ORIGINAL PAPER
Water-use efficiency and whole-plant performance of nine
tropical tree species at two sites with contrasting water
availability in Panama
D. Craven ? J. S. Hall ? M. S. Ashton ? G. P. Berlyn
Received: 16 August 2012 / Revised: 16 November 2012 / Accepted: 19 November 2012 / Published online: 12 December 2012
 Springer-Verlag Berlin Heidelberg 2012
Abstract Across their natural distributions, tropical tree
species are regularly exposed to seasonal droughts of
varying intensities. Their ability to tolerate drought stress
plays a vital role in determining growth and mortality rates,
as well as shaping the functional composition of tropical
forests. In order to assess the ability of species to acclimate
to contrasting levels of drought stress, physiological and
structural traits involved in drought adaptation?wood C
isotope discrimination (d13C), wood specific gravity, and
wood C content?of 2-year-old saplings of nine tropical
tree species were evaluated in common garden experiments
at two study sites in Panama with contrasting seasonality.
We assessed co-variation in wood traits with relative
growth rates (RGRBD), aboveground biomass, and basal
diameter and the plasticity of wood traits across study sites.
Overall, species responded to lower water availability by
increasing intrinsic water-use efficiency, i.e., less negative
wood d13C, but did not exhibit a uniform, directional
response for wood specific gravity or wood C content. Trait
plasticity for all wood traits was independent of RGRBD
and tree size. We found that the adaptive value of intrinsic
water-use efficiency varied with water availability. Intrin-
sic water-use efficiency increased with decreasing RGRBD
at the more seasonal site, facilitating higher survival of
slower growing species. Conversely, intrinsic water-use
efficiency increased with tree size at the less seasonal site,
which conferred a competitive advantage to larger indi-
viduals at the cost of greater susceptibility to drought-
induced mortality. Our results illustrate that acclimation to
water availability has negligible impacts on tree growth
over short periods, but eventually could favor slow-grow-
ing species with conservative water-use strategies in trop-
ical regions experiencing increasingly frequent and severe
droughts.
Keywords Carbon isotope  Wood traits  Wood density 
WUE  Plasticity  Functional traits  Central America
Introduction
Across tropical regions, global change models predict that
droughts will become increasingly frequent and intense
(Cox et al. 2008; Williams et al. 2007). Droughts reduce
tree growth and increase mortality rates, thereby shifting
the competitive advantage toward species with greater
drought stress tolerance (Brenes-Arguedas et al. 2011;
Engelbrecht and Kursar 2003; Kursar et al. 2009; Phillips
et al. 2010). Consequently, changing climatic conditions
threaten to alter species? distributions, as well as the
functional composition of tropical forests (Condit et al.
2004; Davidar et al. 2007; Phillips et al. 2009, 2010).
Plant water-use efficiency, the amount of water used per
carbon gain, explicitly links plant performance with water
availability. At the leaf level, intrinsic water-use efficiency
is expressed as the balance between photosynthetic C fix-
ation (A) and stomatal conductance (gs), which is corre-
lated with the ratio of intercellular to ambient CO2 partial
Communicated by A. Gessler.
D. Craven (&)  M. S. Ashton  G. P. Berlyn
School of Forestry and Environmental Studies, Yale University,
370 Prospect St., New Haven, CT 06511, USA
e-mail: dylan.craven@aya.yale.edu
D. Craven  J. S. Hall
Native Species Reforestation Project (PRORENA),
Center for Tropical Forest Science, Smithsonian Tropical
Research Institute, Av. Roosevelt 401, Balboa,
Anco?n, Panama?, Republic of Panama?
123
Trees (2013) 27:639?653
DOI 10.1007/s00468-012-0818-0
pressures (Ci/Ca) in C3 plants. Therefore, time-integrated,
intrinsic water-use efficiency can be inferred using stable
carbon isotope ratios (d13C) of plant tissues given its
inverse linear relationship with Ci/Ca, whereby high water-
use efficiency is indicated by less negative d13C and low
Ci/Ca and vice versa (Dawson et al. 2002; Farquhar et al.
1982; Farquhar and Richards 1984). Due to interspecific
and temporal variation in mesophyll conductance, intrinsic
water-use efficiency (A/gs) and Ci/Ca can vary indepen-
dently of plant d13C due to interspecific and temporal
variation in mesophyll conductance, which is sensitive to
shifts in environmental conditions (Bonal et al. 2007;
Cernusak et al. 2008, 2009b; Seibt et al. 2008). While foliar
d13C has been used extensively to investigate plant
response to precipitation, seasonal drought, and soil nutri-
ents within and across species and ecosystems (e.g., Ares
and Fownes 1999; Cernusak et al. 2007, 2009b; Craven
et al. 2007, 2010; Diefendorf et al. 2010; Ram??rez-Valiente
et al. 2010), there is growing interest in using wood d13C to
relate patterns of water-use efficiency over multiple years
to variation in tree growth, precipitation, ambient CO2, and
temperature (Brienen et al. 2011; Hietz et al. 2005; Nock
et al. 2011; Rozendaal and Zuidema 2011; Schulze et al.
2006) as wood d13C records the d13C of phloem sap carbon
delivered to the vascular cambium during wood formation
(Cernusak et al. 2005, 2009a).
The strength and direction of the relationship between
water-use efficiency and plant performance can illustrate
interspecific differences in drought tolerance strategies,
ranging from stress tolerance to stress avoidance (see
Aranda et al. 2012; Chaves et al. 2002 and Nicotra et al.
2010 for a more complete discussion of tree responses to
drought). Stress tolerance is associated with higher water-
use efficiency, lower photosynthetic rates, slower growth,
and higher survival, while stress avoidance is associated
with lower water-use efficiency, higher photosynthetic
rates, faster growth, and lower survival (Chaves et al. 2002)
As plant performance underlies variation in life history
strategies (Wright et al. 2010), drought tolerance strategies
also would be expected to vary in coordination with life
history strategies, such that the continuum from pioneer to
late-successional species would parallel that of drought
tolerance strategies, from stress avoidance to stress toler-
ance (Llamb?? et al. 2003).
Wood specific gravity is utilized extensively as a proxy
to capture differences in life history strategies, relative
growth rates, and survival rates of tropical tree species
(Chave et al. 2006, 2009; Kitajima 1994; Wright et al.
2010). These differences in plant performance reflect spe-
cies-specific responses to myriad abiotic and biotic factors,
including water and light availability, soil nutrients,
mechanical damage, pathogens, and disturbance (Chave
et al. 2006; Fan et al. 2012; Kitajima 1994; Muller-Landau
2004; ter Steege and Hammond 2001). Despite the lack of
a consistent, community-level relationship between wood
specific gravity and mean annual precipitation across
Neotropical forests (Muller-Landau 2004), coordination of
wood specific gravity with hydraulic conductivity within
tropical forest communities supports the vital role played
by wood specific gravity in determining drought tolerance
(Markesteijn et al. 2011; Poorter et al. 2010). At the plant
level, variation in wood specific gravity represents varia-
tion in the arrangement and distribution of wood anatom-
ical traits, principally vessels and fibers, which underlie
species? capacity to tolerate the negative pressures in the
xylem associated with drought stress (Chave et al. 2009;
Fan et al. 2012; Hacke et al. 2001; Poorter et al. 2010;
Zanne et al. 2010). The denser spacing of vessels, thicker
vessel walls, greater amount of fiber cells, and greater
thickness of fiber walls found in tree species with high
wood specific gravity suggest a proportionally greater
investment in C-rich tissues (Mart??nez-Cabrera et al. 2009;
Swenson and Enquist 2007), although this hypothesized
relationship has not been supported empirically for tropical
tree species (Martin and Thomas 2011).
The extent to which trees acclimate to their local envi-
ronment, also known as plasticity (Bradshaw 1965), can
confer a competitive advantage to co-occurring individuals
as they compete for limiting resources. Plasticity is thought
to indicate specialization to a particular environment
(Niinemets 2010; Valladares et al. 2007). Hence, plasticity
can be used to identify drought tolerance strategies and
environments in which different strategies are adaptive
(Nicotra et al. 2010; Valladares et al. 2000). Variation in
plasticity among leaf traits suggests that different traits are
adaptive in different light environments (Hulshof and
Swenson 2010; Rozendaal et al. 2006; Valladares et al.
2000). For example, pioneer species (in some cases) have a
higher plasticity in leaf traits related to photosynthesis, while
shade-tolerant species have higher plasticity in leaf traits
related to light interception (Rozendaal et al. 2006). In the
context of drought stress, therefore, species with high trait
plasticity would be expected to exhibit greater performance
where water was not limiting, while species with low trait
plasticity would be expected to perform equally well in
environments with and without drought stress (Comita and
Engelbrecht 2009). The extent to which wood traits accli-
mate to changes in drought stress remains undetermined for
tropical tree species, which could have important implica-
tions for the species and functional composition of tropical
forests (Baltzer et al. 2009; Phillips et al. 2010).
