- By Lois Braun
I've been meaning to write to report on hazelnut harvest for over a month, but once harvest began life got really crazy for me. We started harvesting on August 14th this year, starting with our wild American selections, which tend to ripen first. That was earlier than we normally start. I had heard that small grain harvest was 10 days earlier than usual, due to our earlier than usual spring and warmer than usual summer. We don't know how weather patterns affect hazelnut ripening, so we just decided to pay extra attention this year, especially since we had some research trials for which we could not afford to lose nuts to squirrels and other nut thieves. You would think that after working with hazelnuts for 17 years I'd finally have a better sense of when to harvest them, but I haven't.
It turns out that we probably jumped the gun a bit, judging by how difficult it has been to get some of the nuts out of their husks. The way you tell if a nut is ready is by whether or not it "turns" in the husk. That is, can you pry it out of the husk cleanly, which becomes possible when the plant lays down an "abscission layer" between the shell and the husk. The problem is that not all the nuts on a given plant ripen at the same time, and the squirrels and mice (and blue jays and crows and turkeys and raccoons and what not) are in the field all the time whereas we are not. That was especially a concern at those of our research fields that are far away, such as at Lamberton where we did our controlled crosses this year. So we sometimes pick some of the nut clusters early just to avoid losing any of them.
Ultimately, our harvest season went on longer than usual this year--we even picked a few more late selections yesterday, Sept. 22! Come to think of it, in past years I have noticed that an early spring usually correlates with a more drawn-out harvest, whereas a late spring correlates with a more concentrated harvest. It is as if with an early spring the plants know they have more time, and take it.
Our early and drawn-out harvest was in contrast with the commercial growers, who didn’t start until the beginning of September and who were mostly finished by the 16th. Not only can they afford a little loss, but they need to wait until the nut clusters, not just the nuts within the clusters, are starting to abscise from the stems, which makes them a lot easier to pick. They also need to wait until the majority of plants within their rows are ready, whereas we researchers, picking by hand, pick each plant as it is ready. This is one of the big reasons why getting clonal varieties will be a big help: when we can plant entire rows of the same variety we’ll be able to machine harvest them all at the optimal time, which will reduce a lot of the nut loss.
In spite of the amount of work involved, harvest is my favorite part of the whole annual cycle. The work is outdoors at a time of year when the weather is usually beautiful, though sometimes on the hot side (I like hot weather); it is physical but not strenuous; and you can either listen to the birds or chat with fellow pickers while you do it.
I also like harvest because that is when I get to see the results of all our past hard work. This year was especially gratifying for three reasons.
First, some of the selections that we identified as being superior in past years, the ones we are attempting to micropropagate so we can pass on to you, continued to look very good this year. I have been a little more hesitant than other members of our research team to tout their worth, but with each passing year I get a little more confident of their value. We now have a few solid rows of some of our best selections, and nothing affirms their worth as much as seeing an entire row of good-looking hazelnuts!
As an aside, I’ll pass on some good news, though I'm not sure that I should spread this because I don't want to jinx it: one of our top selections has finally made it out of being coddled in the micropropagation lab into a nursery bed outdoors! Assuming all goes well there, these plants should become available to growers by next fall, in large numbers, with other selections possibly to follow. If the micropropagation lab can figure out how to do one selection, they ought to be able to figure out another!
Second, this was the year in which the plants grown from crosses I made in 2014 finally started to bear nuts, and some of those nuts are notably bigger than most other Midwest hybrids. These are what we’ve been calling our “2nd generation”, with the 1st generation being those I mentioned above. I had been despairing about the progeny from my controlled crosses, because none of the crosses I made in 2012 or 2013 look very desirable. However, not only do the 2014 progeny have larger nuts, but some of them are high yielding too. Often there is a trade-off between nut size and yield, so I have been very reassured to see that some of these plants may capture both.
What makes this work especially interesting is that we have started to observe some clear family patterns. It's gratifying to spot a plant in the row and say "I bet I know what the parents of that plant are", look at my record and see that I'm right! Part of what plant breeding takes is simply learning to recognize the faces in the neighborhood.
