108 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 124. No. 1. March 2012 
FIG. I. Pattern of (A) lek attendance and (B) singing 
rate at the lek of Plovercrests. Sample sues used to 
calculate singing rates (i.e., number of I-min periods when 
songs were counted) are indicaled in parentheses. Means 
(small black squares), and standard errors (vertical bars) 
are indicated. 
returned to the main lek in August 2007; because 
they were unmarked, it is unknown whether all 
were among the original seven studied. 
The mean area encompassed by lek territories 
was 11.4 — 4.4 m 2 (n = 7, range = 7.0-16.0 nr). 
The mean distance between the center of 
territories was 14.8 ± 6.3 m (n = 7, range = 
9.0-25.0 m). Males used from two to five perches 
to sing (3.4 ± 1.0 perches, n = 7) within their 
territories, usually slender horizontal branches 
2.0 ± 0.8 m (n = 28, range = 0.8-3.5 m) above 
ground. 1 did not observe males feeding either 
inside their territories or at the lek area. 
Territory owners received up to four visits 
of conspecifics per observation period of 15 min 
(0.5 ± 0.8 vixits/observation period). Visits were 
slightly more frequent in early morning (0600- 
0700 hrs), but occurred throughout the day. 
Visitation rate varied only slightly through the 
breeding season. Upon arrival of a visiting 
conspecific in a territory, the resident male either 
immediately chased the intruding bird or, less 
olten, the resident male and intruder engaged in a 
stereotyped aerial display. Immediate chases 
occurred when the visiting bird was either an 
adult male or birds in a female-like plumage 
(i.e., females or young males) as far as the brief 
encounters and rapid movements of interacting 
birds permitted me to identify, whereas the aerial 
display was reserved exclusively for adult inales. 
The territory owner remained perched during 
aerial displays while watching the visiting bird 
perform a series of rapid, short (-10 cm) lateral 
flights just in front and slightly above it. Visiting 
birds kept their crests conspicuously erected while 
displaying. The territory owner, after a few 
seconds of displaying, either chased the visiting 
bird or the visiting bird Hew away. 
Males were frequently absent from their lek 
territories, and lek attendance by territory owners 
fluctuated throughout the day (H = 35.3, df = 12. 
P < 0.001; Fig. 2A). Males arrived at the lek area 
approximately at sunrise. Activity slowly dimin¬ 
ished after an initial period of activity after 
arrival, but increased again around 0900 and 
1500 hrs. All males left their territories by 1830 
hrs (Fig. 2A). 
The song given by lek king males was a 
complex, modulated vocalization uttered at a 
mean rate ol 74.8 ± 14.5 songs/min (range = 
46- 101 songs/min, n = 39; Fig. 3). Males sang at 
a similar rate throughout the lekking season (F iM 
= 0.4, P = 0.84: Fig. IB), but not throughout the 
day (Fig, 2B). Singing rates at times increased 
briefly when males returned to their territories 
alter a chase or when a conspecific passed nearby 
Overall, males were present on their territories 
during 35.8% of the observation periods with no 
differences among males ( H = 4.9, df = 6, P = 
0.56; Fig. 4A). I did find a difference among 
males in singing rates (F 6 . 32 = 5.8. P < 0.001; 
Fig. 4B), but singing rate was not correlated with 
lek attendance (r = 0.34, n = 7, P = 0.46). 
DISCUSSION 
The Plovercrest lek studied was similar to those 
ol other lekking hummingbirds. For instance, 
hummingbirds often establish leks in edge 
habitats (Skutch 1958, Atwood et al. 1991) as 
did Plovercrests. Even species typical of dense 
forests, such as hermits, may establish leks in 
areas of secondary vegetation along trails and 
river margins or in forest gaps (Stiles and Wolf 
1979; Ramjohn et al. 2003; M. A. Pizo. pers. 
obs.). Apparent preference for edge habitats may 
be the result of observer bias because these 
habitats are more accessible and more easily 
