CROP IMPROVEMENT DIVISION Plantain/Banana Improvement Program Annual Report 1992 International Institute of Tropical Agriculture October 1993 Acknowledgement Rodomiro Ortiz wrote most of this 1992 annual report of PBIP, with substantial contributions from Dirk Vuylsteke. Sh~un Ferris wrote most of Section 6. Data and information were also provided by Indira Ekanayake (Section 2u), Rony Swennen (Section 5e), Bob Jarret (Section Sf), B. Daramola and D. Ladipo (Section 7b), Mark Winslow and Mwenja Gichuru (site characterization). Gamaliel Harry (Section 5d). Hilde Vandenhout (Section 3h). Anthony Ubi (Tables 88 and 89) and Leo Oragwa (Table 11). Data published by Bob Fullerton and T.L. Olsen were re-analyzed and are presented in Figures 36 and 37. PBIP acknowledges with gratitude the contributions of Sherry Schnapp in the molecular work, the statistical advice of Peter Walker, and the critical review of papers by other liT A colleagues. ii Contents ~. To facilitate crossreferencing, each activity has an alpha-numerical code. A number identifies each project starting from number 2. Each activity within a project is coded with the project number followed by an alphabetical letter. Specific areas of research within each activity are sub-headed in italics and the titles given to additional sub-divisions within a research area are underlined. 1. Introduction ............................................................................................................. 1 2. Developing Muss breeding capability and strategy ............................. 5 2a Musa gene bank ............................................................................................. 5 2b Germplasm characterization and organization ........................................ 5 Musa germplasm ...................................................................................... 6 AAB plantains ........................................................................................... 6 2c Plot technique studies on yield trials of plantains propagated by in vitro methods ......................................................................................... 7 Optimum plot size .......... ........................................................................... 7 Soil heterogeneity at Onne as measured by bunch weight variation in plantains .................................................................. 8 Adequate number of replications .......................................................... 8 2d Production of hybrid seed ............................................................................. 9 AAB plantains x diploid bananas .......................................................... 9 Triploid bananas x diploid bananas ................................................... 1 0 Yearly production ................................................................................... 10 2e Outcrossing, selfing and seed production in diploid bananas ............ 11 2f Variation in male fertility and 2n pollen production in diploid Musa germplasm ......................................................................................... 12 Male fertility ............................. ................................................................ 12 Interchanges and structural hybridity ................................................. 12 2n pollen production .... _ .................................................................. _ ...... 13 Seed set and male fertility ...................................... _ ...................... _ ...... 14 2g In vitro germination of seeds ...................................................................... 14 Monthly variation ...................................................... _ ....................... _ ..... 14 Yearly production ................................................................................... 14 2h Ploidy determination in plantain-banana hybrids ........................... _ ...... 15 Ploidy and stomata size and density ........................................... _ ...... 15 Ploidy predictive model based on stomata measurements ...... _ ..... 16 Causes of variation in stomata size and density ............................... 16 iii 2i Nursery roguing and production of euploid hybrids ............................... 16 Factors affecting recovery of euploid hybrids .................................... 16 Yearly production .................................................................................... 18 2j Field plot techniques and evaluation of plantain-banana hybrids ....... 18 Optimum plot sizes, number of replications and soil heterogeneity .................................. : .............. : .................................. 1 8 Flow of materials in yield trials .... .......................................................... 19 On-farm trials ............................................................................................ 20 2k Selection in early evaluation trials (EET) ................................................. 20 Triploid x diploid crosses .... ................................................................... 21 Population improvement of plantain-derived diploids ..................... 21 Secondary triploids (tetraploid x diploid crosses) ............................. 22 Secondary tetraploids (tetraploid x tetraploid crosses) ................... 22 Inbreeding ............................................................... : ................................ 22 21 Preliminary yield trials (PYT) and selection of plantain hybrids .......... 23 199011992 PYT (plant and ratoon crops) ....... ; ................................... 23 1991/1992 PYT (plant crop) .................................................................. 25 Selected black sigatoka resistant tetraploid hybrids ........................ 26 Plantain-derived diplOid hybrids and germpiasm enhancement at the diploid level .............................. : .......................... 26 2m Multisiteevaluation trials (METs) of selected- Musahybrids .................. 27 Safe movement of germplasm ................ .............................................. 27 Assessment of black sigatoka resistance inmultilocational trials .. 27 MET-1 ........................................................................................................ 28 MET-2 ........................................................................................................ 30 2n Registration of improved plantain germplasm(TMPx) ........................... 30 20 International MusaTesting Program (IMTP-1) ............... : ......................... 31 2p Advanced Musa Yield Trials (AMYTs) ....................................................... 32 2q AMMI analysis of yield trials ........................................................................ 32 2r Taxonomic approaches for banana breeding at the diploid level ....... 34 2s An ideotype for Musa improvement and an index for multitrait selection ......................................................................................... 36 Multitrait selection ................................................................................... 37 2t Breeding the False Horn gene pool using somaclonal variants .......... 38 2u Physiological research to enhance Musa breeding .............................. 39 2v Computerizing PBIP breeding program .................................................... 40 2w Gaining insight into the Musa genome and modeling the breeding strategy .................................................................................... 40 iv Segregation in 4x offspring from 3x x 2x crosses and 2n egg production .................................................................................. 41 Morphological descriptors ..................................................................... 42 Albinism .................................................................................................... 42 Bunch orientation .................................................................................... 43 Pseudostem waxiness ........................................................................... 44 Apical dominance ................................................................................... 44 Dwarfism ................................................................................................... 45 Fruit parthenocarpy ................................................................................. 47 Black sigatoka resistance ..................................................................... 47 Growth and yield parameters with quantitative trait variation (QTV) ............. ............................................................................ 49 Ploidy polymorphism ........................................................................ 49 Ploidy effects and genotype influence .......................................... 49 Combining ability in plantain germ plasm ..................................... 50 Test cross ploidy mean analysis .................................................... 51 Black sigatoka resistance and expression of heterosis ................... 51 Genotype-by-environment interaction (year, location) and broad sense heritability for growth and yield parameters in Musa hybrids ....... .................................................................................. 51 Heritability estimates ........................................................................ 52 Effect of major and minor genes in plant growth and fruit development ..................................................................................... 53 Major genes ....................................................................................... 53 Minor genes ....................................................................................... 54 Trisomic segregation ratios and genome differentiation in AAB plantains .................................................................................................. 55 Genetic sterility in plantain-derived diploid hybrids ............. ..•.......... 56 Linkage groups in Musa using conventional markers ....................... 57 Value of genetic studies in Musa .....................................•..................... 58 2x Efficiency-in Musa breeding ....................................................................... 59 Efficiency of different Musa breeding schemes through computer simulation ........................................................•...................... 60 Genotype dependency and success in tetraploid hybrid production ..................................................... : •............................. 61 Resource allocation and operational costs ........................................ 61 2y Research activities and PBIP breeding program .................................... 62 2z Current status of PBIP conventional breeding program ........................ 63 Ideotype breeding ................................................................................... 63 3. Breeding for disease and pest resistance .............................................. 64 v 3a Genetics and mechanisms of black sigatoka resistance in TMPx germplasm ...................................................................................... 64 How do the three loci controlling black sigatoka resistance interact? ................................................................................ 64 Why are the 3 x plantains susceptible? .............................................. 65 Field evaluation of BS resistance in the TMPx germplasm .. ........... 65 Likelihood of durability of horizontal black sigatoka resistance in TMPx. ................................................................................. 65 Resistance mechanisms ........................................................................ 66 3b Epidemiologyof black sigatoka and the search for durable resistance ....................................................................................... 68 Assessment of pathogen diversity and stability for resistance in M. acuminata ........ ............................................................... 68 Breeding strategy .................................................................................... 70 3c Chemical and host plant resistance strategies to control black sigatoka ................................................................................................ 71 Estimate of costlbenefit ratios for farmers using fungicides or resistant hybrids .................................................................................. 72 4. Banana improvement for the mid " to high altitudes in East Africa .................................................................................................................... 74 4a Screening for female fertility in AAA and other starchy triploid bananas ............................................................................................. 74 4b Mid-altitude cookingJbeer banana breeding ............................................ 75 Hybrids available for East Africa .......................................................... 75 EET at Onne ............................................................................................. 75 5. Biotechnology for Muss breeding ............................................... 77 5a In vitro propagation of selected genotypes for yield trials ..................... 77 5b Somaclonal variation studies ..................................................................... 77 5c Performance of suckers vs. in vitro propagules of plantain ................... 78 Soma clonal variation ............................................................................. 78 Growth and yield parameters ........ ........................................................ 79 5d Embryo rescue after triploid x diploid crosses ......................................... 80 Embryo age effects ................................................................................ 80 Effect of silver nitrate on embryo germination .................................. 81 Hormone effects on embryo germination ......................................... 81 Effect of streptomycin on seed disinfection and embryo germination ........................ ; .................................................................... 81 vi Preliminary results of early screening for black sigatoka using embryo-cultured derived seedlings .................................................... 81 5e Transgenic Musa products ........................................................................ 82 5f Molecular markers ....................................................................................... 83 Taxonomy and germp/asm organization ........................................... 83 Molecular linkage map .......................................................................... 83 Dissecting black sigatoka resistance with molecular markers ...... 84 Marker assisted selection ..................................................................... 8 5 Transmission of chloroplasts in Musa ................................................ 85 Outcrossing rates in diploid bananas ... .............................................. 86 5g Secure and rational management system for in vitro propagated and conserved germ plasm of Musa ............................ 86 6. Postharvest quality of plantains ................................................................. 87 6a Fruit quality evaluation of Musa hybrids ....................... , •.•... ; ................... 87 Dry matter ................................................................................................. 88 Fruit weight.. ....................................................... '" .......... , ....................... 88 Pulp:peel ratios ....................................................................................... 89 Fruit ripening ........................................................................................... 89 6b Sensory evaluation of Musa hybrids ........................................................ 90 Taste panel data ................. .................................................................... 90 Taste panel for Musa clones prepared by bailing green unripe fruit ............................................................................................... 91 Seed set ................................................................................................... 93 Taste panel for Musa clones prepared as fried ripe sections - 'DODO' .................................................................................. 93 Conclusion to taste panel data ............................................................ 94 6c Plantain utilisation ........................................................................................ 9 5 6d On-farm assessment of yield and quality loss from black sigatoka ... 96 6e Genetics of fruit quality ............................................................................... 96 6f Molecular and biochemical studies of Musa fruit ripening .................. 97 7. Genotype-by-cropping systems interaction .......................................... 98 7a The perennial yield potential of plantains ............................................... 98 7b Survey of changes in floristic composition within natural multispecies hedgerows under plantain cropping ................................. 98 8. NARS collaboration, training and extension activities .................. 1 00 8a Multilocational testing of plantain and banana hybrids ........................ 1 00 8b Distribution of bananas ............................................................................... 1 00 8c Training .......................................................................................................... 101 vii Group training ........................................................................................ 101 Individual training ........... ....................................................................... 1 02 Training guides .......... ............................................................................ 1 02 8d. Visitors .......................................................................................................... 102 8e. International travel ....................................................................................... 1 03 References cited In the text .............................................................................. 104 .Tables .......................................................................................................................... 105 Figures ........................................................................................................................ 176 Annexes 1. Publications and seminars about plantain and banana by PBIP staff ......................................................................................................... 205 2. PBIP staff (April 1992-March 1993) ............ ............................................... 207 List of Abbreviations ............................................................................................. 208 About the liT A High Rainfall Station .................................... inside back cover viii 1. INTRODUCTION Plantain and banana (Musa spp.) are important food crops in the humid forest and mid·altitude agroecologies of sub-Saharan Africa, providing more than 25% of the carbohydrates for approximately 70 million people in the region. Plantains are high-yielding and form an integral component of the farming systems in these ecologies. In addition to being a staple food for rural and urban consumers, plantain and banana provide an important source of rural income for the smallholder family farmers, who produce them in compound or home gardens as well as in large fields. The gross value of annual production exceeds that of many other major food'crops, such as maize, rice, cassava, and sweet potato. Most cultivated Musa are triploids (2n = 3x = 33). Being almost completely sterile, they develop fruit by parthenocarpy. The most important cultivars vary in their genomic constitution and are generally categorised as follows: dessert bananas (AAA) , East African highland bananas (AAA), plantains (AAB), and cooking bananas (ABB). The genome of the cultivated types is derived from the diploid wild species M. acuminata and M. balbisiana, which contributed the A and B genomes, respectively. World production of plantain and banana is estimated at 70 million metric tons of which 35% is grown in sub-Saharan Africa. The different cultivated triploid Musa types are cultivated in distinct ecoregions. Thus, the AAB plantains are predominant in the humid lowlands of West and Central Africa, while AAA cooking and beer bananas prevail in the East African highlands. The former region harbours the world's greatest variability of plantains and is thus considered as a secondary centre of plantain diversification. A similar situation is found for the specific East African highland bananas of the Musa AAA group, reflecting a long history of plantain and banana cultivation in Africa. Pest and disease pressure on the crop has increased during the past 15 years. Also, rising population pressure has led to altered farming practices, among which shortening fallow periods are most conspicuous. At present, the major production constraints of plantain and banana in sub-Saharan Africa are diseases (black sigatoka, Fusarium wilt), pests (weevil, nematodes), and the phenomenon of yield decline (mainly in plantains). Declining soil fertility also accounts for decreased yields. liT A included plantain and banana among its mandate crops in 1987 and the Plantain and Banana Improvement Program (PBIP) was created in 1991 as a 1 full part of the Crop Improvement Division (CID). PBIP headquarters are al Onne Station, in the humid forest zone of southeastern Nigeria, which is in the secondary centre of plantain diversification .. During 1992, PBIP has further strengthened its linkages with several ,international, regional and national programs: • the International Network for the Improvement of Banana and Plantain (INIBAP) in the areas of germplasm exchange and evaluation, virus indexing, and training, • the Centre Regional Bananiers et Plantains (CRBP, Cameroon) for evaluation of breeding materials, • the Inter-Ministerial Committee on the Control of Black Sigatoka Disease of Plantain in Nigeria, the National Horticultural Research Institute (NIHORT), the National Root Crops Research Institute (NRCRI), and other Nigerian NARS for evaluation of PBIP breeding materials in Nigeria, • the Nigerian Plant Quarantine Service for movement of germplasm, • the Crop Research Institute (CRI, Ghana), University of Ghana (Agricultural Research Station-Kadel, Institut des Forets (IDEFOR ,Cote d'ivoire), Institut de Recherche Agronomique et Zootechnique (lRAl, Burundi), Uganda National Banana Research Program (UNBRP), International Center of Insect Physiology and Ecology (ICIPE, Kenya), Fundacion de Desarrollo Agropecuario (Dominican Republic), Instituto Nacional de Investigaciones en Viandas Tropicales (INIVIT, Cuba), and Queensland Department of Primary Industries (QDPI, Australia) for evaluation of brE1eding materials, • the Fundacion Hondurena de Investigacion Agricola (FHIA, Honduras) for germplasl11 exchange and testing of FHIA hybrids in Africa, • the Katholieke Universiteit Leuven (KUL, Belgium) in biotechnology, • the Regional Plant Introduction Station of the USDAIARS (Griffin, GA, USA) in the use of molecular markers (RFLPs and RAPDs) in taxonomy, genetics and breeding • the Natural Resources Institute (NRI, England) in post-harvest technology During the past year, PBIP scientists worked along with their cooperators at IITA. NARS and advanced laboratories in six research projects (Table 1). Our main achievements during 1992 are listed below. Hybrid germplasm registration. Fourteen improved tropical Musa plantain hybrids (TMPx) with black sigatoka resistance have been registered in the Journal HortScience to place these in the public domain. 2 Multilocatlonal testing. TMPx genotypes were transferred to NARS for the evaluation/selection process in the multilocational testing. Twenty five public and private NARS from Cote d'ivoire, Ghana, Nigeria, Cameroon, Burundi, Uganda, Kenya, Zanzibar, Cuba, Dominican Republic and Australia are evaluating TMPx genotypes along with one cooking banana hybrid (TMBx) and their respective resistant and susceptible parents, as well as susceptible and resistant reference cultivars. Determination of breeding strategy. The genetics of resistance to black sigatoka, dwarfism, albinism, suckering behaviour, fruit parthenocarpy, bunch orientation, bunch weight and other yield components were established. These findings provided the basis for development of a breeding strategy which entails production of improved secondary triploid germ plasm, the final target in plantain breeding, by intercrossing tetraploid and diploid plantain hybrids with black sigatoka resistance (Fig. 1). A first wave of hundreds of secondary triploid hybrids were field-established in 1992. Molecular linkage map. A molecular linkage map is being developed in collaboration with the USDA laboratory at Griffin, Georgia. Six segregating populations derived from triploid x diploid and diploid x diploid crosses were field-established and characterized for different morphological and agronomic descriptors. DNA was extracted from each clone and fingerprinted with RFLP and RAPD markers for genetic analysis. The development of a saturated Musa map is in progress. Tissue culture and genetic engineering. The production of transgenic Musa products has been the major goal of the common research effort with KUL. A cell suspension protocol for the regeneration of Musa genotypes was developed jOintly by scientists of both institutes. In 1992, our cooperators were able to regenerate plant lets which showed stable transformation with the GUS reporter gene and kanamycin resistance gene, introduced by the particle gun method. Agro1orestry and sustainable production. Previous agronomy research has clearly demonstrated that management of organic matter is essential to maintain long term plantain productivity without causing environmental degradation. We have observed that plantain performs very well in an agroforestry system with spontaneous multispecies hedgerows, in which the 3 regrowth of more than 120 shrub and tree species of the natural fallow is managed by regular coppicing. This system seems promising in terms of maintaining p.roduction over extended periods of cultivation as well as conserving the plant diversity for bush fallows. Research related activities during the 1992 season (April 1992 to March 1993) comprising seven research projects (coded projects 2 through 7) are reported in Sections 2 to 7. The final section (Section 8) reports on collaborative work in respect of testing, training and extension. 4 2. DEVELOPING MUSA BREEDING CAPABILITY AND STRATEGY Plantain and banana breeding continued to be the most important activity in PBIP during 1992. Significant gains in 1991 opened up many possibilities for breeding across ploidy, species and subspecies of Musa. Based on those exciting results. we have decided to intensify and decentralize the breeding work into different agro-ecologies and gene pools. However, there is stili a need at· this stage to consider breeding strategies and determine the most useful methods for improving this little-researched crop for African conditions and constraints. Genetic research and breeding activities indicated below will help to achieve this goal. Recently, we defined PBIP's breeding philosophy as follows: tQ deyelop not only disease and pest reSistant materials. but also better cultivars for the smallholders who produce plantain and banana in compound or home gardens and in large field plantations. 2a. Muss gene bank The main goal of this activity is to coliect, conserve and characterize Musa species and cultivars and evaluate them in the African humid forest environment represented at Onne. Conserved germ plasm (>400 accessions) includes diploid species and cultivars (M. acuminata -AA and M. balbisiana -BB), dessert bananas (AAA. AAB), plantains (AAB), and cooking and beer bananas (AAA and ABB) (Table 2). Over 30 sources of black sigatoka resistance are available in the collection (Table 3). Ali accessions are kept in field genebanks. Most of the plantain accessions (85%) are currently also maintained in vitro in the Tissue Culture laboratory at Onne. The in vitro collection is partly duplicated at the Ibadan tissue culture laboratory for safety reasons, where it is maintained under minimal growth conditions. 2b. Germplasm characterization and organization Descriptor lists are used to evaluate natural Musa germplasm (diploids. triploids and few tetraploids). Multivariate analyses such as prinCipal components (PCA) and cluster can be used for the statistical grouping of the germ plasm. Multivariate statistical techniques, such as PCA were developed for situations where more than one variate was measured on each individual. PCA 5 aims to explain the variance/covariance structure with a few (usually 2 or 3) linear combinations of the original variables. The PCA coefficients (e.g. PRIN 1, PRIN2, etc.) are functions of the eigenvalues (Au and eigenvectors (eD of the covariance/variance matrix. They are basically roots of the matrix. They can be used in the development of a discriminant function to assign clones to each taxonomic group. Qualitative morphological descriptors have been extensively used to group different accessions in taxonomic groups with similar characteristics. However, breeders are interested in agronomic traits which display quantitative variation. The objective of this activity, in support of the breeding program, was to determine whether quantitative descriptors could be used for characterization between and within taxa in Musa germplasm. This may provide a basis to define different gene pools to maximize heterotic responses when hybrids are produced. In 1992 we started the statistical grouping of germplasm accessions in liT A's gene bank through PCA. Musa germp/asm Preliminary principal component analyses on published data (Ref. 1) of an Asian collection from the Philippines banana cultivars using quantitative growth & yield parameters (pseudostem height and girth, leaf length and girth, bunch weight, number of hands, and fingers in the second hand, finger length and diameter, weight of peel and number of suckers at harvest) did not result in any clustering for M. acuminata accessions, which had a wide dispersion in the principal component diagram (Fig. 2). Nevertheless, AA and AAA bananas did not overlap (data not shown). Conversely, most of the AAA bananas overlapped with the other cultivars in the AAB and ABB taxa. This could indicate that PCA based on bunch and vegetative quantitative traits should only be considered to discriminate between ploidy levels in the same taxon. The first principal component, which explained 87% of total variation, was unevenly loaded on plant height and leaf length while the second principal component (7% total variation) was loaded mainly on weight of peel. Therefore, the right upper quadrant included tall cultivars with heavy bunches of large fingers and high weight of peel, e.g. AAB Horn plantains and ABBB natural hybrid. The ABB cooking bananas, which were initially considered as BBB ('Cardaba', 'Saba', etc), were grouped in the right lower quadrant, which was characterized for containing cultivars with large leaves. This demonstrates that some quantitative traits along with PCA could be used to discriminate between taxonomic groups in Musa. 6 AAB plantains There are at least 116 plantain cultivars that have been grouped according to their bunch type (Table 2) and plant size. PCA was used to examine the association between growth and yield parameters in the West African plantain germ plasm , and to find out if the pattern of quantitative variation of bunch and vegetative traits agreed with the taxonomic grouping based on bunch type and plant size. A group of 24 plantain cultivars representing the major variability in West Africa were evaluated for nine quantitative characters. Phenotypic correlations between these traits were calculated (Table 4). Giant cultivars are taller, thicker and flower much later than medium ones and they produce more foliage, resulting in heavier bunches with more hands and fingers. PCA grouping agrees with conventional plantain taxonomy (Figs. 3 and 4). peA was based mail1ly on time to flowering, pseudostem height and number of fingers. The lasttwo traits should be used in combination as discriminant function for grouping plantain cultivars based on the phenotypic correlations and the eigenvectors, which are the weights in the peA (Table 5). The results also suggest that Nigeria has been a transit centre of plantain germ plasm movement in West Africa (Fig. 5). 2c. Plot technique studies on yield trials of plantains propagated by In vitro methods. In vitro techniques are used for conservation, distribution and rapid multiplication of plantains. Phenotypic or somaclonal variation among micropropagated plants of Musa is a common phenomenon.Determination of the yield potential of plantains should thus consider this, as well as proper field plot techniques. The aim of this research was to determine the optimum plot size and adequate number of replications for yield trials of different plantains propagated in vitro. The cultivars used represented the major plantain taxonomic groups: 'Ntanga 2' (Giant French), 'Sob by Tannap' and'Obino I' Ewai' (Medium French), 'Big Ebanga' (Giant False Horn), 'Agbagba' (Medium False Horn) and 'Ubok Iba' (True Horn). Optimum plot size (OPS) A total of 120 true·type plants were sampled and their bunches weighed for each cultivar (Table 6). No data transformations were required because the variance and the standard deviations were independent from the mean (lack of correlation) and the variance of the six plantain cultivars were homogeneous 7 according to Box's rule. The intra-clonal variance for bunch weight was not affected by the reported somaclonal variation in each cultivar (Table 7) . However, a negative trend was observed between both which suggests that selection against somaclonal variants was effective to reduce within clonal variation for bunch weight. OPS was determined using the method of maximum curvature by plotting the coefficient of variation (V axis) versus the number of competing plants per plot (X axis) (Flg_ 6). OPS (without including border plants) consisted of 8-12 (48-72 m2), 15 (90 m2) and 20 (120 m2) competing plants for French, False Horn and Horn plantains, respectively (Table 7). Soil heterogeneity at Onne as measured by bunch weight variation in plantains The coefficient of soil heterogeneity (b) was estimated by the variation of bunch weight within each cultivar as qetermined by solving the following equation: Vx = Vi~ where Vx is the variance (calculated on a per unit basis) of bunch weight per unit area among plots of x units in size, V, is the variance among plots of one unit in size and x is the number of basic units per plot (Fig. 7). Soil heterogeneity was smaller in the French plantain (0.25) plots than in Giant or Horn plantain plots (0.9-1.5) (Table 7). Values of b close to 1 (expected maximal theoretical value) indicate that replicates were completely independent. Therefore, big plots instead of small plots, which will require an increased number of replications, should be used for the evaluation of Giant or Horn plantains. On the contrary, the bunch weight of French plantains may be evaluated in small plots. Adequate number of replications The minimum number of replications (r) was determined using the following equation: r= (MSE x 2) • (tl + tv21 D2, where t, is the critical value of Student's t distribution for a significance level of (J. =0.05, t2 is the tabular Student's value for (1-P ) =0.20, where P is the probability to obtain a significant result, MSE is the error mean square for bunch weight (from a yield trial carried out in a randomized complete block design using 15 plants per plot) and 0 is the desired significant mean difference expressed in kg. The number of replications (reps) needed decreased concomitantly with an increase in the magnitude of 0 (Table 8), e.g. assuming a MSE = 1.85 and 15 degrees of freedom for the t values, 192 reps of 15 plants each were required to detect a sign ificant mean bunch weight difference of 0.5 kg whereas 8 reps of 15 plants each were enough to detect a significant mean bunch weight difference of 2.5 kg between 2 8 plantain cultivars. The detection of significant mean diHerences larger than 4 kg between plantain bunch weights requires only 2 reps of 15 competing plants. 2d. Production of hybrid seed The success of any cross breeding program requires the production of true seed through sexual hybridizations as the first step. This was considered as the most diHicult problem for the genetic manipulation of the triploid cultivated Musa cultivars. However, the identification of 37 seed-fertile plantains (29 French and 8 False Horn cultivars) has opened possibilities for genetic improvement in this gene pool. AAB plantains x diploid bananas Seed set has been found to vary greatly among seed-fertile plantain cultivars, ranging on average from <1 seed to > 20 per bunch. Furthermore, season influenced seed rates. Seed production in two French plantain cultivars, 'Bobby Tannap' (BT) and 'Obino I' Ewai' (Ol), was analyzed over a period of 5 years (1988-1992) to determine the causes of monthly variation in seed set. These two clones were selected because they produced a high frequency of euploid (diploid, triploid and tetraploid) hybrids. The common male parent in all the crosses was 'Calcutta 4' (C4), a wild non-edible diploid banana with high male fertility. Analysis of variance indicated that seed production was not aHected by year. However, significant differences were found between months of a year and between the two plantain cultivars. The interactions, months x years and crosses x months were not significant, which allowed pooling of the results. Fig. 8 illustrates the pooled pattern of monthly variation for seed production (average of 5 years) in each cross combination. Seed set was very high in the early part of the year (peak in February), then declined to a very low level in May and reached a second peak in August-September. The seed set pattern was correlated with monthly weather variation, as measured by rainfall, evaporation, air temperature, wind speed, solar radiation and relative humidity, at Onne. Significant correlations were found between maximum relative humidity and seed set ( r= 0.563, p=0.05) for the cross BT x C4 whereas the variation in seed set was negatively associated with changes in total rainfall (r = -0.600, P < 0.05), and minimum relative humidity ( r = -0.620, P < 0.05) and positively correlated with average relative humidity in the cross OL x C4. 