Hostname: page-component-7c8c6479df-5xszh Total loading time: 0 Render date: 2024-03-18T16:31:03.314Z Has data issue: false hasContentIssue false

A picture is worth a thousand words: the application of camera trapping to the study of birds

Published online by Cambridge University Press:  07 August 2008

Timothy G. O'Brien*
Affiliation:
Wildlife Conservation Society, 2300 Southern Blvd, Bronx, NY 10460, U.S.A., and Mpala Research Centre, P.O. Box 555, Nanyuki, Kenya
Margaret F. Kinnaird
Affiliation:
Wildlife Conservation Society, 2300 Southern Blvd, Bronx, NY 10460, U.S.A., and Mpala Research Centre, P.O. Box 555, Nanyuki, Kenya
*
*Author for correspondance; e-mail: tobrien@wcs.org
Rights & Permissions [Opens in a new window]

Abstract

Core share and HTML view are not available for this content. However, as you have access to this content, a full PDF is available via the ‘Save PDF’ action button.

This study reviews the use of remotely triggered still cameras, known as camera traps, in bird research and suggests new methods useful for analyzing camera trap data. Camera trapping may be most appropriate for large, ground-dwelling birds, such as cracids and pheasants. Recent applications include documentation of occurrence of rare species and new species records, nest predation studies and behavioural studies including nest defence, frugivory, seed dispersal, and activity budgets. If bird postures are analyzed, it may be possible to develop behavioural time budgets. If birds are marked or individually identifiable, abundance may be estimated through capture-recapture methods typically used for mammals. We discourage use of relative abundance indices based on trapping effort because of the difficulty of standardizing surveys over time and space. Using the Great Argus Pheasant Argus argusianus, a cryptic, terrestrial, forest bird as an example, we illustrate applications of occupancy analysis to estimate proportion of occupied habitat and finite mixture models to estimate abundance when individual identification is not possible. These analyses are useful because they incorporate detection probabilities < 1 and covariates that affect the sample site or the observation process. Results are from camera trap surveys in the 3,568 km2 Bukit Barisan Selatan National Park, Indonesia. We confirmed that Great Argus Pheasants prefer primary forest below 500 m. We also find a decline in occupancy (6–8% yr−1). Point estimates of abundance peak in 2000, followed by a sharp decline. We discuss the effects of rarity, detection probability and sampling effort on accuracy and precision of estimates.

Type
Research Article
Copyright
Copyright © Birdlife International 2008