Evaluation of landrace topcross hybrids of pearl millet for arid zone environments

Pearl millet (Pennisetum glaucum (L.) R. Br.) cultivars for marginal, arid environments need to combine the adaptation to stress conditions of indigenous landraces with an improved yield potential and disease resistance, to allow them to both perform well in farmers fields and to meet the requirements for cultivar release. This paper evaluates landrace-based topcross hybrids (adapted landraces crossed on high-yielding male-sterile lines), as a quick and efficient way of achieving this objective. Topcross hybrids showed a consistent increase in biomass production across all test environments, including the harsh arid zone environments. Depending upon the plant type of the male-sterile used to make the hybrid, this was expressed as increased grain yield only, or increased grain and fodder yields. The downy mildew (Sclerospora graminicola) reaction of the topcross hybrids was determined by the reaction of the male-sterile line used, with the resistant male-sterile producing resistant topcross hybrids and vice-versa. Topcrossing adapted landraces on high-yielding male-sterile lines thus provides an opportunity to improve disease resistance and grain and/or fodder yields, with no apparent loss of adaptation to the marginal environments in which the landraces have evolved.ICRISAT Journal Article no. 1575     

Characterization of seed producing pearl millet x elephantgrass hexaploid hybrids

Summary Successful widespread use of a forage or biomass plant depends largely on its ease of establishment. Elephantgrass (Pennisetum purpureum Schum.) is used for both forage and biomass production, due to its high dry matter production, aggressiveness, perenniality, and forage quality. However, lack of high quality seed has limited the use of elephantgrass and its hybrids to pearl millet (P. glaucum L. R.Br.). Development of a seed-propagated pearl millet x elephantgrass hexaploid cultivar, which would combine desirable characteristics from both of these species, would be highly desirable. The objectives of this study were to characterize morphological and seed-related traits from the selfed progeny of seven hexaploid hybrids, and to determine whether seed size had any influence on these traits. Traits studied included number of tillers, height, leaf length and width, panicles per plant, days to flowering, panicle length, seed set, seed production, and weight of 100 seeds. Genetic differences were found among the progeny of the seven hybrids. Differences were also found between plants derived from large- and small-seed lots within families. Plants descended from larger seed had better growth and seed-related characteristics, which resulted in more desirable plants. These hexaploid hybrids showed potential for direct seeding into the field for biomass or forage production.     

Smut reaction of pearl millet hybrids affected by fertility restoration and genetic resistance of parental lines

Summary Pearl millet (Pennisetum glaucum (L.) R. Br.) hybrids based on the A₁ cytoplasmic-nuclear male-sterile (CMS) lines are more susceptible to smut (Tolyposporium penicillariae Bref.) than open-pollinated varieties. Seventy eight pairs of hybrids, made onto male-sterile (A) lines and their counterpart maintainer (B) lines, were evaluated to examine the effects of male sterility and genetic resistance of parental lines on the smut severity of hybrids. The A-line hybrids had higher smut severity and lower selfed seedset than the counterpart B-line hybrids, indicating that it is the CMS-mediated male sterility rather than the A₁ cytoplasm per se that caused greater smut severity of A-line hybrids. However, with the use of resistant parental lines even male-sterile hybrids of A-lines, in several cases, were as resistant as some of the highly resistant male-fertile hybrids of B-lines. It would be possible to produce smut resistant hybrids (< 10% severity) on A-lines, albeit in low frequency, even if only one parent of a hybrid were resistant. However, the probability of producing such hybrids would be higher when both parents were resistant to smut. Thus, improvement in smut resistance of parental lines and fertility restoration ability of pollinators would provide the most effective genetic approach to smut disease management in hybrids.Submitted as JA No 1737 by the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT).     

