- Agronomic: rye
- Crop Production: cover crops
- Education and Training: demonstration, display, extension, farmer to farmer
- Pest Management: allelopathy, competition, mulches - killed
- Production Systems: general crop production
- Soil Management: green manures, soil analysis
Long Branch Farm is 535 acres of undulating clay in SW Ohio. The site has been farmed since the early 1800’s. Over half the farm is woods, creeks and wildlife areas. We have recently completed a project, supported by the Clermont County Soil and Water Conservation District, to fence the waterways from livestock. Of the 220 farmable acres, 125 is crop land and the rest grassland to support 30 Angus beef cows, with calves and followers, plus a flock of 30 ewes. The land is owned by the Cincinnati Nature Center, a local not for profit environmental education organization. This organization is primarily interested in the conservation of the wild areas for native plants and animals. As part of our outreach to the farmers and consumers of tomorrow the farm also hosts field study groups from local elementary and middle schools.
PROJECT DESCRIPTION AND RESULTS
On the farm, and on most neighboring farms, the standard crop rotation is alternating one year corn, on year soybean. We are seeking to introduce a cover crop into this rotation that will improve our crop management without impacting our farm operations or reducing profit.
Our current practice is to leave fields fallow after cropping and to rely on weed germination and crop stubble to provide a winter cover. Our primary weed over winter is wild garlic (Allium canadense). This noxious weed persists despite spraying in the spring.
Wild onion (Allium spp.) is a persistent weed in crop fields. There are approximately 60 species of Allium native to North America. The field weed is generally A. validum or A. canadense. Other varieties, such as A. vineale, have been introduced and become naturalized. All are closely related to cultivated onion. Wild garlic, wild onion, onion grass, etc. are all included in this report under the general term of “wild onion.”
Wild onion is a bulbous herb with a distinct odor. It has slender grass like leaves that can reach about 2 feet tall. Leaves are long and narrow arising from a small underground bulb. Flowers are produced in late summer at the top of a leafless stem. They vary in color from white to pink depending on species. The flower eventually becomes a cluster of bulblets which fall to the ground and propagate.
True wild onions are edible to humans but can be confused with similar looking plants such as Star-of-Bethlehem (Ornithogalum umbellatum) which has caused nausea, diarrhea and death. Wild onions are toxic to cattle, sheep, horses and other animals.
All weeds compete with the desired crop for water, nutrients and sunlight. The primary problem with Allium sp is that their inclusion will taint the harvested crop, giving wheat or soybeans an odor that makes it unfit as flour. It also precludes that crop for sale for seed. Without the options of sale for human consumption or as seed the price received will be much lower. The weed can be carried from field to field or to a neighboring property on tractor wheels, cultivation equipment or combine harvester. It takes time, and therefore expense, to ensure all machinery is free from onions before moving elsewhere.
There is an old country saying that the only way to rid yourself of wild onions is for you to die and go to heaven. A not quite so extreme solution was to sell the farm and move somewhere else. The current solution is to use Atrazine or Bicept (atrazine + duel).
Cover crops are used to provide winter ground cover. They help prevent soil erosion and are used to return nutrients to the soil. Winter cereal rye (Secale cereale) can suppress weeds through competition and allelopathy. Rye residues on the surface release inhibitory substances that affect germination and seedling growth of many weed species.
We tested the competitive and allelopathic effect of cereal rye, at tow different seed rates, on wild onions. We thought that using biological control may reduce our dependence on chemical sprays. Alternatively a combination of biological controls may, over time, eradicate this weed, resulting in a cleaner crop and better price. We also monitored soil nutrient levels during this experiment to see if there was any nutrient cost (or benefit) for the amount of weed control we gained.
During the planning for this project it was important that, if we adopted this cover crop practice for the whole farm, it would not negatively impact on farm income. We accepted that the experiment would cost extra time and money that is what the grant is meant to cover, but that the eventual benefits would exceed the costs.
