Genetic diversity in tropical maize inbred lines: heterotic group assignment and hybrid performance determined by RFLP markers

Tropical maize inbred lines, eight derived from a Thai synthetic population (BR‐105) and 10 from a Brazilian composite population (BR‐106), were assayed for restriction fragment length polymorphisms with 185 clone‐enzyme combinations. The aim of this study was to investigate genetic distances among tropical maize material and their relationship to heterotic group allocation and hybrid performance. Genetic distances (GDs) were on average greater for BR‐105×BR‐106 lines (0.77) than for BR‐106×BR‐106 (0.71) and for BR‐105×BR‐105 (0.69) lines. Cluster analysis resulted in a clear separation of BR‐105 and BR‐106 populations and was according to pedigree information. Correlations of parental GDs with single crosses and their heterosis for grain yield were high for line crosses from the same heterotic group and low for line combinations from different heterotic groups. Our results suggest that RFLP‐based GDs are efficient and reliable to assess and allocate genotypes from tropical maize populations into heterotic groups. However, RFLP‐based GDs are not suitable for predicting the performance of line crosses from genetically different heterotic groups.     

Relationship of intra- and interpopulation tropical maize single cross hybrid performance and genetic distances computed from AFLP and SSR markers

Two sets of tropical maize inbred lines, one derived from the BR-105 population and another derived from the BR-106 population, were assayed for Amplified Fragment Length Polymorphism's (AFLP) and for Simple Sequence Repeat (SSR), in order to investigate genetic distances among lines and their relationship to heterotic group assignment and single cross yield performance. Genetic distances were on average greater for interpopulation than intrapopulation crosses for both AFLP and SSR. Cluster analysis was in agreement with the original assignment for heterotic groups. Inbred line 16, derived from BR-106, was assigned to the BR-105 set, in agreement with single cross yield performance from intra- and interpopulation crosses. However, the same pattern was not observed for SSR where another two lines from BR-106 were also assigned to the BR-105 set. Correlation coefficients of genetic distances (GD) with F₁ grain yield and heterosis were high for BR-106 ×BR-106 crosses (0.91^** and 0.82^** for AFLP and SSR, respectively), moderate for BR-105 × BR-105 crosses (0.52^* for AFLP and SSR) and low for BR-105 × BR-106 crosses (0.29 and 0.16 for AFLP and SSR, respectively). The lower correlation at interpopulation level was due to the smaller range of GD caused probably by a previous selection for combining ability. General results showed that the AFLP molecular marker is efficient in assigning maize lines to heterotic groups and that AFLP-based GD is suitable for predicting the maize single cross performance for intrapopulation crosses of broad-based populations. The efficiency of SSR in assigning lines to heterotic groups and for predicting single cross performance was smaller than AFLP. This revised version was published online in July 2006 with corrections to the Cover Date.     

Relationship between phenotypic and genetic diversity of parental genotypes and the specific combining ability and heterosis effects in wheat (<Emphasis Type="Italic">Triticum aestivum</Emphasis> L.)

The purpose of the study was to evaluate the relationship between the genetic distances (GD) and phenotypic distances (PD) of parents and the specific combining ability (SCA) and heterosis effects. The experiment comprised 18 parental genotypes of wheat (Triticum aestivum L.) and 76 F₂ hybrids, obtained after crossing in a line × tester scheme. Parents and hybrids were examined in a field experiment conducted in a block design with three replications. SCA as well as mid-parent heterosis effects were estimated for selected morphological and technological traits. PDs and GDs were investigated between pairs of parental genotypes. GD between parental genotypes was evaluated by using randomly amplified polymorphic DNA markers. Heterosis was observed in all hybrids, and protein content exhibited the highest heterosis among the seven examined traits. The relationship between PD and GD as well as the SCA and mid-parent heterosis effects were evaluated using correlation coefficient. The correlation between PD, SCA and heterosis were low and not significant for the examined traits, whereas the correlation between SCA, heterosis and GDs were significant for protein content and rheological properties. The results indicate that GDs between parents can be used to predict performance of hybrids for selected technological traits.     

