628

J. Milk Food Techno/. Vol. 39, No.9, Pages 628·646 (,September, 1976)
Copyright © 1976, International Association of Milk, Food, and Environmental Sanitarians

Enrichment and Plating Methodology

for Salmonella Detection in Food. A Review 1
DIANE J. FAGERBERG and JOHNS. A VENS
Department o.f'Animal Sciences
Colorado State University. Fort Collins. Colorado 80523

(Received for publication March 18, 1976)

ABSTRACT Salmonellae are widespread in the environment and

Much research concerned with enrichment and plating methodology appear in a wide variety of foods and food ingredients,
to detect salmonella contamination in foods has been reported by many posing a great problem to the food industry. Salmonella
scientists. This review brings reported findings of major proponents
organisms are transmitted primarily via the fecal-oral
together into one text for greater understanding and appreciation of the
complexity of the problem. Discussed in this review are reported route since the primary reservoir is the intestinal tract of
applications and mechanisms of 11 enrichment media and eight plating man and animals. From this reservoir, organisms are
media, incubation temperatures and times, and serotype specificity spread to air, food, feed, soil and water from ~here the~
involving enrichment and plating media. Also, enrichment emulsifying
may invade other hosts. These . orgams~s . can
agents, agitation during incubation, sample type, and level and
proportion of salmonellae and competitors as related to salmonella
contaminate our food supply by many dtrect and tndtrect
enrichment are discussed. Other factors related to salmonella recovery. routes.
such as multiple media methods, preparation methods, storage of Salmonella organisms have been recovered from
media, and media brand are included in this review, as well as a domestic and wild animals, specifically poultry and other
discussion of the methodology dilemma and some general
livestock, rodents, reptiles, birds, and insects and other
recommendations for future direction.
arthropods (99). Transmission of salmonellae from
animals to humans via food is a serious public health
problem. Foods implicated in salmonellosis cases
The genus Salmonella, within the family
include: poultry meat, eggs, dairy products, red meat
Enterobacteriaceae, is composed of more than 1400
and meat products, sausage, meat pies, fish and fishery
serotypes and bioserotypes. These bacteria are
products, desiccated coconut, cake mixes, custard-filled
gram-negative, non-spore-forming, aerobic rods, most of
bakery products, chocolate and other confections, cereals
which are peritrichously flagellated and naturally inhabit
and grain products, carmine dye, and other products
the intestinal tract of humans and many other animals
which are cooked at a low temperature and are subject to
and have worldwide distribution.
much handling. Frequency of salmonella isolation from
All species are potentially pathogenic for humans.
domestic poultry indicates they may be the largest single
animals, or both (28). A few species are host specific;
reservoir of these organisms among animals (29, 99).
Salmonella typhi causes typhoid fever in man,
More than two-thirds of the salmonellae isolated from
Salmonella gallinarum causes fowl typhoid in chickens
animals involved poultry. Foster (39) indicated
and turkeys and Salmonella pullorum causes pullorum
three-fourths of the salmonella outbreaks in which the
disease in poultry. Most serotypes show little host
vehicle could be identified were attributable to meat,
specificity and can cause the gastrointestinal disease,
poultry, and egg products.
salmonellosis, when ingested by man. Symptoms
Food processors are not allowed to sell produ.cts
associated with the illness are frequently mild and
containing salmonellae. Such food products, accordmg
self-limiting with recovery within 3 or 4 days. Despite
to the Food, Drug and Cosmetics Act (61) are adulterated
usual rapid recovery from the disease, more severe
because they contain harmful or pathogenic micro-
complications often occur among very young, elderly,
organisms. Most food processors make reasonable. ef~ort
and persons with underlying disease problems as the
to achieve a high degree of processing plant samtauon
infection spreads from the intestine to other systems of
and, thereby, control salmonella contamination; many
the body, resulting in serious morbidity and even
have incorporated testing and control programs and have
mortalitv. The number of human salmonellosis cases per
improved and remodeled facilities and equipment.
year in the United States is conservatively estimated to be
two mi!iion (38, 90, 99). Reflecting the economic impact and importance ?f the
salmonella problem, a National Academy of Sctences
'Published with approval of the Director o.f the Colorado State report (90) estimates total cost to th~ Ame~ican_ ec.onomy
University Experiment Station as Scientific Series paper number 2136. to be at least $300 million annually, tncludtng vtcttm and
 DETECTING SALMONELLAE IN FOOD 629

industry costs. The Subcommittee on Food Microbiology of the Food

Public health agencies as well as food processors are Protection Committee of the National Research Council,
searching for solutions to the salmonella problem. National Academy of Sciences, has published a booklet
including determination of sources of salmonella entitled, Reference Methods for the Microbiological
contamination, and most important, remedial measures. Examination of Foods (95). They stated if microbio-
The ultimate goal in food processing plants is to prevent logical criteria (maximum acceptable numbers of
salmonella contamination of food products, which microorganisms) for classes of foods are to be set and
depends on a control program requiring adequate enforced, methodology for sampling and analysis must
sanitation and monitoring of the environment and all be agreed upon. They also said such methods have not
processing steps. been standardized. At least nine different agencies have
Assessment that a salmonella control program is published, or are considering publishing microbiological
effective depends on the reliability of bacteriological methods and criteria for foods. These methods are not in
analytical procedures employed. As Lewis and Hall (99) complete agreement as to analytical procedure. Standard
state, laboratory detection of salmonalle is not an end in methods have not yet been agreed upon for most food
itself toward salmonella control, but it is an important commodities.
and necessary tool for directing and monitoring control The Subcommittee on Food Microbiology (95) has
efforts. It is with this phase of the complex salmonella proposed "reference methods" to serve as guides to the
control effort, laboratory detection of salmonella development of uniform standard methods. They state
contamination, that the present review is concerned. The that " ... a reference method should be the best available
large amount of research reported concerning this topic provided it is practicable. It should not represent
necessitates a literature review to bring the reported compromise between an excellent method and one of
findings of many proponents together into one text for lesser excellence because of expediency. Such limitations
greater understanding and appreciation of the as the reference method has should be recognized and
complexity of the problem. stated." They recommended where a single reference
procedure has not been agreed upon, a study should be
GENERAL DESCRIPTION A1~D DISCUSSION set up in each of several laboratories to compare varia-
OF ANALYTICAL METHODS tions in procedure using specific foods. Acceptable meth-
Identification of sources of food product ods evolved should then be subjected to collaborative
contamination and/or recontamination, and determina- study.
tion that food samples are below detectable levels of The search for more sensitive and rapid salmonella
salmonellae--major problems confronting food analytical techniques has led to a proliferation of
industries, especially poultry, today--can be provided methods which propose new and modified media, both in
only by bacteriological laboratory examination. The the clinical microbiology field and in food and feed
validity of results is directly dependent on laboratory analysis. Research on methodology has been aimed
techniques and media employed in the analyses. toward improvements in specificity, sensitivity,
Numerous salmonella analytical procedures have been simplicity. and rapidity. Although there have been
proposed for food, but not one method has been proven indications one method may be superior to another,
better than all others for all applications. Most accepted Banwart and Ayres (8) stated, in some instances results
methods are cumbersome, time-consuming and obtained from certain samples or from a given quantity of
laborious, and basically involve pre-enrichment, enrich- sample were basis for stating the superiority of a method
ment, selective plating, and biochemical and serological for any sample of any size. Conclusions derived from
tests. Raw foods and fmished products that have not such comparisons do not necessarily mean one method is
been heat processed do not require the pre-enrichment better for all types offood or for all salmonella serotypes.
step and are usually examined by direct enrichment in Another inadequacy of comparative methodology studies
selective broths. was revealed by Taylor and Schelhart (123), who warned
Methods for isolation of salmonellae were first of an inherent lack of correlation between results noted
developed for clinical specimens because of widespread in pure culture studies and observations resulting from
human and animal diseases (42). Later, when food emptrtc practice. They claimed, "nowhere in
became apparent as an important factor in the chain of microbiology is that lack of corroboration any more
salmonellosis infections, microbiologists naturally apparent than in isolation of enteric pathogens from
applied methods shown satisfactory for clinical samples clinical specimens or food products, animal feeds, and
to food samples. It was found these methods did not similar substances." Observations from researcher to
always effectively detect salmonellae in foods and, in researcher, or even analyses by the same person done on
fact, the media were somewhat toxic to the organisms ·the same premises at different times often depict lack of
(53). No ideal procedure has yet been devised for recovery consistency (123).
of all samonella serotypes from all types of food. Also, Many variables have been implicated as important to
there is no single recommended procedure for some salmonella analyses; some are controllable, others are
specific food products, for example, poultry meat which not. These variables, including types of enrichment and
is one of the most frequently incriminated food products. plating media, technical factors causing variations in en-
630 FAGERBERG AND A VENS

richment and plating media, number of pathogens versus as modified by Kauffman (75, 76) and Leit'son's (84)
competitors, salmonella serotypes, sample types, and selenite-F (SF) broth have been used widely to recover
length and temperature ofincubation will be discussed in salmonellae from fecal specimens as well as food
the following literature review. References are made products. They have been widely used to isolate
throughout the review to sample types other than foods; salmonellae from egg products, dried milk, poultry
but it must be realized that findings with other sample products, tlsh meal and other animal feeds, fertilizers,
types, i.e. clinical specimens, are not necessarily exclusive mesenteric glands, sewage, and fecal specimens (70).
for that sample type, but may be quite applicable to food Galton (40) acknowledged that tetrathionate brilliant
products and should not be ignored. green (TETBG), Kauffman's modification of Muller's
tetrathionate broth, has been successfully used to isolate
MECHANISMS OF ENRICHMENT MEDIA
AND REPORTED APPLICATIO~S
salmonellae from a variety of human and animal food
products. Edwards and Ewing (27) endorsed the useful-
Usually if salmonella organisms are present, their
ness oftetrathionate broth for isolating salmonellae from
relative numbers in almost every kind of food product
clinical specimens. Selenite and TETBG were deemed
examined are small compared to greater numbers of
greatly superior to other enrichment media by Smith
non-pathogenic organisms of similar cultural character-
(113). Guinee and Kampelmacher (47) found TETBG
istics. Proof of the presence of salmonellae requires
equal to a modification of selenite, selenite brilliant
isolation of the organism in pure culture from among the green, in isolating salmonellae from porcine feces and
more numerous associated microorganisms. This may be
skin scrapings. Heidrich (56) observed no appreciable
achieved by utilization of enrichment media that through
difference between TETBG and SF in isolating
various mechanisms may favor and enhance growth and salmonellae from fecal and organ samples, whereas
multiplication of salmonellae while other gram-negative Sharma and Packer (III) experimented with artificially
organisms such as coliforms, Proteus, and Pseudomonas contaminated cow and pig feces and found that SF
are suppressed or inhibited. Gram-positive organisms yielded fewer salmonella isolations than did TETBG.
are easily inhibited since they are, in general, quite Contrary to these observations, Smyser and Snoeyenbos
susceptible to the toxic effects of substances in en- (116) observed quite poor TETBG performance with
richment media such as dyes, metals and other selective poultry litter and animal feeds, though an earlier
agents. Galton et al. (42) mentioned it is difficult to publication by Smyser et al. (II 5) indicated tetrathionate
recover salmonellae without selective enrichment if the and TETBG to be better than SF for detecting S.
ratio of normal coliforms to salmonellae is as low as 10:1 typhimurium in artitically contaminated poultry feed
It is generally agreed that enrichment of food samples in and animal by-products. Carlson and Snoeyenbos (JJ)
either selective or non-selective broths is superior to found tetrathionate to be generally better than the
direct plating, especially when very small numbers of selenites for recovery of salmonellae artificially
salmonellae are anticipated (42). An ideal enrichment introduced into chick starter, used poultry litter, and
broth medium, then, should not be too inhibitory to mixed in a test system with fecal microflora. McCullough
support multiplication of salmonellae to detectable levels and Bvrne (87) considered TETBG better than SF for
yet should be selective enough to prevent overgrowth by salmo~ella isolation from artificially infected human
non-pathogens. An enrichment broth, to be acceptable, volunteers, Studies with egg meat, turkey rolls, and
must also meet other criteria. Provided the medium is chicken feces reported by Cox et al. (22) revealed TETBG
prepared and used properly, it must not lose selectivity inferior to other enrichment media tested for salmonella
due to addition of a sample, and should allow recovery of isolation.
all salmonella serotypes. Type of samples inoculated, TETBG is one of the enrichment media recommended
proportion of inoculum to broth, and time and in the Food and Drug Administration's (FDA's)
temperature of incubation must also be considered since Bacteriological Analytical Manual (37) for analysis of
such factors may substantially influence selectivity of raw or highly contaminated meats, animal substances,
enrichment media (42). glandular products, and fish meaL The National
It is apparent from the many modifications in Academy of Sciences (NAS) (95) recommended TETBG
enrichment medium and methodology proposed as one of the enrichments for detection of salmonellae in
throughout the years that no one enrichment medium or foods. Galton et al. (42}, in procedures for meats, other
methodology is perfectly satisfactory (87). No single non-processed food, rendered animal by-products, and
enrichment broth can be depended upon to give feeds suggested TETBG enrichment. The Association of
consistent isolation of salmonellae under all conditions Official Analytical Chemists (AOAC) (5) microbiological
encountered. The literature is tilled with various claims methods for analysis of egg and milk products also
for superiority of one or another medium, with obvious recommended TETBG as one of the acceptable
diversifications, contradictions, and ambiguity of enrichments.
opinions. As stated in the USDA Microbiology Laboratory
Tetrathionate brilliant green and selenite-F media Guidebook (19), TETBG is a peptone medium
Many enrichment media have been proposed for containing bile salts for inhibiting gram-positive
isolation of salmonellae. Muller's (94) tetrathionate broth organisms, brilliant green for inhibiting gram-positive
 DETECTING SALMONELLAE IN FOOD 631

