Cultural methods for the detection and quantification of mould propagules in food and food raw materials have long been used (Jarvis et al., 1983; Beuchat, 1987; Gourama and Bullerman, 1995; Pitt and Hocking, 2009; Samson et al., 2010). They make use of the fact that viable moulds and yeasts can be detected and counted after cultivation of a sample on microbiological culture media, either as numbers of contaminated individual particles or as colony-forming units. In addition to mere detection and enumeration of fungal contamination, exact identification of the species is of prime importance because species identification may provide an indication of possible quality problems associated with the investigated sample. This is particularly important when assessing a possible mycotoxin contamination. Microbiological testing methods have the advantage that they can be easily handled in the laboratory without much equipment, other than a microscope. They can be applied to a variety of different materials to be tested.
The microbiological growth media applied in fungal analysis of food can be divided into general growth media for detection of a wide range of fungi and yeasts and selective media allowing growth of a restricted number of species. Besides the choice of the medium, selective growth of certain species or groups of species can be achieved by the choice of proper incubation conditions. As an example, an incubation temperature of 37°C is applied for the selective cultivation of human pathogenic yeasts from food samples using a non-selective culture medium such as malt extract agar or YPG where the high incubation temperature leads to exclusion of mesophilic yeasts, most of which are non-pathogenic to humans. Also for the investigation of the presence of heat-resistant moulds in pasteurized foods, the selection is done through a pre-treatment of the sample at 70°C before incubation of cultures at elevated temperatures rather than by the use of a selective growth medium (Pitt and Hocking, 2009). Today, non-selective detection and enumeration of a broad spectrum of moulds and yeasts from food sources and from the air is routinely done on Dichloran Rose Bengal Chloramphenicol agar (DRBC agar, King et al., 1979). The medium is selective for ascomyceteous and basidiomycetous filamentous and yeast fungi but growth of bacteria and Zygomycetes is largely suppressed by additives. For the selective investigation of xerophilic fungi as an important group of food- borne and airborne fungi, Dichloran 18% Glycerol agar is often used (DG18 agar, Hocking and Pitt, 1980). DRBC and DG18 have been certified by the International Organization for Standardization (ISO) for enumeration of yeasts and moulds in high water activity food and animal feedstuff's (aw > 0.95, ISO 21527-1) and low water activity foods and animal feedstuff's (aw < 0.95, ISO 215272), respectively. However, both culture media are better suited for general growth and enumeration than for identification of fungi because they do not properly develop the micro- and macro-morphological features that are needed for identification. As a consequence, subcultures have to be prepared on optimal media in order to identify an isolate to the species level. In regard to morphological identification of species, it is particularly important that the morphological features of the moulds are properly expressed. Therefore, different media are used for enumeration, isolation, or identification. Complex media such as malt extract agar, oatmeal agar, maize meal agar, potato dextrose agar are often used for cultivation but also synthetic media, such as SNA have been used and are superior for some groups of fungi. Some species or genera may also need specific media such as clove leaf agar, Czapek yeast autolysate agar or cherry decoction to form the characteristic structures that are important for their identification. Many species have to be incubated at different growth conditions or incubated under UV light or in darkness in order to bring about the typical morphological structures needed for their identification.
In studies focusing on detection and enumeration of single fungal species or groups of physiologically similar species, selective media may provide a tool to circumvent extensive isolation for species identification. Selective growth conditions are created both by specifying certain growth parameters such as aw, pH, salinity or sugar content. Moreover, the selection of certain fungi on the growth medium can also be affected by addition of substances that kill unwanted organisms, or at least strongly inhibit their development. An example for this is addition of substances such as iprodione (Abildgren et al., 1987), dichloran (Andrews and Pitt, 1986; Conner, 1992), rose Bengal (Newhouse and Hunter, 1983) or penta- chloro-nitorbenzole (PCNB, Nash and Snyder, 1962; Nishikawa and Kohgo, 1975; Gyllang et al., 1981; Burgess et al., 1988) to general growth media such as Szapek Dox agar in order to create selective conditions for the examination of Fusarium infestation in cereals. Fusarium spp. are highly resistant to both compounds and can therefore be selectively enumerated in a genus-specific manner. However, isolations still have to be made for species identification. Based on PCNB agar, further development led to species-selective media that can be applied to investigate either F. graminearum (mannitol-PCNB agar, Bohm-Schraml et al., 1993) or F. culmorum (malachite green agar, Bohm-Schraml, 1995) in brewing cereals and malt. The use of dichloran rose bengal yeast extract sucrose agar (DRYS, Frisvad, 1983) or dichloran yeast-extract sucrose 18% glycerol agar (DYSG, Elmholt et al., 1999) proved to be a valuable tool in the microbiological analysis of food commodities for mycotoxin-producing Peni- cillium spp. The latter medium is even useful in the selective differentiation of Penicillium verrucosum from P nordicum, both producers of ochratoxin A. AFPA agar is useful for the selective identification of aflatoxin-producing species such as Aspergillus flavus and A. parasiticus. Aflatoxinogenic species can be recognized by their typical orange-red colour when colonies are observed from their reverse (Pitt et al., 1983).