It's still too early to know if these plants will continue to bear well. Given that we have 10,000 such seedlings in the field, the probability that several of them will combine many of the traits we're looking for and over multiple years, is good. We like to evaluate them for at least three years before selecting them for replicated trials, after which it will take about another seven years to propagate them and grow them out to see which perform well in multiple locations. By about 2030 we should know which ones are worth micropropagating to send out to growers.
Third, this year we harvested a lot of seed from the controlled pollinations that I reported on in my last blog post. I estimate we harvested about 10,000 seeds, so assuming we get 50% germination, a year from now we'll be very busy transplanting. (About half of those will go to the University of Wisconsin, where a post-doctoral student will be studying their genomics in order to develop better breeding strategies.) Then, assuming we chose the right parents to cross, in about 2026 we should be at about the same stage with this 3rd generation of improved Midwestern hazelnuts as we are now with our 2nd generation. That means they ought to be available to growers in about 2036-17 years from when their seed was pollinated! Hazelnut breeding is a slow process, but fortunately, we don’t have to wait until 2036 for the first results to be available. Those might be coming just next year!
female hazelnut flowers are so small that they are barely noticeable in nature, when they are protected inside pollination bags their dark pink stigmas grow several millimeters long and are beautiful to look at. The moment when we take off a pollination bag and reveal a bundle of stems with a flower at each node can be quite thrilling.
The final step was to return a couple of weeks later to remove the bags—and to be sure that all the stems were properly marked as to what kind of pollen was used for each so that when we return to harvest the seed in August we’ll know who is daddy for each one. I like to color code things, so this made for a very colorful field. (We can’t afford to lose these precious seeds to hungry mice, hence the mousetrap.)
This is a long-term project. We will not know whether or not these crosses were successful until late June, and we won’t harvest them until late August. Then we’ll stratify (cold treat) them until March, plant them in the greenhouse, up-pot them in May, keep them an outdoor nursery through the summer, and finally plant them to the field in September 2021. We might harvest a few nuts in August 2025, but won’t know for sure whether they are any good until about 2028. Most won’t be, but those that are might be much better than what we already have. We will propagate those for replicated trials in 2029 and, depending on how they do in those trials, release them to you in… maybe 2037? So don’t hold your breath waiting for them. I expect to be retired by then, but I hope to still be around to enjoy eating them! In spite of the wait, the idea is exciting to me.
Best and be safe, Lois
It is sobering to think that efforts to develop a hazelnut industry in the eastern US go back to the early 1900s and here we are today, 119 years later, still without much of an industry. Why? The main reason is the biology of hazelnuts doesn't operate on the same time scale as we humans. Hazelnuts are found wild throughout the Upper Midwest, but unlike other woody crops, no one has found the perfect plant growing in the wild. If someone had found such a plant we'd long ago had an industry. Instead, we've had to breed a plant that can survive the winter and the native hazelnut pest, eastern filbert blight, while also producing enough nuts of sufficient size to make it worth harvesting and cracking. And therein has been the challenge. It simply takes a long, long time to breed a proven cultivar of a woody plant - 21 years to be exact. Twenty-one years is a long time in human history and many things conspire to get in the way of hazelnut breeding: kids, storms, illness, divorce, budget cuts, old age, you name it.
In 2007, we launched the Upper Midwest Hazelnut Development Initiative with the primary goal to get years of work by private and public hazelnut breeders across the finish line. The easy thing for amateur and professional breeders to do is make crosses between two good plants and grow out the seed. The hard thing is to stick around long enough to evaluate each of those offspring for 7-10 years, find the best one or two, vegetatively propagate them, and evaluate them at multiple locations for another 10 years. Only then do you have a proven cultivar. So, starting in 2007 we set to work finding the best of the offspring and in 2009 we established the first of our replicated performance trials.
At the 1st Annual Upper Midwest Hazelnut Growers Conference in 2010 in LaCrosse, we outlined the two main challenges we needed to address to launch the industry. First, we needed proven hazelnut cultivars. Second, we needed harvesting and post-harvest processing capacity to allow early-adopter growers to get product to market.
