Volume 135, 2018, pp. 15–24
DOI: 10.1642/AUK-17-105.1
RESEARCH ARTICLE
Environmental predictors of nestling condition, postfledging movement,
and postfledging survival in a migratory songbird, the Wood Thrush
(Hylocichla mustelina)
Ben J. Vernasco,1 T. Scott Sillett,2 Peter P. Marra,2 and T. Brandt Ryder2*
1 Department of Biological Sciences, Virginia Tech, Derring Hall, Blacksburg, Virginia, USA
2 Migratory Bird Center, Smithsonian Conservation Biology Institute, Washington, D.C., USA
* Corresponding author: rydert@si.edu
Submitted June 12, 2017; Accepted August 8, 2017; Published October 18, 2017
ABSTRACT
Given that population dynamics of birds are known to be sensitive to high fledgling mortality, a comprehensive
understanding of the environmental factors that drive variation in fledgling survival is essential to avian conservation.
We quantified multiple aspects of the Wood Thrush postfledging period using breeding and radio-telemetry data
collected over the course of 4 breeding seasons in southern Indiana, USA. First, we examined how drought, forest
cover, and brood parasitism affected nestling body condition and brood size. Second, after controlling for the age-
specific increase in survival, we used multimodel inference to examine how brood parasitism, drought, forest cover,
nestling body condition, and nest vegetation structure influenced postfledging survival. Finally, we measured the
relationship between these covariates and fledgling movements. Drought decreased cumulative survival probabilities,
with the youngest age group (,4 days postfledging) being most affected; however, this relationship was dependent
on the amount of mature forest cover. During non-drought years, fledgling survival was lower in study plots with a
high proportion of mature forest cover. By contrast, postfledging survival during a drought year was higher in study
plots with a high proportion of forest cover. This drought year was also associated with lower nestling body condition
and brood size, and with delayed postfledging dispersal from natal territories. Our results suggest that while Wood
Thrush postfledging survival is relatively insensitive to both nest vegetation structure and brood parasitism, forest
cover and breeding-season precipitation interact to affect multiple aspects of the species’ postfledging period.
Keywords: brood parasitism, dispersal, drought, Hylocichla, postfledging survival, Wood Thrush
Predictores ambientales de la condicio´n del polluelo, del movimiento post emplumamiento y de la
supervivencia post emplumamiento en Hylocichla mustelina, un ave canora migratoria
RESUMEN
Dado que las dina´micas poblacionales de las aves son conocidas por ser sensibles a la alta mortalidad de los
volantones, es esencial lograr un entendimiento exhaustivo de los factores ambientales que determinan la
supervivencia de los volantones para la conservacio´n de las aves. En este estudio, cuantificamos mu´ltiples aspectos del
perı´odo post emplumamiento de Hylocichla mustelina usando datos de crı´a y de radio telemetrı´a colectados a lo largo
de 4 estaciones reproductivas en el sur de Indiana. Primero, examinamos como la sequı´a, la cobertura del bosque y el
parasitismo de nidada afectan la condicio´n corporal de los polluelos y el taman˜o de la nidada. Segundo, luego de
controlar por el incremento especı´fico de la supervivencia en relacio´n a la edad, usamos inferencia de modelos
mu´ltiples para examinar como el parasitismo de nidada, la sequı´a, la cobertura del bosque, la condicio´n corporal de los
polluelos y la estructura de la vegetacio´n del nido influencian la supervivencia post emplumamiento. Finalmente,
medimos la relacio´n entre estas covariables y los movimientos de los volantones. La sequı´a disminuyo´ las
probabilidades acumuladas de supervivencia, siendo el ma´s afectado el grupo de edad ma´s joven (,4 dı´as post
emplumamiento); sin embargo, esta relacio´n dependio´ de la cantidad de cobertura de bosque maduro. Durante los
an˜os sin sequı´a, la supervivencia de los volantones fue ma´s baja en las parcelas de estudio con una alta proporcio´n de
cobertura de bosque maduro. En contraste, la supervivencia post emplumamiento durante un an˜o de sequı´a fue ma´s
alta en las parcelas de estudio con una alta proporcio´n de cobertura del bosque. Este an˜o de sequı´a estuvo tambie´n
asociado con una menor condicio´n corporal del polluelo y del taman˜o de la nidada, y con una dispersio´n atrasada post
emplumamiento desde los territorios de nacimiento. Nuestros resultados sugieren que mientras la supervivencia post
emplumamiento de H. mustelina es relativamente insensible a la estructura de la vegetacio´n y al parasitismo de
nidada, la cobertura del bosque y la precipitacio´n de la estacio´n reproductiva interactu´an para afectar mu´ltiples
aspectos del perı´odo post emplumamiento de la especie.
Palabras clave: dispersio´n, Hylocichla mustelina, parasitismo de nidada, sequı´a, supervivencia post emplumamiento
Q 2018 American Ornithological Society. ISSN 0004-8038, electronic ISSN 1938-4254
Direct all requests to reproduce journal content to the AOS Publications Office at pubs@americanornithology.org
INTRODUCTION
The postfledging period in passerines is characterized by
high mortality and affects natal dispersal, immigration,
and, ultimately, recruitment into the breeding population
(Naef-Daenzer et al. 2001, Yackel Adams et al. 2006,
Schaub and Abadi 2011, Robinson et al. 2014, Naef-
Daenzer and Gru¨ebler 2016). Investigations of this phase
of the annual cycle have demonstrated that mortality is
concentrated within the first week after fledging, decreases
nonlinearly with fledgling age, can be negatively correlated
with body condition prior to fledging, and tends to
increase over the breeding season (reviewed by Newton
1998, Cox et al. 2014). We know considerably less about
how environmental factors affect fledging condition as well
as postfledging movement and survival.
Environmental factors such as brood parasitism,
drought, forest cover, habitat structure, and body condition
can influence the postfledging period (Cox et al. 2014).
