NGEE–Tropics

Next-Generation Ecosystem Experiments

What Influences Canopy Tree Mortality in Amazon Forests?

Forest scientists evaluated the influence of crown exposure to light, growth rates, tree size and species wood density in the mortality of canopy trees in the Amazon

Canopy trees at the Amacayacu Forest Dynamics plot seen from a drone image. Location of some canopy trees with their trunk size (diameter at breast height-DBH, cm), species abbreviation and survival status are shown for reference. A standing dead tree of Caryocar glabrum (Caryocaraceae, purple dot) stands out from the canopy.

The Science

Tree mortality influences forest carbon dynamics and composition. However, despite the ecosystem importance of tree mortality, we still don’t clearly understand the factors that influence it. Forest ecologists evaluated the role of crown exposure to light, trunk size, growth rates and species wood density in the mortality of 984 canopy trees in the Amacayacu Forest Dynamics plot, northwestern Amazon. Researchers found that the probability of tree death decreased for low wood density species as trees have higher proportion of their crowns exposed to light and grew more than their species average. On the contrary, high wood density species were slightly more prone to die when their crowns were more exposed to light.

The Impact

These results underscore the varied tree responses to specific local conditions. The inclusion of life history strategies (e.g., proxied by wood density), when predicting forest demography under vegetation dynamic models can help to reduce the uncertainty of forest outcomes. Future research should focus on better integration between remote and ground-based data to improve the understanding of tree mortality.

Summary

Understanding the factors that influence tree mortality is a research priority to predict forest responses to global change. In this study, researchers evaluated the role of light exposure, trunk size, growth rates and species wood density in the mortality of 984 canopy trees in the Amacayacu Forest Dynamics Plot, northwestern Amazon. Light availability was measured as the proportion of crown exposed to sunlight; a new metric that relates the vertical exposed crown area over the potential crown area, and that was derived from the integration of drone photogrammetry and ground-based data. The research found that the probability of death at changing light conditions varied with species wood density; trees of low wood density were less prone to die when their crowns were more exposed to sunlight, whereas trees of high wood density species were slightly more prone to die at the same light conditions. This result highlights the role of life history strategies, proxied by the species wood density, in shaping tree survival at contrasting light conditions; and demonstrates the importance of the inclusion of plant functional types in vegetation dynamic models aiming to predict forest demography under ongoing global changes.

Principal Investigators

Luisa F. Gómez-Correa, MSc.
Smithsonian Tropical Research Institute, Panamá
luifgomezcor@unal.edu.co

Daniel Zuleta, Ph.D.
Forest Global Earth Observatory
Smithsonian Tropical Research Institute, Washington, DC
dfzuleta@gmail.com

Program Manager

Brian Benscoter
U.S. Department of Energy, Biological and Environmental Research (SC-33)
Environmental System Science
brian.benscoter@science.doe.gov

Funding

This project was supported by MinCiencias “Convocatoria 891 2020” and by the Next Generation Ecosystem Experiments–Tropics, funded by the Office of Biological and Environmental Research (BER) within the U.S. Department of Energy’s (DOE) Office of Science. Ground-based data collection was supported by the Forest Global Earth Observatory (ForestGEO) of the Smithsonian Institution. Drone imagery were collected during the 9th Regional SilvaCarbon/GFOI Workshop on Forest Monitoring.

References

Gómez-Correa, L, F., et al. “Canopy tree mortality depends on the proportion of crown exposed to sunlight, but this effect varies with species wood density.” Biotropica, 00, 1–12. https://doi.org/10.1111/btp.13258

The 2015 El Niño severely disrupted the Amazon forest energy and water cycles

Using data from a 65-m tall flux tower in the Amazon, researchers found that a hot drought in 2015 had lasting impacts on leaf abundance and the water cycle.

Instrumented eddy covariance tower (left) at the Tapajós National Forest, Brazil, and the forest as seen from the tower (right). Data from the tower were used to study how water and energy cycles in the Amazon respond to extremely dry and wet years.

Photo credit by Natalia Restrepo-Coupe.

