Penetrance of creeping-rootedness in clonal progenies of lucerne and observations on underground morphology of plants differing for this character

Elaeis oleifera or ‘caiaué’, a close relative of oil palm (E. guineensis), has some agronomic traits of great interest for the oil palm genetic breeding such as slow growth, oil quality (mostly unsaturated) and disease resistance. An analysis of a Brazilian oil palm germplasm collection was carried out using RAPD (Random Amplified Polymorphic DNA) markers with the objective of understanding the genetic variation of ‘caiaué’ accessions collected in the Amazon Forest in the last two decades. A sample of 175 accessions obtained along the Amazon River Basin was analyzed and compared to 17 accessions of oil palm from Africa. Ninety-six RAPD markers were used in the analysis, of which fourteen were shown to be specific to oil palm, while twelve were specific to ‘caiaué’. Results showed that the Brazilian ‘caiaué’ accessions studied have moderate levels of genetic diversity as compared to oil palm accessions. The data allowed the establishment of similarity groups for ‘caiaué’ accessions, which is useful for selecting parental plants for population breeding. Cluster analysis showed that, in general, genetic similarities are not correlated to geographical distances, but consistent with geographical dispersal along the Amazon River network. AMOVA showed that most of the genetic variation is found within populations, as expected for anallogamous and long-lived perennial species. The study provides important information to define strategies for future collection expeditions, for germplasm conservation and for the use of E. oleifera in breeding programs. This revised version was published online in August 2006 with corrections to the Cover Date.     

Assessment of genetic variability of olive varieties by microsatellite and AFLP markers

Fourteen developed microsatellite markers were characterized for their use in genotyping and diversity studies of olive varieties. After optimisation of microsatellite assay and allele sizing, ninety-six alleles were found in nineteen varieties, with an average of 6.8 alleles per locus. The characteristics of the microsatellite markers were used to identify markers that can be reliably applied for variety genotyping. Such features were the generation of complex banding patterns supported by underlying allele sequences, `short allele dominance', an unstable repeat structure and a low number of alleles. AFLP analysis was performed on the same set of olive varieties using eight primer pair combinations. The genetic relationships among nineteen olive varieties were compared on the basis of microsatellite and AFLP polymorphisms. Genetic distances between all pairwise combinations of the varieties were calculated using Jaccard's coefficient of similarity and dendrograms were constructed by the UPGMA method. The results of clustering analysis with both molecular systems showed the common genetic background of Tuscan varieties, and genetic divergence within Slovene olive germplasm. Slovenian varieties ‘Buga’, ‘Štorta’ and ‘Samo’ might represent regionally selected olives, while ‘Zelenjak’ and ‘Črnica’ are probably derived from the Central Italian region. The predominant local ‘Istrska belica’ was introduced to Slovenia independently from the other regional varieties and showed the lowest genetic similarity with the other regional varieties. This revised version was published online in August 2006 with corrections to the Cover Date.     

Genetic diversity of wheat wild relatives in the Near East detected by AFLP

In order to reveal the molecular genetic diversity of wheat wild relatives, an AFLP analysis was conducted with 16 accessions of five Triticum andAegilops species originating from the Near East. Variation within population was studied with at least seven individuals per accession. Four primer combinations were used for selective amplification. Based on the scored bands, we estimated percentage of polymorphic bands, 1 – proportion of shared bands (1-psb) and nucleotide diversity (π). Of the five species used in this study, Ae. speltoides had the highest level of `within population' variation. This species had also the highest value of the variation among populations. As for Triticum species, the level of variation within population was low in diploid species (T. urartu and T. boeoticum),whereas two tetraploid species (T. dicoccoides and T. araraticum) had relatively high levels of variation within population. While the two diploid Triticum indicated a clear interspecific divergence, the two tetraploid wild wheats were not clearly divergent in this study. The variance portioning analysis indicated that the variation detected for diploid Triticum species was mainly composed of `between species' variation, on the other hand that for tetraploid Triticum was mostly composed of `within population' variation. In conclusion, AFLP analysis reveals molecular variation in all accessions used in this study, suggesting a potential genetic diversity of the wheat wild relatives in natural populations. These results have implications for the design of strategies to maintain genetic diversity within genebank collections. This revised version was published online in August 2006 with corrections to the Cover Date.     

