Introduction
Kochia [Bassia scoparia (L.) A.J. Scott] is a resilient and adaptable plant species that has garnered attention for its ecological and agricultural impact (Geddes and Sharpe Reference Geddes and Sharpe2022; Kumar et al. Reference Kumar, Jha, Jugulam, Yadav and Stahlman2019; Qadir et al. Reference Qadir, Tubeileh, Akhtar, Larbi, Minhas and Khan2008). Native to Eurasia, B. scoparia has become widespread across North America, where it is a troublesome weed (Supplementary Figure 1). Historically, B. scoparia has been utilized for various purposes, including erosion control, forage production, and ornamental landscaping. However, its prolific seed production, rapid growth, and ability to thrive in diverse environmental conditions have contributed to its status as a major problematic weed in agricultural as well as non-crop settings.
A key adaptation for B. scoparia’s success is its C4 photosynthetic pathway, coupled with a unique emergence and flowering strategy. As a C4 plant, B. scoparia thrives in hot and dry environments due to efficient water use. This is particularly advantageous when it emerges early in spring, capitalizing on cool, moist conditions for initial growth while tolerating late spring frosts (Friesen et al. Reference Friesen, Beckie, Warwick and Van Acker2009). By delaying flowering until the hottest part of summer, B. scoparia leverages its limited use of moisture in arid conditions to attain its critical reproductive stage. This flexibility makes it a successful competitor in harsh conditions (Friesen et al. Reference Friesen, Beckie, Warwick and Van Acker2009). Genetic studies have revealed the existence of multiple genetic lineages within B. scoparia populations, reflecting both natural selection and human-mediated gene flow (Martin et al. Reference Martin, Benedict, Wei, Sauder, Beckie and Hall2020). This genetic diversity enables acclimation in response to environmental cues such as tolerance to high and low temperatures, low moisture, poor soils with high salt content, and nutrient availability (Kumar et al. Reference Kumar, Jha, Jugulam, Yadav and Stahlman2019; Yadav et al. Reference Yadav, Jha, Kniss, Lawrence and Sbatella2023; Figure 1).
Figure 1. Examples of Bassia scoparia plants growing in Colorado, USA. (A) Seedlings can withstand exposure to freezing temperatures during spring cold snap and (B) resume growth without major injury. (Photos by PW).
Bassia scoparia has been a problematic weed in the High Plains of the United States for years, but its ability to evolve resistance to herbicides has complicated its management (Adeyemi et al. Reference Adeyemi, Westra, Ransom, Creech and Ortiz2025). Various B. scoparia populations have evolved resistance, individually or in combinations, to synthetic auxins, photosystem II inhibitors, acetolactate synthase inhibitors, and glyphosate (Kumar et al. Reference Kumar, Jha, Jugulam, Yadav and Stahlman2019; Varanasi et al. Reference Varanasi, Godar, Currie, Dille, Thompson, Stahlman and Jugulam2015). With respect to glyphosate resistance, most plant species, including B. scoparia, normally have a single copy of 5-enolpyruvyl-3-shikimate phosphate synthase (EPSPS). Cases of resistance involving EPSPS copy number variation (CNV) have arisen by different means (Johnson et al. Reference Johnson, Lemas, Montgomery, Gaines and Patterson2025). For example, in Palmer amaranth (Amaranthus palmeri S. Watson), a 297-kb replicon containing multiple additional open reading frames for additional genes and repetitive DNA elements is amplified as extrachromosomal circular DNA with plants having more than 100 copies of EPSPS (Koo et al. Reference Koo, Molin, Saski, Jiang, Putta, Jugulam, Friebe and Gill2018). In B. scoparia, this process is driven by the presence of an FHY3/FAR1 transposable elements (Jugulam et al. Reference Jugulam, Niehues, Godar, Koo, Danilova, Friebe, Sehgal, Varanasi, Wiersma, Westra, Stahlman and Gill2014; Patterson et al. Reference Patterson, Saski, Sloan, Tranel, Westra and Gaines2019; Wiersma et al. Reference Wiersma, Gaines, Preston, Hamilton, Giacomini, Robin Buell, Leach and Westra2015), with some plants having as many as 20 EPSPS copies (Godar et al. Reference Godar, Stahlman, Jugulam and Dille2015).