The Science

Droughts and very rainy years are both becoming more common in the Amazon because of climate change. Researchers used a long time series of data collected from a tower in the middle of the Amazon to study how water exchange between plants and the atmosphere (evapotranspiration) varies during droughts and rainy periods. They found that when rain is abundant, evapotranspiration barely changes. In contrast, during the El Niño drought of 2015, evapotranspiration was low, because plants lost leaves or stopped transpiring. They also found that recovery after drought was slow.

The Impact

The Amazon forest is very important for the earth’s water and energy cycles. The team found that the recent drought in 2015 was hotter and drier than previous droughts because of climate change. This result is important because the impacts of this drought on leaf abundance were stronger, lasted longer, and changed how the forest interacted with the atmosphere. If droughts become too frequent, the forest may not have enough time to recover before the next drought hits the forest. This could permanently shift the Amazon water cycle. The drought accompanied of an increased warming and drying trend resulted in a strong and long lasting effect on the quantity of leaves and how the forest interacts with the atmosphere

Summary

Over the past decades, the Amazon forests experienced multiple extreme droughts and wet years. This research used data collected between 2001 and 2020 at an eddy covariance tower at the Tapajós National Forest (Brazil) to study how evapotranspiration and sensible heat flux varied during periods of drought and excessive rain.  The team also implemented a model that separates the contribution of plants (transpiration) and soil (evaporation) to the water fluxes. The long time series included a strong La Niña wet event (2008-2009) and a strong El Niño drought (2015-2016).

The team found that the La Niña event did not affect evapotranspiration and sensible heat fluxes. The magnitude and the seasonal cycles of all fluxes were similar between the wet and normal years. However, during the El Niño event, many plants closed their stomata or lost leaves. This impact resulted in lower transpiration and higher sensible heat fluxes, indicating water stress. Moreover, the forest did not return to normal conditions until one year after the drought had ended. The slow recovery suggests that the Amazon forest was close to reaching a tipping point during the 2015 drought, which could have resulted in long-lasting changes in the forest.

Contact

Natalia Restrepo-Coupe
Department of Ecology and Evolutionary Biology, The University of Arizona
nataliacoupe@gmail.com

Marcos Longo
Lawrence Berkeley National Laboratory
mlongo@lbl.gov 

Funding

This research was supported by the Next Generation Ecosystem Experiments-Tropics, funded by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research. The measurements at the Tapajós National Forest were supported by multiple projects, including the Brazilian-led Large scale Biosphere Atmosphere experiment in Amazonia, the National Aeronautics and Space Administration, the GoAmazon project (jointly funded by the U. S. Department of Energy, Office of Science, and the Brazilian science foundations FAPESP and FAPEAM), and the U. S. National Science Foundation (NSF).

Publications

Restrepo-Coupe, N., B. O. Christoffersen, M. Longo, L. Alves, K. S. Campos, A. C. de Araújo, R. C. de Oliveira Jr., N. Prohaska, R. da Silva, R. Tapajós, K. T. Wiedemann, S. C. Wofsy, S. R. Saleska. “Asymmetric response of Amazon forest water and energy fluxes to wet and dry hydrological extremes reveals onset of a local drought-induced tipping point.” Global Change Biology (2023). [DOI:10.1111/gcb.16933]

What is the difference in the energy dynamics and forest structure between intact and degraded forests in the Amazon?

Using satellite data, researchers assessed water stress in degraded forests and found that structure strongly mediates evapotranspiration in these forests.

Examples of intact forests (left) and forests degraded by selective logging (middle) and fires (right) in the Amazon Forest. Forest degradation changes forest structure and the way the forest exchanges energy and water with the atmosphere.

Photos by Ekena Rangel Pinagé.

The Science

Forest degradation through fires and logging is common in the Amazon and changes forest structure. However, little is known on the effects of degradation on the way tropical forests transpire water. Using high-resolution remote sensing of forest structure from spaceborne lidar (GEDI) and evapotranspiration (ET) derived from Landsat, researchers assessed the seasonal water stress and its relationship with forest structure across intact and disturbed forests in the Amazon. They found that forest structure exerts a stronger control on ET in the more disturbed/drier forests than in intact or lightly disturbed forests.