and gram-negative lactose fermenters and tetrathionate studied as extensively as many other enrichment broths,
for toxicity to enterics besides salmonellae. It is highly but United States Department of Agriculture (USDA)
buffered with calcium carbonate so that pH is of no methods (19) recommended TT as the sole enrichment
concern. In their 1970 publication, Palumbo and Alford broth for raw meat, subsequent to lactose pre-enrich-
(98) stated the mechanisms of tetrathionate inhibition ment. In a study to determine a preferred enrichment
were not yet fully known. They stated that the iodides method for egg and diverse food specimens, Montford
probably have no lethal effect. Lethality to bacteria was and Thatcher (92) found TT yielded many coliforms, but
directly related to the concentration of thiosulfate and failed to provide a single colony of Salmonellae.
tetrathionate--only the combination of the two being North and Bartram (96) modified selenite media by
toxic. They stated the lethal effect was a growth-related adjustment of types and amounts of phosphates and
phenomenon--since no killing occurs with non-growing peptones to optimal levels to facilitate isolation of
cells. They related it is known that tetrathionate reacts salmonellae from food products. They also added cystine
with free sulfhydryl groups of enzymes and causes their to enhance growth of salmonell"ae in the presence of large
inactiva:tion and that thiosulfate can also react with amounts of organic material, thus termed selenhe cystine
sulfhydrl groups. They suggested tetrathionate interferes (SC) enrichment broth. SC has been cited in many
with synthesis, activity, or both of sulfur-containing studies and is recommended by the FDA (37) in
enzymes or cell wall and membrane components. conjuction with TETBG for raw meats, by the NAS (95)
Selenite-F is a peptone base medium containing in conjunction with TETBG for isolation of salmonellae
sodium selenite as a selective agent against enterics from foods, also by the AOAC (5) in conjunction with
besides salmonellae. The mechanism of selenite toxicity TETBG for egg and milk products. When SC and
is also not known, but Weiss et al. (128) proposed two TETBG were compared during salmonella analysis of
different general mechanisms, (a) selenite reacts with meat and bone meal, Huhtanen and Naghski (62) found
sulfhydryl groups of cellular components and (b) no difference in total number of positive samples
selenium is incorporated into analogues of sulfur between the two enrichments. SC was found to be one of
compounds (seleno-amino acids may possibly be formed the useful broths for detecting S. typhimurium from
because of metabolic similarity of selenium with sulfur). artificially contaminated poultry feed and animal
They found during early incubation, a sharply increased by-products (115). SC was found to be the preferred
uptake of selenium coincided with high susceptibility to enrichment for egg specimens and diverse foods (92),
selenite broth. They suggested that possibly the relative allowing the most favorable multiplication of
ratios of seleno-amino acids to their sulfur analogues lactose-negative organisms and inhibition of coliforms.
caused the cell protein to become partially or completely
Of the various media tested for selective enrichment of S.
inactive--resulting in less total growth, a reduced growth typhimurium from infected poultry tissues, Yamamoto et
rate or complete cessation of growth. They stated there al. (134) found SC gave the most favorable results. Taylor
was also evidence of a non-metabolic reaction by which and Silliker (126) remarked that in their years of
inorganic selenium compunds exert their toxic effects on experience comparing SC and TETBG, superiority of
organisms.
one or the other could not be claimed because usually
In Leifson's report (84) on his SF medium, he stated one would be superior in one experiment and the other
lactose in the medium served to maintain uniform pH. would prove superior in the next.
He reported that when selenite was reduced by growth of
The Baltimore Biological Laboratory (BBL) Manual of
bacteria, the pH increased and could lessen the toxicity
Products and Laboratory Procedures (6) states TT is an
of the selenite and result in overgrowth of competitors if improved form of TETBG with better defined
the acid produced by fermentation of lactose by
composition than TETBG. TT is greatly enriched with
enterococci and some colon bacilli was not present to additional nutriment by addition of yeast extract.
maintain neutral or slightly lowered pH.
Brilliant green and sodium desoxycholate serve as
inhibitors of gram-positive organisms. Fermentation of
TT and selenite cystine media
dextrose and mannitol, according to Difco (24), aids the
Two important modifications of the two broths, selective property of tetrathionate decomposition in the
Muller-Kauffman's tetrathionate and Leifson's SF were medium.
introduced in the 1950's. Hajna and Damon (51) noted
considerable variation in productivity when tetrathionate Selenite brilliant green and selenite brilliant green sulfa
broth was prepared according to many proposed modifi- media
cations and therefore they made an effort to develop a Significant modifications to the selenite group of
better tetrathionate broth which would include the best media, particularly adapted to the search for salmonellae
features of the broths. They introduced a new tetra- in foods, were introduced by Osborne and Stokes (97,
thionate enrichment, TT broth, that was selective for 120) in 1955. The first modification described. selenite
only the salmonella and Arizona group of bacteria. brilliant green (SBG), was claimed to prevent
Almost twice as many positive salmonella isolations from development of Escherichi and Proteus, yet support
stool specimens were achieved with TT compared to SF luxuriant growth of salmonellae even when the inoculum
(51). Reviewed literature indicates TT broth has not been consisted only of one salmonella cell per ml medium
632 FAGERBERG AND AVENS

(120). Their data were derived from studies with pure Rappaport's medium
cultures having a ratio of 1 salmonella to 100 other
In 1956 Rappaport et al. (106) devised an enrichment
organisms; however, they conceded different results
broth claimed to permit unrestricted development of
might be obtained when isolation of salmonellae from
salmonellae yet inhibit growth of coliforms due to 4 o/o
natural materials was attempted. In fact, they found magnesium chloride and 0.012% malachite green. Mag-
addition of whole egg or egg yolk considerably reduced nesium chloride counteracted the toxic effect of
the selective properties of SBG, and subsequently found malachite green on salmonellae without affecting
the neutralizing effect of egg products on SBG was inhibition of gram-negative contaminants (42). The
eliminated by addition of sulfapyridine (97). Their
originators of Rappaport's (RAP) medium foun? it
selenite brilliant green sulfa (SBGS) enrichment medium
detected almost twice as many cases of salmonellosis as
also claimed successful recovery of salmonellae when
selenite or tetrathionate broths when inoculated with a
only 1 salmonella cell and 100 Proteus or Escherichia
dilute, 1:1,000, suspension of feces (106). Several
were present per ml of medium. SBGS compared to SBG
researchers confirmed the superiority of Rappaport's
and SF, provided best recovery with the widest variety of
magnesium chloride-malachite green medium. Collard
salmonella serotypes from commercial egg whites in tests
and Unwin (18), Hooper and Jenkins (60), Taylor and
conducted by Wells et al (129). Some researchers have
Schelhart (124) Galton et al. (42), and lveson and Kovacs
also tested SBG and SBGS enrichment broths for
(69), in comparative studies with different enrichment
recovery of salmonellae from clinical specimens. Guinee
broths, maintained RAP was superior for isolation of
eta!. (48) determined SBG provided a better medium for
salmonellae from fecal specimens. Since neither SF nor
examination of lymph nodes than did TETBG, whereas
tetrathionate broths were deemed as effective as RAP for
no difference between the two were noted from
detection of salmonellae, Taylor and Schelhart (124)
examination offeces. Kumar et al. (81, 82) had successful
asserted either or both should be replaced by RAP in
results employing SBGS for cloacal swab monitoring of
clinical microbiological laboratories. Taylor and
salmonella-infected breeding flocks, and contaminated
Schelhart also mentioned RAP had become widely used
flock environments and feed. Carlson and Snoeyenbos
in European laboratories, and in food industries even
(13) reported SBGS to be distinctly unsatisfactory for
though it was designed for stool cultures. Iveson et al.
some salmonella strains and allowed major die-off of
(70), sampling contaminated desiccated coconut, and
salmonellae between 24 and 48 h of incubation,
especially at 43 C. Their research involved artificial Anderson and Kennedy (3), working and pure cultures,
also showed superiority of RAP. Burman (10) claimed
mixtures of salmonellae and fecal microflora and
artificial contamination of chick starter and used poultry RAP was superior to selenite broth because it did not
litter. Montford and Thatcher (92) declared negative react with selective plating media to produce inhibitory
results with SBG and SBGS enrichment of frozen whole products or combinations of products which produce a
egg melange. Both enrichments provided many growth inhibition zone at the site of inoculation. Other
coliforms, but no salmonellae on subsequent plating, researcher's results differ: Jacobs et al. (73) compared
possibly due to the high proportion of coliforms to RAP with tetrathionate, TETBG, and SBG in the
salmonellae, 39,000:0.15 cells per gram. Yamamoto et examination of fish meal, and observed no differences in
al. (I 34) considered SBGS too selective and less effective recovery of salmoneiiae among the four enrichments.
than SC and TETBG for salmonella isolation from Sen (1 09) obtained higher salmonellae yield in selenite
turkey tissues and fecal samples. Fagerberg (32) and broth, and Zajc-Satler and Banic (135) found their
Fagerberg and Avens (?3, 34) found SBGS to be superior. preparation of tetrathionate broth better than RAP for
to 11 other enrichments (SC, SF, TETBG, TT, SBG, fecal isolations.
Rappaport's, GN, strontium selenite, strontium chloride, Rappaport et al. (106) devised the peptone-based RAP
brilliant green MacConkey's, and neutral red-lysine-iron- medium with selectivity system of magnesium chloride
cystine) for recovering salmonellae from contaminated and malachite green to inhibit coliforms yet permit
turkey carcasses. development of salmonellae. They stated tha~ hyperto~ic
salt solutions may dehydrate and ki11 bactena and with
Stokes and Osborne (120), originators of SBG and this basis determined ions from magnesium chloride at
SBGS stated that inhibitory properties of the media the 4% level were selective against many organisms
again;t many gram-positive an~ .. gram-nega:ive besides salmonellae. They found malachite green
organisms were the sum of the acttvt:tes of ~ele~tte, (0.012%) was better than other dyes tested for coliform
brilliant green, and sodium taurocholate m combmat10n. inhibition.
They stated the balanced mixture of brilliant green and Gram-negative broth
taurocholate made the medium more selective than SF. Hajna (50) in 1955 devised gram-negative (GN) .broth
Yeast extract was added for additional nutriment and for use as an enrichment medium for stool spectmens
phosphate was added at a level to meet salmonella and reported increased numbers of salmoneiiae and
requirements. They used mannitol instead of lactose to Shigella were isolated. Taylor and Schelhart .(122. 1~3,
help maintain proper pH because salmonellae can !24) conducted several comparative studtes whtch
ferment mannitol. included GN broth. In 1967 they found 97.2% efficacy of
 DETECTING SALMONELLAE IN FOOD 633

GN broth in detection of salmonellae from clinical slightly less intense.

