In both analyses, the five strains

of S. dehoogii were grouped into four different clusters, underlining the pronounced degree of intraspecific variability found in this species. Pseudallescheria apiosperma was dispersed over three clusters. The ten investigated P. boydii strains were recovered in five different clusters using SJ and in four clusters using SSM. Reproducibility testing showed that the methods used were acceptable for analysing the physiological diversity in the Pseudallescheria/Scedosporium complex. Of a total Selleckchem CP 673451 of 570 reactions available in the panel, 254 reactions were polymorphic (44.6%) (Table 2), while 271 reactions (47.5%) were invariant, and a total of 45 reactions (7.9%) were found to be unreliable and were Dinaciclib manufacturer removed from the data set. Reasons for removal were (i) instability or inconsistence (26 reactions; 4.6%) or (ii) turbidity of the medium

or too early colour change of the indicator, i.e. occurring immediately after inoculation (19 reactions; 3.3%). The variability of the test results may be caused by decomposition of the test compound (e.g. with thermolabile components, insufficient solubility or by deviations from optimal pH values). The same problems with these compounds had been encountered in the framework of characterisation of fermentative actinomycetes.23 Several compounds were tested at pH 4.0, pH 7.5 and pH 8.2, in most cases resulting in removal of the obtained results. In contrast, nearly half of the glucosidases and phosphatases reactions proved to show consistent responses

at pH 5.5. Acidification of the medium by Pseudallescheria and Scedosporium strains should be taken into account. Our test results essentially corresponded with those published by de Hoog et al. [1,5], for example, in positive reactions for d-galactose and negative for melezitose in the P. boydii complex, and negative results for creatinine, succinate and positive responses to l-arabinose, l-rhamnose, trehalose, cellobiose and salicin in S. prolificans. The taxonomic separation between purported Pseudallescheria and Petriella/Petriellopsis clades [24] is thus supported by physiological parameters. Within the Pseudallescheria clade, physiological data as assessed by the Taxa Profile system did not fully match with the subdivision of the group into at least five species, as proposed by Miconazole Gilgado et al. [10,12]. Discrepancies were noted with maltose assimilation by P. minutispora and for l-arabinitol assimilation by two out of four S. prolificans strains. Particularly, our d-ribose results differed significantly, underpinning previous observations that testing pentose fermentation by assessment of acid production is highly liable to test errors.23 The results of Gilgado et al. [12] were obtained using macrodilution according to Yarrow [25]; it seems likely that results obtained with different techniques cannot be generalised.