With this in mind we used standard field practices and equipment, namely a 3 point mounted, PTO driven, fertilizer spreader to broadcast the seed, and a bush hog to mow the crop in the spring. The crop was sown on November 4th, 1999 a little past the ideal time but it was literally as the combine left the field. We had to wait until the corn was harvested as we do not have the machinery to sow into a standing crop. Broadcasting was thought to be a quick way to get the seed on. There is only a short time between harvest and the possible onset of wet weather. In SW Ohio the autumn is relatively mild and wet, hence we were able to consider a late October/early November sowing date. The next few days after sowing were in fact wet. I was undecided as to whether I should try to chain harrow or cultivate the field in some way to incorporate the seed or if the harvest trash would ball up and block the harrows. As it was the weather turned wet and I did not harrow the field. The rain helped incorporate the seed among the harvest residue.
According to the Sustainable Agriculture Network Handbook “Managing Cover Crops Profitably” (1991) rye is the only common cover crop that can be sown this late with a reasonable chance of success. It is also tolerant of atrazine. This is useful if there are likely to be traces of it left form the previous crop.
Three different crop fields were selected and in each were marked three adjacent one acre plots (150 x 290 ft each). In each field one plot was the control and received no treatment, one plot was sown at the recommended seed rate of 2 bushels winter rye/acre, and one plot was sown at double this rate (4 bushels/acre). Soil samples were collected from each plot and tested for nutrient levels. This was done to check that the plots were fairly uniform. It is important to know that if the weeds grew differently in one plot it was not down to nutrient level.
Over the winter I watched anxiously as the rye germinated and slowly established itself. Meanwhile the onions seemed to be doing quite well. By springtime there was a good, even establishment of rye. The density of the rye was such that, from a casual observation, it would have appeared as if there were no onions present. This was the key points of the project; does standing rye suppress onion or merely hide it?
To answer this question it is necessary to count the number of onions growing the rye and compare it to the control (untreated) plots. A representative count of the onion population was made rather than count every onion in each one acre plot. A metal square measuring 18” along each side gives a one fourth square yard sampling frame of grid. Starting at one side of a plot, a random transect angle was chosen. Walking along this line the grid was dropped every ten paces. Every onion growing within the square was counted, plus each major rye tiller. This process was repeated for each plot.
The allelopathic effect of rye mulch is well documented (ATTRA literature). The rye exudate has a herbicidal effect on several dicot and monocot plants. No report was found of its effectiveness on Allium species. To test if rye as a field mulch had any effect on onion grasses half of each plot was mown. The mulch was left undisturbed where it lay. A count of onions in both mown and unmown areas was made after 21 days. Another soil sample was also made at this time of all mown and unmown plots.
People without whom this project would not have been possible include:
- Gary Gao, OSU Extension, weed identification
- Steve McKee, OSU Extension, agronomy
- Paul Berringer, Clermont SWCD, experiment design, sampling and outreach
- Steve Ristmiller, Clermont SWCD, GIS
- George Cummings, USDA NRCS, experiment design
- Butch and Scot Foreman, neighboring producers, fieldwork, outreach
- Celeste Baumgartner, Farmweek reporter, outreach
- Matt Reece, Ohio County Journal, outreach
- Cindy O’Donnell, project design
- Fred Schlichter, Agricultural Operations Manager, scouting
- Larry Raper, Vic Feineur, Keith Claverly, helpers with plot layout, etc.
- Mike Young, Agri-Urban, Pleasant Plain, seed and agronomy advice
- Herb Meyer, SW Landmark, outreach
- Jackie Belwood, CNC Scientist, analysis, outreach
- Preston Sullivan, ATTRA, cover crop advice
- Ken Schneider, NCRSARE Program Coordinator for his interest and patience.
The purpose of the experiment was to determine the populations of onion growing in the field and compare that to populations growing under the different treatments. It is not feasible to count every onion in a field so representative sampling was used. Statistical analysis was then used to extrapolate the likely field populations. For comparison this is expressed per square yard.
From the individual 0.25 square yard counts, we calculate the mean (or average) number of onions growing per square yard in each block. Statistics are used to calculate the likely range around this mean that could contain the true population. The accuracy of our mean depends on the number of samples that are counted and the variability between those samples. A large number of samples with little difference between them is likely to give a close approximation to the true population.
