Ecology, 91(10), 2010, pp. 2874?2882
 2010 by the Ecological Society of America
Moisture as a determinant of habitat quality for a nonbreeding
Neotropical migratory songbird
JOSEPH A. M. SMITH,1,3 LEONARD R. REITSMA,2 AND PETER P. MARRA1
1Smithsonian Migratory Bird Center, National Zoological Park, 3001 Connecticut Avenue, Washington, D.C. 20008 USA
2Plymouth State University, Department of Biological Sciences, Plymouth, New Hampshire 03264 USA
Abstract. Identifying the determinants of habitat quality for a species is essential for
understanding how populations are limited and regulated. Spatiotemporal variation in
moisture and its in?uence on food availability may drive patterns of habitat occupancy and
demographic outcomes. Nonbreeding migratory birds in the neotropics occupy a range of
habitat types that vary with respect to moisture. Using carbon isotopes and a satellite-derived
measure of habitat moisture, we identi?ed a moisture gradient across home ranges of radio-
tracked Northern Waterthrush (Seiurus noveboracensis). We used this gradient to classify
habitat types and to examine whether habitat moisture correlates with overwinter mass change
and spring departure schedules of Northern Waterthrush over the late-winter dry season in the
tropics. The two independent indicators of moisture revealed similar gradients that were
directly proportional to body mass change as the dry season progressed. Birds occupying drier
habitats declined in body mass over the study period, while those occupying wetter habitats
increased in body mass. Regardless of habitat, birds lost an average of 7.6% of their mass at
night, and mass recovery during the day trended lower in dry compared with wet habitats.
This suggests that daily incremental shortfalls in mass recovery can lead to considerable
season-long declines in body mass. These patterns resulted in consequences for the
premigratory period, with birds occupying drier habitats having a delayed rate of fat
deposition compared with those in wet habitats. Taken together with the ?nding that males,
which are signi?cantly larger than females, are also in better condition than females regardless
of habitat suggests that high-quality habitats may be limited and that there may be
competition for them. The habitat-linked variation in performance we observed suggests that
habitat limitation could impact individual and population-level processes both during and in
subsequent periods of the annual cycle. The linkage between moisture and habitat quality for a
migratory bird indicates that the availability of high-quality habitats is dynamic due to
variation in precipitation among seasons and years. Understanding this link is critical for
ascertaining the impact of future climate change, particularly in the Caribbean basin, where a
much drier future is predicted.
Key words: body condition; body mass; habitat quality; migratory birds; nonbreeding winter habitat;
Northern Waterthrush; Roosevelt Roads Naval Station, Puerto Rico; Seiurus noveboracensis.
INTRODUCTION
Climate and, more speci?cally, temperature and
moisture, are among the most fundamental parameters
that determine the composition and distribution of
habitats (Holdridge 1967) and play an important role in
determining resource abundances for organisms at
higher trophic levels (Janzen and Schoener 1968, Karr
and Freemark 1983, Bale et al. 2002). In lowland
tropical regions, temperature at a given location is
relatively stable (Janzen 1967). In comparison, moisture
can vary considerably over space, seasonally, and across
years due to geographic and seasonal patterns (Karr and
Freemark 1983, Murphy and Lugo 1986). Such varia-
tion may differentially impact the plant and invertebrate
communities that comprise the habitats higher trophic
level organisms occupy and may ultimately play a role in
driving variation in demographic traits of these species.
Variation in moisture over time and space can have
important impacts on populations by regulating food
supplies (White 2008). In particular, studies in tropical
environments that experience rainfall seasonality suggest
a general pattern that wetter locations have higher food
resources and that these differences in food resources
can have consequences on individual condition and
survival (Dunham et al. 2004, Madsen et al. 2006, Valeix
et al. 2008). While moisture may vary temporally (i.e.,
across years and seasons) and affect all members of a
population equally, moisture availability also varies
spatially (i.e., within and among habitat types) and can
result in consequences for different components of a
population and ultimately play a role in population
Manuscript received 25 November 2009; accepted 25
January 2010. Corresponding Editor: M. Wikelski.
3 Present address: The Nature Conservancy, Delaware
Bayshores Of?ce, 2350 Route 47, Delmont, New Jersey 08314
USA. E-mail: Joseph_Smith@tnc.org
2874
limitation and regulation (Johnson and McIlwee 1997,
Studds and Marra 2007).
Neotropical migratory birds are an excellent taxo-
nomic group to study the role of moisture in determin-
ing habitat quality because a single species often
occupies a broad range of habitats that vary consider-
ably with respect to moisture. Previous studies working
across a range of habitats have demonstrated that
insectivorous Neotropical migratory birds that occupy
drier nonbreeding habitat types show lower overwinter
site ?delity (Wunderle and Latta 2000, Latta and
Faaborg 2001, 2002), declining physical condition and
survival (Marra et al. 1998, Marra and Holmes 2001,
Latta and Faaborg 2002, Brown and Sherry 2006,
Johnson et al. 2006), delayed departure for breeding
grounds (Marra et al. 1998), and delayed breeding
initiation in the subsequent breeding season (Marra et
al. 1998, Norris et al. 2004, Runge and Marra 2005,
Reudink et al. 2009), suggesting that winter habitat
quality plays a critical role in population regulation
(Sherry and Holmes 1996, Runge and Marra 2005).
Habitats are often classi?ed a priori into patches of
homogeneous types. Few studies of nonbreeding migra-
tory birds have classi?ed habitats a posteriori based on
measures of ecologically relevant habitat features to
better re?ect how study organisms perceive habitats.
While moisture is suggested as a determinant of habitat
quality, few of these studies speci?cally measured
moisture at the scale of individual home ranges to
examine patterns of variation within and among
habitats that may in turn in?uence food availability
and an individual?s physical condition.
