Jean Peterson and Al Sterner have found that under certain conditions, flame weeding, combined with mechanical weeding, can be more cost effective than the use of herbicides and can significantly reduce the amount of hand weeding required in vegetable production.  As a result, they now restrict their use of herbicides to occasional, localized weed problems.

Peterson and Sterner make their living growing and marketing up to 50 acres of vegetables full time on their 200 acre farm just east of Delano, Minnesota in western Hennepin County.  They grow a wide variety of vegetables, including asparagus, sweet corn, tomatoes, squash, peppers, beans, onions, and pumpkins.  They also grow a number of acres of hay.  Their sustainable approach includes the use of compost, manure, cover crops, and crop rotation with alfalfa and small grains.

Some of the vegetables Peterson and Sterner grow are compatible with flame weeding.  For example, onions can be flamed around two weeks after the transplants have gotten established. Corn, however, can be flamed at any time, though many flame users will not flame from the six to twelve inch growth period to avoid stressing the plants while they are developing their root systems.  The plants may droop or wilt initially, but they recover by putting out new growth.  However, the weeds are killed.  Green beans, on the other hand, cannot handle exposure to the flame, so flame weeding is used on them only before the bean plants emerge to deal with the initial flush of weeds.
 

Corn Before Weeding

Corn After Flame Weeding

View Along Corn Rows After Flame Weeding

Flame Weeded Row of Onions

Methods of Flame Weeding

Peterson and Sterner have tested two different methods for flame weeding.  The first method uses a hand held propane torch connected to a backpack supported fuel tank.  This manual method enables greater selectivity and accuracy in applying the flame. 

Peterson with "Wand" Flamer

In the second method, four propane burners are attached to a two row, rear mount, tractor-drawn cultivator.  Some farmers use four or six row flamers, depending on the size of their operation. This method is especially useful for large fields and enables flame weeding and mechanical cultivation to be accomplished in one pass.  Burners must be adjusted depending on conditions. 

Tractor-mounted Flamer

Sterner writes, "While we do have our flamer on a cultivator, we have found that it is better not to use the cultivator part too much, we just scratch the center of the rows, as we found that the dirt the shovels throw can interfere with the flame hitting the target weeds and interfere with the ability to travel at the best ground speed for flaming."

Cremating the weeds is not necessary.  When the weeds have been sufficiently heated to destroy them, pressing a weed between finger and thumb will result in an imprint formed by burst surface cells.

Imprints on Flamed Grass Blades

Peterson and Sterner stress the importance of handling fuel, equipment, and flame safely.  Operators should learn and use the safety rules for proper inspection, filling, and use of propane tanks and equipment.  Flaming should be restricted to calm conditions and attention paid to speed and direction of any air movement.  When turning at the end of rows, burners should be adjusted to a pilot setting.
 

Experimental Approach

After several years of experience with flame weeding in their vegetable operation, Peterson and Sterner conducted a two year study, sponsored by the Minnesota Department of Agriculture's Energy and Sustainable Agriculture Program to attempt to quantify the labor of hand hoeing saved by flame weeding over their previous approach, which consisted of hand hoeing and mechanical tillage.  In the second year, they received funding from the Agricultural Utilization Research Institute to expand the study to include the use of herbicides in the comparison.

Demonstration plots for onions and green beans were laid out as in Figures 1 & 2.  Each demonstration plot was divided into subplots, twenty subplots for the onions and sixteen subplots for the green beans.  Each of the subplots contained four twenty five foot rows.  Five alternative treatments were applied to the onions, with each treatment being used in four subplots distributed across the overall plot.  Similarly, four alternative treatments were used for the green beans, with four subplots allocated per treatment.

Alternative treatments for the onion demonstration plot included

1. hand hoeing three times and mechanically cultivating three times
2. manual flame weeding once, hand hoeing twice, and mechanically cultivating three times
3. tractor flame weeding once, hand hoeing twice, and mechanically cultivating three times
4. applying herbicide twice, hand hoeing once, and mechanically cultivating twice
5. null treatment, to serve as an experimental control

In the herbicide treated subplots, Prowl 3.3 EC was applied at a rate of 2 quarts per acre after transplanting the onions.  Roughly six weeks later, Poast 1.5 E was applied at 1.5 pints per acre.  These subplots were hoed once, about two weeks after the second application of herbicide.
 
