From Ben Rees
Paul Clark is quite right to point out that the mark-and-recapture method
depends on all individuals having an equal chance of being recaptured
(Letters, 20 June, p 55).
In fact all ecological census methods incorporate that
assumption in some form or other.
I’m doing a lot of work on passerine bird census methods and they are all
based on the assumption that all species are equally observable. This is clearly
not the case, as some species of passerines (particularly warblers) are
notorious for being very shy and the presence of observers can skew results
because they scare individuals away. Other species have the opposite problem of
being too bold and could be counted twice.
The keys to reducing the errors to a statistically acceptable level are to
use skilled observers and standardise the search method.
For particularly shy/rare species you can use more specialised methods. For
example, British ornithologists use the characteristics of individual calls to
count the number of bitterns.
Advertisement
As with all research you have to admit to the shortcomings of your method and
account for them in the discussion of your results.
Welham Green, Hertfordshire
