Amazon Assessment Report 2021

Chapter 4
Amazonian ecosystems and their ecological
functions

Cipó parasita estrangula árvore (Foto: Leonardo Milano/Amazônia Real)

About the Science Panel for the Amazon (SPA)

The Science Panel for the Amazon is an unprecedented initiative convened under the
auspices of the United Nations Sustainable Development Solutions Network (SDSN). The
SPA is composed of over 200 preeminent scientists and researchers from the eight
Amazonian countries, French Guiana, and global partners. These experts came together
to debate, analyze, and assemble the accumulated knowledge of the scientific
community, Indigenous peoples, and other stakeholders that live and work in the Amazon.

The Panel is inspired by the Leticia Pact for the Amazon. This is a first-of-its-kind Report
which provides a comprehensive, objective, open, transparent, systematic, and rigorous
scientific assessment of the state of the Amazon’s ecosystems, current trends, and their
implications for the long-term well-being of the region, as well as opportunities and policy
relevant options for conservation and sustainable development.

Amazon Assessment Report 2021, Copyright @ 2021, Science Panel for the Amazon.
This report is published under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-
NC-SA 4.0) License. ISBN: 9781734808001

Suggested Citation

Moraes R M, Correa SB, Doria CRC, Duponchelle F, Miranda G, Montoya M, Phillips OL, Salinas N, Silman M, Ulloa Ulloa
C, Zapata-Ríos G, Arieira J, ter Steege H. 2021. Chapter 4: Biodiversity and Ecological Functioning in the Amazon. In:
Nobre C, Encalada A, Anderson E, Roca Alcazar FH, Bustamante M, Mena C, Peña-Claros M, Poveda G, Rodriguez JP,
Saleska S, Trumbore S, Val AL, Villa Nova L, Abramovay R, Alencar A, Rodríguez Alzza C, Armenteras D, Artaxo P,
Athayde S, Barretto Filho HT, Barlow J, Berenguer E, Bortolotto F, Costa FA, Costa MH, Cuvi N, Fearnside PM, Ferreira
J, Flores BM, Frieri S, Gatti LV, Guayasamin JM, Hecht S, Hirota M, Hoorn C, Josse C, Lapola DM, Larrea C, Larrea-
Alcazar DM, Lehm Ardaya Z, Malhi Y, Marengo JA, Melack J, Moraes R M, Moutinho P, Murmis MR, Neves EG, Paez B,
Painter L, Ramos A, Rosero-Peña MC, Schmink M, Sist P, ter Steege H, Val P, van der Voort H, Varese M, Zapata-Ríos G
(Eds). Amazon Assessment Report 2021. United Nations Sustainable Development Solutions Network, New York, USA.
Available from https://www.theamazonwewant.org/spa-reports/. DOI: 10.55161/IKRT9380
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

INDEX

GRAPHICAL ABSTRACT ................................................................................................................................. 2

KEY MESSAGES ............................................................................................................................................... 3
ABSTRACT ....................................................................................................................................................... 4
4.1. AMAZONIAN ECOSYSTEMS: AN INTRODUCTION ............................................................................... 4
4.1.1. VEGETATION TYPES FROM THE HIGH ANDES TO THE ATLANTIC OCEAN ............................................................. 5
4.2. LOWLAND AMAZONIAN ECOSYSTEMS................................................................................................. 7
4.2.1 TERRESTRIAL ECOSYSTEMS ............................................................................................................................... 7
4.2.1.1 Lowland Rainforests ............................................................................................................................................ 8
4.2.1.2 White sand forests .............................................................................................................................................. 10
4.2.1.3. Savannas and grasslands ................................................................................................................................. 10
4.2.2. FRESH WATER BODIES AND WETLANDS .......................................................................................................... 12
4.2.2.1. Rivers, Lakes and Forest streams ..................................................................................................................... 12
4.2.2.2. Freshwater Wetlands ........................................................................................................................................ 16
4.3. ECOSYSTEM FUNCTIONING ................................................................................................................ 18
4.3.1. PRIMARY PRODUCTIVITY, NUTRIENTS, FOREST DYNAMICS AND DECOMPOSITION............................................... 18
4.3.1.1. Terrestrial ecosystems ...................................................................................................................................... 18
4.3.1.2. Freshwater ecosystems...................................................................................................................................... 22
4.3.2. THE FLOOD PULSE AND AQUATIC-TERRESTRIAL TRANSITION ZONE ............................................................... 23
4.4. CONCLUSIONS ....................................................................................................................................... 25
4.5. RECOMMENDATIONS ............................................................................................................................ 26
5. REFERENCES ............................................................................................................................................. 26

