Wild birds are counted for a wide variety of reasons and by a bewildering array of methods. However, detailed descriptions of the techniques used and the rationale adopted are scattered in the literature, and the newcomer to bird census work or the experienced bird counter in search of a wider view, may well have difficulty in coming to grips with the subject as a whole. While not an end in itself, numerical and distributional census work is a fundamental part if many scientific and conservation studies, and one in which the application of given standards is vital if results are not to be distorted or applied in a misleading way.This book provides a concise guide to the various census techniques and to the opportunities and pitfalls which each entails. The common methods are described in detail, and illustrated through an abundance of diagrams showing examples of actual and theoretical census studies. Anyone with a bird census job to plan should be able to select the method best suited to the study in hand, and to apply it to best effect within the limits inherent in it and the constraints of the particular study. The Royal Society for the Protection of Birds and the British Trust for Ornithology have many years pioneered the collaboration of amateurs and professionals in various census studies. Three members of their staff, each with extensive filed experience, now pool the knowledge of these investigations to lay the groundwork for sound census work in future years.
This is perhaps the definitive overview of bird sampling. Extremely comprehensive, it covers most conventional sampling methods and explicitly lists the assumptions, typical procedures, advantages, and disadvantages of each. It doesn't shy away from the issue of detectability bias, though it doesn't go into depth on that topic (I recommend Buckland's book for that), which is an important consideration that other similar publications tend to gloss over (Ralph 1995, Verner).