Weed Population Dynamic Framework

Weed population dynamics under ST systems are more complicated than they are under either CT or NT systems because of the distinct characteristics of untilled BR zone and tilled IR zone, the potential for movement of propagules between these zones, and the potential edaphic and biotic interactions between zones (like movement of soil moisture) (Figure 2). Predictions of species shifts and identification of optimal management strategies under ST may be facilitated by an understanding of the underlying biological processes that affect weeds in these distinct zones. Weeds and their propagules are likely to experience different rates of germination, emergence, survival, fecundity, predation, and decay in BR (Figures 2A–D) compared with IR zones (Figures 2E–H).

To some degree, literature on weed population dynamics under NT and under CT can be applied to explain differences in weed population dynamics in untilled BR zone and the tilled IR zone, respectively. However, weed population dynamics under ST is more than just the sum of its CT and NT components for several important reasons. First, weed propagules under ST can move between adjacent zones, resulting in different propagule densities than would be expected under homogeneous tillage systems. Second, both biotic and abiotic factors influencing weed population dynamics may be strongly influenced by adjacent zones. For example, soil moisture retention in the BR zone is often higher in ST than it is in CT, in part, because of retention of surface mulches (Haramoto and Brainard 2011; Hendrix et al. 2004; Hoyt et al. 1994); because moisture can move freely from BR to IR zones, moisture availability in the IR zone of ST is also sometimes greater than it is under CT (Hendrix et al. 2004). Similarly, biotic factors, including predators of weed seeds, may move between adjacent zones, thereby contributing to greater rates of predation in the IR zone under ST than there are under CT.

Another unique feature of weed population dynamics under ST is the influence of the location of strips from year to year. In any given location in the field, tillage may occur every year, alternate years, every 3 or more years, or never, depending on the relative location and width of disturbance. Variation in the frequency of soil disturbance in a given location may influence weed population dynamics via changes in the vertical distribution of propagules, as well as tillage-mediated changes in soil characteristics that indirectly influence weeds at multiple stages in their life cycles.

Effects of ST on Winter Annual, Biennial, and Perennial Weeds

Because tillage severs, uproots, and buries seedlings, survival and reproduction of weeds that have emerged before tillage will generally be higher in the BR zone of ST than they are in CT systems. Winter annual, biennial, and perennial weeds are often well established at the time of ST in the fall or spring, and in the absence of additional management practices (e.g., herbicides or winter cover cropping) to suppress these species, they can survive and reproduce in the undisturbed BR zone of an ST system. In Michigan, following a 3-yr sweet corn–snap bean–winter squash rotation, the seedbank density of winter annual weed seeds, including henbit (Lamium amplexicaule L.) and common chickweed [Stellaria media (L.) Vill.] was higher under ST than CT systems, presumably because of higher survival and reproduction in the untilled BR zone (Brainard et al. 2012a). Similarly, in a long-term study comparing seedbank densities in continuous corn, winter annuals, biennials, and simple perennials made up 53 to 62% of the seedbank following NT but were virtually absent under CT systems (Cardina et al. 1991).

The overall success of particular winter annual, biennial, and perennial weed species in ST may differ somewhat from that observed in long-term studies in NT depending on the species ability to recolonize tilled zones and the location of strips from 1 yr to the next. Wandering perennial weeds, such as horsenettle (Solanum carolinense L.), are likely to be more problematic in ST systems than are stationary perennials, such as dandelion (Taraxacum officinale G.H. Weber ex Wiggers), because of their ability to rapidly recolonize the disturbed IR zone following tillage via rhizomes. If the location of tilled zones varies systematically from 1 yr to the next, the entire field will receive tillage during the course of 1 to 3 yr, and the capacity of stationary perennials to establish will be more limited than they are in NT systems. After 4 yr in a vegetable cropping-system experiment, comparing CT to ST systems with alternate strip locations, the density of dandelion did not differ between tillage systems, but the ST system had a greater density of horsenettle (D. C. Brainard, unpublished data).

