A systematic map of demographic data from elephant populations throughout Africa: implications for poaching and population analyses

Abstract The crisis facing Africa’s elephant populations is a notorious example of ongoing wildlife declines caused by illegal harvesting. Targeted conservation interventions require detailed knowledge about changes in population sizes and the effect of illegal activities. However, accurately quantifying poaching intensity is a difficult task: commonly calculated from ranger‐based carcass‐encounter data, the proportion of illegally killed elephants (PIKE) is a function of poaching and background mortality. Hence, at constant poaching intensity, PIKE decreases with increasing natural mortality and also with hunting, management interventions, and other anthropogenically induced deaths. Natural mortality is often more difficult to quantify with accuracy than mortality due to illegal killing, as elephants that die naturally are more likely to be missed than those taken by poachers. In recent analyses, constant background mortality rates were assumed. Yet, for example climate‐driven fluctuations in natural mortality, if not quantified and accounted for, may lead to biased estimates of poaching intensity. Varying background mortality rates can be accounted for in the analysis of PIKE, but this requires near‐complete counts of natural and management‐related deaths and hunting records. Carefully developed population models, which simulate population dynamics and demographic changes while accounting for variation in environmental conditions and management strategies, are alternatives. However, successful calibration of such models requires integrating comprehensive demographic data. We systematically review the scientific and ‘grey’ literature on African elephant demography with the objective of facilitating poaching and population analysis possibilities through an inventory of information relevant to demography. Our screening of 10900 publications resulted in the review of relevant information provided by 431 studies from 420 study sites throughout Africa. From these, we extracted demographic data collected between 1900 and 2017, and collated them in the newly created African Elephant Demographic Database (AEDD; 10.6084/m9.figshare.19387085). We found 37 natural mortality estimates from five different study sites. Other mortality data, demographic rates, and age‐ and sex‐structured population data were substantially more abundant, both temporally and spatially. This new collection of demographic rates, age‐ and sex‐structured population data, and cause‐partitioned mortality estimates identifies spatial and temporal data gaps and provides prior information needed for African elephant population models. Closing these data gaps and subsequent analyses of realistic population models may aid elephant conservation via improved policies, legislation, and protection.