The Impact

New satellite data and products allow us to measure forest structure and evapotranspiration across large areas. With these data, we can better understand how human activities in tropical forests are changing the earth’s water and energy cycles. This study found that forest structure influences evapotranspiration more than climate does. In addition, the team found that forest degradation may make the Amazon forests limited by water. This matters because intact forests in the Amazon are normally limited by energy, not water. These findings have important implications for the global water balance and rainfall patterns.

Summary

Deforestation, timber extraction, and forest fires disturb large areas in the Amazon region. These disturbances alter how forests function. Previous work focused on how deforestation affects the water and energy cycles. To understand how degradation changes the water and energy fluxes, this research used many satellite-based data. The research team analyzed evapotranspiration, land surface temperature, and forest structure (tree cover and forest height) data over a region in the Southern Amazon. This region has a mix of deforested, degraded and intact forests, so researchers could study the effects of forest structure on water and energy cycles.

The research team found that water stress conditions started early into the dry in croplands and pastures. They also found that second-growth and burned forests experience stronger water stress than logged and intact forests. Moreover, they found that forest structure is moderately related to evapotranspiration and temperature, but only in the most disturbed forests. This results of this study show the importance of intact forests in maintaining water balance in the Amazon region. Their findings also suggest that disturbed forests may be less able to cope with the changing climate.

Contact

Ekena Rangel Pinagé
College of Forestry, Oregon State University
ekenapinage@gmail.com 

Marcos Longo
Lawrence Berkeley National Laboratory
mlongo@lbl.gov 

Funding

This research was supported by the Next Generation Ecosystem Experiments-Tropics, funded by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research, the Australian Government Research Training Program Scholarship, the USDA Forest Service Pacific Northwest Research Station and International Programs, and the NASA Postdoctoral Program, administered by Universities Space Research Association under contract with NASA. The research carried out at the Jet Propulsion Laboratory, California Institute of Technology, was under a contract with the National Aeronautics and Space Administration.

Publications

Rangel Pinagé, E., D. Bell, M. Longo, C. A. Silva, O. Csillik, A. Huete. “Surface Energy Dynamics and Canopy Structural Properties in Intact and Disturbed Forests in the Southern Amazon.” Journal of Geophysical Research: Biogeosciences, 128(9), e2023JG007465 (2023). [DOI:10.1029/2023JG007465]

 

Large Anthropogenic and Natural Carbon Losses in the Amazon Arc of Deforestation

Airborne laser scanning reveals the impacts of human activities and natural disturbances in a critical region for carbon dynamics in the Earth system.

Image courtesy of KC Cushman (left panel) and Marcos Longo (right panel). Image showing Tropical forests are subject to a range of disturbance types, from small scale mortality from natural processes affecting one or a few trees (left panel) to anthropogenic clearing of large areas (right panel).

Image courtesy of KC Cushman (left) and Marcos Longo (right). Tropical forests are subject to a range of disturbance types, from small scale mortality from natural processes affecting one or a few trees (left) to anthropogenic clearing of large areas (right). 

The Science                                 

The Amazon forest contains globally important carbon stocks, but in recent years, atmospheric measurements suggest that it has been releasing more carbon than it has absorbed because of deforestation and forest degradation. However, attributing the sources of carbon loss to forest degradation and natural disturbances remains a challenge. Using repeated high-resolution airborne laser scanning, we found a greater loss of carbon through forest degradation than through deforestation and a net loss of carbon of 90.5 ± 16.6 Tg C y-1 for the study region attributable to both anthropogenic and natural processes.

The Impact

Accurately attributing the sources of carbon loss to forest degradation and natural disturbances remains a challenge because of the difficulty of classifying disturbances and simultaneously estimating carbon changes This study presents a detailed partitioning of aboveground carbon losses and gains in the Amazon forest, improving our understanding of the relative importance of anthropogenic and natural disturbance types. This study highlights the role of forest degradation in the carbon balance for this critical region in the Earth system. The methodology used here also demonstrates the use of randomized samples with airborne remote sensing for scaling observations from local to regional inference.

