International Journal of Ecology

International Journal of Ecology / 2014 / Article

Research Article | Open Access

Volume 2014 |Article ID 389409 |

Fabien Wagner, Vivien Rossi, Christopher Baraloto, Damien Bonal, Clément Stahl, Bruno Hérault, "Are Commonly Measured Functional Traits Involved in Tropical Tree Responses to Climate?", International Journal of Ecology, vol. 2014, Article ID 389409, 10 pages, 2014.

Are Commonly Measured Functional Traits Involved in Tropical Tree Responses to Climate?

Academic Editor: Pavlos Kassomenos
Received07 Mar 2014
Revised30 May 2014
Accepted30 May 2014
Published25 Jun 2014


Climate models predict significant rainfall reduction in Amazonia, reducing water availability for trees. We present how functional traits modulate the tree growth response to climate. We used data from 3 years of bimestrial growth measurements for 204 trees of 53 species in the forest of Paracou, French Guiana. We integrated climate variables from an eddy covariance tower and functional trait values describing life history, leaf, and stem economics. Our results indicated that the measured functional traits are to some extent linked to the response of trees to climate but they are poor predictors of the tree climate-induced growth variation. Tree growth was affected by water availability for most of the species with different species growth strategies in drought conditions. These strategies were linked to some functional traits, especially maximum height and wood density. These results suggest that (i) trees seem adapted to the dry season at Paracou but they show different growth responses to drought, (ii) drought response is linked to growth strategy and is partly explained by functional traits, and (iii) the limited part of the variation of tree growth explained by functional traits may be a strong limiting factor for the prediction of tree growth response to climate.

1. Introduction

Lower tree growth rates have been reported from all tropical regions in extreme El Niño events, when temperatures reach absolute records and severe droughts occur in many areas [13]. Over Amazonia, temperature increases have already been observed [4] and climate models predict drying for the century [57].

The response of tropical forests to fine climate variations or to particular climatic events has received increasing attention during recent years [1, 810], highlighting the sensitivity of tropical forest ecosystems to the climatic variables for which a change is expected, especially temperature and precipitation. But understanding the diversity of responses to climate variations is needed to predict long-term changes in forest dynamics and community structure. Investigating such a research question is challenging because of the high diversity of tree species in most tropical tree communities, often above 100 species·ha−1.

Functional trait-based approaches offer a promising way to bypass species when modeling dynamics of highly-diverse communities [11, 12]. Recently, Herault et al. [13] showed that the growth of 50 neotropical tree species can be accurately predicted by four functional traits. Crossing from species to functional traits has several advantages, including (i) strongly decreasing the number of model parameters and (ii) allowing biological and ecological interpretations of the growth trajectory. Handling species by their shared morphological, physiological, and phenological characteristics is the central theme of this approach. Several functional traits, known as the Leaf-Height-Seed strategy framework, have been proposed to reflect important functional axes that drive plant performance [14, 15]. An important recent breakthrough was the description of a leaf economics spectrum (LES) which reflects a tradeoff between investments in productive leaves with rapid turnover versus costly physical leaf structure with a longer payback [16]. However, recently Osnas et al. [17] showed that the LES is an artifact of normalization by leaf mass per area (LMA). The normalization independent LES turns out to be very weak. Height reflects the final plant stature and its vertical position in the vegetation, and, thus, its access to light [18]. Seed mass, although not physiologically related to adult plant growth, is an important indicator of the overall species life-history strategy, with fast-growing pioneers tending to have smaller seeds [19]. Baraloto et al. [20] have recently pointed out that wood density is decoupled from leaf traits and they propose a stem economics axis of trait variation. In the following, we detailed the interactions between climate and functional traits expected to affect tree growth. We focused on the trait/climate growth rate interactions in adult trees only, because the role of functional traits in tree growth may change during ontogeny [21].

Among life-history traits, maximum height reflects the adult place in the vertical profile of the canopy, a proxy of light availability [22]. Poorter et al. [23] suggested that tall trees, that is, having crowns in the forest canopy, are exposed to drier conditions more than understorey species. In other words, tall trees often fail to maintain their evaporative demand in dry periods. Granier et al. [24], using sapflow measurements, observed that big trees, that is, trees having large DBH and/or , have higher water demands. For some dominant trees, total tree sapflow reached values as high as 30 kg·h−1 (250–300 kg·day−1). Nepstad et al. [25] used an experimental drought (reduction of 60% of incoming rainfall) to show that, after 3.2 years, the treatment resulted in a 38% increase in mortality rates. Mortality rates increased 4.5-fold among large trees (DBH > 30 cm) and 2-fold among medium trees (10–30 cm dbh). Even if Nepstad et al. [25] did not focus on tree growth, a diminution of tree growth preceding death is often observed [26, 27]. In this paper, we hypothesize that species with high diameter max and height max as well as the biggest individual trees would be much more sensitive to drought events and that drought would negatively affect their growth. Seed mass, even if not directly related to adult plant performance, is an important index of life-history strategy of species, and fast-growing species tend to have small seeds [19, 28]. We hypothesize that species with small seeds could be more sensitive to drought conditions because of their tendency to be fast-growing species, that is, a less conservative resource strategy.

Along the stem economic spectrum, bark thickness may have no direct link with the climate response as this trait is mainly linked to the tree defence strategy [29]. But bark thickness is indirectly linked to diameter increment due to the shrinkage of stems during dry seasons [30]. Trees with the highest bark water contents and thicknesses displayed the strongest decreases in girth during the dry season as well as the highest increases at the start of the wet season [31]. Santiago et al. [32] demonstrated that wood density is well linked to hydraulic efficiency, in turn positively correlated to the photosynthetic potential. The ecological strategy “many narrow conduits” versus “fewer wider conduits” highlights the tradeoff between the hydraulic efficiency of wide conduits and the cavitation resistance of narrow conduits [33]. This cavitation resistance is a potentially great asset to the growth of species with high wood densities when the available soil water content starts to decline at the end of the wet season. This is because species with low densities are known to be more sensitive to water availability and to have a less conservative resource strategy [34]. In this paper, we hypothesize that species with high wood density would be less sensitive to drought events and that drought would negatively affect the growth of softwood species.

