Variable social-ecological indicators across a Tanzanian protected area network

Protected areas (PAs) represent the most dominant conservation model in the world, though existing research on how and why their conservation outcomes vary remain scarce. In contribution to this discourse, we evaluated the densities and temporal trends of seven herbivorous mammal species across a protected area (PA) network in the Tarangire Ecosystem, Tanzania, and assessed human-wildlife interactions for 19 species. We conducted seasonal wildlife surveys inside five PAs and structured questionnaires in adjacent villages. Our study focused on two national parks, two community-based conservation areas, and one game-controlled area with limited management capacity. We extracted several ecological (wildlife densities and their trends over time) and human-wildlife interaction (number of herbivore and carnivore species associated with benefits, costs and illegal use) indicators and compared them across the PAs using generalized linear models. We found that both national parks and community-based conservation areas can support high densities of mammal species with stable or positive annual trends. In some instances, these wildlife population dynamics were associated with negative human-wildlife interactions near the boundaries of the PA, but this did not emerge as a consistent pattern. National parks can also be associated with negative annual population trends and negative human-wildlife interactions near their boundaries. The game-controlled area scored the lowest on both ecological and human-wildlife interaction indicators highlighting the importance of management-in-practice. Our findings suggest that the outcomes of PAs vary considerably, even within the same ecosystem, and that local social-ecological context likely affects their sustainability.