Scinovex

Research article · 2026 · Full text

International Journal of Biology Sciences

Tree community structure and carbon sequestration capacity in tropical wildlife reserves of South America

Lucas Ferreira da Silva

Volume 8(6) · Pages 254–259 · DOI: 10.33545/26649926.2026.v8.i6c.967 ↗

Research areasForest ecology and managementRemote Sensing and LiDAR ApplicationsForest Management and Policy

↓ Read full-text PDF0 citations

Abstract

Author-written

Tropical wildlife reserves in South America contain some of the most species-rich tree communities on Earth, yet their carbon sequestration capacity relative to community structural attributes remains inadequately quantified. This research characterized tree community structure and estimated carbon sequestration capacity in a tropical wildlife reserve in São Paulo State, Brazil. Twenty sampling plots of 20 m × 25 m were established across five habitat types between February and October 2023. All trees with diameter at breast height of 10 cm or greater were measured and identified. Above-ground biomass was computed using pantropical allometric equations incorporating diameter, height, and wood density. A total of 1,456 trees belonging to 62 species across 28 families were enumerated. The Fabaceae family exhibited the greatest species richness with nine species, followed by Bignoniaceae with six species. Shannon-Wiener diversity indices ranged from 2.45 in open woodland to 3.68 in dense forest habitats. Above-ground carbon stocks varied substantially across habitat types, ranging from 45.2 Mg C per hectare in grassland edges to 128.5 Mg C per hectare in dense forest. The overall mean carbon sequestration capacity was estimated at 82.4 Mg C per hectare. Ceiba pentandra exhibited the highest individual carbon contribution due to its exceptional basal area. Correlation analysis revealed strong associations between species diversity, basal area, and carbon storage capacity. Biomass component analysis showed that stem wood constituted the largest carbon pool at approximately 50% of total above-ground carbon. These findings establish that tropical wildlife reserves in South America function as high-capacity carbon sinks whose effectiveness is closely tied to tree community diversity and structural complexity.

Summary

Source-checked

Plain-language reading of the full paper, prepared with AI and checked against the text. Each point links to the authors' own words.

  1. Why it matters

    • Tropical ecosystems are the largest terrestrial carbon reservoirs, holding approximately 470 billion tonnes of carbon in their soils and vegetation. View evidence ↗

      Why it matters

      Tropical ecosystems are the largest terrestrial carbon reservoirs, holding approximately 470 billion tonnes of carbon in their soils and vegetation.

      In the paper's own words

      “Tropical ecosystems represent the single largest terrestrial carbon reservoir, storing an estimated 470 billion tonnes of carbon in vegetation and soils combined”

      Page 1

      Found in: Introduction

      Read it in the full textOpen PDF at page 1 ↗
    • Tropical wildlife reserves are essential for preventing forest loss and providing continuous carbon sequestration services while protecting endangered wildlife. View evidence ↗

      Why it matters

      Tropical wildlife reserves are essential for preventing forest loss and providing continuous carbon sequestration services while protecting endangered wildlife.

      In the paper's own words

      “Wildlife reserves within tropical regions serve as critical bulwarks against forest loss, maintaining intact vegetation communities that provide both habitat for endangered fauna and sustained carbon sequestration services”

      Page 1

      Found in: Introduction

      Read it in the full textOpen PDF at page 1 ↗
    • The ability of a forest stand to accumulate biomass is determined by structural attributes such as stem density, canopy height, basal area distribution, and species richness. View evidence ↗

      Why it matters

      The ability of a forest stand to accumulate biomass is determined by structural attributes such as stem density, canopy height, basal area distribution, and species richness.

      In the paper's own words

      “Community structural attributes including species richness, basal area distribution, canopy height, and stem density collectively determine the biomass accumulation potential of a given forest stand”

      Page 1

      Found in: Introduction

      Read it in the full textOpen PDF at page 1 ↗
  2. The question

    • The study aimed to quantify how carbon storage varies across five habitat types in a Brazilian wildlife reserve and identify which structural and compositional attributes most influence this capacity. View evidence ↗

      The question

      The study aimed to quantify how carbon storage varies across five habitat types in a Brazilian wildlife reserve and identify which structural and compositional attributes most influence this capacity.

      In the paper's own words

      “This research investigated tree community structure and carbon sequestration capacity across five habitat types within a tropical wildlife reserve in São Paulo State, Brazil, aiming to quantify habitat-level variation in carbon storage and identify the structural and compositional attributes that most strongly determine sequestration capacity.”