specimens (122). The following year they found GN Neutral red-lysine-iron-cystine broth
comparable to selenite for recovery of salmonellae from Hargrove et a!. (53), in 1971, developed a medium,
clinical specimens (123). In a later report they considered neutral red-lysine-iron-cystine broth, to differentiate
GN broth the best all-purpose broth for use with stool salmonellae from other Enterobacteriaceae and
cultures. but when specifically concerned with organisms commonly found in dairy products and to
salmonellae, found RAP was better than GN (124). shorten the time required for identification. A positive
Rollender et al. (108) achieved 98% salmonella recovery presumptive test for salmonellae in dairy products was
from fecal and urine samples using GN broth. Cox et al. indicated by a color change from red to yellow and/or
(22) tested egg meat, chicken feces and turkey rolls for production of massive black precipitate of iron sulfide
salmonellae and reported only slightly lower recovery after 24 h of incubation. Absence of salmonellae was
with GN compared to RAP and SC. Edwards and Ewing indicated by no color change or no medium blackening.
(27) suggested a combination of media such as selenite Results from several dairy products tests indicated its
and GN should be used in clinical laboratories when usefulness in rapid screening of these foods. Hoben et al.
possible. (59) evaluated the procedure of Hargrove et al. and modi-
The tryptose in GN is nutriment, the phosphates are fied the technique for routine testing of foods, food
for buffering, and selectivity is attained by incorporation ingredients. and feed materials for the presence of viable
of carbohydrates and salts. The sodium citrate and salmonellae. Samples were first enriched in tetrathionate
desoxycholate are bactericial to gram-positive organisms broth, then transferred to a broth they termed
and inhibit growth of coliforms. The greater lysine-iron-cystine-neutral red, for further enrichment
concentration of mannitol over dextrose is to limit and detection of salmonellae by color change. They
Proteus growth and accelerate growth of salmonellae established lysine-iron-cystine-neutral red was useful as a
(24). screening technique to rapidly eliminate salmonella-
Strontium selenite and strontium chloride media negative samples and presumptively identify salmonella-
Iveson and Mackay-Scollay (71) introduced two media, positive samples.
strontium selenite (SrSe) and strontium chloride (SrCl) MAJOR VARIABLE FACTORS ASSOCIATED
M, modifications of which have replaced enrichment WITH ENRICHMENT METHODOLOGY
procedures previously used in their Australian The course of fluid enrichment, as in many
Salmonella Diagnostic and Reference Laboratory. In biological phenomena, can be determined by studying
1969 they found SrCI M (with malachite green) the complex interactions of variables. Manipulations of
comparable to RAP for recovery of a wide range of variables, conducive to attaining the most efficient and
salmonella serotypes from human feces, sewage, pig accurate methodology, must be made for each specific
feces, and glands (71). In 1971, lveson (67) introduced a situation encountered for salmonella enrichment.
less selective SrCIB (without malachite green) for im- Researchers have reported influence of enrichment
proved isolation of Edwardsiella as well as salmonellae media performance by elevated temperatures of
and Arizona. SrSe, also introduced in 1969, was deter- incubation, addition of detergent emulsifiers, and
mined superior to other enrichment media for recovery of agitation or aeration of incubating enrichments.
S. typhi from human fecal matter (71) and useful for a Incubation time, numbers of salmonellae versus
variety of other products (72). Chau and Huang (16) and competitors, sample material, and salmonella serotypes
Chau and Forrest (15) found SrSe superior to SF for also affect the enriching ability of fluid media for
isolation of S. Typhi from clinical specimens. lveson (68) salmonellae.
modified SrSe and reported improved results with the Incubation temperature
SrSe A modification. He concluded SrCl B and SrSe A Harvey and Thomson in 1953 (55) first reported
enrichments allowed improved isolation of salmonellae, selenite broth incubated at 43 C was superior to 37 C
Arizona. Edwardsiella. and Shigella from human, incubation for isolating salmonellae from human fecal
animal and environmental samples (68). samples. They tested 42, 43, and 44 C and mentioned
Iveson and Mackay-Scollay (71) utilized essentially the 43 C was optimum with the proviso that 43 C might be
same basis for development of SrSe and SrCI as safer, as it possibly represented the upper end of the
Rappaport et al. (106) did in development of RAP-that of useful temperature range. Alteration of incubation
incorporating specific ions into an enrichment capable of temperatures is a well known aid in purifying bacterial
recovering salmonellae and inhibiting non-pathogenic cultures, but elevated temperature techniques have
gram-negative organisms. They found the strontium ion mainly been used to isolate thermophiles (54). Since
in the form of a chloride and also in the form of a selenite the report of Harvey and Thomson (55), other investi-
salt to be favorable. Iveson (68) found it unnecessary to gators have endorsed the advantage of elevated
include fermentable carbohydrates, dyes, antibiotics, or temperatures (41 to 43 C) for recovering salmonellae
accessory growth factors in either SrSe or SrCI. from mixed pure cultures; human, poultry, pig, and
Di-sodium hydrogen phosphate was added to maintain reptile fecal specimens; stream, river, sewage, and
optimum pH. Iveson stated the reduction of selenite to abattoir waters; minced meat, sausage, milk, and other
selenium in the SrSe media was similar to that in SF, but foodstuffs; meat and bone meal, animal by-products,
634 FAGERBERG AND A YENS

pathological tissues, and various other specimens. The ture had a significant eflect with naturally contaminated
advantage most often observed has been suppression of poultry litter. Erdman (30) found 43 C better than 37 C
competing contaminants without inhibition of for incubation of SBGS, SC, and TETBG to recover
salmonellae; the salmonellae multiplying more rapidly salmonellae from artificially contaminated ground beef.
and profusely in an environment of reduced competition, Carlson and Snoeyenbos (12) remarked that the
thus producing a relatively pure culture and providing improved rate of salmonella isolation after higher
an advantage for isolation and identification (12, 14, 93, enrichment incubation temperatures compared to
116. 119). 35-37 C enrichment incubation is probably explained by
Some researchers, to ascertain intrinsic principles greater ease of selecting salmonella colonies from plates
responsible for increased salmonella recovery at elevated showing fewer coliforms but not salmonellae. With pure
temperatures, studied population dynamics of salmon- culture growth curves derived from salmonella and E.
ellae and common competitor contaminants in selective coli in SC, Alford and Knight (2) found 42 C incubation
and non-selective media at higher temperatures. shortened the lag phase of E. coli and lengthened the lag
Georgala and Boothroyd (43) tested three salmonella of S. blockley compared to 37 C, both lag phases lasting
serotypes, two paracolon organisms, Proteus, and 5 hat 42 C. Because there was a better lag differential, S.
Escherichia freundii in SF enrichment broth. Growih blockley 4 h and E. coli 8 h at 37 C, Alford and Knight
curves showed at 37 C salmonellae reached the greatest chose not to use elevated temperatures in their further
population and E. freundii and Proteus developed two studies.
logarithms lower than salmonellae. At 43 C salmonella Incubation temperature comparisons by Morris and
growth still remained greater than the other organisms, Dunn (93) using TETBG inoculated with sausage
but less than at 37 C. They stated incubating SF at 43 samples. Smyser et al. (117) testing animal by-products
instead of 37 C appeared to improve selectivity of the and poultry litter in SBGS and TETBG, Radan et al.
medium for salmonellae likely to be found in foods. (103) isolating salmonellae from food and feeds of animal
Carlson et al. (14) studied pure growth of S. montivideo, origin in tetrathionate and selenite broths, Greenfield
Proteus mirabilis, and E.coli in SBGS broth and and Bankier (45) employing SF inoculated from
non-selective trypticase soy broth incubated at 37 and pathological tissues, foodstuffs and various materials
43 C. Growth of all organisms in trypticase soy broth submitted to their veterinary laboratory, Dixon (25)
incubated at 43 C was equal to 37 C growth. SBGS at employing selenite for enrichment of human feces, and
both temperatures was inhibitory to all three organisms, lveson and Mackay-Scollay (72) using SrCI M and SF
but much more to Proteus and Escherichia. Salmonellae with effluent samples, have all demonstrated improved
optimally developed at 37 C in SBGS, compared to 10 4 salmonella isolation at 43 C incubation temperature
fewer salmonella organisms from 43 C incubation. In compared to 37 C. Spino (119) utilized a temperature of
SBGS enrichment there was no ditierence in Proteus 41.5 C with SBGS and tetrathionate for isolating
population between 37 and 43 C, but the Escherichia salmonellae from stream waters and described consistent
population \vas less at 43 than 37 C. SBGS broth recovery of salmonellae at the elevated temperature,
appeared to be sufficiently selective at 37 C, 10 7 to 10 8 whereas no salmonellae were recovered at 37 C. He, as
more salmonella cells than either competitor. An well as Smyser and Snoeyenbos (116), noted indirectly,
elevated temperature of 43 C was unnecessary and from growth on selective plating media, growth of
actually not an improvement in selectivity since at 43 C competing organisms was markedly reduced in
there were only 10 3 to 10 5 more salmonellae than enrichments incubated at elevated temperatures
competitors. Carlson and Snoeyenbos later devoted an compared to 37 C incubation. Smyser and Snoeyenbos
entire study to population kinetics (12). Their study (J 16) analyzing poultry litter in SBGS at 43 C and in
revealed pure culture populations of S. typhimurium, in TETBG at 37 C, found that of 73 positive samples only
trypticase soy broth, TETBG, and SBGS, were greater at two plates showed Proteus from SBGS at 43 C, where-
37 than 43 C during the stationary growth phase. as from TETBG at 37 C all plates were so over-
Increased temperature lengthened the generation time of grown with Proteus they did not even attempt
salmo.nellae in both TETBG and SBGS. Nearly the same selection of colonies for further analysis. Carlson et at.
observations were noted with simulated natural (14) enriching meat and bone meal and poultry litter
contamination in meat and bone meal, but with samples samples in SBGS found the percent positive samples
of naturally contaminated poultry litter they were higher and there was a decrease in number of
demonstrated the advantageous effects of elevated Proteus. Citrobacter. and coliforms on plating media
temperature incubation. With the latter samples, when 43 C enrichment incubation was used compared to
competitor organisms reached a peak after 16 h then 37 C incubation. Enumeration from their samples
decreased in SBGS and TETBG at 43 C, whereas at 37 C indicated only about 100 salmonella organisms per gram.
the competitor population in TETBG was equal to the They noted the improved isolation rate at 43 C would
salmonella population, and greater than the salmonella probably not have been apparent if samples had been
population in SBGS. The rapid die-off of competitor heavily contaminated with salmonellae. since the
organisms during a period of relative stability of the elevated temperature probably facilitated isolation when
salmonella population demonstrated elevated tempera- the organism was present in low numbers by preventing
 DETECTING SALMONELLAE IN FOOD 635

overgrowth by competitors. Wells et aL (130) reported evaluating the alternative of enriching at 43 C and the
more salmonella recoveries from raw milk were made at "analyst may, at his discretion, incubate enrichment
43 C incubation ofTETBG at salmonella inoculum levels media at 43 C."
of 1000, 100, and 10 cells per liter of milk, than at 37 C. Emuls(fying agents
Also, at 43 C frequent recoveries were made from
When Galton (40) enriched sausage for salmonella
samples containing one salmonella organism per liter,
isolation, after incubation she found the heavy layer of
whereas no recoveries were made at this level at 37 C.
fat on the surface made it difficult to obtain a
Burman (I 0) reported the Metropolitan Water Board
satisfactory loopful of material for streaking plates. To
found for samples such as sewage effluent containing
overcome the problem, she added a wetting agent,
large numbers of interfering organisms, incubation of
Tergitol 7 (f-7) to a level in the enrichment broth, 0.6o/o,
selenite cultures at 43 C gave better recovery of
determined by titration to provide emulsification of the
salmonellae than at 37 C, whereas with cleaner samples
fat and give a distinct head of foam. No evidence of
such as river water 37 C was better. Banffer (7) showed
salmonella inhibition was noted with concentrations up
SF enrichment at 43 C was significantly better than' at
to 0.7o/o, rather T-7 appeared to enhance salmonella
3.5 C and seemed appropriate for isolation of
growth. Similar findings were noted with Tween 80. T-7
salmonellae from human convalescent excretors and car- is now commonly used in salmonella enrichments for
riers having few salmonellae. Banffer (7) and Carlson et products of animal origin and other high fat samples.
al. (14) found more salmonella serotypes were isolated Galton et al. (42), in their recommended procedures for
from convalescent excretors and carriers more frequently isolation of salmonellae from foods and feeds, the FDA
at 43 C. Edel and Kampelmacher (26) found 43 C (37). USDA (19), and NAS (95) recommend use ofT-7 in
enrichment improved salmonella isolation from pig feces
enrichments for fatty samples. Fagerberg and Avens (31,
and minced meat, but also indicated other factors such
33, 35) and Chen (17) have found Tween 80 emulsifier
as the nature of the medium played an important role in
better than T-7 or no emulsification for enriching turkey
addition to temperature. Analyzing reptile feces,
skin samples. Morris and Dunn (93), using T-7 at 0.6o/o in
Koopman and Janssen (79) found tetrathionate
TETBG incubated at 37 C for sausage samples, found
incubated at 43 C better than tetrathionate or selenite at
significantly more salmonella isolations compared to not
37 C, but in analyzing dog and cat feces found 37 C
adding T-7. However, no advantage was derived from
better. adding T-7 when enrichments were incubated at 43 C. At
McCoy (86) in 1962 found tetrathionate incubated at
the higher temperature the fat problem was not so
43 C was lethal to salmonellae and most organisms, and troublesome, but results indicated there may have even
isolation of salmonellae from selenite incubated at 43 C been a disadvantage to adding T-7 and incubating the
was significantly less than at 37 C incubation. Aleksic et
enrichments at 43 C. Some authors reported negative
a!. (1), in 1973, reported distinctly better salmonellae
results on adding T- 7. Carlson et al. (14) found the
isolation from carriers when selenite was incubated at
additon of T-7 at 0.6o/o in SBGS inoculated with meat
37 C compared to 43 C. Harvey and Price (54) suggested
and bone meal, incubated at both 37 and 43 C, did not
lack of success with the elevated temperature of 43 C in
increase the number of salmonella isolations at either
some laboratories probably could be explained on the
temperature. They also indicated it was slightly better
basis of difference in sample types or techniques. They
not to add T -7 and incubate the enrichment at 43 C.
mentioned specimens yielding salmonellae easily at 37 C
Montford and Thatcher (92), analyzing Cheddar cheese
enrichment incubation were not likely to yield more at
(48o/o milkfat) in SC, found T-7 had no effect.
43 C and samples containing minimal numbers of
salmonellae, most in need of resuscitation, may be better Agitation during incubation
examined at 37 C. They also stated media preparation Agitation of enrichment cultures during incubation to
was a relevant factor to utilizing 43 C incubation. They facilitate aeration and/or nutrient availability or other
never heat their SF, but rather filter-sterilize it. advantageous physical conditions has not received much
Fagerberg (32) and Fagerberg and Avens (33) found attention in studies of salmonella enrichment
effects of 35-37 C versus 41-43 C enrichment broth methodology. Acceleration of aerobic growth by aeration
incubation were enrichment media dependent for SC, has been widely used in the fermentation industry and in
SBGS, TETBG, TT, Rappaport's, and strontium gro-wth studies, but its potential for speeding up isolation
selenite. SC and TETBG, compared to the other broths, of salmonellae is reported only minimally (2). Fagerberg
showed the least differences between the two and Avens (33, 35) found stationary, compared to 150
temperatures and lent themselves to incubation at either rpm agitation, enrichment was generally better for 13
temperature better than any of the other broths, yet different enrichment broths tested for efficacy in
neither performed as well at 41-43 C. salmonella detection from turkey skin. Chen (17) found
Methods of analysis of the AOAC, FDA, and USDA agitating SBGS enrichments at 150 rpm yielded
(5, 19. 37), do not currently recommend elevated significantly fewer salmonella recoveries than stationary
temperature incubation of enrichments in their incubation. Insalata and Sunga (65) indicated salmonella
procedures for salmonella analysis. The USDA detection by fluorescent microscopy was accelerated by
Microbiology Laboratory Guidebook (19) states they are aeration of enrichment broths. With the objective of
636 FAGERBERG AND A YENS

