The value to bird conservation of determining the causes of population declines is considered and the diagnostic methods available are reviewed, with examples. Diagnosis of the cause or causes of a decline in bird numbers is likely to be helpful in deciding the priority of conservation actions, though actions which aim to reverse the changes in external conditions which caused the decline need not be the most effective in initiating recovery.
The methods for diagnosing causes of declines in bird numbers with the widest application make use of comparisons between geographical areas or time periods with different trends. Correlations between trends in numbers and measurements of external factors are examined across areas or periods or both. The danger of spurious correlations is minimized by drawing up a list of plausible causes based on studies of the natural history of the species. The effects of all of these candidates should be examined, subject to availability of data. The consistency of observed changes over time, or differences among areas, in survival rate or breeding success with the postulated demographic mechanism of the decline should be examined.
Conclusions based on correlations across geographical areas between trends in numbers and external factors may be misleading if birds are able to move between the areas selected for comparison and if their pattern of settlement depends upon external factors thought to be implicated in the decline.
Manipulative experiments should be carried out to test conclusions drawn from correlative studies. However, it must be recognized that the capacity of birds to move between areas means that experiments may measure effects of manipulations on settlement patterns or distribution rather than population size. Experiments that appear well designed in terms of controls and replication may be misleading when applied to the conservation of bird populations if their geographical scale is inappropriate.