      Page 2

      Found in: Introduction

      Read it in the full textOpen PDF at page 2 ↗
  3. What they did

    • Researchers used a stratified random sampling design, establishing twenty 20 m × 25 m plots distributed across five different habitat types. View evidence ↗

      What they did

      Researchers used a stratified random sampling design, establishing twenty 20 m × 25 m plots distributed across five different habitat types.

      In the paper's own words

      “A stratified random sampling design was employed with four plots of 20 m × 25 m established within each of the five habitat types, totalling twenty sampling plots.”

      Page 2

      Found in: Methods

      Read it in the full textOpen PDF at page 2 ↗
    • All trees with a diameter at breast height (DBH) of 10 cm or greater were identified to the species level. View evidence ↗

      What they did

      All trees with a diameter at breast height (DBH) of 10 cm or greater were identified to the species level.

      In the paper's own words

      “Within each plot, all standing trees with DBH ≥ 10 cm were identified to species level using botanical keys and consultation with taxonomic specialists.”

      Page 2

      Found in: Methods

      Read it in the full textOpen PDF at page 2 ↗
  4. What they found

    • The survey documented 1,456 trees comprising 62 different species from 28 families. View evidence ↗

      What they found

      The survey documented 1,456 trees comprising 62 different species from 28 families.

      In the paper's own words

      “A total of 1,456 trees belonging to 62 species across 28 families were enumerated.”

      Page 1

      Found in: Methods

      Read it in the full textOpen PDF at page 1 ↗
    • The Fabaceae family was the most species-rich with nine species, followed by Bignoniaceae with six. View evidence ↗

      What they found

      The Fabaceae family was the most species-rich with nine species, followed by Bignoniaceae with six.

      In the paper's own words

      “The Fabaceae family exhibited the greatest species richness with nine species, followed by Bignoniaceae with six species.”

      Page 1

      Found in: Methods

      Read it in the full textOpen PDF at page 1 ↗
    • Tree diversity, measured by the Shannon-Wiener index, was lowest in open woodland (2.45) and highest in dense forest (3.68). View evidence ↗

      What they found

      Tree diversity, measured by the Shannon-Wiener index, was lowest in open woodland (2.45) and highest in dense forest (3.68).

      In the paper's own words

      “Shannon-Wiener diversity indices ranged from 2.45 in open woodland to 3.68 in dense forest habitats.”

      Page 1

      Found in: Discussion

      Read it in the full textOpen PDF at page 1 ↗
    • Above-ground carbon stocks varied by habitat, with the lowest in grassland edges (45.2 Mg C/ha) and the highest in dense forests (128.5 Mg C/ha). View evidence ↗

      What they found

      Above-ground carbon stocks varied by habitat, with the lowest in grassland edges (45.2 Mg C/ha) and the highest in dense forests (128.5 Mg C/ha).

      In the paper's own words

      “Above-ground carbon stocks varied substantially across habitat types, ranging from 45.2 Mg C per hectare in grassland edges to 128.5 Mg C per hectare in dense forest.”

      Page 1

      Found in: Introduction

      Read it in the full textOpen PDF at page 1 ↗
    • The estimated overall mean carbon sequestration capacity for the reserve was 82.4 Mg C per hectare. View evidence ↗

      What they found

      The estimated overall mean carbon sequestration capacity for the reserve was 82.4 Mg C per hectare.

      In the paper's own words

      “The overall mean carbon sequestration capacity was estimated at 82.4 Mg C per hectare.”

      Page 1

      Found in: Introduction

      Read it in the full textOpen PDF at page 1 ↗
    • Stem wood was the largest carbon pool, accounting for approximately 50% of total above-ground biomass (145 Mg/ha). View evidence ↗

      What they found

      Stem wood was the largest carbon pool, accounting for approximately 50% of total above-ground biomass (145 Mg/ha).

      In the paper's own words

      “Stem biomass constituted the dominant carbon pool at 145 Mg per hectare, representing approximately 50% of total above-ground biomass.”

      Page 4

      Found in: Biomass allocation patterns

      Read it in the full textOpen PDF at page 4 ↗
    • Dense forest is the primary carbon storage zone and the largest habitat type, covering 42% of the reserve. View evidence ↗

      What they found

      Dense forest is the primary carbon storage zone and the largest habitat type, covering 42% of the reserve.

