Manipulating Weed Seed Banks to Promote their Decline

Introduction

The weed seed bank present at the beginning of a cropping cycle represents the potential for weeds to reduce crop yields, but it does not predict that weeds will do so. Organic growers can employ several strategies to reduce the potential of existing weed seed populations to inflict economic damage. Weed seed banks can be manipulated by:

• Tricking weed seeds into germinating when they can be easily killed
• Conditioning weed seeds or modifying their immediate environment so that they become less likely to germinate during critical phases of crop establishment and production
• Concentrating weed seeds at a position within the soil profile from which they cannot easily emerge, or at which they are most subject to attrition
• Moving germinable weed seeds away from the crop in space (e.g., ridge tillage) or in time (planting date before or after the weeds' peak emergence season)

Choosing the most effective strategies for a particular field requires sufficient knowledge of the field’s weed seed bank, including the major weed species represented and perhaps age and vertical distribution of seeds in the soil profile, as well as a rough estimate of total population density. Weed species differ widely in seed dormancy and longevity, season in which they emerge, depth from which they can emerge, and seed responsiveness to light and other stimuli (Table 1). These characteristics can help the farmer select the best strategies for managing a particular weed seed bank.

Tricking the Weed Seeds: Stale Seedbed and False Seedbed

Many weed seeds, especially smaller-seeded annual broadleaf species, germinate in response to stimuli that indicate that the soil surface has been disturbed and cleared of competing vegetation. Light is the most common germination trigger, though many seeds also respond to temperature and moisture fluctuations, increased aeration, and increased release of nitrate and other soluble nutrients that occur in tilled soil (Table 1). Many weed species also tend to germinate during specific times of year after a certain amount of soil warming. For these weeds, the time of emergence during spring can be approximately predicted by the accumulated growing degree–days (GDD), a summation of the number of degrees that daily average temperatures exceed a base temperature (Table 1).

Table 1. Weed seed germination and emergence characteristics of several weeds of row crops in the north-central United States¹.

Weed Species	Season of Emergence2	Emergence Period (wk)	Emergence Depth3	Half-life4 (yr)	Germination Stimuli
Horseweed	Fall/early spring				L
Shepherds-purse	Fall/early spring	S		2.8	C, L, N, T±
Field pennycress	Fall/early spring		M	5.7	L, T±, A
Giant ragweed	GDD <150	2-3		0.3
Common lambsquarters	GDD <150	3-7	S	7.6	L, T±, N
PA smartweed	GDD <150	3-7		4.0
Annual sunflower	GDD <150	3-7		0.3
Redroot pigweed	GDD 150-300	8-10	S	2.2	L, H, T±, N
Common ragweed	GDD 150-300	3-7	M	1.4	C, L, T±
Velvetleaf	GDD 150-300	8-10	D	2.3	A
Giant foxtail	GDD 150-300	8-10		0.8
Yellow foxtail	GDD 250-400	3-7	D	4.5	C, N
Black nightshade	GDD 250-400	3-7	M-D	L, T±, N
Common cocklebur	GDD 250-400	3-7	D	5.6	C, T±, (D)
Wild proso millet	GDD 250-400	3-7
Large crabgrass	GDD >350	3-7	M	1.2	C, (D)
Fall panicum	GDD >350	3-7			T±
Waterhemp	GDD >350	8-10		2.4
Morningglory	GDD>350	8-10

¹ The information in this table is based on Tables 1 and 3 in Davis (2004), and an extensive literature review of weed seed ecology research by Charles A. Mohler.
² GDD = growing degree-days F (base temperature 48°F). GDD<150 – emerge several weeks before corn planting in the North Central region; GDD 150-300 – emerge shortly before or during corn planting; GDD 250-400 – emerge near the end of corn planting; GDD>350 – emerge after corn emergence.
³ S = shallow, most seeds emerge from surface or top 0.5 in of soil profile; M = medium, most seeds emerge from top inch; D = deep, most seeds emerge from top 2 inches, and a few can emerge from greater depth.
⁴ About 6 to 7 half lives required to eliminate 99% of seed from the weed seedbank.
⁵ A = aeration; C = chilling period; H = high soil temperature; L = light; (D) = not responsive to light; N = nitrate; T± = fluctuating soil temperatures.

