Characterization of Aphid Pathogenic Species in the Genus Pandora by PCR Techniques and Digital Image Analysis

Abstract

Twenty-eight isolates of the insect pathogenic fungus genus Pandora Humber derived from various geographical locations and insect hosts were studied in order to examine both genotypic as well as phenotypic variation. We included sixteen Pandora neoaphidis, two Pandora nouryi, three Pandora kondoiensis, one Pandora bullata and six Pandora delphacis isolates in the study. The sizes of ITS bands and RAPD-PCR patterns were used to investigate the genotypic variations. For phenotypic characterization the length and width of the primary conidia were measured and the size and shape of the conidia were quantified in digital images. The size of ITS divided Pandora isolates from Aphididae into the following four groups: (i) all P. neoaphidis isolates, (ii) two isolates of P. kondoiesis including the ex-neotype, (iii) one isolate of P. nouryi and (iv) one isolate of P. kondoiesis and one isolate of P. nouryi. The analysis of RAPD data fully supported the groupings obtained for the ITS lengths. The RAPD technique was able to distinguish between all isolates except those isolated from the same epizootic. It was shown that, among the P. neoaphidis isolates, a correlation between RAPD-PCR profile and geographical origin of the isolate was present while no correlation between host species and profile was seen. The analysis of digital images was here employed for the first time for entomopathogenic fungi. The digital image analysis supported at a quantitative level the original descriptions of P. neoaphidis and P. delphacis primary conidia as ovoid to ellipsoidal whereas the primary conidia of P. nouryi and P. kondoiensis are ovoid. Within these two clusters information on host insect species is important for distinguishing fungal species. Thus, by combining digital image analysis and information on host insect species the results agreed with the genetic analysis.