compensate small trapliners for the usurpation of rich
food sources by territorialists. Trapliners might fre-
quently gain first access to new, rich nectar sources
(even though they are subsequently excluded by
dominant birds), or they may be able to continuously
locate and use cryptic or low-reward flowers not likely
to be exploited by the larger territorialists.
To test the pointer hypothesis prediction that
dominant species exploit the superior food-finding
abilities of subordinate species, I conducted two ex-
periments to quantify the abilities of Steely-vented
Hummingbirds and Canivet's Hummingbirds to locate
novel food sources. Trials were run using naive soli-
tary birds (experiment 1) and naive heterospecific
pairs (experiment 2) which were introduced for the
first time into an experimental flight cage containing
artificial feeders placed at locations unknown to the
birds. The cage was constructed from a 20-m tube
folded into five segments that were separated from
each other by opaque walls (Fig. 6.17). The feeders
received regular deliveries of sucrose solution at rates
sufficient to keep birds active but unsatiated. For each
experiment, I recorded the time at which each bird
visited each feeder.
Contrary to the pointer hypothesis, the results dem-
onstrated that birds of both species were equally adept
Figure 6.17. Flight cage (top view) to test (top) single
individuals of Amazilia and Chlorostilbon (n = 9 birds per
species, experiment 1) and (bottom) competing hetero-
specific pairs (n = 14 birds per species, experiment 2).at locating novel feeders in this experimental system.
When tested in competing pairs (experiment 2), the
subordinate trapliner was not able to discover signifi-
cantly more new feeders than the territorialist (x = 2.7
and 2.3 feeders discovered first by Cavinet's Hum-
mingbird and Steely-vented Hummingbird, respec-
tively, of 5 total feeders per trial). However, the domi-
nant Steely-vented Hummingbird was significantly
faster at making its first visit to a given feeder once it
had already been discovered by its cage-mate (x = 9.8
vs. 23.0 min elapsed between initial discovery by
cagemate and the first visit for Steely-vented Hum-
mingbird and Canivet's Hummingbird, respectively;
f-test, P = .004). After the trapliner had discovered a
feeder first, the territorialist subsequently made its
first visit to that same feeder in less than 1 min in more
than 60% of the cases (Fig. 6.18). The trapliner was
much less effective in exploiting the discoveries of its
cagemate, visiting the same feeder within 1 min only
30% of the time.
The territorialist's apparent ability to capitalize
on the food-locating efforts of a subordinate are con-
sistent with the pointer hypothesis but inadequate
to support it. The combined results offer stronger
support for alternative hypotheses formulated for
intraspecific competition which propose that domi-
nants can parasitize resources found by subordinates
without suggesting any differences in their respec-Time (min) to Locate a Feeder After Cage-mate
Figure 6.18. The abilities of Amazilia and Chlorostilbon to
locate novel feeders that had previously been discovered by
a cage-mate. The y-axis represents the number of feeders
located by each species within a specified time (x-axis
categories) after having been first discovered by a hetero-
specific cage-mate (experiment 2, n = 14 birds x 5 feeders =
70 feeders located for both species). Time categories are
nonoverlapping and are abbreviated ("< 8" represents 4 <
n < 8, etc.). In the shortest time category (< 2), all feeders
were located within 1 min (see text for analyses).217 Birds