The African lion (Panthera Leo) has suffered drastic population and range declines over the last few decades and is listed by
the IUCN as vulnerable to extinction. Conservation management requires reliable population estimates, however these data
are lacking for many of the continent’s remaining populations. It is possible to estimate lion abundance using a trophic
scaling approach. However, such inferences assume that a predator population is subject only to bottom-up regulation, and
are thus likely to produce biased estimates in systems experiencing top-down anthropogenic pressures. Here we provide
baseline data on the status of lions in a developing National Park in Mozambique that is impacted by humans and livestock.
We compare a direct density estimate with an estimate derived from trophic scaling. We then use replicated detection/nondetection
surveys to estimate the proportion of area occupied by lions, and hierarchical ranking of covariates to provide
inferences on the relative contribution of prey resources and anthropogenic factors influencing lion occurrence. The direct
density estimate was less than 1/3 of the estimate derived from prey resources (0.99 lions/100 km2 vs. 3.05 lions/100 km2).
The proportion of area occupied by lions was Y= 0.439 (SE = 0.121), or approximately 44% of a 2 400 km2 sample of
potential habitat. Although lions were strongly predicted by a greater probability of encountering prey resources, the
greatest contributing factor to lion occurrence was a strong negative association with settlements. Finally, our empirical
abundance estimate is approximately 1/3 of a published abundance estimate derived from opinion surveys. Altogether, our
results describe a lion population held below resource-based carrying capacity by anthropogenic factors and highlight the
limitations of trophic scaling and opinion surveys for estimating predator populations exposed to anthropogenic pressures.
Our study provides the first empirical quantification of a population that future change can be measured against.