Trends in Food Science & Technology 10 (1999) 356±365

Review

The prevalence and

control of spoilage
ke®r, kuomiss, some raw milk smear ripened cheeses,
etc.). However, Fleet [3], mentioning works of the
1980s, refers for the ®rst time, to the need to re-evaluate

yeasts in foods and the impact of food-borne yeasts on public health/safety.

This author cites occasional reports of gastro-enteritis
caused by the ingestion of foods in which the suspected
beverages etiological agents were yeasts. In addition, a report is
cited referring to increasing evidence for the develop-
ment of allergies and negative e€ects in humans due to
yeasts, which has led to the coming up of lay literature
promoting the virtues of `yeast-free' diets. Thomas [4]
V. Loureiro and A. Queroly also cites a reference stating that the exposure to anti-
gens from Saccharomyces cerevisiae may contribute to

Departamento de BotaÃnica e Engenharia BioloÂgica, in¯ammatory bowel disease. In a recent review on new
Instituto Superior de Agronomia, Universidade and emergent yeast pathogens, Hazen [5] states, as well,
TeÂcnica de Lisboa, 1349-017 Lisboa Codex, Portugal that in the medical ®eld, the infections caused by yeasts
(1fax: 351 21 363 50 31; e-mail: vloureiro@isa.utl.pt) are increasingly signi®cant, and the author attributes such
y
Departamento de Biotecnologia, Instituto de situations to, among other factors, immunosuppressive
AgroquõÂmica y Tecnologia de Alimentos (CSIC), P.O. therapeutic regimens, long-term catheterization, broad-
Box 73, 46100 Burjassot, Valencia, Spain (2fax: 34 96 spectrum antibiotic use, and longer survival of immuno-
363 63 01; e-mail: aquerol@iata.csic.es) logically compromised individuals. He states that some
yeasts previously thought as innocuous (e.g. Candida
utilis, Saccharomyces cerevisiae and Candida lipolytica)
Consumers in developed countries have a more critical are capable of damaging the human body. Although the
attitude about what they eat and drink as a consequence of cases reported are rare and occur in extremely peculiar
modern life. Food microbiologists are facing a huge chal- situations, we have to admit the possibility that the
lenge regarding food `freshness' implicit in the consumer's presence of yeasts in foods, either spoilers or not, may
demand for more natural products. Food with less severe assume an increasing protagonism in public health and
processing, that is additive-free, safer, with satisfactory shelf should be a growing concern for microbiologists and
life and easy to prepare is sought because of higher con- food technologists.
sciousness about nutrition and health. Besides the develop- The spoilage yeasts of foods and drinks have also
ment of new methods and preservation concepts, great gained, in recent years, an increasing importance in
advances in the ®elds of chemistry, biochemistry, micro- food technology, being responsible for signi®cant eco-
biology, hygiene and food safety have occurred in recent nomic losses [4]. However, reports on these losses are
decades. But are such developments properly used to still rare. For reasons of commercial con®dentiality, the
respond to the challenge to food mycology? We will sum- incidence and economic cost of industrial outbreaks of
marize the e€ects on food of yeast spoilage activity and yeast spoilage remain unreported [3]. According to this
we will re¯ect on how this activity can be monitored by author and also based on our experience of outbreak
food microbiologists. # 2000 Published by Elsevier Science costs, these are considerably high and involve large legal
Ltd. disputes involving manufacturers, suppliers of raw
materials and packaging, and, often, retailers.
In broad terms, the comprehensive overviews on
The overviews on food spoilage yeasts published spoilage yeasts published in the second half of the 20th
about 30±40 years ago [1,2] reported that food-borne century were focused, essentially, in the qualitative
yeasts had no signi®cance in public health, being regarded description of the main yeast species present in foods
as harmless for the consumers, even when ingested in high and beverages. Barnett et al. [6], listed nearly 120 yeast
amounts through fermented foods (e.g. lambic beer, species, belonging to 30 genera, as being associated with
0924-2244/99/$ - see front matter Copyright # 2000 Published by Elsevier Science Ltd.
PII: S 0 92 4 - 2 24 4 ( 0 0 ) 0 0 02 1 - 2
 V. Loureiro, A. Querol / Trends in Food Science & Technology 10 (1999) 356±365 357

