As a result of analyzing the internal transcribed spacer (ITS) and 5 0 end of the EF-1a sequence of 145 isolates of Metarhizium spp. isolated from soil in Japan using selective agar medium, eight species were identified. ITS sequence analysis divided the isolates into three clades. Two were identified as M. flavoviride var. pemphigi and M. lepidiotae, respectively. EF-1a sequence analysis identified the other clades as six species: M. anisopliae, M. brunneum, M. guizhouense, M. majus, M. pingshaense and M. robertisii. The distribution of M. flavoviride var. pemphigi was restricted to forest or wood soil, and conidial sizes of M. guizhouense and M. majus were incongruent with the phylogeny based on the sequence of the 5 0 end of EF-1a.

Twenty bacteria were isolated from four ant lion larvae. The isolates were classified into three groups by biological characteristics. Since Group I, Group II and Group III were isolated from individual larvae Kuo1, Kuo3, 4 and Kuo2, respectively, with exception of one isolate Kuo2‐1, each ant lion tested had its own dominant bacterial flora. Groups I and II were closer to Serratia liquefaciens and Enterobacter cloacae, respectively, whereas Group III could not be identified by the test used. The phylogenetic analysis of GroEL amino acid sequences revealed that Group I, II and III were related to those of Serratia spp., E. cloacae and Salmonella spp. –Escherichia/Shigella spp., respectively. Among these groups, Group I was highly virulent against Bombyx mori and Periplaneta americana, and caused 100% mortality within 24 h. The other two groups (Group II and III) were avirulent to these insect species. The culture filtrate of Group I caused killing activity to B. mori larvae and the insecticidal substance was purified from culture filtrate of Group I bacterium. Since the insecticidal activity highly correlated with proteolytic activity in the chromatographies, Group I bacterium may secret insecticidal proteinase in vitro.

On August 2017, I observed a Northern Goshawk (Accipiter gentilis) wearing an anklet at Mizumoto Park, Tokyo. Thereafter, I observed such a goshawk eleven times until early April 2018. In all records, the observed individual was a juvenile. Characteristics of plumage and the presence of an anklet suggested that all eleven observation records were derived from one or a few individuals, which had escaped from captivity. If the last observed individual was the same as the first one, it had survived for at least 110 days. Currently, by law most goshawks held in captivity cannot be a subspecies breeding in Japan. If exotic subspecies escaping from captivity bred with Japanese subspecies, native subspecies will be at risk of genetic disturbance due to hybridization.