When we farmers genetically improve our crops, we are usually enhancing potentials already present in at least some of the individuals of that crop. This contrasts strongly with inserting foreign genes, as in GMO corn or soybeans. So we need to set realistic goals based on what we have occasionally seen in our crops in the past, or what we have witnessed with others’ crops, or on printed, unbiased details of what specific crops can do. Realistic goals include disease tolerances, some temperature tolerances, yield potentials, etc. On the farm level, it is not realistic to strive for absurd goals such as hard-freeze tolerance in tomatoes or watermelons; these crops came from a genetic lineage that includes little such history. On the other hand, be open to some of the amazing potentials in plants. It always astounds me that many delicate, flavorful oakleaf and butter lettuces can make it through our winters and then regrow so rapidly with spring. Equally shocking is that our most ancestral currant tomato tolerates mild frost
Choosing an initial line or provenance (roughly, provenance means seed from a specific source) that has some of the characteristics you need can be slow and frustrating. Many gene lines will lack that vital set of potentials that could make them improvable for your site and market. Keep trying. We had to try three different lines of Chinese thick-stem mustard, four lines of San Marzano plum tomatoes, and four lines of Brandywine tomato before we found the right starters for beginning our seed saving and genetic improvement. If you assume that most purchased stock isn’t very adapted to your farm, you can do side-by-side plantings of different provenances of ostensibly the same cultivar, and then rogue out the worst gene lines after screens typical to your farm have done their work. This saves you a lot of time for only a marginally higher cost of seed and labor. But like many farmers, we generally are passive, lulled, and hopeful when reading the seed catalogs, and do same-year provenance trials only after two or three sources have failed us.

Planting the new acquisition under the right conditions to impose your screen(s) is easy. For cold tolerance, plant in the fall so winter will soon test the plants. For good seed germination and emergence under cold, wet conditions, plant in early spring and then irrigate generously to simulate rain. To screen for drought hardiness, plant new acquisitions in a field that is not covered by your irrigation system and then hope for some dry weather. Or to screen for improved disease tolerances, choose a field or management regime that you know will help that disease flourish. An example of this would be planting tomatoes in a field with a higher-than-normal incidence of the wilt fungi the year after you’d planted it to your market tomatoes (i.e., violate your rotation to intentionally stress new gene lines).

After the screen(s) have weeded out the weaker individuals, go in and finish the job before pollen production begins. With plants that are harvested whole, this is easy, because you sell the weak survivors and keep the strong survivors to be mother plants. Yes, we do sell a lot of baby tat soi and baby lettuces every spring. With crops that produce fruits, you will rogue with no such secondary goals. And remember that the most important time to rogue is in the first few years, i.e. with the smaller plantings.

Selecting the best mother plants follows exactly the same procedures as number 5, “Choosing the best mother plants . . .”, in the Seed-Saving section in last month’s article.

To harvest and process seed of this now-improved line, again refer to the Seed Saving section. Annually, repeat steps 3 through 5 with the improving line. Every time you screen for vigor under adverse field conditions, you are investing in an ever-stronger crop that can let you sell product when times are rough. You don’t have to screen and rogue as intensively as the years go on, but remember to take advantage of especially bad weather conditions. In those years, treasure your survivors: they will be superb mother plants for gene lines way better than anything you can buy. The minimum amount of time you need to gauge whether you’ve adequately met your genetic improvement goals is three years. While we usually see improved performance faster than that, with 14 years of experience we have greatly refined the screening suitable for our marketing and management needs. We also readily factor in the year-to-year weather variability when assessing crop performance. If you do see good improvement in your crop after three years, great. If you don’t see this, consider abandoning that gene line, or changing your genetic improvement goals, or tightening up your rogueing or mother-plant selection.