As with other weeds, effective management of glyphosate-resistant (GR) B. scoparia requires integrated weed management strategies that incorporate diverse approaches, such as herbicide rotation, use of alternative herbicides with different modes of action, adoption of cultural practices, and development of nonchemical control methods (Kumar et al. Reference Kumar, Jha, Jugulam, Yadav and Stahlman2019). Proactive monitoring of herbicide resistance and early detection of resistant populations are essential for implementing timely and targeted management interventions to mitigate the spread of resistance and preserve the efficacy of herbicides for weed control in agricultural systems (Torbiak et al. Reference Torbiak, Blackshaw, Brandt, Hamman and Geddes2024).
Herein, we report a survey for (1) frequency of glyphosate resistance in field-collected B. scoparia populations; (2) the inheritance pattern of gene copy number for total EPSPS, the two types of EPSPS CNV, and the associated transposon in a field study; and (3) the frequency of outcrossing in a field study.
Materials and Methods
Pollen Scanning Electron Microscopy
Naturally dried pollen was separated from mature flowers and prepared for scanning electron microscope observation according to Lynch and Webster (Reference Lynch and Webster1975). The grains were sprinkled evenly on double-sided conductive adhesive on gold–palladium metallic stubs and coated with a gold sputter coater (Vacuum Desk II Gold Sputter Coater, Denton North America, Moorestown, NJ USA). The pollen grains were observed using a field emission scanning electron microscope (JSM-6500, JEOL USA, Peabody, MA USA) at the Center for Imaging and Surface Science of Colorado State University (Fort Collins, CO).
Geographic Localization of GR Bassia scoparia and Wind Analysis
The locations and dates of reports of GR B. scoparia were obtained from the International Herbicide-Resistant Weed Database (Heap Reference Heap2024; accessed: July 2024). Twenty-year (January 2001 to December 2020) average monthly wind speed and direction were obtained from the National Aeronautics and Space Administration (NASA) Langley Research Center Prediction of Worldwide Energy Resource (POWER) Project funded through the NASA Earth Science/Applied Science Program (https://power.larc.nasa.gov/).
Plant Material for Field Survey and Treatment
A field survey was conducted in 2015 to assess glyphosate resistance in B. scoparia across Colorado. Bassia scoparia seeds were collected in autumn (October to November) from field and roadside locations along transects, with a minimum distance of 16 km between sites (Supplementary Table 2). Within crop fields, seeds were harvested from individual B. scoparia plants that had survived the entire growing season. To create composite samples for each location, seeds from 5 to 20 plants were pooled. Seeds were then transferred to the Colorado State University greenhouse for screening Seedlings were treated with glyphosate (RoundUp WeatherMax®, Bayer, St. Louis, MO USA) at a rate of 870 g ai ha−1 with 20 g L−1 ammonium sulfate using a moving overhead single-nozzle sprayer (DeVries, Hollandale, MN) calibrated to deliver 187 L ha−1.
Bassia scoparia accessions were classified as susceptible (≤ 20% survival) or resistant (>20% survival) based on their response to the discriminating glyphosate rate corresponding to 870 g ae ha−1. Georeferenced collection sites were mapped using Arc Catalogue and ArcMap (v. 10.2.1) to visualize spatial patterns of glyphosate resistance across Colorado and facilitate comparisons over time (Khater et al. Reference Khater, Ali, Afify, Bayomy and Abbas2022).
