Positive Selection
Part of Seed Saving
Positive selection is the deliberate act of choosing the best plants as seed parents and removing the worst before they can contribute pollen or seed to the next generation. Applied consistently over multiple seasons, it gradually shifts a population toward traits you value — higher yield, better disease resistance, improved storability, or adaptation to your specific climate.
What Positive Selection Achieves
Every time you choose which plants to save seed from, you are acting as a plant breeder. Plants with good traits produce offspring more likely to carry those traits. Plants removed before they can reproduce leave no genetic legacy.
This is the process that created every crop variety humans grow today. Wild ancestors were gradually shaped, over hundreds and thousands of generations, by farmers who noticed which plants performed better and saved their seed preferentially.
In a post-collapse context, you have the same power — and the same responsibility. The varieties you refine now will feed your community for decades. The effort invested in careful selection compounds across generations.
The Three Components of Positive Selection
1. Marking: Identifying superior plants early, before harvest, so you know which to save from
2. Roguing: Removing clearly inferior or off-type plants before they can pollinate or contribute seed
3. Preferential harvest: Collecting seed from marked superior plants first, and setting it aside separately
All three must be practiced together. Marking without roguing allows inferior pollen to contaminate the seed parents. Roguing without marking means your seed harvest is still random. Preferential harvest without the earlier steps produces inconsistent results.
What to Select For: Trait Prioritization
Not all traits are equally valuable, and selecting for too many traits simultaneously slows progress on each one. Prioritize 1–3 traits per generation.
Primary traits (highest priority in survival context):
| Trait | Why It Matters | How to Identify |
|---|---|---|
| Disease resistance | Keeps crop viable under stress | Plants that show no disease when others are affected |
| Drought tolerance | Critical for climate variability | Plants that remain productive with less water |
| Yield consistency | Reliable production across conditions | Plants that yield well regardless of weather variation |
| Maturity date | Allows adaptation to season length | Earliest or latest plants to mature, depending on need |
| Storability | Extends harvest utility | Fruits or roots that hold well after harvest |
Secondary traits (select for after primaries are stable):
| Trait | Notes |
|---|---|
| Flavor | Important for food quality and acceptance |
| Size | Useful but not essential for survival |
| Color | Aesthetic; can be a marker for other traits |
| Uniformity | Important for commercial growing; less so for survival |
Avoid selecting for uniformity at the expense of diversity. A slightly variable population is often more resilient than a very uniform one. Uniformity reduces the population’s ability to respond to novel stresses.
The Marking System
Mark superior plants at the moment you observe the trait you are selecting for — do not rely on memory.
Simple marking methods:
- Colored yarn or ribbon: Tie a piece of brightly colored yarn around the stem or stake of superior plants when you observe the desired trait
- Stakes with flags: Small stakes with cloth flags; different colors for different traits
- Paint marks: Dab the stem with a small amount of paint (non-toxic)
- Exclusion marking: Mark plants you are excluding (rogue targets) rather than plants you are keeping — whichever system is less work
Use a consistent color code each season so records are interpretable years later. Example: red = high yield, blue = disease resistant, yellow = earliest maturing.
Timing of Selection
Different traits must be observed at different growth stages:
| Trait | When to Evaluate |
|---|---|
| Germination vigor | Seedling stage |
| Disease resistance | Throughout season, especially under disease pressure |
| Early maturity | Flowering and early fruit set |
| Yield | At peak harvest |
| Fruit quality | At peak ripeness |
| Storability | After 4–8 weeks in storage |
| Seed quality | At seed harvest |
Walk the planting at each of these stages, marking superior plants as you observe them. A plant that was marked for disease resistance early may prove disappointing at yield. Final seed-saving decisions should incorporate all observations.