Along the leaf economic spectrum, specific leaf area (SLA) indicates the leaf efficiency for light capture per unit biomass invested [35]. SLA is weakly or not related to the growth of large-sized trees [23] for which leaf area and light interception are largely determined by branching patterns, the number of meristems, and tree architecture [36]. Species with high SLA tend to have (i) short-lived leaves with high N and P mass contenst [16, 37], (ii) high mass-based maximum photosynthetic rates [38] and are usually found at the fast payback end of the leaf economic spectrum [20]. This combination of traits is usually found in species with fast growth rates [28]. At the opposite end, leaf traits associated with low SLA (thick leaf blade; small, thick-walled cells) have often been interpreted as ecological adaptations to drought, that is, allowing, at least in evergreen species, continued leaf function (or at least postponing leaf death) under very dry conditions [16, 39]. A tight correlation amongst leaf K and leaf  N or P concentrations was highlighted among neotropical trees [20]. Leaf K concentration is associated with stomatal control. K+ concentration when stomata are open is nearly ten times larger than the K+ concentration when stomata are closed [40]. Potassium limitation has been shown to have a negative effect on the ability of plants to control stomata, often leading to a reduction in water-use efficiency [41, 42]. It may be expected that species with low values of K might be more sensitive to drought as they have less flexibility to control stomata apertures, and, thus, less flexibility to limit the loss of water. Finally, can provide a direct estimate of a tree’s water use efficiency, WUE [43]. Leaf is often thought to be an integrative indicator of stomatal response to atmospheric dryness [44, 45] and is strongly correlated with SLA. Species with low water use efficiency have been found to have small SLA [20, 37, 46]. This trait has not been directly used by the existing coupled vegetation/climate models but would allow assignment of a key parameter for drought adaptation [47]. Here, soft traits referred to traits that are relatively easy and quick to quantify, often good correlates of hard traits, which may be more accurate indicators of plant functions responsible for responses or effects at the ecosystem or biome scale, but which are difficult to quantify for large numbers of species in many regions of the world [38, 48]. In this paper, we hypothesize that drought conditions could first affect species with a high SLA (high water use efficiency, soft leaf trait). Hard leaf traits such as leaf water potential at turgor loss () and leaf water potential at full turgor () are known to give an explanation of species tolerance to drought across biomes at the worldwide scale [49]. Here, we focus on tropical rainforests and make a first analysis of the link between soft traits and tree growth response to drought, that is, the plant performance.

In this paper, we addressed three specific questions which are as follows. (i) Are functional traits involved in tropical tree growth response to seasonal variation of climate? (ii) Which traits are linked to resistance to drought? (iii) Finally, how far can functional traits help in predicting the future of tropical forests in terms of composition toward expected changes of climate? We hypothesize that life-history traits and stem economic spectrum traits, more so than leaf traits, could explain the growth response to drought, that is, the capacity or not to maintain diameter growth during drought periods. Furthermore, we suspect that the tree growth response to climate is associated with the growth strategies of tropical trees, fast growing species being more affected by fast changes of climate conditions in terms of growth.

2. Methods

2.1. Site

The study site is located in Paracou, French Guiana (5°18′N, 52°23′W), a lowland tropical rainforest near Sinnamary [50]. The climate is affected by the north/south movements of the Intertropical Convergence Zone and the site receives nearly two-thirds of its annual 3041 mm of precipitation between mid-March and mid-June, and less than 50 mm per month in September and October. The most common soils in Paracou are the shallow ferralitic soils limited in depth by a more or less transformed loamy saprolite [50]. The site is located approximately 40 m above sea level [50] and is made up of a succession of small hills. The forest is typical to Guianan rainforests [51]. More than 550 woody species attaining 2 cm DBH (diameter at Breast Height, i.e., 130 cm) have been described at the site, with an estimated 160 species of trees > 10 cm DBH per hectare [52]. The dominant families at the site include Leguminosae, Chrysobalanaceae, Lecythidaceae, Sapotaceae, and Burseraceae.

2.2. Data

Seasonal changes in trunk circumference were monitored in 204 trees from 53 species using home-made steel dendrometer bands [31]. Tree growth was censused every c. 40 days from 2007 to 2010 (mean = 39, sd = 19.8). Information on 13 key functional traits (related to leaf economics, stem economics, and life-history) measured on mature trees (DBH > 10 cm) of the 53 focal species was extracted from a trait database of French Guianan tree species [53] (see Table S1.1 in Supplementary Material available online at The values of functional traits used in this study are mean-species values for French Guiana and Paracou is amongst the sites used for functional trait measurements in this database. Details of sample collection and trait measurements are available in Baraloto et al. [20] for the leaf and stem economic traits and in the study by Herault et al. [22] for the life-history traits. The 95th percentile diameters were used as estimates of to minimize the weight of outliers [54]. These traits represent complementary axes of tree ecological strategies in tropical forests (Table 1) and have been proposed to be important for tree performance through ontogeny [14, 23, 55] and for growth responses to climate variations [20, 33, 45]. Due to a strong correlation between bark thickness and on one hand and current tree size on the other [29], bark thickness and were corrected with the method given in Baraloto et al. [20].

