21
2.0
1.0
0.5
plant heigth (cm)
1.5
50
45
40
35
30
25
20
15
10
RS1-R
RS1-S
80
60
40
20
0
0.0
0.0
100
survival (%)
biomass (g)
2.5
0.2
0.5 0.8 1.2 1.7
0.2
0.5
0.8
1.2 1.7
glyphosate (kg AE ha-1)
Figure 3. Upper: Amaranthus tuberculatus seedling dose responses
of the first recurrent material selected at (RS1 –R; gray bars; )
3.2 mM glyphosate and (RS1 –S; white bars; ) 32 µM glyphosate.
Inset: A tuberculatus radicle length (circles) and germination (bars) as
affected by glyphosate. Individual points or bars in the graph
represent the mean of four separate experiments, each containing
three replications (n = 12). Lower: Amaranthus tuberculatus
whole-plant rate response of (gray bars; ) RS1 –R and (white bars;
) RS1 –S. Inset: A tuberculatus height (circles) and mortality (bars)
associated with glyphosate rates. Data points represent the mean of
two experiments with four replications (n = 8). Least significant
difference at the P 0.05 level (LSDα=0.05 ) was estimated by Fisher’s
test. Extensions on circles or bars designate the standard error
associated with individual means (σM ).
°
ž
ž
°
seedling level, the σ 2 of RS1 –R (<0.27) and RS1 –S
(<1.84) was less than the hypocotyl (5.88) and radicle
length (2.17) response in the P0 –EV.
3.5 Evident variability to glyphosate even after
three cycles of recurrent selection
Consistent with the RS1 response, RS3 –S seedlings
did not grow above 1 mM glyphosate (Fig 4).
Conversely, more radicle and hypocotyl growth
was observed at 3.2 mM glyphosate in the RS3 –R
compared with the RS1 –R (Fig 3; Fig 4). Estimates
of hormesis were non–significant for the radicle,
hypocotyl, biomass and plant height data (γ̂ >
0.23), therefore the log–logistic model was used to
estimate fitness to glyphosate. Comparisons based on
GR50 values estimated 3-fold and 14-fold resistance
increases between RS3 –R and RS3 –S for the
biomass and hypocotyl length responses, respectively
(Table 2). Concomitantly, an approximate 2.6-fold
resistance increase was estimated when comparing
mortality as a function of glyphosate concentration
at the population level and endogenous shikimic
acid correlated with the expected A tuberculatus
response (Table 3). The fact that RS3 demonstrated an
enhanced fitness to glyphosate compared with RS1 and
in turn RS1 was better fit to glyphosate than P0 –EV,
suggested that the response of RS3 to glyphosate was
at least in part genetic and heritable. RS1 and RS3
also had less σ 2 in response to glyphosate compared
with P0 –EV, which suggested that the recurrent
selection procedure reduced the inherent variability
to glyphosate observed in P0 –EV. However, fewer
differences in σ 2 were observed when comparing the
RS1 and RS3 populations. For instance, the σ 2 of the
hypocotyl length response to glyphosate was 0.27 for
RS1 –R and 1.84 for RS1 –S, which was similar to that
of 0.27 for RS3 –R and 1.91 RS3 –S. Biomass response
to glyphosate had σ 2 of 0.01, 0.02, 0.50 and 0.24 for
RS1 –R, RS1 –S, RS3 –R and RS3 –S, respectively.
