Distribution and genetic diversity of Pseudocercospora SPP. associated with banana Sigatoka diseases in East Africa

Thesis (PhDFoodSc)--Stellenbosch University, 2020.

ENGLISH ABSTRACT: Sigatoka leaf diseases are a major constraint to banana production worldwide. They are caused by phylogenetically related pathogens belonging to the genus Pseudocercospora. Pseudocercospora fijiensis, the cause of black Sigatoka, is the most widespread and damaging species, causing yield losses of 20-50%. Pseudocercospora fijiensis is heterothallic, and produces infective conidia and ascospores that are dispersed by wind and rain splash. In commercial farms, black Sigatoka is managed by spraying fungicides weekly. This method is not suitable for smallholder farmers who represent most banana producers in Africa. Banana varieties with resistance to black Sigatoka is the most feasible control method for resource poor farmers. An understanding of pathogen distribution, genetic diversity and population dynamics is a prequisite for developing effective and sustainable disease management strategies. A survey was conducted in five banana-growing regions in Tanzania and Uganda to identify the Pseudocercospora spp. associated with Sigatoka leaf spots and determine disease severity. Sigatoka-like symptoms were present in all localities and on all cultivars. Species-specific primers revealed that P. fijiensis was the predominant species in all areas except Kilimanjaro, where Mycosphaerella musae was associated with Sigatoka- like leaf spots. Black Sigatoka was more severe in Uganda, with a mean disease severity index (DSI) of 37.5%, than in Tanzania (DSI=19.9%). Pseudocercospora fijiensis was detected at altitudes of up to 1877 m above sea level, which suggests a habitat range expansion from the previous threshold of <1350 m above sea level in East Africa. This expansion threatens sustainability of banana production in the region. Genetic diversity, population structure and mating type idiomorph distribution was assessed on 319 P. fijiensis single-spore isolates from seven regions, using 16 simple sequence repeat markers and mating type (MAT)-specific primers. The populations were characterised by a high genotypic diversity (296 multi-locus genotypes) and low clonality (7%), with MAT 1 and 2 occurring at a 1:1 ratio in Uganda, while MAT 1 was over- represented at a ratio of 4:1 in Tanzania. The index of association revealed that all populations were at linkage equilibrium (P>0.05), supporting the hypothesis of a random association of alleles. Sub-populations had a moderate level of genetic diversity (Hexp = 0.12-0.31; mean 0.29). These findings are consistent with a pathogen that reproduces both clonally and sexually. Isolates collected at the different locations did not show geographical differentiation, with 90% of the variation occurring among isolates within a subpopulation. This finding suggests a common origin for the isolates and supports the hypothesis of frequent recombination of genotypes. Multi-location evaluation of 21 newly developed East African Highland banana hybrids (NARITA) for resistance to P. fijiensis was conducted in five sites in Uganda and Tanzania. Significant differences in disease severity was observed between the hybrids, test locations, and their interaction (GEI). The environment had the greatest influence (39.1%) on genotypes’ response to P. fijiensis, with GEI accounting for 23.4%. Most of the hybrids exhibited broad adaptability in their response to black Sigatoka. Hybrids with low disease development and a stable response across locations were NARITA hybrids 2, 7, 8, 21 and 23. These can be provided to farmers for managing black Sigatoka in the region. NARITA hybrids 10 and 18 were identified as susceptible, and could be used as susceptible checks in future evaluations. The Mitarula site in Tanzania was identified as a representative test location to evaluate banana hybrids for their response to the black Sigatoka pathogen. To identify potential sources of P. fijiensis resistance, a collection of 95 banana accessions, including selected breeding parents, were evaluated in the field at Sendusu in Uganda. Out of these, 33% of the accessions; belonging to 22 subspecies of Musa acuminata ssp. malaccensis, M. acuminata ssp. zebrina and M. acuminata ssp. Burmannica; were either resistant or partially resistant to P. fijiensis. Symptom progression in these accessions stopped at early lesion development (Stages 2, 3, and 4). Symptom development in Long Tavoy, Pahang, Pisang KRA, 0074 Malaccencis, M.A Truncata, Tani and Balbisiana stopped at Stage 2, like that in the resistant Musa acuminata ssp. burmannicoides, var. Calcutta 4, and these varieties can thus be considered as potential sources of resistance.