 

Figure 1.  Layout of Onion Demonstration Plot

101 - Tractor Flame	301 - Cultivate
102 - Herbicide	302 - Control
103 - Cultivate	303 - Hand Flame
104 - Control	304 - Tractor Flame
105 - Hand Flame	305 - Herbicide
201 - Herbicide	401 - Control
202 - Cultivate	402 - Hand Flame
203 - Control	403 - Tractor Flame
204 - Hand Flame	404 - Herbicide
205 - Tractor Flame	405 - Cultivate

Alternative treatments for the green beans included

1. hand hoeing twice and mechanically cultivating once
2. manual flame weeding once, hand hoeing twice, and mechanically cultivating once
3. tractor flame weeding once, hand hoeing twice, and mechanically cultivating once
4. null treatment, to serve as an experimental control

Figure 2.  Layout of Green Bean Demonstration Plot

404 - Cultivate	204 - Tractor Flame
403 - Tractor Flame	203 - Hand Flame
402 - Hand Flame	202 - Control
401 - Control	201 - Cultivate
304 - Tractor Flame	104 - Hand Flame
303 - Hand Flame	103 - Control
302 - Control	102 - Cultivate
301 - Cultivate	101 - Tractor Flame

Results

For each of the demonstration plots, the amount of time spent hand hoeing was recorded, and the crop yield was measured.  These quantities are plotted on a per acre basis for the onions in Tables 1 and 2 and for the green beans in Tables 3 and 4.  (Since the onions were planted with a three foot spacing between rows, yields per acre may be lower than with alternative spacing practices.)
 
 

Table 1.  Time Required For Hand Hoeing In Onions
 

Table 2.  Onion Yield

In comparison to Peterson and Sterner's previously used combination of hand hoeing and mechanical cultivation, flame weeding substantially reduced the time required for hand hoeing.  However, analysis of the results for the onion plot is complicated by differences in the weather, perhaps the most important variable that could not be controlled.  During 1997, the growing period through June was exceptionally dry followed by a long, continuously wet period in July.  The effectiveness of the application of Prowl seems to have been significantly reduced by the early dry weather.  The wet weather that followed did not seem to cause problems for the treatment of Poast, but it did prevent timely application of flame and cultivation to the other plots, resulting in their being largely overrun by weeds and requiring more time in hand hoeing than otherwise expected.  Despite the vicissitudes of the weather, yields did not seem to suffer for the non herbicide treatments.

Records for the herbicide treatment show the lowest time spent in hand hoeing, since these subplots were hand hoed only once.  However, the herbicide treated subplots showed a substantially lower yield, possibly due to high weed pressure during establishment of the plants.  Substituting an early pass of the flame for the application of Prowl might have prevented this problem and would have cost substantially less than the herbicide.  Further experiments will be needed to determine the relative cost effectiveness of the use of herbicides versus the other approaches under a broader range of conditions.  However, in the final analysis, the differences in the economic value of crops grown with and without herbicides should also be considered.
 
 

Table 3.  Time Required for Hand Hoeing in Green Beans
 

Table 4.  Green Bean Yield (1997)

Yields were not considered for the first year's production of green beans due to substantial losses to deer, which resulted in a change of location of the demonstration plot for the second year.  Under manual flame weeding, the yield was significantly better than that of the other plots.
 

Conclusions

Flame weeding provided substantial benefit in reducing the cost of hand hoeing of herbicide-free onion and green bean crops.  Flame weeding is broad spectrum, with no observed weed resistance, and the fuel and equipment required are relatively inexpensive.  However, the timing and method of application are critical and vary from crop to crop.  Flame weeding is not appropriate for every crop or all weather conditions.  Safety considerations must be carefully followed in use of the propane fuel and equipment.  Given the substantial potential benefits of flame weeding, Peterson and Sterner suggest that interested farmers experiment with the method on a small scale initially to determine how best to integrate the method into their operations.
 

Collaborators

Peterson and Sterner were aided in their investigation of flame weeding by

• George Yokiel, an experienced weed flamer
• Tom Wegner, University of Minnesota Extension
• Eric Hoeft, a University of Minnesota graduate student
• Merlyn Flint, P and H Gas
• Vince Fritz, University of Minnesota Waseca Experiment Station
• Tim King, Sustainable Farming Association of Minnesota
• The members of Peterson's and Sterner's family and many high school students who work on the farm.

Sources

Frazeur, Andrea, (1997). "Weeds Aflame," AURI AG Innovation News, July 1997.

"Peterson Produce," letter. 

"Herbicide Addition to MN Energy and Sustainable Agriculture Grant Study on Onions at Peterson Produce," unpublished article. 

Peterson, Jean, and Sterner, Al, (1998). "Propane Flame Weeding Vegetable Crops," pp. 79-81, Greenbook '98, Energy and Sustainable Agriculture Program, Minnesota Department of Agriculture, St. Paul, MN, August 1998.

Sterner, Al, personal email message to Fred Cox, 18 June 2001. 

Photos courtesy of Jean Peterson and Al Sterner.
 

 Contact Information