determining \vhether aeration would alter the sensitivity 48-h incubation compared to 24-h incubation.
of either E. coli or salmonellae, Alford and Knight (2) Incubation periods of 48 h were found unfavorable for
incubated pure cultures of the organisms in SC at 200 most salmonella strains in SBGS inoculated with egg
rpm on a rotary shaker incubator. Though their report products by Osborne and Stokes (97). SBGS cultures
does not precisely indicate enhancement of salmonella incubated 24 h yielded a billion salmonella cells per ml,
isolation resultant from the agitation, they stated the but after 48 h only a few million viable salmonella cells
normal lag differential between E. coli and salmonellae were detected. and greater growth of competitor
in selenite medium was not affected by their aeration organisms was noted. The salmonella population had
procedure. They maintained though, since toxic started to die off by 48 h after full growth. They found
compounds may be formed from some foods when the only advantage of 48-h incubation was greatly
aerated, specifically chocolate as reported by Busta and increased growth of S. pullorum. Sharma and Packer
Speck (11), agitation should be explored for each (Ill) incubated an inoculum of 15 salmonella organisms
different sample material. in 12 ml of liquid cow feces in TETBG and SF. The
Incubation time optimum length ofTETBG incubation was 24-30 h, after
Recommended incubation times for salmonella which a decrease in salmonella isolation efficiency was
enrichment broths range from 16 to 96 h. The National noted. SF was optimally incubated 24 through 48 h with
Academy of Science's booklet (95) summarized no noticeable falling off in efficiency at 36 or 48 h.
incubation periods recommended by various food Carlson and Snoeyenbos (12) studied times for plating
protection agencies. Association of Food and Drug after TETBG and SBGS were incubated at 43 C. They
Officials ofthe United States (AFDOUS) recommends 24 stated although S. typhimurium populations resulting
h for SC and tetrathionate, FDA recommends 24 ± 2 h from inoculums used were at or near peak in 16-24 h, the
for SC and TETBG, USDA suggests 18-24 h for IT, and subsequent decrease of coliform organisms indicated
the American Public Health Association (APHA) that it would be advantageous to delay plating until the
recommends 16-24 h for SC or tetrathionate. The APHA sample had been incubated 32 to 48 h. Chen (17) found
recommends when small numbers of salmonellae are no significant differences in recovery of salmonellae
present, or when initial plates are negative, enrichment among 20-,24-,30-, and 48-h SBGS enrichment times of
should be incubated another 24 h and streaked again on 15 different serotypes used to artificially contaminate
selective media. They recommend streaking from turkey tails. However. she found increased numbers of
enrichment media up to 3 consecutive days. The National recoveries as time increased even though the amount of
Academy of Science's reference method (95) suggest competitor development was greatest at 30 and 48 h.
18-24 h of incubation for SC and TETBG and add that Few researchers report testing enrichment incubation
although it is not required or recommended, restreaking periods ofless than 16 or 18 h. Sharma and Packer (111)
after an additional 24 h of incubation may produce incubated TETBG and SF, initially containing less than
additional positive samples. Galton et al. (42) stated that one salmonella organism per ml, at 37 C for 0, 6, 9, 12,
subculture after at least two different incubation periods, 15, 18, 24, 30. 36, and 48 h. Salmonellae were not
24 and 48 h or 24 and 72 h, usually resulted in isolation recovered from any broth before 12 h of incubation.
of salmonellae from more samples. Galton et al. (42) Incubation periods. 12, 15, and 18 h, of enrichments
endorsed 24- and 48-h incubation of TETBG in their initially containing very few salmonella organisms,
recommended procedure for meats, other raw revealed recoveries, but not as many positives as were
non-processed foods, rendered animal by-products, and found from incubation periods of 24 h or longer. A rapid
feeds. method for detecting salmonellae, including decreased
time for enrichment incubation. would be very
Morris and Dunn (93), testing sausage samples in
important, but as Sperber and Deibel (118) maintain, to
TETBG for 24- and 48-h incubation periods, found best
be acceptable, any improved salmonella detection
results and recommended plating both after 24 and 48 h
method must not only facilitate a decreased time factor,
because 15.4o/o of the positive samples would have been
but it must also be as sensitive and accurate as possible.
missed if only 24-h streaking had been employed.
Carlson et al. (14) reported plating from SBGS Level and proportion of salmonella and
inoculated with meat and bone meal samples after 48 h competitor-organism contamination
yielded greatest number of positives. Guniee et al. (48) Another important factor that must also be considered
obtained more positive results by subculturing SBGS and in enrichment methodology is the level and proportion of
TETBG to plates after 24 and 72 h, but found plating salmonella and competitor-organism contamination.
from two jars of duplicate enrichments at the same time Prost and Riemann (102) remarked it is a well
resulted in approximately the same increase in number established fact a large number of salmonella organisms
of positive findings as did plating from the same single is necessary to produce food poisoning. Experiments
jar at two different times. Huhtanen and Naghski (62), in with human volunteers have demonstrated that to
tests with meat and bone meal, did not show any produce illness or lesions. the quantity of organisms
advantage of 48-h SBGS incubation as far as detecting necessary varies between hundreds of thousands and
positive samples, but found fewer false positives after millions, dependent on the strain of salmonella (102).
 DETECTING SALMONELLAE IN FOOD 637

Although the level of salmonella contamination in foods salmonellae and competitors constitutes a great
is frequently extremely low, even a low level may present disadvantage to the salmonellae. Jameson (74) and
a potential hazard ifthe food is mishandled, for example McCoy (86) found optimum length of enrichment
time-temperature abuse before consumption (9). incubation is a function of the number of salmonellae
According to the Food, Drug, and Cosmetics Act, no present in the sample being examined. McCoy (86) found
salmonella organisms are allowed in food products; there 96 h was necessary for all possible positive recoveries if
is zero tolerance (61). Lewis and Hall (99) admit it is most there were one or two salmonellae per 100 ml of water, 24
attractive to think in terms of salmonellae-free products, and 48 h were necessary to find all positives when there
but the term "zero tolerance" has little actual meaning were three to nine per 100 mi. When the salmonella
and is a "nightmare" to the laboratory. It is impossible count reached 12 per 100 ml, a single plating after 24 h
to be 100 o/o certain that salmonellae are absent since the was adequate. Various enrichment broths were tested by
entirety of the food products would have to be tested, so a Montford and Thatcher (92) for recovering salmonellae
compromise at some lower level of assurance is a from specimens of frozen whole egg melange with a
practical necessity (39). When there is the pressure of concentration of 0.15 to 2 salmonellae cells per gram in
marketing a "salmonellae-free'' product, the microbio- the presence of diverse contaminants including up to
logists' problem becomes frustrating and the laboratory 40,000 coliforms per gram. Only SC allowed recovery of
is faced with the important consideration of sensitivity salmonellae at that contamination level. Chen (17) found
(99). Lewis and Hall (99) cite that some salmonella that when only 1 to 25 salmonella cells per cm 2 of turkey
detection methods will detect less than 10 salmonellae skin were present, even over 1 million competitors per
per 100 g of specimen, while others are reported to re- cm 2 did not prevent salmonella recovery. Wells et al.
cover 10 per gram. They assert that with variations such (130) reported TETBG incubated at 43 C was capable of
as these, the term "salmonellae-free" obviously can allowing recovery at a level of 10 salmonella organisms
mean quite different things in different situations. per liter of raw milk and frequent recovery at the level of
Though they recognize or mention the difficulty, not 1 salmonella cell per liter. Osborne and Stokes (97)
many researchers have devoted efforts toward or developed SBGS that was determined capable of
reported enrichment methodology for very low numbers recovering 1 salmonella organism per ml from egg
of salmonellae in samples. When salmonellae are present products in the presence of 100 Proteus or Escherichia
in large numbers they are readily recoverable by various per ml. The ideal enrichment technique would detect 1
procedures, but when they are present in small numbers, salmonella organism if it were present in the sample
efforts often fail (92, 116). If two persons examine a food being analyzed.
containing low numbers of salmonellae, it is not unusual Serotype specificity
that one will detect the organisms while the other will not Tendency for some enrichment broths to inhibit
(95). Edel and Kampelmacher (27) sent, for evaluation certain salmonella serotypes noticeably enough to cause
purposes, identically contaminated samples to nine concern has been described by several researchers.
different European laboratories. When there was 1 Salmonella cholerasuis, typhi, and paratyphi occupy
salmonella cell per sample, most labs were able to obtain many of the reports of suppression or toxicity of
high percentages of positives when working with "clean" enrichment media. Smith in 1952 (113) found
material, but when there were many competitors and few tetrathionate broth toxic for S. cholerasuis and in 1959
salmonellae, the results became divergent. Isolation (114) reported selenite toxicity also, but found
using a particular enrichment medium is not governed MacConkey's broth with a 1:5,000 solution of brilliant
solely by number of salmonella organisms present; green added was acceptable for its recovery. Leifson (84)
undoubtedly other factors such as degree of noted high toxicity of selenite for S. cholerasuis when he
contamination with extraneous bacteria affect results first described the SF medium. Greenfield and Bigland
obtained (134). Taylor and Schelhart (123), in a (46) also observed unacceptability of SF for S. cholerasuis
discussion of the effect of microbiological numbers on and Sharma and Packer (111) confirmed these findings
variation of results, said detection of salmonellae is that the organism could not be recovered from SF or
influenced by numbers in two basic ways. One, there TETBG, but could be with brilliant green MacConkey's
must be sufficient "absolute" numbers of salmonellae in broth. Edwards and Ewing (27) stated tetrathionate with
the sample to initiate growth in enrichment broth. Two, added brilliant green and SC were apparently of value
the ratio of salmonellae to competitors must not be so only in isolation of salmonellae other than S. typhi. Chau
disparate that competitors overgrow the salmonellae. and Forrest (15) observed marked inhibition of S. typhi
The absolute number of salmonellae necessary to initiate by RAP and often to some degree by tetrationate,
growth and multiplication in inhibitory broths is higher especially if brilliant green was added. They mentioned
than in non-inhibitory broths. Taylor and Schelhart (123) SF was generally accepted as most suitable for S. typhi,
specified that sensitivity correlates with small absolute but found SrSe-A better than SF. Hajna and Perry (52)
numbers and selectivity is a major factor if both reported SF allowed recovery of S. typhi. but Greenfield
salmonellae and competitors occur in high absolute and Bigland (46) reported the serotype did not do well in
numbers or if disparity between relative numbers of SF. Hobbs and Allison (58) and Cook et al. (22) found
638 FAGERBERG AND AYENS

