Performance of Narita Banana Hybrids in the Preliminary Yield Trial for Three Cycles in Uganda

Abstract

Bananas (Musa spp.) are an indispensable part of life in Eastern Africa providing up to one fifth of total calorie consumption per capita. Unlike many staple crops, bananas deliver food throughout the year, making them an ideal crop for household incomes, food and nutrition security. However, banana yields are low due to several factors amongst others pests and diseases: weevils and nematodes, Fusarium wilt, bacterial wilt and black Sigatoka. There are many potential technology-based interventions for increasing banana yields but host plant resistance is the most appropriate and cost effective intervention given the current stage of development of banana systems in the region. Host-plant resistance also offers significant spill over benefits for human health and positive environmental impacts. Therefore, the Ugandan National Agricultural Research Organization (NARO) and the International Institute of Tropical Agriculture (IITA) jointly breed bananas largely for host-plant resistance to improve banana yields. One of the most important current products of their joint banana breeding efforts is secondary triploid hybrids for food and juice herein referred to as NARITA hybrids. This name specifies the contribution of NARO and IITA. An earlier report (NARITA report 1) presented the results of 25 NARITA hybrids for cycles 1 and 2 combined. The current report presents and discusses the results of the same 25 NARITA hybrids (18 for food and seven for juice) evaluated for three crop cycles at Sendusu in central Uganda and analyzed in combined and separate forms. Results of individual NARITA hybrids within cycles showed high degree of variation for the traits assessed, implying a high potential for selection among the NARITA hybrids evaluated. For example, the bunch weight (BWT) of the individual NARITA hybrids ranged from as low as 5 kg for NARITA 19 to as high as 45 kg for NARITA 24 with a mean of 17.8 kg. Averaged across three cycles, BWT ranged from as low as 8.7 kg for NARITA 19 to as a high as 30.4 kg for NARITA 24. Ninety six per cent of the hybrids had a mean BWT greater than the mean of the local check (Mbwazirume) (11.0 kg). Similarly, NARITA hybrids were better than Mbwazirume for most of the other traits assessed. Eighty four per cent of the NARITA hybrids evaluated were better than the best founder parent (NFUUKA) for bunch yield (t ha-1), indicative of the significant breeding progress made by NARO and IITA in this breeding program. This could be confirmed by the positive better founder parent heterosis for BWT recorded by all NARITA hybrids, with NARITA 17, NARITA 18, NARITA 7 (M9), NARITA 21 and NARITA 14 (all food type) exhibiting highest heterosis. Results of combined analysis of variance (ANOVA) showed significant differences among the NARITA hybrids for all the 14 traits assessed including BWT. This indicated the potential for further selection and improvement of the NARITA hybrids for all the 14 traits. Additionally, results of combined ANOVA showed significant differences among three crop cycles for all the traits assessed except days to bunch maturity (DTM) and number of functional leaves at flowering (NFLF), indicating that the selection of banana hybrids could best be done at certain cycle numbers. The performance of NARITA hybrids for most traits was much higher at cycles 2 and 3 than at cycle 1 with the highest performance observed at cycle 3. However, the difference between cycle 2 and cycle 3 was not significantly different for most traits including BWT. The clear implication of this is that selection for banana hybrids should be done at cycle 2 to reduce costs involved in the management of trials since banana trials are always huge considering the size of bananas as well as spacing of 3 x 3 m or 2 x 3 m commonly used. Also, banana performance data analysis should not be based on a combined evaluation of cycle 1 and 2, as was previously done for NARITA report 1, but on an analysis of individual cycles, preferably cycle 2. The limitation of single site and single line plots is acknowledged. Hence, NARITA hybrids will be evaluated in larger and replicated multi-location trials to ascertain their actual performance, adaptability and stability in comparison with the local EAHB cultivars. Nevertheless, based on these preliminary results, potential high yielding banana hybrids combining resistance to black Sigatoka and farmer-preferred quality traits exist within this NARITA population.