DEMOGRAPHY BEYOND THE POPULATION Functional traits as ... · DEMOGRAPHY BEYOND THE POPULATION Functional traits as predictors of vital rates across the life cycle of tropical trees
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
DEMOGRAPHY BEYOND THE POPULATION
Functional traits as predictors of vital rates across thelife cycle of tropical treesMarco D. Visser*,1,2,3, Marjolein Bruijning1,2, S. Joseph Wright3, Helene C. Muller-Landau3,Eelke Jongejans1,2, Liza S. Comita4 and Hans de Kroon1,2
1Department of Experimental Plant Ecology & Animal Ecology, Institute for Water and Wetland Research, RadboudUniversity, Heyendaalseweg 135, 6525 AJ Nijmegen, The Netherlands; 2Department of Physiology, Institute for Waterand Wetland Research, Radboud University, Heyendaalseweg 135, 6525 AJ, Nijmegen, The Netherlands;3Smithsonian Tropical Research Institute, Box 0843-03092, Balboa, Anc�on, Panama; and 4School of Forestry andEnvironmental Studies, Yale University, New Haven, CT 06511, USA
Summary
1. The ‘functional traits’ of species have been heralded as promising predictors for species’
demographic rates and life history. Multiple studies have linked plant species’ demographic
rates to commonly measured traits. However, predictive power is usually low – raising ques-
tions about the practical usefulness of traits – and analyses have been limited to size-indepen-
dent univariate approaches restricted to a particular life stage.
2. Here we directly evaluated the predictive power of multiple traits simultaneously across the
entire life cycle of 136 tropical tree species from central Panama. Using a model-averaging
approach, we related wood density, seed mass, leaf mass per area and adult stature (maximum
diameter) to onset of reproduction, seed production, seedling establishment, and growth and
survival at seedling, sapling and adult stages.
3. Three of the four traits analysed here (wood density, seed mass and adult stature) typically
explained 20–60% of interspecific variation at a given vital rate and life stage. There were
strong shifts in the importance of different traits throughout the life cycle of trees, with seed
mass and adult stature being most important early in life, and wood density becoming most
important after establishment. Every trait had opposing effects on different vital rates or at dif-
ferent life stages; for example, seed mass was associated with higher seedling establishment and
lower initial survival, wood density with higher survival and lower growth, and adult stature
with decreased juvenile but increased adult growth and survival.
4. Forest dynamics are driven by the combined effects of all demographic processes across the
full life cycle. Application of a multitrait and full-life cycle approach revealed the full role of
key traits, and illuminated how trait effects on demography change through the life cycle. The
effects of traits on one life stage or vital rate were sometimes offset by opposing effects at
another stage, revealing the danger of drawing broad conclusions about functional trait–demography relationships from analysis of a single life stage or vital rate. Robust ecological
and evolutionary conclusions about the roles of functional traits rely on an understanding of
the relationships of traits to vital rates across all life stages.
Key-words: adult stature, leaf mass per area, model averaging, seed size, tree growth, tree
mortality, wood density
Introduction
Functional biology has raised the possibility that morpho-
logical and physiological traits, henceforth functional
traits, might be strongly related to interspecific variation in
provide information on growth and survival for individuals
≥1 cm dbh (hereafter ‘trees’). We analysed data from the 1990
to 2010 censuses, excluding earlier censuses because of small
but important differences in measurement methods (Condit
et al. 1999; R€uger et al. 2009).
2. Seed rain. Seed rain has been recorded in 200 0�5-m2 seed traps
since January 1987 (Wright et al. 2005). Traps are located in a
stratified random manner along trails within the FDP. All
reproductive parts (seeds, flowers, fruits and capsules) are iden-
tified to species and counted weekly (presence is recorded for
flowers). We used seed data from 1993 to 2012, as these years
correspond to records of newly recruiting seedlings (data set 3,
below).
3. Small seedlings. All seedlings and small saplings <1 cm dbh
(with no limits on height) were censused annually in 600 1-m2
seedling plots from 1994 through 2012. These plots are located
2 m from three sides of each of the 200 seed traps (Wright
et al. 2005).
4. Large seedlings. Free-standing woody plants ≥20 cm in height
and<1 cm dbh were censused in 20 000 1-m2 seedling plots each
year from 2001 through 2013, with the exceptions of 2005, 2007
and 2010 (Comita et al. 2007; Comita & Hubbell 2009). These
plots are located in a 5-m grid across the FDP. In each census,
the status (i.e. alive/dead) of previously tagged seedlings is
checked, all individuals are measured for height (except in 2002,
when only new recruits were measured), and new individuals are
tagged and identified to species. In analyses of growth and sur-
vival all census intervals which include missing years were
dropped (i.e. no intervals of>1 year are included for data set 4).
5. Reproductive status. We assessed the reproductive status of
13 358 individual trees to quantify size-dependent probabilities
of reproduction. For each species, a size-stratified sample of
trees was randomly selected and visited during species-specific
reproductive seasons. Reproductive status (fertile or sterile)
was evaluated from the ground using binoculars. For eight
dioecious species, we evaluated sex expression of all individuals
within the FDP. Data were collected between January 1995
and January 1996 for 31 species (Wright et al. 2005), between
2005 and 2007 for 51 wind dispersed species, and between April
2011 and September 2014 for 81 species.
TRA IT DATA
Trait data include seed mass (SM), leaf mass per area (LMA),
adult size (Dmax) and wood density (WD; Table 1; Wright et al.
2010). SM refers to endosperm and embryo dry mass determined
after dissecting diaspores to isolate the endosperm and embryo.
LMA was determined for shade leaves collected from the upper
canopy of the six smallest individuals of each species in the FDP.
We could not use sun-exposed leaves as a basis of comparison
because most FDP species are treelets that complete their entire
life cycle in the shaded understorey (King, Wright & Connell
2006). Dmax is the mean dbh of the six largest individuals in the
FDP (2005 census) and an additional 150 ha of mapped tree
plots located within 30 km and mostly within 10 km of BCI.
Dmax is well correlated with maximum tree height (r = 0�95 on a
log–log scale). Species-specific WD was estimated from tree cores
collected within 15 km of BCI, and was calculated as oven-dried
(60 °C) mass divided by fresh volume (technically wood specific
gravity). Further details can be found in Wright et al. (2010).
The four traits are largely independent of one another, with coef-
ficients of determination (R2 values) of 0�00068, 0�0056, 0�017,0�052, 0�12 and 0�13 for LMA-Dmax, SM-WD, WD-Dmax, SM-
Dmax, LMA-SM and LMA-WD relationships, respectively
(Wright et al. 2010).
We normalized trait values to enable model averaging, and
facilitate comparison of effect sizes among traits with very differ-
ent levels of interspecific variation (Grueber et al. 2011), using all
136 species evaluated here. Species-level trait values were normal-
ized, with SM and Dmax first log-transformed, by subtracting
mean trait values and then dividing by the standard deviation of
the trait values (Table 1).