Summary

Human activities and recent changes in regional climate have caused significant changes to the structure, integrity, and biodiversity of tropical forests. The Brazilian Amazon has experienced severe deforestation and degradation, leading to the region becoming a carbon source rather than a sink in recent decades. However, the relative importance of deforestation, degradation, and natural disturbances for regional carbon dynamics are not well understood. Using randomized, repeated, very high-resolution airborne laser scanning surveys, this study attributed carbon losses among anthropogenic and natural disturbance types, and extrapolated results to the entire Amazonian Arc of Deforestation.

Extrapolating the lidar-based statistics to the study area (544,300 km2), we found that 24.1, 24.2, and 14.5 Tg C y-1 were lost through clearing, fires, and logging, respectively. The losses due to large windthrows (21.5 Tg C y-1) and other disturbances (50.3 Tg C y-1) were partially counterbalanced by forest growth (44.1 Tg C y-1). These results highlight the importance of forest degradation—in addition to more commonly studied deforestation—for regional carbon dynamics.

 Contact

Lead author:
Ovidiu Csillik
Wake Forest University, Winston-Salem, NC (formerly: Jet Propulsion Laboratory, California Institute of Technology)
ovidiu.csillik@gmail.com
213-465-6732

DOE co-author:
K.C. Cushman
Oak Ridge National Laboratory
cushmankc@ornl.gov
865-924-7364

 Funding

The research of O.C., M.K., A.F., and S.S. carried out at the Jet Propulsion Laboratory, California Institute of Technology, was under a contract with the National Aeronautics and Space Administration (80NM0018D0004). The research of K.C.C. was carried out at Oak Ridge National Laboratory, which is managed by the University of Tennessee-Battelle, LLC, under contract DE-AC05-00OR22725 with the U.S. Department of Energy. C., M.K., M.L., and K.C.C were supported by the Next Generation Ecosystem Experiments‐Tropics, funded by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research (DE-AC02-05CH11231). R.P. was supported by a NASA LCLUC Program grant (20-LCLUC2020-0024). Funding for EBA airborne lidar datasets was provided by the Amazon Fund/BNDES (Grant 14.2.0929.1, Improving Biomass Estimation Methods for the Amazon – EBA); Coordenação de Aperfeiçoamento de Pessoal de Nível Superior Brasil (CAPES; Finance Code 001); Conselho Nacional de Desenvolvimento Científico e Tecnológico (Processes 403297/2016-8 and 301661/2019-7). Support to generate carbon calibrations was provided by the Sustainable Landscapes Brazil project supported by the Brazilian Agricultural Research Corporation (EMBRAPA), the US Forest Service, and USAID, and the US Department of State.

Publication

Csillik et al.(NGEE-Tropics Collaboration), “A large net carbon loss attributed to anthropogenic and natural disturbances in the Amazon Arc of Deforestation” Proceedings of the National Academy of Sciences of the United States of America, 121 (33), e2310157121 (2024). [DOI: 10.1073/pnas.2310157121]

 Related Links

Deforestation harms climate less than other types of Amazon degradation, study finds, Reuters, August 5, 2024

Forest degradation releases 5 times more Amazon carbon than deforestation, Mongabay, August 9, 2024

FATES Vegetation-Fire Model Predicts Tropical Ecosystem Biogeography

Simulated feedbacks between fire behavior and vegetation traits drive emergent patterns.

Image courtesy of Jacquelyn K. Shuman, NASA Ames Research Center. Fire is an essential component of many ecosystems, shaping the distribution of trees and grasses. Managed fire, shown implemented in South Africa’s Kruger National Park, can be used to understand fuel moisture thresholds and fire behavior and effects.

Image courtesy of Jacquelyn K. Shuman, NASA Ames Research Center. Fire is an essential component of many ecosystems, shaping the distribution of trees and grasses. Managed fire, shown implemented in South Africa’s Kruger National Park, can be used to understand fuel moisture thresholds and fire behavior and effects. 

