Cultivated common bean germplasm is especially diverse due to the parallel domestication of two genepools in the Mesoamerican and Andean centers of diversity and introgression between these gene pools. Classification into morphological races has helped to provide a framework for utilization of this cultivated germplasm. Meanwhile, core collections along with molecular markers are useful tools for organizing and analyzing representative sets of these genotypes. In this study, we evaluated 604 accessions from the CIAT core germplasm collection representing wide genetic variability from both primary and secondary centers of diversity with a newly developed, fluorescent microsatellite marker set of 36 genomic and gene-based SSRs to determine molecular diversity and with seed protein analysis to determine phaseolin alleles. The entire collection could be divided into two genepools and five predominant races with the division between the Mesoamerica race and the Durango-Jalisco group showing strong support within the Mesoamerican genepool and the Nueva Granada and Peru races showing less diversity overall and some between-group admixture within the Andean genepool. The Chile race could not be distinguished within the Andean genepool but there was support for the Guatemala race within the Mesoamerican genepool and this race was unique in its high level of diversity and distance from other Mesoamerican races. Based on this population structure, significant associations were found between SSR loci and seed size characteristics, some on the same linkage group as the phaseolin locus, which previously had been associated with seed size, or in other regions of the genome. In conclusion, this study has shown that common bean has very significant population structure that can help guide the construction of genetic crosses that maximize diversity as well as serving as a basis for additional association studies.

Studies of diversity permit conclusions about the genetic structure of species. Systematically developed core collections are a particularly good tool for such purposes. The objective of this study was to evaluate the genetic structure between and within gene pools of a core collection of wild Phaseolus vulgaris L. using the technique of amplified restriction polymorphism (AFLP) analysis. DNA fingerprint patterns of 114 genotypes were analyzed by the NTSYS program. Major groups or gene pools were recognized in Mesoamerica, Colombia, the northern Andes of Ecuador and northern Peru, and the southern Andes, although the separation among gene pools was not wide. A unique segment of germplasm from northern Peru showed greater genetic distance than the other gene pools. Although most Mexican wild beans formed a rather homogenous group, a small number of accessions clustered with the Andean gene pool, and accessions from Guatemala tended to cluster apart from Mexican accessions. In the southern Andean gene pool, more discrete groups were formed which were associated with certain regions such as Argentina, possibly reflecting greater geographic and genetic isolation. Colombian accessions appeared to be highly introgressed with germplasm from other areas. AFLP analysis produced a large amount of data in a short period, thus permitting greater insights into the genetic structure of wild beans than had been possible with other methods of analysis.

No abstract

Microsatellite amplification was performed on cassava (Manihot esculenta) and six other different species (all wild) of the Manihot genus. We used ten pairs of microsatellite primers previously developed from cassava, detecting 124 alleles in a sample of 121 accessions of the seven species. The number of alleles per locus ranged from four to 21 alleles, and allelic diversity was greater in the wild species than in cassava. Seventy-nine alleles, including unique ones, were detected in the wild species but were not found in the crop. The lower level of heterozygosity in some wild species probably resulted from a combination of fine-scale differentiation within the species and the presence of null alleles. Overall, microsatellite primers worked across the genus, but, with increasing genetic distance, success in amplifying loci tended to decrease. No accession of M. aesculifolia, M. carthaginensis, and M. brachyloba presented a banding pattern at locus Ga-140; neither did one appear for M. aesculifolia at locus Ga-13. Previous work with amplified fragment length polymorphism (AFLP) markers and this microsatellite analysis show that these three wild taxa are the most distant relatives of the crop, whereas the wild forms M. esculenta subsp. flabellifolia and M. esculenta subsp. peruviana appear to be the closest.

Cassava bacterial blight, caused by Xanthomonas axonopodis pv. manihotis (Xam) is a destructive disease occurring in most cassava growing-areas. Although Colombian isolates of Xam differ in DNA polymorphism and pathogenicity, no suitable host differentials have been identified to demonstrate physiological specialization. A set of 26 Xam isolates from three edaphoclimatic zones (ECZs) in Colombia was selected for inoculation on a set of 17 potential cassava differentials. Leaf inoculation and stem puncture were used in order to detect possible specific interactions between cultivars and isolates. Cultivar £ isolate interaction was highly significant (P , 0´001) after stem inoculation, but not after leaf inoculation. The stem inoculation technique was selected as a method for resistance screening of cassava cultivars for bacterial blight resistance. A highly significant interaction was also detected when cultivar behaviour was rated as area under the disease progress curve (AUDPC) after stem inoculation. Different pathotypes were defined among the 26 isolates and differential cultivars were proposed to define the pathotypic composition of Xam populations in three ECZs in Colombia. The results should help to improve selection of sources of resistance to cassava bacterial blight.

No abstract

Despite the worldwide importance of cultivated cassava (M. esculenta Crantz) its origin and taxonomic relationships with other species in the genus have not been clearly established. We evaluated a representative sample of the crop's diversity and six wild taxa with AFLPs to estimate genetic relationships within the genus. Groupings of accessions of each species by data analysis corresponded largely with their previous taxonomic classifications. A mixed group, consisting of Manihot esculenta subsp. flabellifolia and M. esculenta subsp. peruviana, was most similar to cassava, while M. aesculifolia, M. brachyloba, and M. carthaginensis were more distant. Species-specific markers, which may be useful in germ-plasm classification or introgression studies, were suggested by the unique presence of AFLP products in samples of each of the three wild species. Heterogeneity of similarities among individuals of certain species suggested the existence of intraspecific gene pools, a hypothesis that was supported by morphological or ecogeographic evidence with varying degrees of success. Quantitative assessment of genetic diversity revealed greater homogeneity among cassava accessions than among its closest wild relatives. The demonstration of unique genetic diversity in the two M. esculenta subspecies and their genetic similarity to the crop supports the hypothesis that these materials may be the ancestors of cassava.

Cassava bacterial blight (CBB), caused by Xanthoinonas axonopodis pv. manihotis (Xam), is a major disease of cassava (Manihot esculenta Crantz) in Africa and South America. Planting resistant varieties is the preferred method of disease control. Recent genetic mapping of an F, cross (TMS 30572 x CM 2177-2) led to the construction of the first molecular genetic map of cassava. To better understand the genetics of resistance to CBB, we evaluated individuals of the F, cross for CBB resistance by controlled greenhouse inoculations and visually assessed symptoms on days 7, 15, and 30 days after inoculation, using a scale where O = no disease and 5 = maximum susceptibility. Five Xam strains were used: CIO-84, CIO-1, CIO-136, CIO-295, and ORST X-27. Area under the disease progress curve (AUDPC) was used as a quantitative measure of resistance in QTL analysis by single-marker regression. Based on the AUDPC values, eight QTLs (quantitative trait loci), located on linkage groups B, D, L, N, and X of the femalederived framework map, were found to explain 9-20% of the phenotypic variance of the crop's response to the five Xain strains. With the male-derived framework map, four QTLs on linkage groups G and C explained 10.7-27.1% of the variance. A scheme to confirm the usefulness of these markers in evaluating segregating populations for resistance to CBB is proposed.