August 27th, 2020

1. Introduction to the problem

Crop by region

ICRISAT-Sorghum Sudan

Problem specification

Cycle time is relatively long in Sudan’s pipeline and could see some advantages of being faster and applying additional family selection.

Breeding strategy component tackled

L

Breeders’ equation terms tackled

\(\Delta_g = (i * \sigma_g * r)/L\)

Hypothesis

Recycling at earlier generations can increase the rate of gain.

2. Materials and methods

Treatments

Treatment Description
PERSE_AYT Current line breeding scheme, selecting parents from AYT (~20), using a base index.
PERSE_PYT Recycling parents from PYT (~20), using a base index.
PERSE_PYT_AYT Recycling parents from PYT (~10) + AYT (~10), using a base index.
PERSE_PYT_AYT_FAM Recycling parents from PYT (~10) + AYT (~10), adding family selection at F4 testing, using a base index.
PERSE_PYT_AYT_FAMacc Recycling parents from PYT_F5 (~10) + AYT_F6 (~10), adding family selection at F4 testing but derive panicle at F3, headrows at F4, using a base index.

Simulation procedure

A 20 year burn-in period was modeled using the line breeding schemes. The burn-in was followed by a 30 year evaluation period to measure rates of genetic gain for all treatments. Genetic gain was measured by assessing changes in genetic merit in on Farm lines. Genotype-by-year interaction variance was assumed to be equivalent to genetic variance (based on average correlation between locations being equal to 0.5). 50 replications done. We simulated 1 complex, 3 intermediate complexity traits to be behind the genetic merit.

3.1 Results: Is there an increase by cycling faster?

Based on theory we expect some difference as time evolves. By year 30, the difference are:

Recycling at PYT produces 1 (95% CI: NaN,NaN) times more gain than recycling at AYT. Recycling at PYT & AYT produces 1.06 (95% CI: 1.04,1.08) times more gain. Recycling at PYT & AYT + family selection at F4 produces 1.07 (95% CI: 1.05,1.09) times more gain. Deriving lines earlier produces 1.16 (95% CI: 1.13,1.18) times more gain.

3.2 Results: Are all trait genetic values increasing?

All traits increase depending on the weight. Base on the index we expect this.

SI = 0.2EndospermTexture + 0.25StoverYield + 0.15StayGreen + 0.4GrainYield

4. Conclusion

Given the current levels of accuracy and testing we recommend recycling parents at both stages. If possible, remove one selfing generation to accelerate the cycle time.