Abstract: According to the population of Meriones unguiculatus, flea index and data of seven meteorological factors in Etuoke Banner and Etuokeqian Banner, Inner Mongolia Autonomous Region in 1975-1989, models of linear and curve regression between the flea index and the population of M. unguiculatus were obtained respec tively. The optimum regression subsets of multipic linear regression were conductedbetween the flea index and meteorological factors. And then, the model of standard regression was also obtained. The canonical correlation analysis explained such rela tions of the flea index and their hosts population to the meteorological factors. The conclusions are as follows: The population of M. unguiculatus was the main factoraffecting flea index. The meteorological factors exerted summing-up influence on the flea population, and yearly average relative humidity and temperature in the field were the two main factors with influence. Moreover, the flea index was influenced by the meteorological data greater than the population density of their hosts.

Key words: flea index, Meriones unguiculatus, meteorological factor, mathematical model