 

When several of these conditions occur together—for example, a tillage operation at the weed’s normal time of emergence that creates a light flash and promotes nitrogen mineralization—a high proportion of weed seeds may germinate provided that soil moisture and seed–soil contact are adequate. If the farmer prepares a seedbed and sows a vegetable crop at this time, weed problems will occur. However, if the farmer delays planting until several weeks after seedbed preparation, one or more flushes of weeds can be eliminated by shallow cultivation or flame weeding before planting. If this is done during the weeds’ peak season of emergence, much of the population of readily-germinable weed seeds can be depleted. Two forms of this strategy are called stale seedbed and false seedbed.

In the stale seedbed approach, the soil is tilled to prepare for seeding the crop, then planting is delayed for two or three weeks to allow a flush of weeds to emerge. Just before sowing the crop, emerged weeds are killed with no or minimal soil disturbance. Organic farmers can kill weeds by flaming or cultivating as shallowly as practical, though any cultivation will stimulate some additional weeds to germinate with the crop. Conventional farmers normally use herbicides at the end of the stale seedbed period, an option that may become open to organic producers in the future with the development of natural-product postemergence herbicides that are economically viable at the field scale.

In the false seedbed approach, weeds emerging in response to tillage are killed by two or more additional shallow cultivations at weekly intervals. The crop is planted immediately after the final cultivation. Because small weed seeds germinate better when the soil is firmed to enhance seed–soil contact, rolling is recommended after all except the final cultivation.

Ideally, the final cultivation just before crop planting is done as shallowly as practical to avoid stimulating further weed seed germination, and leaves the soil surface loose and open, forming a dry, crumbly layer from which weed seeds are less able to take up moisture and germinate. Note that good soil tilth promoted by high organic matter and biological activity is essential for these tactics to work effectively. Light duty implements like flexible tine weeders cannot effectively penetrate crusty, cloddy or compacted soils, and stale seedbed can fail to yield weed management benefits in these conditions (Caldwell and Mohler, 2001).

Organic farmers successfully use these approaches to reduce weed pressure in subsequent crops, sometimes realizing weed control commensurate with conventional herbicide applications. Possible drawbacks include yield loss due to delay in planting, increased risk of soil erosion and crusting during the cultivated fallow period, and the risk that dry conditions might inhibit the desired weed seed germination. Another limitation of stale seedbeds is that planting or transplanting equipment can disturb the soil sufficiently to stimulate weed emergence in the crop row (Caldwell and Mohler, 2001). Researchers are now working to develop punch planters, which plant crops with minimal soil disturbance (Rasmussen, 2003).

If the weed seed bank includes weeds that generally emerge after the stale or false seedbed period, the practice may not reduce weed pressure, but only change weed species composition. For example these techniques may work well for soybean in the north-central and northeastern states if the main weeds are common lambsquarters, Pennsylvania smartweed, ragweed, and others that normally emerge before soybean planting. However, if the dominant weeds include later-emerging species like redroot pigweed, common cocklebur, and large crabgrass, a false seedbed would either miss these weeds or entail an unacceptable delay in crop planting. Stale and false seedbed may be especially well suited for late plantings of vegetable crops like lettuce, snap bean, or cucumber, for which sufficient time is available to deplete weed seed populations in the germination zone before vegetable planting. In the southern states, carrots, beets, and some other root crops can be planted in the latter half of July, which allows time for several weeks’ cultivated fallow during the peak emergence period of many weeds.

