No abstract

Due to global biodiversity declines, conservation programmes have increasingly had to consider reintroducing captive animals into the wild. However, reintroductions often fail as captive individuals may be naïve to predators and do not recognise or respond appropriately to predatory cues, contributing to high mortality rates soon after release. This study evaluates differences in predator-response behaviours between individuals from three experimental groups, a captive population (n = 13), a semi-wild population (i.e. raised in captivity and successfully released; n = 6) and a wild population (n = 2) of cheetah (Acinonyx jubatus) to an artificially simulated auditory threat of lions (Panthera leo), a larger, natural predator in South Africa. Such comparisons improve our understanding of differences between captive and wild behaviours and provide an aspect to evaluating the relative success of reintroduction programmes.Changes in the proximal distance, the latency of approach and hesitation towards both control (African bush cricket, Acanthoplus discoidalis) and treatment (lion) auditory cues were observed for 29 cheetah from captive, semi-wild and wild populations in at least three trial replicates each. Overall, captive individuals consistently displayed poor predatory-response behaviours, approaching the treatment as often as the control, as well as spending more time near the stimulus (<10 m) and hesitating more often than semi-wild cheetah, which could distinguish between the control and treatment, consistently fleeing from the latter with little hesitation. Repeatability analyses indicated that these behavioural responses to predatory cues could not be explained by individual personality and between-trial learning comparisons showed no evidence of habituation. Our findings demonstrate how a priori testing for predator naïvety could inform future introductory decisions and thereby increase post-release survival rates, significantly improving the efficacy of reintroduction strategies. We, therefore, emphasise the importance of such research and screening in highly threatened species, such as cheetah, where reintroduction from captivity has become a necessary consideration.

In South Africa, cheetah (Acinonyx jubatus) occur as a relictual, unmanaged population of 'free-roamers', a managed metapopulation across fenced reserves, and in various captive facilities. To ensure that the Cheetah Metapopulation Project (CMP) is not at risk of losing overall genetic variation to drift or inbreeding, we propose various interventions, including exchanges between free-roamers and the metapopulation or supplementation with unrelated individuals from captivity. Simulated trajectories of genetic diversity under such intervention strategies over time could directly inform conservation action and policy towards securing the longterm genetic integrity of the CMP. Single Nucleotide Polymorphisms (SNPs) were genotyped for 172 adult cheetahs across the free-roamer population, the metapopulation, and three major captive facilities. Management intervention trajectory models were tested including, (1) no intervention, (2) genetic exchange between freeroamers and the metapopulation, (3) translocation from a single captive facility and (4) translocation from several captive facilities into the metapopulation. Discriminant Analysis of Principal Components (DAPC) showed that two captive populations are highly differentiated from the metapopulation and each other, whilst the third captive and free-roamer populations are genetically more similar to the metapopulation. Simulated genetic variation over 25 generations indicated that models 1 and 2 show significant losses of heterozygosity due to genetic drift and present a proportional increase in the frequencies of 1st-and 2nd-degree relatives, whilst this variation and pairwise relatedness remain relatively constant under models 3 and 4. We emphasise the potential importance of captive facilities as reservoirs of genetic diversity in metapopulation management and threatened species recovery.