STUDY SPEC IES
For each life stage and vital rate, we analysed all species with trait
data and sufficient demographic data to ensure reasonable preci-
sion of species-specific vital rate estimates. Table 2 gives exact
selection criteria and the number of species in each analysis.
Table S1 (Supporting information) gives the identities of the spe-
cies in each analysis. Figure S1 shows the distribution of trait val-
ues across all species within each analysis.
F ITT ING TRAIT -BASED MODELS FOR VITAL RATES
We evaluated relationships between size-dependent vital rates and
traits, including trait–size interactions, using generalized linear
mixed models (GLMMs), with species and individual included as
random effects. The most complex full model had the following
form:
Table 2. Species selection criteria and sample sizes for each analysis
Analysis Selection criteria Number of species Years
Reproduction (mm dbh) Reproductive status assessed for >20 trees. Species with too wide
confidence intervals were excluded after visual inspection of fit
60 (8891) 1995–2014
Seed production Species had at least 30 seeds captured in traps, and must have been
included in the reproductive analysis to estimate reproductive basal area
38 (NA) 1993–2010
Seedling establishment 30 or more seedling recruits observed between 1995 and 2011, with
>30 seeds observed for the fruiting years corresponding to
1995–2011 seedling recruitment (taking account of species-specific
germination delays)
68 (NA) 1994–2011
Seedling survival (mm height) >100 individuals in data set 80 (93 082) 2001–2013Seedling growth (mm height) >100 individuals in data set 80 (75 990) 2001–2013Tree survival (mm dbh) >100 individuals in data set 117 (267 469) 1990–2010Tree growth (mm2 basal area) >100 individuals in data set 117 (214 373) 1990–2010
Each vital rate under the column ‘Analysis’ is defined in the text, with corresponding units for size given in parenthesis when the analysis
was carried out on individual data. Under ‘Number of species’, the total number of species in each analysis is given with the total number
of individuals between parentheses. This value is ‘NA’ when the corresponding analysis concerns species-level data. The column ‘Years’
gives the time span of data used in each analysis. A total of 136 unique species were included across all analyses.
where y is growth, (logit) survival or (logit) reproductive fraction; and s is
size in mm height, dbh or mm2 basal area for analyses of seedlings, repro-
ductive size or tree survival, and growth, respectively. Trait effects including
their interactions with size are given by the expression in parentheses, where
Ti represents trait i (corresponding to SM, WD, LMA or Dmax). The ran-
dom effects of species and individuals are denoted by esp and eind, respec-tively, and eresidual is the residual error. For each size-dependent vital rate,
we fit 82 possible models including eqn. 1 and all subsets involving different
combinations of the trait and trait by size effects (Table S2).
Two vital rates, seed production and seedling establishment, were mea-
sured and analysed at species level, and we related these to traits directly
using generalized linear models (GLMs) without size effects. Here too, we
evaluated a suite of models including all subsets of SM, LMA, Dmax and
WD (16 models per vital rate; Tables S3 and S4). Details of model fitting
for each vital rate follow.
Reproduction
The size-dependent probability of reproduction was evaluated
with a logistic GLMM (eqn. 1 with binomial error) using data set
5.
Seed production
Species-specific seed production (fseeds, seeds per year per m2 of
reproductive basal area) was quantified as the mean flux of
seeds arriving (seeds per year per m2 of trap area) divided by
mean reproductive basal area density (m2 of reproductive basal
area per m2 of plot area). We used seed trap and tree census
data from 1993 through 2012. Reproductive basal area was cal-
culated from the tree census data in combination with the fitted
logistic models for size-dependent probability of reproduction.
The logistic models predicted each individual’s reproductive
probability as a function of its size. We then weighted each
individual’s basal area by its reproductive probability to calcu-
late total reproductive basal area. Total reproductive basal area
was interpolated between FDP censuses to calculate annual val-
ues of fseeds, which were then averaged over years to obtain a
single mean value for each species. These simple estimates of
seed production were qualitatively similar to more sophisticated
estimates obtained using inverse modelling (Text S1, Fig. S2).
We chose to use the simple estimates because they were avail-
able for more species. Estimates of fseeds were then related to
traits using linear regression.
Seedling establishment
Species-specific mean seed to seedling establishment probabilities
were calculated as the mean flux of newly recruiting seedlings
per year per m2 in seedling plots in years 1995 to 2012 (data set
3) divided by the mean flux of seeds arriving per year per m2 in
seed traps for the corresponding fruiting years after accounting
for germination delays (Wright et al. 2005; data set 2). Seedling
establishment rates were related to traits using GLMs (i.e. logit
transform).
Growth
Growth was modelled as height growth for seedlings (mm per
year) and basal area growth for trees (mm2 per year) using LMMs
(eqn 1). We used basal area growth because general additive mod-
els (GAMs) showed that basal area growth was generally linearly
related to size (mm2 basal area). Growth rates were calculated as
the difference in sizes divided by the time in years between cen-
suses (data sets 1, 3 and 4). For data set 1, we excluded individuals
marked as ‘resprout’, ‘buttressed’, ‘leaning’ and ‘broken above
1�3 m’ in each census, as well as those with growth rates more
than four 4 standard deviations from the mean. These are likely
measurement errors (R€uger et al. 2011).
Survival
The size-dependent (mm height and mm dbh for seedlings and
trees, respectively) probability of survival was evaluated with a
logistic GLMM (eqn 1 with binomial error) using data sets 1, 3
and 4.
We used model averaging to calculate average parameters
(Burnham & Anderson 2002). All models were assigned a weight
based on their AIC score and fitted parameters were averaged
over the full set of models using these weights to obtain a final
average model. The final average model provides a basis to com-
pare effect sizes. Model averaging is superior to selecting the best
model because models with similar fits are not ignored (Burnham
& Anderson 2002; Whittingham et al. 2006; Bolker et al. 2009).
For this reason, model averaging provides a more robust basis for
inference and prediction, reducing bias in estimation of effect sizes,
especially in cases where multiple variables influence the response
variable (Grueber et al. 2011). This contrasts with stepwise multi-
ple regression, which is seen as poorly suited to disentangle contri-
butions of multiple traits to vital rates (Wittingham et al. 2006).
We averaged parameters over all models having AIC weights >0using the ‘zero method’ in which parameters are assigned the value
zero where absent from models. This is a conservative approach
to model averaging (i.e. leading to lower effect sizes) and is recom-
mended when comparing effect sizes among variables (Burnham &
Anderson 2002; Grueber et al. 2011). Confidence intervals for
each weighted parameter were estimated following Buckland,
Burnham & Augustin (1997).
We evaluated model fits to the full data sets, including varia-
tion among individuals, using marginal and conditional R2 val-
ues developed for mixed-effects models (Nakagawa & Schielzeth
2013). These R2 values provide information on how well the
trait-based hierarchical models (the GLMMs) explain individual-
level variation in vital rates over all species in the community.