foods, which represented around 25% of those accepted information available should be useful in developing a
by the authors and 50% of the recognized genera. methodology to evaluate the prevalence of yeasts in
According to Pitt and Hocking [7], most of these 120 foods.
species grow poorly, if at all, in properly formulated A rapid scan of the works published on food altera-
foods, as they are intolerant to reduced water activity, tion by yeasts, since the classical review of Ingram [1] to
heat processing or preservatives. These authors consider the state of the art book of Deak and Beuchat [9], leads
that only about 10 species of yeastsÐDekkera brux- to the conclusion that little has changed, in terms of
ellensis, Issatchenkia orientalis, Debaryomyces hansenii, concepts, methodologies and comprehensiveness of the
Kloeckera apiculata, Pichia membranifaciens, Zygo- biological processes involved. From this, a question
saccharomyces bailii, Zygosaccharomyces bisporus, Sac- arises: is the proportion of yeast-spoiled products higher
charomyces cerevisiae, Zygosaccharomyces rouxii, Schizo- today than in the past? In our opinion the answer is not
saccharomyces pombe and Candida holmii (Saccharomyces easy nor, certainly, consensual. On one hand, because of
exiguus)Ðare responsible for the spoilage of foods that the diculty in obtaining information on yeast spoilage
have been processed and packaged according to the outbreaks and, on the other hand, because the qualita-
normal standards of good manufacturing practice tive analysis of the past (or of the present) do not enable
(GMP). a sound evaluation of the true extension of food
Tudor and Board [8] in their excellent comprehensive spoilage or food depreciation caused by yeasts. A recent
overview on food spoilage yeasts considered, besides the example in the wine industry illustrates well this last
`top-ten' spoilage yeasts listed by Pitt and Hocking [7], a statement. The yeasts of the genus Brettanomyces/
group of `second-division yeasts' constituted by 18 spe- Dekkera have been known for a long time as o€-
cies and one genus of common food contaminant yeasts, ¯avour producers. However, due to their slow growth
which can be opportunist spoilers. These authors con- they were seldom detected in wine microbiological con-
sidered it important to investigate the interactions trol, being regarded as secondary-spoilers or, more
between these species and acid-tolerant spoilage bac- properly, rarely-spoilers. When Chatonnet et al. [10]
teria, considering the production of a wide range of raw identi®ed these yeasts as the main producing agents of
products such as ready-to-use salads, dairy desserts, etc. `horse sweat' taint in red wines, because of their ability
Fleet [3], in his innovative overview, concluded that to produce 4-ethylphenol from p-coumaric acid, the
yeasts occur and grow in a greater range of products detection of Brettanomyces/Dekkera and the analysis of
than was previously thought. He emphasized that further 4-ethylphenol became much more important. As a
advances in understanding the spoilage yeast activities result, these yeasts became unexpected protagonists in
would require a more quantitative approach to study wine microbiology. Nowadays, these yeasts are recog-
their occurrence and growth in speci®c products, and nized as responsible for the depreciation of a signi®cant
more detailed investigations of their biochemical trans- proportion of red wines, particularly those produced
formations. This author compiled, for the ®rst time, the under poor hygiene conditions and/or matured in oak
levels of incidence of yeast contaminants from di€erent barrels, which are a well-known ecological niche for
food commodities, referring to the maximum levels these species.
attained, as well as the species involved, when the
environmental conditions favour their growth and the The extent and e€ects of yeast spoilage
symptoms of spoilage begin to come up. Food spoilage caused by yeasts consists in the visible
Deak and Beauchat [9] were the ®rst authors to, based or detectable alteration of physical and sensorial prop-
on the compilation from the vast literature data, have erties of food as a result of their activity. The most
estimated the frequencies of occurrence of yeast species in known alterations occur in acid drinks, with or without
the major types of foods. The estimation of frequencies sugar, and are characterized by abundant gas production,
has been calculated considering three parameters: (1) which may deform or blow up the packages, cloudiness,
the number of types of food from which a given species sediment or pellicle formation, o€-¯avours dominated by a
is detected; (2) the number of times a given species is slight fermentation smell (alcohol, carbon dioxide and
described from foods; and (3) a commonness index esters) and o€-tastes. Deterioration of food and drinks
value, varying from 1 to 8, based on the number of by yeasts may present other e€ects, more or less evident,
known strains of a given species listed in standard according to the type of food (Table 1). However, in
descriptions and/or which have been isolated from certain cases, yeast spoilage is not so easily de®ned,
foods. The estimated frequency of occurrence is then mainly in fermented foods/drinks (e.g. wines, artisanal
expressed as a percentage of the sum of products of the beers, black olives, soy sauce) where the produced
three entries for all species considered [9]. Although the metabolites contribute to the ¯avour and aroma of the
criteria used by the authors may be controversial and products. As Fleet [3] stated there is only a slight line
the purpose of the work was not to evaluate the inci- between what is perceived as either spoilage or bene-
dence of yeasts in spoiled foods, the huge volume of ®cial activity.
358 V. Loureiro, A. Querol / Trends in Food Science & Technology 10 (1999) 356±365

Table 1. Principal spoilage e€ects caused by yeast activity in foods

Spoilage e€ect
Food Product Surface Discolouration Gas Haze/ Sediments Films O€-¯avours/ Texture
growth production cloudiness souring changes
Fresh vegetables x x x x
Brined vegetables x x x x x x
``Ready-to-eat'' vegetables x x x
Fresh fruits x x x x
Fruit juices x x x x x
``Ready-to-eat'' fruits x x x
Mayonnaise, salad dressings x x x x x
Wine, beer x x x x x
Soft drinks x x x x
Confectionery, jams, jellies x x x x x x x
Syrups, honey, fruit concentrates x x x x
Butter, cream x x
Cheeses x x x
Yoghurts x x x
Sliced bread x x x
Unbaked bread dough x x x
Sausages, meat products x x x x

A good example is the former example of 4-ethylphe- environmental conditions made yeasts stronger and
nol production by Brettanomyces/Dekkera yeasts in red quicker than bacteria and moulds. The prevention of
wines. In fact, when this secondary metabolite is present such spoilage processes, with fewer chemicals and less
at levels less than about 400 g/l it does not depreciate severe preservation processes, requires a clear under-
wine quality. It may, even, contribute to wine complexity standing of the problem and for that, the food micro-
by imparting aromatic notes of spices, leather, smoke or biologist should possess tools to evaluate the whole
hunt, appreciated by many consumers. Nevertheless, at microbiological ecosystem. In fact, many of the prob-
levels higher than 700 g/l the wines are clearly depre- lems caused by yeasts results from the much higher
ciated. attention paid to bacteria and moulds, which are more
The growth of yeast populations in food commodities signi®cant in terms of public health. Their inhibition
to levels as high as 105 to 108 cells/g is not, in many opens the way to yeast activity. But have the food
cases, a clear indicator of product deterioration. In microbiologists the tools that allow them to be su-
many cheeses these levels do not cause detectable ciently aware of the microbiological problems of a food
depreciation, as well as in cloudy beers, where they are commodity?
supposed to cause an appreciable aromatic complexity. According to Mossel et al. [11] in a code of good
The inverse situation also occurs, when yeast counts are manufacturing and distribution practices (GMDPs)
less than 105 cells/g and visible spoilage e€ects may be two ecological distinct groups of measures are essential:
detectable. This is the case when yeast strains are able to (i) limiting contaminants by selecting good quality raw
form cell clusters in bottled white wines, suggesting materials, monitoring the proper functioning of pro-
incorrect ®ltration, or in certain solid foods (e.g. skinless cesses designed to control microbes, and preventing
sausages, sliced bread, etc.) where the presence of yeast cross contamination; and (ii) limiting colonization by
colonies on the surface gives colour spots, white or pink, arresting or retarding microbial growth.
according to the yeast species present. If we take a closer look at these measures, some of
In conclusion, the establishment of microbiological, them have a rather subjective character, as the `good
chemical, physical or sensorial parameters to evaluate quality raw materials' or `monitoring the proper func-
food spoilage caused by yeasts should account for the tioning of processes' that are concepts with a degree of
nature of the food and on the type of e€ect resulting demand directly related with the technological level of
from their activity. the companies and with the scienti®c and cultural level
of their technical sta€. Is the microbial ecology of foods
Limitations of food mycology able to render objective these concepts? The concept of
Food spoilage is a competitive process in which, as in `arresting or retarding growth' depends as well on the
other forms of competition, the race tends to go to the ability of the microbiologist to monitor the growth
strongest at the start and quickest o€ the mark [1]. kinetics of mixed populations in the processing and dis-
When a food commodity is spoiled by yeasts, it means tribution chain. Is this possible? If yes, to what extent
that they were previously present as contaminants and can he do it?
 V. Loureiro, A. Querol / Trends in Food Science & Technology 10 (1999) 356±365 359

