14%. Multiple cohorts were produced between February and October. No-till systems produced higher emergence rates than conventional tillage systems. Seedlings of B. tournefortii did not emerge from 5 cm soil depth; therefore, diligent tillage practices without seedbank replenishment could rapidly reduce the presence of this weed. A soil-moisture study revealed that at 25% of water-holding capacity, B. tournefortii tended to produce sufficient seeds for reinfestation in the field. Brassica tournefortii is a cross-pollinated species, and its wider emergence time and capacity to produce enough seeds in a dry environment enable it to become widespread in Australia. Early cohorts (March) tended to have vigorous growth and high reproduction potential. This study found B. tournefortii to be a poor competitor of wheat (Triticum aestivum L.), having greater capacity to compete with the slow-growing crop chickpea. Therefore, control of early-season cohorts and use of rotations with a more vigorous crop such as wheat may reduce the seedbank. The information gained in this study will be important in developing better understanding of seed ecology of B. tournefortii for the purpose of developing integrated management strategies.
Thus, one strategy for managing the weed seed bank, especially for smaller-seeded weeds, is to maintain seeds at or near the soil surface. It is here that seeds experience the greatest exposure to environmental cues that will encourage germination—the most effective means of debiting the seed bank—as well as greater exposure to seed predators (see Encouraging Weed Seed Predation and Decay). Studies have confirmed that some weed seeds, including velvetleaf, morning glory, and pigweed, germinate in larger numbers in untilled than in tilled soil during the first year after seed shed (Egley and Williams, 1990). It may be tempting to use inversion tillage to place seeds below the depth from which they can emerge. This may be an effective strategy for species with short-lived seeds (see below), but it may simply protect longer-lived seeds from mortality factors like seed feeding animals and decomposer fungi, only to be returned to the soil surface by the next deep plowing event.
This article is part of a series on Twelve Steps Toward Ecological Weed Management in Organic Vegetables. For more on managing the weed seed bank, see:
For a more representative sampling, collect sufficient soil samples to fill several pie dishes, or a seedling flat. The larger the sample, the more closely the observed weed emergence will reflect field populations.
Factors Affecting Weed Seed Longevity
Although seed longevity of agricultural weeds is a cause for notoriety, and a proportion of the population may remain viable for several years or decades, most of the seeds of many weed species will either germinate or die shortly after being dispersed from the parent plant. The seeds of many grasses are particularly short lived. For example, in a field study conducted near Bozeman, MT, wild oat seeds were incorporated into the top four inches of a wheat–fallow field, and approximately 80 percent of them died during the first winter (Harbuck, 2007). It is important to note, however, that postdispersal survival varies widely among weed species.
Weed seeds can have numerous fates after they are dispersed into a field (Fig. 1). Some seeds germinate, emerge, grow, and produce more seeds; others germinate and die, decay in the soil, or fall to predation. The seeds and other propagules of most weeds have evolved mechanisms that render a portion (a large majority in some species) of propagules dormant (alive but not able to germinate) or conditionally dormant (will not germinate unless they receive specific stimuli such as light) for varying periods of time after they are shed. This helps the weed survive in a periodically disturbed, inhospitable, and unpredictable environment. Weed seeds can change from a state of dormancy to nondormancy, in which they can then germinate over a wide range of environmental conditions. Because dormant weed seeds can create future weed problems, weed scientists think of dormancy as a dispersal mechanism through time.
Cultivation efficacy—weed kill—can vary considerably based on equipment, soil conditions, weed growth stage, and operator experience. Eighty percent mortality would be considered quite respectable, a level of weed control far less than that achieved with most herbicides. Therefore, without the “big hammer” of selective herbicides to remove heavy weed populations from standing crops, effective measures to reduce weed seed banks become vital for the organic farmer.
The number of viable seeds remaining from a given year’s weed seed return declines over time as a result of germination (successful or fatal), predation, and decay. The percentage remaining declines in an approximately exponential manner, similar to the decay curve for a radioactive chemical element—the time for the number to decline by 50% is roughly the same, regardless of the initial num. The half-life of weed seeds varies widely among weed species; for example, hairy galinsoga and some annual grass weeds, such as foxtail species, last only one to a few years, whereas some curly dock and common lambsquarters seed can last over 50 years.
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.
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.
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.
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.
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.
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.
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.