9 A simple linear regression model, seed set = -487.7 + 5.3 maximum relative humidity, explained 31.7% of total variation in seed set in the cross BT x C4. The standard error for seed set estimates using this model will be 6.4 seeds. Residuals' analysis for this model (Durbin-Watson statistic = 1.10) however, indicated the presence of serial correlations in the residuals. The standard error for seed set estimates in BT x C4 using this incomplete and inaccurate model will be 6.4 seeds. More than fifty percent of total variation in seed production in the cross OL x C4 was explained by the significant multiple linear regression model (F = 4.76, P < 0.05, R2: 0.514), seed set = -111.9 - 1.6 minimum relative humidity + 2.7 average relative humidity. Moreover, the analysis of residuals . (Durbin-Watson statistic = 1.76) demonstrated the validity of this model for the prediction of seed set in OL x C4 with a level of accuracy of ± 2.5 seeds. Triploid bananas x diploid bananas Similarly, monthly seed set patterns were studied in the crosses between C4 (male parent) with the AAA dessert banana 'Yangambi Km. 5' (Fig. 9) and the AAB cooking banana cv. 'Bluggoe' (Fig. 10) over a two year period (1988- 1989). In these two clones, characterized by a low frequency of tetraploid progenies as a consequence of their low 2n egg production, the monthly variation in seed set followed the changes in climatic factors. Based on seed set in the cross 'Yangambi Km 5' x C4 during the two year period, the following multiple regression model, which explained 62.5% of total variation on monthly seed production, was developed: seed set = 225 + 0.025 total rainfall -0.120 solar radiation· 3.374 maximum temperature - 0.973 average relative humidity. The standard error for the expected seed set in this significant model (F=6.25, p < 0.01) was calculated as 7 seeds per bunch, although the analysis of residuals (Durbin-Watson statistic = 1.99) suggested that this model should be taken with caution due to the serially correlated data structure. In conclusion, it was clear that seed set in triploid bananas was increased by high relative humidity, low irradiation and low temperature which are characteristics of months such as August and September. This information will assist PBIP breeders to decide when to do pollinations based on weather information throughout the year. Consequently more progeny will be produced, thereby improving Musa breeding efficiency. Yearly production In 1992, a total of 4219 bunches of diploid, triploid and tetraploid accessions/cultivars as well as euploid plantain-banana hybrids were pollinated 10 with pollen from diploid or tetraploid parents, resulting in a production of more than 61000 seeds. Table 9 lists the monthly number of bunches pollinated for breeding activities, pollination success (as measured by percentage of bunches with seeds after hand pollination), total seed production, and average seed set (seeds per bunch) in the most important breeding schemes implemented during 1992. Table 9 also reflects PBIP's change of breeding strategy during the past year. Although the main activities were the pollination of plantain cultivars (58% of the crosses) and germ plasm enhancement in the Musa acuminata gene pool (20%), breeding plantains at the diploid level and the production of secondary triploids via tetraploid x diploid crosses were incorporated in the crossing blocks. Moreover, initial results have been encouraging as indicated by the pollination success. High seed sets were obtained in tetraploid x diploid crosses, which challenges the common idea that tetraploid hybrids have low female fertility. The highest percentage of pollinated bunches with seed production were found after tetraploid x diploid crosses (52%) and diploid x diploid crosses (28%) when at least the female parent was a plantain hybrid. Results also demonstrate the limited success of seed production by selfing, sib-mating or cross poliination of tetraploid hybrids with pollen from another tetraploid. There was 15% pollination success with an average seed set per bunch of 1.3 after tetraploid x tetraploid crosses. This may be an indication that tetraploid hybrids havE> low male fertility. Furthermore, production of secondary triploids, which for us could be the fixed target in plantain breeding, and the utilization of diploid plantains for population improvement seem to be the most efficient alternatives for the genetic betterment of the AAB gene pool. This was indicated by the high rate of seed production with low investment in pollination, i.e., 25% and 17% of total seed produced at Onne during 1992 resulted aiter pollinating diploid (10% of total pollinations) or tetraploid plantain hybrids (4% of total pollinations), respectively, with a diploid parent. 2e. Outcrosslng, seltlng and seed production In diploid bananas 'Calcutta 4', the homozygous wild diploid seeded clone was used to determine the effect of inbreeding and outcrossing rate on seed production. Seeds were extracted from selled, open pollinated (OP) and non-pollinated but bagged bunches of C4. A comparison between hand pollination (C4 selled) vs OP was done. Early results (start 01 rainy season) indicated that OP yielded more seeds (an average 01 1400 seed per bunch; 330 with well formed embryos as 11 determined by seed floating test) than selfing (an average of 150 seeds; 14 with well formed embryos). No seeds were obtained in bagged and not pollinated inflorescences. However, increased seed set was observed in later self-hand pollinations which were done at the end of the rainy season. Furthermore, seeds were also obtained in the non-pollinat~d bunches (which were bagged to avoid pollination by bees). The results suggest that apomictic seed development could be possible in the female flowers of C4. This information, along with current investigations on the behaviour of natural pollinators (bats, bees and birds), will allow us to develop a polycross scheme for mass hybrid seed production without hand pollination. , , 2f. Variation of male fertility and 2n pollen production in diploid Muss germplasm Seed set was not only influenced by genotype of the female parent or the pollination season, as indicated in Section 2d., but also by male parent as illustrated in Table 10. The highest number of seeds were always obtained when C4 was the male parent. The low level of male fertility of 'Pisang Iii in' (PI) is the result of its structural hybridity due to an interchange (reciprocal translocation heterozygote), This explains why PI produces about half as much pollen as C4. Male fertility A research activity was started in October 1992 to determine male fertility through microscopic studies and identify the period of maximum pollen viability as measured by its stainability with acetocarmine glycerol. This activity is continuing, Preliminary results of pollen stainability from October 1992 to March 1993 are presented in Fig. 11. C4, as expected, and 'Padri' had the highest and most stable levels of pollen stainability during this six month period while the plantain diploid derivative TMp2x 1297·3, 'Galeo', 'Po jari buaya' (PJB), PI and its hybrid TMb2x 384-30 had a seasonal variation for pollen stainability. Low pollen stainability levels were found at the end of the rainy season at Onne (October/November) but a significant increase in pollen stainability was observed towards the dry season. It seems that high irradiation, high temperature and/or low humidity in the environment favoured male fertility, as measured by pollen stainability, Interchanges and structural hybridity The pattern of pollen stainability variation in PI and 'Galeo' might be the consequence of their structural hybridity. The proportion of unbalanced meiotic 12 products (non-stained pollen grains) is determined by the position and number of chiasmata (in the pairing and interstitial segments), chromosome distribution and centromere orientation. Adjacent distribution, either due to chromosome non-disjunction or disjunction, leads to partial sterility or semi-sterility of translocation heterozygotes. This happens because all or part of the meiotic products are genetically unbalanced due to duplications or deletions. Otherwise, in alternative distribution ,chromosomes bearing non-homologous translocated segments are distributed to the same pole in the first meiotic division. This makes possible the production of functional gametes. With ring formation, alternative distribution must be higher than adjacent segregation, thus more than 50% of viable gametes are expected. However, in the case of chain formation, the two types of distribution have the same frequency which increases the level of sterility. Ring and chain formation depend on chromosome size and chiasma formation. For example, when chiasmata do no form in all the arms or the point of interchange is near the chromosome end, a chain of four chromosomes will be observed. The results, therefore, suggest that crossing over varied in the translocation heterozygotes due to seasonal influence. This information indicates that mapping distances should be taken with caution because recombination frequencies, which are relative measurements, may be influenced by environment. Similar patterns for pollen stainability in TMb2x 384-30, a progeny of M. acuminata x PI, and PJB suggested that they could also be translocation heterozygotes. 2n pollen production The occurrence of giant pollen has been considered as an indicator of 2n pollen production. C4, previously reported as non-2n pollen producer, had normal pollen with a diameter not exceeding 128 microns ().l). Diploid or 2n pollen must have 1.25 times the linear dimensions of haploid or n pollen. Therefore, giant pollen grains having diameters equal to 160).l were classified as 2n pollen. Pollen sizes significantly bigger than the giant pollen size, as defined above indicates more irregular meiosis due to the occurrence of double restitution (4n pollen). PI and its hybrid progeny TMb2x 384-30 had such pollen sizes in varying frequency (Fig. 12) during the six-month period. Also pollen grains bigger than 128).l but smaller than 160).l were observed in PI and TMb2x 384-30. These pollen grains will' be hyperhaploid but also hypodiploids due to unequal distribution of chromosomes to the poles in meiosis, a common feature of some translocation heterozygotes. 13 , It was not surprising that 2n pollen frequency varied between both clones and throughout this study because environmental and genetic factors as well as incomplete penetrance and variable expressivity have been reported to affect 2n pollen production. Production of 2n pollen decreased from October to January or March in 384-30 or PI, respectively. It 'seems, according to preliminary analyses, that 2n gametes in TMb2x 384-30 incr~ased under stressful environments characterized by high temperature and low relative humidity. Also 2n pollen production did not affect male fertility. The correlation coefficient, r = 0.05, was not significant between these two variables in TMb2x 384-30, but a non- significant negative trend (r=-0.60) was observed between pollen stainability and 2n pollen production in 'P. !ilin'. These findings suggest that 2n pollen production was independent of male fertility. Seed set and male fertility Data from controlled crosses (diploid x diploid and triploid x diploid) will be correlated with pollen stainability variation to determine the effect of male fertility on seed production. This study will be completed in 1993. 2g. In vitro germination of seeds Hybrid plant production in the most common triploid Musa clones is hampered by low seed production (see Section 2d) and by low seed germination rates. Seeds of plantain crosses germinate in soil at a rate of about 1 %. Aseptic embryo culture techniques are routinely applied to increase seed germination rates by a factor of 3 to 10. Monthly variation Three years data on embryo culture of plantain seeds obtained after triploid x diploid crosses were analyzed. Seed success (percentage of seedlings germinated over total seed production) and embryo success (percentage of seedlings germinated over total seeds with embryos) were not affected by the year effect (P > 0.05). However, a significant monthly variation was observed (Fig_ 13) for both seed success (P < 0.05) and embryo success (P < 0.01). The interaction year x month was significant ( P < 0.05) only for seed success. Yearly production During 1992, a total of 14570 seeds, produced from crosses on plantain or derived diploid and tetraploid hybrids, were handled in vitro at the Onne tissue culture laboratory, in support of the breeding program. An additional 1530 14 seeds derived from the cross BT x C4 were handled by a research fellow for his embryo culture investigation (see Section 5e). This gave an average of 1342 hybrid seeds handled on a monthly basis. Great fluctuations in germination rate were observed during the year (Table 11). however comparisons with previous years should be done with caution, because in 1992, seeds from secondary triploids and tetraploids as well as diploid-derived plantains were germinated in vitro. In vitro seed handling in 1992 was slightly smaller than in 1991. The average in vitro germination rate of 1992 showed an absolute increase of 0.7% over 1991 results, probably as a consequence of the utilization of euploid parents (either tetraploids or diploids) in the production of secondary triploids or in the population improvement of the plantain gene pool at the diploid level. 2h. Ploidy determination in plantain-banana hybrids Verification of ploidy levels in progenies from triploid x diploid crosses is required due to the production of diploid, triploid, tetraploid, aneuploid and hyperploid hybrids. Ploidy levels are estimated by phenotypic appearance and confirmed either by root-tip chromosome counting or stomata size and density. Protocols for DNA flow cytometry analysis and ploidy determination using in vitro planUets are currently in development. Preliminary results obtained by evaluating plantain-banana hybrids in the seedling nursery indicated that clones varied significantly for both stomata size (P < 0.001) and density (P < 0.001). Variation in pollen sizes was studied in flowering plants. No correlation was found between ploidy level and pollen size. Ploidy and stomata size and density A Belgian research fellow studied the reliability of ploidy determination using stomata measurements in plantain-banana hybrids by correlating these traits with chromosome counts using root tips. In general, size and density of stomata, which are negatively correlated (r =-0.62, P = 0.004), varied according to ploidy level. Diploid hybrids had on average 29 stomata mm-2with an average size (length x width) of 0.125 mm2, while tetraploids had on average 15 stomata mm-2 with an average size of 0.184 mm2. Both traits had a linear relationship with changes in ploidy level. Stomata size and ploidy were positively correlated (r = 0.59, P = 0.006) while stomata density and ploidy were negatively associated (r = -0.57, P = 0.008). The simple linear regression models explained only 33% and 35% of variation in stomata density and stomata size, respectively. 15 Ploidy predictive model based on stomata measurements A multiple regression model to predict ploidy based on stomata measurements was developed. The relationship [ploidy (as measured by chromosome sets) = 1.81+ 10 stomata size - 1.76 x 10-2 stomata density) had a R2 of 43%. Despite this being a statistically significant model (F = 6.39, P = 0.009), the test of residuals (Durbin-Watson statistic = 1.36) was inconclusive. Moreover, the standard error of the estimate (0.7) indicated that it will be difficult to discriminate between diploid and triploid hybrids or triploid and tetraploid hybrids with this equation. For example, the East African cultivar 'Kamaramasenge' had significantly lower stomata density and larger stomata size than C4. Furthermore, its stomata density is lower than that of the triploid plantain cv. 'Bungaoisan', but 'Kamaramasenge' stomata have on average bigger size. This was surprising because 'Kamaramasenge' has been considered as a sterile AB diploid banana, but stomata measurements suggest that this clone could be a triploid. Therefore, chromosome countings are the only accurate proof of ploidy levels in Musa germplasm. Causes for variation in stomata size and density Regression analyses, based on ploidy levels, for stomata size and density measurements were not able to explain more than 35% of their total variation. Therefore, hierarchical analyses of variance (Table 12) were carried out to determine what other factors were involved in stomata polymorphism. These characteristics were influenced not only by ploidy level within each population but also by a significant genotype effect within the same ploidy level. The ploidy effects were higher in the cross OL x C4 than in BT x C4, while the reverse was true for the clone/ploidy/population effects. Table 13 summarizes the evaluation of stomata size and density between ploidies in OL x C4 and BT x C4. 2i. Nursery roguing and production of euploid hybrids Ploidy polymorphism was commonly observed in progenies of triploid x diploid crosses. Aneuploids and hyperploids characterized by their abnormal foliage and/or weakness were rogued at the nursery stage. Factors affecting recovery of euploid hybrids The monthly production of total euploid and tetraploid hybrids in OL x C4 and BT x C4 crosses for the period 1988-1989 is shown in Fig. 14. Seasonal variation followed the same bi-modal pattern as observed for seed set in 16 plantain x banana crosses and 2n gamete production in diploid banana. An overlapping of Fig. 8 (monthly seed production), Fig. 12 (monthly 2n gamete production) and Fig. 13. (monthly seed and embryo success) with the cumulative monthly production of euploid and tetraploid hybrids (Fig. 14) allows us to summarize the results as follows: (i) From the last part of January to mid-March the recovery of tetraploid hybrids at Onne was high, despite the low rate of both seed and embryo success. Seed production was high in this period, but most of the seeds lacked embryos. It seems that the high seed set of February in triploid x diploid crosses had its origin in the high frequency of 2n eggs in the plantain parents as demonstrated by the number of tetraploid hybrids recovered in this month and the poor seed quality (ii) Few seeds were obtained after triploid x diploid crosses from the last days of March to early July_ However, most of the seeds had embryos which resulted in a high rate of seed and embryo germination. Despite the high quality seed production, no tetraploids were recovered. This was due to the production of normal haploid gametes (either n = 2x or n = x) in the plantain parent (iii) There was an unimodal pattern of seasonal variation for seed set, embryo and seed success, and euploid and tetraploid hybrid production from mid- July to December, reaching its peak in September, a month with low temperature, low solar irradiation and high relative humidity. In general, seed had average quality and production of 2n gametes was higher than during the March-July period but lower than in the January-March period (Fig. 14) Based on resource allocation (cosVbenefit ratio), production of tetraploid plantain hybrids, recovered from triploid x diploid crosses, may be more efficient when pollinations are concentrated during the period from end-January to mid- March at Onne or in environments characterized by high temperature, high solar irradiation and low relative humidity. Probably such environments are better suited to 2n gamete production than n or unbalanced (aneuploid) gamete production. The lack of correlation between seed set, embryo and seed success with production of tetraploid hybrids suggests that the fundamental rule of the game of numbers does not always hold: " the larger the population size the higher the chance to recover the desired product". Also, this relationship demonstrates that 'brute force', I.e., more pollinations to recover more tetraploids, does not work in plantain breeding. Production of tetraploid ,hybrids in the period end-July to 17 December requires more investment that in the period end-January to mid- March. The results suggested that the basic rule of the game of numbers may only apply for the production of diploid hybrids. Yearly production From May 1992 to September 1992, about 2700 seedlings were screened in the nursery before field establishment (Table 14). Most of the rogued plants were either aneuploids or had poor vigour. The numbers reflect the emphasis on plantain breeding of PBIP and the reduction of banana breeding efforts at the diploid level. 2i. Field plot techniques and evaluation of plantain-banana hybrids Although plantain and banana have previously been considered intractable to conventional genetic improvement, due to Iriploidy and the high level of sterility, large numbers of high yielding tetraploid hybrids are now being produced at IITA and FHIA. This proliferation of promising hybrids has led to the need to develop a scheme for hybrid selection. Musa breeding consists of two stages: (i) The production of promising hybrids (ii) Testing of elite hybrids before release to farmers. A major limiting factor for testing the yield potential of plantain and banana is its land requirement: each plant requires 6m2. The objective of this research was 10 determine the minimum number of sampled plants needed per clone, and the most efficient field layouts for testing bunch weight at different stages of the breeding program. Optimum plot sizes, number of replications and soil heterogeneity Data from a yield trial, comprising plantain-banana hybrids along with their plantain parent, were used to determine the optimum plot size and number of replications required per location for a yield trial. The method of maximum curvature was used to determine optimum plot size (Fig. 15). The inflection point on the graph indicated that four competing plants constitute the minimum number of plants in the experimental plot for yield trials of French plantain derived hybrids. The negative relationship between the log mean square error of bunch weight and log number of plants sampled per plot (Fig. 16) was expected, based on the empirical relationship indicated in Section 2c. The coefficient of 18 regression of the equation in Fig. 16 (b = 0.51) was the measurement of the index of soil heterogeneity for the experimental field in which the preliminary yield trial was carried out. The coefficient of soil heterogeneity was inconclusive to determine if the plots of any size were serially correlated. This happens when b=O, which indicates that nothing will be gained by increasing the plot size. The probability of having Type I error (a), or false rejection of null hypothesis: 'all the treatments are equal', and Type II error (~), or false acceptance of null hypothesis, and the availability of breeding materials were considered in calculating the plot size and number of replications required for a statistically adequate comparison of bunch weight (Table 15). Flow of materials in yield trials In the early stage of the breeding process, performance of the plantain or banana parents is compared with the performance of selected hybrids. In more advanced stages of the breeding process, the relative performance between hybrids can be used as the selection criterion for the best genotypes. In preliminary stages of the program (Early Evaluation Trials, EET, or Preliminary Yield Trials, PYT) a large number of clones are evaluated (100 or more) to determine which are to be selected for further testing. Planting material is limited at this stage, typically 1-5 plants/clone, and material is only available for unreplicated plots or limited replications in EET or PYT, respectively. In the early selection stages, it is most important not to reject valuable material. We therefore use a probability level of 10% (EET) or 5% (PYT) to select those hybrids with at least equal (EET) or significantly higher (PYT) bunch weight than the parental cultivar. This approach keeps Type II error low. In more advanced stages of the program (e.g. Multilocational testing or Advanced Trials), elite hybrids should be differentiated with a significance level of 1% (highest hybrid mean - 2 or 3 SE) or 0.1% (highest hybrid mean - 3 or 4 SE). This requires an increase in the number of replications to protect against Type II error. Currently, PBIP uses augmented designs in EET on the breeding station; the plantain cultivar is evaluated in replicated plots, but non-replicated plots of 5 plants each are used for the hybrids. Selected materials are evaluated in randomized complete block design (RCBD) with 2 replicates of 4 plants each for PYT in one location, and in RCBD with 2 replicates of 5 plants for multilocational testing in at least 10 different locations of West and Central Africa (Fig. 17). For advanced testing of elite hybrids PBIP recommends 1 of 2 strategies: 19 (i) National programs use a trial design such as a RCBD with 4 replicates of 5 plants for advanced testing of elite hybrids, for at least 2 years (plant crop and ratoon) and at 10 different sites in the targeted ecoregion, or (ii) National programs reduce replications to 1-2 per site but increase the number of sites to be between 30-40 in the target area On-farm trials The development of a cultivar profile to identify proper agronomic practices (alley cropping, N-P-K fertilization, etc) should be fine-tuned by on-farm testing (OFT) of the TMPx along with a local cultivar. OFT requires plots comprising at least six competing plants. Four replications are required to compare 5-7 treatments and to identify as a 'cultivar profile" the treatment which yields 25% more than the control's bunch weight. In OFT maximizing the number of trials is generally more important than replications within sites. Therefore, replications could be decreased within sites and replaced by additional sites. The most common experimental field layouts are completely randomized (CRD) and RCBD. Blocking should be used in sites where a sensible basis for this exists (e.g., fertilizer gradient). CRD are recommended for OFT especially in backyards. PBIP advises to insert a susceptible local cultivar (e.g. 'Valery' or 'Agbagba' for black sigatoka disease) at certain intervals (I.e., every two testing lines) and to plant it at the edge of a yield trial (METs, AMYTs or OFTs) in order to provide a uniform black sigatoka infection pressure. 2k. Selection in early evaluation trials (EET) Newly bred genotypes are initially evaluated in observational plots (1-5 plants in an unreplicated trial) for black sigatoka resistance (BSR), large bunch and fruit size, parthenocarpic fruit development, improved ratooning and dwarfism. The early evaluation of breeding products takes place at the Onne Station and involves observations on several hundred new genotypes each year. Until 1991 most of the early evaluation plots consisted of diploid banana and plantain derived hybrids. In 1992 secondary triploid hybrids, derived from crosses between plantain primary tetraploids and available diploids, as well as offspring of diploid plantain-derived hybrids were evaluated for the first time in the field. The number of clones and seedlings, and their expected ploidy in each breeding scheme are listed in Table 16. Results concerning banana breeding 20 at the diploid level are reported in Section 2s and information regarding selection of cooking and East African bananas are discussed in Section 4b. Triploid x diploid crosses Plantain hybrids planted in the EET of 1992 (92-EET) were derived from crosses of 'Akpakpak', BT, 'Bungaoisan', 'Mbi Egome' and OL with diploid bananas. The AA parents 'Long Tavoy', 'Tjau Lagada', 'Truncata' and 'Zebrina' had a low breeding value, as measured by progeny testing, in the 92-EET. Most of their euploid hybrids had poor bunch characteristics. 'Mbi Egome', a small . sized plantain as measured by its plant height at flowering, produced small but not dwarf (short internodes) hybrids. TMPx 7152-2 (,Mbi Egome' x C4), TMPx 10901-2 ('Akpakpak' x C4) and TMS 548-9 (open pollinated progeny of TMPx 548-9) were the clones selected in the 92-EET due to their adequate bunch characteristics and black sigatoka resistance. They will be included in the PYT of 1993 for more accurate yield testing. Population improvement of plantain-derived diploids Several plantain-derived diploids, disregarding their agronomic attributes or black sigatoka resistance, were the parents in the 92-EET. They were crossed with high male fertile diploid bananas such as SF 247 and SH-3362. Preliminary analyses suggested that only selected parents should be included in the crossing blocks. This was revealed by the poor performance and vigour of several unselected diploids (565-16, 566-32, 1668-7, 2085-1, 2902-43, 4188-2) when crossed with the two diploid i:lananas. Even the breeding value of the selected TMb2x must be checked before using them as regular progenitors in the production of improved diploid plantain germ plasm because selected hybrids such as TMp2x 1518-4 and TMp2x 2625-20 produced offspring with poor vigour. Probably recurrent selection with progeny testing would be the most appropriate tool for breeding in the diploid plantain germplasm. Also the utilization of diploid bananas to breed in this germplasm should be re-examined, especially when fruit quality attributes are considered. We found hyperploids (pentaploids or heptaploids) after crossing diploid derived plantains with SF 247. This may be an indirect evidence of 4n pollen production due to double restitution in SF 247. Secondary trip/aids (tetrap/oid x diploid crosses) Secondary triploids, in the 92-EET, were produced by crossing primary tetraploids (TMPx 548-4, TMPx 548-9, TMPx 582-4, TMPx 1658-4, TMPx 2637- 49, TMPx 2776-20, TMPx 4698-1 and TMPx 6930) with other selected partners (TMp2x 1549-7 and SH-3362) or parthenocarpic bananas (e.g. 'Galeo'). The most promising progenies in the 92-EET were derived from TMPx 548-4. So far, no aneuploids or weak progeny has been obtained with this parent. A high percentage of dwarf secondary hybrids were observed in the crosses between TMPx 582-4 (a dwarf itself) with TMPx 1187-8 and TMPx .6930-1 x TMp2x 1549-7. Some seedlings showed the abnormal growth development associated with aneuploids. This could be the reflection of irregular meiosis in the tetraploid parent leading to the production of aneuploid gametes. Also some hyperploid hybrids were produced by the female parents TMPx 2637-49, TMPx 4698-1 and TMPx 6930-1, which might be considered as an indirect evidence of double restitution in the megasporogenesis of these hybrids. Non-parthenocarpic secondary triploids were obtained by crossing TMPx 4479-1 and SH-3362. This shows that both parents are heterozygous for at least one of the fruit parthenocarpy loci. Most of the secondary triploids derived by crossing the unselected black sigatoka susceptible hybrid 597-4 with SH-3362 were not promising due to their poor vigour. Secondary tetrap/oids (tetrap/oid x tetraploid crosses) Preliminary evaluation of secondary tetraploids showed that some progenies from the cross TMPx 548-4 x 597-4 were abnormal and weak (poor vigour). The comparison of secondary triploids and tetraploids, which were field- established in early evaluation plots, will take place in 1993 and preliminary yield trials will be carried out in 1994 to identify the best type of breeding methodology for further improvement of Musa germplasm. Combining ability of the parents will also be assessed to identify progenitors for the breeding program. Inbreeding Selfed progenies, of both TMp2x and TMPx, showed poor vigour mainly due to an increase of inbreeding, which might indicate that this technique could not be useful in the genetic improvement of Musa . Also we have observed that backcrosses to C4 are not convenient either at diploid or tetraploid level due to an increase number of wild alleles in the cultivated gene pool. 22 An experiment aiming to establish the importance of heterozygosity and inbreeding in bunch weight and other agronomic traits as well as the value of the backcross method was started in 1993. The experimental materials are currently being produced. The experiment will consist of 2-way, 3-way and 4-way tetraploid x diploid, triploid x tetraploid and triploid x diploid crosses between unrelated (8=0) and related parents (8)0). Also backcrosses to the triploid plantains and diploid bananas have been done. An early evaluation trial in 1994 and a preliminary yield trial in 1995 of the hybrid progenies along with their parents will be carried out to determine what could be the best crossing scheme to continue improving Musa germplasrn. 21. Preliminary yield trials (PYT) and selection of plantain hybrids The hybrids combining increased BSR with good bunch qualities are selected in EET for further evaluation in replicated trials at Onne (2 replicates of 4 plants each), which are called PYT by PBIP. In addition to the traits evaluated in EETs, earliness, resistance to other pests/diseases, fruit quality/palatability, and postharvest storage ability are also evaluated in the PVT. From 1989 to early 1991 preliminary yield trials (plant and ratoon crops) were conducted at Onne to evaluate the performance of 25 plantain hybrids (diploids, triploids and tetraploids) alor)g with one primary tetraploid of the cooking banana 'Bluggoe' (see details about this clone in Section 4b). Results of this evaluation were reported in the PBIP 1991 Annual Report. Based on their high yield, adequate bunch characteristics and less susceptibility or partial resistance to black sigatoka the following clones were selected: TMpx 548-4, TMpx 548-9, TMpx 582-4, TMpx 1112-1, TMpx 2796-5, TMp2x 597-2, TMp2x 1518-4, TMp2x 2348-6 and TMp2x 2348-7. The last two clones were derived from the cross 'Bungaoisan' x C4. We also found that 'Bungaoisan' was the best parent in the production of plantain-derived diploid hybrids with high bunch weight and adequate black sigatoka resistance. 199011992 PYT (plant and ratoon crops) At the end of 1990, a preliminary yield trial consisting of the cv. 