Temporal and Spatial Variation of Broadcasted Vocalizations Does not Reduce Lion Panthera leo Habituation

Estimating abundance of large carnivores is often challenging, yet important for their effective conservation. Attracting lions Panthera leo for visual enumeration using broadcasted vocalizations (i.e. call-ins) is commonly used to estimate their abundance. However, call-ins are typically not repeated at sites because of habituation. We evaluated lion response to repeated call-ins varying temporally (1 or 2 weeks) and spatially (1 or 2 km) in Serengeti National Park (SNP), Tanzania, during February-April 2016. We established 30 call-in sites: at 10 sites we used lion and spotted hyena Crocuta crocuta calls while alternating prey distress calls across five consecutive weeks; at 10 sites we alternated calls but conducted call-ins once every two weeks; and at 10 sites we alternated calls weekly for five weeks but moved among three locations separated by 1 or 2 km each week. We used N-mixture models to assess which sampling design would elicit greater lion detections across sessions and to estimate overall abundance. Lions habituated to broadcasted calls within each sampling design, with detectability point estimates overall declining across sessions. Estimated lion abundance at these 30 sites was 198 (95% credible interval = 186-214). Altering time interval and location of call-ins was ineffective at reducing lion habituation. However, altering calls across sessions appeared to reduce lion habituation when compared to a previous survey in SNP that used the same calls across sessions. We recommend that call-in surveys using repeated broadcasts are conducted at the same sites across sessions and use different calls to improve lion response and consequently, estimates of abundance.