The principal aims of this study were (a) to evaluate
wood d13C in relation to whole-plant performance and life
history strategies at two study sites with contrasting annual
precipitation regimes and dry season lengths and (b) to
determine the extent to which saplings of tropical tree
640 Trees (2013) 27:639?653
123
species acclimate wood d13C in response to differences in
water availability between study sites. Nine tropical tree
species were investigated in a common garden experiment
replicated at two study sites with contrasting seasonality,
co-variation in wood d13C, wood specific gravity (a proxy
for life history strategies), and wood C content with whole-
plant performance traits (relative growth rates, basal
diameter, and aboveground biomass). Specifically, the
following three questions were addressed in the present
study:
1. To what extent do wood traits vary across species and
between sites that have contrasting annual precipita-
tion regimes and dry season lengths?
2. How plastic are wood traits between sites with
contrasting annual precipitation regimes and dry
season lengths? Is plasticity of wood traits correlated
with life history strategies and whole-plant
performance?
3. Are wood and tree traits coordinated similarly at sites
with contrasting annual precipitation regimes and dry
season lengths?
Materials and methods
Study sites
The study was conducted on 2-year-old saplings at two
sites in Panama, Rio Hato (Cocle? Province) and Las Lajas
(Chiriqu?? Province). The study sites are located in different
ecological life zones along the Pacific coast of western
Panama: tropical wet (Las Lajas) and tropical dry forest
(Rio Hato) (PIDP 1970). Rio Hato receives less annual
precipitation and has a longer, more severe dry season than
Las Lajas (Table 1; Hijmans et al. 2005). While the study
sites have similar mean annual potential evapotranspiration
and mean annual temperatures, the lower aridity index at
Rio Hato indicates that less precipitation is available for
plant uptake relative to Las Lajas (Zomer et al. 2008). The
study sites have comparable soil characteristics, as both
have similar soil pH and concentrations of total P, total N,
and K (Table 1). At both sites, the natural vegetation was
converted to agricultural fields and cattle pastures, which
were managed actively until the establishment of the
experiments. From here on, Las Lajas will be referred to as
the less seasonal site and Rio Hato as the more seasonal
site.
Species selection
The selected tree species were chosen to represent a range
of differing successional statuses and life history strategies
(Table 2). All species occur naturally in Panama and are
used locally for their timber and firewood, as well as in
silvopastoral systems (Aguilar and Condit 2001; Perez and
Condit 2011). Of the studied species, Dalbergia retusa,
Ormosia macrocalyx, and Zygia longifolia have been
observed to nodulate and likely fix nitrogen (Bryan et al.
1996; de Faria and de Lima 1998; Moreira et al. 1992; Tilki
and Fisher 1998).
Experimental design
Two identical common garden experiments were estab-
lished at both study sites in 2004 (Breugel et al. 2011;
Wishnie et al. 2007). Seeds were collected at three sites
across Panama and, after germination, were cared for in a
nursery for 2?6 months prior to being transplanted. At each
study site, tree seedlings were planted in three completely
Table 1 Climatic and soil characteristics for the two study sites in
Panama
Study site Las Lajas Rio Hato
Site coordinates 8140N
81 520W
8220N
80 090W
AP (mm yr-1) 3375 1172
AP3 (ffim yr-1) 4559.7 (120.6) 1397.9 (103.3)
ET (mm yr-1) 1652 1564
MAT (C) 26.4 27.1
Aridity index 2.048 0.755
Dry season length (months yr-1) 3.33 (0.67) 5.00 (0.00)
pH(in H2O) 4.77 (0.5) 5.38 (0.9)
Total N (%) 0.18 (0.01) 0.08 (0.00)
P (ppm) 3.12 (0.45) 2.44 (0.17)
Ca (ppm) 132.00 (24.8) 440.64 (47.33)
K (ppm) 32.14 (5.41) 53.92 (7.86)
Mg (ppm) 38.85 (5.51) 97.94 (7.70)
Climatic data during the study period (2004?2006) were obtained
from weather stations nearby study sites (Empresa de Transmision
Electrica Panamena SA, ETESA). The global potential evapo-tran-
spiration (Global-PET) and Global Aridity Index (Global-Aridity)
datasets (Zomer et al. 2008) are based on a high resolution (1 km2)
global climate dataset (WorldClim) (Hijmans et al. 2005). AP is mean
annual precipitation (1950?2000, source: WorldClim.), AP3 is mean
annual precipitation during the 3 years of study (2004?2006, adapted
from ETESA), ET is mean annual potential evapotranspiration cal-
culated using the Hargreaves method (1950?2000, source: Global-
PET), MAT is mean annual temperature (1950?2000, source:
WorldClim), Aridity Index is the ratio of mean annual precipitation to
mean annual potential evapotranspiration (1950?2000, source: Glo-
bal-Aridity), and dry season length is number of months per year with
less than 100 mm of precipitation during the 3 years of the study
(2004?2006, adapted from ETESA)
Soils were sampled from 0?15 cm and analyzed using a Mehlich 1
extraction
Standard errors are in parenthesis
Trees (2013) 27:639?653 641
123
randomized blocks. Within each block, three replicate
plots, each containing 20 trees planted at 3 9 3 m, were
established for each species. Plots were cleaned mechani-
cally to reduce competition from grasses, herbaceous
plants, and naturally regenerating seedlings.
Whole-plant performance measurements
For measurement of wood traits and aboveground biomass,
four to six individuals were selected per species at each site
for destructive sampling. Using basal diameter measure-
ments taken at the beginning of the wet season in 2006
(June?August), trees were selected using a stratified ran-
dom sample of individuals from three diameter size classes,
corresponding to the 1?33rd, 34?66th, and 67?100th per-
centiles (Oelmann et al. 2010).
Basal diameter was measured on seedlings at the time of
planting and in subsequent years (2005, 2006). Relative
growth (RGRBD) was calculated following Fisher (1921):
RGRBD ?
ln W2  ln W1
t2  t1
where W2 and W1 are basal diameter at t2 (2006) and t1
(2004), respectively. Time period between measurements
was calculated in days.
Selected individuals were re-measured for basal diam-
eter at the time of destructive sampling. Fresh weight was
determined in the field using either an electronic balance
(400 g capacity, 0.1 g precision) or a spring balance (20 kg
capacity, 0.1 kg precision). Fresh weights of sub-samples
of tree components (leaves, branches, and stem) were also
taken and subsequently dried in an oven at 70 C for
4 days until a constant weight was attained. Total dry
weights were estimated by multiplying the wet:dry ratio of
the sub-sample by the corresponding total fresh weight
(Bastien-Henri et al. 2010).
For determination of oven-dry wood specific gravity of
wood, two uncompressed samples were taken using a
2.8 9 1.0 cm borer attached to a power drill from the
stems of all destructively sampled individuals. All bark was
removed from samples prior to drying. Wood samples were
oven dried to a constant weight at 70 C to avoid binding
water to the sample. Samples were weighed to the nearest
0.0001 g using an electronic balance. To prevent rehydra-
tion, dried samples were coated thinly with melted paraffin
wax and submerged in water to estimate dry volume. A
correction factor (0.9982) was applied to oven-dry weights
following Muller-Landau (2004), as samples were not
dried initially at 100?105 C. Wood specific gravity was
calculated by dividing oven-dry mass (g) by oven-dry
volume (cm3) (Fearnside 1997; Williamson and Wiemann
2010). This method was chosen to ensure comparability of
wood specific gravity values across sites, as plant water
status at the time of sampling can influence green volume.