Recently, experiments to ascertain whether recurrent selection reduced A tuberculatus sensitivity to
glyphosate were initiated with several Midwest populations, including P0 –EV.56 Conversely to our approach
Table 3. Association strength between shikimic acid concentration, plant height, biomass and visual injury, and estimate of glyphosate lethality to
Amaranthus tuberculatus. Numbers in parentheses indicate the probability of the calculated Spearman’s correlation coefficient (r 2 )
Materiala
Plant height
Biomass
Visual injury
LD50 c
FLd
R/Se
P0 –EV
P0 –WT
RS1 –S
RS1 –R
RS3 –S
RS3 –R
DNPb
−0.65 (<0.01)
−0.79 (<0.01)
−0.85 (<0.01)
−0.60 (<0.01)
−0.43 (<0.01)
DNP
−0.74 (<0.01)
−0.83 (<0.01)
−0.79 (<0.01)
−0.62 (<0.01)
−0.80 (<0.01)
DNP
0.74 (<0.01)
0.85 (<0.01)
0.91 (<0.01)
0.70 (<0.01)
0.61 (<0.01)
0.57
0.21
0.55
0.78
0.32
0.83
0.43–0.69
0.11–0.29
0.38–0.68
0.61–0.95
0.22–0.43
0.67–0.99
2.7
1.4
2.6
a P –EV, parental Everly, IA; P –WT, pristine Paint Creek, OH; RS –S, first susceptible recurrent generation; RS –R, first resistant recurrent
0
0
1
1
generation; S3 –S, third susceptible recurrent generation; S3 –R, third resistant recurrent generation.
b Determination not performed.
c Glyphosate rate in kg AE ha−1 that inflicted 50% mortality in the A tuberculatus population.
d
95% lower and upper fiducial limits (FL) of the calculated LD50 estimated by PROC PROBIT.
e
Calculated by dividing the LD50 of the resistant material by that of the susceptible material.
Pest Manag Sci 61:936–950 (2005)
945
IA Zelaya, MDK Owen
16
RS3-R
RS3-S
LSD = 1.66
12
10
6
4
radicle length (mm)
16
8
RS3-R
RS3-S
14
12
120
100
10
80
8
60
6
40
germination (%)
hypocotyl length (mm)
14
20
4
2
0
0.0
0.032 0.1 0.32 1.0 3.2 10.0 32.0
2
0.0
0.032 0.1 0.32 1.0
glyphosate (mM)
3.2 10.0 32.0
3.0
RS3-R
RS3-S
LSD = 0.64
2.0
1.0
plant heigth (cm)
35
1.5
RS3-R
RS3-S
30
25
100
80
60
20
40
15
20
10
survival (%)
biomass (g)
2.5
0
0.0
0.2
0.5 0.8 1.2 1.7
0.5
0.0
0.2
0.5 0.8 1.2 1.7
glyphosate (kg AE ha-1)
Figure 4. Upper: Amaranthus tuberculatus seedling dose responses
of the third recurrent (RS3 ) material selected at (RS3 –R; gray bars; )
8.0 mM glyphosate and (RS3 –S; white bars; ) 10 µM glyphosate.
Inset: A tuberculatus radicle length (circles) and germination (bars) as
affected by glyphosate. Individual points or bars in the graph
represent the mean of four separate experiments, each containing
three replications (n = 12). Lower: Amaranthus tuberculatus
whole-plant rate response of (gray bars; ) RS3 –R and (white bars;
) RS3 –S. Inset: A tuberculatus height (circles) and mortality (bars)
associated with glyphosate rates. Data points represent the mean of
two experiments with four replications (n = 8). Least significant
difference at the P 0.05 level (LSDα=0.05 ) was estimated by Fisher’s
test. Extensions on circles or bars designate the standard error
associated with individual means (σM ).
°
°
ž
ž
of selecting for resistance at the seedling level, recurrent selection in this investigation was performed
on whole-plants by treating 1000 individuals with
0.22–0.33 kg glyphosate AE ha−1 and allowing 100
surviving plants to intercross. Preliminary data suggested a reduction in glyphosate sensitivity after three
cycles of selection and that quantitative genetics
may be implicated in the response to glyphosate.