Short-Term Effects of ST on Summer Annuals

Effects of ST on Vertical Distribution of Weed Propagules

One of the most important factors determining the long-term effects of tillage on weed population dynamics is the vertical distribution of seeds in the soil profile. The depth of a seed in the soil has a major influence on its dormancy status, its susceptibility to predators and decay agents, and its capacity to reach the soil surface following germination (Cousens and Mortimer 1995; Teasdale and Mohler 2000).

Under ST treatment, the vertical distribution of propagules within the IR and BR zones is strongly influenced by both the vertical and horizontal movement of propagules (Figure 3). Studies examining the vertical movement of seeds resulting from different tillage implements (e.g., Mohler et al. 2006) suggest that seeds dispersed to untilled strips are more likely to remain on or near the soil surface, whereas those in tilled strips will be buried to greater depths because of vertical soil movement (Figures 3C and 3D). Therefore, growers transitioning from CT to ST systems are likely to observe a shift toward a seed distribution closer to that observed under NT, where a larger fraction of seeds tend to accumulate near the soil surface over time (Cardina et al. 1991; Roberts and Stokes 1965).

Vertical propagule distribution in ST systems varies by species, based in part, on (1) the zone (IR or BR) that favors their survival and reproduction, (2) the extent of dispersal of their propagules between zones (Figure 3A) before tillage, and (3) the location of future tillage events. For example, winter annuals in the BR zone may escape tillage in the spring and survive to shed seed on the soil surface. Based on the limited data available on seed dispersal distance, Cousins and Mortimer (1995) report that among weeds with “unaided” dispersal, most seeds are deposited within a distance approximately equivalent to the plant height. Many summer annuals that escape control measures may reach heights of 1 m or more; these plants are likely to disperse seeds into multiple zones. However, for smaller-statured weeds—including many winter annuals—dispersal of seeds is likely to cluster them in the zone where seed production occurred. Therefore, winter annuals producing seeds in the BR zone may disperse most of their seeds in the BR zone; when an ST treatment is targeted to the same zone every year, the seeds of those weed species are more likely to be concentrated near the soil surface then are those of tall summer annuals.

Effects on Weed Seed Predation

Tillage affects seed predation potential because (1) weed seeds may be buried beyond the reach of seed predators, (2) seed predators may be killed during tillage, or (3) critical habitat of seed predators may be destroyed (thus reducing survivorship).

Although larval stages of carabids (ground and tiger beetles; Coleoptera: Carabidae) have been shown to consume buried weed seeds (Hartke et al. 1998), most seed predation occurs at or near the soil surface (Saska 2004). When weed seeds remain near the soil surface, seed predation may be a key source of mortality in some cropping systems (Westerman et al. 2003). Therefore, the shallower distribution of seeds in ST, compared with CT, systems may favor greater rates of seed predation. However, even in the absence of tillage, vertical seed drift continues slowly as rainfall, earthworms (Lumbricus terrestris), ants (Hymenoptera: Formicidae), and carabid beetles collect and bury seeds or, sometimes, return buried seeds to the soil surface (Figure 3D) (Seguer and Westerman 2003; Smith et al. 2005).

ST may also affect rates of predation through direct effects on predators or indirect effects on their habitats. Key weed-seed predators in agricultural systems include invertebrates, birds (Tetrapod: Aves), and mammals (Tetrapod: Mammalia), although the significance of these organisms varies from system to system and site to site (Crawley 1992; Janzen 1971; Marino 1997, 2005). Modifications to the tillage system are more likely to affect weed seed predation by small mammals and invertebrates than by birds (Navntoft et al. 2009). CT is often cited as detrimental to seed predators and rates of predation (Brust and House 1988), but the results are far from consistent (Cardina et al. 1996). In some cases, the activity density of predatory carabid beetles has been shown to be greater in CT than in NT systems (Westerman 2003). Additionally, some studies have shown that generalist predators, such as the common black ground beetle (Pterostichus melanarius) are well adapted to disturbance, and there were no differences in activity–density when comparing NT and CT systems (Shearin et al. 2007). Tillage practices within a crop have also been shown to be less significant than is the crop itself in determining seed predator activity–density and seed loss (O'Rourke et al. 2006).