Relationship between molecular marker heterozygosity and hybrid performance in intra- and inter-subspecific crosses of rice

The objective of this study was to investigate the relationship between molecular marker diversity and heterosis in both intra-and inter-sub-specific hybrids of rice to evaluate the feasibility of predicting hybrid performance using molecular markers. Eleven elite lines were intermated resulting in a diallel set including 10 indica × indica, 15 japonica × japonica and 30 indica × japonica crosses. The F₁ hybrids and parents were evaluated for agronomic performance in a replicated field trial. The parental lines were tested for DNA polymorphisms with 113 restriction fragment length polymorphism (RFLP) probes covering the 12 rice chromosomes. Inter-subspecific crosses showed better performance and higher heterosis than intrasubspecific hybrids. Correlations of marker heterozygosity with hybrid performance and heterosis differed considerably between the two subspecies; they were higher in crosses within japonica subspecies than within indica subspecies. Very little correlation was detected in intersubspecific crosses. It was concluded that relationships between marker heterozygosity and hybrid performance were complex owing to germplasm diversity and the complexity of the genetic basis of heterosis. The implications of the results in predicting hybrid performance using molecular markers are discussed.     

Relationship between heterosis and genetic divergence in 'Tongil'-type rice

Improving grain yield and quality of ‘Tongil’‐type rice (indica/japonica) continues to be a major breeding objective in Korea. In this study, genetic divergence among 13‘Tongil’‐type rice cultivars was evaluated and the relationship between genetic distance and hybrid performance in all possible nonreciprocal crosses between them assessed. The 78 F₁ hybrids together with the 13 parents were evaluated for eight traits of agronomic importance, including yield, in a replicated field trial. The 13 parents were examined for DNA polymorphism using 71 micro‐satellite or simple sequence repeats and 46 random decamer oligonucleotide primers. A total of 319 polymorphic variants were generated and, based on the polymorphism data, genetic distances (GDs) ranged from 0.021 to 0.437. Cluster analysis based on GDs revealed associations among cultivars which was in agreement with the pedigree data. Heterosis was observed in hybrids for most of the traits, and yield exhibited the highest heterosis among the eight traits examined. The correlation values of GDs with F₁ performance were mostly nonsignificant, except for yield, culm length and spikelets per panicle. The correlations of GDs with midparent and better‐parent heterosis were not significant enough to be of predictive value. These results indicate that GDs based on the microsatellite and random amplified polymorphic DNA (RAPD) markers may not be useful for predicting heterotic combinations in ‘Tongil’‐type rice and support the idea that the level of correlation between hybrid performance and genetic divergence is dependent on the germplasm used.     

Indica-Japonica differentiation of paddy rice and its relationship with heterosis

The Indica‐Japonica differentiation of three photoperiod‐sensitive and/or thermosensitive genetic male‐sterile rice (PGMS or TGMS, respectively) lines and 47 male parental lines from seven ecotypes were studied for their restriction fragment length polymorphism marker data to determine which ecotype crosses with the three PGMS and/or TGMS lines could lead to higher yield potential, and to estimate the relationship between the Indica‐Japonica differentiation of parents and heterosis in grain yield and its components. The results indicated that hybrids derived from ‘N422s’ and the early‐middle ripening Indica varieties from southern China, and hybrids between ‘Pei'ai64s’ and three Japonica ecotypes, including North‐eastern Japonica varieties, restoring lines of Japonica hybrid rice and north China Japonica varieties, showed the highest grain yields. There was less variation of yield among the F1s between ‘108s’ and the seven ecotypes than among the other F₁s. Highly significant positive correlations between heterosis of the F₁ yield and genetic distance of the parents were detected, although the correlation between F₁ yield performance and genetic distance did not reach a significant level. Considerable variation of correlation between heterosis and genetic distance was also detected in the Indica × Indica crosses and Indica × Japonica crosses. There was much higher correlation (r = 0.63) between the F₁ yield performance and the genetic distance of parents in the Indica × Indica crosses than in the others. It is proposed that a genetic distance of 0.4‐0.8 between the two parents of hybrid rice might be appropriate not only for F₁ performance, but also for heterosis.     

