Thursday, June 18, 2015

Blog Moved!

In order to expand this blog with more topics, and keep it up to date we have moved this blog HERE.

Friday, November 7, 2014

Nutrient Solution Management in Recirculating Systems

Nutrie Solution Management in Recirculating Systems
Summer bibb lettuce yield and quality as impacted by nutrient tank management techniques
By: Natalie Bumgarner, PHD


In recirculating systems producing leafy crops, one of the main factors in the control of the grower is the frequency of tank changes (ie pumping out and turning over the recirculating solution). Since we generally manage the solution based on EC, we are assessing the total amount of solutes in the water. We don’t know the balance of each nutrient individually, so tank changes are carried out to try and maintain necessary levels of nutrients. Essentially, our goal is to change the nutrient solution often enough that we don’t have detrimental buildups of unused ions or depletions of important nutrients. Also keep in mind that different water sources have different background ion levels that can slow or speed up imbalances in the nutrient solution. 

This tank change practice is really based on cost efficiency. There are nutrient solution management systems that have the ability to manage based on individual ions, but these are much higher cost than the typical systems installed in small to mid scale greenhouses. This system cost savings comes at the price of more frequent tank changes that maintain a safe margin of nutrients in solution.

Management methods

Without installing a new nutrient management system, there are methods to try and extend time between reservoir changes. One common method is the addition of small amounts of nutrient concentrate that address the most commonly depleted nutrient ions. Typically, these additions focus on N, K, and P. In this trial, the daily additions were intended to provide 10 ppm N and P, and 15 ppm K. An ion that can increase over time in some solutions is Ca (from source water and calcium nitrate additions). So, potassium nitrate and monopotassium phosphate were the fertilizer materials added to avoid adding Ca. Additionally, some treatments were formulated with a slightly lower initial Ca target to see if this could limit Ca increases in solution over time. The goal of this nutrient tank change test was to investigate how additions of nutrients compared to tank changes in these normal and low calcium nutrient formulations. This trial was carried out in our experimental system where we have the ability to provide four different nutrient solutions. Four treatments were based on these solutions and goals. 1) Normal solution with one tank change at midway point in production (8/25), 2) Normal solution with no tank change but daily additions of KNO3 and KH2PO4 after tank change was carried out in other treatments, 3) Lower calcium solution with one tank change during production (8/25), 4) Lower calcium solution with no tank changes but daily additions of KNO3 and KH2PO4 after tank change was carried out in other treatments.

Plant Management

Seeding was done by hand into pre-moistened 1” x 1” x 1 ½” cubes. Seeds were germinated in 9” nursery channels that were receiving a continuous flow of nutrient solutions set at experimental levels. After 13 days, seedlings were transplanted to the production NFT channels at a spacing of 8” on centers. After transplanting, plants were grown in 4 ¾” channels until harvest. The nutrient solution was automatically and continually adjusted to maintain a target pH of 6.0. Electrical conductivity was maintained at 1.8 for all treatments through the entire experiment from seeding onward. These trials were carried out in a system designed to pull from four different 40 gallon nutrient tanks so that differing solution treatments discussed above could be tested in a randomized block design. At harvest, shoot fresh weight and a visual rating of tipburn was recorded individually for each head.

Timing and Conditions

  • No supplemental lighting of any kind was used and there was no shade cloth on the greenhouse in the summer of 2014. 
  • Environmental data represent only the first 24 days of the experiment.
Biomass Yield by Treatment

Nutrient Solution Data

Final Nutrient Solution Data

Discussion on the trial

There are some interesting points to bring out looking at the data presented on both plant and solution variables.

1)  In this trial, there were not significant differences in final fresh weight yield in the four treatments as is clear by overlapping confidence intervals. Likewise, there were no distinct differences in tipburn ratings (intervals not shown). There was quite a bit of variability in the tipburn rating among treatments, and more work will be needed to see if less frequent tank changes may increase the risk of tipburn. While covering these results, I want to make sure to point out that this is a preliminary study and additional work should follow. It is intended to lay the foundation rather than provide a prescription for current changes to nutrient solutions. This test was designed to replicate as closely as possible the starting point of solution conditions that would be present in the systems of many of the growers. More advanced questions and detailed work would follow to develop recommendations about practical steps that growers could take to prolong the distance between refreshing the solution in you nutrient tank. Keep in mind that individual conditions in source water will have a strong impact of whether trends observed in this work would or could be replicated elsewhere.

