Srbek-Araujo et al. • RED-BILLED CURASSOW BEHAVIOR 
323 
20 sec to take multiple photos of the same animal or 
groups of animals. Cameras were checked and 
maintained even- 30 days. Camera traps were 
attached to trees >15 cm DBH. 45 cm above the 
ground No bait w as used. 
Camera traps w'erc placed along unpaved roads 
in the reserve during the first year. These roads 
were 4 m wide and used solely by internal stall 
and researchers. Three areas were sampled with 
10 trapping stations each: north, south, and west 
totaling 30 trap stations. Every trapping station 
included a pair of cameras, placed facing each 
other, one on each side of the road. The minimum 
distance between trapping stations was 2 km. 
Ten trapping stations with pairs of camera traps 
were placed during the second year inside the 
forest 100-200 m from the nearest road to deter 
theft. The camera pair was relocated to another 
position within 200 m of the original location 
if few photographs were taken after 30 days, 
keeping the same original distance from the 
nearest road. The minimum distance between 
trapping stations was 4 km. The spatial design of 
camera traps during the third year was similar to 
that of the second year; however cameras were 
placed at distances even farther from roads (500 m 
from ihe nearest road). 
Data Analysis .—Photographs from camera 
traps were considered records ( = captures). Only 
die first photograph was considered as a record 
when more than one photograph of the same 
species was taken at the same trapping station 
within a 1 -hr interval. The number of trapping 
stations and sampling effort varied from year to 
year, and we balanced comparisons using capture 
success. Sampling effort (camera-days) was calcu¬ 
lated by multiplying the number of cameras (or pair 
of cameras) by the number of sampling days 
(24 hrs). Capture success was calculated by 
dividing the number of records by the sampling 
effort and expressing the result as a percentage 
(Srbek-Araujo and Chiarello 2005). 
We used Chi-square tests ty; ) lor comparisons, 
corrected for bias when necessary (Yale's correc¬ 
tion: Zar 1996). Expected frequencies were 
calculated respective to sampling effort in each 
case when comparing the record frequency among 
habitats (roads, edges, and forest interior). We 
used only data from the first year to compare 
number of captures among regions (north, south, 
and west), when these regions were subjected to 
similar sampling effort. Wet and dry- seasons were 
contrasted using data from all 3 years and 
TABLE 1. Number of records, sampling effort, and 
capture success of Red-billed Curassows in the Vale 
Natural Reserve. Brazil using camera traps from June 
2005 to October 2008. 
Season 
Variables 
Totals 
Dry Wet 
First Year 
Number of records 
83 
62 21 
Sampling effon (camera-days) 
3.032 
1.563 1.469 
Capture success (%) 
2.74 
3.97 1.43 
Second Year 
Number of records 
64 
19 45 
Sampling effort (camera-days) 
3.468 
1.998 1.470 
Capture success (%) 
1.85 
0.95 3.06 
Third Year 
Number of records 
50 
31 19 
Sampling effort (camera-days) 
3.034 
1.688 1.346 
Capture success (%) 
1.65 
1.84 1.41 
Totals 
Number of records 
197 
112 85 
Sampling effort (camera-days) 
9.534 
5.249 4.285 
Capture success (%) 
2.07 
2.13 1.98 
separately for each year. Captures were classified 
into individual (either male or female), pair (male 
and female) or group (2 or more individuals apart 
from pairs). Males and females were distinguished 
given the dimorphism in plumage. The capture 
frequency of males, females, and associations 
(pairs and groups) was compared for the entire 
period and also yearly (3 X 3 contingency table). 
We examined the time of the photographs to 
define when captures were most frequent using 
2-hr intervals. The daily activity pattern was 
contrasted between males and females using Chi- 
square analysis with a 2 X 8 contingency table, 
considering data from all 3 years. We used 
Program BioEstat (Version 5.0) (http://www. 
mamiraua.org.br/downloads/) for analyses. 
RESULTS 
We logged 9,534 camera trap-days and 197 
records of Red-billed Curassows (Table 1). The 
species was recorded throughout the reserve at 15 
trap stations during the first year, nine during the 
second year, and eight during the last year 
(Fig. 1); 18 (36%) of the 50 trap stations did not 
capture this species. Most records during the first 
year were from the north region (n = 51), 
followed by the south (n = 21). and w-est (n = 
11) (x 2 = 33.0. df = 2. P < 0.005). 