While brood parasitism reduces nest success in many
passerine species (Payne and Payne 1998), its effect on
postfledging survival remains ambiguous, with both
negative and neutral effects reported (Rasmussen and
Sealy 2006, Vitz and Rodewald 2011, Peterson et al. 2012,
Jenkins and Faaborg 2016). Drought, however, has been
shown to lower postfledging survival directly via changes
in resource availability and indirectly via changes in
vegetation structure (Yackel Adams et al. 2006). Moreover,
postfledging survival has been found to be higher in
habitats with dense vegetation structure and greater food
abundance (e.g., successional, riparian, or forest edge
habitat; Fink 2003, White et al. 2005, King et al. 2006, Vitz
and Rodewald 2011). These habitat relationships suggest
that fledglings experience lower survival in areas with a
high proportion of mature forest cover that lack favorable
habitat, but formal tests of how forest cover influences
postfledging survival are rare (but see Rush and Stutchbury
2008, Moore et al. 2010). Furthermore, the effects of body
condition on survival remain unclear, with some studies
reporting a positive relationship between nestling condi-
tion and survival (Cox et al. 2014) and others suggesting
that the hazard type (e.g., predation or exposure)
determines the importance of body condition for post-
fledging survival (Jones et al. 2017).
One of the most studied passerines during the
postfledging period is the Wood Thrush (Hylocichla
mustelina), a Neotropical migrant that has suffered loss
and degradation of breeding habitat and experienced long-
term population declines (Evans et al. 2011, Rushing et al.
2016). The species’ reproductive ecology has been well
documented, and its postfledging period lasts ~21 days
(Anders et al. 1997, Evans et al. 2011). Fledgling movement
is initially restricted to the natal territory and may be
constrained by body condition at fledging, by parental
behavior, and by environmental variation. Independent
juveniles disperse to habitats with dense understories (e.g.,
early- to mid-successional or riparian forests) that provide
both abundant food and protection from predators
(Anders et al. 1998, Fink 2003). Previous studies have
focused on postfledging survival patterns (Anders et al.
1997, Schmidt et al. 2008) but were not designed to test
how environmental factors affect nestling body condition
or how body condition and the environment affect
survival.
Here, we investigate how environmental factors and
body condition influence the survival probability of
fledgling Wood Thrushes. We also examine how brood
parasitism, forest cover, and drought affect nestling
condition, brood size, and fledgling movement. We used
radio telemetry to track fledglings at 12 heterogeneous
forest sites across 4 yr in southern Indiana, USA. After
accounting for a nonlinear increase in survival with
fledgling age (Cox et al. 2014), we built known-fate
survival models to test how brood parasitism, forest cover,
drought, habitat structure, and nestling body condition
affected postfledging survival. On the basis of previous
research, we expected that postfledging survival would be
(1) lower in nests parasitized by Brown-headed Cowbirds
(Molothrus ater), (2) lower during drought conditions, (3)
lower in areas with a high proportion of forest cover, (4)
higher in areas with greater understory vegetation density,
and (5) higher when nestlings were in better body
condition. Our objective was to elucidate the relationship
between environmental predictors and different aspects of
the postfledging period (e.g., nest conditions, movement,
and survival). With a better understanding of this critical
life stage, conservation efforts can more effectively
improve breeding productivity, particularly for declining
species that are strongly limited by breeding-season events
(Cox et al. 2014).
METHODS
Study Area
We tracked fledgling Wood Thrushes on twelve 40–60 ha
plots in southern Indiana in 2011–2014. Plots were located
on Crane naval base (Crane; 38.838N, 86.798W), in Big
Oaks National Wildlife Refuge (BONWR; 38.958N,
85.48W), and within parks operated by the Indiana
Department of Natural Resources in the greater Bloom-
ington area (IDNR; 39.168N, 86.538W). Plots were chosen
to represent a continuum of habitat quality, ranging from
small fragments of secondary growth to continuous old-
growth deciduous forest (see Appendix Table 4). Common
tree species included white ash (Fraxinus americana),
American beech (Fagus grandifolia), flowering dogwood
(Cornus florida), boxelder (Acer negundo), American elm
(Ulmus americana), hickory (Carya spp.), American
The Auk: Ornithological Advances 135:15–24, Q 2018 American Ornithological Society
16 Wood Thrush postfledging survival B. J. Vernasco, T. S. Sillett, P. P. Marra, and T. B. Ryder
hornbeam (Carpinus caroliniana), sugar maple (A. sac-
charum), oak (Quercus spp.), eastern redbud (Cercis
canadensis), American sycamore (Platanus occidentalis),
and tuliptree (Liriodendron tulipifera). The understory was
dominated by spicebush (Lindera benzoin), serviceberry
(Amelanchier arborea), elderberry (Sambucus canadensis),
and multiflora rose (Rosa multiflora).
A severe regional drought affected our study sites during
June and July 2012 (Mallya et al. 2013). Cumulative
precipitation from April to August for 2011, 2013, and
2014 ranged from 24.67 to 27.54 inches (http://www.ncdc.
noaa.gov/cdo-web/; retrieved July 2015). However, cumu-
lative precipitation between April and August 2012 was
substantially lower at 13.33 inches, with consistent deficits
across these months.
Field Sampling
We located Wood Thrush nests using behavioral cues (e.g.,
nest building or provisioning) and systematic searching.
For each nest, we documented stage (incubating, nestlings,
etc.) and the presence of nestling cowbirds and recorded
location to ~10 m accuracy with handheld GPS units.
Nests were monitored every ~4 days until fledging or
failure, following a protocol similar to that described in
Martin and Geupel (1993). All nestling Wood Thrushes
were banded during days 8–12 after hatching with a
numbered federal band and colored leg bands. We
measured tarsus length, body mass, and wing chord. One
randomly selected individual per brood received a VHF
radio transmitter (PipAg 392, 1 g, ~3% of nestling body
mass; Lotek Wireless, Newmarket, Ontario, Canada)
affixed with a leg-loop harness (Rappole and Tipton
1991). We assumed that this randomly selected individual
was representative of the entire brood, given that Wood
Thrushes are not known to divide broods postfledging
(Vega Rivera et al. 2000). The radio-tagged individual
always received a unique color-band combination to
facilitate resighting during tracking sessions. Nest visits
continued until fledging.
We searched for tagged fledglings once every 4 days
using Biotracker programmable receivers and handheld, 3-
element Yagi antennas (Lotek Wireless). Upon detecting a
fledgling, we recorded its location with a handheld GPS
unit and noted the presence of adults or other fledglings.