Plant Material for Inheritance Study
Survivor (parent) plant samples were collected during autumn from different Roundup Ready® sugar beet (Beta vulgaris L. subsp. vulgaris) fields in each year (2018, 2020, 2022). For 2018, parental tissues were randomly collected from a field at coordinates 40.617°N, 105.017°W in northern Colorado, and seeds were brought to Colorado State University. Collection for 2020 was performed within the same field at 40.6°N, 105.017°W, but using a more spatially targeted approach. Mature plant samples 1 to 11 were collected from the west side of the field, samples 12 to 22 were clusters of survivors on the east side of the same field, samples 23 to 31 were isolated from the central part of the field, and samples 32 to 44 were gathered from the northwest part of the field from August through September. In 2022, healthy plants that survived multiple glyphosate applications during the growing season were collected from a sugar beet field at 40.6°N, 105.017°W, transplanted into 22-L pots, and taken to Colorado State University, where they were grown to maturity for seed collection. Samples 1 to 5 were collected from the west side of the field, whereas samples 6 to 10 were collected from the east side. All seeds were placed in 81.3-cm Miracle-Gro® Moisture Control® Potting Mix with Lambert LM-GPS (Scotts, Marysville, OH USA) until reaching the seedling stage in the greenhouse, where plants were watered once a day to maintain field capacity. For evaluating gene copy number variation, a known GR Colorado B. scoparia population with high EPSPS copy number (M32) (Westra et al. Reference Westra, Nissen, Getts, Westra and Gaines2019) and a known glyphosate-susceptible Colorado inbred population (7710) were selected (Patterson et al. Reference Patterson, Saski, Sloan, Tranel, Westra and Gaines2019; Pettinga et al. Reference Pettinga, Ou, Patterson, Jugulam, Westra and Gaines2018; Preston et al. Reference Preston, Belles, Westra, Nissen and Ward2009). Genomic DNA was extracted from GR field populations that had survived a treatment with 900 g ai ha−1 glyphosate during the early growing season.
Extraction of Genomic DNA
Two young rosette leaves were collected from each individual plant at the 2.5-cm seedling height growth stage. Samples were immediately flash-frozen in liquid nitrogen and stored at −80 C until DNA extraction. Three biological replicates were used for DNA extraction from both parental and progeny lines. DNA extraction was carried out using the cetyl trimethylammonium bromide (CTAB) method (Doyle Reference Doyle, Hewitt, Johnston, Young and Techniques1991), followed by quantification using a Thermo ND-1000 model Nanodrop spectrophotometer (Thermo Scientific, Wilmington, NC USA).
Quantitative Real-Time PCR for EPSPS Copy Number Determination
Quantitative real-time PCR (qRT-PCR) was employed to determine the EPSPS copy number variation. Ten nanograms per microliter (ng µl−1) of genomic DNA from three biological replicates of each parental and progeny line were used for the qRT-PCR reactions. Additionally, three technical replicates were performed for each biological replicate to account for variation during the sample preparation for qRT-PCR. The reactions were performed using a Power SYBR Green PCR Master Mix (Applied Biosystems, Warrington, UK) on a Bio-Rad CFX Real-time PCR System (Bio-Rad Technologies, Hercules, CA USA). The thermal cycling conditions included an initial 3-min denaturation step at 95 C, followed by 40 amplification cycles consisting of denaturation at 95 C for 15 s, and a combined annealing/extension step at 70 C for 30 s, as described elsewhere (Patterson et al. Reference Patterson, Saski, Sloan, Tranel, Westra and Gaines2019; Wiersma et al. Reference Wiersma, Gaines, Preston, Hamilton, Giacomini, Robin Buell, Leach and Westra2015).
Specific primer sets targeted the full-length Type I EPSPS segment (56.1 kb), the shorter Type II segment (32.9 kb), and a flanking mobile genetic element, FAR1 (15 kb) (Table 1) (Patterson et al. Reference Patterson, Saski, Sloan, Tranel, Westra and Gaines2019; Ravet et al. Reference Ravet, Sparks, Dixon, Küpper, Westra, Pettinga, Tranel, Felix, Morishita, Jha, Kniss, Stahlman, Neve, Patterson, Westra and Gaines2021). The acetolactate synthase (ALS) reference gene was used to normalize EPSPS copy number using the ΔΔCt method (Gaines et al. Reference Gaines, Barker, Patterson, Westra, Westra, Wilson, Jha, Kumar and Kniss2016; Wiersma et al. Reference Wiersma, Gaines, Preston, Hamilton, Giacomini, Robin Buell, Leach and Westra2015). Briefly, the relative expression ratio (R) was calculated as R = 2−(ΔCt(sample) − ΔCt(calibrator)) (Schmittgen and Livak Reference Schmittgen and Livak2008), where ΔCt(sample) represents the difference in threshold cycle (Ct) values between the target gene (EPSPS) and the reference gene (ALS) for a given sample, and ΔCt(calibrator) represents the average ΔCt value for the parental line used as the calibrator. The technical replicates were averaged first; and then biological replicates were averaged for the analysis, followed by calculation of standard deviation to assess variation within each treatment group.
