Much variation can occur when using different of Salmonella plating methods.

Culturing for Salmonella

When culturing for naturally occurring Salmonella, food microbiology laboratories may use two or more selective enrichment broths and two or more selective plating media to reduce the likelihood of false negative results. Many laboratories are only interested in determining the absence or presence of Salmonella and may test as few as one typical isolate. 

Comparing plating media

The objective of this study was to compare different plating media for cultivation of broiler carcass Salmonella and detection of serotypes. 

We collected broiler carcasses just prior to chilling in a commercial processing plant, rinsed the carcasses and cultured for Salmonella. In total, 49 whole broiler carcass rinsates confirmed to be Salmonella positive were used to inoculate selective broths which were streaked onto selective agar plates after overnight incubation. 

The two selective agar plating media used in this study were brilliant green sulfa (BGS) and Xylose Lysine Tergitol 4 (XLT-4).  We picked as many as 12 isolated Salmonella suspect colonies from each carcass (six colonies per plating medium). 

All isolates were subjected to classical serotyping using Salmonella antisera. To limit variation, the laboratory work in this study was done by one technician with more than 20 years of experience in isolating Salmonella from poultry.

Experimental results

From the 49 carcass rinses, we confirmed a total of 236 Salmonella isolates. Four of the serotypes (Kentucky, Berta, Kiambu and Mbandanka) accounted for more than 80% of the Salmonella isolated with Kentucky being the predominant serovar.

It was not the intention of this study to show that one plating medium was better than the other, but rather to determine if the same or different assortment of serovars were detected on the two different plating media.

 Four serotypes accounted for 80% of the 236 Salmonella isolates detected in the study.


The table shown here depicts the number of times the two plating media had the same or different assortment of serotypes. For these two plating media, significantly more


S. Kentuckywas recovered from XLT-4 than from BGS. AlsoS. Kiambuwas detected 15 times on XLT-4 but went completely undetected on BGS.

In another study, we had observed a similar variability with these two plating media. In that study, S. Lille was isolated 34 times from BGS but only once from XLT-4 plates. Therefore, we decided to take an in-depth look at why this was happening.

When S. Lille and S. Kiambu were streaked and studied on these two plating media, an explanation of this variation was determined. S. Lille was a weak hydrogen sulfide producer and therefore did not appear to be a typical, black-centered Salmonella colony on XLT-4 and was overlooked. S. Kiambu was a strong hydrogen sulfide producer so it appeared to be a typical colony on XLT-4 but formed a smaller than average Salmonella colony on BGS and hence went undetected on those plates.


If this much variation is seen with just two plating media, one can imagine how much variation could exist when considering the dozens of media combinations recommended by various domestic and international regulatory agencies. Other factors such as multiple rinsing of a carcass, use of neck skin sample instead of whole carcass rinse, and the possibility of encountering stressed Salmonella could affect Salmonella detection.

Therefore, regulatory agencies should proceed with caution when trying to compare or correlate Salmonella isolates detected from multiple sources, in multiple laboratories or using multiple methods.

Authors: N.A. Cox, M.E. Berrang, S.L. House, A. Hinton, Jr., J.E. Line, and L.T. Wiggins; U.S. National Poultry Research Center.