Stale and false seedbed practices require favorable soil temperature and adequate soil moisture as well as sufficient time to obtain maximum weed emergence prior to crop planting. If soils are dry, some growers irrigate the newly-prepared seed bed to encourage weed emergence. In organic systems that utilize floating row covers for season extension or pest control, placing the row cover over the seedbed at the beginning of the fallow period can accelerate weed emergence by increasing soil temperature and moisture, and thereby enhance efficacy of the stale seedbed (Brainard et al., 2007). This practice reduces the need for removal of row cover for weed management following crop planting, and allows more timely planting of crops.

Certain summer annual weeds that have a long period of germination, or little or no seed dormancy, can be tricked into emerging late enough in the season that fall frosts or fall tillage will kill them before they can set seed. For example, if the final cultivation in late snap beans or fall broccoli stimulates emergence of morningglories, pigweeds, or foxtails, the weeds will not much affect crop yield, and the first frost may turn them into harmless organic matter before they flower. The short-lived, nondormant seeds of galinsoga can be triggered to germinate quite close to the fall frost date, for example during seedbed preparation for a fall cover crop. However, since the timing of frost is unpredictable, and many summer annuals produce seeds very rapidly under the short days of fall (as little as 25–30 days after emergence for galinsoga), this strategy entails risks, and may need to be supplemented by subsequent cultivation or manual weeding if frost is late. Prompt tillage or mowing after harvest of warm season crops like snap bean can prevent weed seed set if frost does not do the job.

Depleting the Seed Bank by Stimulating Germination

Theoretically, one should be able to deplete the soil's weed seed bank through repeated cultivation or other tactics that provide seed germination stimuli, combined with stringent year-round weed control that prevents weeds from setting seeds or otherwise reproducing. Timely cultivation combined with other measures to eliminate all weed seed set can draw most weed seed banks down to perhaps 5–10% of their initial population densities within several years; however the remaining seeds can be much more difficult to eliminate through such tactics (Egley, 1986). In some cases, the seeds are hard—they do not imbibe moisture even at high soil moisture content; in others they are in a state of deep dormancy that requires multiple environmental cues to break. Critical factors may include any or most of the following:

• Seasonal changes or daily fluctuations in soil temperature and moisture
• Light—presence or absence, and quality (full spectrum daylight or filtered through green foliage)
• Concentrations of nitrate and nitrite in the soil
• Concentrations of oxygen, carbon dioxide, and ethylene (C₂H₄) in the soil
• Presence or absence of specific germination stimulants or inhibitors released into the soil by plant roots or plant residues

Multiple stimuli can break the dormancy of at least some seeds in the more persistent weed seed bank, and researchers continue to explore means by which these can be effectively delivered in the field to lower weed seed banks further (Egley, 1986). Whereas some of the methods investigated are not appropriate for organic systems (for example, applications of soluble N fertilizers or the synthetic ethylene-generating compound ethephon), other strategies may emerge that utilize organic soil management practices to provide multiple seed germination stimuli.

Conditioning Weed Seeds: Putting Them to Sleep

A different strategy is to  avoid soil disturbance and other weed seed germination stimuli before and during spring planting. Keeping weed seeds dormant until the crop is well established can substantially reduce potential weed competition. This is a major objective in the no-till cover crop management and vegetable planting systems that some organic farmers and researchers are developing, in which winter annual cover crops are grown to maturity (heading in grasses, flowering in legumes and other broadleaf crops) then rolled or mowed to create an in situ mulch.

However, no-till can be challenging and risky for organic growers, and other practices have been proposed for keeping weed seeds dormant during crop establishment. While some farmers utilize cultivated fallow in spring, others do primary tillage in fall and avoid or minimize soil disturbance in spring, utilizing a flame weeder to kill emerging seedlings while letting dormant weed seeds remain dormant (Davis, 2004). Many European growers believe that tilling their soil at night in late fall can condition the weed seed bank to reduce weed emergence in the following spring. Compared to daytime tillage, lightless tillage can reduce emergence of light-sensitive species like common lambsquarters, pigweeds, and barnyard grass by 50% or more, although results have not always been consistent (Buhler, 1997; Scopel et al., 1994). Since lightless tillage is effective primarily against small seeded broadleaf weeds (Buhler, 1997), continued use of night cultivation may select for species shifts that reduce its effectiveness over time (Dyer, 1995).

