Weed Management in Organic Conservation Tillage/No Tillage

Objective 1. Develop agronomic practices for managing allelopathic cover crops to control weeds while enhancing soil quality.

We conducted successful field and laboratory experiments to develop organic no-till systems for soybean production. In the first two years, brassica, hairy vetch, and rye cover crops were planted in fall and soybeans drilled into standing cover crops, with appropriate no-cover crop controls. Weed control was excellent in rye plots, but brassica cover crops failed to establish both years in spite of planting times ranging from August 15 to October 1, during which ideal establishment conditions prevailed. In two years of field trials at two locations, only rye cover crops gave sufficient season-long weed suppression. We attempted to combine rye with brassica, but the optimum seeding dates for the two species differs greatly in our area, so we did not get a good stand of both species. Therefore, subsequent experiments focused on small grains as a cover crop.

Comparisons of small grain crops (rye, barley, oat, spelt, wheat, triticale) showed that rye was most effective, followed by wheat and spelt; barley and triticale were not effective. A novel design for a cover crop roller was built and tested for two growing seasons on several cover crops. Our roller is simpler to construct than those described from elsewhere, but it could be improved by using 2-inch rather than 4-inch angle-iron for crimping the cover crops. However, comparison studies showed that when weed populations are not high, the need for the roller is questionable, and simply drilling into standing rye is equally effective (one year’s data only).

Rye was planted in October and no side-dress fertilizer was applied in spring, contrary to usual practice. Food-grade soybeans were planted at 90 lb/acre with a no-till drill in early June, when rye was heading. There were three treatments for suppressing rye: Roller (see figure), mower (brush-hog), and untreated (plant directly into standing rye). More weed pressure was recorded in plots that were rolled or mowed than those where rye was left standing (Table 1), however this level of weed pressure is very low, and all treatments were considered successful compared to the control. There were no significant differences between soybean yields (the “control” plot had neither soybeans nor cover crop, and thus indicated the potential for weed growth in the plots).

This system was successful during 3 years of studies. We speculate that success depends on a high level of cover crop biomass. We tested several rye varieties, and only those that produced high levels of cover gave adequate weed suppression. Rye provides early spring cover and suppresses spring weeds. Since no nitrogen was applied in spring, the rye takes up soil N, making it unavailable to N-loving weeds. Rye was under some nutrient stress in the plots, but this might actually have been helpful, since such stress might have increased production of allelopathic chemicals that help suppress weeds. We speculate that the N deprivation, allelopathy, and physical impedance of emergence work together to suppress weeds. Soybeans were planted at a high seeding rate, they fix their own N, and there was ample soil moisture, so they grew quickly to form a canopy, which further helped to suppresses weeds. When all data from cover crops and weeds were combined, the relationship between cover crop biomass and weed cover showed a fast decline in weed cover as cover crop biomass increased, with about 1500 g/m2 of biomass needed to achieve excellent weed suppression. This relationship held for all four rye varieties tested.

Objective 2. Evaluate the effectiveness of various natural products for suppressing weeds.

The tests of natural product effects on preemergence activity showed significant germination inhibition from extracts of several crop seeds. Similar results were found for crabgrass and pigweed seed germination. About 50% inhibition has found for some extracts, but other extracts did not reduce germination sufficiently to pursue use of those species in subsequent tests. In laboratory studies, we found several seed extracts with significant suppression of weed seed germination, but no effect on photosynthesis or growth in postemergence applications.

We tested extracts in soil media as well as in petri dishes with filter paper to determine if the active compounds would be bound to soil and lose activity. Results showed that activity was not lost in soil. We focused on wild carrot seed extracts since this was the species that showed activity and it is one that is easily cultured and readily available. It also represents a case of making a potentially useful product from a species that is considered a weed. It is also considered a useful species for providing habitat for beneficials, and so would fit in well with farmscapes that leave borders for such habitat.

The LD50 represents a value that can be used for comparison among products and formulations. The LD50 for pigweed seed germination was a 19.2% dilution and for crabgrass seed germination the LD50 was a 4.2% dilution when seeds were treated with wild carrot extract (Figure 1). This LC50 is for the crude extract, and fractionation is not yet complete to determine the LC50 for the active fractions. Pigweed, a dicotyledonous species, is very sensitive to the compounds in the extract, whereas crabgrass, a monocotyledon, was slightly more tolerant. Bioactivity at low LC50 levels is a favorable aspect for soil-applied bioherbicides. Inhibition at low concentrations is environmentally and economically friendly. Low concentrations of bioactive compounds reduce the chances of leaching into the surrounding environment and groundwater. It is also economically favorable since lower amounts of these compounds would be needed for effective weed control.