In this study, we examine the role of moisture in
driving arthropod availability and the physical condition
and performance of the Northern Waterthrush (Seiurus
noveboracensis) across four habitat types and over
diurnal and seasonal timescales. We quantify moisture
levels on the home ranges of radio-tracked birds
occupying the four different habitat types using satellite
imagery and a stable carbon isotope index of habitat
moisture assessed from a blood sample from these radio-
marked birds. To assess changes in body condition we
quantify (1) overnight mass loss and daytime body mass
recovery and (2) seasonal changes in body mass of
Northern Waterthrush across the habitat moisture
gradient. Finally, we assessed impacts of changes in
body condition on the timing of spring migration.
METHODS
Species description and study area
The Northern Waterthrush is a paruline warbler that
breeds throughout Canada, the northeastern United
States, and Alaska and spends the nonbreeding period in
the neotropics in Central America, northern South
America, and the Caribbean (Eaton 1995). During the
nonbreeding period the Northern Waterthrush is site
faithful and solitary during the day (Schwartz 1964,
Lefebvre et al. 1994, Reitsma et al. 2002) and is
considered to be a moist-habitat specialist, occurring
in its highest recorded densities in mangroves (Wunderle
and Waide 1993). The occupation of exclusive non-
breeding territories was documented in one study
(Schwartz 1964) but others have found that home
ranges overlap (Lefebvre et al. 1994, Reitsma et al.
2002). Evidence from our study sites indicates that at
least some individuals maintain exclusive territories
(Smith 2008). Foraging occurs primarily at ground level
(Schwartz 1964, Lefebvre et al. 1992), and winter diet is
composed of invertebrates (Lefebvre et al. 1992, 1994).
We conducted this research at Roosevelt Roads Naval
Station (188200 N, 658600 W) near Ceiba, on the east
coast of Puerto Rico during January?April 2002?2004.
Of the total 3464 ha that compose the station, 1612 ha
are second-growth dry forest, 769 ha are mangrove
forest, and 1083 ha are developed (GIS data; Eagan,
McAllister and Associates, unpublished data). Developed
areas are composed of roads, an airstrip, residential
areas, of?ce complexes, commercial properties, and a
ship port.
We worked in four habitat types including second-
growth dry forest and three mangrove forest types. Each
of the mangrove forest plots was dominated by a single
species: black mangrove (Avicennia germinans), white
mangrove (Languncularia racemosa), and red mangrove
(Rhizophora mangle). Dry forest was a heterogeneous
habitat with a more complex plant community domi-
nated by Bursera simaruba, Leucaena glauca, Prosopis
juli?ora, and Guaiacum of?cinale (Ewel and Whitmore
1973). These forests developed from agricultural lands
that were primarily sugar cane or pasture as late as the
1930s (Thomlinson et al. 1996). Red mangrove areas
were directly adjacent to the coast, lagoons, and major
drainages and occurred in pure stands, whereas black
and white mangrove formed mixed- and single-species
stands inland from red mangrove areas. Dry forest sites
were still further inland, adjacent to mangroves.
Additional mangrove areas present at Roosevelt Roads
Naval Station that were only rarely used by Northern
Waterthrush included short-stature (,2 m) stands of
black mangrove and stands of dead mangrove.
Study sites consisted of two dry-forest sites 1.5 km
apart, two white mangrove sites 6.5 km apart, two black
mangrove sites 1.5 km apart, and one red mangrove site
that was separated from the nearest dry-forest site by a
300-m band of salt ?ats interspersed with sparse, short-
stature black mangrove. Canopy heights within each site
ranged from 3 to 20 m. Dry-forest sites did not have
standing water. The dry season in this region lasts from
January through April, when mean monthly precipita-
tion is 7.5 cm or less (Daly et al. 2003). All mangrove
sites had standing water, usually ,1 m in January, but
water depth gradually decreased as the dry season
progressed toward April and May. Daily ?uctuations in
moisture levels infrequently occurred in coastal red
mangrove areas during periods of exceptional tide
?uctuations.
October 2010 2875MOISTURE AND WINTERING MIGRATORY BIRDS
Bird sampling and radiotelemetry
Northern Waterthrush were captured at two-week
intervals from late January to early April. We erected
10?15 12-m mist nets at each study site for a two-day
period from 06:00 to 18:00. Waterthrush were captured
without the use of playbacks or other enticements to
avoid biases toward territorial or behaviorally dominant
individuals. At the time of capture all individuals were
banded with a unique combination of two colored leg
bands and an aluminum U.S. Fish and Wildlife Service
band. Each individual was measured (un?attened wing
chord, tarsus length, and tail length to 60.5 mm),
weighed to the nearest 60.1 g using a digital scale
(Ohaus, Pine Brook, New Jersey, USA), and given a
furcular fat score (categories: 0, none visible; 1, trace; 2,
fat forming a solid sheet across the bottom of the
furculum; and 3, fat ?lling furculum [Holmes et al.
1989]). Age was not determined due to the unreliability
of aging methodology for this species during the winter
period (Pyle 1997). A 75-lL blood sample was taken for
isotope analysis and to determine sex using a molecular
technique (Grif?ths et al. 1998). Previous studies have
demonstrated that blood sampling does not affect mass
of small passerines (Stangel 1986) or their survival
(Hoysak and Weatherhead 1991). We sexed 270 birds in
total (2002, n ? 153; 2003, n ? 63; 2004, n ? 54).
We analyzed carbon isotopes in blood samples as a
measure of habitat use (Marra et al. 1998) for all
Northern Waterthrush radio tracked in 2003 and 2004
as well as a random sample of birds captured over the
course of the study (n?229). Radio-tracked birds served
as a reference group with known habitat af?liations so
that birds that were not tracked could be reliably
assigned to a habitat category using discriminant
function analysis. Habitat use could be inferred from
these data because carbon isotope ratios in plant tissues
are determined by both the water use ef?ciency and the
photosynthetic pathways of plants (O?Leary 1981,
Farquhar et al. 1989). Stable carbon isotope ratios are
transferred up the food chain to invertebrates feeding on
the plants and eventually to invertebrate predators such
as the Northern Waterthrush (Peterson and Fry 1987).
Because it re?ects the spatial scale at which birds are
feeding, isotope ratio can be used as an index of habitat
moisture within and between habitats (Marra et al.