In addition, the authors purported an association
between plant architecture and the observed differential response of A tuberculatus to glyphosate.56
While the variability to glyphosate was reduced in
RS1 and RS3 , segregation for glyphosate resistance was
still evident in the S3 –R and S3 –S populations. This
observed persistent variability to glyphosate could be
attributed to the several factors, including the reproductive nature of A tuberculatus and the genetics of
946
glyphosate resistance. Genetic diversity is an inherent
characteristic among species of Amaranthaceae.57,58
Recently, A rudis JD Sauer and A tuberculatus were
grouped in a single-polymorphic species, where populations with more genetic diversity occur in Ohio.52
Because A tuberculatus is an obligate-allogamous
species, genetic recombination during mitosis could
explain the phenotypic variability observed in RS1 and
RS3 . Alternatively, glyphosate resistance in A tuberculatus could be determined by a polygenic event.
Under this hypothesis, alleles associated with resistance could recombine during mitosis, produce different allele combinations, and generate a progeny
with different levels of resistance as observed in the
segregant RS3 . While glyphosate resistance in Lolium
rigidum Gaudin and Conyza canadensis (L) Cronq is
monofactorial,59,60 at least two resistance mechanisms
have been reported in Eleusine indica (L) Gaertner
with potentially independent alleles.61,62 Glyphosate
resistance in Ceratopteris richardii (L) Brongn is governed by two independent loci that are inherited in an
incompletely dominant and recessive fashion.63 Moreover, quantitative genetics of glyphosate tolerance has
been reported in other plant species.64,65 These cases
exemplify that enhanced fitness to glyphosate can
be mediated by monogenic and polygenic events in
plants.
Finally, variation to glyphosate could be explained
by the differential expression of EPSPS homologs at
various A tuberculatus phenological stages. The EGO
database was used to identify EPSPS orthologs across
plant species (Section 2.7). Individual TC numbers
comprise the compiled annotated consensus sequence
of individual expressed sequence tags (ESTs) and
thus potentially represent a unique gene. Therefore,
comparison of EPSPS TCs within a species allowed
for estimation of the minimum number of expressed
EPSPS homologs in the species (Table 4). Homologs
in Pinus spp and S bicolor failed to bear any similarity,
yet the TC sequences demonstrated high homology to
previously characterized EPSPS genes, thus suggesting
that the TCs possibly represented opposite ends of the
same gene. Overall, the analysis confirmed that plant
genomes contained 1–3 copies of EPSPS; nonetheless,
additional EPSPS homologs may be identified as
plant genome projects are further sequenced. The
data also suggested that most EPSPS homologs were
tissue specific. For example, M truncatula TC77645
was composed of nine ESTs, six of which were
differentially expressed in roots. Conversely, eight of
the 13 ESTs comprising TC77644 were expressed
in above-ground organs. It has been suggested that
the expression of EPSPS isoenzymes could represent
an additional level of regulation of the shikimic acid
pathway.66,67
3.6 Several generations of selection may be
required for resistance to evolve in the field
Collectively the data herein presented confirmed that
variability in response to glyphosate exists naturally in
Pest Manag Sci 61:936–950 (2005)
Differential response of Amaranthus tuberculatus to glyphosate
2.0
2.4
0.8
1.8
0.6
1.6
dosage
1.6
1.4
0.4 (10)
(10)
92 93 94 95 96 97 98 99 00 01 02
1.4
(9)
(7)
1.2
(8)
(5)
1.8
1.2
(10)
(9)
1.0
(7)
(7)
0.8
(7)
1.0
0.6
0.8
0.4
0.6
0.2
0.4
kg EA glyphosate per application per ha
number of glyphosate applications per year
1.0
2.0
frequency
kg AE glyphosate
per ha per year
1.2
2.2
0.0
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
growing season
Figure 5. Glyphosate use in glyphosate-resistant soybean within USA Midwest states (IL, IN, IA, KS, MI, MN, MO, NE, OH, WI). Gray bars:
producer average number of applications per year; dark gray bars: average kg AE glyphosate per application per ha. Inset: ( ) Total kg AE
glyphosate applied per ha per year among Midwest states; the solid line represents the expected total use as estimated by the four-parameter
sigmoidal model. Averages were estimated from the number of states as indicated in parenthesis above bars in the main plot. Extensions on circles
and bars designate the 95% confidence interval (CI) associated with the average. Data were compiled from the US Department of Agriculture
(USDA) National Agricultural Statistics Service (NASS) database.4
ž
Table 4. Frequency of 3-phosphoshikimate 1-carboxyvinyltransferase (EPSPS; EC 2.5.1.19) homologs in ten plant species with the largest number
of expressed sequence tags (dbEST)
Highest homologous species in GenBankd
Putative
EPSPS
homologsa
TC
numbersb
BLASTN 2
sequences (%)c
Arabidopsis thaliana
2
Glycine max
2
Hordeum vulgare
Medicago truncatula
1
2
TC253016
TC272122
TC226805
TC226804
TC140189
TC77645
TC77644
TC250561
TC53076
TC53450
TC66045
TC66044
TC94681
TC96059
TC98271
TC192224
TC270442
Species
Oryza sativa
Pinus spp.