The number of samples counted in each block is statistically small. The “Student’s t-distribution” gives the best indication of population distribution around the mean under these conditions. Using a 95% confidence level the likely range that contains the true population number is calculated. For example, plot A had a mean of 1.42 wild onions per 0.25 square yard. The student’s t-distribution tells us that we can be 95% certain that the true population is within plus or minus 0.7 of the mean. Hence the true population of onions in plot A is between 0.72 and 2.12 onions per 0.25 square yard. This is multiplied by four to give the population per square yard of between 2.86 and 8.47. This process was repeated for plots B and C that make Block 1.
Table 1. Wild onion population range in Block 1.
Plot, Minimum, Mean, Maximum
A, control, 2.86, 5.67, 8.47
B, rye x 1, 1.88, 4.33, 6.79
C, rye x 2, 0.00, 1.00, 2.02
The plots A, B and C in Block 1 are united in similar location, soil type, aspect, etc. For the purpose of this experiment was assume this similarity exists and that any differences in results are due to the different treatments applied to that plot, namely:
- rye sown at the recommended seed rate of 2 bushels/acre, (rye x 1)
- rye sown at twice the recommended seed rate, (rye x 2)
- no rye sown, (control)
So in Block 1 we have calculated that the true population of wild onions is between 2.86 and 8.47 onion plants per square yard. If rye is grown at the recommended seed rate of 2 bu/acre then, in this block, the population of onions becomes somewhere between 1.88 and 6.79 onion plants per square yard. This looks like a reduction in onions but in reality is not. The possible population counts between 2.86 and 6.79 are shared by both plots. For example, the true population could be 5 onions per square yard in both of these plots, indicating that the rye had no effect on onion population.
Rye sown at twice the recommended seed rate gave a population range of zero to 2.02 onions per square yard. The population ranges between control and rye x 1 overlap, so too does rye x 2 and rye x 1. However there is no overlap between control and rye x 2. From this we infer that the likely true population of onion is different and that applying rye at twice the recommended seed rate does have an effect on wild onion in this block.
This process was then repeated for the other blocks.
Table 2. Wild onion population range in Block 2
Plot, Minimum, Mean, Maximum
D, control, 4.66, 6.74, 8.82
F, rye x 1, 2.86, 5.26, 7.67
E, rye x 2, 1.31, 4.42, 7.53
The population ranges between control, rye x 1, and rye x 2 all overlap. There is no statistical population difference in Block 2.
Table 3. Wild onion population range in Block 3.
Plot, Minimum, Mean, Maximum
G, control, 19.47, 25.50, 31.53
H, rye x 1, 13.26, 19.82, 26.37
I, rye x 2, 6.45, 10.17, 13.88
The population ranges between control and rye x 1 overlap, so too does rye x 2 and rye x 1. However there is no overlap between control and rye x 2. From this we infer that the likely true population of onion is different and that applying rye at twice the recommended seed rate does have an effect on wild onion in this block.
Each block is looked at separately so that any unusual results can be spotted. If one block was very different from the others it could help identify an experiment error. In this experiment all three blocks show a similar pattern. The overall results are amalgamated in table 4.
Table 4. Wild onion population range, all blocks
Minimum, Mean, Maximum
Control, 13.02, 17.13, 21.23
Rye x 1, 7.60, 11.09, 14.59
Rye x 2, 4.01, 6.18, 8.35
The population range between control and rye x 1 overlap, but there is no overlap between control and rye x 2. We therefore surmise that there is a difference between control and rye x 2 and that difference is caused by sowing rye at twice the recommended seed rate.
Cut and Mulch:
The second hypothesis is that rye, cut and left as a mulch, inhibits wild onion re-growth. To test this hypothesis part of each plot was mown and the cut material left to mulch. These became a sub-section of each plot designated with “c.” For example plot Ac was the control with no rye sown in it, Cc had 2 bu/acre of rye sown, and Bc had 4 bu/acre of rye sown. All were mown and left for 3 weeks. They were then sampled using the 0.25 square yard grid. All onions growing in each random placement of the grid were counted.
The results were analyzed as before. The statistical range of the population mean in each block was calculated.
Table 5. Wild onion population range in blocks 1, 2, and 3 after cutting and mulching.