Isotopic and elemental analysis
For wood d13C, two wood samples were taken from all
individuals in the manner described for wood specific
gravity. All samples were ground using a sample mill with
a 40 lm mesh and homogenized and weighed
566.09 ? 4.59 lg (mean ? standard error). Samples were
analyzed using a continuous flow mass spectrometer
(ThermoFinnigan DeltaPlus Advantage, Costech Analyti-
cal Technologies Inc., Valencia, CA, USA). Carbon iso-
tope ratios were calculated as 13C/12C ratio relative to
PeeDee belemnite with an average precision of 0.14 %. At
each site, two leaf samples were also collected per species
and analyzed for carbon isotope ratios, to enable compar-
isons between leaf and wood tissue. Relative to leaves,
wood tissue is typically more enriched in 13C and we
observed a similar pattern for the studied species. On
Table 2 Scientific name, family, successional status, wood specific gravity, N-fixation capacity, and identification code for the nine studied
species
Species Family Succession status Wood specific gravity (g cm-3) N-fixation capacity Species code
Hura crepitans Euphorbiaceae Early 0.38 (0.36?0.42) No Hc
Byrsonima crassifolia Malpighiaceae 0.56 (0.54?0.60) No Bc
Terminalia amazonia Combretaceae 0.62 (0.57?0.65) No Ta
Swietenia macrophylla Meliaceae 0.63 (0.61?0.66) No Sm
Ormosia macrocalyx Fabaceae 0.63 (0.59?0.67) Yes Om
Zygia longifolia Fabaceae 0.67 (0.64?0.69) Yes Zl
Dalbergia retusa Fabaceae 0.70 (0.65?0.73) Yes Dr
Tabebuia impetiginosa Bignoniaceae 0.72 (0.68?0.74) No Ti
Sapindus saponaria Sapindaceae Late 0.75 (0.71?0.74) No Ss
Successional status was determined from a literature review (Perez and Condit 2011), wood specific gravity values are taken from the present
study (95 % confidence intervals in parenthesis), and N-fixation capacity was determined from a literature review (Moreira et al. 1992; de Faria
and de Lima 1998 and Tilki and Fisher 1998)
642 Trees (2013) 27:639?653
123
average, wood was 1.81 % more enriched in 13C than
leaves, which is within the range of values reported in other
studies (Cernusak et al. 2009a). While there are multiple
hypotheses for this pattern (see Cernusak et al. 2009a), we
found a high correlation of d13C between wood and leaf
tissues for the studied species (ordinary least squares
regression, R2 = 89.99 %, p value\0.001), thus indicating
that variation in wood d13C accurately reflected leaf-level
physiological processes during the 2-year period of this
study. Therefore, we use wood d13C as a proxy for time-
integrated, whole-plant intrinsic water-use efficiency in the
present study (Cernusak et al. 2007).
Using the same wood samples as for wood d13C, wood
C content was analyzed using a combustion analyzer (Flash
EA 1112 Series NC Soil Analyzer, Thermo Electronic
Corporation, Waltham, MA, USA). Wood samples
weighed 8.26 ? 0.91 mg. Concentrations were calculated
directly, using atropine as a standard reference material
(70.56 % C).
Wood trait plasticity
For all wood traits, plasticity was calculated to assess
species-specific across-site plasticity. Across-site trait
plasticity represents the expressed variability of a trait in
response to the contrasting seasonality at the two study
sites and was calculated in the following manner, for each
species:
PlasticityTrait ?
mean trait valueDRY  mean trait valueWET
maximum trait mean
where ?dry? represents the more seasonal site and ?wet?
represents the less seasonal site (Rozendaal et al. 2006;
Valladares et al. 2000). The resulting values for plasticity
range from negative to positive one in order to indicate the
direction of plasticity, such that positive values denote
mean trait values being greater at the more seasonal site
and negative values denote mean trait values being greater
at the less seasonal site. As wood d13C values are negative,
the sign of the plasticity values was switched for compar-
ative purposes.
Statistical analysis
Prior to analysis, all wood traits?wood specific gravity,
wood d13C, and wood C content?and whole-plant per-
formance traits?basal diameter, aboveground biomass,
and RGRBD?were evaluated for normality and homoge-
neity of variance across groups. All whole-plant perfor-
mance traits were natural-log transformed and wood C
content was arc-sine transformed to meet normality
assumptions (Warton and Hui 2011). All variables had
homogeneous variance across groups, which were
evaluated with Levene?s test (p value[0.05). For all wood
and tree traits, two-way ANOVAs with Type III sum of
squares were used to test for differences between species,
site, and all interactions. The proportion of the explained
variance was calculated by dividing the sum of squares
separately of each main factor and the interaction of both
main factors by the total sum of squares of the model
(Quero et al. 2006). Welch?s two-sample t tests were per-
formed separately for all wood and tree traits for each
species. For all wood traits, site-specific averages and 95 %
confidence intervals were calculated from 10,000 boot-
strapped replicate samples, using the bias-adjusted cor-
rection method (Venables and Ripley 2002). Estimates of
mean trait plasticity and its corresponding 95 % confidence
intervals were calculated for each wood trait in the same
manner.
To determine the degree to which species-specific wood
trait values at one site correlated with those at the other
site, major axis (MA) regression was performed for each
wood trait as both variables contained measurement error
(Warton et al. 2006). The null expectation of no across-site
plasticity was that the slope of the regression would be 1
and the elevation (sample mean of residual scores) would
be 0. The null hypothesis was rejected if (a) the trait values
were not significantly correlated or if (b) the slope and
intercept varied significantly from 1 and 0, respectively.
The relationships between plasticity of wood traits and
whole-plant performance traits were assessed by Pearson?s
correlation coefficients. MA regression analysis was per-
formed using the R package ?smatr?. Relationships between
wood characteristics and tree traits at each study were
evaluated using Pearson?s correlation coefficients. As
wood d13C values are negative (range 23.5 % to
-28.8 %), they were natural-log transformed after adding
an offset of 30 to make values positive. All analyses were
performed with R 2.14.1 (R Development Core Team
2012).
Results
Interspecific variation in whole-plant performance
and wood traits across sites
Across the nine studied tree species, differences in growth
were strongly determined by species identity (Fig. 1;
Table 3). For basal diameter and aboveground biomass,
between 34 and 38 % of model variation was explained by
species identity, respectively. Although statistically sig-
nificant, site was practically inconsequential in explaining
variation in aboveground biomass and basal diameter.
Significant species 9 site interactions for both variables
showed that species did not respond uniformly to
Trees (2013) 27:639?653 643
123
contrasting environmental conditions at the study sites.
Specifically, sampled individuals of Byrsonima crassifolia
and Terminalia amazonia exhibited significantly greater
basal diameter (Welch?s two-sample T test, p values\0.05,
Fig. 1) and aboveground biomass (Welch?s two-sample
T test, p values\0.05, Fig. 1) at the less seasonal site than
at the more seasonal site. A high proportion of model
variation (47 %) also was explained by species identity for
RGRBD, which did not differ at statistically significant
levels between sites or for the interaction of species and
site (Table 3). While most species grew at similar rates at
both study sites (Fig. 1), Zygia longifolia and Tabebuia
impetiginosa had statistically significant higher RGRBD at
the more seasonal site than at the less seasonal site
(Welch?s two-sample T test, p values \0.05, Fig. 1).
To a greater extent than whole-plant performance traits,
variation in all wood traits was concentrated at the species
level (Table 3). Interspecific differences in wood traits
explained between 35 and 76 % of model variability. Site
was not a significant predictor for any of the wood traits
and, consequently, explained a low proportion of model
variability (0?16 %) (Table 3). Still, average wood d13C
across all studied species was significantly higher at the
more seasonal site (mean -26.10 %, 95 % confidence
intervals -25.78, -26.40 %) than at the less seasonal site
(mean -27.06 %, 95 % confidence intervals -26.77,
-27.31 %). For wood d13C, there were significant inter-
actions between species and site, indicating that species
responded differently to contrasting site conditions. Four of
the studied species, D. retusa, Z. longifolia, Sapindus
saponaria, and T. amazonia, exhibited significantly higher
wood d13C at the more seasonal site than at the less seasonal
site (Welch?s two-sample T test, p value\0.05; Fig. 2). For
wood specific gravity and wood C content, fewer species
demonstrated significant differences between sites. Only
T. amazonia had significantly greater wood C content at the
less seasonal site than at the more seasonal site (Welch?s
two-sample T test, p value \0.05; Fig. 2) and only one
species, Sapindus saponaria, had higher wood specific
gravity at the more seasonal site than at the less seasonal site
(Welch?s two-sample T test, p value = 0.05; Fig. 2).
Wood trait plasticity
The studied species demonstrated a low amount of across-
site plasticity for wood traits, with the exception of wood
Fig. 1 Whole-plant performance traits of nine tropical tree species in
Panama. Asterisks indicate statistically significant differences
between sites for each species (Welch?s two-sample T test, p value
\0.05). Black-filled bars correspond to the less seasonal site (Las
Lajas) and white-filled bars to the more seasonal site (Rio Hato).
Whisker bars are standard errors
b
644 Trees (2013) 27:639?653
123
d13C. For wood specific gravity and wood C content, trait
values were strongly and significantly correlated between
sites (range of R2 values: 61.5?85.9 %) (Fig. 3). Further-
more, the slopes and elevations of the across-site regres-
sions for the aforementioned wood traits did not vary
significantly from 1 or 0, respectively. Wood d13C values
were not correlated significantly between sites, indicating
that across-site plasticity was not uniform across the
studied species.
