62 
THE WILSON JOURNAL OF ORNITHOLOGY • Voi 124. No. 1. March 2012 
TABLE 2. Habitat selection of Serra Finches in Serra do Cipo National Park, Brazil. The proportion expected was 
calculated from the relative proportion of each habitat available within the study area, and the proportion observed in 
relation to 318 locations obtained for 17 pairs. Habitat selection is indicated for differences <0.05 significance; positive{+) 
for expected proportions below the Bonferroni confidence interval, and negative (-) for expected proportions above 
the interval. 
Habilal 
Proportion expected 
Proportion observed P, 
Confidence interval 
Selection 
Candeial 
0.003 
0.053 
0.039 < P, < 0.068 
+ 
Dry grassland 
0.793 
0.019 
0.010 < P, < 0.028 
Wet grassland 
0.079 
0.025 
0.015 < P } < 0.352 
_ 
Rocky outcrop 
0.047 
0.366 
0.334 < P 4 < 0.397 
+ 
Dry scrubland 
0.016 
0.072 
0.055 < / J , < 0.089 
-f 
Wet scrubland 
0.038 
0.234 
0.207 < P f , < 0.262 
+ 
Riparian woodland 
0.022 
0.231 
0.204 < P 1 < 0.259 
+ 
Domingues and Rodrigues 2007), Shrike-like Tan- 
ager (Neothraupis fascia ta) (4.3 ha; Alves 1990), 
Henna-capped Foliage-gleaner (Hylocryptus rectir- 
ostris) (2.9 ± 1.4 ha; Faria el al. 2007), C.ray- 
hacked Tachuri (4.3 ± 1.2 ha; Hoffmann 2006), and 
Cipo Canastero (Astltenes luizae) (4.0 ± 2.6 ha; 
Freitas 2011). These studies all used similar 
methods, without radiotelemetry (except Freitas 
2011). The home range area of those species could 
be larger than reported (Anich et al. 2009). 
The Serra Finch appears to be similar to most 
tropical bird species, maintaining a long-term pair 
bond and a territory throughout (he year (Stuehbury 
and Morton 2001), at times making it difficult to 
discern the territory space of the total home range 
area (Lopes and Marini 2006). The territorial 
behavior was illustrated by the annual site fidelity 
associated with duet song performance, and mini¬ 
mum overlap of the used area visualized by the MCP. 
Habitat Selection. —Johnson (1980) proposed a 
natural order to the process of habitat selection 
according to the scale of observation. First-order 
selection identifies the physical or geographical 
distribution of a species which, for the Serra Finch, 
is the mountaintop habitat complex that occurs 
throughout eastern Brazil (Vasconcelos 2008). 
Second-order selection delineates the home- 
range area of an individual or social group. Serra 
Finches did not establish territories at random and 
second-order selection was visually demonstrated 
by the localization of the 17 Serra Finch territories 
closely associated with watercourses (Fig. 1). We 
also found that wet scrubland and riparian wood¬ 
land habitats delineated establishment of the 
N Jinn B u p\ rr <%) in relali ° n ,0 locations of lhe sampled Pairs of Serra Finches in Serra do Cip4 
National Park. Braz.l, and the habitat composition within each 95<7r kernel territory, area. 
Habilal 
% locations 
Candeial 16.28 
Dry grassland 6.98 
Wet grassland 0.00 
Rocky outcrop 0.00 
Dry scrubland 25.58 
Wet scrubland 18.60 
Riparian woodland 32.56 
% territory 
Candeial 3 34 
Dry grassland 38.77 
Wet grassland 130 
Rocky outcrop Q.QQ 
Dry scrubland 27.24 
Wet scrubland 19.56 
Riparian woodland 9.«o 
0.00 
0.00 
3.51 
8.77 
0.00 
50.88 
36.84 
0.00 
42.32 
10.54 
5.30 
0.00 
35.54 
6.30 
5.77 
3.85 
3.85 
78.85 
0.00 
7.69 
0.00 
1.83 
30.73 
6.41 
53.94 
0.00 
7.09 
0.00 
0.(K) 
0.00 
0.00 
75.00 
0.00 
9.09 
15.91 
0.00 
26.46 
1.41 
62.55 
0.00 
7.09 
2.49 
14.29 
0.00 
7.14 
28.57 
14.29 
21.43 
14.29 
1.03 
62.93 
10.98 
3.46 
4.44 
13.57 
3.60 
0.00 
0.00 
0.00 
69.23 
0.00 
15.38 
15.38 
0.00 
67.31 
5.47 
22.78 
0.00 
2.67 
1.78 
6.67 
0.00 
6.67 
33.33 
3.33 
23.33 
26.67 
0.72 
56.73 
8.24 
14.56 
0.61 
7.54 
11.60 
16.67 
0.00 
0.00 
16.67 
25.00 
16.67 
25.00 
1.27 
73.22 
4.74 
1.77 
4.88 
4.21 
9.92 
0.00 
0.00 
6.25 
25.00 
0.00 
37.50 
31.25 
0.00 
51.84 
18.95 
7.69 
0.00 
16.90 
4.62 