The negative effects of tillage on seed predator populations may be moderated by reducing or eliminating tillage (Brust and House 1988; Cardina et al. 1996; Stinner and House 1990) or by providing refuge habitat, such as herbaceous filter strips (Carmona and Landis 1999; Menalled et al. 2001), beetle banks, and hedgerows, where tillage is eliminated altogether (Collins et al. 2003; MacLeod et al. 2004). ST typically spares two-thirds or more of the field from disturbance, with the presumption that seed predation will be enhanced in these systems as well. These untilled BR zones emulate refuge strips on a smaller spatial and temporal scale and may provide essential habitat and refuge for carabid beetles during soil disturbance events.

To date, there is little data showing the effects of ST on seed predation potential. As with seedling emergence, attempts to predict outcomes are confounded by the interactions between the many system components, including surface residues, which are often inherent in ST systems. Cover crop residues may ameliorate mortality caused by tillage and cause higher-than-expected seed predation rates even under intensive tillage (Cromar et al. 1999). Trophic interactions, including those between significant seed predators, compound the difficulty of predicting outcomes. For instance, some rodents are known to be active predators of carabid beetles and other granivorous ground beetles, and any practice that reduces tillage is likely to cause an increase in rodent populations as well. One recent study contrasted seed predation potential of ST and CT plots at 2 sites for 4 yr. In contrast to expectations, carabid beetle activity density (primarily, P. melanarius) did not increase in ST plots compared with CT plots (Green 2010). Estimates of weed seed predation (using exclosures that allowed ground beetle access to weed seeds) were more than 50% greater in CT, compared with ST, systems, but only when broadcast insecticides were applied to these treatments.

Long-Term Effects on Weed Communities

Greater concentration of seeds at the soil surface under ST, compared with CT, treatments may result in weed community shifts away from species with seed characteristics that promote survival and emergence of buried seeds (e.g., dormancy, capacity to emerge from depth) and toward those that promote survival and emergence at the soil surface (e.g., tolerance to desiccation, resistance to predation). For example, greater concentration of seeds near the soil surface may favor weed species with (1) small seeds because they are often less susceptible to predation and more sensitive to burial depth for successful emergence, (2) less-dormant seeds (e.g., many grasses) because dormancy is less important for avoiding fatal germination when seeds are not buried (Cousens and Mortimer 1995), and (3) wind-dispersed seeds (R. E. Peachey, personal observation). Long-term tillage studies in agronomic cropping systems provide evidence for shifts toward small-seeded weeds (Buhler and Daniel 1988) and grass species (e.g., Buhler 1995; Cardina et al. 1991; Pollard and Cussans 1981) when tillage is reduced. After 3 yr of a sweet corn–snap bean–winter squash rotation, the ST system has resulted in a shift toward grass species, especially large crabgrass [Digitaria sanguinalis (L.) Scop.], when compared with a CT system (D. C. Brainard, unpublished data). However, interpretation of such studies is often complicated by potential interactions of tillage with other management practices, including herbicides (Cardina et al. 1991).

Evidence from various studies suggests that weed species diversity may increase under ST, compared with CT, treatments. For example, following 25 yr of continuous corn, Cardina et al. (1991) reported greater diversity in NT systems than found with CT systems, in part, because of the greater prevalence of winter annual, biennial, and stationary, perennial species. Although the long-term effects of ST management on species diversity have not been evaluated, an ST system may support a greater diversity of weed species then either a CT or NT system does because the two zones of ST treatments represent distinct niches, each supporting species best adapted to that zone. The capacity for movement between zones will be an important determinant of species' fitness under ST management. For example, annuals with limited seed-dispersal distances and perennials with noncreeping perennation are less likely to recolonize tilled zones following ST treatment. Conversely, species with wind-dispersed seeds or wandering perennials can rapidly recolonize. However, these long-term shifts in community diversity may not be realized or may take longer to develop if strip location varies from year to year because disturbance will still be a factor, albeit a less frequent one.