Precaution was taken to never flush fledglings during
tracking, and when a visual was not possible, we projected
GPS waypoints to the estimated location. Fledglings were
tracked until dispersal from study site (~21 days),
predation, transmitter failure, or the end of the study
(early to mid-August). A bird was considered depredated if
remains (e.g., feather, tarsi, or color bands) or a damaged
transmitter (e.g., bent antenna or broken harness) was
found. When a signal could not be obtained for a bird that
was too young to leave the natal territory (based on
previous published estimates of age at independence), the
area was searched for adults and/or fledglings. We
assumed predation if we could not detect any signs of a
family group (e.g., alarming calling by adults and/or
fledglings begging) and resight the fledgling. Individuals
that prematurely lost their tags (i.e. a tag was found with
no obvious signs of remains, transmitter damage, and an
intact harness) or had tag malfunctions were censored
from the dataset at the estimated date of tag failure.
Vegetation Sampling and Forest Cover Calculation
We sampled vegetation composition within a circular plot
(radius ¼ 5 m) centered on each Wood Thrush nest.
Vegetation metrics within each plot were (1) frequency of
trees in different size classes and (2) vertical foliage density.
The frequency of tree size classes was measured with a
forester’s prism, and each tree was classified into 4 size
classes (2 to ,5 cm, 5 to ,20 cm, 20 to ,30 cm, and 30
cm) using a tree DBH (diameter at breast height) tape.
Vertical foliage measurements were taken at each cardinal
direction on the plot edge, using a 5.5 m extendable pole.
The extended pole was subdivided into 0.25 m increments
from 0 to 3 m, and observers recorded whether or not any
live woody vegetation (including stems, leaves, needles,
branches, and tree trunks) contacted the pole in each
increment.
To measure forest cover, we centered a circle (radius¼ 2
km, total area ¼ 1,257 ha) on each study plot. This 2 km
radius encompasses likely dispersal distances and home-
range sizes for fledgling Wood Thrushes (Anders et al.
1998, Vega Riveria et al. 1998). Within each circle, we
quantified the proportion of area classified as mature
forest using the National Land Cover Database (NLCD; Fry
et al. 2011).
Data on nestling condition, brood size, nest parasitism
rates, vertical foliage density, and number of trees .30 cm
DBH are summarized in Appendix Table 5.
Data Analysis
We used the RMark package (Laake 2013) in R (R Core
Team 2015) to build known-fate models of postfledging
survival. Known-fate models estimate daily survival
probabilities given a detection probability of 1 and have
the ability to examine the effect of covariates on survival
(White and Burnham 1999). For our analysis, we used a 2-
stage modeling approach in which the first model set
accounted for the age-specific increase in postfledging
survival and the second examined the relationship between
4 predictors and postfledging survival. For our second
model set, we used the top model from our first model set
as the base model to control for the temporal structure in
fledgling survival. Candidate model sets were symmetrical
with respect to all predictors (Doherty et al. 2012). We
ranked models using Akaike’s Information Criterion
The Auk: Ornithological Advances 135:15–24, Q 2018 American Ornithological Society
B. J. Vernasco, T. S. Sillett, P. P. Marra, and T. B. Ryder Wood Thrush postfledging survival 17
corrected for small sample size (AICc) and determined the
relative likelihood of each model using model weights (wi;
Burnham and Anderson 2002). Parameters were consid-
ered informative based on the weight of evidence (i.e. Rwi
from relevant candidate models) and if the confidence
interval of the estimate did not overlap the origin (Arnold
2010). Model results are given as standardized beta
parameter estimates and 95% confidence intervals.
Our first model set included 7 different parameteriza-
tions of survival probability: (1) constant, (2) linearly
increasing with fledgling age (Age), (3) nonlinearly
increasing with fledgling age (Age2), (4) fully time
dependent (time), and (5–7) three models with different
age-specific survival periods. For each of these parameter-
izations, a duplicate model was made that also incorpo-
rated seasonal changes (represented by fledge date) in
fledgling survival. The number of survival periods and the
number of days within each survival period in a given
model were based on the survival estimates of other
postfledging survival studies (e.g., Ausprey and Rodewald
2011) and on the developmental stages of fledgling Wood
Thrushes (Anders et al. 1998, Evans et al. 2011). Survival
periods included either 2 periods (days 1–7 and 8–30), 3
periods (days 1–7, 8–21, 22–30), or 4 periods (days 1–3,
4–7, 8–21, 22–30).
The second model set used the 4-age-class parameter-
ization from the first set in all candidate models and
considered 6 additional variables that we predicted would
affect survival probability: (1) presence or absence of brood
parasites (binary), (2) vertical foliage density surrounding
the nest, (3) number of trees .30 cm in DBH, (4) amount
of mature forest cover within a circle (radius ¼ 2 km)
around each study plot, (5) a dummy variable representing
a drought year, and (6) nestling body condition prior to
fledging. We considered models with 1 or 2 predictors in
addition to age-dependent effects on survival. Candidate
models included both additive and factorial combinations
of predictors. A nest was considered to be parasitized if a
nestling cowbird was detected during any nest check. We
included vertical foliage density around the nest because
shrub density has previously been shown to be an
important predictor of postfledging survival (Fink 2003,
Cox et al. 2014).We also included the number of trees.30
cm in DBH because this habitat feature also predictsWood
Thrush fecundity (Rushing et al. 2017). Furthermore, we
used a dummy variable (rather than monthly or weekly
values) to represent the drought year because of the
consistent monthly deficits across the 2012 breeding
season (Mallya et al. 2013). We calculated Peig and Green’s
(2009) scaled mass index using mass and tarsus length
measured shortly before fledging. Nestlings were weighed
at different stages of development, but allometric growth
patterns showed that tarsus length, an index of structural
size, was positively correlated with age (r1368 ¼ 0.69, P ,
0.001). This mass index controlled for age-based size
variation.
To further understand the effects of brood parasitism,
drought, and forest cover on the postfledging period, we
also tested the effects of these covariates on nestling brood
size, nestling condition, and postfledging movement. We
compared nestling condition (among all nestlings banded)
and brood size between parasitized and nonparasitized
nests and across drought and non-drought years using
Student’s t-test. We also tested the effect of forest cover on
nestling condition and brood size using linear regression.
For our movement analysis, we calculated the Euclidean
distance between each fledgling detection and its nest. We
did not quantify daily movements because our tracking
schedule confounds estimates of an individual’s daily
movements (i.e. birds were tracked once every 3–4 days).
We used AICc to compare linear mixed models with
additive effects of each variable (condition, drought, forest
cover, and brood parasitism) and age class (as defined in
our first model set), an interactive effect of each variable
and age class, a model with just age class, and a null model.