Table 1. Forward and reverse primer sequences for ALS (control), EPSPS, Type I, Type II repeats, and mobile genetic element (FAR1).
Simple Sequence Repeat Genotyping Using Bio-analyzer
Simple sequence repeats (SSRs) were used to study inheritance in selected GR B. scoparia populations. From the set of 11 SSR primers mentioned in a study by Ravet et al. (Reference Ravet, Sparks, Dixon, Küpper, Westra, Pettinga, Tranel, Felix, Morishita, Jha, Kniss, Stahlman, Neve, Patterson, Westra and Gaines2021), three specific markers were selected for further analysis because of their respective locations in the genome and because they showed allelic polymorphism in preliminary analysis of parental samples (Table 2). Specifically, SSR 1792 is on chromosome 4 (bp 103692085), SSR 2656 is on chromosome 6 (bp 46032154), and SSR 2896 is on chromosome 6 (bp 13718386). On the other hand, the mobile genetic element (MGE) is present as a single copy on chromosome 9 in the analyzed glyphosate-susceptible B. scoparia genome, while a nearly identical single-copy MGE is present next to EPSPS on chromosome 1 in the analyzed GR B. scoparia genome (Hall et al. Reference Hall, Montgomery, Chen, Saski, Matzrafi, Westra, Gaines and Patterson2025). Therefore, the SSR markers can be considered neutral with respect to glyphosate resistance and are not in linkage disequilibrium with the EPSPS gene duplication or the MGE.
Table 2. List of simple sequence repeat (SSR) primers used for genotyping.
qRT-PCR was performed utilizing genomic DNA diluted to a concentration of 5 ng μl−1 with EconoTaq PLUS master mix. The process starts with a 2-min denaturation step at 94 C, followed by denaturation at 94 C for 30 s, a 30-s annealing step at temperatures shown in Table 2, a 45-s extension step at 72 C for 37 cycles, and a final extension of 2 min at 72 C. The PCR products were then multiplexed using the three SSR markers (Table 2) and processed for fragment analysis using a 2100 bioanalyzer (Agilent Technologies, Santa Clara, CA USA) at the biotechnology core facilities at the University of Nebraska (Lincoln, NE USA). A high-sensitivity DNA kit for the sizing and quantitation of fragmented DNA was utilized for sample preparation before fragment analysis (dsDNA 905 Reagent Kit, Agilent Technologies, Santa Clara, CA USA).
Statistical Analysis
Statistical significance of gene copy numbers was determined by ANOVA followed by Fisher’s LSD test (Williams and Abdi Reference Williams, Abdi and Salkind2010) at the α = 0.05 level using module agricolae in RStudio (2024.04.2 Build 764) and R v. 4.3.1 (2023-06-16 ucrt).
The homogeneity of the gene CNV across years for parental and progeny EPSPS was assessed with the Bartlett’s test (Bartlett and Fowler Reference Bartlett and Fowler1937) using the stats module in R. This test was used to determine whether the data from the various collection years could be pooled. The null hypothesis (H0) assumes equal variances across groups, while the alternative hypothesis (Ha) states that at least two groups have unequal variances. Statistical analysis and means separation of the original data for each year are available in Supplementary Table 3.
Results and Discussion
Pollination Dynamics and Inheritance
As part of the Amaranthaceae family, B. scoparia is characterized by its small, inconspicuous greenish-white to pinkish flowers lacking showy petals or sepals arranged in dense clusters along slender stems (Friesen et al. Reference Friesen, Beckie, Warwick and Van Acker2009; Figure 2A). Additionally, it is considered an outcrossing species, because its flowers are protogynous, where stigmas are receptive 1 wk before dehiscence of anthers on the same flower (Friesen et al. Reference Friesen, Beckie, Warwick and Van Acker2009). Furthermore, its pollen grains are spheroidal with a diameter of 20 to 40 µm, having a granular surface due to 100 to 130 pores (Stallings et al. Reference Stallings, Thill, Mallory-Smith and Bahman1995a; Figure 2B), which is typical for a anemophilous pollen compared with pollen from biotically pollinated plants that is more than 200 µm (Harder Reference Harder1998). These small spherical and dimpled pollen grains have structural similarities to a golf ball whose structure reduces air resistance and increases lift, maximizing their ability to remain suspended in the air (Ackerman Reference Ackerman, Dafni, Hesse and Pacini2000).