1
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Graphical Abstract

Figure 4A Graphical Abstract

2
Science Panel for the Amazon
Biodiversity and Ecological Functioning in the Amazon

Mónica Moraes R.*a, Sandra Bibiana Correab, Carolina Rodrigues da Costa Doriac, Fabrice Duponchelled, Guido Mirandae, Mariana
Montoyaf, Oliver L. Phillipsg, Norma Salinash, Miles Silmani, Carmen Ulloa Ulloaj, Galo Zapata-Ríosk, Julia Arieiral,m, Hans ter Steege*n

Key Messages

● Between the Andean mountains and the Amazon plain, a diverse mosaic of ecosystems and vegetation
is represented by forest, savanna, and swamp biomes. The key to understanding the ecology of the Am-
azon region is to integrate functional processes between terrestrial and aquatic components, across
multiple biophysical gradients, from the continental divide to the ocean.
● Amazonian lowland forests, covering 5.79 M km2, is likely the richest forest area globally, holding an
estimated 16,000 tree species and perhaps over 50,000 plant species, many of which are still unknown.
With close to 400 billion trees, the Amazon is home to 13% of all trees worldwide.
● Species composition is not evenly distributed across the basin but is determined by soil geology and
climate. The most diverse forests are found in the western Amazon; however, protected areas are re-
quired across the basin for comprehensive conservation. Forests in the western Amazon cover rela-
tively fertile soil, are species-rich, have high stem turnover, and have somewhat lower above-ground
biomass. Forests in the central and eastern Amazon, mainly found on poor soils, are less dynamic and
have high biomass.
● The Amazon River Basin holds the largest tropical wetland area on Earth, and a vast number of rivers,
comprising not only the world’s largest store of freshwater, but also 15% of all fish species.
● Forest composition is already being affected by climate change, with the mortality of wet-soil affiliated
genera having increased in places where the dry season has strengthened the most. Given climate
change projections for this century, such changes are likely to intensify.
● Amazonian ecosystems result from a mixture of terrestrial and aquatic landscapes in often extensive
floodplains, whose dynamics are affected by the tectonic uplift of the eastern Andean slopes and the
much less geologically active lowland Amazon River Basin. The contact areas, or ecotones, between ter-
restrial and aquatic ecosystems (fresh and marine waters) are of critical importance for the dynamics
of the whole region. They contribute to the movement of animals, plant propagules, and nutrients be-
tween the floodplain and adjacent terra firme forests, and promote habitat heterogeneity.

a Herbario Nacional de Bolivia, Instituto de Ecología, Universidad Mayor de San Andrés, Av. Andrés Bello y calle 27 - Zona Sur, La
Paz, Bolivia, [email protected]
b Department of Wildlife, Fisheries and Aquaculture, Mississippi State University, 775 Stone Blvd, Mississippi State, MS 39762, USA

c Universidade Federal de Rondônia, Av. Pres. Dutra, 2965 - Olaria, Porto Velho - RO, 76801-058, Brazil

d Institut de Recherche pour le Développement, UMR MARBEC, Marseille, France

e Wildlife Conservation Society, Casilla 3-35181 SM, La Paz, Bolivia

f Wildlife Conservation Society, Av. Roosevelt Nº 6360, Miraflores, Lima, Peru

g School of Geography, University of Leeds, Woodhouse Lane, Leeds, LS2 9JT, United Kingdom

h Pontificia Universidad Católica del Perú, Av. Universitaria 1801, San Miguel 15088, Peru

i Wake Forest University, Department of Biology, 1834 Wake Forest Road, Winston-Salem NC 27109, USA

j Missouri Botanical Garden, 4344 Shaw Blvd, St. Louis MO 63110, USA

k Wildlife Conservation Society, Avenida de los Granados N40-53 y París, Quito, Ecuador

l Universidade Federal do Espírito Santo, Instituto de Estudos Climáticos, Av. Fernando Ferrari, 514, Goiabeiras, Vitória, ES, Brazil.

m Instituto Nacional de Ciência e Tecnologia em Áreas Úmidas (INAU), Universidade Federal de Mato Grosso, R. Quarenta e Nove,

