Comparison of two methods of counting birds at sea

In an attempt to improve techniques for surveying marine birds, the author compares methods for estimating relative abundance and density. Quantitative relationships exist between estimates of relative abundance and density for most marine bird species. Estimates of abundance can be derived from values of density by regression models when the following factors are considered: bird behavior (flying, sitting, and ship-following), duration of the count, and ship speed. The author's data suggest a minimum numerical relationship of 2:1 between estimates of abundance and density for sitting birds and 1:1 for flying birds. Neither method of counting birds at sea eliminates the problem of ship-followers. However, the density method minimizes an inflationary effect caused by ship-followers that is apparent with the abundance estimate, especially when the observer's vessel is associated with fishing activities. Differences in abilities of observers to count birds is a principal, but unavoidable, source of variability with both counting methods. The density method also minimizes this problem by focusing the observer's view in an area with known boundaries. This standardization decreases variability in bird detection due to weather, size and color differences between species, flock sizes, and bird behavior. These factors are not controlled in estimates of abundance.