Figure 2. Bassia scoparia flower and pollen grain. (A) Photograph of a B. scoparia flower and (B) scanning electron micrograph of a B. scoparia pollen grain. Micrograph was obtained at 3,000× magnification and with 5 kV (line = 10 µm).
Being prone to wind pollination, B. scoparia relies on abiotic factors such as wind speed and direction for pollen dispersal and pollination (Chen et al. Reference Chen, Burns, Fleming and Patterson2020). The absence of showy petals or nectar rewards suggests that B. scoparia is not dependent on insect pollinators for reproduction (Gill Reference Gill2004).
Glyphosate Resistance Survey
Dispersal of herbicide-resistant alleles via pollen-mediated and/or seed-mediated means can drive the movement of herbicide-resistant populations across an agroecoregion. GR B. scoparia was first reported in Kansas in 2007 and is now causing problems in 10 western states (Heap Reference Heap2024; Waite et al. Reference Waite, Thompson, Peterson, Currie, Olson, Stahlman and Al-Khatib2013; Figure 3). This weed is highly fecund, releasing abundant amounts of pollen and producing large numbers of seeds that can be dispersed by wind as the mature plants tumble across a field. Earlier studies demonstrated that pollen-mediated dispersal is strongly dependent on prevailing wind direction during pollination (Beckie et al. Reference Beckie, Blackshaw, Hall and Johnson2016). Similarly, tumbling mature B. scoparia plants can produce100,000 seeds or more over 1 km with wind speed as low as 0.3 m s−1 (Beckie et al. Reference Beckie, Blackshaw, Hall and Johnson2016). These combined methods of dispersion are expected to lead to a rapid expansion of the resistance trait (Martin et al. Reference Martin, Benedict, Wei, Sauder, Beckie and Hall2020).
Figure 3. States with reported glyphosate-resistant (GR) Bassia scoparia populations (and year of first report) based on data from Heap (Reference Heap2024). The map was generated with MapChart (https://www.mapchart.net/usa.html). Light green arrows indicate 20-yr (January 2001–December 2020) predominant wind direction during flowering ranged from northeast to southeast, and small lines represent the range of wind direction during that period of time. Data from National Aeronautics and Space Administration (NASA) Langley Research Center Prediction of Worldwide Energy Resource (POWER) Project (https://power.larc.nasa.gov/).
Interestingly, identification of GR B. scoparia populations has expanded primarily westward and against the predominant wind direction in this region of the United States (Figure 3; Supplementary Table 1). Indeed, analysis of the wind pattern from 2001 to 2020 reveals that this region received wind from the west or southwest with an average wind speed around 5 m s−1 (Supplementary Table 1). Therefore, the westward movement of GR B. scoparia over long distances may not be solely the result of wind dispersal. On the other hand, the lack of eastward expansion may be due to the wetter midwestern conditions that are not conducive to good B. scoparia survival and expansion. Consequently, the emergence of GR B. scoparia in other states may be the result of seed transport by farming equipment or birds or the result of independent selection of new resistant biotypes (Kumar and Jha Reference Kumar and Jha2015b).
GR B. scoparia poses a great challenge to agricultural weed management strategies, particularly in regions where this weed species has become widespread (Kumar et al. Reference Kumar, Jha, Jugulam, Yadav and Stahlman2019; Figure 4A). GR B. scoparia seeds can spread during harvesting or while tumbling across large areas, resulting in movement of GR B. scoparia in the field and beyond (Figure 4B).
Figure 4. Impact of Bassia scoparia on agroecosystem. (A) glyphosate-resistant (GR) B. scoparia plants growing in a Colorado sugar beet field. (Photo by André Araujo.) (B) Path taken by a GR B. scoparia plants tumbling across a field, dropping seeds that later emerge and grow as GR weeds. (Photo by PW).
