Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Temperature rising would slow down tropical forest dynamic in the Guiana Shield

Abstract

Increasing evidence shows that the functioning of the tropical forest biome is intimately related to the climate variability with some variables such as annual precipitation, temperature or seasonal water stress identified as key drivers of ecosystem dynamics. How tropical tree communities will respond to the future climate change is hard to predict primarily because several demographic processes act together to shape the forest ecosystem general behavior. To overcome this limitation, we used a joint individual-based model to simulate, over the next century, a tropical forest community experiencing the climate change expected in the Guiana Shield. The model is climate dependent: temperature, precipitation and water stress are used as predictors of the joint growth and mortality rates. We ran simulations for the next century using predictions of the IPCC 5AR, building three different climate scenarios (optimistic RCP2.6, intermediate, pessimistic RCP8.5) and a control (current climate). The basal area, above-ground fresh biomass, quadratic diameter, tree growth and mortality rates were then computed as summary statistics to characterize the resulting forest ecosystem. Whatever the scenario, all ecosystem process and structure variables exhibited decreasing values as compared to the control. A sensitivity analysis identified the temperature as the strongest climate driver of this behavior, highlighting a possible temperature-driven drop of 40% in average forest growth. This conclusion is alarming, as temperature rises have been consensually predicted by all climate scenarios of the IPCC 5AR. Our study highlights the potential slow-down danger that tropical forests will face in the Guiana Shield during the next century.

Introduction

The tropical forests cover accounts for 25% of the terrestrial carbon pool, and therefore plays an essential role on carbon cycle and storage1,2. Higher atmospheric CO2 concentration might increase carbon uptake, maintaining the carbon sink historical role of tropical forests3. But recent droughts linked to El Nino phenomenon have weakened this carbon sink4,5,6,7, highlighting the dependence of tropical forest dynamics on the global Earth climate. On the other hand, tropical forest dynamic, through tree growth and mortality, itself impacts carbon storage and cycle, and provides important feedbacks on climate change. In this context, more and more efforts are being made to describe the long-term impact interplays between climate change and tropical forest functioning8,9,10,11,12,13. Recently, the impacts of exceptional droughts have been coaching more attention, first because droughts are predicted to be more frequent and severe in the tropics14, and second because tropical forests have already suffered from past severe droughts15,16,17. Massive tree mortality have been observed after droughts18,19, potentially caused by hydraulic failure and/or carbon starvation20, and affecting more severely large trees19,21. Beyond exceptional droughts and other long-term changes in water availability, temperatures are also expected to rise and the dry season length to increase over the next century in Amazonia14,22. These changes will likely impact tree dynamics23,24, and dynamic global vegetation models (DVGMs) sometimes predict a shift toward drier forests or even savannas25.

Coarse scale DGVMs allow efficient large-scale carbon cycle prediction with little input data, relying on a wide set of mechanistic assumptions26. These models were initially developed to simulate ecosystem carbon fluxes, they develop fast and are now used among other things to model nitrogen cycle27 or land management28, but also plant range shift29 or forest mortality30. However, DVGMs failed to predict observed regional patterns of tropical forest dynamics31 for two reasons. First, although DGVMs may model different major species or plant functional traits, they do not account for the huge tree diversity found in tropical forests so that they neglect the diverse functional strategies and the equally-diverse demographic strategies that shape tropical forest response to climate-induced disturbances32,33,34,35. Second, they are not demographic-explicit30. And we do know that it is essential to disentangle the ecosystem trajectory in a comprehensive process-based approach, i.e. by segregating the climate control on each demographic processes (growth, recruitment, mortality) as opposed to an all-in-one model in which only the ecosystem response is modeled, to reveal mechanisms underlying tropical forest response to disturbance and to make more robust predictions of the future trajectories32,36,37. To overcome these limitations, individual-based vegetation models provide a good framework to explore how climate and individual tree demographic strategy may interact and impact community tree dynamics. Managing diversity in these models can be done with functional traits that provide good proxies of the demographical strategies38,39,40,41 and at the same times reflects physiological differences in response to climate variations42,43,44,45.

In this paper, we investigate the potential impacts of climate change on long-term forest dynamics using an individual-based model calibrated with data from the Paracou long-term disturbance experiment, in the Guiana Shield. We simulated a tropical forest community under projected future climate scenarios. These simulations allow us to identify (1) the climate variables that will likely be responsible for most of the changes in forest dynamics, (2) the sensitive ecosystem processes and attributes that will be impacted, and (3) the way the forest structure will consequently change.