RolesUnitsSample rangeaFG rangebWETcDRYdReferences

 DBHmaxArchitecture and structuremm190–98010–1254+*−*Granier et al. [24]; Nepstad et al. [25, 70]; Monserud [26]; Wyckoff and Clark [27]
 HeightmaxArchitecture and structurem18–508–56+Poorter et al. [23]; Herault et al. [13]
 Seed massReproductive investmentg<0.01–>20<0.01–>20+Moles et al. [19]; Poorter and Bongers [28]

Stem economic spectrum
 Wood densityTransport structure defenseg·cm−30.38–0.840.28–0.88+Santiago et al. [32]; Chave et al. [33]; Markesteijn et al. [34]
 Bark thicknessDefensemm(−4.7)–13(−6.2)–13.4Stahl et al. [31]; Baker et al. [30]

Leaf economic spectrum
 SLAResource capture, defensecm2·g−15.41–164.02–24.78+Poorter et al. [23, 35]; Sterck and Bongers [36]; Wright et al. [16]; Fyllas et al. [37]; Cornelissen et al. [38]; Poorter and Bongers [28]
 KResource captureg·g−1 +Roelfsema and Hedrich [40]; Marschner [41]; Santiago and Wright [42]
 NResource captureg·g−1 +Baraloto et al. [20]; Poorter and Bongers [28]
 PResource captureg·g−1 +Baraloto et al. [20]; Poorter and Bongers [28]
 C/NResource capture, defenseg·g−113–3411–47+Baraloto et al. [20]; Poorter and Bongers [28]
 LTDResource capture, defenseg·cm−3 +Wright et al. [16]; Donovan et al. [39]
  ChlorophyllResource capture g·mm−248–10021–150+Wright et al. [16]; Baraloto et al. [20]; Cornelissen et al. [38]
 ToughnessDefenseN0.85–3.40.2–11.4+Wright et al. [16]; Donovan et al. [39]
δ13CResource capture (−1.7)–2.5(−2.9)–4.8+Farquhar et al. [43]; Prentice et al. [44]; Lamont et al. [46]; Fyllas et al. [37]; Baraloto et al. [20]

Range of traits values of the Paracou sample, brange of traits values of French Guiana (FG) extract from the BRIDGE database [20], cexpected growth response to wet conditions, dexpected growth response to drought conditions, *(+) trees with high values of a given functional trait are expected to grow faster, and (−) trees with high values of a given functional trait are expected to grow slower.
2.3. Meteorological Data

In 2003, a 55 m self-supporting metallic eddy covariance flux tower, Guyaflux, was built in the Paracou forest in a natural 100 m2 gap, with minimal disturbance to the upper canopy. This location covers a range of more than 1 km of forest in the direction of the prevailing winds. The top of the tower is about 20 m higher than the overall canopy. Meteorological and eddy flux sensors (full details in Bonal et al. [56]) were mounted 3 m above the tower. A large set of climatic variables were recorded at a daily time step (supporting information, Table S2.1). Most climatic variables exhibited strong seasonal changes, highlighting the north/south movements of the Intertropical Convergence Zone (supporting information, Figure S2.1).

We used a water balance model explicitly designed for tropical forests [57] to estimate water availability for trees. The model computes daily water fluxes (tree transpiration, understorey evapotranspiration, rainfall interception, and drainage), soil water content at different layers, and relative extractable water for trees for the entire soil (REW). REW is a daily value between 0 and 1; when REW = 1, the amount of extractable water by the tree is at its maximum, and when REW = 0, no water is available for trees. We created with REW a variable of water stress, SWD for soil water deficit, SWD = 1 − REW.

2.4. Data Analysis

As an initial exploration of the covariation among functional traits, we investigated their associations through a principle component analysis (PCA) on the normalized traits dataset. We conducted this analysis to support the interpretation of functional traits implication in tree growth response to climate. We modelled the link between tree growth and the functional traits/climate interactions in a linear regression framework. The model was built at a daily time step. We first included a discrete variable tree in the model to account for an individual tree effect in tree growth so as not to bias our result. Next we included the discrete variable period which estimates a model parameter for each census period. This variable captured the variance due to the climatic effect common to all individuals from all species. Tree and period are treated as fixed factors and estimated in the model. We then explored the interaction between functional traits and climatic variables one by one. The interaction between the current DBH and climate variables was also investigated to test the hypothesis of a size-mediated response to climate stress. All interactions were successively tested in univariate models to lower multicollinearity problems associated with multivariate models. Prior to the analysis, functional traits were scaled, that is, each value was divided by the trait standard error. The growth model was fitted with weighted least squares, with weights calculated to correct the facts that (i) growth, climatic, and traits data were summed over different period lengths and (ii) the number of studied trees varied among species. Consider: where is the growth in diameter and in mm of the tree over the period , is the individual daily tree effect, is the daily period effect, quantifies the effect of the interaction between the trait value and the climate variable , and is the number of days of the period . The weight is the product of the number of days of the period, , with the number of individuals from each species in the data, .

To test the significance of each trait/climate interaction, we computed for each univariate model the Bayesian information criterion, BIC [58]. We compared this BIC to the Bayesian information criterion value of the model that includes the fixed trait variable and the fixed climate variable but that omits their interaction, . A BIC lower than means that the information carried by the interaction improved the model even if it added a parameter.

All analyses were performed using the R project software (

3. Results

The first axis of the principal component analysis (Figure 1) explained 31.15% of the variance in the trait dataset and was positively correlated with N, P, SLA, K, , and negatively correlated to C/N, toughness, LTD, and wood density. The second axis explained 12.63% of variance and was positively correlated with bark thickness, toughness, and K.

More than 33.5% of the observed variation in tree growth could be attributed to individual tree behaviour, with the period effect explaining 15.0% of the tree growth variance. This means that climate variables alone could explain up to 15% of the variance of tree growth. The interaction term between a single trait and a single climate variable captured a maximum of 0.38% of the variance (Table 2). The BIC selection procedure retained 24 trait/climate interactions (Table 2). With the exception of relative air humidity and atmospheric pressure, all climate variables showed interactions with traits for predicting tree growth. Six of the 13 functional traits , Bark, SLA, K, P, and were not associated to any climate variables for predicting tree growth.


REWa 0.19*** (+) 0.12** (+)
Rain0.24*** (+) 0.34*** (+) 0.13** (+)
SWDb0.38*** (−)0.11** (−) 0.08** (+)0.15** (−) 0.13** (+) 0.08* (−) 0.15** (+)
Irradiance0.12** (−) 0.11** (+)0.08* (−) 0.12** (+)
Tmax0.08* (−)
Tmin0.08* (−)
Wind.speed0.09* (−) 0.09* (+) 0.08* (+)
VPDc0.08* (−)0.09* (−) 0.08* (+)

Relative extractable water, bsoil water deficit (1-REW), and cvapour pressure deficit.
3.1. Life-History

Tree species with a high grew more slowly when the soil water deficit was high, and larger trees also grew more slowly in drought conditions (high SWD and high irradiance) and when vapour pressure was high. Furthermore, trees with large DBH grew faster during the wet period.

