BackgroundRabies is a serious yet neglected public health threat in resource-limited communities in Africa, where the virus is maintained in populations of owned, free-roaming domestic dogs. Rabies elimination can be achieved through the mass vaccination of dogs, but maintaining the critical threshold of vaccination coverage for herd immunity in these populations is hampered by their rapid turnover. Knowledge of the population dynamics of free-roaming dog populations can inform effective planning and implementation of mass dog vaccination campaigns to control rabies.Methodology/Principal FindingsWe implemented a health and demographic surveillance system in dogs that monitored the entire owned dog population within a defined geographic area in a community in Mpumalanga Province, South Africa. We quantified demographic rates over a 24-month period, from 1st January 2012 through 1st January 2014, and assessed their implications for rabies control by simulating the decline in vaccination coverage over time. During this period, the population declined by 10%. Annual population growth rates were +18.6% in 2012 and -24.5% in 2013. Crude annual birth rates (per 1,000 dog-years of observation) were 451 in 2012 and 313 in 2013. Crude annual death rates were 406 in 2012 and 568 in 2013. Females suffered a significantly higher mortality rate in 2013 than males (mortality rate ratio [MRR] = 1.54, 95% CI = 1.28–1.85). In the age class 0–3 months, the mortality rate of dogs vaccinated against rabies was significantly lower than that of unvaccinated dogs (2012: MRR = 0.11, 95% CI = 0.05–0.21; 2013: MRR = 0.31, 95% CI = 0.11–0.69). The results of the simulation showed that achieving a 70% vaccination coverage during annual campaigns would maintain coverage above the critical threshold for at least 12 months.Conclusions and SignificanceOur findings provide an evidence base for the World Health Organization’s empirically-derived target of 70% vaccination coverage during annual campaigns. Achieving this will be effective even in highly dynamic populations with extremely high growth rates and rapid turnover. This increases confidence in the feasibility of dog rabies elimination in Africa through mass vaccination.

Dogs (Canis familiaris) are often free-roaming in sub-Saharan African countries. Rabies virus circulates in many of these populations and presents a public health issue. Mass vaccination of dog populations is the recommended method to decrease the number of dog and human rabies cases. We describe and compare four populations of dogs and their vaccination coverage in four different villages (Hluvukani, Athol, Utah and Dixie) in Bushbuckridge Municipality, Mpumalanga province, South Africa. Cross-sectional surveys were conducted in the villages of Athol, Utah and Dixie, while data from a Health and Demographic Surveillance System were used to describe the dog population in Hluvukani village. All households of the villages were visited to obtain information on the number, sex, age and rabies vaccination status of dogs. From May to October 2013, 2969 households were visited in the four villages and 942 owned dogs were reported. The populations were all young and skewed towards males. No differences were observed in the sex and age distributions (puppies 0–3 months excluded) among the villages. Athol had a higher proportion of dog-owning households than Hluvukani and Utah. Vaccination coverages were all above the 20% – 40% threshold required for herd immunity to rabies (38% in Hluvukani, 51% in Athol, 65% in Dixie and 74% in Utah). For the preparation of vaccination campaigns, we recommend the use of the relatively stable dog:human ratio (between 1:12 and 1:16) to estimate the number of dogs per village in Bushbuckridge Municipality.