Blake and Loiselle • VARIATION IN GLYPHORYNCHUS ABUNDANCE 
439 
TABLE 1. Numbers of Glyphorynchus captured (Cap.), recaptured (Rec.), and observed (Obs.) on two 100-ha plots. 
Tiputini Biodiversity Station. Ecuador, 2001 2010. CR = capture rate expressed as captures/100 mn/hr (one mist net open 1 hr 
= 1 mn/hr). Numbers are for February/March when iwo values are listed for Obs. but for February il only one value is given. 
Year 
Harpia 
Puma 
Harpia 
Puma 
Cap. 
CR 
Rec. 
‘Tr Rec. 
Cap. 
CR 
Rec. 
<3> Rec. 
Obs, 
Obs. 
2001* 
55 
9.2 
1 
2 
53 
8.1 
2 
4 
2002 
89 
7.6 
41 
46 
85 
7.9 
42 
49 
61/61 
50/38 
2003 
80 
6.3 
49 
61 
83 
6.7 
42 
51 
63/66 
32/54 
2004 
109 
7.9 
63 
58 
107 
9.0 
58 
56 
64/80 
52/61 
2005 
98 
8.1 
52 
53 
121 
10.2 
76 
63 
62 
82 
2006 
91 
7.9 
52 
57 
95 
8.2 
55 
58 
93 
62 
2007 
86 
7.4 
40 
47 
77 
6.8 
45 
58 
2008 
87 
7.7 
46 
53 
123 
10.8 
76 
62 
2009 
88 
7.3 
57 
65 
119 
10.2 
66 
55 
82 
49 
2010 
78 
7.3 
46 
59 
100 
8.7 
78 
78 
65 
57 
* Only one misi-net sample. March. 
common way to characterize point patterns; tests 
for deviation from random are based on Ripley’s 
L-hat (0 = random. <0 = clumped. >0 = 
uniform). Significance of E-hat values were based 
on permutation tests implemented with PAS- 
SaGE. Calculation of the K statistic was corrected 
for edge effects (Rosenberg and Anderson 2011). 
We compared numbers of observations within 
25-m radius circles around net sites based on the 
habitat types associated with each net site, as 
described for captures. Results are given as means 
and SE. unless otherwise noted. 
RESULTS 
Capture Data. —We recorded 861 captures of 
Glyphorynchus on Harpia (414 individuals, 447 
recaptures) and 963 captures (423 individuals, 540 
recaptures) on Puma (Table 1). Overall capture rate 
per year was slightly higher on Puma (mean = 8.7 — 
± 0.45) than on Harpia (7.7 ± 0.23; paired /-test, zl 
r = 2.23, df = 9, P = 0.052). The difference in ^ 
capture rates was largely a result of greater numbers t 5 
of recaptures on Puma (569r of total) than on Harpia ™ 
(5257-); number of individuals captured per year did ^ 
not differ between plots (paired /-test, / - 0.21. df = ^ 
9. P = 0.84). Capture rates were fairly consistent ££ 
across years but were somewhat more variable on q! 
Puma (CV = 16.3) than on Harpia (9.6; variance 
ratio test for capture rate y = 3.67. P = 0.082). 
Capture rates were not well correlated between plots 
(/' = 0.27, P = 0.45). 
Most individuals were captured only once or 
twice (Fig. 1) with decreasing numbers captured 
multiple times. Most had moved 0 to 150 m from 
the previous capture (i.e.. recaptured in the same 
net or in nets up to three sites away; Fig. 2); 
few individuals were recaptured >250 m from 
the previous location although occasional, longer 
recapture distances were recorded. The mean dis¬ 
tance between capture sites within a year (i.e., 
within the same month or between months) was 
less on Harpia than on Puma (/ = 2.85, df = 448, 
P < 0.005; Fig. 3); mean recapture distance from 
I year to the next had less variation between plots 
(/ = 1.16, df = 512 ,P = 0.25). 
Captures were not evenly spread among all nets 
within a plot (Fig. 4) with most nets capturing ~4 
to 14 birds. All nets on each plot captured at least 
one individual, indicating Glyphorynchus oc¬ 
curred throughout most of each plot. Captures 
Number of captures per individual 
FIG. I. Number of captures per individual: approxi¬ 
mately half of all Glyphorynchus were captured only once 
on two study plots at Tiputini Biodiversity Station. Ecuador. 
