Application of mark-recapture methods to lions: Satisfying assumptions by using covariates to explain heterogeneity

Accurate estimation of wildlife population size is central to effective management and conservation but is notoriously difficult for rare and cryptic carnivores. Mark-recapture methods can reliably estimate the abundance of large carnivores but are sensitive to departures from the assumption of equal catchability, which is probable for territorial and social carnivores. We empirically analysed how pride area, age and sex, group size and pride size of lions Panthera leo and possession of a kill affect the performance of the sight-resight method relative to concurrent total counts. We stratified the population by pride area to obtain a better sample size for reliable population size estimation, and determined the composition and size of each pride. Marking and recapture involved merely identifying and re-identifying individual lions. The per cent of the population marked in each pride area varied from 33.3 to 100. Twenty-six estimates of pride size for 22 different pride areas and repeat surveys in five pride areas were made between September 1990 and April 1992. The second samples involved sampling with replacement, so b5Bailey's (1952) bias-corrected estimator for closed populations was applied. Increasing the per cent of the population marked improved accuracy only when the number of lions marked in a pride area was at least seven. Estimates were somewhat poorer when the number marked was less than seven but improved with more markings, but no estimate of population size was biased high. Age, sex, group size, pride size and feeding events all influenced the sightability of lions. Notably, females (>1 year) were sighted or resighted in proportion to their presence in the population, whereas similarly aged males, adults or subadults fluctuated erratically about their population proportions. Sight-resight surveys of territorial and social carnivores are reliable but can be costly and time consuming