2)  In the area of essential nutrients, we see that some of our original assumptions were not necessarily correct. Reviewing the solution data, we can see that calcium levels actually did not increase to undesirable levels over the course of the experiment. In fact, these solution values indicate that Ca in the normal recipe was probably more in line than in the low Ca solution. Mg was present in fairly consistent and generally adequate levels throughout the trial. By the end of the trial, N was present in all solutions, but certainly not in excessive amounts. K and P were present is quite sufficient amounts. With K and P, it is likely that adjustments could even be made to our daily addition recipe to optimize these levels. P could be lower in treatments 3 and 4. Probably the most detrimental attributes of the solution that are revealed by these samples was the increase in S and Na content by the end of the trial. One key way that several of these issues could be addressed is by the use of a nitric acid solution instead of a sulfuric acid solution to lower the pH. This would increase the level of N without needing to add K (as is necessary in the addition of KNO3) and also reduce the buildup of S in solution. The increasing levels of Na in solution were caused by source water conditions and are much harder to change. More testing will be necessary to determine of the increased Na levels over time were detrimental to lettuce quality. Finally, it is important to note that lower levels of some micronutrients, especially Mn were observed in this study. If tank change intervals are to be increased, then daily additions should likely be including the most highly used micronutrients, and higher initial targets likely needed to be started to prevent deficiencies even under regular tank change intervals.

Tuesday, October 21, 2014

Cucumber Production- An Overview of OH numbers in 2013 and 2014

Cucumber Production- An Overview of OH numbers in 2013 and 2014
By: Natalie Bumgarner, PHD

Overview and Data Considerations

  • This data covers a couple of years of data (2013, 2014). So, the goal is to present to you an overview of crop yields and schedules. 
  • Several of the main cultivars that we carry and that our customers use were trialed. However, these data do not represent all cultivars at all times of year, so comparisons across years are limited, 
  • One of the most important things to keep in mind is that these trials were carried out in relatively small blocks. Our vine crop greenhouse is mostly dedicated to tomato production, so cucumber trials took place in units of 10-20 buckets on the western side of the greenhouse. 
  • In some respects, these small sections of space dedicated to cucumber production are similar to many grower houses, but it is important to remember that yields can vary according to light in different locations within the greenhouse. 
  • These trials were all carried out in Lodi, OH, so we cannot account for location variation that may be seen in other areas or seasons. 
  • All of these trials were run with the plant maintained in an umbrella system. The majority of the crops were produced without pinching and were removed when the main leader reached the floor after traveling up to the wire and back down.
Environmental Overview

*40% white shade cloth was installed July 16th and removed August 20th, so these solar radiation averages reflect the loss of light in the greenhouse due to shading. Additionally, in the warmest and most humid portions of the summer, some condensation on the plastic also reduced incoming radiation.
** Data not available

*No shade cloth was installed in the greenhouse in 2014, so all light as represented above is the full amount allowed through the greenhouse covering.

Spring 2013- Manar Cucumber

Manar cucumbers are Beit Alpha or mini cucumbers. This cultivar tends to produce a mid-sized fruit. We often harvested at 6 to 8 inches in length. Manar has high resistance to powdery mildew, so it is recommended for spring through summer production. It is a generative plant that will produce multiple fruit per node; however, we generally only harvested one fruit per node in this trial. In this crop, plants were grown with 8 ft2 per plant ( one plant per Bato bucket). Fruit were removed on the first 8 -10 nodes and that is one of the reasons for the long period of time from transplant to first pick.

Spring/Summer 2014- Eldora Cucumber

Eldora cucumbers are a long English cultivar. Fruit are generally harvested at 12 to 15 inches in length. Eldora has high resistance to powdery mildew, so it is recommended for spring through summer production. In this crop, plants were grown with 8 ft2 per plant ( one plant per Bato bucket). Fruit were removed on approximately the first 5 nodes, and only one fruit per node was allowed to mature to harvestable size.