That completes the detailed tasks we use in seed-saving and genetic improvement. I’ll now use two examples to show how the process can work: arugula and San Marzano plum tomatoes.
We grow many acres of winter greens. While many of these crops are for typical American consumers, our restaurant accounts want a lot of arugula, a leafy green from the Mediterranean basin that is like a nutty-flavored cress. When we first started growing this, we got seed from my Italian mother that probably originated, years before, from a U.S. seed company, labeled “rocket” or “roquette” (this was before the arugula craze hit mainstream America). The gene line had excellent flavor but initially was adequately (not wonderfully) winter-hardy, with freeze-induced mortality rates of 30-40%. After four years of selection work, I got mortality down to 10%, and one of those winters was abnormally harsh, with an ice storm, two major blizzards, and frozen ground for nine weeks. I nonetheless sensed a slight but real decline in plant size and regrowth rates. As I had been saving seed from only about 15 mother plants, and had no idea about the original source of this line, I suspected inbreeding depression. I set a new goal for our arugula of broadening the genetic base while retaining that good winter-hardiness, and got 1 ounce of seed from a company in Turkey that sells mostly to farmers. This was planted in September in beds alongside our old arugula. We got another really bad winter, this one with many warm spells followed by super-fast freezes and ice ( the kiss of death for many annuals), as well as insulating snow. Survival in both beds was still very good: 90-95%. I further rogued about 5%, but those were the only scrawny ones of the lot. Once the plants started to flower in April, I snipped flower heads of the earliest individuals in order to get all plants to flower at once (this took a total of only 40 minutes). Pollinators freely roamed the adjacent beds. Finally, I selected the best 60% (about 80 individuals) as mother plants for this new adapted land race, and bulked all seed together.

We have never looked back. We grow outstanding arugula that, for our upscale restaurant accounts, sets a standard that the produce wholesalers have never been able to meet. This isn’t because it is certified organic. This is because our arugula is slow-grown, in the winter, when full flavors --some subtle and some powerful-- can develop. That is what our chefs, with rather jaded but extraordinarily experienced palates, want. That is the market goal. The production goal is growing acres of the stuff outside and not in treasured greenhouse space. And the genetic goal is winter hardiness. We have had 100% survival for at least 5 winters, have strong drought tolerance, viral incidences of only about 3 plants per acre, and zero black-leg even in the wettest springs. We do no roguing whatsoever, and mother plant selection and harvesting takes only about four hours to get enough seed for 3 or so acres of sowing (seed processing takes another 6 hours). The land race is established and now takes little care. And no, we don’t sell the seed, unless you’ll put on your Carhartts and harvest with me for a week in those biting winter winds.

San Marzano tomato
San Marzano plum tomatoes can illustrate summer cropping. In the Washington D.C. metro area, Italian influences are strong in the cooking of many of the better restaurants. The San Marzano is an excellent calling card for new accounts that know that that heirloom is one of the elite plum tomatoes of the world. So we had a market goal of growing them in our hot and humid part of Maryland (very different from the dry Mediterranean climate in which they were originally developed), and under organic production constraints as well. Our genetic goal was to develop a high-yielding land race that was tolerant of the three microbial pests of tomato rampant in our area, and that was indeterminant as well. We tried hard to get Italian seed initially but failed. So over two years we bought one hybrid and two non-hybrid “Super-Marzano” or “San Marzano” from reputable U.S. seed companies. The trials of these were dismal and frustrating failures, with little disease tolerances (in spite of advertised claims to the contrary), persistent blossom end rot even when none of our other cultivars had such, very poor flavor, and hence no marketable yield. Nothing is more expensive than bad seed: we had thrown away not just the dollars spent for those claimants to the San Marzano name, but also all the time, labor, and supplies involved in growing 1 acres of those tomatoes. And they tasted like junk. So I decided that the initial line or provenance for our market and production goals had to be better. I contacted a old family friend from Italy, and I filed the appropriate paperwork with USDA for an seed import permit. Both parties came through superbly, and in 1999 we received non-hybrid, non-hyped San Marzano seed from a farm-specialized seed company in Italy. The differences even in that year of the worst drought in state history were amazing. We had very high marketable yields of finely-flavored plum tomatoes even under a water rationing irrigation scheme and with a bone-dry subsoil. Since then, we have successfully screened for further tolerance to TMV and the wilt fungi. We have had no success, though, in developing this adapted San Marzano land race into one that will give us fruits over extended periods. The line remains stubbornly determinant. So we have taken that genetic goal away and substitute management—sequential plantings—to attain the same marketing goal.

