Thuringiensin produced by Bacillus thuringiensis subsp. darmstadiensis was further subjected to a two-phase aqueous separation system. A modiÐed HPLC method and a test for quantitative pathogenicity using the house Ñy Musca domestica were used for analysis of thuringiensin. Within a realistic range of dosages, more e †ect was observed in the pupal stage than in the larval stage.
The percentage e †ective control rate (ECR) was calculated by (100 [ percentage emergence) ; malformed and non-reproductive adults were considered as emerged. Pupal mortality, pupal weight, and ECR after feeding the three-day-old larvae were the measured response criteria for bioassay. The of thuringien-EC 50 sin for pupae mortality was 1É64 kg ml~1 diet, and 0É83 kg ml~1 for mortality of adults. Insecticidal activity of the broth increased with fermentation time-course from 9th to 21st hour. The bioassay curve constructed with three-hour sampling interval during the fermentation course had good correlation to thuringiensin content as determined by the HPLC method. In the two-phase aqueous separation system, a maximum of 96É7% ECR was achieved with the bottom salt layer, compared to a value of 46É7% with the upper PEG layer. These results suggest that thuringiensin, prepared through a fermentation and recovery process, is suitable for pest control.