Standardising distance sampling surveys of parrots in New Caledonia

Abstract

Standardised surveys are essential for monitoring populations and identifying areas that are critical for conservation. With the aim of developing a standardised method of surveying parrots in the rainforests of New Caledonia, we used distance sampling to estimate densities of New Caledonian Parakeets (Cyanoramphus saisseti), Horned Parakeets (Eunymphicus cornutus), Ouvea Parakeets (E. uvaeensis), and New Caledonian Rainbow Lorikeets (Trichoglossus haematodus deplanchii). We carried out surveys in the early morning and late afternoon, when parrots were easiest to detect. To minimise errors associated with estimating distances and flock sizes by ear, we conducted brief searches to locate parrots, then measured their distance from the transect line. We recorded birds in flight and consider these records to be important when estimating parakeet populations. In agreement with existing knowledge on distance sampling, we found line transects to be more efficient than point transects for estimating the density of parakeets. Our results indicate that parrots located beyond 50–70 m from the transect line have little influence upon density estimates. In addition, surveys on roads are likely to underestimate densities if not corrected for road width. We generated relatively stable and precise density estimates (CV < 0.25) with approximately 40–50 detections, yet additional effort may be warranted under different study conditions. Although we aimed to improve parrot surveys in New Caledonia, our suggestions may be useful to other researchers studying rainforest birds, and can be adapted to suit different species or environments.