Fig. 7.

Results of a detail-rich simulation of the spread of a costly DEG through a population of the mosquito Anopheles gambiae (sensu stricto) based on population parameters and weather at a site in Tanzania (from [61]). The colours of the stars are the results of simulations for the particular cost and homing frequency combinations: green, DEG fails to establish; yellow, DEG goes to fixation and population persists; black, equilibrium with both DEG and wild-type allele; red, population elimination. Also shown is the degree to which simple models explain these results. Genetic models predict the DEG always invades and is fixed for parameter values below the thick black line [20]. Because of stochastic effects, when homing is very weak the DEG is often lost by chance and fails to establish (green stars at left). Also, when homing is very weak, it may take a very long time for fixation to occur, which explains the black stars to lower left. Population models predict elimination in a top right sector whose size depends on the growth rate of the population when rare [30]. This was not estimated explicitly from the data but a value of around 4 (geometric mean over time) best fits the simulation. When the population is small and near elimination the DEG may get fixed through random effects, which explains the yellow stars above the line near the region of elimination in the top right. Overall there is a pleasing concordance between the results of the simple and complex modelling exercises