A total of 51 B. scoparia samples were collected from Colorado and assessed for glyphosate resistance frequency (Supplementary Table 2). Most of the samples were collected from the fallow phase of wheat (Triticum aestivum L.) or corn (Zea mays L.) cropping systems. However, some samples were also from roadsides and field margins, with a few from alfalfa (Medicago sativa L.) fields. Due to the reliance on glyphosate for weed management in no-till fallow, 68% of the progenies had some GR individuals.
Overall, glyphosate resistance frequency in the samples collected (Supplementary Table 2) ranged from 0% to 63%. Mapping the B. scoparia resistance survey data on 51 accessions collected in 2015 and screened in 2016 indicates GR B. scoparia georeferenced clustering within an 80-km radius of this survey (Figure 5). A 80 km radius was selected as a practical limit for sample collection logistics. Identifying GR B. scoparia clusters is important to focus weed control efforts in areas with the highest GR B. scoparia pressure. Resources can be directed toward these areas to prevent further spread and ensure effective weed management.
Figure 5. Georeferenced collection sites. Each collection site was mapped using Arc Catalogue and ArcGis (v. 10.2.1) (Maguire Reference Maguire, Shekhar and Xiong2008) to visualize spatial patterns of glyphosate resistance across Colorado and facilitate comparisons over time.
Glyphosate Resistance in Bassia scoparia
The mechanism of glyphosate resistance in B. scoparia is associated with EPSPS gene CNV (Gaines et al. Reference Gaines, Barker, Patterson, Westra, Westra, Wilson, Jha, Kumar and Kniss2016; Godar et al. Reference Godar, Stahlman, Jugulam and Dille2015; Jugulam et al. Reference Jugulam, Niehues, Godar, Koo, Danilova, Friebe, Sehgal, Varanasi, Wiersma, Westra, Stahlman and Gill2014). While sensitive B. scoparia individuals have a single EPSPS copy, GR B. scoparia plants with multiple copies of EPSPS have reduced sensitivity to glyphosate and can survive and reproduce in glyphosate-treated fields, leading to the proliferation of resistant populations and exacerbating weed management challenges (Lim et al. Reference Lim, Jha, Kumar and Dyer2021).
The origin of the EPSPS gene duplication event and the evolution of glyphosate resistance are attributed to an MGE containing a FAR1-like transposon (Hall et al. Reference Hall, Montgomery, Chen, Saski, Matzrafi, Westra, Gaines and Patterson2025). Amplified EPSPS copies are typically positioned in tandem in the GR B. scoparia genome. Researchers have performed FISH (fluorescence in situ hybridization) analysis to study these tandem repeats and found the sizes of repeat to be ∼45 kb and ∼66 kb (Jugulam et al. Reference Jugulam, Niehues, Godar, Koo, Danilova, Friebe, Sehgal, Varanasi, Wiersma, Westra, Stahlman and Gill2014). There are two dominant repeats upstream and downstream of CNV boundaries known as Type I, which has a full-length 56.1-kb repeat, and Type II, which has a smaller 32.9-kb repeat, as well as a large MGE of ∼15 kb interspersed in the repeat structure (Patterson et al. Reference Patterson, Saski, Sloan, Tranel, Westra and Gaines2019). A typical MGE containing AR1 DNA binding, zinc finger, and SWIM domains is present both upstream and downstream of the CNV repeats in GR populations. Unequal crossing-over events, facilitated by a transposable element insertion, have resulted in amplification of the EPSPS gene. A recently discovered repetitive element called Muntjac, with a mutator Don-Robertson transposase that has transcription factor-like activity, provides perspectives of GR B. scoparia via transduplication processes (Dupeyron et al. Reference Dupeyron, Singh, Bass and Hayward2019; Hall et al. Reference Hall, Montgomery, Chen, Saski, Matzrafi, Westra, Gaines and Patterson2025). In B. scoparia, the larger Type I EPSPS locus contains seven other co-duplicated genes, whereas the smaller Type II EPSPS (32.9 kb) repeat contains only four genes (Patterson et al. Reference Patterson, Saski, Sloan, Tranel, Westra and Gaines2019). Because the gene duplication of EPSPS involves a structural variant, the number of EPSPS gene copies inherited by progenies may differ from the numbers in the parents due to unequal crossing over during meiosis (Hall et al. Reference Hall, Montgomery, Chen, Saski, Matzrafi, Westra, Gaines and Patterson2025; Jugulam et al. Reference Jugulam, Niehues, Godar, Koo, Danilova, Friebe, Sehgal, Varanasi, Wiersma, Westra, Stahlman and Gill2014).
