Author:

Abstract:

Banana (Musa spp.) production ranks among the top 15 crops cultivated worldwide. Optimal production requires continuous and abundant water, which in many cases is scarce and becoming unavailable. With climate change, water requirements in banana-growing regions will even further increase. Improving banana performance for future climate conditions is therefore one of the priorities of banana breeding. The banana wild relatives have been suggested as new sources of allelic diversity for drought tolerance as they contain naturally acquired tolerances to abiotic stresses. However, the characterisation and evaluation of wild banana species are still in their infancy and their potential for drought tolerance has not been evaluated yet. The general aim of this PhD was to identify drought tolerance traits among the wild banana diversity. Drought tolerance is a complex phenomenon, consisting of the integrated response of many drought tolerance traits of which the impact varies across environments. Therefore, we focused in this PhD on the identification of traits related to specific environmental dynamics. More specifically, we targeted transpiration and stomatal responses to dynamics in light intensity, vapour pressure deficit (VPD) and soil water potential. Our main focus was on the ancestors of the commonly cultivated banana, Musa acuminata and Musa balbisiana, as these have the highest relevance for breeding programs. In a first step, we evaluated the impact of fluctuating light on stomatal responses (Chapter 3). In the field light intensity fluctuates within seconds, requiring continuous adjustments in stomatal aperture for optimal functioning. The kinetic responses of stomatal conductance to these fluctuations are key to photosynthesis and water use efficiency. Using step-changes in light intensity, we studied the diversity of light-induced stomatal kinetics in five banana genotypes. Significant differences in stomatal speed were identified and genotypes with slower stomatal responses were characterized by higher photosynthesis limitations. In two contrasting genotypes, the impact of differential stomatal kinetics was further investigated at the whole-plant level and under diurnally fluctuating light conditions. Genotype-specific stomatal kinetics observed at the leaf level were corroborated at the whole-plant level, validating that genotype-specific responses are still maintained despite differences in stomatal control at different locations of the leaf and across leaves. We discovered that under diurnally fluctuating light conditions the impact of stomatal speediness on photosynthesis and intrinsic water use efficiency depended on the time of the day. During the afternoon there was a setback in kinetics: absolute stomatal conductance and its responses to light were damped, which strongly limited photosynthesis and impacted diurnal water use efficiency. Secondly, we investigated the stomatal and transpiration response under high VPD (Chapter 4). With climate change VPDs are predicted to rise, increasing the crop water demand. Transpiration rate restrictions at high VPD are a relevant water-saving trait for breeding as soil water is retained for later phases of the crop cycle. Eight wild banana subspecies were exposed to step-increases in VPD at both leaf and whole-plant level. All banana genotypes showed significant reductions in their transpiration rate response to increasing VPD. The VPD at which these significant transpiration rate reductions occurred ranged across genotypes between 1.6 and 2.5 kPa. Three subspecies, Musa acuminata ssp. errans, M. acuminata ssp. zebrina and M. balbisiana, showed significantly higher stomatal closure at high VPD, which resulted in higher limitations in photosynthesis. Musa acuminata ssp. zebrina and M. balbisiana showed strong reductions in stomatal conductance with increasing VPD, while M. acuminata ssp. errans showed an overall reduced stomatal conductance. The reductions in transpiration rate measured at the leaf level were in line with these measured at the whole-plant level. These dynamic transpiration responses to light intensity and VPD, were combined with decreasing soil water potential in a high-throughput phenotyping experiment under fluctuating greenhouse conditions (Chapter 5). By applying continuous high-throughput phenotyping in combination with continuous environmental monitoring, we were able to construct genotype-specific transpiration models in response to light, VPD and soil water potential. All investigated wild (sub)species already showed stomatal conductance reductions at a relatively high soil water potential, but significant genotypic differences in the onset and intensity of stomatal closure were observed. Based on their transpiration responses we discerned four phenotypic clusters. Musa balbisiana experienced a stomatal based transpiration rate reduction at a significantly lower soil water content threshold compared to the M. acuminata subspecies. Once this genotype-specific threshold was achieved, the decrease in transpiration rate was highest in M. balbisiana and M. acuminata ssp. errans. Musa acuminata ssp. errans responded to decreasing soil water potential by a transpiration rate reduction at a significantly higher soil water content threshold than all other genotypes. Because of its strong stomatal limitation, Musa acuminata ssp. errans had the highest water use efficiency, but also the slowest growth. The M. acuminata ssp. banksii genotypes clustered together with M. acuminata ssp. microcarpa, and showed high transpiration rates at high light intensities and high VPDs. The M. acuminata ssp. malaccensis genotypes and ssp. burmannicoides, on the other hand, were characterized by the lowest transpiration rate under high light intensities and high VPDs. These phenotypic clusters were largely in line with the genetic characterization and a validation experiment confirmed the phenotypes of M. acuminata ssp. banksii, ssp. malaccensis and M. balbisiana. The contrasting water balance regulations of these three wild banana (sub)species were validated by continuous turgor and leaf movement measurements (Chapter 6). Transpiration rates were combined with leaf movements measured by top view cameras and with turgor dynamics measured by leaf patch pressure probes. Both M. acuminata subspecies showed significant laminae folding under both high VPD and low soil water potential, while these effects were not observed for M. balbisiana. Similarly, both M. acuminata subspecies showed at severe water deficit a significant increase in the time needed for nocturnal turgor recovery, while no increase was observed for M. balbisiana. These responses suggest a better water uptake and transport in M. balbisiana, enabling it to maintain transpiration and growth under high VPDs and low soil water potentials. Finally, we focused on the diversity within a single subspecies, M. acuminata ssp. banksii (Chapter 7). Musa acuminata ssp. banksii is one of the main ancestors of the commonly cultivated banana and is indigenous to Papua New Guinea. Genotypes of M. acuminata ssp. banksii were acquired through two collection missions to Papua New Guinea. The genomic dissimilarity of the collected M. acuminata ssp. banksii genotypes was 4 % and they differed at least 5% from genotypes present in the International Transit Centre, the world's largest banana gene bank. Across two M. acuminata ssp. banksii genotypes from distinct provinces in Papua New Guinea, significant differences in root : shoot ratio and transpiration rate behaviour were observed. These differences suggest that phenotypic plasticity for drought tolerance traits exists not only across wild banana (sub)species, but also within (sub)species. Based on the research performed in this PhD, we conclude that the banana wild species contain diversity in drought tolerance traits, both within and across (sub)species. These results highlight the importance of their conservation and their phenotypic evaluation for breeding programs. Future research should address the impact of these traits by crop modelling and field experiments. Alleles underlying these traits should be identified to increase the relevance for breeding. This PhD forms a first step in the identification of drought tolerance sources in wild bananas.