CHANGES OVER 12 YEARS IN POPULATIONS OF GLOSSINA-PALLIDIPES AND GLOSSINA-LONGIPENNIS (DIPTERA, GLOSSINIDAE) SUBJECT TO VARYING TRAPPING PRESSURE AT NGURUMAN, SOUTH-WEST KENYA

Long term changes in the size of populations of the tsetse Glossina pa llidipes Austen and G. longipennis Corti were monitored over a 12 year period at Nguruman in south-western Kenya. Tsetse populations were su bject to droughts of varying intensity and, from 1987, to trapping, in itially by a research organization, and later by a community-based dev elopment project. Populations were mainly sampled using odour-baited b iconical traps, with data from other monitoring traps corrected accord ingly. Mark-release-recapture studies were carried out to relate trap catches to absolute population size, and to quantify movement between subpopulations. Trypanosomiasis incidence rates in a herd of local cat tle were also monitored for much of this period. Trap catches were sho wn to be well correlated with estimates of absolute population size, w ith no evidence of any seasonal change in trap efficiency. The intensi ty of trapping and level of seasonal immigration appeared to be the ma in determinants of population trends, with effective control being ach ieved when traps were well maintained. Movement between the two lowlan d subpopulations was shown to be greater for females, and to be invers ely related to temperature. An analytical model was used to investigat e the responses of a partially isolated population to trapping pressur e. Predictions of a deterministic simulation model demonstrated that t he observed changes are consistent with an adult trapping mortality of 4-8% per day, and immigration of 100,000 G. pallidipes females per mo nth in the long rains (April and May), 5000 per month in the short rai ns (November), and about 500 per month during the dry seasons. Trypano somiasis incidence in local cattle was greatly reduced during the peri od of community-based tsetse control. When tsetse were sampled exactly where the cattle were grazing, disease incidence was shown to be line arly related to G. pallidipes catches. Arguments for tr ap resistance and residual populations were examined, and found to be inconsistent w ith the data. The future for tsetse control by the Nguruman community is considered.