Final Report for FW09-049
Navajo tea greenthread is a native herb, with an extensive history of use among Native, Hispanic and Anglo residents of the Colorado Plateau as a favorite herbal tea, dye plant and medicinal herb. Though it is commonly wildcrafted, as an alternative herb crop it is unfamiliar among growers of the region. At my farm, however, it has proven to be a high-value crop requiring few agricultural inputs in order to thrive in the harsh, marginal agricultural lands near Gallup, NM. Once established, the herb’s demands are low for fertility, irrigation and protection from extremes of temperature, aridity and wind, resulting in a low-input crop capable of sustainably producing harvests unattainable in the wild.
But how labor-intensive is it? As with any perennial crop, organic weed control is an issue, the default approach being hand-weeding. This project explores the effectiveness of three organic, in-row weed control methods implemented to reduce hand weeding required: the Williams Tool System’s spring tine weeder, flame weeding and the application of organic corn gluten meal.
The spring tine implement used in the trials, the Williams Tool System, resembles a Lely finger weeder (85? bent tines) for in-row cultivation, but with the addition of a tool bar for between-row cultivators and gauge wheels. It requires the use of a tractor moving at relatively high field speed, though horsepower requirements are very low. Flame weeding was done with a backpack flamer moving at a walking pace down the row. The corn gluten meal was applied and mixed either by hand or with a drop spreader and the spring tine implement.
My project resulted in data that compares the amount of time spent implementing each method on sample rows of the crop and time spent on the remaining hand-weeding needed before harvest. The most complete data came from the second season of the 16-month project. The spring tine implement proved as effective as properly applied corn gluten meal in keeping down the weed population and preventing extensive hand-weeding. Both reduced hand-weeding time by as much as 45%. The equipment cost of tractor and implement is much greater than the up-front investments required of the other methods, but its flexibility in timing is likely to be more attractive to the average small-scale grower than corn gluten meal. I plan to rely mostly on the Williams Tool System as my primary weed control method for my greenthread crop in the coming years.
To conduct field trials to compare the effectiveness of three methods of organic weed control in an established greenthread field, including:
(a) flame weeding
(b) corn gluten meal as pre-emergence organic herbicide
(c) mechanical, in-row cultivation using a spring tine weeding implement
The perennial greenthread field was in its second and third season during the trials. Because the density of the original planting was very light, I expected the crop to multiply in the field over the years, allowing the plants to spread by rhizomes and by seed into a 30-inch wide bed to increase yields. For these trials, the row is considered a 12-inch wide area under which a single sub-surface drip line is buried about 4 inches deep.
Four roughly 800-foot rows were selected for trials at the start of the project in 2009 with the help of technical advisor, Charles Martin. These four rows represent approximately 20% of the rows planted with wild seed in 2008. They were chosen for their relative consistency in plant spacing. The four rows were labeled and designated as attached.
Use of the three weed control methods is limited by crop height.
At the time the weed control tools/materials were available this season, the crop was too close to harvest to implement them. For this reason, the trials began after the first harvest.
Set-up time is a consideration when comparing methods, but I have separated it from the per-row labor. Learning to use the unfamiliar methods requires considerable experimentation and time. Once familiar, none of the methods require much set-up time. The amounts of time spent in setting up, adjusting and trying out each method are considered separately from the amount of time each takes per row in comparison of the methods.
Comparing Time for Applying Method with Time Hand-weeding for Clean Harvest
No trial methods were ready to apply for the first harvest of 2009, and no hand-weeding was required preceding the second harvest due to (1) lack of water, which prevented weeds from thriving enough to interfere with cutting the crop, and (2) time of harvest, after the first frost, which withered weeds but left the hardy crop standing. This season did not produce useful data to compare the methods other than the amount of time each application required.
The second season data do show what the trials were intended to explore; that is, to what extent the weed control method reduces the amount of hand-weeding necessary to get a clean harvest.
Before the first harvest of 2010, the control row required 61 minutes of hand-weeding, more than the other rows, which shows that the other methods did make some difference in the 2009 season:
*The spring tine implement took four minutes to apply and left 43 minutes of hand-weeding before harvest, for a total of 47 minutes of labor, or 23% improvement over hand-weeding alone.
*The flame weeder took 13 minutes to apply and left 46 minutes of hand weeding to do before harvest, for a total of 59 minutes, a 3% improvement over hand-weeding alone.
*The corn gluten meal took 19 minutes to apply by hand and left 38 minutes of hand-weeding to do before the harvest, a total of 57 minutes of labor, a 7% improvement over hand-weeding alone.
Before the second harvest, hand-weeding the control row took 42 minutes, considerably longer in combination with any of the trial methods. This is evidence of improvement in the timing of and/or the way the methods were applied:
*The spring tine implement was applied twice during the early stages of crop regrowth after the first harvest. Each time, the implement was pulled both ways, a double pass. This took a total of 13 minutes and left 11 minutes of hand-weeding prior to harvest, for a total of 24 minutes, or a 45% improvement over hand-weeding alone.