2367, Boa Esperança, Cuiabá, MT, Brazil

n Naturalis Biodiversity Center, Darwinweg 2, 2333 CR Leiden, The Netherlands; Systems Ecology, Vrije Universiteit Amsterdam, De

Boelelaan 1105, 1081 HV Amsterdam, The Netherlands

Science Panel for the Amazon 3

Chapter 4: Biodiversity and Ecological Functioning in the Amazon

● Because of its size and the carbon density of its ecosystems, the Amazon forest is a huge carbon store.
Spatial variation in Amazonian biomass, carbon stocks, and biomass dynamics is driven more by soil
conditions than climate and more by spatial variation in tree mortality than productivity.
● Amazonian wetlands also store large amounts of carbon due to the extensive and deep accumulation of
below-ground peat deposits (e.g., >3 Pg C in north-western Amazonian swamps). Hence, their conser-
vation also plays a crucial role in modulating global warming.

Abstract

Amazonian lowland tropical rainforests cover ~5.79 million km2. Based on geology, the Amazon lowland
forest area can be divided into six regions. The Guiana Shield and Brazilian Shield (in the southern Ama-
zon) are on very old, nutrient-poor soils, while the Western Amazonian regions (northern and southern)
and the regions along the Amazon River are mainly built from more recent sediments of Andean origin
and of variable nutrient richness. The six regions are characterized by differences in soil fertility and rain-
fall, causing differences in above-ground biomass, productivity, and tree turnover. There is still intense
debate concerning the total plant species richness of the the Amazon. A well-supported estimate for trees
(diameter >10 cm) is 16,000 species, ~11,000 of which have been collected and described. Estimates of the
total flora range from 15,000 to 55,000 species. As in much of the tropics, Fabaceae (the bean family) are
the most species-rich of the major woody groups in the Amazon. South America and the Amazon are also
renowned for the abundance and diversity of palms. While most ecosystem vegetation models emphasize
climate and carbon production processes, these are not sufficient to understand how Amazonian forest
ecosystems vary spatially. In particular, long-term observations with plots show that spatial variation in
Amazonian forest biomass and stem dynamics are driven more by soil conditions than climate, while car-
bon stocks are constrained as much by soil physical features and tree floristic composition as by produc-
tivity. The key effects of soil on the Amazon’s ecosystem function also extend to animals and their im-
portant functions, including herbivory, seed dispersal, and insect activity. Soil and geology influence Am-
azonian rivers too, which are distinguished as being either white-water (carrying sediments from the An-
des), clear-water (draining the ancient Shields), or black-water (draining white sand areas). The nutrients
associated with each major river class strongly determine the floodplain forest ecology and species, with
igapó in sediment-poor clear and black-waters, and várzea (known as tahuampa in Peru) with white, sedi-
ment-rich waters. Climate impacts become stronger towards the margins, and some Amazon forests are
already close to the thermal and hydrological limits of sustaining productive forest ecosystems. Amazo-
nian tree mortality rates are already increasing in many intact forests, Amazonian forest composition has
been affected by recent droughts, and the mortality of wet-affiliated Amazonian tree genera has increased
in places where the dry season has intensified. Key areas of uncertainty include understanding the extent
to which recent climate change has caused a slowing of the carbon sink in intact Amazonian forests, and
whether intact forests will now lose carbon, or whether the shallow water tables and rich biodiversity of
many Amazonian forests will buffer against climate change, especially in the western part of the basin.

Keywords: Amazonian ecosystems, aquatic ecosystems, forest dynamics, ecological features, ecosystem processes, inter-
actions, river systems, terrestrial ecosystems.

4.1. Amazonian Ecosystems: An Introduction Brazilian Shields, covers about 8.4 million km2 of
northern South America (see Chapter 2). The
The Amazonian biogeographical region, includ- Amazon River basin (7.3 million km2), including
ing the lowland Amazon and Orinoco River Ba- the Tocantins and Araguaia Basins, covers 41%
sins and adjacent upland areas of the Guiana and of South America, encompassing two of the