3.2. Stem Economics

Trees with high wood densities grew more rapidly during drought events (high SWD and high irradiance).

3.3. Leaf Economics

Trees with high foliar N (and low CN ratio) grew more slowly during drought events (high SWD and high irradiance), or when vapour pressure was high and temperatures and wind speed were maximum. Trees with high leaf tissue densities had higher growths in wet seasons (high REW and high precipitation), when wind speed is high, and lower growths were observed when soil water deficit was high. Trees with high leaf chlorophyll content had a higher growth in wet conditions (high REW and high precipitation) and when wind speed was high. Finally, trees with high leaf toughness had a higher growth in drought conditions (high SWD and high irradiance).

4. Discussion

We observed direct links between functional traits and growth response to climate variations, an important step to improve global vegetation models in predicting tropical forest responses to global change [45, 59]. However, we found that the interaction term between a single trait and a single climate variable captured a significant but very thin part of the variance of tree growth, Table 2. Effects of functional traits may be noticeable throughout the whole year, as found for leaf tissue density (LTD) or current diameter (DBH) for which interactions were significant both in wet and in dry seasons (Table 2). However, most of the effects we detected were seasonal. Positive interactions with REW and/or rain indicated that a high value of a given trait was correlated with more rapid growth in wet seasons while positive interactions with irradiance, soil water, deficit and/or vapour pressure deficit highlighted better growth in dry periods. Overall, water (REW, SWD, and rain) and light (irradiance) shaped the functional trait responses. Wind speed and VPD are proxies of water and light, respectively (supporting information, Figure S2.1). Wind speed reaches its maximum value at the start of the rain season and VPD is strongly correlated to irradiance and soil water deficit (supporting information, Table S3.1). Contrary to the results of Clark et al. [1], temperature variations captured a negligible proportion of growth variance. We propose two related explanations. First, in French Guiana, temperatures are very stable throughout the year unlike other climate variables (supporting information, Figure S2.1). Second, photosynthesis requires light and water directly, whereas temperature only changes the kinetics of the reaction. As a result, within the relatively narrow range of average minimum and maximum temperatures of our study (23.4–28.4 °C), only a minor change in photosynthesis kinetics can be expected [9].

4.1. Life-History Traits

Big trees (DBH) in our study were clearly more affected by drought events than smaller ones. Big trees have higher water demands, are more exposed to light irradiance, and have their crowns in drier environments than the forest understorey [25]. Our results are therefore consistent with general trends from 18 mortality studies in Amazonia, 12 of which have demonstrated higher mortality rates for larger trees during drought events, three found no size-related effect, and three indicated higher mortality for smaller trees [60]. We also found that taller species () were more affected by drought than smaller-statured species. However, species of larger girth () did not exhibit this pattern. This highlights the fact that most of the individual trees in our study were closer to their maximum height than to their maximum diameter due to the power relationship between height and DBH. This also explains why is a better index than for placing a tree along its ontogenetic trajectory [13]. Considering this result we suggest that individual tree DBH, and to a lesser extent the species , can be used as an index of tree growth response to water stress.

4.2. Stem Economics

Consistent with the assumption that dense-wood species have smaller vessels that prevent them from embolism and that allow them to maintain baseline growth under drought conditions [33, 34], we found that trees with the densest wood were the least affected by drought. In contrast, fast-growing species with large vessels and lighter wood appeared to be more sensitive to drought. And we should keep in mind that this growth response to drought may be coupled with a buckling resistance and a biomechanic strength due to more fibres that constitute themselves as important survival attributes for the trees with the densest wood. We found no relationship between bark thickness and tree growth response. Even if this trait did not mediate the tree response to drought, bark thickness is well known as a good proxy for resistance to fire [29] which may increase under future climate-change scenarios for the Amazon basin [61].

4.3. Leaf Economics

Recent studies have suggested that leaf traits do not explain differential growth of adult trees in tropical forests [13, 23]. However, our results showed that some leaf traits could be implicated in tree growth response to climate (N, CN, LTD, Chlo, and Tough). Species with high nutrient concentrations are usually found at the fast payback end of the “economic spectrum” [16]. On the other hand, leaf traits associated with low SLA (thick leaf blade; small, thick-walled cells) have often been interpreted as adaptations that allow continued leaf functioning (or at least postponing leaf death) under very dry conditions [16, 39]. Our results are not consistent with this hypothesis. In our dataset, N leaf concentration was more discriminating than SLA for tree growth response to drought. At a global scale, there is a strong negative correlation between leaf N content and leaf lifespan [16, 39]. It may be expected that trees with long-lived leaves suffer less from regular drought events while trees with short-lived leaves senesce their leaves when stressed and, consequently, stop their secondary growth (Table 2) [62]. Leaf toughness, chlorophyll content, and leaf tissue density are intimately correlated (Table 1) but their effect seems to be associated with different seasons. In the wet season, species with high chlorophyll content and high leaf tissue densities have more rapid growth (Table 2). This could be linked to the capacity of species with high value of chlorophyll to have better capacity to use the lower light levels of the wet season. Leaf toughness traits are uncorrelated with relative growth rates but correlated positively with leaf lifespan [63, 64]. Furthermore, leaf toughness is negatively correlated with other leaf traits favouring fast growth [20]. Species with a high toughness could have a higher growth in drought conditions, that is, higher temperatures and irradiance, due to their position on the leaf economic spectrum, as opposed to fast growing species. was not correlated with growth response to climate. Foliar is an indicator of leaf-level water-use efficiency reflecting a tradeoff between photosynthetic rates and stomatal conductance [65]. With the complete French Guianan tree functional trait database (668 species), Baraloto et al. [20] have found that foliar was weakly correlated with LES and SES, when corrected for stature. They conclude that the sensitivity of this measure to microhabitat variation may preclude its utility as a plant functional trait at the community scale in systems with a complex vertical structure such as tropical forests.