Relationship between hybrid performance and AFLP based genetic distance in highland maize inbred lines

The objectives of this study were to determine the crossing performance of highland maize inbred lines for grain yield, days to silk and plant height; estimate genetic distance (GD) among the inbred lines and in association with tester parents, and to investigate the relationship of GD with hybrid performance and midparent heterosis (MPH). A total of 26 inbred lines were crossed with six (population and line) testers in a factorial-mating scheme. The F_1’s and the parents were evaluated at five locations in Ethiopia. Nine amplified fragment length polymorphism (AFLP) primer pairs were used to genotype all the parents. The F_1’s were found to vary widely for grain yield and other traits measured. Yield superiority of more than 30% over the best hybrid check was obtained for some testcross hybrids. Midparent heterosis on average was moderate for grain yield and, plant height. And for days to silking, MPH values were mostly negative. Mean GD values determined from the inbred lines by population tester (0.680) and line tester (0.661) combinations were not significantly different. Cluster analysis separated the tester parents from the corresponding inbred lines. AFLP grouping of the inbred lines was in agreement with their pedigree records. Genetic distances derived from the inbred lines × all testers and from the population testers’ sub-group were not positively correlated with hybrid performance and MPH for most traits. In contrast, correlations of GDs involving the line testers’ sub-group with F_1’s and MPH were significantly positive but with low magnitude to be of predictive value.     

Association between AFLP-based genetic distance and hybrid performance in tropical maize

Identifying the best inbred combinations for the development of commercial hybrid maize varieties remains the main challenge to maize breeders. The aim of this work was to study associations between the genetic distance (GD) of 21 inbreds and the corresponding F₁ phenotypic data. Furthermore, the impact of grouping lines into genetically similar clusters was investigated. The 21 inbred lines were fingerprinted using amplified fragment length polymorphism markers. Parents and 210 F₁ progeny were evaluated in the field. Joint data analysis mostly revealed a tighter association between GD and the F₁ performance or mid parent heterosis in the intergroup than in the intragroup crosses. Despite these correlations, intergoup crosses should always be field‐tested before their release. Crosses showing low GD values should be discarded to avoid field‐testing costs. Better F₁ hybrid performance predictions can be achieved by integrating molecular and F₁ phenotypic data.     

Heterosis in Primary Hexaploid Triticale with Heterozygous Wheat or Rye Genome*)

Twelve primary hexaploid triticale (X Triticosecale Wittmack), synthesized from, three lines of tetraploid wheat (Triticum durum L., T. turgidum L.) and four inbred lines of rye (Secale cereale L.), were used to produce 18 crosses with homozygous wheat and heterozygous rye genome and 12 crosses with heterozygous wheat and homozygous rye genome. Parents and crosses of triticale, wheat, and rye were tested for two years (rye for one year only) in two-replicate block designs with 1 m²-plots. Data were assessed for plant height, grain yield and for yield-related traits. Performance of triticale crosses was considerably lower than that of the wheat and rye crosses. The amount of heterosis varied greatly between years. Positive and mainly significant heterosis was revealed in triticale generations F₁ and F₂. The average values were closer to those in wheat than to those in rye. For most characters a high level of heterosis was retained in tnucalt1 generation F₂. Heterozygosity of the wheat and rye genome both contributed to heterosis in triticale. However, gene action of the rye genome strongly depended on the homozygous wheat background: one wheat line almost completely suppressed and another greatly stimulated the heterotic effect of the rye genome. In the later case, the amount of heterosis was related to that in rye per se. Information from hybrid rye breeding may therefore be used when establishing gene pools for hybrid breeding in triticale.     

