Characterisation, identification and integration of drought tolerance in banana (Musa spp.).

Abstract

Banana (Musa spp.) is one of the major staple foods grown in the subtropical and tropical parts of the world. Given the increasing drought occurrences in sub-Saharan Africa and high susceptibility of banana to water stress, the improvement of cultivars through breeding offers the most effective and long-term solution to protect the crop against the daunting effects of climate change. Therefore, the objectives of this study were: (1) to assess the effects of drought on banana production and identify management strategies deployed by farmers (2) to determine the response of a diverse set of Musa genotypes to water stress based on phenotypic and physiological traits and select promising genotypes for use in banana drought tolerance breeding (3) to analyse the hybridisation success of the two selected drought-tolerant candidate male parental lines based on pollination success, seed production, embryo recovery and embryo germination rates (4) to determine the genetic relationships and diversity among 55 banana genotypes using DArT-based SNP markers and (5) to determine the variability of water usage and assess the growth behaviour and transpiration responses of secondary banana hybrids and their parental lines to declining soil water content in function of vapour pressure deficit and light intensity. A baseline participatory study involving 120 banana farmers selected from eight major banana growing districts in the cattle corridor of Uganda showed that the majority of the farmers were small-scale holders utilizing intercropping systems and growing mostly East African Highland (EAHB) cooking-type bananas. Fifteen drought stress effects were reported, with reduced bunch weight (90%) being the most prevalent. Amongst the different drought mitigation practices, mulching (56%) was used the most, although irrigation (10%) was the most effective option, but the high cost of water pumps and water scarcity limited its deployment. The extent of deployment of mitigation practices was mostly low (72%), with farmers applying only 1-3 practices. The second study evaluated 16 genetically diverse banana genotypes in the screen-house at the National Agricultural Research Laboratories, Uganda, under well-watered and water stress conditions. Water stress significantly reduced the plant height (PH), total leaf area (TLA), number of functional leaves (FL), total dry matter (TDM), chlorophyll content (CC) and relative water content (RWC). Genotype by water treatment interaction effects were significant for TLA, PH, FL, number of new leaf cigars (LC) and RWC (p < 0.01; p < 0.05). The water use efficiency (WUE) of 12 genotypes increased under water stress conditions. Stomatal conductance was also affected by the genotype x water treatment interaction (p < 0.05). ‘ITC.0987’ was the most tolerant, whilst among the improved diploids, ‘TMB2x9722-1’ had the least TDM reduction and the highest WUE, and ‘TMB2x9172’ showed the least decrease in RWC and highest root-shoot ratio (RSR) under stress conditions. Thus, ‘ITC.0987’, ‘TMB2x9722-1’ and ‘TMB2x9172’ are essential drought-tolerant candidates that may be utilized in breeding. In the third study, pollination success, seed set rate, embryo recovery and germination success were determined for crosses between four tetraploid Musa hybrids, including ‘660K-1’, ‘917K-2’, ‘1201K-1’ and ‘222K-1’ (female parents) with two diploid droughttolerant candidates, ‘TMB2x9722-1’ and ‘TMB2x9172’ (male parents). Seed set (p < 0.05), embryo recovery (p < 0.001) and embryo germination (p < 0.001) were significantly affected by female-male parent interaction effects. Crosses with ‘TMB2x9172’ exhibited relatively higher pollination success, seed set, and embryo recovery than those with ‘TMB2x9722-1’. Thus, ’TMB2x9172’ was the more ideal male parent for generating progeny populations for subsequent drought tolerance studies. The fourth study determined the genetic relationships and diversity among 55 Musa spp. genotypes using 1551 high-quality DArTseq-based SNP markers. The study population comprised breeding lines, Musa progenies, wild species and landraces. Neighbour-joining (NJ) cluster analysis divided the population into two major clusters, with one large cluster consisting of all the breeding lines, progenies, ‘Mpologoma’ (AAA) and ‘Calcutta-4’ (AA) and a smaller cluster consisting of ‘Musa balbisiana’ and three landraces, all with a ‘B’ genome. We zoomed into the genetic relationships of 16 selected Musa