SHORT COMMUNICATIONS 
405 
from the saki’s left side, grabbing the head and 
chest with its left foot and the lower half of the 
animal with its right foot. The anterior end of the 
monkey contained four punctures: the outer toe 
punctured the dorsal side of the neck at the spinal 
column, the middle toe entered the right side of 
the neck at the jugular, the inner toe entered the 
right side of the ventral thorax, and the hallux 
pierced the left side of the ventral thorax. The two 
punctures in the thorax formed a mirror image 
30 mm below the collar bone at the location of the 
heart. Harpy Eagles typically make between two 
and four punctures to secure the posterior end of 
their prey, but the posterior end contained only 
one visible puncture on the right flank that was 
60 mm anterior to the hip joint. We may have 
missed a posterior puncture(s) as they can be 
difficult to locate quickly because they can close 
without external hemorrhaging (Tania Sunaiotti, 
pers. comm.). 
We noticed the Harpy Eagle staring intently at 
us during our examination of the saki. moving its 
head side-to-side and up-and-down, from a perch 
15 m high and 10 m distant. We hid in hope of 
seeing the eagle return to the kill but abandoned 
this effort after 20 min. We relocated the eagle 
perched 8-10 m up in the understory and 20-25 m 
from the carcass as we left the area, The Harpy 
Eagle flushed when we were within 15-20 m and 
flew ahead of us in the understory in the same 
direction in which we were walking. Alter moving 
forward 40 m in the understory, the Harpy 
suddenly flew into the canopy and made a 180- 
degree turn, reluming to the area of the kill. 
The saki was attacked while traveling in the 
smaller (5 individuals) of two subgroups (—-18 
individuals total). Typical black-bearded saki 
alarm calls began following the first loud scream 
at 0911 hrs. The smaller subgroup moved to the 
location of the larger subgroup. 40 m from the 
location of the attack, where together they alarm 
called vigorously until 0935 hrs (24 min). Indi¬ 
viduals then gradually moved away until all 
alarms ceased (0946 hrs) and all individuals 
appeared to have left the area. 
We infer the first scream was emitted by the 
individual that was attacked during an unsuccess¬ 
ful arboreal attempt. As it screamed we believe 
the monkey jumped 20-25 m to the ground, 
successfully eluding the eagle's grasp. We did not 
see the apparent terrestrial strike, but it appeared 
to be accompanied by another scream from the 
saki because it was at the location of this scream 
that we saw the sapling shaking immediately 
following the second scream and from which the 
Harpy Eagle flew from the ground. We believe 
that all wounds occurred during the second 
attempt and that this attempt was fatal. 
Predation Attempts with Uncertain Outcomes .— 
The second and third Harpy Eagle-primate 
interactions, involving another black-bearded saki 
and a red howler monkey ( Alouatta seniculus), 
occurred in the last 25 years while one of the 
authors (dos Reis) worked on a variety of projects 
at the BDFEP. The second interaction was a 
Harpy Eagle attack on a black-bearded saki where 
the eagle attempted to capture an arboreal saki at 
one of the BDFFP’s continuous forest research 
sites, Cabo Frio. The monkey screamed at the 
moment of attack, jumped to the ground, and 
climbed 1-2 m up an understory tree where it 
remained. The eagle landed 15-25 m distant and 
surveyed the area while also studying the 
observer. As the observer left he heard another 
scream during what was likely a second attack 
with an unknown outcome. 
During the third incident, an attempt on an 
arboreal red howler monkey at the continuous 
forest research site of Km41, the howler made an 
atypical vocalization at the moment of the attack 
and swung beneath the horizontal branch on 
which it was standing, briefly hanging upside 
down while the eagle pussed over the top. The 
howler then returned to the top of the branch and 
ran towards its terminal end, moving to a 
neighboring vine-filled tree where it hid in thick 
vegetation with the rest of its group. The eagle 
perched 30 m distant and again surveyed the area, 
including the observer. As the observer left the 
area he heard another atypical howler vocaliza¬ 
tion, likely during a second attempt. The two 
vocalizations made by the howler during this 
encounter have not been heard by the observer 
(dos Reis) before or since (a total of 25 years). 
The outcome of this Interaction is also unknown. 
Post-capture Observations. —The final two 
observations, made from a truck on the large dirt 
road (ZF-3) used to access many of the research 
camps, were of Harpy Eagles that already had 
monkey carcasses. In the first, a Harpy Eagle was 
observed on top of a dead red howler monkey, 
using its beak to remove the howler’s hair, which 
was placed in a single pile next to Ihe body. The 
eagle flushed upon the approach of the vehicle but 
likely returned as the carcass was gone when the 
truck returned. The second observation was of a 
