A civil war is being fought in Darfur in western Sudan that has resulted in a humanitarian disaster with large numbers of Sudanese refugees seeking a safe haven in neighbouring Chad. The conflict is largely ecological in origin and is based on competition for natural resources, including plant genetic resources. Western Sudan has relatively low rainfall and supports rainfed sedentary agriculture and nomadic pastoralism, and a range of domesticated plant species is grown under irrigation by seasonal watercourses. Nomadic pastoralist populations move their herds according to established routes of seasonally available range and pasture species. There has always been a delicate balance between the sedentary and nomadic ways of life that through careful use of water and plant genetic resources has maintained the ecology of the region in balance. Drought, deforestation, overpopulation and overgrazing have accelerated desert encroachment and have exerted huge pressures on the ecology of the region. Indigenous plant genetic resources, including many wild species used as famine foods, are the key to livelihoods in the region. The upsurge in violence has resulted in many farmers abandoning their farms and seed stocks. Social systems are disrupted and rehabilitation of agriculture will be very difficult. This article describes domesticated and wild plant genetic resources of southern Darfur and details their importance to livelihoods in the region.

As a member of the tribe Brassiceae, Eruca sativa, although a minor crop worldwide, is considered a valuable genetic resource for cabbage, rapeseed and other Brassica crops. Self-incompatibility (SI) in Brassica has been extensively studied, but information on SI in E. sativa is limited. Of six chemicals used to treat the stigmas to overcome SI in five E. sativa lines, gibberellin was the most effective. As gibberellin is well known for its ability to break dormancy and to promote cell elongation, its effectiveness may help to understand the mechanisms of SI. Urea and ammonium sulphate were also effective. These two chemicals are known to affect protein stability, which may help explain their effects on SI. Although table salt has been reported as being effective in overcoming SI in B. rapa, B. oleracea and B. napu, it was not effective in E. sativa. Sucrose and alcohol also had negligible effect. There was significant variation among the genotypes in SI intensity and response to chemicals, but the genotype–chemical interaction was not significant. The data presented in this paper add to our understanding of SI in E. sativa and may lead to a better use of this genetic resource.

The Sorghum genus is extremely diverse both morphologically and geographically, however, relatively few of the 25 recognized species have been evaluated genetically. The apparent lack of basic knowledge pertaining to the levels of genetic diversity both within and between the 17 Australian wild species is a major obstacle to both their effective conservation and potential use in breeding programmes. Twelve Sorghum bicolor-derived simple sequence repeat (SSR) markers were evaluated for cross-species amplification in all 25 Sorghum species. The SSR markers were highly polymorphic, with diversity indices ranging from 0.59 to 0.99 with mean of 0.91. Five markers combined were able to differentiate 24 of the 25 Sorghum species, with intra-species polymorphism apparent. Sorghum bicolor-derived SSRs have proven to be an efficient source of markers for genetic diversity studies of the relatively poorly characterized Australian indigenous Sorghum species.

In order to describe the phenotypic diversity in a core collection of grasspea, 47 progenies of single plants (pure lines) selected from 25 accessions were evaluated. Sixteen morpho-agronomic characters including seed 3-(β-N-oxalyl)-l-2,3 diaminopropionic acid (β-ODAP) content were studied. Univariate and multivariate analyses (principal component analysis and cluster analysis) were performed to estimate differences between progenies. Some high-yielding lines could be used directly for cultivation as new varieties and/or as parental lines in crossing programmes. Seed β-ODAP concentration varied from 0.24% to 0.64%. However, a medium to high seed β-ODAP level was recorded for most of the lines, and no significant correlation was observed between this and other morphological traits.

The phenotypic diversity of 32 landrace populations of tetraploid wheats originating from the Bale and Wello regions of Ethiopia was studied. Eight heritable qualitative traits (glume hairiness, glume colour, awnedness, awn length, beak awn, awn colour, spike density, seed colour) were measured on 2453 individual plants (45–110 plants per landrace). The frequencies of each phenotypic class were used to estimate and analyse the diversity at different levels (population, altitude, region). Beak awn and seed colour showed the highest diversity index, and glume hairiness the lowest. Glume hairiness, glume colour, beak awn and awn colour were regionally variable, while gradients across altitude were observed for glume hairiness, glume colour and beak awn. Variation was, however, largely due to the differences in the level of the different characters within populations. On a regional basis, a higher mean diversity index was observed for materials from Wello than from Bale. No drastic change in the overall diversity between these collections and those analysed in the 1970s was evident.

A major challenge for the molecular characterization and evaluation of minor grass species germplasm is the lack of sufficient DNA markers. A set of 210 simple sequence repeat (SSR) markers developed from major cereal crops (self-pollinated wheat and rice, mainly self-pollinated sorghum and out-crossing maize) were evaluated for their transferability to minor grass species (finger millet, Eleusine coracana; seashore paspalum, Paspalum vaginatum; and bermudagrass, Cynodon dactylon). In total, 412 cross-species polymorphic amplicons were identified. Over half of the primers generated reproducible cross-species or cross-genus amplicons. The transfer rate of SSR markers was correlated with the phylogenetic relationship (or genetic relatedness) of these species. The average transfer rate of genomic SSR markers was different from the average transfer rate of expressed sequence tag (EST)-SSR markers. The level of polymorphism was significantly higher among species (67%) than within species (34%), and was related to the degree of out-crossing for each species. The level of polymorphism detected within species was 57% from self-incompatible species, 39% from out-crossing species and 20% from self-pollinated species. Genomic SSRs detected a higher level of polymorphism than EST-SSRs. The use of transferred polymorphic SSR markers for the characterization and evaluation of germplasm is discussed.

Fifty lines of Medicago truncatula, derived from ecotypes or cultivars of diverse geographical origin, were grown under uniform conditions, and variation in seed protein composition and quantity was investigated. One-dimensional electrophoretic profiles revealed 46 major seed polypeptides, of which 26 were polymorphic within the collection. The vicilin/convicilin (7S) and the legumin (11S) type proteins were identified by immunoblotting using antibodies raised against the homologous pea proteins. The polymorphism for the major seed protein classes allowed the clustering of the genotypes into four groups. There was no evidence of clustering according to geographical origin of the lines. However, all lines not belonging to either M. truncatula ssp. truncatula or ssp. longispina were clustered in a single group, demonstrating the value of seed protein profiles in delimiting species boundaries. The Jemalong line group was differentiated early in the dendrogram, and thus represents an ancient clade in the seed diversity of M. truncatula. Within-accession variation was investigated for one-dimensional seed profiles, with additional lines obtained from the same ecotypes. As expected for an autogamous species, within-accession variation was low. Seed protein content was highly variable among the 50 lines examined. Lines contrasting for qualitative traits and seed protein content were identified to allow for the genetic determination of these characters.

We successfully generated clear amplified fragment length polymorphism (AFLP) profiles of Anemone coronaria using only a single restriction enzyme (EcoR I). Four primer combinations generated 571 bands in 37 anemone accessions with an average 142.75 bands per primer combination. Polymorphism among the 37 samples was very high (98.15%). This high polymorphism showed that the one-enzyme AFLP method is effective for the assessment of genetic diversity as well as cultivar differentiation of anemone. We observed 26 bands which differentiated between fingerprints derived from leaf and corm DNA of eight plant individuals. However, the low percentage of such bands showed that anemone corms can also be used to evaluate anemone germplasm using AFLP markers.