Barley mutant heterosis and fixation of'F1-performance'in doubled haploid lines

The phenomenon of mutant heterosis was studied in barley using F₁ hybrids developed from crosses involving two mutants derived from the same parent variety (067AR× 032AR), or from crosses of mutants with their parent varieties (Diva × 228DV; 437MG × Mg 4170). The agronomic performance (plant height, tillers per plant, kernel number and weight per plant) of F₁ hybrids was evaluated in relation to the corresponding parent variety used for mutant development. Wide-space sowing (0.O6 m²/plant) was applied to allow the full expression of hybrid vigour. The hybrids significantly exceeded their parent varieties in various yield parameters in each year of the experiment, Depending on the cross, the level of heterosis ranged from 39.45% to 82.1% for kernel number and from 44.45% to 91.2% for kernel weight, as an average of observations over three years. The increased yield of hybrids was accompanied by an increase in spikes/plant, without significant changes, in most cases, in yield components such as kernels/spike and 1000 kernel weight. The height of F₁ plants remained similar to the corresponding parent variety. Doubled haploicls (DH) were developed through anther culture from three F₁ hybrids to determine whether the heterotic effect could be fixed, even partly, in homozygous lines. Twenty-eight, 42 and 68 DH lines per cross were produced, and evaluated in the field, space-planted at 0.06 m²/plant, together with hybrids and parent varieties. The best-performing DH lines were evaluated again the following season in a conventional field trial. We were able to select 1.5 -14.3% of‘F₁,-performing’DH lines per hetorotic cross combination, which, in two years of studies, reached the yield of heterotic hybrids and significantly out-yielded the parent variety in plot. The yield potential of DH lines was better expressed under a wide sowing density, but even under normal growing conditions, the yield of the best selected DH lines surpassed their corresponding parent varieties by 17.77ndash;30.1%.     

Heterotic groups based on yield-specific combining ability data and phylogenetic relationship determined by RAPD markers for 28 tropical maize open pollinated varieties

Twenty eight maize open pollinated varieties (OPVs) were crossed in a diallel scheme and the 378 F₁'s were evaluated in 10 environments in Brazil. Based on yield-specific combining ability data (SCA), these varieties were classified in four heterotic groups. The consistence of the proposed heterotic groups was confirmed comparing intra- and inter-group F₁ values and midparent heterosis. Superior OPVs combinations for use as a source of inbreds in hybrid breeding programs were determined. RAPD markers were used to genotype these varieties. A UPGMA dendogram, based on marker data from 50 primers and 178 polymorphic bands, was obtained. Phylogeny obtained with RAPD markers agreed with known pedigree data. Dent germplasm tended to group separately from flint germplasm. Multidimensional Scaling Analysis on marker data and morphological data showed a higher degree of genetic divergence among the dent germplasm than among the flint germplasm used in this study. Correlation between RAPD marker estimated genetic distance and SCA for yield was low and positive (r = 0.16^**). This revised version was published online in August 2006 with corrections to the Cover Date.     

Combining ability,heterosis and genetic diversity in tropical maize (<Emphasis Type="Italic">Zea mays</Emphasis> L.) under stress and non-stress conditions

Drought and low soil fertility are considered the most important abiotic stresses limiting maize production in sub-Saharan Africa. Knowledge of the combining ability and diversity of inbred lines with tolerance to the two stresses and for those used as testers would be beneficial in setting breeding strategies for stress and nonstress environments. We used 15 tropical maize inbred lines to (i) evaluate the combining ability for grain yield (GY), (ii) assess the genetic diversity of this set of inbred lines using RFLP, SSR, and AFLP markers, (iii) estimate heterosis and assess the relationship between F₁ hybrid performance, genetic diversity and heterosis, and (iv) assess genotype × environment interaction of inbred lines and their hybrids. The F₁ diallel hybrids and parental inbreds were evaluated under drought stress, low N stress, and well-watered conditions at six locations in three countries. General combining ability (GCA) effects were highly significant (P < 0.01) for GY across stresses and well-watered environments. Inbred lines CML258, CML339, CML341, and CML343 had the best GCA effects for GY across environments. Additive genetic effects were more important for GY under drought stress and well-watered conditions but not under low N stress, suggesting different gene action in control of GY. Clustering based on genetic distance (GD) calculated using combined marker data grouped lines according to pedigree. Positive correlation was found between midparent heterosis (MPH) and specific combining ability (SCA), GD and GY. Hybrid breeding program targeting stress environments would benefit from the accumulation of favorable alleles for drought tolerance in both parental lines.     