野生榆叶梅遗传多样性的AFLP分析

本研究利用AFLP分子标记技术,对野生榆叶梅6个居群进行分析.利用已筛选出的8对引物组合进行扩增,共产生256条谱带,其中206条谱带具有遗传多态性,占总谱带的80.47％;平均每对引物组合扩增产生129条多态条带,平均多态率为50.26％.6个居群观察等位基因数Na平均值为1.502 6,总Na值为1.804 7;有...     

Molecular mapping of quantitative trait loci (QTLs) controlling aluminium tolerance in bread wheat

Aluminium (Al) toxicity is a major constraint to crop productivity in acidic soils. A quantitative trait locus (QTL) analysis was performed to identify the genetic basis of Al tolerance in the wheat cultivar ‘Chinese Spring’. A nutrient solution culture approach was undertaken with the root tolerance index (RTI) and hematoxylin staining method as parameters to assess the Al tolerance. Using a set of D genome introgression lines, a major Al tolerance QTL was located on chromosome arm 4DL, explaining 31% of the phenotypic variance present in the population. A doubled haploid population was used to map a second major Al tolerance QTL to chromosome arm 3BL. This major QTL (Qalt _CS .ipk-3B) in ‘Chinese Spring’ accounted for 49% of the phenotypic variation. Linkage of this latter QTL to SSR markers opens the possibility to apply marker-assisted selection (MAS) and pyramiding of this new QTL to improve the Al tolerance of wheat cultivars in breeding programmes.     

Genetic diversity of natural populations of Atriplex halimus L. in Morocco: An isoenzyme-based overview

Based on polyacrylamide gel electrophoresis, nine natural populations of Atriplex halimus L., a perennial shrub, collected in different regions of Morocco, were studied for their genetic variation using isoenzyme polymorphism of the highly active enzyme systems: esterases (EST), acid phosphatases (ACP) and glutamate oxaloacetate transaminase (GOT). Different allozyme frequencies from 7 different loci were obtained for all populations of this halophyte species. High levels of genetic diversity were revealed. The mean number of alleles per locus (A = 1.9–2.0), the percentage of polymorphic loci (p = 71.4–85.7) and the mean expected heterozygosity (H_e = 0.339–0.385) showed an important variability in all populations. Gene diversity was essentially explained by the within population component. The between populations differentiation accounted for 8% of the whole diversity (F_ST, averaged over all loci, is 0.08). The relationships among the 9 populations were inferred from the Nei’s genetic distances. Four major groups were formed. The northern population ‘Tanger’, forming a unique group, was highly divergent from the other groups. It appeared that the genetic distance between all groups was related to the geographic distance that separates them. This revised version was published online in August 2006 with corrections to the Cover Date.     

Genetic diversity among elite varieties of the aromatic grasses, Cymbopogon martinii

The elite and popular cultivars of Cymbopogon martinii were examined for genomic and expressed molecular diversity through RAPD, enzyme and SDS-PAGE protein polymorphisms. The allelic score at each locus of the enzymes as well as presence and absence profiling in RAPDs, overall occurrence of band types etc. were subjected to computation of gene diversity, expected heterozygosity, allele number per locus, and similarity matrix. These, in turn, provide inputs to derive primary account of allelic variability, genetic bases of the cultivated germplasm, putative need for gene/trait introgression from the wild or geographically diverse habitat etc. in elite selections. ‘PRC1’ possessed highest number of unique bands based on RAPD polymorphism. In variety ‘IW31245E’, diaphorase and glutamate oxaloacetate transaminase isozymes generated two unique bands as dia-III ² and got-II ⁴. ‘RRL(B)77’ exhibited three unique bands; one produced by esterase as allele est-II ¹ and two by malic enzyme (me-III ^1,3). Only one unique band was generated by malic enzyme in variety ‘Trishna’. But sofia had three unique bands, two contributed by diaphorase (dia-II ³ and dia-II ⁴ and one by glutamate oxaloacetate transaminase (got-II ²). SDS-PAGE analysis revealed presence of unique polypeptide fragments (97.7 kDa to 31.6 kDa) in varieties ‘IW31245E’, ‘RRL(B)77’, ‘Tripta’, ‘Trishna’, ‘PRC1’ and var. sofia, generated as a diagnostic marker. In general, molecular distinctions associated with var. motia and var. sofia have been clearly noticed in C. martinii. This revised version was published online in July 2006 with corrections to the Cover Date.     