The above mentioned questions are not usually Selective and/or di€erential media: a promising future
raised by the food microbiologists or technologists The most comprehensive approaches on the use of
working in the industry, particularly when it concerns selective and/or di€erential media for isolation or enu-
yeast populations, who rely on the usual micro- meration of particular yeast species have been conducted
biological techniques and on the implemented HACCP in the brewing and wine industries, where the goal is to
systems. Their doubts only come up when an outbreak develop media that selectively di€erentiate Saccharo-
occurs. The limitations of the techniques are then myces cerevisiae, spoilage species within the genus Sac-
recognized and they become concerned, improving the charomyces, and spoilage species in other genera [3].
methodologies temporarily. But once the e€ect of the The best example of this approach is the use of the
outbreak is gone the usual unworried attitude comes classical Lysine agar, largely used in brewing industry,
back. to assess the contamination level of beer with non-
Despite the scienti®c innovation that occurred in the Saccharomyces species, since the former cannot utilise
past years on methods to identify microorganisms, the lysine as a nitrogen source. Other media with similar
majority of the standard methods of food analyses, as purposes have been developed, such as Schwarz di€er-
well as other microbiological methods used in the food ential medium and Lin's wild yeast di€erential agar.
industry, keep using the classical steps established The most used strategy to design speci®c media to
several decades ago: (i) maceration/blending of the isolate/enumerate yeast resistant to a given compound
sample; (ii) decimal dilution of the homogenate; (iii) has been to supplement a general-purpose medium with
plating of the dilutions onto appropriate agar media or that compound. Thus, selective media have been devel-
inoculation into multiple tubes containing liquid media; oped to assess the presence in foods of yeasts resistant
and (iv) isolation, puri®cation, and identi®cation of to ethanol, preservatives and reduced water activity.
colonies. Sometimes, in addition, selective or pre- However, these media have, generally, the disadvantage
enrichment techniques are necessary prior to plating to require longer incubation periods (up to one week)
(low levels of contamination or stressed cells) but with a than the normal media, given the stress induced on the
more frequent use for bacteria than for yeast or moulds. yeast cells. Among these media, standard malt agar
Nevertheless, these procedures do not provide a `real supplemented with 0.5% of acetic acid is currently used
image' of the microbial ecosystem of foods. A recent to detect preservative-resistant yeasts. Numerous solid
review [12] made for microorganisms in general, pre- media have been formulated for isolation and enu-
sents a detailed discussion of the main limitations and meration of fungi capable of growing at low aw values.
errors of the classical procedures which, as a rule, are However, none is completely selective for xerotolerant
not generally questioned. For this reason we will focus yeasts and only one, dichloran 18% glycerol (DG18)
only on the main technical limitations to isolate and agar, is commercially available [13]. Di€erent approa-
enumerate yeast that, in our opinion, do really occur in ches based on the evaluation of product susceptibility to
food and beverage industries. spoilage yeast have been proposed to design shelf life
tests [14]. These are based on pre-enrichment proce-
Isolation and enumeration media for food-borne dures, which promote the emergence of dangerous
yeasts yeasts to the product stability.
As a rule, the general isolation and enumeration Di€erential media seem to be much more promising
media for food-borne yeasts are the same as those used than selective media, because they do not involve, gen-
for food-borne moulds. Basically, these are complex erally, extended incubation times. However, there are
media, nutritionally rich, containing an energy source very few di€erential media for food-borne yeasts and
(e.g. glucose, fructose, sucrose), a digested protein (e.g. much less than those for food-borne bacteria. One
peptone, tryptone, casitone), a complex vitamin supple- strategy used to design yeast di€erential media is based
ment (e.g. yeast extract, malt extract), one or more on the capacity of the yeasts to hydrolyse certain poly-
antibiotics against bacteria (oxytetracycline, chlortetra- mers such as polysaccharides, proteins, pectins and
cycline, chloramphenicol), a compound to inhibit the lipids. Some of these media, although with small speci-
most rapidly spreading moulds (e.g. rose Bengal, ®city, have been utilized to detect the presence of
dichloran), sometimes a pH indicator (e.g. bromocresol hydrolytic yeasts in food where these compounds are
green), agar or not, depending on its use as solid or present in high-levels [3].
broth media. Many studies have demonstrated that Perhaps the best succeeded di€erential medium for
these media are generally superior to the earlier acidi®ed yeast ever designed has been the medium of Chaskes
media (with organic or inorganic acids to get a pH value and Tyndall [15] to distinguish the non-food pathogen
around 3.5) in terms of yeast recovery [13]. Unfortu- Cryptococcus neoformans. The authors' strategy was
nately, they have a great limitation: they do not distin- based on the species unique capacity to produce, under
guish between the dangerous spoilage yeasts and the appropriate environmental conditions, dark pigments
others, which we call `innocent yeasts'. from l-2,3-dihydroxyphenylalanine (l-Dopa). This
360 V. Loureiro, A. Querol / Trends in Food Science & Technology 10 (1999) 356±365