Methods

The SELVA individual-based model

The simulator SELVA is a an individual-based forest simulator set-up on the CAPSIS 4.0 Java platform46,47. In the simulator, individual growth, mortality and recruitment are described by sub-models on a two-year time step. Each tree i is described with the diameter at breast height (DBHi), the species (si), a set of functional traits associated with each species (Table 1), and an individual vigor estimate. The simulator implements an already-parameterized joint growth-mortality model described earlier48,49,50, and a neutral recruitment model, based on the neutral assumption that each dead tree is replaced by a new recruited tree, respecting the proportion of each species in the metacommunity. The growth-mortality model used the individual tree parameters and climatic variables (Table 2) to compute individual growth and mortality probability at each time step. Details can be found in the Supplementary Information. The calibration of a precise recruitment model would necessitate more information about the small trees (diameter < 10 cm) and seedlings51, such information is often lacking in tropical forests. In the study site of Paracou, where no information is recorded for trees with DBH < 10 cm, a good modelling framework of recruitment is lacking. Therefore, we made the simplistic assumption of a neutral recruitment.

Table 1 The five functional traits used as proxies of ecological strategies in order to simulate a hyperdiverse tropical forest of the Guiana Shield under future climate scenarios.
Table 2 The climate variables included in the growth-mortality model.

Accounting for the individual vigor

The tree vigor was defined at the individual tree level and reflects the individual tree growth effect on the mortality model parameters, acknowledging that trees of a given species growing less than expected (as compared to individuals of the same species) have a far higher probability of dying and vice-versa49, the so-called dominance of the suppressed52. In our simulations, we used the individual vigor in two ways, reflecting two ways of seeing this intraspecific diversity in tree performance. First, we assume that the individual tree vigor is an endogenous property of a given tree and thus we sampled tree vigor once before starting the simulations. In this way, the individual tree vigor value will not be impacted by the climate-induced growth changes (model 1). Second, we assume that tree vigor is also under environmental control so that climate changes, by modifying the average growth of a given species, will also impact the individual vigor. In this way, we recalculated the individual tree vigor at each time step as the difference between the individual growth and the average species growth (model 2) and modified the mortality probability accordingly. Two versions of the model corresponding to these two hypotheses were used in this study. See Supplementary Information for mathematical details.

Model inputs

To initialize the tree population, we used the tree inventories of the experimental site of Paracou, French Guiana, collected in 2001. The experimental site of Paracou (5°18′N,52°55′W) is a lowland tropical forest near Sinnamary, French Guiana. The forest is a typical Guiana shield forest, with dominant tree families including Fabaceae, Chrysobalanaceae, Lecythidaceae, and Sapotaceae. There are more than 700 woody species attaining 10 cm diameter at breast height (DBH) at the site. Six undisturbed plots of 6.25 hectares each totalizing 22,401 individual trees were used to constitute the initial population in the forest simulator. The functional traits used in this study are extracted from a large database collected in the Guiana Shield and described earlier53,54.

Three climate variables are needed to run the model48 (Table 2): a water stress estimator (Aunder), the total precipitation over two years (Pre) and the mean temperature (tmp). The water stress estimate Aunder was based on a water balance model developed at our study site and taking the daily precipitation from the CRU as input data55,56. Four climate scenarios were investigated based on the scenarios of the IPCC report14. The first scenario (A) is equivalent to the RCP2.6, the second (B) is an intermediary scenario, and the third (C) is equivalent to the RCP8.5. The last scenario (BASE) is a control scenario that uses the current values of the climatic variables and assumes that they will remain stable over time.

At each time step, climate variables were sampled in a normal distribution where the mean changed over time, while the standard deviation remained the same, equal to the historical standard deviation. Historical values were computed between 1991 and 2011 using climatic data from the Climatic Research Unit (CRU) at the University of East Anglia57. The predicted mean temperatures (Temp) and rain (Pre) for the next century were computed using the prediction of the IPCC report14. The water stress estimator Aunder was computed using an estimated change of the dry season length of plus two weeks over a century for the RCP8.522 (Table 2). Details about the climatic scenarios can be found in the Supplementary Information.

Model outputs

At each time step, we computed the community growth and mortality rates to track forest dynamics in time. To characterize the community structure at the end of the simulations, we computed the basal area per hectare (BA), the quadratic diameter (QD) and the above-ground fresh biomass (AGFB) with a local equation58.

Sensitivity analysis

Different climate variables are used as drivers of the forest dynamics in our model, and these variables are predicted to evolve more or less drastically in the future. To disentangle which variables might be responsible for the forest dynamics evolution, we performed a variance based sensitivity analysis. This analysis consists in repetitions of simulation with varying intputs (climate variables) and study of the varying outputs (growth and mortality rates, BA, QD, and AGFB) with a sensitivity index computed with the variances of the outputs. The sensitivity analysis on the climate variables was conducted using a complete factorial design of 27 scenarios (3 scenarios, 3 climate variables). We ran the 27 scenarios 50 times and computed the first-order sensitivity index of Sobol (Si) for each climate variable i59:

$${S}_{i}=\frac{V[E({Y}_{j}|{X}_{i})]}{V({Y}_{j})},$$

where Xi is an input variable from the vector \(X=({A}_{under},{Pre},TMP)\), and Yj is an output variable from the vector \(Y=(BA,morta,growth,AGFB,QD)\), \(V[E({Y}_{j}|{X}_{i})]\) is the variance of the expected value (E) of the output variable Yj knowing the input variable Xi, and V(Yj) is the variance of the output variable Yj. The higher the sobol index, the higher the input variable impact on the output variable.

