Organic production methods for Belgian endive: Boosting small organic farmers’ winter income

OBTAINING SEED

            We planned to test two varieties of Belgian endive.  Our search for seeds took well over two months.  We learned that the folks at small seed companies were wonderful and all extremely helpful. 

Several companies, including High Mowing and Johnny’s Selected Seeds offer Totem F1, but trying to locate another hybrid variety was challenging and time-consuming.  Kristin Dubard at Johnny’s did everything she could to help, including giving us a list of seed companies that might carry one, and we searched the catalogues of a number of seed companies we found on-line.  We were able to obtain Witloof Bruxelles (Cicoria Witloff de bruxelles) from Seeds from Italy, an open pollinated variety we tried in the past with only modest success.  Dam Seed Company in Ontario carried “Flash,” one of the varieties recommended in our literature search but unfortunately we learned Canadian seed companies are prohibited from shipping to the U.S.   Dan Nagengast at “Seeds from Italy: in Lawrence KS informed us of a radicchio variety “Triviso” grown just like Belgian Endive which we included in our trials.  Martin Walsh at Baker Creek seeds was kind enough to request his purchaser to contact the Dutch grower (PeterPik) who supplied “Flash” to the Canadian seed company.  We contact PeterPik in the Netherlands who in turn referred to their supplier.  Eventually we contacted a Netherland’s producer who on their own kind initiative provided us with 100,000 seeds of the variety they recommended, “Atlas”, free of charge.  As of this writing (December 2018), Totem is still the only readily available hybrid variety in the U.S. for the small scale grower.          

 GROWING THE ENDIVE ROOT

            We found growing the endive root to be simple and easy.  We prepared the field as normal and fertilized according to a soil test (we used Kinsey Ag lab in Charleston, Missouri).  We had read in many places to avoid excess nitrogen, as this would encourage excess leaf growth at the expense of root development; however, we used 400 pounds/ acre of 13% protein meal in both 2016 and 2017 and did not experience excess leaf growth at all.  It should be pointed out that this protein meal—containing feather meal, blood meal, and meatmeal—is a quick release fertilizer and by the time the roots really began to put on girth in late Summer, it will have mostly have disappeared.

            We planted on 30” raised beds made with an even-flo raised bed plastic layer, without the plastic.  This created a raised be about 5 inches high.  We then watered the beds, waited for the first flush of weeds to sprout (7 to 10 days) and then flame weeded the beds.  Only then did we plant the endive seed.  This is important because endive seed can germinate very quickly and can emerge before all but the most precocious weeds.  The O.P. variety we planted had emerged from the ground in 2-3 days, while the Atlas took 5 and the totem 7.  We seeded with a Jang precision seeder, xy24 seed cartridge, front sprocket 11 rear sprocket 11.  This proved to be much too thick, and we subsequently spent a lot of time thinning the crop.    In 2017 we switched to a xy12 cartridge, effectively halving the seeding rate.  This still proved too thick for the Atlas, but the Totem did not germinate as well and so did not require thinning in 2017.   We seeded two rows in two of the beds, and three in three beds.  Before seeding in 2017, we did a germination test on seed held over from 2016.  Atlas seed held over from 2016 and planted in 2017 suffered no loss in viability, while totem seed held over from 2016 did not germinate reliably in 2017. 

            The endive seed was planted in 2016 on June 4^th and June 7^th.  This gave us plenty of time for the roots to grow to full size before the end of the growing season.  Good weed control is essential in Belgian endive production as the plant does not compete well against weeds.  In addition to pre emergent flame weeding, each bed required hoeing once (about two weeks after planting) and two moderate hand weedings (about five weeks and two months after planting).  After that, the endive foliage formed a dense enough canopy that no new weed seedlings emerged.  Due to the seeding rate being too thick in 2016, we spent a lot of time hand thinning at the time of the first hand weeding in order to attain the 4 to 6 inch spacing between each plant required for full sized roots.   We also thinned part of one bed to only a three inch spacing, just to test whether the literature was correct in asserting that a minimum of four inches in necessary for full sized root development.  Indeed, the literature was correct.  Total time spent in 2016 hoeing, thinning and hand weeding the crop (five beds, at 200 feet long, three rows in 3 beds and 2 rows in two beds) was 26 hours!  Ouch!

Endive grows slowly at first and the roots develop girth slowly.