Genetic characterization and identification of new accessions from Syria in an olive germplasm bank by means of RAPD markers

Thirty-two olive cultivar accessions from Syria, most of them obtained from collecting expeditions, were characterized by means of RAPD markers before being introduced in the World Germplasm Bank of Cordoba. A total of 79 polymorphic bands(6.1 polymorphisms per primer) out of 93(7.1 bands per primer) were scored for the13 primers used, corresponding to 84.9% of the amplification products. Thirty-one different genotypes were clearly discriminated. Differences were not found among the amplification profiles from different individuals of the same cultivar. Only two cases of mislabeling or errors of planting were found. Fourteen accessions corresponding to 6 homonyms were discriminated by RAPDs as different genotypes. The dendrogram obtained by RAPD analysis included three major groups. Some evidence of relationships of the Syrian accessions studied according to their geographic origin and/or diffusion was found. For instance, cultivars from the Central Syria (Tadmur/Palmyra)such as Toffahi', ‘Abbadi Abo Gabra’-1033,‘Abo Kanani’, ‘Shami’-1041, ‘Abbadi Shalal’ ‘Adgam’-844 and ‘Majhol’-1013 clustered in Group 1 and 2. Six cultivars from Northern Syria clustered in Group 2. But it was not found a geographic structure for the cultivars from South and West of Syria. These results agree with the hypothesis of autochthonous origin of most of the olive cultivars. Some associations between cultivars from Central Syria and their fruit size were observed. This suggests that fruit size was a criterion of local selection in olive cultivars of this area. This revised version was published online in July 2006 with corrections to the Cover Date.     

Cultivar identification and genetic map of mango (Mangifera indica)

Amplified Fragment Length Polymorphism (AFLP) information was used for identification of mango (Mangifera indica L.) cultivars, for studying the genetic relationship among 16 mango cultivars and seven mango rootstocks and for the construction of a genetic linkage map. Six AFLP primer combinations produced 204 clear bands and on the average 34 bands for each combination. The average Band-Sharing between cultivars and rootstocks was 83% and 80%, respectively. The average Band-Sharing for mango is 81%. The probability of obtaining a similar pattern for two different mango cultivars and rootstocks is 6 × 10⁻³and 2 × 10⁻³, respectively. A preliminary genetic linkage map of the mango genome was constructed, based on the progeny of a cross between ‘Keitt’ and ‘Tommy-Atkins’. This linkage map consists of 13 linkage groups and covers 161.5 cm defined by 34 AFLP markers. This revised version was published online in August 2006 with corrections to the Cover Date.     

Environmental, phenotypic and genetic variation of wild barley (Hordeum spontaneum) from Israel

Wild relatives of crop plants offer an attractive gene pool for cultivar improvement. We evaluated genetic and phenotypic variation for a set of 72 Israeli accessions of wild barley from 21 populations. These populations were grouped further into four ecotypes. In addition, environmental variables describing the local conditions for the populations were used to infer the environmental divergence. Genetic, phenotypic and environmental distances were estimated from the data and UPGMA dendrograms constructed. The results showed that genetic variation was larger between populations than within them, whereas for phenotypic measurements variation was larger within populations than between them. No significant correlation was found between genetic and phenotypic similarities, or between phenotypic and environmental similarities, whereas a weak correlation between genetic and environmental similarities was detected. Twenty-three AFLP-markers were identified to be ecotype specific. Chromosomal location was known for five of these markers. Four of the five ecotype specific markers were correlated with both phenotypic traits and environmental variables.     

Bulked AFLP analysis for the assessment of genetic diversity in white clover (Trifolium repens L.)

The use of bulked leaf samples from individual plants for amplified fragment length polymorphism (AFLP) analysis was evaluated as a tool for assessment of genetic diversity in white clover (Trifolium repens L.). Bulking of leaf samples produced slightly simpler AFLP profiles compared to the combined profiles of individual plants from the same cultivar. Approximately 90% of bands which were present in individual plants were present in bulked samples of the same cultivar. The majority of those absent were rare bands, shared by less than 25% of individual plants. Replicate bulk samples gave almost identical banding patterns, demonstrating the robustness of the bulked AFLP technique. Cluster analysis of AFLP data derived from individual plants resulted in a phenogram similar to that produced from data derived from bulked samples of the same plants. AFLP analysis of bulked samples detected significant amounts of genetic variability among 52 cultivars and accessions with genetic similarity values ranging from 0.42 to 0.92. However, cluster analysis of AFLP data only partially reflected the geographic origin of cultivars and accessions and was not congruent with cluster analysis based on variation for morphophysiological characters. Bulked AFLP analysis provides a powerful tool for rapid assessment of genetic variability in white clover and may also be used for cultivar identification. This revised version was published online in August 2006 with corrections to the Cover Date.     

