The emergence of freckle disease as an economically important disease of banana has necessitated more detailed studies of the causal fungal species within the genus Phyllosticta. Various aspects of pathogen biology including morphological and molecular taxonomy, population genetic structure, morphogenesis and infection events and molecular identification of these causal species are investigated in detail in this study.
First, the identities of species causing freckle disease on banana are investigated. A polyphasic approach of morphological and molecular taxonomy was applied on a collection of primarily infected leaf specimens and some pure cultures. Detailed examinations, measurements and descriptions of various morphological characters of the fungal species revealed the presence of multiple species. Nucleotide sequences of the internal transcribed spacer (ITS) rRNA and the LSU rRNA genes identified three different, but closely related, species of Phyllosticta causing freckle disease as well as the presence of an endophytic species, Phyllosticta capitalensis on banana. A novel species, named Phyllosticta cavendishii, showed pathogenicity towards both Cavendish and non-Cavendish cultivars of the AAA, AAB and ABB genomic groups. A new name, Phyllosticta maculata was introduced for Guignardia musae and this species only infects non-Cavendish cultivars of the AAB and ABB genomic groups. Both newly named species have a wide geographic distribution. The third species, Phyllosticta musarum which infects only the AAB and ABB genomic groups has a more restrictive distribution.
The population genetic structure of the three Phyllosticta species was investigated through analysis of a collection of 100 samples belonging to three host genomic groups, AAA, AAB and ABB from four geographic regions, Southeast Asia, Oceania, Australia and India/Sri Lanka. DNA sequence data of the ITS rRNA, actin and translation elongation factor 1-α genes was used for this analysis. Allelic networks based on this data showed clustering of the three species in the ITS but alleles for geographic region and host genomic group were shared in all loci. Common alleles shared across regions were in accordance with historical gene flow. Hierarchical analysis revealed significant genetic variation among populations irrespective of species, geographic region and host genomic group. The genetic variance among individuals within population contributes to most of the diversity (P ≤ 0.001). Significant geographic region and host genomic group structuring (P ≤ 0.001) reflect contemporary restricted gene flow. Both Oceanic and Australian populations were possibly introduced in multiple events from Southeast Asia, the likely centre of diversity and centre of origin of the three Phyllosticta species.
The ultra-structural changes and mode of infection of P. maculata on banana plant surfaces after artificial inoculation were studied in sequential stages using scanning electron microscopy. Germination of conidia on the leaf surface commenced within 3 h after inoculation, producing long and slender germ tubes. At the apex of germ tubes, appressoria of various shapes were formed within 18 h after inoculation. Conidial germination on the fruit surface also occurred within 3 h post-inoculation, but forming predominantly short, swollen germ tubes that functioned as lateral appressoria. The pathogen penetrated the leaf epidermis directly, with or without appressoria, and no stomatal penetration was apparent. Direct penetration was further evident by wax dissolution or enzymatic degradation of the plant cuticle around the penetration site. Examination of naturally infected leaf lesions showed each lesion was studded with groups of pycnidia surrounded by prolific conidia. The pycnidial cavity was found to be lined with conidiogenous cells that bear conidia. These asexual spores are one of the sources of inoculum, mainly responsible for secondary infection of banana leaves and fruit.
The final part of the thesis describes the development of a high resolution melting (HRM) analysis assay to identify and differentiate the three Phyllosticta species – P. musarum, P. maculata and P. cavendishii in a rapid single assay. A specific primer pair for the assay was designed based on the ITS of the nuclear rDNA operon. The specificity of the assay was tested and no non-specific amplification of other Phyllosticta species nor other foliar fungal pathogens of banana or various host cultivars occurred. The assay reliably detected the presence of pathogen DNA down to 0.1 ng in a background of host plant DNA. The assay generated three distinct melt profiles representing each of the three Phyllosticta species, and a deviant melt curve for a mixed DNA sample of either any two species or all three species combined, with two and three melting domains, respectively. Validation of the assay was conducted using 35 known primarily infected leaf specimens and in all cases the assay correctly identified and differentiated the species involved. Further validation was conducted by testing 18 uncharacterized infected leaf specimens and in all cases the correct species was identified as confirmed by subsequent nucleotide sequencing. The advantages of HRM analysis in genotyping fungal plant pathogens were highlighted.
By presenting two new identities of Phyllosticta species involved in banana freckle disease, providing insights into the genetic structure of these pathogen populations, unravelling the process and mode of penetration, in addition to developing a molecular assay for distinguishing the three fungal species, the research findings in this thesis will have significant implications in future management of freckle disease in areas such as the implementation of quarantine measures and breeding and selection for disease resistant cultivars.