A dense, shading plant canopy can also deepen the dormancy of some weed seeds. The dim green light under such a canopy can actually be more effective than continuous darkness in inhibiting the germination of light-responsive seeds (Mohler, 2001a). Reduced weed emergence has been observed in the season after a dense-canopy crop such as a cereal grain/clover intercrop, or August-planted forage radish. Martens and Martens (2008) suggested that the light quality under such foliage may have rendered weed seeds more dormant. Dense crop canopies may also reduce subsequent weed emergence by reducing seed production or increasing seed mortality. For example, dense crop or cover crop canopies can provide favorable habitat for seed predators, resulting in reductions in the seed bank and subsequent weed emergence (Davis and Liebman, 2003).

This dormancy strategy works best for annual weeds whose seeds often show conditional, light-mediated dormancy. Some larger-seeded weeds—and especially the vegetative propagules of wandering perennial weeds like yellow nutsedge and Bermuda grass—may be difficult or impossible to condition for delayed emergence.

Since soluble N can stimulate germination of seeds of many weeds including pigweed and lambsquarters (Table 1), manipulation of soil fertility has been extensively explored as a tool for reducing weed density (Dyer, 1995). Practices that avoid large pulses of soluble N early in crop development, such as delayed or split N applications, or use of slow-releasing N sources such as mature compost, can delay weed emergence and reduce weed density in the crop. Conversely, incorporation of leguminous cover crops or applications of materials that release N rapidly, such as chicken manure, can promote weed emergence and growth. For example, ammonium released from decomposing hairy vetch can stimulate germination of smooth pigweed (Teasdale and Pillai, 2006). Mixing such legumes with grasses can reduce the concentration of soluble N released after incorporation and may be beneficial for weed management.

The effects of N fertilization on weed emergence are variable, owing in part to complex interactions between N and other factors such as ethylene concentration in soil, light, and genetic differences in responsiveness both between and within weed species (Dyer, 1995; Brainard et al., 2006). Nonetheless, practices that improve the synchrony of crop N demand and N supply not only reduce the risk of N losses but may aid weed management by reducing weed emergence during crop establishment.

Seeds require adequate seed–soil contact in order to take up the moisture needed to initiate germination. For small seeds such as those of lambsquarters, galinsoga, or Canada thistle, a fine tilth and firmed soil surface optimizes seed soil contact and promotes germination, whereas a coarse, loose seedbed can significantly reduce their germination. This is why higher weed densities sometimes occur within rows of crops seeded by mechanical planters with press wheels (Caldwell and Mohler, 2001), and is one of the mechanisms by which incorporated cover crop residues can reduce weed emergence (Gallandt, 2006). Gallandt et al. (1999) further suggest that moving loose soil over planting rows after mechanical seeding can reduce within-row populations of small-seeded weeds.

Positioning Weed Seeds in the Soil Profile

Is it better to leave recently-shed weed seeds on the soil surface, or to plow them under? The answer: “It depends.” Burying the short-lived, small, nondormant seeds of weeds such as galinsoga and kochia to a depth of three or four inches can virtually eliminate viable seed within a year or so after it was shed (Mohler, pers. observation; Schwinghamer and van Acker, 2008). Longer-lived seeds like pigweeds, lambsquarters, and velvetleaf, however, may remain viable and dormant at this depth for several years, during which additional tillage may bring them back to the surface and trigger rapid germination and growth. Seeds lying on or near the soil surface are more subject to predation, and may dry out and die after beginning to germinate. Seeds on the soil surface, however, are also more likely to germinate successfully and grow than are most buried seeds. Because weed seeds at or near the soil surface are generally more likely to germinate, deteriorate, or become food for seed predators than seeds buried an inch or deeper, delaying tillage after the annual weed seed rain until the following spring generally reduces the number of seeds added to the long-term seed bank (Egley and Williams, 1990).