4
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

major South America biomes, tropical moist for- the Andes, followed by a more detailed descrip-
ests and tropical savannas (Coe et al. 2008). The tion of the lowland Amazonian terrestrial vegeta-
Amazon region is considered one of the most im- tion types, and conclude with the vast wetlands
portant ecological regions in the world, because included in the area. We continue with an analy-
it includes the largest area of continuous tropical sis of the main ecosystem functions (e.g., terres-
moist forests, estimated to cover 5.79 million trial and aquatic), with an emphasis on produc-
km2 (Ter Steege et al. 2015) and an estimated tivity and carbon sequestration. The aim of this
>10% of all known species of vertebrates and vas- chapter is to reveal the enormous variation of
cular plants on Earth are estimated to live there vegetation types, their diversity and functioning,
(Chapter 2). It also contains by far the largest and how this is affected by soil, climate, and
tropical floodplain system (Keddy et al. 2009), flooding dynamics.
constituted by a rich mosaic of terrestrial,
aquatic, and transitional ecosystems subjected 4.1.1. Vegetation types from the High Andes to
to seasonal or permanent waterlogging (Salo et al. the Atlantic Ocean
1986) (Figure 4. 1).
Alexander von Humboldt’s Tableau Physique
The ability of ecosystems to capture, process, (Humboldt 1805) is, arguably, the first published
and store carbon and other nutrients is deter- overview of plant composition in northern South
mined by key climatic, edaphic, and biological America as a region (Figure 4.2). His travels ex-
factors. The Amazon, with the largest tropical tended from the Pacific to the Atlantic Oceans
rainforest on the planet, encompasses signifi- and passed Chimborazo, the highest equatorial
cant differences in precipitation regimes but volcano in Ecuador (Ulloa Ulloa and Jørgensen
even greater differences in the geological origin, 2018).
age, and nutrient richness of the soils that sup-
port its ecosystems (see Chapter 1). Here we re- Humboldt depicted the biotic and physical char-
view the role of these factors in controlling forest acteristics, and changes in vegetation structure
composition and processes, especially those re- and composition along an elevation gradient,
lated to productivity and forest dynamics. For ex- from the tree-dominated lowlands to the treeless
ample, Amazonian forest biological, structural, páramo bordering the snow line.
and functional diversity is fundamentally af-
fected by water and nutrients. Hydrology defines Plant communities in the high Andes (above
their higher-level classification as terra firme for- 3,000 m) are known as ‘páramo’ in the more hu-
ests, seasonally flooded forests (várzea, igapó), mid areas of the northern Andes of Venezuela,
and swamp forests. Freshwater ecosystems Colombia, and Ecuador, and ‘jalca’ in northern
cover more than 1 million km2, consisting of Peru (Madriñán et al. 2013); ‘puna’ is found in the
three main water types: white, black, and clear southern, drier Altiplano of Peru and Bolivia
waters, which differ in their origin and sediment (Sánchez-Vega and Dillon 2006).
composition. Within the extensive non-flooded
forests, distinctive and extremely poor white Páramos and punas are grass-dominated ecosys-
sand forests may be found, especially in the up- tems with plants uniquely adapted to these ex-
per Rio Negro area and the Guianas (see Adeney treme environments of cold temperatures, low
et al. 2016). pressure, and extreme solar radiation, with
prominent rosette forming plants, such as those
In this chapter we summarize information on in the genera Espeletia and Puya. Only a few spe-
Amazonian ecosystems and their ecological cies of trees, such as those in the genera Buddleja,
functions, with a primary focus on trees. We start Gynoxys, and Polylepis, reach the highest eleva-
with a short description of the vegetation types of tions, up to 4,700 m (Hoch and Körner 2005).

5
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Figure 4.1 Map of Amazonian vegetation and ecosystems (Source: Comer et al. 2020). The solid gray box highlights the high richness
of vegetation and ecosystems found in the latitudinal and altitudinal gradients in the Amazon (see Figure 4.4 for detail).

6
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Upper montane forests traverse humid sites submontane forests gradually change into Ama-
from 2,500 to 3,900 m elevation. Montane forests zonian lowland forests, defined here as those be-
are among the most species rich vegetation types low 500 m, which cover most of the basin. Over-
to be found in the tropical high Andes (Gentry all, the Andes mountains are extraordinarily di-
1988). These forests are 5 to 20 m tall with emer- verse due to their climatic and topographic com-
gent trees reaching 35 m or more, but with plexity, their size, and their position spanning
smaller individuals at the treeline, in places the Equator, northern, and southern tropical
where soils are shallow, or where disturbances zones. Including the forests of the eastern flanks
altered past vegetation. Lower-Montane forests of the Andes, which merge into the Amazonian
are found at middle elevations, between 1,000 lowlands, they have exceptional levels of diver-
and 2,500 m, and can be as diverse and complex sity and endemism, combined with ongoing
as forests found in humid tropical lowlands. In- rapid deforestation and land use changes (e.g.,
termontane valleys cut through the tropical An- Young et al. 2007).
des, reaching as low as 2,000 m. Andean and Am-
azonian species and ecosystems form spatial 4.2. Lowland Amazonian Ecosystems
mosaics in the alluvial valleys above 1,000 m,
surrounded by slopes covered by montane for- 4.2.1 Terrestrial Ecosystems
ests (Josse et al. 2009). Below 1,000 m, Andean