Chromosomal regions associated with green plant regeneration in wheat (Triticum aestivum L.) anther culture

Genetic capacity for green plant regeneration in anther culture were mapped in a population comprising 50 doubled haploid lines from a cross between two wheat varieties ‘Ciano’ and ‘Walter’ with widely different capacity for green plant regeneration. Bulked segregant analysis with AFLP markers and composite interval mapping detected four QTLs for green plant percentage on chromosomes 2AL (QGpp.kvl-2A), 2BL (QGpp.kvl-2B.1 and QGpp.kvl-2B.2) and 5BL (QGpp.kvl-5B).The three QTLs detected on chromosome 2AL and 2BL all derived their alleles favouring green plant formation from the responsive parent ‘Ciano’.The remaining QTL on chromosome 5BL had the allele favouring green plants from the low responding parent ‘Walter’. In a multiple regression analysis the four QTLs could explain a total of 80% of the genotypic variation for green plant percentage. None of the chromosomal regions with QTLs for green plant percentage showed significant influence on either embryo formation or regeneration frequencies from the anther culture. The three major QTLs located on group two chromosomes were fixed in a second DH population derived from two parents ‘Ciano’ and ‘Benoist’,both with high capacity to produce green plants. A QTL explaining31.5% of the genetic variation for green plant formation were detected on chromosome 5BL in this cross as well. This revised version was published online in July 2006 with corrections to the Cover Date.     

Yield and disease resistance of plantain (Musa spp., AAB group) somaclones in Nigeria

This study examined the potential of somaclonal variation for the improvement of plantain. Approximately 500 somaclones each of ‘Agbagba’ (False Horn plantain) and ‘Bise Egome’ (French plantain) were field evaluated for their agronomic performance and response to the black Sigatoka disease. The micropropagated populations were independently generated from a number of suckers from each accession. Significant differences between micropropagated accessions and crop cycles were observed. Differences between plants derived from suckers of the same accession were also expressed, indicating the chimerical nature of variation in the traits studied. None of the plants of the micropropagated populations from ‘Bise Egome’ exhibited significantly better disease tolerance and agronomic performance than the source accession. However,one somaclonal variant of ‘Agbagba’ (‘AO 2B2-2’) expressed lower susceptibility to the black Sigatoka disease. Compared to ‘Agbagba’, ‘AO 2B2-2’ had a higher bunch weight, more fruits per bunch with higher average weight, greater average length, and greater average girth. These data clearly show that, in contrast to previous reports, it is possible to recover superior somaclonal mutants in Musa. This revised version was published online in July 2006 with corrections to the Cover Date.     

Pear mutagenesis: In vitro treatment with gamma-rays and field selection for productivity and fruit traits

In vitro shoots of six pear (Pyrus communis L.) cultivars, ‘Conference’, ‘Doyennéd'Hiver’, ‘Passe Crassane’, ‘Bartlett’, ‘AbbéFetel’ and ‘Butirra Precoce Morettini’ were irradiated with gamma rays (3.5 Gy). After three subcultures, microcuttings from the irradiated shoots and from additional non-irradiated microcuttings were rooted to establish plants for survey orchards. All trees were individually observed for variation in fruit traits and for productivity. Trees were selected for improved characters related to production such as early bearing and consistent productivity. Variations observed in fruit appearance concerned degree of russeting, fruit shape and size. The frequencies of the observed variations in fruit traits depended on the cultivar, ranging from 0.81% in ‘Doyennéd'Hiver’ to 3.64% in ‘Passe Crassane’. Of the 97 variants selected, only two showed chimeral behavior. This revised version was published online in July 2006 with corrections to the Cover Date.     

Marker selection for <Emphasis Type="Italic">Fusarium</Emphasis> head blight resistance based on quantitative trait loci (QTL) from two European sources compared to phenotypic selection in winter wheat

