138 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123, No. 1, March 2011 
10-m diameter circular plots and >50 cm in 20-m 
diameter circular plots. We compared these data 
to equivalent measurements from 87 points at 
200-m intervals along 17.5 km of trails in Bilsa 
that we used to survey for Aramides nests. We 
classified these 87 points as being in primary, 
altered, or secondary forest based on visual 
inspection and knowledge of land use history, 
and used a discriminant analysis to build a 
predictive model of group membership based on 
habitat characteristics. This model correctly 
assigned 85% of the 87 training points as primary 
(n = 41), selectively-logged (n = 23) or 
secondary (n = 23) forest. The two discriminant 
analysis functions were significant (Function 1, 
Wilks’ Lambda = 0.280, P < 0.001; Function 2, 
Wilks’ Lambda = 0.866, P = 0.008) and were 
subsequently used to classify the type of forest 
where Brown Wood Rail nests were found. 
We captured three adult Brown Wood Rails in 
mist nets between March 2007 and January 2009, 
and took morphological measurements and ap¬ 
plied three colored leg bands. We applied a 
lightweight radio transmitter (model PD-2; Holo- 
hil Systems, Carp, ON, Canada) using a Rappole 
harness (Rappole and Tipton 1991) to a breeding 
individual of unknown gender captured on 12 
March 2008. The 3.8-g radio weighed <1% of the 
bird's total body mass. We tracked the individual 
using a TR4 receiver and a RA-2AK “H” 
antenna (Telonics, Mesa, AZ, USA) until the 
radio battery failed in November 2008. We 
suspended radio-tracking during nesting to mini¬ 
mize disturbance. We obtained locations of the 
bird at 30-min intervals during each radio¬ 
tracking session and recorded UTM coordinates 
using a handheld GPS unit. We plotted these 
cooidinates using the Animal Movement exten¬ 
sion in ArcView GIS 3.2 (ESRI 2006) and 
describe home ranges as minimum convex 
polygons (MCP’s) (Mohr 1947), and 95 and 
50% fixed kernel isopleths using least-squares 
cross validation (Worton 1989, Seaman and 
Powell 1996). Means ± SD are provided for all 
measurements. 
RESULTS 
We found nine active nests and seven addition¬ 
al nests that had evidence of recent activity but 
which were not active when discovered. Nests 
were found in February (n = 3 ), March (n = 10 ), 
and April (/? = 3). Median clutch size was four 
eggs (mean = 3.7 ± 0.7, range = 2-4). Eggs were 
oval in shape and cream-colored with brown 
spotting at the ends; dimensions of one were 
4.7 X 3.5 cm (Fig. 1 A). At least two adults shared 
incubation duties with replacement triggered by 
sharp, cracking vocalizations by the arriving adult. 
At most only a single adult was banded at any given 
nest, and we could not confirm whether more than 
two birds incubated. Maximum incubation period 
observed was 19 days, which should be considered 
the minimum for this species because all active 
nests had full clutches with discovered. 
Two of nine active Wood Rail nests were 
apparently depredated, one was abandoned, and 
six successfully fledged four young each. Hatch¬ 
ing was synchronous (on the same day) and young 
left the nest within 24 hrs of hatching. Chicks 
hatched with eyes open and were brooded almost 
continuously until departing the nest; we observed 
no feeding while the chicks were still in the 
nest. Chick plumage was dark brown with light 
brown longitudinal streaking and highly cryptic 
(Fig. IB), similar to that described for other 
Rallidae. At least two adults continued to care 
for the young for up to 10 days of age. Young 
chicks stayed together and were twice observed 
among the roots of a palm (Iriartea deltoidea) 
with stilt roots but were cryptic, moved rapidly, 
and difficult to observe. 
Nests were open cups atop stumps of fallen 
trees (n = 5 cases; mean tree DBH = 31.1 ± 
16.2 cm, mean tree height = 1.5 ± 0.6 m), at the 
intersection of multiple trunks and/or lianas (n = 
3 cases; DBH = 5.03 ± 1.0 cm, height = 4.1 ± 
2.9 m), or in understory shrubs (n = 8 cases; DBH 
= 6.6 ± 3.6 cm, height = 2.8 ± 1.4 m). Average 
nest height was 1.8 ± 0.5 m (range = 1 .2-2.6) 
above the ground. Nests were constructed primar¬ 
ily of large, dead leaves (e.g., Araceae, Cecropia- 
ceae, Piperaceae, and ferns) and a few small 
pieces of dried vine, and were relatively bulk}' 
(exterior dimensions; 26.8 ± 8.3 X 28.2 ± 6.0 X 
12.3 ± 4.0 cm; interior dimensions: 12.0 ± 1-7 X 
20.0 ± 2.6 X 3.8 ± 1.0 cm). The interior was 
lined with a mixture of live and dead, smaller 
leaves (primarily Melastomataceae). Nests were 
constructed beneath leaves and ferns in low light 
environments, making them relatively cryptic 
despite their large size (Fig. 1C). 
Nests were found in forest areas where 
elevation averaged 551 ± 31 m asl (range = 
448—587), canopy height averaged 15.2 ± 6.5 m, 
densiometer measures of canopy openness aver¬ 
aged 14.1 ± 6.9%, and there were 3.13 ± 2.1 
