addition to information on bird movements. Chapter 5 covers these issues
in detail.
The principle behind standard effort capture-mark-recaptureis that, if birds are
caught and individually marked (e.g. with rings or bands), then from the ratio of
marked to unmarked birds subsequently recaptured, population size can be esti-
mated. Imagine that on the first day of capture at a site, 100 birds of a particular
species were caught in nets, marked, and released. A week later, the nets were put
back up. This time 50 of the same species were caught, 25 of which had been
marked on the first day. If we assume that the population is closed and the original
100 birds caught had become fully mixed back in the population over the inter-
vening week, then the total population size of the species on the site is 200. That is,
we assume that the proportion of birds caught on the second date that were marked
(25/50 or 50%) is the same as that in the total population on the site. Because we
know that the number of marked birds is 100, then the total population is twice
that, that is, 200. Expressed mathematically: the total population size, Pn 1 n 2 /m 2
where n 1 is the number caught, marked, and released on the first date, n 2 the
number caught on the second date, and m 2 is the number of those caught on the
second date that were marked. In practice, there is no need to actually catch birds
on the second date, as they could be recorded by walking around the site trying
to see as many birds as possible and recording those that were marked.
While the capture-mark-recapture approach may seem simple, it is in practice
fraught with problems because it relies on a suite of assumptions, many of which
may be untrue. For example:
- It assumes that birds mix freely within the population and this may rarely
be the case. - It assumes that the population is closed and that no birds enter or leave the
population, either through births, deaths, or movements. - It assumes that marking does not affect the probability that a bird will
be recaptured, and that marked birds have the same probability of survival
as unmarked birds. - It assumes that marks do not fall off or become less visible.
While many of these assumptions may be broken, it is possible to plan fieldwork
to minimize their influence on the results. For example, if the first and second
capture dates are reasonably close together, the study site is well defined, and
the study is undertaken outside of the breeding and migration periods, then the
population will more approximate a closed one.
An array of mathematical models has been developed to analyze data from
capture-mark-recapture studies. While it is not within the scope of this chapter
48 |Bird census and survey techniques