Bernardo et al. • BLACK-FRONTED PIPING GUAN DENSITY ESTIMATES 
693 
TABLE 2. Total effort (km), number of encounters, relative abundance (encounters/10 km) and mean density as 
individuals/km 2 and groups/km 2 (95% confidence interval) of Black-fronted Piping Guans in 11 Atlantic rainforest 
remnants in Sao Paulo State, Brazil (study areas correspond to Table 1). 
Study area 
Total number 
of transects 
Length of 
transects tm) 
Total km 
walked 
Number of 
encounter. 
Relative abundance 
(encounter,s/10 km) 
Density (Range) 
(imiividual/km 2 ) 
Density (Range) 
(groups/knr) 
Islands 
PEIC 
14 
700-4,000 
273 
10 
0.4 
3 (2.4—3.5) 
2 (1.6-2.3) 
PEI 
11 
900-5.500 
188 
29 
1.5 
13.3 (11-16.2) 
8 (6.7—9.8) 
Paranapiacaba Massif 
PECB 9 
1,200-5,400 
237 
17 
0.7 
4.4 (3.6—5.3) 
3.7 (3.1-4.5) 
PETAR 
7 
1,400-8.000 
256 
10 
0.4 
3 (2.5-3.7) 
2 (1.7-2.5) 
PEJa 
5 
700-4,500 
103 
I 
0.1 
0.5 (0.4-0.6) 
0.5 (0.4-0.6) 
PEJu 
6 
1,500-4,500 
222 
0 
Not found 
Not found 
Not found 
Serra do Mar Massif 
EEJI 5 
700-3,350 
190 
1 
0.1 
0.27 (0.23-0.33) 
0.27 (0.23-0.33) 
Car 
5 
700-3.000 
138 
1 
0.1 
0.4 (0.3-0.5) 
0.4 (0.3-0.5) 
Cun 
8 
1,050-8,450 
218 
1 
0.05 
0.5 (0.4-0.6) 
0.24 (0.20-0.29) 
SV 
3 
700-5,300 
210 
0 
Not found 
Not found 
Not found 
Pic 
3 
1,250-8,000 
210 
0 
Not found 
Not found 
Not found 
Data Analyses .—Distance data were analyzed 
using Program DISTANCE Version 5.1 (Thomas 
et al. 2006). Key function selection was evaluated 
using Akaike’s Information Criteria (AIC) and a 
Chi-square statistic was used to assess the 'good¬ 
ness of fit’ of each function (Buckland et al. 1993, 
2001). Repeated line transects increased sample 
sizes at each site, providing more precise estimates 
of variance and increasing the reliability of the 
detection function (Buckland et al. 2001, Lee and 
Marsden 2008). We had small sample sizes 
(between 1 and 29 encounters in each area. 
Table 2), and used a generalized effective strip 
width (ESW = 9.6 in, 95% confidence interval = 
7.9-11.7 m) obtained by pooling all observation 
data of Black-fronted Piping Guans. Pooling all 
records generated a more precise modeling of 
piping guan detection because the combined AIC 
values of the separate forest site detection functions 
were greater than the AIC value for the pooled 
detection function (Buckland et al. 200J). This 
procedure allowed us to calculate a single detection 
curve (half-normal key distribution with cosine 
adjustments) and provided a more reliable estimate 
of species density (individuals/km 2 ) (Buckland et 
al. 1993). Encounter rates (number of encounters/ 
10-km transect) were calculated for all sites as an 
estimate of relative abundance. 
RESULTS 
We recorded 70 observations of Black-fronted 
Piping Guans across the 11 study sites from 2001 
to 2007 during 2,246 km of line transects. All 
individuals were recorded in trees at heights 
between 5 to 10 m and, as rare events, some 
birds were observed 20 m above the ground, 
frequently on isolated heart-of-palm trees (Euter¬ 
pe edulis). The birds were recorded close to each 
transect (95% confidence limits = 7.9-11.7 m), as 
shown by the effective strip width value (mean = 
9.6 m, CV = 10%). Detection probability 
decreased with increasing distance (Fig. 2), 
mainly because the dense understory and abun¬ 
dant vines made observations of the species 
difficult at longer distances. 
Most birds observed were either solitary (60% 
of all records) or in pairs (33% of all records) and 
were only observed in larger groups (range = 3-5 
individuals) in 7% of the occasions. Pairs were 
seen <5 m from each other over all years. The 
majority of visual records (51%) occurred during 
the morning (28% of the records between 1000 
and 1200 hrs and 23% between 0800 and 1000 
hrs). 
The number of observations recorded and 
density estimates varied among sites (Table 2). 
The low overall mean density estimate (1.6 
individuals/km 2 . range = 1.2-2.2 individuals/ 
km 2 ) demonstrated how difficult it was to record 
the species in the largest Brazilian Atlantic Forest 
remnants. At least four populations were found in 
Sao Paulo State: two in the inshore islands, a 
small population in Serra do Mar massif, and a 
larger one in Serra de Paranapiacaba massif 
