Mathematical modeling, spatial complexity, and critical decisions in tsetse control

The tsetse fly complex (Glossina spp.) is widely recognized as a key contributor to the African continent's continuing struggle to emerge from deep economic, social, and political problems. Vector control, the backbone of intensive efforts to remove the human and livestock trypanosomosis problem, has been typified by spectacular successes and failures. There is widespread agreement that integrated vector control, combined with direct disease treatment and prevention, has to play a major role in alleviating the tsetse burden in Africa. Mathematical and computer-based simulation models have been extensively used to try to understand how best to manage these control efforts. Such models in ecology have been helpful in giving broad generalizations about population dynamics and control. Unfortunately, in many ways they have inadequately addressed key aspects of the fly's biology and ecology, particularly the spatio-temporal variability of its habitats. These too must factor in any control efforts. Mathematical models have inherent limitations that must be considered in their use for control programs. In this review, we consider some of the controversies being debated within the field of ecology and evolution about the use of mathematical models and critically review several models that have been influential in structuring tsetse control efforts. We also make recommendations on the appropriate role that mathematical and simulation models should play when used for these purposes. Management programs are often vulnerable to naively using these models inappropriately. The questions raised in this review will apply broadly to many conservation and area-wide pest control programs with an ecological component relying on mathematical and computer simulation models to inform their decisions.