Saccharum officinarum
1
1–2
2
Sorghum bicolor
1–3
Triticum aestivum
Zea mays
1
1
E value
Species
94
0.0
0.0
Arabidopsis thaliana
93
e−127
8e−83
0.0
0.0
e−134
0.0
7e−41
3e−12
0.0
0.0
0.0
0.0
1e−10
0.0
0.0
Sarracenia purpurea
Vitis vinifera
Oryza sativa
Medicago truncatula
—
91
—
NSSFe
95
NSSF
—
—
Oryza sativa
Petunia hybrida
Arabidopsis thaliana
Zea mays
Zea mays
Zea mays
Oryza sativa
Eleusine indica
Zea mays
Accession
number
NM 130093
NM 103780
AY545663
AY150520
XM 550613
AC119419
AB016765
M21084
BX815732
X63374
AY106729
AY106729
X63374
AP008208
AJ417034
AY106729
a
Additional EPSPS homologs may be elucidated as plant genome projects are further sequenced.
Annotated tentative consensus sequence (TC); available at TIGR database (http://www.tigr.org/).
c Used to determine if multiple homologs identified from the same species truly represented different EPSPS genes.
d Homology of TC sequences to sequences in the GenBank non–redundant database was estimated with the BLASTX algorithm.
e No significant similarity was found.
b
A tuberculatus populations, and that the potential for
the evolution of glyphosate resistance is significant.
The frequency of resistant individuals within P0 –EV
Pest Manag Sci 61:936–950 (2005)
was extremely low (<4.0 × 10−4 ) and thus required
high selection pressure to isolate resistant individuals
(Table 1). However, selection pressure under field
947
IA Zelaya, MDK Owen
conditions may be less stringent, since survival of
A tuberculatus plants depends on several biotic and
abiotic factors. Therefore, several generations of
glyphosate selection may be necessary to isolate A
tuberculatus genotypes with a stable trait for resistance
in the field. At least 8–15 years of glyphosate selection
pressure in the field was necessary for resistance
to evolve in other weed species.3 This suggested
that (1) the frequency of resistant individuals within
the population is extremely low, (2) several genes
are involved in the trait and recombination plus
selection may be required for the accumulation of
resistant alleles, or that (3) the trait for resistance
imposed a physiological penalty on plants. Evidence
for the later hypothesis was reported in petunia
(Petunia sp), where a single point mutation of
glycyl101 to alanyl encoded for an EPSPS with
increased apparent dissociation constant (Ki app ) for
glyphosate and apparent Michaelis–Menten constant
(Km app ) for PEP.68 Models predicted that the low
frequency of evolved glyphosate resistance in the
field was associated with the application timing of
the herbicide, the frequency of resistant individuals
within the population, the dominance of the resistance
allele(s) and the fitness of the resistant weed biotype.69
At present, glyphosate resistance has only been confirmed in Ambrosia artemisiifolia L and C canadensis
in Midwest cropping systems.18 However, increases
in the area planted with glyphosate-resistant crops
may enhance the selection pressure and lead to
the evolution of additional resistant biotypes.3 Historic estimates of pesticide use allude to the fact
that weed populations have increased fitness to
glyphosate since the introduction of glyphosateresistant crops in 1996. Midwest farmers on average applied 0.64 kg glyphosate AE ha−1 on a single
application to manage weeds in 1995 (Fig 5). By
2002, the dose and frequency had increased to an
average 0.91 kg glyphosate AE ha−1 and 1.3 applications year−1 , respectively. This translated to an