selenite media superior ot tetrathionate for isolation of S. diminution of salmonella numbers was affected more by
paratyphi B. Rappaport and Konforti (105) and Iveson the kind of food than by either strain of salmonellae
et al. (70) determined RAP supported growth of S. inoculated or enrichment employed. Neither an
para(vphi B. Banwart and Ayres (8), studying S. overwhelming number of coliforms nor unfavorable
paratyphi A in pure culture, found definite inhibition in balance in ratio of coliforms to salmonellae affected
tetrathionate broth; fewer ceils were detected after selectivity of tetrathionate or selenite broths as much as
incubation than immediately after inoculation. Smith the addition of foods including gelatin, albumen, egg
(1 13) noted tetrathionate toxicity to S. aborius-ovis, yolk, or dried beef. Galton et al. (42) indicated addition
Greenfield and Bigland (46) found S. gallinarum and S. of relatively large amounts of organic matter, such as
pullorum did not do well in SF, Smyser and Snoeyenbos food samples, may have an adverse effect on selectivity
(116) determined SBGS was better for S. senftenberg and enrichment quality of certain media, depending on
than TETBG, and Osborne and Stokes (97) indicated the type of food. Lewis and Hall (99) stated methods for
SBGS did not support luxuriant growth of S. pullorum in salmonella isolation must be modified slightly or
24 h, but allowed sufficient growth to permit isolation. markedly, depending on the product being analyzed.
Galton et a!. (42) mentioned SF and tetrathionate According to experience of Prost and Riemann (102),
enrichments were toxic for certain serotypes. Erdman technique for recovery of salmonellae must be adapted to
(30) stated the serotype involved influenced choice of material that is being examined. Hurley and Ayres (63)
enrichment broth, since in comparisons of SBGS, SC, reported addition of egg reduced selectivity of SF,
and TETBG for enrichment of artificially contaminated tetrathionate, and TETBG. SBGS broth was developed
ground beef SBGS was found best for S. by Osborne and Stokes (97) because additon of egg
schwartzengrund. S. typhimurium, and S. dublin; SC markedly reduced effectiveness of SBG. Diversity of
best for S. senftenberg: and TETBG and SBGS equally sample materials relative to acidity, salinity, sugar, fat,
effective of S. newport and S. worthington. Significant moisture. etc., causing various alterations in enrichment
differences in recovery of serotypes, dependent on medium pH, nutritional composition, or other physical
whether SC or tetrathionate was used, where shown in a factors that selectivity and/or sensitivity of the medium
study by Huhtanen and Naghski (62). They found relies upon. can result in variable success in subsequent
signiticantly higher isolation of serogrou p C 1 from isolation of salmonellae. Galton et al. (42), USDA (19),
tetrationate and of G, 35, and poly D from selenite. and FDA (37) all recognize different samples require
Fagerberg and Avens (36) artificially contaminated different methodology for isolation of salmonellae and
turkey skin with less than 50 organisms per ml of 61 list recommended procedures for specific products.
different serotypes, used SBGS enrichment plated to MECHANISMS OF PLATING MEDIA
BGS, and were able to recover all serotypes tested except AND REPORTED APPLICATIONS
S. paratyphi C. Low numbers of positive recoveries were
noted with S. cholerasuis, S. gallinarum, S. missouri, and Plating media employed for salmonella isolation
S. typhi. Banwart and Ayres (8) studied pure culture should, as with enrichment media, exert minimum
growth cures of S. paratyphi A, S. bredeny. S. inhibition toward salmonellae and maximum bacterio-
typhimurium, S. oranienburg, S. pullorum, S. anatum, static action toward competitor organisms. While
S. give, and S. worthingion. in SF and found a decrease enrichment media allow salmonellae an opportunity to
in number of viable cells during initial incubation with grow and multiply, the resulting culture usually includes
all organisms; the decrease was significant with S. a mixed flora and thus selective plating media have been
anatum. S. give was inhibited approximately 3 h in devised to allow isolation of salmonellae from other
tetrathionate during initial incubation. The SF inhibition organisms (19). Many such media have been described
of all the organisms was deemed by Banwart and Ayres and numerous modifications have been proposed. These
(8) undesirable in the case of samples with low counts of generally consist, as outlined in the USDA Microbiology
salmonellae since destruction during the lag growth Laboraiory Guidebook (19), of a basic nutritional
phase may result in failure to isolate the organisms. medium with dyes, antibiotics, bile salts, and/or other
Carlson and Snoeyenbos (13) reported that host-adapted chemicals to inhibit growth of undesirable micro-
serotypes grew poorly or not at all in TETBG, organisms and an indicator system to reveal by
tetrathionate, SC, and SBGS. Certain media may characteristic color. colonies likely to be salmonellae.
function better than others in certain aspects of Differentiation of salmonellae is usually by one of two
salmonella recovery, but literature reports indicated no characteristics, hydrogen sulfide production or the
ability to ferment a certain carbohydrate such as lactose
one medium functions maximally for all serotypes.
or sucrose. Media range from only slightly selective, such
Sample type as MacConkey agar, through moderately selective and
The deleterious impact of addition of certain food differential, including Salmonella-Shigella (SS), Desoxy-
materials to enrichment broths was demonstrated by cholate Citrate (DC), Hektoen Enteric (HE), and Xylose
Silliker and Taylor (1 12). They studied effects on Lysine Desoxycholate (XLD) agars, to highly selective
salmonella isolation of adding various food products to and differential, for example Bismuth Sulfite (BS),
selenite and tetrathionate enrichments and found the Brilliant Green (BG), and Brilliant Green Sulfa (BGS)
 DETECTING SALMONELLAE IN FOOD 639

agars. The highly selective plating media are said by coliforms when BG is prepared properly. The FDA (37).
Galton (42) to be more effective, in both clinical and food USDA (19), NAS (95), APHA (110), AFDOUS (4), AOAC
microbiology, in the isolation of salmonellae from (5), and Galton et al. (42), in their recommended methods
enrichment broths since they are designed to inhibit for detection of salmonellae, all include BG or BGS as
many gram-negative organisms that are frequently one, if not the only, plating media.
present and not sufficiently suppressed on less selective BG and BGS are basically peptone and yeast extract
media to prevent overgrowth. media with brilliant green dye to suppress gram-positive
Brilliant green and brilliant green sulfa agars organisms and coliforms. Sulfapyradine in BGS
suppresses Proteus multiplication to a greater extent
Brilliant green agar, first developed in 1925 by
than salmonella. The selectivity and differential system is
Kristensen et al. (80) and later modified by Kauffman
dependent on two fermentable carbohydrates, lactose
(76). and BGS (41. 97) have been found highly
and sucrose, and an acid indicator system utilizing
satisfactory as plating media for salmonellae by many
phenol red. Salmonellae are unable to ferment lactose or
researchers. Banwart and Ayres (8) and Georgala and
sucrose and their metabolism of peptone results in
Boothroyd (43) with pure salmonella cultures; Goo et al.
alkaline end products, therefore, since phenol red
(44), Taylor et al. (127), Montford and Thatcher (92), and
produces red color in the alkaline state, salmonellae
Morris and Dunn (93) testing various foods; Yamamoto
colonies are red (18). Sugar fermenters such as coliforms
et al. (134) testing turkey tissues; and Smyser et al. (115)
are yellowish since their production of acid causes phenol
testing poultry feed and animal by-products described
red to turn yellow.
best salmonella recoveries from BG and BGS agars.
Guinee and Kampelmacher (47) and Galton (40) Bismuth su(fite agar
reported excellent suppression of other enteric organisms Bismuth Sulfite agar (BS), another one of the most
on BG agar. Galton et al. (42) mentioned Escherichia widely known and used plating media, was developed by
and Enterobacter types grow on BG, but are often Wilson and Blair in 1927 (132). In the early 1930's,
suppressed as much as or more than on other selective Wilson and Blair (133), Hajna and Perry (52), and
agars. Montford and Thatcher (92) and Yamamoto et al. Gunther and Tuft (49) acknowledged the superiority of
(134) reported that even though coliform growth was high BS for salmonella detection. Several recommended
on BG, the salmonella colonies were readily procedures for salmonella isolation include BS as one of
differentiated from Proteus, Pseudomonas, and other the plating media: AOAC (5), FDA (37), NAS (95),
extraneous lactose-negative organisms. Sulfadiazine or APHA (110), and Galton et al. (42). Edwards and Ewing
sulfapyridine added to BG agar reportedly inhibits (27) stated salmonella usually grow well on BS and it is a
Proteus and pseudomonads (19, 41, 42, 95, 97). good media for their isolation, including isolation of a
Yamamoto et al. (134) found Proteus was inhibited when rare strain of salmonella that ferments lactose rapidly .
sulfapyridine was added to BG, though the rate of The indicator system in BS is not based on fermentation
salmonella recovery was not improved. Montford and of any carbohydrate, rather glucose, fermentable by all
Thatcher (92) found BGS highly dependable and more Enterobacteriaceae, is included. Most salmonellae
advantageous to use than BG because of increased appear as black colonies on this medium because of
coliform inhibition and high percentage of apparent hydrogen sulfide production. BS is often recommended
salmonella colonies picked that were confirmed as for use in conjunction with one of the plating media
salmonellae. Smyser et al. (117) noted essentially no reliant on the inability of salmonellae to ferment lactose,
difference between BG and BGS for isolating for example BG agar. so as not to miss the types that
salmonellae. They found growth of competitors was ferment lactose rapidly (19). There are occasional
occasionally more restricted on BGS than BG; however, aberrant salmonella and Arizona serotypes that ferment
at times salmonella growth was sufficiently restricted on lactose, and as well, serotypes that do not produce
BGS to make colony selection ditlicult. In another study hydrogen sulfide. BS allows growth of many
Smyser et al. (115) preferred BG because BGS was too gram-negative enteric commensal bacteria, but is an
restrictive. Wells et al. (129) noticed smaller salmonella especially effective medium for isolation of S. typhi (42).
colonies on BGS compared to BG. Fagerberg (32) and Wells et al. (129) reported BS was more inhibitory than
Fagerberg and Avens (33) tested seven different plating BG to organisms other than salmonellae. Montford and
media and 12 different enrichment broths in all Thatcher (92), on the other hand, found BS allowed
combinations and found BGS was superior to all other development of high numbers of coliforms and yielded
plating media regardless of the enrichment type. Despite high numbers of false positives (apparent salmonella
the excellent record, some problems have been noted in colonies picked that could not be confirmed as
controlling the selectivity of BG and BGS. Hobbs (57) salmonellae). Erdman (30) found in research with
noted BG was only slightly selective and Read and Reyes artificially contaminated ground beef that BS was
(!07) found some lots of commercial BG (to which they successful when samples were salmonella-positive, but
added sulfadiazine) would not support satisfactory gave large numbers of false-positives when samples were
growth of salmonellae. Galton et al. (42) claimed negative-an occurrence not found with brilliant green
maximal yields of salmonellae and ~nhibition of agars. BS was deemed unsatisfactory by Yamamoto et ai.
640 FAGERBERG AND AVENS

(134) because of frequent formation of ill-defined extract. DCLS utilizes sodium desoxycholate for
colonies and rapid decomposition of the medium. Cox et inhibitory action against gram-positive organisms and
al. (23) reported very low positive recovery percentages coliforms, while bile salts, rather than purified chemicals,
with BS compared to other media. are incorporated into SS for the same purpose. Sodium
The BS medium does not depend on any fermentative desoxycholate has strong solvent action on some
action for recognition of salmonellae, but rather utilizes gram-positive bacteria similar to bile but more powerful
a hydrogen sultide indicator system. Wilson and Blair (83). The sodium chloride in DCLS, brilliant green in SS,
(132) stated that the principle of BS medium rested in the and sodium citrate in both DCLS and SS exert inhibitory
ability of some salmonellae to reduce sulfite to sulfide in action except toward gram-negative pathogenic bacilli.
the presence of glucose. For the hydrogen sulfide Both media utilize carbohydrates which change the
indicator system to yield black salmonella colonies, thev neutral red in the media to a red color when fermented.
stated that a source of iron and phosphate wer~ Salmonellae colonies are colorless on these media. DCLS
necessary. They related that the inhibitory action to employs both lactose and sucrose while SS incorporates
coliforms was due to the combined action of the bismuth only lactose. A hydrogen sulfide selective system is also
sulfite precipitate and sodium sulfite solution. employed in the SS medium by incorporation of ferric
Desoxycholate citrate lactose sucrose, and sodium citrates and sodium thiosulfate.
Salmonella-Shigella, and MacConkey agars MacConkey's is a simple peptone-based medium
which has a small amount of bile salts and crystal violet
Desoxycholate citrate agar (DC) introduced by Leifson
to inhibit gram-positive organisms. Lactose is
(83), modified by Hynes (64) and Salmonella-Shigella
incorporated as a fermentable carbohydrate and neutral
~gar (SS), essentially a modification of the DC agar (6),
red as an indicator. Lactose fermenters are red in color
two moderately selective and differential plating media,
due to neutral red reacting in acidic conditions, but since
and MacConkey agar (MAC) (88) a slightly selective agar,
salmonellae do not ferment lactose, colonies are
have been used often for isolation of salmonella
uncolored and transparent.
organisms. MAC, being less inhibitory than most other
plating media for salmonella isolation, is not considered Xylose lysine desoxycholate and Hektoen
useful for streaking from enrichment broths where the enteric agars
flora is likely to be highly mixed and contain Proteus, Generally, if a plating media is highly inhibitory to
but is useful at particular times when a relatively some Enterobacteriaceae, i.e. coliforms, there is a
non-inhibitory medium is desired (19). The FDA (37), concomitant loss of sensitivity for fastidious pathogens
AFDOUS (4), APHA (I 10), and AOAC (5) include SS as such as salmonellae; thus, plating media embodying the
one of the plating media in their recommended extremes of sensitivity and selectivity have been chosen
procedures. Cox et al. (22) and McCullough and Byrne and recommended for standard methodology (125). Two
(87) found SS superior to other agars tested for isolating recently introduced plating media, xylose lysine
salmonellae. Late lactose fermenters or lactose-negative desoxycholate (XLD) developed by Taylor (121) and
organisms such as Proteus and Pseudomonas can Hektoen enteric (HE) developed by King and Metzger
develop on DC and SS media, and cannot always be (77), were formulated to retain necessary sensitivity for
differentiated from salmonellae, causing relatively high less hardy pathogens, yet be selective enough to prevent
numbers of false-positives-the major objection most overgrowth by the coliform majority. Isenberg et al. (66)
authors express. Pollock and Dahlgren (101) found the found the performance of XLD and HE very similar,
more inhibitory SS yielded more false-positives than the both media readily permitting recovery of salmonellae.
less inhibitory MAC medium, generating more wasted Pollock and Dahlgren (101) found XLD and HE superior
effort and fewer isolates than any other plating media toSS and MAC agar, yielding more salmonella isolations
they tested. Rollender et al. (108) and Yamamoto et al. and fewer false-positives. McCarthy (85) found XLD
(134) reported difficult differentiation between salmonel- equaled the best performances of traditional plating
lae, Proteus. and other non-lactose fermenting organisms media for salmonella isolation. Taylor and Schelhart in
and high incidence of false-positives with SS agar. 1967 (122) reported XLD was better than BG and SS
Montford and Thatcher (92) found good coliform agars and in 1969 (124) found superiority of XLD over SS
inhibition with SS, but high coliform growth on DC agar. and BS agars. In 1971 (125) they found XLD and HE
Differentially, though, approximately 60% of apparent quite similar, though XLD showed slightly fewer
salmonella colonies picked from either of these media false-positives than HE. Rollender et al. (108) claimed
were confirmed as not salmonellae. Testing diluted, pure supenonty of XLD and mentioned salmonella
salmonella enrichment cultures plated onto SS, identification was greatly facilitated by the distinctive
desoxycholate citrate lactose sucrose agar (DCLS), a morphological appearance. King and Metzger (77) found
modification of the DC formula, BS and BG, Banwart good growth of salmonellae and Shigella, inhibition of
and Ayres (8) reported SS and DCLS showed many non-pathogens and good colonial differentiation
significantly lower numbers of colonies of every between major groups on HE agar and reported
salmonella serotype tested. superiority over SS agar (78). Goo et al. (44), comparing
DtLS and SS are media based on peptone and meat BG and HE, found BG detected significantly more
 DETECTING SALMONELLAE IN FOOD 641

