through their large home ranges, picking up and drop-
ping off members of other species ("facultative join-
ers") whose smaller territories are nested within the
large home ranges (Buskirk 1972, 1976, Powell 1979,
Shopland 1985, J. Shopland, unpubl. data). In the
MCFP, the nuclear species are the Common Bush-
Tanager and the Three-striped Warbler. Nuclear
species in lower elevation woodland flocks are the
Golden-crowned Warbler and the Lesser Greenlet.
Three questions were central to my studies: (1)
does foraging with a flock affect the foraging success
of an individual flock member, (2) what are the
mechanisms for these changes, and (3) what do they
suggest about the evolution of flocking behavior? I
studied two facultative joiners, the Slate-throated
Redstart (Fig. 6.7) and the Collared Redstart, in
MCFP flocks and one nuclear species, the Golden-
crowned Warbler, in woodland flocks. I used focal-
animal samples (Altmann 1974) of foraging behavior,
following one individual and recording how much
insect mass was consumed per minute. Intake rate was
measured as a composite of the number of foraging
maneuvers per minute, the proportion of foraging
maneuvers that were successful, and the size of prey
caught. I also recorded the bird's rates of social and
territorial behaviors, which diminish potential for-
aging time. I sampled behavior within and outside
flocks. To compare success, I pooled events per
minute into before-flock, with-flock, and after-flock
blocks following Shopland (1985).
Contrary to the predictions of the foraging-success
hypothesis, both species of redstarts ate less food per
unit time in the company of mixed-species flocks
than they could have found on their own. When they
joined flocks, their intake rates decreased by one-third
(Slate-throated Redstarts) and two-thirds (Collared
Redstarts). Neither foraging maneuver rate nor propor-
tion of success could account for these losses; prey
size was decisive. Both species caught fewer large
insects when they foraged with flocks. Outside flocks,
Slate-throats were predominantly aerial hawkers, but
in flocks they shifted to striking at and gleaning from
leaves and branches. Collareds took up hawking in
flocks versus their usual striking-and-gleaning style.
Both redstarts' activity levels (perch changes per
minute) and rates of social interaction and territo-
rial display ("strutting") increased when they joined
flocks. "Floaters" and neighboring pairs of redstarts
often intruded as members of a mixed-species flock,
which provoked increased chasing and display. Thus,
rather than enjoying greater foraging success in flocks,
these facultative joiners suffered two kinds of losses:
(1) large insects (which may have been taken by indi-
viduals of more specialized flock species), and (2)
time and energy spent in chasing invading members
of their own species. Nevertheless, redstarts actively
joined flocks, making the best of a bad lot.
Golden-crowned Warblers have home ranges that
are three times the size of redstart territories (Buskirk
1972) and are found in 97% of woodland flocks. These
warblers fared no better in flocks than did the red-
starts; their intake rate decreased by 40%. As with
redstarts, the loss of large insects was the source of
decrease. Unlike the two facultative joiners, however,
Golden-crowns did not switch to uncharacteristic
foraging methods in the presence of flocks (they con-
tinued to glean), nor did they show higher levels of
social interactions.
These results challenge the view that multispecies
flocking has evolved to improve foraging efficiency.
Predation may have been a stronger selective pressure
shaping the tendency of these species to flock. Results
suggest that feeding with mixed-species flocks can
have a large impact on within-species territoriality by
increasing the costs of defense. Over evolutionary
time, these costs could lead to coincident territories,
joint defense, and obligate flocking, as in lowland
flocks in South America (Munn and Terborgh 1979,
Powell 1985).
These studies raised questions that could be directly
explored by focusing on a limited number of gridded
territories with color-marked individuals (owners,
neighbors, and floaters). Although neotropical mi-
grants' attendance of mixed flocks has been extensively
documented, the effects of this sociality on their for-
aging success (and thus potentially on their nonbreed-
ing season survival) has not been investigated. Mixed-
species flocks are a conspicuous part of avifaunas in
many regions, but little is known of the effects of habi-
tat degradation and fragmentation on flock formation
and the behavioral flexibility of individual flocking
species. Monteverde, with its patchwork of pastures,
second growth, relict woodlands, and primary cloud
forest, is an ideal setting for such research.207 Birds