All covariates were modeled as fixed effects, and individual
was included as a random effect. We log-transformed the
response, average distance from nest, to meet model
assumptions. These statistical analyses were performed
using R and the ‘‘lme4’’ package (Bates et al. 2015).
RESULTS
Of the 210 fledgling Wood Thrushes tracked from 2011 to
2014 (2011, n ¼ 47; 2012, n ¼ 53; 2013, n ¼ 66; 2014, n ¼
44), 110 individuals survived for 21 days, the average age of
independence. Radio-tagged individuals were tracked from
1 to 48 days (mean ¼ 14.39 6 10.1). Among the 61
individuals with known mortalities, 27 were due to
predation, 5 were due to exposure, and 29 died of
unknown causes. Of the 39 censored individuals, 8
individuals were due to tag failure, 28 had unknown fates,
and 3 were tracked until the end of the field season.
Our first model set revealed strong statistical support for
4 age-specific survival periods (0–3, 4–7, 8–20, and 21
days since fledging;
P
wi¼0.84; Table 1). Fledgling survival
probability tended to decline with ordinal day, but this
relationship had weak statistical support compared to
constant survival over the breeding season (Table 1).
Therefore, we used 4 age classes and constant survival over
the breeding season as the base model parameterization in
the second candidate set.
The top model of our second model set included an
interaction between drought and the amount of forest at
the landscape scale (bdroughtxforest¼ 0.003, 95% CI: 2.39e-4
to 6.97e-3), with drought effects receiving more support
(
P
wi ¼ 0.63) than forest cover (
P
wi ¼ 0.4). Cumulative
survival (i.e. the probability of surviving to 30 days after
The Auk: Ornithological Advances 135:15–24, Q 2018 American Ornithological Society
18 Wood Thrush postfledging survival B. J. Vernasco, T. S. Sillett, P. P. Marra, and T. B. Ryder
fledging) was lower in 2012, the drought year (0.53, 95%
CI: 0.35–0.6; bdrought ¼ 4.24, 95% CI: 7.69 to 0.8),
compared to the cumulative survival probability for 2011,
2013, and 2014 (0.69, 95% CI: 0.58–0.78; Figure 1). The
relationship between forest cover and fledgling survival
was weak (bforest ¼0.002, 95% CI: 9.5e-05 to 0.004),
with birds in less forested landscapes having higher
survival. Decreases in survival during the drought year
were most pronounced among fledglings 0–3 days old;
and, during drought years, birds from study plots with
lower forest cover showed the steepest decline in survival.
By contrast, fledglings 0–3 days old in more forested
landscapes had roughly equivalent survival during drought
and non-drought years (Figure 1). Furthermore, brood
parasitism and nestling condition were not strongly related
to the survival of fledgling Wood Thrushes on our study
sites (Table 2). Although the second-ranked model
suggests support for an effect of vertical foliage density,
the beta confidence intervals include zero (0.08, 95% CI:
0.16 to 0.00). Similarly, we found no evidence that the
number of trees .30 cm DBH near nest sites affected
postfledging survival.
Nestling condition and brood size were both influenced
by drought and brood parasitism, but not by forest cover.
Specifically, nestlings were in significantly poorer body
condition during the drought year compared to the 3 non-
drought years (ldrought¼ 26.52 6 0.34 g [n¼ 263 nestlings,
n ¼ 112 nests], lnon-drought ¼ 28.88 6 0.11 g [n ¼ 1,116
nestlings, n¼ 431 nests], t315¼8.81, P , 0.001; Figure 2),
and nestlings from parasitized nests had higher body-
condition indices (lparasitized¼ 29.15 6 0.24 g, lnonparasitized
TABLE 1. Akaike’s Information Criterion (AIC) rankings of our 10
best-fitting known-fate models examining the temporal struc-
ture of fledgling Wood Thrush survival on our study sites in
southern Indiana, USA. Daily survivorship rates change between
2 (~AgeClass2), 3 (~AgeClass3), and 4 (~Ageclass4) survival
periods, remain constant (~1), vary daily (~time), are a linear
(Age) or quadratic function (AgeþAge2), and depend on fledge
date (FD).
Rank Model K a DAICc
b wi
c Deviance
1 ~AgeClass4 4 0 0.59 89.99
2 ~AgeClass4 þ FD 5 1.89 0.23 557.64
3 ~Age þ Age2 3 4.12 0.08 96.12
4 ~(AgeþAge2) þ FD 4 6.03 0.03 563.79
5 ~Age 2 6.54 0.02 100.54
6 ~AgeClass4*FD 8 7.85 0.01 557.57
7 ~Age þ FD 3 8.45 0.01 568.21
8 ~(Age þ Age2)*FD 6 8.88 0.01 562.62
9 ~Age*FD 4 10.43 0 568.19
10 ~AgeClass2 2 11.09 0 105.09
a Number of parameters in each model.
b Difference in AIC score (corrected for small sample size, AICc)
between each model and the best-fitting model (AICc¼ 565.77
for the top-ranked model).
c Model weight.
FIGURE 1. Wood Thrush postfledging survival on our study sites in southern Indiana, USA, as a function of the interaction between
drought and the amount of mature forest cover within a circle centered on each study plot (radius¼ 2 km, maximum forest cover¼
1,257 ha). Lines represent estimates of survival probability (6 SE) from our top-ranked known-fate survival model. The highest
mortality occurred in the first 3 days postfledging in areas with low forest cover during a drought year. Note that y-axis scales vary by
age class.
The Auk: Ornithological Advances 135:15–24, Q 2018 American Ornithological Society
B. J. Vernasco, T. S. Sillett, P. P. Marra, and T. B. Ryder Wood Thrush postfledging survival 19
¼ 28.15 6 0.14 g, t221¼3.60, P , 0.001). Brood size was
significantly lower in parasitized nests (lparasitized¼ 2.35 6
0.08, lnonparasitized¼ 2.61 6 0.05, t221¼ 2.66, P , 0.01) and
during a drought year (ldrought¼ 2.35 6 0.09, lnon-drought¼
2.60 6 0.05, t168 ¼ 2.4, P , 0.02). Drought also
influenced fledgling movements, with the best-supported
model of fledgling movement including an interaction
between age class and drought (wi ¼ 0.74; Table 3).
