Prevention

Prevention is one of the most basic of weed control methods (Buhler 2002). This tactic, which uses methods to prevent the introduction of a weed into an area or to prevent its spread, is a vital part of an IWM program. Laws regulating seed purity and prohibiting the spread of noxious weeds contribute to prevention. Other methods include planting crops that are not contaminated by weed seed; cleaning field machinery, so that weed seeds and propagules will not be transported to other fields; and preventing the spread of weed propagules by transport of livestock, manures, or compost, or through irrigation or drainage waters (Walker 1995). Timely scouting of fields, another preventive method, allows early detection of uncontrolled, potentially resistant weeds or of shifts to difficult-to-control weeds when the weeds are small and there are still effective control options.

Cultural Control

Cultural weed management refers to agronomic practices that use competitiveness of the crop to maximize its growth while diminishing the growth and subsequent competitiveness of associated weeds. Interest in cultural weed control was renewed in the 1980s and 1990s because of mandatory, government-imposed reductions in pesticide use in European countries and growing public support for similar reductions in North America (Hansen and Zeljkovich 1982; King and Buchanan 1993; Liebman and Dyck 1993b). Cultural strategies include such tactics as rotating crops, improving crop competition through crop variety selection and planting date, and optimizing seeding rates (Beckie and Gill 2006; O'Donovan et al. 2007).

Crop Rotation

Before the introduction of modern herbicides in the 1940s, crop rotation was a primary method of weed control. Crop rotations are one method of preventing weed shifts (change in the composition or relative frequencies of weeds in a plant population or community because of environment or agronomic practices that favor one species over another). For example, a rotation from rice (Oryza sativa L.) to crops such as cotton, tomato (Lycopersicon esculentum L.), safflower (Carthamus tinctorius L.), grain sorghum [Sorghum bicolor (L.) Moench. ssp. bicolor], or wheat (Triticum aestivum L.) reduced aquatic weed populations in flooded rice production (Hill and Bayer 1990). A green foxtail [Setaria viridis (L.) Beauv.] population was 960 plants m⁻² in monoculture wheat vs. 3.5 m⁻² in wheat with a summer fallow rotation (Liebman and Dyck 1993a). In a corn monoculture, a single weed species comprised 94% of weeds present, whereas in a corn–wheat rotation, no single species contributed more than 43% of the total weeds present (Liebman and Dyck 1993a).

Crop rotations improve weed control by periodically changing the weed community because various crops differ in planting and harvest dates, growth habit, competitive ability, fertility requirements, and associated production practices, thereby favoring different weed associations (Buhler 2002; Forcella et al. 1993; O'Donovan et al. 2007). Weeds can adapt to a specific crop (Labrada 2006), and a rotation between different crop types can help break the cycle of adapted weeds (Buhler 2002). Ecologically, the rotation crop may inhabit niches that weeds occupied in the previous crop, or it may be a highly competitive crop that prevents weeds from thriving and producing seeds (Melander and Rasmussen 2001). Herbicide use in different crops may also result in control of different species. For example, a grass crop (corn, grain sorghum, wheat, or rice) can be rotated with a broadleaf crop, such as soybean; herbicides used in the broadleaf crop can effectively control grass weeds that were not controlled well in the grass crop. Before more-effective herbicides were introduced for grass control in rice, one of the recommendations for control of red rice (Oryza sativa L.) in commercial rice was to rotate to soybean, where herbicides to control grasses could be used more effectively.

For resistance management, crop rotation allows the introduction of herbicides having different MOAs to avoid successive use of a single MOA (Anderson et al. 1999; Buhler 2002; HRAC 2009a). Additionally, the life cycle of the weed can be disrupted or avoided by rotating crops (HRAC 2009a). Crops with different times of planting and different production practices allow a variety of cultural techniques to be used to optimize crop competitiveness at the expense of weed growth and reproduction.

Cover Crops, Intercropping, and Mulches

Cover crops are a way of minimizing weed populations while maintaining seasonal vegetative cover to prevent soil erosion (Moore et al. 1994). A cover crop is usually a “noncash” crop that can be grown before, or in the case of a living mulch or smother crop, with, a cash crop so that vegetative cover remains on the field for as long as possible during the year (Melander et al. 2005). Several advantages accrue from cover crops. They help producers meet conservation-tillage requirements for year-round vegetation cover; aid in soil erosion prevention; improve soil structure and, often, organic matter content; protect plants in sandy areas from sand-blow injury; fix nitrogen if the cover crop is a legume; and possibly suppress weed emergence and growth (Akemo et al. 2000; Krutz et al. 2009; Melander et al. 2005; Norsworthy et al. 2011; Snapp et al. 2005; Teasdale 1998). Some research has shown that cover crops can provide enough early season weed control that a PRE application of herbicide can be eliminated (Ateh and Doll 1996; Fisk et al. 2001; Isik et al. 2009; Malik et al. 2008; Reddy 2001). Suppression of weeds by cover crops depends partly on biomass production of the crop. In a comparison of nine cover crops in Kentucky, biomass from cereal rye (Secale cereale L.) and wheat was greater than that from fescue (Festuca) species and legumes, provided a more-compatible planting situation for seedling establishment of soybean, and provided greater suppression of weed growth. Some cover crops also have allelopathic potential (discussed in a later section). Cover crop systems that included tillage to stimulate weed seed germination, followed by a cover crop to suppress weed growth, resulted in a significant decrease in the weed seedbank (Mirsky et al. 2010). Although tillage may not be desirable for many row-crop producers, the Mirsky et al. (2010) results demonstrate the success of integrated methods in reducing the weed seedbank.

Intercropping (growing two cash crops simultaneously) has also been used to reduce weed growth (Liebman and Davis 2000; Liebman and Dyck 1993a; Melander et al. 2005). Intercropping combinations allow for the exploitation of more available resources compared with one crop and may suppress weed growth by use of those resources (Ballare and Casal 2000; Liebman and Dyck 1993b). Intercropping grain sorghum with cowpea [Vigna unguiculata (L.) Walpers] resulted in lower weed densities and less weed dry matter than did grain sorghum alone because the cowpea intercepted light and used N, P, and K that were then unavailable for weed growth (Abraham and Singh 1984). Weed seed populations have also decreased because of intercropping (Wilson and Phipps 1985), which is also a tool for maintaining ecological diversity (Barberi 2002). Intercropping is often accepted by producers because of income from two cash crops (Barberi 2002; Melander et al. 2005).

Nonliving mulches are another means of reducing weed infestations. Mulches can consist of organic materials, such as tree bark, straw, or litter composts (Niggli et al. 1990), or of inorganic materials, such as plastic mulches. Black plastic mulches are widely used in high-value crops, such as ornamentals and vegetables (Bangarwa et al. 2009). They prevent light from reaching the soil surface, thereby inhibiting weed growth, and can conserve soil moisture and reduce herbicide dissipation (Bangarwa et al. 2009, 2011). Black polyethylene mulch is often used in tomato, bell pepper (Capsicum annuum L.), and some herbs. It effectively controlled weeds in sweet basil (Ocimum basilicum L.) and rosemary (Rosmarinus officinalis L.) but not in parsley [Petroselinum crispum (Mill.) Nyman ex A.W. Hill](Richards and Whytock 1993).

Although these methods of cultural control add expense to total production costs (Snapp et al. 2005), the cost for establishing cover crops and buying mulches may ultimately be a better alternative than the cost of controlling weeds that have evolved resistance.

Planting Date

Crop planting date can affect the severity of a weed infestation. Rapid and consistent emergence of the crop is critical to its success and competitive advantage over associated weeds. Crops sown very early may emerge slowly and have uneven establishment, which makes them more susceptible to weed competition (Lutman 1991; Radosevich et al. 1997). However, crop planting date can be manipulated to provide the crop a competitive advantage over weeds (Walsh and Powles 2007; Williams 2006). Delayed seeding is used in IWM strategies in Australia to help manage herbicide-resistant ryegrass (Lolium spp.)(Walsh and Powles 2007). Delaying crop seeding by 2 wk allows emergence of annual weeds that can be controlled with a nonselective herbicide before planting (Walsh and Powles 2007). O'Donovan et al. (2007) suggest that late crop seeding, with control of weeds by tillage or herbicides before planting, could be an advantage in areas with early emerging weeds, such as wild oat (Avena fatua L.), as long as the growing season is not too short to risk a delay in planting. Williams (2006) reported that weed management tactics in June-planted sweet corn were less intense than were those in May-planted corn because of the greater weed pressure in May.