Relationship between molecular marker heterozygosity and hybrid performance in intra- and interspecific hybrids of cotton

The present study was conducted to investigate the relationship between parental molecular marker diversity and hybrid performance in both intra‐ and interspecific hybrids of cotton to evaluate the feasibility of predicting hybrid performance using molecular markers. Three cytoplasmic male sterile (CMS) lines were crossed with 10 restorer lines to produce 22 F₁ hybrids during 2003. Of 22 F₁s, 14 hybrids were intraspecific (Gossypium hirsutum × G. hirsutum) and eight interspecific (G. hirsutum × G. barbadense). These 22 F₁ hybrids and their parents were evaluated for yield and fibre quality traits at Zhejiang University, Hangzhou, China during 2004 and 2005. Genetic distances (GD) among the parents were calculated from 56 random‐amplified polymorphic DNAs (RAPD) and 66 simple sequence repeat (SSR) marker data, and their correlation with hybrid performance and heterosis were analysed. The parents could be discriminated into G. hirsutum and G. barbadense clusters by cluster analysis based on both RAPD and SSR markers data. The correlation (r = 0.503, P ≤ 0.05) was calculated between GD_rapd (GD based on RAPD markers) and GD_ssr (GD based on SSR markers). Correlation of GD with hybrid performance and heterosis differed considerably between intra‐ and interspecific hybrids. The correlation between GD and hybrid performance was non‐significant for most of traits within the hybrids of G. hirsutum species. However, it was significantly and positively correlated for fibre length, fibre strength and elongation in interspecific hybrids. The relationship between GD and heterosis was observed to be positively significant for boll weight within hybrids of G. hirsutum with significant and negative correlations for fibre length and elongation. In conclusion, the power of predicting hybrid performance using molecular markers in cotton is low. But, the relationship between SSR marker heterozygosity and hybrid performance can be used to predict fibre length during interspecific hybrid cotton breeding.     

Prospects for hybrid breeding in winter triticale: II. Relationship between parental genetic distance and specific combining ability

Significant relative midparent heterosis (MPH%) for grain yield in triticale (×Triticosecale Wittm.) has generated interest in the development of hybrid cultivars. The objectives of this study were to (i) examine the association between parental genetic distance (GD) and specific combining ability (SCA), (ii) investigate the existence of genetically distant heterotic groups in elite germplasm, and (iii) draw conclusions for future hybrid breeding in winter triticale. Genetic distance between 61 lines was estimated, based on 93 polymorphic simple sequence repeat (SSR) marker loci and 10 AFLP (amplified fragment length polymorphism) primer‐enzyme combinations (PEC). Agronomic data of 206 F₁ crosses and their 61 parental lines grown in six German environments were published recently in a companion study. Correlations were made between SCA for grain yield, number of spikes/m², 1000‐kernel weight and number of kernels per spike with GD estimates of the 56 female and five male parents (testers). Principal co‐ordinate analyses (PCoA) based on SSR data revealed no distinct subgroups in the germplasm. Correlations between GD and SCA were low for all traits (|r| ≤ 0.31), which hampers the prediction of SCA from molecular data. A multi‐stage procedure is recommended for future hybrid breeding in triticale by applying a pragmatic approach for the grouping of germplasm following the early history of hybrid breeding of maize.     

