NATURAL HABITAT 151 



various soil types collected on the desert near San Carlos, Arizona, 

 distant from human habitations but near numerous rodent burrows. 

 Hundreds of other attempts to isolate Coccidioides from the upper 

 layers of soil within endemic areas have failed. 



Recent studies in southern Arizona have suggested the possibility 

 of a rodent reservoir of the disease.- '• ^ Among small rodents 

 trapped in this area the white-footed mouse, grasshopper mouse, and 

 wood rat were never or very rarely infected; but 15 per cent of three 

 species of pocket mice and 17 per cent of one species of kangaroo rat 

 trapped had pulmonary coccidioidomycosis. Surprisingly, a second 

 fungus, Haplosporangium parvum, which resembles Coccidioides in 

 its parasitic phase but is different in culture, was found with even 

 greater frequency (66 per cent of pocket mice) causing a similar 

 pulmonary disease in rodents. No human infections with the latter 

 fungus have yet been recognized, but many individuals who react to 

 the intradermal injection of coccidioidin also react to an antigen 

 (haplosporangin) prepared from this fungus. 



Coccidioidomycosis in these rodents seemed to be a slowly progres- 

 sive chronic disease which did not exterminate the species and, in 

 fact, appeared to interfere httle with normal development and repro- 

 duction. Further investigations will be required to determine whether 

 these rodents are infected because of their intimate exposure to in- 

 fested soil or whether the fungus is primarily a pathogen of rodents 

 and is present in soil that has been contaminated by infected rodents. 

 Certain circumstances seem to suggest the latter as probable. The 

 disease in rodents is known in species of Perognathus and Dipodomys 

 which are desert species with ranges coinciding generally with a part 

 of the known geographical range of coccidioidomycosis. Species of 

 Peromyscus, however, although more common than Perognathus, and 

 living in adjacent burrows, were rarely found infected under field 

 conditions. Their exposure to the fungus would seem to be equal 

 to that of Perognathus, and they are very susceptible to experimental 

 infections, but they apparently do not serve as hosts under the field 

 conditions investigated. There are no doubt other rodent hosts of 

 the fungus. There appears to be the sort of adjustment between 

 pathogen and the Perognathus host which is compatible with the 

 theory of a rodent reservoir. The difficulty of isolating Coccidioides 

 from desert soil in which it seems to have a spotty distribution not 

 correlated with any recognized differences in vegetation or soil types 

 makes the hypothesis of a rodent reservoir attractive. 



Coccidioidomycosis occurs also in other animals within the en- 

 demic area. Stiles and Davis ^^ have described the focalized infec- 



