For 6 consecutive years (2013 to 2019), we performed an annual rodent elimination by placing rodenticide in 3 villages and measuring rodent abundances. Simultaneously, we monitored the same outcome in 3 different villages without rodent elimination. This dual monitoring allowed us to evaluate the long-term impact of such an approach. We found the effect to be strong but short-lived: rodents re-invaded houses within 2-3 months after elimination, leading to a similar abundance one year after elimination (figure 2).

Figure 2: Dynamics of M. natalensis measured by the trapping success inside houses in 3 treated villages (Brissa, Dalafilani and Yarawalia), and in 3 control villages (Damania, Sokourala and Sonkonia) over 6 years. AR indicates the period of elimination by anticoagulant rodenticide.

There are several reasons for the return of rodents after the end of the elimination:

1. the fertility of this species, whose mean litter size is 9.2;
2. the survival of a few animals in the fields surrounding the houses, which allows recolonization of the human habitat;
3. several closed houses which we were unable to enter and deposit rodenticide baits, and which may have served as shelters from where the rodents could recolonize surrounding houses;
4. very attractive foods storage available for rodents in the houses;
5. the porosity of walls and roofs, allowing rodents to enter houses very easily;
6. the low prevalence of domestic predators, such as cats and dogs.

We concluded that once-yearly rodent elimination was not sufficient to maintain a low abundance of the M. natalensis reservoir. We therefore developed a modelling approach to investigate the efficiency of a rodent control.