Results

Forest structure and dynamics

Average growth and mortality rates consistently decreased as the scenario became pessimistic (most pessimistic scenario is C), with the community mortality rate falling from 2 to 1.4% per 2 years and a community growth rate going from 0.25 to 0.16 mm per 2 years for the scenario C (Table 3). The forest structure variables BA, QD, and AGBF also decreased between scenario BASE and scenario C, but these reductions are less substantial than for the forest dynamic variables: BA mean is 30.7 in the scenario BASE and 30.1 in the scenario C, QD mean is 25.6 in the scenario BASE and 25.3 in the scenario C, and AGBF mean is 456 in the scenario BASE and 444 in the scenario C (Table 3).

Table 3 Summary statistics of the simulated model (versions 1 and 2), names used in the paper, definition, units and values.

On the role of individual tree vigor

The two versions of the model correspond to two different individual tree vigor estimators (fixed at the beginning or updated during simulations). The reduction in growth is almost the same for models 1 and 2, and is quite progressive between 2001 and 2100 (Fig. 1). The reduction in mortality is much clearer for model 1 than for model 2, with a minimum for scenario C observed at 1.4% per 2 years for model 1 and a minimum of 1.7% per 2 years for model 2 (Fig. 1).

Figure 1
figure1

Evolution of the community-averaged growth and mortality rates for four climate scenarios and the two forest dynamic models. Growth rates (a and c) and mortality rates (b and d) for model 1 on the left (a and b) and model 2 on the right (c and d). Colored areas represent the 95% confidence interval. In model 1, we assumed that the vigor estimator is not impacted by climatic variables that impact the growth, whereas in model 2, we assumed that climatic variables that impact the community growth also impact the vigor and, consequently, the mortality. Scenario A is equivalent to the RCP2.6, B is an intermediary scenario, and C is equivalent to the RCP8.5. BASE is a control scenario that uses the current values of the climatic variables and assumes that they will remain stable over time (Table 2).

Sensitivity analysis

Sensitivity analyses of models version 1 and 2 are very similar (Fig. 2). Growth was primarily controlled by changes in temperature, whereas mortality patterns were driven by precipitation. All the forest structure variables BA, QD and AGFB were mostly impacted by temperature (on average 67% of variance) and less by precipitation (between 29 and 31% of variance). Almost no effect of the drought estimator Aunder was observed (0.7% of variance).

Figure 2
figure2

Results of the sensitivity analysis. Mean of the 50 Sobol indexes computed for each input and output variable. Inputs: QD: quadratic diameter, AGBF: above ground fresh biomass, growth: average growth rate, morta: mortality rate, BA: basal area. Outputs: Aunder: Area over REW and <0.4, Pre: precipitation, TMP: mean temperature, and interactions. Results of model 1 are on the left and model 2 on the right. Almost all outputs are primarily impacted by the temperature changes. Only mortality is strongly impacted by the precipitation changes.

Discussion

We used an individual-based forest model, where species diversity is approximated by functional traits and demographic processes are explicit, to simulate the future dynamics of the Paracou forest for the next century using predictions of the IPCC 5AR for three different climate scenarios (optimistic RCP2.6, intermediate, pessimistic RCP8.5) and a control (current climate). Whatever the scenario, all ecosystem processes and structure variables exhibited decreasing values as compared to the control, suggesting a general slow-down of the forest under climate change. A sensitivity analysis identified the temperature as the stronger climate driver of this behavior, highlighting a temperature-driven drop of 40% in average forest growth for the most pessimistic scenario (from 0.25 to 0.16 mm per .2 years−1, Table 3).

Modeling limitations

As any forest simulators, the SELVA individual-based model is based on simplified assumptions. In our simulated communities, we took into account two major ecological processes, i.e. competition and response to stress, using the individual vigor. Indeed, the individual vigor can be seen as competitive vigor, the quality of how a tree is able to compete for resources, or it may also be used as capability to react to environmental stresses49. In model 2, the individual vigor is under environmental control so that climate changes, by modifying the average growth of a given species, will also impact the individual vigor and, then, forest dynamics. A major shortcoming of our approach is that, apart from the investigated climate drivers, other potentially important environmental variables were not explicitly modeled. Among others, the nutrient availability has often been highlighted as a major driver of forest dynamic in tropical forests60. In the Guiana Shield however, recent studies have concluded to a low control of soil nutrient availability on forest dynamics and suggested that nutrient-recycling mechanisms other than the direct absorption from soil (e.g. the nutrient uptake from litter, the resorption, or the storage of nutrients in the biomass), may be more important for forest functioning61. Hence we do recognize that SELVA present some limitations to study the future forest functioning but, because our modeling framework succeeded in reproducing the current forest structure and dynamics from real data (see Supplementary Information), we are quite confident in the model ability to explore their future evolution.

