Food safety and security of sago starch in rural Papua New Guinea

Sago starch is an important source of dietary carbohydrate in lowland and coastal areas of Papua New Guinea (PNG). There have been sporadic reports of severe haemolytic illness resulting from sago starch consumption, termed sago haemolytic disease (SHD), with most reports coming from the Western Province. Despite the occurrence of SHD, and a high likelihood of less severe foodborne illness resulting from consumption of indigenous foods in general in PNG, there have been no detailed studies of the microbiology of sago starch. The aim of this study was to establish a broad basal knowledge of the microbiology of traditionally prepared sago starch in PNG. Sago starch samples and sociological information were collected in two of the main sago eating regions of PNG, the East Sepik Province and the Western Province. Sago starch samples were collected predominantly from the houses of sago starch producers in rural areas of the two provinces, and to a lesser degree from markets in some villages in the East Sepik Province. In addition to these samples considered ‘fit for consumption’, two samples of sago starch that had been associated with outbreaks of SHD were also analysed. Analysis of the sago starch for common bacterial pathogens was done using accepted methods, and where possible was based on the relevant Australian Standards. The findings suggest that faecal contamination of sago starch is widespread, with over three-quarters of all samples tested for faecal coliforms at the upper limit of detection. The human pathogen Salmonella spp. was isolated from approximately 7% of samples tested. The presence of emerging human pathogens such as Citrobacter freundii and Enterobacter sakazakii was tested, with the former being present in a low percentage of samples tested. Other important bacterial food pathogens such as Staphylococcus aureus and Bacillus cereus were also detected in sago starch, but none of the 57 samples tested for Listeria monocytogenes was positive. Mycological analysis of sago starch revealed a variety of fungal contaminants. Commonly occurring genera of filamentous fungi included Penicillium, Scytalidium, Aspergillus, and Acremonium. Mycotoxin analysis of sago starch revealed that the common mycotoxins such as aflatoxins, ochratoxin A, cyclopiazonic acid, sterigmatocystin, zearalenone and citrinin were not present. Selected fungal isolates were tested for the presence of mycotoxin production in pure culture, with two-thirds found to be capable of citrinin production and one isolate capable of sterigmatocystin synthesis. In an attempt to determine the aetiological agent of SHD, bacterial and fungal isolates were screened for haemolytic activity on blood agar. A suitable semiquantitative assay was developed, and extracts from bacterial and fungal cultures were tested. Particular attention was paid to the haemolytic activity of fungal extracts, given the long speculated role of mycotoxins in the aetiology of SHD. The haemolytic activity of numerous fungal species has been demonstrated for the first time, and steps in the optimised extraction and purification of the haemolytic component of some isolates has been completed. Further work was conducted on extracts from Penicillium steckii, a common contaminant of sago starch. The chemical properties of the extract suggest that a novel fungal metabolite is responsible for haemolytic activity. Detailed studies on the microbial ecology of stored sago starch have been conducted, primarily to garner a greater understanding of factors that contribute to the microbial safety of the food. The study has established that sago starch is a naturally fermented product, and this fermentation process contributes greatly to the food safety of the product. Bacterial pathogens such as B. cereus, L. monocytogenes, S. aureus and Salmonella sp. do not survive well in actively fermenting sago starch, primarily due to the production of weak acids. Furthermore, numbers of filamentous fungi are low in actively fermenting sago starch, presumably due to the reduced oxygen availability. On the basis of the sociological data and microbial studies, a hazard analysis and critical control point (HACCP) plan was devised that was considered appropriate for application in village based production of sago starch in rural PNG. Through a greater understanding of the microbiology of sago starch, together with the development of an appropriate HACCP plan, this research will lead to increased food safety and food security for sago consumers in rural and remote lowland areas of PNG. Moreover, studies of the haemolytic metabolites of fungi isolated from sago starch pave the way for further research to determine the aetiology of SHD