Harvest Period in Detail- Mini 2014

*Total harvest weight per picking for the entire crop cycle

A few closing thoughts

This overview has thrown quite a bit of data at you the reader, but hopefully some of it has been useful. I would like to close with a few comments and certainly feel free to email or call with any further questions. 
  • Due to management and environmental factors, these yield numbers are not appropriate for basing production estimates in your operation. 
  • That being said, there are some general trends that can be useful to you in planning. Typically, growers can produce from 3 to 5 (or even more if all mini) crops in the greenhouse per year. It is clear from these numbers, though, that those crops are not likely to have the same number of days. It was common for the fall and winter crops to be in the buckets longer to produce similar yield per plant. 
  • Comparing the smaller cucumber data to the long English data is also useful. It is clear that the plants will often have harvestable fruit at a younger age and provide more harvestable fruit per plant. This more rapid maturation and heavier fruit load (by #) is much of the reason that the cropping cycle is typically shorter for the mini cucumber. 
  • One element not covered in detail in this trial is the fact that training methods can be important. We typically use a version of the umbrella method that allow the plant to grow up and over the support cable and then back down. More intensive training systems and pruning methods (we did not allow fruit to bear on lateral branches) would impact fruit load and yield. Likewise, we used a low plant density to reduce crop management burden and optimize light interception and yield.

Friday, September 12, 2014

Summer 2014 Oakleaf Trial

Summer 2014 Oakleaf Trial
Natalie Bumgarner, PHD

Methods and Management

Seeding was done by hand into pre-moistened 1” x 1” x 1 ½” cubes of three different media (Grodan 200 ct rockwool, Oasis XL 162, and GrowTech 162). Seeds were germinated in clear water in seeding trays, and were transferred to the nursery and nutrient solution 3 to 5 days after seeding. Seedlings were produced in flowing nutrient solution in the nursery for an additional 10 to 12 days before transplanting. Due to the season, no supplemental lighting was provided during the seedling phase. After transplanting, plants were grown out in the channels for 25 to 28 days until harvest. The nutrient solution was continually cycled through the Fertroller where automatic pH and EC adjustments met programmed solution set points. The pH was maintained at 5.8 by the addition of dilute sulfuric acid. EC was maintained at 1.7 by the addition of concentrated fertilizer solution and source water. Tank changes were carried out every two weeks.

Timing and Conditions

Biomass Yield

Media Impacts

Some comments on the trial

Looking over the environmental and yield data, there are some interesting points and items that I would like to pull out and discuss a bit.

  • I haven’t really focused often on the environmental data in previous blogs- mostly saying that through a four run trial, there was quite a bit of environmental variability. This trial, though, I want to point out that over a four month period in northeastern OH, we probably won’t find many data sets with this similar a set of conditions. Carbon dioxide stayed quite consistent and especially across the last three runs, the temperature and light were also quite similar. Humidity did increase under warmer summer conditions as the wet wall ran for longer periods of time to reach cooling set points. Due to similar environmental conditions, we also followed fairly consistent transplant and harvest schedules through the four runs I comment on the environmental conditions because it is clear that some variability still exists in crop growth rates and final yield even with reasonably similar temperature and light conditions (similar for a greenhouse meaning outside a growth chamber, etc.).
  • Having discussed the reasonably similar environments, this leads to some serious questions about why there was so much variability in the weights reported in earlier tables. One key reason is that we experienced some seed quality and germination issues. This was most clear in the Panisse and Oscarde cultivars where germination was inconsistent in the first three runs (we switched to a new seed lot in the 4th run). This led to uneven growth or not even having enough seedlings to provide the 15 plants of each cultivar in each media. A glance at the standard deviations in the previous slides illustrates that these two cultivars were generally more variable than the other three. Germination and plant stand likely had a strong influence on those deviations.
  • The third and final thing I want to talk about is related to the other two points of discussion. The use of three different growing media had some interesting general impacts. I inserted runs 1 and 4 into the above table because they represent the most complete data sets. Overall, for the Kireve, Rouxai, and Rutiali cultivars, peat cubes tended to show the potential for yield increases. These trends were not repeated and were in some cases reversed in the cultivars where germination and plant stand were less consistent (Panisse and Oscarde). Does that suggest that peat performs differently in some cultivars over others? I cannot say that this is not the case, but I can put forth another idea. It may be that in seed lots and cultivars with strong vigor and rapid germination, peat can be an asset to seedling production and plant growth. However, rockwool and/or oasis may be providing better environments for germination and early growth in less high quality seeds. More investigation is certainly needed as there is definitely room for improvement in our seedling production, and further work will help us better understand whether we are dealing with a seed, media, or environmental issue.
Plant Images

Kireve- Run 4

Oscarde- Run 4

Panisse- Run 4

Rutilai- Run 4

Rouxai- Run 4

Monday, August 4, 2014

Branching out with Brassicas - Summer trial in NFT production in Ohio

Branching out with Brassicas
Summer trial in NFT production in Ohio
Natalie Bumgarner, PHD


In the greenhouses that I visit and crops I discuss with growers, it is clear that lettuce still fills a majority of plant spaces in the NFT system. However, we field an increasing number of questions about the many other leafy crop possibilities. Many of the other leafy options are in the Brassica family - cabbage cousins, essentially. These include kale, mustard, mizuna, and pac choi most commonly.