EPSPS and FAR1 Copy Number Variation
The homogeneity of the gene CNV across years in parental and progeny for EPSPS CNV was assessed with Bartlett’s test in a location under constant glyphosate selection. This test was used to determine whether there were significant differences in these populations between years under conditions of glyphosate selection. Bartlett’s test revealed no statistically significant difference in variance for EPSPS copy number across the three collection years (2018, 2020, and 2022) in the parental lines (Bartlett’s K-squared = 1.948, df = 2, P-value = 0.3776) and progenies (Bartlett’s K-squared = 0.45504, df = 2, P-value = 0.7965). The pooled data had normal distribution in both parents and progeny (Figure 6), with an average 12 to 13 EPSPS copy numbers in both parents and progenies. While the number of copies ranged from 7 to 19 and 6 to 20 in the parent and progeny, respectively, both populations include 95% of the distribution.
Figure 6. Distribution of EPSPS copy number in parent (blue) and progeny (red) samples.
There was no difference (P-value = 0.769) between the EPSPS copy number of parental and progeny populations (Figure 7A). This suggests that the EPSPS copy number distribution was constant in the parental and the progeny populations over the 2018 to 2022 study period (Figure 7A).
Figure 7. EPSPS gene copy number. (A) Total EPSPS copy number comparison between parents (blue) and progeny (red) for the years 2018, 2020, and 2022. (B and C) Pooled parent and pooled progeny Type I, Type II EPSPS and FAR1 copy numbers. Each data point represents individual samples. Means with same letters are not statistically different at the α = 0.05 level, using Fisher’s protected LSD test.
The copy number variations of the Type I and Type II EPSPS segments and the FAR1 transposable element were investigated. The range in copy numbers of Type I EPSPS was similar in the parental and progeny lines (Figure 7B). While the same is true with Type II EPSPS, this smaller duplication event is not as common as the Type I duplication (Figure 7B), which accounts for most of the EPSPS copies. This provides evidence for a stable parental EPSPS copy number being maintained under selection at a population level and is consistent with the trait maintaining Hardy–Weinberg equilibrium even under selection (Rousset Reference Rousset2008). It appears that a stable CNV with at least 6 to 7 copies is sufficient for full resistance, and thus there was no selection to further increase the CNV over consistent selection pressures in these glyphosate-treated sugar beet fields. Finally, there were more FAR1 copy numbers than the total EPSPS copy numbers (Figure 7A and 7C), suggesting that FAR1 is also present in other parts of the genome (Hall et al. Reference Hall, Montgomery, Chen, Saski, Matzrafi, Westra, Gaines and Patterson2025). Interestingly, based on qRT-PCR markers, Ravet et al. (Reference Ravet, Sparks, Dixon, Küpper, Westra, Pettinga, Tranel, Felix, Morishita, Jha, Kniss, Stahlman, Neve, Patterson, Westra and Gaines2021) defined three genotypes of EPSPS gene duplication. Our samples from all 3 yr fall into the category of genotype I, characterized by an increase in EPSPS, Type I and II repeats, and MGE copy numbers corresponding to more than 10 EPSPS copies (Ravet et al. Reference Ravet, Sparks, Dixon, Küpper, Westra, Pettinga, Tranel, Felix, Morishita, Jha, Kniss, Stahlman, Neve, Patterson, Westra and Gaines2021).
The inheritance of glyphosate resistance was investigated using three specific markers SSRs (2656, 2896, and 1792) across the parent and progeny from the samples collected in 2018, 2020, and 2022. SSR marker analysis requires some interpretation of the peaks. In this study, peaks within ±5-bp from an SSR marker were considered to match that marker, whereas those with a size beyond the ±5-bp threshold were considered different (as illustrated in Figure 8). Plant samples with amplicon size outside this threshold were considered to be outcrossing events. For example, the parent (4) had amplicons at 170, 177, 267, and 278 bases. One of its progenies (4a) had amplicons at 169, 177, 263, 272, 278, and 303, which indicates an outcrossing event; another progeny (4c) had 177, 263, 272, and 278 amplicons, which indicates either a self-pollinating event or an outcrossing event between individuals with identical genotypes for these markers.