On the importance of tree vigor

The two investigated models differed in the ways the tree individual vigor was implemented. In model 1, the reduction of growth due to higher temperature in time did not influence mortality rates so that the decreasing mortality rates was only due to rain diminution. In model 2, the reduction of growth due to higher temperature induced a reduction of the tree vigor which increased mortality rates. This compensates the effect of rain diminution itself and, all in all, leads to a less marked decrease in mortality rates than in model 1. This result highlights the key role of the individual tree vigor49, a component still insufficiently taken into consideration in forest models52. Model 1 looks better adapted to simulate the actual dynamics observed in our study site in French Guiana, as no evident correlation has been empirically found between temperature and mortality rates in our studied forests48. This means that the rise in temperature would solely impacts the growth. However, strong links between growth slow-down and mortality risks are already well documented62, and past growth, a surrogate of our tree vigor, is sometimes used as a predictor of mortality in forest models63. During an experimental throughfall exclusion in Brazil, a decrease in growth was observed64, and followed a few years later by an increase in mortality rates65. These experimental results are more consistent with model 2, i.e. where a decrease in tree vigor translates, at next time step, into an increase in mortality risk. This makes the choice between model 1 and 2 difficult and we have to admit that we almost ignore how this tree vigor component will behave in the next century under the climate pressures that will be different from those currently observed. The reality will probably fall between models 1 and 2, and therefore these two models are useful to explore the possible futures and to measure the impacts of the different hypotheses we put forward to construct our simulations.

Temperature is the main driver of future forest dynamic

Temperature rise is by far the strongest driver of almost all summary statistics while precipitation variability primarily influences mortality rates only. First and foremost, our results must be considered with caution because the simulated ranges of climatic variables solely depends on the IPCC 5AR predictions. According to the latter, the relative changes in temperature values will be higher than the relative changes for precipitation and water stress, and this clearly underlies our results (Fig. 2). Nevertheless, our results highlight the important role of future temperature rises in tropical forest dynamic and structure, confirming previous studies66,67. In our simulations, growth is the most impacted demographic process, and this slowing-down dynamics implies, all else being equal, a substantial reduction in above-ground biomass, quadratic diameter and basal area. If, as highlighted by the results from model 2, the temperature-driven growth reduction leads to higher mortality rates, the forest community structure will significantly change with few large old canopy trees and more small slow-growing trees, with possible consequences for e.g. ecosystem water uptake from deep soil layers during dry season68. This community change will impact the basal area (from 30.1 m2.ha−1 for scenario C with model 1 to 26 m2.ha−1 for scenario C with model 2) and the above-ground fresh biomass (from 444 t.ha−1 for scenario C with model 1 to 369 t.ha−1 for scenario C with model 2). In order to be concrete, temperature is expected to rise of 4.5 °C during the next century in the Guiana Shield. Such temperature can drastically affect photosynthesis by causing irreversible damage to the functioning of leaves4 and we have to admit we are in uncharted ground because, currently, no forests in the world exist in areas with mean temperatures of 31 °C. Nevertheless, we do know, from a leaf physiologist perspective, that as the temperature rises, the velocity of reacting molecules increases, leading to more rapid reaction rates but also to damage of the tertiary structures of the enzymes69. These two processes lead to the well-known bell-shaped curve of growth response to temperature70. Temperature also affect photosynthesis in a more indirect manner, through leaf temperatures defining the magnitude of the leaf-to-air vapor pressure difference, a key factor influencing stomatal conductances69. In the tropical environment of the Guiana Shield, as temperatures are already very high, rising temperatures will imply lower growth.

Uncertain impacts of precipitation changes

The predicted reduction of precipitation spearheads a noticeable reduction in mortality rates. This counter-intuitive results is however supported by a growing common understanding that strong winds and heavy rainfalls associated with severe convective storms are the dominant natural drivers of tree mortality in the Amazon71,72. This precipitation-driven mortality is obvious at Paracou where the proportion of fallen trees, relatively to standing death, is higher during the most rainy years48, trees being more vulnerable to uprooting when soil is water-saturated73. Consequently, the predicted decrease of precipitation implies a decrease in mortality rates in the simulated forest communities. But the IPCC AR5 also forecasts an intensification of abundant rain events in the tropics14, that may play the inverse role, increasing mortality rates. The problem is that such punctual and rare events are currently not well quantified, and relations between mortality and extreme events are statistically complex to model20,74. This makes mortality a crucial demographic process upon which we need to focus our research effort.

Conclusion

Our study highlights the potential slow-down danger that tropical forests will face in the Guiana Shield during the next century and this conclusion is alarming, as temperature rises have been consensually predicted by all climate scenarios of the IPCC 5AR.

References

  1. 1.

    Bonan, G. B. Forests and climate change: forcings, feedbacks, and the climate benefits of forests. Science 320, 1444–1449, http://www.ncbi.nlm.nih.gov/pubmed/18556546 (2008).

    ADS  CAS  Article  Google Scholar 

  2. 2.

    van der Sleen, P. et al. No growth stimulation of tropical trees by 150 years of CO2 fertilization but water-use efficiency increased. Nature Geoscience 8, 24–28 (2015).

