1Cooperative Research Centre for Tropical Plant Pathology, University of Queensland, St Lucia, 4072 Australia; 2Department of Botany, University of Queensland, St Lucia, 4072 Australia; 3Horticulture Institute, DPI, Indooroopilly, 4068 Australia; 4Farming Systems Institute, DPI, Indooroopilly, 4068 Australia

Background and objectives
Fusarium wilt (or Panama disease) is regarded as one of the most significant threats to banana production in the world. The disease is caused by the fungus Fusarium oxysporum f. ;sp. cubense (Foc). Options for management of this disease are limited. Commercially acceptable varieties with resistance to Foc are not widely available and production is often maintained by shifting to new land once populations of Foc reach economically unsustainable levels. Two races of the pathogen are of greatest concern: race 1 which attacks Lady finger, Silk and Pisang Awak cultivars, and race 4 which attacks Cavendish cultivars as well as all cultivars susceptible to race 1. If race 4 becomes established in Central and South America then the world banana export industry may again be under threat, since there are no resistant Cavendish replacements available at present.

Banana plantations where Foc is present but not causing significant loss of production have been described as having 'suppressive soils'. The suppressive nature of these soils may be due to the presence of organisms in the banana roots or rhizosphere. Specific populations of organisms may suppress the pathogen via direct competition in the rhizosphere for root exudates, or by inducing host systemic resistance prior to infection by the pathogen. Previous research into the organisms causing Fusarium wilt suppression in other crops has indicated that nonpathogenic endophytic populations of F. ;oxysporum may provide plants with protection from pathogenic strains [1].

A higher incidence of fusarium wilt has been observed in banana plants derived from tissue culture planting material in comparison with conventional vegetative planting material, i.e. rhizome pieces and suckers. Beneficial isolates may induce resistance in tissue culture plants against pathogenic strains and be applied as a pre-planting treatment. The aims of this study are to isolate endophytic strains of F. ;oxysporum from the roots of banana plants growing in suppressive soils, and to test the potential use of these isolates as a pre-planting treatment for tissue culture plants to boost resistance to Foc.

Materials and methods
Symptomless banana roots were randomly sampled from Australian plantations which were reported to possess fusarium wilt-suppressive soils. Roots were surface-sterilized and endophytic isolates of F> ;oxysporum were cultured from the vascular tissue. Tissue-cultured banana plants (Lady finger and Cavendish cv. Williams varieties) were inoculated with individual isolates of F> ;oxysporum and exposed to Foc 2 ;weeks later. Lady finger plants were inoculated with race 1 Foc and Cavendish cv. Williams plants with race 4 Foc. The protection given by each endophyte was evaluated by the assessment of external and internal symptoms of Fusarium wilt. Both the endophyte and the pathogen were reisolated from plant roots to check infection and persistence of the endophyte.

Results and conclusions
Three isolates of F> ;oxysporum tested provided levels of protection against fusarium wilt for Cavendish cv. Williams plants. These isolates were recovered from the roots of the Cavendish plants. These plants expressed either no symptoms or mild symptoms of fusarium wilt. Since the endophytes were originally recovered and were also reisolated from banana roots, it is possible that the mode of protection provided is via induced systemic resistance as opposed to antagonistic suppression of the pathogen. Further studies are being conducted to ascertain the mode of protection afforded by the endophytic F> ;oxysporum isolates and their potential use as biological control agents of fusarium wilt.

1. Alabouvette C, Lemanceau P, Steinberg C, 1993. Pesticide Science 37, 365-373.