As a result of a plague epizootic in the San Bruno Mountains, a mammalectoparasite survey was undertaken in the vicinity in spring, 1954. Trapping was conducted in 25 designated areas, using lines of 25 or 50 Museum Special snap traps. The principal vegetation types represented were chaparral, coastal scrub, woodland and grassland. Small mammals were most abundant in chaparral and coastal scrub, very few in dense woodland or grazed grassland. There were often high populations in ungrazed grassland with shrub clumps or tall forbs.
Peromyscus maniculatus was the most abundant and widespread species. Reithrodontomys megalotis occurred infrequently but was negatively correlated with the abundance of P. maniculatus. The overall population of Microtus californicus was low and colonies were scattered; greatest abundance was in ungrazed grassland with forbs or shrub clumps. Other species taken were Peromyscus californicus, Mus musculus, Thomomys bottae, Rattus norvegicus, Sorex trowbridgei, and Sorex vagrans. There were many Rattus norvegicus on hog farms adjoining the U. S. P. H. S. study area.
Fifteen species of fleas were obtained in the survey. Opisodasys keeni was predominant on P. maniculatus. Malareus telchinum was predominant on Microtus californicus but was also abundant on P. maniculatus and was generally the most ubiquitous flea species. Catallagia wymani occurred frequently on P. maniculatus and M. californicus. Hystrichopsylla sp. and Atyphloceras multidentatus were important primarily as nest fleas of M. californicus and probably of Peromyscus. Rattus norvegicus trapped from hog farms bore almost entirely Nosopsyllus fasciatus.
No effect of habitat differences was apparent for fleas on P. maniculatus. Differences in the degree of non-randomness of flea distribution on various host species are discussed. Frequency indexes for the presence of flea species on different host species are calculated as a measure of host specificity. The specificity of most of the prominent fleas was low. Microtus californicus is considered a key species in promoting flea exchange.
Two positive pools were obtained by us, one in Rivera Canyon about ¾ mile from the study area, and one following control operations on the ridge above the study area. It is suggested that the epizootic must have been essentially limited to this vicinity. After the initial survey, poison grain was distributed by air over about 500 acres. The results were evaluated by the use of repeated traplines. Apparent reductions of from 25 to 80 per cent were registered in four different areas. Microtus was most severely affected. Trapping about two months later showed that populations had generally regained their previous levels. Reithrodontomys displayed a striking and unexplained increase. The density of fleas on hosts increased sharply immediately after poisoning. The poisoning operation was apparently ineffective in reducing the plague transmission potential among wild rodents. Rats were successfully controlled on the hog farms.
The role of habitat distribution is emphasized in governing interspecific host and flea contacts, continuity of distribution of host populations and potential for the contact of native mammals with humans and domestic rats.