1CRC for Tropical Plant Pathology, University of Queensland, QLD 4072, Australia; 2Plant Protection Unit, QDPI, Meiers Rd, Indooroopilly, QLD 4068, Australia.

Background and objectives
The ascomycete fungus Mycosphaerella fijiensis is the causal agent of black Sigatoka, the most destructive foliar disease of banana and plantain. M. ;fijiensis is present on many islands of the Torres Strait, which separates mainland Australia from Papua New Guinea (PNG) by 150 ;km. Australian banana plantations are free of black Sigatoka, but threatened by continual disease outbreaks, in remote regions of Queensland, and close to the major banana production region of Tully-Innisfail. Since this disease was first discovered in Australia in 1981, exclusion and quarantine measures have stopped disease establishment. The disease currently exists in the Torres Strait and PNG, one of the hypothesised centres of origin of M. ;fijiensis, perhaps providing a continual inoculum source for disease epidemics on the Australian mainland.

M. ;fijiensis isolates were collected from five islands in the Torres Strait in 1996 with the aim of examining the genetic structure of populations in this region and their relationship to M.fijiensis populations from the nearby banana growing regions of PNG and the Pacific, previously examined in a global population study [1].

Results and conclusions

A preliminary analysis of Southern blots was done for four probes before further probing continued. Three populations: Murray, Badu and the group of Boigu, Saibai and Dauan, were identified within the isolates collected from the Torres Strait on the basis of allele and genotype distribution. Isolates from PNG and the Pacific were identified as different populations in the global study [1] and were distinguishable as different from the three Torres Strait populations.

Alleles were defined by the DNA restriction fragments present with each probe-enzyme combination. Three of the four single-copy probes revealed polymorphisms with 2-5 alleles per RFLP locus. Multilocus genotypes from haplotype data, were analysed using the NT-SYS computer program. Allele frequency was calculated as well as gene diversity (H) and Gst [2].

Twenty-two genotypes were identified for the 159 isolates examined. Seven genotypes (31%) occurred only once in the populations examined. The high number of genotypes is as expected for a heterothallic, sexually reproducing species such as M. ;fijiensis. Several genotypes were widespread, occurring in the PNG, Pacific and Murray populations (4%), shared between PNG and three Torres Straits populations (31%), or shared between Pacific and Murray populations (9%). Uncommon genotypes that appeared region specific were identified for the populations of PNG (18%), Murray (22%) and Pacific (9%).

Gene diversity (H) was highest in the PNG population (H=0.469), but also very high in the Murray population (H=0.433). Lower values were observed for the Pacific (H=0.377), and Badu (H=0.331) populations. The Boigu group population had a very low gene diversity, H=0.131, a result of some loci being fixed or near fixed. Gst for the whole population was moderate (Gst=0.453), indicating substantial variation in both sub-populations and the entire population.

Genotypic diversity was sample size corrected as G/N, to calculate the percentage of maximum possible diversity. PNG had the highest genotypic diversity (G/N=0.37) but all other populations were much lower, ranging from G/N=0.13-0.19.

Of the Torres Straits populations, diversity was greatest in the Murray population. It has been hypothesized that M. ;fijiensis spread from its south-east Asian centre of origin to other regions including the Pacific as discrete founder events [1]. Genotype distribution of M. ;fijiensis in this research shows how the disease could have spread through the Torres Strait to the Pacific. There is also evidence of past or low level current gene flow between the populations of PNG and Murray Island and between the Pacific and Murray Island populations. This is being investigated further in work now in progress.

1. Carlier J, Lebrun MH, Zapater MF, Dubois C, Mourichon X, 1996. Molecular Ecology 5, 499-510.
2. Nei M, 1973. Proceedings of the National Academy of Sciences USA 70, 3321-3323.