average increase from 0.65 to 1.2 kg glyphosate AE
ha−1 year−1 . This change is in part due to the different glyphosate formulations and recommendations
available to farmers at the time. In 1996, the commercial formulation (Roundup Ultra ) recommended
rates of 0.62 kg AE ha−1 for control of most annual
weeds, including 15-cm tall A tuberculatus. Both
commercial formulations (Roundup UltraMAX and
WeatherMAX ) available to farmers at present recommend 0.84–0.87 kg AE ha−1 for 15-cm tall A tuberculatus. Regardless, anecdotal accounts of increased
tolerance in C communis, A theophrasti, C album and
other species have been reported by USA farmers
in the past few years.3 More research is needed to
understand the impact of continuous glyphosate use
on the dynamics of weed populations and the evolution of resistance. This information may be relevant
in developing effective weed-control strategies and
maintaining the productivity of current USA cropping
systems.
948
ACKNOWLEDGMENTS
We thank Jonathan Gressel for his comments and
recommendations to this investigation. Duane Wilcox,
Jofiel Girón, Ignacio Pimentel, and Lu ı́s Ortiz assisted
with seed screening and greenhouse endeavors.
This work was possible in part by the economic
and technical support of Monsanto Company and
Syngenta Crop Protection.
REFERENCES
1 Shaner DL, The impact of glyphosate-tolerant crops on the
use of other herbicides and on resistance management. Pest
Manag Sci 56:320–326 (2000).
2 Reddy KN, Glyphosate-resistant soybean as a weed management tool: opportunities and challenges. Weed Biol Manag
1:193–202 (2001s).
3 Owen MDK and Zelaya IA, Herbicide-resistant crops and weed
resistance to herbicides. Pest Manag Sci 61:301–311 (2005).
4 Anonymous, Agricultural chemical use database. National Agricultural Statistics Service (NASS), USDA, http://www.pestmanagement.info/nass/ (2005).
5 Watkinson AR, Freckleton RP, Robinson RA and Sutherland WJ, Predictions of biodiversity response to genetically modified herbicide-tolerant crops. Science (Washington)
289:1554–1557 (2000).
6 Freckleton RP, Stephens PA, Sutherland WJ and Watkinson AR, Amelioration of biodiversity impacts of genetically
modified crops: predicting transient versus long-term effects.
Proc R Soc Lond B Biol Sci 271:325–331 (2004).
7 Heard MS, Hawes C, Champion GT, Clark SJ, Firbank LG,
Haughton AJ, Parish AM, Perry JN, Rothery P, Roy DB,
Scott RJ, Skellern MP, Squire GR and Hill MO, Weeds in
fields with contrasting conventional and genetically modified
herbicide-tolerant crops. II Effects on individual species. Phil
Trans R Soc Lond B Biol Sci 358:1833–1846 (2003).
8 Hilgenfeld KL, Martin AR, Mortensen DA and Mason SC,
Weed management in a glyphosate -resistant soybean system:
weed species shifts. Weed Technol 18:284–291 (2004).
9 Kwon YW, Kim DS and Yim KO, Herbicide-resistant genetically modified crops: assessment and management of gene
flow. Weed Biol Manag 1:96–107 (2001).
10 Meagher TR, Belanger FC and Day PR, Using empirical data
to model transgene dispersal. Phil Trans R Soc Lond B Biol Sci
358:1157–1162 (2003).
11 Haygood R, Ives AR and Andow DA, Consequences of recurrent gene flow from crops to wild relatives. Proc R Soc Lond B
Biol Sci 270:1879–1886 (2003).
12 Stewart CN, Halfhill MD and Warwick SI, Genetic modification: transgene introgression from genetically modified crops
to their wild relatives. Nat Rev Genet 4:806–817 (2003).