To evaluate linearity, we plotted model residuals against size for
each model (Fig. S3). Residuals deviated from linearity for seed-
lings taller than 2�5 m and for trees with dbh >50 cm in the
growth and survival analyses. To ensure linearity, we therefore
excluded trees with dbh >50 cm and seedlings taller than 2�5 m
(corresponding to 0�41% and 0�47% of the data, respectively).
Nonlinearity was not detected in our reproduction analysis. All
analyses were performed in R 3.1.1 (R-core 2014), making use
of the LME4 package for mixed-effects models (Bates et al.
2015). An example R-code is provided for model averaging
(Appendix S2).
THE POWER OF TRA ITS TO EXPLA IN INTERSPEC IF IC
VAR IAT ION
We performed a second set of analyses to estimate the contribu-
tions of traits to explaining interspecific variation in demographic
rates at particular sizes, and thereby to enable more direct com-
parisons of our results with earlier studies based on species-level
estimates of vital rates (Poorter et al. 2008; Wright et al. 2010;
Iida et al. 2012). We first calculated trait-based predictions for
each species and each vital rate based on the fitted average mod-
els. We then compared these predictions with observed mean
SI: Functional traits as predictors of vital rates
across the life-cycle of tropical trees.
Marco D. Visser∗1,2, Marjolein Bruijning†1, Stuart J. Wright‡2,Helene C. Muller-Landau§2, Eelke Jongejans¶1, Liza S. Comita‖3,
and Hans de Kroon∗∗1
1Radboud University, Departments of Experimental Plant Ecology& Animal Ecology and Physiology, The Netherlands
2Smithsonian Tropical Research Institute, Republic of Panama3Yale University, School of Forestry and Environmental Studies
December 1, 2015
Contents
S1 TEXT: Comparison of results using inverse modeling 2
S2 Tables 4
S3 Figures 21
S1 TEXT: Comparison of results using
inverse modeling
When multiple adult trees contribute seeds to a single location it becomes com-plicated to assign individual seeds to their respective sources. A now commonlyused method pioneered by Ribbens et al (1994) to overcome this problem isinverse modeling (IM) of seed shadows (Clark et al, 1999; Muller-Landau etal, 2008). A seed shadow describes the distribution of seeds with distance from
S1 TEXT: COMPARISON OF RESULTS USING INVERSE MODELING
their parent, and consists of two elements; (1) a component describing the fecun-dity of a tree, usually as a function of size and (2) a dispersal kernel describingthe distribution of seeds as a function of distance from the source. The seedshadow is the product of these two elements (Clark et al., 1999).
Seed shadows from individual trees will overlap within a stand and the seedrain of these multiple sources can be viewed as a smoothed version of the individ-ual shadows (Clark et al., 1999). This smoothed shadow can be expressed as thesummed contributions of each individual seed shadow (hereafter the summedseed shadow SSS). The number of seeds in a certain quadrat is predicted by sum-ming the densities the quadrat receives from each of the surrounding sources, asprojected by each individual seed shadow. IM has been shown to be a promisingestimate of dispersal distances, as it is not inherently biased to long or shortdistance dispersal (Jones and Muller-Landau, 2008; Muller-Landau et al, 2008).With IM the predicted seed density in a sample quadrat at location j is givenby the summed seed shadow of all N trees at location j (SSSj);
SSSj = α ·T∑
i=1
F (pibi|β) ·Q(rij |µ, σ) (1)
where SSSj is the predicted seed number in trap j, which is the sum of seedcontributions from T trees multiplied by trap size α (0.5 m2). For each treei, the seed shadow is a function of the fecundity F , and a two-dimensionaldispersal kernel Q. Q depends on the distance rij between tree i and trapj, while F is a function of basal area bi (calculated as [dbh/2]2π) and seedproduction probabilities pi (obtained from the reproductive census as describedin the main text). Only females were included for dioecious species.
The parameters β, µ and σ were fitted per species. We tested four differentdispersal kernels: the Exponential, 2dt (Clark et al., 1999), 2dt1k (which is the2dt distribution with the degrees of freedom parameter set to 3) (Muller-Landauet al., 2008) and the Cauchy density distribution.
Off plot integrationTraps within 20 meter from the edge of the plot were excluded since they couldbe highly influenced by trees outside the plot (Muller-Landau et al., 2008). Datafrom the remaining 188 traps were used. We then estimated the contribution tothe seed rain from unknown sources outside the mapped plot. Here it is assumedthat the per unit area seed production off plot is uniform to that within theplot (see Muller-Landau et al., 2008, for details). The total expected seed count
ˆSSSj in trap j then becomes the sum of the expected contributions from treesinside the plot SSSj and the expected contributions from trees outside the plot
(immigrant seeds) Ij to seed trap j.
ˆSSSj = SSSj + Ij (2)
2
S2 TABLES
Parameter estimationModels were fit using maximum likelihood, assuming seed counts in traps weredistributed according to a negative binomial distribution, with an additinalclumping parameter k (Muller-Landau et al., 2008), as previous studies haveshown that seed arrival tends to be clumped (Muller-Landau et al., 2008).
The results from the above described inverse modeling approach are comparedto the alternative estimates (see maintext) in figure S2.
S2 Tables
3
S2
TABLES
Table S1. Overview of species used in each analysis, ordered alphabetically. True indicates a species was included.
Species Reproduction Seed production Seedlingestablishment
Table S2. All evaluated forms of the models for demographic rates (Y) that arepotentially size-dependent. The variable Y refers to seedling growth, seedlingsurvival, tree growth, tree survival, or tree reproduction. The variable size refersto seedling height, tree dbh (when Y is tree reproduction or survival) or treebasal area (when Y is tree growth).The variables wd, sm, lma and dmax referto wood density, seed mass, leaf mass per area and maximum dbh, respectively.Random effects are for species (1 | species) and individual (1 | tag).
