Why should any of us save seed of our crops? Why should we add to our already lengthy lists of tasks and responsibilities? Because we can get much higher quality seed for replanting. Because we can save a good deal of money. Because we can guarantee our seed source for key crops and don’t have to cope with other seed producers’ failures or shortfalls. And because for the certified organic farmers among us, we’ll soon be required to use organic seed or to extensively document why we didn’t.

And why should we genetically manage our own crops? Because most of us can do a much better job of it than can large national seed companies. Because we can build in outstanding degrees of local adaptability, disease tolerance, and weather hardiness, all without sacrificing flavor. And because year after incremental year, genetic management lets the rest of the farming operation develop and profit from always improving, superior seed.

These days the words “genetic management” and “crops” easily evoke dialog about the GMOs, genetically modified organisms, that many consumers fear and that organic farmers are prohibited from using. This article isn’t about GMOs. Instead, I outline the straightforward techniques of selection and seed saving that farmers have done for millennia across the globe. For all the hype about this year’s hottest new hybrid variety of lettuce, tomato, or melon, many directly marketing farmers find that consumers really want old-fashioned, distinctively flavored varieties. Our farm needs to buy some seed every year. But we have also developed our own open-pollinated, fully adapted, disease-resistant, and vigorous vegetable lines. These have come from three sources: heirlooms, open-pollinated newer varieties, and hybrid newer varieties. In doing so, we have gratefully stood on the shoulders of the countless farmers who selected and saved seed over many thousands of years. These ancient and more recent farmers, and not big seed companies, are primarily responsible for the tomatoes, corn, beans, lettuces, melons, et alia, that grace our fields and our tables. While seed companies help nearly all of us, we don’t have to depend on them for each and every crop that we plant. We can do the job better.

Sharpen your competitive edge
Farmers do not need a lot of time or a team of breeders to improve our competitive edge. We typically see unadapted starting materials evolve into noticeably better seed and crops after only one year of intensive screening and management;truly superior crops after 3 years of management; and vastly superior, utterly adapted lines after 5-6 years.

This has helped our farm to compete and to grow in a number of ways. We have attained:
•tolerance of or resistance to Fusarium and Verticillium wilts, and to tobacco mosaic virus, in 25 lines of tomatoes and 10 lines of peppers;
•tolerance of Rhizoctonia in all of 10 lines of Brassicaceae (cole crops);
•drought hardiness in the newly planted stage of 5 of our brassicas (in the drought of September 2001 through March 2002, we had 5 acres of these that got no irrigation or rain, only periodic fog, for the first two months after germination, but that held on long enough to produce marketable yields once the rains came);
•cold hardiness in our 10 brassica lines that lets us direct-sow in September and harvest, whenever the ground isn’t frozen, from October through April. We are practically alone in the marketplace in the cold months, and this makes for strong sales. While our winters in Maryland are pretty mild, our farm is extraordinarily windy and has freezes nearly every night in deep winter, and at least one period of 5-15 degree temperatures practically every winter. Not Vermont, but not Florida, either. I’ll cover cold hardiness in depth later.
•greatly reduced splitting in cherry and teardrop tomatoes following rains in the heat of summer. This gives us premium product when a lot of other local farmers are short.
•much higher and more uniform germination rates and seedling vigor than with purchased seed.

In sum, we have gotten – as can nearly any other farmer - high yields under real and often adverse field conditions. The key to all of this is selecting for local adaptability, for vigor, and of course for flavor and easy sales.

Definition of terms
There is a small amount of vocabulary required for this outline of farm-level crop genetics. Here it goes, and I apologize to the scientific community for perhaps over-simplifying:
Genetic line (usually shortened to line): any group of plants sharing parents (at some point) and many recognizable traits. A variety, or cultivar, is a genetic line. A line can also be a smaller group, such as the particular strain of Brandywine tomato saved by a farm family over the years.
Genetic management is managing the presence and frequency of certain genes in a breeding population over time, with the tools of selection, breeding, or whatever. It sounds complex but needn’t be: the ancient farmers who bequeathed to us nearly all vegetables and grains never read Mendel, never knew what a chromosome was, and never patented a variety. And they were outstanding genetic managers.

Hybrid: the offspring of two relatively unrelated parents. If in a planned breeding program, hybrids can be exceedingly uniform. Developing your own lines from hybrids can be frustratingly time-intensive but is sometimes unavoidable.

Inbreeding: in most cases, a bad result of using too small an initial population of parents to retain the genetic diversity needed for long-term crop vigor. In some cases (not here), inbreeding is intentionally used prior to creating hybrid offspring.

Inbreeding depression is the decline in vigor or yield caused by excessive inbreeding, and can be insidious and hard to notice on the farm level. Increasing the number of mother plants from which you collect seed helps greatly in avoiding inbreeding; see the later outline on Seed Saving.

Land race: a locally adapted, heterogenous, genetically diverse breeding population. Such a group is reliable and resilient under adverse field conditions and is vigorous when times are good. This, and not the hybrid in the seed catalog sense, is the goal for our farm and for this article.

Mother plants are the individuals the farmer selects for seed saving. Most crops have both female and male parts. As pollen is usually very mobile, when we choose a mother plant of a crop that doesn’t self-pollinate we are generally ignorant of which plant(s) fathered the seed. This is called maternal selection. While you can get the fastest genetic improvement by actively choosing both parents, this takes more work and maternal selection works very well on the farm level. See the definition of rogueing.