4.4. Rethinking the Tree Growth Response

We present a unique dataset integrating climatic variables and tree growth records with trait values at high resolution. Yet at best our model explains 48.9% of the variance of tree growth. Surprisingly, the interactions between traits and climate, even if some were significant and followed the general frame of the expected results, explained only a limited part of the variation in tree growth, no more than 0.4% each. This may be a strong limiting factor for use of a functional trait modeling framework to predict future growth in a climate change context. Further research will be needed to explore the functional link between traits and tree growth response to climate, including the integration of some hard traits that are not typically included in functional trait screenings, including leaf and stem anatomies and leaf turgor pressure [49, 66, 67]. During the 3 years of the study, climatic conditions have followed the classical pattern of wet season and dry season successions. Even the dry season can be considered as particularly dry for tropical forest with 3-4 months with less than 50 mm; no extreme changes in climate were observed on the period of the study. The dry period could be usual for the investigated trees, and they may not have used extreme strategies of growth which are assumed to be mediated by some functional traits. The majority of trees stopped their growth in the dry season. But some species are able to maintain their growth more days than others (Figure 2) and we point out in this study that fast growing species growth rates could be more impacted than slow growing species due to their less conservative resource-use strategy. Anyway, it might be interesting to add memory in modelling annual growth. Indeed, our results suggest that growth is at its maximum at the beginning of the rainy season, when water and light are available, and the leaves are new. Similar climatic conditions are frequent during the long rainy season (January to June) but the leaves age [68] and secondary growth decreases symmetrically (Figure 2). A model taking into account the ageing of leaves, for example, through a decreasing exponential, should be explored. In order to develop predictive model of CO2 fluxes between the forest and the atmosphere; further investigations are needed as tree growth is one of the demographic variables, and trees may respond to climate change in other demographic dimensions, for example, mortality and reproduction, that may or may not show similar climate interactions.

4.5. Conclusion

With the absence of extreme climatic events during our study, we cannot conclude that a composition change is expected. However, our results join the recent finding that smaller, denser-wooded trees are able to maintain their functioning in drought conditions [33, 60]. In such cases where climate variation is regular and implies quite thin variance explanation, using traits in analysis must be done with the traits closer to the processes. For example, N, P, and K are very correlated but only N improves the model, probably because this trait is closest to the process linking tree growth and climate, as N is a proxy of the maximum photosynthetic rate and maximum growth rate [38]. Among the climate variables, we found no effect of trait interaction with temperature on tree growth. The mean temperature at our site is near the optimal range for photosynthesis and is stable throughout the year. But, if temperature follows the expected increase for the 21st century [69], temperature will impact tree growth by reducing photosynthesis kinetics [9].

In this study, each growth measurement is assumed to be independent of others. This is a strong, but common, assumption for growth modeling. Our results suggest that a single heavy rain in the dry season does not have the same effect on tree growth as a heavy rain in the wet season. This practically compels one to analyse tree growth with dynamic models. In the absence of extreme climate events, trait-based community dynamic models appear to be the only way to investigate extreme climatic scenarios, such as those predicted for the 21st century. Such modeling approaches, combined with experiments on unusual extreme events [70], are urgently needed to test the sensitivity of tropical forests to climate change.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.


Funding came from the project Climfor (Fondation pour la Recherche sur la Biodiversité,, from the project Guyasim (European structural fundings, PO-Feder, task=view id=14 Itemid=206), and from the Fapesp (Fundação de Amparo à Pesquisa do Estado de São Paulo, processo 13/14520-6, This work has also benefited from an “Investissement d’Avenir” Grant managed by Agence Nationale de la Recherche (CEBA, ref. ANR-10-LABX-0025). This study has been possible thanks to the important field work realized at Paracou by B. Burban, J.-Y. Goret, J. Cazal, F. Bompy, J. Sermage, and S. Thibaudeau. The authors thank all participants of the project BRIDGE who participated in field and laboratory collection and treatment of specimens. They thank two anonymous referees for useful comments on previous drafts of the paper.

Supplementary Materials

The Supplementary Material presents: (i) descriptive statistics of the functional traits and climate variables used to study seasonal tree growth community response to seasonal climate variations (Supplementary Material S1); and (ii) correlation between climate data (Supplementary Material S2).