Ergot reaction of pearl millet hybrids affected by fertility restoration and genetic resistance of parental lines

Summary High ergot (Claviceps fusiformis Loveless) susceptibility of pearl millet (Pennisetum glaucum (L.) R. Br.) hybrids has often been associated with the A₁ cytoplasm of male-sterile lines (A-lines). To understand the underlying basis of this association and to examine the prospects of breeding ergot-resistant hybrids, we evaluated 56 hybrids and their 15 parental lines for ergot reaction and selfed seedset for 2 years in disease nurseries at ICRISAT Asia Center. Hybrids were made by crossing seven pollen parents (2 susceptible and 5 resistant) onto two resistant and two susceptible A-lines, and their four corresponding maintainer lines (B-lines). A-lines had no selfed seedset while B-lines had 32–75% selfed seedset. Hybrids of A-lines had significantly less selfed seedset than the hybrids of the corresponding B-lines. The reduced seedset of A-lines and their hybrids, however, was not always accompanied by significantly higher ergot susceptibility. Highly resistant hybrids were obtained where both A-lines and pollen parents were highly resistant, regardless of male fertility levels of the hybrids. Thus, although the A₁ cytoplasm, by its reduction of male fertility, had a large and significant effect in increasing ergot severity of hybrids, the contribution of nuclear genetic factors of female parents was about 1.8 times larger than that of the cytoplasm.Submitted as JA No. 1776 by the International Crops Research Institute for the Semi Arid Tropies.     

Smut severity of pearl millet hybrids with male sterile and fertile cytoplasm

Summary Effect of A₁ male sterile cytoplasm on smut severity in pearl millet (Pennisetum glaucum (L.) R.Br.) was studied by comparing 35 pairs of F₁ hybrids, each pair carrying male sterile and normal cytoplasm. Mean smut severity was not significantly different in the hybrids carrying male sterile or normal cytoplasm. This suggests that in pearl millet male sterile cytoplasm is not associated with higher smut susceptibility. Partitioning of variance into different components showed that pollinators, A/B line pairs and their interaction primarily influenced smut severity of hybrids. Smut susceptibility might be attributed to effects of cytoplasm × nuclear interactions.     

Effect of d2 dwarfing gene on grain yield and yield components in pearl millet near-isogenic lines

Summary A d₂ dwarfing gene in pearl millet [Pennisetum glaucum (L.) R. Br.] is currently being extensively used for the development of hybrid parents. Its effect on grain yield and yield components is poorly understood. Twelve pairs of tall and dwarf near-isogenic lines developed in the diverse genetic background of three composites were evaluated for grain yield and yield components for 2 years at two locations in southern India. The d₂ gene or the genes linked to it, on an average, reduced plant height by 42%, grain yield by 14%, and head girth by 8% but increased head length and number of tillers per plant by about 5–6%. Large variations were observed among pairs (genetic background) for the difference between tall and dwarf near-isogenic lines for all of the above yield components resulting in no significant difference in five pairs and 17–35% less yield in dwarfs as compared to their tall counterparts in six pairs. Days to 50% flowering and seed weight were least affected by the d₂ gene with the average difference between tall and dwarf groups of near-isogenic lines being of the order of 1–2%. These results indicate that the advantageous effects of d₂ dwarfing gene can be effectively exploited by manipulating the genetic background. The difference between the average grain yields of tall and dwarf groups of near-isogenic lines showed considerable variation across environments with the dwarfs yielding as much as tall group in one environment and up to 30% less than the tall group in the other, thus, indicating that the d₂ gene effect may be substantially modified by the environments.Submitted as JA No. 979 by the International Crop Research Institute for the Semi-Arid Tropics (ICRISAT).     