      In the paper's own words

      “Dense forest comprised 42% of the reserve area, representing the single largest habitat type and the primary carbon storage zone.”

      Page 4

      Found in: Introduction

      Read it in the full textOpen PDF at page 4 ↗
  5. What it means

    • The authors conclude that basal area is a practical and reliable metric for rapid carbon assessment in forest monitoring programs. View evidence ↗

      What it means

      The authors conclude that basal area is a practical and reliable metric for rapid carbon assessment in forest monitoring programs.

      In the paper's own words

      “The strong correlation between basal area and carbon stock identifies a practical metric for rapid carbon assessment in monitoring programs.”

      Page 5

      Found in: Conclusion

      Read it in the full textOpen PDF at page 5 ↗
    • The study recommends that policy makers and conservationists integrate carbon monitoring into the management frameworks of wildlife reserves. View evidence ↗

      What it means

      The study recommends that policy makers and conservationists integrate carbon monitoring into the management frameworks of wildlife reserves.

      In the paper's own words

      “For conservation practitioners and policy makers, these results recommend integrating carbon monitoring into reserve management frameworks.”

      Page 5

      Found in: Conclusion

      Read it in the full textOpen PDF at page 5 ↗

Visual summary

2 visuals

Illustrations explain the paper; charts are replotted from the paper's own table values.

AI-generated illustration
AI-generated conceptual illustration (Tropical wildlife reserve landscape in South America). It explains the idea and does not show study data.↗ Enlarge
Habitat Carbon Distribution
AI-generated conceptual illustration (Tropical wildlife reserve landscape in South America). It explains the idea and does not show study data.

AI-generated conceptual illustration (Tropical wildlife reserve landscape in South America). It explains the idea and does not show study data.

Open full-size image ↗

Habitat Carbon Distribution

AI-generated conceptual illustration (Tropical wildlife reserve landscape in South America). It explains the idea and does not show study data.

View source evidence ↗
Replotted from source data
Replotted from Table 2 (page 2) of the original paper. Every bar is the paper's own table value.↗ Enlarge
Total C (Mg/ha): Ceiba pentandra vs Swietenia macrophylla vs Hymenaea courbaril vs Cedrela odorata vs Tabebuia rosea
Replotted from Table 2 (page 2) of the original paper. Every bar is the paper's own table value.

Replotted from Table 2 (page 2) of the original paper. Every bar is the paper's own table value.

Open full-size image ↗

Total C (Mg/ha): Ceiba pentandra vs Swietenia macrophylla vs Hymenaea courbaril vs Cedrela odorata vs Tabebuia rosea

Replotted from Table 2 (page 2) of the original paper. Every bar is the paper's own table value.

View source evidence ↗

Figures & tables

From the original paper
Correlation matrix between community structural variables and carbon stockTable 1Page 2Ranking of top ten species by carbon contribution with importance value indicesTable 2Page 2Correlation heatmap illustrating associations between tree species carbon stocksFigure 1Page 3Box plot comparison of individual tree carbon stocks across dominant speciesFigure 2Page 3Biomass component allocation across tree structural compartmentsFigure 3Page 4Proportional land cover distribution within the tropical wildlife reserveFigure 4Page 4
FULL TEXT Original article · reflowed for reading

Introduction

Tropical ecosystems represent the single largest terrestrial carbon reservoir, storing an estimated 470 billion tonnes of carbon in vegetation and soils combined, yet deforestation and degradation continue to release substantial quantities of this stored carbon into the atmosphere annually [1,2] . Wildlife reserves within tropical regions serve as critical bulwarks against forest loss, maintaining intact vegetation communities that provide both habitat for endangered fauna and sustained carbon sequestration services [3] . The relationship between tree community structure and carbon sequestration capacity has emerged as a central research question in conservation ecology. Community structural attributes including species richness, basal area distribution, canopy height, and stem density collectively determine the biomass accumulation potential of a given forest stand [4,5] . In the Neotropical region, exceptionally high tree diversity translates into complex community structures with multiple canopy layers, diverse growth forms, and wide-ranging wood density values, all of which influence carbon storage dynamics [6] . South American tropical wildlife reserves encompass a diversity of habitat types from closed-canopy forest to open savannah, each characterized by distinct tree community compositions and consequently different carbon storage capacities [7] . Previous investigations in Amazonian and Atlantic Forest remnants have documented above-ground carbon stocks ranging from 80 to 250 Mg C per hectare in mature forest stands [2,8] . However, carbon stock assessments within designated wildlife reserves remain comparatively scarce, particularly for the transitional vegetation types that characterize many reserve landscapes.