  1. Supplementary Material


  1. D. B. Clark, D. A. Clark, and S. F. Oberbauer, “Annual wood production in a tropical rain forest in NE Costa Rica linked to climatic variation but not to increasing CO2,” Global Change Biology, vol. 16, no. 2, pp. 747–759, 2010. View at: Publisher Site | Google Scholar
  2. D. A. Clark, “Sources or sinks? The responses of tropical forests to current and future climate and atmospheric composition,” Philosophical Transactions of the Royal Society B: Biological Sciences, vol. 359, no. 1443, pp. 477–491, 2004. View at: Publisher Site | Google Scholar
  3. N. McDowell, W. T. Pockman, C. D. Allen et al., “Mechanisms of plant survival and mortality during drought: why do some plants survive while others succumb to drought?” New Phytologist, vol. 178, no. 4, pp. 719–739, 2008. View at: Publisher Site | Google Scholar
  4. Y. Malhi and J. Wright, “Spatial patterns and recent trends in the climate of tropical rainforest regions,” Philosophical Transactions of the Royal Society B: Biological Sciences, vol. 359, no. 1443, pp. 311–329, 2004. View at: Publisher Site | Google Scholar
  5. Y. Malhi, L. E. O. C. Aragao, D. Galbraith et al., “Exploring the likelihood and mechanism of a climate-change-induced dieback of the Amazon rainforest,” Proceedings of the National Academy of Sciences of the United States of America, vol. 106, no. 49, pp. 20610–20615, 2009. View at: Publisher Site | Google Scholar
  6. E. J. Burke, S. J. Brown, and N. Christidis, “Modelling the recent evolution of global drought and projections for the twenty-first century with the Hadley Centre climate model,” Journal of Hydrometeorology, vol. 7, no. 5, pp. 1113–1125, 2006. View at: Publisher Site | Google Scholar
  7. T. C. Johns, J. M. Gregory, W. J. Ingram et al., “Anthropogenic climate change for 1860 to 2100 simulated with the HadCM3 model under updated emissions scenarios,” Climate Dynamics, vol. 20, no. 6, pp. 583–612, 2003. View at: Google Scholar
  8. P. M. Brando, D. C. Nepstad, E. A. Davidson, S. E. Trumbore, D. Ray, and P. Camargo, “Drought effects on litterfall, wood production and belowground carbon cycling in an Amazon forest: results of a throughfall reduction experiment,” Philosophical Transactions of the Royal Society B: Biological Sciences, vol. 363, no. 1498, pp. 1839–1848, 2008. View at: Publisher Site | Google Scholar
  9. J. Lloyd and G. D. Farquhar, “Effects of rising temperatures and [CO2] on the physiology of tropical forest trees,” Philosophical Transactions of the Royal Society B: Biological Sciences, vol. 363, no. 1498, pp. 1811–1817, 2008. View at: Publisher Site | Google Scholar
  10. F. Wagner, V. Rossi, C. Stahl, D. Bonal, and B. Herault, “Water availability is the main climate driver of neotropical tree growth,” PLoS ONE, vol. 7, no. 4, Article ID e34074, 2012. View at: Publisher Site | Google Scholar
  11. S. Diaz, J. G. Hodgson, K. Thompson et al., “The plant traits that drive ecosystems: evidence from three continents,” Journal of Vegetation Science, vol. 15, no. 3, pp. 295–304, 2004. View at: Google Scholar
  12. S. Lavorel and E. Garnier, “Predicting changes in community composition and ecosystem functioning from plant traits: revisiting the Holy Grail,” Functional Ecology, vol. 16, no. 5, pp. 545–556, 2002. View at: Publisher Site | Google Scholar
  13. B. Herault, B. Bachelot, L. Poorter et al., “Functional traits shape ontogenetic growth trajectories of rain forest tree species,” Journal of Ecology, vol. 99, no. 6, pp. 1431–1440, 2011. View at: Publisher Site | Google Scholar
  14. M. Westoby, “A leaf-height-seed (LHS) plant ecology strategy scheme,” Plant and Soil, vol. 199, no. 2, pp. 213–227, 1998. View at: Publisher Site | Google Scholar
  15. M. Westoby, D. S. Falster, A. T. Moles, P. A. Vesk, and I. J. Wright, “Plant ecological strategies: some leading dimensions of variation between species,” Annual Review of Ecology and Systematics, vol. 33, pp. 125–159, 2002. View at: Publisher Site | Google Scholar
  16. I. J. Wright, P. B. Reich, M. Westoby et al., “The worldwide leaf economics spectrum,” Nature, vol. 428, no. 6985, pp. 821–827, 2004. View at: Publisher Site | Google Scholar
  17. J. L. D. Osnas, J. W. Lichstein, P. B. Reich, and S. W. Pacala, “Global leaf trait relationships: mass, area, and the leaf economics spectrum,” Science, vol. 340, no. 6133, pp. 741–744, 2013. View at: Publisher Site | Google Scholar
  18. T. Hirose and M. J. A. Werger, “Maximizing daily canopy photosynthesis with respect to the leaf nitrogen allocation pattern in the canopy,” Oecologia, vol. 72, no. 4, pp. 520–526, 1987. View at: Publisher Site | Google Scholar
  19. A. T. Moles, D. S. Falster, M. R. Leishman, and M. Westoby, “Small-seeded species produce more seeds per square metre of canopy per year, but not per individual per lifetime,” Journal of Ecology, vol. 92, no. 3, pp. 384–396, 2004. View at: Publisher Site | Google Scholar
  20. C. Baraloto, C. E. T. Paine, L. Poorter et al., “Decoupled leaf and stem economics in rain forest trees,” Ecology Letters, vol. 13, no. 11, pp. 1338–1347, 2010. View at: Publisher Site | Google Scholar
  21. S. J. Wright, K. Kitajima, N. J. B. Kraft et al., “Functional traits and the growth-mortality trade-off in tropical trees,” Ecology, vol. 91, no. 12, pp. 3664–3674, 2010. View at: Publisher Site | Google Scholar
  22. B. Herault, J. Ouallet, L. Blanc, F. Wagner, and C. Baraloto, “Growth responses of neotropical trees to logging gaps,” Journal of Applied Ecology, vol. 47, no. 4, pp. 821–831, 2010. View at: Publisher Site | Google Scholar
  23. L. Poorter, S. J. Wright, H. Paz et al., “Are functional traits good predictors of demographic rates? Evidence from five neotropical forests,” Ecology, vol. 89, no. 7, pp. 1908–1920, 2008. View at: Publisher Site | Google Scholar
