Numerical taxonomy of staphylococci

Abstract

Over 200 staphylococci from human and animal sources and representatives of established species of Staphylococcus, Micrococcus and Planococcus were compared in a numerical phenetic survey using 115 unit characters. Data were analyzed using the Jaccard coefficient and the unweighted pair group method with averages algorithm. Cluster composition was not markedly affected by test error, estimated as 3.49%. The staphylococci were assigned to eighteen clusters containing four or more strains and to three single member clusters. Most of the clusters were distinct and homogeneous though two were divided into subclusters. Some of the clusters and subclusters were equated with the established taxa S. aureus, S. capitis, S. cohnii, S. epidermidis, S. haemolyticus, S. hominis, S. hyicus, S. saprophyticus, S. sciuri subspecies lentus, S. sciuri subspecies sciuri, S. simulans, S. warneri and S. xylosus, the remaining ones may represent the nuclei of additional centres of variation. The numerical data also cast doubts upon the reliability of some of the tests recommended for the identification of coagulase-negative staphylococci.