Preventing Reproduction

Perhaps the most important element of successful, long-term weed management is prevention of weed propagule (seed and vegetative reproductive tissue) production. Because many weed propagules are highly persistent in soils, failure to prevent propagule production can greatly increase future weed populations and the long-term costs of weed management (Buhler et al. 1997). Prevention is often less expensive then treatment, thus “zero seed rain” strategies are sometimes advocated (e.g., Gallandt 2006). Although zero seed rain may not be a realistic or optimal approach in all situations, minimizing seed rain is an important goal, and the extra initial costs associated with reducing weed propagule production may often be justified by lower weed management costs in future years. In ST production systems, the importance of preventing weed reproduction is heightened because weed management is often more expensive when tillage is not an option. This has been observed in a northern organic vegetable ST system, which had to be plowed after 4 yr because of extensive weed pressure (A. Rangarajan, personal observation).

Cover Crops for Weed Suppression in ST

Under ST treatments, cover crops can be used to enhance weed suppression through effects on virtually all phases of weed life cycles, including germination, emergence, growth and fecundity, and seed predation and decay. Surface cover crop and crop residues present in the BR zone in an ST system can help reduce weed emergence either by reducing seed germination or by increasing postgermination seedling mortality below the residue surface. Several mechanisms may contribute to reduced emergence where surface cover crop residues are present, including reduction in light penetration to the soil (Teasdale and Mohler 1993), physical obstruction resulting in seed-reserve depletion before emergence (Teasdale and Mohler 2000), and increased seed predation or decay (Cromar et al. 1999).

Cover crops in MT systems may be helpful for suppressing weeds, but care must be taken to avoid crop interference (Teasdale 1998). Indeed, one of the primary rationales for ST systems is to avoid problems with stand establishment and growth, where heavy crop or cover crop residues are present (Luna et al. 2012). Even in ST systems in which cover crops are removed from the IR zone via row-cleaners, interference can occur through a variety of mechanisms, including reductions in soil temperatures, immobilization of soil N, or allelopathic effects (Hoyt et al. 1994). For example, in ST cabbage production, bare soil treatments out-yielded treatments with wheat (Triticum aestivum L.) or cereal rye (Secale cereale L.) cover crops (Hoyt 1999; Mochizuki et al. 2007; Rangarajan 2008). In snap beans, yields under NT were reduced by 63% compared with CT, and when rye was combined with ST management, yields improved relative to the NT system but were still reduced by 20% compared with CT treatment (Bottenberg et al. 1999).

Under ST management, cover crops offer unique challenges and opportunities that are distinct from those encountered in either CT or NT systems. In particular, the optimal choice of cover crop species and management is likely to vary between the two distinct zones under ST treatments. For example, to maximize weed suppression and moisture retention in the BR zone, a thick mulch from a high carbon cover crop, such as winter rye may be desirable. However, that same mulch may be detrimental to crop establishment and growth in the IR zone. Recent research in ST pepper production found that moving rye mulch residue from the BR zone into the IR zone after crop establishment reduced IR weeds and improved the effectiveness of BR weed management with a high-residue cultivator (Rostampour 2010). Such physical movement of crop residue after crop establishment is one promising approach to overcoming barriers to cover crop use in ST vegetable systems.

To tailor cover crop services to the unique requirements of BR and IR zones, different cover crop species can be planted in these zones. For example, winter rye can be planted in the BR zone to maximize the amount of weed-suppressive residue produced, whereas cover crops such as winter-killed oats and peas (Pisum sativum L.), or hairy vetch (Vicia villosa Roth) can be planted in the IR zone to improve soil tilth or provide N where the crop will be planted. Innovative growers have begun adopting segregated strips of cover crops in ST systems in parts of the Midwest (Gruver 2011); however, adoption in vegetable cropping systems has been limited. In a recent study, yield of ST broccoli grown with an oat–pea mixture planted IR and a rye–hairy vetch mixture planted BR had similar yields to bare-soil controls and higher yields than produced in a rye–hairy vetch mixture planted both IR and BR (A. Rangarajan, personal communication). Similarly, in ST sweet corn, targeted planting of hairy vetch IR and rye BR can provide N to crops while minimizing N availability to weeds in the BR zone; this approach may also help minimize the potential problem of hairy vetch regrowth often seen in MT systems in which herbicides are not used (C. J. Lowry and D. C. Brainard, unpublished data). Mustard species (e.g., radish, Raphanus sativus L.) planted in the IR zone can be incorporated with an ST implement to maximize the suppressive effect of their glucosinolate by-products (Gruver 2011). With the advent of GPS tractor-guidance systems, such targeted cover-crop placement is a more realistic approach now than it has been in the past.