HARVESTING THE ENDIVE ROOT AND YIELDS OF ROOTS

We harvested the roots by undercutting them with a bar attached to the tractor’s three point hitch for this purpose (the same implement we use to harvest carrots, parsnips, and other similar crops common on small vegetable farms).  Roots were then pulled from the ground, the tops trimmed to within 1 inch of the root, and placed in plastic lined crates or boxes for storage.  This process took some time—approximately 5 to 8 person- hours per 200 foot bed, depending on who was working.  Endive root holds tight to the soil and, even after undercutting, requires some effort to pull out.  Trimming the tops (we used handheld bypass pruners) seemed to take the most time.  In 2017, we tried mowing off the tops both with a standard rotary lawn mower and with a high wheeled model (Billy Goat).  Both machines ruined too many roots, and it became clear the machines’ time savings were not worth replacing hand trimming. Also, many of the tops were trimmed incompletely or inadequately, necessitating further hand labor.   Part of the reason so many roots were damaged was due to the unevenness of the beds and the fact that neither mower was able to fully straddle a bed.  Given a different planting system, (perhaps planting on narrow ridges or, given a powerful enough digger, simply planting in flat ground?) mechanical mowing might be a better option.  We did, however, significantly reduce harvest time by trimming the tops with a hedge trimmer before the roots were dug.  While this was a little hard on the back, it only took 20 to thirty five minutes, depending on who was doing it, to trim a two hundred foot bed (two rows in the bed).

Yields of roots per bed varied.  The lowest yielding bed (200 feet long) produced 160 pounds of roots, and the most we got from a 220 foot bed was 550 pounds of roots.  Why such variability?  Mostly it had to do with the spacing of the plants.  Too close together, and the roots do not grow big enough to produce a marketable chicon, too far apart, and the root grows extra large to no purpose.  As is covered in greater detail under “Yield of Chicons”, a jumbo sized root does not necessarily correspond to a jumbo sized chicon!  Therefore, to measure success merely by weight of roots harvested is misleading. 

Based on our experiences is the forcing chamber, the ideal diameter of roots is 1 ½ inches at its point of greatest girth.  We found that to achieve this, a spacing of 4 inches within the row is ideal.  Less than this seriously reduces the number of usable roots, and greater than 4 inches causes the roots to grow large to no purpose, thus reducing the yield of marketable chicons (but not weight of roots).   However, it is better to have the roots spaced further apart than too tight:  5 or 6 inch spacing within the row is perfectly acceptable.

We experimented with both 2 rows per 30” raised bed and 3 rows per bed.  The extra row created extra work in weeding and thinning, cost extra seed, and did not yield more usable roots.  Thus, two rows per bed are recommended.  Growers not using raised beds will want to space the rows a minimum 18 inches apart.  

We learned that it is extremely important to harvest the roots on time! (See section “Yield of Chicons”).  In 2016 we harvested from 140 to 145 days after planting, slightly later than the seed producer’s recommended days to maturity, with only a few roots showing signs of overage or bolting.  In 2017 harvest was delayed until mid November due to labor shortages.  As will be discussed later, this extra delay in harvesting led to many over mature roots and significant losses in the forcing chamber.

2016: We monitored moisture closely and irrigated when necessary.  Growth looked good, but the roots really began to size up from about mid-September to mid-October, a little later than we anticipated.  We began harvesting on Oct. 24^th, and finished Nov. 5^th.  Harvesting took place between 140 and 154 days from seeding, a little longer than the 140 day maximum time we said in our proposal.  A bar digger pulled by the tractor went under the raised beds to facilitate harvest.  This was very successful.  Almost no roots were broken, and harvesting was easy.  We discarded roots too small for forcing (i.e. less than the 1.25 inches diameter minimum required), trimmed the tops to between a half inch and an inch from the crown, packed the roots in plastic bags in wooden crates, and stored the crates in the walk-in cooler.  We harvested approximately 80 boxes of roots suitable for forcing, each weighing approximately 18 pounds each for a total of 1,440 pounds.  Our grant narrative spoke in terms of number of roots (not pounds) with a goal of having 2400 roots. 

CONSTRUCTING THE FORCING CHAMBER

             The two crucial factors in a forcing chamber are complete darkness and temperature control; therefore, we constructed the chamber with those in mind.  We built the forcing chamber in a corner of our packing room.  This is inside an unheated barn, originally built as a confinement veal facility.  The floor is concrete and already had a drain.  We framed out an area 10 feet by 19 feet and covered the inside walls with metal siding.  We chose metal in order to prevent rodents from gaining entry to the forcing chamber.  Fiberglass insulation was put in between the wall studs, then the entire outside walls were covered with an 8 mil sheet of black plastic, in order to exclude light from the chamber.  Any gaps in the exposed framing were filled with expanding foam insulation, and the plastic edge was taped to the framing top plate, also to exclude light.  The ceiling was formed by placing rigid polystyrene panels over the barn attic truss bottoms, then fiberglass insulation over that.  Underneath the polystyrene, the entire ceiling was again covered with 8 mil black plastic to exclude light.  The door opening was made 4 feet wide and the “door” consisted of a sheet of 3 inch polystyrene, wedged against the frame with a heavy object.  Probably a real door would be better, and the farmer intends to install one later.  This basic construction was effective in blocking all light from entering the forcing chamber.  We also insulated the floor before commencing forcing in 2017.  For our needs, this chamber was much too big.  Even at peak production, we did not require more than two  towers of forcing trays measuring roughly 3 ½ feet by 3 ½ feet by five feet high, usually we needed only one such tower.  Add the water reservoir and piping (see below) and space to work in and our chamber was still about twice as large as it needed to be.  This made constructing and heating the chamber more difficult and expensive than it should been.  In year two we installed a curtain halfway across the chamber, this helped somewhat to reduce heating costs and provide an even temperature for forcing.