In the blink of an eye, here we are ten years later. The 10th Annual Upper Midwest Hazelnut Growers Conference will be held this March 8-9 in Eau Claire, WI and it's going to be an exciting conference....a celebration really. We will formally introduce the top 12 selections from the performance trials and present information on how to grow, harvest, and process hazelnuts. For early adopters that have been at every conference since 2010 this will be a chance to celebrate the work that we've all done. For those that have been lurking and waiting for the right time to get involved with hazelnuts, now is the time. Take a look at the conference brochure and register today!
Nothing happens very quickly in the world of woody crop breeding. It's a long slow slog that takes patience and perseverance, which is what makes weeks like last week so enjoyable. Both myself in Wisconsin and my colleague, Lois Braun, in Minnesota spent pretty much all of last week planting hazelnuts and not just run of the mill hazelnuts. We were planting clonal layers from our 1st Generation selections in replicated field trials. After more than a decade working on these selections we were finally able to plant larger-scale field trials. Shown above is one of six Joint Performance Trials (JPT) established in the Upper Midwest in 2017. This trial at the West Madison Agricultural Research Station in Verona, WI now includes five selections from the Grimo breeding program (Marion, Frank, Aldara, Andrew, Northern Blais), two selections from the Consortium program (OSU 541.147, Gran Traverse), and 9 of our UMHDI 1st Generation selections (Arb 7-1, Shep Rosy, Cuddy 2-28, Heas B, Arb 4-2, Rose 9-2, Price W41, SPC-2D5, and Price W41). Once available, our plan is to add the rest of our selections plus some additional selections from Grimo, the Consortium, and Rutgers in 2019.
These six Joint Performance Trials in the Upper Midwest (Verona, WI, Bayfield, WI, St. Paul, MN, Staples, MN, Fenton, IA, Cedar Rapids, IA) are intended to compare the very best germplasm being developed for the colder regions of the Eastern US and Canada to help growers make decisions when purchasing plant material in the future. We are excited to announce a new location being planned for 2019 in Saint Camille, Quebec!
Contributed by Jason Fischbach
Anyone that has hand-harvested hazelnuts, or worse yet, paid someone to hand-harvest hazelnuts knows it is not a viable option for commercial nut production. Even the fastest pickers are too slow and too expensive. In Oregon, hazelnuts drop to the orchard floor and are mechanically harvested by blowing them into windrows and using vacuums or sweepers to pick up the nuts. This is not an ideal method as it requires considerable management of the orchard floor (herbicides and scraping) and depends on a dry harvest season. The production model being used in the Upper Midwest is based on harvesting the nut clusters directly from hedgerows of shrubs with over-the-top harvesters as is done with many different small fruits like blueberries and aronia.
Some early-adopter growers have been using BEI blueberry harvesters using slapper-type harvesting heads with good results. To support these efforts, our research team was recently awarded grant funding from the USDA to evaluate different harvesting equipment and design and build a prototype optimized to hedgerow hazelnuts. The ideal machine will remove only the ripe clusters from the bushes and do little to no damage to the wood or catkins. In addition, the machine will separate the in-shell nuts from the husks in the field as part of the harvest process. Oh, and the machine will be as safe, fast, and cheap as possible.
Most of the work on this grant project will be done in 2019, but thanks to some gracious collaborators we were able to evaluate some equipment this week on plantings in Wisconsin. These are intended as the first of many harvester evaluations. It is important to note these are only preliminary observations of two possible harvest systems and are not intended as an endorsement or criticism of any single harvester unit. We have not yet conducted formal trials. If you have harvester ideas or suggestions, we'd love to hear about them. Contact Jason Fischbach at email@example.com.
Weremczuk Joanna 4
The first machine we looked at was a Weremczuk Joanna 4 harvester primarily used with aronia berries. It works by bending the shrub to roughly a 45 degree angle into the rotary shaker head that uses both fingers and vibration to remove the nut clusters. We were most interested in this unit because it is designed to harvest half a row at a time and, thus, can handle wide hedgerows. It does so with a wedge-nose out front of the unit that wedges half of the stems away from the center of the row and into the shaker head.