feature is so speci®c and consistent among the strains of serve to describe and to identify species and, to a very
this species that it has been adopted by yeast taxono- minor extent, genera. The tests most used for routine
mists as a relevant characteristic of the species. The identi®cation purposes are fermentation of, and growth
same authors, recently succeeded in distinguishing several on, carbon sources, growth on nitrogen sources,
species of Cryptococcus based on their capacity to pro- requirements of vitamins, growth at various tempera-
duce brown pigments from several aminophenols and tures and on media with a high content of sugar or
diaminobenzenes. Surprisingly this promising strategy sodium chloride, hydrolysis of urea and resistance to
to design di€erential culture media for yeast, to our antibiotics [22]. This author emphasized that there is not
knowledge, was never exploited again, until it was retaken a single standardized method for many of these tests
by the work of Carreira and Loureiro [16]. They di€er- and their results are often dependent on the techniques
entiated Yarrowia lipolytica based on the production of employed. In addition to this weakness, many results of
brown pigments from tyrosine, the detection of this the tests are variable for di€erent strains of the same
species being possible within 24 hours. species, giving rise to frequent misidenti®cation. As
The selective properties of triarylmethane dyes and referred by von Arx [23], this type of approach resulted
di€erential responses of yeasts isolated from meat prod- in heterogeneous taxa, as well as the description of
ucts have been also explored [17]. The results presented identical species under di€erent names. This author also
by these authors suggest that this approach is promising referred to the description of numerous super¯uous
for developing new di€erential media for yeast isolation `species' distinguished mainly by the assimilation pat-
and presumptive identi®cation. terns resulting in a classi®cation, which could only be
The use of conjugated substrates for speci®c detection applied by a few specialists. Besides all the mentioned
of b-glucosidase enzyme on solid media was recently limitations, it is generally necessary to conduct about
exploited to design a chromogenic medium for the 50±100 tests in order to reliably identify yeasts at the
detection of Debaryomyces hansenii from intermediate species level, and one to two weeks, at least, are required
moisture foods [18]. Using a similar strategy the same to obtain a ®nal result. Obviously, this identi®cation
research group had already developed a di€erential scheme cannot be routinely utilized by the food industry
medium (under patent protection) to detect Kluyver- and various miniaturized and simpli®ed identi®cation
omyces marxianus and K. lactis, based on the expression methods, based on yeast response to a few carefully
of the enzyme b-galactosidase [19]. selected tests, have been developed [9]. However, they
The di€erent ability of various spoilage yeasts to grow are based on the same approach of the classical method
in a mineral medium containing glucose and formic of yeast identi®cation, being time consuming, even when
acid, as the only carbon and energy sources, was procedures are automated and computerized, and con-
recently exploited to design a selective/di€erential medium ducting, often, to false or equivocal identi®cations. To
(under patent registration) to Zygossacharomyces bailii overcome these diculties, alternative faster typing
and Z. bisporus [20]. The nature of the acid and its methods have been developed, based, among other
concentration as well as the pH of the medium, asso- things, on analysis of total proteins or long-chain fatty
ciated with the incorporation of an acid-base indicator acids of the yeast cell or based on its isoenzyme pro®le.
ensured that only dangerous species gave positive Besides, recent progress in molecular biology has led to
responses within 48 hours. Another selective and di€er- the development of new techniques for yeast identi®ca-
ential medium (also under patent registration), designed tion and characterization. These include, for example,
for species of Dekkera/Brettanomyces, is based on the RFLP of mitochondrial DNA, chromosomal DNA
capacity of these species to produce 4-ethylphenol from electrophoresis, ribosomal DNA restriction analysis,
p-coumaric acid, which is very easily detectable by its and RAPDs. However, most of these techniques are
peculiar odour. In this medium, ethanol is the carbon useless for the routine identi®cation of yeasts and,
and energy source and the typical production of acetic usually, there is no available database to identify a high
acid by these species is detected by the presence of an number of species. The majority of these techniques
acid-base indicator [21]. were applied to characterize only some genera, so we
The new strategies developed recently create the pos- cannot use them as a general method to yeast identi®-
sibility to build, in the near future, a set of di€erential cation. Because of the relevance that such methods can
media for the main spoilage yeasts found in foods and have to the food industry in the future, some of them
provide excellent opportunities to improve the control will be described in general terms.
of these contaminant microorganisms by the industry.
Long-chain fatty acids
Yeast typing/identi®cation The use of fatty acid pro®ling as a biomarker has not
Yeasts have been classi®ed on the basis of morphologi- been successfully used to type spoilage yeasts in the
cal, sexual, biochemical, and physiological properties for food industry. Certain problems have been suggested:
more than 50 years. Physiological properties, primarily, dependence of total fatty acid composition on yeast
 V. Loureiro, A. Querol / Trends in Food Science & Technology 10 (1999) 356±365 361

growth conditions, low variability of yeast fatty-acid are more apparent than real. In fact, the number of
composition, time-consuming and technique costs, need dominating species in various food commodities seldom
of highly skilled sta€ to perform it, old fashioned tech- exceeds three and the number of minor species is below
nique, etc. However, such prejudices are not real, 10±12 species [9]. Under these circumstances it is not
because it is possible to standardize growth conditions dicult to select yeast colonies on agar plates, based on
(even in plate media) in order to minimize the variation morphological characteristics, and to estimate, satisfac-
of total lipid composition, the results can be obtained torily, the relative proportion of the di€erent species of
within two days after yeast isolation and puri®cation (it a given food. The single signi®cant diculty for the
may take longer if the strains are slow-growing species), food industry is the absence of a database of long-chain
the technique is not more expensive than most of other fatty acid pro®les for the main food contaminant yeasts.
rapid typing techniques available [24]. We consider that However, several research laboratories already have
this technique has the potential to be used in the food such data; the diculty is easy to overcome.
industry, because fatty acid pro®ling permits the typing
of most of the relevant species in the food industry, to Electrophoretic isoenzyme pro®les
separate species according to their spoilage potential The di€erences found in amino acid sequences of
and to trace yeast contamination in bottling plants, as enzymes from di€erent organisms re¯ect organismal
shown in the wine industry Malfeito-Ferreira et al. [23]. genetic divergence. Amino acid substitutions can often
The principal advantage of this approach is to separate be detected from the extent of migration shown by
the food-borne yeasts into three major groups based on enzymes on electrophoretic gels and the corresponding
the presence of linoleic (C18:2) and linolenic (C18:3) patterns are termed zymograms. These zymograms, also
fatty acids (see Table 2), which have, for most of the termed electrophoretic isoenzyme pro®les, have been
food industry, a technological meaning. Thus, the pres- shown to be a useful yeast taxonomic tool and have
ence of most potential spoilage yeasts can be indicated been used to type clinical isolates of yeast pathogens.
by signi®cant amounts of C18:2 and absence of C18:3 With this approach, Duarte et al. [25] grouped 35 yeast
fatty acids in yeast cells; the presence of C18:2 and strains into the four currently recognized species of
C18:3 fatty acids in yeast cells probably re¯ects the Saccharomyces sensu stricto (S. cerevisiae, S. bayanus, S.
presence of yeast species normally associated with poor pastorianus/S. carlsbergensis and S. paradoxus), demon-
hygiene or insucient use of preservatives; the absence strating its value for identi®cation purposes and showing
of polyunsaturated C18 acids suggests the presence of the usefulness of the technique for a rapid and sensitive
fermentative strains which can be spoilers, such as those identi®cation of strains from this industrially important
belonging to the Saccharomyces genus. To those who group of yeasts.
are not familiar with this technique, it may seem dicult to Bearing in mind that, after strain puri®cation, this
implement it at industrial level. However, most diculties technique takes about 48 hours to be performed, does