The Science
Frequent fire can prevent trees from growing in drier regions of the tropics. In wet tropical forests, fire is infrequent and low intensity. Researchers adapted a model of fire behavior for use with a model of vegetation dynamics to capture this interaction between fire and trees in predictions of tropical ecosystems across climate gradients. The predictions captured the gradient from high to low biomass along with a transition from wet tropical forest to drier savanna and grassland across South America. This new modeling capability is essential to predicting how Amazon forest could change under future drier conditions.

The Impact
The dynamic interaction between fire and vegetation is not well represented in most global model predictions of the future climate and biosphere. Yet it is expected to be important to both. This work demonstrates an important new modeling capability that expands the potential applications for the Functionally Assembled Terrestrial Ecosystem Simulator (FATES), a vanguard vegetation model supported by the DOE. The model captures the feedback and interactions between fire, changing fuels, surviving tree canopies and local conditions, and can be used to project where tropical trees and grasses are likely to survive under future climate and fuel conditions.

Summary
To accurately project changes in fire and its impacts requires fire-vegetation models that integrate plant properties that affect fire behaviors and the response of vegetation distribution and structure to fire. We adapted the fire behavior and effects module, Spread and Intensity of Fire (SPITFIRE), for use with the Functionally Assembled Terrestrial Ecosystem Simulator (FATES), a size-structured vegetation demographic model, and tested model predictions of tropical forest and grassland biogeography. In the simulations, three types of vegetation competed for resources: a fire-vulnerable tree, a fire-tolerant tree, and a fire-promoting grass. We found the model sensitive to a parameter governing fuel moisture, with drier fuels expanding grass, fire-tolerant trees, and fire-burned area. Fire mortality was elevated for small and fire-vulnerable trees, as in observations. The model captured productivity and spatial biomass patterns in fire-disturbed forests, but was biased in less disturbed areas. Although burned fraction was predicted as greater than observed in grass-dominated areas, biogeography of fire-tolerant and fire-vulnerable trees corresponded to observations across the tropics. The results reflect a positive grass–fire feedback and suggest that infrequent-fire forests may be vulnerable to higher fire intensities. With SPITFIRE, FATES is useful for assessing the vulnerability and resilience of tropical forests to fire.

Contact
Jacquelyn Shuman
NASA Ames Research Center
jacquelyn.k.shuman@nasa.gov, 303-319-1509 

Funding
This research was supported as part of the Next Generation Ecosystem Experiments-Tropics (NGEE-Tropics), funded by the U.S. Department of Energy, Office of Science. 

Publications
J.K. Shuman, et al. “Dynamic ecosystem assembly and escaping the “fire-trap” in the tropics: Insights from FATES_15.0.0.” Geoscientific Model Development 17, 4643–4671 (2024) [DOI: 10.5194/gmd-17-4643-2024]

Hysteresis area at the canopy level during and after a drought event in the Central Amazon

The changes in hysteresis area during the 2015 ENSO drought compared to the post-drought period highlight the risk of hydraulic failure of tropical species under extreme heat stress.

 

Canopy-level transpiration during the 2015 ENSO-driven drought in the Central Amazon exhibited significant deviations due to exceptionally high Vapor Pressure Deficit (VPD) conditions and increased temporal differences between the peaks of stomatal conductance (gs) and VPD. This resulted in an increased hysteresis effect, as evidenced by the expanded hysteresis area (Harea) across multiple ecophysiological variables compared to the post-drought period. Image courtesy of authors.

Canopy-level transpiration during the 2015 ENSO-driven drought in the Central Amazon exhibited significant deviations due to exceptionally high Vapor Pressure Deficit (VPD) conditions and increased temporal differences between the peaks of stomatal conductance (gs) and VPD. This resulted in an increased hysteresis effect, as evidenced by the expanded hysteresis area (Harea) across multiple ecophysiological variables compared to the post-drought period. Image courtesy of authors.