1998), with depleted d13C values being a signature of
moist habitat and enriched values indicative of a drier
habitat. We chose to use blood plasma since it has a
quicker turnover rate than other blood components (2?3
days; Hobson and Clark 1993) and thus would best
re?ect the habitat that the bird was using just prior to
the time of capture. All samples were collected within a
two-week interval during January. Blood samples (75
lL) were collected in capillary tubes via brachial
venipuncture, placed on ice, and later (,4 h) centrifuged
at 1675.8 m/s2 for 8 min to separate red blood cells from
plasma. The blood components were separated using a
Hamilton syringe and frozen at158C. We freeze-dried,
powdered, and loaded 1 mg of each sample into tin
capsules for combustion (Hobson et al. 1997) in a
continuous ?ow isotope ratio mass spectrophotometer
(Europa Scienti?c, Cheshire, UK; University of Cal-
ifornia Davis, Stable Isotope Facility). Results of stable
isotope analyses are expressed as the deviation (in parts
per thousand) from the Pee Dee belemnite standard
(Ehleringer and Osmond 1989).
From 2002 to 2004 we used radio transmitters
attached to a subset of Northern Waterthrush to
establish a reference group of stable carbon isotope
samples from birds with known habitat af?liations. We
radio tracked a total of 124 Northern Waterthrush over
three seasons (2002, n? 33 females, 16 males; 2003, n?
21 males, 25 females; 2004, n ? 19 males, 18 females).
Each individual was ?tted with a radio transmitter
that had a unique frequency (n ? 75; Holohil BD2-A,
Carp, Ontario, Canada; 0.74 g) using a leg harness
technique (Rappole and Tipton 1991). Transmitter life
ranged from three to four weeks. Each year we deployed
30 transmitters in late January, 30 in mid-February, and
15 in early March. During the February and March
capture intervals we attempted to recapture as many
birds as possible that were initially captured and tracked
during January to measure change in body mass and to
replace failing transmitters with new units to extend the
tracking period of these individuals. As a result, birds
were tracked over a variable period ranging from 7 to 70
days (mean ? 32.3, median ? 26 days).
To quantify space use, one location per day of each
individual was acquired during its tracking period. In
2002, individuals were found three times per day and
little movement occurred within a day once a bird
arrived on a feeding area from its roost site (within-day
minimum convex polygon area ? 0.07 ha 6 0.01 [mean
6 SE], n? 25) thus little information was lost by ?nding
birds once per day and we could increase our sample
sizes of independent individuals. Location timing was
varied randomly to re?ect the full span of daylight
hours, except for dawn (before 07:00) and dusk (after
17:00) periods when the majority of birds were
commuting to distant roost sites (Smith et al. 2008).
We used radio receivers (Fieldmaster 16, Advance
Telemetry Sytems, Isanti, Minnesota, USA) and three
element Yagi antennas (Advanced Telemetry Systems)
to relocate individuals with transmitters via homing. We
approached as close as possible without disturbing the
bird (to ;10 m). The observer location was then marked
with a GPS (GPS 12, Garmin, Olathe, Kansas, USA),
and the distance and bearing from this point to the bird
were noted.
Home range moisture
We used satellite imagery to quantify wetness on
radio-tracked Northern Waterthrush home ranges. A
representative scene of 30-m resolution Landsat En-
hanced Thematic Mapper (?ETM) imagery was used to
quantify habitat conditions on 9 January 2001, which
JOSEPH A. M. SMITH ET AL.2876 Ecology, Vol. 91, No. 10
corresponds to the time interval re?ected in blood
samples taken for isotope analysis of radio-tracked
birds. Although the scene used does not correspond to a
year during which ?eldwork was conducted, habitat
coverage within the study sites is identical (GIS data,
Eagan, McAllister and Associates, unpublished data),
and the scene used represents a year with typical annual
rainfall patterns in which a pronounced dry season
occurred between January and March following a much
wetter period between September and December.
To derive an index of wetness from the Landsat scene,
the tasseled cap transformation was performed using
ENVI 4.3 (ITT Corporation 2006). This transformation
uses a standard series of coef?cients to create weighted
sums of spectral bands that result in a reduced number
of bands that explain .95% of image variation (Crist
and Cicone 1984). The ?rst three resultant bands are the
most frequently used and correspond to brightness,
greenness, and wetness (Crist and Kauth 1986). Here we
use the wetness component as a relative measure of soil
moisture within Northern Waterthrush home ranges.
This component estimates the amount of moisture held
by vegetation and soil and is calculated by contrasting
infrared spectral bands with the sum of visible and near-
infrared spectra (Crist and Cicone 1984). Previous work
has shown that this index corresponds well to temporal
and spatial variation of soil (Ordoyne and Friedl 2008)
and canopy foliar moisture (Toomey and Vierling 2005).
To calculate wetness on Northern Waterthrush home
ranges we extracted these values for all raster cells
corresponding to telemetry locations and calculated a
mean wetness value for each home range.
Food availability
Each time a bird was located, prey availability in the
bird?s foraging area was quanti?ed using a timed, direct
search of the ground substrate within two 0.25-m2
quadrats (Johnson and Strong 2000). We placed the
sample site as close as possible to the location without
disturbing the bird. Each search was divided into a
passive and active period. The ?rst 3 min were spent
visually searching the sample area and during the
following 2 min the substrate (usually leaf litter) was
overturned to expose hidden prey because the Northern
Waterthrush is an active leaf-tosser during foraging
(Schwartz 1964, Post 1978). All invertebrates encoun-
tered were classi?ed by taxonomic order, and length was
measured. Biomass was calculated using published
length?mass regressions for Jamaican arthropods (John-
son and Strong 2000). Previous research using emetic
sampling documented the winter diet of Northern
Waterthrush in Panama at a similar coastal study site
(Lefebvre et al. 1992, Poulin et al. 1994). We constrained
our samples to arthropod orders and size classes that
were a known component of waterthrush diets. Home
range food availability was measured for 74 individuals,
and samples taken per territory ranged from 5 to 36
(mean ? 14, median ? 12).