In the evaluation of postemergence activity, most seed extracts did not show any post emergence herbicidal activity on pigweed or crabgrass plants (data not shown). Only one extract showed some suppression of crabgrass height. This result was somewhat unexpected since the extract showed little suppression of germination. The study was repeated with the same result; therefore, further studies are being conducted to explore the possible mode of action of this extract.

The field application test was not successful due to very dry conditions just after application. This study will be repeated if a no-cost extension is approved.

Some promising extracts have been found, from seeds of carrot and parsley, with variable activity from seeds of rye, buckwheat, mustards, and clovers. The appeal of this approach is that seeds of these species are available on the market and the quality is expected to be fairly consistent, as opposed to allelopathic chemicals that might be found in stems or roots.

Results suggest that seed extracts may be a source of an effective ‘natural’ soil applied herbicide. Extracts inhibited pigweed and crabgrass germination in soil, at low concentrations. The activity of the extracts was stable over time, and not affected by temperature. Some seed extracts exhibited selectivity by significantly reducing crabgrass germination without significant impact on pigweed, a broadleaf weed.

General Methods

Experiments were conducted in northeast Ohio on a research farm and by local growers. The soils in this part of the country are generally forest-based silt loams with organic matter around 2% and a pH between 5.5 and 6.0. The fields are transitional or certified organic under the USDA guidelines, and had been in forage, wheat, soybean and corn rotations. Four-row planter and two-row harvest machinery were used throughout. Field experiments were arranged in a randomized complete block design, with a minimum of four replications of plots that are typically 20 ft wide and 40 ft long. All experiments were conducted at least two times, in different fields, with some adjustment to treatments to accommodate new ideas and discard treatments that had no chance of working.

All data were subjected to analysis of variance, with treatment type, timing, and/or application rate as independent variables contributing to the variation in dependent variables (percent control, weed density, crop yield etc). Interactions, where present, were examined by graphing the response curve for the relationship between treatment application time and effectiveness of control for various combinations of treatment type and application rate.

The brassica cover crop used in these studies was ‘Ida Gold’ yellow mustard (Sinapis alba), hereafter referred to as ‘brassica’. This variety possesses a glucosinolate profile that had been shown – in greenhouse experiments - to be effective in suppressing weeds. The oat variety was ‘Bob’, a southern spring oat with superior cold tolerance. In the middle of the project, rye varieties were introduced, and these are described below.

Results were open to public comment and evaluation at field days and at grower meetings, and we are in the process of revising our project web-site to include pictures of plots and details of experimental procedures. The plots have been made available to many growers, who have been encouraged to visit the experimental plots as often as they wish, and specific meetings were arranged to make critical management decisions and to evaluate treatment responses.

Objective 1. Develop agronomic practices for managing allelopathic cover crops.

a. Evaluate planting time and methods for establishment.
Experiments were established in fields rotating from wheat to soybean, silage corn to soybean, and soybean to corn. Following winter wheat, the brassica cover crop were seeded in mid-July (just after wheat harvest) and mid-August using a no-till drill. For the soybean crop rotating to corn, we broadcast the mustard seed into the standing crop at the same times as above. In all of these experiments, brassica was seeded at rates of 10, 30, and 50 lbs/acre. The factorial set of main–plot treatments in each experiment were arranged in randomized complete blocks as described above, so that main effects and interactions of planting time and rate could be evaluated. A no-brassica check plot was included in each replication.

In addition, there was a three-year experiment in a wheat-soybean-corn rotation where we intended to plant mustard every year in order to monitor cumulative effects of the cover crop. Since the mustard cover crop proved to be unsuccessful, this aspect of the project was abandoned. Instead, we put our effort into other cover crops, especially rye, which proved very successful.

After a killing frost, cover crop growth was harvested from two 1-m2 quadrats per plot. The total biomass was weighed to calculate total dry weight production. The following spring, soybean and corn crops were planted into the cover crop residue using a standard no-till soybean drill and corn planter as soon as practical after April 15, or after rye heading (early June). Crop varieties (mostly organically produced and all Non-GMO) were chosen to match those used by local farmers, and planted at standard seeding rates.