*The flame was applied once after the first harvest, taking 11 minutes and leaving 16 minutes of hand-weeding to do before the harvest, a total of 27 minutes and an improvement of 36% over hand-weeding alone.
*Corn gluten meal was applied with a drop spreader and mixed with the spring tine implement, greatly reducing the time needed to apply it, but bringing up the question of whether its effectiveness is due to the raking in by the implement, disrupting weed growth or to the chemistry of the material itself. The spring tine implement does a very good job of incorporating powder into the top inch of the soil. Total application time was only seven minutes with this approach, leaving 17 minutes of hand-weeding prior to harvest, for a total of 24 minutes, a 45% improvement over hand-weeding alone. These results are much better than those from earlier in the season, most likely because a light rain fell (measured as 0.1 inch) the day after application of corn gluten meal, which is within the window recommended for maximum effectiveness.
Overall Analysis of Weed Control Methods
The Williams’ Tool System is an implement that combines blind-cultivating Lely-type spring tines with a tool bar for sweeps, knives, hilling disks, raised-bed guidance cones and gauge wheels to achieve both in-row weed control and between-row cultivation if needed. The long, heavy, spring-metal tines act as a special kind of harrow. They are spaced out across the width of the bed and along the length of the implement with considerable space between each point in the array. The pressure each tine exerts on the soil may be adjusted individually. The angle at which all the tines engage the soil is variable, depending on the height of the frame above the ground, determined by the gauge wheels. The ideal pressure for an established greenthread crop seemed to be the third notch from maximum pressure, and the ideal angle for the tip of the spring tine seemed to be a few degrees from vertical, achieved with the gauge wheels set to hold the frame as low as possible to the soil surface. A more aggressive angle was tried with the spring tines late in 2009, resulting in a regrettable percentage of crop loss as the tine chopped the tap root of some plants and dragged them across the surface.
How the spring tine implement works: The length of the spring tine, which places the vertical tip of the spring tine pressing into the soil far from the point of attachment to the frame, makes the tip bounce and wiggle as it is pulled through the soil at a speed of about six mph. This result is a stirring action on the soil across the entire bed width and to a depth of 1-2 inches. Any seedling with shallow roots seems to be disturbed enough by this stirring motion that it weakens and gradually dies or is pulled out and dragged along the surface. Meanwhile, the tap-rooted, established greenthread plants simply shift out of the way without losing their hold on the deeper soil.
The use of subsurface irrigation also may encourage deep root systems in the perennial crop, making them more resistant to the stirring action of the spring tine weeder. It is not clear if the spring tine implement reduces the vegetative spread of the plant laterally. The lack of rain or irrigation during much of the trial period may be the cause of a lack of observed vegetative propagation throughout the field.
The flame weeder’s success depends on timing and the susceptibility of the weeds to flame. In these trials, the backpack flamer did show good results on weed populations when applied after the first harvest when the greenthread plants are under four inches high and the weeds are newly sprouted from rains and/or irrigation.
Corn Gluten Meal
Proper application involves the use of some kind of rake or harrow to mix the corn gluten meal into the top inch of the soil. The short-term effectiveness of the application may possibly be attributed to the physical movement of the soil to mix it in. However, observation over a 4-6-week period of rain and/or irrigation may demonstrate that the corn gluten meal has a longer-term effect on weed populations than the mixing action alone would account for. The most successful approach was to apply corn gluten meal at rates of about 40 lbs/1000 sf using drop spreader (22-inche wide push model), then mix in the top inch of soil with rake or spring tine implement.
My choice for continued in-row weed control in an established field will be the Williams Tool System because it is fast and flexible. It works when the soil is slightly moist to very dry in drought conditions, and when the weed seedlings are slightly more developed. It leaves a loose soil mulch over the sub-surface drip rows, which helps to retain moisture deep below the surface. Properly set, it does not damage the established crop, and it can reduce weed control time by as much as six hours per acre. It does have a higher equipment cost up front than the other weed control methods, but the long useful life of the implement and tractor make the per-season cost much less than OMRI certified corn gluten meal, which is of limited availability and high price. Flame weeding must be done at a walking pace down each row, and as the row widens into a bed, more flame and fuel will be required. The spring tine weeder covers the entire bed width with each pass without additional fuel cost.
