Weed communities in agronomic fields are dominated by annual species. Summer annuals initiate growth each spring from seeds found in the upper soil profile (Figure 1). In most fields, a small percentage of the emerging plants survive and contribute new seeds to the soil seedbank. Historically, most research of the annual weed life cycle has focused on seed dormancy and emergence (A), effect of control tactics on weed survival (B), and weed seed production (C). The fate of seeds between the time of maturation on the plant and entering the seedbank (D) has largely been ignored. Now is the time to think about controlling summer annual weeds prior to seed set in cropping situations where possible. Preventing seed production is important for driving down the weed seed bank and reducing the need for weed control inputs (i.e. herbici The fate of weed seeds in the soil has been an area of much research in recent years. Most studies have focused on the seeds that successfully produce seedlings since these are the seeds that cause immediate problems for farmers. In most studies, annual emergence typically accounts for 1 to 30% of the weed seed in the soil. Thus, the majority of seeds found in the soil seed bank fail to produce seedlings in any given year.
Weed seed predation in agricultural fields
Weed communities in agronomic fields are dominated by annual species. Summer annuals initiate growth each spring from seeds found in the upper soil profile (Figure 1). In most fields, a small percentage of the emerging plants survive and contribute new seeds to the soil seedbank. Historically, most research of the annual weed life cycle has focused on seed dormancy and emergence (A), effect of control tactics on weed survival (B), and weed seed production (C). The fate of seeds between the time of maturation on the plant and entering the seedbank (D) has largely been ignored. However, current research at Iowa State University and other organizations has shown that significant seed losses routinely occur in agronomic fields, and these losses may influence the effectiveness of weed management programs. This article will provide a brief summary of some of the current research in this area and the potential importance of seed predation to weed management.
Prairie deer mouse – a common seed predator.
Plant seeds are storage organs for high energy compounds that supply plant embryos the resources needed to germinate and develop into seedlings. These energy reserves are an excellent food source for a variety of animals that live in or near agricultural fields, including ground beetles (carabid beetles), crickets, mice and others. Estimates of cumulative seed losses due to seed predators have ranged from 20% for barnyardgrass and lambsquarter in a chisel plow system (Cromar et al. 1999) to 88% for giant ragweed in no-tillage (Harrison et al. 2003).
A common method of measuring seed predation involves lightly attaching seeds to sandpaper or a similar material and placing the seed cards in the field. After a few days the card is retrieved and the percentage of seeds removed is determined (Westerman et al. 2005). Averaged over 12 sampling periods from May through November, seed losses ranged from 7 to 22% per day depending on the crop present in the field in a study conducted near Boone, IA (Figure 2). The higher predation rates in small grain and alfalfa compared to corn and soybean may be due to differences in crop canopy development. The rate of seed predation typically increases as a crop canopy develops within a field. Corn and soybean canopies provide little protection for predators early in the growing season compared to small grain or alfalfa, and thus predators may seek other habitats when little canopy is present. Later in the season, predator activity is typically similar in corn and soybeans as in other field crops.
Insect predators (field crickets, ground beetles, etc.) are active during the growing season when temperatures are favorable for cold-blooded species, whereas field mice are active year round. Seed predators have a remarkable ability to locate seeds on the soil surface; however, once seeds move into the soil profile the threat of predation is greatly reduced. The highest rates of seed predation likely occur in late summer and early fall when weed seeds are shed from plants onto the soil surface. Tillage buries the majority of seeds at depths where predation is minimal. Avoiding or delaying fall tillage following harvest should increase seed losses due to predation. Seeds can also enter the profile due to the impact of rain droplets, by falling into cracks, or due to freezing/thawing cycles during the winter. Ongoing research at ISU is evaluating the fate of seeds on the soil surface and how long they remain available to predators.
Field crickets on seed card.
The preference of predators for different species of weed seeds in the field is poorly understood. When given a choice, seed predators often will feed preferentially on one species over another (van der Laat et al. 2006; Figure 3). A common question is whether seed predators pose a threat to crop seed. Seed size and depth of planting minimize risks of corn and soybean seed losses to predators. Small-seeded legumes and grasses are at greater risk for predation losses, but proper planting where the majority of seed are placed under the soil surface should minimize losses.
Significant numbers of weed seeds are consumed by predators in agronomic fields, but the full impact of seed predation on weed densities and weed management is poorly documented. Clearly, destruction of a significant percentage of the weed seeds produced in a field will impact the following year’s weed density. The impact of giant foxtail seed rain and seed predation on giant foxtail densities was evaluated near Boone, IA (Figure 4). Giant foxtail seed (750 / sq ft) were spread on the soil surface in standing corn in late September 2004. The field was planted to no-till soybean in 2005 and foxtail emergence monitored throughout the season. The experimental area had a history of good weed control, thus foxtail densities were very low (
Modeling efforts at ISU have shown that seed predation can significantly affect long-term weed population dynamics within agricultural fields. For example, in a 4-year crop rotation (corn/soybean/small grain+alfalfa/alfalfa) the seed bank of giant foxtail rapidly increased from 2000 seed/m 2 to 4.3 million seed / sq m over an 18 year simulation period in the absence of predation (Figure 5). However, allowing for 25% seed predation resulted in a static seed bank, whereas any seed predation in access of 25% resulted in a decline in the seed bank density. The diverse rotation required 80% less herbicide than a conventionally managed corn-soybean rotation.