Statistical analyses
We used ANCOVA to examine variation in body
mass during the day and across the span of the dry
season. For this analysis, we excluded all birds with
visible fat scores .1 (i.e., birds that had begun
premigratory fattening) and entered a body size index
variable (principal component 1 [PC1] from a principal
components analysis [PCA] of wing, tail, and tarsus that
explained 80.2% of variance) as a covariate. For birds
captured multiple times within a season, we used only
the ?rst capture incident during a season, but considered
captures of the same individual between years to be
independent since the environmental conditions an
individual experienced between years were independent.
We added the following explanatory variables into an
ANCOVA model for 229 Northern Waterthrush cap-
tures to examine their effects on body mass: body size,
time of day, season, year, and habitat. The season
component condensed banding periods into the early
and late dry season. We also included habitat3 time of
day and habitat 3 season interactions to examine
whether habitat affected mass change across these two
temporal scales. All interactions including covariates
were not signi?cant and were removed from the ?nal
model. The habitat component was a binary variable
representing the wet/dry dichotomy revealed by a
discriminate function analysis of carbon isotopes using
a single categorical habitat variable. The discriminant
function was 93.5% accurate at assigning radio-tracked
birds (n ? 78) with known habitat af?liations to the
correct category. We used the resulting classi?cation
coef?cients to assign birds that were captured by mist
nets that were not radio tracked to one of these habitat
categories for subsequent analysis.
We used repeated-measures ANOVA to analyze rates
of fattening of individual birds between habitats during
the premigratory period. For this analysis, we con-
strained the data set to a one-month period spanning
between mid-March and early April and divided the set
into two sampling periods of equal length. Only
individuals captured in both periods were included in
the analysis. If they were captured more than once in a
period, a mean fat score for that individual was used.
Time of day was included as a covariate.
RESULTS
Satellite moisture patterns by habitat
The satellite wetness index produces a unitless
measure of wetness with increasing values corresponding
with increasing wetness. Radio-tracked Northern Wa-
terthrush with home ranges in black mangrove (51.80
6 4.08 [mean 6 SE]; n ? 23) and dry forest (61.04 6
3.41; n? 30) had signi?cantly drier values than birds in
white mangrove (33.97 6 4.54; n ? 17) and red
mangrove (20.04 6 5.00; n? 14; ANOVA with Tukey
post hoc test, F? 18.45, df? 3, 80, P , 0.0001; Fig. 1).
October 2010 2877MOISTURE AND WINTERING MIGRATORY BIRDS
Stable carbon isotopes as a measure of habitat moisture
Stable carbon isotope signatures from the blood of
radio-tracked birds with known habitat af?nities mir-
rored the moisture gradient found using satellite imagery
(Fig. 1). Northern Waterthrush inhabiting black man-
grove (24.25% 6 0.27%, n? 22; Fig. 1) and dry forest
(24.40% 6 0.25%, n ? 27) had signi?cantly more
enriched (i.e., drier) stable carbon isotope values
(ANOVA with Tukey post hoc test, F ? 21.83, df ?
3, 74, P , 0.0001) compared to those occupying white
(27.18% 6 0.33%, n ? 15) and red mangrove
(26.04% 6 0.34%, n ? 14).
Habitat moisture and food availability
Northern Waterthrush home ranges in the dry
habitats had signi?cantly lower food availability (6.59
6 0.78 g dry mass/0.25 m2, n ? 46; Wilcoxon signed-
rank test, z ? 3.52, P ? 0.0004) compared to the home
ranges of birds occupying wet habitat (10.63 6 0.93 g
dry mass/0.25 m2, n? 31).
Seasonal mass change of radio-tracked birds and indices
of moisture
We found a strong linear relationship between stable
carbon isotope values in blood plasma and mass change
of radio-tracked Northern Waterthrush over the dry
season (adjusted R2 ? 0.45, df ? 1, P , 0.0001, n ? 32;
Fig. 2A). Birds with more enriched stable carbon
signatures, representing drier habitat conditions and
lower food availability, lost body mass whereas birds
with more depleted stable carbon isotope signatures
(wetter habitat conditions) gained mass as the season
progressed.
The same pattern was evident when we examined the
relationship between mass change and wetness of
individual home ranges derived from satellite imagery
(adjusted R2? 0.27, df? 1, P? 0.0007, n? 35; Fig. 2B).
Once again birds occupying drier home ranges (repre-
sented by more negative values) tended to lose mass
while those using wetter sites tended to gain mass as the
dry season progressed.
Variation in habitat use and mass change by sex
For radio-tracked birds there was also a sex-speci?c
pattern in mass change. An ANCOVA model (F? 6.65,
df ? 3, 267, P ? 0.0012) including sex, the wet/dry
habitat dichotomy, and a body size covariate (PC1 from
a PCA of wing, tail, and tarsus) continued to show that
birds in wet habitats had more positive trajectories of
mass change (F ? 11.63, df ? 1, 267, P ? 0.0017; wet
least-squares mean mass change, 0.66 6 0.15 g; dry
least-squares mass change, 0.008 6 0.12 g) but
regardless of habitat type, males showed signi?cantly
better performance than females (F? 6.65, df? 1, 267, P
? 0.014; male least-squares mass change, 0.66 6 0.17 g;
female least-squares mass change, 0.011 6 0.15 g).
We found no evidence that sexes segregated by
habitat. Based on a sample of 106 known-sex Northern
Waterthrush, with habitat use inferred from carbon
isotope results, sex ratio did not vary between the wet
and dry habitat dichotomy (v2 ? 0.202, P ? 0.65).
FIG. 1. Stable carbon isotope values (mean 6 SE) in blood plasma (d13C) and mean tasseled cap wetness index values of home
ranges for radio-marked Northern Waterthrush (Seiurus noveboracensis) in Puerto Rico. Results of stable isotope analyses are
expressed as the deviation (in parts per thousand) from the Pee Dee belemnite standard (Ehleringer and Osmond 1989). See
Methods: Home range moisture for an explanation of the wetness index. More negative stable isotope values represent wetter
habitat conditions, whereas increasing wetness index values represent wetter habitat conditions. Birds occupying white mangrove
(Languncularia racemosa) and red mangrove (Rhizophora mangle) showed signi?cantly wetter values for both measures compared
to birds using black mangrove and dry forest. The shaded section of the ?gure indicates the drier habitats.