Inheritance of threshing percentage in pearl millet

Summary The threshing percentage (TH%) has been suggested as a selection criterion to identify the pearl millet (Pennisetum glaucum (L.) R.Br.) lines with improved ability to fill and set grains under water limiting conditions. In this study, eight genetically diverse pearl millet inbreds and their 28 crosses produced by half diallel crossing design were used to examine range in TH%, to evaluate general combining ability (GCA) of parents and specific combining ability (SCA) effects of crosses. The results showed significant variation among parental lines for TH%. The inbreds differed for their GCA effects and crosses for their SCA effects. Parents with high TH% and positive GCA effects were identified. The results of the study revealed that both additive and dominance components, with the preponderance of later, were important in the inheritance of TH%. Heritability in narrow sense was moderate (55%) indicating that selection for high TH% might be effective.     

Effect of water availability pattern on yield of pearl millet in semi-arid tropical environments

Summary Throughout much of the semi-arid tropics, fluctuations in grain yield can largely be attributed to differences in timing and intensity of drought stress. Since seasonal rainfall in these environments is often poorly related to grain yield, the aim of this paper was to establish a relationship between water availability and grain yield for pearl millet (Pennisetum glaucum (L.) R. Br.), grown across 24 semi-arid tropical environments in India. We used a simple soil water budget to calculate a water satisfaction index (WSI) throughout the season. The cumulative WSI at maturity explained 76% of the variance in grain yield. This was three times as much as explained by actual rainfall, because WSI accounted for differences in water losses and pan evaporation. A classification of environments into four groups of water availability patterns explained 75% of the environmental sum of squares for grain yield. For a subset of 13 environments, environmental differences in grain number could also be explained by water availability patterns, whereas differences in grain mass were related to both water availability and temperature. Our results indicate that cumulative WSI, which is an integrated measure of plant-available water, can provide an adequate estimation of the environmental potential for yield in environments where grain yield is mainly limited by variable availability of water.ICRISAT Journal Article 1637     

Heterosis in landrace-based topcross hybrids of pearl millet across arid environments

This study quantified the magnitude of heterosis in pearl millet (Pennisetum glaucum) topcross hybrids produced by crossing 16 diverse landraces and three high yielding open-pollinating varieties on two homozygous male-sterile lines. Hybrids and pollinators were grown in 12 year ×;location combinations in India that were grouped into three zones. Genetic components of variance quantifying the differences among these hybrids were estimated. The hybrids showed a conspicuous heterosis for grain yield, earliness and biomass yield but not for straw yield. The level and direction of heterosis for time to flowering depended strongly on the earliness of the male-sterile line. In the terminal drought stress zone hybrids made on the early maturing male-sterile line 843A had the highest level of heterosis for grain yield (88%). This was partly due to escape from terminal stress. In the other two zones the heterosis for grain yield was on average 30%. Heterosis for biomass yield and biomass yield per day was on average also positive in all three zones. For all traits, except time to flowering and biomass yield per day, pollinator effects were the only significant source of variation. Differences between hybrids were mostly caused by additive genetic effects. Significant amount of heterosis observed in landrace-based topcross hybrids for grain yield and other productivity-related traits suggested that substantial improvement in pearl millet productivity in and environments can be obtained by topcrossing locally adapted landraces on suitable male-sterile lines. This revised version was published online in July 2006 with corrections to the Cover Date.     

Commercial viability of alternative cytoplasmic-nuclear male-sterility systems in pearl millet

Commercial viability of three cytoplasmic-nuclear male sterility (CMS) systems (A₄, A₅ and A_v) as potential alternatives to the most widely used A₁ system in pearl millet (Pennisetum glaucum (L.) R.Br.) was evaluated in terms of stability of complete male sterility of four isonuclear A-lines (81A₁, 81A₄, 81A₅ and 81A_v) and the level and stability of male fertility restoration of their 44 single-cross hybrids. Lines 81A₄ and 81A₅ had no pollen shedders (PS), and there were very low frequency of non-PS plants of these A-lines that had a maximum of 1–5% selfed seedset (SSS). In 81A₁ and 81A_v,there were, albeit low frequency (<1%) of PS plants, and relatively higher frequency of the non-PS plants in these two lines, the more so in 81A_v,had 1–5% and even greater SSS. Some hybrids made on each of the three A-lines (81A₁, 81A₄ and 81A_v) had high and stable male fertility, while others made on the same three A-lines displayed large variation in SSS across the environments, the more so in case of hybrids made on 81A_v. These results indicate that the A₄ CMS system provides a better alternative to the A₁ CMS system, while the A_v system does not. On the basis of highly stable male sterility and the highest frequency of pollinators behaving as maintainers, the A₅ CMS system appeared to be the best for A-line breeding. The commercial viability of this CMS system in breeding R-lines of grain hybrids, however, still remains to be ascertained as no hybrid on it was fully male fertile in any environment. This revised version was published online in August 2006 with corrections to the Cover Date.     

