A comprehensive and quantitative analysis of the mycoflora on the surface of commercial fruit was performed. Nine kinds of fruits grown in Japan were tested. Overall fungal counts on the fruits ranged from 3.1 to 6.5 log CFU/g. The mean percentages of the total yeast counts were higher than those of molds in samples of apples, Japanese pears, and strawberries, ranging from 58.5 to 67.0%, and were lower than those of molds in samples of the other six fruits, ranging from 9.8 to 48.3%. Cladosporium was the most frequent fungus and was found in samples of all nine types of fruits, followed by Penicillium found in eight types of fruits. The fungi with the highest total counts in samples of the various fruits were Acremonium in cantaloupe melons (47.6% of the total fungal count), Aspergillus in grapes (32.2%), Aureobasidium in apples (21.3%), blueberries (63.6%), and peaches (33.6%), Cladosporium in strawberries (38.4%), Cryptococcus in Japanese pears (37.6%), Penicillium in mandarins (22.3%), and Sporobolomyces in lemons (26.9%). These results demonstrated that the mycoflora on the surfaces of these fruits mainly consists of common pre- and postharvest inhabitants of the plants or in the environment; fungi that produce mycotoxins or cause market diseases were not prominent in the mycoflora of healthy fruits. These findings suggest fruits should be handled carefully with consideration given to fungal contaminants, including nonpathogenic fungi, to control the quality of fruits and processed fruit products.

In this study, enumeration methods for fungi in foods were evaluated using fruits that are often contaminated by fungi in the field and rot because of fungal contaminants. As the test methods, we used the standard most probable number (MPN) method with liquid medium in test tubes, which is traditionally used as the enumeration method for bacteria, and the plate-MPN method with agar plate media, in addition to the surface plating method as the traditional enumeration method for fungi. We tested 27 samples of 9 commercial domestic fruits using their surface skin. The results indicated that the standard MPN method showed slow recovery of fungi in test tubes and lower counts than the surface plating method and the plate-MPN method in almost all samples. The fungal count on the 4th d of incubation was approximately the same as on the 10th d by the surface plating method or the plate-MPN method, indicating no significant differences between the fungal counts by these 2 methods. This result indicated that the plate-MPN method had a number agreement with the traditional enumeration method. Moreover, the plate-MPN method has a little laborious without counting colonies, because fungal counts are estimated based on the number of plates with growing colonies. These advantages demonstrated that the plate-MPN method is a comparatively superior and rapid method for enumeration of fungi.