JOSEPH A. M. SMITH ET AL.2878 Ecology, Vol. 91, No. 10
Daily and seasonal mass change of birds that were not
radio tracked
Nocturnal mass change.?Individual birds captured
while moving between roosts and diurnal habitats
(Smith et al. 2008) in the evening and recaptured again
the following morning (n ? 29) lost 7.6% of body mass
on average (repeated-measures ANOVA, F? 27.63, df?
1, 28, P , 0.0001; evening mean, 16.2 6 0.17 g; morning
mean, 15.1 6 0.16 g). We observed no signi?cant
difference in nocturnal mass loss of birds using wet vs.
dry diurnal habitats (F? 0.910, df? 2, 28, P? 0.35; wet,
n? 17, change,0.061% 6 0.011%; dry, n? 12, change,
0.075% 6 0.013%).
Diurnal mass change.?An analysis (ANCOVA, Table
1) of body mass for single captures of Northern
Waterthrush during the day (n ? 229) revealed
signi?cant effects of body size, time of capture, year,
and a signi?cant habitat3 season interaction (Table 1).
For the year effect, birds in 2003 (least-squares mean,
15.20 6 0.114 g) were signi?cantly lighter overall than
birds from 2002 (least-squares mean, 15.63 6 0.103 g)
and 2004 (least-squares mean, 16.02 6 0.092 g). The
analysis also showed that Northern Waterthrush gained
mass linearly throughout the day. The rate of gain was
similar between habitats (wet habitats, 0.10 g/h, 95% CI,
0.048?0.147; dry habitats, 0.07 g/h, 95% CI, 0.057?
0.083).
The habitat 3 season interaction term revealed that
differences in mass between wet and dry habitats became
highly signi?cant over the season (Fig. 3). At the
beginning of the dry season, birds in both habitats had
similar masses but by mid-March, at the peak of the dry
season, body mass of birds occupying wet habitats had
increased (early, least-squares mean, 15.48 6 0.15 g;
late, least-squares mean, 16.05 6 0.13 g), whereas birds
in dry habitats declined in mass during the progression
of the season (early, least-squares mean, 15.606 0.083 g;
late, least-squares mean, 15.33 6 0.12 g).
FIG. 2. (A) The relationship between stable carbon isotope
values from blood plasma of Northern Waterthrush and change
in body mass between the early and late dry season. (B) The
relationship between satellite-derived wetness of Northern
Waterthrush home ranges and change in body mass between
the early and late dry season. The shaded sections of the plots
indicate drier habitats identi?ed using satellite imagery and
stable carbon isotopes. See Methods: Home range moisture for
an explanation of the wetness index.
TABLE 1. Results from ANCOVA of factors predicting mass
variation in the Northern Waterthrush (Seiurus novebora-
censis) in Puerto Rico.
Source of variation
Sum of
squares df F P
Habitat (wet or dry) 3.99 1, 222 6.05 0.015
Body size 42.99 1, 222 65.15 ,0.0001
Time of day (hour) 32.22 1, 222 51.85 ,0.0001
Season (early or late) 1.30 1, 222 1.30 0.26
Year 20.87 2, 222 15.81 ,0.0001
Habitat 3 time of day 1.51 1, 222 2.29 0.13
Habitat 3 season 7.39 1, 222 11.19 0.001
Notes: The dependent variable is Northern Waterthrush
mass (n ? 229); adjusted R2 ? 0.41. Habitat represents a
dichotomy identi?ed using carbon isotopes. Body size is
principal component 1 of a principal components analysis of
wing, tail, and tarsus length. Season represents the early
(January?February) vs. late (March?April) dry season.
FIG. 3. Northern Waterthrush mass corrected for body size
and time of day (least-squares mean 6 SE) as an estimate of
physical condition during early (January?February) and late
(March?April) dry-season periods.
October 2010 2879MOISTURE AND WINTERING MIGRATORY BIRDS
Rates of fat deposition
Based on repeated captures of individuals, we found
that rates of premigratory fat deposition just prior to
spring migration to the breeding grounds from mid-
March to early April varied by habitat. Northern
Waterthrush using dry habitats (n ? 12) signi?cantly
delayed fat deposition (repeated-measures ANOVA, F?
23.26, df?1, 11, P, 0.0001; late-March mean fat score,
0.56 6 0.20; early-April mean fat score, 1.21 6 0.22)
compared with those using wet habitats (n ? 25; late-
March mean fat score, 1.34 6 0.17; early-April mean fat
score, 2.15 6 0.21).
DISCUSSION
Seasonal changes in moisture availability characterize
tropical biomes throughout the world, and the manner
in which animals, resident and migratory, respond and
cope with such changes is poorly understood (Carson
and Schnitzer 2008). Here, using radio tracking, along
with stable carbon isotope and satellite wetness indices,
we demonstrate that moisture determined food avail-
ability and consequent overwinter condition of a wet-
habitat specialist, the Northern Waterthrush.
Our ?nding that males were signi?cantly heavier than
females regardless of habitat type provides evidence that
there may be competition for high-quality home ranges.
Males in both wet and dry habitats showed more
positive rates of mass change over the course of the dry
season when compared to females in the same habitat.
This suggests that along the wet?dry continuum within
these habitats, males may be securing higher quality
home ranges. This is consistent with sex-speci?c patterns
of spatial and social behavior we observed in a
concurrent study (Smith 2008) in which we found that
birds showing high levels of aggression (predominantly
males) are more likely to be territorial and that birds
maintaining territories occupied wetter sites with higher
food availability than birds that used alternative space
use strategies. The larger body size of males we observed
may offer a competitive advantage in effort to secure
and defend high-quality home ranges that could be
limited in availability.
