Accurate abundance estimates can contribute to effective management of large carnivore populations. Lion, Panthera leo, and spotted hyena, Crocuta crocuta, populations are frequently estimated at night by eliciting their approach using broadcasted vocalizations. Spotlights are typically used to observe these species on approach but can disturb animals and adversely affect counts. We compared the efficacy of spotlight with red filters and forward looking infrared (FLIR) thermal monocular to enumerate lions and spotted hyenas in Serengeti National Park (SNP; non-hunted area) and Maswa Game Reserve (MGR; hunted area), Tanzania, during 2015−2017. We established 119 call-in sites in SNP and 20 in MGR and conducted repeated call-ins at 1–2 week intervals. During call-ins we conducted systematic paired counts using both devices. We assessed the influence of device order, species, hunting regime, and land cover on species counts. We found that FLIR was more efficacious for counting hyenas in MGR and spotlight for counting lions in SNP. We found evidence for temporary artificial light disturbance in MGR, as counts were higher when FLIR was used as the second device. Habitat type within 200 m of call-in sites did not influence device performances. Greater spotlight efficacy in SNP is a likely consequence of lower perceived risk and less anthropogenic disturbance compared to MGR. To improve accuracy of counts and subsequent population estimates for lions and spotted hyenas, we recommend consideration of variation in device efficacy, based on species surveyed and management regime.

We established 119 call-in sites in SNP and 20 in MGR and conducted repeated call-ins at 1–2 week intervals. During call-ins we conducted systematic paired counts using both spotlight and a thermal camera.