    ADS  Article  Google Scholar 

  3. 3.

    Lapola, D. M., Oyama, M. D. & Nobre, C. A. Exploring the range of climate biome projections for tropical South America: The role of CO2 fertilization and seasonality. Global Biogeochemical Cycles 23, 1–16 (2009).

    Article  Google Scholar 

  4. 4.

    Doughty, C. E. et al. Drought impact on forest carbon dynamics and fluxes in Amazonia. Nature 519, 78–82, https://doi.org/10.1038/nature14213 (2015).

    ADS  CAS  Article  PubMed  Google Scholar 

  5. 5.

    Brienen, R. J. W. et al. Long-term decline of the Amazon carbon sink. Nature 519, 344–348, https://doi.org/10.1038/nature14283 (2015).

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  6. 6.

    Gatti, L. V. et al. Drought sensitivity of Amazonian carbon balance revealed by atmospheric measurements. Nature 506, 76–80, https://doi.org/10.1038/nature12957 (2014).

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  7. 7.

    van der Laan-Luijkx, I. R. et al. Response of the Amazon carbon balance to the 2010 drought derived with CarbonTracker South America. Global Biogeochemical Cycles 29, 1092–1108 (2015).

    ADS  Article  Google Scholar 

  8. 8.

    Aguilos, M. et al. Interannual and Seasonal Variations in Ecosystem Transpiration and Water Use Efficiency in a Tropical Rainforest. Forests 10, 14, http://www.mdpi.com/1999-4907/10/1/14 (2018).

    Article  Google Scholar 

  9. 9.

    Aguilos, M., Hérault, B., Burban, B., Wagner, F. & Bonal, D. What drives long-term variations in carbon flux and balance in a tropical rainforest in French Guiana? Agricultural and Forest Meteorology 253254, 114–123, http://linkinghub.elsevier.com/retrieve/pii/S0168192318300595 (2018).

  10. 10.

    Pillet, M. et al. Disentangling competitive vs. climatic drivers of tropical forest mortality. Journal of Ecology 106, 1165–1179, https://doi.org/10.1111/1365-2745.12876 (2018).

    Article  Google Scholar 

  11. 11.

    Wagner, F. H. et al. Climate drivers of the Amazon forest greening. PLoS One 12, e0180932, https://doi.org/10.1371/journal.pone.0180932 (2017).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  12. 12.

    Wagner, F. H. et al. Climate seasonality limits leaf carbon assimilation and wood productivity in tropical forests. Biogeosciences 13, 2537–2562, http://www.biogeosciences.net/13/2537/2016/ (2016).

  13. 13.

    Wagner, F., Rossi, V., Stahl, C., Bonal, D. & Hérault, B. Asynchronism in leaf and wood production in tropical forests: a study combining satellite and ground-based measurements. Biogeosciences 10, 7307–7321, http://www.biogeosciences.net/10/7307/2013/ (2013).

  14. 14.

    Stocker, T. F. et al. IPCC, 2013: Climate Change 2013: The Physical Science Basis. Contributi on of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, http://centaur.reading.ac.uk/1019/ (2013).

  15. 15.

    Bonal, D., Burban, B., Stahl, C., Wagner, F. & Hérault, B. The response of tropical rainforests to drought—lessons from recent research and future prospects. Annals of Forest Science 73, 27–44, https://doi.org/10.1007/s13595-015-0522-5 (2016).

    Article  Google Scholar 

  16. 16.

    Corlett, R. T. The Impacts of Droughts in Tropical Forests. Trends in Plant Science 21, 584–593, https://doi.org/10.1016/j.tplants.2016.02.003 (2016).

    CAS  Article  PubMed  Google Scholar 

  17. 17.

    Hérault, B. & Gourlet-Fleury, S. Will Tropical Rainforests Survive Climate Change? In Torquebiau, E. (ed.) Climate Change and Agriculture Worldwide, chap. 14, 183–196, https://doi.org/10.1007/978-94-017-7462-8_14 (Springer Netherlands, Dordrecht, 2016).

    Google Scholar 

  18. 18.

    Phillips, O. L. et al. Drought sensitivity of the Amazon Rainforest. Science 323, 1344–1347, http://www.sciencemag.org/cgi/content/abstract/323/5919/1344 (2009).

  19. 19.

    Phillips, O. L. et al. Drought-mortality relationships for tropical forests. New Phytologist 187, 631–646 (2010).

    Article  Google Scholar 

  20. 20.

    Hartmann, H., Adams, H. D., Anderegg, W. R. L., Jansen, S. & Zeppel, M. J. B. Research frontiers in drought-induced tree mortality: Crossing scales and disciplines. New Phytologist 205, 965–969 (2015).

    Article  Google Scholar 

  21. 21.

    Bennett, A. C., McDowell, N. G., Allen, C. D. & Anderson-Teixeira, K. J. Larger trees suffer most during drought in forests worldwide. Nature Plants 1, 15139, http://www.nature.com/articles/nplants2015139 (2015).