  24. A. Granier, R. Hue, and S. T. Barigah, “Transpiration of natural rain forest and its dependence on climatic factors,” Agricultural and Forest Meteorology, vol. 78, no. 1-2, pp. 19–29, 1996. View at: Google Scholar
  25. D. C. Nepstad, I. M. Tohver, R. David, P. Moutinho, and G. Cardinot, “Mortality of large trees and lianas following experimental drought in an amazon forest,” Ecology, vol. 88, no. 9, pp. 2259–2269, 2007. View at: Publisher Site | Google Scholar
  26. R. Monserud, “Simulation of forest tree mortality,” Forest Science, vol. 22, pp. 438–444, 1976. View at: Google Scholar
  27. P. H. Wyckoff and J. S. Clark, “The relationship between growth and mortality for seven co-occurring tree species in the southern Appalachian Mountains,” Journal of Ecology, vol. 90, no. 4, pp. 604–615, 2002. View at: Publisher Site | Google Scholar
  28. L. Poorter and F. Bongers, “Leaf traits are good predictors of plant performance across 53 rain forest species,” Ecology, vol. 87, no. 7, pp. 1733–1743, 2006. View at: Google Scholar
  29. C. E. T. Paine, C. Stahl, E. A. Courtois, S. Patino, C. Sarmiento, and C. Baraloto, “Functional explanations for variation in bark thickness in tropical rain forest trees,” Functional Ecology, vol. 24, no. 6, pp. 1202–1210, 2010. View at: Publisher Site | Google Scholar
  30. T. R. Baker, D. F. R. P. Burslem, and M. D. Swaine, “Associations between tree growth, soil fertility and water availability at local and regional scales in Ghanaian tropical rain forest,” Journal of Tropical Ecology, vol. 19, no. 2, pp. 109–125, 2003. View at: Publisher Site | Google Scholar
  31. C. Stahl, B. Burban, F. Bompy, Z. B. Jolin, J. Sermage, and D. Bonal, “Seasonal variation in atmospheric relative humidity contributes to explaining seasonal variation in trunk circumference of tropical rain-forest trees in French Guiana,” Journal of Tropical Ecology, vol. 26, no. 4, pp. 393–405, 2010. View at: Publisher Site | Google Scholar
  32. L. S. Santiago, G. Goldstein, F. C. Meinzer et al., “Leaf photosynthetic traits scale with hydraulic conductivity and wood density in Panamanian forest canopy trees,” Oecologia, vol. 140, no. 4, pp. 543–550, 2004. View at: Publisher Site | Google Scholar
  33. J. Chave, D. Coomes, S. Jansen, S. L. Lewis, N. G. Swenson, and A. E. Zanne, “Towards a worldwide wood economics spectrum,” Ecology Letters, vol. 12, no. 4, pp. 351–366, 2009. View at: Publisher Site | Google Scholar
  34. L. Markesteijn, L. Poorter, H. Paz, L. Sack, and F. Bongers, “Ecological differentiation in xylem cavitation resistance is associated with stem and leaf structural traits,” Plant, Cell and Environment, vol. 34, no. 1, pp. 137–148, 2011. View at: Publisher Site | Google Scholar
  35. H. Poorter, U. Niinemets, L. Poorter, I. J. Wright, and R. Villar, “Causes and consequences of variation in leaf mass per area (LMA): a meta-analysis,” New Phytologist, vol. 182, no. 3, pp. 565–588, 2009. View at: Publisher Site | Google Scholar
  36. F. J. Sterck and F. Bongers, “Crown development in tropical rain forest trees: patterns with tree height and light availability,” Journal of Ecology, vol. 89, no. 1, pp. 1–13, 2001. View at: Publisher Site | Google Scholar
  37. N. M. Fyllas, S. Patino, T. R. Baker et al., “Basin-wide variations in foliar properties of Amazonian forest: phylogeny, soils and climate,” Biogeosciences, vol. 6, no. 11, pp. 2677–2708, 2009. View at: Google Scholar
  38. J. H. C. Cornelissen, S. Lavorel, E. Garnier et al., “A handbook of protocols for standardised and easy measurement of plant functional traits worldwide,” Australian Journal of Botany, vol. 51, no. 4, pp. 335–380, 2003. View at: Publisher Site | Google Scholar
  39. L. A. Donovan, H. Maherali, C. M. Caruso, H. Huber, and H. de Kroon, “The evolution of the worldwide leaf economics spectrum,” Trends in Ecology and Evolution, vol. 26, no. 2, pp. 88–95, 2011. View at: Publisher Site | Google Scholar
  40. M. R. G. Roelfsema and R. Hedrich, “In the light of stomatal opening: new insights into 'the Watergate',” New Phytologist, vol. 167, no. 3, pp. 665–691, 2005. View at: Publisher Site | Google Scholar
  41. H. Marschner, Mineral Nutrition in Higher Plants, Academic Press, London, UK, 1986.
  42. L. S. Santiago and S. J. Wright, “Leaf functional traits of tropical forest plants in relation to growth form,” Functional Ecology, vol. 21, no. 1, pp. 19–27, 2007. View at: Publisher Site | Google Scholar
  43. G. Farquhar, J. Ehleringer, and K. Hubick, “Carbon isotope discrimination and photosynthesis,” Annual Review of Plant Physiology and Plant Molecular Biology, vol. 40, pp. 503–537, 1989. View at: Publisher Site | Google Scholar
  44. I. C. Prentice, T. Meng, H. Wang, S. P. Harrison, J. Ni, and G. Wang, “Evidence of a universal scaling relationship for leaf CO2 drawdown along an aridity gradient,” New Phytologist, vol. 190, no. 1, pp. 169–180, 2011. View at: Publisher Site | Google Scholar
  45. S. M. McMahon, S. P. Harrison, W. S. Armbruster et al., “Improving assessment and modelling of climate change impacts on global terrestrial biodiversity,” Trends in Ecology and Evolution, vol. 26, no. 5, pp. 249–259, 2011. View at: Publisher Site | Google Scholar
  46. B. B. Lamont, P. K. Groom, and R. M. Cowling, “High leaf mass per area of related species assemblages may reflect low rainfall and carbon isotope discrimination rather than low phosphorus and nitrogen concentrations,” Functional Ecology, vol. 16, no. 3, pp. 403–412, 2002. View at: Publisher Site | Google Scholar
  47. B. E. Medlyn, R. A. Duursma, D. Eamus et al., “Reconciling the optimal and empirical approaches to modelling stomatal conductance,” Global Change Biology, vol. 17, no. 6, pp. 2134–2144, 2011. View at: Publisher Site | Google Scholar
  48. J. G. Hodgson, P. J. Wilson, R. Hunt, J. P. Grime, and K. Thompson, “Allocating C-S-R plant functional types: a soft approach to a hard problem,” Oikos, vol. 85, no. 2, pp. 282–294, 1999. View at: Google Scholar