FORCING THE ROOTS HYDROPONICALLY

             Endive must be stored for a minimum of two to three weeks before it is ready to force.  Forcing must be performed in absolute darkness; otherwise the chicons will be green, hairy and unpleasantly bitter.  Our measures for eliminating light (see section on constructing the forcing chamber, above) were successful and the chicons grew up pale yellow and smooth, without a hint of green. 

            The roots were packed tightly into three foot by three foot hydroponic flood trays.  These were nestled inside stackable wooden forms, with 12 inches headspace between the top of the tray and the bottom of the next one.   This gave ample headspace for packing the trays, growth of chicons, and harvest.  We stacked the forms in towers of three.

            We used a 15 gallon tub for a water reservoir.  As the growing chicons take up a lot of water, a smaller reservoir would be inadvisable.  When replenishing the water supply, we added wrist temperature water rather than using unheated well water, to avoid uneven temperatures on the roots and stressing the heating system.     Into the reservoir we placed a simple aquarium pump, the outlet of which led to a PVC pipe which emptied out into the southeast corner of the top tray.  We then drilled a hole in the northwest corner of the top tray, just below the lip, and inserted a vinyl tube in it.  This fed into the northwest corner of the middle tray.  We then drilled a hole in the southeast corner of the middle tray and inserted a tube in in which led to the corner of the bottom tray, and from the bottom tray back to the reservoir.  Thus, water was continuously circulating across the trays.  All the materials needed for this setup are readily available at any hydroponics supply store.

            The temperature of the forcing chamber is supposed to be kept between 58 and 64 degrees Fahrenheit.  We used two portable electric space heaters to heat the chamber.  However, we discovered that this was insufficient to maintain the temperature during very cold (below 15 degrees) weather during the first year.  We had fewer problems maintaining the temperature the second year after we insulated the floor and installed the thermal curtain, but still had issues when there was even a momentary power interruption, causing the heaters to shut off, or when they did so out of mere capriciousness.  The main effect of lowered temperature was slow growth of the chicons.  Once temperature was brought back up, growth resumed at the normal rate, and we observed no loss in quality compared to those chicons which enjoyed adequate temperatures for their entire growth.

            Much of the literature on Belgian endive production recommends keeping the water temperature five degrees warmer than the air temperature.  We therefore placed a 100, later a 200, watt aquarium heater in the water reservoir.  However, we discovered that this was insufficient to raise the water temperature even slightly above the air temperature.  It also cost us a lot in electricity and succeeded in fusing the extension cord to the heater plug, indicating a safety hazard.  After we abandoned the attempt to heat the water, no differences in chicon quality or speed of forcing were observed.  

            Ideally, the roots spend 21 days in the forcing chamber before the chicons are harvested.  Our experiences in 2016- 17 did not bear this out, with times of 4 weeks to a month being the norm.  Only by the end of the forcing season in 2018 did we achieve a 21 day time frame.  Here it must be noted that, due to warmer weather, the temperature in the forcing chamber was higher than recommended, ranging from 62 to 70 degrees Fahrenheit.   Surprisingly, the differences in chicon quality and yield at these temperatures were minor.       In 2016-17, we experimented with adding OMRI approved fertilizer to the circulating water.  We added 1 oz Neptune’s harvest liquid fish/ gallon of water and 10 fl/ gallon of Ful-power, a fluvic acid product.  We saw no increase in size of chicon/ size of root, nor any increase in overall yield.  We know that large scale producers do add fertilizer to their hydroponic solution, so more research in this area would be beneficial.

Conclusion: Forcing the roots was a much more forgiving process than we anticipated.  That is to say, given good quality roots, sufficient water, and complete darkness, high quality chicons resulted even if the temperature in the forcing chamber was inconsistent.  No added fertilizers to the forcing medium were needed.

SOIL FORCING

            Traditional endive production is done in soil, not hydroponically.  The available literature is very sparse when it comes to detailed descriptions of this process, but from what we could tell from the YouTube video we watched on the topic, the roots are placed in a large underground outdoor pit, then French people shovel sand on them.  Presumably the climate is such in Northern France and Flanders that this creates a warm enough environment for forcing, and the sand provides both top pressure and darkness.  As we do not have these conditions in Northeastern North America, we tried soil forcing the roots in boxes in the same forcing chamber we used for the hydroponic forcing.  The boxes (13”x17” by 12” deep) were filled with organic potting mix and then the roots were inserted.  Yields were comparable to hydroponically forced roots (8 pounds of chicons from 24 pounds of roots for Totem, 10 pounds of chicons from 26 pounds of roots for Atlas).  However, it was hard to keep the soil off the chicons and trimming took longer.  Also, it took time to cart in the soil and filling the boxes took longer.  I do not recommend this method of forcing.