Abstract: The rate of infection of Nosema locustae decreased with the decrease of densities of the complex population of rangeland grasshoppers in the treated season (as Plot 1).However, the rate of infection of N. locustae in Myrmeleotettix palpalis, one of the dominant species of grasshoppers occurred lately in the season, increased with its increase in number in the complex population. The spatial distribution of infected grasshoppers by N. locus-tae in the treated season, varied with sampling time. On the 7th days after treatment with N. locustae bait (7.5 x 10⁹ spores/1.5 kg wheat bran/ha), the spatial distributions fixed to the Poisson distribution. But on the 28th and 40th days fixed to aggregate distributions. The densities of complex grasshoppers population were correlated with the percentages of infection of N. locustae in August of 1993 and August of 1994 in the plots, where were treated with N. locustae in 1988 and 1989 respectively (as Plot 2 and Plot 3). Further analysis of the data from the plots showed that both densities and infections in 1993 were significantly higher than that in 1994. And N. locustae still distributed widely in most of dominant species of grasshoppers, Oedaleus asiaticus, M. palpalis, Bryodema luctuosum luctuosum, Angaracris spp. And Dasyhippus barbipes. The spatial distribution of infected grasshoppers by N. locustae both in 1993 and 1994 were suitable to Poisson distribution in Plot I and Plot 2. The dispersal distance and directions of N. locustae were investigated in1993 and 1994 in the two plots too. The results indicated that the dispersal distance of N.Locustae was at least 1000 meters, and the dispersal direction might be affected by the geographic features and the prevalent wind etc.

Key words: grassland grasshoppers, Nosema locustae, spatial distribution, dispersal pat- tern