The value of intercepting weed seed before they enter the seed bank is somewhat of a forgotten control tactic. In the 1930’s and 40’s, combines were commonly equipped with a weed seed collector that separated and collected weed seed from chaff as the crop was harvested. When modern herbicides were introduced in the 1950’s, it was considered less expensive and more convenient to control weeds with chemicals, and these accessories quickly disappeared from combines. In Australia, seed collectors are again being used on combines due to widespread herbicide resistance and the loss of effective herbicides. Rigid ryegrass infestations have been reduced by as much as 70% through use of weed seed collectors during harvest (Gill, 1995). The effectiveness of weed seed collectors varies among weed species depending on timing of seed shed. Weed species that drop the majority of their seed prior to crop harvest would not be impacted significantly by use of weed seed collectors.
Weed seeds are an important food source for a variety of organisms that live within or adjacent to agricultural fields. It is clear that seed predation is an important form of biological control that influences weed communities within agricultural fields. Yet to be defined is how cropping systems can be manipulated to enhance the activity of seed predators and maximize their benefit, therefore allowing reductions in other more disruptive control tactics.
ISU research cited in this article was partially funded by:
The Leopold Center for Sustainable Agriculture
USDA National Research Initiative
Cromar, H.E, S.D. Murphyand C.J. Swanton. 1999. Influence of tillage and crop residue on postdispersal predation of weed seeds . Weed Sci. 47:184-194
Gill, G.S. 1005. Development of herbicide resistance in annual ryegrass in the cropping belt of Western Australia. Aust. J. Exp. Agric. 35:67-72.
Harrison , S.K., E.E. Regnier and J.T. Schmoll. 2003. Postdispersal predation of giant ragweed seed in no-tillage corn. Weed Sci. 51:955-964.
van der Laat, R., M. D.K. Owen and M. Liebman. 2006. Quantification of post-dispersal weed seed predation and invertebrate activity-density in three tillage regimes. J. Agric. Ecosys. Envir. Under review.
Westerman, P.R., M. Liebman, F.D. Menalled, A.H. Heggenstaller, R.G. Hartzler and P.M. Dixon. 2005. Are many little hammers effective? – Velvetleaf population dynamics in two- and four-year crop rotation systems. Weed Sci. 53:382-392.
Weed Seeds This Fall Means More Weeds Next Spring
Control annual weeds now in fallow areas to prevent seed set. Also, now is the time to start considering ways to manage perennials in small grain stubble.
Control weeds before seed set. Photo credit: Penn State Weed Science, D. Lingenfelter
Now is the time to think about controlling summer annual weeds prior to seed set in cropping situations where possible. Preventing seed production is important for driving down the weed seed bank and reducing the need for weed control inputs (i.e. herbicides). It is rather easy to prevent weed seed production following a cereal grain such as wheat, barley, or oats as well as some vegetable crops such as sweet corn or snap beans. Proper timing of the control practice is essential in preventing seed production. In general, below is a summary of estimated seed drop for various weed species:
- Giant foxtail: late August and peak seed rain usually occurs from late September through the month of October
- Yellow foxtail: early August and continues into late October
- Pigweed species: begin to produce mature seed by mid-August
- Lambsquarters and ragweed: generally, do not mature until the month of September
- Palmer amaranth or waterhemp: make sure to monitor them routinely over the next couple months and control any regrowth or new seedlings before they set seed. Palmer amaranth plants notoriously retain their seeds late into the summer and fall and thus seeds don’t necessarily fall to the ground upon maturity but are usually spread via the combine.
To prevent seed production, fields can be sprayed with an effective herbicide or mowed once or twice. Glyphosate is particularly effective at stopping grass growth and reproduction. The plant growth regulators (2,4-D and dicamba) would probably be a better choice for broadleaf weeds. With giant foxtail, even treating the field by mid-September can greatly reduce seed production. If seed heads are present, check suspect fields to determine how advanced flowering and seed rain are and time control practices accordingly. In alfalfa or pastures, if weeds are taller than the forage, consider running a brush-hog at a high setting to clip off the immature weed seed heads above the forage canopy. Taking the time to prevent seed production this year can make a big difference next year. About 80% of weeds next season come from weed seed this fall. For more information on weed emergence, weed seed set, and seedbank dynamics, refer to “A Practical Guide for Integrated Weed Management in Mid-Atlantic Grain Crops.”
In addition, many perennial broadleaves are evident in these same small grain stubble fields. The challenge with perennial weeds at this time of year is the fact they are still in the vegetative and reproductive phases. Therefore, most of the plant sugars are not being significantly transported to the roots and a herbicide application now will mostly only impact the top-growth. One consideration would be to mow those fields soon to prevent seed production and allow regrowth to occur. Then apply an effective systemic herbicide (ie, glyphosate, 2,4-D, dicamba) in late September or early October so the herbicide will be transported to the roots for more effective control.
