Age dependent sampling biases in tsetse flies (Glossina): Problems associated with estimating mortality from sample age distributions

For a closed (island) population of G. morsitans morsitans Westwood, the probability per week of capturing females on ox fly rounds was about 0.3 in the first week of life, less than 0.2 for 27 to 35-d-old flies and greater than 0.4 for flies more than 80 d old. For open populations, the relative changes in capture probability were measured from the ovarian age distributions of trap and ox fly round samples. They were used (with island data) to show that the age dependent sampling bias of traps for female G. m. morsitans increased more than sixfold over the first 80 d of life. The age dependent bias for G. pallidipes Austen taken from odour baited traps is probably at least as serious as for G. m. morsitans. Estimates of daily mortality from the mark-recapture studies were always (up to 20 times) higher than estimates from ovarian age samples taken at the same times. The mortalities recalculated from samples adjusted for sampling biases were closer to, but still lower than, the mark-recapture estimates. Odour baited targets are successful in controlling tsetse populations, despite the relatively low probability of treating young females. If sterilants instead of insecticides were used on the targets, young females could be treated indirectly via treated males, which transfer the sterilant to virgin females during copulation. (author). 15 refs, 2 figs