The impact of cover crop treatments on weed suppression was measured by sampling each crop in the rotation. All weeds within four randomly located quadrats (0.25 m2) per plot were counted and weighed by species for measures of weed density and biomass. For soybean and corn, weeds were sampled 2 weeks after planting, then at first flower for soybean, and just before tasseling for corn. All plots were harvested using a plot combine and yields (on a dry weight basis) recorded.

b. Determine optimum proportions of mixed cover crop species for weed suppression.
We planted fall-seeded spring oats, hairy vetch, rye, and brassica to provide a mixed-species cover crop. Part of the rationale was to include a mix of species that have allelopathic activity in order to provide superior weed suppression. The experiment was a 2 x 4 factorial. Various species mixes were no-till planted at two times (mid-July, mid-August) after wheat harvest. Cover crop biomass was measured as described above. Soybeans was planted into all plots in the spring and weed density, and weed biomass measured as above.

c. Evaluate grower experience in on-farm trials using cover crops.
The grower cooperators defined specific questions regarding the implementation of reduced tillage and cover crops. These were the basis of their own on-farm experiments. For example, some of the farmers indicated an interest in testing the brassica in strip plantings. Standard and alternative management strategies were tested in strips, using the growers’ standard or modified equipment. A similar set of measurements on cover crops, weed populations, and crop performance were made as described above. Over time, the growers modified their approaches and added other species, such as sunflower, sorghum-sudan, and tef to cover crop mixes.

Objective 2. Evaluate the effectiveness of various natural products for suppressing weeds.

a. Screen essential oils, vinegar products, organic soaps, etc. against typical grass and broadleaf weeds in greenhouse studies.
Preliminary studies with natural products, such as salts of fatty acids and oils showed some toxicity to several weeds. In addition, or preliminary studies showed that vinegar solutions were effective against newly emerging seedlings. Our intention was to develop the potential of such products that would be acceptable in organic systems.

We compared the phytotoxic potential of several natural products that are mostly nontoxic, aromatic liquids extracted and distilled from plants. We focused initially on products that are readily available and that have shown biological activity in preliminary studies. These include essential oils in addition to various types of vinegar and salts of fatty acids. Eventually, we focused on extracts derived from seeds of various crop species. The rationale is that these are readily available and the extractions could be done by farmers themselves.

Greenhouse screenings used typical plant species that vary in susceptibility to most conventional herbicides (pigweeds, foxtail, and crabgrass). These were grown in square flats using a standard greenhouse growth medium. Various concentrations of each extract plus a water control, with or without a surfactant, were applied to 2-leaf seedlings. Visible injury symptoms were recorded daily for 10 days. Treatments were replicated 5or 6 times and experiments were repeated. Products showing significant injury (necrosis, discoloration, or stunting) were included in follow-up tests using a larger range of concentrations and additional test plants and growth stages.

b. Evaluate effectiveness of promising natural materials for suppressing weeds that appear in a cover crop system.
Natural products showing promise in the greenhouse screening study were tested in field plots. Each product was applied at three application times and three application rates, plus a water control. For example, cinnamon oil was tested 7.5%, 15%, and 30% v/v. Plots were arranged in a randomized complete block design with four replications. All products were applied with a hand-held sprayer in water at recommended volumes. Visible injury symptoms on weeds and crops were recorded 2 and 7 days after application. Treatments that appeared to be effective were repeated in subsequent growing seasons.

Since few of the products showed sufficient activity to be useful for farmers, we initiated a test of seed extracts. Water and ethanol extracts were prepared from ground seeds and filtered. Ethanol extracts were evaporated to dryness and reconstituted with water. Standard germination bioassays were conducted using smooth pigweed (Amaranthus hybridus L.) and large crabgrass (Digitaria sanguinalis (L.) Scop.).

In follow-up studies we estimated the LC50, evaluated chemical stability, and compared extraction techniques, focusing on species known to accumulate a large diversity of biologically active chemical compounds. Because we detected significant weed suppression from these extracts, and because others have reported bioactivity of interest for pest management, we focused initially on these species. We used petri dishes containing Wooster silt loam soil. We planted 50 seeds of four weed species. Four grams of ground seeds were spread uniformly, followed by 15 ml H2O. For extracts, 15 ml of extract (plus a water control) were applied per dish. The dishes were covered, sealed in plastic bags and incubated at 30/25°C (12-hr light/12-hr dark). Germinated seeds were counted and removed daily for 14 days.

c. Evaluate grower experience using natural products on organic farms.
The grower cooperators were given access to greenhouse and field studies of the various natural products. Some used vinegar as a pre-plant treatment, to kill a cover crop before planting, or as a postemergence rescue treatment when weeds get away from them. The original plan was to conduct replicated strip trials on the farms, but this became complicated due to safety concerns about most products. Therefore, we let growers use whatever commercial products they had an interest in, which included vinegar solutions and salts of fatty acids. Concerns arose about the concentration of vinegar that are acceptable for organic standards. In most cases, the lower concentrations were not effective against weeds in these experiments.