birds there depend largely on aban- 
doned woodpecker holes. 
Purple martins migrate through 
Central America and the Caribbean 
to the Amazon River basin of Brazil, 
where they winter. Their annual ar- 
rival in the United States begins in 
late January in Florida and Louisiana 
and continues through May, farther 
north. Early arrival time may be re- 
lated to finding a nesting cavity, a 
problem encountered by hole nesters 
that do no{ excavate their own cavities. 
Early arrival probably resulted from 
intraspecific competition for nesting 
sites, a behavior that evolved before 
humans provided martins with bird- 
houses. When the only nesting sites 
available were unused tree cavities, 
competition for sites was probably in- 
tense. If early arriving birds got the 
best or the only cavities, and as a 
result raised the most young, then ar- 
riving early could be selected for. A 
counterbalance to an increasingly 
early arrival in breeding areas is the 
probability that early birds will face 
periods of cold weather and starvation, 
and thus extremely early arrival could 
be selected against. 
When purple martins nested largely 
in abandoned woodpecker holes (as 
the small western populations still do), 
they were probably mainly solitary 
nesters. This statement is supported 
by two kinds of evidence. First, before 
men and birdhouses came along, colo- 
nies of martins would have been pos- 
sible only where woodpeckers had ex- 
cavated a number of holes close 
together. However, most woodpeckers 
in the eastern United States that ex- 
cavate cavities large enough for purple 
martins to use are solitary-nesting 
woodpeckers, and it is improbable that 
they bore and then abandon enough 
holes in any single tree to allow colo- 
nies of martins to form. Acorn wood- 
peckers ( Melanerpes formicivorous) 
in the West are colonial and do dig 
many holes close together, but their 
extreme aggressiveness around their 
nesting sites prevents martin colonies 
from forming there. 
Second, purple martins are not able 
to recognize their own young. This 
is strong evidence that they originally 
were solitary nesters. Biologists be- 
lieve that whenever there is a chance 
that young of different parents may 
get mixed up, such as in dense colonies 
of birds, selection is intense for chick 
recognition so that an individual par- 
66 