Figure 8. Example of an simple sequence repeat (SSR) analysis for assessing whether progenies were the result of a self-pollinating or outcrossing event. The amplicon sizes for the selected SSR markers were 174, 267, and 277. The range considered as matching the markers was ±5-bp, as shown by bars.
Consequently, the outcrossing rate was approximately 78%, whereas the estimated self-pollination rate was 22% (Figure 9). However, the self-pollination rate may be an overestimation, as it may combine true self-pollination events as well as outcrossing events that yielded progenies reporting the same markers as the parents for the three SSR markers evaluated.
Figure 9. Percentages of progeny resulting from outcrossing (blue) and self-pollinating (red) events (n = 58). Error bars are ±1 SE of the mean. Means with different letters are statistically different at the α = 0.05 level, using Fisher’s protected LSD test.
GR B. scoparia biotypes have a transposon-associated CNV of EPSPS. On average, parents and progenies have 12 to 13 copies of EPSPS, and most of these duplicated genes have the larger Type I genetic structure. Inheritance of the resistant trait is stable under constant selection, with the number of copies being similar in the parents and progenies over years, suggesting that populations have sufficient resistance to glyphosate and there has not been selection for greater copy numbers. The wide range in the distribution of the number of EPSPS copies can be accounted for by the fact that B. scoparia favors outcrossing, which is encouraged by its protogynous flower structure, with stigmas being receptive several days before anthers release their pollen. Additionally, the open configuration of the flower promotes wind dispersal of B. scoparia’s small spheroidal anemophilous pollen grains. The high fecundity and wind dispersal of the GR trait through pollen likely promote the local spread of GR B. scoparia plants (Geddes and Pittman Reference Geddes and Pittman2022). However, the westward spread of GR B. scoparia from Kansas to Oregon is against prevailing wind, suggesting that these seeds are moved by other means, and poses significant challenges for weed management in agricultural systems. Strong winds can easily spread B. scoparia seeds in multiple directions, highlighting the pattern of seed dispersal from tumbling mother plants (Stallings et al. Reference Stallings, Thill, Mallory-Smith and Lass1995b). Effective monitoring of pollen movement could be beneficial in controlling the distribution of resistant traits across fields. Unchecked seed dispersal of this tumbling weed has potential for contributing to further herbicide-resistance development.
The elevated EPSPS gene copy number imparting glyphosate resistance in B. scoparia is stably inherited across generations and growing seasons. This indicates that once resistance is established in a B. scoparia population, it is likely to persist without continued glyphosate application. Thus, resistance management requires long-term, integrated approaches. The discovery of a high outcrossing rate (78%) under field conditions highlights the rapid potential for gene flow among B. scoparia populations. Combined with the weed’s protogynous flowering structure, wind-dispersed pollen, and ability to disperse seeds over long distances, these traits facilitate the widespread dissemination of resistance. Immediate management strategies should focus on preventing seed production and minimizing pollen escape from known GR populations. This can be achieved by using preemergence soil-residual herbicides with different modes of action, rotating crops to disrupt B. scoparia life cycles, and enhancing crop competitiveness through narrow row spacing or cover crops. Given that many GR populations possess EPSPS copy numbers above the threshold for field-level resistance, reliance on postemergence glyphosate alone is no longer viable. Therefore, integrating herbicides with different modes of action, especially preemergence herbicides with good residual activities, and crop rotation may prevent seed deposition and infestation of GR B. scoparia in soil (Kumar and Jha Reference Kumar and Jha2015a, Reference Kumar and Jha2015b).
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/wsc.2025.10049
Acknowledgments
We are thankful for the support provided by undergraduate students and other colleagues in the Weed Research Laboratory at Colorado State University.
Funding statement
We are grateful for financial support from the USDA National Institute of Food and Agriculture, Hatch Projects 7005022/ COL00785A (FED) and COL00783 (TAG).
Competing interests
The authors declare that they have no conflicts of interest.
Open access
