and problems is considerable. On the other hand, the literature on bird numbers,
habitats, distribution, and trends is increasingly full of interesting and practically
important findings. Many have been produced by some of the simpler methods
described in this chapter. Superficially this is paradoxical. I believe the explanation is
that many of the things we need to know about for practical conservation purposes
are sufficiently plain as to be revealed by studies with quite simple methodology.
It is clear to me, on a global scale, we have so far done insufficient to document
the basic parameters of distribution and relative abundance of birds. This is
especially true in the species-rich tropics where there is a great need for better data
to understand and respond to change. It is also clear that simple methods can be
very powerful. I suspect that some of the more sophisticated methods, especially
those used for monitoring in wealthy northern countries, have deterred people
from seeing the potential of simpler approaches for primary exploration and
documentation. The earlier we lay down such baseline information, the quicker
future generations will be able to use it to promote conservation.
It would be nice if there were a standard approach, because this would
make data more comparable across studies and capable of being amalgamated for
different purposes. While individual studies will continue to need specific design
elements, some general standards can be recommended (Table 1.4).
12 |Bird diversity survey methods
Table 1.4Key points for designing bird surveys
1 The selection of fieldwork locations should be designed (systematic, random, or
stratified) and the design should be a documented part of any publication or
database
2 The recommended basic field method is the collection of multiple complete species lists
each within a defined period of time and area of ground
3 The time period should be 20 min, 1 h, or 1 day with the smallest possible chosen to be
consistent with aims, observer acceptability and data handling consequences
4 Individual timed lists should be collected from the smallest reasonable area (2 ha, 1 km^2 ,
10 k m^2 , or quarter degree cell), which should be defined and located by map reference, or
GPS coordinate. Separate lists are preferable to amalgamation across major changes of
habitat or altitude
5 If evidence on breeding status is collected it should follow EOAC standards (above)
6 The completeness of lists should be assured by use of a stopping rule whereby single
occurrences are as frequent or less than doubles. Completeness should be explicitly
considered and declared in published analyses
7 For dispersed species, the frequency of occurrence on lists is a measure of relative
abundance. This does not make sense for congregatory species where numbers should be
estimated
8 Survey data should be deposited in appropriate secure and persistent electronic archives
to maximize their potential value