Roguing: Removing Off-Types and Inferior Plants
Roguing is the removal of plants that do not belong in your seed population. This includes:
- Off-types: Plants that visibly differ from the variety in leaf shape, flower color, growth habit, or other characteristic traits — likely crosses or mutations
- Diseased plants: Any plant showing virus symptoms, severe fungal infection, or bacterial blight
- Chronically underperformers: Plants that consistently lag in growth, yield, or vigor without obvious environmental explanation
- Late-bolting failures in biennials: In crops selected for late bolting (lettuce, brassicas), remove any plant that bolts prematurely
Rogue Before Flowering
Roguing must happen before plants flower, or diseased/off-type plants will contribute pollen to your seed parents. In cross-pollinating crops, even a single off-type plant can contaminate the pollen pool. For self-pollinating crops, timing is more forgiving, but early removal is always better.
Roguing intensity: Do not remove more than 20–25% of a population in any one generation for any trait. Heavy roguing reduces effective population size and can cause unintended loss of rare but valuable genetics. If a population has severe problems, spread correction across 3–5 generations.
Selecting Seed-Bearing Plants
From the remaining population (after roguing), designate your seed parents. Choose:
- The top 20–30% of the population by your primary trait criteria
- Plants that have also survived to full maturity without problems
- Plants spread across the planting area (not clustered in one corner, which may have better soil or microclimate)
- Plants representing the full range of acceptable variation, not just the single “best” individual
How many plants to designate as seed parents: See minimum population size guidelines. Never designate fewer than the minimum, even if you find fewer plants that fully meet your criteria. Accept the best of what you have.
Preferential Harvest
Harvest seed from designated seed parents first, before any other harvest activity disrupts the planting. Harvest from non-seed plants afterward, for eating.
If seed from designated plants is not clearly separated from the rest, the selection work is wasted.
Color-coded buckets or labeled bags at harvest time ensure seed lots remain distinct. In the field, mark harvested seed plants with a different colored tag so you know they have been collected.
Family Selection (Advanced)
Mass selection (choosing the best individuals) is the simplest approach. Family selection is more powerful but requires more record keeping:
- Select superior plants and keep their seed separately — each plant’s seed is a “family”
- Plant each family in a separate row the next season
- Evaluate families as units — which row performs best overall?
- Save seed from the best-performing rows (families), not just the best individuals within any row
- Continue for 3–5 generations
Family selection is more effective because a plant’s individual performance can be partly due to its position (better soil patch, less competition). Family performance averages out micro-environmental variation and better reveals true genetic quality.
Tracking Progress
Keep a season-by-season log:
| Season | Trait Selected For | % Rogued | # Seed Parents | Notes |
|---|---|---|---|---|
| Year 1 | Disease resistance | 15% | 18 plants | Severe blight pressure |
| Year 2 | Yield | 10% | 22 plants | Good season — easy selection |
| Year 3 | Early maturity | 12% | 20 plants | Shortened season; selected earliest |
Compare germination rates, yield, and disease incidence year-over-year. Improvement should be measurable within 3–5 generations of consistent selection.
Common Mistakes
Selecting only from the largest fruits or heaviest plants: Size may reflect soil position rather than genetics. A large fruit from a plant in the best corner of the garden may produce average offspring. Select based on performance across the whole plant and the full season.
Selecting the same trait every year: If you always select for largest fruit, you will eventually lose other qualities (disease resistance, flavor, storability). Rotate selection emphasis.
Discarding the seed from “ordinary” plants: In a diverse population, ordinary-looking plants often carry valuable recessive traits. Keep the seed from non-seed-parent plants for eating, but do not assume they have nothing to offer genetically.
Neglecting roguing in a rush season: When time is short, roguing is the first thing skipped. It should be the last. Five minutes of roguing before flowering is worth more than hours of other seed-saving work.
Observe Under Stress
The most informative selection environment is stress — drought, disease pressure, cold snap. Under good conditions, most plants perform adequately. Under stress, differences in genetics become visible. If you have the opportunity to observe plants under difficult conditions, use that as your primary selection event.
Positive Selection Summary
Positive selection improves a variety over generations by marking superior plants, removing off-types and underperformers before flowering (roguing), and collecting seed exclusively from the best representatives of the population. Prioritize 1–3 traits per season, observe at multiple growth stages, and maintain adequate population sizes throughout. Family selection offers more precision than mass selection for complex traits. Consistent records allow you to track progress and refine selection strategy over years.