Select seed for adaptation, not for uniformity: save from the plants that survive and produce in your own soil, weather and pest conditions, and keep enough plants to hold variation in the stock. Pushing a variety toward a tidy uniform look speeds the loss of genes it will need in a bad year.
Part of Community Seed Bank Management
Seed selection chooses which plants will parent the next crop; the choice shapes what the variety becomes.
Uniformity Serves The Market, Adaptation Serves The Table
A commercial variety is bred to look the same in every field, so that machines and buyers can handle it. A community that eats its own crop needs something else: plants that crop in its soil in a dry year, a wet year and a diseased year. Variation is the working material for that. A variety with some plants early and some late, some tall and some short, has insurance against any one season.
Selection narrows variation by design. The skill is in narrowing the traits that matter while leaving the rest alone.
Select On Traits That You Can See And That Pass On
| Trait | How easy to improve | Notes |
|---|---|---|
| Time to flower and ripen | Easy | Highly inherited; moves quickly |
| Plant height, seed colour, pod shape | Easy | Visible; select against off-types |
| Disease resistance | Moderate | Save from plants that stayed healthy when neighbours failed |
| Drought tolerance | Moderate | Save from plants that produced in a dry spell |
| Total yield | Hard | Strongly affected by soil and position, little by genes |
The rule: a trait improves quickly when most of the difference between plants is inherited. Yield is the poorest case, because a plant on a wet corner outyields one on a thin slope without being better. Do not choose the top yielder in an uneven field. Choose plants whose neighbours in the same conditions did worse.
Method
Step 1: Decide the one or two traits that matter this year. For example, survival through a disease outbreak and earliness. More than two dilutes the effect.
Step 2: Mark plants before harvest. Walk the field when the trait shows (the disease year, the dry spell). Tie a strip of cloth on survivors.
Step 3: Include average plants. The seed bank carrier’s advice is to save also from average plants. Extreme selection narrows the gene pool, and a selection of only a handful of top plants leaves too few parents (Minimum Population Sizes).
Step 4: Remove only clear off-types. The Genetic Diversity guide advises not removing more than about a tenth of a population per trait per generation; heavier culling should be spread over several generations (Genetic Diversity).
Step 5: Mix the seed from the marked plants in equal amounts. Equal contributions keep more variation.
Step 6: Record it. Write down what you selected for, how many plants, and the season (Seed Inventory and Catalog System).
Selection Can Go Wrong
- Selecting the bad season’s loser. One year’s weather can favour a variant that fails the next year. Select across several seasons before narrowing.
- Selecting a tidy look. Big, uniform, perfect fruit may be the result of the best soil on the plot.
- Selecting against taste or storage. A disease-resistant tomato nobody likes to eat has no future. Keep tasting and storage tests in the loop.
- Selecting from too few parents. After a hard cull, check the number of plants that remain.
Common Mistakes
- Treating the whole field as one environment. Compare plants that grew side by side.
- Selecting on every trait at once. Progress on any one slows.
- Not recording the criteria. The next steward cannot repeat or reverse the choice.
What’s Next
- Learn next: Landrace Development — years of local selection
- Combine with: Seed Selection — the Phase 2 treatment of the same decision