progenies and their drought-sensitive female (‘917K-2’) and tolerant male (‘TMB2x9172’) parents and identified sub-clusters of uniquely and genetically similar progenies. Principal component analysis was consistent with the NJ tree and predicted that three progenies, ‘NM101F1’, ‘MNK-17-11’ and ‘MNK-17-12’, were more genetically identical to their male parent, while the rest of the progenies were more identical to the female parent. Analysis of molecular variance showed that 88% and 12% of the total genetic variation were within and between the gene pools, respectively. The SNP markers had a mean proportion of polymorphic loci and observed heterozygosity of 0.65 and 0.36, respectively, indicating the existence of substantial genetic diversity across the study population. The fifth study investigated the water usage, plant growth and transpiration rates (Erate) of 18 Musa spp. hybrids, including 16 progenies and their two parental genotypes. Traits recorded were duration of soil water content (SWC) decline from 2.1 g g-1 (-0.01 MPa) to 0.7 g g-1 (-1.92 MPa), functional total leaf area (TLA), leaf damage, and transpiration rates (Erate). Erates were modelled in relation to SWC, vapour pressure deficit (VPD) and light intensity according to the Jarvis-Stewart model. The number of days taken for SWC to drop to 0.7 g g-1 (-1.92 MPa) differed significantly (p < 0.05) among the genotypes. The TLA, leaf damage and Erate were significantly affected by the genotype, water treatment and genotype x water treatment interaction effects. The female parent (‘917K-2’) and four triploid progenies, ‘MNK-16-3’, ‘MNK-16-16’, ‘MNK-16-8’ and ‘MNK-17-5’, exhibited good growth potential under optimum conditions, but their growth was significantly reduced during short periods of water stress (17-22 days). The male parent (‘TMB2x9172’) and two diploid progenies, ‘MNK-17-11’ and ‘MNK-17-12’’ took longer to deplete their soil moisture (28-35 days) and sustain relatively good growth under water stress conditions. ‘MNK-16-2’, ‘MNK-16-4’, ‘MNK-16-5’ and ‘MNK-17-6’ (all triploids) had relatively good growth under both well-watered and stress conditions but relatively fast soil water depletion. Genotype-specific critical thresholds (SWCcrit) were observed, implying varying stomatal and or hydraulic control levels. Based on the daily Erate model, ‘MNK-17-6’ had the highest SWCcrit (4.92 g g-1, -0.0002 MPa), indicating a conservative, drought-avoiding response to a declining SWC. Among the eight genotypes assessed under gradual increase in VPD and light, ‘MNK-17-4’ and ‘MNK-17-12’ had the fastest and slowest Erate increases, respectively. In conclusion, drought is a major threat to sustainable banana production in the East African Great Lakes region. Therefore, there is an urgent need to develop and deploy varieties with improved tolerance to drought. Also, farmers need to prioritise preventive drought mitigation practices. The selected drought-tolerant candidate banana genotypes (diploids) are recommended for further field testing and future use in crosses as male parents to generate larger segregating populations and or improve drought tolerance in cultivars that are susceptible but possess desirable agronomic and taste attributes, e.g. the EAHBs. The causes of low hybridisation success and poor embryo germination in crosses with the drought-tolerant candidate, ‘TMB2x9722-1’, should be investigated to enhance its utilisation in banana drought tolerance crossbreeding. The study also recommends crossing ‘TMB2x9172’and ‘TMB2x9722-1’ to generate even more drought-tolerant hybrids. Understanding the genetic relations of the progenies is valuable in providing insight into the genetic diversity underlying the phenotypic diversity, especially regarding their responses to water stress. The great phenotypic variation observed in the water usage, growth, and transpiration behaviour of the tested progenies in relation to the fluctuating environment suggests that the suitability of a given progeny will depend on the prevailing drought scenario. The relatively drought-tolerant diploid progenies, ‘MNK-17-11’ and ‘MNK-17-12’, should be further evaluated for their growth and yield potential under optimum and drought field conditions. The contrasting phenotypic behaviour among genetically similar progenies highlighted the significance of the environmental impact on a plant’s phenotypic expression. Despite the challenges posed by the complexity of banana drought tolerance, this work illustrates that banana drought tolerance improvement through crossbreeding is possible.