13 DeGennaro FP and Weller SC, Differential susceptibility of
field bindweed (Convolvulus arvensis) biotypes to glyphosate.
Weed Sci 32:472–476 (1984).
14 Hartzler RG and Battles BA, Reduced fitness of velvetleaf
(Abutilon theophrasti) surviving glyphosate. Weed Technol
15:492–496 (2001).
15 Patzoldt WL, Tranel PJ and Hager AG, Variable herbicide
responses among Illinois waterhemp (Amaranthus rudis and A
tuberculatus) populations. Crop Prot 21:707–712 (2002).
16 Westwood JH and Weller SC, Cellular mechanisms influence
differential glyphosate sensitivity in field bindweed (Convolvulus arvensis) biotypes. Weed Sci 45:2–11 (1997).
17 Norsworthy JK, Burgos NR and Oliver LR, Differences in weed
tolerance to glyphosate involve different mechanisms. Weed
Technol 15:725–731 (2001).
18 Heap I, The international survey of herbicide resistant weeds.
http://www.weedscience.com/ (2005).
19 Baerson SR, Rodriguez DJ, Tran M, Feng Y, Biest NA and
Dill GM, Glyphosate-resistant goosegrass. Identification of
Pest Manag Sci 61:936–950 (2005)
Differential response of Amaranthus tuberculatus to glyphosate
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
a mutation in the target enzyme 5-enolpyruvylshikimate-3phosphate synthase. Plant Physiol 129:1265–1275 (2002).
Lorraine-Colwill DF, Powles SB, Hawkes TR, Hollinshead PH,
Warner SAJ and Preston C, Investigations into the mechanism of glyphosate resistance in Lolium rigidum. Pestic Biochem
Physiol 74:62–72 (2003).
Feng PCC, Tran M, Chiu T, Sammons RD, Heck GR and
CaJacob CA, Investigations into glyphosate-resistant horseweed (Conyza canadensis): retention, uptake, translocation,
and metabolism. Weed Sci 52:498–505 (2004).
Baylis AD, Why glyphosate is a global herbicide: strengths,
weaknesses and prospects. Pest Manag Sci 56:299–308
(2000).
McGregor RL, Barkley TM, Baker WT, Brooks RE, Churchill
SP, Kaul RB, Kolstad OA, Sutherland DM, Van Bruggen T,
Weedon RR and Wilson JS, Atlas of the flora of the great plains,
Iowa State University Press, Ames, IA, pp 70–72 (1977).
Mayo CM, Horak MJ, Peterson DE and Boyer JE, Differential
control of four Amaranthus species by six postemergence
herbicides in soybean (Glycine max). Weed Technol 9:141–147
(1995).
Foes MJ, Liu L, Tranel PJ, Wax LM and Stoller EW, A biotype
of common waterhemp (Amaranthus rudis) resistant to triazine
and ALS herbicides. Weed Sci 46:514–520 (1998).
Patzoldt WL, Tranel PJ and Hagerb AG, A waterhemp (Amaranthus tuberculatus) biotype with multiple resistance across
three herbicide sites of action. Weed Sci 53:30–36 (2005).
Trucco F, Jeschke MR, Rayburn AL and Tranel PJ, Promiscuity in weedy amaranths: high frequency of female tall
waterhemp (Amaranthus tuberculatus) × smooth pigweed
(A hybridus) hybridization under field conditions. Weed Sci
53:46–54 (2005).
Zelaya IA and Owen MDK, Amaranthus tuberculatus (Mq ex
DC) JD Sauer: potential for selection of glyphosate resistance,
in 13th Australian Weeds Conference, ed by Spafford-Jacob H,
Dodd J and Moore JH, Shannon Books, Perth, Australia,
pp 630–633 (2002).
Smeda RJ and Schuster CL, Differential sensitivity to glyphosate
among biotypes of common waterhemp (Amaranthus rudis
Sauer), in 13th Australian Weeds Conference, ed by SpaffordJacob H, Dodd J and Moore JH, Shannon Books, Perth,
Australia, pp 642–644 (2002).