These options present growers a chance to diversify to attract new customers as well as provide more product to existing customers. One of the challenges with any new crop is understanding its production capacity to assist in pricing and tailoring production to anticipated demand. Also, there is the potential for higher light and temperature to negatively impact quality in the summer as is the case for some lettuce and other leafy crops. This trial was designed to evaluate a selection of Brassica crops as well as amaranth (not a Brassica, by the way) under summer conditions. Rather than extensively trialing multiple cultivars of the same crop, the goal was to evaluate production capacity and crop quality of a selection of alternative leafy crop
options to lay a broad foundation for future work.

Methods and Management
Seeding was done by hand into pre-moistened cubes. Three media were compared in this trial- rockwool (25 x 40 mm), Oasis (162 count Horticube XL), and a peat media (162 count, Grow-Tech). Seeds were germinated in clear water in seeding trays, and were transferred to the nursery and nutrient solution 3 to 5 days after seeding. Seedlings were produced in flowing nutrient solution in the nursery for an additional week to two weeks before transplanting (no supplemental lighting was provided during the seedling phase). After transplanting, plants were grown out in the channel until harvest. The nutrient solution was continually cycled through the Fertroller where automatic pH and EC adjustments met programmed solution set points. The pH was maintained at 5.8 by the addition of dilute sulfuric acid. EC was maintained at 1.7 to 1.8 by the addition of concentrated fertilizer solution and source water.

* This trial measured single harvest yields to produce the most accurate and comparable yield totals. However, some growers may harvest single leaves or leaflets from kale or amaranth plants. This could increase the total yield per plant but require additional time in the channel and quite honestly make comparisons much more challenging.

Timing and Conditions

Biomass Yield
* Amaranth was seeded with multiple seeds per cube as is typical in production, but this increased the yield variability.

Some thoughts on the trial

After considering these trials, there are a few things that I would like to bring up for discussion.

  • First, I should report that there were very few quality issues with any of these crops in these trials. They grew through the OH spring and summer conditions quite well with no losses or issues to speak of.
  • Second, and most clear in the data, we can see that the yield potential of these crops is wide ranging. That is really an understatement. In fact, we had to use two different scales to weigh these trials. Besides our scale challenge, there are two key grower impacts. One is the fact that timing of transplanting and harvest really should be varied. The kale maybe could have been grown a bit longer and the WinWin Choi should have been harvested earlier for highest quality. The other important point is that it will be important for any grower going into sales with such crops to do a few trials before setting prices. We can often count on bibb lettuce to finish out at predictable weights at predictable times, so prices can be set and costs calculated simply. When selling and pricing kale, pac choi and the like, be aware that yields and therefore input costs per weight of produce vary. Don’t undersell yourself early in the process of growing a new crop.
  • Thirdly, the impact of our different growing media was not clear in these trials. This is certainly my least favorite point as unclear results frustrate every researcher. In looking at the yield trends between the two runs, the differences (or lack thereof) between rockwool, oasis, and the Grow-Tech cubes were not consistent. In run 1, the peat cubes tended to perform better while that was not seen in run 2. Environmental conditions were reasonably consistent in these two trials, so that is unlikely to be the primary cause for these inconsistencies. I am hesitant to draw too strong a conclusion about this on early trials, but I will say that it may be differences in germination speed, moisture content (and therefore fertility), temperature and the like early in crop growth that led to these variations. We generally deal with pelleted lettuce seed that produces very consistent germination and early growth. Because these Brassica crops are not bred for controlled environment production, their response to small differences in conditions may not be as well understood at this point in time. More focus on the seedling aspect may well be needed.
Notice the difference in germination rate and seedling size between the five crops in this trial. Tiny amaranth seedlings are in the back of the tray.
Plant Images from Run 1

Toscano Kale- At harvest (44 days after seeding)

Red Giant Mustard- 44 days after seeding

Red Choi- 44 days after seeding

WinWin Choi- 44 days after seeding

Red Leaf Amaranth- 44 days after seeding