Table 2. Separation of the yeast species in three groups according to their polyunsaturated C18 fatty acids composition
Yeast speciesa
Group I [C18:2 (ÿ); C18:3 (ÿ)] GroupII [C18:2 (+); C18:3 (ÿ)] Group III [C18:2 (+); C18:3 (+)]
Hanseniaspora uvarum Dekkera anomala Candida catenulata Kluyveromyces lactis
Hanseniaspora guilliermondi Dekkera bruxellensis Candida diddensiae Kluyveromyces marxianus
Hanseniaspora vineae Dekkera naardenensis Candida guilliermondii Kluyveromyces thermotolerans
Saccharomyces bayanus Pichia etchellsii Candida norvegica Lodderomyces elongisporus
Saccharomyces cerevisiae Saccharomyces dairensis Candida parapsilosis Metschnikowia pulcherrima
Saccharomyces chevalieri Torulaspora delbrueckii Candida sake Pichia anomala
Saccharomyces exiguus Torulaspora globosa Candida rugosa Pichia norvergensis
Saccharomyces pastorianus Yarrowia lipolytica Candida tropicalis Pichia fermentans
Saccharomyces unisporus Zygosaccharomyces bailii Candida utilis Pichia guilliermondii
Saccharomycodes ludwigii Zygosaccharomyces bisporus Candida zeylanoides Pichia jadini
Schizosaccharomyces octosporus Zygosaccharomyces ¯orentinus Cryptococcus albidus Pichia membranifaciens
Schizosaccharomyces pombe Zygosaccharomyces mellis Cryptococcus humicolus Rhodotorula glutinis
Zygosaccharomyces microellipsoides Cryptococcus ¯avus Rhodotorula mucilaginosa
Zygosaccharomyces rouxii Cryptococcus laurentii Saccharomyces kluyveri
Cryptococcus terreus Zygosaccharomyces fermentati
Debaryomyces hansenii
Debaryomyces polymorphus
Issatchenkia orientalis
Issatchenkia terricola
a
Teleomorphs of the species described in the 4th edition of ``The Yeasts. A Taxonomic Study'' [22].
362 V. Loureiro, A. Querol / Trends in Food Science & Technology 10 (1999) 356±365

not require expensive or sophisticated equipment, is DNA and only the M13 sequence GAGGGTGGCGG-
cheaper than other rapid typing methods and is extremely TTCT have been used for yeast identi®cation and charac-
sensitive and reliable, we do not understand why it has terization. This technique is more reproducible than
been forgotten by food microbiologists. In our opinion, RAPD based methods since the annealing temperature
this technique has high potential applicability in food is higher, 55 C instead of 37 C [29]. Using this technique
industry. For this purpose it is necessary to create a Baleiras Couto et al. [31] identi®ed strains of Z. bailli
representative database of the electrophoretic isoenzyme and Z. bisporus present in mayonnaise and salad dressings.
patterns of the most relevant yeast species in food tech- The usefulness of PCR for detection of yeasts involved
nology. in food spoilage includes assays for Dekkera/Brettano-
myces [32] and Z. bailii [33].
Molecular biology techniques Gene probe in a variety of in situ hybridization or
In several hemiascomycetous yeasts, rRNA genes are PCR formats have been used for the detection, identi®-
located in a single genomic region composed of 100±150 cation, and characterization of speci®c DNA segments.
tandem repeats of a fragment of 9 Kb. These fragments The versatility of gene probe analyses is demonstrated
contain two transcriptional units, one of them (7 Kb) is by the many clinical and environmental applications
a cluster of the genes coding for the 18S, 5.8S and 25S that have been performed. Constantly new gene probes
rRNAs and two internal transcribed spacers, ITS1 and for food analyses are potential tools in improving the
ITS2. The second unit, which is transcribed in the rate of identi®cation and characterization of food-borne
opposite direction, corresponds to the 5S rRNA. or spoilage microorganisms.
Recently, Esteve-Zarzoso et al. [26], proposed a new Another application of gene probes is in situ hybridi-
rapid and easy method of routine yeast identi®cation zation. Fluorescent in situ hybridization of whole cells
based on the restriction analysis of the 5.8S rRNA gene with rRNA-targeted oligonucleotide probes has
and the internal transcribed spacers (ITS). They pre- increasingly become a valuable tool for the speci®c
sented an initial database to identify a total of 132 yeast detection of microorganisms without previous cultiva-
species belonging to 25 di€erent genera. Using this tion. The use of probes that speci®cally binds to a target
method it is possible to identify yeast from isolated region in the 18S rRNA was applied in the identi®cation
colonies or directly from food samples. This is the ®rst of fungal pathogens such as Candida species [34] and
available molecular method that presented information yoghurt spoiling yeast [35]. The dependence of this
to identify a high number of yeast species. Using the method from sequences deposited in databases for
same methodology, but amplifying di€erent regions probe design and comparison constitutes, frequently, a
(18S rRNA and ITS1), Dlauchy et al. [27] constructed a limitation of this technique. The 18S rRNA and 28S
database of restriction fragment patterns to identify 128 rRNA gene sequences were applied in yeast taxonomy
species associated mainly with food, wine, beer, and soft studies. However, the homology of this region among
drinks. species is very high and in some cases it is dicult to
Di€erent approaches are applied using PCR. The distinguish closely related species [35]. Therefore, the
Random Ampli®ed Polymorphic DNA (RAPD) tech- application of probes for other regions could be more
nique is based on the ampli®cation of random segments promising.
of DNA with a single and short (5±15 mer) primer of an Since the ®rst karyotyping performed by orthogonal-
arbitrary nucleotide sequence. Due to the low annealing ®eld-alternation gel electrophoresis, pulsed ®eld electro-
temperature, the primer hybridizes at loci distributed at phoresis techniques have made possible the separation
random throughout the genome, allowing the ampli®- of the chromosomes of S. cerevisae and other yeasts (for
cation of polymorphic DNA fragments (for a schematic a review and food industry application, see Deak et al.
diagram see Refs 28 and 29). The level of di€erentiation, [28]). Chromosomal polymorphisms are due to dele-
either interspecies or intraspecies, depended highly on tions, insertions and translocations of long enough
the primers used. However, one of the most important DNA fragments to be detectable by electrophoresis and
problems with this technique is the low stable pattern. have proved to be useful for the di€erentiation of species
Otherwise, the RAPD assay is a less time-consuming belonging to several genera, e.g. Candida, Kluyver-
tool principally for typing organisms and has also been omyces, Saccharomyces, and Zygosaccharomyces (for a
shown to be suitable for the identi®cation of yeast species. review see Esteve-Zarzoso et al. [36]). However, this
This technique has been used to discriminate between technique is complex and time-consuming and does not
some of the typical spoilage yeast such as Candida lipo- permit the analysis of a large number of samples; this
lytica, C. valida, S. cerevisiae, Z. bailii and Z. rouxii [30]. could be the reason why no applications to study spoilage
Another PCR-based approach consists in the ampli®- yeasts have been published until now.
cation of fragments with oligonucleotides speci®c for Finally, mitochondrial DNA (mtDNA) polymor-
simple repetitive DNA sequences, named microsatellites. phisms have been extensively used to detect changes in the
Primers (GTC)5, (GAC)5, (GACA)4, the phage M13 organization of this genome and, as well, as an important
 V. Loureiro, A. Querol / Trends in Food Science & Technology 10 (1999) 356±365 363

