Duration of survey changes interpretation of habitat preferences: an example of an endemic tropical songbird, the Bangwa Forest Warbler

Understanding the distribution, habitat preferences and population trends is crucial to effectively manage and protect birds. Currently, in this area, there are disproportionally more studies on temperate than on tropical species. Consequently managing, protecting and predicting population trends in tropical birds can be difficult. In our study we investigated detectability, distribution and habitat preferences of an endemic songbird, the Bangwa Forest Warbler Bradypterus bangwaensis (BFW) in the Bamenda Highlands (NW Cameroon). We used autonomous sound recorders to detect the BFW at 192 points and described habitat around these points. The survey was conducted from one hour before sunrise to three hours after sunrise. BFW was widespread within the study area, recorded at 65% of surveyed points. The probability of detecting BFW ranged from 26% during the first hour before sunrise to 93% during three-hour survey (from one hour before sunrise to two hours after sunrise). Three of ten habitat characteristics significantly affected BFW occurrence: altitude (negatively), vegetation cover at 3–5 m (positively) and at 5–10 m (negatively) above ground. Varying survey duration (one, two, three or four hours) and then comparing distribution patterns obtained from species occurrence of each survey duration illustrated that different habitat features were identified as relevant to BFW occurrence. We showed that BFW is a habitat generalist, inhabiting different types of montane bush. Deforestation may cause population declines of this species. We suggest that the counting methods based on short-time survey duration may be inaccurate in tropical environments, at least for detecting species similar to the BFW, which sings irregularly, and is detected mostly only aurally. Autonomous sound recording might be an effective alternative to the traditional approach to determination of bird distribution in such cases, because it enables increasing survey duration, appropriate sampling in time and increased probability of species detection.