  22. 22.

    Joetzjer, E., Douville, H., Delire, C. & Ciais, P. Present-day and future Amazonian precipitation in global climate models: CMIP5 versus CMIP3. Climate Dynamics 41, 2921–2936 (2013).

    ADS  Article  Google Scholar 

  23. 23.

    Fargeon, H. et al. Vulnerability of Commercial Tree Species to Water Stress in Logged Forests of the Guiana Shield. Forests 7, 105, http://www.mdpi.com/1999-4907/7/5/105 (2016).

    Article  Google Scholar 

  24. 24.

    Wagner, F. et al. Pan-tropical analysis of climate effects on seasonal tree growth. PLoS One 9, e92337, http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3966775{&}tool=pmcentrez{&}rendertype=abstract (2014).

  25. 25.

    Chave, J. Floristic shifts versus critical transitions in Amazonian forest systems. Forests and Global Change 131–160, http://ebooks.cambridge.org/ref/id/CBO9781107323506A016 (2014).

  26. 26.

    Sitch, S. et al. Evaluation of ecosystem dynamics, plant geography and terrestrial carbon cycling in the LPJ dynamic global vegetation model. Global Change Biology 9, 161–185 (2003).

    ADS  Article  Google Scholar 

  27. 27.

    Smith, B. et al. Implications of incorporating N cycling and N limitations on primary production in an individual-based dynamic vegetation model. Biogeosciences 11, 2027–2054 (2014).

    ADS  Article  Google Scholar 

  28. 28.

    Lindeskog, M. et al. Implications of accounting for land use in simulations of ecosystem carbon cycling in Africa. Earth System Dynamics 4, 385–407 (2013).

    ADS  Article  Google Scholar 

  29. 29.

    Snell, R. S. et al. Using dynamic vegetation models to simulate plant range shifts. Ecography 37, 1184–1197 (2014).

    Article  Google Scholar 

  30. 30.

    Hartmann, H. et al. Research frontiers for improving our understanding of drought-induced tree and forest mortality. New Phytologist 218, 15–28 (2018).

    Article  Google Scholar 

  31. 31.

    Johnson, M. O. et al. Variation in stem mortality rates determines patterns of above-ground biomass in Amazonian forests: implications for dynamic global vegetation models. Global Change Biology 22, 3996–4013 (2016).

    ADS  Article  Google Scholar 

  32. 32.

    Hérault, B. & Piponiot, C. Key drivers of ecosystem recovery after disturbance in a neotropical forest. Forest Ecosystems 5, 2, https://doi.org/10.1186/s40663-017-0126-7 (2018).

    Article  Google Scholar 

  33. 33.

    Herault, B., Ouallet, J., Blanc, L., Wagner, F. & Baraloto, C. Growth responses of neotropical trees to logging gaps. Journal of Applied Ecology 47, 821–831, https://doi.org/10.1111/j.1365-2664.2010.01826.x (2010).

    Article  Google Scholar 

  34. 34.

    Guitet, S. et al. Disturbance Regimes Drive The Diversity of Regional Floristic Pools Across Guianan Rainforest Landscapes. Scientific Reports 8, 3872, http://www.nature.com/articles/s41598-018-22209-9 (2018).

  35. 35.

    Flores, O., Hérault, B., Delcamp, M., Garnier, É. & Gourlet-Fleury, S. Functional Traits Help Predict Post-Disturbance Demography of Tropical Trees. PLoS One 9, e105022, http://www.ncbi.nlm.nih.gov/pubmed/25226586, https://doi.org/10.1371/journal.pone.0105022 (2014).

    ADS  Article  Google Scholar 

  36. 36.

    Piponiot, C. et al. Carbon recovery dynamics following disturbance by selective logging in Amazonian forests. eLife 5, e21394, https://doi.org/10.7554/eLife.21394 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  37. 37.

    Claeys, F. et al. Climate change would lead to a sharp acceleration of Central African forests dynamics by the end of the century. Environmental Research Letters 14, 044002, http://stacks.iop.org/1748-9326/14/i=4/a=044002?key=crossref.3370e853c05e1f685902deadb1a0d7a4 (2019).

    ADS  Article  Google Scholar 

  38. 38.

    Hérault, B. et al. Functional traits shape ontogenetic growth trajectories of rain forest tree species. Journal of Ecology 99, 1431–1440, https://doi.org/10.1111/j.1365-2745.2011.01883.x (2011).

    Article  Google Scholar 

  39. 39.

    Uriarte, M. et al. Impacts of climate variability on tree demography in second growth tropical forests: the importance of regional context for predicting successional trajectories. Biotropica 48, 780–797 (2016).

    Article  Google Scholar 

  40. 40.

    Mirabel, A. et al. A whole-plant functional scheme predicting the early growth of tropical tree species: evidence from 15 tree species in Central Africa. Trees 33, 491–505, https://doi.org/10.1007/s00468-018-1795-8 (2019).

    Article  Google Scholar 

  41. 41.