Among wood traits, wood d13C was the most plastic
wood trait between study sites (mean: 0.037, 95 % confi-
dence intervals: 0.02?0.06), followed by wood specific
gravity (mean: 0.0049, 95 % confidence intervals
0.04?0.05) and wood C content (mean 0.001, 95 % confi-
dence intervals 0.007?0.007). Wood d13C was the only
wood trait whose 95 % confidence intervals did not overlap
with zero, which indicates that the acclimation response to
the seasonality gradient was negligible for wood specific
gravity and wood C content. Across-site plasticity for all
wood traits was not correlated significantly with whole-
plant performance traits, nor with wood specific gravity
(p value [0.05).
Relationships between wood traits and whole-plant
performance
Distinct relationships among wood and whole-plant per-
formance traits were found at each study site (Table 4). At
the more seasonal site, wood d13C was negatively corre-
lated with RGRBD and wood C content, while at the less
seasonal site wood d13C only was negatively correlated
with wood specific gravity. In contrast, relationships
among tree growth traits were largely consistent at both
study sites.
Discussion
By assessing young saplings from the same species in
replicated common garden experiments located in sites
with contrasting seasonality, the current study allows us to
evaluate the extent to which water availability modulates
relationships between wood d13C and whole-plant perfor-
mance. We found strong evidence that the adaptive value
of high intrinsic water-use efficiency (less negative wood
d13C) is restricted to environments with low water avail-
ability and that plasticity of wood traits to contrasting
levels of water availability was mostly neutral with regard
to whole-plant performance.
Interspecific variation in wood traits across seasonality
gradient
Overwhelmingly, variation in wood traits was conserved at
the species level. The significant interspecific variation in
wood d13C confirmed expectations that species would
exhibit a differential capacity to compete for water. Across
the tropical regions, interspecific differences in wood and
foliar d13C have been reported extensively (Bonal et al.
2000, 2007; Cernusak et al. 2007; Craven et al. 2007;
Guehl et al. 1998; Huc et al. 1994; Jackson et al. 1996) and
are determined largely by variation in gs and hydraulic
conductivity, not A, related to differences in successional
status, shade tolerance, and leaf phenology. For example,
evergreen species in a tropical dry forest in Mexico com-
pensated for low intrinsic water-use efficiency by using
water from deeper in the soil to avoid seasonal drought
stress, while deciduous species tolerated drought stress by
reducing gs, which increased intrinsic water-use efficiency
(Hasselquist et al. 2010). In the French Guiana, similar
Table 3 Results of linear models for wood and whole-plant performance traits of nine tree species at two study sites in Panama
Wood traits Adjusted R2 (%) Species Site Species x site
SSx/SStotal p value SSx/SStotal p value SSx/SStotal p value
Wood d13C (%) 51.92 0.35 \0.001 0.16 0.722 0.09 0.018
Wood specific gravity (g cm-3) 75.36 0.76 \0.001 0.00 0.280 0.03 0.094
Wood C (%)b 44.04 0.47 \0.001 0.00 0.732 0.06 0.226
Whole-plant performance
Aboveground biomass (g)a 42.0 0.34 \0.001 0.02 0.001 0.16 0.002
Basal diameter (cm)a 39.8 0.38 \0.001 0.00 0.012 0.11 0.021
RGRBD (cm day
-1)a 48.3 0.47 \0.001 0.04 0.794 0.06 0.168
For all variables, two-way ANOVAs using Type III sum of squares were performed (95 % confidence interval, a = 0.05, degrees of freedom for
site = 1 and for species = 8) to test for differences across species, sites, and the combination of the main effects. The proportion of explained
variance (SSx/SStotal) and the p value are provided for each main effect and the combination of both main effects. All variables had homogeneous
variance; all tree traits were natural-log transformed and wood C and wood N were arc-sine transformed to meet normality assumptions. Adjusted
R2 is the proportion of total variance explained by the model, adjusted by the number of terms included in the model
a Natural-log transformed to meet normality assumptions
b Arc-sine transformed to meet normality assumptions
Trees (2013) 27:639?653 645
123
differences were observed for shade-tolerant and intolerant
species, where contrasting hydraulic conductance between
the two functional groups determined variation in d13C
(Bonal et al. 2000). The observed differences in wood d13C
of the studied species likely correspond to different drought
tolerance strategies, whose impacts on whole-plant per-
formance will be discussed below.
Across-site variation of wood d13C was moderate relative
to interspecific variation, yet revealed the capacity for the
studied species to acclimate long-term intrinsic water-use
efficiency to changes in water availability. Average wood
d13C was significantly lower at the more seasonal site,
indicating that the studied species?except for B. crassifolia
and S. macrophylla?acclimated to lower annual precipita-
tion and a longer dry season by increasing water-use effi-
ciency. The aridity index, which integrates annual
precipitation and potential evapotranspiration, is 2.7 times
higher at the less seasonal site than at the more seasonal site
and, thus, further supports the idea that the sampled indi-
viduals responded to lower water availability by increasing
water-use efficiency (Diefendorf et al. 2010; Donovan et al.
2007; Gouveia and Freitas 2009; Ram??rez-Valiente et al.
2010). In general, our results corroborate previous studies
along water availability gradients in tropical, temperate, and
Mediterranean ecosystems, which have found increasing
water-use efficiency with decreasing water availability (e.g.,
Ares and Fownes 1999; Damesin et al. 1997; Keitel et al.
2006; Ram??rez-Valiente et al. 2010; Schulze et al. 2006;
Sobrado 2010). At the species level, the response of wood
d13C to contrasting seasonality varied considerably as evi-
denced by the significant interaction of species and site,
where four of the nine studied species (D. retusa, S. sap-
onaria, T. amazonia, and Z. longifolia) exhibited significant
differences between sites for wood d13C. These results
suggest that changes in water availability altered which
species competed most effectively for water.
Whereas previous research on inter- and intra-specific
variation of wood specific gravity of tropical tree species has
focused on micro-habitat variation within the same forest
community (Sungpalee et al. 2009) or across large geo-
graphical regions (Chave et al. 2006; Mart??nez-Cabrera et al.
2009; Muller-Landau 2004); the present study is unique in
that we compared wood specific gravity of the same nine tree
species across a seasonality gradient under the same man-
agement and using the same seed sources, while controlling
for soil fertility and land-use history. Variation of wood
specific gravity was strongly conserved at the species level,
Fig. 2 Wood traits of nine tropical tree species in Panama. Asterisks
indicate statistically significant differences between sites for each
species (Welch?s two-sample T test, p value\0.05). Black-filled bars
correspond to the less seasonal site (Las Lajas) and white-filled bars
to the more seasonal site (Rio Hato). Whisker bars are standard errors
b
646 Trees (2013) 27:639?653
123
Table 4 Results of correlation analysis of wood and whole-plant performance traits of nine tree species in Panama at each study site
Dry (Rio Hato) Basal diameter
(cm)
RGRBD
(cm day-1)
Aboveground
Biomass (g)
Wood specific
Gravity (g cm-3)
Wood C (%) Wood d13 C (%)
Basal diameter (cm)
RGRBD (cm day
-1) -0.10
Aboveground biomass (g) 0.70 0.08
Wood specific gravity (g cm-3) 20.76 0.33 -0.20
Wood C (%) -0.35 0.34 0.09 0.31
Wood d13 C (%) 0.12 20.80 -0.17 -0.31 -0.67
Wet (Las Lajas) Basal
diameter (cm)
RGRBD
(cm day-1)
Aboveground
biomass (g)
Wood specific
gravity (g cm-3)
Wood C (%) Wood d13C ( %)
Basal diameter (cm)
RGRBD (cm day
-1) 0.58
Aboveground biomass (g) 0.87 0.68
Wood specific gravity (g cm-3) 20.70 -0.06 -0.32
Wood C (%) -0.13 0.08 0.25 0.42
Wood d13C (%) 0.11 -0.40 -0.22 20.68 -0.31
Pearson correlation coefficients were calculated using natural-log transformed values for all variables. As wood d13C values are negative, they
were natural-log transformed after adding an offset of 30 to make values positive. In the lower left corner, Pearson correlation coefficients are
presented; values in bold are statistically significant at a = 0.05
Fig. 3 Mean wood traits at two
sites of nine tropical tree species
in Panama. Mean values for
each wood trait per species at
each site are plotted against one
another. On the X and Y axes,
?wet? and ?dry? correspond to
the less seasonal and more
seasonal sites, respectively. The
solid black line is the major axis
(MA) regression line (p value
\0.05). The dotted gray line
represents a 1:1 relationship;
species that occur along this line
have equivalent trait values at
both study sites. Species codes
correspond to Table 2
Trees (2013) 27:639?653 647
123
as 76 % of its variation was attributable to interspecific
differences. This value compares favorably with values
reported by Chave et al. (2006) of 74 % for 2,456 species in
Central and South America and by Sungpalee et al. (2009) of
80 % for 72 species in Thailand. Despite expectations that
wood specific gravity would be higher at the drier site due to
slower growth and preferential allocation to wood structural
traits associated with drought tolerance (Hacke et al. 2001;
Swenson and Enquist 2007), wood specific gravity of the
studied species, except S. saponaria, did not vary signifi-
cantly between sites. Our results support findings from pre-
vious studies, in which differences in species composition, as
well as soil fertility, were identified as being the principal
drivers of variation in wood specific gravity in neotropical
forests (Baker et al. 2004; Muller-Landau 2004; ter Steege
and Hammond 2001).