formula
1 Y ∼ 1 + (1|species) + (1|tag)2 Y ∼ size + (1|species) + (1|tag)3 Y ∼ size + lma + (1|species) + (1|tag)4 Y ∼ size ∗ lma + (1|species) + (1|tag)5 Y ∼ size + sm + (1|species) + (1|tag)6 Y ∼ size + sm + lma + (1|species) + (1|tag)7 Y ∼ size + sm + size ∗ lma + (1|species) + (1|tag)8 Y ∼ size ∗ sm + (1|species) + (1|tag)9 Y ∼ size ∗ sm + lma + (1|species) + (1|tag)
10 Y ∼ size ∗ sm + size ∗ lma + (1|species) + (1|tag)11 Y ∼ size + wd + (1|species) + (1|tag)12 Y ∼ size + wd + lma + (1|species) + (1|tag)13 Y ∼ size + wd + size ∗ lma + (1|species) + (1|tag)14 Y ∼ size + wd + sm + (1|species) + (1|tag)15 Y ∼ size + wd + sm + lma + (1|species) + (1|tag)16 Y ∼ size + wd + sm + size ∗ lma + (1|species) + (1|tag)17 Y ∼ size + wd + size ∗ sm + (1|species) + (1|tag)18 Y ∼ size + wd + size ∗ sm + lma + (1|species) + (1|tag)19 Y ∼ size + wd + size ∗ sm + size ∗ lma + (1|species) + (1|tag)20 Y ∼ size ∗ wd + (1|species) + (1|tag)21 Y ∼ size ∗ wd + lma + (1|species) + (1|tag)22 Y ∼ size ∗ wd + size ∗ lma + (1|species) + (1|tag)23 Y ∼ size ∗ wd + sm + (1|species) + (1|tag)24 Y ∼ size ∗ wd + sm + lma + (1|species) + (1|tag)25 Y ∼ size ∗ wd + sm + size ∗ lma + (1|species) + (1|tag)26 Y ∼ size ∗ wd + size ∗ sm + (1|species) + (1|tag)27 Y ∼ size ∗ wd + size ∗ sm + lma + (1|species) + (1|tag)28 Y ∼ size ∗ wd + size ∗ sm + size ∗ lma + (1|species) + (1|tag)29 Y ∼ size + dmax + (1|species) + (1|tag)30 Y ∼ size + lma + dmax + (1|species) + (1|tag)31 Y ∼ size ∗ lma + dmax + (1|species) + (1|tag)32 Y ∼ size + sm + dmax + (1|species) + (1|tag)33 Y ∼ size + sm + lma + dmax + (1|species) + (1|tag)34 Y ∼ size + sm + size ∗ lma + dmax + (1|species) + (1|tag)35 Y ∼ size ∗ sm + dmax + (1|species) + (1|tag)36 Y ∼ size ∗ sm + lma + dmax + (1|species) + (1|tag)37 Y ∼ size ∗ sm + size ∗ lma + dmax + (1|species) + (1|tag)38 Y ∼ size + wd + dmax + (1|species) + (1|tag)39 Y ∼ size + wd + lma + dmax + (1|species) + (1|tag)40 Y ∼ size + wd + size ∗ lma + dmax + (1|species) + (1|tag)41 Y ∼ size + wd + sm + dmax + (1|species) + (1|tag)42 Y ∼ size + wd + sm + lma + dmax + (1|species) + (1|tag)43 Y ∼ size + wd + sm + size ∗ lma + dmax + (1|species) + (1|tag)44 Y ∼ size + wd + size ∗ sm + dmax + (1|species) + (1|tag)45 Y ∼ size + wd + size ∗ sm + lma + dmax + (1|species) + (1|tag)46 Y ∼ size + wd + size ∗ sm + size ∗ lma + dmax + (1|species) + (1|tag)47 Y ∼ size ∗ wd + dmax + (1|species) + (1|tag)48 Y ∼ size ∗ wd + lma + dmax + (1|species) + (1|tag)49 Y ∼ size ∗ wd + size ∗ lma + dmax + (1|species) + (1|tag)50 Y ∼ size ∗ wd + sm + dmax + (1|species) + (1|tag)51 Y ∼ size ∗ wd + sm + lma + dmax + (1|species) + (1|tag)52 Y ∼ size ∗ wd + sm + size ∗ lma + dmax + (1|species) + (1|tag)53 Y ∼ size ∗ wd + size ∗ sm + dmax + (1|species) + (1|tag)54 Y ∼ size ∗ wd + size ∗ sm + lma + dmax + (1|species) + (1|tag)55 Y ∼ size ∗ wd + size ∗ sm + size ∗ lma + dmax + (1|species) + (1|tag)56 Y ∼ size ∗ dmax + (1|species) + (1|tag)57 Y ∼ size + lma + size ∗ dmax + (1|species) + (1|tag)58 Y ∼ size ∗ lma + size ∗ dmax + (1|species) + (1|tag)59 Y ∼ size + sm + size ∗ dmax + (1|species) + (1|tag)60 Y ∼ size + sm + lma + size ∗ dmax + (1|species) + (1|tag)61 Y ∼ size + sm + size ∗ lma + size ∗ dmax + (1|species) + (1|tag)62 Y ∼ size ∗ sm + size ∗ dmax + (1|species) + (1|tag)63 Y ∼ size ∗ sm + lma + size ∗ dmax + (1|species) + (1|tag)64 Y ∼ size ∗ sm + size ∗ lma + size ∗ dmax + (1|species) + (1|tag)65 Y ∼ size + wd + size ∗ dmax + (1|species) + (1|tag)66 Y ∼ size + wd + lma + size ∗ dmax + (1|species) + (1|tag)67 Y ∼ size + wd + size ∗ lma + size ∗ dmax + (1|species) + (1|tag)68 Y ∼ size + wd + sm + size ∗ dmax + (1|species) + (1|tag)69 Y ∼ size + wd + sm + lma + size ∗ dmax + (1|species) + (1|tag)70 Y ∼ size + wd + sm + size ∗ lma + size ∗ dmax + (1|species) + (1|tag)71 Y ∼ size + wd + size ∗ sm + size ∗ dmax + (1|species) + (1|tag)72 Y ∼ size + wd + size ∗ sm + lma + size ∗ dmax + (1|species) + (1|tag)73 Y ∼ size + wd + size ∗ sm + size ∗ lma + size ∗ dmax + (1|species) + (1|tag)74 Y ∼ size ∗ wd + size ∗ dmax + (1|species) + (1|tag)75 Y ∼ size ∗ wd + lma + size ∗ dmax + (1|species) + (1|tag)76 Y ∼ size ∗ wd + size ∗ lma + size ∗ dmax + (1|species) + (1|tag)77 Y ∼ size ∗ wd + sm + size ∗ dmax + (1|species) + (1|tag)78 Y ∼ size ∗ wd + sm + lma + size ∗ dmax + (1|species) + (1|tag)79 Y ∼ size ∗ wd + sm + size ∗ lma + size ∗ dmax + (1|species) + (1|tag)80 Y ∼ size ∗ wd + size ∗ sm + size ∗ dmax + (1|species) + (1|tag)81 Y ∼ size ∗ wd + size ∗ sm + lma + size ∗ dmax + (1|species) + (1|tag)82 Y ∼ size ∗ wd + size ∗ sm + size ∗ lma + size ∗ dmax + (1|species) + (1|tag)
9
S2
TABLES
Table S3. All evaluated model forms, model weights, and ∆AIC values for the seed production analysis (fseeds).
Table S5. Evaluated model forms, model weights, and ∆AIC values for the seedling growth analysis, for the top 30 models(out of 82). A full list of model forms can be found in table S2.
Table S6. Evaluated model forms, model weights, and ∆AIC values for the seedling survival analysis, for the top 30 models(out of 82). A full list of model forms can be found in table S2.
