Antimicrobial resistance in the Serengeti region of Tanzania: role of food and pharmaceutical value chains


Jonathan Rushton, Ruth Rushton




Advances in antibiotic treatment are continually challenged by the evolution and dissemination of antibiotic resistance and this leaves medical practitioners with dwindling options for cost-effective therapies.

Project details

Antibiotic use per se directly favors resistant organisms, but many resistance traits persist without antibiotic selection pressure. For bacteria, resistance genes are typically carried on horizontally transmissible elements and persistence in the absence of antibiotic selection probably arises through genetic linkage between resistance traits and genes that confer other selective advantages.

Hence, resistance genes comprise a ‘horizontal gene pool’ that can colonize multiple bacterial populations and species to maintain and disseminate resistance traits in the absence of antibiotic selection pressure. The transmission of antibiotic resistance therefore transcends the ecological and epidemiological boundaries of any single pathogen or host bacterium.

The mechanisms of antibiotic resistance are generally understood, but the current understanding of how antibiotic resistance persists at population scales is weak. This lack of knowledge hinders development of effective intervention strategies that would benefit greatly from a unified understanding of the overall ecology of antibiotic resistance seen in the context of the broader ecological community.  

Our long-term goal is to identify the ecological and socio-economic drivers that contribute to  maintenance and dissemination of antibiotic resistance.

The objectives of this project are:

(i) to develop  a community scale model of antibiotic resistance epidemiology that integrates phenotypic and molecular genetic data with ecological modeling,

and (ii) to correlate ecological patterns of antibiotic  resistance with socio-economic data. This strategy of combining both ecological and socio-economic drivers will be applied to study antibiotic resistance traits amongst three host populations (human, livestock, and wildlife) and across three distinct ecological zones in Tanzania. We selected the greater Serengeti ecosystem for our study in part because (i) the close proximity and contact between potential reservoir populations provides a tractable system for developing models to test hypotheses that are relevant to both industrialized and resource resource constrained countries; (ii) the local unregulated access to antibiotics provides a robust opportunity to test our central hypothesis (see below) in the presence of drug selection pressure;

(iii) socio-economic conditions vary across space and time with ongoing changes occurring in the region regarding adoption of new livestock production systems, greater reliance on tourism and growing human populations alongside antibiotic use patterns and human-animal interactions in rural communities;

and (iv) the spread of antibiotic resistance in Tanzania is directly relevant to local communities. Because Tanzania is undergoing rapid urbanization, our findings will have implications for other countries experiencing similar socio-economic changes. We will determine the relative contribution of transmission pathways and ecological reservoirs to the persistence of antibiotic resistance in bacteria from humans and animal populations and integrate the contribution of community knowledge, attitudes and practices to model the socio-economic contribution to antibiotic resistance.

Our central hypothesis is that the prevalence and diversity of antibiotic resistance in human and livestock populations is primarily explained by the dissemination and persistence of resistance within and between various host populations, rather than by direct use of antibiotics, leading to similar antibiotic resistance patterns and gene pools in human and livestock populations.

This hypothesis predicts that antibiotic resistance is correlated with patterns of connectivity within and between communities, arising from shared resources (e.g., water), population density (interactions between human and livestock), and animal movements via trade between communities. Our analysis will also consider the alternative hypothesis that the prevalence and diversity of antibiotic resistance in human and livestock populations is primarily explained by antibiotic use patterns with no significant correlation with ecological variables.