Simple deterministic modeling can guide the design of breeding pipelines for self-pollinated crops

Genetic gains delivered by plant breeding programs in the developing world must increase to deliver adaptation to a changing climate. Programs should be planned and optimized, with respect to gains per year and per dollar invested, as cyclic population improvement pipelines whose selection response is described by the breeder's equation (BE). We present a spreadsheet-implemented model for the BE that includes operating costs. The model can be easily used to compare predicted selection response per cycle for different pipeline designs and resource allocation scenarios for self-pollinated crops or hybrid crops that use inbred lines as parents. We modeled breeding programs that generate and select among F-2- to F-6-derived or doubled-haploid (DH) lines, advance populations from one to three generations annually, and conduct agronomic testing in one to three stages during the main cropping season. Reducing cycle time is the most efficient method of increasing genetic gain under most circumstances. Although gains increase with increased population size and selection intensity, clear optima exist, in terms of the cost of a unit of gain, at relatively small population sizes and moderate selection intensities. Investments in reducing cycle length to three or even two years generate large increases in genetic gain and are more cost-effective than increasing population size and standardized selection differential. High rates of genetic gain can be delivered by small breeding programs that complete cycles quickly.