Table S7. Evaluated model forms, model weights, and ∆AIC values for the tree growth analysis, for the top 30 models (outof 82). A full list of model forms can be found in table S2. The variable BA refers to basal area (mm2).
Table S8. Evaluated model forms, model weights, and ∆AIC values for the tree survival analysis, for the top 30 models (outof 82). A full list of model forms can be found in table S2.
Table S9. Evaluated model forms, model weights, and ∆AIC values for the reproduction analysis, for the top 30 models (outof 82). A full list of model forms can be found in table S2.
Table S11. R2 values at different sizes for average models including only single traits for seedlings. Dmax, LMA, SM and WDrepresent adult stature, leaf mass per area, seed mass and wood density.
Table S12. R2 values at different sizes from average models for growth, survival and reproduction including only single traitsfor trees (>1cm dbh). Dmax, LMA, SM and WD represent adult stature, leaf mass per area, seed mass and wood density.
Figure S1. Boxplots showing the distribution of trait values (in standardizedunits) for each vital rate analysis. Each box plot within a panel corresponds tothe distribution of trait values for the species used within that specific analysis.
21
S3 FIGURES
Ful
l
SM
WD
LMA
Dm
ax
A
R2 fo
r se
ed p
rodu
ctio
n
0.0
0.2
0.4
0.6
0.8
1.0
0.64
0.54
0.04
0.09
0.06
Ful
l
SM
WD
LMA
Dm
ax
B
0.0
0.2
0.4
0.6
0.8
1.0
0.55
0.54 0
0.05
0.04
Figure S2. R2 values for interspecific variation in seed production from averagedmodels for a single trait and for the full averaged model for all four traits,compared between seed production estimates obtained through inverse modeling(A, 18 species) or by estimating seed production per unit basal area as describedin the main text (B, 38 species).
22
S3 FIGURES
Figure S3. Plots of residuals of the full averaged model against size for treegrowth (A,B), tree survival (C,D), seedling growth (E,F), seedling survival (G,H,and reproduction (I). Black dots indicate residual values from each model-averaged mixed-effect model (predicted - observed), while the red line is themoving average. The first column (A,C,E,G,I) shows residuals for models fit tothe full datasets, while the second column (B,D,F,H) shows residuals for modelsfit to datasets truncated above at 500 mm diameter (B,D) or 2500 mm height(F, H) to avoid non-linearities with size.
23
S3
FIG
URES
0 100 200 300 400 500
0.0
0.4
0.8
Alchornea costaricensi
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Alseis blackiana
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Apeiba membranacea
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Apeiba tibourbou
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Aspidosperma spruceanum
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Astronium graveolens
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Beilschmiedi pendula
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Brosimum alicastrum
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Casearia aculeata
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Cassipourea elliptica
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Chrysophyllu argenteum
Dbh (mm)R
epro
duct
ion
prob
abili
ty
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Chrysophyllu cainito
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Cordia alliodora
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Cordia bicolor
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Coussarea curvigemmia
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Desmopsis panamensis
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Figure S4. Observed and fitted proportion of individuals that are reproductive as a function of tree diameter (mm) for eachspecies. Black dots show observations, red dashed lines show species-specific moving averages (from a Generalized AdditiveModel using a loess smoother), blue dashed lines show the fitted trait based average model, and grey lines show a fitted modelbased only on size (not on traits, thus identical for all species).
24
S3
FIG
URES
0 100 200 300 400 500
0.0
0.4
0.8
Dipteryx oleifera
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Eugenia oerstediana
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Faramea occidentalis
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Garcinia intermedia
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Guarea guidonia
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Guatteria dumetorum
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Gustavia superba
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Hasseltia floribunda
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Heisteria concinna
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Hirtella triandra
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Hieronyma alchorneoide
Dbh (mm)R
epro
duct
ion
prob
abili
ty
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Inga marginata
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Jacaranda copaia
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Laetia thamnia
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Lonchocarpus heptaphyllus
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Luehea seemannii
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Figure S4. Continued.
25
S3
FIG
URES
0 100 200 300 400 500
0.0
0.4
0.8
Mosannona garwoodii
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Mouriri myrtilloides
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Platypodium elegans
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Poulsenia armata
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Pouteria reticulata
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Prioria copaifera
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Protium panamense
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Protium tenuifolium
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Pseudobombax septenatum
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Pterocarpus rohrii
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Quararibea asterolepis
Dbh (mm)R
epro
duct
ion
prob
abili
ty
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Randia armata
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Simarouba amara
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Sloanea terniflora
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Tabebuia rosea
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Tabernaemont arborea
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Figure S4. Continued.
26
S3
FIG
URES
0 100 200 300 400 500
0.0
0.4
0.8
Talisia nervosa
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Terminalia amazonia
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Terminalia oblonga
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Tetragastris panamensis
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Trattinnicki aspera
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Trichilia pallida
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Trichilia tuberculata
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Triplaris cumingiana
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Unonopsis pittieri
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Virola sebifera
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Vochysia ferruginea
Dbh (mm)R
epro
duct
ion
prob
abili
ty
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
0 100 200 300 400 500
0.0
0.4
0.8
Xylopia macrantha
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Dbh (mm)
Rep
rodu
ctio
n pr
obab
ility
Figure S4. Continued.
27
S3
FIG
URES
500 1000 1500 2000 2500
−20
00
100
Alchornea costaricensi
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Alibertia edulis
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Alseis blackiana
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Anacardium excelsum
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Annona acuminata
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Annona spraguei
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Apeiba membranacea
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Aspidosperma spruceanum
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Beilschmiedi pendula
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Brosimum alicastrum
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Calophyllum longifolium
Height (mm)Ye
arly
gro
wth
(m
m h
eigh
t)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Capparis frondosa
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Cassipourea elliptica
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Chrysophyllu argenteum
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Chrysophyllu cainito
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Cordia alliodora
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Figure S5. Observed and predicted annual seedling height growth (mm/yr) as a function of initial height (mm). Black dotsshow observations, red dashed lines show species-specific moving averages (from a Generalized Additive Model using a loesssmoother), blue dashed lines show the fitted trait based average model, and grey lines show a fitted model based only on size(not on traits, thus identical for all species).
28
S3
FIG
URES
500 1000 1500 2000 2500
−20
00
100
Cordia bicolor
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Coussarea curvigemmia
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Desmopsis panamensis
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Dipteryx oleifera
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Drypetes standleyi
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Eugenia coloradoensi
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Eugenia galalonensis
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Eugenia nesiotica
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Eugenia oerstediana
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Faramea occidentalis
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Garcinia intermedia
Height (mm)Ye
arly
gro
wth
(m
m h
eigh
t)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Guapira standleyana
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Guarea guidonia
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Guatteria dumetorum
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Gustavia superba
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Hamelia axillaris
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Figure S5. Continued.