Fledglings stayed closer to the nest and remained within
the natal territory longer during the drought year (Figure
3). By contrast, we found no evidence for a relationship
between brood parasitism, forest cover, or nestling body
condition and movement.
DISCUSSION
Postfledging survival is a key component of recruitment
and predictor of avian population growth (Sæther and
Bakke 2000, Robinson et al. 2014), but we lack a
comprehensive understanding of how environmental
heterogeneity affects postfledging movement, dispersal
dynamics, and survival probability. Here, we used 4 yr of
postfledging data to examine how brood parasitism,
drought, nestling condition, forest cover, and local habitat
structure influence postfledging survival. Consistent with
findings from other studies (Naef-Daenzer and Gru¨ebler
2016), our results show strong age-specific patterns of
survival, with the majority of mortality occurring during
the first 4 days postfledging. Furthermore, of the 5
predictors considered, drought received the most support
and had a negative effect on cumulative survival probabil-
ity.
Cumulative fledgling survival estimates (i.e. probability
of surviving to 21 days) for Wood Thrushes in southern
Indiana were similar to those documented from other
parts of the breeding range (/¼ 0.42 to 0.79; Anders et al.
1997, Schmidt et al. 2008) and to estimates for similar-
sized species (/ ¼ 0.62 to 0.87; Cohen and Lindell 2004,
Moore et al. 2010, Tarof et al. 2011). However, we did not
find an association with fledge date, in contrast to several
studies of postfledging survival, but consistent with other
investigations of Wood Thrushes (reviewed in Cox et al.
2014). Unlike other songbirds, fledgling Wood Thrushes
may not be affected by within-breeding-season trends in
predator behavior or resource availability (Tarof et al.
2011).
A drought-associated reduction in passerine postfledg-
ing survival has been documented by only one other study
TABLE 2. Akaike’s Information Criterion (AIC) rankings of our 10 best-fitting known-fate models examining the effects of drought;
forest cover within 2 km of study plot (Forest); 2 measurements of local vegetation structure, vertical foliage density (VFD) and count
of trees with DBH .30 cm (DBH); Brown-headed Cowbird nest parasitism (BHCO); and nestling body condition (BCI) on Wood Thrush
postfledging survival on our study sites in southern Indiana, USA.
Rank Model K a DAICc
b wi
c Deviance
1 ~AgeClass4 þ Drought*Forest 7 0.00 0.17 513.05
2 ~AgeClass4 þ VFD*Drought 7 0.27 0.15 513.32
3 ~AgeClass4 þ BCI*Forest 7 1.14 0.10 514.19
4 ~AgeClass4 þ Drought 5 1.63 0.08 81.14
5 ~AgeClass4 þ Forest þ Drought 6 2.48 0.05 517.54
6 ~AgeClass4 þ DBH þ Drought 6 3.10 0.04 518.16
7 ~AgeClass4 4 3.16 0.04 84.68
8 ~AgeClass4 þ VFD þ Drought 6 3.19 0.03 518.25
9 ~AgeClass4 þ Drought þ BCI 6 3.30 0.03 518.36
10 ~AgeClass4 þ Drought þ BHCO 6 3.40 0.03 80.90
a Number of parameters in each model.
b Difference in AIC score (corrected for small sample size, AICc) between each model and the best-fitting model (AICc¼527.09 for the
top-ranked model).
c Model weight.
FIGURE 2. Scaled mass indices from nestling Wood Thrushes
measured shortly before fledging on our study sites in southern
Indiana, USA. (A) Nestlings from parasitized nests were in
significantly better body condition. (B) During a drought year,
nestlings were in significantly poorer body condition.
The Auk: Ornithological Advances 135:15–24, Q 2018 American Ornithological Society
20 Wood Thrush postfledging survival B. J. Vernasco, T. S. Sillett, P. P. Marra, and T. B. Ryder
(Yackel Adams et al. 2006). Our study design did not allow
us to determine the mechanism, but the effects of drought
were likely mediated by resource limitation; both nestling
condition and brood size were significantly lower during
the drought. This hypothesis is supported by previous
work showing that Wood Thrush prey abundance is
positively correlated with soil moisture, and that individ-
uals in poorer condition are more susceptible to predation
or starvation early in the postfledging period (Cox et al.
2014, Jirinec et al. 2016). We did not find support for a
relationship between nestling body condition and post-
fledging survival. However, our measure of nestling
condition may not be a useful predictor of fledgling body
condition and, hence, survival probability (e.g., Streby et al.
2014). The effect of drought on fledgling survival may also
be mediated by landscape-scale habitat features.
Although we did not find strong associations between
nest vegetation structure and postfledging survival of
Wood Thrushes in southern Indiana, our analyses revealed
that the amount of mature forest cover and drought
interacted to affect survival probability. Daily survival
probability in non-drought years declined as the amount of
forest cover increased, particularly for the youngest
fledglings (Figure 1). By contrast, less forested landscapes
exhibited declines in fledgling survival in the drought year,
whereas forested landscapes appeared to be buffered.
Areas of high mature forest cover have understory
microclimates that are more resilient to climatic pertur-
bations because canopy cover mediates understory soil
moisture, a predictor of Wood Thrush prey availability
(Chen et al. 1993, von Arx et al. 2012, Jirinec et al. 2016).
The negative relationship between forest cover and
fledgling survival during non-drought years is likely due
to differences in nestling development and habitat
availability. Nestling Wood Thrush growth rates are
greater closer to forest edges (Kaiser and Lindell 2007),
and more developed nestlings have higher postfledging
survival (Jones et al. 2017). The negative relationship
between forest cover and vertical foliage density (r1966 ¼
0.19, P , 0.001) also suggests that favorable fledgling
habitat may be limited in areas with extensive, mature
forest.
Fledglings also made shorter movements from the nest
and delayed dispersal from their natal territory during the
2012 drought. This relationship was strongest in the 2
oldest age groups, after the majority of mortality occurred,
FIGURE 3. Average distance (6 SE) from the nest moved by
fledgling Wood Thrushes, as a function of age (days since
fledging) and environmental conditions (drought) on our study
sites in southern Indiana, USA. During drought years, fledglings
stayed, on average, closer to the nest from which they fledged
and appeared to delay dispersal.
TABLE 3. Akaike’s Information Criterion (AIC) rankings of our 10 best-fitting models examining the effects of drought, brood
parasitism (BHCO), forest cover within 2 km of study plot (Forest), nestling condition (BCI), and age (Age; i.e. the 4 age groups used in
Table 2) on Wood Thrush postfledging movement on our study sites in southern Indiana, USA.