Similar to wood specific gravity and wood d13C, wood C
content also varied significantly across species. Average
non-volatile wood C content for all species was 47.4 %,
which is very similar to the global value of non-volatile
wood C content (47.5 %) and is considerably below the
expected value of 50 % (Martin and Thomas 2011;
Thomas and Malczewski 2007). This result illustrates the
importance of using actual wood C content values when
estimating tropical forest C storage, which can result in
significant overestimations when using the assumed value
of 50 % to scale up C storage estimates at large spatial
scales (Martin and Thomas 2011).
Acclimation to variation in water availability: plasticity
in wood traits
We found that the plastic response of three wood traits to
contrasting levels of water availability varied across wood
traits. Wood d13C exhibited the most plasticity, while the
plastic responses of wood specific gravity and wood C con-
tent were effectively neutral. Our results are consistent with
previous studies on leaf trait plasticity in tropical forests,
which reported greater plasticity for physiological traits than
for morphological traits (Valladares et al. 2000). The dif-
ferences in plasticity across wood traits possibly reflect the
relative costs of acclimation to changes in water availability
for particular traits. The energetic costs of increasing
intrinsic water-use efficiency are minimal as no additional
structures are needed to regulate stomatal movement. The
costs of plasticity of wood specific gravity and wood C
content to environmental variation are relatively high, as
species would have to allocate proportionally greater
amounts of C to increase cell wall thickness and fiber density
(Mart??nez-Cabrera et al. 2009). Furthermore, structural
investments to increase drought stress tolerance are not
reversible and have long-term consequences that could
decrease the competitive advantage of structural adaptations
in the future (Nicotra and Davidson 2010). Therefore, the
low plasticity of wood specific gravity and wood C content
points to genetic constraints, not variation in water avail-
ability, as being the principal determinants of variation for
these traits (Baltzer et al. 2007; Chave et al. 2009).
We expected that plasticity in wood traits would be
associated with life history strategies and whole-plant per-
formance traits. As early successional species experience
more stressful and variable environmental conditions in
recently cleared areas and young secondary forests (Duff
et al. 1997; Holl 1999; Lebrija-Trejos et al. 2011), we
anticipated that these species would exhibit lower plasticity
in wood traits than late-successional species. Counter to
expectations, trait plasticity for all wood traits was statisti-
cally independent of life history strategies, which suggests
that the ability to acclimate to changes in water availability is
equally adaptive in both younger and older tropical forests
(Engelbrecht et al. 2006; Slot and Poorter 2007).
In general, plasticity of all wood traits was neutral as the
studied species achieved similar levels of whole-plant per-
formance across sites despite variation in their physiological
response to contrasting seasonality. These results conform
with other studies that have found neutral or maladaptive
plasticity of drought tolerance traits in response to water
limitation for annual and tree species in desert and arid
ecosystems (Donovan et al. 2007; Pohlman et al. 2005;
Ram??rez-Valiente et al. 2010). However, we encountered an
exception to the overall pattern of neutral plasticity in
T. amazonia, which exhibited an adaptive plastic response to
the seasonality gradient. At the more seasonal site, the sig-
nificant decreases in aboveground biomass and basal diam-
eter of T. amazonia were matched by increases in water-use
efficiency and wood C content. The adaptive plastic response
of T. amazonia is consistent with the performance of habitat
specialists, which exhibit better performance in more
favorable environments, thus indicating that this species
could be limited to areas with shorter, less intense dry sea-
sons within the geographical range of its distribution
(Comita and Engelbrecht 2009). This finding is further
substantiated by the significant higher mortality rates of
T. amazonia saplings at the more seasonal site reported by
Breugel et al. (2011). Conversely, the neutral plastic
response of the other studied species is comparable to the
performance of habitat generalists (no differences across
environments), which suggests that the habitat preferences
of these species are not strongly influenced by water avail-
ability (Baltzer et al. 2009; Comita and Engelbrecht 2009).
Relationships of wood- and whole-plant performance
traits in contrasting environments
We also explored relationships among wood and whole-
plant performance traits to evaluate the adaptive value of
648 Trees (2013) 27:639?653
123
wood traits in sites with contrasting seasonality and found
support for the prediction of Nicotra et al. (2010) that the
adaptive value of wood d13C is context specific. At the
more seasonal site, the studied species demonstrated a
trade-off between water-use efficiency and growth, where
wood d13C was significantly and negatively correlated with
RGRBD. The coordination of wood d
13C with RGRBD
illustrates the consequences of water-use efficiency for
species with contrasting drought tolerance strategies where
water availability is limiting; fast-growing species had low
intrinsic water-use efficiency, which is consistent with the
stress avoidance strategy, while slow-growing species had
high intrinsic water-use efficiency, which is associated with
the stress tolerance strategy (Bonal et al. 2007; Chaves
et al. 2002). While high intrinsic water-use efficiency did
not enhance whole-plant performance in terms of greater
tree size at the more seasonal site, the slower growth of
species was adaptive given its well-established association
with higher survival (King et al. 2006; Wright et al. 2010).
However, high intrinsic water-use efficiency was not sim-
ilarly coordinated with wood traits related to hydraulic
function. Rather, the inverse relationship between wood
d13C and wood C content at the more seasonal site suggests
that species with low intrinsic water-use efficiency
increased hydraulic safety by allocating more resources to
C-rich structures (Hacke et al. 2001; Mart??nez-Cabrera
et al. 2009; Swenson and Enquist 2007).
The relatively favorable growing conditions at the less
seasonal site substantively altered the relationship of wood
d13C to whole-plant performance traits. At the less seasonal
site, we found that wood d13C, through its inverse rela-
tionship with wood specific gravity, decreased with tree
size at the less seasonal site. Despite having focused on
saplings, the relationship between wood d13C and tree size
found in the present study broadly coincides with previous
studies on adult-sized trees in tropical forests, where
increasing physiological efficiency with tree size also has
been reported (Martinelli et al. 1998; Nock et al. 2011;
Rijkers et al. 2000). With increasing tree stature, which is
highly correlated with tree diameter (see King et al. 2005),
trees adjust to the greater hydraulic resistance to water flow
in the xylem by closing stomata and increasing leaf mass
area to limit water loss (Rijkers et al. 2000; Thomas and
Winner 2002). The greater drought-induced mortality suf-
fered by large trees with low wood specific gravity in
Amazonian forests suggests that, despite being more con-
servative in their water use, species with larger individuals
and low wood specific gravity in the present study are more
susceptible to drought stress (Nepstad et al. 2007; Phillips
et al. 2010). While able to buffer against water stress
because of greater stem water storage capacity, species
with low wood specific gravity have a lower resistance to
drought-induced embolism than species with higher wood
specific gravity due to larger diameter vessels and fibers
(Borchert and Pockman 2005; Hacke et al. 2001; McCulloh
et al. 2012). Thus, the adaptive value of intrinsic high
water-use efficiency at the less seasonal site is conditional
upon water availability; species with large-statured indi-
viduals would have a competitive advantage during periods
of abundant precipitation (Coomes and Allen 2007; Potvin
and Dutilleul 2009), but would likely experience dispro-
portionally higher mortality following droughts.
Conclusions
To our knowledge, this is one of the few studies where the
adaptive value of traits that underlie drought tolerance
strategies in areas with contrasting levels of water avail-
ability have been explicitly tested in the tropics. Variation
in wood traits of tree saplings was strongly determined by
species identity, greatly exceeding variation attributable to
differences in water availability between study sites. This
variation in wood traits suggests that the response of
tropical forests to severe droughts will depend greatly upon
species composition. Perhaps due to genetic constraints,
plasticity of wood traits to water availability was largely
neutral, indicating that most species at younger life stages
are able to maintain similar levels of growth across habitats
that vary in water availability. Relationships of water-use
efficiency with whole-plant performance and other wood
traits also revealed that the adaptive value of water-use
efficiency changed with water availability. While lower
water availability can induce acclimation responses of
physiological processes, our results suggest that?absent
direct competition from other individuals?the effects of
increased drought stress on growth of 2-year-old saplings
were negligible for a majority of the studied species. Over
longer periods of time, the prolonged impact of longer,
more intense dry seasons might favor slow-growing species
with conservative water-use strategies, as they might suffer
less drought-induced mortality than less water-use efficient
species.