'Obino I' Ewai' (OL, under fungicide and non-fungicide treatments), and its tetraploid hybrids TMpx 548-4, TMPx 548-9 and 597-4 was planted at Onne. The experimental design was a randomized complete block with 4 replications of 25 plants, but observations and data were taken in 9 intra-clonal competing plants. Results of the plant crop were discussed in PBIP 1991 Annual Report. The tetraploid hybrid TMPx 597-4, which initially was selected as black sigatoka 23 resistant, showed a partial resistance to black sigatoka until flowering, but became susceptible during the fruit filling period. The results of analyses of variance (ANOVAs), based on production cycle means, assuming the statistically strict random model over the two production cycle (plant and ratocn crops) are given in Table 17. This AN OVA requires F values (mean square ratios) of 53.4 and 161 to detect statistical significance (P = 0.05) for treatments and production cycle, respectively. This is due to the number of degrees of freedom for the F-test (1,1 and 1,4 for production cycle and treatment respectively). Therefore AN OVA significance for these two sources of variation should be regarded as conservative. In general the bunch weights of the ratoon cycle (9.3 kg) were inferior to those previously reported in the plant crop (13.2 kg) due to a shorter cycle in the ratoon (337 days) than in the plant crop (380 days). Also the suckers developed faster in the ratoon cycle (RC) than in the plant crop (PC) as measured by the height of the tallest sucker at harvest (208 cm and 154 cm for the RC and PC, respectively). Possibly source-sink relationships favoured the fast growth of the suckers and decreased the allocation of resources (nutrients) for fruit filling. The significance of treatment x production cycle interaction was assessed by Tukey's test of non-additivity. Only significant interactions were found for days to flowering and days to fruit filling. However, a change of magnitudes but not on ranking was the cause for this statistical interaction. For example, there was a sharp reduction in days to fruit filling for the ratoon cycle of the plantain, especially when non-chemical protection was provided, whereas the hybrids used about 120 days for fruit filling in both cycles. Due to this non-cross order interaction, it was statistically correct to make comparisons of means over the two production cycles (Table 18). Student-Neuman-Keul's multiple comparison tests at the 5% level were carried out for statistical mean separation. This conservative test protects against Type I error but increase the chance of Type 1/ error (false acceptance of null hypothesis, i.e. no differences between mean treatments). Three one degree of freedom linear contrasts were carried out of: (i) The performance of the plantain cultivar, OL under fungicide and non- fungicide treatments, this provided information about gains for each traits when chemical protection was used (ii) The average performance of the black sigatoka reSistant TMPx 548-4 and TMPx 548-9 vs. that of 597-4, which measures the gain by incorporating resistance genes in almost similar genetic background and at the same ploidy level 24 (iii) The fungicide-treated plantain, Ol vs. TMPx 548-4 and TMPx 548-9 together The last contrast should be regarded as the closest approximation to the comparison of host plant resistance with chemical strategies to control black sigatoka in plantain. Bunch weigths of Ol increased by 57% under chemical treatment. The major factor for the increase was the gain of fruit weight (43%) due to ten more days for fruit filling. Both fruit length and circumference contributed equally (17% and 16%, respectively) to this gain in fruit size. The tetraplOid hybrids only differed in bunch weight and its components. The susceptible 597-4 had inferior bunches relative to the black sigatoka resistant TMPx 548-4 and TMPx 548-9 arising mainly from differences in fruit weights, of which the yield components (although non-significant) were increased fruit length and circumference in the black sigatoka resistant hybrids. The black sigatoka resistant hybrids relative to their fungicide treated plantain parent had greater bunch weights, improved suckering behaviour, shorter production cycle, due to earlier flowering, and longer fruit filling period (Table 18). The increase in bunch weight was due to both more and larger fruits in the hybrids than in Ol-FT, rather than increases only in individual fruit sizes. Differences in the fruit length and circumference between the BSR hybrids and the plantain parent did not attain statistical significance. Nevertheless, the fruit circumferences of both BSR hybrids were greater by 14% and 7%, for TMPx 548-9 and TMPx 548-4 repectively, than that of the fungicide treated plantain. Similarly the fruit length of TMPx 548-9 had a slight increase (5%) as compared with Ol-FT. These attributes analysis explain the higher bunch weight of the hybrids when compared with Ol-FT. A more detailed comparison between host plant resistance and chemical control strategies is discussed in Section 3c. 199111992 PYT (plant crop) A preliminary yield trial comprising the selected hybrids from Ol x C4 (TMPx 548-4, TMPx 548-9, TMp2x 597-2) and BT x C4 (TMPx 582-4, TMp2x 1518-4) as well as other tetraplOid hybrids produced from both crosses (period 1989 and mid-1990) and between both BT and Ol with PI was started in mid- 1991. A maximum of five replicates was used for the evaluation of each clone. Results of growth and yield parameters in the plant crop (1991/1992 season) for BT x C4 and Ol x C4 progenies are summarized in Table 19. 25 Progeny testing of the plantain cultivars was carried out to assess their combining ability for bunch weight and black sigatoka resistance before using them as elite progenitors in PBIP breeding program. In this regard, plantain cv. 'Obino l'Ewai' had greater breeding value than cv. 'Bobby Tannap', not only because the former had a higher black sigatoka resistance inheritance rate (see PBIP 1991 Annual Report), but also because its progenies had significantly better average bunch weight (Table 19). Both traits, black sigatoka resistance and yield, are currently the most important parameters used in the early evaluation of breeding materials, again emphasizing that 'Obino l'Ewai' seems -to be a more valuable maternal source for plantain improvement as compared with 'Bobby Tannap'. Data analyses from the ratoon crop will be included in 1993 PBIP Annual Report. Genetic analyses and black sigatoka (BS) reactions are discussed in detail in Section 2w. and Section 3a. Selected black sigatoka resistant tetraploid hybrids Based on these PYT results, 12 hybrids of OL x C4, 3 of BT x C4, and one each from BT x PI and OL x PI were selected (Table 20) due to their equal or higher bunch weight than their fungicide-treated plantain parent as well as their decreased susceptibility or partial resistance to BS. Taking into consideration that the tetraploids constituted the reference breeding population, the rates of selection were very high: 75% for OL x C4 and 50% for BT x C4. If all the euploids are considered for this calculation, the selection rates decrease for OL x C4 and BT x C4 respectively to 32% and 5%. Plantain-derived diploid hybrids and germplasm enhancement at diploid level Plantain-derived diploids recovered from the triploid x diploid crosses prove to be very advantageous to plantain breeding, because they provide opportunity for germplasm enhancement at the diploid level and simplify genetic analysis due to disomic inheritance. This facilitates and speeds up the breeding work. They have been used in the extension of the initial breeding approach to include the production of secondary triploids by further crossing the primary tetraploid hybrids with these diploids. Twenty four selected diploids (Table 21), based on their performance in EETs and PYTs, are being evaluated for fruit quality traits prior their utilization as regular progenitors in the breeding program. Selection rates, over the total number of diploids, were 12%, 23%, 27% and 29% in 'French Reversion' x C4, BT x C4, OL x C4 and 'Bungaoisan' x C4, respectively. 26 2m. Multlslte evaluation trials (METs) of selected Muss hybrids The TMP(B)x which incorporate most of the desirable traits in PYT are included in Multilocational Evaluation Trials (MET). Each MET is carried out in a minimum of 3 different locations using 2 replications of 5 plants each of 8-15 genotypes. The main objective of MET is to test the performance of the TMP(B)x under a wider range of agro-ecological conditions in sub-Saharan Africa, thereby assessing the genotype by environment interaction for specific traits and the stability of both yield and BSR. The involvement of key NARS in this relatively early stage of hybrid evaluation is important for feedback to improve selection procedures. Conducting METs also allows IITA's breeders to decide what material might be further evaluated by NARS in the target area. MET is an integral activity of the breeding program of PBIP in which the collaboration with NARS seeks to assist the strengthening of their research capacity. An additionally outcome should be the release and distribution of materials through the collaborating NARS. Safe movement of germplasm Aseptic shoot-tip culture is used as a vehicle for the safe exchange of banana/plantain germplasm. Shoot-tip cultures confer considerable advantages for the international transfer of germ plasm because: (i) The mass of plant material involved in the movement is greatly reduced (ii) The plant material is contained (iii) They overcome nearly all of the problems associated with pests and pathogens, other than in some expectional circumstances (iv) They are amenable to rapid multiplication The Nigerian Plant Quarantine Service visits the PBIP in vitro Tissue Culture Laboratory once every year to check the health status of the stocks. After this inspection, phytosanitary certificates are officially printed to export planting materials for the METs and AMYTs planted outside Nigeria. Each non-Nigerian cooperator of the MET and AMYT networks also sends import permits, in advance, which are issued by the respective national plant quarantine authority. This process, therefore, follows all the international, regional and local quarantine regulations for the implementation of yield trials out of Onne Station. Assessment of blaok sigatoka resistance in multilocational trials PBIP has selected improved germplasm with non-immunity or less susceptibility to black sigatoka (see Section 3a). Our resistance criteria are 27 defined on a scale relative to known susceptibility standards. In Section 2w a detailed description is presented on how this scale facilitated elucidation of the inheritance of black sigatoka resistance. Due to different surveyors in each location, the youngest leaf spotted (YLS) at flowering, which correlates significantly with disease development time, is defined as the common criterion to measure the black sigatoka response of each clone across environments. Scoring for this trait is easy. It consists of recording the number of the youngest leaf, counting down from the first (top) unfurled leaf, that showed mature spots, i.e. YLS. The inclusion of resistant and susceptible standards in MET-l makes it possible to compare results collected in different locations and to assess potential pathotype differentiation. MET-1 The objective of this set of replicated trials was to test the performance of the improved TMPx, selected in 1990-pYT, under different agro-ecological conditions in West and Central Africa relative to their triploid and diploid parents and black sigatoka susceptible checks. MET had a total of eight TMP(B)x entries and was planted in 1991 at four sites: three IITA stations in Ibadan (transition zone), Onne (High Rainfall Station in the humid forest), M'Balmayo (Humid Forest Stalion) and one in a national program in Ibadan, Nigeria (NIHORT). The latter experiment was lost due to drought. Data from IITA-Ibadan are being collected and will be reported in the 1993 PBIP Annual Report. Mixed analyses of variance combined over locations were carried out for the METs after testing the homogeneity of variances. The locations were considered as random samples of the humid forest zone with the clones as fixed factors. Therefore, conclusions are relevant to this agro-ecology but restricted to the entries tested. Preliminary analyses of plant CfOp results collected at Onne and M'Balmayo (Table 22), indicate that there were significant differences between locations for black sigatoka resistance, number of leaves, bunch weight, and fruit length, circumference, and weight. In general the clones had more leaves, heavier bunches with larger fruits at Onne than in M'Balmayo. Also the clones showed less susceptibility to black sigatoka disease at Onne than at M'Balmayo. The higher bunch and fruit weights at Onne might reflect more photosynthetic activity at this location than at M'Balmayo, due to more leaf area. Furthermore, the genotype by location interaction was significant for plant height, days to flowering, black sigatoka resistance, fruit filling period, bunch weight, number of fingers, fruit circumference and weight. This indicates that the clones had different maturation times at each location. The results emphasise the value of multilocation testing before cultivar release, even in the same agro- ecological region. To make a more accurate assessment of black sigatoka responses all the clones were compared with known susceptible cultivars ('Agbagba' and 'Valery') and the hybrids against their triploid parents ('Bluggoe', BT and Ol). TMP(B)x with more or significantly more leaves without necrotic spots than the susceptible parent were considered as having less susceptibility or partial resistance to black sigatoka disease, respectively. Hybrid and cultivar means of black sigatoka resistance, growth and yield parameters evaluated at each location are presented independently in Table 23 (Onne) and Table 24 (M'Balmayo) because the genotype-by-environment interaction affected most of the traits evaluated. Skewness and kurtosis were the statistics employed for the analyses of frequency distribution for all the traits in each location. We found that all traits follow almost a quasi-normal distribution at M'Balmayo and that only number of fingers, fruit length and circumference did not follow a normal distribution at Onne. This was indicated by the significant value of skewness and kurtosis for the last three traits at Onne. Transformation were not done because these non-normal frequency distributions could be related to the excessive number of triploid or tetraploid clones with high value for such traits as compared with the diploids, which are characterized for having low values for fruit characteristics. In general, the results agree with those published in PBIP 1991 Annual Report. TMP(B)x had less severe black sigatoka reactions and equal or greater bunch weight relative to their plantain parents. Also the tetraploids had shorter plant stature and better suckering behaviour, which are desirable traits combined with earlier flowering and longer fruit filling periods when compared with their plantain parents. Only 'Agbagba', the False Horn plantain, and the ABB cooking bananas had fruit sizes (weighVlength and circumference, respectively) or greater than the tetraploid hybrids. This was expected because large and heavy fruits, and big circumferences are distinctive characteristics of the False Horn and ABB cooking banana gene pools, respectively. The results from the Humid Forest Stations of IITA indicated that TMPx 1658-4 (hereafter PITA-7) and TMPx 548-4 (PITA-2) were high and stable yielding clones in this agro-ecology. These clones out yielded not only their susceptible plantain parent OL but also other ABB cooking bananas with less susceptibility to black sigatoka. Also both clones and TMPx 2796-5 (PITA-5) were stable and consistent in their black sigatoka response across locations. 29 MET-2 The MET-2. planted in 1992. tests 12 promising hybrids (Table 25) in IITA's stations at Onne and M'Balmayo, at Namulonge Station, Uganda and involves the participation of NARS in Australia, Cameroon, Cuba, Dominican Republic, Ghana and Nigeria (Table 26). The field layout used at Onne Station, which is similar to that of the other sites but with different randomization, is illustrated in Fig. 18. Preliminary reports of our cooperators are encouraging. For example, the host response to black sigatoka in the MET-1 at Onne (Table 23) was similar to . the one in the MET-2 at Nyombe, Cameroon. In the CRBP 1992 Annual Report, our collaborators noted from preliminary observations that all TMPx showed high partial resistance, while TMBx 612·74 was rated as highly resistant (no necrosis). The highest levels of partial resistance were shown by PITA-5 (YLS 12.0 and 9.1. at Nyombe and Onne, respectively). YLS results for other entries at Nyombe and Onne respectively were: PITA 2 9.5, 8.8; TMPx 5484 10.2, 7.5 PITA 7 9.8 7.5. The susceptible reference banana cv. 'Valery' had a YLS of 6.2. and 4.2 at Nyombe and Onne, respectively. Complete results of the MET·2 will be analyzed and published in PBIP 1993 Annual Report. 2n. Registration of Improved plantain germplasm (TMPx) Improved genetic material of Musa, developed by IITA, is released to national programs without restriction whenever it proves useful. The improved germ plasm developed by IITA breeders has been submitted for registration in order to place the material in the public domain. This official registration provides an accurate description of the material and establishes a bench·mark of advancement. The tetraploid hybrids, listed in Table 27, were included in this first registration of IITA improved Musa germplasm with black sigatoka resistance in Hortscience (Ref. 2, paper currently in press). The paper contains information on the origin, description, availability and comparative data of the TMPx germplasm. A limited amount of in vitro plantlets are available upon request. Recipients are asked to give appropriate recognition of the source of germ plasm if it is used in the development of a new germplasm, parental line or cultivar. 30 20. International Muss Testing Program (IMTP-1) The IMTP, set up by iNiBAP, aims at the identification of germplasm (natural or improved) with resistance/tolerance to a specific disease or pest through testing on a global scale. The first IMTP concentrates on black sigatoka resistance and involves six selection sites in two continents (Africa and tropical America). IITA participates as an equal partner in the IMTP that is carried out in Africa. The IMTP-1 trial focused on the evaluation of seven hybrids of banana and plantain produced by FHIA (Fundacion Hondurena de Investigacion Agricola, La Lima, Honduras), along with 11 black sigatoka standard hosts and reference cultivars in observation (unreplicated) plots. This research is part of IITA's collaboration with iNIBAP and FHIA. Preliminary results (Table 28) ~howed that FHIA-1 (a 'Prata' banana hybrid), FHIA-2 ( a 'Cavendish' banana hybrid) and FHIA-3 (a 'Cardaba' derivative) had partial resistance to black sigatoka leaf spot disease while the remaining plantain hybrids (derived from either Maqueno or AVP-67) were susceptible or less susceptible to BS. FHIA-4 (a plantain hybrid derived from AVP-67 French plantain) showed resistance to black sigatoka until flowering, but became susceptible towards the harvest period. Comparisons among cultivars should be made only with caution because they were planted randomly in unreplicated plots, following INIBAP's guidelines for this trial. In general tall plants had a wider girth at flowering than shorther plants (r = 0.74, P < 0.01), as well as more leaves ( r = 0.75; P < 0.01). However there were not significant correlations between plant height and number of leaves with bunch weight (r = 0.47 and 0.46, respectively). Therefore, the development of high yielding cultivars with short stature should be possible. This information is being considered for the development of an ideotype in Muss breeding (see Section 2s). 'Pahang', PI and C4 showed lack of apical dominance, which resulted in a fast development of the sucker (follower for the ratoon cycle) as measured by the ratoon growth index (RGi) or the ratio between the height of tallest sucker at harvest and pseudostem height in the plant crop. None of FHIA hybrids had improved suckering behaviour, as measured by RGi. This index was smaller or equal to those of 'Valery' or 'Agbagba' for dessert bananas and plantains, respectively.'Tuu Gia' and 'Pisang Berlin' were other diploid bananas with fast sucker growth whereas SF 247 and 'Niyarma yik', early maturity clones, showed low RGi probably due to high apical dominance. This would question previous ideas indicating fixation of genes for non-apical dominance in the AA gene pool. 31 20. International Muss Testing Program (IMTP-1) The IMTP, set up by INIBAP, aims at the identification of germ plasm (natural or improved) with resistance/tolerance to a specific disease or pest through testing on a global scale. The first IMTP concentrates on black sigatoka resistance and involves six selection sites in two continents (Africa and tropical America). liT A participates as an equal partner in the IMTP that is carried out in Africa. The IMTP-1 trial focused on the evaluation of seven hybrids of banana and plantain produced by FHIA (Fundacion Hondurena de Investigacion Agricola, La Lima, Honduras), along with 11 black sigatoka standard hosts and reference cultivars in observation (unreplicated) plots. This research is part of liT A's collaboration with INIBAP and FHIA. Preliminary results (Table 28) showed that FHIA-1 (a 'Prata' banana hybrid), FHIA-2 ( a 'Cavendish' banana hybrid) and FHIA-3 (a 'Cardaba' derivative) had partial resistance to black sigatoka leaf spot disease while the remaining plantain hybrids (derived from either Maqueno or AVP-67) were susceptible or less susceptible to BS. FHIA-4 (a plantain hybrid derived from AVP-67 French plantain) showed resistance to black sigatoka until flowering, but became susceptible towards the harvest period. Comparisons among cultivars should be made only with caution because they were planted randomly in unreplicated plots. following INIBAP's guidelines for this trial. In general tall plants had a wider girth at flowering than shorther plants (r = 0.74, P < 0.01). as well as more leaves ( r = 0.75; P < 0.01). However there were not significant correlations between plant height and number of leaves with bunch weight (r = 0.47 and 0.46, respectively). Therefore, the development of high yielding cultivars with short stature should be possible. This information is being considered for the development of an ideotype in Muss breeding (see Section 2s). 'Pahang', PI and C4 showed lack of apical dominance, which resulted in a fast development of the sucker (follower for the ratoon cycle) as measured by the ratoon growth index (RGi) or the ratio between the height of tallest sucker at harvest and pseudostem height in the plant crop. None of FHIA hybrids had improved suckering behaviour, as measured by RGi. This index was smaller or equal to those of 'Valery' or 'Agbagba' for dessert bananas and plantains, respectively.'Tuu Gia' and 'Pisang Berlin' were other diplOid bananas with fast sucker growth whereas SF 247 and 'Niyarma yik', early maturity clones, showed low RGi probably due to high apical dominance. This would question previous ideas indicating fixation of genes for non-apical dominance in the AA gene pool. 31 The cooking banana hybrid FHIA-3 was selected for thorough testing in sub-Saharan Africa. FHIA granted authorization to include this hybrid in our AMYT and for further genetic manipulations in PBIP's breeding programme. PBIP has put forward the black sigatoka-resistant plantain hybrids PITA-2, PITA-3, PITA-4, PITA-5 and PITA-6 (see Table 20) as entries in the IMTP-2 along with breeding materials from FHIA, EMBRAPA (Brazil), CIRAD/IRFA (Guadaloupe) and INIVIT (Cuba). 2p. Advanced Muss Yield Trials (AMYT) There is a genuine need within a breeding program to evaluate thoroughly the experimental materials through a sequence of trials, which will result in the selection of promising genotypes as potential new cultivars for distribution by NARS. Because of the existence of several programs which produce candidate breeding lines for release to African NARS, there is a need for agreement on, and coordination of, the efforts aimed at evaluating these products. Selected materials from multilocational testing (MET), along with local cultivars used as checks, are included in an Advanced Muss Yield Trial (AMYT) to be undertaken simultaneously in several local evaluation sites of various countries (Table 29) and in which the NARS would be the executing agencies. The results of such trials would thus be of regional relevance. The objective of the AMYT-1 is to identify elite genotypes for potential release as new cultivars by NARS according to the specific regulations in each country. AMYT-1 evaluates eight selected promising hybrids (Table 30) over a period of at least two years in a RCB design with 4 replications of 5 plants each. Field layouts (Fig. 19) are similar to those of the METs with independent randomizations for each location to avoid correlation of errors. Criteria for selection should not only include disease resistance and productivity, but also local preferences as they affect consumer acceptability (see Section 6). 2q. AMMI analysis of yield trials Plant breeders are interested in the selection of genotypes which perform well over a wide range of environments. To identify such genotypes, yield trials are carried out over years in the breeding station and at different locations. In Muss the genotype-by-year (GxY) interaction seems less important than the genotype-by-Iocation (GxL) interaction, even in the same agro-ecological zone (see Sections 2m and 2w). 32 In the presence of a significant interaction, both the stratification of environments according to their agro-climatological similarities and the determination of stability parameters for genotypes across environments are important tools for the management of the genotype-by-environment (GxE) interaction. Several techniques have been developed to determine what was the most stable genotype in a set of replicated trials across years, locations or combinations of both (namely environments). However, these postdictive models are not useful in the identification of what genotypes and environments contribute to the GxE interaction. Moreover, a breeder may be interested to identify what genotypes are adapted to specific environments and/or to predict their performance in a specific location. The additive main eHects and multiplicative interaction (AMMI) model was developed to provide answers for such questions. AMMI uses the analysiS of variance (ANOVA) to study the main effects of genotypes and environments and principal component analysis (PCA) for the residual multiplicative interaction. Five tetraploid hybrids, which cllrrently are undergoing evaluation in the multilocation trials, were chosen to determine the validity of AMMI analysis of factorial (genotype-by-environment) designs for Musa yield trials. Data from early evaluation trials (EET at Onne 1989/1990), preliminary yield trials at Onne in the plant crop (PYT-PC 1990/1991) and ratoon (PYT-R 1991/1992), and multilocational trials (MET) at Onne, M'8almayo and lbadan (1992/1993) were used for this analysis. Fig. 20, the AMM11 model for bunch weight of five tetraploid hybrids evaluated in six trials, shows the genotype (G) and trial (T) main effects on the abscissa and the interaction PCA 1 score (IPCA 1) for G and T in the ordinate. G and T are indicated with open and filled symbols, respectively. Imaginary lines parallel to the axes must be drawn to ihdicate the grand mean (perpendicular to the abscissa point 12.3 kg) and zero in the ordinate axis, respectively. Among pOints of the same kind (either between G's or between T's) displacements along the bunch weight axis indicate differences between hybrids or trials (main effects) while displacements along the IPCAl axis indicate different interaction effects. The most stable, but average yielding, genotype was the black sigatoka (8S) resistant TMPx 582-4. PYTs and METs at Onne and M'Salmayo differed only in interaction effects. For points of different kinds, the AMMI model equation provides the expected bunch weights. For example, the AMMI expected yield for TMPx 548-9 in EET-Onne, [TMPx 548-9 mean + EET-Onne mean - Grand mean + (IPCA1 33 EET)* (IPCA1 TMPx 548-9)]. should be equal 10 17.B, which was not significanlly different from the observed bunch weight (17.9) of TMPx 548-9 in the EET-Onne. This means that the AMMI1 model left a residual of 1 ~Og. This biplot (Fig. 20) gives also additional interesting information. The main effects for genotypes may indicate their reaction (either susceptible or resistance) to BS, with the smallest genotype mean for the BS susceptible tetraploid clone 597-4. Similarly the main effects for environment represented overall site quality, with EET-Onne having the best weather and soil conditions in 1989-1990, while the poor bunch weights of the MET-·lbadan could be the . result of the lack of adaptation of the tetraploid hybrids to the dry season of this Nigerian location in the savanna-humid forest transition zone. The AMMI2 model (interaction analysis) for bunch weight is illustrated in Fig. 21. Points near the origin (0,0), e.f:). TMPx 582-4 or MET-Onne, have little interaction and should be well fitled by the additive submodel. Points near each other have similar interaction patterns (e.g. the full-sibs TMPx 548-4 and TMPx 548-9) while points distant to each other are indeed different (e.g., the unrelated TMPx 2796-5 and 597-4). The IPCA 2 also helps in the identification of those trials or genotypes with the worse fit, i.e. farther from zero (origin). For example, TMPx 2976-5 in EET-Onne was least fit by the AMMI1 model, i.e. its expected bunch weight was 21.3 kg instead of the 25.9 observed bunch weight in this trial. This also suggests that bunch weights from unreplicated EETs should be taken with caution. In conclusion, AMMI seems to be a powerful tool to analyze yield trials with a factorial structure, e.g. Locations x Genotypes. AMMI can analyze trials with within-site replications (1ITA's METs) or trials with only 1 replication per location (INIBAP's IMTP). 2r. Taxonomic approaches for banana breeding at the diploid level Plantain and banana breeding has long been considered essentially as diploid breeding due to the triploid nature of most of the cultivated types. Despite the success in the direct utilization of a wild non-related diploid banana such as 'Calcutta 4' in the production of black sigatoka resistant plantain hybrids, diploid breeding is still considered as an important component for the genetic improvement of Musa spp. In this regard, an important step in the breeding program is the production of improved diploid material to be used as male par~nts in both triploid x diploid crosses (to produce primary tetraploids) and tetraploid x diploid crosses (to produce secondary triploids). The selection of 34 diploid material is based on both female and male fertility, black sigatoka resistance and good bunch characteristics. The genome composition of triploid banana and plantain cultivars was considered as a crucial point to develop breeding strategies by CIRAD/IRFA and IITA (1987-1990). This information was used by breeders in these institutes, as a means of choosing diploid materials to transfer desired characteristics to cultivars. This taxonomic approach doss not seem to be genetically correct because it ignores the occurrence of segregation and recombination during meiosis. In other words, genes and not genotypes are carried to the next generation by the gametes of the diploid parents. Important genotypic combinations are broken due to the independent assortment of genes or the occurrence of crossing overs between loci. Nevertheless, selection gains in a breeding program rely on change of gene frequency, which results in the production of improved genotypes. Inreresting diploid materials, which have been obtained at IITA using this approach, will be incorporated in the regular breeding program of PSIP to test their breeding value. The intensity of selection at the diploid level was 0.7 %, the lowest across PBIP breeding schemes in 1992. Eleven improved diploid banana clones (Table 31) combining BSR, a compact and pendant high-yielding bunch with long parthenocarpic fingers, were identified. For example TMb2x 9128-3 (a progeny from the cross 'Tjau Lagada' x PI) had a bunch of 12.2 kg with 19 hands and 316 fingers. It had higher bunch weight but smaller fingers than SH- 3362 (improved diploid clone from FHIA) at Onne. An exceptional selection, due to its origin, was TMb2x 6142-1, which was derived from the cross between the AM East African cooking banana 'Nyamwihogora" and 'Calcutta 4'. This clone may be useful for breeding in the East'African gene pool (see Section 4). The improved diploids are being multiplied for their use on a large scale in the crossing blocks. In this way their breeding values will be assessed. We determined the breeding value of some diploid accessions, by their progeny testing, as well as their genotype for specific loci. 'Djum Metek' 'Galeo', Higa 111-4, M. a. subssp. malaccensis, M. 8. subssp. zebrina, M. balbisiana accessions, 'Po madu" 'Po rotan', 'Pitu', 'Putewey' and Truncata' have limited value as indicated by their nil selected offspring. We also found that 'Padri' , SH- 3362 and unknown 'Dibit' are heterozygous for the fruit parthenocarpy P, locus (see Section 2w). Segregation for this trait was observed in test-crosses with C4 (homozygous recessive for this locus). A diploid derived by crossing 'Valery', as female, and C4 , as male, was discarded because it had non-parthenocarpic fruits. This indicated that 'Valery' is heterozygous for the P, locus. 35 2s. An Ideotype for Muss improvement and an index for multitrait selection Results in Table 4 indicated that pseudostem height and bunch weight are correlated with the leaf area index. This suggests that tall plantains with big leaves have high yielding bunches because they have more leaf area for photosynthesis. However, giant plantains flowered late and therefore required a longer production cycle (planting to harvest). Moreover, adjusted yields per unit time did not differ among medium and giant plantains in the False Horn gene pool and actual real yields, taking into consideration the incidence of non- productive plants, were similar between giant and medium French Plantains (see PBIP 1991 Annual Report). This demonstrates the importance of correlated characters to define an ideotype for plantain breeding. Similarly the contributions of diploid bananas for ideotype breeding in Musa were analyzed based on the IMTP results (Section 20) using different biometrical approaches. Correlation analyses have been discussed in Section 20. Due to the significant correlation b~tween plant height and girth at flowering we recommend only plant height as a breeding record. However, both traits seem to be useful for taxonomic purposes as illustrated in Fig_ 22. The IMTP entries had an almost perfect clustering according to their taxonomic designations based on these two measurements. All the diploid bananas were characterized for their shorter stature and thinner pseudostem when compared with the triploid or tetraploid clones. Principal component analysis (PCA) gave more information about what diploid parents have characteristics which may be incorporated into our ideotype. PRIN1, which explained 77.2% of the total variation, was unevenly and positively loaded on plant height and days to flowering and negatively loaded on height of tallest sucker. Similarly PRIN2, was positively and unevenly loaded on days to flowering, fruit filling period and height of tallest sucker and explained another 14.3% of the total variation. Thus, the PCA diagram (FIg_ 23) was divided in quadrants: in the upper right side, the two plantain hybrids (FHIA- 4 and FHIA-5) and 'Agbagba' were clustered together and closer to FHIA-3 and 'Valery'. With the exception of FHIA-3, they were characterized for being tall (except 'Valery'). late flowering, black sigatoka susceptible as well as having high apical dominance. which slows sucker development. Close to this group was the AAB derived banana hybrid. FHIA-1. In the centre of the graph and close to 'Valery' were 'Po Berlin' and FHIA-2. a hybrid of 'Williams'(Cavendish group). They had short stature and faster sucker growth than the other clones mentioned above. In the lower right quadrant. the black sigatoka susceptible diploid 36 bananas SF 215 and 'Niyarma yik' were grouped due to their early flowering, short fruit filling time and potential apical dominance. In the upper left side we found the black sigatoka resistant AA bananas 'Tuu Gia', 'Pahang" and C4. The other black sigatoka resistant banana PI was closer to this group. While 'Tuu Gia' and 'Pahang' were tall clones both C4 and PI had short stature and early flowering. The black sigatoka resistan~ bananas had also very fast sucker growth rates and long fruit filling period. This information combined with the comparison of TMPx germ plasm with their parents (Sections 21, 2m, 2n) allowed us to determine that the short stature, . early flowering, fast sucker growth and long fruit filling time were the most important traits contributed by the AA banana gene pool for the improvement of the AAB plantain gene pool. They have the genes (Section 2w) for dwarfism, earliness, non-apical dominance and increased time for fruit filling which made possible the production of TMPx germplasm with these attributes. Hence, these are the four main characteristiCs, besides disease/pest resistance, that Musa breeders should look for in the AA bananas before utilizing them in the crossing block. Multitrait selection We know that the breeding product, an improved cultivar with sustainable production, is a combination of several traits, e.g. a high and stable yielding plantain with black sigatoka resistance and fruit quality. This clearly reflects that the economic value of the final product depends on more than one trait. Therefore, simultaneous selection for several traits has been put in practice at I ITA. Several methods were implemented in 1992 by PBIP for multitrait selection: (i) Tandem selection in which one trait was used as selection criteria at any one time, but the chosen trait varied from cycle to cycle of selection. For example, in the roguing nursery all the aneuploids and hyperploids were eliminated due to their poor vigour and stunted growth, later (at flowering) selection took place only in the euploid population; and (ii) Independent culling, or when a separate threshold is set for each trait, and the selected individuals are those which are simultaneously above the threshold for each trait. This happened in the PYTs where the selected TMPx had both less susceptibility to black sigatoka and higher bunch weight than their respective parents. 37 Despite the successful utilization of these multitrait selection methods, we consider that simultaneous selection based on an index combining more than two traits may speed the accumulation of desirable traits in the breeding population. We are developing such an index taking into consideration the variance/covariance structure, genetic and phenotypic correlations and economic weights that the breeder gives to the different traits according to their economic importance. Data from the plant and ratoon crops (1991-1993 PYTs at Onne) are being analyzed to calculate selection indices for each diploid and tetraploid breeding population. Parents for further utilization in crossing blocks . wi II be selected in 1993/1994 based on thei r index. 