The allometric estimation of tree biomass in tropical forests has benefited from the development of pantropical equations that incorporate diameter, height, and wood density as predictive variables [9] . These equations have substantially improved the accuracy of biomass estimates compared to earlier diameter-only models, though species-level and sitelevel variation remains a source of uncertainty [10] . Understanding how individual species contribute to total stand-level carbon storage has practical implications for conservation management, as it identifies which species warrant particular protection to maintain ecosystem carbon functions. This research investigated tree community structure and carbon sequestration capacity across five habitat types within a tropical wildlife reserve in São Paulo State, Brazil, aiming to quantify habitat-level variation in carbon storage and identify the structural and compositional attributes that most strongly determine sequestration capacity.

Material and Methods

Material and Methods

Material

Material

Methods

Field data collection was performed between February and October 2023. A stratified random sampling design was employed with four plots of 20 m × 25 m established within each of the five habitat types, totalling twenty sampling plots. Within each plot, all standing trees with DBH ≥ 10 cm were identified to species level using botanical keys and consultation with taxonomic specialists. Voucher specimens were deposited at the Herbarium of the University of São Paulo. For each tree, DBH was recorded using a diameter tape at 1.3 m, total height was measured with a Vertex IV hypsometer, and crown projection area was estimated from two perpendicular crown diameter measurements. Above-ground biomass was calculated using the pantropical allometric equation of Chave et al. (2014): ln(AGB) = -1.803 -0.976E + 0.976 ln(ρ) + 2.673 ln(D) -0.0299(ln(D))², where E is the environmental stress variable, ρ is wood density, and D is DBH. Below-ground biomass was estimated at 24% of AGB following IPCC guidelines for tropical moist forests. Carbon content was assumed to be 47% of dry biomass. Biomass component partitioning was estimated using published allocation fractions for Neotropical tree species. Diversity was quantified using the Shannon-Wiener index (H'), Simpson diversity (1-D), and Pielou evenness (J'). Correlation matrices were generated using Pearson coefficients. Statistical analyses used R version 4.3.2 with the vegan package for community analyses. The correlation matrix revealed strong significant positive associations between all structural variables and carbon stock. Basal area showed the strongest correlation with carbon stock (r = 0.94, p<0.001), followed by species richness (r = 0.82, p<0.001).

Results

Results

Discussion

The documentation of 62 tree species within the wildlife reserve represents a moderately rich flora for Atlantic Forest transitional vegetation in southeastern Brazil. Comparable investigations in Atlantic Forest remnants of similar size have reported species counts ranging from 45 to 120, with variation attributable to altitude, fragment isolation, and successional stage [2,7] . The dominance of Fabaceae is

International Journal of Biology Sciences https://www.biologyjournal.net consistent with broader patterns documented across Neotropical forests, where this family frequently ranks among the most species-rich and ecologically important [6] . [8] , while Nogueira et al. (2020) reported values of 90 to 230 Mg C per hectare in more mature Amazonian formations [1] . The relatively moderate carbon stocks in the present research likely reflect the transitional character of the vegetation and the reserve's establishment only 18 years ago, indicating that forest stands are still in active carbon accumulation phases. The strong correlation between basal area and carbon stock (r = 0.94) aligns with the findings from multiple tropical forest assessments and confirms basal area as the most reliable field-measurable predictor of stand-level carbon storage [4,9] . This relationship has practical significance for rapid carbon estimation in monitoring programs, as basal area can be measured more efficiently than the full suite of variables required by complex allometric equations. The disproportionate carbon contribution of large-diameter species, particularly Ceiba pentandra, demonstrates the importance of protecting individual large trees within conservation landscapes. Similar findings have been reported from Central African forests and Southeast Asian rainforests, suggesting that this pattern is a general feature of tropical tree communities [3,10] . The substantial variation in carbon stocks across habitat types within the reserve highlights the importance of maintaining habitat heterogeneity for maximizing landscape-level carbon storage. While dense forest habitats store the most carbon per unit area, the mosaic of habitat types present in the reserve collectively supports higher total diversity than any single habitat would in isolation, potentially enhancing the stability and resilience of the overall carbon pool.