Overnight mass trajectories of Northern Waterthrush
declined at equal rates and increased during the day,
with a trend toward a slower rate of gain in drier
habitats. Body mass loss overnight was on average 0.12
g/h, representing ;7.6% of total mass. This is similar to
that observed for other small passerines (e.g., Parus
spp., House Sparrow) wintering in temperate climates
(Clark 1979, Lehikoinen 1987) but to the best of our
knowledge this is the ?rst estimate of overnight mass
loss for a Neotropical migrant on tropical nonbreeding
grounds. That birds using both wet and dry habitats
during the day experienced similar rates of overnight
mass loss while using the same roosting habitat (Smith et
al. 2008) suggests that the metabolic constraints of
roosting are relatively constant regardless of diurnal
habitat. Thus, the differences we detected in seasonal
mass change between wet and dry habitats are due to
patterns of diurnal habitat occupancy and the food
availability within each of those habitats.
Although we did not observe a signi?cant difference
in the rate of mass gain during the day between wet and
dry habitats, the slope estimate of diurnal mass gain was
less steep in dry habitats. The strong seasonal differences
we observed in mass change suggest that each day
presents a challenge to recover the mass lost during the
previous evening. Even slight differences in rates of gain
between habitats result in daily shortfalls in mass
recovery that, over days and weeks, accumulate to
become signi?cant season-long declines in body mass, as
we observed in Northern Waterthrush that occupied
drier habitats. Habitat-speci?c differences in seasonal
mass change have been reported for other Neotropical
migrant species such as American Redstart (Setophaga
ruticilla), Prairie Warbler (Dendroica discolor), Cape
May Warbler (Dendroica tigrina), and Ovenbird (Seiurus
aurocapilla) (Marra et al. 1998, Strong and Sherry 2000,
Latta and Faaborg 2001, 2002, Studds and Marra 2005,
Brown and Sherry 2006, Johnson et al. 2006), but ours is
the ?rst to link habitat moisture with habitat quality of
home ranges to body condition at both daily and
seasonal timescales.
We further found evidence that winter habitat
occupancy may have in?uenced migration schedules of
individuals. Rates of fattening differed signi?cantly
between wet and dry habitats, suggesting that before
building fat reserves, birds in dry habitats ?rst need to
recover from season-long de?cits of fat-free mass
resulting from protein catabolism of muscle tissue. We
did not document such a mechanism in this study but
experimental results from several species of migratory
passerines indicate that a disproportionate amount of
protein mass loss under food-limited conditions is from
digestive organs (Karasov et al. 2004). Smooth muscle
mass of the digestive tract may need to be rebuilt before
storage of fat reserves for migration progresses (Gannes
2002, Karasov et al. 2004, Pierce and McWilliams 2004).
Delays in fattening likely contributed to a delayed
departure and arrival on breeding areas and possibly to
lower reproductive success, similar to what has been
demonstrated in the American Redstart (Marra et al.
1998).
Rainfall and its impact on moisture availability to
plants and resulting arthropod productivity varies
annually (Dugger et al. 2004, Studds and Marra 2007).
Since the Northern Waterthrush is an extreme-wet-
habitat specialist, this species may be especially sensitive
to even subtle deviations of annual precipitation and
may be impacted heavily during El Nin?o/La Nin?a cycles
(Giannini et al. 2000). Given that waterthrush forage
primarily at the land?water edge, sites in anomalous
years could be either too wet (i.e., under water) or too
dry for optimal foraging. This could in part explain our
result that birds in 2003 had signi?cantly lower mass
than the other two years of the study. Interestingly, 2003
JOSEPH A. M. SMITH ET AL.2880 Ecology, Vol. 91, No. 10
corresponded with modest El Nin?o conditions, while
2002 and 2004 were normal years (NOAA, available
online).4 This pattern may be evident because annual
climate variation induces corresponding variation in the
availability of high-quality habitats.
Of particular concern are predictions that levels of
rainfall in the Caribbean basin will signi?cantly decline
over the next 50 years (Neelin et al. 2006, Rauscher et al.
2008). Such changes will have devastating effects on
extant vegetative communities and their fauna. Our
?nding that a satellite-derived index of habitat moisture
corresponds to an underlying gradient of habitat quality
is a ?rst step in developing tools to remotely map habitat
quality over large areas and multiple time periods to
make predictions about how changing conditions may
affect populations. Such an advance would play an
important role in our attempts to understand and
predict how declining precipitation may affect this
species and possibly others at broader spatial scales.
Future research on climate change should emphasize not
only how plant and animal species adapt to increasing
temperature but also adapt to a drier future, especially
in the Caribbean basin.
ACKNOWLEDGMENTS
This research was supported by a grant from the DoD
Legacy Resources Management Program. We thank Daniel
Brown, Ram Papish, Ryan Peters, Mark Pollock, T. J.
Robinson, Jared Woodcock, Alicia Byrd, Sara Campbell, Bill
Deluca, Brian Gibbons, and Karin Roux for their excellent and
dedicated work in the ?eld. We thank Winston Martinez and
Oscar Diaz for facilitating our work at Roosevelt Roads Naval
Station and the U.S. Forest Service Sabana Field Station and
its staff for providing lodging and laboratory space. Thanks to
William Gould and Gary Potts for help with acquiring imagery.
Lee Weight at the Smithsonian Museum Support Center
provided laboratory space and indispensable guidance during
analysis of blood samples. We thank two anonymous reviewers
for their thoughtful comments on the manuscript. Finally we
thank Jean Lodge and Joe Wunderle for all of the assistance
and hospitality they extended during our stay in Puerto Rico.
LITERATURE CITED
Bale, J. S., et al. 2002. Herbivory in global climate change
research: direct effects of rising temperature on insect
herbivores. Global Change Biology 8:1?16.
Brown, D. R., and T. W. Sherry. 2006. Food supply controls
the body condition of a migrant bird wintering in the tropics.
Oecologia 149:22?32.
Carson, W., and S. Schnitzer. 2008. Tropical forest community
ecology. Wiley-Blackwell, Oxford, UK.
Clark, G. A. 1979. Body weights of birds: a review. Condor 81:
193?202.