In addition to dead cover-crop mulches, ST systems can accommodate preestablished living mulches that may provide multiple ecosystem services including weed suppression. For example, in ST carrot production, growers in western Michigan grow wheat or barley (Hordeum vulgare L.) cover crops before carrot planting and retain those covers between crop rows as a windbreak until carrots are well established (Brainard and Noyes 2012). This system has no adverse effect on crop yields, and in some cases, contributes to weed suppression relative to CT management. Living mulches have also been tested in production of cabbage and other vegetable crops in combination with MT or ST treatment, but these systems have not gained widespread acceptance by growers, in part, because they often compete too aggressively with the crop (Andow et al. 1986; Grubinger and Minotti 1990; Teasdale 1998).

Although an important tool for weed suppression, cover crops alone are unlikely to eliminate the need for other weed management approaches in northern vegetable cropping systems. In northern climates, the shorter growing season makes it more challenging to produce sufficient cover crop biomass (> 8,000 kg ha⁻¹) to achieve full-season weed suppression (Mohler and Teasdale 1993). Poorly managed cover crops can also interfere with crop establishment, promote insects and diseases, and entail costs for seeds, establishment, and maintenance—all factors that reduce their practicality in some vegetable production systems (Luna et al. 2012).

Tillage and Crop Rotations

Rotation of crops, herbicides, and tillage intensity is important for managing many pests. Purposefully rotating tillage intensity is perhaps not common practice, but tillage can be strategically positioned within a given crop sequence to restrain pest populations. Rotational tillage sequences (variation of primary tillage practices within a crop rotation) are often implicit in crop rotations, but typically, the emphasis is on nutrient release and alleviating compaction, although they can also be specifically designed to minimize weed fecundity or weed seed survival in the soil. Some have proposed the use of rotational tillage systems to improve short-term nitrogen use efficiency and crop growth in predominately NT systems (Doran 1987), although this may reverse gains in soil quality (Grandy and Robertson 2006).

The effect of alternating ST and CT treatments on weed management has not been determined. Differences in weed seed drift and vertical distribution of weed seeds within the soil of the ST environment (Figure 3) make it difficult to predict the outcome of any tillage rotation sequence. An experiment over 6 yr in a three-crop rotation indicated that summer-annual weed density is reduced when direct seeding (using a slot planter that causes very little in-row disturbance) is alternated with full-width tillage. However, the outcome was dependent on the crop that was paired with the specific tillage system used and the sequence of the tillage rotation (Peachey et al. 2006).

The relative location of strips from 1 yr to the next may have a major impact on weed communities by altering both vertical seed distribution and the mortality of perennial weeds (Fig. 4). The most common practice among growers in continuous ST systems is to offset the tillage strip by one-half row each year so that crops are grown between rows in alternating years. Assuming that seeds are shed and dispersed uniformly on the soil surface (see discussion above), this would result in greater burial of seeds than would occur if strips were in the same location (Figure 4A vs. Figure 4B). Conversely, when strips are targeted to the same location, a greater proportion of seeds remain on the soil surface, resulting in shifts in weed communities in the BR zone similar to those observed under NT management.

Locating strips in the same location from year to year would likely provide soil health benefits relative to alternating-zone systems by restricting tractor traffic to noncropped zones and by facilitating increases in soil organic matter (Figure 4). Grandy and Robertson (2006) demonstrated that a single tillage event following years of NT production can substantially and permanently reduce desirable soil characteristics and suggested that NT soils need to be continuously maintained to protect aggregation and physically stabilized C pools. Improvements in soil physical, biological, and chemical characteristics in the BR zone of systems where tillage is restricted to the same location may favor growth and development of weeds in those zones. The extent to which such changes will alter weed–crop competition is difficult to predict but may be expected to favor weeds because crops are restricted to the tilled IR zone where soil health improvements may be more limited. As previously noted, restriction of tillage to the same location is also likely to favor buildup of perennial species, particularly in the BR zone (Figure 4).