Unlike the arugula illustrated above, neither this San Marzano plum nor any of our lines of heirloom beefsteaks can be considered secure: we still rigorously rogue all tomatoes that show susceptibility to TMV or the wilt fungi, and as rigorously and intensively select for vigor in mother plants. I believe that the microbial pests of tomato are quite capable of evolving to overcome host defenses, especially here where winter is briefer than in the northern tier of the U.S. So maintaining our tomato land races takes work. But I’ve seen a big spike in wholesale tomato seed prices recently, so our labor seems well-justified. This is especially so for the intensely-flavored cherry tomatoes, which for us require little genetic care. They are closer to wild-type, and so produce generously almost regardless of field conditions

Disadvantages of seed saving
To present the other side I’ll now outline the most relevant disadvantages I see in seed saving and genetic management.

One reason to not save your own seed is that it takes a lot of time, especially in the beginning year or two. We are very diverse market farmers. We have loss leaders, as well as crops which aren’t profitable in themselves but which add the diversity we depend on to attract both wholesale and retail customers. It seems crazy to save seed of these crops; I feel I’m throwing good money, and labor, after bad. Turnips are a case in point: They can be hard to sell and conventional wholesale seed is really cheap. Further, they are fully winter-hardy. I see little room for improvement. We now buy organic seed because of the new national regulations about certified growers using organic seed. And I’ll only allocate time once again for the very simple task of turnip seed-saving when the price of organic seed goes above my cost threshold, or when I can sell a lot more of these lovely roots at a greater profit.

Another reason to not undertake seed saving is that it is hard to keep cross-pollination under control, especially on small farms. Many cultivars love to donate pollen and it is hard to stop insect or wind pollinators. If you slip up, you can end up with a mishmash of crops, cherry tomatoes interspersed with beefsteaks and Brandywines. This makes harvesting a challenge. I believe that the temporal and spatial approaches I detailed in the seed-saving section help immeasurably, and farmers can use our own creativity to come up with other ways to keep pollen away from where it doesn’t belong. But recall too that many a bag of purchased seed has disconcertingly high levels of contaminating seeds that we discover only when they are growing on our farms.

A third reason to not save seed is added equipment costs. We spend $200-400 annually on cover crop seed. If I inherited a thresher and cleaner to process our own clover and vetch seed, I’d be pretty happy. But my parents aren’t farmers and the equipment isn’t cheap. So I’ll be buying such seed into the distant future because green manures are vital investments in our long-term soil health and productivity.

The final reason to avoid adding this responsibility to your chores is that it can be a nuisance. I actively resent any new task that draws my thoughts away from the constant hustle inherent in being a successful, market-driven farmer. I don’t want to devote brain cells or field labor to anything except the pressing needs of getting all those cases out of the fields and into the truck. Seed companies might be able simplify my life.

Seed-saving and genetic management, though, are like very good soils: once you’ve experienced what they can do for your production, you may never want to go back to, like poor soils, inadequate purchased seed. Not all bought seed is bad. But it is rarely fully adapted to your farm and often delivers less than promised. We compete at the wholesale level by offering produce that is outstanding and by offering service and reliability that the produce wholesalers cannot match. Our restaurant customers don’t want to hear about crop failures due to disease outbreaks, drought, or frigid temperatures. The cases coming in the back door are how chefs measure us. And at farmers’ markets, we build bases of loyal retail customers that know that our production is more reliable than that of other farms.

Using ultimately adapted gene lines lets us produce, and thereby compete, when others can’t. We still buy a small amount of seed every year. But I overwhelmingly choose the added hassles of seed-saving and genetic management over the alternative, unreliable production from low-quality purchased seed. I recommend that you consider the same.

Brett Grohsgal owns and operates Even Star Farm in Lexington Park, Maryland, with his wife, Dr. Christine Bergmark, who is also director of the Tobacco Transition Program in southern Maryland. Brett has a Master’s Degree in Soil Science from North Carolina State University and Christine has a Doctorate in Plant Physiology. They sell to about 14 restaurants and three grocers and at two farmers’ markets.