Figure 4.2 Alexander von Humboldt’s Tableau Physique (Humboldt 1805), a graphic overview of plant communities, from the Pacific
to the Atlantic Ocean and passing over the Andean mountains. Reproduced with permission from the Peter H. Raven Library at
the Missouri Botanical Garden (https://www.biodiversitylibrary.org/page/9869921).

7
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

4.2.1.1 Lowland Rainforests The Amazonian forest holds approximately 392

billion individual trees with a diameter of over 10
Amazonian lowland rainforests cover approxi- cm (dbh) (ter Steege et al. 2013), amounting to
mately 5.79 million km2 over nine countries (ter 13% of all trees on earth (Crowther et al. 2015). If
Steege et al. 2013, ter Steege et al. 2015). Mean an- trees over 2.5 cm dbh are chosen (Draper et al.
nual rainfall varies from especially humid forests 2021) the number of 392 billion may easily dou-
in the northwestern Amazon (over 3,000 mm) to ble. The average density is approximately 570 in-
drier, more seasonal systems in the south (1,500 dividual trees per hectare, with the highest den-
mm) (Espinoza-Villar et al. 2009). Based on the sities in the wettest parts, notably the northwest-
maximum geological age of the soil producing ern Amazon (ter Steege et al. 2003).
materials, the area has been divided into six re-
gions (Quesada et al. 2011, ter Steege et al. 2013). The composition of Amazonian forests is deter-
These regions and their tree diversity are dis- mined primarily by soil fertility (ter Steege et al.
played in Figure 4.3. 2006, Tuomisto et al. 2019, Chapter 1), and an-
nual rainfall (ter Steege et al. 2006, Esquivel
Soils in the northwestern and southwestern Am- Muelbert et al. 2016). At the southern climatic
azon (parts of Colombia, Ecuador, Peru, and also margins of the Amazon the forest gradually
extending into western Brazil and parts of Bo- changes into cerrado (a tree savanna).
livia), originate from recent (Holocene and Qua-
ternary) Andean riverine sediments or Tertiary Cardoso et al. (2017) recorded 14,003 species,
estuarine deposits. These are typically more nu- 1,788 genera, and 188 families of seed plants in
trient-rich than the much older clays of the east- Amazonian lowland rain forest, with one-half of
ern Amazon, and soils derived from the ancient these trees capable of reaching ≥10 cm dbh
Precambrian Guiana and Brazilian Shields (6,727 species, 48% of the total flora; 803 genera,
(Quesada et al. 2010, 2011). However, the western 45% of the total genera). More than one-half of
soils are often less physically favorable to trees, seed plant species diversity in the Amazonian
being often shallower, with poorer structure and rain forests comprises shrubs, small trees, lia-
more prone to water-logging. Overall, therefore, nas, vines, and herbs (7,276 species, 52% of total
a rainfall gradient runs from the northwest (wet) flora). Three of these top 10 families are exclu-
to the south and southeast (drier), while a more sively herbaceous (Araceae, Orchidaceae, and
complex soil gradient runs almost orthogonal to Poaceae, except for bamboos such as Guadua
this, from the west and south-west (more fertile) species). Although a large proportion of its whole
to the east and northeast (less fertile). As a result, diversity is still not known, ter Steege et al. (2013,
the lowland forests of the southwestern Amazon 2020) estimated that the Amazon may hold close
have hot, moist, and somewhat seasonal climates to 16,000 tree species alone – from an estimated
very similar to the distant forests of the Guianas, total flora that ranges from 15,000 to 50,000 spe-
yet soils which are more fertile and, in terms of cies – of which 10,000 tree species have been col-
physical structure and rooting depth, often much lected in the area (ter Steege et al. 2016, 2019b).
less favorable. In spite of the similar climates, Truly core Amazonian species may be less than
there is almost complete turnover of dominant this, as many species from the cerrado or higher
tree species (ter Steege et al. 2006). Overlaid on elevations in the Andes are found in the edges of
these large-scale basin-wide patterns are com- the Amazon, which may largely explain the dif-
plex regional-scale and landscape-scale geomor- ference with the estimate of Cardoso et al. (ter
phological, fluvial, edaphic, and hydrological Steege et al. 2020). Regardless of the true total,
variations which help create the great biological Amazonian forests, especially in the western
richness and diversity of Amazonian ecosys- Amazon, include many of the most tree-species-
tems. rich ecosystems in the world (Sullivan et al. 2017).