Based on work in the north-central United States, Davis (2004) recommends a flexible approach to tillage based on existing weed seed species composition. Whereas chisel plowing leaves many seeds fairly near the surface, the moldboard plow sends the majority of seeds to deeper layers. No-till leaves weed seeds at the surface until natural processes incorporate them into the upper layers. As seeds are placed deeper in the soil profile, they become less likely to germinate, but they are also less exposed to attrition through predation and weathering (Table 2). If a weed species with small, relatively short-lived seed, such as giant foxtail and galinsoga, gets out of control and produces a heavy seed rain one season, a moldboard plow can be used to place the seeds deep in the soil profile. It is essential to till shallowly for the following year or two to avoid bringing up the buried seeds before they have lost viability.

Table 2. Effect of weed seed depth placement on dormancy, germination, and mortality¹.

Depth (inches)	Dormancy	Germination	Mortality
0–0.5	Low2	High2	High
0.5–2	Low	High	Medium
2–5	Medium	Low	Low
5–10	High3	Very low3	low

¹ From Davis (2004).
² Dormancy and germination at 0–0.5 inches depth depends on seed size. Small seeded weeds tend to have low dormancy and high germination at the soil surface, whereas large seeded weeds have medium–high dormancy and low germination at the soil surface.
³ Note, however, that shorter-lived seed with no or little dormancy mechanism tend to undergo fatal germination at this depth (C. L. Mohler, personal communication, 2008).

Moving Weed Seeds Away From the Crop

Knowing what weed species are most abundantly represented in the seed bank can allow the farmer to estimate when the most weeds are likely to emerge, and adjust the timing of stale or false seedbed practices, crop planting dates, and the choice of crop to reduce weed pressure. Simply delaying final seedbed preparation and crop planting until after most of the season's weed emergence has occurred is often recommended to organic growers as a means to move weeds away from the crop in time. However, delayed planting can reduce yield potential, especially for long-season summer crops like corn and cotton. Diversified vegetable cropping systems, in which different vegetables are planted from March to September, and rotated with winter and summer annual cover crops, offer greater flexibility and more opportunities to “dodge” the major weeds. For example, if common lambsquarters, giant foxtail, and common ragweed have become major problems in summer vegetables, the grower might try planting a vigorous, smothering cover crop in late spring, followed by a fall vegetable. The cover crop will restrict the growth and reproductive capability of the weeds, and fewer additional weeds will emerge in the vegetable.

Usually, weed seed banks are sufficiently diverse that some weeds will emerge in crops planted almost any time. However, the varied planting dates of a diversified vegetable rotation can help prevent any one weed species from building its seed bank up to unmanageably high levels. For example, including both fall planted crops (like garlic, fall greens, and cereal grains) and spring planted crops (like sweet corn, cucumbers, and tomatoes) in a rotation can help reduce both spring and fall germinating weed species.

Ridge tillage is one system that can physically move weed seeds out of crop rows, provided that the seeds are concentrated near or at the soil surface. Just before planting, the tops of the ridges are removed by sweeps that push residues and the top inch or two of soil into the interrow valleys. This moves surface layer weed seeds to the interrow, where they will not compete as intensely with the crop, and they are easier to cultivate out. Later-season cultivations move soil back into the crop row to bury those weeds that emerge within the crop row, and also rebuild the ridges. Late-germinating weeds in the crop row may not significantly affect crop yields, but they can replenish the seed bank. Removing enough soil (about 2 in.) to leave a wide (12-in.), flat bed top for planting is recommended to ensure satisfactory weed control with this strategy (Mohler, 2001b).