tool in taxonomic studies. Yeasts are a group of organ- (e.g. faecal coliformes, enterococci, etc.) but with eco-
isms with a high degree of variability in the size and logical habits similar to the pathogens are detected. They
form of their mtDNA, and several methods have been are called `index organisms' [11] and each one includes,
developed for the isolation of yeast mitochondrial generally, several bacterial genera and species. Like
DNA. However, Querol et al. [37] have developed a these markers, others are also used to evaluate the pre-
mtDNA restriction analysis method for Saccharomyces, sence or activity of spoilage microorganisms in foods,
which does not require previous isolation of mitochon- generally named as `spoilage indicators'. When measur-
dria or puri®cation of mtDNA. This technique has also ing directly the presence of spoilage microorganisms
been used successfully to characterize strains of species they are called `microbiological markers' and include
other than Saccharomyces; Candida zeylanoydes and groups without taxonomical signi®cance, but with one
Debaryomyces hansenii and species of the genera Brettano- or more relevant phenotypical characteristics (e.g. spore
myces, Kluyveromyces and Zygosaccharomyces [36]. formers, thermophiles, psychrotrophs, proteolytics,
The rapid identi®cation of spoilage microorganisms is lipolytics, pseudomonads, micrococci, etc.); when mea-
of eminent importance to the food industry. Rapid suring the metabolic activity, through metabolite deter-
PCR-based detection techniques or in situ hybridization mination (e.g. carbon dioxide, butanediol, trimethyl-
can be used in the food industry as an integral part of amine, etc.) or through the composition of cellular biomass
quality control to detect or identify yeasts. The major (e.g. long-chain fatty acids, electrophoretic isoenzyme
problem of these techniques is the direct application to pro®les, etc.) they are named as `chemical markers';
the sample without previous yeast isolation due to the when measuring physical parameters dependent on
impossibility to discriminate non-viable cells. For this microbial activity (e.g. pH, changes in the electrical
reason a negative result is conclusive but not a positive resistance, redox potential, etc.) they are called `physical
one. Another important problem asociated with the markers'. Most `spoilage indicators' used in food indus-
identi®cation of foodborne yeast consists in the speci®c try were developed to characterize bacterial spoilage. To
detection of strains. Discrimination at subspecies or characterize the activity of spoilage yeast there are,
strain levels and the analyses of within-species genetic however, fewer indicators, revealing the lower impor-
polymorphisms and natural population variability have tance paid to yeast by food microbiologists and, also,
been shown as very helpful to determine contamination the lower biochemical versatility of yeast. This fact
sources and processes. After the critical analysis of the makes the development of di€erential tests to distin-
application of di€erent methods currently available for guish the di€erent groups of yeasts relevant in food
yeast strain characterization, we conclude that the use technology more dicult. We will now discuss some of
of chromosomal pro®le and mtDNA restriction pattern the main `zymological indicators' (zymo=yeast) used,
analyses are the most useful to di€erentiate and dis- or with potential to be used, in food industry.
criminate a higher number of yeast strains.
Zymological indicators
Spoilage indicators The most common indicator is the classical `yeast and
There are di€erent reasons to perform micro- moulds', resulting from the utilization of a general pur-
biological analysis of a food product. The most rele- pose plating medium to enumerate microfungi (yeasts
vant, without any doubt, is to estimate the presence of and moulds) responsible for the contamination of food.
pathogenic microorganisms in foods which may put in As with bacteria, there are certain foods (e.g. meat, ®sh
jeopardy the health of the consumers. However, from the and chilled products) in which, occasionally, it is useful
industrial point of view, other reasons must be considered to detect/enumerate `psychrotroph yeasts' using the
as well: the estimation of the product shelf life, the eval- same culture medium for `yeasts and moulds' but incu-
uation of the quality of raw materials, the tracing of the bating at temperatures lower than 25 C (generally 5 or
route of contamination in the processing lines, etc. The 10 C).
choice of the analysis to be made, that is, the micro- Other common indicators are formed by groups of
biological criteria to be adopted, depends on the purpose yeasts tolerant to environmental stress, which are
and on the diculty of the techniques to perform the detected/enumerated by modi®ed general purpose media,
required analysis. This fact allowed the emergence of according to the stress factor involved. Thus, `acid-
the concept of `microbiological indicator', as a means of resistant yeasts' is a zymological indicator used to
avoiding the diculties associated with the detection of characterize dangerous yeast in acid food industries,
pathogenic bacteria. In general, pathogens are present in such as acetic and lactic acid preserved pickles, olives,
foods in low numbers and their identi®cation requires mayonnaise, sauces, fruit juices, beverages stabilized
the utilisation of numerous biochemical and physiologi- with preservatives, wines, etc. `Xerotolerant (osmophilic)
cal tests, turning impracticable their detection in yeasts' is used for grouping yeast particularly dangerous in
industrial routine. Alternatively, other microorganisms or industries that process low aw foods, such as sugar syrups,
groups of microorganisms, without taxonomic signi®cance fruit concentrates, jams, jellies, confectionery, bakery,
364 V. Loureiro, A. Querol / Trends in Food Science & Technology 10 (1999) 356±365