Acknowledgments This article is a scientific contribution of the
Native Species Reforestation Project (PRORENA), a collaborative
research program of the Center for Tropical Forest Science (CTFS) at
the Smithsonian Tropical Research Institute (STRI) and the School of
Forestry and Environmental Studies at Yale University. The species
selection trial at Rio Hato was managed by PRORENA through an
agreement with Panama?s National Environmental Authority
(ANAM) and the land at Las Lajas was made available by Forest
Finance. Export permits for wood and leaf samples were obtained
from ANAM. We thank the field technicians and support staff of
PRORENA and CTFS for their considerable efforts in establishing
and maintaining the species selection trials at Rio Hato and Las Lajas.
We are especially grateful for the help of the following people at
different stages of this study: Yael Camacho, Mark Wishnie, Marcos
Garcia, Julio Nunez, Leidy Martinez, Anabel Rivas, Norma Cedeno,
Trees (2013) 27:639?653 649
123
Rivieth de Leon, Adriana Sautu, Marla Diaz, Rolando Perez, Gerard
Olack, Helmut Ernstberger, Jonas Karosas, Mark Bradford, Christian
Salas, Charles Nock, and Nathaly Guerrero. This research was made
possible through the generous financial support of the Frank Levinson
Family Foundation, the Frank Levinson Donor-Advised Fund at the
Peninsula Community Foundation, Yale University, and STRI. DC
was supported by the Lewis B. Cullman Fellowship for Dissertation
Research and the Tropical Research Institute.
References
Aguilar S, Condit R (2001) Use of native tree species by an Hispanic
community in Panama. Econ Bot 55:223?235
Aranda I, Gil-Pelegr??n E, Gasco? A, Guevara M, Cano J, Miguel M,
Ram??rez-Valiente J, Peguero-Pina J, Perdiguero P, Soto A (2012)
Drought Response in forest trees: from the species to the gene.
Plant Responses to Drought Stress: 293?333
Ares A, Fownes JH (1999) Water supply regulates structure,
productivity, and water use efficiency of Acacia koa forest in
Hawaii. Oecologia 121:458?466
Baker TR, Phillips OL, Malhi Y, Almeida S, Arroyo L, Di Fiore A,
Erwin T, Killeen TJ, Laurance SG, Laurance WF, Lewis SL,
Lloyd J, Monteagudo A, Neill DA, Patino S, Pitman NCA, Silva
MJN, Vasquez Martinez R (2004) Variation in wood density
determines spatial patterns in Amazonian forest biomass. Glob
Change Biol 10:545?562
Baltzer JL, Davies SJ, Noor NSM, Kassim AR, Lafrankie JV (2007)
Geographical distributions in tropical trees: can geographical
range predict performance and habitat association in co-occur-
ring tree species? J Biogeogr: 1?11
Baltzer JL, Gregoire DM, Bunyavejchewin S (2009) Coordination of
foliar and wood anatomical traits contributes to tropical tree
distributions and productivity along the Malay-Thai Peninsula.
Am J Bot 96:2214?2223
Bastien-Henri S, Park A, Ashton M, Messier C (2010) Biomass
distribution among tropical tree species grown under differing
regional climates. For Ecol Manage 260:403?410
Bonal D, Sabatier D, Montpied P, Tremeaux D, Guehl J (2000)
Interspecific variability of d 13 C among trees in rainforests of
French Guiana: functional groups and canopy integration.
Oecologia 124:454?468
Bonal D, Born C, Brechet C, Coste S, Marcon E, Roggy J-C, Guehl
J-M (2007) The successional status of tropical rainforest tree
species is associated with differences in leaf carbon isotope
discrimination and functional traits. Ann For Sci 64:169?176
Borchert R, Pockman WT (2005) Water storage capacitance and
xylem tension in isolated branches of temperate and tropical
trees. Tree Physiol 25:457?466
Bradshaw AD (1965) Evolutionary significance of phenotypic
plasticity in plants. In: Caspari EW, Thoday JM (eds) Advances
in genetics, vol 13. Academic Press, London pp 115?155
Brenes-Arguedas T, Roddy A, Coley P, Kursar T (2011) Do
differences in understory light contribute to species distributions
along a tropical rainfall gradient? Oecologia 166:443?456
Brienen R, Wanek W, Hietz P (2011) Stable carbon isotopes in tree rings
indicate improved water use efficiency and drought responses of a
tropical dry forest tree species. Trees-Str Funct 25:103?113
Bryan JA, Berlyn GP, Gordon JC (1996) Toward a new concept of the
evolution of symbiotic nitrogen fixation in the Leguminosae.
Plant Soil 186:151?159
Cernusak LA, Farquhar GD, Pate JS (2005) Environmental and
physiological controls over oxygen and carbon isotope compo-
sition of Tasmanian blue gum, Eucalyptus globulus. Tree Physiol
25:129?146
Cernusak LA, Aranda J, Marshall JD, Winter K (2007) Large
variation in whole-plant water-use efficiency among tropical tree
species. New Phytol 173:294?305
Cernusak LA, Winter K, Aranda J, Turner BL (2008) Conifers,
angiosperm trees, and lianas: growth, whole-plant water and
nitrogen use efficiency, and stable isotope composition (d13C
and d18O) of seedlings grown in a tropical environment. Plant
Physiol 148:642?659
Cernusak LA, Tcherkez G, Keitel C, Cornwell WK, Santiago LS,
Knohl A, Barbour MM, Williams DG, Reich PB, Ellsworth DS,
Dawson TE, Griffiths HG, Farquhar GD, Wright IJ (2009a) Why
are non-photosynthetic tissues generally 13C enriched compared
with leaves in C3 plants? Review and synthesis of current
hypotheses. Funct Plant Biol 36:199?213
Cernusak LA, Winter K, Turner BL (2009b) Physiological and isotopic
(d13C and d18O) responses of three tropical tree species to water
and nutrient availability. Plant, Cell Environ, pp 1441?1455
Chave J, Muller-Landau HC, Baker TR, Easdale TA, Steege Ht,
Webb CO (2006) Regional and phylogenetic variation of wood
density across 2,456 Neotropical tree species. Ecol Appl
16:2356?2367
Chave J, Coomes D, Jansen S, Lewis SL, Swenson NG, Zanne AE
(2009) Towards a worldwide wood economics spectrum. Ecol
Lett 12:351?366
Chaves MM, Pereira JS, Maroco J, Rodrigues ML, Ricardo CPP,
Osorio ML, Carvalho I, Faria T, Pinheiro C (2002) How plants
cope with water stress in the field? Photosynthesis and growth.
Ann Bot 89:907?916
Comita LS, Engelbrecht BMJ (2009) Seasonal and spatial variation in
water availability drive habitat associations in a tropical forest.
Ecology 90:2755?2765
Condit R, Aguilar S, Hernandez A, Pe?rez R, Lao S, Angehr G,
Hubbell SP, Foster RB (2004) Tropical forest dynamics across a
rainfall gradient and the impact of an El Nin?o dry season. J Trop
Ecol 20:51?72
Coomes DA, Allen RB (2007) Mortality and tree-size distributions in
natural mixed-age forests. J Ecol 95:27?40
Cox PM, Harris PP, Huntingford C, Betts RA, Collins M, Jones CD,
Jupp TE, Marengo JA, Nobre CA (2008) Increasing risk of
Amazonian drought due to decreasing aerosol pollution. Nature
453:212?215
Craven D, Braden D, Ashton MS, Berlyn GP, Wishnie M, Dent D
(2007) Between and within-site comparisons of structural and
physiological characteristics and foliar nutrient content of 14 tree
species at a wet, fertile site and a dry, infertile site in Panama.
For Ecol Manage 238:335?346
Craven D, Gulamhussein S, Berlyn GP (2010) Physiological and
anatomical responses of Acacia koa (Gray) seedlings to varying
light and drought conditions. Environ Exp Bot 69:205?213
Damesin C, Rambal S, Joffre R (1997) Between-tree variations in leaf
d13C of Quercus pubescens and Quercus ilex among Mediter-
ranean habitats with different water availability. Oecologia
111:26?35
Davidar P, Rajagopal B, Mohandass D, Puyravaud JP, Condit R,
Wright SJ, Leigh EG (2007) The effect of climatic gradients,
topographic variation and species traits on the beta diversity of
rain forest trees. Glob Ecol Biogeogr 16:510?518
Dawson TE, Mambelli S, Plamboeck AH, Templer PH, Tu KP (2002)
Stable isotopes in plant ecology. Annu Rev Ecol Syst 33:507
de Faria SM, de Lima HC (1998) Additional studies of the nodulation
status of legume species in Brazil. Plant Soil 200:185?192
Diefendorf AF, Mueller KE, Wing SL, Koch PL, Freeman KH (2010)
Global patterns in leaf 13C discrimination and implications for
studies of past and future climate. Proc Natl Acad Sci 107:5738
Donovan LA, Dudley SA, Rosenthal DM, Ludwig F (2007) Pheno-
typic selection on leaf water use efficiency and related
650 Trees (2013) 27:639?653
123
ecophysiological traits for natural populations of desert sun-
flowers. Oecologia 152:13?25
Duff GA, Myers BA, Williams RJ, Eamus D, O?Grady A, Fordyce IR
(1997) Seasonal patterns in soil moisture, vapour pressure
deficit, tree canopy cover and pre-dawn water potential in a
Northern Australian Savanna. Aust J Bot 45:211?224
Engelbrecht BMJ, Kursar TA (2003) Comparative drought-resistance
of seedlings of 28 species of co-occurring tropical woody plants.