The harvesting action was very effective at removing the nut clusters, but it did break some branches and remove some catkins. The wedge-nose that bends over the branches also caused some damage to the bark along the lower stems. Mechanical harvest by any means will result in some damage to the plant, and such damage is offset by the cost-savings of the mechanized-harvest compared to hand-harvest. However, it seemed this unit was causing too much damage. Some slight modifications to the unit, such as a wider throat opening would likely reduce damage. We operated the unit at roughly 0.3 mph, which is also slow.
The plants we were harvesting are located in Stoughton, WI and are a mix of full-sibling hybrids and pure American hazelnut seedlings planted in 2011. As such, every plant is genetically and phenotypically unique, as is evident in the videos. In addition, these plants have not yet been pruned. Commercial hedgerow hazelnut production will likely have more uniform plant material and pruning will be implemented to facilitate the specific harvest system, meaning any harvest unit will work better when the plants are managed specific to the unit.
BEI Slapper-Type Harvester
There are at least 2 BEI harvesters being used by hazelnut growers in the Upper Midwest and we were able to view one in action this week. The BEI harvesters are still being used by blueberry and aronia growers in the Upper Midwest though BEI is no longer in business. These harvesters work by shaking or slapping the nut clusters off the branches as they pass over the shrubs. The nut clusters are collected at the bottom of the outside walls and moved out the back of the machine into bins.
The slapper-type head in the BEI unit is much less aggressive and results in less damage to the plants, but it did not remove all the clusters, particularly in the middle of the wider plants. A more aggressive setting on the slapper bars would likely improve the detachment percentage. At 48" tall and 72" high, the tunnel is able to accommodate fairly large plants, as long as the stems are supple enough to bend and not break when passing through the tunnel. The unit shown was travelling between 0.8 and 1.0 mph. This particular planting is roughly 12 years old and had much lower nut production this year compared to previous years. The plantings in the Upper Midwest (as with the one shown in the video) consist of genetically unique seedlings resulting in a wide variation of ripening times among the plants. In addition, within a single plant, the cluster detachment forces can vary significantly among the clusters as they ripen. For this reason, the plants are harvested 2-3 different times with the slappers set at a fairly gentle setting to harvest only the ripe clusters while minimizing damage to the plants. Although the aggressiveness of the Joanna 4 unit can be adjusted, it likely is best used as a single-pass unit given the damage that it causes. The slapper-type action appears better suited for the multiple-pass approach required in the diverse seedling plantings. More uniform plant material would likely reduce the number of necessary passes, and, as with the Joanna 4, using plant material and implementing pruning strategies specific to the harvest unit will increase effectiveness of the harvester.
With over-the-top harvesting in mind, our UMHDI breeding program has been focused on developing relatively compact plants with early high yields. Combined, those two attributes will help keep the plant size manageable and easier to mechanically harvest.
- Contributed by J. Fischbach
Hazelnut harvest started in earnest in Bayfield, WI last week and we anticipate finishing up early next week. It certainly has been a compressed harvest season compared to last year. Typically our plants ripen over a four week period, but this year it'll be no more than 14 days. The video is from one of our hazelnut performance trials located in Bayfield, WI. We have been tracking the performance of the individual plants in these trials since 2010 with the goal of finding roughly 10 1st generation genotypes capable of supporting a viable hazelnut industry in the Upper Midwest. The genotypes being evaluated at Bayfield are also in trials at St. Paul, MN, Tomahawk, WI, Lake City, MN, and Lamberton, MN. Each plant is harvested individually in order to determine annual kernel yields. We also evaluate growth form, resistance to Eastern Filbert Blight, kernel quality, and resistance to big bud mite.
The plants at Bayfield are grown with 6' in-row spacing and 15' between-row spacing, which is likely too much space at the Bayfield location where the soils are sandy and not very fertile. We envision hazelnuts in the Upper Midwest being grown in hedgerows and key for economic viability will be filling the hedgerows quickly, and for Bayfield that will likely mean a higher initial planting density. We are just mowing the herbaceous vegetation between the rows at Bayfield, but in a production system that growers might use, the row-middles would be used during the establishment phase for alley-cropping to generate income while the hazelnuts are maturing.
- Contributed by J. Fischbach