This research was granted by FAPESP (process N° 2006/61171-3) and, CAPES-PROSUP.

Malassezia pachydermatis is an etiological agent of superficial mycosis in dogs and cats, causing otitis and dermatitis. Since M. pachydermatis is the only non-lipodependent species in the genus, veterinary laboratories usually employ mycological culture medium without lipids to isolate this microorganism. Nevertheless, at the research level, lipodependent Malassezia species have been isolated from both healthy and diseased animals. This study's objective was to determine the presence of lipo and-non-lipodependent Malassezia species in the external ear canal of dogs with and without otitis, characterizing strains using phenotypic and genotypic techniques. Fifty dogs were studied, 25 with and 25 without otitis. A total of 100 samples were studied--48 of otic secretion and 52 of cerumen originated from the ear canal with and without infection, respectively. Samples were cultured on Mycosel agar supplemented with glycerol and olive oil. The isolates were characterized phenotypically by means of macro- and micromorphology, catalase tests and growth on Tween 20, 40, 60 and 80, and Cremophor-EL. Karyotyping was performed via Pulsed Field Gel Electrophoresis (PFGE) and Restriction Fragment Length Polymorphism (RFLP). Malassezia spp. were isolated from 19/25 (76%) and 18/25 (72%) dogs without and with otitis, respectively. Malassezia spp. were isolated respectively from 30/52 (58%) and 35/48 (73%) of the cerumen and otic secretion samples. M. pachydermatis was the most prevalent species, having been isolated from 100% and 94% of the strains' originated from cerumen and otic secretion. Four lipodependent strains (4/33-12%) were isolated from 3/19 (16%) animals with otitis, and characterized phenotypically as M. furfur. Through both genetic molecular methods employed in this research, all strains were identified as M. pachydermatis. Therefore, the 4 strains identified as M. furfur by phenotypical methods really were lipodependent M. pachydermatis, a fact already verified by other researchers. Concordance between phenotypical and genotypical techniques was 94%. It concludes that phenotypic methods permit a satisfactory identification scheme for Malassezia spp., but if commercial veterinary laboratories do not use media with the addition of lipids for isolation of these yeasts, it might yield false-negative results in veterinary clinical samples, and probably therapeutic unsuccessful.