salmonella isolations than HE, though about 11 o/o of the and 37 C among most authors or any food protection
total positives would not have been detected without HE agency method, but length of incubation varies among
and HE was far more selective than BG in inhibiting type of media and the agency method involved.
non-lactose fermenters. The USDA (19) recommends Generally, 18 to 24 h of incubation is recommended for
employing XLD as one of the plating media for most plating media, except BS, for which 48 h is
salmonella isolation from foods. advisable. Both FDA (37) and AOAC (5) suggest if BG,
XLD medium differentiates salmonellae from other SS, or BS do not have typical or suspicious salmonella
enterics on the basis of xylose fermentation, lysine colonies or do not have growth within 24 h, they should
decarboxylation, and hydrogen sulfide production (121). be incubated an additional 24 h. Yamamoto et al. (134)
Three fermentable sugars, xylose, sucrose, and lactose could not readily detect S. typhimurium growth on BG
are incorporated. After salmonellae exhaust the xylose, until after 48 h of incubation and it is mentioned in the
they decarboxylate lysine and cause a reversion to BBL manual (6) that BG should be incubated 48 h.
alkaline pH which is indicated by red color with the MacConkey agar, according to the Difco Manual (23),
phenol red indicator. Lysine-positive coliforms do not should be incubated 16 to 18 h because prolonged
revert to the red color because fermentation of lactose incubation may lead to confusion of results due to
and sucrose in the medium produces excess acid. excessive colony growth.
Desoxycholate is added to prevent Proteus swarming.
HE agar employs a gram-positive and coliform SECONDARY TECHNICAL FACTORS ASSOCIATED
inhibitor system of bile salts and sodium desoxycholate, WITH ENRICHMENT AND PLATING MEDIA
fairly high amounts of carbohydrates (sucrose, lactose, Multiple methods and media
and salicin) and peptones for growth promotion and a
Several other factors involved in salmonella
hydrogen sulfide indicator system. King and Metzger
enrichment or plating methodology reportedly influence
(77) related that this medium was different from existing
recoverability of salmonellae, i.e. the use of multiple
media in that the indicator system was of minimal
media. media preparation techniques, and storage and
toxicity to enteric pathogens and the amounts of
tempering of media. In the opinion of McCullough and
peptones and carbohydrates were greater. Bromothymol
Byrne (87), Jameson (74), Malo and Cousineau (89), Sen
blue and Anchade's are used as indicators and the
(109), Ramadan et al. (104), BBL (6), and Difco (23), the
non-lactose fermenting salmonellae appear blue-green to
more methods and media employed, the greater the
blue. Arizona and Proteus also appear blue-green to
success in isolating salmonellae. Huhtanen and Naghski
blue, Shigella and Providencia appear green, Pseudomo- (62) comparing SC and tetrathionate, Taylor and Silliker
nas appear green or brown, and coliforms appear salmon (126) using TETBG and SF, Radan et al. (103)
colored.
considering tetrathionate and selenite broths, Edwards
and Ewing (27) citing GN and selenite, and Hajna and
SEROTYPE SPECIFICITY INVOLVING Damon (51) indicating SF and TT, recommended
PLATING MEDIA concurrent use of two different enrichment media
The efficacy of some plating media employed for because of the inability of one alone to detect all positive
salmonella isolation, as with some enrichment media, is samples. Huhtanen and Naghski (62) felt it important to
dependent on the salmonella serotype involved. Banwart use two different enrichment broths because of serotype
and Ayres (8) testing S. paratyphi A, S. bredeny, S. specificity and Taylor and Silliker (126) stated to use
typhimurium, S. oranienburg, S. pullorum. S. anatum, either TETBG or SF alone must be considered a
S. give, and S. worthington, observed DCLS was calculated risk. Koopman and Janssen (79) conceded the
somewhat inhibitory to all organisms, SS significantly use of more than one enrichment medium would give
inhibited all organisms, and all organisms were further increases in yield, but felt this increase would not
significantly inhibited on BS, exceptS. paratyphi A. and counterbalance the additional work and materials. They
S. pullorum. BG agar was found least inhibitory to all indicated a single specific standard method should be
organisms. Galton (40) said in her experience, use of BG devised for each different specific specimen to avoid the
was indicated for maximum yield of salmonellae other need for parallel use of different enrichment media.
than S. typhi following enrichment of clinical and food The situation is similar for plating media; Banwar-t
specimens. Edwards and Ewing (27) also maintained S. and Ayres (8) and King and Metzger (78) conceded to the
typhi was not expected to be isolated on BG agar. They advantage of plating on multiple media, but alleged that
claimed BS was the most effective medium yet devised time and materials consumed in such a process were
for isolating S. typhi. In their comparative study, Chau deterrants. Edwards and Ewing (27) found it advisable
and Forrest (15) found XLD was clearly less efficient to employ a variety of plating media; at least one plate of
than SS for isolation of S. typhi. each of slightly selective, selective, and highly selective
media. The USDA (19) maintained the most thorough
INCUBATION TIME OF PLATING MEDIA job of salmonella isolation can be done when two plating
The recommended incubation temperature for any media are used, one reliant on hydrogen sulfide
selective plating media does not vary from between 35 production and one employing lactose for a differential
642 FAGERBERG AND A YENS

sugar. They mentioned since highly selective media tend the other ingredients. Moats and Kinner (91) reported
to be inhibitory for salmonellae also, usually a less commercial BG agar was quite variable in selectivity and
selective medium additionally is employed, the choice mentioned variations in autoclave performance and
being governed by the flora likely to be found in the geometry of containers may alter effective heat treatment
product being examined. King and Metzger (78), the medium receives, which in turn may affect inhibitory
attempting to develop an agar, HE, that could singly yet properties of BG if brilliant green dye is added before
realiably facilitate identification of salmonellae with a autoclaving.
minimum of further investigation, found even though References to preparation of BS agar are quite
HE was superior toSS, without the SS agar some positive ambiguous and contradictory. Cook (20) and Hobbs (57)
samples would have been missed. observed BS was too inhibitory for salmonella serotypes
The general opinion that multiple media offer greater other than S. typhi unless it was refrigerated at 4 C for at
reliability toward salmonella recovery than single least 24 h before streaking. To isolateS. typhi they found
enrichment and plating media, permeates through the medium should be streaked immediately after
recommendations of various food protection agencies. pouring. The USDA (19) stated BS should be poured and
The FDA (37), AFDOUS (4), and NAS (95) recommend used promptly, but could be stored refrigerated up to 3
concurrent use of tetrathionate and SC enrichments. days. Edwards and Ewing (28) stated BS was too
The FDA (37) and AOAC (,5) recommend BG, SS, and inhibitory after refrigeration storage for more than 24 to
BS agars, AFDOUS (4) suggests BG and SS, the NAS 36 h. McCoy (86) said for S. typhi freshly poured BS
(95) recommends BG and BS and USDA (19) re- plates were satisfactory, but they were too inhibitory for
commends XLD and BGS, used simultaneously for other serotypes. He stated many serotypes grow well on
recovering salmonellae after specimen enrichinent. the medium stored at 4 C for 4 to 5 days, inhibition
Taylor and Schelhart (122) assert it is axiomatic that lessening with age. He claimed aging was essential for
maximal isolations of salmonellae result from use of production of characteristic salmonella colonies, adding
multiple enrichment broths and plating media. They cite that salmonella colonies could be easily recognized on
two contributing factors: one factor is quantitative in properly aged media after 18 h of incubation, but freshly
that replication reduces the effect of sampling error and poured or insufficiently aged plates required 48 h.
the other is qualitative in that since each specialized Galton eta!. (42) did not find it necessary to "age" BS.
medium, enrichment or plating, represents a
Miscellaneous technicalfactors
compromise between desired sensitivity and selectivity,
no one is completely adequate. Jameson (74) commented Taylor and Schelhart (123) discussing factors causing
on multiple isolation procedures: variations in plating media, mentioned variations due to
different brands of media in which different peptones
"Since a limit must always be set on the number of different media and bile salts or dyes, or even variations from one lot
and procedures which can be used during routine examinations, a number to another, may be great sources of
corresponding limit has also to be set on the combined efficiency of dissimilarities in results. They stressed since plating
procedures used. Not even a very laborious combination of methods
can be expected to yield quite 100 %of the isolations that are tech·
media attempt an effective compromise between
nically possible. When this rather unpalatable conclusion is ac- sensitivity for salmonellae and selectivity against com-
cpeted, a way is opened up for compromises between endeavor and petitors, extremely minute chemical changes "wreak
reward, in terms of salmonella isolations. Judicious application of havoc" on delicate sensitivity-selectivity balance. When
principles and methods assessed in these papers might be expected any of the media are allowed to dehydrate, the Eh is
sometimes to lead to modifications in techniques which increase their
efficiency. and in other cases to modifications which reduce labor
changed and inhibitor concentration is increased, but if
without diminishing efficiency." freshly poured plates are replaced into the mylar bags
and resealed, aberrant results can be deterred. SBGS
Preparation methods broths, one self-prepared from basic ingredients
Galton et al. (42) stated the problem of finding according to formula and two brands available in
overgrowth and even Proteus swarming on BG agar may dehydrated form. were compared by Fagerberg
be due to improper sterilization-a critical factor in the (unpublished) for effectiveness in salmonella recovery
preparation of this medium. They designated 15 min at and distinctively different results were achieved from
15 psi for BG agar. Difco (23) and BBL (6) recommended each.
BG sterilization at 121 C for 15 min. Poelma (100), Huhranen and Naghski (62), Smyser et al. (117), and
USDA (19), and FDA (37) suggest 1-liter portions of Carlson and Snoeyenbos (12) recommended tempering
BG agar be autoclaved 12 min at 121 C. Galton et al. (42) enrichment broths to respective incubation temperatures
Difco (23), and FDA (37) mentioned that additional before inoculation on the basis that a major heating time
heating tends to decn:ase selectivity, FDA also lag may occur if the incubation period is started with
mentioned that less heating increased selectivity of the media at temperatures much below the desired incuba-
media. Guinee and Kampeimacher (47} attributed tion temperature. Utilizing enrichment media directly
excellent suppression of coliform organisms and Proteus from refrigerator storage (3 C) was found comparable by
and selection of salmonellae on BG agar in part to Fagerberg and Avens (31) to media tempered to
adding brilliant green dye solution after sterilization of incubation temperatures before inoculation of SF, SC,
 DETECTING SALMONELLAE IN FOOD 643