21. Breeding the False Horn gene pool using somaclonal variants Occurrence of French reversion variants among in vitro propagated populations of False Horn plantains provides opportunity to breed this highly sterile gene pool. Indeed, the French variant is more fertile that the original 'Agbagba' cultivar,the number of seeds set being 2.5 and 0.03 per bunch, respectively (see PBIP 1991 Annual Report). The off-type did not only show a different inflorescence morphology, but also shorter stature and fruit filling period, lower bunch and fruit weight, fewer hands, but more fingers than the true~type (Table 32). The French reversion variant was also significantly different from the naturally occurring French cultivars for all growth and yield parameters. Moreover, preliminary results for fruit quality analyses (see Section 6), showed that TMPx 1112-1, a triploid black sigatoka resistant hybrid derived from crossing the French reversion variant of 'Agbagba' with C4, had closer fruit characteristics to 'Agbagba' than the other French plantain hybrids. This would indicate that somaclonal variation partially affected the characteristics of the 'Agbagba' cultivar. Therefore, the utilization of the French reversion variant as parent in triploid x diploid crosses could be considered as a way for the genetic improvement of the highly sterile False Horn plantains. Selection for larger and fewer parthenocarpic fruits, i.e., the False Horn ideotype, was done in the segregating diploid populations comprising euploid hybrids (diploids, triploids and tetraploids). We selectedTMPx 1297-3 among the diploid derived clones (selection rate of 12.5%) of French Reversion x C4. This clone has a pendant bunch with parthenocarpic fruits while its other full-sibs and their progenies had very poor,bunch appeBj'ance and!l1ost of them developed non-parthenocarpic fruits. Directional selection for reduced number of fruits may result in larger fruit size and concomitant bunch weight increase. 38 This scheme has been put in practice by crossing TMPx 1112-1 (PIT A-9) and TMp2x 1297-3. Their offspring as well as that derived by intermating their other full-sibs diploid plantain hybrids were evaluated in the 1992-EET (see Section 2k). 2u. Physiological research to enhance Muss breeding There are several physiological traits that can be used as indirect criteria for selection. Preliminary results indicate that high values for stomatal resistance to water vapour flux (sec cm-') in the afternoon could be a good indicator of potential drought tolerance. Small resistance values in the afternoon indicate that stomata remain open and leaves have increased water loss. Conductance to water vapour (the reciprocal of stomatal resistance) in the morning has been suggested as an indirect measurement of photosynthesis. High resistance values (stomata closing) in the morning could indicate less photosynthetic activity and therefore less productivity_ Leaf age effects on the leaf conductance to water vapour on both the upper and lower leaf surfaces were measured in the morning and in the afternoon on 17 different plantain and banana genotypes included in the MET-1 at Ibadan. The irradiance level increased 3-fold from morning to afternoon while leaf to air vapour pressure deficit levels increased 2 to 4 fold during the sampling period (Table 33). Conduct~nce values were higher for the lower leaf surface and in the afternoon than in thl'! morning, with some genotypic differences (Table 34). Low values of leaf conductance to water vapour in the afternoon under a short dry spell was sufficiently variable (P 3 in 1-5 scale, and clustered together in the upper right of the PCA graph, Fig. 37) indicates that only one M. acuminata susceptible accession could be included as reference cultivar for the e.ssessment of black sigatoka variability (both severity and incidence) in a multilocational trial such as the IMTP. Currently, PBIP uses 'Valery' and 'Agbagba' as their standard reference susceptible banana and plantain cultivars, respectively. in its multilocational trials. 3c. Chemical and host plant resistance strategies to control black sigatoka M. fijiensis was introduced to Africa about two decades ago and spread rapidly through all production zones. This fungus causes severe leaf necrosis and reduces yield by 30-50% (see PBIP 1991 Annual Report and Section 21). All plantain germplasm at Onne (115 cultivars). collected from West and Central Africa, tropical America and the Philippines. was found to be equally susceptible to black sigatoka (see Table 36 in TRIPP 1988 Annual Report). 71 Chemical control strategies exist, but are socio-economically inappropriate in the framework of the resource-poor smallholders which grow the crop in Africa. In addition, fungicide applications are hazardous to health in the village homesteads. It was pointed out recently that "fungicides are the most abundant pollutants in water downstream from banana plantations, where they can kill fish or prevent them from reprocjucing. Anything to reduce pesticide use in the tropics is strongly to be welcomed' (New Scientist, 17 April 1993 p. 9). Therefore, resistance breeding was considered by IITA as the most appropriate component intervention to control sigatoka leaf spot. A genetic improvement program was started in 1987, targeting the incorporation of durable host plant resistance to black sigatoka in plantain and banana. The results speak for themselves (see Section 2n) . Estimate of costlbenefit ratios for farmers using fungiCides or resistant hybrids We decided to quantify potential farmer gains by using two different black sigatoka control strategies: fungicides treatments and improved black sigatoka resistant hybrids. Results reported in the preliminary evaluation of hybrids along with their susceptible plantain parent under fungicide and non-fungicide protection were used to compare both control strategies (see PBIP 1991 Annual Report and Section 21 in this report). For a proper assessment of gains we estimated the potential yields (MT ha-' year-') of the TMPx black sigatoka resistant plantains and those obtained by their plantain parent 'Obino I' Ewai' under fungicide (Ol-FT) and non-fungiCide treatments (Ol-NT). Then, the potential yield of Ol-NT was used as the reference pOint (Fig. 38). The best hybrid, TMPx 548-9, had a 2.25 fold increase in yield than the plantain whereas fungicide control only increased plantain yields by about 70%. Moreover, farmers require an investment of at least US $ 750 ha-' year-' to provide chemical protection against black sigatoka disease while few suckers of the hybrids may be provided free-of Charge or at a minimum cost by NARS for further multiplication. Taking into consideration the cost!benefit ratio under two market periods at rural markets in the neighbouring area to Onne Station, we estimated roughly the advantages of host plant resistance over chemical protection to black sigatoka leaf spot disease. In normal periods 1 kg of plantain costs about N 5 (30 N = 1 US$), while in scarcity periods the cost rises to around N8 per 1 kg. Fig. 39 illustrates this comparison. At scarcity period farmer's gains by increasing production due to chemical protection might be up to US$ 1111. This may decrease to US$ 367 in normal periods when plantains are abundant in the 72 rural markets. With the TMPx black sigatoka resistant germplasm the gains may be about US$ 3633 and US$ 6055 in the normal and scarcity periods, respectively. Therefore, host plant resistance through the utilization of favourable genes could have a comparative advantage over antifungal chemicals of 10:1 or 5.5:1 in normal or scarcity periods of plantain production in southeastern Nigeria. In conclusion, genetic improvement of plantain and banana, by a modified but conventional cross-breeding approach (see Section 2), is an important component in the scientific path towards sustainable production of Muss cultivars in Africa. Its final impact will depend on how useful the new breeding materials are to national researchers and farmers, and their success in making these an integral part of plantain and banana production systems. 73 4. BANANA IMPROVEMENT FOR MID TO HIGH ALTITUDES IN EAST AFRICA The world's highest per capita consumption of bananas is in the East African highlands. Pest pressure is higher in this region as compared to West and Central Africa, because of the additional presence of Fusarium wilt disease and banana bunchy-top virus. Genetic improvement of the highland cooking and beer banana cultivars (AAA and ABB) that are adapted to this agro-ecology seems possible because seeds have been produced in a number of these cultivars. Methods developed by PBIP (see Section 2) for the genetic improvement of plantains can be extended to AAA and ABB bananas. Breeding, to have success, must be carried out in similar conditions of the targeted eco-region. Hence, /ITA will expand its Muss breeding work by posting a breeder at Namulonge Station in Uganda, as a staff member of the IITA East and Southern Africa (ESA) Regional Centre for the Improvement of Cassava, Banana and Plantain. 4a. Screening for female fertility in AAA and other starchy triploid bananas A preliminary screening for seed-fertility in this gene pool was done in the last five years using the cultivars planted in the IITA gene bank (Table 2) at Onne Station. Table 77 indicates the average seed production in AAA and other starchy triploid bananas. We were able to identify 5 sterile and 19 seed- fertile East African (EA) cultivars. 'Makara' an EA banana, once produced up to 67 seeds per bunch when pOllinated with C4. Also we found that the AB cv. 'Kamaramasenge' was seed-fertile and we have obtained several hundred hybrids seeds using this clone as female parent. The two most common cultivars in that region ('Igitsiri' and 'Igisahira gisanzwe') were multiplied in vitro and later field-established in a larger pollination block at Onne station in 1992. Hybrids seeds will be produced during 1993 by crossing these AAA bananas with diploid resistant accessions such a:> C4 (black sigatoka resistance), PI (black sigatoka resistance and heterozygous carrier of dominant gene for Fusarium wilt resistant), SH-3362 (with potential genes for Fusarium wilt and nematode resistance). Due to its male fertility the AAA dessert banana 'Yangambi Km.S' may be used in crosses with the AAA cooking/beer bananas because it has resistant/tolerant genes to protect against banana weevil. 74 4b. Mid-altitude cooking/beer banana breeding It is envisaged that hybrid materials, produced at Onne, will be sent in seed form to the !ITA ESA Regional Centre. True seed produced at the Onne station will be examined to determine which microorganisms are present and, if required, how health status can be improved, before seed is sent to Uganda. For example the AB cv. "Kamaramasenge' has good female fertility at Onne and several hundred hybrid seeds may be produced and sent to East Africa for hybrid production and testing. There, true seeds will be germinated for clonal production, evaluation and selection for improved yield, plant type and resistance to sigatoka diseases, weevil and nematode pests in this highland environment. Hybrids available for East Africa A black sigatoka resistant hybrid of the ABB cv. 'Bluggoe' TMBx 612-74 (hereafter BITA-l) has already been selected due to its high bunch weight (> 17 kg) and big fruits (>200g) at Onne and is being tested in the AMYT-East Africa (Burundi, Kenya, Uganda and Zanzibar) and the MET-2 in Uganda. TMBx 1378 ('Fougamou' x M. balbisiana), a hybrid developed at Onne Station, may be a potential hybrid for this area. Other potential hybrids are PITA-5 (included in the AMYT-l set) and PITA-7 (in MET set) because they have PI as male parent and the plantain germ plasm always has been considered as resistant/tolerant to Fusarium wilt. EETatOnne In the 1992-EET, we observed that hybrids derived from AM cvs. 'Inkira', 'Ingumba', 'Kibungu', 'Nyamwihogora', 'Yangambi Km. 5', were mostly diploids with limited value. A tetraploid hybrid derived by crossing 'Ingumba' with 'Tjau Lagada' had a poor yielding bunch with few hands. The only tetraploid from the cross 'Kibungu' x C4 was black sigatoka susceptible. Also, hybrids with non· parthenocarpic fruits were recovered in the cross 'Yangambi Km. 5' x C4, which indicated that 'Yangambi Km. 5' is also heterozygous for the P, locus. The diploid AB cv. 'Kamaramasenge' did not produce interesting hybrids. Its hybrids with 'Truncata' had poor appearance and others with 'Guyod' had droopy leaves which could indicate tetraploidy. In our experience breeding in AAB cooking banana, we found that most of the 'Pelipita' x C4 derived triploid hybrids had bunches which resembled that of their parental line, which might suggest apomictic seed development in the cv. 75 'Pelipita'. Hybrids from the cross 'Pelipita' x 'Maraw' showed poor vigour. Thus, 'Pelipita' cross breeding seems to be difficult. Also derived diploids of 'Bluggoe' x C4 and 'Bluggoe' x 'Tjau Lagada' were eliminated due to their poor bunches in the EETs at Onne Station. 76 5. BIOTECHNOLOGY FOR MUSA BREEDING Recent concern over rising pest and disease pressure on banana and plantain has spurred genetic improvement programmes. Hence attention has focused on the collection, movement and conservation of Musa germplasm. However, efforts to propagate, conserve and breed cultivated Muss are fraught with many obstacles (slow propagation, low reproductive fertility, lack of genetic variability) specific to the biology of this vegetatively propagated crop. Several plant tissue culture and molecular genetic techniques are increasingly being used as enabling and enhancing technologies for the handling and improvement of Musa germplasm. Sa. In vitro propagation of selected genotypes for yield trials Shoot-tip culture is a well-established, adequate and relatively simple in vitro method for the rapid propagation of selected Muss materials and the production of clean planting material. Multiplication rates are several orders of magnitude higher than those obtained with conventional methods, which is of great value for the multiplication of newly bred genotypes or to speed up the testing of selections by plant breeders. Micropropagation continued to be an important tool in the rapid deployment of new breeding schemes and trials by supplying large numbers of plants of female and male parents for the crossing blocks and of TMPxs for the evaluation trials. In the last 12 months, the TC Lab supported the breeding and agronomy work by providing 800, 2760 and 1600 plantlets for 1993 pollination blocks, on- and off-station trials, respectively. The TC Lab at Onne also produced about 9000 in vitro plantlets for further distribution to our NARS cooperators for the multilocational trials in the period March 1992 to March 1993 (see Section 8a). Micropropagated plants of black sigatoka resistant cooking bananas were distributed to Nigerian national programs (see Section 8b). An additional 2500 plantlets of 'Cardaba' were produced for distribution in 1993. Technical support on in vitro propagation of Musa was provided to tissue culture laboratories of Nigerian NARS. 5b. Somaclonal variation studies The increased genetic variation among plants regenerated from in vitro culture has been termed as somaclonal variation, a ubiquitous phenomenon. The frequent use of in vitro culture techniques for the handling of Musa 77 germplasm warrants investigations into the occurrence of somaclonal variation in this genus (for a thorough discussion see Ref. 6.) Table 78 summarizes the number of somaclonal variants per phenotypic trait in 13 different cultivars of plantain. It was clear that plantain cultivars varied widely in terms of their in vitro stability as determined by the occurrence of variation in the phenotype of regenerated plants. Also the results suggested that cultivars differed in the characters affected by phenotypic variation. To investigate the extent to which the genotypic component affects somaclonal variation, based on phenotypic evaluation, we carried out an experiment using four plantain cultivars, either False or True Hom, which were evaluated in two independent trials (Table 79). The results were not consistent over the two trials. For example the cv. 'Big Ebanga' had 35% of somaclonal variants in trial 1 and only 4.3% in trial 2. This might indicate that the environment plays an important role in the occurrence of somaclonal variation. We were able to determine that there were significant differences between cultivars in the rate of somaclonal variation in each independent trial. This was revealed by an analysis of frequency ~istribution using X2 tests (bottom of Table 79). This should be considered as evidence of genotype dependence in the occurrence of somaclonal variation. Also the genotype-by-environment interaction seems 10 affect the somaclonal variation rate. For example when cullivars 'Big Ebanga' and 'Bise Egome' (unstable clone) were compared, the rates of somaclonal variation changed in opposite directions in the two independent trials (cross order interaction). In conclusion it was clear that somaclonal variation occurs in high frequencies after in vitro propagation of False or True Hom plantain cultivars. However this phenomenon is affected by the cultivar, the environment and the interaction between both factors. 5c. Performance of suckers vs. In vitro propagules of plantain This research is of significance to the rapid multiplication of seed stocks in plantain for which in vitro techniques have been advocated. Somac/onal variation The frequencies of somaclonal variants according to the plant propagation method in the False Horn plantain 'Agbagba' were studied at Onne Station in 1992 (Table 80). The test for independence in a contingency table indicated that somaclonal variation and propagation method (conventional vs. in vitro shoot-tip culture) are independent criteria, i.e., the ratio of true-to-type to off- 78 types plants remains the same for both techniques of plantain propagation. The phenotypic variation observed in in vitro propagules of 'Agbagba' is thus not significantly higher than in conventional propagules. This conclusion, however, only is valid for 'Agbagba' until similar 'comparisons are made with other cultivars which exhibit higher frequencies of somaclonal variation (see Table 78). Growth and yield parameters The two propagation methods were compared for commercial production of plantain. Only the true-to-type were included in the analysis (Table 81) of an unbalanced data set. The sampling or within plot error was used to determine the significance of the replication x treatment interaction (experimental error). This was significant for traits which showed a coefficient of variability (CV) > 13%; plant height six months after planting, height of tallest sucker at harvest, YLS, bunch and average fruit weight, and number of hands and fingers. Concerns about potential non-phenotypic variants which were not detected by visual inspection were ruled out by an analysis of frequency distributions. Third (skewness) and fourth (kurtosis) degree statistics were employed to detect non-normal distributions. Skewness measures the degree of departure of a distribution from symmetry while kurtosis determines the peakedness of a distribution. Surprisingly the two traits with high CVs, bunch weight and height of tallest suckers, showed a non-skewed distribution. The significant skewness for plant height, total number of leaves, youngest leaf spotted at flowering, fruit filling time, number of hands and number of fingers in the conventional propagules was explained by the shift of the distribution to low values. High values for days to flowering, days to harvest, number of hands and average fruit weight explained the skewed distribution in the in vitro propagules. These were inherent characteristics of each population rather than the result of somaclonal variants which were not visually detected. Analyses of homogeneity of variance (F-tests) indicated that both populations had similar variability for all the traits except plant height at six months after planting, days to flowering, days to harvest and number of fingers. Growth parameters exhibited more variability in the in vitro propagules while number of fingers had a greater variation in conventional propagules. This demonstrated that an increase variability was not dependent on type of propagation. In general in vitro propagules had their means skewed toward high values. There were significant differences in plant height six months after 79 planting, which is an indirect measurement of plant vigor, plant height at flowering and height of tallest sucker at harvest between conventional and in vitro propagules. The latter suggested that chemical compounds present in the in vitro propagation medium may have residual effects affecting sucker development. In vitro propagules grew vigorously and faster that conventional propagules. However, black sigatoka affected both equally, by reducing their yields. This demonstrated that an improved biotechnology such as tissue culture cannot solve the problem of plantain production alone. In this regard, in vitro propagation has a complementary rol~, to PBIP's conventional breeding program, by disseminating improved TMPx germ plasm which is not affected by black sigatoka disease. 5d. Embryo rescue after triploid x diploid crosses A research fellow is investigating factors that influence immature embryo germination to increase further in vitro germination rates, a crucial aspect in PBIP's breeding program (see Section 2f). Conventional statistical analyses (ANOVAs, regression approaches) were unable to provide meaningful data interpretation. Therefore, non-parametrics statistics for the analysis of frequency distribution were also employed to analyze experiments in this section when the coefficients of variability (CV) were high (Table 82). High CVs were the result of: (i) Unequal sample sizes, i.e. the number of seeds used at a time (ii) Inherent variability in the experimental material, i.e. each embryo is a distinct genetic entity (iii) The marked difference in environmental factors at the different periods when the embryos were excised, and the last but not the least important (iv) The very sensitive nature of the experiments Embryo ages effects The aim of this research was to investigate embryo rescue at different ages (days after pollination). In this way the best embryo age was identified for in vitro germination. Fig. 40 illustrates the differences between four different ages. Younger embryos (50-55 days) had a significantly higher success in embryo rescue than older embryos. 80 Effect of silver nitrate in embryo rescue Silver nitrate (AgN03) promotes regeneration and somatic embryogenesis in tobacco, Hevea, and several monocots such as pearl millet. We investigated whether the germination of plantain mature embryos could be increased by the addition of AgN03 to the basal Murashige & Skoog medium. We found that only the lowest concentration of AgN03 increased significantly the percentage of embryo germination when compared with the control (Fig. 41). Hormone effects in embryo rescue and embryo germination This experiment was set to evaluate the combined effects of embryo ages (50 and 80 days after pollination) and plant growth hormones on the in vitro germination of plantain hybrid seeds. The hormones were 6-benzyladenine (BA) and (X-naphthaleneacetic acid (NAA) at 11lM. No significant interaction between age x hormones was found. Therefore the conclusions applied to both embryo ages. The combined effects of NAA + BA significantly increased the germination of plantain hybrid seeds, however their effects alone were not significantly different from the control (Fig_ 42). Another experiment in which gibberellic acid was added as a third growth regulator to enhance the germination of plantain hybrids did not provide conclusive results. Effect of streptomycin on seed disinfection and embryo rescue We evaluated the effectiveness of antibiotics, e.g. streptomycin, in reducing the rate of bacterial or fungal contamination. Also its secondary effect on in vitro embryo germination was investigated. The antibiotic increased significantly from 7% to 29% the percentage of healthy plantiets recovered after embryo rescue. This antibiotic, however, did not affect the germination rates. Preliminary results of early screening for black sigatoka using embryo-cultured derived seedlings. This on-going research aims to identify procedures for early detection of black sigatoka resistant hybrids obtained after embryo rescue. Progenies from BT x C4 were in vitro germinated, and seedlings along with the parents were exposed to the disease in Onne Station. Preliminary analysis revealed that there were differences between the hybrids with respect to their susceptible parent in black sigatoka incubation times, or the time between fungus penetration and appearance of the first disease symptom. No differences, however, were found between the hybrids and the parents in the symptom evolution time, i.e. the 81 period following the appearance of the first symptom until the formation of the necrotic spot (last stage of disease development). 5e. Transgenic Muss products In collaboration with the Catholic University of Leuven (KUL) the genetic stability of cooking banana plants (Musa spp, ABB group, cv. 'Bluggoe') regenerated from cell suspension cultures by somatic embryogenesis was investigated (see PBIP 1991 Annual Report). In a continued joint effort with this university, plants obtained from cryopreserved cell suspensions were also screened for somaclonal variation at Onne station. Recently, the team led by Dr R. Swennen has developed a protocol for the production of transgenic Musa plants. It consists of shooting tungsten particles coated with plasm ids containing the GUS marker gene and an antibiotic resistance gene into embryogenic cell suspension cultures. Many cells with transient expression were recovered. After a culture of about three months in a liquid medium containing an antibiotic, a few cells were selected. These cells were Jransferred onto a semi-solid medium where they multiplied. All cells coloured blue within 2 hours when treated with X- gluc. This is a confirmation of the first ~table transformation in Muss and that it was possible to select those cells among non-transformed cells. Recent research at KUL by Dr ~. Cammue has led to the discovery of plant-seed antifungal proteins that inhibit Mycosphaerella fijiensis and Fusarium I spp. We (KUL and IITA) propose that the introduction of genes encoding antifungal proteins into the genome of banana and plantain using recombinant DNA techniques as described above ~hould be pursued. The utilization of different resistance genes also provides a means to increase the durability of the resistant genotypes. The proposed recombinant technology seems to be the only means for the genetic improvement of certain groups of Muss such as the Horn plantains and the 'Cavendish' bananas. We have written a jOint project, awaiting funding, which focuses on th~ direct production of transgenic cultivars and the incorporation of antifungal pro!eins in IITA's Musa breeding populations for further manipulations by conventional methods. This proposal combines the comparative advantages in Musa genetics and improvement of the institutions involved. The hybrid Musa germ plasm (TMPx or TMBx) developed by PBIP/IIT A as well as its expertise in field testing and selection are the intended contributions of liT A to this project. 82 Sf. Molecular markers This is a collaborative project which is being carried out through a contract with Dr R. L. Jarret (USDAIARS, Georgia, USA). The Musa genome is investigated using restriction fragment length polymorph isms (RFLPs) and random amplified polymorphic DNA (RAPDs). Taxonomy and germp/asm organization The ability to accurately estimate genetic relatedness in germplasm collections will be an important tool in Ihe conservation and management of the genetic diversity in gene·banks. The inability to estimate genetic diversity results in gross over· and under·estimations of the need for additional collections. As a consequence of the inadequate germ plasm characterization, unique materials may not be added to the collection or potentially valuable materials may be discarded. Molecular markers (MM) aFe the most important tools available for DNA fingerprinting. MM are very useful to estimate genetic diversity and genetic identity in the genus Musa . The polymerase chain reaction (PCR) was used to detect polymorphisms among 29 diploid clones of Musa acuminata Colla collected in Papua New Guinea (PNG, Fig. 43). Primer sequences were derived from a 520bp highly repetitive and dispersed sequence isolated from M. acuminata subssp. ma/accens/s (Fig. 44). Primers (Table 83) , used individually, detected a total of 48 polymorph isms that were scored as unit characters and used to generate a Jaccard's similarity index. The relative abundance of variability within the PNG diploid bananas reflects the extreme variability within the M. acum/nata gene pool. Principal coordinate analysis was used to cluster clones (Figs. 458 and 45b) and the unweighted paired·group method of analysis (UPGMA) was used to compute genetic distance among the materials examined (Fig. 46). PCR with primers from a highly repetitive sequence is a rapid and reliable means of detecting genetic diversity in M. acum/nata. Molecular linkage map The development and utilization of molecular markers of Musa could help to obtain basic genetic knowledge and to apply it to plantain and banana breeding. For example, they could help to elucidate the genetic nature and origin of the diploids obtained from 3x x 2x crosses involving plantains or the genetic mechanism involved in the modified megasporogenesis leading to the formation of 2n eggs in plantain (see Section 2w). 83 We (USDAIARS and PBIP) are developing the molecular linkage map of Musa using RFLPs, RAPDs, and hypervariable sequences to detect variable number tandem repeats (VNTRs). RFL.,Ps are the differences in the size of DNA fragments generated upon digestion of genomic DNA of different genotypes with restriction endonucleases. RAPDs reslJlt from differential annealing and extension of primer DNA to template (Qenomic) DNA in the presence of Taq polymerase. Tandem repeats are dispersed throughout the plant genome of most organisms and, when minisatellite repeats are used as probes, they detect individual-specific hypervariable sequences in humans, animals and plants. Various random genomic probes (> 1aOO), random primers and cloned minisatellites, including the M13 bacteriophage repeat, are being evaluated for their ability to detect polymorph isms in segregating Musa populations. In 1992 we extracted at IITA, DNA from different Musa accessions, somaclonal variants and segregating ~opulations (2 testcrosses and 1 F2)' Test crosses were derived by intermating p!antains and C4, while the F2 was the result of selfing a TMp2x (Tabla 68). Additional segregating populations were developed in 1992 for mapping purposes and to identify useful markers for Musa breeding across different genetic backgrounds. We have almost finished with the morphological characterization using the Musa descriptor list of five of these segregating populations. Preliminary reports about the inheritance of morphological markers were presented in Section 2w. Several hundreds of molecular markers have been and are still being produced by our cooperators in USDAIARS and are used for the construction of a highly saturated Musa molecular map. This provides means to select the most informative probe enzyme combinations, arbitrary primers (Table 84) or tandem repeats, for detecting molecular polymorph isms and hypervariable sequences in Musa segregating populations. In this way, the identification of chromosome regions carrying loci of interest may be possible for further genetic manipulations using non-conventional or modified breeding approaches: genetic engineering and/or marker assisted selection. Dissecting black sigatoka resistance with molecular markers Molecular marker loci linked to tre loci for black sigatoka resistance (see Section 2w) are also being investigated. As such, the quantitative trait variation (QTV) for black sigatoka resistance may be dissected into its components and the suggested main types of gene action controlling black sigatoka phenotypic expression in Muss (see Section 2w) may be confirmed. In this regard, the two test-cross populations derived from triploid-diploid crosses and an F2 generated B4 by selting a susceptible TMP2x (see Section 2W.) have been evaluated for the disease and are being scored using molecular markers. Marker assisted selection A project, still pending approval, has been proposed to FAO to initiate this research in collaboration with FHIA and USDAIARS. The project focuses on , black sigatoka and burrowing nematode resistance in plantain and banana. The most informative markers (probe enzyme combinations, arbitrary primers or tandem repeats linked to loci coding for a specific resistance) are being identified by combining field evaluation of UTA and FHIA breeding materials with DNA information. They will be used as 'landmarks' for marker assisted selection (MAS) in early stages of plant development and for introgression of exotic germplasm into the cultivated Musa gene pool. We expect that the markers linked to BSR could be used for MAS in the roguing nursery to identify promising diploid, triploid and tetraploid breeding materials. The efficiency of MAS will depend of several factors such as linkage disequilibria, number of marker loci, population size, heritability of the trait of interest, and the percentage of additive variance explained by the marker. An improvement of such technology for plantain and banana breeding should be achieved by developing an optimum selection index which combines information on molecular with phenotypic measurements. Therefore, both molecular and field work are required to achieve this goal. Alternatively the identification of functional gene sequences and/or protein-derived products providing resistance to this disease could be useful for the production of diagnostic kits for the in vitro selection and rapid multiplication of black sigatoka resistant genotypes. Transmission of chloroplasts in Musa The determination of the mode of inheritance (uniparental vs biparental) of chloroplasts in Musa spp. will be carried out by using cpDNA probes and intra- and inter-specific crosses. 'Calcutta 4' (M. acuminata ssp. burmannica) has been crossed as male and female with M. a. ssp. banksii acc. No. 237. Several thousand seeds and hundreds of seedlings were obtained from these reciprocal crosses at Onne in 1992. Seeds or in vitro plantlets will be sent to USDAIARS for DNA extraction and fingerprinting. 85 Outcrossing rates in diploid bananas 'Calcutta 4', the homozygous, wild diploid seeded clone was used to determine the outcrossing rate in diploid germplasm. Seeds extracted from selfed and open pollinated bunches of C4 (see Section 2e.) were sent to USDAIARS for DNA analysis. This information will allow us to determine the rate of outcrossing, mainly done by bees and bats, in this clone. 5g. Secure and rational management system for In vitro propagated and conserved germplasm of Musil In vitro culture is a valuable component of current propagation, conservation and improvement strategies for Musa. However, the full exploitation of in vitro culture is hindered py the occurrence of somaclonal variation. Many somaclonal variants are not apparent until in the field, often only at fruiting. This has serious practical and economic consequences for banana/plantain production at the farm level, threatens the security of in vitro conservation, and hampers the efficient application of biotechnological approaches to genetic improvement. Research over recent years has advanced towards characterizing somaclonal variants and understanding some of the underlying factors. However, there is a need to develop a strategy for the rational management of in vitro collections for Musa germplasm conservation. A project under FAO's Associate Professional Officer (APO) scheme has been proposed to address this issue in collaboration with the Intemational Plant Genetic Resources Institute (former IBPGR, Rome, Italy), KUL and others. The project mainly aims at investigating factors that influence the nature and extent of somaclonal variation in plantains and bananas in order to develop improved in vitro culture management procedures for the control of instability in culture. The project will also explore methods for the early detection and characterization of somaclonal variants with a focus on biochemical and molecular markers. This endeavour may eventually result in a genetic map of somaclonal variation in Musa, which could elucidate the underlying causes and origins of this phenomenon. 86 6. POST HARVEST QUALITY ,OF PLANTAINS Several promising black sigatoka-resistant hybrids have arisen from crosses between wild diploid bananas and cultivated plantains. These hybrids are not plantains with resistance, but should be considered as novel genetic combinations. Hybrids may therefore ~ave different fruit ripening and quality characteristics compared with plantain parents and this may affect consumer acceptability. This research is aimed to determine criteria associated with plantain fruit quality and to provide breeders with rapid simple methods for efficient post harvest screening. The screening techniques being investigated include physico-chemical tests and taste panel analysis to determine fruit quality. The major criteria are fruit durability and palatability. This project is also conducting work on socio-economic aspects of plantain marketing, product development and developing methods for NARS to test the acceptability of new plantain hybrids at the farm level. 6a. Fruit quality evaluation of Muss hybrids The plantain breeding program has produced many promiSing black sigatoka resistant plantains (TMPx) which are currently being field tested for resistance durability and agronomic performance. Post harvest research is focusing on fruit ripening characteristics of the hybrids compared with the parents and plantain from the favoured False Horn gene pool. Fruit materials are obtained from METs and PYTs (most recent promising selections) at Onne Station. Key criteria which are assessed include: fruit ripening period, fruit weight loss during ripening, initial sugar, sugar:starch ratio's, starch hydrolysis during ripening, pulp: peel ratios, rate of tissue softening, dry matter content, and mineral composition. These data will be used to obtain a ripening profile specifically for the hybrids but also to provide data of ripening characteristics across the gene pool, for further use by the breeding team. Studies on dry matter content will also be used to determine the effect of desiccating period on seed germination. This will only apply to pollinated bunches. The data in Table 85 show three of the post harvest parameters which will be used by the breeding team. The parameters of dry matter, fruit weight and pulp: peel ratios were taken from clones in the MET-1 trial located at Onne station. The data also indicate the rank order of the clones for these parameters. It should be noted that the data sets for the postharvest section are on-going, 67 additional data are required to provide more statistically valid data. Data from the starch and sugar analysis, tissue softening and mineral composition are on- going and are not presented at this stage. Dry matter Dry matter has become increasingly important as a quality criterion, especially when developing new plant varieties. The dry matter immediately provides a measure of water content and clearly a breeder first needs to determine whether increased yield is due to higher water retention or a genuine increase in harvested product. Dry matter is also important for correlations with storage period and cooking quality. The ranking for dry matter (Table 85) showed that 'Pelipita', a cooking banana, had the highest dry matter content (OM) at 41% and the dessert banana 'Valery' had the lowest dry matter at 23%. According to previous reports, 'Pelipita' has good cooking qualities which may be related to high dry matter content. it is generally accepted that dessert bananas are eaten raw and have poor cooking qualities for which low dry matter may be a contributing factor. When dry matter contents were pooled and averaged according to the four distinct types of Musa fruit, the order from high to low of dry matter content was plantain, cooking banana, plantain hybrid and dessert banana respectively. The dry matter for the plantains ranged from between 36-37% OM, cooking bananas ranged from 31-41% OM, plantain hybrids ranged from 31-35% OM, and bananas from 23-29% OM. These early results indicate that plantain hybrids have a dry matter which is 2-3% lower than true plantains but higher than the cooking bananas 'Bluggoe' and 'Cardaba', and the dessert bananas. When percentage of dry matter for the Musa fruits are compared with other perishable staples (Table 86), then it is clear that dry matter content for the plantains, cooking bananas and plantain hybrids are favourable relative to other starchy staple foods. An important aspect of OM analysis is that although the plantain hybrid is derived from a plantain crossed with a banana type, the hybrid offspring retain a closer association to their plantain parent. Fruit weight The fruit weight data were taken from the second hand of the bunch. The second hand was selected as it typically provides less variability than fruits from the first hand. The ranking order for the fruit weights in this experiment (Table 85) showed that the cooking banana 'Bluggoe' produced the heaviest fruit followed by the False Horn plantain 'Agbagba'. Excluding the hybrid clone 597- 88 4, which the breeders have rejected due to its black sigatoka susceptibility, the fruit weights of the other seven hybrids ranged from 160-191 g, which was clearly higher than the fruit weights from the ~ybrid plantain parents OL (151g) and BT (114g). Individual fruit weight is important in terms of market acceptability. In the market place fruits are typically sold individually and consumers prefer to purchase large fruits. In this respect despite having a high dry matter, the lower weights of the cooking bananas 'Pelipita' and 'Fougamou' may be less acceptable in the market. Pulp:peel ratios The pulp: peel ratio data (Table 85) provides further information on the value of yield. A high pulp: peel ration indicates the fruit has higher proportion of pulp compared with the peel. The trends in this data set are not apparent but it is clear from the rankingr that PI and 'Fougamou' have a very thin peel compared with the hybrids, 597-4 and 612-74 offtype. The dessert banana 'Valery' also has a low pulp: peel ratio and the hybrids were generally lower than the plantain parents. Fruit ripening Fruit ripening in this experiment was conducted on individual fruits selected randomly from the second and third hands of the bunch. The bunch was harvested when the fruits were fully mature and taken immediately to the laboratory where the fruits were cut prior to selection. Temperature in the experimental area was maintained at approximately 250 C and was vented every 2 hours to prevent build up of ethylene. The ripening period was measured in terms of changes in the peel colour ov~r time. This data set is the period from harvest until the fruit peel reached colour stage 7, when the peel was fully yellow with presence of black spots. The data set in Table 87 has been analysed because the number of replications was acceptable but this data set is also on- going. Fruit ripening period is of considerable commercial interest to farmers, retailers and the consumer. The ripening period of the hybrid should therefore be consistent with the plantain parent and the breeder will be interested in the performance of the hybrid relative to the market standard. It was interesting to find that the cooking banana 'Pelipita' had the longest ripening period of 18 days. 'Agbagba' had a significantly lower ripening period of 12 days compared with 'Pelipita'. However for the purposes of this experiment 'Agbagba' is used 89 as the reference clone, because 'Agbagba' is a False Horn plantain grown extensively in the plantain belt of South East Nigeria which attains a high market price. The result of most interest to the breeder in this data set was that the plantain hybrid TMPx 1112·1 had a ripening period of 12 days which was equal to the reference clone. All the hybrids had a significantly longer ripening period than the cooking bananas and more importantly had a longer ripening period than the plantain female parents. The female plantain parent of most of the hybrids, OL, was found to have a ripening period of 7 days which corresponded to 63% of the Agbagba ripening period. This result was consistent with other previous studies. Once more of the physically measured characteristics have been evaluated, these data will be assessed relative to the Ideotype (see Section 2z). At a later stage when more defined data profiles are available a set of criteria can be developed by the breeders to determine a threshold value for each characteristic for selection using independent culling levels. 6b. Sensory evaluation of Muss hybrids Objective physiological measurements (see Section 6a) are complemented with subjective studies on fruit palatability for which taste panels are used. These data provide information on important fruit quality characteristics for consumers and indicates potential for consumer acceptability. At Onne Station taste panels are used for sensory evaluation of boiled unripe and fried ripe fruit. Criteria for testing include taste, colour, sweetness, texture and acceptability. Fruit samples are also ranked for preference to provide information which compares clones. Although the initial studies are conducted at Onne Station, they will be extended to other locations which have the MET or AMYT cional set when fruit are available. Taste panel testing will be set up with NARS for these sites. It is hoped that studies on 'matoke' and beer preparation can be made in Uganda and taste panel analysis of 'fufu' will be conducted in Ghana. Post harvest training was given to trainees during the plantain technology training course (see Section 8c.) and this information will be followed by further training at key sites before hybrids are harvested. Taste panel data The taste panel consistS of a pool of staff from Onne station. The pool comprises 26 people, 7 women and 19 men, although this profile changes with 90 staff changes. The taste panels are conducted on a weekly basis when fruit is available and each panel consists of at least 10 people. Panelists are initially presented with either bunches or cut fingers which they score for preference. This test attempts to determine consumer preference for the fruit as sold in the market. Panelists are then presented wi\h samples of the same fruit clones which have been prepared by boiling or frying. Each panelist is presented with 5 different numbered samples and a form to complete. Tests have a duration of approximately 45 minutes. Individuals are also randomly selected from the taste panel and subjected to an in depth interview. These interviews provide further information on how people prepare plantain, quantity of plantain consumed by family, market price and their reaction to plantain hybrids. Taste panel for Musa clones prepared by boiling green unripe fruit The data in Table 88 show the .results from taste panel comparisons of 18 Musa clones which were prepared as boiled sections. The table is constructed to show the relative preferences for plantain, IITA plantain hybrids, cooking bananas, FHIA hybrids and a dessert banana. The table provides an overall ranking of clones and shows the percentage preference for particular clones against other clones. For examples, the first data point for 'Ubok Iba' indicates that when 'Ubok Iba' and 'Agbagba' were compared, 89% of the responses favoured 'Ubok Iba'. This elata set therefore provides the breeding team with comparative data on how individual clones performed against standard clones and other hybrids. The mean indicates the overall preference for a clone in all comparisons and is accompanied by the number of observations for that clone. The final rank is based on the mean value. The most preferred fruit were the three plantains, 'Ubok Iba' (True Horn), 'Agbagba' (False Horn) and OL (French). The next most preferred clone was TMPx 1112-1, followed by 'Pelipita'. Preferences ratings for TMPx 548-4 and 548-9 were ranked at 6 and 7 respectively. The IITA clone that was distinctly not preferred was 612-74 offtype, ranked at 16 out of the 18 entries. Data from the cooking bananas suggested that 'Pelipita' and 'Bluggoe' were relatively favoured and were in the top 8 clones. However the cooking bananas 'Cardaba' and 'Fougamou' wl3re ranked 12 and 13 and were therefore not considered favourably. The FHIA hybrids FHIA-2 and FHIA-3 ranked considerably worse than the plantains and plantain hybrids against which they were tested. FHIA-2 is a candidate for the dessert banana market and a low ranking was not considered unusual. However FHIA-3 is reputedly derived from 91 a cooking banana and is aimed as a potential resistant plantain like product for Africa. The tests from the boiled taste panel data do not suggest that FHIA-3 is a favoured candidate. The basis of clonal preferences is explained to some extent by the information in Table 89 which lists the quality criteria that were assessed by the panelists on the individual samples and their mean response. For each criterion the panelists were required to state whether the fruit was good or bad and then the panelist was given a number of choices to explain this quality assessment. The first two columns in Table 89 report colour quality and colour rating. Clearly the plantain cultivars were scored as having a very good quality with a dark yellow-orange colour after cooking. The fruits which were scored as poor included Tuu Gia, FHIA-1, FHIA-2, FHIA-3 , TMBx612-74 and 'Valery'. The colours of these fruits were a light yellow or white as in the case of FHIA-3. The white pulp colour is more typical of the cooking banana from which FHIA-3 is derived. The taste score data show that the plantain and 'Cardaba' were of good quality with a good plantain like flavour. The other cooking bananas along with TMPx 1112-1 and TMPx 548-4 had an acceptable flavour. The remaining clones were considered to have a poor flavour except for FHIA-2, FHIA-3 and 'Tuu Gia' which were scored as having an unacceptable flavour. The sweetness data were somewhat surprising in that the dessert bananas were scored as not sweet enough. This was unusual because dessert bananas commonly have a higher sugar content than plantains. One possible explanation for this was that the sugar present in the fruit was leached during boiling which gave the pulp a bland flavour. Texture indicated that the IITA hybrids were of about the right firmness except for TMBx 612-74. This cooking banana hybrid is derived from a cross between 'Bluggoe' and C4. 'Bluggoe' has a 5-6% lower dry matter than plantain and this may reduce the bOiling quality. The acceptance data provide a summary of the taste panel information. For ease of comparison this data set has been put in rank order in Table 90 and the order corroborates well with the findings in Table 88. In the quality column the plantains were scored as very good. 'Pelipita', TMPx 548-4 and TMPx 1112-1 were considered as good. The remaining IITA hybrids bar one were rated as fair. The dessert banana 'Valery', 'Tuu Gia', FHIA-1, FHIA-3 and TMBx612-74 were poor. Similarly in the ranking overall, only the top seven clones were considered as fully acceptable. 92 The data set is useful because it can provide the breeding team with a ready made screen. Although the clone TMPx 548-9 had shown well in the ratings (Table 88) it failed to attain acceptance rating. The main reasons for this were that TMPx 548-9 had a poor flavour many respondents stating that it was not sweet enough. A probable overiding factor however was that many respondents stated that it was of good quality but contained seeds which were undesirable. The high number of seeds was also the reason for the low ranking of TMBx 612-74. Seed set One of the common problems that has arisen from the taste panel analysis concerning the hybrids and cooking bananas is the occurrence of seeds in the pulp. For example the triploid-fertile c\/.. 'Pelipita' produces on average 233 ± 54 open pollinated seeds. Musa seeds are pea sized, hard, black objects which distinctly reduce the quality perception of the fruit. Seed set may be a problem with the more fertile tetraploid plantain hybrids. How important seed set will be in the future is at present uncertain. Seed set at Onne Station is believed to be high due to the levels of pollen being wind dispersed from the pollination blocks and from the male diploid lines being evaluated in the same experiment. Preliminary results from hand pollination (Section 2d) suggested that tetraploids have reduced male fertility. Moreover, in the farmers' fields there are no diploid pollinators and therefore there will be very little viable pollen to cause seed set. We are now planting a selected hybrid into farmers fields to assess the fruit quality as perceived by the farming community. Seed set will be one of the criteria evaluated. Alternatively the hybrid will be supplied with information that the male bud should be removed once the bunch has formed. This action will eliminate the pollen source and is the standard practice for cooking banana production in southeast Asia and for export banana production in tropical America. Taste panel for Musa clones prepared as fried ripe sections - (local name, dodo) The data in Table 91 shows the results from taste panel comparisons of 17 Musa clones which have been prepared as fried seclions from ripe fruit. The mean indicates the overall preference for a clone in all comparisons and is accompanied by the number of observations for that clone. The final rank is based on the mean value. The preference ranking for the clones show that 'Ubok Iba' and 'Agbagba' retain positions 1 and 2. Positions 3,4 and 5 have changed in the order 93 compared with boiled data but retain the same clones, viz rank orders 3,4,and 5 to clones Pelipita, OL and TMPx 1112-1 respectively. A note of caution is needed with the data set because 'Pelipita' was not compared directly with OL or TMPx 1112-1. The absence of these comparisons may effect the overall ran kings. The data in Tables 92 and 93 indicate the basis for the preference ranking. The dodo data show similar information to the boiled data but with some clear changes in clonal performance. The most preferred colour for dodo was a deep golden colour. The clones TMPx 1658-4, FHIA-2 and 'Valery' produced a light yellow product and were therefore scored as poor. Almost all clones provided a reasonable flavour when fried. This common feature was probably due to the flavour imparted from the refined oil used to fry the sliced pulp. The exceptions were 'Fougamou', FHIA-2 and 'Valery'. Sweetness data indicated that 'Bobby Tannap', 'Fougamou' and FHIA-2 were not sweet enough, the plantains, 'Pelipita', TMPx 548-9, TMPx 2796-5, TMPx 1112·1, TMBx 612·74 and FHIA-1 were acceptably sweet. The remaining clones were too sweet. Texture data indicated that the plantains, TMPx 548-9, TMPx 1112-1, 'Bluggoe' and 'Pelipita' were of about the right firmness. The remaining clones being considered slightly too soft and FHIA·2 and 'Valery' disintegrated on frying. Table 93 shows the preference of the clones according to quality and acceptability score. Ten clones were considered as good quality and 11 clones had acceptable quality. These data reconfirm the information from the previous data that the plantains were the most acceptable clones. However more importantly the data indicate that for the dodo test the IITA hybrids TMPx 548-4, TMPx 548-9, TMPx 2796-5 and TMPx 1112-1 were consistently highly rated. The data also suggest that the IITA hybrids 597-4, TMPx 582-4, TMPx 1658-4 and TMBx 612-74 were not preferred. Similarly 'Fougamou', FHIA-2 and 'Valery' were not clone with potential as dodo. Conclusion to Taste panel data The data from the taste panel tests remain at an early stage and the number of blank spaces in the tables shows that far more data are required. However at this stage an overall picture is emerging based on the information from the two data sets and this is valuable. Tthe liT A hybrids TMPx 548-4, TMPx 548-9, TMPx 2796-5 and TMPx 1112-1 were consistently highly rated and therefore should be considered for continued advanced testing. The data also 94 suggest that the IITA hybrids 597-4, TMPx 582-4, TMPx 1658-4 and TMBx 612- 74 were overall not preferred and may not serve as potential candidates for replacing black sigatoka susceptible plantain. Once the clones have been assessed more fully using several physical parameters and evaluated via the taste panel, a detailed profile of a clone can be made available to the breeding team for selection and genetic analysis based on fruit quality. Other projects related to fruit quality include the evaluation of fruit damage susceptibility and fruit storage. Data on these topics are being collected but are insufficient for reporting at this time. 6c. Plantain utilisation Hybrid plantains closely resemble but are not identical to parent plantains. Hence there is a need to compare the utilisation characteristics of hybrids with parents as differences between hybrids and parents may confer advantage, e.g. in terms of cooking time or sugar content. Initial studies are investigating preparation times for hybrids in traditional dishes. Later studies will investigate simple processing techniques such as vinegar yields from high sugar content hybrids, for storage purposes. Projects are on-going on a number of these topics: (i) Cooking time The cooking time of the clones included in the MET sets is being assessed in terms of changes in texture. Fruit are cut and placed into boiling water. Three sections are removed every 2 minutes and these sections are tested for texture using a penetrometer. This data will be used to correlate with the starch levels and texture softening information. (ii) Wine production Overripe plantain pulp appears to be a good medium for wine production. The initial brews to which sugar was added produced an alcohol yield of 16 %. (iii) Vinegar production Studies are continuing to brew different clones for wine production and then attempt to convert alcohol into vinegar. Acetobacter will be sent from Belgium to enable conversion. (iv) Plantain chips The use of sealed polyethylene bags to store plantain chips has proven successful. Chips are of high quality after 3 weeks compared with 2 days using local thin plastics as used in the markets .. Optimum chip size has been selected and now the product is being produced on a commercial basis to test in the market place whether this type of product 95 is commercially viable as a cottage industry. Marginal profit analysis will follow. The period from May -August is the famine period for plantain and the price of the initial product may determine viability. 6d. On-farm assessment of yield and quality loss from black slgatoka This project will serve several functions: (i) To determine the effect of black sigatoka disease on production (ii) Determine effect of sigatoka on fruit quality, with special emphaSis on seed set (iii) To develop methods for on-farm post harvest evaluation Present interest from Imo State-ADP (southeastern Nigeria) on post harvest evaluation at the on-farm level will be followed up and incorporated in the next on-farm assessment. Plants for this project were generated during the year and will be planted in the rainy season of 1993. A second plan for future on-farm work is being developed which will concern on-farm varietal testing. Along with these activities PBIP has supplied equipment for cassava chipping, developed by engineers of IITA's Post Harvest Unit, to farms in the neighbouring area to Onne Station. These farms will receive both new cassava as well plantain hybrids in the 1993 planting season. 6e. GenetiCS of fruit quality Preliminary analyses based on results shown in Table 85 suggested that dry matter content was under the control of genes exhibiting additivity. Most of the hybrids had values in between plantain (high dry matter content) and diploid banana (low dry matter) parents. Ther~fore, phenotypic recurrent selection may be an efficient method to increase the frequency of favourable alleles for high dry matter in the breeding populations. Epistatic interactions arising from the interactions between the plantain genes and the banana genes (as discussed in Section 2w), rather than overdominance, explained transgressive segregation for fruit weight. Most of the hybrids not only exceeded significantly the mid-parent value but also the fruit weight of the plantain parent. Preliminary analysis of storage duration of fruit (Table 87) revealed that different genetic systems may be operating in PI (an structural heterozygote) and C4 (a true-breeding line). C4's hybrids either with plantains or the cooking banana 'Bluggoe' exhibited transgressive segregation for storage duration of the 96 fruit probably due to complementation between the parental genes. While in PI the results suggested that partially dominant genes (towards the high values of the banana parent) were controlling storage duration of fruit in TMPx 2796-5 and TMPx 1658-4. Data for pulp:peel ratio did not give any conclusive evidence about the genetic system controlling this trait. More accurate genetic analyses about quality traits will be carried out using the TMp2x germplasm. Currently we are assessing different fruit characteristics in this germ plasm. 6f. Molecular and biochemical studies of Muss fruit ripening Work being conducted by Drs. Tucker and Vendrell (Univ. of Nottingham, UK) is concerned with the effect of polygalacturonase (PG) on cell wall degradation which causes the fruit tissue softening associated with fruit ripening. Drs. Tucker and Vendrell are looking at polygalacturonase in banana and have started to screen mRNAs for PG message. Identification of mRNAs and cDNAs for the key cell wall degrading enzymes would provide good RFLP probes to assist with breeding work (see Section 5f). It is intended to use the IITA Muss collection (see Section 2a) as the base from which to isolate a range of ripening response material for testing following primary screening at Onne Station. This collaborative project is subject to funding from European Economic Community. The project will be finalised at Nottingham University by Dr. G. Tucker to be submitted for the 1993 funding round. 97 7. GENOTYPE-BY-CROP MANAGEMENT SYSTEMS INTERACTION Careful management of organic matter is essential to achieve sustained perennial productivity of plantain under large scale field production conditions. Agroforestry systems such as alley cropping and the management of regrowth of natural bush fallow species in plantain fields are being investigated for their capacity to maintain productivity over long periods of cultivation without degradation of the resource base. AI~o, the performance of improved genotypes relative to the crop management system should be assessed. On-site cultural practices used in breeding trials may differ considerably from typical farmer practice and could influence genotype selection. Research on this project is carried out in cooperation with agronomists/soil scientists of the Resource and Crop Management Division (RCMD) of IITA. 7a. The perennial yield potent,al of plantains This research aims to determine the long term yield potential of six plantain cultivars representing different taxonomic groups. The experiment, planted in 1987 with in vitro produced plants, will continue until 1993. The trial • was established in a 10-year old bus~ fallow which was cleared manually and not burnt. In the 3 m wide alleys between the rows of plantain, the natural shrub and tree species of the bush fallow were allowed to regrow in order to form a spontaneous, muHispecies alley cropping system. Through regular coppicing, the prunings of the multispecies hedgerows serve as mulch for the plantain crop, as such providing the organic matter required for sustained plantain productivity. In 1991, a research scholar from Belgium performed the data analysis of the first 2 production cycles, the plant crop and the first ratoon. The analyses were published in the PBIP 1991 Annual Report. The analYSis for the six year period (1987-1993) will be done considering the effects of years (Y), plantain cultivars (P), and the YxP interaction in the following traits: pseudostem height, total number of leaves, number of leaves at flowering, length/width ratio of the seventh leaf, height of tallest sucker atflowering and at harvest, days to harvest, bunch weight, number of hands and fingers, and length of production cycle. 7b. Survey of changes In floristic composition within natural multlspecles hedgerows under plantain cropping A botanical survey was conducted to determine the number of plant species and the changes occurring in the floristic composition of natural hedgerows established from fallow species that were allowed to regrow in order 98 to form a spontaneous multispecies hedgerow under plantain cropping. The study was conducted in a 4 ha, 5-year old plantain field which was planted in a manually cleared 10-year old bush fallow. About 120 different plant species were identified in these natural hedgerows, which were coppiced regularly to provide mulch to the plantain. An increase in the number of plants per unit area was observed with increasing plot age, i.e., the longer the plot was cultivated with plantain, the higher the number of plants in the hedgerows. The most frequent species were Anthonata macrophylla, Mil/etia arborensis and the aggressive shrubby weed Chromo/aena odorata. The presence of highly nodulating tree species and the high Qiomass produced by plants such as Chromo/aena odorata, which provide a source of mulch, could explain the good perennial performance of plantain in this field. Research by the RCMD and the International Center for Research in Agroforestry (ICRAF) continues on the management of the multispecies hedgerows, to determine the rate and pattern of increase in the population of shrubs, trees and herbaceous annual species found in this field. The effects 01 the multispecies hedgerows on the nutrient status of this site will also be monitored. 99 8. NARS COLLABORATION, TRAINING AND EXTENSION ACTIVITIES Because of the recent expansi9n of UTA's plantainlbanana improvement work, NARS collaboration is less developed than for other UTA crops. PBIP aims to strengthen NARS through cooperative research, consultancy support, individual training (scientific visits and PhD students), group training. transfer of improved Musa genetic materials and production packages, and publication of production manuals, research guides and other training materials. 8a. Multllocational testing of plantain and banana hybrids The Nigerian Interministerial Committee for the Control of Black Sigatoka Disease requested the concourse of PBIP for the development of new breeding materials and technology in this area. In 1992. both programs started a joint effort for the testing of TMPx germ plasm in 11 Nigerian sites to identify promising genotypes for future cultivar release (see Section 2m.). Also CRBP (Cameroon) and CRI (Ghana) joined the MET-2 nelwork. A total of 12 MET sets, consisting of 200 test plants of 20 clones (Table 2~) and 200 'Valery' borders, were sent to the 13 African cooperators. Also MET-2 sets, consisting of 2 plants of each entry (for further rapid in vitro multiplication and later field testing), were sent to our cooperators in Cuba, Dominican Republic and Australia. The AMYT-sets (Table 27), comprising only the testing clones (360 plants per experimental site) were sent to nine cooperators of the AMYT-network (Table 28). Two experimental protocols were developed and written versions of each as well as data sheet notebooks were distributed. The booklets: . Vuylsteke, D. & R. Ortiz. 1993. Advanced Musa Yield Trial 1 in East Africa. Plantain and Banana Improvement Program, International Institute of Tropical Agriculture, Onne Station, Nigeria. 4 pp - Vuylsteke. D. & R. Ortiz. 1993. Advanced Musa Yield Trial I in West Africa. Plantain and Banana Improvement Program, International Institute of Tropical Agriculture, Onne Station, Nigeria. 4 pp describe experimental procedures and field layouts and are available upon request for NARS researchers interested in the evaluation of Musa germplasm in their own environments. The popular reference manual: - Swennen, R. 1990. Plantain production under West African conditions. lITA, [badan, Nigeria. 24 pp was also included in the shipment. ab. Distribution of bananas PBIP cooperation was requested by the Commissioner of Agriculture of Rivers State, Nigeria, to continue providing planting materials (suckers and in 100 vitro plantlets) of the cooking banana cv. 'Cardaba', which has been released as a new cultivar for this area. In 1992, PBIP provided support to the local government in the distribution of this cooking banana to farmers. More than 2500 seedlings/suckers of the cooking banana cvs. 'Cardaba'(176S), 'Bluggoe' (70), 'Pelipita' (339), and the black sigatoka resistant dessert banana 'Yangambi Km. 5' (366) were distributed as basic seed not only to the Ministry of Agriculture of Rivers State but also to Akwa Ibon-AOP, Rison Palm Ltd. (Yenagoa), Prison Farm (Elele) and the extension services of Shell Co. and NAOC (AGIP Ltd). In vitro virus indexed plants of the bananas 'Yangambi Km. 5', 'Fougamou' as well as the plantains 'Agbagba' and 'Muracho' were sent to Seychelles Islands. The black sigatoka resistant banana 'Yangambi Km. 5' was also sent to Cuba. Bc. Training Group training IITA and INIBAP organized a three-week training course in Plantain and Banana Research and Technology Transfer for research workers and extension supervisors in the humid and sub-humid regions of Africa. The majority of course activities took place at Onne Station. The course aimed to provide basic knowledge and skills in plantain and banana cultivation to individuals responsible for establishing and managing field research. Training was offered in November 1992 to 20 young African scientists and technicians. IITA and Nigerian scientists participated as resource persons on specific topics. This training course provided a forum to discuss the most important limiting factors for the production of plantain and banana in sub-Saharan Africa. Course participants from 10 countries (Cote d'ivoire, Ghana, Togo, Nigeria, Cameroon, Sao Tome & Principe, Zaire, Uganda, Tanzania/Zanzibar and Malawi) indicated that disease complex (mainly sigatoka, banana weevil and nematodes), yield decline, and post-harvest losses are the major constraints to banana/plantain production in Africa. Group training for Nigerian technicians evaluating the MET-2 (see Section 2m) was also provided in May, 1992. IITA scientists and PBIP senior staff based at Onne station were the resource persons. 101 Individual training Individual degree-training was provided to Mr. G. I. Harry who is investigating in vitro embryo rescue for the enhancement of hybrid plantain production (see Section 5c), and to Ms. H. Vandenhout , whose research focused on the determination of ploidy levels and ploidy effects on growth and yield parameters in plantain-banana hybrids (see Section 2h).Short-periOcl Individual non-degree training was also given to technicians froin Cameroon, Ghana, Nigeria and Rwanda in 1992 by IITA's Onne staff. Training guides In cooperation with the IITA's Training Materials Unit, PBIP scientists are publishing Crop Production and Research Guides. Crop Production Guides contain information for crop production, training and research. Although the information is directed at an intermediate level, it can be adapted for communicating with farmers. Research guides are based on documents produced for training. Their objective is to provide information and guidance to researchers and technicians conducting research essential to agricultural development. The Research Guides can be used in both research and training. The following are the guides which were produced in 1992 and are currently in press: Swennen, R. & R. Ortiz. Morphology and growth of plantain. IITA Crop Production Guide Gauhl, F., C. PRsberg-GRuhl, D. Vuylsteke & R. Ortiz. Multilocational eval uation trial of UTA Mus a Hybrids. Protocol and guidelines for field management and data collection. UTA Research Guide ad. Visitors Among the many visitors during 1992, were: • His Excellency Mr Rufus Ada-George, Governor of Rivers State, Nigeria, who joined Dr Lukas Brader, our Director General, and allllTA's staff based at Onne to celebrate the silver jubilee of the institute in June 23, 1992 • Drs Randolph Baker (Cornell University, USA), Vittorio Delucchi (Swiss Federal Institute of Technology, Switzerland), Keiko Nakamura (Waseda University, Tokyo), Joseph Mukibi (National Agricultural Research Organization, Uganda), who are also current members of IITA's Board of Trustees • Ms Anjuba Oruambo, Hon. Commisioner of Agriculture and Natural Resources of Rivers State, Nigeria • Dr Edmond E. De Langhe, who came to Ibadan as Director General of INIBAP for the annuaIIlTA·INIBAP consultation meeting 102 • Drs David R. MacKenzie (USDA, USA) and Mandivamba Rukuni (University of Zimbabwe), and Ms. Kerri Wright Platais (CGIAR Secretariat), who visited us as INIBAP's external panel review members • Drs Rafael Ortiz-Quesada and Rafael Perez-Duverge, Fundacion Desarrollo Agropecuario, Dominican Republic • Dr C. C. Anojulu, FACU, Nigeria and other Nigerian cooperators of MET-2; • Dr Maria Ayodele and Mr. G. Adejara of the Nigerian Plant Quarantine Service based at Ibadan • Dr Franklin Rosales, FHIA Breeder, Honduras • Drs Bruno Cammue and Rony Swennen, Catholic University of Leuven, Belgium • other local authorities,University Professors, NARS researchers, ADP and private extension agents, journalists, and farmers of Rivers and neighbouring States in southeastern Nigeria 8e. International Travel PBIP staff visited Belgium, Benin Republic, Burundi, Cameroon, Costa Rica, Cote d'ivoire, France, Ghana, Kenya, several Nigerian states, Peru, Republic of China (Taiwan), Uganda, UK, USA and ZanzibarlTanzania to participate in consultation meetings, present oral papers and posters in invited seminars, workshops and conferences, and deliver and evaluate breeding materials being tested in the METs (Section 2m) and AMYTs (Section 2p). The visits to NARS in sub-Saharan Africa were very informative and provided means to develop a more targeted breeding program, especially for East Africa. 103 REFERENCES CITED IN THE TEXT 1. Pascua, O. C., M. C. Sabomido & N. D. Beltran. 1984. Philippine banana cultivars: their morphological and agronomic characters. IBPGRISEAN Newsletter 8(4):10-20 2. Vuylsteke, D., R. Swennen & R.. Ortiz. 1993. Registration of 14 improved tropical Musa plantain hybrids with black sigatoka resistance. HortScience, Vol. 28, (in press) 3 IBPGR, 1984. Revised banana descriptors. Rome, Italy. Intemational Board for Plant Genetic Resources 4. Simmonds, N. W. 1953. Segregations in some diploid bananas. J. Genetics. 51: 458-469 5. Fullerton, R. A. & T. L. Olsen. 1991. Pathogenic variability in Mycosphaerella fijiensis Morele!. In: Valmayor, R. V., B. E. Umali & C. P. Bejosano (eds.) Banana diseases in Asia and the Pacific: proceedings of a technical meeting on diseases affecting banana and plantain in Asia and the Pacific, Brisbane, Australia, 15-18 April 1991. Intemational Network for the Improvement of Banana and Plantain, Montpellier, France. pp. 105-114 6. Vuylsteke, D. & R. Swennen. 1990. Somaclonal variation in African plantains. liT A Research 1: 4-10 104 TABLES ...... o ()1 Table 1. Research projects of PBIP in 1992 Plantain/Banana * Developing Musa breeding capability and strategy * Plantainlbanana breeding for durable host plant resistance to black sigatoka * Banana improvement for the mid-altitude * Biotechnology for Musa breeding Funding core core/BADC core/USAID core * Post harvest quality of plantains core * Genotype-by-cropping systems interaction core BADC .. Belgian Administration for Development Cooperation, Belgium INIBAP .. International Network for the Improvement of Banana and Plantain, France FHIA - Fundacion Hondurena de Investigacion Agricola, Honduras CRBP '" Centre Regional Bananiers et Plantains, Cameroon KU leuven • Katholieke Universiteit leuven, Belgium NIHORT .. National Horticultural Research Institute, Nigeria CRI - Crops Research Institute, Ghana IDEFOR _ Institut de' Forets, "COte d'ivoire IRAZ - Institut de Recherche Agronomique et Zootechnique, Burundi NARS - National Agricultural Research System UNBRP - Uganda National Banana Research Program USAID • US Agency for International Development USDAIARS • US Department of Agriculture/Agricultural Research Service Cooperating Location Duration Institutions USDAIARS, KU Leuven Nigeria 1992- INIBAP, FHIA, CRBP Nigeria 1987- KU Leuven, NIHORT, Cameroon CRI, IDEFOR, IRAZ, Uganda UNBRP & other NARS NARS Nigeria 1992- Uganda USDAIARS, KU Leuven Nigeria 1983- NARS Nigeria 1992- Nigeria 1980- ~ o 0> Table 2. Musa accessions maintained at the IITA gene-bank (duplicates kept in vitro at Onne Te lab are underlined). Ensete ventricosum, M. basjoo (Eumusa), M. jackeyi (BS 588) (Austranmusa), M.laterita (Rhodochlamys), M. peekelii (Australmusa). M. schizocarpa (?), unknown 'Mainz' (135) (?). Musa acuminata AA: 'A 366615 (?), 'Akondro Mainty', 'Bandjim', diploid 'Basilan' (11-403), 'Beram', 'Bie yeng Obi' (SF 247), 'Borneo', 'Calcutta 4" 'Colatino Ouro', unknown 'Oibit', 'Djum metek', 'Djum tau', 'Dole', 'Figue sucree' (BS 107), 'Po madu' (BS 507), 'Po mas' (BS 653), 'Gabah Gabah' (?), 'Galeo' (SF21 0), 'Guyod' (11-33 = BS 255). 'Gwanhou" (SF 227 or NBE 14), 'Heva' (SF 232 or NBE19), 'Inamibal', 'Long Tavoy', 'Mambee Thu' (SF223 = SF209 or Wudi yalu yaJua'), 'Madang', 'Manang', 'Monjet', 'Morong datu', M. acuminata 120, M. a hybrid (BS 302), M. a. subspp. malaccensis, M. acuminata subspp. malaccensis 'Selango", M. a. subspp. malaccensis holotype, M. a. subsDQ. ma1ac;censis 'Pahang'. M. acurninata subspp. rnicrocarpa 'Zebrina' (,Maioa'), 'Niyarma yik' (SF 248), No. 110, 'Pa' ('Musore'), 'Pa' (,Patthalong'). 'Pa' (,Songkla'), 'Paka', 'Palang', 'Pamoti-on', 'Pisang Bernn', 'Po buntar, "P. gigi buaya', 'Gu nin chio', 'Po jari buaya (11-320)', 'P. jari buaya (111-33)', 'P.liNn', 'P. madu' (BS 258 or BS 276), 'Po mas aye", 'Po muHk', 'Po tongat', 'Palen berry', 'Pitu', 'Pu-te-Ia-bum', 'Pu-te- wey', 'Racadag', 'Saing Todloh' (?), SF 215 (NBA14) ,SF 265 (NBB11), SH-3362 (FHIA il1l>roved banana). 'Siang Hil', SN2, 'Thong Dok mak', 'Tjau lagada', 'lopata' (?), 'Toowoolee' (SF237), 'Tuu Gja', 'Truncata', 'Truncata (BS 393), 'Undu Jamau', 'Uwati', 'Waigu' (?), Wariam (?), Wh-O-Gu' (SF21 0 or NBA10), 'Zebrina' (G.F.), 163?? M. a subsp. banskii: (112), (237), (BS 467), (BS 606), (BS 607), (BS 619), (BS 620), (BS 621), (BS 623), (BS 624), Higa 111-4 (BS 428), Higa 111-97 (BS 464). Hybrids: (BS 622 = hybrid with M. schizocarpa), Higa 111-4 (BS 378 = unknown male parent). AAA Dessert bananas Cavendish: 'Dwarf Cavendish', 'Giant Cavendish', 'lacatan' (BS 176), 'Po masak hijau', 'Poyo', 'Robusta', '~'. Gros Michel: 'Gros Michel', 'Highgate'. Others: 'Biu Ketip', 'Green Red', 'Hungtu', 'Jawaka', 'lai', 'leite', 'Mambu' (111-181), 'Marauw', 'Muga', 'Mun', 'Ouro Mer, 'Padri', 'Po nangka', 'Red', 'Vangambi ~' ("Klue Hom Bao). Cooking and beer bananas(*) of East Africa AAA: 'Bakurura', 'Chibulangombe"', 'GashuHe*', Guineo' 'Igihuni', 'Igisahira Gizanzwe', '!Qi.t§Ui*', 'Ikigeregere*', 'Ikimaga', 'Imbogo', 'Indemera y'lmbihire*', 'Ingarara', 'Ingarama', 'Ingororooka*', 'Ingumba y' Inyarrunyo', 'Inkira*', 'I mama', 'Intariho', 'Intokatoke*', 'Inyonya', 'Inzirabahima', 'Isha*', 'Kagera', 'Kibungo 1', 'Kitawira', 'Makara*', 'Mbirabire', 'Mbwazirurne', 'Nakjtengwa', 'Nshika"', 'Nsira', 'Nyamwihogora', 'Rugondo'. AS: Kamaramasenge (dessert banana). Kisubi*. AAB starchy bananas 'Come plantain'('Maia Maon'), 'Klue roi wee', 'laknau' (AVP-66), 'laknau' (11-22), 'Maqueno', 'Maritu' ('Sao Dorringos' 0 'Resplando"),'Mbeta l' (Mysore), 'P. kelar, 'P. raja', 'Porne', 'Popou/ou', 'Rajapuri India', 'Silk', 'Thong Det'. continued over page Table 2 continued AAB plantains (reference collection for West & Central Africa). Grouped according to inflorescence type. French: 'AApakpals', 'Amml' 'wm Onniaba', '&>em pa', 'Bind Immosendp' 'Bise Egome 2' (unstable), 'Bobby Tannap', 'Bungaoisan' (from the Philippines), 'Cantebalon', 'Congo 2', 'Dwarf Frencb' (286), 'W, 'Egiooa', 'French Royge', 'Gabon 2', 'Gabon 4', 'Kar Ngoy', 'Kelong Mekintu', 'Kwa', 'lifongo Uko', 'Madre del Platana", 'Maiden Plamain' (AVP-67), 'Mbi Egome 1', 'Mbi Eoome 3', 'Moungeli', 'Moutoka 1', '~', 'Mu\oloy', 'Muracho', '~', '~', 'Ndjngo Liko', 'Niabang', 'Njock Kan', 'Nselouka', 'Ntanga 2', 'Ntanoa 3', 'Ntanga 4', 'Ntanga 5', 'Ntanga 6', '~', 'Nyornbe 2', 'Obino I' fUl', '0. Ntanga 1', '0. Ntanoa 2', '0. Ntaoga green !TlJtant', 'Okele', 'Ql!ang', 'Oyoko Akpan', 'Plantain No.2', 'Plantain No. 17', 'Pumle Plantain', '~ plantain', 'Red Plantain Hembra', 'Red Vade', 'Rose d' Ekona', 'Rouge de loum', 'Walunau 8', Walyngy 15', 'Wine Plantain', 'Zye Ekon', 75-19S, a5..Qa. French Horn:'aataar, '.Qi.bx.2' (off-type), 'Mbang Olson'. False Horn: 'Abomieou', 'Agbagba', 'Apantu', 'AtaJi Kjogo', 'Big Ebanga', 'Bo-Ahiu Abue', 'Come 4', 'Currare', 'Currare enano', 'Pidiede', 'pwarf False Hmn', 'Eba Oboikpa', 'Eberedia Ukom', 'Essang', '~', 'Mbjrioyoog', 'Mbjdnyoog green mutant', 'Mbouroykou 3', 'Mimi Abue', 'Monganga', 'M21Q Ebaoga', 'Moutouka 2', 'Ngok Egorne', 'Njangafe/o', 'Qbubjt Ukom', 'Okoyo Ukom', 'Odshele', 'Red Ukom', 'UkQm',1.6.22. True Horn: 'Asamiensa', 'Balsa', 'paluyao', 'Elar Icon', '3 Hand Planty', 'Horse Plantain', 'Ihjtjsjm', 'Lysoka', 'Mbomo' (= 'Nothing but Red'), 'f&1ml' (='Nothing but green'),'Nothing but Green', 'Nothing but Red', 'Osoaboaso', 'Pisang Lang' & 'Pisang tandok' (from Malaysia),'Ubok fba'. Awaiting evaluation: 'Banana Serpent', 'Ngougou', 'Red madre', Tsambuny'. o Latin-American introductions: 'Oominico Harton', 'Dominico Harton Roja', 'Dominico Macho', 'Dominico Roja', 'Dominico 500', 'Harton Birracimo', ..... 'Harton Maqueno' ('Ngok Egome'), 'Harton Tigre', 'Platano Harton'. ABB Cooking bananas 'Blue Torres strait I', 'Bluggoe' (,Cachaco', Esperrno' or 'Kivuw'*), 'Cardaba', 'Fougamou', 'Foulah 4', 'fce-cream', 'Kinkala', 'Maduranga', 'Monthan' ('Saba'), 'Nzizi', 'Pelipita' (or 'Pelipita l' or 'PeUpita Majoncho'), 'P. awak' ('Kayinja'*), 'Sabra', 'Simili Radjah'. Natural tetraploids MAB: 'Oura da mata', 'Ngem', 'Platina'. AABB: 'Kluai Tiparof. Artificial tetraploid hybrids IC2 (= 'Gros Micher. x AA, Trinidad), EHIA-2 (='Willams' x AA, Honduras), 'Goldfinger' or EHIA-1 (= 'Prata' x AA, Honduras). EHIA-3 (tertiary tetraploid hybrid, Honduras). M. balbisiana 'Butohan 1', 'Butohan 2', 'Eli kehel', M. balbisiana (10852), M. babisiana (11-63), M. balbisiana du Cameroon, M. balbisiana Honduras, M. balbisiana los Banos, M. balbisiana MontpelUer, M. balbisiana Singapur, M. balbisiana Tanj'. ? Indicates doubts .about taxonomic assignment. () indicates synonymous name or identification in other countries/collections. SF accessions were collected in Papua New Guinea. Table 3. Levels of black sigatoka (B8) resistance in the IITA Musa gene-bank. ReactionS HR PR LS Description Symptoms restricted to a specific leaf area because disease development stops (hypersensitivity?). B8 does not reach final stages. Disease developments slow down but can reach final stages; however, leaf spots do not coalesce (horizontal & durable resistance?) complete disease development but slower than in susceptible cultivars. Cultivars 'Calcutta 4', 'P. madu', 'P. !ilin', 'Tuu Gia.,'Truncata', 'Truncata' (B8 393) & 'Yangambi Km. 5'. 'Fougamou', 'Foulah 4', 'Heva', 'IC2', 'Long Tavoy', M. acuminata 120 & hybrid (B8382), M. a. subspp. malaccensis & holotype, M. balbisiana (11-63 & 10852), 'Pahang', 'Po jari buaya' (11-320 & 111-33), SF 265. 'Cardaba', 'Igisahira Gizanzwe', 'Igitsiri', 'Intokatoke', 'Nyawihogora', 'Nzizi', 'Ouro mel', 'Pelipita', 'P. mas", 'P. mas ayer" & 'Pitu'. BHR = highly resistant, PR=partially resistant, L8 = less susceptible. 108 ..... 0 co Table 4. Phenotypic correlations between bunch and vegetative traits in the AAB West African plantain germplasm. The significance level, (x, for false rejection of null hypothesis: p =0, is indicated in brackets. --.-----~--------------------------------------------------------------------------------------------------.-------------------------------------------------~------------ Trait 2 3 4 5 6 7 8 9 1. Bunch weight 0.556 0.887 -0.316 0.473 0.782 0.801 0.752 0.086 (0.004) (0.000) (0.123) (0.017) (0.000) (0.000) (0.000) (0.690) 2. No. of hands 0.490 -0.398 0.380 0.336 0.450 0.403 0.096 (0.013) (0.048) (0.060) (0.100) (0.027) (0.045) (0.656) 3. No. of fingers -0.652 0.363 0.578 0.603 0.520 0.103 (0.000) (0.074) (0.002) (0.002) (0.008) (0.673) 4. Average fruit weight 0.141 0.119 0.087 0.166 0.070 (0.501) (0.571) (0.686) (0.429) (0.746) 5. Pseudostem height 0.802 0.802 0.737 0.745 (0.000) (0.000) (0.000) (0.000) 6. Pseudostem circumference 0.918 0.912 0.357 (0.000) (0.000) (0.086) 7. Time of flowering 0.943 0.368 (0.000) (0.084) 8. Number of leaves 0.452 (0.026) 9. Length/width leaf Ratio -------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Probability values> 0.05 indicate a non-significant correlation coefficient. Table 5. Eigen vector values for principal components using bunch and vegetative traits in AAB plantains. Trait PRIN1 PRIN2 Bunch weight 0.087 0.123 No. of hands 0.015 0.016 No. of fingers 0.352 0.850 Average fruit weight -0.000 -0.002 Pseudostem height 0.537 -0.510 Pseudostem circumference 0.145 0.003 Time of Flowering 0.746 -0.049 Number of leaves 0.057 0.005 Length/width leaf ratio 0.002 -0.002 ----.----------------------------------------------------------------------------------------------------- % of total variation explained 79.409 14.239 110 Table 6. Bunch weight (kg plant-1) and somaclonal variation (%) of six plantain cultivars propagated by in vitro shoot-tip culture Plantain Bunch Weight1 (kg) Somaclonal cultivar Type2 mean var sd CV(%) variation Ntanga 2 FG 20.99 12.74 3.57 17.00 0.5 Bobby Tannap FM 14.01 8.50 2.91 20.77 0.0 Obino l'Ewai FM 12.48 10.64 3.26 26.10 1.8 Big Ebanga FHG 11.90 3.76 1.94 16.30 35.0 Agbagba FHM 9.07 4.20 2.05 22.60 5.7 Ubok Iba H 7.80 6.38 2.53 32.40 12.5 1 Data from 120 random true-to-type plants. 2 FG = French giant, FM = French medium, FHG = false Horn giant, FHM = false Hom medium, H = Horn. Table 7. Optimum plot sizes for bunch weight evaluation and soil heterogeneity index, as measured by bunch weight variation, of six plantain cultivars using in vitro propagated plants. Cultivar No. Plants Plot-1 Soil Heterogeneity VI CV (b) Ntanga 2 30 12 1.45 Bobby Tannap 30 8 0.24 Obino l' EWai 20 12 0.26 Big Ebanga 24 15 1.12 Agbagba 30 15 1.14 Ubok Iba 12 20 0.94 a This was determined through the method of maximum curvature by plotting either variance index (VI) or coefficient of variation (CV) (Y axis) vs. number of plants per plot (X axis) (see Fig. 6) Table 8. Number of replications required to detect a significant (P < 0.05) mean difference (expressed in kg plant-1) for bunch weight between 2 plantain cultivars with 80% statistical confidence Mean difference (kg plant-1) 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 Recommended number of replications (15 plants plot-1) 111 192 48 21 12 8 5 4 3 2 2 2 1 ~ .... I\) Table 9. Bunches pollinated, pollination success (100 x bunches with seeds/bunches pOllinated, total seed production and average seed set (seeds per bunch) in different types of mating during 1992 at Onne. Type of cross Jan Feb Mar April May Jun Jul Aug Sep Oct Nov Dec Total Parents with high seed seta I. Germplasm enhancement of M. acuminata gene pool bunches poUinated 82 103 74 84 72 82 45 66 48 52 60 51 819 pollination success (%) 15 10 7 16 25 21 29 40 27 19 35 29 20 total seed production 349 411 68 124 600 548 1184 1206 1286 average seed set 4.3 4.0 0.9 1.5 7.7 6.7 26.3 18.3 26.8 3147 1951 60.5 32.5 596 11470 11.7 14.0 II. Female fertility in East African Highland bananas (to complete screening of previous years) bunches pollinated 11 8 1 7 1 1 1 3 8 4 4 6 11 6 pollination success (%) 18 12 0 0 0 0 33 25 25 17 0 33 total seed production 3 2 0 0 0 0 203 25 62 49 0 133 average seed set 0.3 0.3 0.0 0.0 0.0 0.0 25.4 6.3 15.5 8.2 0 22.2 III. Breeding in the MB gene pool a. triploid x diploid crosses bunches pollinated 329 324 200 131 147 209 188 164 179 184 165 195 pollination success (%) 30 10 20 11 11 19 26 21 32 38 39 42 total seed production 1 025 466 466 48 73 216 586 293 607 782 559 625 average seed set 3.1 1.4 2.3 0.4 0.5 1.0 3.1 1.8 3.4 4.3 3.4 3.2 b. tetraploid x tetraploid crosses bunches pollinated 24 45 24 24 3 2 2 poUination success (%) 8 18 12 17 67 0 0 total seed production 32 45 1 65 16 0 0 average seed set 1.3 1.0 0.0 2.7 5.3 0.0 0.0 90 11 472 5.2 2415 25 5746 2.4 124 15 159 1.3 Biu-Ke-Tip, Calcutta 4, Dibit, Djum Metek, Galeo, Gu-nin- chio, Guyod, Klue Hom Boa, M. a hybrid, M. a. subssp. banski4 Pa (Pathalong). Padri, P. madu, P. mas ayer, Pitu, SH3362, Siang Hil, Toowoolee, Topala, Truncata, Tuu Gia. Kamaramasenge. Bobby Tannap, Bungaoisan, Didiede (French Reversion), Corne PlantainIMaia Maoli, Obino I' Ewai. TMPx 548-4, TMPx 597-4, TMPx 1187-8, TMPx 4479-1. Continued over page .... .... U> Table 9 continued Type of cross Jan Feb Mar April May Jun Jul Aug Sep Oct Nov Dec Total Parents with high seed set1 c. tetraploid x diploid crosses bunches pollinated 6 6 pollination success (%) 1 7 1 7 total seed production 66 158 average seed set 11.0 26.3 10 30 78 7.8 16 8 19 88 22 191 1.4 23.9 23 16 10 8 48 62 100 75 993 2282 1197 106 43.1 142.6 119.7 13.3 d. triploid x diploid crosses in False Hom gene pool bunches pollinated 1 1 1 pollination success (%) 0 0 0 total seed production 0 0 0 average seed set 0 0 0 e. enhancement of diploid plantain germplasm 1 o o o 2 - 100 4 2 bunches pollinated 25 32 39 35 41 35 30 25 50 pollination success (%) 8 6 20 9 34 29 40 48 36 total seed production 82 71 524 156 125 average seed set 3.3 2.2 13.4 4.4 3.0 502 1094 14.3 36.5 982 1423 39.3 28.5 IV. Improving ABB cooking and other starchy bananas bunches pollinated 6 6 14 9 1 0 1 6 pollination success (%) 17 33 29 22 60 75 total seed production 27 20 1 0 10 122 823 average seed set 4.5 3.3 0.7 1.1 12.2 51.4 aMore than 10 seeds per bunch at least once. 2 9 13 o 67 92 o 647 1038 0.0 71.9 79.8 10 18 50 67 164 1048 16.4 58.2 1 o o o 45 37 47 32 3037 1073 67.5 29.0 11 11 64 82 473 603 43.0 54.8 16 147 25 52 216 6527 13.5 44.4 7 28.5 4 0.6 38 432 16 28 248 9317 6.5 21.6 13 120 15 52 75 3848 5.8 32.1 TMPx 548-4, 548-9, 582-4, 1621-1,2637-49,2776-20, 2796-5,4479-1,4698-1, 4744-1,5511-2,6562-1, 6930-1, 7002-1. TMp2x 566-32, 1135-1, 1199-1 (3x!), 1297-3, 1448-1, 1489-3, 1518-4, 1549-5,1549-7,1586-2, 1637-1.1659-9,1668-7. 2085-1.2625-5.2625-20, 2829-62. 4231-2, 4250-1. 4281-2.4364-1,4400-3. 4400-6,4400-7,4400-8, 4600-12,4600-15,4800-1, 4885-1, 5020-1, 5233-1. 5837-1. 6048-1, 7579-1. 8223-1,9243-2,9007-4, 9593-1,9636-1,9722-1. Bluggoe, Cardaba. Ouro da Mata, Platina. Table 10. Seed set in crosses between AAA and AA bananas: 'Yangambi Km. 5' x 'Calcutta 4' (Km 5 x C4) and 'Yangambi Km. 5' x 'Pisang lilin' (Km 5 x PI) (1988-1989) Cross Jan Feb Mar Apr May Jun Jul Aug Sap Oct Nov Dec Year 1988 Km5xC4 1.0 0.7 0.0 0.3 3.3 4.3 12.5 33.9 17.0 19.7 11.5 6.9 Km5 x PI 0.0 0.0 0.5 1.1 1.3 0.1 1.4 1.5 0.8 Year 1989 Km5 x C4 7.0 2.4 1.0 0.0 1.7 11.0 20.6 29.3 18.0 10.3 - Km5x PI 0.1 0.1 0.0 0.2 0.2 0.0 0.0 2.1 4.5 1.5 1.8 - Table 11. Monthly overview of plantain-derived hybrid seeds handled by the Plantain/Banana Tissue Culture Laboratory (Onne station) in 1992 Number Number Number Seed Embryo Pollination of of of success germination month seeds embryos germinations rate(%) rate (%) January 731 287 26 3.5 9.1 February 1007 313 26 2.6 8.3 March 1037 427 72 6.9 16.9 April 203 120 9 4.4 7.5 May 431 209 23 5.3 11.0 June 1491 675 99 6.6 14.7 July 2028 813 100 4.9 12.3 August 2178 1227 279 12.8 22.7 September 1321 638 39 3.0 6.1 October 1708 736 119 7.0 16.2 November 1848 868 63 3.4 7.38 December 587 185 22 3.8 12.0 --------------------------------------------------------------------------------------------------- Total 14570 6498 877 6.0 13.5 8 Most of the seeds were treated late and might have contributed to low germination rate for this month 114 Table 12. Hier~rchical analyses of variance for stomata size and density between and within two plantain-banana hybrid populations Source of variation degrees of freedom % of sum of squares size density Mean squares size density Population(P) 1 6.4 0.6 0.107 ns 3025.4 ns Ploidy(X)/P 4 31.7 16.3 0.132 *** 19011.5 ** CloneslXlP 106 37.0 75.6 0.006 *** 3329.0 *** Error(within clones) 1008 24.9 7.5 0.0004 34.8 Coefficient of variation (%) 14.7 23.3 Table 13. Stomata size and density in two plantain-banana hybrid populations comprising clones with different ploidy levels Population Ploidy Number Stomata size Stomata density of hybrids mean range mean range mm2 mm2 mm-2 mm-2 'Obino \' Ewai' diploids 25 0.125 0.05-0.18 38.5 1-109 x 'Calcutta 4' triploids 2 0.146 0.13-0.16 3.1 1-5 tetraploids 18 0.185 0.14-0.23 14.5 1-36 'Bobby Tannap' diploids 59 0.125 0.08-0.17 24.8 1-67 x 'Calcutta 4' triploids 4 0.157 0.14-0.19 20.4 9-33 tetraploids 4 0.181 0.15-0.20 15.7 1-26 Table 14. Seedlings evaluated in the roguing nursery before field establishment at Onne in 1992 Cross Plantain or derivatives a x AA AAxAA ABB/AAAx AA Bunches pollinated Seedlings produced Plants to field 775 1400 1083 91 1166b 180 49 92 21 a Primary tetraploids (TMPx) or plantain derived diploids (TMp2x). /:Mainly 'Calcutta 4' x 'Valery', which was discarded because it was found that the seed produced in this cross was mainly the result of premature selfing of 'Calcutta 4'. Also several hundred seedlings of the cross 'Calcutta 4' x M. acuminata subspp. banksii, originally produced for genetic studies, are included in this total number 115 ..... ..... 0) Table 15. Total number of plants (p) evaluated in different number of replications (r) at different stages in the Plantain and Banana Improvement Program of "T A. Triala Criteria cx EmQirical Values Theoretical Values/Loc Real Values/Loc 3x= Plantain (# SE) P MSE SE #TMPx p/r r Locs tCX(dferror)b p/r r Total p Total area (m2) EET TMPx~3x 0.10(1) 5 6.0 1.1 100 5 1 1 1.645(400) 5 1 500 3000 PYT TMPx>3x 0.05(2) 8 4.6 0.8 30 4 2 1 1.960(180) 4 2 240 1440 MET TMPxs>4x 0.01 (2-3) 10 4.4 0.7 20 5 2 10 2.336(1600) 4-5 2 200 1200 AMYT TMPxs>4x 0.001(3-4) 20 2.9 0.4 5 5 4 10 3.050(800) 5-6 4 100 600 aEET = early evaluation trial, PYT = preliminary yield trial, MET =multilocational evaluation trial, AMYT: advanced Musa yield trial b dferror = (p-1) * R * # of TMPx * Locs . Table 16. Number of clones and seedlings, according to the breeding scheme, planted in the 1992 Early Evaluation trials (92-EET). Breeding scheme AABxAA Primary 4x (II 4x) x 2x II 4x x 114x 2x-plantain x 2x TMPx 1112-1 xTMp2x 1297-3 AAxAA AAAxAA ABBxAA Total No. of clones 154 248 28 265 1 1 147 13 15 881 117 No. of seedlings Ploidy outcome 179 270 34 304 16 160 16 16 1118 2x,3x or4x mainly 3x mainly 4x mainly 2x 2x,3x or4x mainly 2x 2x,3x or4x 2x,3x or4x ..... ... 00 Table 17. Analyses of variance, based on production cycle means, for growth and yield parameters in 4x progenies (TMPx) of 'Obino I' Ewai' x 'Calcutta 4' along with the fungicide treated (OL-FT) and non-treated (OL-NT) at Onne, Nigeria (1990-1992). -------------------------------------------------------------------------------------Source of degrees of Mean squares variation freedom H OF FFP O-H HTSh BW H/B FIB FW FV FL Fe -------------------------------------------------------------------------------------Production cycle (PC) 1 ns ns ns * * * ns ns ns ns ns ns Treatments (T) 4 ns ** ** * ns * ns ns * * * * OL-FT vs OL-NT 1 ns ns ns ns ns ns ns ns ns ns ns ns TMPx-BSRa vs BSSb..4x 1 ns ns ns ns ns ** ns ns ** ** * * TMPx-BSR vs OL-FT 1 ns ** *** * * * ns ns * * ns ns PCxT 4 ns * * ns ns ns ns ns ns ns ns NS Coefficient of variation (%) 6.3 8.6 4.5 5.1 16.9 17.1 5.3 8.2 16.3 16.0 6.1 7.3 aBSR = black sigatoka resistant. b BSS = black sigatoka susceptible . H = plant height at flowering (em), OF = days to flowering, FFP = fruit filling period (days from flowering to harvest), D-H = days to harvest, HTSh = height of tallest sucker at harvest (cm), BW = bunch weight (kg), H/B = number of hands per bunch, FIB = number of fingers (fruits) per bunch, FW = average fruit weight (g), FV = fruit volume (ml), FL = finger length (cm), FC = finger circumference. ns, *, **, *** indicate non significance, significant at the 0.05,0.01 and 0.001 levels, respectively. .... .... <0 Table 18. Growth and yield parameters in 4x progenies (TMPx) of 'Obi no I' Ewai' x 'Calcutta 4' along with the fungicide treated (OL-FT) and non-treated (OL-NT) at Onne, Nigeria (1990-1992). -------------------------------------------------------------------------------------Treatment! Black sigatoka Mean squares clone reaction H OF FFP O-H HTSh BW HIS FIB FW FV FL FC -------------------------------------------------------------------------------------OL-NT susceptible 347 342 c 79 c 422 c 116 6.7 b 6.2 70 82 b 82 b 15.4 a 10.2 a OL-FT 340 304 b 87 b 390 b 152 10.5 ab 6.4 76 118 ab 118 ab 18.0 a 11.8 a TMPx 548-4 partially resistant 324 208 a 124 a 330 a 252 14.4 ab 6.0 88 152 ab 153 ab 18.0 a 12.6 a TMPx 548-9 partially resistant 334 221 a 118 a 340 a 206 17.6 a 6.0 87 182 a 182 a 18.9 a 13.4 a TMPx 597-4 susceptible 281 192 a 119 a 310 a 178 7.0 b 6.0 81 84 b 84 b 14.4 a 9.8 a -------------------------------------------------------------------------------------H = plant height at flowering (cm), OF = days to flowering, FFP = fruit filling period (days from flowering to harvest), O-H = days to harvest, HTSh = height of tallest sucker at harvest (cm), BW = bunch weight (kg), HIS = number of hands per bunch, FIB = number of fingers (fruits) per bunch, FW = average fruit weight (g), FV = fruit volume (ml), FL = finger length (cm), FC = finger circumference . Mean columns without letters indicated that the source of variation treatment was not significant in the ANOVA (Table 17) . Means followed by the same letter were not statistically different according to Student-Neuman-Keul's test at the 5%. Mean square of the production cycle x treatment was used for the estimation of the least significant difference. .... N o Table 19. Preliminary evaluationa of growth and yield parameters in 2x I 3x and 4x progenies of 'Bobby Tannap' x 'Calcutta 4' (BT x C4) and 'Obi no I' Ewaj' x 'Calcutta 4 (OL x C4). (Onne, Nigeria 1991-1992) .. -------------------------------------------------------------------------------------Cross 2n N H HTSf HTSh FFP BW H/B FIB FW FL FC -------------------------------------------------------------------------------------BTxC4 2x 52 243±6 155± 7 238± 8 134± 6 3.3± 0.3 6.4 ± 0.2 92 ± 4 36±3 10.6 ± 0.4 7.2 ± 0.2 3x 4 283 ± 21 167 ± 24 277 ± 28 119±22 5.0± 1.1 7.0 ± 0.7 104 ± 15 51 ± 11 11.5 ± 1.5 9.7 ± 0.8 4x 6 258 ± 17 120 ± 20 205± 23 125 ± 18 6.7 ± 0.9 5.9± 0.6 75± 12 83± 9 13.5 ± 1.2 11.0 ± 0.7 F-test between ploidies ns ns ns ns it* ns ns *** ns *** OLx C4 2x 19 252±9 139 ± 12 229 ± 13 132± 4 3.2 ± 0.7 6.3 ± 0.3 87 ± 7 38± 6 11.6 ± 0.6 7.4 ± 0.3 3x 1 293 285 308 114 3.0 5 78 38 12.6 8.8 4x 14 338± 11 146 ± 13 238± 15 125± 5 13.8 ± 0.8 6.6± 0.3 96±8 144± 7 18.6 ± 0.6 12.1 ± 0.4 F-test between ploidies ** *b ns ns *** ns ns *** *** *** arv1ean ± standard errors. bBut not significant contrast between 2x and 4x ploidy levels . 2n = ploidy level; N = sample size, H = plant height at flowering (cm), HTSf = height of tallest sucker at flowering of plant crop (cm), HTSh = height of tallest sucker at harvest (cm), BW = bunch weight (kg), FFP = fruit filling period (days from flowering to harvest), H/B = number of hands per bunch, FIB = number of fingers (fruits) per bunch, FW = average fruit weight (g), FL = finger length (cm), FC = finger circumference. ns, *, **, *** indicate non significance, significant at the 0.05,0.01 and 0.001 levels, respectively. Table 20. Black sigatoka resistant selected tetraploid and triploid plantain hybrids and current status of evaluation. Cross # Female Male Ploidy CMV Most advanced IITA (TMPx) parent parent level indexB evaluationsb code 548-4*C 'Obino I' Ewai' 'Calcutta 4' 4x METs, AMYTs PITA-1 548-9*c 'Obino I' Ewaj' 'Calcutta 4' 4x METs, AMYTs PITA-2 582-4*c 'Bobby Tannap' 'Calcutta 4' 4x METs, AMYT-EA PITA-4 1112-1* 'French reversion' 'Calcutta 4' 3x MET1 1621-1*c 'Obi no I' Ewai' 'Calcutta 4' 4x not yet MET2 1658-4*C 'Obino I' Ewaj' 'P.Ulin' 4x METs PITA-7 2481* 'Obi no I' Ewai' TMb2x 366-8 4x MET2 2637-49*c 'Obino I' Ewai' 'Calcutta 4' 4x soon MET2 2776-20C 'Bobby Tannap' 'Calcutta 4' 4x PYT2 2796-5*C 'Bobby Tannap' 'P.lilin' 4x METs, AMYTs PITA-5 4479-1 C 'Bobby Tannap' 'Calcutta 4' 4x not yet PYT 4698-1 *C 'Bobby Tannap' 'Calcutta 4' 4x soon MET2, AMYT·EA PITA-6 4744-1 * 'Obino I' Ewai' 'Calcutta 4' 4x soon PYT 5511-2*c 'Obino I' Ewai' 'Calcutta 4' 4x MET2, AMYTs PITA-3 5706-1 *C 'Obino I' Ewaj' 'Calcutta 4' 4x soon PYT 5860-1 * 'Obino I' Ewai' 'Calcutta 4' 4x PYT 6930-1*c 'Obino I' Ewai' 'Calcutta 4' 4x MET2 7002-1 *C 'Obi no I' Ewai' 'Calcutta 4' 4x MET2, AMYTs PITA-8 * Deposited in INIBAP Transit Center at KUL, Belgium. alnformation provided by INIBAP. bMET =Multilocational Evaluation Trial, AMYT = Advanced Musa Yield Trial, AMYTEA = AMYT in East Africa, AMYTWA = AMYT in West Africa. cGermplasm registered in HortScience (see Section 2n) 121 Table 2'1. Selected parthenocarpic diploid derived plantain hybrids. TMp2x # Female parent 597 -2 'Obino I' Ewai' 1199-6 'Obino I' Ewai' 1297 -3 'French Reversion' 1448-1 'Obino I' Ewai' 1489-3 'Bobby Tannap' 1518-4 'Bobby Tannap' 1549-5 'Bobby Tannap' 1 549-7 'Obino I' Ewai' 1586-2 'Obino I' Ewai' 1605-1 'Ntanga l' 1657-4 ,'Bobby Tannap' 1659-13 'Bobby Tannap' 2348-6 'Bungaoisan' 2348-7 'Bungaoisan' 2625-5 'Bobby Tannap' 2625-20 'Bobby Tannap' 2829-62 'Bobby Tannap' 4281-2 'Bobby Tannap' 4400-8 'Bobby Tannap' 4600-12 'Bobby Tannap' 4600-15 'Bobby Tannap' 4729-22 'Bobby Tannap' 5837-1 'Bobby Tannap' Male Parent 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' 'Calcutta 4' AttributesB BSPR, LAD, PB, MF, FS(?). BSPR, RAD, SHB, MF, FF. BSPR, LAD, PB, MF, FF. BSPR, LAD, SHB, MF, FF. BSLS, LAD, PB, MF, FF. BSPR, LAD, SHB, MF, FF. BSPR, LAD, SHB, MF, FF. BSPR, LAD, SHB, MF, FF. BSLS, RAD, PB, MF, FF. BSPR, PB, MS. BSPR, LAD, SHB, MF, FS. BSLS, LAD, SHB, MF, FS. BSPR, RAD, PB, HBW. BSPR,RAD,SHB,HBW. BSPR, LAD, PB, FF. BSPR, RAD, SHB, MF, FF. BSPR, RAD, PB, MF, FF. BSLS, LAD, SHB, MF, FF. BSLS, LAD, SHB, MF, FF. BSPR, LAD, HB, MF, FF. BSPR, LAD, PB, MS, FF. BSS, HAD, SHB, FS. BSLS, LAD, PB, MF, FF. asSPR = black sigatoka partially resistant, BSLS = black sigatoka less susceptible, BSS = black sigatoka susceptible, LAD = low apical dominance, RAD = regulated suckering behaviour, HAD = high apical dominance, HB = horizontal bunch, SHB = subhorizontal bunch, PB = pendulous bunch, dw = dwarf, MF = male fertile, MS = male sterile, FF = female fertile, FS = female sterile, HBW = high bunch weight. 122 Table 22. Analyses of variance for growth and yield parameters combined over two locations of the Humid Forest Zone of Nigeria (Onne) and Cameroon (M'Balmayo) (1991- until March 31,1993). Source of degrees of Mean Squares variation freedom H YLS TNLf DF HTSh FFP BW H/B FIB FW FL FC ------------------------------------------------------------------------------------- Location (L) L-1 = 1 ns * ** ns ns ns * ns ns ** ** ** Reps/L (R-1)L = 2 ns ns *' * ns ns ns ns ns * ns ns Clonesa(C) C-1 = 15 *** *** *'** *** *** **- -** *** * .. * .. * .. .. *'* *** CxL (C-1)(1-1) = 15 *' .. *** ns **' ns * ** ns *' *' ns - Error (R-1)(C-1)L = 30 Total RCL-1= 63 Coefficient of variation (%) 4.8 11.0 5.8 6.7 12.5 4.3 12.8 8.3 10.6 13.0 7.4 4.9 a 'Pisang Lilin' and 'Agbagba' were missing at M'Balmayo due to somaclonal variation. Degrees of freedom decreased for C and CxL, error and total in all the traits evaluated at harvest because no data was collected for 'Calcutta 4' and 'Fougamou' at .- M'Balmayo. This last cultivar was not harvested yet when this report was written). ~ H = plant height at flowering (cm), BSR= black sigatoka resistance as measured by youngest leaf spotted at harvest, TNLf = total number of leaves at flowering of plant crop (cm), DF=days to flowering, HTSh = height of tallest sucker at harvest (cm), BW = bunch weight (kg), FFP = fruit filling period (days from flowering to harvest), H/B = number of hands per bunch, FIB = number of fingers (fruits) per bunch, FW = average fruit weight (g), FL = finger length (cm), FC = finger circumference. *, **, *** indicate non significance, significant at the 0.05,0.01 and 0.001 levels, respectively. Table 23. Clonal mean of growth and yield parameters evaluated in 18 genotypes at ')nne Station (1991-~' .(ch 1993) --------------------------------------------------------------- ---------- -----------Clone BSR H TNLf OF HTSh FFP BW H/6 FIB FW FL FC --------------------------------------------------------------------------------------TMPx548-4 PR 311c 37e-h 328e-g 314bc 118a-e 14.2bc 6c 86c-g ;49 cd 19b-d13de TMPx 548-9 PR 328 be 38 d-g 341 d-g 286 b-d 113 c-t 15.9 b 6c 88 e-f 161 cd 20 b 14 cd 597-4 S 248 de 33 h 2839 187 t9 110 ef 5.4 f9 6 e 67 h-j 71 fg 15 e-g 10 g TMPx582-4 PR 261d ·35gh 318fg 247d-f 115c-f 9.9 de 6c 74f-i 124de 16ef 130p TMPx 2796-5 PR 313 e 40 e-f 368 c-f 282 b-d 117 b-e 14.6 be 6 c 84 dog 168 cd 20 be 13 de TMPx 1112-1 PR 312 c 38 dog 405 a-c 258 c-e 111 et 15.5 b 7 b 99 b-d 147 cd 18 c-e 12 ef TMPx1658-4 LS 369 a 39d-g 382b-e 244d-f 112d-t 19.4a 8a l11ab 162cd 20b 13de 612-74 off type A 302 c 36 f-h 417 a-c 241 d-t 116 c-t 10.2 de 5 d 58 ij 163 cd 18 c-e 15 be 'Obino I' Ewai' S 370 a 44 a-c 412 a-c 174 9 99 gh 12.2 cd 7 b 78 e-h 147 cd 20 be 13 de 'Bobby Tannap' S 318 c 42 b-d 375 c-t 162 9 91 hi 9.7 de 7 b 87 e-t 99 et 16 de 12 et 'Valery' S 222 et 36 f-h 329 e-g 174 9 108 e-g 7.4 ef 7 b 102 be 59 fg 16 et 9 9 'Bluggoe' LS 357ab 41e-e 420a-c 322b 123a-c 14.0 be 6e 59ij 217b 19b-d16a 'Cardaba' PR 313e 40c-f 397a-d 268b-e 122a-d 14.2bc 6c 70g-j 184be 18e-e 16a ... 'Pelipita' PR 390 a 47 a 458 a 328 b 127 ab 15.4 b 8a 99 bod 143 cd 16 e 14 cd ~ 'Fougamou' PR 368 a 46 ab 423 a-c 414 a 128 a 11.7 cd 8a 118 a 83 efg 13 9 12ef 'Calcutta 4' HR 196 f 24 i 203 h 245 d-t 106 fg 0.3 h 7 b 92 c-e 2 h 6 h 4 h 'Agbagba' S 327 be 42 bod 440 ab 213 e-g 83 i 3.8 9 4 e 12 k 304 a 26 a 15 be 'Pisang lilin' A 199 f 23 i 221 h 207 e-g 118 a-e 3.7 9 5 d 56 j 50 9 14 f9 9 9 -------------------------------------------------------------------------------------C. V. (%) 11.2 7.6 9.8 13.2 19.5 10.1 29.9 12.7 16.4 25.2 9.8 8.5 ------------------------------------------------------------------------------------- aprobably an off-type after in vitro propagation. Means followed by the same letter were not statistically different at the 5 % level according to Duncan's Multiple Range Test. BSR= black sigatoka reaction: HR = highly resistant, R=resistant, PR=partially resistant, LS = less susceptible and S = susceptible. This was determined by comparing the youngest leaf spotted of each clone with respect to the susceptible checks and the hybrids with their triploid parents, H = plant height at flowering (em), TNLf = total number of leaves at flowering of plant crop (em), OF=days to flowering. HTSh = height of tallest sucker at harvest (cm). BW:;:: bunch weight (kg), cFP = fruit filling period (days from flowering to harvest), HlB = number of hands per bunch. FIB = number of fingers (fruits) pt~r bunch, FW = average fruit weight (g), FL = finger length (cm), FC = finger circumference. Table 24. Clonal mean of growth and yield parameters evaluated in 17 genotypes at M'Balmayo Station (1991-March 1993) -------------------------------------------------------------------------------------Clone BSR H TNLf DF HTSh FFP BW H/B FIB PN FL FC -------------------------------------------------------------------------------------TMPx548-4 PR 319de 27a-c 317 cd 251c-e 112b-e 13.8b 6d 96 cd 139abc 16a 13b TMPx 548-9 PR 342 bod 28 ab 322 cd 230 d-t 109 c-e 14.5 ab 6 d 100 c 144 ab 16 a 13 b 597-4 S 294 tg 27 abc 338 c 188 t-h 108 de 7.0 gh 7 c 92 c-t 62 h 13 c 10 d TMPx582-4 PR 293t9 22 abc 284d 259a-e 114b-d 11.0cd 6d 90c-t 109d-t 16a 12c TMPx 2796-5 PR 340 bod 26 abc 323 cd 306 a 114 bod 12.6 bc 6 d 88 c-t 136 a-c 16 a 13 b TMPx1112-1 PR 259h 29ab 352c 213e-g 104e 9.8 de 6d 85c-t 106ef 14b 12c TMPx1658-4 LS 359ab 27 abc 322 cd 241c-e 112b-e 16.4 a 7c 116b 148a 16a 13b 612-74 ofttype R 288 tg 26 abc 323 cd 256 b-e 125 a 6.2 h 5 e 60 9 93 tg 12 c 13 b 'Obi no I' Ewai' S 347 a-c 35 a 406 b 153 h 86 f 9.8 de 7 c 76 t 113 de 15 a 12 c 'Bobby Tannap' S 278 gh 34 a 412 b 163 h 87 f 8.4 e-g 7 c 82 d-f 78 gh 13 c 10 d 'Valery' S 216i 25 abc 295d 176fg 119ab 9.6 de 11a 137a 71h 13c ge 'Bluggoe' S 339c-e 29 abc 353c 266a-d 118ab 7.4f-h 5e 49g 126b-d 13c 14a 'Cardaba' PR 311 et 27 abc 336 c 283 a-c 117 bc 9.2 d-f 6 c 78 et 122 c-e 13 c 13 b 'Pelipita' lS 360 ab 31 ab 397 b 302 ab 112 b-e 11.2 cd 8 b 94 c-e 110 d-t 13 c 12 c ~ 'Fougamou' LS 366 ab 33 ab 462 a not harvested yet 0'1 'Calcutta 4' HR 206 j 16 c 239 e data was not collected 'Agbagba' off-type 432 ab 'Pisang liIin' off-type -------------------------------------------------------------------------------------C. V. (%) 8.6 5.3 5.4 6.4 14.4 6.6 21.7 12.6 14.1 9.2 6.6 19.5 MeansfOltowedbyttle-sarrieletter-werenotstatisticatiYdlffererltatthe5%lev9laccordingtoi5uncan75-MultiPieRangeTeSt~ BSR= black sigatoka reaction: HR = highly resistant (no leaf spots). R=resistant, PR=partially resistant, LS = less susceptible and S = susceptible. This was determined by comparing the youngest leaf spotted of each clone with respect to the susceptible checks and the hybrids with their triploid parents, H = plant height at flowering (em), TNLf = total number of leaves at flowering of plant crop (em), DF=days to flowering, HTSh = height of tallest sucker at harvest (em), BW = bunch weight (kg), FFP = fruit filling period (days from flowering to harvest), H/B = number of hands per bunch, FIB = number of fingers (fruits) per bunch, FW = average fruit weight (g), FL = finger length (em), FC = finger circumference. Table 25. Entries in MET-2 distributed for evaluation in 3 IITA stations and 15 NARS of sub-Saharan Africa, tropical America and Australia Tetraploid hybrids with black sigatoka resistance: TMPx 548-4, TMPx 548-9, TMPx 582-4, TMPx 1621-1, TMPx 1658-4, TMPx 2481, TMPx 2637-49, TMPx 2796-5, TMPx 4698-1, TMPx 5511-2, TMPx 6930-1 and TMBx 612-74. Triploid females parents susceptible to black sigatoka: 'Obino l'Ewai', 'Bobby Tannap' and 'Bluggoe'. Male diploid parents with black sigatoka resistance 'Calcutta 4' and 'Pisang Jilin' Reference cultivars: 'Agbagba' (susceptible to sigatoka), 'Cardaba' (less susceptible) and 'Valery' (susceptible). Table 26. Locations and officers of the MET-2 Location Onne M'Salrnayo Namulonge Nyombe Assin-Fosu Obrikom Bori Ibadan Umudike Ibadan Akure Oquo Ibono Calabar Ugwuoba-Oji River Benin Santo Domingo Villa Clara, S. Domingo South Johnstone Country Nigeria Cameroon Uganda Cameroon Ghana Nigeria Nigeria Nigeria Nigeria Nigeria Nigeria Nigeria Nigeria Nigeria Nigeria Domican Republic Cuba Australia Institution In-situ coordinator UTA PBIP UTA PBIP UTA PBIP/PHMD CRSP Mr. C. Jenny CRI Dr. O. B. Hemeng NAOC Dr. S. Akele Shell Mr. A. Lawrence NIHORT" Dr. B. A. Adelaja NRCRI* Dr. O. Arene PAMOL * Dr. C. Anojulu ADP* Dr. Kayode BSDP* Mr. B. Umo-Etuk A&RD* Mr. A. N. Agbor TIS/FDA* Mr. A. A. Okoho UNIBEN*a Dr. F. E. O. Ikiediugwu Fund. Des. Agr. Dr. A. Rivera de Castillo INIVIT Dr. A. Rodriguez-Nodals CDPI Dr. J. Daniells * With financial support of the Federal Government of Nigeria,through the Interministerial Committee on the ContrOl of Black Sigatoka Disease, to carry out this MET·2 trial with Nigerian NARS. a to be planted in 1993· mid-rainy season. 126 Table 27. Agronomic evaluation of 14 selected TMPx clones at Onne, Nigeria, from 1990 to 1992a. Youngest leaf Plant HTSc at Fruit Bunch spotted at height harvest filling time weight Number Number Genotype Parentsb flowering (cm) (cm) (days) (kg) of hands of fruits TMPx 548-4 OLxC4 10.0** 330** 280** 126** 16.7*" 6.5** 103** TMPx 548-9 OLxC4 11.0** 290** 310** 131 ** 16.5** 6.0*" 83** TMPx 582-4 BTxC4 11.0** 300** 270** 135** 14.3ns 7.5ns 124** TMPx 1621-1 OLxC4 7.5ns 340** 280** 120** 13.8** 6.5** 90ns TMPx 1658-4 OLxPI 9.3** 320** 185ns 134** 21.5** 7.5ns 123** TMPx 2637-49 OLxC4 11.6** 360* 225** 127** 16.6** 6.4** 95ns TMPx 2796-5 BTxPI 10.0** 335ns 205** 123** 21.3** 6.5** 101 ns TMPx 4479-1 BTxC4 9.0** 295** 290** 114** 13.2ns 6.0** 88** TMPx 4698-1 OLxC4 10.5** 345** 275** 130** 20.0** ** 124** ~ 8.5 I\) -...J TMPx 4744-1 OLxC4 10.0** 360* 235** 125** 11.5* 6.0** 89ns TMPx 5511-2 OLxC4 10.0** 345** 165* 118** 17.8** 6.0** 87ns TMPx 5706-1 OLxC4 7.5ns 340** 250** 133** 13.6** 6.8ns 96** TMPx 6930-1 OLxC4 10.8** 330** 240** 138** 17.5** 7.5ns 121 ** TMPx 7002-1 OLxC4 13.0** 340** 270** 120** 17.5** 6.0** 115** 'Bobby Tannap' d 7.0 ± 0.29 340 ±2 171 ± 3 92 ±O 14.0 ± 0.2 7.8 ± 0.1 100 ± 1 'Obino I' Ewai' d 7.0 ±0.2 370±2 188 ± 5 92 ±O 12.4 ± 0.2 7.2 ± 0.1 91 ± 1 a This consists of recording the number of the youngest leaf, counting down from the first (top) unfurled or cigar leaf, that showed mature spots. bel = 'Obino l'Ewai'; BT = 'Bobby Tannap'; C4 = 'Calcutta 4'; PI = 'Pisang lilin'. C Height of tallest sucker. d Fungicide-treated plots. e Mean ± standard error (n=200). ns, * I ** Nonsignificant or significant at P=O.OS or 0.01. respectively, for multiple comparison within columns of TMPx genotypes with the respective plantain parent. Critical values of Student's "t" distribution were adjusted by Sidak's multiplicative inequality based on a'=1-(1-a) 1/c, in which a' is the level of significance required for statistical testing and c is the number of comparisons between the hybrids and the plantain parents (11 and 3 for Ol and BT, respectively) . Table 28. Clonal mean and standard deviations for growth and yield parameters evaluated in the IMTP-1 at Onne Station (December 1991-March 1993). -------------------------------------------------------------------------------------Clone N BSR H G TNLf Li OF HTSh RGi FFP BW YP RYP -------------------------------------------------------------------------------------Tetraploid Hybrids: FHIA-1 (dessert) 8 PR 202± 17 47±5 45±4 2.4±0.1 368±48 126±26 62% 124±4 15.1 ±2.1 18.7 18.7 FHIA-2(dessert) 9 PR 184±12 39±2 37±2 2.3±0.2 296±27 110±25 60% 117±5 11.7±1.817.2 15.5 FHIA-3(cooking) 8 PR 224± 16 59±3 44±3 2.6±0.1 382±44 74±42 33% 109±6 12.1 ± 1.8 15.0 15.0 FHIA-4(plantain) 3 S 298±4 50±5 48±6 2.8±0.1 430±83 0 0% 114 18.0 20.1 12.1 FHIA-5 (plantain) 8 S 280 ± 20 60 ± 9 50 ± 5 2.8 ± 0.3 426 ± 51 48 ± 67 17% 98 ± 8 11.7 ± 1.8 13.6 12.2 FHIA-6 (cooking) - S not evaluated because showed virus-like symptoms at flowering and INIBAP recommended its elimination. FHIA-7 (cooking) - S not evaluated because showed virus-like symptoms at flowering and INIBAP recommended its elimination. Reference cultivars: Iriploids cultivars 'Valery' (dessert) 10 S 175 ± 11 40 ± 3 32 ± 2 2.5 ± 0.1 368 ± 60 98 ± 10 56% 82 ± 10 7.2 ± 1.9 9.7 9.7 'Agbagba' (plantain) 4 S 246 ± 17 44 ± 4 47 ± 8 2.4 ± 0.2 417 ± 48 82 ± 16 33 % 60 ± 11 0.6 ± 0.2 0.8 0.8 N Diploid wild and cultivated bananas 0) 'Calcutta4'(wild) 10 HR 180±13 28±3 32±22.2±0.1 223±24 204±18 113% 107±23 0.6±0.2 1.1 1.1 'Niyarmayik' 10 S 233±11 38±2 43±4 2.4±0.2 330±62 48±28 21% 56±4 3.3±0.8 5.2 5.2 'Pahang'(wild) 3 PR 192±10 26±0 40±22.7±0.1 297±42 190±9 99% 114±29 0.6±0.3 0.9 0.9 'Pisang Berlin' 7 PR 174 ± 14 33 ± 3 41 ± 4 3.1 ± 0.2 323 ± 42 121 ± 27 70% 65 ± 3 5.2 ± 2.7 8.3 7.5 'Pisanglilin' 10 R 118±8 23±2 22±22.6±0.2 197±20 169±19 143% 122±10 3.4±1.0 6.5 6.5 Pisang mas' 1 S 220 31 56 3.8 430 not harvested SF-215 10 S 196±5 33±1 39±22.2±0.1 268±17 51±6 26% 46±2 3.4±0.5 6.6 6.6 'I ani' (wild) S not flowered yet 'IuuGia" 10 HR 200± 19 31 ±2 40±3 2.6±0.2 337±52 182± 13 91% 107±8 4.0±0.8 5.5 5.5 N~num1;erofPiantsevaiuited,-BSR;-blackSigatokareaction:HR-~hi9hiYreSiStant~R~resiStarlt;-PR;-ParticiliyreSiStant,-CS-~- less susceptible and S = susceptible, H = plant height at flowering (cm), G = plant girth at flowering (cm) TNLf = total number of leaves at flowering of plant crop (cm), li= leaf index, ratio of leaf length/width, DF=days to flowering, HTSh = height of tallest sucker at harvest (cm), RGi= rate of growth for ratoon cycle -> [100 x HTSh/H], BW = bunch weight (kg), FFP = fruit filling period (days from flowering to harvest), YP= yield potential (MT ha-1) -> [BW x 365 x 1667/1000/(OF+FFP)], RYP= real yield potential (MT ha-1) -> % survival x YP. Table 29. Locations and officers of the AMYT-l Location West Africa: Irno-State Ibadan Abia State Abidjan Kade East Africa: Rugombo Kawanda Kizirnbani Ungoye (Lake Victoria shores) Country Nigeria Nigeria Nigeria Cote d'ivoire Ghana Burundi Uganda Zanzibar (Tanzania) Kenya Institution In-situ coordinator(s) ADP Mr. T. O. Echebiri NPQS Dr. M. Ayodele ADP Dr. E. Okoro IDEFOR Dr. A. N' Guessan Univ. of Ghana Dr. K. Afreh-Nuarnah IRAZ Dr. J. Kafurera UNBRP Dr. Kararnura Min. of Agric. Mr. MbereklMr. B. Aalders ICIPE Dr. S. Reddy Table 30. Hybrids and cultivars to be evaluated in AMYT-l IITAID Cross 10 West Africa East Africa Black sigatoka resistant hybrids: PITA 1 TMPx 548-4 x x PITA 2 TMPx 548-9 x x PITA 3 TMPx 5511-2 x x PITA 4 TMPx 582-4 x PITA 5 TMPx 2796-5 x x PITA 6 TMPx 4698-1 x PITA 8 TMPx 7002-1 x x BITA 1 TMBx 612-74 x x FHIA3 SH-3565 (FHIAB) x x Reference cultivars: 'Km 5' AAA banana, BSRb x x 'Agbagba' AAB plantain, Bsse x 'Obi no I ' Ewai' AAB plantain, BSS x x 'Bluggoe' ABB banana, BSS x 'Cardaba' ABB banana, BSS x 'Nakitengwa' AAA banana, BSS x .:selected Hybrid of Fundacion Hondurena de Investigacion Agricola. SH-3565 = {[,Cardaba' x BB] x SH-27 41} x AA b Black sigatoka resistant. C Black sigatoka susceptible 129 Table 31. Selected diploid banana hybrids (TMb2x) for further use in breeding in banana breeding TMb2x Female parent Male parent 5105-1 'Pisang Jilin' 'Calcutta 4' 5265-1 'Tjau Lagada' 'Calcutta 4' 6142-1 'Nyamwihogora' 'Long Tavoy' 7197-2 SH-3362 'Long Tavoy' 8075-7 SH·3362 'Calcutta 4' 8532·1 'Heva' 'Calcutta 4' 9128-3 'Tjau Lagada' 'Calcutta 4' 9719-7 'Manang' 'Calcutta 4' 9839-1 'Calcutta 4' 'Padri' 9869-1 'Long Tavoy' 'unknown Dibit' Table 32. Comparison (mean ± s.e.) of 'Agbagba' (fungicide-treated) and its French Reversion in vitro induced somaclonal variant Trait Agbagba French Reversion Comparison mean s.e. mean s.e. Plant height (cm) 358 2.0 333 5.0 •• Bunch weight (kg) 9 0.1 7 0.3 •• Hands bunch-1 8 0.1 6 0.1 .* Fingers bunch-1 28 0.4 61 0.1 •• Fruit weight 323 6.0 116 5.5 •• Fruit filling period 90 0.3 85 2.3 * ••• *. indicate that the mean comparison was significant at the 5 and 1 % levels respectively using t-tests Table 33. Percentage increase in mean leaf area conductance to water vapor from morning to the afternoon on young and older leaves at two different sampling periods after a dry spell and after rains (Ibadan, 1992) % increase in conductance Leaf surface Young leaves Older leaves Sampling period 1 (after dry spell): Upper -52.9 Lower 45.2 Sampling period 2 (after rains) Upper 47.2 Lower 58.0 -130.0 7.4 39.3 67.4 % increase in conductance = [(afternoon-morning)/afternoonJ 130 Table 34. Leaf stomata conductance (s.mm-1) of Musa genotypes measured in the afternoon on the lower surfaces of young leaves and in the morning on the upper surfaces of older leaves taken at two sampling periods and the percentage increase in conductance of lower leaf surface from morning to afternoon). Hybrid Lower surface Upper surface % increase in code/ ~oung leaves older I~§v~s conduct§nc~a Cultivar 1b 2 1 2 1 2 PITA-1 1.97 8.83 0.14 0.10 43.2 89.9 PITA-2 5.11 6.84 0.18 0.06 82.0 78.8 597-4 1.97 3.04 0.22 0.19 9.6 33.2 PITA-4 6.04 6.27 0.33 0.05 80.8 27.4 PITA-5 2.91 4.60 0.30 0.11 56.7 73.0 PITA-7 1.23 3.39 0.21 0.26 -56.1 1.2 PITA-9 2.53 2.43 0.19 0.18 40.7 49.4 612-74 offtype 5.48 6.06 0.41 0.15 -9.7 78.4 'Bluggoe' 2.33 2.51 0.38 0.15 39.1 55.8 'Pelipita' 3.41 4.61 0.22 0.20 57.5 79.3 'Fougamou' 1.90 2.46 0.24 0.25 -14.3 77.2 'Cardaba' 4.79 3.86 0.14 0.18 69.7 -60.8 'Valery' 1.74 10.83 0.20 0.12 56.3 58.6 'Calcutta 4' 2.84 3.98 0.19 0.06 78.2 20.9 'Agbagba' 3.12 5.73 0.20 0.17 80.8 28.8 'Bobby Tannap' 11.90 18.42 0.01 0.06 95.1 95.1 'Obino I' Ewai' 0.54 3.36 0.27 0.55 -26.4 5.7 LSD OeE (combined over sampling periods) 0.63 0.01 aMeans of the lower surface of young leaves. bSampling period 1 indicates measurements after a short dry spell and the sampling period 2 indicates measurements after rains. 131 Table 35. Phenotypic appearance of parents, F1 and F2 plantain-banana hybrids. Vegetative charaQter Ph~not~~ic a~~earanceElantain C4 F1 .E2 Leaf orientation Erect x x 13 Intermediate x 20 Drooping 2 Leaf color Green x x x 31 Green with red 4 Lamina shape Intermediate x x x 34 Broad 1 Suckering Inhibited x 4 Non-inhibited x x 30 Pseudostem blotches Slight 7 Moderate x x 24 Extensive x 4 Petiole margins Spreading x x 32 Enrolled x 3 Color petiole margins Red x x 35 Green x 0 Wax in pseudostem Absent x x 29 Present x 6 132 Table 36. Tests of goodness of fit and types of genes action for vegetative morphological descriptors. Vegetative character .... G::..>;e<.:..n=e........,a=ct=io=n'"'--____________ X2 l§.§1 Leaf orientation Leaf color Lamina shape Suckering Pseudostem blotches Petioles margin Wax in pseudostem erect ->complementary interaction of 3 dominant genes 0.40 drooping -> interaction of another 2 recessive alleles in intermediate genetic background (1 :15) 0.27 red -> modifier recessive suppresor (3:13) 1.23 intermediate ->duplicate loci with dominance of both (15:1) 0.63 inhibition -> one recessive allele (1 :3) 3.18 dominant epistasis -> 12 moderate: 3 slight: 1 extensive 1.72 spreading -> duplicate loci with dominance of both (15:1) 0.32 absence -> 1 dominant aUele (3:1) or 2 loci with recessive suppressor interaction (13:3) 133 1.15 0.90 Table 37. Segregation for reproductive traits in a test-cross segregating population. InflQreSQenge trait PhenQtygig gggea~ PL C4 TC ~ene agliQD X2 1e§1 Peduncle hairiness finely hairy x x 40 several genes with coarsely hairy 2 additive effects - Angle female axis subhorizontal x 23 threshold trait with (bunch orientation) horizontal 18 additive effects of pendulous x 2 dominant alleles ->subhoriz.(4:3: 1) 2.50 Persistence neutral deciduous x x 27 2 loci but only 1 flowers& male bracts persistent 16 segregates (1: 1) 2.32 Bunch density lax x 37 2 duplicate loci compact x 6 (3:1) 2.95 Finger shape erect 6 3 dominant alleles curved x x 37 -> erect (1 :3) 2.95 Peduncle length long 2 4 dominant alleles short/absent x x 40 -> long (1 :7) 1.64 Fruit cross section rounded 10 3 dominant alleles angular x x 33 -> rounded (1 :3) 0.01 Apex of male bract acute x 34 2 dominant alleles intermediate/obtuse x 7 -> acute (3: 1) 0.98 Internal colour male bract fading x x 39 4 dominant alleles bright crimson 2 -> bright crimson 1.53 Imbrication of male bud convolute 8 3 dominant alleles imbricated x x 34 -> convolute (1 :3) 0.51 male bracts behaviour reflexed not revolute 3 4 dominant alleles revolute x x 39 -> reflexed (1 :7) 0.67 external colour red to purple x 33 2 dominant alleles male bracts brown bronze x 9 -> brown 0.29 Tepal compound tepal absent x 25 1 recessive allele purple tinge present x 16 -> absent (1:1) 1.98 Colour style present x 2 4 dominant alleles purple tinge absent x 39 -> present (1 :7) 2.18 134 Table 38. Segregation ratios for bunch orientation in plantain derived diploid hybrids. Bunch orientation Cross horizontal or erect , sub-horizontal pendulous X2 testa 'Bobby Tannap' x 'Calcutta 4' 'Obi no I' Ewai' x 'Calcutta 4' 25 8 23 6 8 8 ~hi square tests adjusted by Yates' correction for small sample sizes. 2.71 nsb 1.07 nsC . bFit an expected test cross trisomic segregation ratio (13:10:4) for a duplex parent for 1 loci and simplex for the other two, under the assumption of dominant genes with additive effects controlling a threshold trait. C Fit an expected test cross trisomic segregation ratio (12:7:8) for a simplex parent for two loci and nulliplex for the other, under the assumptiom of dominant genes with additive effects controlling a threshold trait Table 39. Genetic model explaining bunch orientation in diploid and triploid Musa. Genotype Additive effects of Bo; loc; Diploid leyel: 801_B02_B03_a B01_B02_b03 b03 B01_b02 b02 B03_ b01b01B02_B03_ B01_b02 b02 b03 b03 b01b01b02 b02 803_ b01b01B02_b03 b03 b01b01b02 b02 b03 b03 b TriplQid leyel: 3 2 2 2 1 1 1 o 801801bo1802 b02 b02 803 b03 b03c 2 80 1b01bo 1802 bo2 b02 b03 b03 b03 d 0 a 'Pisang lilin'. b 'Calcutta 4'. Bunch orientation sub-horizontal sub-horizontal sub-horizontal sub-horizontal horizontal horizontal horizontal pendulous su b-horizontal pendulous C'Bobby Tannap', (2 copies of B01 , and nil effects of B02 and B03 in respective simplex loci due to dosage effects). d" Obino I' Ewai', (nil effects of B01 and 802 in respective simplex loci due to dosage effects) 135 Table 40. Segregation for pseudostem waxiness in plantain-banana hybrid progenies. Cross Calcutta 4 (C4) selfedB SH-3362 x C48,C SH-3362 x Long Tavoy8,C Pisang Lilin (PI) x C48,C Bluggoe x C48,C Bluggoe x C48,C Bobby Tannap x C48,C Bobby Tannap x C4b,c Obi no L' Ewai x C48,C Obino L' Ewai x C4b,c Obi no L' Ewai x PI8,c Obino L' Ewai x Plb,c French Reversion x C48,C French Reversion x C4b,c French Reversion x C48,d B2x. b4x. cF1. dF2- Parental Phenotype. NW selfed WxNW WxNW WxNW WxNW WxNW WxNW WxNW WxNW WxNW WxW WxW WxNW WxNW NW selted eWith slight pseudostem waxiness. Non-waxiness Waxiness (NW) (W) 46 0 40 2 10 8 31 15 14 5 0 1 26 12 0 5 15 2 6 5 2 1 0 1 6 0 3 0 29 6e 136 Table 41. Degree of expression, mid-parent value and progeny mean of pseudostem waxiness in plantain and banana 2x, and 4x hybrid progenies. Cross Pseudostem waxiness in progenye Mid-parent ------------------------------------------------------ SH-3362 x C4B SH-3362 x Long TavoyB Pisang Lilin (PL) x C4B Bluggoe x C4B Bluggoe x C4b Bobby Tannap x C4B Bobby Tannap x C4b Obino L' Ewai x C4B Obino L' Ewai x C4b French Reversion x C48,C French Reversion x C4~c B2x. b4x. cF1. valued 1.000 1.000 1.000 1.000 1.500 1.000 1.500 0.750 1.125 0.500 0.750 dMPV 4x= (3 * 3x+ 1 * 2x)/4 none slight moderate mean 40 2 0.095 10 6 2 0.556 31 15 0.326 14 5 0.263 all 1.000 26 10 2 0.368 all 1.000 15 2 0.118 6 5 0.454 all 0.000 all 0.000 e Mean = [0 x non waxy individuals + 1 x individuals with slight pseudostem waxiness + 2 x individuals with moderate pseudostem waxiness]/total number of individuals. 137 ~ Col (X) Table 42. Segregation ratios for apical dominance in intraspecific AA x AA and interspecific-interploidy AAB x AA crosses. Cross Genotypic cross Dominance 'Calcutta 4' (C4) selted Ad/Ad x Ad/Ad AAcax AAwb 'Pi sang lilin' x C4 Ad/Ad x Ad/Ad SH-3362 x C4 Ad/Ad x Ad/Ad SH-3362 x 'Long T avoy' Ad/Ad x Ad/Ad AABcxAAw 'French Reversion' x C4 ad/ad/ad x Ad/Ad F1 -2x F1- 3x F2--2x Ad/ad x Ad/ad 'Bobby Tannap' x C4 ad/ad/ad x Ad/Ad 2x 3x 4x 'Obino J'Ewai' x C4 ad/ad/ad x Ad/Ad 2x 3x 4x a cultivated 2x Musa acuminata. b wild 2x Musa acuminata. C cultivated 3x plantain. Apical Non-apical Expected dominance Ratio 0 46 0:1 0 46 0:1 0 42 0:1 0 18 0:1 0 8 0:1 0 1 0:1 4 30 1:3 7d 52 0:1 0 1 0:1 1d 7 0:1 2d 16 0:1 0 1 0:1 3d 12 0:1 d high apical dominance at flowering but during fruit filling time 1 sucker develops freely. Statistics X2 = 3.18 n.s. EMe= 21.82 EMe= 1.18 EMe=4.76 EMe=3.62 e Expected minimun according to binomial distribution assuming that the 2x parent was heterozygous. If Ad had incomplete penetrance, the EM should be higher than the observed in a cross between a recessive homozygous and a dominant homozygous. Table 43. Pseudostem height and height of tallest sucker at harvest of 2x and 4x plantain-banana hybrids and plantain parent. Plantain Pseudostem Height of Ratio IHybrid height (1) tallest sucker (2) (211 ) cm cm % 'Bobby Tannap' (BT) 340 171 50 BT x 'Calcutta 4' (C4) 2x (BT2x) 233 235 101 BT x C4 4x ( BT 4x) 281 249 89 'Obino I' Ewai' (OL) 370 188 51 OL x C4 2x (OL2x) 254 238 94 OL x C4 4x (OL4x) 336 243 72 Comparisonsa BT2x vs. BT 4x ** ns BTvs. BT4x ** ** BTvs. BT2x ** ** OL2x vs. OL4x ** ns OL vs OL4x ** ** OL vs OL2x ** ** BTvs OL ** ns BT4x vs OL4x ** ns BT2x vs OL2x ns ns a Mean comparisons using critical values of Student's t distribution based on Sidak's multiplicative inequality at the significance level p' = 1-(1-p)1A<, where p is theType I Error rate and k is the number of comparisons intended. ns, * , ** indicates non significance, and significance at the 0.05 and 0.01 significance levels, respectively. 139 Table 44. Effect of dw gene in pseudostem height at flowering and for the plant crop of diploid and tetraploid plantain-banana hybrids . Diploid population Phenotype a Genotype Plant height normal dwarf Contrast Dw vsdw Dw/dw dw/dw em 246 210 ** Tetraploid population Genotype Plant height Dw/Dw/Dw/dw Dw/dw/dw/dw em 299 274 ** a Normal plantain-banana hybrids have large internode length (> 9 em) whereas dwarf plantain- banana hybrids have short internode length « 9 em) . •• , .... indicate that the mean comparisons between normal vs dwarf phenotypic classes are significantly different according to the F test for the single degree of freedom contrast at the 0.01 and 0.001 significance levels, respectively Table 45. Segregation for fruit parthenoearpy in diploid derived plantains Cross OLxC4 BTxC4 Parthenocarpjc fruits Non-parthenocarpjc frujts 14 4 !O. 54 12. 16 G-statsa = 0.06 a Log likelihood test to compare the homogeneity of frequency distributions between both populations 140 Table 46. Analysis of variance for host response to black sigatoka as measured by the number of youngest leaf spotted (YLS) in the susceptible reference plantain cv. 'Agbagba' Source of Degrees of Sum of Mean F P level variation freedom squares square in F table Blocks (B) b-1=1 1.750 1.750 9.831 > 0.05 ns RepslB (r-1 )b=2 0.357 0.178 0.126 > 0.05 ns Residual (p-1 )br=24 34.000 1.417 Total pbr-1 36.107 Coefficient of variation =19.2% b = number of blocks = 2 r = number of reps per block = 2 P = number of plants evaluated per rep = 7 Table 47. Experimental unit size (EUS = number of plants), mean, variance, variance index and coefficient of variation (CV) for number of youngest leaf spotted (YLS) evaluated in a uniform evaluation trial of the reference plantain cv. 'Agbagba' EUS Means Variance Variance indexb CV (%) 1 6.2 1.34 1.34 18.7 2 6.2 0.68 1.37 13.2 3 6.3 0.26 0.76 8.0 4 6.2 0.14 0.54 5.9 5 6.3 0.12 0.58 5.5 6 6.3 0.09 0.52 4.7 7 6.2 0.12 0.85 5.7 14 6.1 0.12 1.75 5.8 a Mean over EUS b Variance index = variance x EUS 141 Table 48. Host response to black sigatoka disease according to the evaluation of number of youngest leaf spotted (YlS) in the susceptible reference plantain cv. 'Agbagba' Host response Susceptible less susceptible Partially resistant Code S lS PR 'Agbagba' YlS meana + n standard deviationsb <8 8 - 10 >10 a Digits were rounded to the nearest integer. b 2 and 3 standard deviations added to the 'Agbagba' mean for lS and PR classes, respectively Table 49. Segregation of host response to black sigatoka among 2x and 4x plantain-banana progenies obtained from crosses of the susceptible 3x plantain cultivars 'Obino l'Ewai" (Ol) and 'Bobby Tannap' (BT) with the wild 2x banana 'Calcutta 4' (C4) Cross PRa lS S G-stats ~ OlxC4 8 3 2 BTxC4 3 2 3 Total4x 11 5 5 1.56 ns ~ OlxC4 6 7 7 BTxC4 10 23 25 Total2x 16 30 32 1.41 ns 4xvs 2x 7.81 * apR = partially resistant; lS = less susceptible; S = susceptible ns, * Nonsignificant or significant at P=0.05, respectively (2 df) 142 ..... ~ (,) Table 50. Genotypes, types of gene action, expected segregation ratios (ESR), and X 2 tests of goodness of fit in 2x and 4x plantain-banana hybridsa Genotype [liJllgiQs Tetm~lgids (n=2 for 2x; ESRb Gene actionC Phenotyped ESRb Gene action Phenotype n=4 for 4x) d t BS 1 bsr2 bsr3 BS1 bsr2 bsr3 BSLbsr2-.bsr:L 1 1 0 0 0 S 1 0 0 0 S BSLbsri'bsr:L 1 2 0 2 0 S 1 0 4 0 LS BSLbsr2-.bsri' 1 2 0 0 2 S 1 0 0 4 LS BSLbsri'bsri' 1 4 0 2 2 S 1 0 4 4 LS bS1"bsr2-.bsr:L 1 2 R 0 0 LS 1 R 0 0 PR bS1 "bsri'bsr:L 1 4 R 2 0 LS 1 R 4 0 PR bS1"bsr2-.bSr3n 1 4 R 0 2 LS 1 R 0 4 PR bS1f'bsri'bsri' 1 8 R 2 2 PR 1 R 4 4 PR X 2 test (2 df) 5.2ns3.6ns 3.2ns ns Nonsignificant a Derived from crosses between a simplex plantain (BS1bs-,2bsr#bss;tJsribssj and the homozygous recessive 'Calcutta 4' banana (bs-,2bsr;1bsr:l> b Disomic (d) and trisomic (t) expected segregation ratios were tested at the 2x level. The expected segregation ratio at the 4x level is based upon production of SDR 2n eggs and the assumption that segregating loci are between the centromere and the first cross over C D = dominant susceptible (BS 1), R = recessive resistant (bs 1). Each bsr20r bsr3 has a (+1) effect and each bSS20r bss3allele a (-1) effect d S = susceptible, LS = less susceptible, PR = partially resistant Table 51. Hierarchical (or nested) univariate analyis of variance of euploid populations showing ploidy polymorphism considering one environment Population (P) Ploidy (X)/P Clone (C)IXIP Within clone d. f. P-1 (X-1 )P (C-1)XP (R-1)CXP Expected mean square a 2e + Ra2c, XI p + RCa2x,r F* RCXa2p a 2e + Ra2c, XI P + RCa2x,r P a2e + Ra'o XI p a~ Table 52. Effects of genotype and ploidy level on quatitative trait variation in 'Obino I' Ewai' (OL) and 'Bobby Tannap' (BT) hybrids Trait ParentPloidy (OUBT) 4x(OUBT) 2x(OUBT) Plant Height *** **INS **INS *** Height sucker NS NS ** *** Days fruit filling NS NS ***INS NSI*** Bunch weight *** *** ***/NS *** No. hands NS NS *** *** No. fingers NS NS *** *** Fruit weight *** *** ***INS *** Fruit length *** ***INS *** *** Fruit circumference *** *** ***INS *** NS, ** and *** indicate non significance, significant at the 0.01 and 0.001 levels, respectively 144 Table 53. Theoretical expected values in "Test-cross" ploidy mean analysis Ploidy ICiglil~ EacilOl IJUglil~ Eacilol Sil11glil~ Eacilol A 0 H A 0 H A 0 H 4x popa 2 1 1 0 1 1 -2 1 1 3x pOpb 1 1 1 -1/3 1 1 -5/3 213 213 2x popc 0 1 1 -2/3 213 2/3 -4/3 113 1/3 3x parent 3 1 0 1 1 1 -1 1 1 2x parent -2 0 0 -2 0 0 -2 0 0 asy SDR 2n eggs. ~ssuming chromosome segregation. CAssuming trisomic segregation ratios. A: gddjtjvjty. 0: dominance. H: degree of heterozygosity. Note that epistatic interactions AA. AD. and DO can be included in the model. Their coefficient will be the product of the respective columns (e.g. AD column A x column D) Table 54. Complementation of alleles controlling black sigatoka resistance (BSR) and high yield in heterosis for bunch weight. (Onne. 1991-1993) TMPx clone 1658-4 2796-5 548-9 548-4 582-4 597-4 Black sigatoka less susceptible partially resistant partially resistant partially resistant partially resistant susceptible Hetilrosjs (%) for buoch weight mid-parent highest parent 210 192 154 127 98 n ~ aJl 16 2 High yield = complementation of BSR + yield alleles 145 Ta~le. 55 .. Partition of total. ph~notypic variability according to the sources of vanatlon In the. ANOVA, significance of the ANOVA variance components and the macro-environmental (year) effect .and the genotype-by-year interaction (GxY) and broad sense heritability (H2) of growth parameters, bunch characteristics and black sigatoka response of plantain hybrids (Onne 1990 1991). ' Phenotypic COmr;2Q[]~Qts Qf v§(i§nQ~ SigQific§nc~ Trait PopulationS Geneticb Micro-environmente Year GxY H2 H 57.0% *** 30.1% ** 12.9% * * 0.85 HTSf 2.2% NS 60.0% NS 37.8% NS NS 0.44 HTSh 12.9% NS 51.5% NS 35.6% NS * 0.32 NSLF 19.8% NS 43.2% NS 37.0% NS NS 0.50 OF 29.9% * 42.7% NS 27.4% NS * 0.06 BW 41.9% *** 49.1% *** 10.0 % * NS 0.84 H/B 12.1% NS 54.1% NS 33.8% NS NS 0.27 F/H 19.7% NS 57.7% ** 22.6% NS NS 0.34 FW 24.8% * 57.6% *** 17.6% NS NS 0.84 FL 46.6% * 45.2% *** 8.2% NS NS 0.93 FC 30.2% * 66.5% *** 3.3% * NS 0.94 YLS 7.0 % NS 37.2% NS 59.8% * NS 0.63 TSSLI 11.5% NS 30.2% NS 58.3% NS NS 0.43 H= plant height at flowering (m), YLS= youngest leaf with necrotic spot due to black sigatoka, NSLF= total number of standing leaves at flowering, TBSLI = total black sigatoka leaf index score; from 0 (immune) to 4 (very susceptible), HTSf=height of tallest sucker at flowering (m), HTSh= at harvest (m), BW = bunch weight (kg), DF=days from flowering to harvest (fruit filling period), HIS = number of hands per bunch, F/H number of fingers (fruits) per bunch, FW = average fruit weight (g), FL=finger lenght (cm), FC=finger circumference. ns, *, ** , *** indicate non-significance and statistically significant at the 0.05, 0.01 and 0.001 levels. a Differences between plantain parents because the male parent was the diploid true breeding accession 'Calcutta 4'. b Differences between clones within the same population cross. e Within clone variance. 146 Table 56. Effect of major genes in quantitative traits Recessive genes: Plant height: dwarf gene - short internodes Height tallest sucker: apical dominance gene - sucker inhibition Black slgatoka: resistance gene for slower symptom development Dominant gene: Fruit parthenocarpy: increase in volume = length x circumference Table 57. Effect of allele substitution (additivity) and ploidy effect in the fruit parthenocarpy locus (P1), as measured by regression (b) in fruit characteristics of two euploid plantain-banana populations in the plant crop at Onne (1991-1992). The coefficient of determination (R2) indicates the total of phenotypic variation explained by the regression model Y = a + bx~ 'Bgbb)! Iaooag .l$ CalQuUa ~' ,~=atn 'QbiOQ I' ~ai x Cab.lUa ~' (N=34) Trait a b R2 a b R2 Effect of allele substitution (X 1): Bunch weight (kg) 1.10 2.68 *** 31.3% -2.79 7.55 *. 60.8% Fruit filling period (days) 128.00 6.02 0.5% 138.91 -6.51 4.0% Fruit weight (g) 14.55 29.23*" 32.4% -21.18 75.66·** 65.6% Fruit length (cm) 6.96 3.99 * •• 36.0% 6.30 5.79·* 59.3% Fruit circumference (cm) 4.50 3.20 *.* 54.8% 3.74 4.00 ••• 68.9% Ploidy effect (X2J: Bunch weight (kg) -0.30 1."78 *.* 17.6% -6.99 5.00 * •• 65.6% Fruit filling period (days) 144.07 -4.76 0.5% 145.67 -5.40 6.2% Fruit weight (g) -10.59 23.61 .** 28.0% -62.65 49.90 ••• 70.6% Fruit length (cm) 7.03 1.63 • 7.8% 3.93 3.54 * •• 54.8% Fruit circumference (cm) 3.10 1.97 .** 27.0% 2.32 2.37·· 59.3% •••• , .** indicates Significance of the coefficient of regression at the 5%, 1 % and 0.1 % levels 147 Table 58. Multiple regression models to explain the combined effects of allele substitution (# of copies of P1 allele) and ploidy (X) in the fruit parthenocarpy locus which affected fruit characteristics (Y) of two euploid plantain-banana populations in the plant crop at Onne (1991-1992). The coefficient of determination (R2) indicates the total of phenotypic variation explained by the regression model Y = a +~bjXi Multiple regression equation s.e. prediction R2 Fc (P) Durbin-Watson % ANOVA stats (d)S 'Bobby Tannap' x 'Calcutta 4' (N = 60) [dl= 1.51, du =1.65) Bunch weight (kg) = 0.70 + 0.27 X + 2.46 P1 2.16 31 12.82 (P < 0.001) 1.96 Fruit weight (g) = -2.32 + 11.39 X + 20.30 P1 21.93 36 16.27 (P < 0.001) 2.07 Fruit length (cm) = 9.11 - 1.46 X + 5.13 P1 2.76 40 18.90 (P < 0.001) 1.77 Fruit circumference (cm) • 4.38 + 0.08 X + 3.14 P1 1.56 55 35.16 (P<0.001) 1.56 'Obino I' Ewai' x 'Calcutta 4' (N= 34) [dl= 1.33 ,du= 1.58] Bunch weight (kg) .. -6.13 + 3.43 X + 2.71 P1 3.67 67 31.90 (P < 0.001) 2.44 Fruit weight (g) == -53.56 + 33.23 X + 28.80 P1 32.97 72 39.38 (P < 0.001) 2.82 Fruit length (cm) = 5.29 + 1.04 X + 4.33 P1 3.15 60 23.63 (P < 0.001) 2.38 Fruit circumference (cm) .. 3.46 + 0.29 X + 3.59P1 1.80 69 33.89 (P < 0.001) 2.21 S Residual analysis If dU < d < 4-dU the null hypothesis "data not serially correlated" is not rejected, however if d < dl or d > 4-dL the null hypothesis is rejected. When dl < d < dU or 4 -dU < d < 4 - dL the test is inconclusive Table 59. Genetic parameters estimated for black sigatoka reaction as measured by youngest leaf spotted in 3 plantain x banana hybrid families using 'Calcutta 4" as common resistance 2x parent. (Original data from PBIP 1991 Annual Report) Plantain parent Mean Additivity Recessiveness 'Obino L'Ewai' 7.8 1.0 1.8 'Bobby Tannap' 6.4 0.5 0.4 'French Reversion' 5.4 1.9 0.3 148 Table 60. Segregation in diploid test-crosses derived from matings between heterozygous triploid plantain and a homozygous diploid wild banana Trait Black sigatoka resistance (bS1) Dwarfism (dw) Fruit Parthenocarpy (P1) Ob~erved Yes' No 46 32 24 41 54 16 X2 calculated on assumption of: disomic trisomic 2.17 1.75 4.37 ** 0.43 19.60 ** 2.99 ** Deviated highly significantly from expected 1:1 test cross ratio for disomic inheritance according to X2 test adjusted by Yate's correction Table 61. Segregation for male (MF) and female fertility (FF) in diploid derived plantain-banana hybrids Cross Fertility (separate sex) Both sexes MaIil Female MF andl Both (MF) Sterile (MS) (FE) Sterile (FS) or FM Sterile 'Bobby Tannap' x 'Calcutta 4' 35 'Obino I' Ewai' x 'Calcutta 4' 16 Pooled results 51 15 5 20 x2test of homogeneity 0.08 ns x2testing 2:1 ratio 0.77 ns MS vs FS (x2test ~ = 1 .59 ns Sadjusted by Yates's correction 149 27 13 40 20 6 26 0.51 ns 0.83 ns 47 9 17 4 64 13 0.00 ns 9.07 ** Table 62. Non-independent assortment of Pt (Fruit parthenocarpy , FP) and BSt (black sigatoka response, BSR) loci in AAB plantain E~ aSB-> ~B LS S Yes 16 21 17 bStPt bStPt BStPt P P R No 0 7 9 bStpt bStpt BStPt R R P G=10.2** p .. parental (non-recombinants); R • recombinant types Table 63. Co-segregation for male and female fertility in diploid derived plantain- banana hybrids. Number of individuals showing the trait combination in each independent cross are separated by diagonal (I) symbol in the following order: ['Obi no I' Ewai' x 'Calcutta 4']/['Bobby Tannap' x 'Calcutta 4'] Male fertile Male steri Ie Female fertile 12/20 1/5 X21df.Homogeneity (pooled results): X23df.lndependent assortment: 1.60 ns 7.63 ns Female sterile 3/9 3/9 Expected values for X2 test of independence calculated assuming the ratio 4:2:2:1 instead of the 1: 1 :1:1 ratio for diploids due to duplex genotypes of the parents for these loci as well as trisomic segregation ratios. Also x2was adjusted by Yates' correction for small sample sizes. 150 Table 64. Co-segregation for male fertility and black sigatoka resistance in diploid derived plantain-banana hybrids. Number of individuals showing the trait combination in each independent cross are separated by diagonal (I) symbol in the following order: ('Obi no I' Ewai' x 'Calcutta 4']/['Bobby Tannap' x 'Calcutta 4'] bS1bs1 (Resistant or less susceptible) Bstbs;. susceptible) Male fertile Male sterile 11/24 316 x23d.f.lndependent assortment (pooled results): 3.65 ns 4/12 2/9 Expected values for X2 test of independence calculated assuming the ratio 4:2:2:1 instead of the 1:1:1:1 ratio for diploids due to duplex genotypes of the parents for these loci as well as trisomic segregation ratios. Also x2was adjusted by Yates' correction for small sample sizes Table 65. Co-segregation for female fertility and black sigatoka resistance in diploid derived plantain-banana hybrids. Number of individuals showing the trait combination in each independent cross are separated by diagonal (I) symbol in the following order: ['Obi no " Ewai' x 'Calcutta 4']/['Bobby Tannap' x 'Calcutta 4'] bS1bs1 (Resistant or less susceptible) BS1bs;. susceptible) Female fertile Female sterile 9/22 4/9 x23d.f.lndependent assortment (pooled results): 11.01 ** 3/5 2/13 Expected values for X2 test of independence calculated assuming the ratio 4:2:2:1 instead of the 1:1:1:1 ratio for diploids due to duplex genotypes of the parents for these loci as well as trisomic segregation ratios. Also x2was adjusted by Yates' correction for small sample sizes 151 Table 66. Co-segregation for male fertility and fruit parthenocarpy in diploid derived plantain-banana hybrids. Number of individuals showing the trait combination in each independent cross are separated by diagonal (I) symbol in the following order: [,Obi no I' Ewai' x 'Calcutta 4']/['Bobby Tannap' x 'Calcutta 4'] P1/J1 (parthenocarpic fruits) P1Pt< non parthenocarpic fruits) Male fertile Male sterile 14/30 3112 X21df.Homogeneity (pooled results): 3.11 ns X23d.f.lndependent assortment (pooled results): 11.85 ** 0/5 2/3 Expected values for '1.2 test of independence calculated assuming the ratio 4:2:2:1 instead of the 1:1:1:1 ratio for diploids due to duplex genotypes of the parents for these loci as well as trisomic segregation ratios. Also X2was adjusted by Yates' correction for small sample sizes Table 67. Co-segregation for female fertility and fruit parthenocarpy in diploid derived plantain-banana hybrids. Number of individuals showing the trait combination in each independent cross are separated by diagonal (I) symbol in the following order: ['Obino I' Ewai' x 'Calcutta 4']/['Bobby Tannap' x 'Calcutta 4'] P'P' (parthenocarpic fruits) P1P1( non parthenocarpic fruits) Female fertile Female sterile 12/24 3115 '1.21df.Homogeneity (pooled results): 1.66 ns x23d.t.lndependent assortment (pooled results): 14.30 ** 011 2/5 Expected values for X2 test of independence calculated assuming the ratio 4:2:2:1 instead of the 1:1:1:1 ratio for diploids due to duplex genotypes of the parents for these loci as well as trisomic segregation ratios. Also X2was adjusted by Yates' correction for small sample sizes 152 Table 68. Segregating populations for genetic studies and DNA fingerprinting which were produced until 1992 in PBIP-IITA. Crossa ploidy Number Qf ingivigual§ Total Characterized DNA-fingerprinted AA Bananas: SH-3362 x 'Calcutta 4' (C4*) 2x >50 42 0 'Pisang lilin'(PI*)xC4*(or reciprocal) 2x >50 46 0 C4 * X M. a. subspp. banksiib 2x 100 0 AAB plantain derivatives: {'French Reversion'* x C4*} F2 2x 40 36 33 'Bobby Tannap' (BT*) x C4* 2x 67 61 64 3x 2 2 2 4x 8 8 7 Total 77 71 73 'Obino I' Ewai' *x C4* 2x 32 23 30 3x 2 1 2 4x 16 15 15 Total 50 39 47 {BT* x [PI*xC4 *]} F~ 2x 90 0 0 aparents previously characterized and DNA fingerprinted are indicated with * bField established in 1992. 153 Table 69. Gene symbols, phenotypic trait and type of gene action in Muss . .. -_ ... -- ...... _ .... --.............. -.... --_ .............. -.. -_ ... _ .... -......... - .. _ ......... __ ................ _------ ... -.. _-_ .......... _ .......... -..... _--.... _--_ .... ---......... _-.............. _ ............ _-...... _-- Gene symbol Phenotypic trait Type of gene action 8 1. 82 albinism complementary recessives ad apical dominance major recessive gene b black pseudostem blotches modifier gene interaction due to recessive suppresor Bo; sub· horizontal bunch dominante gene action of 3 (angle of female axis) loci with threshold effect bS1 black sigatoka resistance major recessive gene with dosage effects in 4x bsr2. bsr3 black sigatoka resistance additive minor genes with dosage effects in 4x ? burrowing nematode resist. 1 or more dominant genes ? dwarfism in AAA bananas dominant but modifiers chang!3 dwarfism level dw short false internodes major ,recessive but dwarfism in 'Bobby Tannap' modififlrs affect plant height ms, fs male sterililty, female sterility recest'jive alleles interact P" P2.P3 fruit parthenocarpy Pmct , Pmc2 petioles margin clasping ? Fusarium wilt resistance ? Moko disease resistance pb 1. pb2 extensive pseudostem blotching pb3 extensive pseudostem blotching Pmb,. Pmb2 persistence of male bracts Pnf 1. Pnf2 persistence of hermaphrodite flowers r red pigmentation in leaves wx pseudostem waxiness ? yellow sigatoka resistance ? ? with s!,nsitive cytoplasm complementary dominants Dupli~te dominant loci major dominant for race 1 polygenic for race 4 several recessive genes complementary recessives in plantain additional complementary recessive in acuminata dominant complementary dominant complementary modifier gene interaction due to recessive supressor recessive but additive alleles modify expressivity recessive genes in burmanicca dominant genes in malaccensis multiple genes with dosage effects in microcarpa and errans. 154 Reference Ortiz and Vuylsteke (subm.) Ortiz and Vuylsteke (1992a) Ortiz (this report) Ortiz (this report) Ortiz and Vuylsteke (1992b) Ortiz and Vuylsteke (1992b) Rowe and Rosales (1992) Rowe and Richardson (1975) Ortiz and Vuylsteke (1992a) Ortiz (this report) Simmonds (1953) for AA Ortiz and Vuylsteke (1992b) for AAB plantains. Ortiz (this report) Larter (1947). Vakili (1965) Rowe and Rosales (1992) Rowe and Richardson (1975). Vakili (1965) Ortiz (this report) Ortiz (this report) Simmonds (1953), Ortiz (unp.) Simmonds (1953). Ortiz (unp.) Ortiz (this report) Ortiz (this report) Shepherd (1990) Rowe (1984) Vakili (1968) Table 70. Comparison of different breeding schemes, based on population sizes, for production of improved triploid hybrids (a) and population improvement at the diploid and tetraploid levels (b) Project/scheme Population size (N)a Land requirementsb (ha) a. Production of secondary triploids resembling the original parent plus trait of interest Plantain TMPx (4x) x TMp2x Chromosome segregationc 362.8 million 217,635 Chromatid segregationd 693.0 million 415,741 3x cv AAB x TMPx (4x) Chromosome segregation Chromatid segregation Banana TMBx (4x) x TMb2x Chromosome segregation Chromatid segregation 3x cv AAA x TMBx (4x) Chromosome segregation Chromatid segregation 1.3 million 2.5 million 4.2 million 2.0 million 2048 954 789 1,508 2,516 1,178 1.23 0.57 b. Recovery of desired homozygous additive/recessive genotype One·gene Diploid level 4 0.002 Tetraploid level Chromosome segregation 36 0.022 Chromatid segregation 22 0.013 Two·genes Diploid level Tetraploid level Chromosome segregation Chromatid segregation Mixed segregation 16 1296 474 784 0.010 0.777 0.284 0.470 aBased on computer simulation under the following assumptions: production of second division restitution 2n eggs in the triploid parent, segregation of only 11 loci (one for each putative linkage group), diploid parent is a homozygous recessive line, the primary tetraploid obtained will be at least duplex (2 copies of dom inant gene) for each of the 11 loci. b Plant density= 1667 ha·1 cLocus close to the centromere. dLocus after first crossing over, normally far away from the centromere. 155 Table 71. Efficiency of euploid hybrid production in 'Obino I' Ewai' x' Calcutta 4' (OL x C4) and 'Bobby Tannap' x 'Calcutta 4' (BT x C4) at Onne (1988-1990). Year/Cross Bunches EUplQids IelraplQids Seeds EuplQids T elraplQids pollinated in 100 pollinated bunches produced in 100 seeds produced 1988 OLxC4 152 12.5 2.6 1490 12.8 2.7 BTxC4 171 10.5 0.6 4543 4.0 0.2 1989 OLxC4 332 3.9 1.5 1519 8.6 3.3 BTxC4 313 4.1 1.0 5938 2.2 0.8 1990 OLxC4 514 1.6 1.1 1862 4.3 3.2 BTxC4 217 4.1 0.5 3703 2.4 0.3 . --------------------.-----_.---------------------------------------------------------------------------.----- 1988-1990 OLxC4 998 4.0 1.5 4871 8.2 3.1 BTxC4 701 5.7 0.7 14184 2.8 0.4 156 Table 72. Operational costs to produce one primary tetraploid hybrid and early evaluation based on PBIP 5 years' experience. Activity cost rate for 1 tetraploid OLxC4 BTxC4 Staff hours Staff hours Labour cost 1. Pollination (1 bunch) 1.5 100 200 2. Seed extraction (1 bunch) 1.5 8 17 3. In vitro seed germination (20 seeds) 1 1.5 12.5 4. In vitro multiplication (4 plantlets) 0.1 0.1 0.1 5. Roguing nursery(4 seedlings) 1 ?a ?a Sub-Total +110 +230 6. Early evaluation trial a. Field establisrnent and maintenance (4 plants)b 2 2 2 b. Field evaluation and record keeping (4 pl~nts) 3 3 3 c. Harvest & sucker preparation for next trial (4 plants) 5 5 5 Materials and equipment (costs are not indicated due to price fluctuation in local markets): 1. Pollination: plastic bags to cover female inflorescence (avoid unwanted natural pollinations by bats or bees, ,cheese clothe bag to cover male bud (avoid pollen contamination), small plastic bag to collect male flowers for further use in pollination, paper tags, plastic labels, water proof markers, pollination books, ladders (especially for giant cultivars). 2. Seed extraction: mechanical seed extractor, buckets, ethylene to speed ripening, seed envelopes, markers, seed production noteboOk. 3. In vitro seed germination: (see Vuylsteke, D.,t Ill. 1990. Tissue culture technology for the improvement of African plantains. In Sigatoka leaf spot diseases on bananas. INIBAP. pp. 316-337). 