Kuwait-Backed Breeding Plan Sets 1.5% Genetic-Gain Target for Arab Drylands
A new Kuwait-backed programme aims to move climate-resilient crop breeding in the Arab region from scattered research projects toward a shared delivery system, with an unusually explicit performance target: annual genetic gains above 1.5%. The target matters because it can be measured. It is also easy to misread. Genetic gain describes improvement in selected breeding material over successive cycles; it is not a promise that farmers’ yields will rise by the same percentage each year.
The Kuwait Fund for Arab Economic Development and the International Center for Agricultural Research in the Dry Areas, or ICARDA, signed the BredMod 2.0 grant agreement in Kuwait City on 21 October 2026. ICARDA says it will implement the programme with the Kuwait Institute for Scientific Research and national research institutions in multiple Arab countries. The plan combines genomics, artificial intelligence, accelerated breeding, field and laboratory testing, plant physiology and research into roots and soil microbiomes.
Its stated outputs are climate-resilient germplasm for seven priority crops, breeding programmes capable of exceeding 1.5% annual genetic gain, and stronger local capacity to test and release varieties adapted to drought and heat. The announcement did not name the seven crops, identify the participating countries, disclose the grant value or provide a country-by-country delivery schedule. Those omissions limit what can yet be said about scale.
Illustration: Agrilinkage original editorial artwork. It visualises accelerated dryland-crop breeding and does not depict a specific facility, researcher or signing event.
What the 1.5% target would mean
If a breeding population improved at exactly 1.5% a year and gains compounded, the cumulative change would be about 4.6% after three years: (1.015³ − 1) × 100. That is a useful benchmark for programme management, but it should not be converted directly into a harvest forecast. Farm output also depends on seed multiplication, variety registration, rainfall, irrigation, soil fertility, pests, management and whether farmers can obtain the seed at an affordable price.
The target is therefore best read as a test of the breeding pipeline. A credible evaluation would need a declared baseline, trait-specific measures, comparable multi-location trials and a clear distinction between genetic gain under controlled selection and realised gain in farmers’ fields. Yield is only one possible target; heat tolerance, maturity, disease resistance, grain quality and stability across environments may be equally important.
Speed breeding can remove years, not every bottleneck
ICARDA says the programme will draw on experience from its African Breeding Accelerator in Morocco. The centre says conventional variety development can take as long as 12 years and that accelerated methods can save four to six years. Such methods can advance more plant generations per year and combine them with genomic selection, controlled environments and faster phenotyping.
That does not mean a finished variety can automatically move from a cross to widespread cultivation in a few seasons. Candidate lines still need multi-environment evaluation, national performance testing, registration where required, breeder- and foundation-seed multiplication, commercial or public seed production, and distribution. A fast breeding cycle is valuable precisely because it gives those downstream steps a better chance of starting earlier; it does not eliminate them.
A separate ICARDA description of the Morocco accelerator lists wheat, barley, faba bean, lentil, grass pea, chickpea and cactus among its focus crops. That list shows the centre’s dryland portfolio, but it should not be treated as the confirmed BredMod 2.0 crop list until the new programme publishes its own selection.
Why regional coordination could be the real asset
The technology is only one part of the proposition. National breeding programmes often hold local germplasm, trial networks and knowledge of farmer preferences, while international centres provide wider genetic resources, specialist laboratories and cross-border data. Connecting those capabilities can reduce duplicated work and allow candidate material to be tested across contrasting heat, drought, soil and disease environments.
That regional design could also make results more transferable. A line that performs well at one site may fail under a different combination of temperature, water stress and planting date. Coordinated trials can reveal whether a trait is broadly stable or narrowly adapted. For seed-system planning, that distinction is as important as the headline yield result.
The Kuwait Fund’s role is relevant beyond the grant itself. The development institution says agriculture and irrigation are among its core financing sectors and that it provides grants and technical assistance as well as concessional loans. BredMod 2.0 places agricultural research capacity inside that development-finance mandate, rather than treating breeding as a stand-alone scientific exercise.
The announcement leaves a timetable question
ICARDA’s 21 October release contains an internal timing inconsistency. It first calls BredMod 2.0 a two-year initiative, then says it is designed to deliver results within three years. Agrilinkage has therefore not assigned a definitive end date. The project partners should clarify whether the grant period, implementation period and results window are different, or whether one of the statements is an error.
The next useful disclosures would be the grant amount, participating countries, seven crops, baseline breeding-cycle lengths, target traits, trial network, release milestones and seed-multiplication plan. Without those details, the agreement is a significant regional research commitment, but not yet evidence of varieties delivered to farmers.
What to watch
Three indicators will show whether the programme is moving from institutional ambition to practical delivery. First, partners should publish a crop-and-country work plan with baseline and target genetic gains. Second, candidate lines should enter transparent multi-location trials with results reported by trait and environment. Third, national seed systems should begin preparing multiplication and release pathways before the breeding work is complete.
If those elements appear, BredMod 2.0 could become a useful test of whether AI, genomics and speed breeding can be organised as a regional public-research system for dryland crops. If they do not, the 1.5% target will remain difficult for outsiders—and ultimately farmers—to verify.
Comments