differentfoods, as not all contaminants havethe sameability to causefood
spoilage. Thiscan clearly be seenin meatproduction as it has beenestimated
that only10% of the totalbacterialloadcontaminatingthe carcasses is capable
of growing duringrefrigerated storage(Nychaset al., 1988).Moreover,the meat
is usuallypackaged undervacuumor modifiedatmosphere(MA), creatinga
furtherhurdlefor contaminating bacteria.Thisresultsin the growth of certain
bacterialgroups, suchas psychrotrophiclactic acid bacteria (LAB).In a recent
study(Bjo»rkrothet al., 2005)concerningMA packaged,marinated broiler legs,
it was shown that the majorityof initialLABcontaminants (enterococci)were
overgrownduringstorageat 6 ÎC by mainly heterofermentative rods spoilingthe
product. Only27%of the initial LAB contaminants,forming just a minor
fractionof the totalbacterialload,would thenbe relevant to traceduring a
contaminationanalysis of this product. Thesesituations showthe limitationsof
the approaches monitoringtotalbacterialcounts or evenLABcounts. In a
contaminationanalysis of specificspoilage bacteria,we are quiteoften facinga
`needlein a haystack' situation, whichis challenging our methodology.In the
worstcase, the mainspoilage-causingbacterialgroupmaynot evenbe known.
Monitoring of pathogenic bacterialcontamination in foodprocessingsets
different challenges from the corresponding analyses of spoilage bacteria.
Sometimesdetection analysesservewell in tracingcontaminationsitesor source
monitoring.However, the natureof pathogen contaminationis quitevariable
depending on the properties of the contaminant.Sometimes traditional or
polymerase chainreaction(PCR)-basedspeciesdetectionapproaches do not
showthe whole picture of the problem. In the caseofL. monocytogenes,
persistentstrains maycauseprolongedproductcontamination, whichshouldbe
revealed by strain typing techniques.Theseaspects are dealtwithin more detail
in Section 33.3.
Despitethe limitations of conventionalenumerationand detection,these
techniques still formthe cornerstone of contaminationstudies. Sincethe majority
of bacterialtypingtechniques still rely on culturingof the organisms,however,
goodknowledgeof thesemethodsshould be maintained. We are currentlyquite
far fromthe situation whereculture-independenttechniques maybe applied in
routinemonitoringof microbialcontaminantsin industrial foodprocessing.
33.2.2 Molecular-based contaminationanalysis
Thedevelopment of molecular-basedtypingmethodshas revolutionizedthe
fieldof industrialcontamination analysis. Before the era of thesemethods,
contaminationanalysiswas basedon isolation and detection, whichleft many
questions openconcerningcontamination routes.The characterization of species
to strain level has enabled researchers to analyse the relatedness and
epidemiologyof pathogenic and spoilage microorganisms(Bjo»rkrothet al.,
1996,1998;Miettinenet al., 1999;Lunde¬net al., 2003a). Thisinformation has
increased our understanding of contaminationroutes to and withinthe food
processing industry.
Contaminationroutes and analysis in foodprocessing environments 541