29
S3
FIG
URES
500 1000 1500 2000 2500
−20
00
100
Hampea appendiculat
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Heisteria concinna
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Hirtella triandra
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Hybanthus prunifolius
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Inga marginata
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Inga acuminata
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Inga thibaudiana
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Inga umbellifera
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Lacistema aggregatum
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Lacmellea panamensis
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Lonchocarpus heptaphyllus
Height (mm)Ye
arly
gro
wth
(m
m h
eigh
t)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Luehea seemannii
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Mouriri myrtilloides
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Ocotea cernua
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Ocotea puberula
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Trophis caucana
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Figure S5. Continued.
30
S3
FIG
URES
500 1000 1500 2000 2500
−20
00
100
Ouratea lucens
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Palicourea guianensis
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Cinnamomum triplinerve
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Picramnia latifolia
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Platymiscium pinnatum
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Pouteria reticulata
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Prioria copaifera
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Protium panamense
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Protium tenuifolium
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Psychotria acuminata
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Psychotria deflexa
Height (mm)Ye
arly
gro
wth
(m
m h
eigh
t)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Psychotria horizontalis
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Psychotria marginata
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Psychotria racemosa
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Quararibea asterolepis
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Randia armata
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Figure S5. Continued.
31
S3
FIG
URES
500 1000 1500 2000 2500
−20
00
100
Simarouba amara
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Sorocea affinis
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Spondias radlkoferi
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Sterculia apetala
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Stylogyne turbacensis
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Swartzia simplex_var.
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Tabebuia rosea
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Tabernaemont arborea
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Tachigali versicolor
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Talisia croatii
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Tetragastris panamensis
Height (mm)Ye
arly
gro
wth
(m
m h
eigh
t)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Trichilia tuberculata
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Unonopsis pittieri
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Virola sebifera
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Xylopia macrantha
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
500 1000 1500 2000 2500
−20
00
100
Zanthoxylum panamense
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Height (mm)
Year
ly g
row
th (
mm
hei
ght)
Figure S5. Continued.
32
S3
FIG
URES
500 1000 1500 2000 2500
0.0
0.4
0.8
Alchornea costaricensi
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Alibertia edulis
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Alseis blackiana
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Anacardium excelsum
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Annona acuminata
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Annona spraguei
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Apeiba membranacea
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Aspidosperma spruceanum
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Beilschmiedi pendula
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Brosimum alicastrum
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Calophyllum longifolium
Height (mm)Ye
arly
sur
viva
l pro
babi
lity
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Capparis frondosa
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Cassipourea elliptica
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Chrysophyllu argenteum
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Chrysophyllu cainito
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Cordia alliodora
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Figure S6. Observed and predicted annual seedling survival as a function of initial height (mm). Black dots show observations,red dashed lines show species-specific moving averages (from a Generalized Additive Model using a loess smoother), bluedashed lines show the fitted trait based average model, and grey lines show a fitted model based only on size (not on traits,thus identical for all species).
33
S3
FIG
URES
500 1000 1500 2000 2500
0.0
0.4
0.8
Cordia bicolor
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Coussarea curvigemmia
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Desmopsis panamensis
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Dipteryx oleifera
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Drypetes standleyi
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Eugenia coloradoensi
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Eugenia galalonensis
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Eugenia nesiotica
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Eugenia oerstediana
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Faramea occidentalis
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Garcinia intermedia
Height (mm)Ye
arly
sur
viva
l pro
babi
lity
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Guapira standleyana
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Guarea guidonia
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Guatteria dumetorum
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Gustavia superba
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Hamelia axillaris
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Figure S6. Continued.
34
S3
FIG
URES
500 1000 1500 2000 2500
0.0
0.4
0.8
Hampea appendiculat
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Heisteria concinna
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Hirtella triandra
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Hybanthus prunifolius
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Inga marginata
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Inga acuminata
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Inga thibaudiana
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Inga umbellifera
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Lacistema aggregatum
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Lacmellea panamensis
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Lonchocarpus heptaphyllus
Height (mm)Ye
arly
sur
viva
l pro
babi
lity
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Luehea seemannii
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Mouriri myrtilloides
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Ocotea cernua
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Ocotea puberula
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Trophis caucana
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Figure S6. Continued.
35
S3
FIG
URES
500 1000 1500 2000 2500
0.0
0.4
0.8
Ouratea lucens
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Palicourea guianensis
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Cinnamomum triplinerve
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Picramnia latifolia
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Platymiscium pinnatum
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Pouteria reticulata
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Prioria copaifera
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Protium panamense
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Protium tenuifolium
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Psychotria acuminata
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Psychotria deflexa
Height (mm)Ye
arly
sur
viva
l pro
babi
lity
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Psychotria horizontalis
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Psychotria marginata
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Psychotria racemosa
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Quararibea asterolepis
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Randia armata
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Figure S6. Continued.
36
S3
FIG
URES
500 1000 1500 2000 2500
0.0
0.4
0.8
Simarouba amara
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Sorocea affinis
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Spondias radlkoferi
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Sterculia apetala
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Stylogyne turbacensis
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Swartzia simplex_var.
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Tabebuia rosea
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Tabernaemont arborea
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Tachigali versicolor
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Talisia croatii
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Tetragastris panamensis
Height (mm)Ye
arly
sur
viva
l pro
babi
lity
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Trichilia tuberculata
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Unonopsis pittieri
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Virola sebifera
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Xylopia macrantha
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
500 1000 1500 2000 2500
0.0
0.4
0.8
Zanthoxylum panamense
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Height (mm)
Year
ly s
urvi
val p
roba
bilit
y
Figure S6. Continued.
37
S3
FIG
URES
0 10000 20000 30000
−10
000
500
Adelia triloba
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 20000 40000 60000
−10
000
500
Alchornea costaricensi
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Alibertia edulis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Allophylus
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Alseis blackiana
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 4000 6000 8000
−10
000
500
Andira inermis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Annona acuminata
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 10000 20000 30000
−10
000
500
Annona spraguei
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0e+00 4e+04 8e+04
−10
000
500
Apeiba membranacea
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 4000 6000 8000
−10
000
500
Aspidosperma spruceanum
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 20000 40000 60000
−10
000
500
Astronium graveolens
Basal area (mm2)Ye
arly
gro
wth
(m
m2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Beilschmiedi pendula
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Brosimum alicastrum
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Calophyllum longifolium
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Capparis frondosa
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Casearia aculeata
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Figure S7. Observed and predicted annual tree growth (mm2 basal area/yr) as a function of initial size (mm2 basal area).Black dots show observations, red dashed lines show species-specific moving averages (from a Generalized Additive Model usinga loess smoother), blue dashed lines show the fitted trait based average model, and grey lines show a fitted model based onlyon size (not on traits, thus identical for all species).
