Cowpea is a drought-tolerant food legume grown in the savannah regions of the tropics and subtropics. The International Institute for Tropical Agriculture (IITA) holds the world collection of 15,003 cultivated cowpea from 89 countries in its genebank. In excess of 12,000 accessions were characterized for 28 agrobotanical descriptors. The entire collection was first stratified by country of origin and biological status. Land race accessions (10,227) with information on origin and characterization data were grouped using clustering procedures. The clustering analysis was based on Euclidean distances between and among accessions. Accessions within each country were then grouped hierarchically, according to their similarity. The number of clusters selected for countries varied from 2 to 20, depending on the size of collection for that country. A percentage number of accessions (5–25%) was chosen from each country, based on the size of the collection and its proximity to the centre of diversity. The number of accessions from each cluster in a country was then chosen randomly, based on the predetermined percentage. In countries where the numbers of accessions were few ( < 10), the percentage chosen from those countries would be higher and may go up to 100% to ensure that at least one accession was chosen from every country. Accessions with no characterization information were treated as a group, and accessions were chosen randomly, based on their geographical distribution. In the process of selection, 200 accessions that are known to be resistant to pests and diseases and not originally chosen through the selection processes were, nevertheless, kept as part of the core collection. Following these procedures, a total of 1701 accessions of landraces were chosen from the entire collection. Using the same procedures, a subset of 225 accessions was chosen from 1422 advanced cultivars and breeding or research lines. A third subset of 130 accessions was chosen from 838 accessions that either lack information on origin or biological status, and six accessions of wild and weedy forms from the available 64 wild and weed accessions were added. Thus a core collection of 2062 accessions of cowpea was constituted. The diversity in the core collection was similar to that of the entire collection and correlated traits that may be linked were also preserved in the core collection. This core collection of cowpea provides an opportunity for further exploitation of the cowpea germplasm for improvement of this crop.

The landraces of rice (Oryza sativa L.) possess wide diversity, which needs to be properly characterized for their use in genetic improvement. Replicated field studies were conducted in 1998, 1999 and 2000 at two sites in Nepal to determine diversity in 183 landraces of rice adapted to the lowlands and the hills in Nepal. Fourteen improved genotypes were also used for comparison. Thirteen agronomic traits were investigated. Shannon–Weaver diversity index (H) and Simpson's index of diversity (D) were estimated to determine the level of genetic richness among the landraces. The landraces differed significantly for all traits. Except for plant height and maturity, at least one of the landraces compared well with the performance of improved cultivars. A principal component analysis separated the lowland- and hill-adapted landraces into two broad groups.

Spelt (Triticum spelta L.) is hulled wheat which was widely cultivated in Spain until the first half of the 20th century and is still cultivated in Asturias (northern Spain). Much of the variability of this species has been preserved, but little evaluation has been carried out. We describe the characterization of some of these spelt lines for several quantitative traits, including plant height and awnedness.

Fifty morphological characteristics, fruit production over 3 years (from 1999 to 2002) and the amplified fragment length polymorphism (AFLP) technique were used to analyse a set of 48 guava (Psidium guajava L.) accessions cultivated in Mexico, in order to characterize their genetic relationships. Germplasm was collected from the Calvillo-Cañones region and planted in Huanusco, Mexico. The study included two P. cattleianum (Sabine) and two P. friedrichsthalianum (Berg-Niedenzu) accessions from Costa Rica as outgroups. Principal component analysis (PCA) explained less than 30% of total variation and 14 characteristics from trees (1), leaves (2) and fruits (11) were the most informative. PCA analysis separated the germplasm into three major groups of accessions based on fruit size and weight, stem diameter and leaf size. Significant differences in fruit yield were detected among accessions and years, where P. guajava produced 36 kg/year/tree of fresh fruit while P. cattleianum and P. friedrichsthalianum showed fruit yield lower than 7 kg/year/tree. The fruit yield broad sense heritability was 0.25. The AFLP analysis produced two clusters of Psidium accessions, the first included P. cattleianum and P. friedrichsthalianum, and the second P. guajava accessions. This is the first report about the use of AFLP marker methodology for the genetic characterization of Mexican native guava germplasm and the results based on phenotypic and productive characteristics suggest that germplasm was selected from open pollinated trees.

Eruca vesicaria subsp. sativa (syn. E. sativa) is a cruciferous vegetable and oilseed crop that is high in erucic acid. It occurs throughout the Mediterranean region and western Asia, and has been naturalized elsewhere as a crop/weed escape. It is closely related to subsp. vesicaria and subsp. pinnatifida, which are endemic to Spain and north-western Africa, respectively. This study evaluated patterns and levels of diversity in the three subspecies based on 234 amplified fragment length polymorphisms (AFLP), and evaluated agronomic and seed quality data in a field trial in western Canada. AFLP data revealed three main clusters: ‘Sativa’ (33 accessions of subsp. sativa), ‘Vesicaria’ (nine accessions of subsp. vesicaria) and a ‘Pinnatifida’ cluster (one accession of subsp. pinnatifida and three Moroccan accessions of subsp. sativa). The Sativa cluster separated into Mediterranean and Asian groups, likely reflecting differences in origin (wild versus cultivated) or primary usage, vegetable versus seed oil. The origin of the introduced Mexican population was confirmed as subsp. sativa. The highest levels of diversity were found in the Sativa cluster (88% AFLP polymorphisms) and the least in the Vesicaria (56%) and Pinnatifida (39%) clusters. Extensive variation was observed among the 159 subsp. sativa accessions evaluated in the field trial, and overall findings indicated a favourable agronomic potential.

Molecular markers have been used to study genetic diversity within a set of Lablab purpureus accessions collected from the southern states of India. Amplified fragment length polymorphism (AFLP) molecular marker studies using a total of 78 L. purpureus accessions with nine primer combinations showed there was very little genetic diversity within the L. purpureus accessions from the southern Indian germplasm collection as compared to a set of 15 accessions from other international germplasm collections that included African accessions. The set of 15 were selected from a random amplified length polymorphism (RAPD) marker study and chosen on the basis of widest genetic distance. Further molecular analysis with polymerase chain reaction (PCR) markers from 97 expressed sequence tag (EST) and gene-specific primer pairs, designed from a range of legume sequences, concurred with the AFLP analyses. Both of these approaches provide a wealth of markers for diversity and mapping studies. The 97 sequence-specific primer pairs tested in L. purpureus resulted in 70% amplification success, with 44% of primer pairs amplifying single bands and 10% double bands. Markers generated from these EST and genomic sequences provide useful cross-reference to comparative legume genomics that will potentially have long-term benefit to legume plant breeding.