Undell et al. • NEST SUCCESS OF NEOTROPICAL THRUSHES 
503 
investigate influences besides land-cover type on 
nest success. We documented nest site character¬ 
istics that made nesting attempts more or less 
likely to fail by identifying sources of variation in 
nest success (e.g., Aguilar et al. 2008). 
METHODS 
Study Sites.— Nests of White-throated and 
Clay-colored thrushes were located from 1909 to 
2002 and monitored throughout Las Alturas. a 
private forest reserve (08 c 56' N, 82 51' W) at 
-1,300 m asl in southern Costa Rica. Systematic 
nest searching was conducted in 1999 in two 
abandoned coffee plots, each of 6 ha. in large 
fields. We expanded the two coffee plots in 2000 
to 10.3 and 13.3 ha. We also added two pasture 
plots, one of 17.9 ha and one of 13,2 ha. We 
searched the four plots from 2000 to 2002. Coffee 
plots were in plantations that had been abandoned 
in the mid-1990s and consisted of 1 to 3-m tall 
coffee plants with a light overstory consisting 
mainly of 5 to 7-m tall Erthyrina spp. The 
understory in the plantations was variable al¬ 
though it tended to be sparse between the coffee 
plants, i.e.. there was little vegetation to impede 
the movements of nest searchers. Pasture plots 
were in active fields used by dairy and beef cattle 
that had a few scattered tall (> 8 m) trees and 
shorter woody vegetation. Both coffee and pasture 
plots were adjacent to, or within 200 m of old 
growth forest. We also monitored nests found 
outside of plots. 
We also searched an active coffee plantation, 
Rio Negro, in 2001, 12 km from Las Alturas (08 
52' N, 82 c 51’ W) at -1,100 m asl. The coffee 
plants at Rio Negro were cultivated and tended to 
he shorter (usually < 2 m) than those at Las 
Alturas. The Erythrina overstory was also shorter 
iusually < 5 m), because of active pruning. 
Nest Searching and Monitoring. —Nest search¬ 
ing started in late March-early April and continued 
through August from 1999 to 2002, coinciding with 
•he nesting seasons of White-throated and Clay- 
colored thrushes. Once found, nests were checked 
eveiy 2-4 days. Nests were not approached when 
potential predators were present. We only re¬ 
mained near nests for sufficient lime to ascertain 
•heir contents. A nest was considered to have failed 
when nestlings were missing before the age at 
which they could have fledged. Successful nests 
were those at which nestlings could have reached 
the age of fledging and other evidence ot fledging 
was present including a flattened nest rim, feces on 
the rim, or fledglings in the vicinity of the nest. 
Nests were considered to have uncertain fate when 
nestlings could have reached the age of fledging 
but there was no other evidence of a successful 
nesting attempt. 
Nest Site Characteristics.—We investigated 
characteristics that were likely to be important 
in our system including nest concealment, nest 
height, nest substrate (ground vs. vegetation), nest 
initiation interval, and year. Our expectations 
were that greater concealment (e.g., Martin 1993) 
would positively influence the likelihood of nest 
success. We also expected greater success for 
nests on the ground compared to those in 
vegetation (e.g., Yaliner and Scott 1988) and for 
nests at lower heights. These expectations were 
based partly on findings from other studies (e.g., 
Soderstrom et al. 1998) and partly from our 
previous work indicating that toucans and raptors 
arc important predators in this system (Cohen and 
l.indcll 2004; C. A. Lindell, pers. obs.) and might 
be more likely to detect high than low nests. We 
also expected greater success for nests initiated 
earlier in the season (e.g., Nur et al. 2004). Nest 
height and concealment were measured for each 
nesting attempt. 
Nest height (m) was measured trom the ground 
to the bottom of the nest. Nests built on the 
ground or in a bank had a height of zero. Nests 
were grouped into three height intervals: low - 
0-1.5 m, medium — 1 .6-2.4 m, and high = 2.5- 
3.5 m. Nests >3.5 m could not be checked and 
were not monitored. Concealment was catego¬ 
rized on a 1 to 4 scale (1 = <25% concealment, 2 
- 26-50%, 3 = 51-75%, and 4 - 76-100%). 
Concealment was estimated by an observer 1 m 
from the north, south, east, and west sides of the 
nest. Concealment estimates were also made from 
1 m above and below nests when possible. Nests 
on the ground were given a concealment estimate 
of 4 from below. Wc calculated a mean of the six 
readings for each nest to generate one value 
between one and four for each nest. These values 
were divided into three categories; low conceal¬ 
ment (0-1.99). medium concealment (2-2.99), 
and high concealment (3-3.99). 
Data Analysis. —White-throated and Clay-col¬ 
ored thrush nests were found in pasture, coffee, 
forest fragments, and along roads at Las Alturas. 
Sample sizes were small for both species in forest 
fragments and for Clay-colored Thrush nests 
along roads; nests from these land-cover types 
were excluded from analyses. 