Rank Model K a DAICc
b wi
c Deviance
1 ~Age*Drought 10 0.00 0.74 1,601.17
2 ~Age þ Drought þ BCI 8 2.27 0.24 1,450.65
3 ~Age þ Drought 7 8.96 0.01 1,592.28
4 ~Age þ BCI 7 10.73 0.00 1,597.92
5 ~Age þ Drought þ BHCO 8 10.80 0.00 1,601.12
6 ~Age þ Drought þ Forest 8 11.00 0.00 1,601.16
7 ~Age þ BCI þ BHCO 8 12.38 0.00 1,450.50
8 ~Age þ BCI þ Forest 8 12.77 0.00 1,446.19
9 ~Age*BCI 10 13.09 0.00 1,444.42
10 ~Age 6 13.15 0.00 1,450.65
a Number of parameters in each model.
b Difference in AIC score (corrected for small sample size, AICc) between each model and the best-fitting model (AICc¼ 1,449.63 for
the top-ranked model).
c Model weight.
The Auk: Ornithological Advances 135:15–24, Q 2018 American Ornithological Society
B. J. Vernasco, T. S. Sillett, P. P. Marra, and T. B. Ryder Wood Thrush postfledging survival 21
which suggests that the observed changes in movement
behavior did not affect fledgling survival. We did not find
an effect of condition on postfledging movement, but our
condition measurement was taken prior to fledging.
Nestling condition appears to be predictive of passerine
fledgling movements only during the first few days after
leaving the nest (Cox et al. 2014). A reduction in food
availability during the drought year could have extended
the postfledging period by slowing fledgling development.
Such condition-dependent natal dispersal (Ims and
Hjermann 2001) affects fitness because these individuals
disperse into poorer-quality habitat and have lower
lifetime reproductive success (Lens and Dhondt 1994,
Verhulst et al. 1997, Visser and Verboven 1999, Barbraud et
al. 2003). Alternatively, the distribution of dispersal
distances during 2012 may have been bimodal, with some
birds dispersing short distances and others dispersing long
distances (i.e. outside of the study area and thus
undetectable). However, mean postfledging dispersal in
Wood Thrushes is ,2 km (Anders et al. 1998, Vega Rivera
et al. 1998). All individuals (n¼ 9) that survived to 21 days
postfledging in 2012, presumed independent, exhibited
delayed dispersal. Therefore, bimodal dispersal in 2012
seems unlikely.
Brood parasitism is another environmental factor that
has been hypothesized to reduce postfledging survival
(Cox et al. 2014). Counter to our expectation, but
consistent with the results of other studies (e.g., Fink
2003, Jenkins and Faaborg 2016), brood parasitism did not
influence postfledging survival or movement of Wood
Thrushes in our study area. The negative effects of brood
parasitism found in other passerine species have been
attributed to asynchronous hatching (i.e. the parasite
hatching before the host) or to a size advantage of the
parasite nestling (Hauber 2003). Wood Thrushes are
unlikely to be affected by these factors because their
incubation period and nestling size are similar to those of
Brown-headed Cowbirds. Nestlings from parasitized nests
in southern Indiana were in significantly better condition,
but the effect size was relatively small (lparasitized¼ 29.15 6
0.24 g, lnonparasitized ¼ 28.15 6 0.14 g). Higher nestling
condition in parasitized nests could be explained by a
combination of brood-size reduction, increased provision-
ing by adults (e.g., Ursino et al. 2011), and increased
parasite begging (Bolopo et al. 2015). Although nestlings
from parasitized nests were in better condition, an increase
in survival or fledgling movements was not expected,
because we found no evidence that body condition was
important for these parameters.
In conclusion, this research advances our understanding
of how body condition, habitat structure, drought, and
brood parasitism affect the postfledging period of migra-
tory passerines. We found that drought, in particular, can
influence nestling condition, brood size, and postfledging
movement and survival, but its effect on survival depends
on the amount of mature forest cover. Additional research
is needed to discover how drought effects are mediated by
resource availability, the predator community, and parental
behavior. Our results also suggest that brood parasitism
has a negligible effect on the postfledging period in Wood
Thrushes. Investigations of how hosts select habitat and
invest in fledgling Brown-headed Cowbirds may reveal
effects of brood parasitism not captured by our study.
Ultimately, determining the sensitivity of migratory birds’
population dynamics to the postfledging period is essential
for conservation planning and for mitigating the effects of
future environmental change.
ACKNOWLEDGMENTS
We thank the many field assistants who collected data for this
project, C. Stanley and J. Valente for field support and useful
discussions, and B. Brown for helpful analytical advice.
Funding statement: Financial support was provided by the
Strategic Environmental Resource and Development Program
under award no. RC-2121 to the Smithsonian Migratory Bird
Center, University of Oregon, U.S. Army Engineer Research
and Development Center, and the Institute for Bird Popula-
tions. The funders had no input into the content of the
manuscript and did not require approval prior to submission
or publication.
Ethics statement: All research was approved by the National
Zoological Park’s Animal Care and Use Committee (protocol
nos.11-04 and 14-04).
Author contributions: T.B.R., T.S.S., and P.P.M. designed the
research project and acquired funding. B.J.V. and T.B.R.
collected the data. B.J.V. analyzed the data. B.J.V., T.B.R., and
T.S.S. wrote the manuscript.
Data deposits: Data will be deposited online upon publica-
tion (https://nationalzoo.si.edu/migratory-birds/data).
LITERATURE CITED
Anders, A. D., D. C. Dearborn, J. Faaborg, and F. R. Thompson III
(1997). Juvenile survival in a population of Neotropical
migrant birds. Conservation Biology 11:698–707.
Anders, A. D., J. Faaborg, and F. R. Thompson III (1998).
Postfledging dispersal, habitat use, and home-range size of
juvenile Wood Thrushes. The Auk 115:349–358.
Arnold, T. W. (2010). Uninformative parameters and model
selection using Akaike’s Information Criterion. Journal of
Wildlife Management 74:1175–1178.
Ausprey, I. J., and A. D. Rodewald (2011). Postfledging
survivorship and habitat selection across a rural-to-urban
landscape gradient. The Auk 128:293–302.
