INTRODUCTION
The origins and spread of herbicide resistance in a landscape context may result from a single (or few) independent sources followed by dispersal via gene flow versus multiple, localized, independent evolutionary events (Baucom, 2019). A summary of previous research indicates that gene flow often contributes similarly to or more than independent evolution to the occurrence of resistance among populations (Beckie, Busi, Bagavathiannan, & Martin, 2019). Use of highly effective herbicides, such as glyphosate, across large areas generates a stark selection pressure in which individuals with resistance alleles exhibit high fitness. This widespread selection intensity increases the probability that rare resistance mutations that arise independently will increase in frequency rather than be lost via genetic drift, resulting in a rapid increase in the frequency of a resistance allele following movement to a new location via pollen or seed-mediated dispersal (Beckie et al., 2019). Genomic-based epidemiology enables identification of unique resistance alleles as well as tracking movement of individuals across a landscape when the mutations associated with herbicide resistance are known, providing insight into the origins and spread of resistance mechanisms (Comont & Neve, 2020). Understanding these patterns of independent origins and movement will provide insights for improved management practices, such as focusing area-wide management approaches to prevent sources of seed introduction to new areas when gene flow is moving resistance, or managing selection pressure on a field-by-field basis when resistance is evolving independently and frequently.
To investigate the question of gene flow and independent evolution for herbicide resistance using genomic-based epidemiology, we used kochia [Bassia scoparia (L.) A. J. Scott, synonymous with Kochia scoparia (L.) Schrad.], an introduced weed that occurs in the semi-arid arable lands of the western United States and Canada (Friesen, Beckie, Warwick, & Van Acker, 2009). Kochia collections from around North America exhibit high levels of genetic diversity in the invaded range and a lack of true population structure (Kumar, Jha, Jugulam, Yadav, & Stahlman, 2019; Mengistu & Messersmith, 2002). This lack of structure is most likely due to several reproductive traits that emphasize cross pollination as well as long-range dispersal of both pollen and seed (Beckie, Blackshaw, Hall, & Johnson, 2016). Kochia has protogynous flowers in which the stigmas emerge first and are receptive to pollen from other flowers before pollen production within the same flower occurs, reducing the self-pollination rate (Guttieri, Eberlein, & Thill, 1995). Additionally, kochia is a well-known tumbleweed species, with some plants dispersing seeds for dozens or even hundreds of miles (Kumar et al., 2019). This dispersal mechanism greatly increases the spread of herbicide-resistance alleles and makes containment extremely difficult (Beckie et al., 2016; Kumar et al., 2019; Stallings, Thill, Mallory-Smith, & Shafii, 1995).
Herbicide-resistance mechanisms can be classified as either target-site (mutations or changes in expression of the gene encoding the protein inhibited by the herbicide) or non-target-site (mechanisms that reduce the concentration of active herbicide reaching the target site protein) (Gaines et al., 2020). Target-site mechanisms can be considered specialist adaptations while non-target-site mechanisms can be considered generalist mechanisms, as they can sometimes confer resistance across different herbicide modes of action (Baucom, 2019). Evolution of resistance to the herbicide glyphosate has included both target-site and non-target-site mechanisms, including the parallel evolution of increased copy number of the gene encoding the glyphosate target enzyme 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS) in multiple species (Gaines, Patterson, & Neve, 2019; Patterson, Pettinga, Ravet, Neve, & Gaines, 2018).
The first report of glyphosate-resistant (GR) kochia was from Kansas in 2007 (Waite et al., 2013) and reports have since confirmed glyphosate resistance in multiple US states and Canadian provinces (Kumar et al., 2019). The widespread regional evolution of GR kochia has negatively impacted the sustainability of reduced-tillage weed management and moisture and soil conservation during fallow periods in the Central Great Plains of North America (Kumar et al., 2019). The mechanism of glyphosate resistance has been thoroughly investigated in kochia, in terms of physiology and fitness penalty as well as the genetic mechanisms that cause resistance (Beckie et al., 2018; Kumar & Jha, 2015; Martin et al., 2017; Osipitan & Dille, 2017; Wiersma et al., 2015). Increased copy number of the EPSPS gene has been identified as the resistance mechanism in all studied kochia populations to date from across seven US states (Montana, Wyoming, Oregon, Idaho, Nebraska, Kansas, and Colorado) (Gaines et al., 2016; Godar, Stahlman, Jugulam, & Dille, 2015; Kumar, Felix, Morishita, & Jha, 2018; Kumar, Jha, Giacomini, Westra, & Westra, 2015; Wiersma et al., 2015). In kochia, EPSPS gene duplications are tandem and occur at a single locus (Jugulam et al., 2014; Patterson et al., 2019).
The EPSPS locus has been sequenced from a single GR kochia individual using BAC libraries. The EPSPS repeat unit was variable with two units being most common; 1) a full-length repeat containing EPSPS and six other flanking genes and 2) a less frequent form containing EPSPS and only three other flanking genes (Patterson et al., 2019). A ~15 kb mobile genetic element (MGE) was found inserted flanking both upstream and downstream of the entire tandem duplication. Additionally, a copy of the MGE was found between every repeat unit, indicating the MGE has been co-duplicated after subsequent crossing over events (Patterson et al., 2019). Once the structure of the repeat was determined, quantitative PCR markers that specifically amplify the two types of repeats and the MGE were developed to confirm the sequence and measure the copy number of each part of the repeat structure.
Due to long distance gene dispersal in kochia and its prevalence in cropping systems where it can be carried in/on agricultural equipment, we tested the hypothesis that a single EPSPS gene duplication event occurred initially in Kansas and then subsequently spread to initiate all other GR kochia populations and its alternative, that gene duplication may have occurred multiple times in independent events in a case of parallel evolution. In this study, we used two approaches to address our question. First, we used a genomic-based epidemiology approach with the qPCR markers developed for measuring the various units of the EPSPS tandem duplication to investigate whether all GR plants had the same EPSPS repeat structure. Second, we used a population genetics approach with simple sequence repeat (SSR) markers to determine the relatedness of GR and glyphosate-susceptible (GS) kochia populations from across the Central Great Plains, Northern Plains, and the Pacific Northwest. Combined, these approaches provide insight into the origins and spread of glyphosate resistance alleles in kochia.