This numerical taxonomic study examined 222 bacterial strains representing members of the Actinomycetaceae family and related taxa using 124 biochemical and physiological characters. Strains were analyzed using numerical taxonomic techniques with various similarity coefficients and average linkage clustering. Most recognized species formed distinct phenetic groups with useful differential characters. Actinomyces israelii was particularly well-defined as a discrete cluster containing 59 strains. Actinomyces naeslundii and A. viscosus grouped together but showed internal differentiation into subclusters, suggesting possible subspecific variation. Notably, A. bovis strains clustered atypically with representatives from other genera including Bifidobacterium, Corynebacterium pyogenes, and Erysipelothrix rhusiopathiae, rather than with other Actinomyces species. Bacterionema matruchotii and Rothia dentocariosa formed distinct tight clusters that associated only distantly with other Actinomycetaceae family members. Arachnia propionica formed a well-defined phenon separating into two subclusters corresponding to serotypes 1 and 2. The study provides practical identification tests and clarifies taxonomic relationships within these clinically important fermentative actinomycetes.

Key findings

• Actinomyces israelii formed a highly distinct cluster with 59 strains showing good internal homogeneity across multiple similarity coefficients
• A. naeslundii and A. viscosus grouped together but subdivided into multiple subclusters, suggesting subspecific or serotype-associated variation
• A. bovis clustered anomalously with non-Actinomyces genera (Bifidobacterium, Corynebacterium pyogenes, Erysipelothrix) rather than with other Actinomyces species
• Bacterionema matruchotii and Rothia dentocariosa formed distinct, tight clusters that associated only at low similarity levels with other Actinomycetaceae members
• Arachnia propionica phenon subdivided into two serologically distinct subgroups despite maintaining overall phenetic coherence