Barbraud, C., A. R. Johnson, and G. Bertault (2003). Phenotypic
correlates of post-fledging dispersal in a population of
greater flamingos: The importance of body condition. Journal
of Animal Ecology 72:246–257.
Bates, D., M. Ma¨chler, B. Bolker, and S. Walker (2015). Fitting
linear mixed-effects models using lme4. Journal of Statistical
Software 67(1).
The Auk: Ornithological Advances 135:15–24, Q 2018 American Ornithological Society
22 Wood Thrush postfledging survival B. J. Vernasco, T. S. Sillett, P. P. Marra, and T. B. Ryder
Bolopo, D., D. Canestrari, M. Rolda´n, V. Baglione, and M. Soler
(2015). High begging intensity of Great Spotted Cuckoo
nestlings favours larger-size crow nest mates. Behavioral
Ecology and Sociobiology 69:873–882.
Burnham, K. P., and D. R. Anderson (2002). Model Selection and
Multimodel Inference: A Practical Information–Theoretic
Approach, second edition. Springer, New York, NY, USA.
Chen, J., J. F. Franklin, and T. A. Spies (1993). Contrasting
microclimates among clearcut, edge, and interior of old-
growth Douglas-fir forest. Agricultural and Forest Meteorol-
ogy 63:219–237.
Cohen, E. B., and C. A. Lindell (2004). Survival, habitat use, and
movements of fledgling White-throated Robins (Turdus
assimilis) in a Costa Rican agricultural landscape. The Auk
121:404–414.
Cox, W. A., F. R. Thompson III, A. S. Cox, and J. Faaborg (2014).
Post-fledging survival in passerine birds and the value of
post-fledging studies to conservation. The Journal of Wildlife
Management 78:183–193.
Doherty, P. F., G. C. White, and K. P. Burnham (2012). Comparison
of model building and selection strategies. Journal of
Ornithology 152:317–323.
Evans, M., E. Gow, R. R. Roth, M. S. Johnson, and T. J. Underwood
(2011). Wood Thrush (Hylocichla mustelina). In Birds of North
America Online (P. G. Rodewald, Editor). Cornell Lab of
Ornithology, Ithaca, NY, USA. https://birdsna.org/Species-
Account/bna/species/woothr
Fink, M. L. (2003). Post-fledging ecology of juvenile Wood Thrush
in fragmented and contiguous landscapes. Ph.D. dissertation,
University of Missouri, Columbia, MO, USA.
Fry, J., G. Xian, S. Jin, J. Dewitz, C. Homer, L. Yang, C. Barnes, N.
Herold, and J. Wickham (2011). Completion of the 2006
National Land Cover Database for the conterminous United
States. Photogrammetric Engineering and Remote Sensing
77:858–864.
Hauber, M. E. (2003). Hatching asynchrony, nestling competition,
and the cost of interspecific brood parasitism. Behavioral
Ecology 14:227–235.
Ims, R. A., and D. Ø. Hjermann (2001). Condition-dependent
dispersal. In Dispersal (J. Clobert, E. Danchin, A. A. Dhondt,
and J. D. Nichols, Editors). Oxford University Press, Oxford, UK.
pp. 203–216.
Jenkins, J. M. A., and J. Faaborg (2016). Potential effects of
Brown-headed Cowbirds (Molothrus ater) on host postfledg-
ing dispersal and survival. The Wilson Journal of Ornithology
128:404–411.
Jirinec, V., R. E. Isdell, and M. Leu (2016). Prey availability and habitat
structure explain breeding space use of a migratory songbird.
The Condor: Ornithological Applications 118:309–328.
Jones, T. M., M. P. Ward, T. J. Benson, and J. D. Brawn (2017).
Variation in nestling body condition and wing development
predict cause-specific mortality in fledgling Dickcissels.
Journal of Avian Biology 48:439–447.
Kaiser, S. A., and C. A. Lindell (2007). Effects of distance to edge
and edge type on nestling growth and nest survival in the
Wood Thrush. The Condor 109:288–303.
King, D. I., R. M. Degraaf, M.-L. Smith, and J. P. Buonaccorsi
(2006). Habitat selection and habitat-specific survival of
fledgling Ovenbirds (Seiurus aurocapilla). Journal of Zoology
269:414–421.
Laake, J. L. (2013). RMark: An R interface for analysis of capture–
recapture data with MARK. AFSC Processed Report 2013-01.
Alaska Fisheries Science Center and NOAA National Marine
Fisheries Service, Seattle, WA, USA.
Lens, L., and A. A. Dhondt (1994). Effects of habitat fragmen-
tation on the timing of Crested Tit Parus cristatus natal
dispersal. Ibis 136:147–152.
Mallya, G., L. Zhao, X. Song, D. Niyogi, and R. Govindaraju (2013).
2012 Midwest drought in the United States. Journal of
Hydrologic Engineering 18:737–745.
Martin, T. E., and G. R. Geupel (1993). Nest-monitoring plots:
Methods for locating nests and monitoring success. Journal
of Field Ornithology 64:507–519.
Moore, L. C., B. J. M. Stutchbury, D. M. Burke, and K. A. Elliott
(2010). Effects of forest management on postfledging survival
of Rose-breasted Grosbeaks (Pheucticus ludovicianus). The
Auk 127:185–194.
Naef-Daenzer, B., and M. U. Gru¨ebler (2016). Post-fledging
survival of altricial birds: Ecological determinants and
adaptation. Journal of Field Ornithology 87:227–250.
Naef-Daenzer, B., F. Widmer, and M. Nuber (2001). Differential
post-fledging survival of great and coal tits in relation to their
condition and fledging date. Journal of Animal Ecology 70:
730–738.
Newton, I. (1998). Population Limitation in Birds. Academic Press,
San Diego, CA, USA.
Payne, R. B., and L. L. Payne (1998). Brood parasitism by
cowbirds: Risks and effects on reproductive success and
survival in Indigo Buntings. Behavioral Ecology 9:64–73.
Peig, J., and A. J. Green (2009). New perspectives for estimating
body condition from mass/length data: The scaled mass
index as an alternative method. Oikos 118:1883–1891.
Peterson, S. M., H. M. Streby, and D. E. Andersen (2012). Effects of
parasitism by Brown-headed Cowbirds may persist into post-
fledging. The Wilson Journal of Ornithology 124:179–183.
