692 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123, No. 4. December 2011 
TABLE 1. Characteristics of study sites in Atlantic rainforest remnants of Sao Paulo State, Brazil. 
Study area 
Size" (ha) 
Park 
creation' 1 
(year) 
Connectivity 
Altitudinal 
range 
(m)* 
Study 
period 
Number 
of park 
rangers' 
Islands 
Parque Estadual Ilha do Cardoso (PEIC) 
11.100 
1962 
Island 
0-784 
2001-2004 
5 
Parque Estadual Ilhabela (PEI) 
27,025 
1977 
Island 
0-1,375 
2004-2005 
11 
Serra de Paranapiacaba Massif 
Parque Estadual Carlos Botelho (PECB) 
37,644 
1982 
Continuous 
50-975 
2004-2006 
12 
Parque Estadual Turistico Alto Ribeira (PETAR) 
35,712 
1958 
Continuous 
600-1,200 
2006-2007 
19 
Parque Estadual Jacupiranga (PEJa) 
150,000 
1969 
Isolated 
10—1.310 
2005-2006 
13 
Parque Estadual Jurupara (PEJu) 
26,250 
1988 
Isolated 
400-900 
2004 
7 
Serra do Mar Massif 
Estate Ecoldgica de Jureia-ltatins (EEJI) 
79,270 
1986 
Isolated 
0-1.369 
2001-2002/ 
2005-2006 
8 
Parque Estadual Serra do Mar - Caraguaiatuba (Car) 
13,770 
1977 
Continuous 
500-1,298 
2004-2006 
8 
Parque Esladual Serra do Mar - Santa Virginia (SV) 
13,000 
1989 
Continuous 
860-1,500 
2002-2003 
5 
Parque Estadual Serra do Mar - Pieinguaba (Pic) 
47,500 
1979 
Continuous 
0-1.670 
2002-2003 
4 
Parque Estadual Serra do Mar - Cunha (Cun) 
11,660 
1974 
Continuous 
1,100-1,820 
2002-2003 
5 
“ According to IF (20111 and park managers. 
METHODS 
Study Areas .—We surveyed nine large areas 
ranging from 111 to 1,500 km 2 of mainland 
remnant Atlantic rainforest (6 continuous and 3 
non-connected forests), and two forested inshore 
islands in Sao Paulo State, Brazil (Table 1). These 
legally protected areas comprise some of the few 
largest Atlantic rainforest remnants, including the 
Serra do Mar and Serra de Paranapiacaba 
mountain ranges (Table 1). The inshore island of 
Ilha do Cardoso is a land-bridge island and its 
closest point to the continent is 300 m. The 
oceanic inshore island of Ilhabela is 1.7 km from 
the nearest point of the continent. All areas were 
dominated by three different types of forest 
habitat, which vary according to the latitude and 
altitude: lowland rainforest, and submontane and 
lower montane rainforest (Oliveira-Filho and 
Fontes 2000). Mean annual temperatures of these 
areas vary between 17 and 25 C (IF 2011). 
All study areas have experienced palm-heart 
{Euterpe edulis) harvesting and illegal hunting 
(Galetti and Fernandez 1998, Aguiar et al. 2003. 
Olmos et al. 2004, Galetti et al. 2009). All areas 
are part of two Endemic Bird Areas (BirdLifc 
International 2003): Atlantic Forest lowlands and 
Atlantic Forest mountains. They are also consid¬ 
ered Important Bird Areas (IBA), except llha do 
Cardoso and Jurupara (Staitersfield et al. 1998). 
Bird Surveys .—We used a variable-distance 
line-transeel method (e.g., Buckland et al. 1993) 
to survey populations of Black-fronted Piping 
Guans following standardized methodology de¬ 
rived by Peres (1999) and used by Galetti et al. 
(2009). Between three and 14 transects were 
systematically positioned at each site to be 
representative of the surrounding habitat (Peres 
1999). Transect locations were selected mainly 
by vegetation type, elevation and distance from 
rivers and roads, and the variation in these 
habitat features were well represented in our 
study areas. Transect length varied from 0.7 to 
8.4 km, depending upon local topography and 
forest patch size (Galetti et al. 2009) (Table 2). 
Ten observers, who were fully trained in distance 
sampling techniques and cracid identification, 
were responsible for conducting the bird surveys 
al all sites during 2001 to 2007. Transects were 
surveyed systematically during 0615 and 1730 
hrs by one or two observers, once a month at 
each site, with direction of travel along each 
transect rotated between subsequent surveys to 
minimize the effect of time of day. W'e recorded 
the date. time, global positioning system (GPS) 
co-ordinates, number of individuals seen and 
perpendicular distance to the line transect 
(accurately obtained with a measuring tape) for 
each piping guan detected. Survey effort at each 
site ranged from 103.1 to 273 km (Table 2). 
which is greater than the sampling effort required 
for reliable abundance estimates of large birds 
(Thoisy et al. 2008). 
