Quinoa Farming Report
Sam Madge, Berggren Demo Farm
The planting site for the BDF Quinoa experiment was a 1/5 acre portion of a clover field on a flat clayish loam soil type. The farm is approximately halfway between Springfield and Walterville on Camp Creek Rd, at the downstream end of the McKenzie River valley, approximately 500′ elevation.
Most of the seed came from Wild Garden Seed in Philomath, OR, with some also from Adaptive Seeds outside Marcola, OR. The cultivars trialled were “Red Head” (6oz) and “Brightest Brilliant Rainbow” (1oz) from WGS and CO407D(1 gram) from Adaptive. Wild Gardens’ varieties have all been bred from the Colorado 407 Dave, which was developed from a Chilean variety “Linares”.
Preparation & Planting:
Approximately half the land used was worked over once in the fall, and left fallow, uncovered thru the winter which was a relatively dry, cold one. This half was then again tilled in early April at the tail end of a dry spell, and once more on May 2nd, using the same Land Pride 5′ wide rotary tiller, 5 days prior to planting 3oz Redhead variety in 16 (120ft) rows with 18” between row spacing and an average of 2” in row spacing. Seeds were sown using a Earthway garden seed drill, and also by hand, furrowing a 1/4-1/2” deep row. After sowing, 5 lbs of organic bloodmeal powder was scattered over area, some was lost to the breeze. On May 10th 16ct of 128 cell flats were seeded with mostly Redhead, and one flat CO407D.
On May 12th, the other half of the plot which was covered in clover, with some reed canary grass peeking through, was mowed down, and the next day turned over with the rotary tiller. May 16th, this section was tilled once more, and planted the next day, 9 rows of RH and 2 rows of BBR(120ft), with same in row, but 30” between row spacing. About 3 lbs of pellitized organic chicken manure was broadcast over entire 2nd area.
These processes took a combined total of 20 hours work.
Emergence and Vegetative growth cycle:
Emergence of both WGS varieties was relatively poor throughout field, probably due to uneven seed planting depth, but also poor in uniform 3/16” depth of greenhouse flat sowing. The CO407D emerged at about 75% which was noticeably better than Redhead (50%) in the same conditions. The best theory I have to explain this is that Adaptive Seed might have used a gravity (density) and aperture (screen) separator, and Wild Garden might have only winnowed, as the CO407D seed seemed uniformly large whereas the RH had more variation in size. Wild Garden, however, is confident in they’re 80% germ rate in the batch.
Of the direct field-seeded, the first sowing seemed to have better emergence than the second, probably a result of the weather. The first sowing was immediately followed with 2 days of heavy rain, then 2 days of scattered showers, whereas the 2nd saw only some rain the day after, followed by some heat that dried the soil down about 3 inches deep. This is when weekly irrigation began in order to combat the relatively dry month of June. Both sections completed emergence about 10 days after planting.
Once the seedlings in greenhouse had 2 sets of true leaves, on June 4th, they were transplanted out into the field, in a section of the first sown area that had particularly sparse emergence for no discernible reason. They were spaced 1 foot apart each other in 5 rows 36” apart, with 1tbsp pellitized chicken manure mixed around in each hole. The CO407D was mixed in randomly in this section. At this point the first sown direct seeded population was on average twice the size.
By June 12th, at 4-6”, the first sown section seemed large enough to cultivate, which was accomplished by a 10” wide rotary tiller, run twice between each row. Following this, about 10 lbs pellitized chicken manure was applied only near quinoa plants throughout whole field.
A 24” rotary tiller was run through to cultivate the second sown section on June 21st. These processes took a combined 20 hours of work.
Inflorescence, pollination, and seed head development:
By June 30th, inflorescence were visible throughout crop, with an average plant height of 2′. At this point there were a large population of cucumber beetles, flea beetles, and lygus present, having very minimal visual impact on the crop. July 8th, with an average height of 3′, one last pass through the whole field to cultivate with the rotary tillers was taken. At this point pollination had began, and seemingly peaked for about a week mid July, during unusually hot weather(90+hi, 60+lo).
An average irrigation schedule of 1 inch 3x a week was kept from mid june through the end of July. Seed heads began forming mid July and seemed for the mostly formed throughout field, plants reaching an average peak height of 6′ by July 28th. There were issues with plants toppling throughout field, worst in areas with wider spacing. Some of these toppled over plants became moldy when in close contact with ground(especially CO407D). July 30th, irrigation ceased. After this, flea and cucumber beetle populations slowly declined but lygus continued to flourish, and an aphid population blossomed simultaneous to the drying of lower foliage. Maturation took about a month, and started with an evident pigmentation intensity increase in the seed head, and moved down the stalk and into lower foliage. Some damage from lygus sporadically throughout field became apparent during maturation with small dark spots on seed heads, that contained shriveled up “empty” seeds. In a few toppled over plants, this damage led to faster molding throughout head.
10 hours of work was put in during this period of time, mostly in observation.
Harvest, Processing, Yield, Cost of Production:
Harvest began on August 28th in the first sown area where a peak in pigmentation was evident throughout. This consisted of bundling some select few whole plants to be kept for subsequent plantings, and hanging them upside down in a barn to dry. The rest of the crop, seed heads were cut off, and spread out on a tarp in the sun to dry. Sept 10th with the 1st section complete, harvest began in the 2nd, and was fully completed on the 16th.
Once dry, processing began with threshing, the bulk of which was accomplished thru dancing.
Separation of chaff and stem from seed was accomplished thru gravity screen, followed by winnowing.
Several prototype machines were fabricated to automate this procedure. The bulk of the gravity screening was powered by a reciprocating jigsaw. Winnowing was accomplished mostly using a 150 watt 6″ in line duct fan, with a motor rated dimmer switch and some flexible ducting. Improvements to this machine should include a large filter chamber and rigid ducting to minimize debris build up, and a bigger feed hopper with an auger to prevent clogging. At this point, not including an estimated 25% crop loss due to rotting seed heads laying on ground, and crudeness in harvest and processing technique, the total yield weighed in at 280 lbs. This equates to a 1400 lb/ per acre rate(1700 incl. 25% loss), which is above standard yields indicated by other reports of North American growers, but far below some Bolivian reports.
After threshed and winnowed, desaponification was accomplished by washing with cold water in a small concrete mixer, then dried in pillowcases in a clothes drier, and sent through the winnowing machine once more. This resulted in a product no more bitter than most bolivian imports available at retail outlets, but with a more rich and nutty flavor. Improvements to this process should include a larger (55 gal) washing machine that can be also used as a dryer, and/or a dry process friction desaponifier.
Excluding time spent designing and building various machines, harvest and processing took about 40 hours of work. In total, the project took about 90 hours of actual physical labor. Land and equipment rental cost totaled almost $500, fertilizer and fuel cost combined was about $100, and material cost for processing equipment totaled around $300.
Considering labor valued at $20 per hour, the total cost of production was about $2,700. Amortized down, the cost of production was about $9.64/ lb of quinoa. This is far above current wholesale price of bolivian imports. In order for this crop to be viable for farmers in the area, that production cost should be cut almost in half. Some suggestions to help achieve this would be denser plant spacing, using a seed drill, more like wheat or other grains, so it may be harvested by combine. A higher level of mechanization in the processing would cut down on labor costs as well. Of course, as with many crops, the economy of scale is something to be considered. Many acres would need to be put into production to justify use of seed drill, harvest combine, professional seed cleaning machinery, etc.