Utility of pearl millet landraces in breeding dual-purpose hybrids for arid zone environments of India

In pearl millet (Pennisetum glaucum (L.) R. Br.), single cross hybrids based largely on exotic germplasm, have found very limited adoption in drought-prone arid regions mainly due to their inadequate adaptation to prevalent extremely harsh agro-climatic conditions. This study tested the hypothesis that the use of restorer parents derived from adapted landrace germplasm could be an alternative strategy in developing suitable hybrids for arid regions. Forty inbred restorer lines developed from two phenotypically diverse landraces were used to develop hybrids evaluated in this study for their performance in five typical arid zone environments between 2003 and 2007. Inbred lines from both landrace populations differed significantly in their combining ability for biomass, harvest index and grain and stover yields. A larger proportion of experimental hybrids, based on selected inbred lines, significantly outperformed the commercial checks for grain and stover yields. Increase in grain and stover yields was achieved primarily due to their higher biomass productivity, with no decline in harvest index. Choice of landrace determined the relative grain and stover productivity of their hybrids. Phenotypic differences observed in parental landraces in tillering and panicle length were also visible in their respective hybrids, indicating that characteristics of landrace parental populations were transmitted in their hybrids.     

Stability of male-sterile sources and fertility restoration of their hybrids in pearl millet

Genetic background has a significant effect on the expression of cytoplasmic-nuclear male sterility (CMS) in pearl millet (Pennisetum glaucum (L.) R. Br.). Therefore, a reliable characterization of CMS sources requires the use of near-isonuclear lines and their hybrids. We used this approach to characterize five CMS sources (A₁, A₂, A₃, A_v, and A₄). Male sterility of 81A₄ was the most stable, followed by 81A₁ and 81A_v, indicating the relative merits of these CMS sources in breeding stable male-sterile lines. Male-sterile lines 81A₂ and 81A₃ were highly unstable, indicating their minimal value. Differential male fertility restoration patterns of hybrids made on 81A_V and 81A₄ suggest that the A_v and A₄ cytoplasms represent CMS systems that are different from each other and from A₁, A₂ and A₃. An evaluation of topcross hybrids of 15 diverse populations made on 81A₁ and 81A₄ showed that each population had restorers and maintainers of both cytoplasms and that the frequency of maintainers of A₄ was as high as, or higher than, that of the A₁ cytoplasm. Thus, use of the A₄ cytoplasm can substantially increase the probability of breeding stable male-sterile lines based on inbreds developed from diverse genetic backgrounds, and also provide the opportunity for breeding restorers from each of these diverse genetic sources.     