TETBG, or TT with turkey skin samples. of salmonella from human faecal specimens in selenite enrich-
The experience of Whitehill and Gardener (131) ment medium at incubation temperatures of 37° C. or 43° C. [in
German, English abstract]. Zbl. Bakt. Hyg.,l. Abt. Orig. A 225:
indicated once a dehydrated SF medium bottle is 27-33.
opened, it should be stored over calcium chloride in a 2. Alford, J. A., and N. L. Knight. 1969. Applicability of aeration
dessicator. Their experiments showed broths prepared and delayed addition of selenite to the isolation fo salmonellae.
from exposed SF cannot support growth of small Appl. Microbial. 18:1060-1064.
3. Anderson, K., and H. Kennedy. 1965. Comparison of selective
numbers of S. typhimurium and media freshly prepared
media for the isolation of salmonellae. J. Clin. Pathol. 18:747-
from unopened stock gave best recovery. They 749.
emphasized that commercial producers of media warn it 4. Association of Food and Drug Officials of the United States.
should be stored in a cool place. Zajc-Satler et al. (135) 1966. AFDOUS Quart. Bull., Suppl. Issue.
found tetrathionate broth freshly prepared from 5. Association of Official Analytical Chemists. 1970. Methods of
analysis. J. Assoc. Off. Anal. Chern. 53:845-852.
infusion broth supported salmonella growth better than
6. Baltimore Biological Laboratory. 1973. BBL manual of products
dehydrated media. North and Bartram (96) felt greater and laboratory procedures. BBL Division of Becton, Dickinson
emphasis should be placed on the necessity of and Company, Cockeysville. Maryland.
determining the efficiency of enrichment broths when 7. Banffer, J. R. J. 1970. A method for the isolation of Salmonella
using new batches. from scanty positive human faeces. Antonie van Leeuwenhoek
The point of concern with these seemingly minor J. Microbial. Serol. 36:586.
8. Banwart, G. J., and J. C. Ayres. 1953. Effect of various enrich-
details in laboratory tests can only be as accurate as the ment broths and selective agars upon the growth of several
techniques employed allow. Not all laboratories are species of Salmonella. Appl. Microbiol. 1:296-301.
equally proficient at isolating salmonellae, because of 9. Blood, R. M. 1969. Salmonellae in foods. B.F.M.l.R.A. Scien-
differences in attention to technical details (99). tific and Technical Surveys #60.
10. Burman, N. P. 1967. Recent advances in the bacteriological
Frequently, workers neglect to report technical details of
examination of water, p. 185-212. In C. H. Collins (ed.), Progress
their studies. making it difficult to assess the effects or in microbilogical technique. Plenum Press. New York.
import of various secondary techniques, and possibly 11. Busta, F. F., and M. L. Speck. 1968. Antimicrobial effect of
causing some of the inconsistency observed among cocoa on Salmonella. Appl. Microbial. 16:424-425.
authors attempting to duplicate and confirm reported 12. Carlson, V. L., and G. H. Snoeyenbos. 1972. Relationship ·of
population kinetics of Salmonella typhimurium and cultural
research results.
methodology. Am. J. Vet. Res. 33:177-184.
CONCLUDING REMARKS 13. Carlson, V. L., and G. H. Snoeyenbos. 1974. Comparative
Methodology optimum for salmonella enrichment and efficacies of selenite and tetrathionate enrichment broths for
the isolation of salmonella serotpes. Am. J. Vet. Res.
plating should be determined for every different food
35:711-718.
product. For any given food product the best enrichment 14. Carlson, V. L., G. H. Snoeyenbos, B. A. McKie, and C. F.
medium and plating medium combination for the Smyser. 1967. A comparison of incubation time and tempera-
greatest positive recovery of serotypes expected to be ture for the isolation of Salmonella. Avian Dis. 11:217-225.
encountered should be found. All variables including 15. Chau. P. Y .. and C. R. Forrest. 1972. Further study of strontium
selenite and selenite-F broths for the isolation of S. typhi J. Clin.
enrichment emulsification, time and temperature of
Pathol. 25:966-969.
incubation, agitation during incubation, plating media 16. Chau, P. Y., and C. T. Huang. 1970. Carriage rate of salmonella
inoculation methods, and time and temperature of serotypes in hospital patients and comparison of enrichment
incubation should be tested and decisions made as to the media for their isolation. Trop. Med. 13:94-101.
most effective combinations for detecting very low 17. Chen, A. Ai-Ti. 1975. Improved enrichment and plating
methodology for detecting salmonella in meat. Ph.D. Disserta·
numbers of salmonellae amidst high levels of expected
tion. Colorado State University.
competitors. Determining optimum methodology for 18. Collard. P., and M. Unwin. 1958. A trial of Rappaport's
individual types of food products is a tedious and medium. J. Clin. Pathol. 11:426-427.
laborious task, but such research is indeed an essential 19. Consumer and Marketing Service. 1969. Microbiology labora-
part of the complex salmonella control efforts. Once tory guidebook. U.S. Dept. of Agriculture, Washington, D.C.
20. Cook, C. T. 1952. Comparison of two modifications of bismuth-
researchers determine optimum methods for any given sulphite agar for the isolation and growth of Salmonella typhi
food, they should submit their results through proper and Salmonella typhimurium. J. Pathol. Bacterial. 64:559-566.
channels so collaborative and/or pilot studies can 21. Cook, G. T., B. R. Frisby, and W. H. H. Jebb. 1951. The routine
confirm the proposed method as superior to others. The use of selective and enrichment media for the isolation of
proven superior method for any specific food should be salmonellae. Monthly Bulletin of the Ministry of Health and· the
Public Health Laboratory Service 10:89-95.
adopted as a "standard method" and recommended for 22. Cox, N. A., B. H. Davis, J. H. Kendall, A. B. Watts, and A. R.
use by all analytical laboratories. Public health agencies Colmer. 1972. Salmonella in the laying hen. 3. a comparison of
and organizations should work together and support various enrichment broths and plating media for the isolation
analytical methodology investigators and help them of Salmonella from poultry feces and poultry food products.
channel results appropriately to ultimately derive Poultry Sci. 1312-1316.
23. Difco Laboratories, Inc. 1953. Difco manual of dehydrated cul-
standardized methods useful to all food analysts for most ture media and reagents for microbiological and clinical lab-
effectively protecting public health. oratory procedures. Detroit, Michigan.
REFERENCES 24. Difco Laboratories, Inc. 1972. Difco supplementary literature.
1. Aleksic, S., R. Rohde, and A. QuddusKhan. 1973. The isolation Detroit, Michigan.
644 FAGERBERG AND AVENS

25. Dixon. J. M. S. 1961. Rapid isolation of salmonellae from of variations of the enrichment method for the detection of
faeces. J. Clin. Pathol. 14:397-399. Salmonella, Antonie van Leeuwenhoek J, Microbiol. Serol.
26. Edel. W., and E. H. Kampelmacher. 1969. Salmonella isolation 28:417-427.
in nine European laboratories using a standardized technique. 48. Guinee, P. A. M., E. H. Kampelmaeher. and H. H. M.
Bull. WHO 41:297-306. Hoejenbos-Spitout. 1965. Further studies on the influence of
27. Edwards, P. R .. and W. H. Ewing. 1972. Identification of variations in the enrichment method for the detection of
enterobacteriaceae. 3rd. ed. Burgess Publishing Co., salmonellae. Antonie van Leeuwenhoek 1. Microbiol. Serol.
Minneapolis. 31:1-10.
28. Edwards, P. R., and M. M. Galton. 1967. Salmonellosis. 49. Gunther, C. B., and L. Tuft, 1939, A comparative study of
Adv. Vet. Sci. 11:1-63. media employed in the isolation of typhoid bacilli from feces
29. Edwards, P. R., M. M. Galton, P. S. Bachman, and C. E. and urines. J. Lal,l. Clin. Med, 24:461-471.
McCall. 1965. A perspective of salmonellosis. Communicable SO. Hajna, A. A.1955. A new enrichment broth medium for gram-
Disease Center, Atlanta. negative organisms of the intestinal group. Public Health Lab.
13:83-89.
30. Erdman, E. E. 1974. ICMSF methods studies. IV. International 51. Hajna, A. A .. and S. R. Damon. 1956. New enrichment and
collaborative assay for the detection of Salmonella in raw meat. plating media for the isolation of Salmonella and Shigella orga-
Can. J. Microbiol. 20:715-720. nisms, Appl. Microbiol. 4:341-345.
31. Fagerberg, D. J., and J. S. Avens. 1973. Optimum salmonella 52. Hajna, A. A., and C. A. Perry. 1938. A compartive study of sel-
enrichment method for poultry meat. Poultrv Sci. 52:2026. ective media for the isolation of typhoid bacilli from stool speci-
(Abstr.) - mens. L Lab. Clin. Med. 23:1185-1193.
32. Fagerberg, D. J. 1974. Optimum enrichment and plating 53. Hargrove, R. E .. F. E. McDonough, and R. H. Reamer. 1971. A
methods for detecting low levels of salmonellae in poultry meat. selective medium and presumptive procedure for detection of
M.S. Tesis. Colorado State University. Salmonella in dairy products. J. Milk Food Techno!. 34:6-12.
33. Fagerberg. D. J., and J. S. Avens. 1974. Optimum enrichment 54. Harvey. R. W. S., and T. H. Price. 1968. Elevated temperature
incubation of enrichment media for the isolation of salmonellas
and plating methods for salmonella recovery from poultry
from heavily contaminated materials. L Hyg. Camb. 66:377-
carcasses. XV World's Poultry Congress and Exposition Pro- 381.
ceedings and Abstracts. p. 445-446. 55. Harvey, R. W. S., and S. Thomson. 1953. Optimum tempera-
34. Fagerberg. D. J., and J. S. Avens. 1974. Superior enrichment ture of incubation for isolation of Salmonella. Mon. Bull. Min.
and plating methods for salmonella recovery from poultry Health Public Health Lab. Serv. 12:149-150.
carcasses. Poultry Sci. 53:1923. (Abstr.) 56. Heidrich, D. 1963. Comparative evaluation of tetrathionate
35. Fagerberg. D. J.. and J. S. Avens. 1974. Improving salmonella broth and selenite broth for the demonstration of Salmonella in
detection on turkey meat. Research Highlights of the Animal bovine feces and in bacteriological meat inspection [in German,
Sciences Department, Colorado State University Experiment English summary]. Deut. Tieraerztl. Wochenschr. 70:595-601.
Station. General Series 938:41-42. 57. Hobbs, B. C. 1963. Techniques for the isolation of salmonellae
36. Fagerberg. D. J., and J. S. Avens. 1976. Recovery of 61 from eggs and egg products. Ann. Inst. Pasteur 104:621-637.
salmonella serotypes from artificially contaminated turkey 58. Hobbs, B. C., and V. D. Allison. 1945. Studies on the isolation
of Bact. ryphosum and Bact. paratyphosum B. Gt. Brit. Minis-
skin. Avian Dis. 20:195-198.
try of Health Public Lab. Serv. Monthly Bull. 4:40-46.
37. Food and Drug Administration. 1972. Bacteriological analytical 59. Hoben, D. A., D. H. Ashton, and A. C. Peterson. 1973.
manual. 3rd ed. li .S. Dept. of Health, Education and Welfare, A rapid, presumptive procedure for the detection of Salmonella
Washington, D.C. in foods and food ingredients. Appl. Microbiol. 25:123-129.
38. Foster, E. M. 1969. The problem of salmonellae in foods. 60. Hooper, W. L., and H. R. Jenkins. 1965. An evaluation of
Food Techno!. 23:1178-1182. Rappaport's magnesium chloride/malachite green medium in
39. Foster. E. M. 1971. The control of salmonellae in processed the routine examination of faeces. J. Hyg. 63:491-495.
foods: a classification and sampling plan. J. Assoc. Off. Anal. 61. House Committee on Interstate and Foreign Commerce and the
Chern. 54:259-266. Senate Committee on Labor and Public Welfare. 1973. Com-
40. Galton, M. M. 1961. Laboratory procedures for the isolation of pilation of selected public health laws (Joint committee print,
93rd Congress, 1st session). Vol. 2 Federal Food, Drug and
Salmonella from human and animal food products. Proceedings Cosmetics Act. U.S. Government Printing Office, Washington.
65th Annual Meeting, l:.S. Livestock Sanitary Assoc. 435-438. 62. Huhtanen, C. N., and J. Naghski. 1972. Effect of type of en-
41. Galton. M. M .. W. D. Lowery, and A. V. Hardy. 1954. richment and duration of incubation on Salmonella recovery
Salmonella in fresh and smoked pork sausage. J. Infect. Dis. from meat-and-bone meal. Appl. Microbiol. 23:578-585.
95:232-235. 63. Hurley, N. A., and J. C. Ayres. 1953. A comparison of six enrich-
42. Galton, M. M., G. K. Morris, and W. T. Manin. 1968. ment media for isolating Salmonella pullorum from egg
products. Appl. Microbiol. 1:302-306.
Salmonellae in foods and feeds. Review of isolation methods
64. Hynes, M. 1942. The isolation of intestinal pathogens by selec-
and recommended procedures. li .S. Dept. of Health, Educa- tive media. J. Pathol. Bacteriol. 54:193-207.
tion. and Welfare/Public Health Service. Communicable Dis- 65. Insalata, N. F. and F. C. A. Sunga. 1969. Improved immuno-
ease Center, Atlanta. Georgia. fluorescence technique for the detection of salmonellae in foods.
43. Georgala, D. L., and M. A. Boothroyd. !965. A system for Bacteriol. Proc. p. 2.
detecting salmonellae in meat and meat products. J. Appl. 66. Isenberg. H. D., S. Kominos, and M. Siegel. 1969. Isolation of
Bacteriol. 28:206-212. of salmonellae and shigellae from an artificial mixture of fecal
bacteria. Appl. Microbiol. 18:656-659.
44. Goo, V. Y. L.. G. Q. L Ching. and J. .M. Gooch. 1973. 67. Iveson, J. B. 1971. Strontium chloride B and E. E. enrichment
Comparison of brilliant green agar and Hektoen enteric agar broth media for the isolation of Edwardsiella, Salmonella, and
media in the isolation of salmonellae from food products. Appl. Arizona species from tiger snakes. J. Hyg. Camb. 69:323-330.
Microbiol. 26:288-292. 68. Iveson, J. B. 1973. Enrichment procedures for the isolation of
45. Greenfield. J., and J. C. Bankier. 1969. Isolation of Sal- Salmonella. Arizona and Shigella from faeces. J. Hyg. Camb.
monella and Arizona using enrichment media incubated at 35 71:34'1-361.
69. lveson, J. B., and N. Kovacs. 1967. A comparative trial of Rap-
and 43° C Avian Dis. 13:864-871.
paport enrichment medium for the isolation of salmonellae
46. Greenfield. J., and C. H. Bigland. 1970. Selective inhibition of from faeces. J. Clin. Pathol. 20:290-293.
certain enteric bacteria by selenite media incubated at 35 and 70. Iveson, J. B., N. Kovacs. and W. Laurie. 1964. An improved
43° C. Can. J. Microbiol. 16:1267-1271. method of isolating salmonellae from contaminated desiccated
47. Guinee. P. A. M .. and E. H. Kampelmacher. 1962. Influence coconut. J. Clin. Pathol. 17:75-78.
 DETECTI~G SALMO~ELLAE IN FOOD 645