dairy desserts, dried fruits, cured meats and other meat As yeast and moulds can produce 1,3-pentadiene much
products, etc. The medium chosen for detection or enu- faster in laboratory conditions than in the product, this
meration of these yeasts should re¯ect the content of the compound can be considered an excellent `spoilage pre-
food being analysedÐglucose, fructose, sucrose, sodium dictor' to this type of food products.
chloride [9]. Ethanol (11.4% v/v) has been successfully As already mentioned, it is possible to type (identify)
used as a selective agent in a medium developed to the most relevant yeast to food technology based on
detect `wine-spoiling yeast' [38]. `Wild yeasts' is an their total biomass fatty acid pro®les and to separate
indicator commonly used in brewing, based on the fact them into three groups with technological signi®cance
that brewer's yeast (and most other Saccharomyces species), according to the presence or absence of linoleic (C18:2)
unlike other yeast genera that occur as contaminants, or linolenic (C18:3) acids. A careful examination of
cannot utilize lysine as a sole nitrogen source. Although these groups (see Table 2) shows that these acids can be
the absence of yeast growth on lysine agar does not used as chemical markers of the microbiological quality
necessarily mean an absence of wild yeast, this medium of raw materials and ®nal products in many food
can, nevertheless, be used to monitor the presence of industries. Besides, taking into account that these acids
contaminants in the beer processing. are detected through a chromatogram giving the total
Occasionally, it is of interest to detect the presence of long-chain fatty acid composition, it is possible, in
yeasts with speci®c hydrolytic capacities, namely `lipo- many cases, to type the species contaminating the food.
lytic yeasts', `pectinolytic yeasts' and `proteolytic According to the experience of one of us (V.L.) this
yeasts', using appropriate culture media [9]. chemical marker can be especially useful in the charac-
The species of spoilage yeasts are seldom used as terization of the microbial quality of processed raw
zymological indicators, given the lack of selective or materials, such as sugar syrups and fruit concentrates.
di€erential media to quantify them in a rapid way. A few Here the presence of group III yeasts is not usually
of the known examples have already been referred to in dangerous, but the presence of group I and, particularly,
the description of selective and di€erential culture media. group II yeasts are extremely dangerous to the industries
An alternative approach to the zymological indica- reprocessing them (e.g. confectionery, soft drinks,
tors, which has been extensively studied in bacteria bakery, etc.).
during the past 80 years, is to examine food samples for
chemical (or sensorial) evidence of past microbial activity. Conclusions
The use of metabolites as indicators of spoilage is often In the last few decades the relevance of yeasts as food
more convenient and faster than using microbiological spoilage agents has become greater. Nevertheless, the
methods of examination. However, very few have zymological indicators used in food industry, to assess
received acceptance as a means of assessing the degree of the quality of raw materials, processing and ®nal products,
yeast spoilage of foods. Ethanol and acetoin levels pro- are not, at present, completely satisfactory and re¯ect
vide reliable indexes of the quality of the fruit on arrival the still limited tools available in food mycology to
at the factory and of hygiene in the processing plant, enumerate and to type food contaminant yeast. The
respectively [11]; analysis of carbon dioxide in the head strategies to be developed to improve the present indi-
space of sealed culture vials was proposed for rapid cators should consider the following aspects: (i) rather
enumeration of fermentative yeasts in food, using a than `spoilage indicators' it is necessary to develop
selective medium and gas-chromatographic analysis `spoilage predictors'; (ii) rather than identifying food
[39]; 4-ethyl-phenol can be used as a sensory or chemical contaminant yeasts it is necessary to know the yeasts
marker for Dekkera/Brettanomyces-infected wines. This properties or, as stated by Davenport [14], ``micro-
chemical marker may be regarded as a `spoilage pre- organism behaviour patterns and traits are paramount
dictor' because, as already mentioned, in low levels the over correct names''; (iii) the development of new
wine is not spoiled and allows the detection of Dekkera/ zymological indicators should be based, as a matter of
Brettanomyces yeasts in contaminated wines before they priority, on methods that lead to a `coarse distinction'
deteriorate. Recently, Casas et al. [40] proposed a sim- of the yeasts; the `®ne distinction', usually made by
pli®ed test to detect, in 48 hours, `petroleum-like' o€- molecular methods at an intraspeci®c level, should
odour-producing fungi for marzipan based products complement the former ones, especially to trace routes
preserved with sorbate. The principle of the test is based of contamination in processing or distribution lines; and
on the ability of some yeasts (Z. rouxii and D. hansenii) (iv) the de®nition of spoilage indicators/predictors
and moulds (Penicillium chrysogenum, P. simplicissimum, should always have in mind the concept of `ecology of
P. crustosum and Aspergillus niger) to convert sorbate in zero growth', being based on two or more determina-
1,3-pentadiene, which is easily detectable by nose. In tions separated in time. This last condition leads,
addition, the test allows distinguishing Z. rouxiiÐthe necessarily, to predictive mycology which, for food
most dangerous and frequent spoilage yeast from this spoilage yeast, is still virtually terra incognita or it is
type of productsÐbecause it is the only gas producer. used con®dentially by large companies.
 V. Loureiro, A. Querol / Trends in Food Science & Technology 10 (1999) 356±365 365