Another option within continuous ST systems is to shift the location of strips from year to year so that the entire field surface undergoes primary tillage over the course of 3 yr. The effect of this system on weed communities is difficult to predict and has not been studied, to our knowledge, but may suppress the expected buildup of noncreeping perennial weeds.

In the Pacific Northwest, crop rotation requirements have limited ST systems to annual tillage rotations in both vegetables (Luna et al. 2012) and sugarbeets (Strausbaugh and Eujayl 2012). Continuous ST is not possible in these systems because these crops are often rotated with cereal crops with narrow row spacings that cannot accommodate an ST treatment. After vegetable crop harvest following a single season of ST treatment, fields are commonly left fallow over the winter or tilled with MT practices and planted with cover crops or cereal crops such as wheat. Although a few growers have experimented with NT seeding of cover crops following vegetable crop harvest, success has been limited (J. M. Luna, personal communication).

Resource Placement and Timing

Concentration of fertilizers in the crop-rooting zone can significantly reduce weed competition while maintaining or enhancing crop yields. For example, in CT systems, banding fertilizers in the crop row has been shown to improve yields while reducing weed density and biomass relative to broadcast fertilization applications (DiTomaso 1995). Banding of fertilizer at depths of 5 cm or more may offer weed management benefits, especially for large-seeded crops or transplants that can more rapidly access deep placed fertilizer than many smaller-seeded weed competitors (Liebman and Davis 2000; Rasmusen and Peterson 1996). This approach is most successful where background surface fertility levels are low, so that weeds have insufficient initial fertility to rapidly compete with crops (Cochran et al. 1990; Liebman and Davis 2000).

Under ST management, fertilizers can be conveniently placed at desired depths directly below the crop row via adjustment of fertilizer applicator tubes installed behind the ST shank. This approach has been adopted by ST users in both field and sweet corn production (Rangarajan 2008) and holds promise for reducing weed interference. Maintenance of high-carbon crop residues on the soil surface under ST systems may enhance the benefits of deep-fertilizer placement by supporting microbial populations in the surface soils that serve to reduce soluble soil N and thereby fertility available to weeds.

In organic production systems, targeted placement of organic amendments, including cover crops, composts, or mulches, may also promote crop growth and suppress weeds. For example, because high rates of compost can suppress germination and emergence of some weeds without adversely affecting crops (Lowry and Brainard 2012; Menalled et al. 2005), surface placement of compost in the IR zone of ST treatments may suppress weed emergence while maintaining or improving crop yields (Lowry and Brainard 2012). However, weed response to compost is variable, and some combinations of weed species and compost rates may increase weed emergence. For example, Brainard and Noyes (2012) observed a threefold increase in Powell amaranth emergence where compost was used in ST carrot production. In such cases, incorporation of compost in the IR zone of ST would be preferable from a weed management perspective. Clearly, a better understanding of the mechanisms by which compost and other soil amendments influence emergence and growth of specific crops and weeds would be helpful for developing strategies for selectively suppressing weeds in ST systems.

Mechanical Cultivation

The rapid development and commercialization of GPS and RTK tractor-steering systems in the 1990s has greatly expanded the opportunities for precision weed-control technology, although adoption rates among vegetable growers are not known. These technologies make mechanical cultivation and other physical means of weed control easier and more effective through straighter crop rows. They are particularly well suited to ST systems where conditions in the BR zone and IR zone may require different tools for optimal management.

In traditional CT, the BR zone has been managed relatively easily using sweeps, rolling cultivators, or other equipment. Cultivator shields can be used to protect the crop row from moving soil, and the previously tilled soil easily accommodates the cultivation equipment. In ST systems, however, the BR zone frequently contains living cover crops, weeds, or residue from previous cropping and is often not cultivated until many weeks after the crop is planted, if at all. Cover crops residues that suppress or delay weed emergence allow cultivation to be delayed so that crops are less at risk and may reduce the number of cultivations needed to control weeds and cover crop regrowth. However, crop residues are often an impediment to near-row cultivation, particularly to stationary sweeps that collect and drag residue. The BR vegetation may be managed by mowing (sometimes several times) or by using an undercutter (Creamer and Dabney 2002) or a high-residue cultivator to kill the standing vegetation (Luna et al. 2012). The Buffalo Cultivator (Fleischer Manufacturing Inc., Columbus, NE) employs heavy-duty sweeps with hardened points for cultivation in untilled soil. These high-residue sweep cultivators can undercut cover crops and weeds growing in the BR zone; however, additional passes with power takeoff–driven rotovators or other cultivation equipment are sometimes needed for adequate weed control. This intensive cultivation may conflict with the goals of MT system if farmers adopt MT treatments to help maintain or improve BR soil quality.