Biomass allocation patterns

The partitioning of above-ground biomass among tree components revealed that stem wood constituted approximately 50% of total above-ground carbon, a proportion consistent with published values for tropical broadleaf forests globally. The relatively high branch biomass fraction of 21% reflects the open-canopy architecture of many trees in the transitional vegetation zones, where lateral crown expansion is favoured over vertical growth. Understanding these allocation patterns is essential for accurate carbon accounting, as remote sensing methods that estimate total biomass from canopy measurements must account for the variable relationship between crown dimensions and total tree biomass across different vegetation types.

Climate change projections

Under current climate change projections for southeastern Brazil, increased drought frequency and elevated temperatures may alter tree community composition and carbon storage trajectories within the reserve. Species with lower drought tolerance may experience elevated mortality rates, potentially shifting community composition toward more drought-adapted assemblages with different carbon storage characteristics. Long-term monitoring of tree mortality, recruitment, and growth rates is needed to assess whether the carbon accumulation trends documented in this research will persist under changing climatic conditions.

Conclusion

In the broader context of tropical conservation in South America, this research contributes to a growing body of evidence demonstrating that protected wildlife reserves function as important carbon sinks. The documentation of 62 tree species maintaining a mean carbon stock of 82.4 Mg C per hectare within an 18-year-old reserve indicates substantial carbon accumulation potential even in relatively young protected landscapes.

Several key findings merit emphasis. The strong correlation between basal area and carbon stock identifies a practical metric for rapid carbon assessment in monitoring programs.

The disproportionate carbon contribution of Ceiba pentandra and other large-diameter species underscores the conservation priority of protecting mature individuals. The three-fold variation in carbon stocks across habitat types demonstrates the importance of maintaining diverse habitat mosaics within reserves.

For conservation practitioners and policy makers, these results recommend integrating carbon monitoring into reserve management frameworks. Carbon credit mechanisms could provide supplementary funding for reserve operations, while species-specific management targeting high-carbon-value trees could enhance the climate mitigation function of existing protected areas. Looking ahead, research priorities should include establishing permanent monitoring plots for tracking longterm carbon dynamics, investigating the influence of edge effects on carbon stocks in reserve margins, and evaluating the potential of degraded habitats within reserves for carbon enhancement through assisted natural regeneration.

Acknowledgement

AcknowledgementsRAC and MOS contributed to fieldwork and data collection; LFS designed the research framework and led analysis. The Instituto Chico Mendes de Conservação da Biodiversidade provided institutional support and research permits. Financial support was received from the São Paulo Research Foundation.

Figures and tables from the paper

Table 1. Correlation matrix between community structural variables and carbon stock
VariableSpecies RichnessBasal AreaMean DBHMean HeightCarbon Stock
Species Richness1.000.78**0.52*0.61**0.82**
Basal Area0.78**1.000.85**0.72**0.94**
Mean DBH0.52*0.85**1.000.68**0.79**
Mean Height0.61**0.72**0.68**1.000.75**
Carbon Stock0.82**0.94**0.79**0.75**1.00

Table text extracted from the original PDF; superscript letters are shown inline. The original table image is below.

View original table image
Correlation matrix between community structural variables and carbon stock↗ Enlarge
Table 1
Correlation matrix between community structural variables and carbon stock

Correlation matrix between community structural variables and carbon stock

Open full-size image ↗
Original paper tableSource PDF · page 2 ↗
Table 2. Ranking of top ten species by carbon contribution with importance value indices
RankSpeciesIVIMean C (kg/tree)Total C (Mg/ha)
1Ceiba pentandra38.5485.218.5
2Swietenia macrophylla32.1342.814.2
3Hymenaea courbaril28.6298.512.8
4Cedrela odorata25.4265.310.5
5Tabebuia rosea22.8218.69.2
6Cariniana estrellensis19.5195.47.8
7Aspidosperma polyneuron16.2168.26.5
8Piptadenia gonoacantha14.8142.55.8
9Gallesia integrifolia12.5128.64.2
10Ficus guaranitica10.2112.43.8

Table text extracted from the original PDF; superscript letters are shown inline. The original table image is below.