8
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Figure 4.3 Map of tree α-diversity of the Amazon (http://atdn.myspecies.info), based on an interpolation of Fisher’s α of 2,282 plots
of mostly 1-ha. Black dots: Fisher’s α of individual plots. Green background color: the interpolated values calculated for 565 Amazo-
nian 1-degree grid cells (~111 km). In gray the six regions of the Amazon as used in this chapter (Quesada et al. 2011, ter Steege et
al. 2013).

More generally, even with a large proportion of From mathematical models it can be estimated
its diversity still not described, the Amazon that over 10,000 species number less than 1 mil-
houses a remarkable share of currently docu- lion individuals, while over 5,000 number less
mented global biodiversity, holding at least 18% than 5,000 individuals. The Amazon thus com-
of vascular plant species, 14% of birds, 9% of bines hyper-diversity with hyper-dominance
mammals, and 8% of amphibians found in the and hyper-rarity.
Tropics. As one example of the level of mamma-
lian diversity, of the 18 New World primate gen- Ten families contribute 65% of all trees in the
era, 14 occur in Amazon, and eight are endemic Amazon; Fabaceae (47 billion), Arecaceae (26 bil-
to the region (da Silva et al. 2005). lion), and Lecythidaceae (20 billion) are the most
abundant. The ten most abundant species are
While the forests are exceptionally diverse, the Eschweilera coriacea (4.7 billion), Euterpe precatoria
tree communities at large scale are dominated by (3.9 billion), Oenocarpus bataua (2.8 billion), Pseu-
relatively few species, and several of these domi- dolmedia laevis (2.8 billion), Protium altissimum (2.8
nants are widespread. As a result, a little over 200 billion), Iriartea deltoidea (2.6 billion), Mauritia
tree species (out of the estimated 16,000) account flexuosa (1.9 billion), Socratea exorrhiza (1.9 bil-
for half of all trees over 10 cm dbh (ter Steege et lion), Astrocaryum murumuru (1.8 billion), and Pen-
al. 2013, ter Steege et al. 2020). taclethra macroloba (1.7 billion) (ter Steege et al.
2020). It is interesting to note that palms

9
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

(Arecaceae) are the second most abundant family nutrient poorness but more likely a consequence
and contribute seven of the ten most abundant of their small, fragmented area (ter Steege et al.
species, yet consist of very few species compared 2000, 2019a). Because of the stark soil differ-
to the most abundant family, Fabaceae. The lat- ences between white sand forests and terra firme
ter have 789 species in the plot data of ter Steege forests, white sand forests are characterized by
et al. (2020), while Arecaceae have only 74. In fact, high levels of endemism (Adeney et al. 2016).
Arecaceae are five times more likely to be among
the ~220 hyperdominants than would be ex- Tree genera typically found in white sand forests
pected on the basis of its species richness. Faba- include Eperua, Micrandra, Clathrotropis, Dicymbe,
ceae are also the family with the highest tree spe- Hevea, Aspidosperma, Protium, Licania, Pouteria,
cies richness in the Amazon with 1,386 collected Swartzia (ter Steege et al. 2013). Impoverished ar-
species (ter Steege et al. 2019b), For all seed eas (often due to burning) tend to have more
plants the majority of the species-rich families scrub-like vegetation (locally called campina,
are small statured or herbaceous, except Faba- bana, muri scrub), often dominated by Humiria
ceae (Cardoso et al. 2017). balsamifera, and in the Guianas by Dimorphandra
conjugata as well (Lindeman and Molenaar 1959).
Tree species diversity is not evenly distributed Because of their isolation in small patches, white
across the Amazon (Figure 4.3). The highest di- sand forests may never recover species that have
versity is found in the northwestern Amazon and been lost (Álvarez Alonso et al. 2013). White-sand
central Amazon where single plots of one hectare ecosystems in the central Amazon still remain
may have over 300 tree species (Amaral et al. inaccessible and poorly studied (Adeney et al.
2000, Gentry 1988). Much lower diversity is the 2016).
Brazilian and Guiana shields, especially towards
the edges of the Amazonian forest. 4.2.1.3. Savannas and grasslands