Managing weeds in the IR zone of ST systems involves many of the same strategies and tactics used in CT systems because the strip for the crop row is tilled before planting. These methods typically kill weeds by uprooting, severing, or burying (Kurstjens and Kropff 2001) and are usually focused on the weed seedling stage when they are susceptible to mechanical damage. Unlike full-width CT systems, however, ST systems require mechanical weed cultivation to be performed in narrow, tilled strips, frequently 25 to 30 cm wide. As in CT systems, IR cultivation methods in ST systems usually depend on a size differential between the crop and weeds that allows the weed seedlings to be removed selectively by the mechanical device. Transplanted vegetable crops clearly offer a large size differential between the crop and the newly emerged weeds, providing some options that are not available with direct-seeded crops (Ascard and Fogelberg 2008). Compared to direct seeded crops, transplants also have a competitive advantage over weeds that may have escaped control from cultivation (Melander et al. 2005).

An array of tools are used in IR weeding, alone or in combination with others, including the rotary hoe (Gunsolus, 1990; Leblanc et al. 2006; Mohler et al. 1997), spring tine harrow (Peruzzi at al 2007), torsion weeder (Duerinckx et al. 2005), finger weeder (Barberi et al. 2000; Bowman 1997), brush weeder (Melander, 1997), the spiked-disk weeder (Raffaeli et al. 2010), and the pneumatic weed blower (van der Weide et al. 2008) Each tool offers distinct advantages and disadvantages for varying soil conditions, crops and stages of growth of crops and weeds.

Flame Weeding, Steam Weeding, and Sand Blasting

Several alternatives to cultivation may be helpful for integrated weed management in ST systems. Flame weeding is an alternative to cultivation that has been widely adopted on organic farms. However, because of the large amount of residue often encountered in ST systems, few growers (if any) have attempted to use it. The prospect of igniting residues is a strong deterrent to use. Equipment innovators have designed a variety of shielded flamers that protect crops. The two-row flamer (Steam Weeding Ltd, Lakeside Rd 3, Leeston, Canterbury, New Zealand) (Figure 5) was specifically built for use in high-residue ST systems (C. Merfield, personal communication). Flame is confined to the crop row with row covers, and microjets deliver water/organic herbicide in a curtain in front of the flamer to minimize the potential of the residue to ignite (R. E. Peachey, unpublished data). The effect of the curtain of spray on weed control efficacy requires further testing.

Steam weeding is an alternative to flame weeding and eliminates the possibility of igniting residues. These systems are highly efficacious but require large amounts of water (up to 500 L h⁻¹). Steam can also be applied to pasteurize the soil and kill weeds seeds and disease propagules, which was done in high-value crops for more than a century. Recent innovations are being tested to determine whether steam can be applied in narrow bands, just wide enough for the seed row (Melander et al. 2002; Melander and Kristensen 2011). This alternative is well suited for ST systems. Again, a large amount of water is needed, and the process is slow, but water use is much less than if applied to the entire field, and the system minimizes the risk of residue ignition.

An alternative tactic that can potentially help manage weeds under high-residue conditions involves propelling abrasive particles through a sand blaster to shred and kill weed seedlings (Forcella 2009; Nørremark et al. 2006). Forcella (2009) found that abrasive grit made from corn cobs and expelled at 517 kPa pressure killed small seedlings of common lambsquarters without adversely affecting corn. These results suggest that air-propelled grit may be an effective POST weed-management tactic for targeted control of weeds at either IR or BR zones in ST systems.