References tamination in Bottling Plants' in Int. J. Food Microbiol. 38, 143±

155
1 Ingram, M. (1958) `Yeasts in Food Spoilage' in The Chemistry 25 Duarte, F.L., Pais, C., Spencer-Martins, I. and LeaÄo, C. (1999)
and Biology of Yeasts (Cook, A.H., ed.), pp. 603±633, Academic `Distinctive Electrophoretic Isoenzyme Pro®les in Sacchar-
Press omyces cerevisiae sensu stricto' in Int. J. System. Bacteriol. 49,
2 Walker, H.W. and Ayres, J.C. (1970) `Yeasts as Spoilage 1907±1913
Organisms' in The Yeasts Vol. 5, (Rose, A.H. and Harrison, J.S. 26 Esteve-Zarzoso, B., Belloch, C., Uruburu, F. and Querol, A.
eds), pp. 435±516, Academic Press (1999) `Identi®cation of Yeasts by RFLP Analysis of the 58S
3 Fleet, G.H. (1992) `Spoilage Yeasts' in Crit. Rev. Biotechnol. 12, rRNA Gene and the Two Ribosomal Internal Transcribed
1±44 Spacers' in Int. J. Syst. Bacteriol. 49, 329±337
4 Thomas, D.S. (1993) `Yeasts as Spoilage Organisms in Bev- 27 Dlauchy, D., Tornai-Lehoczli, J. and PeÂter, G. (1999) `Restriction
erages' in The Yeasts, Vol. 5, 2nd edition (Rose, A.H. and Har- Enzyme Analysis of PCR Ampli®ed rDNA as a Taxonomic Tool
rison, J.S. eds), pp. 517±561, Academic Press in Yeast Identi®cation' in System. Appl. Microbiol. 22, 445±453
5 Hazen, K.C. (1995) `New and Emerging Yeast Pathogens' in 28 Deak, T. (1995) `Methods for the Detection and Identi®cation
Clin. Microbiol. Rev. 8, 462±478 of Yeasts in Foods' in Trends Food Sci. Tech. 6, 287±292
6 Barnett, J., Payne, R.W. and Yarrow, D. (1983) Yeasts: Charac- 29 Van der Vossen, J.M.B.M. and Hofstra, H. (1996) `DNA Based
teristics and Identi®cation, Cambridge University Press Typing, Identi®cation and Detection Systems for Food Spoilage
7 Pitt, J.I. and Hocking, A.D. (1985) Fungi and Food Spoilage, Microorganisms: Development and Implementation' in Int. J.
Academic Press Food Microbiol. 33, 35±49
8 Tudor, E.A. and Board, R.G. (1993) `Food Spoilage Yeasts' in The 30 Baleiras Couto, M.M., Vogels, J.T.W.E., Hofstra, H., Huis in't
Yeasts, Vol. 5, 2nd ed. (Rose, A.H. and Harrison, J.S. eds), pp. Veld, J.H.J. and van der Vossen, J.M.B.M. (1995) `Random
435±516, Academic Press Ampli®ed Polymorphic DNA and Restriction Enzyme Analysis
9 Deak, T. and Beuchat, L.R. (1996) Handbook of Food Spoilage of PCR Ampli®ed rDNA in Taxonomy: Two Identi®cation
Yeasts, CRC Press Techniques for Food-Borne Yeasts' in J. Appl. Bacteriol. 79,
10 Chatonnet, P., Dubourdieu, D., Boidron, J.N. and Pons, M. 525±535
(1992) `The Origin of Ethylphenols in Wines' in J. Sci. Food 31 Baleiras Couto, M.M., Hartog, B.J., Huis in't Veld, J.H.J., Hofstra,
Agric. 60, 165±178 H. and van der Vossen, J.M.B.M. (1996) `Identi®cation of Spoi-
11 Mossel, D.A.A., Corry, J.E.L., Struijk, C.B. and Baird, R.M. (1995) lage Yeasts in a Food-Production Chain by Microsatellite
Essentials of the Microbiology of Foods, John Wiley & Sons Polymerase Chain Reaction Fingerprinting' in Food. Microbiol.
12 Fleet, G.H. (1999) `Microorganisms in Food Ecosystems' in Int. J. 13, 59±67
Food Microbiol. 50, 101±117 32 Ibeas, J.I., Lozano, I., Perdigones, F. and JimeÂnez, J. (1996)
13 Beuchat, L.R. (1993) `Selective Media for Detecting and Enu- `Detection of Dekkera-Brettanomyces Strains in Sherry by a
merating Foodborne Yeasts' in Int. J. Food Microbiol. 19, 1±14 Nested PCR Method' in Appl. Environ. Microbiol. 62, 998±1003
14 Davenport, R.R. (1986) `Forensic Microbiology of Soft Drinks 33 Stubbs, S., Hutson, R., James, S. and Collins, M.D. (1994) `Dif-
Business' in Soft Drinks Management Int. 4, 31±32 ferentiation of the Spoilage Yeast Zygossacharomyces bailii
15 Chaskes, S. and Tyndall, R.L. (1975) `Pigment Production by from Other Zygossacharomyces Species Using 18S rDNA as
Cryptococcus neoformans from para- and ortho-diphenols: Target for a Non-Radioactive Ligase Detection Reaction' in
E€ect of the Nitrogen Source' in J. Clin. Microbiol. 1, 509±514 Lett. Appl. Microbiol. 19, 268±272
16 Carreira, A. and Loureiro, V. (1998) `A Di€erential Medium to 34 Lischewiski, A., Kretschmar, M., Hof, H., Amann, R., Hacker, J.
Detect Yarrowia Lipolytica in 24 Hours' in J. Food Mycol. 1, 3±12 and MorschhaÈuser, J. (1997) `Detection and Identi®cation of
17 Fung, D.Y.C. and Liang, C. (1990) `Critical Review of Isolation, Candida Species Inexperimentally Infected Tissue and Human
Detection, and Identi®cation of Yeasts from Meat Products' in Blood by rRNA-Speci®c Fluorescent in situ Hybridization' in J.
Crit. Rev. Food Sci. Nutr., 337±341 Clin. Microbiol. 35, 2943±2948
18 SiloÂniz, M.I., Valderrama, M.J. and Peinado, J.M. (2000) `A 35 Kosse, D., Deiler, H., Amann, R., Ludwig, W. and Scherer, S.
Chromogenic Medium for the Detection of Yeasts with b- (1997) `Identi®cation of Yoghurt-Spoiling Yeasts With 18S
galactosidase and b-glucosidase Activities from Intermediate rRNA-Targeted Oligonucleotide Probes' in System. Appl.
Moisture Foods' in J. Food Prot. (in press) Microbiol. 29, 468±480
19 Valderrama, M.J., SiloÂniz, M.I., Gonzalo, P. and Peinado, J.M. 36 Esteve-Zarzoso, B., Manzanares, P., RamoÂn, D. and Querol, A.
(1999) `A Di€erential Medium for the Isolation of Kluyver- (1998) `The Role of Non-Saccharomyces Yeasts in Industrial
omyces marxianus and Kluyveromyces lactis from Dairy Prod- Winemaking' in Int. Microbiol. 1, 143±149
ucts' in J. Food Prot. 62, 189±193 37 Querol, A., Barrio, E. and RamoÂn, E. (1992) `A Comparative
20 LeaÄo, C., Corte-Real, M. and Schuller, D.E. (1999) `Culture Study of Di€erent Methods of Yeast Strains of Saccharomyces
Medium for the Detection of Zygosaccharomyces bailii and cerevisae Characterization' in Syst. Appl. Microbiol. 58, 2948±
Zygosaccharomyces bisporus' Portuguese Patent (Provisional) 2953
12480V (in Portuguese) 38 Thomas, D.S. and Ackerman, J.C. (1988) `A Selective Medium
21 Rodrigues, N.F., GoncËalves, M.G. and Loureiro, V. (1999) `Cul- for Detecting Yeasts Capable of Spoiling Wine' in J. Appl.
ture Medium for the Detection and Identi®cation of Yeasts of Bacteriol. 65, 299±308
the Genera Dekkera/Brettanomyces' Portuguese Patent (Provi- 39 Guerzoni, M.E., Piva, M. and Gardini, F. (1985) `Proposal of a
sional) 12481X (in Portuguese) Rapid HS-GLC Method for Microbiological Control of Food' in
22 Yarrow, D. (1998) `Methods for the Isolation, Maintenance and Lebbensm.-Wiss. Technol. 18, 128±132
Identi®cation of Yeasts' in The Yeasts. A Taxonomic Study, 4th 40 Casas, E., Valderrama, M.J. and Peinado, J.M. (1999) `Simpli®ed
edition (Kurtzman, C.P. and Fell, J.W., eds), Elsevier Test for the Rapid Detection of Dangerous Yeasts and Fila-
23 Von Arx, J.A. (1980) `A Mycologist's View of Yeasts' in Biology mentous Fungi in Marzipan Products' in Proceedings of the
and Activities of Yeasts (Skinner, F.A., Passmore, S.M. and International Congress on Improved Traditional Foods for the
Davenport, R.R., eds), pp. 53±61, Academic Press Next Century, DG XII-European Commission and the Instituto
24 Malfeito-Ferreira, M., Tareco, M. and Loureiro, V. (1997) `Fatty- de AgroquõÂmica y Tecnologia de Alimentos, pp. 304±307,
Acid Pro®ling: A Feasible Typing System to Trace Yeast Con- Valencia, Spain