Educational & Outreach Activities
The materials I created for outreach for this project have been sent to the contact list generated at the 2010 Organic Grower’s Conference and others who have inquired about the project through my grower’s blog: www.highdesertfarmers.com. The results are also available for download from the blog, which has had continuous updates throughout the two seasons, showing the activities of the project. I am scheduled to present at a workshop at New Mexico State University’s Southwest Medicinal Herbs Production and Marketing Seminar, December 9, 2010 at San Juan Community College in Farmington. This event is geared toward Native and Hispanic herb growers and practitioners. I have been asked by Charles Martin, organizer of the event, to emphasize cultivation over wild collection and discuss entrepreneurship, online marketing and financial management/financial risk management. In a recent email, Martin provided the following encouragement and guidance for this presentation:
“The idea is to present cota as a model/example of a native herb brought under cultivation for marketing as a value-added product by an individual or small collective of growers. What you have done is successfully brought all the pieces together singlehandedly, so it is a good example of what other small-scale growers can do, and in a way that can complement an Indo-Hispano approach.”
Attached is a brochure that has been emailed to 17 grower contacts from the 2010 Organic Grower’s Conference and that also will be emailed to any contacts generated from the seminar in Farmington next month.
Three potential growers in the region visited my farm in 2010 to see how I produce organic greenthread. Two men, Tom and Jeff, who inherited land at Zuni Pueblo, though currently living in San Francisco, visited the farm July 18 before planning their own field of “Zuni tea” on tribal land near Dowa Yalaanne (Corn Mountain) at the heart of Zuni land. They were very interested in small-farm scale production methods I had been working on and eager to hear the results of the trials. I have been in closer touch with Colleen, from Hesperus, CO, who visited the farm this summer. This fall she disked an acre of land for spring sowing to start a dry-land greenthread field. Although she is a new grower on a new piece of ground, she has the advice of a local wheat farmer with 30 years of dry-land growing experience. Steve and Ann, from northeast Arizona, visited the farm in August and commented that they felt inspired to convert a portion of their land to greenthread production. Taking advantage of late summer rains, they sowed a small dry-land field with uncertain results the first season.
McKinley County, NM, has very few farms, and the size of farms is expanding each decade as the number of farmers dwindles. Economics and access to water have been the determining factors in this trend. It is possible that existing farms will begin to produce greenthread as an alternative crop, but considering that these growers have good sources for irrigation water and can grow standard crops fairly easily, it is more likely that greenthread will be seen as an option for new farms in places with marginal crop land and poor water sources.
Reactions from producers
Besides sharing my project with those who have visited my farm with interest in growing greenthread, in February 2010 I presented at a workshop at the New Mexico Organic Grower’s Conference in Albuquerque. My slide show documented evidence of greenthread as an appropriate alternative organic crop for the Colorado Plateau, and it was both well attended and well received, resulting in a list of email 37 contacts, 17 of which noted they were interested in growing the herb.
Greenthread is an appropriate, high-value, alternative crop for growers on the Colorado Plateau, many of whom are yet to be introduced to it. When growers are introduced to the crop, it will be more appealing if presented in conjunction with effective organic cultivation options to reduce the amount of hand-weeding required to bring in a clean, weed-free harvest. Growers may refer to these trials in deciding which methods fit their existing operation.
My fields of perennial greenthread produced over 900 lbs dry on 1.5 acres in the third season. This seems to be just a starting point for these plants, which have multiplied and filled out the field in 2010. It is reasonable to expect next year’s harvest to double. I have sold my harvest at an average of $15/lb, which could result in $9,000/acre income from the third season’s harvest, more in subsequent harvests. Another organic grower with even a small field of marginal agricultural land and limited water supply who puts in a field of greenthread may benefit from the herb and the trials of weed control methods in this project. It is also possible that rural residents of the region, including those living in isolated communities on Indian reservations, could put into sustainable agricultural production small parcels of range land, producing and cooperatively marketing their greenthread crop. This would create another sustainable income source for rural residents, contributing to the stabilization of rural communities.
There seems to be very little information available online or in print about organic weed control in perennial crops like greenthread, which do not thrive under plastic mulch. The use of blind cultivation, which I found effective with greenthread, is common among organic field crops. Other common blind cultivators that would be worth testing include various spring or spike-toothed harrows, rotary hoes and even the Lilliston rolling cultivator. This project did not explore methods of weed control on first season greenthread plants, which are very slow to establish and often lose in competition with fast growing annual weeds. In-row cultivation of crop seedlings is a highly skilled organic practice carried out by very few farmers these days, but some form of in-row cultivation may be appropriate for a first season field. Planting on raised beds and allowing the cultivators to steer on guidance cones at the sides of the bed may be one approach. Also, sterile seed beds, which require extra planning and time to begin, may successfully reduce hand-weeding needed in a first season crop of greenthread. Use of a front or belly-mounted cultivator may assist a grower in cultivation around young plants. Perennial grasses are one type of weed that seems to interfere with crop yield in a greenthread field, and for which none of the weed control methods tested here are helpful. Organic control of western wheat grass, especially, is a topic of interest to me at this time.