Skip to main content
Dryad logo

Density estimates of African lions in Queen Elizabeth National Park


Braczkowski, Aleksander (2020), Density estimates of African lions in Queen Elizabeth National Park , Dryad, Dataset,


African lions are declining across much of their range, yet robust measures of population densities remain rare. The Queen Elizabeth Conservation Area (QECA; 2400 km2) in East Africa’s Albertine Rift has potential to support a significant lion population. However, QECA lions are threatened, and information on the status of lions in the region is lacking. Here, we use a spatially explicit search encounter approach to estimate key population parameters of lions in the QECA. We then compare home range sizes estimated from our models to those from a radio-collaring study implemented a decade earlier. We recorded 8243.5 km of search effort over 93 days, detecting 30 individual lions (16 female and 14 male) on 165 occasions at a rate of 2 lion detections/100 km2. Lion density in the QECA was 2.70 adult lions/100 km2(SD=0.47), while mean abundance was 71 individuals (SD=11.05). Worryingly, the movement parameter for male lions was 3.27 km and 2.22 km for females, suggesting >400%, and >100% increases in home range size, respectively, compared to a decade earlier. Sex ratio of lions in the QECA was lower (1 male: 0.75 females), when compared to a previously published review (mean=1:2.33). The large movements and skewed sex ratios we report on in this paper are likely a result of human driven prey depletion. Our results suggest lions in the QECA are in a precarious state and the lion densities are significantly lower than what they could be. As lions are under pressure throughout much of Africa, our study presents the utility of a census technique that could be used elsewhere as an early warning of lion declines. 


The following document provides the user with the R code to generate African lion densities and other state variables for the QECA (and accompanying raw outputs from R) using the csv input files in Supplementary Information 1. We initially ran models at 11000 iterations (with a burn in of 1000) and achieved convergence for the Global (most parameterized), Xand Y models. The full (ie. Model with most support but where theta was 1), SexthetaSex and Theta models were run for 39000 iterations with a burn in of 1000 to achieve convergence. 

*This code is tested and correct as of 12 August 2019. If you experience any problems replicating these results, or have queries about the code, please contact the lead author on  

Usage Notes

Please use the accompanying manual (with R code) and RAW csv files to generate the results we produce in our paper. 


National Geographic Society, Award: WW-106ER-17

Siemiatkowski Foundation

University of Queensland

Siemiatkowski Foundation