    Hogan, J. A. et al. Understanding the recruitment response of juvenile Neotropical trees to logging intensity using functional traits. Ecological Applications 28, 1998–2010, https://doi.org/10.1002/eap.1776 (2018).

    MathSciNet  Article  PubMed  Google Scholar 

  42. 42.

    Fyllas, N. et al. Solar radiation and functional traits explain the decline of forest primary productivity along a tropical elevation gradient. Ecology Letters 20, 730–740 (2017).

    Article  Google Scholar 

  43. 43.

    Poorter, L. et al. Are functional traits good predictors of demographic rates? Evidence from five neotropical forests. Ecology 89, 1908–1920 (2008).

    CAS  Article  Google Scholar 

  44. 44.

    Santiago, L. S. et al. Coordination and trade-offs among hydraulic safety, efficiency and drought avoidance traits in Amazonian rainforest canopy tree species. New Phytologist 218, 1015–1024, https://doi.org/10.1111/nph.15058 (2018).

    Article  PubMed  Google Scholar 

  45. 45.

    Bonal, D. et al. Leaf functional response to increasing atmospheric CO2 concentrations over the last century in two northern Amazonian tree species: a historical δ13C and δ18O approach using herbarium samples. Plant, Cell & Environment 34, 1332–1344, https://doi.org/10.1111/j.1365-3040.2011.02333.x (2011).

    Article  Google Scholar 

  46. 46.

    Coligny, F. D. et al. CAPSIS: Computer-Aided Projection for Strategies In Silviculture: Advantages of a shared forest-modelling platform. International Workshop of IUFRO working party 4, 4–7 (2003).

    Google Scholar 

  47. 47.

    Dufour-Kowalski, S., Courbaud, B., Dreyfus, P., Meredieu, C. & De Coligny, F. Capsis: An open software framework and community for forest growth modelling. Annals of Forest Science 69, 221–233 (2012).

    Article  Google Scholar 

  48. 48.

    Aubry-Kientz, M., Rossi, V., Wagner, F. & Hérault, B. Identifying climatic drivers of tropical forest dynamics. Biogeosciences 12, 5583–5596, http://www.biogeosciences.net/12/5583/2015/ (2015).

  49. 49.

    Aubry-Kientz, M., Rossi, V., Boreux, J.-J. & Hérault, B. A joint individual-based model coupling growth and mortality reveals that tree vigor is a key component of tropical forest dynamics. Ecology and Evolution 5, 2457–2465, https://doi.org/10.1002/ece3.1532 (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  50. 50.

    Aubry-Kientz, M., Hérault, B., Ayotte-Trépanier, C., Baraloto, C. & Rossi, V. Toward trait-based mortality models for tropical forests. PLoS One 8, e63678 (2013).

    ADS  CAS  Article  Google Scholar 

  51. 51.

    Ameztegui, A., Coll, L. & Messier, C. Modelling the effect of climate-induced changes in recruitment and juvenile growth on mixed-forest dynamics: The case of montane-subalpine Pyrenean ecotones. Ecological Modelling 313, 84–93, https://doi.org/10.1016/j.ecolmodel.2015.06.029 (2015).

    Article  Google Scholar 

  52. 52.

    Farrior, C. E., Bohlman, S. A., Hubbell, S. P. & Pacala, S. W. Dominance of the suppressed: Power-law size structure in tropical forests. Science 351, 1–14 (2016).

    Article  Google Scholar 

  53. 53.

    Baraloto, C. et al. Functional trait variation and sampling strategies in species-rich plant communities. Functional Ecology 24, 208–216, https://doi.org/10.1111/j.1365-2435.2009.01600.x (2010).

    Article  Google Scholar 

  54. 54.

    Baraloto, C. et al. Decoupled leaf and stem economics in rain forest trees. Ecology Letters 13, 1338–1347, https://doi.org/10.1111/j.1461-0248.2010.01517.x (2010).

    Article  PubMed  Google Scholar 

  55. 55.

    Wagner, F., Hérault, B., Stahl, C., Bonal, D. & Rossi, V. Modeling water availability for trees in tropical forests. Agricultural and Forest Meteorology 151, 1202–1213, https://doi.org/10.1016/j.agrformet.2011.04.012 (2011).

    ADS  Article  Google Scholar 

  56. 56.

    Wagner, F., Rossi, V., Stahl, C., Bonal, D. & Hérault, B. Water Availability Is the Main Climate Driver of Neotropical Tree Growth. PLoS One 7, e34074, https://doi.org/10.1371/journal.pone.0034074 (2012).

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  57. 57.

    Mitchell, T. D. & Jones, P. D. An improved method of constructing a database of monthly climate observations and associated high-resolution grids. International Journal of Climatology 25, 693–712 (2005).

    ADS  Article  Google Scholar 

  58. 58.

    Molto, Q. et al. Predicting tree heights for biomass estimates in tropical forests &ndash; a test from French Guiana. Biogeosciences 11, 3121–3130, http://www.biogeosciences.net/11/3121/2014/ (2014).