Oecologia 136:383?393
Engelbrecht BMJ, Dalling JW, Pearson TRH, Wolf RL, Ga?lvez DA,
Koehler T, Tyree MT, Kursar TA (2006) Short dry spells in the
wet season increase mortality of tropical pioneer seedlings.
Oecologia 148:258?269
Fan Z-X, Zhang S-B, Hao G-Y, Ferry Slik JW, Cao K-F (2012)
Hydraulic conductivity traits predict growth rates and adult
stature of 40 Asian tropical tree species better than wood density.
J Ecol 100:732?741
Farquhar GD, Richards RA (1984) Isotopic composition of plant
carbon correlates with water-use efficiency of wheat genotypes
11:539?552
Farquhar GD, O?Leary MH, Berry JA (1982) On the relationship
between carbon isotope discrimination and the intercellular
carbon dioxide concentration in leaves. Funct Plant Biol
9:121?137
Fearnside PM (1997) Wood density for estimating forest biomass in
Brazilian Amazonia. For Ecol Manage 90:59?87
Fisher RA (1921) Some remarks on the methods formulated in a
recent article on ??the quantitative analysis of plant growth??. Ann
Appl Biol 7:367?372
Gouveia A, Freitas H (2009) Modulation of leaf attributes and water
use efficiency in Quercus suber along a rainfall gradient. Trees-
Str Funct 23:267?275
Guehl J, Domenach A, Bereau M, Barigah T, Casabianca H, Ferhi A,
Garbaye J (1998) Functional diversity in an Amazonian rainfor-
est of French Guyana: a dual isotope approach (d15N and d13C).
Oecologia 116:316?330
Hacke UG, Sperry JS, Pockman WT, Davis SD, McCulloh (2001)
Trends in wood density and structure are linked to prevention of
xylem implosion by negative pressure. Oecologia 126:457?461
Hasselquist N, Allen M, Santiago L (2010) Water relations of
evergreen and drought-deciduous trees along a seasonally dry
tropical forest chronosequence. Oecologia 164:881?890
Hietz P, Wanek W, Du?nisch O (2005) Long-term trends in cellulose
d13C and water-use efficiency of tropical Cedrela and Swietenia
from Brazil. Tree Physiol 25:745?752
Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A (2005) Very
high resolution interpolated climate surfaces for global land
areas. Int J Climatol 25:1965?1978
Holl KD (1999) Factors limiting tropical rain forest regeneration in
abandoned pasture: seed rain, seed germination, microclimate,
and soil. Biotropica 31:229?242
Huc R, Ferhi A, Guehl JM (1994) Pioneer and late stage tropical
rainforest tree species (French Guiana) growing under common
conditions differ in leaf gas exchange regulation, carbon isotope
discrimination and leaf water potential. Oecologia 99:297?305
Hulshof CM, Swenson NG (2010) Variation in leaf functional trait
values within and across individuals and species: an example
from a Costa Rican dry forest. Funct Ecol 24:217?223
Jackson RB, Canadell J, Ehleringer JR, Mooney HA, Sala OE,
Schulze ED (1996) A global analysis of root distributions for
terrestrial biomes. Oecologia 108:389?411
Keitel C, Matzarakis A, Rennenberg H, Gessler A (2006) Carbon
isotopic composition and oxygen isotopic enrichment in phloem
and total leaf organic matter of European beech (Fagus sylvatica
L.) along a climate gradient. Plant, Cell Environ 29:1492?1507
King DA, Davies SJ, Supardi Noor MN, Tan S (2005) Tree growth is
related to light interception and wood density in two mixed
dipterocarp forests of Malaysia. Funct Ecol 19:445?453
King DA, Davies SJ, Noor NSM (2006) Growth and mortality are
related to adult tree size in a Malaysian mixed dipterocarp forest.
For Ecol Manage 223:152?158
Kitajima K (1994) Relative importance of photosynthetic traits and
allocation patterns as correlates of seedling shade tolerance of 13
tropical trees. Oecologia 98:419?428
Kursar TA, Engelbrecht BMJ, Burke A, Tyree MT, Ei Omari B,
Giraldo JP (2009) Tolerance to low leaf water status of tropical
tree seedlings is related to drought performance and distribution.
Funct Ecol 23:93?102
Lebrija-Trejos E, Pe?rez-Garc??a EA, Meave JA, Poorter L, Bongers F
(2011) Environmental changes during secondary succession in a
tropical dry forest in Mexico. J Trop Ecol 27:477?489
Llamb?? LD, Fontaine M, Rada F, Saugier B, Sarmiento L (2003)
Ecophysiology of dominant plant species during old-field
succession in a high tropical Andean ecosystem. Arct Antarct
Alp Res 35:447?453
Markesteijn L, Poorter L, Bongers F, Paz H, Sack L (2011)
Hydraulics and life history of tropical dry forest tree species:
coordination of species? drought and shade tolerance. New
Phytol 191:480?495
Martin AR, Thomas SC (2011) A reassessment of carbon content in
tropical trees. PLoS ONE 6:e23533
Martinelli L, Almeida S, Brown I, Moreira M, Victoria R, Sternberg
L, Ferreira C, Thomas W (1998) Stable carbon isotope ratio of
tree leaves, boles and fine litter in a tropical forest in Rondonia,
Brazil. Oecologia 114:170?179
Mart??nez-Cabrera HI, Jones CS, Espino S, Schenk HJ (2009) Wood
anatomy and wood density in shrubs: responses to varying
aridity along transcontinental transects. Am J Bot 96:1388?1398
McCulloh KA, Johnson DM, Meinzer FC, Voelker SL, Lachenbruch
B, Domec J-C (2012) Hydraulic architecture of two species
differing in wood density: opposing strategies in co-occurring
tropical pioneer trees. Plant, Cell Environ 35:116?125
Moreira FMDS, Silva MFD, Faria SMD (1992) Occurrence of
nodulation in legume species in the Amazon region of Brazil.
New Phytol 121:563?570
Muller-Landau HC (2004) Interspecific and inter-site variation in
wood specific gravity of tropical trees. Biotropica 36:20?32
Mv Breugel, Hall JS, Craven DJ, Gregoire TG, Park A, Dent DH,
Wishnie MH, Mariscal E, Deago J, Ibarra D, Ceden?o N, Ashton
MS (2011) Early growth and survival of 49 tropical tree species
across sites differing in soil fertility and rainfall in Panama. For
Ecol Manage 261:1580?1589
Nepstad DC, Tohver IM, Ray D, Moutinho P, Cardinot G (2007)
Mortality of large trees and lianas following experimental
drought in an Amazon forest. Ecology 88:2259?2269
Nicotra AB, Davidson A (2010) Adaptive phenotypic plasticity and
plant water use. Funct Plant Biol 37:117?127
Nicotra AB, Atkin OK, Bonser SP, Davidson AM, Finnegan EJ,
Mathesius U, Poot P, Purugganan MD, Richards CL, Valladares
F, van Kleunen M (2010) Plant phenotypic plasticity in a
changing climate. Trends Plant Sci 15:684?692
Niinemets U? (2010) Responses of forest trees to single and multiple
environmental stresses from seedlings to mature plants: past
stress history, stress interactions, tolerance and acclimation. For
Ecol Manage 260:1623?1639
Nock CA, Baker PJ, Wanek W, Leis A, Grabner M, Bunyavejchewin
S, Hietz P (2011) Long-term increases in intrinsic water-use
efficiency do not lead to increased stem growth in a tropical
monsoon forest in western Thailand. Glob Change Biol
17:1049?1063
Trees (2013) 27:639?653 651
123
Oelmann Y, Potvin C, Mark T, Werther L, Tapernon S, Wilcke W
(2010) Tree mixture effects on aboveground nutrient pools of
trees in an experimental plantation in Panama. Plant Soil
326:199?212
Perez R, Condit R (2011) Tree atlas of Panama. http://ctfs.
arnarb.harvard.edu/webatlas/maintreeatlas.php. January 23, 2011
Phillips OL, Aragao LE, Lewis SL, Fisher JB, Lloyd J, Lopez-
Gonzalez G, Malhi Y, Monteagudo A, Peacock J, Quesada CA,
van der Heijden G, Almeida S, Amaral I, Arroyo L, Aymard G,
Baker TR, Banki O, Blanc L, Bonal D, Brando P, Chave J, de
Oliveira AC, Cardozo ND, Czimczik CI, Feldpausch TR, Freitas
MA, Gloor E, Higuchi N, Jimenez E, Lloyd G, Meir P, Mendoza
C, Morel A, Neill DA, Nepstad D, Patino S, Penuela MC, Prieto
A, Ramirez F, Schwarz M, Silva J, Silveira M, Thomas AS,
Steege HT, Stropp J, Vasquez R, Zelazowski P, Alvarez Davila
E, Andelman S, Andrade A, Chao KJ, Erwin T, Di Fiore A,
Honorio CE, Keeling H, Killeen TJ, Laurance WF, Pena Cruz A,
Pitman NC, Nunez Vargas P, Ramirez-Angulo H, Rudas A,
Salamao R, Silva N, Terborgh J, Torres-Lezama A (2009)
Drought sensitivity of the Amazon rainforest. Science
323:1344?1347
Phillips OL, van der Heijden G, Lewis SL, Lo?pez-Gonza?lez G,
Araga?o LEOC, Lloyd J, Malhi Y, Monteagudo A, Almeida S,
Da?vila EA, Amaral I, Andelman S, Andrade A, Arroyo L,
Aymard G, Baker TR, Blanc L, Bonal D, de Oliveira A?CA, Chao
K-J, Cardozo ND, da Costa L, Feldpausch TR, Fisher JB, Fyllas
NM, Freitas MA, Galbraith D, Gloor E, Higuchi N, Honorio E,
Jime?nez E, Keeling H, Killeen TJ, Lovett JC, Meir P, Mendoza
C, Morel A, Vargas PN, Patin?o S, Peh KSH, Cruz AP, Prieto A,
Quesada CA, Ram??rez F, Ram??rez H, Rudas A, Salama?o R,
Schwarz M, Silva J, Silveira M, Ferry Slik JW, Sonke? B,
Thomas AS, Stropp J, Taplin JRD, Va?squez R, Vilanova E
(2010) Drought-mortality relationships for tropical forests. New
Phytol 187:631?646
Pohlman CL, Nicotra AB, Murray BR (2005) Geographic range size,
seedling ecophysiology and phenotypic plasticity in Australian
Acacia species. J Biogeogr 32:341?351
Poorter L, McDonald I, Alarco?n A, Fichtler E, Licona J-C, Pen?a-
Claros M, Sterck F, Villegas Z, Sass-Klaassen U (2010) The
importance of wood traits and hydraulic conductance for the
performance and life history strategies of 42 rainforest tree
species. New Phytol 185:481?492
Potvin C, Dutilleul P (2009) Neighborhood effects and size-
asymmetric competition in a tree plantation varying in diversity.
Ecology 90:321?327
Proyecto de Inventario y Demostraciones Forestales (PIDP) (1970)
Mapa ecolo?gico de Panama?. El Proyecto, Panama?
Quero JL, Villar R, Maran?o?n T, Zamora R (2006) Interactions of
drought and shade effects on seedlings of four Quercus species:
physiological and structural leaf responses. New Phytol
170:819?834
Ram??rez-Valiente JA, Sa?nchez-Go?mez D, Aranda I, Valladares F
(2010) Phenotypic plasticity and local adaptation in leaf
ecophysiological traits of 13 contrasting cork oak populations
under different water availabilities. Tree Physiol 30:618?627
R Development Core Team (2011) R: a language and environment for
statistical computing. R foundation for statistical computing,
Vienna, Austria. ISBN 3-900051-07-0, http://www.R-project.
org/
Rijkers T, Pons TL, Bongers F (2000) The effect of tree height and
light availability on photosynthetic leaf traits of four Neotropical
species differing in shade tolerance. Funct Ecol 14:77?86
Rozendaal D, Zuidema P (2011) Dendroecology in the tropics: a
review. Trees-Str Funct 25:3?16
Rozendaal DMA, Hurtado VH, Poorter L (2006) Plasticity in leaf
traits of 38 tropical tree species in response to light; relationships
with light demand and adult stature. Funct Ecol 20:207?216
Schulze E-D, Turner NC, Nicolle D, Schumacher J (2006) Leaf and
wood carbon isotope ratios, specific leaf areas and wood growth
of Eucalyptus species across a rainfall gradient in Australia. Tree
Physiol 26:479?492
Seibt U, Rajabi A, Griffiths H, Berry JA (2008) Carbon isotopes and
water use efficiency: sense and sensitivity. Oecologia
155:441?454
Slot M, Poorter L (2007) Diversity of tropical tree seedling responses
to drought. Biotropica 39:683?690
Sobrado MA (2010) Leaf characteristics, wood anatomy and
hydraulic properties in tree species from contrasting habitats
within upper Rio Negro forests in the Amazon region. J Trop
Ecol 26:215?226
Sungpalee W, Itoh A, Kanzaki M, Sri-ngernyuang K, Noguchi H,
Mizuno T, Teejuntuk S, Hara M, Chai-udom K, Ohkubo T,
Sahunalu P, Dhanmmanonda P, Nanami S, Yamakura T, Sorn-
ngai A (2009) Intra- and interspecific variation in wood density
and fine-scale spatial distribution of stand-level wood density in
a northern Thai tropical montane forest. J Trop Ecol 25:359?370
Swenson NG, Enquist BJ (2007) Ecological and evolutionary
determinants of a key plant functional trait: wood density and
its community-wide variation across latitude and elevation. Am J
Bot 94:451?459
ter Steege H, Hammond DS (2001) Character convergence, diversity,
and disturbance in tropical rain forest in Guyana. Ecology
82:3197?3212
Thomas SC, Malczewski G (2007) Wood carbon content of tree species
in Eastern China: interspecific variability and the importance of
the volatile fraction. J Environ Manage 85:659?662
Thomas SC, Winner WE (2002) Photosynthetic differences between
saplings and adult trees: an integration of field results by meta-
analysis. Tree Physiol 22:117?127
Tilki F, Fisher RF (1998) Tropical leguminous species for acid soils:
studies on plant form and growth in Costa Rica. For Ecol
Manage 108:175?192
Valladares F, Wright SJ, Lasso E, Kitajima K, Pearcy RW (2000)
Plastic phenotypic response to light of 16 congeneric shrubs
from a Panamanian rainforest. Ecology 81:1925?1936
Valladares F, Gianoli E, Go?mez JM (2007) Ecological limits to plant
phenotypic plasticity. New Phytol 176:749?763
Venables WN, Ripley BD (2002) Modern applied statistics with S.
Springer, Verlag
Warton DI, Hui FKC (2011) The arcsine is asinine: the analysis of
proportions in ecology. Ecology 92:3?10
Warton DI, Wright IJ, Falster DS, Westoby M (2006) Bivariate line-
fitting methods for allometry. Biol Rev 81:259?291
Williams JW, Jackson ST, Kutzbach JE (2007) Projected distributions
of novel and disappearing climates by 2100 AD. Proc Natl Acad
Sci 104:5738
Williamson GB, Wiemann MC (2010) Measuring wood specific
gravity?correctly. Am J Bot 97:519?524
Wishnie MH, Dent DH, Mariscal E, Deago J, Cedeno N, Ibarra D,
Condit R, Ashton PMS (2007) Initial performance and refores-
tation potential of 24 tropical tree species planted across a
precipitation gradient in the Republic of Panama. For Ecol
Manage 243:39?49
Wright SJ, Kitajima K, Kraft NJB, Reich PB, Wright IJ, Bunker DE,
Condit R, Dalling JW, Davies SJ, Diaz S, Engelbrecht BMJ,
Harms KE, Hubbell SP, Marks CO, Ruiz-Jaen MC, Salvador
CM, Zanne AE (2010) Functional traits and the growth-mortality
trade-off in tropical trees. Ecology 91:3664?3674
652 Trees (2013) 27:639?653
123
Zanne AE, Westoby M, Falster DS, Ackerly DD, Loarie SR, Arnold
SE, Coomes DA (2010) Angiosperm wood structure: global
patterns in vessel anatomy and their relation to wood density and
potential conductivity. Am J Bot 97:207?215
Zomer RJ, Trabucco A, Bossio DA, Verchot LV (2008) Climate
change mitigation: a spatial analysis of global land suitability for
clean development mechanism afforestation and reforestation.
Agric Ecosyst Environ 126:67?80
Trees (2013) 27:639?653 653
123