Hinz JRR and Owen MDK, Acetolactate synthase resistance in
a common waterhemp (Amaranthus rudis) population. Weed
Technol 11:13–18 (1997).
Anderson DD, Roeth FW and Martin AR, Occurrence and
control of triazine-resistant common waterhemp (Amaranthus
rudis) in field corn (Zea mays). Weed Technol 10:570–575
(1996).
Anonymous, Crop Protection Reference, C & P Press, Inc, New
York, NY, pp 1836–1855 (2004).
Robertson JL, Preisler HK, Ng SS, Hinkle LA and Gelernter WD, Natural variation: a complicating factor in bioassays with chemical and microbial pesticides. J Econ Entomol
88:1–10 (1995).
Schaub L, Sardy S and Capkun G, Natural variation in baseline
data: when do we call a new sample ‘resistant’? Pest Manag
Sci 58:959–963 (2002).
Brenner DM and Widrlechner MP, Amaranthus seed regeneration in plastic tents in greenhouses. Plant Genet Resour Newsl
116:1–4 (1998).
Hallauer AR, Compendium of recurrent selection methods and
their application. Crit Rev Plant Sci 3:1–33 (1985).
Boguski MS, Lowe TM and Tolstoshev CM, dbEST—database
for ‘‘expressed sequence tags’’. Nat Genet 4:332–333 (1993).
Lee Y, Tsai J, Sunkara S, Karamycheva S, Pertea G, Sultana R,
Antonescu V, Chan A, Cheung F and Quackenbush J, The
TIGR Gene Indices: clustering and assembling EST and
known genes and integration with eukaryotic genomes. Nucleic
Acids Res 33:D71–D74 (2005).
Anonymous, TIGR database: The Institute for Genomic
Research, http://www.tigr.org/ (2005).
Pest Manag Sci 61:936–950 (2005)
40 Tatusova TA and Madden TL, Blast 2 sequences—a new tool
for comparing protein and nucleotide sequences. FEMS
Microbiol Lett 174:247–250 (1999).
41 Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z,
Miller W and Lipman DJ, Gapped BLAST and PSI–BLAST:
a new generation of protein database search programs. Nucleic
Acids Res 25:3389–3402 (1997).
42 SAS, SAS procedures guide, version 8, SAS Institute, Cary, NC,
p 3884 (2000).
43 Huynh H and Feldt LS, Conditions under which mean
square ratios in repeated measurements designs have exact
F –distributions. J Am Stat Assoc 65:1582–1589 (1970).
44 Schabenberger O, Tharp BE, Kells JJ and Penner D, Statistical
tests for hormesis and effective dosages in herbicide dose
response. Agron J 91:713–721 (1999).
45 Brain P and Cousens R, An equation to describe dose responses
where there is stimulation of growth at low doses. Weed Res
29:93–96 (1989).
46 Seefeldt SS, Jensen JE and Fuerst EP, Log–logistic analysis
of herbicide dose–response relationships. Weed Technol
9:218–227 (1995).
47 Collett D, Modelling binary data, CRC Press, Boca Raton, FL,
p 408 (2002).
48 Paterson EA, Shenton ZL and Straszewski AE, Establishment
of the baseline sensitivity and monitoring response of Papaver
rhoeas populations to florasulam. Pest Manag Sci 58:964–966
(2002).
49 Moulton JK, Pepper DA, Jansson RK and Dennehy TJ, Proactive management of beet armyworm (Lepidoptera: Noctuidae) resistance to tebufenozide and methoxyfenozide: baseline monitoring, risk assessment, and isolation of resistance.
J Econ Entomol 95:414–424 (2002).
50 Jutsum AR, Heaney SP, Perrin BM and Wege PJ, Pesticide
resistance: assessment of risk and the development and
implementation of effective management strategies. Pestic Sci
54:435–446 (1998).
