LAUSR

Abstract Context. Estimating population abundance can be plagued by the violation of methodological assumptions, which can be overcome with standardised protocols. The black caiman (BC) is considered a conservation-dependent species, and previous abundance estimates are surrounded by uncertainty and flaws in the survey (e.g. different survey design and efforts) and analytical approach used (e.g. relative abundance index, which ignores imperfect detection). Its conservation status assessment demands the implementation of a standardised monitoring protocol. Aims. The protocol provides guidelines to collect and analyse data in a consistent manner to survey BC. Besides accounting for imperfect detection to produce reliable abundance estimates, the protocol aimed to be easily implemented by park rangers, and to fit field observations into a hierarchical modelling approach to assess how environmental variables affects detectability and abundance. Methods. The protocol subdivides a 20-km transect into 10 2-km segments; each transect is surveyed four consecutive nights, starting at 1900 hours and finishing when the 20 km are completed. For each caiman detected, the observers estimated head size to classify the individual by age. We tested the protocol in Ecuador during January and December 2017, and field data were analysed using N-mixture models. We compared abundance estimates derived with this protocol with commonly used relative abundance indexes. Key results. We surveyed 460 km that resulted in 177 detections. Percentage of moonlight and distance from human settlement best explained detectability and abundance respectively. Mean detection probability was 0.14 (95% BCI: 0.1–0.18), whereas absolute abundance was 196 (95% BCI: 147–370). The overall adult to immature ratio was 1 : 1.3. Conclusions. This is the first estimate of detectability and absolute abundance for BC by using a standardised survey with a clearly defined and repeatable survey and analysis methods. Relative abundance indexes did not reflect absolute abundance estimates. We recommend the use of this protocol in future surveys across the Amazon region to effectively evaluate BC conservation status. Implications. Population size cannot be estimated from relative abundance indexes; they lead to bias estimates for ignoring imperfect detection. We discourage the use of relative abundance indexes to evaluate the conservation status of this species.