Fusarium head blight (FHB) infects all cereals including maize and is considered a major wheat disease, causing yield losses and mycotoxin contamination. This study aimed to compare the realized selection gain from marker and phenotypic selection in European winter wheat. A double cross (DC) combined three FHB resistance donor-QTL alleles (Qfhs.lfl-6AL and Qfhs.lfl-7BS from ‘Dream’, and one QTL on chromosome 2BL from ‘G16-92’) with two high yielding, susceptible winter wheats, ‘Brando’ and ‘LP235.1’. The base population of 600 DC derived F₁ lines was on one hand selected for the respective QTLs by SSR markers (marker-selected cycle, CM), resulting in 35 progeny possessing different combinations of beneficial donor-QTL alleles. On the other hand it was selected phenotypically, only by FHB rating, and the best 20 lines were recombined and selfed (phenotypically selected cycle, CP). The variants CP, CM, and an unselected variant (C0) were tested at four locations by inoculation of Fusarium culmorum. Resistance was measured as the mean of multiple FHB ratings (0–100%). FHB severity was reduced through both phenotypic and marker selection by 6.2 vs. 5.0%, respectively. On a per-year basis, marker selection by 2.5% was slightly superior to phenotypic selection with 2.1%, because the first variant saved 1 year. Marker-selected lines were on average 8.6 cm taller than phenotypically selected lines. A high genetic variation within the marker-selected variant for FHB resistance and the high effect of a resistance-QTL allele on straw length indicate that additional phenotypic selection will further enhance selection gain.     

Genetic diversity evolution through participatory maize breeding in Portugal

Natural, and in particular, artificial (human) selection may pose a danger to the existing crop genetic diversity. Nevertheless, on-farm breeding systems seem to achieve phenotypic improvements even though preserving variability. Using SSR markers, we analysed several selection cycles, over a 20 years period, of a Portuguese on-farm participatory maize OPV-‘Pigarro’ breeding project. No significant differences in allelic richness (N_ar), observed heterozygosity (H_O), expected heterozygosity (or gene diversity; H_E) or inbreeding coefficient (f) were detected among the selection cycles. 58 out of 107 alleles were common to all the selection cycles studied. The analysis of molecular variance showed that the variation among selection cycles represented only 7% of the total molecular variation. However, the number of private alleles varied among the selection cycles, being the highest detected at the beginning of the selection project. These findings demonstrate that an allele flow took place during the on-farm selection process of ‘Pigarro’ but the level of genetic diversity was not significantly influenced. Since interesting phenotypic improvements were also achieved, on-farm breeding projects, like this one, should be valued as a way to preserve unique Portuguese maize landraces in risk of disappearing.     

Heritability and gain from selection for field resistance against multiple root-knot nematode species (Meloidogyne incognita race 1 and M. javanica) in carrot

Heritability and gain from selection of traits associated with field resistance to multiple root-knot nematode species were estimated in carrot (Daucus carota L.). Experimental plots were uniformly and simultaneously infested with Meloidogyne incognita race 1 and M. javanica. Forty-seven half-sib families derived from the cv. ‘Brasília’ were evaluated for: (1) percentage of commercial roots with non-galling symptoms (%RNG); (2) percentage of commercial roots with gall symptoms (%RGS) and (3) percentage of non-commercial roots (%NCR). The cultivars ‘Brasília’ (resistant), ‘Carandaí’ (intermediate) and ‘Nova Kuroda’ (susceptible) were employed as standards. Broad-sense heritability estimates were 61.9 % for %RNG; 30.6 % for %RGS and 67.9 % for %NCR. However, the gains from selection were very small for all traits. The genotypic correlation between %RGS and %NCR was –0.38 and the correlation between %RNG and %RGS was 0.13. Selection on increasing %RNG resulted in a negative correlated response on %NCR (genotypic correlation = –0.99) indicating repulsion phase linkage(s) between the genetic factor(s) controlling these traits. The phenotypic correlation of –0.88 between %RNG and %NCR suggested that incomplete penetrance and dosage effects might be involved in the expression of these traits. Selection of superior genotypes (i.e. high %RNG, low %RGS, and low %NCR) can be achieved. ‘Brasília’ was among the genotypes selected for the following cycle of recombination reinforcing the notion that this cultivar is one of the most promising sources of stable, wide-spectrum field resistance to Meloidogyne species in D. carota. This revised version was published online in July 2006 with corrections to the Cover Date.     

Genetic variation in South Korean natural populations of wild soybean (Glycine soja)

Summary Wild relatives are valuable genetic resources for crop improvement. Evaluating genetic variation in these species is not only important for their use in breeding programs, but will also provide information about evolution of crops. Seeds representing six natural populations were used to study the level of variation in the South Korean wild soybean. Electrophoretic assays of the seeds on horizontal slab gels were conducted to determine the genotypes of each natural plant at 35 loci in 17 isozymes and one protein. The results indicated a surprisingly high variation. The number of alleles at each locus was as high as four. Seventy two of the 94 reported alleles for the 35 loci were present in these populations. The average number of alleles per locus, 99% polymorphism and the expected heterozygosity in the total population were 2.1, 77.1% and 0.215, respectively. This amount of variation was not only higher than that reported for 857 soybean cultivars and wild soybean populations from other geographic regions, but also higher than the average for 123 self-fertilized plant species and 473 plant species of all mating systems. The high variation in the South Korean wild soybean as well as cultivated soybean indicated in this and other population genetic studies prompts us to propose that South Korea is one of the major soybean gene centers.     