  49. M. K. Bartlett, C. Scoffoni, and L. Sack, “The determinants of leaf turgor loss point and prediction of drought tolerance of species and biomes: a global meta-analysis,” Ecology Letters, vol. 15, no. 5, pp. 393–405, 2012. View at: Publisher Site | Google Scholar
  50. S. Gourlet-Fleury, J. M. Guehl, and O. Laroussinie, Ecology and Management of a Neotropical Rainforest: Lessons Drawn from Paracou, a Long-Term Experimental Research Site in French Guiana, Elsevier, 2004.
  51. H. ter Steege, N. C. A. Pitman, O. L. Phillips et al., “Continental-scale patterns of canopy tree composition and function across Amazonia,” Nature, vol. 443, no. 7110, pp. 444–447, 2006. View at: Publisher Site | Google Scholar
  52. C. Baraloto, O. J. Hardy, C. E. T. Paine et al., “Using functional traits and phylogenetic trees to examine the assembly of tropical tree communities,” Journal of Ecology, vol. 100, no. 3, pp. 690–701, 2012. View at: Publisher Site | Google Scholar
  53. M. Ollivier, C. Baraloto, and E. Marcon, “A trait database for Guianan rain forest trees permits intra- and inter-specific contrasts,” Annals of Forest Science, vol. 64, no. 7, pp. 781–786, 2007. View at: Publisher Site | Google Scholar
  54. D. A. King, S. J. Davies, and N. S. M. Noor, “Growth and mortality are related to adult tree size in a Malaysian mixed dipterocarp forest,” Forest Ecology and Management, vol. 223, pp. 152–158, 2006. View at: Publisher Site | Google Scholar
  55. E. Weiher, A. van der Werf, K. Thompson, M. Roderick, E. Garnier, and O. Eriksson, “Challenging theophrastus: a common core list of plant traits for functional ecology,” Journal of Vegetation Science, vol. 10, no. 5, pp. 609–620, 1999. View at: Google Scholar
  56. D. Bonal, A. Bosc, S. Ponton et al., “Impact of severe dry season on net ecosystem exchange in the Neotropical rainforest of French Guiana,” Global Change Biology, vol. 14, no. 8, pp. 1917–1933, 2008. View at: Publisher Site | Google Scholar
  57. F. Wagner, B. Hérault, C. Stahl, D. Bonal, and V. Rossi, “Modeling water availability for trees in tropical forests,” Agricultural and Forest Meteorology, vol. 151, no. 9, pp. 1202–1213, 2011. View at: Publisher Site | Google Scholar
  58. G. Schwarz, “Estimating the dimension of a model,” Annals of Statistics, vol. 6, no. 2, pp. 461–464, 1978. View at: Google Scholar
  59. S. Diaz, S. Lavorel, F. de Bello, F. Quetier, K. Grigulis, and T. M. Robson, “Incorporating plant functional diversity effects in ecosystem service assessments,” Proceedings of the National Academy of Sciences of the United States of America, vol. 104, no. 52, pp. 20684–20689, 2007. View at: Publisher Site | Google Scholar
  60. O. L. Phillips, G. van der Heijden, S. L. Lewis et al., “Drought-mortality relationships for tropical forests,” New Phytologist, vol. 187, no. 3, pp. 631–646, 2010. View at: Publisher Site | Google Scholar
  61. D. Nepstad, P. Lefebvre, U. L. Da Silva et al., “Amazon drought and its implications for forest flammability and tree growth: a basin-wide analysis,” Global Change Biology, vol. 10, no. 5, pp. 704–717, 2004. View at: Publisher Site | Google Scholar
  62. G. Wang, C. Alo, R. Mei, and S. Sun, “Droughts, hydraulic redistribution, and their impact on vegetation composition in the Amazon forest,” Plant Ecology, vol. 212, no. 4, pp. 663–673, 2011. View at: Publisher Site | Google Scholar
  63. J. W. Westbrook, K. Kitajima, J. G. Burleigh, W. J. Kress, D. L. Erickson, and S. J. Wright, “What makes a leaf tough? Patterns of correlated evolutionbetween leaf toughness traits and demographic rates among 197 shade-tolerant woody species in a neotropical forest,” American Naturalist, vol. 177, no. 6, pp. 800–811, 2011. View at: Publisher Site | Google Scholar
  64. K. Kitajima and L. Poorter, “Tissue-level leaf toughness, but not lamina thickness, predicts sapling leaf lifespan and shade tolerance of tropical tree species,” New Phytologist, vol. 186, no. 3, pp. 708–721, 2010. View at: Publisher Site | Google Scholar
  65. U. Seibt, A. Rajabi, H. Griffiths, and J. A. Berry, “Carbon isotopes and water use efficiency: sense and sensitivity,” Oecologia, vol. 155, no. 3, pp. 441–454, 2008. View at: Publisher Site | Google Scholar
  66. L. Sack, C. Scoffoni, A. D. McKown et al., “Developmentally based scaling of leaf venation architecture explains global ecological patterns,” Nature Communications, vol. 3, p. 837, 2012. View at: Publisher Site | Google Scholar
  67. J. L. Baltzer, D. M. Gregoire, S. Bunyavejchewin, N. S. M. Noor, and S. J. Davies, “Coordination of foliar and wood anatomical traits contributes to tropical tree distributions and productivity along the malay-thai peninsula,” American Journal of Botany, vol. 96, no. 12, pp. 2214–2223, 2009. View at: Publisher Site | Google Scholar
  68. A. Pennec, V. R. Gond, and D. Sabatier, “Tropical forest phenology in French Guiana from MODIS time series,” Remote Sensing Letters, vol. 2, no. 4, pp. 337–345, 2011. View at: Publisher Site | Google Scholar
  69. IPCC, Climate Change 2007, the Fourth Assessment Report (AR4), 2007.
  70. D. Nepstad, P. Moutinho, M. Dias et al., “The effects of partial throughfall exclusion on canopy processes, aboveground production, and biogeochemistry of an Amazon forest,” Journal of Geophysical Research-Atmospheres, vol. 107, no. 20, 2002. View at: Publisher Site | Google Scholar

Copyright © 2014 Fabien Wagner et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Related articles

No related content is available yet for this article.
 PDF Download Citation Citation
 Download other formatsMore
 Order printed copiesOrder

Related articles

No related content is available yet for this article.

Article of the Year Award: Outstanding research contributions of 2021, as selected by our Chief Editors. Read the winning articles.