38
S3
FIG
URES
0 20000 60000
−10
000
500
Casearia arborea
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Cassipourea elliptica
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Chrysochlamy eclipes
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Chrysophyllu argenteum
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 4000 6000 8000
−10
000
500
Chrysophyllu cainito
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Coccoloba coronata
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Coccoloba manzinellens
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0e+00 4e+04 8e+04
−10
000
500
Cordia alliodora
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 20000 40000 60000
−10
000
500
Cordia bicolor
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Coussarea curvigemmia
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Cupania rufescens
Basal area (mm2)Ye
arly
gro
wth
(m
m2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Cupania seemannii
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Desmopsis panamensis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Drypetes standleyi
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Erythroxylum panamense
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Eugenia coloradoensi
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Figure S7. Continued.
39
S3
FIG
URES
200 400 600 800 1200
−10
000
500
Eugenia galalonensis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Eugenia nesiotica
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 4000 6000 8000
−10
000
500
Eugenia oerstediana
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 4000 6000 8000
−10
000
500
Faramea occidentalis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Garcinia intermedia
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Garcinia madruno
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 20000 40000 60000
−10
000
500
Guapira standleyana
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 10000 20000 30000
−10
000
500
Guarea guidonia
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Guatteria dumetorum
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 10000 30000 50000
−10
000
500
Gustavia superba
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
50 60 70 80 90 100
−10
000
500
Hamelia axillaris
Basal area (mm2)Ye
arly
gro
wth
(m
m2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 10000 30000 50000
−10
000
500
Hampea appendiculat
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 15000 25000
−10
000
500
Hasseltia floribunda
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 4000 6000 8000
−10
000
500
Heisteria acuminata
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 10000 30000 50000
−10
000
500
Heisteria concinna
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Herrania purpurea
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Figure S7. Continued.
40
S3
FIG
URES
0 5000 10000 15000 20000
−10
000
500
Hirtella triandra
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Hybanthus prunifolius
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
5000 15000 25000
−10
000
500
Hieronyma alchorneoide
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 4000 6000 8000
−10
000
500
Inga goldmanii
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Inga marginata
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 4000 6000 8000
−10
000
500
Inga nobilis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Inga acuminata
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Inga sapindoides
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 15000 25000
−10
000
500
Inga thibaudiana
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Inga umbellifera
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 50000 100000 150000
−10
000
500
Jacaranda copaia
Basal area (mm2)Ye
arly
gro
wth
(m
m2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Lacistema aggregatum
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 10000 30000 50000
−10
000
500
Lacmellea panamensis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 4000 6000 8000
−10
000
500
Laetia thamnia
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Licania hypoleuca
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Licania platypus
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Figure S7. Continued.
41
S3
FIG
URES
0 5000 10000 15000 20000
−10
000
500
Lonchocarpus heptaphyllus
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 15000 25000
−10
000
500
Luehea seemannii
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Mosannona garwoodii
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 4000 6000 8000
−10
000
500
Maquira guianensis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
50 60 70 80 90 100
−10
000
500
Mouriri myrtilloides
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Nectandra cissiflora
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Ocotea cernua
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Ocotea puberula
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 4000 6000 8000
−10
000
500
Trophis caucana
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Ormosia coccinea
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Ormosia macrocalyx
Basal area (mm2)Ye
arly
gro
wth
(m
m2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Ouratea lucens
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Palicourea guianensis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Pentagonia macrophylla
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 4000 6000 8000
−10
000
500
Perebea xanthochyma
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Picramnia latifolia
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Figure S7. Continued.
42
S3
FIG
URES
0 10000 30000 50000
−10
000
500
Platymiscium pinnatum
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Platypodium elegans
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Poulsenia armata
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 15000 25000
−10
000
500
Pourouma bicolor
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Pouteria reticulata
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Prioria copaifera
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Protium panamense
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Protium tenuifolium
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
50 60 70 80 90 100
−10
000
500
Psychotria horizontalis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
50 60 70 80 90 100
−10
000
500
Psychotria marginata
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Pterocarpus rohrii
Basal area (mm2)Ye
arly
gro
wth
(m
m2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Quararibea asterolepis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
1500 2500 3500
−10
000
500
Quassia amara
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Randia armata
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Simarouba amara
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Sloanea terniflora
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Figure S7. Continued.
43
S3
FIG
URES
0 1000 2000 3000
−10
000
500
Sorocea affinis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Spondias mombin
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 15000 25000
−10
000
500
Spondias radlkoferi
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Stylogyne turbacensis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 4000 6000 8000
−10
000
500
Symphonia globulifera
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Tabebuia rosea
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Tabernaemont arborea
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Tachigali versicolor
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Talisia nervosa
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
200 400 600 800 1200
−10
000
500
Talisia croatii
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Tetragastris panamensis
Basal area (mm2)Ye
arly
gro
wth
(m
m2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 1000 2000 3000
−10
000
500
Thevetia ahouai
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 10000 15000 20000
−10
000
500
Trichilia pallida
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 5000 15000 25000
−10
000
500
Trichilia tuberculata
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 10000 20000 30000
−10
000
500
Triplaris cumingiana
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 10000 20000 30000
−10
000
500
Turpinia occidentalis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Figure S7. Continued.
44
S3
FIG
URES
0 5000 15000 25000
−10
000
500
Unonopsis pittieri
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 10000 30000 50000
−10
000
500
Virola sebifera
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 10000 30000 50000
−10
000
500
Virola surinamensis
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 2000 6000 10000
−10
000
500
Xylopia macrantha
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
0 10000 30000 50000
−10
000
500
Zanthoxylum panamense
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Basal area (mm2)
Year
ly g
row
th (
mm
2 )
Figure S7. Continued.
45
S3
FIG
URES
0 100 200 300 400 500
0.0
0.4
0.8
Adelia triloba
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Alchornea costaricensi
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Alibertia edulis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Allophylus
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Alseis blackiana
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Andira inermis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Annona acuminata
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Annona spraguei
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Apeiba membranacea
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Aspidosperma spruceanum
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Astronium graveolens
Dbh (mm)Ye
arly
sur
viva
l pro
babi
lity
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Beilschmiedi pendula
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Brosimum alicastrum
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Calophyllum longifolium
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Capparis frondosa
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Casearia aculeata
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Figure S8. Observed and predicted annual tree survival as a function of initial dbh (mm). Black dots show observations, reddashed lines show species-specific moving averages (from a Generalized Additive Model using a loess smoother), blue dashedlines show the fitted trait based average model, and grey lines show a fitted model based only on size (not on traits, thusidentical for all species).
46
S3
FIG
URES
0 100 200 300 400 500
0.0
0.4
0.8
Casearia arborea
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Cassipourea elliptica
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Chrysochlamy eclipes
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Chrysophyllu argenteum
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Chrysophyllu cainito
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Coccoloba coronata
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Coccoloba manzinellens
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Cordia alliodora
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Cordia bicolor
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Coussarea curvigemmia
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Cupania rufescens
Dbh (mm)Ye
arly
sur
viva
l pro
babi
lity
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Cupania seemannii
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Desmopsis panamensis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Drypetes standleyi
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Erythroxylum panamense
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Eugenia coloradoensi
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Figure S8. Continued.