Crist, E. P., and R. C. Cicone. 1984. A physically-based
transformation of thematic mapper data: the TM tasseled
cap. IEEE Transactions on Geoscience and Remote Sensing
22:256?263.
Crist, E. P., and R. J. Kauth. 1986. The tasseled cap de-
mysti?ed. Photogrammetric Engineering and Remote Sens-
ing 52:81?86.
Daly, C., E. H. Helmer, and M. Quinones. 2003. Mapping the
climate of Puerto Rico, Vieques and Culebra. International
Journal of Climatology 23:1359?1381.
Dugger, K. M., J. Faaborg, W. J. Arendt, and K. A. Hobson.
2004. Understanding survival and abundance of overwinter-
ing warblers: Does rainfall matter? Condor 106:744?760.
Dunham, K. M., E. F. Robertson, and C. C. Grant. 2004.
Rainfall and the decline of a rare antelope, the tsessebe
(Damaliscus lunatus lunatus), in Kruger National Park, South
Africa. Biological Conservation 117:83?94.
Eaton, S. W. 1995. Northern Waterthrush (Seiurus novebor-
acensis). In A. Poole, editor. The birds of North America.
Academy of Natural Sciences, Philadelphia, Pennsylvania,
USA.
Ehleringer, J. R., and C. B. Osmond. 1989. Stable isotopes.
Pages 281?290 in R. W. Pearcy, J. R. Ehleringer, H. A.
Mooney, and P. W. Rundel, editors. Plant physiological
ecology: ?eld methods and instrumentation. Chapman and
Hall, London, UK.
Ewel, J. J., and J. L. Whitmore. 1973. The ecological life zones
of Puerto Rico and the U.S. Virgin Islands. U.S. Forest
Service Research Paper ITF-18. Institute of Tropical
Forestry, Rio Piedras, Puerto Rico.
Farquhar, G. D., J. R. Ehleringer, and K. T. Hubick. 1989.
Carbon isotope discrimination and photosynthesis. Annual
Review of Plant Biology 40:503?537.
Gannes, L. Z. 2002. Mass change pattern of blackcaps refueling
during spring migration: evidence for physiological limita-
tions to food assimilation. Condor 104:231?239.
Giannini, A., Y. Kushnir, and M. A. Cane. 2000. Interannual
variability of Caribbean rainfall, ENSO, and the Atlantic
Ocean. Journal of Climate 13:297?311.
Grif?ths, R., M. C. Double, K. Orr, and R. J. G. Dawson.
1998. A DNA test to sex most birds. Molecular Ecology 7:
1071?1075.
Hobson, K. A., and R. G. Clark. 1993. Turnover of 13C in
cellular and plasma fractions of blood: implications for
nondestructive sampling in avian dietary studies. Auk 110:
638?641.
Hobson, K. A., H. L. Gibbs, and M. L. Gloutney. 1997.
Preservation of blood and tissue samples for stable-carbon
and stable-nitrogen isotope analysis. Canadian Journal of
Zoology 75:1720?1723.
Holdridge, L. R. 1967. Life zone ecology. Tropical Science
Center, San Jose, Costa Rica.
Holmes, R. T., T. W. Sherry, and L. Reitsma. 1989.
Population-structure, territoriality and overwinter survival
of 2 migrant warbler species in Jamaica. Condor 91:545?561.
Hoysak, D. J., and P. J. Weatherhead. 1991. Sampling blood
from birds: a technique and an assessment of its effect.
Condor 93:746?752.
ITT Corporation. 2006. ENVI. Version 4.6. ITT Corporation,
White Plains, New York, USA.
Janzen, D. H. 1967. Why mountain passes are higher in the
tropics. American Naturalist 101:233.
Janzen, D. H., and T. W. Schoener. 1968. Differences in insect
abundance and diversity between wetter and drier sites
during a tropical dry season. Ecology 49:96?110.
Johnson, C. N., and A. P. McIlwee. 1997. Ecology of the
northern bettong, Bettongia tropica, a tropical mycophagist.
Wildlife Research 24:549?559.
Johnson, M. D., T. W. Sherry, R. T. Holmes, and P. P. Marra.
2006. Assessing habitat quality for a migratory songbird
wintering in natural and agricultural habitats. Conservation
Biology 20:1433?1444.
Johnson, M. D., and A. M. Strong. 2000. Length?weight
relationships of Jamaican arthropods. Entomological News
111:270?281.
Karasov, W. H., B. Pinshow, J. M. Starck, and D. A?k. 2004.
Anatomical and histological changes in the alimentary tract
of migrating blackcaps (Sylvia atricapilla): a comparison
among fed, fasted, food-restricted, and refed birds. Physio-
logical and Biochemical Zoology 77:149?160.4 hwww.elnino.noaa.govi
October 2010 2881MOISTURE AND WINTERING MIGRATORY BIRDS
Karr, J. R., and K. E. Freemark. 1983. Habitat selection and
environmental gradients: dynamics in the ??stable?? tropics.
Ecology 64:1481?1494.
Latta, S. C., and J. Faaborg. 2001. Winter site ?delity of prairie
warblers in the Dominican Republic. Condor 103:455?468.
Latta, S. C., and J. Faaborg. 2002. Demographic and
population responses of Cape May warblers wintering in
multiple habitats. Ecology 83:2502?2515.
Lefebvre, G., B. Poulin, and R. McNeil. 1992. Abundance,
feeding-behavior, and body condition of Nearctic warblers
wintering in Venezuelan mangroves. Wilson Bulletin 104:
400?412.
Lefebvre, G., B. Poulin, and R. McNeil. 1994. Spatial and
social-behavior of Nearctic warblers wintering in Venezuelan
mangroves. Canadian Journal of Zoology 72:757?764.
Lehikoinen, E. 1987. Seasonality of the daily weight cycle in
wintering passerines and its consequences. Ornis Scandinav-
ica 18:216?226.
Madsen, T., B. Ujvari, R. Shine, and M. Olsson. 2006. Rain,
rats and pythons: climate-driven population dynamics of
predators and prey in tropical Australia. Austral Ecology 31:
30?37.
