Breeding for resistance to fusarium wilt and internal fruit quality traits in Sukali Ndizi (Musa genome-AAB)

Abstract

Apple banana (Musa species, AAB genome group) is the most widely distributed dessert banana cultivar in Uganda. It is locally known as Sukali Ndizi in central region and Kabaragara in Western regions of Uganda and in Rwanda, it is known as Kamaramasenge. Farmers grow landraces that have low yield potential and are susceptible to various stresses such as Fusarium wilt disease. Introduction of improved dessert banana varieties, with stress tolerance traits such as the FHIA-17, significantly increased productivity but the attempt to introduce these high yielding disease and pest resistant varieties into the dominantly landrace system in Great lakes region was not successful because farmers claimed that the varieties were lacking in fruit quality traits that they preferred. The aims of the study reported in this thesis were 1) investigate diversity aspects of resistance to Fusarium wilt disease and fruit quality traits in Sukali Ndizi. 2) Assess the feasibility of getting a consumer acceptable Ndizi hybrid, with improved yield and resistant to Fusarium oxysporum f. sp. Cubense race 1 through cross pollination. 3) Determine the seggregative parterns these traits. 4) Developing near infra-red calibration models based on internal fruit quality traits. Nineteen Sukali Ndizi landraces were collected from nine key Sukali Ndizi growing districts of Uganda. They were then, evaluated in the field and laboratory for their fruit quality attributes and response to Fusarium wilt. Results showed a wide diversity among Sukali Ndizi landraces for fruit-quality traits (fruit pulp texture, flavor and taste) but no significant difference for susceptibility to FOC race 1. Results of cross pollination, two new Sukali Ndizi hybrids (SDH1 and SDH2) with bunch yields higher than those of check cultivars by >50%. SDH1 and Yangambi-KM5 showed no symptoms of Foc race 1, whereas SDH2 showed mild symptoms on its corms. The consumer acceptability of SDH1 and SDH2 were not significantly different from Sukali ndiizi, thus perfect substitutes. The results of the seggregative parterns this traits.The cross Sukali Ndizi and TMB2X8075-7 showed high frequency of transgressive segregation for internal fruit quality traits in F1 generation meaning the crossing between Sukali Ndizi and TMB2X8075-7 provided a base for the selection of superior progeny. While a subset of families were analyzed in detail, the data indicated that the important internal fruit quality aspects such as fruit pulp texture, fruit acidity, fruit sugars and sugar/acid ratio were quantitatively controlled. Good calibration equations of Sukali Ndizi internal fruit quality were achieved with modified partial least square (MPLS) method. Calibration model performance for the attributes of TSS, texture and S/A ratio (typical R2c of 0.73, 0.69 and 0.7) respectively and typical standard error of calibration of 0.73, 0.25 and 5.36 respectively. Ability to automate the processes and the possibility of simultaneously predicting multiple quality parameters is of a great advantage towards development of a high throughput phenotyping platform.