Fate of weed seeds in the soil
The fate of weed seeds in the soil has been an area of much research in recent years. Most studies have focused on the seeds that successfully produce seedlings since these are the seeds that cause immediate problems for farmers. In most studies, annual emergence typically accounts for 1 to 30% of the weed seed in the soil. Thus, the majority of seeds found in the soil seed bank fail to produce seedlings in any given year. The fate of seeds that fail to germinate and emerge is poorly understood. While some of these seeds are simply dormant and will remain viable until the following year, others are lost due to decay or consumed by insects or small animals. This article will describe results of an experiment that monitored the fate of seeds for the first four years following introduction into the soil.
Methods: Seeds of velvetleaf, waterhemp, woolly cupgrass and giant foxtail were harvested from mature plants during the 1994 growing season. The seeds were cleaned and counted and then buried in the upper two inches of soil on October 21, 1994. Two thousand seeds were buried within a 3 sq ft frame to allow recovery during the course of the experiment. Weed emergence was determined by counting seedlings weekly during the growing season. Emerged seedlings were pulled by hand after counting. In the fall of each year one quarter of the soil within a frame was excavated and the remaining seeds were extracted and counted. Corn or soybeans were planted between the frames during the course of the experiment to simulate agronomic conditions.
Results: The emergence patterns of the four species were described in an earlier article (see emergence patterns). The fate of the seeds (emergence, loss or survival in soil) during the first four years after burial is shown in Figure 1. In the first year following burial waterhemp had the lowest emergence (5%) whereas greatest emergence was seen with woolly cupgrass (40%). Total emergence over the four years ranged from 300 seedlings (15% of seed) for waterhemp to 1020 seedlings (51%) for woolly cupgrass. More than three times as many seedlings emerged in the first year than in subsequent years for velvetleaf, woolly cupgrass and giant foxtail, whereas 140 waterhemp seedlings emerged in 1996 compared to only 100 in 1995.
Figure 1. Fate of seeds during the four years following burial in the upper two inches of soil. Two thousand seeds of each species were buried in the fall of 1994. The area in white represents the number of intact seeds present in the fall of each year, green represents the total number of seeds that produced seedlings during the four years, and the blue represents the total number of seeds lost. Buhler and Hartzler, 1999, USDA/ARS and ISU, Ames, IA.
Seeds of the two grass species were shorter lived than those of velvetleaf or waterhemp. At the end of the third year (1997) no grass seeds were recovered. Somewhat surprising is that waterhemp seed was more persistent than velvetleaf in this study. Velvetleaf has long been used as the example of a weed with long-lived seeds. In the fourth year of the study four times more waterhemp seedlings than velvetleaf emerged and four times more waterhemp seed than velvetleaf seed (240 vs 60) remained in the seed bank.
For all species except woolly cupgrass the majority of seeds were unaccounted for (the blue portion of the graph) in this experiment. Determining the fate of the ‘lost’ seeds is a difficult task. A seed basically is a storage organ of high energy compounds, thus they are a favorite food source of insects and other organisms. In natural settings more than 50% of seeds are consumed by animals. The importance of seed predation in agricultural fields is poorly understood, but recent studies have shown that predation can be a significant source of seed loss. Another important mechanism of seed loss likely is fatal germination. This occurs when a seed initiates germination but the seedling is killed before it becomes established. Fatal germination probably is more important with small-seeded weeds such as waterhemp and lambsquarters than with large-seeded weeds, but is poorly understood. A better understanding of the factors that influence seed losses might allow these processes to be manipulated in order to increase seed losses.
So what does this mean as far as managing weeds in Iowa. First, consider how the methods used in this experiment might influence the results. The seeds were buried in the upper two inches of soil, the zone most favorable for germination. Most long term studies investigating the persistence of seeds have buried the seeds at greater depths than used here in order to minimize germination. If the seeds were buried deeper one might expect less emergence and greater persistence since the seeds would be at a soil depth with less biological activity. If the seeds had been placed on the soil surface it is likely that there would be more predation, less emergence and shorter persistence.
The results indicate that the seed bank of giant foxtail and woolly cupgrass should be able to be depleted much quicker than that of the two broadleaves. Maintaining a high level of weed control for two years should greatly diminish populations of these weeds in future years and simplify weed management. Unfortunately, a single plant escaping control can produce more seed than was introduced to the soil in these experiments, thus the seed bank can be rapidly replenished any time weed control practices fail to provide complete control. Finally, over 50% of velvetleaf and waterhemp seed was lost in the first two years following burial. However, significant numbers of seed of these species remained four years after burial. This will make populations of these two species more stable over time than those of woolly cupgrass and giant foxtail.
Doug Buhler is a Research Agronomist at the National Soil Tilth Laboratory, USDA/ARS, Ames, IA.