Rappole, J. H., and A. R. Tipton (1991). New harness design for
attachment of radio transmitters to small passerines. Journal
of Field Ornithology 62:335–337.
Rasmussen, J. L., and S. G. Sealy (2006). Hosts feeding only
Brown-headed Cowbird fledglings: Where are the host
fledglings? Journal of Field Ornithology 77:269–279.
R Core Team (2015). R: A Language and Environment for
Statistical Computing. R Foundation for Statistical Comput-
ing, Vienna, Austria. https://www.R-project.org/
Robinson, R. A., C. A. Morrison, and S. R. Baillie (2014). Integrating
demographic data: Towards a framework for monitoring
wildlife populations at large spatial scales. Methods in
Ecology & Evolution 5:1361–1372.
Rush, S. A., and B. J. M. Stutchbury (2008). Survival of fledgling
Hooded Warblers (Wilsonia citrina) in small and large forest
fragments. The Auk 125:183–191.
Rushing, C. S., J. A. Hostetler, T. S. Sillett, P. P. Marra, J. A.
Rotenberg, and T. B. Ryder (2017). Spatial and temporal
drivers of avian population dynamics across the annual cycle.
Ecology 98. In press.
Rushing, C. S., T. B. Ryder, and P. P. Marra (2016). Quantifying
drivers of population dynamics for a migratory bird
throughout the annual cycle. Proceedings of the Royal
Society B 283:20152846.
Sæther, B.-E., and Ø. Bakke (2000). Avian life history variation and
contribution of demographic traits to the population growth
rate. Ecology 81:642–653.
Schaub, M., and F. Abadi (2011). Integrated population models:
A novel analysis framework for deeper insights into
The Auk: Ornithological Advances 135:15–24, Q 2018 American Ornithological Society
B. J. Vernasco, T. S. Sillett, P. P. Marra, and T. B. Ryder Wood Thrush postfledging survival 23
population dynamics. Journal of Ornithology 152 (Supple-
ment 1):227–237.
Schmidt, K. A., S. A. Rush, and R. S. Ostfeld (2008). Wood Thrush
nest success and post-fledging survival across a temporal
pulse of small mammal abundance in an oak forest. Journal
of Animal Ecology 77:830–837.
Streby, H. M., S. M. Peterson, J. A. Lehman, G. R. Kramer, B. J.
Vernasco, and D. E. Andersen (2014). Do digestive
contents confound body mass as a measure of relative
condition in nestling songbirds? Wildlife Society Bulletin
38:305–310.
Tarof, S. A., P. M. Kramer, J. R. Hill III, J. Tautin, and B. J. M.
Stutchbury (2011). Brood size and late breeding are
negatively related to juvenile survival in a Neotropical
migratory songbird. The Auk 128:716–725.
Ursino, C. A., M. C. De Ma´rsico, M. Sued, A. Farall, and J. C.
Reboreda (2011). Brood parasitism disproportionately in-
creases nest provisioning and helper recruitment in a
cooperatively breeding bird. Behavioral Ecology and Socio-
biology 65:2279–2286.
Vega Rivera, J. H., C. A. Hass, J. H. Rappole, and W. J. McShea
(2000). Parental care of fledgling Wood Thrushes. The Wilson
Bulletin 112:233–237.
Vega Rivera, J. H., J. H. Rappole, W. J. McShea, and C. A. Haas
(1998). Wood Thrush postfledging movements and habitat
use in northern Virginia. The Condor 100:69–78.
Verhulst, S., C. M. Perrins, and R. Riddington (1997). Natal dispersal
of Great Tits in a patchy environment. Ecology 78:864–872.
Visser, M. E., and N. Verboven (1999). Long-term fitness effects of
fledging date in Great Tits. Oikos 85:445–450.
Vitz, A. C., and A. D. Rodewald (2011). Influence of condition and
habitat use on survival of post-fledging songbirds. The
Condor 113:400–411.
von Arx, G., M. Dobbertin, and M. Rebetez (2012). Spatio-
temporal effects of forest canopy on understory microclimate
in a long-term experiment in Switzerland. Agricultural and
Forest Meteorology 166–167:144–155.
White, G. C., and K. P. Burnham (1999). Program MARK: Survival
estimation from populations of marked animals. Bird Study
46 (Supplement):S120–S139.
White, J. D., T. Gardali, F. R. Thompson III, and J. Faaborg (2005).
Resource selection by juvenile Swainson’s Thrushes during
the post-fledging period. The Condor 107:388–401.
Yackel Adams, A. A., S. K. Skagen, and J. A. Savidge (2006).
Modeling post-fledging survival of Lark Buntings in response
to ecological and biological factors. Ecology 87:178–188.
APPENDIX TABLE 4. Size of study plots in southern Indiana,
USA, and number of transmitters (n) deployed on each plot.
Plot Region a Size (ha) n
AR07 BONWR 54 21
AR27 BONWR 54 38
AR41 BONWR 53 15
AR58 BONWR 48 15
EABO Crane 57 20
FIRS Crane 36 3
MART IDNR 52 14
MCCR IDNR 48 22
OWEN IDNR 47 15
SEED Crane 84 10
SPMI IDNR 43 18
SULP Crane 40 19
a BONWR ¼ Big Oaks National Wildlife Refuge, Crane ¼ Crane
naval base, and IDNR ¼ Indiana Department of Natural
Resources.
APPENDIX TABLE 5. Annual mean (6 SE) values for nestling Wood Thrush condition, brood size, vertical foliage density, and
number of trees .30 cm DBH, and proportion of nests parasitized, on our study sites in southern Indiana, USA, for the 4 yr of the
study.
2011 2012 2013 2014
Mean nestling condition 29.55 6 0.29 26.52 6 0.34 28.7 6 0.17 28.7 6 0.15
Mean brood size 2.56 6 0.1 2.35 6 0.09 2.58 6 0.08 2.64 6 0.07
Proportion of nests parasitized 0.34 0.28 0.18 0.18
Mean vertical foliage density 3.1 6 0.09 2.86 6 0.07 3.02 6 0.07 2.55 6 0.05
Mean number of trees .30 cm DBH 15.02 6 0.43 13.74 6 0.37 11.6 6 0.24 19.07 6 0.34
The Auk: Ornithological Advances 135:15–24, Q 2018 American Ornithological Society
24 Wood Thrush postfledging survival B. J. Vernasco, T. S. Sillett, P. P. Marra, and T. B. Ryder