Inheritance of partial rust resistance in pearl millet

Quantitative disease resistance should be exploited to complement the use of genes for qualitative or hypersensitive resistance. The expression and inheritance of partial rust resistance of pearl millet inbreds 700481-21-8 and ‘ICMP 501’ crossed to moderately susceptible Tift 383’ were evaluated in seedling assays in the greenhouse and in generation mean and single-seed descent populations in the field. Uredinium sizes on seedling leaves of hybrids were generally intermediate to those of the parental inbreds and consistent differences could be discerned in uredinium lengths. Area under the disease progress curves (AUDPCs) of individual plants of the parents, F₁, F₂, and backcross F₁S to each parent were determined from field trials. Broad-sense heritability estimates for both crosses were 43%. In generation mean analyses, additive genetic effects were significant in the cross of 700481–21–8 × Tift 383′, whereas additive, dominance, and dominance × dominance epistatic effects were significant for ‘ICMP 501’בTift 383’. The number of genes conferring partial resistance was estimated to be two for 700481–21–8 and 2.5 for ‘ICMP 501’. A hierarchical single-seed descent analysis revealed significant differences in AUDPC among F₃-derived F₄ progenies in the F₆ generation. Selection for progenies with greater resistance should be possible among F₄ families. Higher levels of resistance were observed in progeny derived from ‘ICMP 501’. Because segregation of resistance differed among progeny derived from 700481–21–8 and ‘ICMP 501’, the genetic basis for resistance probably differs between the two inbreds.     

Structure of variation among morphological and physiological traits in three pearl millet composites

Summary The plant breeder's task of improving and stabilizing many plant traits simultaneously is complicated by interrelationships that occur among the traits. Factor analyses were conducted on three phenotypically diverse pearl millet [Pennisetum glaucum (L.) R. Br.] composites to describe the structure of relationships among yield, morphological, and physiological traits. Approximately 1000 S₀ spaced-plants from each composite were evaluated for 20 traits, and random samples of 289 S₁ progenies from each composite were evaluated for 18 of these traits. Factors extracted in S₀ and S₁ populations identified unique sets of traits that were interrelated along axes of (a) biological yield, (b) panicle size, (c) dry matter partitioning and (d) compensation between number and size of seeds. Several plant traits had large loading coefficients on the Biological Yield and also, but with opposite signs, on the Dry Matter Partitioning factor. The traits having large loading on these two factors differed between space-planted and normal-density stands, showing that environmental conditions contributed to the associations observed among traits. Correlations of S₁ with parental S₀ factor scores for the Biological Yield, Panicle Size and Seed Paramters factors produced significant correlation coefficients, indicating that these trait complexes had a genetic basis. The implications of these results for millet breeding are discussed.     

Genotypic differences in callus growth and organogenesis of eight pearl millet lines

Summary Calli were obtained from mesocotyl explants of eight varieties of pearl millet using MS medium supplemented with 2,4-D. Variable frequencies of callus initiation were observed among the varieties. Plantlets were regenerated from the calli by transferring on to the medium devoid of auxins. Incorporation of coconut milk, Kinetin, IAA or NAA did not enhance either the callus initiation or growth rate or plantlet regeneration. The eight genotypes differed among themselves in callus growth rate and frequency of shoot bud production.     

Cytoplasmic differentiation using near-isonuclear polycytoplasmic male sterile lines in pearl millet

Summary Five near-isonuclear polycytoplasmic versions of 81A₁ and two of Pb 402A₃ male sterile lines of pearl millet (Pennisetum typhoides), along with their corresponding maintainer lines (81B and Pb 402B) were studied for 14 agronomic and two disease traits, and for isozymes of peroxidase in anthers and leaves for assessing cytoplasmic differences. Significant differences among isonuclear polycytoplasmic lines of both 81A₁ and Pb 402A₃, having the same genome and variable cytoplasms, were observed for several agronomic traits. Banding pattern of peroxidase isozymes revealed clear cut differences among cytoplasms. The results provide evidence for the influence of cytoplasmic factors on the phenotypic expression of nuclear genes.     

Seed parent breeding efficiency of three diverse cytoplasmic-nuclear male-sterility systems in pearl millet

Pearl millet (Pennisetum glaucum (L.) R. Br.) hybrids, grown widely in India and to some extent in the US, are all based on an A₁ CMS source, leaving the pearl millet hybrids vulnerable to potential disease or insect pest epidemics. A comparison of this CMS system with two additional CMS systems (A₄ and A₅) in the present study based on isonuclear A-lines (seed parents) and their isonuclear hybrids showed that A-lines with the A₄ cytoplasm had much fewer pollen shedders and much reduced selfed seed set in visually assessed non-shedding plants as compared to those with the A₁ cytoplasm. A-lines with the A₅ cytoplasm had neither any pollen shedders nor did they set any seed when selfed. This showed that the A₅ CMS system imparts complete and most stable male sterility, followed by the A₄ and A₁ CMS systems. The frequency of maintainers, averaged across a diverse range of 26 populations, was highest for the A₅ CMS system (98%), followed by the A₄ (59%) and the A₁ (34%) system indicating the greatest prospects for genetic diversification of A-lines lies with the A₅ cytoplasm, and the least with the A₁ cytoplasm. Mean grain yield of hybrids with the A₁ cytoplasm was 5% more than the A₄-system hybrids, while there was no difference between the mean grain yield of hybrids based on A₁ and A₅ CMS systems. Based on these results, it is suggested that seed parents breeding efficiency will be the greatest with the A₅ CMS system, followed by the A₄ CMS system, and least with the currently commercial A₁ CMS system.     

Production of 27-, 28-, and 56-chromosome apomictic hybrid derivatives between pearl millet (2n=14) and Pennisetum squamulatum (2n=54)

Summary An interspecific hybridization program designed to transfer gene(s) controlling apomixis from Pennisetum squamulatum Fresen. (2n=6x=54) to induced tetraploid (2n=4x=28) cultivated pearl millet, Pennisetum americanum (L.) Leeke resulted in four offtype plants, two with 27 chromosomes and two with 28 chromosomes. These plants were found among 217 spaced plants established from open-pollinated seed of an apomictic 21-chromosome polyhaploid (2n=21) plant derived from an apomictic interspecific hybrid (2n=41) between tetraploid pearl millet and Pennisetum squamulatum. It appeared that a 21- (or 20-) chromosome unreduced egg from the apomictic polyhaploid united with a 7-chromosome pearl millet (2n=2x=14) gamete to produce a 28- (or 27-) chromosome offspring. Meiotic chromosome behavior was irregular averaging from 3.60 to 4.05 bivalents per microsporocyte in the 27- and 28-chromosome hybrids. The 27- or 28-chromosome hybrids, like the 21-chromosome female parent, shed no pollen, but set from 1.8 to 28 seed per panicle when allowed to outcross with pearl millet. Progeny of the 28-chromosome hybrids were uniform and identical to their respective female parents, indicating that apomixis had been effectively transferred through the egg. In addition, a 56-chromosome plant resulting from chromosome doubling of a 28-chromosome hybrid was identified. Pollen was 68 per cent stainable and the plant averaged 2.3 selfed seeds per panicle. Chromosomes of the 56-chromosome plant paired as bivalents (x=10.67) or associated in multivalents. Three to nine chromosomes remained unpaired at metaphase I. Multiple four-nucleate embryo sacs indicated the 56-chromosome hybrid was an obligate apomict. The production of 27-, 28-, and 56-chromosome hybrid derivatives were the results of interspecific hybridization, haploidization, fertilization of unreduced apomictic eggs, and spontaneous chromosome doubling. These mechanisms resulted in new unique genome combinations between x=7 and x=9 Pennisetum species.     

Genetics of induced dwarf mutants in pearl millet,Pennisetum americanum (L.)Leeke

Summary Inheritance pattern of seven dwarf mutants revealed each of them to be controlled by a single recessive gene. Tests for allelism indicated existence of two groups of dwarfs, group 1 comprising six mutants and group 2 a single dwarf. The dwarfs of group 1 were found to be allelic to D₂ of the dwarf types already known in this crop and the dwarf of group 2 was allelic to D₁ of these types.Botany Department, R.V.V.N. College, Dharanikota, India.