Marra, P. P., K. A. Hobson, and R. T. Holmes. 1998. Linking
winter and summer events in a migratory bird by using
stable-carbon isotopes. Science 282:1884?1886.
Marra, P. P., and R. T. Holmes. 2001. Consequences of
dominance-mediated habitat segregation in American Red-
starts during the nonbreeding season. Auk 118:92?104.
Murphy, P. G., and A. E. Lugo. 1986. Ecology of tropical dry
forest. Annual Review of Ecology and Systematics 17:67?88.
Neelin, J. D., M. Mu?nnich, H. Su, J. E. Meyerson, and C. E.
Holloway. 2006. Tropical drying trends in global warming
models and observations. Proceedings of the National
Academy of Sciences USA 103:6110?6115.
Norris, D. R., P. P. Marra, T. K. Kyser, T. W. Sherry, and
L. M. Ratcliffe. 2004. Tropical winter habitat limits
reproductive success on the temperate breeding grounds in
a migratory bird. Proceedings of the Royal Society B 271:59?
64.
O?Leary, M. H. 1981. Carbon isotope fractionation in plants.
Phytochemistry 20:553?567.
Ordoyne, C., and M. Friedl. 2008. Using MODIS data to
characterize seasonal inundation patterns in the Florida
Everglades. Remote Sensing of Environment 112:4107?4119.
Peterson, B. J., and B. Fry. 1987. Stable isotopes in ecosystem
studies. Annual Review of Ecology and Systematics 18:293?
320.
Pierce, B. J., and S. R. McWilliams. 2004. Diet quality and food
limitation affect the dynamics of body composition and
digestive organs in a migratory songbird (Zonotrichia
albicollis). Physiological and Biochemical Zoology 77:471?
483.
Post, W. 1978. Social and foraging behavior of warblers
wintering in Puerto Rican coastal scrub. Wilson Bulletin
90:197?214.
Poulin, B., G. Lefebvre, and R. McNeil. 1994. Diets of land
birds from northeastern Venezuela. Condor 96:354?367.
Pyle, P. 1997. Identi?cation guide to North American birds.
Part 1. Slate Creek Press, Bolinas, California, USA.
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.
Rauscher, S., F. Giorgi, N. Diffenbaugh, and A. Seth. 2008.
Extension and intensi?cation of the meso-American mid-
summer drought in the twenty-?rst century. Climate Dy-
namics 31:551?571.
Reitsma, L., P. Hunt, S. L. Burson, and B. B. Steele. 2002. Site
?delity and ephemeral habitat occupancy: Northern Water-
thrush use of Puerto Rican black mangroves during the
nonbreeding season. Wilson Bulletin 114:99?105.
Reudink, M. W., P. P. Marra, T. K. Kyser, P. T. Boag, K. M.
Langin, and L. M. Ratcliffe. 2009. Non-breeding season
events in?uence sexual selection in a long-distance migratory
bird. Proceedings of the Royal Society B 276:1619?1626.
Runge, M. C., and P. P. Marra. 2005. Modeling seasonal
interactions in the population dynamics of migratory birds.
Pages 375?389 in R. Greenberg and P. P. Marra, editors.
Birds of two worlds: the ecology and evolution of migration.
Johns Hopkins University Press, Baltimore, Maryland, USA.
Schwartz, P. 1964. The northern waterthrush in Venezuela.
Living Bird 3:169?184.
Sherry, T. W., and R. T. Holmes. 1996. Winter habitat quality,
population limitation, and conservation of Neotropical
nearctic migrant birds. Ecology 77:36?48.
Smith, J. 2008. The non-breeding ecology of a Neotropical
migratory songbird: the in?uence of habitat quality on
spatial and social behavior and consequences for individual
performance. Dissertation. George Mason University, Fair-
fax, Virginia, USA.
Smith, J. A. M., L. R. Reitsma, and P. P. Marra. 2008.
Roosting behavior of a Neotropical migrant songbird, the
northern waterthrush Seiurus noveboracensis, during the non-
breeding season. Journal of Avian Biology 39:460?465.
Stangel, P. W. 1986. Lack of effects from sampling blood from
small birds. Condor 88:244?245.
Strong, A. M., and T. W. Sherry. 2000. Habitat-speci?c effects
of food abundance on the condition of ovenbirds wintering in
Jamaica. Journal of Animal Ecology 69:883?895.
Studds, C. E., and P. P. Marra. 2005. Nonbreeding habitat
occupancy and population processes: an upgrade experiment
with a migratory bird. Ecology 86:2380?2385.
Studds, C. E., and P. P. Marra. 2007. Linking ?uctuations in
rainfall to nonbreeding season performance in a long-
distance migratory bird, Setophaga ruticilla. Climate Re-
search 35:115?122.
Thomlinson, J. R., M. I. Serrano, T. d. M. Lopez, T. M. Aide,
and J. K. Zimmerman. 1996. Land-use dynamics in a post-
agricultural Puerto Rican landscape (1936?1988). Biotropica
28:525?536.
Toomey, M., and L. A. Vierling. 2005. Multispectral remote
sensing of landscape level foliar moisture: techniques and
applications for forest ecosystem monitoring. Canadian
Journal of Forest Research 35:1087?1097.
Valeix, M., H. Fritz, S. Chamaille-Jammes, M. Bourgarel, and
F. Murindagomo. 2008. Fluctuations in abundance of large
herbivore populations: insights into the in?uence of dry
season rainfall and elephant numbers from long-term data.
Animal Conservation 11:391?400.
White, T. C. R. 2008. The role of food, weather and climate in
limiting the abundance of animals. Biological Reviews 83:
227?248.
Wunderle, J. M., and S. C. Latta. 2000. Winter site ?delity of
Nearctic migrants in shade coffee plantations of different
sizes in the Dominican Republic. Auk 117:596?614.
Wunderle, J. M., and R. B. Waide. 1993. Distribution of
overwintering Nearctic migrants in the Bahamas and
Greater-Antilles. Condor 95:904?933.
JOSEPH A. M. SMITH ET AL.2882 Ecology, Vol. 91, No. 10