20 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123, No. 1, March 2011 
aggressively to playback, but were not associated 
with a known territory. Three (2 juveniles) of the 
floater males (4 juveniles, 3 adults) were seen 
again elsewhere on the plot in subsequent years 
with females on known territories. One juvenile 
male traveled through the focal plot with an 
unbanded female in 2006 and attempted to 
establish a territory in the center of the study 
area, but both birds disappeared within 1 week. 
This “territory” had not been previously occu¬ 
pied, and was not occupied thereafter. 
DISCUSSION 
Chestnut-backed Antbirds, once on a territory, 
exhibit restricted movement. The few individuals 
that switched territories typically moved to an 
adjacent territory, and all movements observed 
were <850 m. Mean life span of adult Chestnut- 
backed Antbirds at La Selva is —4.24 years (mean 
life span = l/-ln(s), where 5 = annual survival; 
Brownie et al. 1985). Thus, breeding dispersal, 
even over the course of a lifetime, may be 
insufficient to maintain direct demographic or 
genetic connectivity between populations separat¬ 
ed by more than a few kilometers, especially if 
barriers (e.g., large rivers, habitat discontinuities) 
are present. 
Apparent Annual Survival and Detection Prob¬ 
ability .—Our estimates of (p for adult Chestnut- 
backed Antbirds are at the high end of the range of 
estimates for other thamnophilids summarized by 
Blake and Loiselle (2008) (range = 0.45-0.87). 
We found no evidence for differential apparent 
survival between adult males and females, and 
high detection probabilities suggest that our 
estimates of (p are close to true survival, at least 
for adult males (Jones et al. 2004, Ruiz-Gutierrez 
et al. 2008). Apparent survival of all juveniles was 
poorly estimated largely due to small sample size, 
as juveniles by definition only contribute to a 
single capture period, and afterwards contribute to 
adult survival estimates. However, p for juvenile 
males was relatively high, and the large confi¬ 
dence intervals around estimates of juvenile male 
cp can be interpreted as evidence of greater 
variability in apparent survival. 
Detection probabilities (p ) of males and fe¬ 
males were strikingly different with females 
having lower and more variable values. The 
unequal detection rates of age and male or female 
classes highlight ecologically important differenc¬ 
es in the behavior of individuals (e.g., Crespin et 
al. 2008), and also have methodological implica¬ 
tions. For example, in sexually monomorphic 
species, or those with delayed male plumage 
maturation (e.g., some Pipridae; DuVal 2005), 
important differences in apparent survival among 
groups may go undetected, and lead to high 
variance of parameter estimates. 
Turnover Rate and Territory Switching.—Over¬ 
all rate of turnover (disappearing + switching) by 
Chestnut-backed Antbirds at La Selva was 
comparable to other antbird species. Turnover 
was higher for females, and turnover among 
female Chestnut-backed Antbirds was the highest 
reported for any species (Table 3). Territory 
switching by Chestnut-backed Antbirds was 
markedly less common than for several other 
species studied (e.g.. White-bellied [Myrmeciza 
longipes] and Dusky [Cercomacra tyrannina ] 
antbirds). The few territory switches of Chest- 
nut-backed Antbirds observed generally were on 
the scale of 1-2 territory-widths, and all switches 
involved distances of <1 km, consistent with 
patterns in other small tropical understory resident 
insectivores (Greenberg and Gradwohl 1997, 
Morton et al. 2000, Robinson 2000, Gill and 
Stutchbury 2010). Low switching rates and short 
breeding dispersal distances have been reported in 
a wide variety of bird species, both tropical and 
temperate, and may be a general characteristic of 
monogamous species with year-round, multipur¬ 
pose territories (e.g., Woolfenden and Fitzpatrick 
1989, Bried and Jouventin 1998, Komdeur and 
Edelaar 2001, Thorstrom et al. 2001, VanderWerf 
2004, Gill and Stutchbury 2006, Eikenaar et al. 
2008, Gill and Stutchbury 2010). 
The abundance of floaters ( sensu Winker 1998) 
is difficult to quantify, especially for females. 
Most territorial vacancies (male and female) were 
filled from year to year, mainly by unbanded 
birds, and both adults and juveniles were 
represented among the replacements. However, 
not all vacancies were filled, and some territories 
remained vacant at the end of our study. How 
floaters may affect long-term population trends or 
estimates of dispersal patterns is not clear, but 
should be considered in future studies (Zack and 
Stutchbury 1992, Winker 1998). 
CONSERVATION IMPLICATIONS 
Our findings have important implications for 
conservation of tropical understory forest birds in 
fragmented landscapes. Rare and small-scale 
lifetime breeding dispersal by Chestnut-backed 
Antbirds in a contiguous forest setting agrees with 
