Microbial contaminants associated with sago processing and storage in Papua New Guinea

• Queensland Department of Primary Industries and Fisheries, Australia
• University of Technology, Papua New Guinea
• University of Papua New Guinea, Papua New Guinea

Sago harvesting and consumption form an important part of the staple diet in selected areas of Papua New Guinea (PNG). However, this resource (in excess of 1 million hectares) is under-utilised. There is an estimated market for sago starch in PNG in excess of 12,000 tonnes (some of which is presently imported), and part of the international trade in starch, (15,000-20,000 tonnes into Australia alone) at $A150-200 per tonne could be filled by PNG sago. However the ability to satisfy these markets with high quality starch, essentially free of mycotoxins and food borne pathogens is not assured.

Contaminant problems relate to the presence of mycotoxins in the starch, which may lead to sago haemolytic disease. This is prevalent in the main sago growing areas in the south of the country; however, its cause has never been resolved. Contamination of food with pathogenic bacteria and viruses is likely to be of a widespread nature, since the incidence of diarrhoeal disease is extensive and caused by some of the bacterial agents associated with sago contamination (determined by experimental sampling from selected market sites). The extent and seriousness of sago contamination cannot be fully understood until a more extensive investigation is mounted. Both the definition of contamination problems (fungal and bacterial) and the development of suitable solutions are of equal importance to the development of PNG sago as a village food and as a trade commodity.

This project determined the causes and extent of health risks to consumers caused by contamination in village-produced sago in Papua New Guinea, to identify options for reducing those risks and improving marketability of the processed sago.

This two-year project assessed the microbiological problem at a spectrum of field and market sites in the Sepik, Western and Gulf provinces of PNG, with targeted training components for scientists working in PNG.

Scientists collected field data on harvesting, preparation and storage of sago in the principal sago districts. Samples of good and inferior quality sago were collected at different periods of the year. These samples were analysed for the presence of fungi, yeasts and bacteria, and the fungi, yeast and total bacterial loads were enumerated.

Particular attention was given to what fungal species were involved, in order to identify potential mycotoxin producers. Selected isolates were cultured to monitor their potential for producing mycotoxins under different environmental conditions. At the same time, field samples were analysed for a range of mycotoxins, moving well beyond the small group for which assays had been previously conducted. The analysis looked for indicator mycotoxins such as aflatoxins, but moved to other groups that had not been considered previously or which had been recently identified as of significance to human and animal health.

The same food samples were analysed for the presence of at least four significant bacterial pathogens involved in diarrhoeal diseases and abortions. These included Salmonella and Listeria, organisms of considerable interest in international trade negotiations that will also be of increasing significance in domestic marketing arrangements as development proceeds.

The project collected data on the techniques in use in two regions of PNG (Sepik and Western Provinces) for the production of sago starch, identified potential bacterial agents of foodborne illness associated with these products, and assessed the degree of fungal contamination and potential for mycotoxin production in these products - to attempt to elucidate the causative agent/compound responsible for sago haemolytic disease, and to train PNG scientists in the culture of foodborne pathogens and fungi using Australian Standard Methods for the detection of mycotoxins. These data will be used to inform production processes for sago starch and to develop recommendations for reducing sago health risks, leading to improvements in the health of consumers and the development of an export industry.

Tandem studies of storage method and contaminants together with the HACCP analyses indicated that fermentation and storage of sago in palm leaf bundles (in Western province) and clay pots (in East Sepik) produced product of relatively low contaminant risk when consumed within a few weeks. By contrast storage in baskets resulted in inferior fermentation and greater risk of contamination from soil or faeces and subsequent chance of food poisoning or SHD.

Other outcomes included:
The project identified a potential novel compound from a fungal isolate derived from sago starch that shows strong haemolytic activity with human red blood cells.
A simple process using washed red blood cells was developed to assay novel compounds derived from fungal cultures following extraction for haemolytic activity.
PNG scientists were trained in the formal culture and identification of a range of bacteria and fungi of public health significance, as well as the analysis of foods for mycotoxins.

Further work is now needed to characterise the novel compound isolated, demonstrate that Koch's postulates can be verified with this compound, and complete the analysis of other fungal isolates for novel compounds with haemolytic capability.

The project team also organised a symposium in Port Moresby during September 2005 that involved team members, health workers, provincial and national government representatives, the Governor of Western province and the member for Middle Fly to report project findings and those of complementary medical research and propose recommendations for reducing sago health risks.