Safety of Bacillus thuringiensis for earthworms

Abstract

The effects of two insecticides, diflubenzuron (Dimilin® 25 WP) and Bacillus thuringiensis var. kurstaki (B.t.k.; Dipel ES®) on the soil fauna of a deciduous forest in the Upper Rhine Valley were investigated in a field experiment. The four sampling plots were located in an oak–hornbeam forest in South-West Germany. One plot remained untreated (control), two plots (DIM-1 and B.t.k.) each received 50% of the amount used in aerial applications of the insecticides, and one plot (DIM-10) a 10 times the recommended concentration of diflubenzuron. One year before application the composition and abundance of selected soil fauna groups (earthworms, enchytraeids, oribatids, gamasids, and collembolans) were surveyed to assess seasonal and spatial variations. The annelids were not affected by either insecticide, all arthropod groups investigated reacted to diflubenzuron in a concentration of 25 g ha–1. Most clearly the dominant collembolan species Folsomia quadrioculata (Tullberg, 1871) was affected: even 1.5 years after application the population had not recovered. The abundance and dominance of some mite species decreased, the dominance of others increased, and some species migrated from the litter to deeper soil layers. Populations recovered slowly in relation to the rapid degradation of diflubenzuron. The only effect of B.t.k. was a reduction in the dominance rank of Veigaia nemorensis (CL. Koch, 1892) (Gamasida). In summary, the soil dwelling arthropods were affected by diflubenzuron, not by B.t.k.