71. Iveson, J. B., and E. M. MacKay-Scollay, 1969. Strontium recherche du Bacille typhique et des paratyphiques. Comp.
chloride and strontium selenite enrichment broth media in the Rendus Soc. Bioi. 89:434-436.
isolation of Salmonella. J. Hyg. Camb. 67:457-464. 95. National Academv of Sciences. 1971. Reference methods for the
72. Iveson, J. B., and E. M. MacKay-Scollay. 1972. An evaluation of microbiological e~amination of foods. Food Protection Com-
strontium chloride, Rappaport and strontium selenite enrich- mittee, Food and Nutrition Board. Division of Biology and Agri-
ment for the isolation of salmonellas from man, animals, meat culture, National Research Council.
products and abattoir effluents. J. Hyg. Camb. 70:367-384. 96. North, W. R., and M. T. Bartram. 1953. The efficiency of
73. Jacobs, J., P. A. M. Guinee, E. H. Kampelmacher, and A. van selenite broth of different compositions in the isolation of
Keulen. 1963. Studies on the incidence of salmonella in im- Salmonella. Appl. Microbial. 1:130-134.
ported fish meal [in English]. Zentralb. fur Veterinltrmedizin 97. Osborne, W. W., and J. L. Stokes. 1955. A modified selenite
10:542-550. brilliant-green medium for the isolation of Salmonella from egg
74. Jameson, J. E. 1962. A discussion of the dynamics of salmonella products. Appl. Microbiol. 3:295-299.
enrichment. J. Hyg. Camb. 60:193-205. Palumbo, S. A., and J. A. Alford. 1970. Inhibitory action of
98.
75. Kauffman, F.1930. Die Technik der Typhenbestimmung in der tetrathionate enrichment broth. Appl. Microbial. 20:970-976.
Typhus-Paratyphus-gruppe [in German, English summary.] 99. Pennsylvania State University. 1968. Proceedings: Salmonellae
Zentralb. Bakteriol. Parasitenk. Infektionskr. Hyg. Abt. Orig. in foods conference. College of Agric., Univ. Park, Pa.
119:152-160. 100. Poelma. P. L. 196 7. Detection and identification of Salmonella
76. Kauftman, F. 1935. Weitere Erfahrnngen mit der kombinierten from dried whole egg. dried egg yolk. and dried egg white:
Anreicherungsverfahren fur Salmonella-bacillen [in German, report of a collaborative study. J. Assoc. Off. Anal. Chern.
English summary]. Z. Hyg. Infektionskr. 117:26-32. 50:753-763.
77. King, S., and W. I. Metzger. 1968. A new plating medium for 101. Pollock, H. M., and B. J. Dahlgren. 1974. Clinical evaluation of
the isolation of enteric pathogens. I. Hektoen Enteric agar. enteric media in the primary isolation of Salmonella and
Appl. Microbial. 16:577-578. Shigella. Appl. Microbial. 27:197-201.
78. King, S., and W. I. Metzger. 1968. A new plating medium for 102. Prost, E., and H. Riemann. 1967. Food-home salmonellosis.
the isolation of enteric pathogens. II. Comparison of Hektoen Ann. Rev. Microbial. 21:495-528.
Enteric agar with SS and EMB agar. Appl. Microbial. 16:579- 103. Radan, M., V. Fuchs, S. Bejerano, and S. Finsterbusch. 1970.
581. The comparative efficacy of selenite and tetrathionate media in
79. Koopman, J.P., and F. C. J. Janssen. 1973. The occurrence of the isolation of Salmonella from some foods and feeds of animal
salmonellas and lactose-negative Arizonas in reptiles in the origin [in English]. Refuah Veterinarith 27:14-18.
Netherlands, and a comparison of three enrichment methods 104. Ramadan. F. M., S. El-Hawaary, F. A. Saleh, and D. W.
used in their isolation. J. Hyg. Camb. 71:363-371. Darweesh. 1971. Salmonella versus Providencia on specific
80. Kristensen, M., V. Lester, and A. Jurgens. 1925. Use of tryp· media. J. Hyg. Epid., Microbial. Immunol. 15:227-228.
sinized casein, brom-thymol blue, brom-cresol-purple, phe- Rappaport, F., and N. Konforti. 1959. Selective enrichment me-
105.
nol-red and brilliant green for bacteriological nutriet media. dium for paratyphoid bacteria. Inhibitory and growth promot-
Brit. J. Exp. Pathol. 6:291-299.
ing factors. Appl. Microbiol. 7:63-66.
81. Kumar, M. C., H. R. Olson, L. T. Ausherman, W. B. Thurber, Rappaport, F., N. Konforti, and B. Navon. 1956. A new en-
106.
M. Field, W. H. Hohlstein, and B. S. Pomeroy. 1972. Evalua- richment medium for certain salmonellae. J. Clin. Pathol.
tion of monitoring programs for Salmonella infection in turkey
9:261-266.
breeding tlocks. Avian Dis. 16:644-648. 107. Read, R. B., Jr., and A. L. Reyes. 1968. Variation in plating
82. Kumar, M. C., M.D. York, J. R. McDowell, and B.S. Pomeroy. efficiency of salmonellae on eight lots of brilliant green agar.
1971. Dynamics of Salmonella infection in fryer roaster turkeys. Appl. Microbial. 16:746-748.
Avian Dis. 15:221-232. 108. Rollender. W., 0. Beckford, R. D. Belsky, and B. Kostroff.
83. Leifson, E. 1935. New culture media based on sodium desoxy- 1969. Comparison of xylose lysine desoxycholate agar and
cholate for the isolation of intestinal pathogens and for the MacConkey agar for the isolation of Salmonella and Shigella
enumeration of colon bacilli in milk and water. J. Pathol. from clinical specimens. Am. J. Clin. Pathol. 51:285-287.
Bacterial. 40:581-599. 109. Sen. R. 1964. An assessment of the relative efficiency of several
84. Leifson, E. 1936. New selenite enrichment media for the isola- media in the isolation of shigellae and salmonellae from faeces.
tion of typhoid and paratyphoid <Salmonella) bacilli. Am. J. Indian J. Pathol. Bacterial. 7:93-101.
Hyg. 24:423-432. 110. Sharf, N. M., ed. 1966. Recommended methods for the
85. McCarthy. M.D. 1966. The use of xylose lysine desoxycholate microbiological examination of foods, 2nd ed. American Public
agar as a medium for the isolation of intestinal pathogens. Health Association, Inc., New York.
N. Z. J. Med. Lab. Techno!. 20. Sharma. R. M., and R. A. Packer. 1969. Evaluation of culture
86. McCoy, J. H. 1962. The isolation of salmonellae. J. Appl. 111.
media for the isolation of salmonellae from feces. Appl.
Bacteriol. 25:213-224.
87. McCullough,~- B., and A. F. Byrne.1952. Relative etliciency of Microbial. 18:589-595.
different culture mediums in isolation of cenain members of 112. Silliker, J. H., and W. I. Taylor. 1958. Isolation of salmonellae
from foods samples. II. The effect of added food samples on
the Salmonella group. J. lnf. Dis. 90:71-75.
88. MacConkey, A. 1905. Lactose-fermenting bacteria in faeces. J. the performance of enrichment broths. Appl. Microbial. 6:228-
Hyg. 5:333-379. 232.
113. Smith, H. W. 1952. The evaluation of culture media for
89. Malo, R., and J. G. Cousineau. 1970. Evaluation offour enrich-
the isolation of salmonellae from feces. J. Hyg. SO: 21-36.
ment media for the demonstration of avian salmonella [in
French, English summary]. Canad. J. Comp. Med. 34:111-114. 114. Smith, H. W. 1959. The isolation of salmonellae from mesen-
90. Microbiological Subgroup of the USDA Food Safety Commit- teric lymph nodes and faeces of pigs, cattle, sheep, dogs and cats
tee. 1970. Food protection by the department of Agriculture. and from other organs of poultry. J. Hyg. 57:266-273. ·
A review of the NAS-NRC report, an evaluation of the salmonel- 115. Smvser, C. F., J. Bacharz, and H. Van Rockel. 1963. Detection
la problem. U.S. Depanment of Agriculture, Washington, D.C. of Salmonella typhimurium from artificially contaminated poul-
91. Moats, W. A., and J. A. Kinner. 1974. Factors affecting select- try feed and animal by-products. Avian Dis. 7:423-434.
116. Smyser, C. F., and G. H. Snoeyenbos. 1969. Evaluation of
ivity of brilliant green-phenol red agar for salmonellae. Appl.
Microbial. 27:118-123. several methods of isolating salmonellae from poultry litter and
92. Montford, J., and F. S. Thatcher. 1961. Comparison of four animal feedingstuffs. Avian Dis. 13:134-141.
methods of isolating salmonellae from foods, and elaboration 117. Smyser, C. F., G. H. Snoeyenbos, and B. McKie. 1970. Isola-
of a preferred procedure. J. Food Sci. 26:510-517. tion of salmonellae from rendered by-products and poultry litter
93. Morris, G. K., and C. G. Dunn. 1970. Influence of incubation cultured in enrichment media incubated at elevated tempera-
temperature and sodium heptadecyl sulfate (Tergitol No. 7) on tures. Avian Dis. 14:248-254.
the isolation of salmonellae from pork sausage. Appl. Microbiol. 118. Sperber, W. H., and R. H. Deibel. 1969. Accelerated procedure
20:192-195. for Salmonella detection in dried foods and feeds involving only
94. Muller, L. 1923. Un nouveau milieu d'enrichissement pour Ia broth cultures and serological reactions. Appl. Microbial.
646 FAGERBERG AND AVENS

17:533-539. choice of media for the detection and enumeration of Sal-

119. Spino. D. F. 1966. Elevated temperature technique for the monella. Appl. MicrobioL 6:189-193.
isolation of Salmonella from streams. Appl. Microbiol. 14:591- 128. Weiss, K. F.. J. C. Ayres, and A. A. Kraft. 1965. Inhibitory ac-
596. tion of selenite on E. cob: P. vulgaris. S. thompson. J. BacterioL
120. Stokes, J. L., and W. W. Osborne. 1955. A selenite brilliant 90:857-862.
green medium for the isolation of Salmonella. Appl. Microbiol. 129. Wells, F., D. Bergquist, and R. H. Forsythe. 1958. A compari-
3:217-220. son of selective media for the isolation of Salmonella from com-
12L Taylor, W. I. 1965. Isolation of shigellae I. Xylose lysine agars; mercial egg white solids. AppL MicrobioL 6:198-200.
new media for isolation of enteric pathogens. Am. J. Clin. 130. Wells, J. G., G. K. Morris, and P. S. Brachman. 1971. New
Pathol. 44:471-475. method of isolating salmonellae from milk. Appl. Microbiol.
122. Taylor, W.l., and D. Schelhart. 1967. Isolation of Shigellae. IV. 21:235-239.
Comparison of plating media with stools. Am. J. Clin. Pathol. 131. Whitehill, A. R., and B. L. Gardner. 1970. Evaluation of
48:356-362.
selenite-F broth as a selective medium for Salmonella. Dev.
123. Taylor, W. L, and D. Schelhart. 1968. Isolation of shigellae.
VI. Performance of media with stool specimens. Appl. Ind. MicrobioL 12:249-252.
Microbial. 16:1387-1393. 132. Wilson, W. J., and E. M. MeV. Blair. 1927. Use of a glucose
bisumth sulphite iron medium for the isolation of B. typhosus
124. Taylor, W. !., and D. Schelhart. 1969. Isolation of shigellae.
and B. proteus. J. Hyg. 26:374-391.
VII. Comparison of Gram-Negative broth with Rappaport's en-
richment broth. Appl. Microbiol. 18:393-395. 133. Wilson, W. J., and E. M. MeV. Blair. 1931. Further experience
125. Taylor, W. I., and D. Schelhart. 1971. Isolation of shigellae. of the bismuth sulphite media in the isolation of Bacillus
VIII. Comparison of xylose lysine desoxycholate agar, Hektoen typhosus and B. paratyphosus B from faeces, sewage and water.
Enteric agar, Salmonella-Shigella agar, and eosin methylene J. Hyg. 31:139-161.
blue agar with stool specimens. Appl. Microbiol. 21:32-37. 134. Yamamoto, R., W. W. Sadler, H. E. Adler, and G. F. Stewart.
126. Taylor, W.I., and J. H. Silliker. 1962.Isolation of salmonellae 1961. Comparison of media and methods for recovering Sal-
from food samples. IV. Comparison of methods of enrich- monella typhimurium from turkeys. Appl. Microbiol. 9:76-80.
ment. Appl. Microbiol. 9:484-486. 135. Zajc-Satler, J., and S. Banic. 1965. Efficacy of ditierent selec-
127. Taylor, W. I., J. H. Silliker, and H. P. Andrews. 1958. Isolation tive media for the isolation of salmonellae from faeces. J. Clin.
of salmonellae from food samples. I. Factors atiecting the Pathol. 18:750-751.

Now
Simple! available
0 0,002 0,003 0,006 0,007 0,008 in the U.S.!
NEGAT£VE POSITIVE

Arapid, accurate test for Penicillin residue in milk

~p
0 81VOL8SL
I -h
I Standard diffusion test for the determination of antibiotic residues in milk
A complete self-contained kit. Contains nutrient-indicator tablets,
dosing syringe, disposable syringe tips and ampules containing
seeded agar.
It's easy to use. Ampules are broken at the neck, a nutrient-indi-
cator tablet is added, 0.1 ml sample added via syringe, ampule
placed in a water bath of 64" (±) 2'C.
It's rapid. Readings arc taken after 2v2 hours or after I Y2 hour
with I hour '·pre-incubation" period.
It's sensitive. Penicillin concentrations of 0.003 I. C or less per
ml. of mi!k always yield a negative test result (entirely yellow),
while levels of 0.006 l.U or higher give a positive result (entirely
purple). In-between concentrations give yellow-purple and purple
colorations.
It's economical. Ideal for screening of loads and practical to per-
form for one or many determinations.
It's diverse. Also sensitive to most antibiotics used in lactating
cattle.
Manufactured bv

I
, Exclusive distributors for the USA
Gist-Brocades nv Enzyme Development Corporation
r• Postbus 1 Delft
IIIII The Netherlands
2 Penn Plaza, New York, N.Y. 10001
Telephone: (212) 736-1580