  59. 59.

    Sobol, I. M. On sensitivity estimation for nonlinear mathematical models. Matem. Mod. 2, 112–118 (1990).

    MATH  Google Scholar 

  60. 60.

    Turner, B. L., Brenes-Arguedas, T. & Condit, R. Pervasive phosphorus limitation of tree species but not communities in tropical forests. Nature 555, 367 (2018).

    ADS  CAS  Article  Google Scholar 

  61. 61.

    Grau, O. et al. Nutrient-cycling mechanisms other than the direct absorption from soil may control forest structure and dynamics in poor Amazonian soils. Scientific Reports 7, 45017, http://www.nature.com/articles/srep45017 (2017).

  62. 62.

    Chao, K.-J. et al. Growth and wood density predict tree mortality in Amazon forests. Journal of Ecology 96, 281–292 (2008).

    Article  Google Scholar 

  63. 63.

    Rüger, N., Huth, A., Hubbell, S. P. & Condit, R. Determinants of mortality across a tropical lowland rainforest community. Oikos 120, 1047–1056 (2011).

    Article  Google Scholar 

  64. 64.

    Nepstad, D. C. The effects of partial throughfall exclusion on canopy processes, aboveground production, and biogeochemistry of an Amazon forest. Journal of Geophysical Research 107, 1–18, http://www.agu.org/pubs/crossref/2002/2001JD000360.shtml (2002).

  65. 65.

    Nepstad, D. C., Tohver, I. M., Ray, D., Moutinho, P. & Cardinot, G. Mortality of Large Trees and Lianas Following Experimental Drought in an Amazon Forest. Ecology 88, 2259–2269, https://doi.org/10.1890/06-1046.1 (2007).

    Article  PubMed  PubMed Central  Google Scholar 

  66. 66.

    Nemani, R. R. et al. Climate-driven increases in global terrestrial net primary production from 1982 to 1999. Science 300, 1560–1563 (2003).

    ADS  CAS  Article  Google Scholar 

  67. 67.

    Feeley, K. J., Joseph Wright, S., Nur Supardi, M., Kassim, A. R. & Davies, S. J. Decelerating growth in tropical forest trees. Ecology letters 10, 461–469 (2007).

    Article  Google Scholar 

  68. 68.

    Stahl, C. et al. Depth of soil water uptake by tropical rainforest trees during dry periods: does tree dimension matter? Oecologia 173, 1191–1201, https://doi.org/10.1007/s00442-013-2724-6 (2013).

    ADS  Article  PubMed  PubMed Central  Google Scholar 

  69. 69.

    Lloyd, J. & Farquhar, G. D. Effects of rising temperatures and [co2] on the physiology of tropical forest trees. Philosophical Transactions of the Royal Society B: Biological Sciences 363, 1811–1817 (2008).

    CAS  Article  Google Scholar 

  70. 70.

    Fitter, A. H. & Hay, R. K. Environmental physiology of plants (Academic press, 2012).

  71. 71.

    Negrón-Juárez, R. I. et al. Vulnerability of Amazon forests to storm-driven tree mortality. Environmental Research Letters 13, 054021, http://stacks.iop.org/1748-9326/13/i=5/a=054021?key=crossref.3897552e1ee1116652eb036016730341 (2018).

    ADS  Article  Google Scholar 

  72. 72.

    Esprito-Santo, F. D. et al. Storm intensity and old-growth forest disturbances in the Amazon region. Geophysical Research Letters 37, 1–6 (2010).

    Google Scholar 

  73. 73.

    Ferry, B., Morneau, F., Bontemps, J. D., Blanc, L. & Freycon, V. Higher treefall rates on slopes and waterlogged soils result in lower stand biomass and productivity in a tropical rain forest. Journal of Ecology 98, 106–116 (2010).

    Article  Google Scholar 

  74. 74.

    Niu, S. et al. Plant growth and mortality under climatic extremes: An overview. Environmental and Experimental Botany 98, 13–19 (2014).

    Article  Google Scholar 

Download references

Acknowledgements

Funding came from the the GFclim project (European Structural Funds, FEDER 2014–2020, Project GY0006894). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. This work also benefited from an ‘Investissement d’Avenir’ grant managed by the Agence Nationale de la Recherche (CEBA, ref ANR-10-LABX-0025) and from a grant from the Centre de coopération Internationale en Recherche Agronomique pour le Développement.

Author information

Affiliations

Authors

Contributions

M.A.-K. and B.H. conceived and designed the experiments; M.A.-K. and G.C. performed the simulations; V.R. and F.W. contributed analysis and discussion; M.A.-K. and B.H. wrote the manuscript.

Corresponding author

Correspondence to Bruno Hérault.

Ethics declarations

Competing Interests

The authors declare no competing interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Aubry-Kientz, M., Rossi, V., Cornu, G. et al. Temperature rising would slow down tropical forest dynamic in the Guiana Shield. Sci Rep 9, 10235 (2019). https://doi.org/10.1038/s41598-019-46597-8

Download citation

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing