to resistcooking in the core of the product bar. The air in the slicingdepartment
and adjacentcoldroomwas shownto contain veryfewLABbut surface-
mediatedcontaminationwas detectedduringslicing/packagingstages.
In this case, molecular typing providedusefulinformationrevealingthe LAB
contaminationsources/sites of the product.Theairborne contamination site
preceding the slicingand packagingphaseswas not anticipatedbeforethese
analyses. Instead, the slicingmachine, whichwas difficult to clean,was thought
to be the source of LABcontamination.After theseanalysesit was concluded
thatthe strains, which were detectedon the finished productsurfacewere
already present beforeslicingtook place.Theresults also confirmed that
following traditionalquantificationof the LABassociated withthe production
phases, no conclusions of the relevant contamination sitescan be made.Indeed,
surfaces werefoundto containmanystrainsthat wereneverdetected in the
product. Apparently these LABwerenot ableto growin the product.
Thesecontamination analyseswerebasedonlyon the ribopattern differences
and no speciesidentification was donein the study. Later(Bjo»rkrothet al.,
1998),an identificationstudy was undertakenin whichthe mostpotentspoilage
LABwas shown to be aLeuconostoc carnosumstrain. Duringthis work an LAB
identificationdatabaseemployingnumericalanalysis ofHindIII ribopatterns
was established. In addition,macro-restriction patterns separated by pulsed-field
gel electrophoresis (PFGE) showed that one majorPFGEtypewas responsible
for the mostabundant spoilage changes. Depending on the levelof information
needed, typingtechniquespossessinghighdiscriminatory powermay haveto be
usedfor obtainingthoroughcomprehensionof the LABcontaminantsand the
contaminationprocess.
33.4.3 Tracing spoilageLABat poultryslaughterhouse and adjacent
productmanufacture
Vihavainenet al.(2005)enumerated and identifiedLABassociated withMAP,
non-marinated,broiler meat at the end of producer-definedshelf-lives.In order
to reveal how spoilage-associated LAB were connected with poultry and
processing environmentcontamination,broiler chickenhandledduring the early
stagesof slaughter and air fromprocessingphasesrelatedto carcass-chilling,
cuttingand packagingwerestudied. The LABcounts in the late shelf-life
products variedfrom 104 to 10^8 cfu/g.A totalof 447,86 and 122 isolates
originatingfrombroiler products, broiler carcassesand processingplant air,
respectively,weresubjectedto identification using a 16 and 23S rRNAgene
RFLPdatabasebasedonHindIII patterns.
CarnobacteriumdivergensandCarnobacteriummaltaromaticum(piscicola)
weredominating the developing spoilage LABpopulations,forming63%of the
LABisolated. OthermajorLABspeciesdetectedin the productsbelongedto the
generaofLactococcus, LeuconostocandLactobacillus.LABassociated withthe
late shelf-life broiler products were recovered fromthe productionplantair
adjacentto the finalcutting and packagingstagesbut not fromthe broiler
Contaminationroutes and analysis in foodprocessing environments 549