References and Citations

• Brainard, D. C., R. R. Bellinder, and V. Kumar. 2007. Effects of floating row cover on weed emergence and stale seed bed performance. p. 63. In H. A. Sandler (ed.) Proceedings of the Sixty-first Annual Meeting of the Northeastern Weed Science Society, 2–5 Jan. 2007, Baltimore, MD. Northeastern Weed Science Society. Available online at: http://www.newss.org/proceedings/proceedings_2007.pdf (verified 11 March 2010).
• Brainard, D.C., A. DiTommaso, and C.L. Mohler. 2006. Intraspecific variation in germination response to ammonium nitrate of Amaranthus powellii originating from organic versus conventional vegetable farms. Weed Science 54: 435–442. (Available online at: http://dx.doi.org/10.1614/WS-05-162R1.1) (verified 23 March 2010).
• Buhler, D. 1997. Effects of tillage and light environment on emergence of 13 annual weeds. Weed Technology 11: 496–501. (Available online at: http://www.jstor.org/stable/3988427) (verified 23 March 2010).
• Caldwell, B., and C. L. Mohler. 2001. Stale seed bed practices for vegetable production. HortScience 36: 703–705.
• Davis, A. S. 2004. Managing weed seedbanks throughout the growing season. New Agriculture Network Vol. 1, No. 2.
• Davis, A., and M. Liebman. 2003. Cropping systems effects on giant foxtail (Setaria faberi) demography: 1. Green manure and tillage timing. Weed Science 51: 919–929. (Available online at: http://dx.doi.org/10.1614/P2002-133A) (verified 23 March 2010).
• Dyer, W. E. 1995. Exploiting weed seed dormancy and germination requirements through agronomic practices. Weed Science 43: 498–503. (Available online at: http://www.jstor.org/stable/4045586) (verified 23 March 2010).
• Egley, G. H. 1986. Stimulation of weed seed germination in soil. Reviews of Weed Science 2: 67–89.
• Egley, G. H., and Robert D. Williams. 1990. Decline of weed seeds and seedling emergence over five years as affected by soil disturbance. Weed Science 38: 504–510. (Available online at: http://www.jstor.org/stable/4045064) (verified 23 March 2010).
• Gallandt, E. R. 2006. How can we target the weed seedbank? Weed Science 54: 588–596. (Available online at: http://dx.doi.org/10.1614/WS-05-063R.1) (verified 23 March 2010).
• Gallandt, E. R., M. Liebman, and D. R. Huggins. 1999. Improving soil quality: Implications for weed management. Journal of Crop Production 2: 95–121. (Available online at: http://dx.doi.org/10.1300/J144v02n01_06) (verified 23 March 2010).
• Martens, K., and M. Martens. 2008. Weed control in row crops. Oral presentation, Pennsylvania Association for Sustainable Agriculture Conference, Feb. 7–9, 2008, State College, PA.
• Mohler, C. L. 2001a. Weed life history: identifying vulnerabilities. p. 40–98. In M. Liebman et al. Ecological management of agricultural weeds. Cambridge University Press, New York.
• Mohler, C. L. 2001b. Mechanical management of weeds. p. 139–209. In M. Liebman et al. Ecological management of agricultural weeds. Cambridge University Press, New York.
• Rasmussen, J. 2003. Punch planting, flame weeding and stale seedbed for weed control in row crops. Weed Research 43: 393–403. (Available online at: http://dx.doi.org/10.1046/j.0043-1737.2003.00357.x) (verified 23 March 2010).
• Schwinghamer, T. D., and R. C. Van Acker. 2008. Emergence timing and persistence of kochia (Kochia scoparia). Weed Science 56: 37–41. (Available online at: http://dx.doi.org/10.1614/WS-07-098.1) (verified 23 March 2010).
• Scopel, A. L., C. L. Balare, and S. R. Radosevich. 1994. Photostimulation of seed germination during soil tillage. New Phytologist 126: 145–152. (Available online at: http://dx.doi.org/10.1111/j.1469-8137.1994.tb07540.x) (verified 23 March 2010).
• Teasdale, J. R., and P. Pillai. 2005. Contribution of ammonium to stimulation of smooth pigweed (Amaranthus hybridus L.) germination by extracts of hairy vetch (Vicia villosa Roth) residue. Weed Biology and Management 5: 19–25. (Available online at: http://dx.doi.org/10.1111/j.1445-6664.2005.00155.x ) (verified 23 March 2010).