Crop Cultivar Choice

Crop cultivars that mature at different rates or have different competitive abilities may be used to suppress weed populations and weed growth (Froud-Williams 1988; Richards 1989). Use of competitive crop cultivars has been examined in crops as varied as vegetables and grains in organic crop production (Barberi 2002), soybean (Nordby et al. 2007), canola (Beckie et al. 2008; Blackshaw et al. 2008; Harker et al. 2003), wheat (Wicks et al. 2004), peanut (Arachis hypogaea L.)(Place et al. 2010), rice (de Vida et al. 2006; Gealy et al. 2003), and potatoes (Solanum tuberosum L.)(Colquhoun et al. 2009). Factors affecting competitive ability include height, density, leaf area, tillering, canopy type, allelopathic potential, and cultivar differences in competition for light, water, and soil nutrients (Grundy et al. 1992; Lotz et al. 1991; Moss 1985; Pyšek and Lepš 1991; Richards and Whytock 1993; Standifer and Beste 1985). For example, early released varieties of canola were poor competitors, but hybrid cultivars were more competitive with weeds than were the earlier cultivars and had higher yields (Blackshaw et al. 2008). Barnyardgrass [Echinochloa crus-galli (L.) Beauv.] was consistently suppressed to a greater extent by highly competitive, high-yielding Asian rice cultivars than it was by U.S. cultivars ‘Starbonnet’, ‘Kaybonnet’, ‘Lemont’, and ‘Cypress’, suggesting that growing weed-suppressive rice cultivars could be a component of an effective and economical weed management strategy for rice (Gealy et al. 2003). Even though cultivars may not differentially suppress weeds, their yield potentials may differ in the presence of weeds (Colquhoun et al. 2009).

Seeding Rate

Increased crop population density can improve the competitive ability of a crop against weeds because of rapid canopy development. Increased seeding rates are used as a weed management strategy in Australia's dryland wheat to manage herbicide-resistant rigid ryegrass (Lolium rigidum Gaudin.)(Gill and Holmes 1997; Walsh and Powles 2007). Lemerle et al. (2004) reported reduced ryegrass biomass with wheat seeding rates that increased wheat density from 100 to 200 plants m⁻². Increased seeding rates of winter wheat also reduced the number of flowering heads of blackgrass (Alopecurus myosuroides Huds.) and the total biomass of sterile oat (Avena sterilis L.)(Anderson 1986; Froud-Williams 1988). Increased crop density and narrow row widths were used to reduce the competitive ability of wild oat in spring barley (Hordeum vulgare L.)(Anderson 1986). Greater seeding rates of safflower hastened the formation of dense canopies and improved crop competitiveness with associated weed species (Blackshaw and O'Donovan 1993). Increasing wheat seeding rate from approximately 60 to 90 kg ha⁻¹ helped combat resistant weeds and increase yields (Llewellyn et al. 2004). Because of the benefits, increased seeding rate is now a standard practice in Australia's dryland crop production (Walsh and Powles 2007).

Row Spacing

In most crops, a narrower row spacing can increase the competitiveness of a crop by allowing the crop to form a canopy more quickly and intercept more light relative to the associated weeds (Arce et al. 2009; Norsworthy and Oliveira 2004; Norsworthy et al. 2007). Reduced row crop width has been shown to favor crop development at the expense of weeds in soybean (Anaele and Bishnoi 1992; Freed et al. 1987; Harder et al. 2007), cotton (Vories et al. 2001), peanut, and corn (Froud-Williams 1988). Gunsolus (1990) reported that soybean was better suited than corn to benefit from the competitive advantages offered by narrow row spacing and late planting because it normally reaches canopy closure more slowly than corn does. Cotton planted in 53-cm rows produced greater yields than did cotton planted in rows of 79 or 106 cm and required only 6 wk of weed-free maintenance for maximum yield, whereas cotton planted in 79- and 106-cm rows required 10 and 14 wk, respectively, of weed-free conditions to obtain optimum yield (Rogers et al. 1976).

Soil Fertility

Competition in the rhizosphere for nutrients and moisture is particularly important for crop vigor and competitiveness with associated weed species. The relative efficiencies of nutrient acquisition by crops and weeds may be responsible for differences in aboveground competition for light (Aldrich 1984; Siddiqi et al. 1985). Soil fertility can affect weed management by increasing crop vigor, which improves competitiveness of the crop with associated weeds. However, species composition and competitive ability of weeds can also be affected. Weeds usually take up fertilizer more rapidly than do crops (Alkämper 1976). Corn plants growing with pigweed (Amaranthus spp.) contained only 58% as much nitrogen (N) as weed-free corn plants (Vengris et al. 1955). Crop varieties vary in their relative competitiveness for N. In competition with yellow foxtail [Setaria pumila (Poir.) Roemer & J.A. Schultes], early maturing corn hybrids were more competitive for N than were late-maturing hybrids (Standifer and Beste 1985).

Nutrient management can be used to directly manipulate weed populations (Pyšek and Lepš 1991). Fertilizer treatments may be used in fallow years to promote weed seed germination; a grower then controls the seedlings, thereby reducing the amount of seed in the soil seedbank. Placement and timing of fertilizer can be manipulated to reduce weed interference in crops (Blackshaw et al. 2003, 2004a,b; Kirkland and Beckie 1998; Melander et al. 2003; Rasmussen et al. 1996). Blackshaw et al. (2004b, 2008) and Kirkland and Beckie (1998) reported that N fertilizer applied as a subsurface band, rather than broadcast, reduced competition from several weed species, including wild oat, green foxtail, wild mustard (Sinapis arvensis L.), and common lambsquarters (Chenopodium album L.). Fertilizer applied in spring, rather than in fall, often reduced weed biomass and increased yield of barley, wheat, and garden pea (Pisum sativum L.) (Blackshaw et al. 2004b, 2005). Not all weeds, however, respond equally to N. Competitiveness of redroot pigweed (Amaranthus retroflexus L.), an N-responsive species, increased as N rate increased. However, wild oat competitiveness was unaffected by N fertilizer rate (Blackshaw and Brandt 2008). This result suggests that fertilizer management strategies should be based on weed management as well as crop yield.

Irrigation

Water early in the season is important as a weed management tool to promote healthy crop growth and improve the ability of the crop to compete with associated weeds (Zimdahl 1971). Water management practices are especially important for weed control in rice production (Caton et al. 2002). Flooding rice fields is primarily for weed control because rice can grow under flooded conditions where water-saturated soil limits oxygen availability, but many weeds cannot grow under these low-oxygen conditions. Effective water management is critical in IWM programs for rice (Bhagat et al. 1996; Gealy et al. 2003), and control of many weeds depends on flood timing and depth (Rao et al. 2007; Williams et al. 1990), as well as on weed size and herbicide use. For example, for red rice control, flooded conditions needed to be established within 14 d after early POST application of imazethapyr but within 7 d after late POST application (Avila et al. 2005b).

The type and amount of irrigation water can also influence weed seed germination. Weed emergence in tomato was 46 to 96% lower with subsurface drip irrigation than it was with furrow irrigation, and weed emergence in furrows with the drip irrigation system was almost eliminated (Shrestha et al. 2007). In garden lettuce (Lactuca sativa L.), preplant irrigation and shallow tillage 14 d later reduced in-crop weed density by up to 77% and allowed effective use of lower herbicide rates (Shem-Tov et al. 2006).

Allelopathy

The term allelopathy has been used since 1937, when Molisch (1937) first described the phenomenon of plants affecting one another through the release of toxic chemical agents. A well-documented case of allelopathy is the production of juglone by black walnut (Juglans nigra L.) to eliminate surrounding vegetation (Weston and Duke 2003). Allelopathy has been recognized as having untapped potential as a weed management tool (Crutchfield et al. 1985; de Almeida 1985; Purvis et al. 1985; Putnam 1994; Smeda and Putnam 1988). Allelopathic potential exists in both weeds and crops. Rice (Dilday et al. 1991), wheat, barley, sorghum (Belz 2007; Putnam and DeFrank 1985), rye (Burgos and Talbert 2000; Barnes and Putnam 1986; Dhima 2006), wild radish (Raphanus raphanistrum L.)(Norsworthy 2003), and several Brassicaceae species (Haramoto and Gallandt 2005: Norsworthy et al. 2011) have been shown to have allelopathic potential. Species in the Brassicaceae family have received attention because of the known suppressive ability of isothiocyanates, which are derived from glucosinolates, produced by these species after the plants are crushed or decomposed. Norsworthy and Meehan (2005) tested isothiocyanates for control of Palmer amaranth (Amaranthus palmeri S. Wats.), pitted morningglory (Ipomoea lacunosa L.), and yellow nutsedge (Cyperus esculentus L.) and found several that suppressed emergence of weeds, reducing even yellow nutsedge emergence 95%. In other studies, barley, oat, wheat, and cereal rye residues reduced total weed biomass and the weight of several indicator species (Putnam and DeFrank 1985). Allelopathic activity could be attained through use of allelopathic cover crops, allelochemicals as natural herbicides, or allelopathic crops (Weston and Inderjit 2007). Although much is known about allelopathy, future research is needed on the mechanisms of allelochemical selectivity, the modes of release, and the environmental and fertility effects on activity, persistence, and potential for synthesizing bioactive products as herbicides. Allelopathic crops open the potential for development of higher levels of weed suppression through conventional breeding or biotechnology (Singh et al. 2003).

Mechanical Methods with an Emphasis on Tillage

Before crop emergence, the type of tillage employed by a grower, whether no tillage, ridge tillage (cleaning 3 to 6 cm of soil off the top of the planting bed, or ridge, that was formed in the autumn), harrowing (shallow disturbance of the soil with one of several types of harrows), moldboard plowing (deep cutting and inverting the soil layer), or some other form, has a profound effect on crop and weed interactions. More specifically, tillage affects the crop's competitive ability and the type of weed species present, and it defines the agroecosystem in which the crop and weeds exist (Malhi et al. 1988). Disadvantages of tillage as a weed control method are potential crop injury with in-season tillage, increased disease incidence, and lack of residual control (Wiese and Chandler 1988). Additionally, selective cultivation is dependent on the weather and if not performed in a timely manner, ineffective. Tillage also is a major reason for loss of soil moisture and increased water and wind erosion. For many years, however, tillage was the primary method of weed control. Moss (1979, 1980, 1985) compared cultivation systems (plowing, tine cultivation, and direct drilling) for control of blackgrass in spring barley. Blackgrass infestations in tine-cultivated wheat were similar to those in direct-drilled wheat. In contrast, plowing buried blackgrass seeds from the previous crop, reducing blackgrass populations by burying the seed to a depth from which it could not emerge. In a similar experiment, shallow cultivation reduced weed seed populations in cereals by 34%, and plowing to a depth of 20 cm eliminated blackgrass infestations (Froud-Williams 1981).

The effect of conservation-tillage systems on weed population dynamics has been evaluated in numerous studies. Wicks et al. (1994) reported the results of a 40-yr study in North Platte, NE, in which tillage shifted from plowing to disking to ridge tillage. In the plow system, kochia [Kochia scoparia (L.) Schrad.], redroot pigweed, and green foxtail were predominant. Annual grasses dominated when disking supplanted plowing as the tillage system, whereas ridge tillage favored the emergence of winter annual broadleaf weeds. Similarly, in conservation-tillage fields, there was an increase in the density of perennial weeds, annual grasses, windblown weeds, and native plant species that were not normally found in cultivated fields (Donaghy 1980). Biennial and perennial weeds increased in conservation-tillage sorghum. In Brazil and Argentina, annual weed populations declined in zero-tillage systems (Ferrando et al. 1982), but Zentner et al. (1988) found that annual grass weeds were favored by conservation-tillage systems.

In 1983, Froud-Williams et al. hypothesized that windblown weed species may be favored by conservation-tillage systems, and this has since been shown in several studies. Derksen et al. (1993) found wind-dispersed weeds to be more numerous in no-till treatments. Fay (1990) reported that weeds in the Poaceae (grass) family with wind-borne seeds that were not well adapted to burial were favored in conservation-tillage systems. It was theorized that larger and heavier seeds have greater biological reserves to emerge from greater depths in the soil compared with smaller seeds.

Hartmann and Nezadal (1990) observed that weed cover in the field was reduced from 80% when tillage was performed in daylight to 2% after 7 yr when all tillage operations were performed between 1 h after sunset and 1 h before sunrise. The success of this system is attributed to the weed seeds requiring light for germination by means of a response to the photoreceptor, phytochrome (Buhler 1997; Mancinelli 1994; Smith 1995). The germination of many crop plants, however, is light independent. The authors hypothesized that this type of system could reduce herbicide inputs dramatically.

Stale seedbed is a conservation-tillage technique that combines cultural, mechanical, and chemical weed control methods. It involves disking or harrowing a field well before seeding, applying a nonselective herbicide to weeds that germinate, and planting the crop with minimal soil disturbance to reduce weed seed germination (Elmore and Moorman 1988). The technique incorporates the use of timely planting, irrigation, and effective weed control with preplant herbicides. Unless continual weed control without soil disturbance is used, a stale seedbed is a means of creating a weed-suppressing mulch by leaving vegetative residue on the surface and reducing light penetration to the soil surface. Large crabgrass [Digitaria sanguinalis (L.) Scop.] was eliminated from the upper 2 cm of soil, and populations of flatsedge (Cyperus ssp.) and bluegrass (Poa ssp.) species were significantly reduced using stale seedbed techniques (Standifer and Beste 1985).

The “judicious use of tillage” is a tactic that can delay or manage resistance (Beckie and Gill 2006). However, in some countries, especially the United States, Argentina, and Brazil, the adoption of glyphosate-resistant crops made it possible for producers to adopt conservation-tillage production, including no-tillage (Duke and Powles 2009; Givens et al. 2009). Conservation tillage has many environmental and economic advantages, but the reliance on glyphosate as a sole method of weed control reduced the use of mechanical weed control, which reduced weed-control diversity, and, in turn, placed great selection pressure on weeds in glyphosate-resistant crops (Duke and Powles 2009). The many benefits from conservation tillage, however, make it even more important to adopt integrated management practices so that conservation tillage will be a sustainable production practice.

Biological Weed Control

Biological control of weeds is broadly defined as the use of a biological agent, a complex of agents, or biological processes to bring about weed suppression (WSSA 2007). Biological control has been used successfully as a practical and economically affordable weed control method in many situations. However, it should be emphasized that biological control has seldom been used in annual agricultural situations so will likely not have a role to play in controlling herbicide-resistant weeds in herbicide-resistant crops. Classical biological control, which is biological control of nonnative, invasive weeds with natural enemies originating from the native range of the weed, has proven a viable strategy for managing weeds in areas subjected to low-intensity management, such as rangelands, forests, preserved natural areas, and some waterways. The use of an inundative method, also called the bioherbicide strategy, where an organism is applied to achieve rapid reduction in weed populations, has also proven successful in some instances. In the future, pathogens may also be used to introduce or alter specific genes to control growth, flowering, seed set, and competitiveness of weeds.

Appleby (2005) provided an excellent history of biological weed control in his history of weed control in the United States and Canada. The cases of the cactus moth (Cactoblastis cactorum Berg) on pricklypear cactus (Opuntia spp.) and the success of the Klamath weed beetle (Chrysolina quadrigemina Suffrian) in reducing a population of common St. Johnswort (Hypericum perforatum L.) are classic cases of biological weed control. The cinnabar moth [Tyria jacobaeae L.) was marginally successful against tansy ragwort (Senecio jacobaea L.) in western North America, but the addition of the ragwort flea beetle (Longitarsus jacobaeae Waterhouse) to help the cinnabar moth has been dramatically successful. Other success stories include alligatorweed flea beetle (Agasicles hygrophila ) on alligatorweed [Alternanthera philoxeroides (Mart.) Griseb], mottled waterhyacinth weevil (Neochetina eichhorniae) on water hyacinth [Eichhornia crassipes (Mart.) Solms], and the salvinia weevil (Cyrtobagous salviniae Calder and Sands) on giant salvinia (Salvinia molesta Mitchell). Many other insects have been introduced with varying degrees of success.

The first commercially available mycoherbicide, Phytophthora palmivora Butl. (trade name, Devine), was introduced in 1981 for control of stranglervine [Morrenia odorata (Hook. & Arn.) Lindl.] in citrus (Citrus spp.). In 1982, the causal agent of mango anthracnose (Colletotrichum gloeosporioides Penz., f . sp. aeschynomene [trade name, Collego]) was introduced for control of northern jointvetch [Aeschynomene virginica (L.) B.S.P.] in rice and soybean. Considerable work continues in the search for mycoherbicides, and a number of promising leads have developed, but biological effectiveness is only one requirement for commercial success. BioMal [Colletotrichum gloeosporioides (Penz.) Penz. and Sacc. f. sp. malvae), for control of common mallow (Malva neglecta Wallr.), was the first bioherbicide introduced in Canada, but production was discontinued because production costs were too high for commercial acceptance. In addition, generalist fish, such as grass carp (Ctenopharyngodon idella Val.), have been used to graze down biomass of aquatic plants without specific target weeds. A few other vertebrate species, such as geese (Anseriformes; Anserinae), goats, and sheep have been used to remove weeds from localized areas (DeBruin and Bork 2006). More than 1,000 releases of more than 350 biological control agents have been used against more than 100 target weed species around the world since the late 1800s (Julien and Griffiths 1998).

Herbicide Classification Based on Timing of Application: Soil-Applied Herbicides

Soil-applied herbicides generally affect seed emergence or the growth of weed seedlings and must persist in the soil to be effective. When applied before the crop is planted, these herbicides are referred to as preplant or presowing herbicides. Some preplant herbicides must be incorporated into the soil to be effective and are referred to as preplant incorporated (PPI) herbicides. Preplant herbicides are applied from a few days to several months before crop sowing, depending on their soil persistence and tolerance of the crop to be planted. Herbicides applied at planting or within a few days before crop emergence are referred to as preemergence (PRE) herbicides. In row crops, preplant and PRE herbicides can be applied in a band over the crop row to reduce herbicide costs, especially if cultivation will be used to control weeds between the rows. A soil-applied herbicide, in some cases can also be applied after the crop is established (POST) to lengthen residual weed control in the crop. In glyphosate- or glufosinate-resistant crops, for example, metolachlor can be sprayed POST, usually in a tank mixture with glyphosate or glufosinate. Metolachlor has almost no activity on emerged plants but provides residual control between applications of the broad-spectrum herbicides glyphosate or glufosinate, which have no residual activity.

Soil-applied herbicides are important as assurance that weeds will not emerge with the crop and be too large to control with the first POST application. Before the release of glyphosate-resistant cotton, soil-applied herbicides were especially important in cotton because there was no broad-spectrum POST herbicide that could be applied over-the-top of the crop. Because of resistant weeds, the use of residual soil-applied herbicides in corn and soybean has increased over the past 5 to 10 yr (Owen et al. 2011). Modeling of resistance in Palmer amaranth indicated that resistance could be delayed with the use of soil-applied herbicides (Neve et al. 2010, 2011). A soil-applied herbicide can also introduce another MOA into an integrated resistance-management program.

Foliar-Applied Herbicides

Foliar-applied herbicides are applied to weed foliage, with or without contact of the spray with the crop, and are effective generally against young weed seedlings. POST herbicides are generally considered to be those applied after crop emergence. The spray can be applied broadcast over the crop and weeds, directed to the weeds at the base of the crop if there is limited crop selectivity, or applied under shields if there is no crop selectivity. Foliar sprays also are used for controlling emerged weeds present at planting in conservation-tillage systems, referred to as burndown herbicides. Foliar-applied herbicides are referred to as contact herbicides when only the treated part of the plant is affected and are called systemic or translocated when the herbicide enters the plant and moves within it to the site of herbicide action. Translocation can be either through the phloem, which carries the herbicides to aboveground and belowground growing points, or through the xylem, where they move with the transpiration stream and accumulate at leaf margins.

Herbicide Classification Based on MOA

Good resistance-management programs recommend diversification of herbicide MOAs as a key resistance-management strategy. In 1997, herbicides were classified by MOA (Retzinger and Mallory-Smith 1997) with the idea that if herbicides with a similar MOA were placed in groups, it would be easier to recommend and use appropriate herbicides for resistance management. Each MOA group is assigned a number. Within each MOA group are the families of herbicides with that MOA. Families are groups of herbicides with similar chemical properties and activity; e.g., pendimethalin, trifluralin, and ethalfluralin are members of the dinitroaniline family). A complete list of MOA groups with site of action and the number of resistant weed biotypes and species is included in Table 1. A complete list of weed species that have been documented to have herbicide-resistant biotypes, with the corresponding list of herbicide MOAs to which resistance has been documented and the number of states in which they occur, is included in Table 2. Below is a summary of herbicide MOAs (WSSA 2012).

Table 1.

Site of action and Weed Science Society of America mode of action group for herbicide-resistant weeds in the United States based on the number of resistant weed biotypes (summed across states) and number of the weed species (Heap 2012).

Table 2.

Weed species documented to have herbicide-resistant biotypes in the United States, the herbicide site of action for which resistance has been documented, and the number of states where the species have been documented (Heap 2010).

Table 2.

Continued.

Group 1: Inhibitors of Acetyl-Coenzyme A Carboxylase

The acetyl-coenzyme A carboxylase (ACCase)-inhibiting herbicides include the cyclohexanedione, phenoxyproprionate, and phenylpyrazolin herbicides. These herbicides block the first committed step in de novo fatty acid synthesis (Burton et al. 1989; Focke and Lichtenthaler 1987). The growth of susceptible plants is halted, and plants gradually die because of the absence of phospholipids for building new cell membranes.

Group 2: Inhibitors of ALS

ALS inhibitors are also called acetohydroxy acid synthase inhibitors. These herbicides are inhibitors of a common enzyme leading to the synthesis of the branch-chain amino acids leucine, valine, and isoleucine (Devine et al. 1993). ALS inhibitors are used in all of the major agronomic crops (i.e., corn, cotton, pastures, peanut, small grains, soybean, turf, and wheat) and typically have residual activity. Symptoms of these herbicides include growth cessation, internodal shortening, purple foliage, and shortened lateral roots (“bottle-brush” roots). Five herbicide families comprise the ALS-inhibiting herbicides: the sulfonylureas, imidazolinones, triazolopyrimidine sulfonanilides, pyrimidinylthiobenzoic acids, and sulfonylaminocarbonyltriazolinones.

Groups 5, 6, and 7: Photosynthesis Inhibitors

The phenylcarbamates, pyridazinones, triazines, triazinones, and uracils (group 5); benzothiadiazinones, nitriles, and phenylpyridazines (group 6); and amides and ureas (group 7) include herbicides that inhibit photosynthesis by binding to the Qb-binding niche of the D1 protein of the photosystem II complex in chloroplast thylakoid membranes. A distinction among groups within the photosynthetic inhibitors refers to the different pockets where they bind in the Qb-binding site. Herbicide binding in this site negatively affects processes and products necessary for the transport of chemical energy. Death in susceptible plants generally occurs via cell membrane disintegration because of the creation of triplet state chlorophyll, singlet oxygen, and lipid peroxidation.

Group 22: Photosystem I inhibitors (Electron Diverters)

The bipyridyliums (e.g., paraquat, diquat) are examples of herbicides that accept electrons from photosystem I and reduce them to form a herbicide radical. The radical then reduces molecular oxygen to form superoxide radicals. The superoxide radicals lead to the formation of hydrogen peroxide and hydroxy radicals that destroy lipid membrane fatty acids and chlorophyll, thereby causing membrane destruction, cellular leakage, and a very rapid plant death in sunlight.

Group 14: Inhibitors of Protoporphyrinogen Oxidase

These chlorophyll synthesis inhibitors (PPO-inhibitors) comprise several herbicide families: the diphenylethers (e.g., fomesafen), N-phenylphthalimides (e.g., flumioxazin), oxadiazoles (e.g., oxadiargyl), oxazolidinediones, phenylpyrazoles (e.g., flufenpyr-ethyl), pyrimidindiones (e.g., butafenacil), thiadiazoles, and triazolinones (e.g., azafenidin, sulfentrazone). Herbicides in this category inhibit protoporphyrinogen oxidase (PPG oxidase or Protox), an enzyme involved in chlorophyll and heme synthesis. Inhibition leads to an accumulation of PPIX, the first light-absorbing chlorophyll precursor. Light absorption by PPIX apparently produces triplet state PPIX that interacts with ground state oxygen to form singlet oxygen. Triplet PPIX and singlet oxygen lead to the formation of lipid radicals that initiate a chain reaction of lipid peroxidation and cause cellular leakage and plant death (Duke et al. 1991).

Groups 11, 12, 13, and 27: Carotenoid Biosynthesis Inhibitors

Herbicides in group 12 include the amides, anilidex, furanones, pheno`xybutan-amides, pyridazinones, and pyridines that block carotenoid synthesis by inhibition of phytoene desaturase (Bartels and Watson 1978; Sandmann and Böger 1989). Carotenoids play an important role in dissipating the oxidative energy of singlet oxygen and other radicals. Callistemon, isoxazole, pyrazole, and triketone herbicides (group 27) inhibit a key step in plastoquinone biosynthesis; its inhibition gives rise to bleaching symptoms on new growth. Symptoms on plants result from an indirect inhibition of carotenoid synthesis because of the involvement of plastoquinone as a cofactor of phytoene desaturase. Clomazone (group 13) is metabolized to the 5-keto form of clomazone that is herbicidally active. The 5-keto form inhibits 1-deoxy-d-xylulose 5-phosphate synthase, a key component of plastid isoprenoid synthesis for production of carotenoids (Ferhatoglu and Barrett 2006). Amitrol and aclonifen make up group 11. Amitrol inhibits accumulation of chlorophyll and carotenoids in light, and aclonifen acts similarly to carotenoid inhibiting/bleaching herbicides, although the specific mechanisms of action of these herbicides have not been determined.

Group 9: Enolpyruvyl Shikimate-3-Phosphate Synthase

Group 9, the glycines, includes only glyphosate. Glyphosate inhibits enolpyruvyl shikimate-3-phosphate synthase (EPSPS)(Amrhein et al. 1980), which leads to depletion of the aromatic amino acids tryptophan, tyrosine, and phenylalanine, all needed for protein synthesis or for biosynthetic pathways leading to growth.

Group 10: Glutamine Synthetase Inhibitors

Herbicides in group 10 are glufosinate and bialaphos, the phosphinic acids. They inhibit the activity of glutamine synthetase (Lea 1984), the enzyme that converts glutamate and ammonia to glutamine. Accumulation of ammonia in the plant (Tachibana et al. 1986) destroys cells and directly inhibits photosystem I and photosystem II reactions (Sauer 1987). Ammonia also reduces the pH gradient across the membrane that can uncouple photophosphorylation.

Group 18: Dihydropteroate Synthase Inhibitor

A unique carbamate herbicide, asulam, appears to inhibit cell division and expansion in plant meristems, perhaps by interfering with microtubule assembly or function (Fedtke 1982; Sterrett and Fretz 1975). Asulam also inhibits 7,8-dihydropteroate synthase, an enzyme involved in folic acid synthesis and crucial for purine nucleotide biosynthesis (Kidd et al. 1982; Veerasekaran et al. 1981).

Groups 3, 15, and 23: Inhibitors of Mitosis

Benzamide, dinitroaniline, phosphoramidate, and pyridine compounds are group 3 herbicides that bind to tubulin, a major microtubule protein. The herbicide–tubulin complex inhibits polymerization of microtubules at the assembly end of the protein-based microtubule leading to a loss of microtubule structure and function, and cell wall formation is negatively affected. Dinitroaniline herbicides are volatile, readily dissipating into the atmosphere and. therefore. often require soil incorporation (either disked into the soil or through irrigation) for maximum efficacy. The carbamate herbicides, carbetamide, chlorpropham, and propham are group 23 compounds and inhibit cell division and microtubule organization and polymerization. Acetamide, chloroacetamide, oxyacetamide, and tetrazolinone compounds are group 15 herbicides that are thought to inhibit very long chain fatty acid synthesis (Böger et al. 2000; Husted et al. 1966). These compounds affect susceptible weeds before emergence but do not inhibit seed germination.

Groups 20, 21, 28, and 29: Cellulose Inhibitors

The benzamides (group 21), nitriles (group 20), and triazolocarboxamides (group 28) are herbicides that inhibit cellulose and cell wall synthesis in susceptible weeds (Heim et al. 1990). Indaziflam is an alkylazine compound (group 29) that disrupts cellulose biosynthesis (Myers et al. 2009).

Groups 8 and 16: Fatty Acid and Lipid Biosynthesis Inhibitors

The benzofuranes (group 16) and the phosphorodithioates and thiocarbamates (group 8) are herbicides known to inhibit several plant processes, including biosynthesis of fatty acids and lipids, proteins, isoprenoids, flavonoids, and gibberellins. It is currently thought that these effects may be linked by the conjugation of acetyl-coenzyme A and other sulfhydryl-containing molecules by thiocarbamate sulfoxides (Casida et al. 1974; Fuerst 1987). Herbicides with this MOA may be referred to as the lipid and secondary biosynthesis inhibitors. To maximize efficacy in controlling weeds, thiocarbamates are typically incorporated into the soil. Seedlings of susceptible weed species fail to emerge, and phytotoxic symptoms are characterized by tightly whorled leaves that do not unroll in a normal fashion.

Group 4: Synthetic Auxins

The benzoic acids, phenoxycarboxylic acids, pyridine carboxylic acids, and quinoline carboxylic acids collectively form the group of synthetic auxin herbicides. Herbicides in this group have activity similar to that of endogenous auxin, although the true mechanism is not well understood. Symptoms include abnormal growth and eventual plant death. The oldest selective herbicide, 2,4-D, is categorized in this group.

Group 19: Auxin Transport Inhibitors

The phthalamates (e.g., naptalam) and semicarbazones (e.g., diflufenzopyr) are group 19 herbicides. These herbicides inhibit polar transport of naturally occurring auxin, indoleacetic acid, and synthetic auxin-mimicking herbicides in sensitive plants. Auxin-transport inhibition causes an abnormal accumulation of auxin in the meristematic regions, thereby disrupting the auxin balance needed for plant growth (Grossmann 2010).

Groups 17, 25, and 26: Potential Nucleic Acid Inhibitors or Unclassified Herbicides

Herbicides in these groups have an unknown MOA.

Knowledge of basic weed science principles and practices are needed for understanding herbicide resistance and for planning and implementing sound IWM programs. The last new herbicide MOA discovery was inhibition of the enzyme hydroxyphenylpyruvate dioxygenase, which was discovered more than 20 yr ago (Lee et al. 1997) and from which the herbicide mesotrione was developed. Before this, a new MOA was introduced approximately every 3 yr (Duke 2012). The search for new herbicide chemistry has slowed to the extent that producers can no longer count on “the next new herbicide” to control resistant weeds. Instead, herbicide programs must be integrated with preventive, cultural, biological, and mechanical weed control practices to develop diversified resistance management programs.

PSII-Inhibiting Herbicides

Photosystem II-inhibiting herbicides, such as the triazines, block photosynthetic electron transfer from chlorophylls associated with the P680 chlorophyll center to the initial electron acceptor, a compound known as plastoquinone. In a normal functioning plant, plastoquinone accepts an electron from the chlorophyll in the initial step of the electron-transfer process. This transfer occurs at the Qb binding site of a 32-kDa protein known as D1 of photosynthesis. Triazines and other PSII herbicides compete with plastoquinone for the binding site on the Qb protein. There are several mutations to the Qb protein that prevent PSII herbicides from binding while allowing plastoquinone to bind, albeit with reduced efficiency. The most prevalent mutation occurs at the Ser 264 and involves an amino acid change from serine to glycine. Other mutations conferring herbicide resistance are Ser 264 to Thr and Asn 266 to Thr (Oettmeier 1999; Powles and Yu 2010).

ACCase-Inhibiting Herbicides

Acetyl-coenzyme A carboxylase is the key enzyme in lipid biosynthesis. In plants, there are two forms of ACCase, a prokaryotic form made up of multiple subunits and eukaryotic ACCase that is a large multidomain protein. Plants contain cytosolic as well as plastidic forms of ACCase. In grasses, the plastids contain the eukaryotic form of ACCase and are sensitive to three chemical classes of herbicides known as the graminicides. Most dicot plant species contain the prokaryotic form of the enzyme that is insensitive to graminicide herbicides.

There are eight mutations to the ACCase enzyme leading to target-site resistance to ACCase-inhibiting herbicides (Powles and Yu 2010). The most common is a Leu to Ile substitution at residue 1781. Different substitutions in the ACCase enzyme lead to differing patterns of resistance among the graminicide herbicides.

Mitotic Inhibitors

Herbicides such as the dinitroanilines inhibit mitosis through the disruption of microtubule assembly. Specifically, microtubules comprise α-tubulin and β-tubulin dimers. In mitosis, the microtubules are involved in chromosomal migration to the daughter cell. Mitotic-inhibiting herbicides bind to the α-tubulin, preventing it from binding to the β-tubulin dimer and inhibiting dimer formation. In dinitroaniline-resistant weed species, a substitution of Ile for Thr at the 239 base pair prevents the dinitroaniline herbicide from binding to the α-tubulin (Smeda and Vaughn 1994). Other less-common mutations at amino acid position 268 (Met to Thr) and 136 (Leu to Phe) have been reported (Deleye et al. 2004).

PPO-Inhibiting Herbicides

Protoporphyrinogen IX oxidase is a key enzyme in the synthesis of chlorophyll and heme in plants. PPO catalyzes the oxidation of protoporphyrinogen (protogen) to protoporphyrin IX (Proto IX). In plants, there are two nuclear-encoded isoforms (PPO1 and PPO2). The PPO1 product is targeted to the chloroplast while the PPO2 product is targeted to the mitochondria. Several groups of herbicides, including diphenyl ethers, inhibit PPO. PPO resistance has been limited to one species, tall waterhemp [Amaranthus tuberculatus (Moq.) Sauer], with a novel target-site–resistance mechanism. In the PPO gene, there is a codon deletion at position 210 leading to the loss of a Gly residue conferring herbicide resistance (Patzoldt et al. 2006).

Enolpyruvyl shikimate-3-phosphate synthase

The target-site mechanism of glyphosate is inhibition of the enzyme EPSPS, which is a key enzyme in the shikimic acid pathway leading to the aromatic amino acids of phenylalanine, tyrosine, and tryptophan. Glyphosate is a competitive inhibitor of EPSPS, competing with phosphoenolpyruvate (PEP) at the catalytic site. Thus, there are few mutations that confer glyphosate resistance without leading to enzyme failure and a lethal mutation. There is only one reported mutation (a Pro to Ser mutation at amino acid position 106) in weeds that confer resistance to glyphosate via a target-site mutation, whereas most glyphosate-resistant weeds are resistant via nontarget-site resistance mechanisms. A special case of target-site resistance has been associated with glyphosate-resistance in some weed species where an amplification of the target site is noted (Gaines et al. 2010).

Target-Site Resistant Crops

Herbicide-resistant crops can be developed by whole-cell selection, mutagenesis, selection, and breeding from plants with natural resistance or by genetic engineering (Duke 2005; Tan and Bowe 2009). Imidazolinone-resistant crops are based on target-site mutations that reduce the sensitivity of the ALS enzyme to those herbicides. Most of these crops were created by chemical mutagenesis, where cells were mutated by chemical treatment, and then, crops were commercialized as Clearfield crops (BASF, 26 Davis Drive, Research Triangle Park, NC 27709) using conventional breeding methods. Imidazolinone crops include Clearfield corn (maize), canola, wheat, rice, and common sunflower (Helianthus annuus L.). Clearfield common sunflower was developed by breeding cultivated sunflowers with a resistant wild sunflower (Tan and Bowe 2009). Another example of breeding a crop from a wild relative is triazine-resistant canola, in which resistance from a triazine-resistant weed, birdsrape mustard (Brassica rapa L.), was moved into canola, which is also a Brassica species, by breeding (Hall et al. 1996).

Transgenic crops that have target-site resistance have been created by artificially inserting a gene or genes from another organism into the crop to give it a desirable characteristic that it does not naturally possess. Transgenic crops for herbicide resistance have been developed for glyphosate, glufosinate, and bromoxynil. (Bromoxynil-resistant crops were removed from the market because bromoxynil is not a broad-spectrum herbicide and the crops competed poorly in the market [Duke 2005].) Most glyphosate-resistant crops were developed by inserting a bacterial EPSPS known as CP4 from a Agrobacterium sp. bacterium, which encodes a glyphosate-insensitive form of EPSPS (Duke 2005; Powles and Preston 2006). The binding of glyphosate is excluded by conformational changes resulting from those amino acid sequence changes in CP4 outside the glyphosate/PEP binding region (Dill 2005).

Alfalfa

Alfalfa is a perennial, mainly outcrossing, insect-pollinated crop with no known compatible wild relatives in the United States (St. Amand et al. 2000; Van Deynze et al. 2004). However, feral alfalfa populations are common in areas where alfalfa is grown. The removal of feral populations could reduce the likelihood of gene flow, but that is not easily accomplished because of its perennial life cycle.

In the United States, alfalfa seed is produced primarily in the western states, on approximately 40,500 ha (Van Deynze et al. 2004). Insect-mediated pollination is necessary for alfalfa seed production (Fitzpatrick et al. 2003). In a study that evaluated gene flow from alfalfa fields, gene dispersal via pollen beyond the current isolation distances was reported (St. Amand et al. 2000). Outcrossing rates in seed production fields were 38%, whereas rates in hay fields were lower, but still greater than 25%.

Alfalfa seed is small, approximately 500 seed g⁻¹, which may increase the risk of commingling among alfalfa seed lots. Hard seeds, which are common in alfalfa, may lie dormant for years before germinating (Gunn 1972). Seed dormancy allows alfalfa to persist in the seedbank and produce volunteers in subsequent crops.

Alfalfa can be propagated by stem cuttings and alfalfa crowns can persist and re-grow (Busbice et al. 1972). Alfalfa crowns moved by machinery within and between fields could result in gene flow. Although no studies have reported vegetative gene flow in alfalfa, this mechanism of gene movement needs to be investigated.

Canola

Canola is an annual, self-fertile, and outcrossing species that is both insect- and wind-pollinated and has the potential to establish outside of cultivation. Canola can be either rapeseed (Brassica napus L.) or birdsrape mustard (Brassica rapa L.)(formerly Brassica campestris L.). In North America, most of the canola grown, and all of the herbicide-resistant canola, is B. napus.

Gene flow via pollen in canola is significant. Outcrossing rates as high as 47% have been reported (Williams et al. 1986). Canola pollen dispersal ranges from a few meters to 1.5 km (Timmons et al. 1995). Pollen movement depends on wind direction and speed, surrounding vegetation, and topography (Gliddon et al. 1994; Thompson et al. 1999). Bees are known to pollinate canola. Most bees forage close to their hive, but there are reports of bee movement up to 4 km (Mesquida and Renard 1982; Thompson et al. 1999). Because loose pollen grains can be picked up in a hive, a 4-km flying distance could result in pollen being moved 8 km. In Canada, gene movement between two genetically engineered canola lines was found at 800 m, which was the boundary of the study (Beckie et al. 2003). In a Canadian field, volunteer canola plants were identified that contained transgenes for both glyphosate- and glufosinate-resistant canola (Hall et al. 2000).

Although canola does not generally survive in undisturbed habitats, it can establish in areas adjacent to agricultural sites, roadsides, and field edges (Warwick et al. 1999). The occurrence of glyphosate- and glufosinate-resistant canola plants along railways and roadways in Canada was measured in 2005 (Yoshimura et al. 2006). In Saskatchewan, 34% of 300 canola plants tested were glyphosate-resistant; in British Columbia, 43% of 81 plants tested were resistant. One hybrid between B. rapa and B. napus was identified as glyphosate-resistant.

Volunteer canola plants can be a significant weed problem in subsequent crops (Kaminski 2001; Thomas et al. 1998). In 35 sampled fields, harvest seed loss ranged from 3 to 10%, with an average of about 6% or an equivalent of 107 kg ha⁻¹ (Gulden et al. 2003a). In general, canola seedbanks decline quickly, but some seed may persist for several years (Gulden et al 2003b). Canola seeds are reported to survive longer when buried (Pekrun and Lutman 1998).

Many studies have addressed gene flow via pollen from transgenic or conventional canola to weedy or wild relatives (Bing et al. 1996; Chèvre et al. 1998, 2000; Darmency et al. 1998; Jørgenson and Andersen 1994; Jørgenson et al. 1994, 1996; Lefol et al. 1995, 1996; Rieger et al. 2001; Warwick et al. 2003). The movement of a herbicide-resistance gene would lead to herbicide-resistant individuals and the likelihood that resistant populations could occur both within and outside of cultivated fields.

Corn (Maize)

Corn is an annual, highly outcrossing species that produces abundant pollen and is primarily wind-pollinated (Brittan 2006; Halsey et al. 2005). Corn has no compatible relatives in the United States, and there is no risk of gene flow that could result in a herbicide-resistant weed by hybridization. Because corn does not persist outside of cultivation, the main dispersal mechanisms for a herbicide-resistance gene are via pollen among neighboring corn fields and seed commingling.

Corn seed does not shatter, and corn rarely sheds its seed. However, corn seed are often scattered by harvest equipment, mature corn plants can lodge and drop ears, intact ears and individual kernels can be removed from fields by small mammals that forage for food, and corn seed can be scattered along commercial transportation routes. Corn grains or ears left in the field after harvest often create volunteer corn plants the following year (Tolstrup et al. 2003). Management of herbicide-resistant volunteer corn plants in continuous corn or soybean resistant to the same herbicide requires the use of additional or alternative herbicides and could become a significant economic problem in areas where corn is planted in rotation with soybean and cotton (Deem et al. 2006; Clewis et al. 2008).

Cotton

Cotton can be either self-pollinated or cross-pollinated by insects. Two species of cotton, upland (G. hirsutum) and Pima (Gossypium barbadense L.), are grown in the United States, with the largest proportion being upland cotton.

A range of outcrossing rates and a potential for insect pollination provide opportunity for some gene flow via pollen among cotton fields (Meredith and Bridge 1973). Reported outcrossing rates vary widely and depend on the number of pollinators present in a field and, possibly, on the cotton cultivar (Green and Jones 1953; Simpson and Duncan 1956). The introduction of the transgenic, insect-resistant Bacillus thuringiensis (Bt) cotton may lead to an increase in outcrossing rates because of the greater survival of pollinators as a result of fewer insecticide applications.

Volunteer cotton plants occasionally establish in subsequent crops, but they generally do not survive the winter, except in unusually mild winters in the southernmost U.S. states (Wozniak 2002; York et al. 2004). Because so much of the cotton grown in the United States is herbicide-resistant, the greatest potential for commingling of herbicide-resistant cotton seeds with non–herbicide-resistant cotton seeds may occur during ginning if both types are processed at the same facility.

Grain Sorghum

The Sorghum genus includes (Sorghum versicolor Anderss.), a wild African grass (sometimes called bruinsaadgras); Sorghum bicolor (L.) Moench., which includes all sorghums, shattercane, and Sudangrass; sorghum-almum (Sorghum almum Parodi), a weak perennial weed; and johnsongrass [Sorghum halepense (L.) Pers.], a robust perennial and formidable weed throughout its range across the humid areas and irrigation ditches of the southern United States and Ohio River Valley.

The sorghums are wind-pollinated. All the cultivated sorghums are interfertile and spontaneously cross in the field. Arriola and Ellstrand (1996) reported that johnsongrass and grain sorghum cross-pollination rates can be as high as 100% at distances of 0.5 to 100 m. Grain sorghum cultivars with resistance to the ALS- and ACCase-inhibitor herbicides have been developed and licensed by Kansas State University. These herbicide-resistant cultivars are expected to be commercialized. Biotypes of johnsongrass that are resistant to ALS- and ACCase inhibitors, the dinitroanilines, and glyphosate have been reported in the midsouth region of the United States. The emergence and spread of herbicide-resistant populations of johnsongrass is a matter of economic concern in all row crops and along roadsides and other noncrop areas in the South, especially in the midsouth and southern plains. In this respect, the release of herbicide-resistant cultivars of sorghum that may contribute to evolution of new populations of herbicide-resistant johnsongrass is a concern.

Rice

Rice is an annual, predominantly self-pollinated crop with minimal cross-pollination potential; yet, gene flow of herbicide-resistance traits is a global issue because of the widespread occurrence of weedy rice in rice-producing regions and wild relatives in Asia, Africa, and Central America (Delouche et al. 2007). The herbicide-resistant rice technology was developed primarily to control weedy rice in rice production.

Gene flow by seed is a concern because of seed movement from field to field via irrigation systems, farm machinery, and grain trucks. Elimination of the risk of commingling between non–herbicide-resistant and herbicide-resistant rice cultivars is practically impossible because of the bulk handling of rice grains from the farm to the shipping docks. The rice complex comprises seven wild and weedy species, five of which—rice (Oryza sativa L.), brownbeard rice/Indian wild rice (Oryza rufipogon Griffiths/Oryza nivara S.D. Sharma & Shastry), African rice (Oryza glaberrima Steud.), wild rice (Oryza barthii A. Chev.), and longstamen rice (Oryza longistaminata A. Chev. & Roehr.)—are weeds in rice fields worldwide (Vaughan et al. 2003). The Oryza species complex can interbreed (Vaughan et al. 2005); thus, their coexistence with cultivated rice results in gene flow via pollen and has produced morphologically and biologically diverse weedy rice populations (Shivrain et al. 2010a; Shivrain et al. 2010b), which reduces the overall efficacy of management tactics. To manage this most difficult weed problem in rice (that is, weedy rice), herbicide-resistant rice was developed. The nontransgenic, herbicide-resistant rice (Clearfield), which is resistant to the imidazolinone herbicides among the ALS inhibitor MOA group, is now widely adopted in the southern United States, with growers in Central and South America following closely behind. This technology is highly effective in controlling weedy rice with control generally above 95% (Avila et al. 2005a; Levy et al. 2006; Ottis et al. 2004; Steele et al. 2002). The bottleneck of this technology in rice are twofold: (1) the escape of herbicide-resistant trait to the weedy or wild relative, which produces herbicide-resistant weedy rice, and (2) selection for herbicide-resistant biotypes of not only weedy rice but also other major weed species in rice that are otherwise controlled by the ALS-inhibitor herbicides (e.g., flatsedge [Cyperus L.] spp., cockspur/barnyardgrass [Echinochloa P. Beauv.] spp., fringerush [Fimbristylis Vahl] spp.). Gene flow from herbicide-resistant rice (Clearfield and Liberty Link [Bayer CropScience, Monheim am Rein, Germany]) to weedy rice is well documented in Asia, Europe, and the Americas, showing pollen flow rates from 0.003% to 0.25% (Chen et al. 2004; Gealy 2005; Lentini and Espinosa 2005; Messeguer et al. 2004; Noldin et al. 2002; Sankula et al. 1998; Shivrain et al. 2006, 2007, 2008; Song et al. 2003; Zhang et al. 2003). These seemingly low numbers, however, have resulted in the presence of herbicide-resistant weedy rice in Arkansas rice fields with at least 2 yr of history with Clearfield rice (Singh et al. 2012). Herbicide-resistant weedy rice resulting from gene flow seems more prevalent in the tropical regions of Central America and South America (B. E. Valverde, personal communication), where there is no winter kill, and rice is generally planted twice a year.

Now that the acreage of hybrid rice in the southern United States has increased significantly, it is important to note that, whereas the highest outcrossing rate with Clearfield nonhybrid rice is 0.25%, with Clearfield hybrid rice, it is up to 1.26% (Shivrain et al. 2009b). Prolonging the utility of this technology rests on the following basic weed management principles: (1) adoption of BMPs that ensure maximum efficacy of the herbicide, (2) minimizing the synchronization of flowering between herbicide-resistant rice and weedy rice (by adjusting planting dates and knowing the phenology of weedy rice relative to the rice cultivars), (3) preventing escaped weedy rice in a Clearfield rice field from producing seed, and (4) preventing volunteer rice or weedy rice from producing seed in the next crop cycle by controlling it in a rotational crop, such as soybean. It is important to consider that once the resistance trait is introgressed into the weedy rice population, it can be transferred to other weedy rice populations or to non–herbicide-resistant rice cultivars (Shivrain et al. 2009a). Thus, weedy rice needs to be controlled not only in the rice paddies but also along field edges and irrigation ditch banks.

How far should rice fields or weedy rice populations be from each other to prevent pollen-mediated gene flow? In general, effective pollen flow is detectable within 1 m and declines significantly beyond that (Messeguer et al. 2001; Shivrain et al. 2007; also, see review by Gealy 2005). Effective pollen flow between cultivated rice and Indian wild rice (O. rufipogon) was detected up to 40 m (Song et al. 2003). As the population size of the pollen donor increases, effective pollination distance also increases. In field-scale, Clearfield rice-pollen flow experiments, outcrossing was detected up to 297 m (Burgos et al. 2010).

Soybean

Soybean is an annual, highly self-fertile, self-pollinating species (Caviness 1966). Pollen-mediated gene flow from herbicide-resistant soybean is considered a low risk (Gealy et al. 2007). Soybean is generally not found outside of cultivation (OECD 2000), and it has no compatible relatives in the United States. Nevertheless, cross-pollination occurred at low levels between non–herbicide-resistant cultivars at short distances (Caviness 1966), which demonstrates the potential for gene flow between adjacent fields.

Most of the soybeans in the United States are grown for oil extraction and animal feed. Therefore, gene flow could occur during seed transport. Because of the widespread adoption of glyphosate-resistant cultivars, commingling has not generally been a concern. This could change with the introduction of the glufosinate-resistant cultivars. Volunteer soybean plants are common in subsequent crops and are generally well-controlled with herbicides and tillage, except in cotton, where conservation-tillage practices are common and glyphosate may be the only POST herbicide used. Weed management is a greater challenge when herbicide-resistant soybeans are grown in rotation with other crops resistant to the same herbicide; i.e., glyphosate-resistant corn grown in rotation with glyphosate-resistant soybean. Volunteer glyphosate-resistant soybean plants also have been reported in glyphosate-resistant cotton (York et al. 2005), although commingling of soybean and cotton seeds during harvest has not been reported, most likely because of the differences in growth and production practices of the two crop species.

The movement of herbicide resistance genes via either pollen or seeds is of greatest concern for soybean growers that produce non–herbicide-resistant soybeans and where volunteer glyphosate-resistant soybean plants require different management strategies, especially in minimum- or no-tillage environments that depend heavily on the use of glyphosate.

Sugarbeet

Sugarbeet is an outcrossing, wind-pollinated species. Although sugarbeet is a biennial, it is grown as a winter annual crop for seed or as a summer annual for root production. When sugarbeets are grown for their roots, they are harvested before the plants bolt. Occasionally, there will be some plants that set seed before harvest. On the occasion that fields cannot be harvested before winter, they are left in the field. Plants that remain in fields may flower the next cropping season if not killed by winter conditions; however, gene flow via pollen or seed in root production fields generally is not an issue.

All major sugarbeet seed companies have addressed the issue of pollen flow from glyphosate-resistant transgenic sugarbeet to compatible crops by voluntarily increasing the isolation distances from compatible species (G. Burt and J. R. Standard, personal communication). Still, it is possible for a seed-producing field to be planted closer than the recommended isolation distances. At least one seed company has placed the herbicide-resistant trait on the female seed plants, nearly eliminating herbicide-resistant pollen movement in these types of fields. To prevent seed commingling, seed producers are not allowed to grow conventional and herbicide-resistant sugarbeet seed on their farm in the same year and herbicide-resistant seed is cleaned and stored separately (G. Burt and J. R. Standard, personal communication). Seed shattering can occur during harvest, and volunteer sugarbeet plants require control in subsequent crops. Sugarbeet can produce a persistent seedbank if seeds are buried during tillage.

Seed production fields are established by using either seeds or transplants. Viable herbicide-resistant sugarbeet roots that were left after planting a seed field were found in compost sold for use in flower and vegetable gardens (C. A. Mallory-Smith, personal observation). The roots were not disposed of in a manner consistent with the company's prescribed protocol. The use of transplants provides another potential mechanism for gene flow.

Sugarbeet does not produce feral populations in the United States, but it has two compatible relatives, beet [Beta vulgaris L. ssp. macrocarpa (Guss.) Thell.] and wild beet [Beta vulgaris ssp. maritima (L.) Arcang.], in California (USDA plant database). Hybridization and introgression of sugarbeet alleles have been reported in an accession of the B. vulgaris L. ssp. macrocarpa, which is a widespread weed in and near sugarbeet fields in the Imperial Valley, CA (Bartsch and Ellstrand 1999). The authors also suggested that sugarbeet can hybridize with B. vulgaris spp. maritima. The conspecific crops red beet (Beta vulgaris ssp. vulgaris var. conditiva Alef.) and Swiss chard [Beta vulgaris L. ssp. cicla (L.) W.D.J. Koch) freely hybridize with sugarbeet. Therefore, there is potential for a herbicide resistance gene to persist in the environment outside of production fields.

Wheat

Wheat is a predominately self-pollinated species. Wheat does not produce a persistent seedbank but does produce volunteers in following crops, and volunteer wheat is a major weed in no-till or minimum-tillage systems. A herbicide-resistant wheat could have a negative effect on volunteer management, especially if it were glyphosate resistant because glyphosate is the foundation for weed control in no-tillage or minimum-tillage wheat production systems.

Wheat hybridizes with jointed goatgrass, a weed found throughout wheat-growing regions in the Pacific Northwest and the Great Plains of the United States. The F₁ hybrids produce very few seeds because they are male-sterile with low female fertility (Zemetra 1998). However, each successive backcross to either parent increases fertility. Selective control of jointed goatgrass with a herbicide was not possible until the introduction of Clearfield (imazamox)-resistant cultivars. In 2008, hybrids carrying the herbicide-resistance gene were found in commercial wheat fields in Oregon (Perez-Jones et al. 2010). In subsequent field surveys in 2009 and 2010, resistant hybrids and putative backcross progenies were widespread across the wheat-growing region in Eastern Oregon (C. Mallory-Smith, unpublished data). The hybrids are not being controlled in the wheat fields with the application of imazamox so remain in the field with the potential for backcrossing to wheat or jointed goatgrass. The number of resistant plants has increased along with a decrease in the benefit derived from planting the herbicide-resistant wheat.