Characterization of segregation distortion on chromosome 3 induced in wide hybridization between indica and japonica type rice varieties

We previously surveyed chromosomal regions showing segregation distortion of RFLP markers in the F₂ population from the cross between a japonica type variety ‘Nipponbare’ and an indica type variety ‘Milyang23’, and showed that the most skewed segregation appeared on the short arm of chromosome 3. By comparison with the marker loci where distortion factors were previously identified, this region was assumed to be a gametophytic selection-2 (ga2) gene region. To evaluate this region, two near isogenic lines (NILs) were developed. One NIL had the ‘Nipponbare’ segment of this region on the genetic background of ‘Milyang23’ (NIL9-23), and the other NIL had the ‘Milyang23’ segment on the genetic background of ‘Nipponbare’ (NIL33-18). NIL9-23 and ‘Milyang23’, NIL33-18 and ‘Nipponbare’, and ‘Nipponbare’ and ‘Milyang23’ were respectively crossed to produce F₁ and F₂ populations. The F₁ plants of NIL9-23 × ‘Milyang23’ and NIL33-18 × ‘Nipponbare’ showed high seed fertility and the same pollen fertility as their parental cultivars, indicating that ga2 does not reduce seed and pollen fertility. Segregation ratio of a molecular marker on the ga2 region in the three F₂ populations was investigated to clarify whether segregation distortion occurred on the different genetic backgrounds. Segregation distortion of the ga2 region appeared in the both F₂ populations from the NIL9-23 and ‘Milyang23’ cross (background was ‘Milyang23’ homozygote) and the ‘Nipponbare’ and ‘Milyang23’ cross (background was heterozygote), but did notin the F₂ population from the NIL33-18 and ‘Nipponbare’ cross (background was ‘Nipponbare’ homozygote). This result indicates that ga2 interacts with a ‘Milyang23’ allele(s) on the different chromosomal region(s) to cause skewed segregation of the ga2 region. In addition, segregation ratio was the same between the F₂ populations from NIL9-23 × ‘Milyang23’ and ‘Nipponbare’ × ‘Milyang23’ crosses, suggesting that the both genotypes, ‘Milyang23’ homozygote and heterozygote, of gene(s) located on the different chromosomal region(s) have the same effect on the segregation distortion. This revised version was published online in July 2006 with corrections to the Cover Date.     

Population genetic responses of wild forage species to grazing along a rainfall gradient in the Sahel: A study combining phenotypic and molecular analyses

Genetic diversity was studied in wild forage species subjected to grazing along a rainfall gradient in West Niger, within the Sahel with aim of identifying adapted genotypes, which could be used to reclaim degraded land. Two legumes (Alysicarpus ovalifolius and Zornia glochidiata) and two grasses (Brachiaria xantholeuca and Cenchrus biflorus) were selected to relate phenotypic adaptation to genetic diversity in response to grazing and rainfall. Populations of each species were sampled from both heavily grazed and ungrazed sites along a rainfall gradient, approximating 200 mm yr⁻¹ to 800 mm yr⁻¹ rainfall isohyets. The adaptative phenotypic expressions to aridity and grazing of the populations from each of the species were characterised by morphological measures performed on the plants sampled in the field. These analyses were then compared with the results from genetic analyses using the PCR-based techniques of amplified fragment length polymorphism (AFLP) and random amplified polymorphic DNA (RAPD). Analyses of molecular data using cluster analysis (UPGMA), principal co-ordinates analysis (PCO), Mantel tests and an Analysis of Molecular Variance (AMOVA), revealed genotypic distinction between populations subjected to both differing aridity and grazing. The majority of the total genotypic variation sampled in all species occurred among individuals within a population. The significant morphological differentiation found among populations subjected to varying grazing and aridity stresses, determined through linear regression analyses, did not correlate significantly with the genotypic differentiation, as revealed by Mantel tests. The results suggest that grazing does not cause a loss of genetic diversity in the wild forage species studied, although with increasing aridity the impact of grazing on the genetic diversity of populations may increase.R. Jamnadass and A. Muchugi Equal contribution to this work by both authors.