Signatures of divergent selection for cold tolerance in maize

Divergently selected genotypes can be used for detecting the genomic regions affecting the selected trait (selection signature). Moreover, the genetic distances (GDs) among divergently selected lines can be correlated with the agronomic performances of the crosses among them. Using as source the maize F₂ of B73?×?IABO78, we previously conducted four cycles of divergent recurrent selection and three cycles of divergent selection in inbreeding for cold tolerance at germination. We finally obtained 10 lines selected for low (L) and 10 lines selected for high (H) cold tolerance, which exhibited a notable divergence for both the selected and associated traits. Herein, we investigated the 20 lines and the 28 single diallel crosses among eight random lines (four L and four H); the main objectives were to identify the putative regions controlling the selected and associated traits and to study the relationships between crosses performances and GDs among their parental lines. Allele frequencies at 932 recombination blocks based on 19,220 polymorphic SNPs were obtained for the two lines’ groups; the F _ST calculated across sliding windows indicated 18 regions highly divergent between groups. The increasing alleles for cold tolerance were contributed by both parents, consistently with the transgressive segregations previously found. Several regions associated to DG also affected various agronomic traits. The cross performances showed some relationships with the genetic distances among parental lines for traits affected by dominance, provided that all crosses were considered, while these relationships vanished when only L?×?H crosses were examined.     

Reciprocal differences for forage traits in single and three-way crosses of maize

Eleven flint inbreds of maize were crossed with 11 dent inbreds according to a balanced incomplete factorial mating design to produce 66 single crosses (2W) in both reciprocal forms. Additionally, six three-way crosses (3W) were also produced in both reciprocal forms. These hybrids were evaluated for 11 forage yield and quality traits in 1983 and 1984 at three sites in Germany. Highly significant reciprocal differences were observed for nine traits in the 2W. The respective variance component estimates amounted to 2–33% of the variance component due to crosses. On average, flint × dent 2W had lower dry-matter content (DMC) than dent × flint 2W and a 1 % greater dry-matter yield (DMY) of forage. Three-way crosses produced on singles exhibited greater DMC (3%) and DMY (5%) of the ear, metabolizable energy content (MEC) of stover (1%), and DMC and DMY of forage (2%) than their reciprocal forms produced on lines. Based on this study, because of their minor importance, routine testing for reciprocal differences among 2W hybrids is not recommended.     

Diallel analysis of combining ability and heterosis for yield and yield components in rice by using positive loci

Diallel crosses have been widely used for analysis of general combining ability (GCA), specific combining ability (SCA) and heterosis. In the present research, 12 lines were selected from 60 inbred rice bred by International Rice Research Institute with extremely-high or -low yielding records according to the previous three consecutive seasons of yield trial experiments, to construct a half diallel cross. The genetic distances (GDs) revealed by molecular markers for the 12 lines ranged from 0.2288 to 0.7169, averaging at 0.5882; clustering analysis showed the 12 lines were divided into four groups maintaining the original cluster structure of the 60 lines. The positive loci (PLs) including effective-increasing loci (ILs) and effective-decreasing loci (DLs) were screened. The results showed that hybrids derived from those parents with higher GCA effects had better performance for traits of yield and yield components. The SCA effects were strongly correlated to F₁ performance as well as heterosis; the GDs based on ILs were significantly positive correlated to SCA effects and heterosis for yield and yield components while the GDs based on DLs showed the significant negative correlations. Based on this research, a new conception, i.e. general sum of combining ability (GSCA) was conceived, which is defined as the sum of GCAs for two parents of a hybrid. The GSCA and SCA showed similar correlations with traits of yield and yield components. The results illustrated that ILs could be used for further study on prediction of heterosis for traits of yield and yield components; and GSCA may be another considerable parameter combined with ILs for breeders in selecting elite hybrid.     

Genetic analysis of heterotic loci detected in a cross between indica and japonica rice (Oryza sativa L.)