AFRIKAANSE OPSOMMING: Sigatoka blaarsiektes beperk piesangproduksie wêreldwyd. Die siektes word deur filogeneties-verwante patogene in die genus Pseudocercospora veroorsaak. Pseudocercospora fijiensis, die oorsaak van swart Sigatoka, is die mees wyd-verspreide en skadelike spesies, en veroorsaak oesverliese van 20-50%. Pseudocercospora fijiensis is heterotallies en produseer konidia en askospore wat deur die wind en spattende reën versprei word, en infeksie kan veroorsaak. In kommersiële plase word swart Sigatoka deur die weeklikse spuit van fungisiedes bestuur. Hierdie metode is nie geskik vir kleinboere wat die grootste deel van piesangprodusente in Afrika verteenwoordig nie. Piesangvariëteite met weerstand teen swart Sigatoka is die mees uitvoerbare beheermetode vir hulpbron-arm boere. Kennis van patogeen-verspreiding, genetiese diversiteit en populasie-dinamika is ʼn voorvereiste vir die ontwikkeling van effektiewe en volhoubare siektebestuurstrategieë. ʼn Opname is in vyf piesang-produserende streke in Tanzanië en Uganda uitgevoer ten einde Pseudocercospora spp. wat met Sigatoka blaarvlekke geassosieer word, te identifiseer, en siekte-intensiteit vas te stel. Sigatoka-agtige simptome was in alle lokaliteite en op alle kultivars teenwoordig. Spesies-spesifieke inleiers het getoon dat P. fijiensis die oorheersende spesies in alle areas was, behalwe in Kilimanjaro, waar Mycosphaerella musae met Sigatoka-agtige blaarvlekke geassosieer is. Swart Sigatoka was meer ernstig in Uganda, met ʼn gemiddelde siekte-intensiteit-indeks (DSI) van 37.5%, as in Tanzanië (DSI=19.9%). Pseudocercospora fijiensis is op hoogtes van tot 1877 m bo seevlak waargeneem, wat op ʼn habitatreeks uitbreiding van die vorige drumpelwaarde van <1350 m bo seevlak in Oos-Afrika, dui. Hierdie uitbreiding bedreig volhoubaarheid van piesangproduksie in die streek. Genetiese diversiteit, populasie-struktuur en paringstipe idiomorf verspreiding is op 319 P. fijiensis enkelspoor isolate vanaf sewe streke bepaal, deur gebruik te maak van 16 eenvoudige basispaar-volgorde herhaling merkers en paringstipe (MAT)-spesifieke inleiers. Die populasies is deur ʼn hoë genotipiese diversiteit gekarakteriseer (296 multi- lokus genotipes) en lae klonaliteit (7%), met MAT-1 en -2 wat teen ʼn 1:1 verhouding in Uganda voorkom, terwyl MAT-1 óór-verteenwoordigend teen ʼn verhouding van 4:1 in Tanzanië was. Die indeks van assosiasie het getoon dat alle populasies by ‘linkage equilibrium’ (P>0.05) was, wat die hipotese van ʼn ewekansige assosiasie van allele ondersteun. Sub-populasies het ʼn matige vlak van genetiese diversiteit gehad (Hexp = 0.12-0.31; gemiddelde 0.29). Hierdie bevindinge is konsekwent met ʼn patogeen wat beide klonaal en geslagtelik reproduseer. Isolate wat by die verskillende liggings versamel is, het nie geografiese differensiasie getoon nie, met 90% van die variasie wat tussen isolate binne ʼn sub-populasie voorkom. Hierdie bevinding dui op ʼn algemene oorsprong vir die isolate en ondersteun die hipotese van gereelde rekombinasie van genotipes. Die multi-omgewing evaluasie van 21 Oos-Afrika Hoogland piesanghibriede (NARITA) vir weerstand teen P. fijiensis is in vyf plekke in Uganda en Tanzanië uitgevoer. Betekenisvolle verskille in siekte-intensiteit is tussen die hibriede, toetsomgewings, en hul interaksie (GEI), waargeneem. Die omgewing het die grootste invloed (39.1%) op genotipe se reaksie op P. fijiensis gehad, met GEI wat 23.4% uitgemaak het. Meeste van die hibriede het wye aanpasbaarheid in hul reaksie op swart Sigatoka getoon. Hibriede met lae siekte-ontwikkeling en ʼn stabiele reaksie óór liggings, was NARITA hibriede 2, 7, 8, 21 en 23. Hierdie kan aan boere verskaf word vir die bestuur van swart Sigatoka in die streek. NARITA hibriede 10 en 18 is as vatbaar geïdentifiseer, en kan as vatbare kontroles in toekomstige evaluasies gebruik word. Mitarula in Tanzanië is as ʼn verteenwoordigende toetsligging geïdentifiseer vir die evaluasie van piesanghibriede vir hul reaksie op die swart Sigatoka patogeen. Ten einde potensiële bronne van P. fijiensis weerstand te identifiseer is ʼn versameling van 95 piesang inskrywings, insluitend geselekteerde teel-ouers, in die veld by Sendusu in Uganda geëvalueer. Uit hierdie was 33% van die inskrywings, wat tot 22 subspesies van Musa acuminata ssp. malaccensis, M. acuminata ssp. zebrina en M. acuminata ssp. Burmannica behoort, weerstandbiedend of gedeeltelik weerstandbiedend teen P. fijiensis. Simptoom-ontwikkeling in hierdie inskrywings het by vroeë letsel- ontwikkeling (Fases 2, 3 en 4) gestop. Simptoom-ontwikkeling in Long Tavoy, Pahang, Pisang KRA, 0074 Malaccencis, M.A Truncata, Tani en Balbisiana het by Fase 2 gestop, soos in die weerstandbiedende Musa acuminata ssp. burmannicoides, var. Calcutta 4. Hierdie variëteite kan dus as potensiële bronne van weerstand beskou word.

Doctorate