The Science
During the 2015 ENSO-driven drought in the Central Amazon, canopy-level transpiration exhibited significant deviations due to exceptionally high Vapor Pressure Deficit (VPD) conditions and increased temporal differences between the peaks of stomatal conductance (gs) and VPD. This resulted in an increased hysteresis effect, as evidenced by the expanded hysteresis area (Harea) across multiple ecophysiological variables compared to the post-drought period.

These canopy-level observations were developed using two distinct methods due to the difficulty of accessing canopy layers in the Amazon forest. First, species selection was determined by the proximity of the tree crown to the K-34 flux tower. The K-34 tower provided access to the crowns of a few species for leaf-level measurements. Additionally, we conducted leaf gas exchange experiments using, for the first time in the Amazon forest, a 26.0 m telescopic boom lift (Genie® Z-80/60) to access the canopy of multiple species.

The Impact
There are several challenges to performing in situ observations at the canopy level in tropical ecosystems such as the Amazon forest. This difficulty in conducting leaf-level measurements in the tropics is being overcome by the use of towers, climbing techniques, canopy walkways, canopy-access cranes (e.g., Panama and Daintree Forest in Australia), and, for the first time in the Amazon forest, a telescopic boom lift (“cherry picker”) used as reported by this study (Manaus, ZF-2). This significant effort in the Amazon forest has provided new insights into leaf gas exchange processes and the access of multiple species at different canopy levels. From this perspective, understanding the patterns of stomatal conductance and leaf water potential across a range of species, along with canopy temperature and vapor pressure deficit measurements as demonstrated by this study, is crucial for improving the climate models in the 21st century such as FATES, especially during extreme drought events.

Summary
Understanding forest water limitation during droughts within a warming climate is essential for accurate predictions of forest-climate interactions. In hyperdiverse ecosystems like the Amazon forest, the mechanisms shaping hysteresis patterns in transpiration relative to environmental factors are not well understood. From this perspective, we investigated these dynamics by conducting in situ leaf-level measurements throughout and after the 2015 El Niño-Southern Oscillation (ENSO) drought. Our findings indicate a substantial increase in the hysteresis area (Harea) among transpiration (E), vapor pressure deficit (VPD), and stomatal conductance (gs) at canopy level during the ENSO peak, attributed to both temporal lag and differences in magnitude between gs and VPD peaks. Specifically, the canopy species Pouteria anomala exhibited an increased Harea, due to earlier maximum gs rates leading to a greater temporal lag with VPD compared to the post-drought period. Additionally, leaf water potential (ΨL) and canopy temperature (Tcanopy) showed larger Harea during the ENSO peak compared to post-drought conditions across all studied species, suggesting that stomatal closure, particularly during the afternoon, acts to minimize water loss and may explain the counterclockwise hysteresis observed between ΨL and Tcanopy. The pronounced Harea during the drought points to a potential imbalance between water supply and demand, underlining the role of stomatal behavior of isohydric species in response to drought.

Contact
Bruno O. Gimenez
University of California – Berkeley (UCB)
b_gimenez@berkeley.edu

Funding
This work  is based upon work supported as part of the Next Generation Ecosystem Experiments-Tropics (NGEE-Tropics), as part of DOE’s Terrestrial Ecosystem Science Program – Contract No. DE-AC02–05CH11231. Additional funding for this research was provided by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), INCT – Madeiras da Amazônia, Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) – Processo: 403839/2021-1 Chamada CNPq/MCTI/FNDCT Nº 18/2021 – Faixa A – Grupos Emergentes Universal 2021, and Fundação de Amparo à Pesquisa do Estado do Amazonas (FAPEAM).

Publications
Gimenez, B. O., Souza, D. C., Higuchi, N., Negrón-Juárez, R. I., de Jesus Sampaio-Filho, I., Araújo, A. C., … & Chambers, J. Q. (2024). “Hysteresis area at the canopy level during and after a drought event in the Central Amazon” Agricultural and Forest Meteorology, 353, 110052. https://doi.org/10.1016/j.agrformet.2024.110052