47
S3
FIG
URES
0 100 200 300 400 500
0.0
0.4
0.8
Eugenia galalonensis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Eugenia nesiotica
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Eugenia oerstediana
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Faramea occidentalis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Garcinia intermedia
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Garcinia madruno
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Guapira standleyana
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Guarea guidonia
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Guatteria dumetorum
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Gustavia superba
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Hamelia axillaris
Dbh (mm)Ye
arly
sur
viva
l pro
babi
lity
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Hampea appendiculat
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Hasseltia floribunda
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Heisteria acuminata
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Heisteria concinna
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Herrania purpurea
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Figure S8. Continued.
48
S3
FIG
URES
0 100 200 300 400 500
0.0
0.4
0.8
Hirtella triandra
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Hybanthus prunifolius
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Hieronyma alchorneoide
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Inga goldmanii
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Inga marginata
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Inga nobilis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Inga acuminata
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Inga sapindoides
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Inga thibaudiana
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Inga umbellifera
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Jacaranda copaia
Dbh (mm)Ye
arly
sur
viva
l pro
babi
lity
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Lacistema aggregatum
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Lacmellea panamensis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Laetia thamnia
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Licania hypoleuca
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Licania platypus
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Figure S8. Continued.
49
S3
FIG
URES
0 100 200 300 400 500
0.0
0.4
0.8
Lonchocarpus heptaphyllus
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Luehea seemannii
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Mosannona garwoodii
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Maquira guianensis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Mouriri myrtilloides
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Nectandra cissiflora
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Ocotea cernua
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Ocotea puberula
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Trophis caucana
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Ormosia coccinea
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Ormosia macrocalyx
Dbh (mm)Ye
arly
sur
viva
l pro
babi
lity
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Ouratea lucens
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Palicourea guianensis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Pentagonia macrophylla
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Perebea xanthochyma
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Picramnia latifolia
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Figure S8. Continued.
50
S3
FIG
URES
0 100 200 300 400 500
0.0
0.4
0.8
Platymiscium pinnatum
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Platypodium elegans
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Poulsenia armata
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Pourouma bicolor
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Pouteria reticulata
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Prioria copaifera
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Protium panamense
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Protium tenuifolium
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Psychotria horizontalis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Psychotria marginata
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Pterocarpus rohrii
Dbh (mm)Ye
arly
sur
viva
l pro
babi
lity
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Quararibea asterolepis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Quassia amara
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Randia armata
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Simarouba amara
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Sloanea terniflora
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Figure S8. Continued.
51
S3
FIG
URES
0 100 200 300 400 500
0.0
0.4
0.8
Sorocea affinis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Spondias mombin
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Spondias radlkoferi
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Stylogyne turbacensis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Symphonia globulifera
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Tabebuia rosea
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Tabernaemont arborea
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Tachigali versicolor
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Talisia nervosa
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Talisia croatii
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Tetragastris panamensis
Dbh (mm)Ye
arly
sur
viva
l pro
babi
lity
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Thevetia ahouai
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Trichilia pallida
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Trichilia tuberculata
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Triplaris cumingiana
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Turpinia occidentalis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Figure S8. Continued.
52
S3
FIG
URES
0 100 200 300 400 500
0.0
0.4
0.8
Unonopsis pittieri
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Virola sebifera
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Virola surinamensis
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Xylopia macrantha
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
0 100 200 300 400 500
0.0
0.4
0.8
Zanthoxylum panamense
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Dbh (mm)
Year
ly s
urvi
val p
roba
bilit
y
Figure S8. Continued.
53
S3 FIGURES
●
●
●
●
●
●●
●●●
●
●●
●
●
● ●
●
●●
●
●●
●
●●
● ● ●
●●
● ●●
●●●
●
●●
●
●
●
●
●●
●
●
●
●
●● ●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●●
●●
●
●●
●
●
●
●
●●
●
●
●
●
●●
50 100
200
500
1000
2000
5
10
20
50
100
200
500
1000
Maximum diameter (mm)
Rep
rodu
ctiv
e th
resh
old
size
(m
m d
iam
eter
)
R2=0.807
Figure S9. Relationship between species’ maximum observed tree diameter(Dmax) and the onset of reproduction (50 percent quantile of reproduction;R50). The dashed line indicates the relationship R50 = 0.5Dmax).
54
S3
FIG
URES
200 400 600 800 1200
2040
60SM
size
sdl g
rowt
h −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−++++++++++++++++
Figure S10. Separate Analyses at Every Size. Fitted effects of each trait (columns) on size-dependent vital rates (rows), asin Figure 2, but based on separate models for each size (the robustness analysis). The black lines present the vital rate-sizerelationships with all traits set to their mean values (with the corresponding predictions from the hierarchical model in grey).The blue and red lines present the same relationships with one trait set to its mean plus or minus one standard deviation,respectively, and the two remaining traits set to their mean values (with exception of Dmax, see below). The trait whose valuevaries among the blue, black and red lines is named at the top of each column. Actual mean values and standard deviations foreach trait are given in table 1. An important difference from figure 3 is that for this analysis here, small statured species dropout at larger sizes and thus a different subset of species are used at each size (mean Dmax increases); in contrast, the result infigure 3 based on the hierarchical model simply shows the main effects as fitted across sizes. The size ranges of the axis differfrom Fig. 2 also because species drop out with increasing size, and we limited the analysis include at least 15 species.
55
S3
FIG
URES
Figure S11. Separate Analyses at Every Size. Among-species variance explained by traits (as measured by R2 values) throughout thelife-cycle of tree species from Barro Colorado Island, as in Figure 3, but based on separate models for each size (the robustness analysis).The rows show results for (top to bottom) size-dependent growth, size-dependent survival, and vital rates associated with reproduction.The R2 value at the upper edge of the stacked colors represents the proportion of the total variation among species explained by traits,as reflect in the averaged model when species-specific values for that vital rate are regressed on all four traits (i.e. including traits Dmax,LMA, SM and WD) and model averaging is applied, where the observed values for each species are based on species-specific GAMs. Therelative importance of different traits is indicated by the relative height of each color band as a proportion of the total, with height scaledto the R2 values for averaged models including only one trait (i.e. including only Dmax, LMA, SM or WD). The number of speciesincluded in the analyses depends on size; for reference, species numbers are shown in solid circles above each graph at fixed intervals,with their diameters proportional to log species number.
56
REFERENCES
References
James. S Clark, Miles R Silman, Ruth. Kern, Eric. Macklin, and Janneke. Hil-leRisLambers. Seed dispersal near and far: patterns across temperate andtropical forests. Ecology, 80(5):1475–1494, 1999.
H C Muller-Landau, S J Wright, O Calderon, R Condit, and S P Hubbell.Interspecific variation in primary seed dispersal in a tropical forest. Journalof Ecology, 96:653–667, 2008.