51 Murray MJ, The genetics of sex determination in the family
Amaranthaceae. Genetics 25:409–431 (1940).
52 Pratt BD and Clark LG, Amaranthus rudis and A tuberculatus—one species or two? J Torrey Bot Soc 128:282–296
(2001).
53 Menalled FD, Liebman M and Buhler DD, Impact of composted swine manure and tillage on common waterhemp
(Amaranthus rudis) competition with soybean. Weed Sci
52:605–613 (2004).
54 Bierzychudek P and Eckhart V, Spatial segregation of the sexes
of dioecious plants. Am Nat 132:34–43 (1988).
55 Bram MR and Quinn JA, Sex expression, sex–specific traits,
and the effects of salinity on growth and reproduction of
Amaranthus cannabinus (Amaranthaceae), a dioecious annual.
Am J Bot 87:1609–1618 (2000).
56 Tranel PJ, Patzoldt WL and Volenberg DS, Waterhemp’s
response to glyphosate selection. Proc North Cent Weed Sci
Soc Abstr 59:83 (2004).
57 Chan KF and Sun M, Genetic diversity and relationships
detected by isozyme and RAPD analysis of crop and wild
species of Amaranthus. Theor Appl Genet 95:865–873 (1997).
58 Sun M, Chen H and Leung FC, Low-Cot DNA sequences
for fingerprinting analysis of germplasm diversity and
relationships in Amaranthus. Theor Appl Genet 99:464–472
(1999).
59 Lorraine-Colwill DF, Powles SB, Hawkes TR and Preston C,
Inheritance of evolved glyphosate resistance in Lolium rigidum
(Gaud). Theor Appl Genet 102:545–550 (2001).
60 Zelaya IA, Owen MDK and VanGessel MJ, Inheritance of
evolved glyphosate resistance in Conyza canadensis (L) Cronq.
Theor Appl Genet 110:58–70 (2004).
61 Ng CH, Wickneswary R, Salmijah S, Teng YT and Ismail BS,
Glyphosate resistance in Eleusine indica (L) Gaertn from
different origins and polymerase chain reaction amplification of specific alleles. Aust J Agric Res 55:407–414
(2004).
949
IA Zelaya, MDK Owen
62 Ng CH, Wickneswari R, Salmijah S, Teng YT and Ismail BS,
Gene polymorphisms in glyphosate-resistant and -susceptible
biotypes of Eleusine indica from Malaysia. Weed Res
43:108–115 (2003).
63 Chun PT and Hickok LG, Inheritance of two mutations
conferring glyphosate tolerance in the fern Ceratopteris
richardii. Can J Bot 70:1097–1099 (1992).
64 Racchi ML, Stefanini F, Camussi A and Forlani G, Inheritance
of glyphosate tolerance among maize somaclones. Maydica
42:275–280 (1997).
65 Duncan CN and Weller SC, Heritability of glyphosate susceptibility among biotypes of field bindweed. J Hered 78:257–260
(1987).
66 Forlani G, Parisi B and Nielsen E, 5-enol-Pyruvyl-shikimate-3phosphate synthase from Zea mays cultured cells: purification
950
and properties. Plant Physiol 105:1107–1114 (1994).
67 Forlani G, Nielsen E and Racchi ML, A glyphosate-resistant
5-enol-pyruvyl-shikimate-3-phosphate synthase confers tolerance to a maize cell line. Plant Sci 85:9–15 (1992).
68 Padgette SR, Re DB, Gasser CS, Eichholtz DA, Frazier RB,
Hironaka CM, Levine EB, Shah DM, Fraley RT and Kishore
GM, Site-directed mutagenesis of a conserved region of the 5enolpyruvylshikimate-3-phosphate synthase active site. J Biol
Chem 266:22 364–22 369 (1991).
69 Neve P, Diggle AJ, Smith FP and Powles SB, Simulating
evolution of glyphosate resistance in Lolium rigidum I:
population biology of a rare resistance trait. Weed Res
43:404–417 (2003).
Pest Manag Sci 61:936–950 (2005)