CONSTRUCTION OF A GENETIC MAP FOR ARABICA COFFEE

Authors: PEARL, HEATHER - HARC; NAGAI, CHIFUMI - HARC; Moore, Paul; STEIGER, DENISE - HARC; OSGOOD, ROBERT - HARC; MING, RAY - HARC

Submitted to: Journal of Theoretical and Applied Genetics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/10/2003
Publication Date: 11/5/2004
Citation: Pearl, H., Nagai, C., Moore, P.H., Steiger, D., Osgood, R., Ming, R.. Construction of a genetic map for arabica coffee. Journal of Theoretical and Applied Genetics. 108:829-835. 2004.

Interpretive Summary: Coffee is an important agricultural commodity with a world-wide production of more than 16.6 billion pounds annually priced from $0.35/lb for robusta coffee beans to $0.60/lb for Arabica beans. There are numerous production and quality problems with both types of coffee but the genetic tools are not as well developed for Arabica because of its narrow tetraploid germplasm base. Genetic maps, needed for a scientifically directed crop improvement programs, have not existed prior to the one that we now report. It is expected that the strategy we used for developing this genetic map will allow us to develop markers linked to quantitative traits contributing to yield and that it will become a framework for mapping agronomically important loci for plant breeder production of improved cultivars.

Technical Abstract: Molecular marker linkage maps are being developed as scaffolds for phenotype mapping in many crop plants to assist directed germplasm improvement through marker-assisted technologies. We are using amplified fragment length polymorphisms (AFLPs) to construct a genetic linkage map on a pseudo F2 population of arabica coffee (Coffea arabica L.) derived from a cross between the cultivars Mokka hybrid and Catimor. Sixty trees from this population were selected on the basis of plant height distribution to construct a linkage map. A total of 456 dominant markers and eight co-dominant markers were generated from 288 AFLP primer combinations. 68% of the total number of markers generated were from Catimor, 30% from Mokka hybrid, and 2% co-dominant. This distribution suggests that the heterozygosity within the Catimor sub-genome was twice that as within the Mokka hybrid sub-genome. Chi-square analysis revealed that 188 markers (41%) fit a 1:1 segregation ratio and 160 (34%) fit a 3:1 ratio. The remaining 117 (25%) showed segregation distortion within the population. Markers segregating in a 1:1 or 3:1 ratio were from heterozygous loci in one or both sub-genomes. Linkage groups were constructed using Mapmaker v.3.0 resulting in 16 major linkage groups, containing 4 to 21 markers, and 15 small linkage groups, consisting of 2 to 3 linked markers each. The total length of the map was 1802.8 cM with an average distance of 10.2 cM between adjacent markers. This genetic map will serve as the framework for mapping quantitative trait loci (QTL) controlling source-sink traits in the same population.