Plot, Status, Minimum, Mean per square yard, Maximum
Ac, control, 1.21, 3.33, 5.45
Cc, rye x 1, 0.00, 0.33, 0.99
Bc, rye x 2, 0.00, 0.00, 0.00
Dc, control, 0.22, 1.33, 2.45
Fc, rye x 1, 0.00, 0.00, 0.00
Ec, rye x 2, 0.00, 0.00, 0.00
Gc, control, 1.86, 2.88, 3.89
Hc, rye x 1, 0.11, 0.63, 1.14
Ic, rye x 2, 0.00, 0.00, 0.00
In each block there is a difference between the statistical population mean for each plot.
The cut and mulch results were amalgamated. The results are show in Table 6.
Table 6. Wild onion population ranges, all cut and mulched blocks.
Cut and Mulched, Minimum, Mean, Maximum
Control, 1.86, 2.64, 3.42
Rye x 1, 0.10, 0.42, 0.76
Rye x 2, 0.00, 0.00, 0.00
This shows that there is a significant reduction in wild onions when rye is cut and mulched.
The number of major rye tillers was recorded in each uncut sample grid as well as the number of wild onions. This allowed a test for correlation; does more rye result in less onion in a predictable manner? The correlation was -0.26 which means that there was some correlation but it is not a strong or directly related equation. A correlation of 1 or -1 means that they are directly related and the result can always be predicted.
Soil samples were collected from each plot in November 1999 before the rye cover crop started to grow. In June 2000 further samples were collected from each uncut plot and form each cut and mulched plot. The November figure give a fall baseline, the June figures are uncut and show the summer growing season nutrient status, and the June cut figures are after the rye and weeds have been cut and mulched. June and June cut samples were collected on the same date. No fertilizer was applied during this experiment. The average available nutrient increases uniformly in all blocks between November and June. This is probably due to warmer soil increasing the bacterial decomposition of soil organic matter.
If cutting and mulching has no immediate effect on soil analysis then the June cut figure should be the same as the June figure. Phosphate showed a possible increase. Magnesium and Cation Exchange Capacity (CEC) showed variable results. The soil pH remained within the 5.6 to 6.0 range.
These soils analysis results seem to indicate that growing a cover crop of winter rye did not significantly alter the soil nutrient levels. Cutting and mulching did cause some changes in soil nutrient levels but the extent and duration of these changes are not clear. The sample was taken only 3 weeks after mowing. Further study is needed to investigate the long term effects of cover crop mulching on soil nutrient status.
This experiment has prompted me and my colleagues to look more closely at how we farm and the impact our management can have on both our crops and on the environment. The project has identified that a winter rye cover crop does have some effect on wild onions but for a lasting effect it will need to become a regular farm practice. Cover crops used in conjunction with no till (to prevent mechanical dispersal of onions) may reduce the number of wild onions over time.
Cover crops take commitment. It takes time, effort and money to get back out in the fields immediately after harvest. There is only a small (or even non-existent!) period of time between harvest and the onset of winter weather. Seed must be purchased and another pass of the fields made. In the spring a timely cut and mulch must be made and there is the possibility of surface trash blocking the seed drill when trying to plant the following cash crop. The alternative of just leaving the field and then spraying with atrazine in the spring seems appealing. However the potential environmental dangers of atrazine and the added benefits of cover crops mean that this is a relevant alternative production method.
On the farm these results have given us another option to consider. We are seeking to extend our crop rotation beyond corn and beans. Winter rye cut in the spring for silage could give us extra forage while still having some effect on the wild onions, even if it is less effective than leaving it as mulch.
This project prompts further questions: What is the long term effect of annual cover crops on wild onions? What effect does mulch have on soil nutrient status? I was surprised at the amount of time and physical and mental energy this seemingly simple project demanded. It has been interesting and worthwhile and we will probably commit ourselves to another SARE project again at a later date.
The full report is rather bulky for a quick read! A flyer or poster copy was produced for distribution at the Ohio Ecological Food and Farming Association (OEFFA) Annual Conference. It is also available through the Clermont SWCD. Details have also been sent to the OSU Extension Service and a press release sent to various farming publications.
A copy of the full experiment, including all the data, graphs, photographs, etc. will also be given to Clermont SWCD and OSU Extension. Details are due to be included in their various newsletters, etc. Copies of both the full report and the flyer will be available from Long Branch Farm.
Cincinnati Nature Center has distributed the results through its Agricultural Operations sub-committee, an advisory group of local farmers, landowners and interested parties.