Growth of aflatoxigenic moulds and aflatoxin formation in Brazil nuts

The present study aimed at gaining more knowledge of the growth of aflatoxigenic moulds and aflatoxin production in Brazil nuts in relation to humidity conditions and storage time. For this purpose, the growth of aflatoxigenic moulds and the increase in aflatoxin levels in Brazil nuts was studied in the laboratory at temperature and humidity conditions that are relevant for the Amazon region. Fresh unprocessed Brazil nuts in shell were inoculated with an aflatoxin producing strain of Aspergillus nomius previously isolated from Brazil nuts. The nuts were stored at 27 °C in combination with 97, 90 or 80% surrounding relative humidity in a respirometer for up to 3 months. The General Linear Model (GLM) was used for evaluation of the effect of water activity and time on aflatoxigenic mould levels and on aflatoxin levels, as well as the relationship between mould and aflatoxin levels. During storage at the highest relative humidity (97%) aflatoxin formation occurred rapidly, whereas storage at 90% relative humidity resulted in slower aflatoxin formation. At the lowest relative humidity (80%), aflatoxin formation occurred sporadically during storage. The increase in mould and aflatoxin levels along the production chain is also described, using field data collected in the state of Para, Brazil. The growth of aflatoxigenic moulds and aflatoxin formation increased rapidly between 40-90 days following collection of the nuts, before the nuts reached the final drying stage at the processing plant. In addition, a logistic regression model predicting the probability that the European legislative limit of 4 µg/kg for aflatoxins in nuts will be exceeded in relation to colony counts of either one selected aflatoxigenic mould strain (laboratory experiments) or of a mixture of aflatoxigenic strains (field data) was developed. The probability that total aflatoxin levels will exceed the European legislative limit of 4 µg/kg increased rapidly from approx. 30% to above 80% for both experimental and field samples at mould levels between 2 and 3 log cfu/g.