Selecting in a wilt-sick plot: F4 chickpea families under a maturity and seed-size penalty
A rainfed pulse breeder holds 250 F4 families from a cross between an early, wilt-susceptible desi cultivar and a wilt-resistant kabuli donor. The station has one wilt-sick plot, where the susceptible check dies at 35-40%, and one clean plot. The target environment is rainfed, must be harvested by the first week of March to escape terminal heat, and the kabuli grain has to reach 9 mm for the premium market. Seed and labour allow only about 180 plots this season, and each family has 40 seeds left after storage losses. Disease screening cannot be repeated in the same plot within the season, and the sick plot has shown uneven patches in past years. The station chief wants at least 20 advanced lines in the initial varietal trial two years from now. The breeder must submit the selection and crossing plan next week.
1. Choose the selection scheme for this season - bulk, single-seed descent, or single-plant selection within families - and justify it against the station's two-year deadline.
Read the worked answer
Recommendation: single-plant selection within families, screened in the sick plot, with confirmation in the clean plot.
Reasoning:
- At F4, families are about 87.5% homozygous, so a selected single plant will breed reasonably true; you gain a generation instead of losing one.
- The sick plot supplies information that cannot be bought elsewhere. Bulk harvesting would carry susceptible segregants forward unseen, and single-seed descent advances generations without any disease exposure - you would only learn about wilt in the varietal trial, which is too late for a two-year deadline.
- Within-family selection exploits both between-family and within-family variance for wilt, which matters when only 180 of 250 families can be sown.
Practical design: sow 10 plants per family in two rows, with a susceptible check every tenth row, and randomise families. Score per-family mortality as a percentage, not just survival of your chosen plants. Select only within families showing under 10% mortality. Mark selected plants at flowering for earliness and pod load, then harvest single plants individually.
Guard against the uneven patch: if check mortality in any block falls below 25%, treat that block as an escape and do not select in it.
2. How do you stop disease screening from silently selecting against kabuli seed size and earliness, and how do you avoid calling escapes 'resistant'?
Read the worked answer
The trap: a sick plot rewards survival, and the plants that survive longest are often late and small-seeded - exactly the traits the market and the March deadline forbid. Selection on survival alone will produce resistant, late, bold-seeded rejects.
Design fixes:
- Select on family mean wilt incidence first, then select individual plants only within acceptable families - this keeps the disease filter at family level, where it is well estimated.
- Score days to 50% flowering and pod load at flowering, before any disease death, and tag early plants; harvest only tagged plants.
- Do not discard large-seeded families that show moderate resistance at F4. Seed size is quantitatively inherited and is poorly expressed under disease stress, so it is judged in the clean plot and the next generation.
- Run one clean-plot replication of survivors to separate genetic seed size from stress-induced shrivelling.
Avoiding false resistance: escapes occur where inoculum is patchy, so (i) keep the susceptible check every tenth row, (ii) discard blocks where the check shows under 25% mortality, (iii) treat this season as a screen, not a decision - survivors are re-screened in a second season or a second sick plot before being called resistant. Wilt is soil-borne, so late plants do not escape by phenology; they escape by position. Position is what you must randomise against.
3. Twenty lines survive, each with about 30 g of seed. Lay out the next two seasons to reach the initial varietal trial, stating how you use that little seed.
Read the worked answer
Season 1 - preliminary yield trial, two locations. Chickpea sowing at roughly 30-40 kg/ha is about 3.5-4 g per 3 m row. Thirty grams therefore buys about three replicated 3 m rows: a three-replication single-row trial per location, with the check and the resistant parent included. Keep 5 g of each line in cold storage as a reserve for DNA fingerprinting and uniqueness testing, and never sow the reserve.
Test at the station plus one hotspot or farmer field, because one season at one site cannot separate line from site.
Season 1 selection: from 20 lines, advance 8-10. Criteria in order: wilt reaction across the two sites, days to maturity compatible with the first week of March, seed size on the harvested produce, then yield. Do not advance all 20 - the varietal trial has limited plots and weak lines consume them.
Season 2 - initial varietal trial. The harvested seed from 8-10 lines (roughly 150-250 g each) is enough for 3-4 m plots with three replications. Include the check and one local resistant check; record wilt, maturity, 100-seed weight, and yield.
Risk management: if season 1 yields less than about 100 g per line, run a seed-increase row rather than the trial and enter the varietal trial one season later - better a late entry than an unreplicated one.