4. In vitro multiplication: (see Vuylsteke, D. 1989. Shoot-tip culture for the propagation, conservation and exchange of MflSa germplasm. IBPGR. 56 pp.) 5. Roguing nursery: seedling bags, labels, soil, manure, fiber. 6. Field trial: fertilizers (dosage depens on loCal conditions), herbicides (could be replaced by complete hand weeding), insectices (until production of banana weevil resistant hybrids). nematicides (until production of nematode resistant hybrids). "machete". Capital investment and other optional equipment: Rustic cold-frame or greenhouse. building for tissue-culture lab and office, ripening room. workshop, furniture for office, labs and ripening room. Optional: computer (aid for data analysis) and microscope (cytological wOrk). aTo be calculated, depends on aneuploidy frequency. b Ferilization. thining. weeding, pnJnning alleys. Costs for land preparation are not included. 157 Table 73. Flow of improved genetic materials and breeding/genetics research at PBJP. (In brackets are indicated institutions carrying the trials). Trial Knowledge Research EET (unrep.) ~ Segregation data collected for qualitative traits BASIC ~ Elucidate genetic systems (liT A) PYT (reps) ~ Assessment variability for quantitative traits STRATEGIC ~ Determine breeding methodology (IITA) M ET ~ GxE interactions and yield stability (reps & Iocs) ~ Select clones for targeted area AMYT/OFT ~ Identify clones for cultivar release (rep, Ioc, yrs)~ Develop Cultivar Profile 158 APPLIED (IITA/NARS) ADAPTIVE (N ARS/IIT A) -" 0'1 25 MT ha-1 year1 ! <3m. 210-240 days (7-8 months) 90-120 days (3-4 months) 365 days (1 year) >15 kg 6-8 bunch-1 75-100 bunch-1 >150-2'00 g > 75% sucker growth at harvest not quantified yet pendant no seep production parthenocarpic development >33% > 8 days from harvest until fruit yellows with black spots under natural conditions > 1.3 'dodo', 'tutu', boiled, chips 'matoke' (cooking), 'urwagwa' (beer) pest/disease resjstanceltolerance Black sigatoka* youngest leaf spotted> 10 Fusarium wilt* not quantified yet Nematode(s)* not quantified yet; specification required Banana weevil not quantified yet Abiotic stresses Drought tolerance low leaf conductance to water vapor in afternoon Sdepends of local consumption patterns in targeted area. * indicates trait in which genetic systems have been investigated. 161 Table 77. Seed-fertility in dessert, cooking and beer bananas at Onne (1988-1992). Most female fertile AAA East African bananas are indicated in bold. Non East African bananas are underlined. Seedset Cultivar -~------------------------------------------------- > 100 'Pelipita', 'Cardaba' 10 -100 'Bluggoe', 'Fougamou' CPisang Awak') .. 'Yangambi KmS', 5 - 10 1 - 5 < 1 o 'Makara' 'Ikimaga', 'Ingoromoka', 'Intama', 'Kibungo' 'Bakurura', 'Guineo', 'Igitsiri', 'Ingumbay', 'Inkira', 'Mbirabire' 'Chibulangombe', 'Igisahira', 'Imbogo', 'Ingarama', 'Intokatoke', 'Isha', 'Nyamwihogora', 'Rugondo', 'Highgate', 'Valery , 'Gashulie', 'Igihuni', 'Indemera', 'Nakitengwa', 'Nshika' 162 Table 78. Number of variants per phenotypic trait aHected by somaclonal variation in 13 different cultivars of plantain (Musa spp, AAB group). Cultivar No. of Inflorescence Plant Bunch Pseudosteml Bract Fruit AQnQrmal fQligge Total Frequency plants morphology stature orientation a petiole colour colour apex Distorted Drooping no. of of variation screened leaves leaves variantsb (%) French plantains Bobby Tannap 713 0 0 0 0 0 0 0 0 0 0 Mbi Egome 1 346 0 0 0 0 0 0 0 0 0 0 French reversionc 197 0 0 0 0 0 0 0 0 0 0 Ntanga 2 476 0 1 0 0 0 0 1 0 2 0.4 Akpakpak 322 1d 0 0 1 0 0 1 0 3 0.9 Obi no l'Ewai 2624 0 0 13 25 11 17 26 7 78 3.0 False Horn plantains Mimi Abue 178 3 0 0 3 0 0 4 0 7 3.9 Agbagba 1916 60 0 0 0 0 0 14 29 103 5.4 Didiede 315 29 0 0 0 0 0 0 0 29 9.2 Big Ebanga 532 135 0 0 0 0 0 1 0 136 25.6 ~ Ol Horn plantains U) Ihitisim 182 7d 0 0 0 0 0 0 0 7 3.8 Ubok lba 408 11 0 0 35 0 0 9 0 45 11.0 Uncertain Bise Egome 2 221 164 0 0 0 0 0 1 0 165 74.7 Total 8430 410 1 13 64 11 17 57 36 575 6.8 Frequency of variants (%) 4.86 0.01 0.15 0.76 0.13 0.20 0.68 0.43 aVariation in bunch orientation should be considered with caution as it was invariably associated with plants lacking vigour: stems were weak, bunches small and fruits short, approaching a diplOid-like phenotype. bTotal number of variants are not necessarily equal to the sum of variants per trait, because a Single off-type may be variant for several traits. cSomaclonal variant of the False Hom cv. 'Agbagba' with a French inflorescence type. d The stability of these variants in the ratoon could not be confirmed. Table 79. Frequencies of somaclonal variation in four plantain (Musa spp.) cultivars and in two independent trials Cultivar/ Total Qb§~Dl~g l£a.lu~§ Somaclonal 'X,2 Trial No. plants True-type Off-type variation (%) valuesa 'Agbagba'/1 520 497 23 4.4 Agbagbal2 244 235 9 3.7 Agbagbaltotal 764 732 32 4.2 0.22 ns 'Ubok Iba'/1 248 217 31 12.5 'Ubok Iba'/2 160 146 14 8.8 'Ubok Iba'/total 408 363 45 11.0 1.6 ns 'Big Ebanga'/1 369 240 129 35.0 'Big Ebanga'/2 163 156 7 4.3 'Big Ebanga'/total 532 396 136 25.6 55.95·* 'Bise Egome'/1 181 56 125 69.1 'Bise Egome'/2 40 0 40 100.0 'Bise Egome'/total 221 56 165 74.7 16.49 ** ------------------------------------------------------------------------------------------------------------- Cultivar differences/1 st trial 359.33 ** Cultivar differences/2nd trial 332.11 *. a'X,2test for indendence in a contingency table, ns, ., •• non significant or significant at P = 0.05 or 0.01, respectively Table 80. Frequencies of somaclonal variation according to plant propagation method in the False Horn plantain cv. 'Agbagba' Total Qb§~O!~g l£a.lu~s Somaclonal 'X,2 Propagation method plants True-type Off-type variation (%) value suckers 195 192 3 1.5 in vitro 244 235 9 3.7 1.84 ns 164 .... 0'> C1'I Table 81. Comparison of growth development and yield of two methods of propagation in cultivar 'Agbagba' Propagule H~ Hf OF TNLf YLSf FFP OH BW H/B FIB FW HTSh ANOVA (unbalanced two-way analysis of variance with sampling within experimental plot). Replications ns ns ns ns ns ns ns ns ns ns ns ns Propagation ** * ns ns ns ns ns ns ns ns ns * Exp. error ** ns ns ns ** ns ns ** * ** ** ns CV(%) 13.3 8.0 13.0 12.6 14.8 7.8 10.4 36.1 20.6 32.6 24.0 33.0 ------------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------------Sucker (N=86) mean 173 293 357 11 7 90 446 5.2 5 20 0.263 160 variance (a2a) 565 678 1006 2 1 52 1034 3.6 2 54 0.005 1944 standard error 2 3 3 0.2 0.1 1 3 0.2 0.1 1 0.008 5 mininum 105 200 278 6 4 64 363 0.0 0 0 0.000 40 maximum 230 355 443 14 10 112 537 9.6 8 52 0.450 275 skewness ns * ns ns ns * ns ns *** ** * ns kurtosis ns ** ns ** ** *** ns ns *** *** *** ns -------------------------------------------------------------------------------------------------------------------------------------------------------------- In vitro (N=1 04) mean 226 316 324 11 8 87 411 5.4 6 20 0.264 118 variance (a2w) 995 587 3010 2 1 43 3084 4.4 2 42 0.004 2275 standard error 3 2 5 0.1 0.1 1 5 0.2 0.1 1 0.006 5 mininum 120 240 241 7 5 70 327 0.4 2 4 0.067 36 maximum 290 390 514 15 11 102 593 12.0 10 39 0.500 430 skewness * ns *** ns ns *** *** ns ns ns ns ns kurtosis ns ** ** ns ns ns ns ns ns ns ** ** -------------------------------------------------------------------------------------------------------------------------------------------------------------- F-test~ ~a2w * ns ** ns ns ns ** ns ns * ** ns Hff.MP = plant height six months after planting (vigor), Hf = plant height at flowering (em), TNLf = total number of leaves at flowering of plant crop (em), DF=days to flowering, YLSf = youngest leaf spotted with black sigatoka at flowering, DH = days to harvest, BW = bunch weight (kg), FFP = fruit filling period (days from flowering to harvest), HIB = number of hands per bunch, FIB = number of fingers (fruits) per bunch, FW == average fruit weight (kg), HTSh == height of tallest sucker at harvest (an). aor viceversa when 02;..> 02s, ns, ., .. indicate non significant or significant at the 0.05, 0.01 or 0.001 level, respectively. Table 82. Effect of embryo age on germination of immature embryos of 'Bobby Tannap' x 'Calcutta 4'. Month 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Means Total plantlets Total seeds Embryo age (days after pollination) 80-85 (control) 70-75 60-65 0.0 0.0 0.0 0.0 0.0 6.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 12.8 4.2 0.0 12.8 4.0 14 348 Germination (%) 0.0 0.0 17.1 0.0 0.0 0.0 0.0 23.9 2.0 0.0 0.0 0.0 0.0 0.0 0.0 7.2 6.9 0.0 0.0 28.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.5 10 406 4.0 19 474 X2 for homogeneity of frequency distribution = 24.95 *** *** Significant at P = 0.001. 166 50-55 3.6 87.5 0.0 16.7 73.9 0.0 0.0 2.6 0.0 50.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 10.0 30 302 Table 83. Primer sequences used to detect polymorphisms in Papuan diploid M. acuminata clones. 01 5' -AAA-TGA-ATC-ACC-GGC- 3' 02 5' -GCA-GCG-CGG-CGG-TAC- 3' 03 5' GCC-GGC-TTG-CCG-TAC- 3' 04 5' -TCA-TGG-TTT-CTC-TTG- 3' 05 5' -CAC-AAG-AAA-AAT -CAT- 3' 06 5' -CGA-TTC-ATG-CAT-TTG- 3' 07 5' -AAT-CTT-TTT-GTC-AAA- 3' 08 5' -GAC-AGA-CAG-ACA-GAC- 3' 09 5' -TCT-TTT-CGC-TAT-TTG- 3' 10 5' -GAA-TAC-AGG-AAA-TAC- 3' 11 5' -CAC-TTC-AAC-TCA-TCA- 3' 12 5' -GAA-TAC-AAA-GAA-TTG- 3 ' Table 84. Twenty 10-mer DNA primers that were used to fingerprint Musa clones. ACGGATCCTG GAGGATCCCT CCTGATCACC GGTGATCAGG CCGAATTCCC GGGAATTCGG CCGATATCCC GGGATATCGG CCAAGCTTCC GGGAATTCGG CCGATATCCC CCAAGCTTCC GGAAGCTTGG CTTCCCCATT GGGAGACATC ACTCCTGCGA AAGCGGCCTC AGGTTGCCAG TGATICGATC GTCGCTAGGG 167 ONGOING DATA SET Table 85 . Dry matter content, fruit weights and pulp: peel ratios of in MET-1 (Onne, 1992-1993) Fruit Type %Dry matter Fruit weight (g) Pulp:peel Clone name content Baok 2nd. hand Baok ratio Bank PH TMPx 548-4 34.8 7 169 7 1.53 8 PH TMPx 548-9 33.1 1 1 184 6 1.58 7 PH 597-4 34.4 8 77 16 1.32 15 PH TMPx 582-4 35.1 6 130 12 1.60 6 PH TMPx 2796-5 31.1 12 191 4 1.52 9 PH TMPx 1112-1 33.8 10 160 9 1.36 14 PH TMPx 1658-4 33.8 9 186 5 1.45 1 1 PH 612-74-offtype 30.8 14 165 8 1.14 17 CB 'Fougamou' 36.9 3 94 14 1.96 1 CB 'Pelipita' 41.4 1 142 1 1 1.38 13 CB 'Cardaba' 31.0 13 198 3 1.61 5 CB 'Bluggoe' 30.7 15 243 1 1.39 12 P 'Obino I'ewai' 36.2 5 151 10 1.70 3 P 'Bobby Tannap' 37.2 2 114 12 1.47 10 P 'Agbagba' 36.3 4 212 2 1.66 4 DP 'Calcutta 4' 4 18 DP 'Pi sang Iillin' 29.6 16 52 17 1.73 2 DB 'Valery' 23.4 17 82 15 1.20 16 PH = Plantain hybrid, CB = Cooking banana, P = Plantain ,DP = Diploid banana DB =Dessert banana Table 86. Dry matter content of starchy crops Crop Plantain (Musa spp. AAB group) Cooking banana (Musa spp. ABB group) Plantain hybrids (tetraploids =AAB x AA) Cocoyam (Colocasia esculenta) Cassava (Manihot esculenta) White yam (Dioscorea rotundata) . Water yam (Dioscorea alata) Desert banana (Musa spp. AAA group) Irish potato (Solanum tuberosum) Sweet potato (Ipomoea batatas) Dry matter (%) 36-37 31-41 31-35 28-34 25-38 25-35 25-35 23 23-29 20-40 Data other than Musa crops from liT A annual reports and from H. D. TIndall (1983) "Vegetables in the Tropics". ELBSlMacmillian Education Ltd., UK, for Irish potato; and from J. Bradbury and W. Holloway (1988), ·Chemistry of tropical roots crops" ACIAR, Austratia for cocoyam and taro 168 ONGOING DATA SET Table 87. Storage duration of fruit: Analysis of days from harvest until fruit peel reaches colour stage 7a Significance according to . Clone type Days to peel % of standard Duncan's multiple range Rank order colour stage 7 Agbagba test at P= 0.05 % 'Pelipita' 17.9 'Agbagba' 11.6 TMPx 1112-1 11.5 612-740fftype 10.8 'Valery' 10.7 TMPx 582-4 10.2 'Pisang Lillin' 9.5 TMPx'1658-4' 8.7 597-4 8.7 TMPx 548-4 8.4 TMPx 548-9 8.3 TMPx 2796-5 7.6 'Calcutta 4' 7.4 'Bluggoe' 7.4 'Obino I' Ewai' 7.3 'Bobby Tannap' 6.5 'Cardaba' 6.4 'Fougamou' 5.1 100 99 93 92 88 81 75 75 72 72 66 64 64 63 56 55 44 a b bc cd d d e f f fg fg gh h h h a Peel colour stage 7 relates to a peel which is yellow with black spots. 169 ONGOING data set Table 88. Panel test ranking order of preference for BOILED samples of 18 Musa clones UI AGB Ol 548-4 548-9 597-4 582-4 2796-5 1112-1 1658-4 612-74 PEL BlU CAR FOU FHIA-2 FHIA-3 Valery Mean Nobs Rank Ubok Iba 89 100 100 100 100 97.8 45 1 Agbagba 11 97 100 100 100 97 100 100 100 100 100 100 100 95.3 192 2 'Obino l'Ewai' 92 87 100 93 100 75 80 85.9 99 3 TMPx548-4 3 8 70 73 52 50 100 100 90 53.5 129 6 TMPx548-9 0 0 13 30 48 80 100 73 100 78 100 80 48.8 170 7 597-4 0 67 33.3 18 10 TMPx582-4 0 27 33 9 100 82 37.0 73 9 TMPx2796-5 TMPx 1112-1 3 0 48 52 91 73 100 70 73 100 100 80 59.4 180 4 :::j TMPx 1658-4 7 20 27 100 67 30.7 75 11 0 61217 4-offtype -- 0 0 0 0 0 0 0 33 2.9 102 16 'Pelipita' 0 0 25 50 27 30 100 78 100 100 90 54.8 146 5 'Bluggoe' 0 20 27 33 100 58 42.4 99 8 'Cardaba' 0 0 0 0 0 100 22 91 100 27 29.9 127 12 'Fougamou' 0 22 67 0 42 11 23.5 51 13 FHlA2 0 0 0 0 0 0.0 45 17 FHIA3 0 0 18 0 73 18.2 55 14 'Valery' 0 10 20 20 10 12.0 50 15 Mean ... % preference for clone in all comparisons Nabs = Total number of comparisons involving this clone Table 89. Scored descriptive analysis for taste panel data from 18 Musa clones prepared as BOILED fruit Clone Colour Colour Taste Taste Sweetness Sweetness Texture Texture Acceptability Acceptabil ity quality rating quality rating quality rating quality rating quality rating Ubok Iba very good dark yellow good good flavour good acceptably sweet verygood slightly hard very good good quality Agbagba very good orange very good good flavour good acceptably sweet verygood correct texture very good good quality Obino l'Ewai very good dark yellow very good good flavour good acceptably sweet very good correct texture very good good quality 548-4 good dark yellow fair accept. flavour fair not sweet enough good correct texture good acceptable quality 548-9 fair dark yellow fair poor flavour fair not sweet enough good correct texture fair requires improvement 597-4 good dark yellow fair poor flavour fair not sweet enough fair correct texture fair requires improvement 582-4 good yellow fair poor flavour fair not sweet enough good correct texture fair requires improvement 1112-1 good dark yellow fair accept. flavour good acceptably sweet good correct texture good acceptable quality 1658-4 fair dark yellow fair poor flavour fair not sweet enough fair correct texture fair requires improvement ...... ...... ...... 612-74 poor brown poor poorflavou fair not sweet enough poor too soft poor not acceptable Pelipita good dark yellow good accept. flavour good acceptably sweet good slightly hard good acceptable quality Bluggoe fair dark yellow fair accept. flavour fair not sweet enough good correct texture fair requires improvement Cardaba fair dark yellow fair good flavour fair not sweet enough fair slightly too soft fair acceptable quality Fougamou fair yellow good accept. flavour good acceptably sweet fair slightly too soft fair requires improvement FHlA2 poor yellow poor unaccept. flavour poor no taste fair slightly too soft poor not acceptable FHlA3 poor white poor unaccept. flavour poor not sweet enough fair slightly too soft poor not acceptable TuuGia poor light yellow poor unaccept. flavour poor no taste fair correct texture poor not acceptable Valery poor light yellow poor poor flavour poor no taste poor too soft poor not acceptable Table 90. Rank order of clones for acceptability based on quality and acceptance rating of boiled samples -----------------------------------------------------Rank Clone Quality grade Rank Clone Acceptability score ----------------------------------------------------- 1 Ubok Iba very good 1. Uboklba good quality 2 Agbagba very good 2. Agbagba good quality 3 Obino l'Ewai very good 3. Obino l'Ewai good quality 4 Pelipita good 4. Pelipita acceptable quality 5 1112-1 good 5. 1112-1 acceptable quality 6 548-4 good 6. 548/4 acceptable quality 7 548-9 fair 7. Cardaba acceptable quality 8 Cardaba fair 8. 548-9 requires improvement 9 1658-4 fair 9. Bluggoe requires improvement 10 Fougamou fair 10. 582-4 requires improvement 11 Bluggoe fair 11. 1658-4 requires improvement 12 582-4 fair 12. Fougamou requires improvement 13 FHIA 1 fair 13. 597-4 requires improvement 14 597-4 fair 14. FHIA 1 requires improvement 15 FHIA2 poor 15. 612-74 not acceptable 16 Tuu Gia poor 16. FHIA 3 not acceptable 17 FHIA3 poor 17. Tuu Gia not acceptable 18 612-74 poor 18. Valery not acceptable 19 Valery poor 19. FHIA 2 not acceptable ------------------------------~---------------------- 172 ONGOING data set Table 91. Panel test ranking order of preference for DODO samples of 17 clones UI AGB Ol 548-4 548-9 597-4 582-4 2796-5 1112-1 1658-4 612-74 PEL BlU CAR FOU FHIA-2 Valery Mean Nobs Rank 'Ubok Iba' 74 71 100 100 100 100 100 86.1 114 'Agbagba' 26 65 88 80 90 96 90 100 100 100 29 100 100 96 100 100 80.4 352 2 'Obino I 'Ewai' 29 35 85 74 90 100 71 68 80 95 55 91 100 66.7 318 4 TMPx548-4 0 12 15 41 18 48 36 55 14 65 58 91 100 36.9 240 9 TMPx548-9 0 20 26 59 20 33 18 80 86 43 73 86 100 100 45.1 282 7 597-4 0 10 10 80 70 34.0 50 11 TMPx582-4 0 4 0 82 17.5 57 14 TMPx2796-5 0 10 29 52 67 30 36 61 33 37.9 153 8 TMPx 1112-1 0 32 64 82 64 64 91 57.3 110 5 .... TMPx 1658-4 0 20 20 60 80 36.0 50 10 -...I Co) 612-740fftype -- 0 5 45 14 40 91 80 29.5 105 13 'Pelipita' 71 86 57 100 78.6 28 3 'Bluggoe' 0 45 35 27 35 45 17 30.1 103 6 'Cardaba' 0 42 67 80 47.9 48 12 'Fougamou' 4 9 9 14 20 20 11.1 99 15 FHIA-2 0 0 0 2 2 8.0 50 16 'Valery' 0 0 0 0 0.0 28 17 Mean - % preference for clone in all comparisons absolute mean Nabs = Total number of comparisons involving this clone ONGOING data set Table 92. Scored descriptive analysis for taste panel data from 19 Musa clones prepared as DODO Clone Colour Colour Taste Taste Sweetness Sweetness Texture Texture Acceptability Acceptability quality rating quality rating quality rating quality rating quality rating Ubok Iba very good golden very good good flavour good acceptably sweet verygood correct texture very good good quality Agbagba very good golden very good good flavour very good slightly sweet verygood correct texture very good good quality Obino l'Ewai very good golden very good good flavour good acceptably sweet good correct texture very good good quality BobbyTannap good golden good accept. flavour good not sweet enough good correct texture good acceptable quality 548-4 fair golden good accept. flavour very good slightly sweet fair slightly too soft good acceptable quality 548-9 good golden good accept. flavour good acceptably sweet good correct texture good acceptable quality ... 597-4 fair golden fair accept. flavour good acceptably sweet fair slightly too soft fair requires improvement ....., .;. 582-4 fair golden fair accept. flavour very good slightly sweet fair slightly too soft fair requires improvement 2796-5 fair golden good accept. flavour good acceptably sweet fair slightly too soft fair acceptable quality 1112-1 good golden good accept. flavour good acceptably sweet good correct texture good acceptable quality 1658-4 fair golden good accept. flavour good Slightly sweet good slightly too soft fair requires improvement 612-74 fair golden good accept. flavour fair acceptably sweet fair slightly too soft fair requires improvement Pelipita very good golden very good good flavour very good slightly sweet verygood correct texture very good good quality Bluggoe good golden fair accept. flavour fair not sweet enough good correct texture fair requires improvement Cardaba good golden good accept. flavour fair acceptably sweet good slightly too soft good acceptable quality Fougamou poor light yellow fair poor flavour fair not sweet enough fair slightly too soft poor requires improvement FHIA 1 fair golden fair accept. flavour good acceptably sweet fair slightly too soft fair acceptable quality FHIA2 poor light yellow fair poor flavour good slightly sweet poor definitely too soft poor, not acceptable Valery poor light yellow poor poor flavour good slightly sweet very poor definitely too soft poor, not acceptable Table 93. Rank order of clones for acceptability based on quality and acceptance rating of fried ripened samples Rank Clone Quality grade Rank Clone 1 Pelipita 2 Ubok /ba 3 Agbagba 4 Obino L'ewai 5 Bobby Tannap 6 1112-1 7 548-9 8 548-4 9 2796-5 10 Cardaba 11 FHIA 1 12 Bluggoe 13 612-74 14 1658-4 15 Fougamou 16 582-4 17 FHIA2 18 Valery very good very good very good very good good good good good good good fair fair fair fair fair poor poor poor 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. Agbagba Uboklba Pelipita Obi no L'ewai Bobby Tannap 1112/1 54819 548/4 2796/5 Cardaba FHIA 1 Bluggoe 612/74 1658/4 Fougamou 582/4 FHIA2 Valery . Acceptability score good quality good quality good quality good quality acceptable quality acceptable quality acceptable quality acceptable quality acceptable quality acceptable quality acceptable quality requires improvement requires improvement requires improvement requires improvement requires improvement not acceptable not acceptable ----------------------------------------------------- 175 FIGURES Fig.1 . Plantain breeding strategy using diploid and tetraploid TMPx. 3x AAB BSS x TMP2x1 BSR 2n=3x J, n=x 3x AAB BSS x 4x BSR x TMP2x 2 BSR (MF->seeds) n=x J.- n=2x J.- n=x 3x BSR (MS->unseeded) ass = black sigatoka susceptible plantain. BSR = black sigatoka resistant hybrid MF = male fertile MS = male sterile. TMP2x1 and TMP2x2 should be unrelated diploid hybrids. Fig 2. Principal component analysis of Musa germ plasm based on growth and yield parameters N Z -a: a. 50 o -50 -:iJ.AA + o 0 +-tr + + # I:lt-+ :~8)( -SOD +00 )(Iiik 0 n a 0 + +.t: 'l- 0' +-fA""'" o 0 0 +0 "k<)+ + + 0)(0 )( + -100 -+--...,....--,--~--T'---.....---r---.----. 150 250 350 PRIN1 450 550 176 o microcarpa [] banksli A zebr!na + malaccensis )( errans o starchy bananas A French plantains o Horn plantains + cooking bananas Fig 3. Plot of the first (PRIN1) and second (PRIN 2) principal components from analysis of vegetative and bunch traits of West African plantains. Symbols based on their bunch type. 0 0 French M 0 [J French Horn Z 0 • False Horn ~ -100 0 ~ cJJ 00 0 6 Horn [J c ~ • [J 0 ... ·6 6 • • 6 • 0 -200 300 400 500 600 700 Fig 4. Plot of the first (PRIN1) and second (PRIN 2) principal components from analysis of vegetative and bunch traits of West African plantains. Symbols according to their plant stature. o M Z II I:l. Giant -100 - dl) Ii Medium ~ ~ .. Ii I:l. .. Small ~ .. I:l. /Jila llil I:l. Ii I:l. I:l. -200 I 300 400 500 600 700 PRIN 1 177 Fig. 5. Plot of the first (PRIN1) and second (PRIN 2) principal components from analysis of vegetative and bunch traits of West African plantains. Symbols based on their geographical distribution. 0 0 NIGERIA M + • GHANA Z + + CAMEROON ~ ·100 - .i' ~ + ~ • ~ct> + o· 0 0 + -200 I I 300 400 500 600 700 PRINt Fig. 6. Estimates of coefficient of variation for bunch weight of the cv. 'Obino l'Ewai' when different number of plants were included in the experimental plot. - 30 ~ 0 -c .. 0 25 - .: C'CI • 'i: C'CI > • - 20 • 0 ... • • c ••• CP • 'u 15 - • •• = • 8 0 10 I I I I I 0 10 20 30 40 50 60 70 plants per plot 178 Fig 7. The logarithmic relationship between the number of plants in the experimental plot and the variance of bunch weight in cv. 'Big Ebanga'. Q) u c: m a- m > en o 1 • 0- -1 o 0 o o o o o y = 0.64269 - 1.1187x RA2 = 0.927 o ·2+----------------~----------~~----------~---------------~, 012 log number of plants per plot Fig. 8 Monthy average seed set in the French plantain cultivars 'Bobby Tannap' and 'Obino I' Ewai' pollinated with the wild non-edible diploid banana 'Calcutta 4'. (Average of five consecutive years: 1988-1992). 30 - .t: 0 c: :I .c a- 20 Cl) Co f/) 'C Q) Q) 10 f/) -0 0 z 0 , , , , , , ~ , r:I ~ ... ~ , m • [?] h ~ n ~I " Bobb y Tannap l'Ewa! Ob!no ~ n ~ / / Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month 179 Fig 9. Monthly variation in seed set in the AAA banana 'Yangambi Km. 5' when pollinated with the wild diploid non-edible banana 'Calcutta 4' during two consecutive years (1988-1989). -o . o z 40Tr=======~-------------------------------~ 30 20 10 • 1988 Q 1989 , , , , , , , , , , , , , o~~~~~~~~~~~~'~~~~ Jan Feb Mar Apr May Jun Jut Aug Sep Oct Nov Dec Month Fig 10. Monthly variation in seed set in the ASS cooking banana 'Sluggoe' when pollinated with the wild diploid non-edible banana 'Calcutta 4' during two consecutive years (1988-1989). en "C Q) Q) tn -o . o z 400~======~-----=~----------1 300 200 100 • 1988 o 1989 , , , , , , , , , , , , , , , , , , , , , , , , , , o~~~~~~~~~~~~'~~~~ Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month 180 Fig. 11. Pollen stainability variation in five banana accessions, one banana hybrid and one plantain hybrid evaluated at Onne (October 92 to March 93). 100 -~ ~ 80 • 'Calcutta 4' >- .'!:: • 'P. lI11n' .t.:I 60 tV • TMb2x 384-30 C EJ 'Padrl' tV 40 ... III 0 'P. Jarl Buaya' C .! 20 0 • 'Galea' fa TMp2x 1297-3 Q. o OCT NOV DEC JAN FEB MAR Month Fig. 12. Variation in male fertility, as measured by pollen stainability, and in 2n pollen production in 'Pisang lilin' and its banana hybrid TMb2x 384-30 evaluated at Onne (October 92 to March 93). -~ - CI.I en tV ~ CI.I > tV ~ .c .... c 0 ::e 100~----------------------------------------------------------------------~ r------------------, 80 60 40 20 o OCT NOV DEC JAN FEB MAR Month 181 • 'Po IIl1n' (PI) II giant pollen PI • TMb2x 384-30 t:.::3 giant pollen 384-30 Fig 13. Monthly embryo and seed success for in vitro germination of plantain seeds at Onne (average of three years: 1989-1991). -~ 0 - en en CD u u :::J en 30 20 10 o • Seed success .. Embryo success Jan Feb Mar Apr May Jun Jut Aug Sap OCI Nov Dec Month Fig 14. Cumulative monthly production of euploid and tetraploid hybrids in 'Obino I' Ewai' x 'Calcutta 4' and 'Bobby Tannap' x 'Calcutta 4' (1988-1990). 30 "" CD 20 .c E :::J C «J 10 -0 t- o • Euploid hybrids 13 Tetraploid hybrids Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month 182 Fig 15. Estimates of variance index for bunch weight in a preliminary yield trial including the plantain hybrid and its tetraploid hybrids when different number of plants were included in the experimental plot. 1.8 - • >< Q) 1.6 ..; "C C 1.4~ •• 1.2 Q) (.) 1.0 ..; C • ca 0.8 ... -ca 0.6 --. > 0.4 J J J J J I I I J I I I I I 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Number of plants to be evaluated Fig 16. The logarithmic relationship between the number (#) of plants in the experimental plot and the mean square error (MSE) for bunch weight in a preliminary yield trial including the plantain and its tetraploid hybrids. 0.9 Y - 1.1275 - 0.5d709x RJ\2 - 0.966 - - 0.8 W en 0.7 :i 0.6 C) 0 0.5 ..J 0.4 0.3 0.6 0.8 1.0 1.2 1.4 1 .6 Log # plants 183 Figure 17. Flow of IITA Musa breeding materials. Step Duration Type of trial 1 2 3a 12-18 months EET: Early Evaluation Trial (Data recorded on BSR, bunch size, fruit parthenocarpy, dwarfness) J, 1-2 years PYT: Preliminary Yield Trial (Data recorded idem step 1 + earliness, ratooning, postharvest quality & durability, and other pest/disease resistance) J, 2 years MET: Multilocational Evaluation Trial (Data recorded to assess yield and BSR stability) Numbers involved > 100 clone~ 1-5 plants/l rep. 25-30 EET selected ~ RCBD or lattices with 2 reps. of 4 plants each 8-15 PYT selected ~+parents& local cvs as checks. RCBD using 5 plants! 2 reps/5-15 locs. 4 LET: Local Evaluation Trial Rapid testing by NARS for release to farmers (only of local relevance) 3b Other breeding materials coming from conventional or biotech programs through global network testing (IMTPIINIBAP) 5 2-3 years 6 # years J, J, AMYT: Advanced Musa Yield Trial (Data recorded idem 3 + standards of local preferences of each NARS NARS to supply local cultivars) J, OFf: On-Farm Trials (to release new cultivars & develop agronomic practices to maximize its yield: "cultivar profile") 3-5 advanced selected METlIMTP clones RCBD using 5 plants! 2-4 reps/ ? locs. 1-2 new cvs. with traditional cvs. # plants, # reps, # treatments (density, mulching, N-P-K, etc.) depending on NARS Steps 1 & 2 done at Onne (Nigeria) by the Plantain & Banana Improvement Program of IITA. Step 3a done by IITA in collaboration with National Agricultural Research Systems (NARS) Step 3b done by the Intenational Network for the Improvement of Plantain and Banana (INIBAP) in collaboration with liT A & other breeding programs and NARS Step 4 done by NARS with liT A & INIBAP support Step 5 done by NARS with IITA & INIBAP support Step 6 done by NARS (if required with inputs from liT A & INIBAP) 184 Fig 18. Field layout of MET-2 IReplication 1 (5 ~Iants each clone) IRe~lication 2 (5 plants each clone) VALERY VALERY VALERY VALERY V 1 TMPx 582-4 V 1 1 TM Px 2796/5 V 1 TMPx 6930-1 V 11 Agbagba A 2 TMBx 612-74 A 12 TMPx 1621-1 A 2 Pisang Lilin A 12 TMPx 2481 L VALER Y L VALER Y L VALER Y L VALER Y E 3 Obino l' Ewai E 13 Bluggoe E 3 TMPx 2637-49 E 13 Cardaba R 4 TMPx 6930-1 R 14 Valery R 4 TMPx 582-4 R 1 4 TMPx 548-4 Y VALER Y Y VALER Y Y VALER Y Y VALER Y 5 TMPx 2481 15 TMPx 2637-49 5 TMPx 548-9 1 5 Calcutta 4 ffilv 6 Pi sang Lilin 1 6 Calcutta 4 6 TMBx 612-74 16 TMPx 5511-2 VALER Y V VALER Y V VALER Y V VALER Y A 7 TMPx 1658-4 A 17 Cardaba A 7 Obi no l' Ewai A 17 Valery L 8 TMPx 4698-1 L 18 Bobby Tannap L 8 Bluggoe L 18 Bobby Tannap E VALER Y E VALER Y E VALER Y E VALER Y R 9 TMPx 548-9 R 19 TMPx 548-4 R 9 TMPx 1621-1 R 19 TMPx 2796-5 Y 10 Agbagba Y 20 TMPx 5511-2 Y 10 TMPx 4698-1 Y 20 TMPx 1658-4 VALER Y VALER Y VALER Y VALER Y width = 48 m distance between plants within a row = 2 m * randomized complete block * separate randomization in each location test plants: 200 borderplants: 200 V A L E R YI length = 45 m distance between rows = 3m V A L E R Y Fig. '9. Schematic field layout of AMYT·' x x x x x x x x x x x x x x x x x x x x x x x x x x x 1 1 x , 1 x 1 , x 1 1 x , 1 x , , x x 0 0 x 0 0 x 0 0 x 0 0 x 0 1 x , , x x , 2 x 3 4 x 5 6 x 7 8 x 9 0 x , 2 x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x 2 2 x 2 2 x 2 2 x 2 2 x 2 2 x 2 2 x x 0 0 x 0 0 x 0 0 x 0 0 x 0 , x , , x x , 2 x 3 4 x 5 6 x 7 8 x 9 0 x , 2 x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x 3 3 x 3 3 x 3 3 x 3 3 x 3 3 x 3 3 x x 0 0 x 0 0 x 0 0 x 0 0 x 0 , x 1 , x x , 2 x 3 4 x 5 6 x 7 8 x 9 0 x 1 2 x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x 4 4 x 4 4 x 4 4 x 4 4 x 4 4 x 4 4 x x 0 0 x 0 0 x 0 0 x 0 0 x 0 , x , , x x , 2 x 3 4 x 5 6 x 7 8 x 9 0 x 1 2 x x x x x x x x x x x x x x x x x x x x x x x x x x x Plot numbers are read downwards; first digit indicates the block and next 2 digits the experimental plot of 5 plants each. The border plants. a susceptible cultivar. are indicated with a "x" 186 Fig. 20. AMMI1 Model for bunch weight in TMPx germplasm 7.0 3.5 ,.. C3 0.0 0. -3.5 - -7.0 5 • )C • 0 41 0 + ~ • • I I 10. 15 20 Bunch weight mean (kg) o TMPx548-4 o TMPx548-9 )C 597-4 /). TMPx 582-4 • + TMPx 2796-5 • EET-Onne • PYT-PC Onne • PYT-R Onne • MET-Onne • MET-M'balmayo • MET-Ibadan Fig. 21. AMMI2 model for bunch weight in TMPx germplasm N 6 4 2 0.05). bChi-square tests for goodness of fit were calculated using Yates' corrections for continuity in small sample sizes, where X2 = [lObs. -Exp.1 - 0.5]2fExp. 190 Fig 26. Components of height of tallest sucker phenotypic variation at flowering (a) and at harvest (b) according to the sources of the analysis of variance and measured as % of total sum of squares. a. Height of tallest sucker a~, flowering , • Plantain parent (0.03%) • Ploidy (7.41%) • Between clones (68.11%) fa Within clones (24.25%) b. Height of tallest sucker at,' harvest 191 • Plantain parent (0.29%) • Ploidy (1.62%) • Between clones (72.62%) IZI Within clones (25.47%) Fig. 27. Selection schemes to produce non-apical plantains at different ploidy levels using AAB plantains and wild (AAw) or cultivated (AAc) Musa acuminata. AAB x (ad/ad/ad) J, 4xa or (Ad/ad/ad/ad) x J, 3xc 75% Ad/_/ad e a Primary tetraploid hybrid b Primary triploid hybrid c Secondary triploid hybrid d Phenotypic recurrent selection Increase AAw or AAc (Ad/Ad) 2x or3xb (Ad/ad) (Ad/ad/ad) J,PRSd Improved 2x 75% Ad/_ J,PRS f(Ad) in 2x population e Assuming chromosome segregation (no cross over between centromere and locus Ad). 192 Fig 28. Dwarfism in 'Bobby Tannap' (medium AAB French plantain) x 'Calcutta 4' (AA wild banana) segregating population. 'Bobby Tannap' x 'Calcutta 4' SDR 2n= 3x or n=2x eggs or n=x eggs J, n= x pollen J, J, J, 4x 3x 2x 2 normal normal 41 normal 6 dwarf 24 dwarf a a Test cross ratios tested: 1:1 (disomic inheritance), X2:: 4.37. * (Reject null hypothesis); 2:1 (trisomic inheritance due to pairing between A and B 'genomes' ), X? = 0.43 n.s. 'Fig 29. Components of pseudostem height phenotypic variation according to the sources of the analysis of variance and measured as % of total sum of squares. 12% 68% 193 • Ploidy II Genotype • Environment Fig 30. Breeding schemes to incorporate dwarfism in medium duplex and giant triplex AAB seed-fertile plantains utilizing 'Calcutta 4' (dw/dw). a. Medium duplex AA6 plantain( Ow/OW/dw) 1. AAB Ow/Ow/dw 4x Ow/Ow/Ow/dw (normal) x (select against normal) 2. OW/dw/dw/dw (dwarf) x dw/dw/dw (50%) Ow/dw/dw(50% ) (select and intercross) -L 3x -> -> b. Giant triplex AAB plantain (Ow/Ow/Ow) 1. AAB Ow/Ow/dw 4x 2. Ow/Ow/Ow/dw (normal) (intercross) Ow/Ow/Ow (25%) Ow/Ow/dw (50%) Ow/dw/dw (25%) If no dosage effect: 3. Dw/dw/dw Ow/dw/dw/dw x x 3x -> -> -> x -L -> AA (,Calcuta 4') dw/dw 2x Ow/dw (normal) dw/dw (dwarf) dwarf secondary triploid dwarf if dosage effect of the dw allele at the 3x level AA (,Calcuta 4') dw/dw 2x Ow/dw normal normal normal dwarf if dosage effect of dw allele at the 3x level dw/dw tertiary dwarf tetraploid 4.Production of tertiary dwarf triploids will be similar than in a2 194 Fig. 31. Principal component diagrams based on quantitative traits in 2 euploid plantain-banana hybrid populations: 'Bobby Tannap' (BT) x 'Calcutta 4' (C4) and 'Obino I ' Ewai (OL) x C4. . PCA on quantitative traits in BT x C4 200 0 • • 00· p • • • • D • • 0 • D N • • 0 4x • D C 100 • .1:] • • D 3x 'i: • •• • .. ., • c... • • • • 2x • • • • 0 • • \ • • • • .. • • • • • • o I I , , , I , I I I 190220250280310340370400430460490520 Prin 1 PCA on quantitative traits in OL x C4 300 -., • • 200 .. , .. N • •• • 4x c .. .. .. • 3x 'i: .. .. .. 2x c... .. .. 100 - .. • .. • • .. . • • .. • • 0 , 200 300 400 500 600 Prin 1 195 Fig. 32. Frequency distribution of youngest leaf spotted in euploid hybrids of IObino II Ewail x ICalcutta 41 (OL x C4) and 180bby Tannapl x ICalcutta 41 (8T x C4). Number on x axis indicate the maximum value for the rangel e.g. clones with YLS of 7 or 8 are grouped under 8. - ~ .! E :::J C ->-u c Q) :::J CT Q) ~ LL 20 10 0 4 6 8 10 YLS • BTxC42x II BTxC44x • OLxC42x ~ OLxC44x 12 Number on x axis indicates maximum value for the ran e, e. . clones with YLS of 7 or 8 are rou ed under 8. Fig. 33. Repulsion phase linkage between recessive black sigatoka resistance (bS1) and dominant fruit parthenocarpy (Pt}genes in AAB plantain. O_Bs L _________________ P1 o bs . P _ 1__________________ 1 o bs P _ 1__________________ 1 ----------34 eM---------- 196 Fig. 34. First linkage group in plantain. O-BSt---------FF--------Pt--MF--NFPI Locus designation: as 1, major locus for black sigatoka resistance. FF, female fertility. P1, fruit parthenocarpy. MF, male fertility. NFPi , persistence of neutral flowers and male bracts. 0, centromere. Fig. 35. Genetics of host response to black sigatoka disease in plantainlbanana (Musa spp.). Resistance is under the control of 1 recessive allele (bS1) and modifier gene action of another 2 independent additive resistance alleles (bsr2 and bsr3). S = susceptible reaction, LS = less susceptible, MR = moderately resistant 2X 4X ®-------r~----~ 197 Fig. 36. Black sigatoka resistance stability in M. acuminata accesions. 2.5 -.c 2.0 - >- :t= 1.5 -.- .c '" -til 1.0 a: en m 0.5 0.0 0 - 0 0 + 0 .q-, U - 0 1 2 3 4 5 Black slgatoka reaction in host cultivar o (avirulent) to 5(hlghly virulent) o SF215 o Grande Nalne [J Salmaea o PI sang mas • KokodJa + Plsang Berlin o Pahang o Calcutta 4 o TU8 o TuuGla Fig. 37. peA model for black sigatoka response in banana across different isolates. o SF215 6 0 Grande Naine N 0 Salmaea 3 - • t!f 0 0 Plsang mas 0 ". • KokodJa •• .. . ". + ·0 CT'1 + Pisang Berlin - x C\I 0 w- 0 Pahang • II( 0 Calcutta 4 0 0 Q. -3 - o TU8 o TuuGla • Pacific Islands -6 • Papua-N.Gulnea 0 • Philippines -9 , , It Central Amer. 4 6 8 10 12 14 16 18 peA 1 20 22 24 26 )( Nigeria 198 Fig. 38. Gains from fungicide application vs. host plant resistance to control black sigatoka. Non fungicide treated plantain IObino II Ewai' was the reference point (9.7 MT ha-1 year1 =100%) to determine gains in potential yield with respect to its derived tetraploid hybrids TMPx 548-4 and TMPx 548-9 as well as the fungicide treatment. Treatment TMPx 548-9 TMPx 548-4 Fungicide o o Obino I' Ewai • Gains (%) 100 200 300 400 Gains (%) Fig. 39. Potential cash benefit (US$ ha-1 year1) from fungicide application vs. host plant resistance to control black sigatoka disease in plantains under two rural market conditions, (based on prices for plantain in areas near Onne, S. E. Nigeria and a fungicide cost of US$ 750 ha-1year1). Period Scarcity ~§ Abundance o 2000 • Host plant resistance 8 Chemical control 4000 6000 8000 Profit for surplus production (US $ per ha) 199 Fig. 40 Effect of seed age in immature embryo rescue of plantain hybrid seeds obtained by triploid diploid crosses. -fn >-CIS ,., - 50-55 (8) 60-65 (b) 70-75 (b) 80-85 (b) o 246 % Germination 8 10 Ages followed by different letters (in brackets) were significantly different at the 0.01 probability level (X2 test for homogeneity of frequency distributions). Fig. 41. Effect of AgN03 on the germination of mature plantain hybrid embryos obtained by triploid diploid crosses. ~ " ~ , , , , 10 (ns) , , , , , , , , , , , , , , -E 3 (ns) -Cl ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~t ' , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , - (t) 1 (*) 0 Z ',',',',',',',',',',',',',',',',',',',',',',',',',',','~ ',',',',',',',',',',',',',',',',',',',',',',',',',',','~ ',',',',',',',',',',',',',',',',',',',',',',',',',',','~ Cl