Spatial-temporal population dynamics of Cotesia flavipes (Hymenoptera : Braconidae) in Kenya

The spatial-temporal population dynamics of an imported parasitoid of stemb orers, Cotesia flavipes Cameron, were analyzed. A two-step spatial interpol ation method, spatial splines,,vith Kriging of the residuals, was developed for interpolating the distribution of C. flavipes in Kenya. The results in dicate that C. flavipes has become established throughout the whole souther n part of Kenya and northern part of Tanzania, and that the population dens ity of C. flavipes in maize fields is still increasing. The maximum C. flav ipes density was about one parasitized borer per plant in the first growing season of 1999. The mean C. flavipes densities varied from region to regio n in 1999, with the highest density (five parasitized borers per 20 plants) in the southeast and the lowest parasitism in central Kenya (one parasitiz ed borer per 20 plants). The percentage of sites occupied by C. flavipes wa s 90% in the southeastern area during the first growing season of 1999. In coastal Kenya, C. flavipes occupied > 70% of the sampling sites, compared w ith approximate to 50% in both central and western Kenya. Over the whole sa mpling area, C. flavipes increased its distribution from 60% of sites occup ied in 1994-80% in 1999. Temporally, C. flavipes population density remaine d low, with an average of less than one parasitized borer per 20 plants, un til the second growing season of 1997-1998, when there was as harp, increas e. The suppression of stemborer populations by C. flavipes appeared in sout heastern and coastal Kenya from 1998. Analyses indicated that C. flavipes w as a minor parasitoid before 1995, but had become the predominant parasitoi d after 1998. The results of the spatial interpolation showed that the C. f lavipes population did not spread much from the release sites before the fi rst growing season of 1996, but expanded tremendously after the first growi ng season of 1997. The spatial interpolation model was validated with field data from 1999. The model predicted the C. flavipes density well at the zo nal level but underestimated C. flavipes density country-wide. The possible displacement of an ecologically similar native congener, C. sesamiae, is d iscussed.