In this study, three isolates of Streptomyces isolated from Egyptian soil at Giza Governorate and belonging to red series were identified based on their cultural, morphological and physiological characters. According to the key proposed by Pridham and Tresner (1974), the experimental isolates ST55, ST71 and ST86 ap-peared to be related to S. lincolnensis, S. venezuelae and S. umbrinus, respectively. In addition, random amplified polymorphic DNA (RAPD)-polymerase chain reac-tion (PCR) was used to amplify the DNA genome of the three applied isolates and an identified strain to determine their fingerprints. This was carried out using 10-decamer oligonucleotides, i.e., OP-A02, OP-D01, OP-D02, OP-D05, OP-D06, OP-D07, OP-D08, OP-D11, OP-D18 and OP-D20. The PCR amplified products were detected by electrophoresing on 1.5% agarose gel and visualized by staining in eth-idium bromide and UV trasilluminator. Results showed that the same oligonucleo-tide flanking fragments differed from one isolate to another with different sizes. Monomorphic and polymorphic fragments were amplified confirming the molecular relationships as well as the genetic diversity between the four applied strains of Streptomyces. Based on the statistical analysis of RAPD-PCR polymorphisms a de-gree of similarities ranged from 33.9 to 59.2 % was found. Furthermore, frag-ment(s), whatever absent or present, were found to be unique molecular markers and could be used for characterization of some isolates. This result supported the idea for using the RAPD-PCR technique for classification of Streptomyces isolates.

A numerical method was suggested for the identification of eight local halotolerant streptomycete isolates. Cultural, morphological, physiological and antagonistic characteristics of those isolates were determined. Arbitrary scoring of these characters for the eight unknown isolates and those of the more related known Streptomyces species in the keyof Pridham and Tresner (1974) was given and resulted 58 characteristic units. The degree of similarity was determined using the Dice coefficient method and clustering was achieved using the unweighted pair group method average (UPGMA) algorithm. Using this numerical method, six out of the eight Streptomyces isolates, i.e., ST08, ST09, ST11, ST13, ST14 and ST15 were identified as S. longisporus, S. janthinus, S. griseochromogenes, S. antibioticus S. baarnensis and S. albolongus with 100, 95.8, 95.0, 92.8, 95.0 and 91.1%, respectively. Isolates ST10 and ST12 were duplicate of S. echinatus with similarities of 93.3 and 94.6%, respectively. Application of the suggested numerical taxonomy on the 14 known Streptomyces species revealed that these species fell into three major clusters based on their color of aerial mycelia.