Comprehensive overview of coffee breeding programs integrating marker-assisted selection, gene pyramiding for disease resistance, QTL mapping, and international collaborative networks for Coffea arabica and Coffea canephora improvement.
Traditional coffee breeding is a slow process, requiring approximately 25 years to develop new varieties due to the long generation time (5-6 years) of this perennial plant. Marker-assisted selection accelerates the identification and concentration of target alleles, being essential for developing cultivars resistant to multiple diseases [2][3][5][9].
The development and use of resistant cultivars have proven to be the most suitable methods for sanitary control in crops, due to the cost–benefit ratio, efficacy, easy adoption by producers, as well as the low environmental impact [5][9]. In coffee improvement programs, interspecific and intraspecific crossings have been carried out to introgress resistance genes into cultivars with agronomic characteristics of commercial interest.
The importance of gene stacking is to obtain cultivars with durable multiple resistance to different pathogens as well as optimal beverage quality, high productivity, and morphoagronomic characteristics that facilitate phytotechnical management [5][9].
Marker-assisted selection reduces the number of generations required for cultivar development [2][3][5][9].
A landmark 2025 study demonstrates the power of marker-assisted selection for pyramiding resistance genes in arabica coffee [2][3][5][9].
Population with resistance alleles to CLR and CBD [2][3][5][9]
Pyramiding of 5 resistance genes [2][3][5][9]
Morphoagronomic traits evaluated over 4 years [2][3][5][9]
| Locus | Gene/Marker | Homozygous Dominant (%) | Heterozygous (%) | Recessive Homozygous (%) |
|---|---|---|---|---|
| Locus B | Resistance allele | 57.04% | 33.80% | 9.15% |
| Locus C | Resistance allele (C_) | 59.15% (C_ presence) | 40.85% | |
| Locus D | Presence | 74.65% | 25.35% | |
| Locus E | Presence | 71.13% | 28.87% | |
| Locus F | Ck-1 (CBD resistance) | 56.34% | 35.21% | 8.45% |
Resistance to leaf miner in pyramiding genotypes [2][3][5][9]
Resistance to cercospora in pyramiding genotypes [2][3][5][9]
Significant morphoagronomic traits identified [2][3][5][9]
At least nine dominant resistance genes to CLR are present in coffee plants of different species, which can act together or individually [5][9].
| Gene | Source Species | Characteristics | Current Status |
|---|---|---|---|
| SH1-SH5 | Coffea arabica | Identified in arabica coffee | Already replaced by CLR in several coffee cultivation areas [5][9] |
| SH3 | Coffea liberica | Introgression from liberica | Absent in studied F2 population [5][9] |
| SH6-SH9 | Coffea canephora | Detected in robusta coffee | Important for durable resistance |
The HdT is the only natural cross between C. arabica and C. canephora, and it possesses [5][9]:
Studies suggest the existence of two additional main resistance genes that have not yet been characterized, along with several others of lesser effect, which may or may not be associated with the genes SH1–SH9. These genes theoretically confer resistance to more than 50 races of H. vastatrix [5][9].
The CBD resistance in C. arabica is governed by three genes [5][9]:
The T and R genes are dominant while the k-gene is recessive and only confers partial resistance to CBD in a homozygous state. The R locus has been reported to have multiple alleles (R1R1) in C. arabica variety Rume Sudan [5][9].
In response to industry demand, WCR launched its robusta research and breeding efforts in 2023, with significant progress reported in 2025 [1].
Global coffee production (robusta), up from 25% in early 1990s [1]
First clones distributed for multi-location trials [1]
Confirmed partner countries: Ghana, Uganda, Vietnam [1]
WCR launched robusta research and breeding efforts in response to industry demand [1]
Controlled crosses initiated using elite parent plants; pollen collected and transferred onto emasculated flowers [1]
Seeds harvested from individual trees; seedlings transferred to specialized propagation facility [1]
Global robusta breeding network official launch [1]
Several thousand plantlets distributed to partners across robusta breeding network for six-year performance trials [1]
WCR's robusta breeding efforts are structured as a collaborative global network, modeled on the Innovea Global Coffee Breeding Network for arabica. The robusta network brings together national coffee institutes from key producing countries, each of which will contribute to and benefit from shared breeding resources, performance data, and modern breeding technologies [1].
Confirmed partner countries include Ghana, Uganda, and Vietnam. Partners will have the ability to select high-performing clones and integrate them into their own breeding programs, helping to modernize robusta production across diverse origins [1].
A 2025 USDA study investigated the genetic architecture of key traits in two Coffea canephora populations using single-SNP association analysis and machine learning [6].
| Trait | Population | Candidate Genes |
|---|---|---|
| Leaf rust incidence | Premature | RPP13-like, NB-ARC, CERK1 (plant defense mechanisms) |
| Leaf rust incidence | Intermediate | TPR_REGION-containing protein, nitrate regulatory gene2 protein |
| Coffee bean production | Premature | Alpha/beta-hydrolase superfamily protein, IPPc domain-containing protein (chromosome 6 region) |
| Green bean yield | Premature | Putative caffeine synthase 3 gene (Cc09t06990.1) [6] |
Quantitative trait loci (QTL) mapping enables the genetic dissection of complex agronomic traits in coffee [7].
| Study | Population | Markers | Map Length | QTLs Detected |
|---|---|---|---|---|
| Moncada et al. (2016) | Caturra × CCC1046 F2 (278 individuals) | 848 SSR and SNP | 3800 cM (22 LGs) | Yield, plant height, bean size QTLs [4][9] |
| C. canephora study | Populations A and B (Congolese × Guinean crosses) | 249 SSRs | 1201 cM | 143 total QTLs; 60 shared between models [7] |
Two segregating populations were used to characterize QTLs involved in agronomic and biochemical traits [7]:
Total QTLs detected: 143 total; 60 shared between models; 28 found with two models; 2, 13, and 40 specific from models I, II, and III respectively [7].
EU-funded project (2017-2022) using Arabica coffee F1 hybrids to design varieties better adapted to agroforestry systems and climate change [8].
Design and test coffee varieties, better adapted to AFS and CC, maintaining a robust defense system to biotic and abiotic stresses [8].
Comprehensive database containing data from studies on tropical plants for markers, QTLs, genotypes, phenotypes, and genetic maps [4].
(2025). Plants 14(3):391 [2][3][5][9]
Timor Hybrid × Tupi Amarelo population; 98.6% resistance alleles; 29% five-gene pyramiding; 100% leaf miner resistance; 90% cercospora resistance.
View AbstractWorld Coffee Research (2025) [1]
Global robusta breeding network; 1,000 unique individuals; 2027 trial distribution; partners: Ghana, Uganda, Vietnam.
View ReportAhn E.J., et al. (2025). Plants 14(23):3675 [6]
RPP13-like, NB-ARC, CERK1 for rust resistance; caffeine synthase gene for green bean yield; population-specific genetic architecture.
View AbstractMoncada M.P., et al. (2016). Tree Genetics & Genomes 12(1):5 [4][9]
848 SSR and SNP markers; 3800 cM map length; QTLs for yield, plant height, and bean size.
View AbstractEU H2020 Project (2017-2022) [8]
F1 hybrids in 8 countries; GxE assessment; metabolomic and transcriptomic analysis; farmer and roaster participation.
View Project(2012). INRAE/CIRAD [7]
143 QTLs detected; 60 shared between models; connected population approach more efficient for low-variance QTLs.
View AbstractPeer-reviewed sources and official reports cited in this research
* Additional references available in the complete Publications Database. All sources have been peer-reviewed and are accessible through academic databases.