Population: genetically, this refers to all the individuals in your fields that readily cross-pollinate each other. Pollen doesn’t recognize field boundaries, though, so your sweet corn and your neighbor’s GMO feed corn could form one population, if they tassel and silk simultaneously.
Rogueing is the act of destroying inferior or diseased members of your breeding population before they can make pollen or seed. As most of our crops are high-value and hand-harvested, we routinely rogue while harvesting greens or at the second walk-through on fruit crops like tomatoes. This limits pollen-making to stronger individuals and is a form of paternal selection. Many of our crops are only intensively rogued the first year - i.e., with plants grown directly from purchased seed.

Screening is the imposition of one or more harsh environmental conditions in order to destroy the weaker members of your breeding population. Screening is the single most powerful tool we use in genetic management. Screens need to be rigorous enough to eliminate good numbers (e.g., 30%) of an unimproved starting population but not so harsh that you are left with too few surviving parents.

Selection in this article has two meanings: a) genetic selection, in which the environment favors some individuals’ breeding abilities, and hurts others’, after the imposition of a screen, and b) mother plant selection, in which the farmer or a trusted employee chooses the most vigorous plants with the desired crop characteristics (e.g., large fruits on heirloom tomatoes) from which to harvest seed.

If all this sounds excessively complicated, please forgive me. It truly isn’t. If you are already running a market-driven farm, you have the backbone to manage your own crop genetics. And if you are a surviving farmer in this competitive and corporate era, you’ve got more than enough brain cells to manage your crop genetics very well. Most importantly, seed saving and genetic management can be readily integrated into the seasonal operations of most market farms.
The way to begin seed saving and genetic management is to start small –choose only one or two cultivars whose purchased seed cost is very high, or has unsatisfactory germination rates, or that would fit your marketing or production needs better if it were more tolerant of some disease or of freezes or whatever. And don’t try to develop an adapted land race from hybrid initial stock until you’ve got some experience. There are plenty of excellent non-hybrid varieties.
The remainder of this article has 4 parts: an overview of the main tasks in seed saving and genetic management, details on how we do these tasks, two examples of very successful seed saving and genetic management (arugula and San Marzano plum tomatoes), and a small counter-argument – some reasons to consider not saving seed or managing your crops’ genetics.

Seed saving
Decide what crop or cultivar has purchased seed that is unsatisfactory in quality or cost.
Plant crop in blocks or otherwise isolate that crop to minimize accidental cross-pollination.
Treat crop normally in fertilization, irrigation, etc.: don’t baby it.
Determine number of mother plants needed to avoid inbreeding depression.
Decide on how you’ll choose superior mother plants (i.e., what does “best” look like?).
Select best mother plants and harvest seed pods or fruit from these.
Process seed correctly and air-dry.
Store seed at right temperature after double-wrapping.

Genetic management
1. Set goals for crop improvement (e.g., cold tolerance).
2. Choose an initial line or provenance that has the right group of characteristics.
3. Plant it under the right conditions to impose the screen(s).
4. After screen(s) has done its job, if your time permits, destroy the weakest to prevent their pollen from fertilizing better mother plants.
5. Also after screen(s) has done its job, further select for the best surviving mother plants.
6. Harvest seed of this now-improved line.
Annually, repeat steps 3 through 6 with the new improved line.
Give yourself three years or crop cycles to assess success.

The tasks detailed
Deciding what crop or cultivar will be your first try in seed saving should be easy. As a farmer you are very aware of seed costs. Hybrid seed justifiably costs more because there is generally more labor invested before you even receive those seeds. But if you flinch when purchasing non-hybrid, easily grown seed like watermelon, squashes, heirloom tomatoes, or greens, then choose one of those. Alternatively, seed or source quality may be irksome. Our farm had real trouble finding good quality seed of San Marzano plum tomatoes; we tried U.S. lines and found them woefully inadequate both on the table and in the field. So when we finally got a small seed shipment directly from Italy, and found that it produced an outstanding crop, we chose to save seed from that line to guarantee our source in the future.

Accidental cross-pollination in seed saving can be a real problem for either farmers limited by space or for farmers who forget that pollen is extremely mobile. There are at least two tools we use to minimize cross-pollination: spatial separation and temporal separation. We plant in blocks of rows (i.e., beds), never in isolated rows, to better manage both our soils and our crop genetics. When we plan our crop rotation scheme annually, we separate varieties likely to accept each other’s pollen by large blocks of other, more distantly related cash crops, by whole fields, and/or by tree lines. We further enhance this by never choosing mother plants that are on that edge of a planting block that is closest to a potential contaminator. These are spatial tools and are most apt for larger farms. Timing to avoid accidental cross-pollination can be done a few ways: a) plant the seed crop particularly early, or late, to avoid the presence of pollen of the contaminator; and b) mow or plow down the potential contaminator right before it develops pollen. These and other temporal tools are particularly useful for smaller farms. We use all of the above, especially with certain lines that love to donate unwanted pollen (e.g. turnips).
The whole point in developing your own ultra-adapted lines is to get seed that will perform superbly for you under your normal field conditions. That means that you need to avoid babying your seed-producing crops. Make the strong survive. The only time we give special treatment to seed crops is when (twice in the history of our farm) we start with too little seed and we need to maximize how much seed we get for the next planting. One problem with many seed sellers is that they need to produce as much seed as possible every year, so seed crops are often given the irrigation, fertilizers, and pesticides to make for consistently lush crops. There is no screening or selection for durability, for reliability, for tolerance to adversity. Don’t take that path.