The study on the genetic basis of heterosis has received significant attention in recent years. In this study, using a set of introgression lines (ILs) and corresponding testcross F₁ populations, we investigated heterotic loci (HL) associated with six yield-related traits in both Oryza sativa L. subsp. indica and japonica. A total of 41 HL were detected on the basis of mid-parent heterosis values with single-point analysis. The F₁ test-cross population showed superiority in most yield-related traits and was characterized by a high frequency of overdominant HL. Thirty-eight of the 41 HL were overdominant, and in the absence of epistasis, three HL were dominant, suggesting that heterotic effects at the single-locus level mainly appeared to be overdominant in rice. Twenty-four HL had a real positive effect, suggesting that they are viable candidates for the improvement of rice yield potential. Compared with the quantitative trait loci (QTLs) detected in the ILs, only six out of the 41 (14.6%) HL were detected in QTL analysis under the same statistical threshold, indicating that heterosis and trait performance may be conditioned by different sets of loci.     

Derivation of superior F5 lines from heterotic hybrids in pea

Summary F₅ seeds from six highly heterotic F₁ hybrids were produced to determine whether the superior performance of F1 hyrids could be fixed in pure-line derivatives of pea (Pisum sativum L.). For each cross, 24–31 F₅ lines derived from single seed descent were compared with F₁ hybrids and their parents in two environments. The F₁ hybrids out-yielded the best parent by up to 11%. All crosses produced F₅ lines which were as high in yield as the F₁ hybrids indicating that pure-line derivatives equivalent in yield to the heterotic F₁ hybrid could be developed by conventional breeding. Heterosis of the F₁ hybrids over the best commercial cultivar was thus a useful indicator of the future performance of the pure-line derivatives. These results indicate that overdominance was not an important component of heterosis in peas.     

Heterosis,combining ability and genetic effect,and relationship with genetic distance based on a diallel of hybrids from five diverse <Emphasis Type="Italic">Gossypium barbadense</Emphasis> cotton genotypes

The cultivated tetraploid Gossypium barbadense L. cotton produces superior natural fibers for the textile industry in the world. However, the possibility in utilization of heterosis to further increase its lint yield has not been extensively explored. In this study, two commercial US Pima cotton cultivars and three exotic G. barbadense lines, together with all of their possible hybrids in F₁ and F₂ progeny without reciprocals, were tested for lint yield, yield components, and fiber quality traits in four environments in 2005–2007. With a few exceptions, genotype (G), environment (E), and G × E were all significant or highly significant for all the traits studied. General combining ability (GCA) variances for all the traits in both F₁ and F₂ were also significant, while specific combining ability (SCA) variances were detected only for lint yield, fiber length, and micronaire in both generations and boll weight in F₁. GCA × E was also detected for lint percent, seed index, and fiber length in both F₁ and F₂, and boll weight in F₁, but none of the traits had significant SCA × E. As a group, F₁ and F₂ out-yielded the parent group by 20–40% and 6–10%, respectively. Mid-parent heterosis (MPH) for lint yield in F₁ was generally positive, ranging from ?4.7 to 116.4% with an average of 21.2–48.7%, while lint yield MPH in F₂ ranged from ?23.3 to 69.4% with an average of 6.4–12.4%. However, useful heterosis in lint yield was only detected in the hybrid between the two US commercial cultivars Pima S-7 and DP 340. MPH for other traits was low or not detected. MPH in F₂ was lower than that in F₁ but they were generally positively correlated. The genetic distances (GD) of the parents (based on 467 polymorphic RAPD and AFLP markers) between the five parents was not consistently correlated with MPH and SCA of their hybrids and dominant effects for lint yield and other traits. However, significant and positive correlations between GD of parents and the performance of their hybrids were detected for lint yield, lint percentage, and lint index in both F₁ and F₂ in most of the tests. GD of parents was also correlated with their GCA and additive effects in lint yield, lint percent, lint index, micronaire, plant height, and elongation. The results suggest that the close correlation between GD and hybrid performance per se was mainly due to the existence of GCA and additive effects from parents.