View original table image
Ranking of top ten species by carbon contribution with importance value indices↗ Enlarge
Table 2
Ranking of top ten species by carbon contribution with importance value indices

Ranking of top ten species by carbon contribution with importance value indices

Open full-size image ↗
Original paper tableSource PDF · page 2 ↗
Figure 1. Correlation heatmap illustrating associations between tree species carbon stocks
Correlation heatmap illustrating associations between tree species carbon stocks↗ Enlarge
Figure 1
Correlation heatmap illustrating associations between tree species carbon stocks

Correlation heatmap illustrating associations between tree species carbon stocks

Open full-size image ↗
Original paper figureSource PDF · page 3 ↗
Figure 2. Box plot comparison of individual tree carbon stocks across dominant species
Box plot comparison of individual tree carbon stocks across dominant species↗ Enlarge
Figure 2
Box plot comparison of individual tree carbon stocks across dominant species

Box plot comparison of individual tree carbon stocks across dominant species

Open full-size image ↗
Original paper figureSource PDF · page 3 ↗
Figure 3. Biomass component allocation across tree structural compartments
Biomass component allocation across tree structural compartments↗ Enlarge
Figure 3
Biomass component allocation across tree structural compartments

Biomass component allocation across tree structural compartments

Open full-size image ↗
Original paper figureSource PDF · page 4 ↗
Figure 4. Proportional land cover distribution within the tropical wildlife reserve
Proportional land cover distribution within the tropical wildlife reserve↗ Enlarge
Figure 4
Proportional land cover distribution within the tropical wildlife reserve

Proportional land cover distribution within the tropical wildlife reserve

Open full-size image ↗
Original paper figureSource PDF · page 4 ↗

References

16 in the original paper
  1. Nogueira EM, Yanai AM, de Vasconcelos SS, Graça PM, Fearnside PM. Carbon stock loss from deforestation through 2013 in Brazilian Amazonia. Glob Change Biol. 2020;26(4):2530-2546.
  2. Vieira SA, de Camargo PB, Selhorst D. Forest structure and carbon dynamics in Amazonian tropical rain forests. Oecologia. 2021;140(3):468-479.
  3. Lewis SL, Sonké B, Sunderland T. Above-ground biomass and structure of 260 African tropical forests. Philos Trans R Soc B Biol Sci. 2013;368(1625):20120295.
  4. Poorter L, van der Sande MT, Thompson J. Diversity enhances carbon storage in tropical forests. Glob Ecol Biogeogr. 2015;24(11):1314-1328.
  5. Dongs YD, Saka MG, Adedotun A. Species tree diversity and carbon sequestration potentials in Jos Wildlife Park, Jos, Plateau State, Nigeria. South Asian J Agric Sci. 2024;4(1):28-36.
  6. Chave J, Coomes D, Jansen S. Towards a worldwide wood economics spectrum. Ecol Lett. 2009;12(4):351366.
  7. ter Steege H, Pitman NCA, Sabatier D, et al. Hyperdominance in the Amazonian tree flora. Science. 2013;342(6156):1243092.
  8. Mello RA, Toppa RH, Cardoso-Leite E. Priority areas for forest conservation in an urbanized landscape at the transition between Atlantic Forest and Cerrado. Bol Mus Para Emílio Goeldi Cienc Nat. 2016;11(3):357370.
  9. Chave J, Réjou-Méchain M, Búrquez A, et al. Improved allometric models to estimate the aboveground biomass of tropical trees. Glob Change Biol. 2014;20(10):3177-3190.
  10. Feldpausch TR, Lloyd J, Lewis SL, et al. Tree height integrated into pantropical forest biomass estimates. Biogeosciences. 2012;9(8):3381-3403.
  11. Bustamante MM, Roitman I, Aide TM, et al. Toward an integrated monitoring framework to assess the effects of tropical forest degradation and recovery on carbon stocks. Glob Change Biol. 2016;22(1):92-109.
  12. Sullivan MJ, Talbot J, Lewis SL, et al. Diversity and carbon storage across the tropical forest biome. Sci Rep. 2017;7(1):1-12.
  13. Malhi Y, Roberts JT, Betts RA, Killeen TJ. Climate change, deforestation, and the fate of the Amazon. Science. 2008;319(5860):169-172.
  14. Brienen RJ, Phillips OL, Feldpausch TR, et al. Longterm decline of the Amazon carbon sink. Nature. 2015;519(7543):344-348.
  15. Slik JW, Paoli G, McGuire K, et al. Large trees drive forest aboveground biomass variation in moist lowland forests across the tropics. Glob Ecol Biogeogr. 2013;22(12):1261-1271.
  16. Saatchi SS, Harris NL, Brown S, et al. Benchmark map of forest carbon stocks in tropical regions across three continents. Proc Natl Acad Sci U S A. 2011;108(24):9899-9904.