Species richness is highest in Dryland (terra Savanna vegetation is characterized by the pres-
firme) forests (Figure 4.4), especially those of the ence of up to 40% tree cover, often less than 8 m
more fertile western Amazon, and lowest in tall, with a graminoid layer. Savanna occupies
flooded forests (várzea, igapó), swamp forests, and 14% of the Amazon basin (including the Tocan-
white sands. Although fertility and flooding may tins-Araguaia basin) and is distributed in terra
affect species richness, tree diversity (and its in- firme in the southeast of the Brazilian Amazon,
verse – dominance) is also linked to the total area and in permanently or seasonally flooded sites,
a particular system makes up in the Amazon (ter as in Beni savanna in Bolivia, in patches of open
Steege et al. 2000, ter Steege et al. 2019a). savanna under washed white sand across the
Amazon, or on degraded lands subject to fire.
4.2.1.2 White sand forests White sand savannas are mainly found in the up-
per Rio Negro area and the Guianas (see above).
White sand forests (known by common names Savannas extend over sandy-clay substrates and
like campinarana, Amazonian caatinga, varillar) eventually form forest islands – around 0.3 to 1.5
are found on pockets of highly leached deposits km2 – mixed with swamps in depressions and
of podzolized white-sand (Adeney et al. 2016). gallery forests within the basin, which are part of
the drainage system of the whole landscape.
White sand forests occupy roughly 3-5% of the Woody savannas on terra firme or slighty higher-
Amazon, with major occurrence in the upper Rio relief terraces of the alluvial plain are formations
Negro area and the Guianas (Adeney et al. 2016). with species of Curatella americana, Anacardium
They are generally species poor, especially in the microcarpum, Hancornia speciosa, Qualea grandi-
Guianas, a feature often attributed to their flora, Byrsonima crassifolia, and Tabebuia spp., as

10
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Figure 4.4 A. Key ecosystems are found in Amazonian lowland rainforests, such as floodplain forests, Amazon savanna, white-
sand savanna, and seasonally dry forest. B. The ten most encountered tree species on ~2,000 plots across the Amazon by forest
type (IG – igapó, PZ – white sand forest, SW – swamp forest, TF – terra firme forest, VA – várzea forest. Top lines: total species
encountered in plots in these forest systems and the percentage compared to the 5,058 species in all 2,000 plots (data: ter Steege
et al. 2015).

