Title: Probability of identifying different salmonella serotypes in poultry samples Authors
Submitted to: Poultry Science Association Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: March 15, 2010
Publication Date: July 11, 2010
Citation: Cason Jr, J.A., Cox Jr, N.A., Buhr, R.J., Bourassa, D.V., Richardson, L.J. 2010. Probability of identifying different salmonella serotypes in poultry samples. Poultry Science Association Meeting Abstract. Technical Abstract: Recent work has called attention to the unequal competitive abilities of different Salmonella serotypes in standard broth culture and plating media. Such serotypes include Enteritidis and Typhimurium that are specifically targeted in some regulatory and certification programs because they cause a large proportion of human salmonellosis. Common lab methods recommend selecting 3 to 5 colonies per plate, but surveys show that many laboratories pick and identify only 1 probable Salmonella colony from poultry samples. To explore the implications of Salmonella serotypes surviving and growing at different rates during culture and isolation, spreadsheet formulas were used to calculate binomial and multinomial probabilities of picking serotypes present at various ratios on plates, assuming that 100% of picked colonies are Salmonella. When 2 serotypes are present in equal numbers, 6 colonies must be picked to have a 95% probability of finding both serotypes. To identify 3 serotypes under the same conditions, 11 colonies must be picked. If a serotype is outnumbered 10 to 1 by another serotype on a plate, a ratio that has been reported in the scientific literature, 32 colonies must be picked to have a 95% probability of finding the minority serotype. Relatively small survival and growth rate differences can produce large changes in the likelihood of picking a colony of a particular serotype even when that serotype was present in the original sample in equal numbers, so picking one colony per plate can give a distorted picture of what serotypes are in samples. Given the labor and expense of isolating and serotyping suspect Salmonella colonies, methods are needed for culturing specific serotypes of interest.