Advancing Wheat Lines from a Six-Location Trial with a Yellow Rust Hotspot
The All-India varietal trial for timely-sown wheat ran at six locations in 2023-24. Twelve advanced lines (F7, pedigree-derived from HD-type parents) were tested in a randomised complete block design with four replications. Mean trial yield was 5.20 t/ha and the check HD 3086 yielded 5.02 t/ha. Line L-7 gave 6.11 t/ha pooled, with regression coefficient (bi) 1.42 and deviation mean square 0.19, and topped three of six locations. Line L-11 gave 5.44 t/ha with bi 0.94 and deviation mean square 0.03. At Ludhiana, natural yellow rust appeared as a hot spot; L-7 scored 40S on the modified Cobb scale at 95 days, L-11 scored 5MS, and the susceptible check 80S. At Karnal, L-7 was the highest yielder. The breeder has seed for one more season of testing and must nominate two lines for the national seed chain, where certified seed must come from a nucleus with no more than 0.10% off-types and rust history at the seed farm is a documented risk factor.
1. Using both stability parameters and the disease data, advise which lines to advance and which to nominate now for breeder seed, and justify the trade-off.
Read the worked answer
Read the two stability parameters separately before deciding. bi = 1.42 says L-7 responds strongly to favourable environments; deviation MS 0.19 against 0.03 for L-11 says its yield is also far less predictable. Its 1.09 t/ha advantage over the check is real but concentrated in good sites, and Ludhiana shows the price of that response: 40S is a susceptible reaction at high severity, not a field escape. L-11 is only 0.42 t/ha above the check, but bi close to 1 with near-zero deviation MS describes a line that holds rank across environments, and 5MS at the same hot spot is a moderate, possibly adult-plant-type response. Decision: advance both to a second year; nominate only L-11 to the seed chain now. Send L-7 to an artificial epiphytotic nursery and to spot testing in the north-western plains before any breeder seed is multiplied. A 0.10% off-type limit is meaningless if the nucleus itself places a rust-susceptible, high-input line in a district with documented rust history.
2. The pathologist argues that one natural rust score cannot support a release decision. Design a screening protocol and an explicit decision rule that the breeders' committee can adopt before seeing next season's data.
Read the worked answer
A single natural reading confounds escape with resistance: at Ludhiana, inoculum load, dew period and growth stage were not recorded, and one score at 95 days misses the shape of the epidemic. Replace it with: (i) two seasons of artificial inoculation using a local mixed isolate, with susceptible spreader rows, scoring per cent severity and response type at three dates to give an AUDPC; (ii) a susceptible and a resistant check in every block so scores are adjusted to the check; (iii) a decision rule fixed in advance - advance a line only if its AUDPC is at most 25% of the susceptible check, final severity is at most 10S, and it is not in the top quartile of severity at any location. Keep L-7 in the trial but flag it 'high risk, conditional advancement'. Writing the threshold before the data are seen is the point: otherwise a breeder who likes the yield simply reinterprets the score afterwards.
3. If L-7 is eventually released, what seed-production safeguards follow directly from these data?
Read the worked answer
The data dictate where and how the nucleus is produced, not just how it is tested. First, produce breeder seed at a rust-free location (north-hill or off-season site) rather than at a farm in the rust hotspot, and record the rust history of the seed farm as part of site selection. Second, run three field inspections - pre-flowering, flowering and pre-harvest - with the 0.10% off-type limit applied at each, and rogue diseased plants at the first inspection, not the last. Third, keep a distinct nucleus block with isolation distance from other wheat and from any rust-susceptible variety. Fourth, attach a rust-risk advisory to the release package telling the seed chain which districts should not multiply the variety. Fifth, retain a sample of each breeder seed lot for a post-release verification grow-out. Off-type limits protect varietal identity; they do nothing about a susceptibility that is fixed in the genome.