The study of limiting factors in nature is very
difficult because ideally it requires changing the
amount of the factor alone and observing
whether this change affects the size of the
population. Theoretically, if more than one
factor changes, the analysis can still be per
formed, but in practice, if more changes of known
nature occur, more of an unknown nature usually
also occur. Limiting factors have been studied
in two ways. The best way is artificially to modify
single factors in the environment, observing the
effect upon the birds. MacKenzie (1946) reviews
some experiments of this type. The most notable
was the increase from zero to abundant of pied
flycatchers (M uscicapa hypoleuca) when nest
boxes were introduced in the Forest of Dean.
This showed conclusively that lack of nesting sites
had limited the population. Such simple modifica
tions are not always feasible. For instance, chang
ing the food supply of an insectivorous bird is
nearly impossible. The most feasible approach
in such a case is to compare the bird populations
in two regions which differ in the abundance of
the factor being considered. Ideally, the two
regions should differ only in this respect, but this
is very improbable. A good example of this
method of study is the work of Breckenridge
(1956) which showed that the least flycatchers
(Empidona.l' minimus'y were more abundant in
a given wood wherever the wood was more open.