11
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

well as grasses such as Trachypogon, Paspalum, Cy- swamps, and sustains the greatest freshwater
peraceae, and others (Pires and Prance 1985). fish diversity on Earth; an ichthyofauna that is
equivalent to 15% of all freshwater species cur-
Among the animal species characteristic of the rently described (Junk et al. 2011, Tedesco et al.
savannas are the White-Tailed Deer (Odocoileus 2017). In the animal communities associated
virginianus), Greater rhea (Rhea americana), with aquatic ecosystems there are numerous fish
Southern screamer (Chauna torquata), Banded ar- species, and iconic species such as Capybara (Hy-
madillo (Dasypus novemcinctus), and maned wolf drochoerus hydrochaeris), Neotropical otter (Lutra
(Crysocyon brachyurus). longicaudis), Giant otter (Pteronura brasiliensis),
Amazon River Dolphins (Inia spp.), Yellow-Spot-
4.2.2. Fresh Water bodies and Wetlands ted River Turtle (Podocnemis unifilis), Matamata
(Chelus fimbriatus), Anaconda (Eunectes murinus),
Freshwater ecosystems in the lowland basin (el- Black Caiman (Melanosuchus niger), and other spe-
evations below 500 m) include rivers, lakes, and cies of crocodilians, among others.
streams, in addition to areas with permanent,
temporary, or seasonal standing or flowing wa- The Amazonian fluvial network is made up of dif-
ter, or with saturated soils, such as swamps, ferent types of waters (Figure 4.5). Amazonian
flooded forests, and marshes. These ecosystems rivers generally are classified into white-water,
are a fundamental part of the large fluvial system clear-water, and black-water, based on the color
of the Amazon and occupy >800,000 km2, or 14% of the water, which is related to transparency,
of the drainage area (Melack and Hess 2010; Hess acidity (pH), and electrical conductivity (Sioli
et al. 2015). Aquatic ecosystems in the Amazon 1984, Bogota-Gregory et al. 2020, Table 4.1).
are connected through the annual flood pulse, the These water characteristics also correlate to the
periodic fluctuation in water level that connects geological and geomorphological properties of
lowland rivers with their floodplains and allows the river catchments and their origins (McClain
the exchange of water, organic and inorganic ma- and Naiman 2008). The catchment properties di-
terials, and organisms (Junk and Wantzen 2003, rectly influence the composition and amount of
Junk et al. 2015; see 4.3.2 below). Depending suspended sediments in the water and, in turn,
upon classification criteria (e.g., scale, floristic the productivity of rivers and floodplain lakes (Si-
composition, geomorphology, the pattern of in- oli 1984). The fish communities in rivers and as-
undation, and water chemistry), aquatic ecosys- sociated floodplains also are influenced by water
tems and freshwater wetlands may vary from a characteristics. Conductivity and turbidity, in
few general types to more than 30 distinctive eco- particular, seem to be major drivers shaping Am-
systems (Comer et al. 2020). azonian fish communities (Bogota-Gregory et al.
2020).
4.2.2.1. Rivers, Lakes and Forest streams
White-water rivers (such as the Amazon main
The Amazon drainage basin is formed by the Am- stem, Caquetá-Japurá, Marañón, Ucayali, and
azon River and approximately 269 sub-basin Madeira) originate in the Andes, or, in the case of
tributaries with catchment areas between 300- the Jurua and Purus Rivers, in the hilly, rugged
1,000 km2 (Venticinque et al. 2016). The largest moderate elevations below 1,000 m in the Uca-
tributary systems that join the Amazon are the yali region in Peru. The Andean mountains sup-
Madeira, Negro, Japurá, Tapajos, Purus, and ply most of the terrestrial sediments, organic
other rivers that are among the 20 largest rivers matter, and mineral nutrients influencing the
on the planet. With more than 7,000,000 km2, the hydrology, geomorphology, biochemistry, ecol-
Amazon is the most extensive hydrographic net- ogy, and productivity of white-water rivers and
work in the world, bordered by riparian forests or their floodplains, all the way to the Amazon River

12
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Figure 4.5 Amazon River Network across the largest tributary systems and the entire Amazon Basin (source: Venticinque et al.
2016), indicating the distribution of flooded environments (modified from Hess et al. 2015). Wetland areas cover ~14 % of the
basin (nor considering Tocantins-Araguaia drainage and estuarine coastal areas) (5.83×106 km2) and 16 % of the lowland basin
(5.06×106 km2) (Hess et al. 2015).

13
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Table 4.1 Ranges of physico-chemical properties in blackwater, clearwater, and white-water for rivers and floodplain lakes across
the basin (gray text) (Source: Bogotá-Gregory et al. 2020). Conductivity (EC), dissolved organic carbon (DOC), dissolved oxygen (DO),
Inorganic (Inorg.), Herbaceous (Herb.). a Periodic phytoplankton (including cyanobacteria) blooms induce DO supersaturation (ca.
8–15 mg L− 1) and color clearwater green. b Precipitation of suspended silt due to reduced flow in white-water floodplain lakes sub-
stantially increases transparency relative to the parent white-water rivers. c High water hypoxia results from litter decomposition
in inundation forests; this effect is greater in large white-water floodplains. d Shallow white-water lakes reach extreme high low-
water temperatures.

Whitewater Clearwater Blackwater

Water Chemistry

pH High (6.5-7.5) (near neutral) Intermediate (EC 5.5-8.0) Low (3.5-6.0) (acidic)

Color Turbid, Cafe con Leche Clear or blue-greenish Reddish or brownish

Nutrient High (EC 40-300 µS cm-1) Low (EC 5-40 µS cm-1) Low (EC 5-20 µS cm-1)

Dominant cations Na+/K+ Variable Ca2+/Mg2+

Dominant anions CO3 2-/NO3 -/PO4 3- Variable SO4 2-/ Cl-

DOC High Low High

Transparency Low (0.1-0.6 [usually < 0.3] m) High (1-3 m) High (0.6-4 m)

Variable (LW <0.6, HW 0.5-3 m)b

DOa High (2-8 mg L-1) High (2-8 mg L-1) High (2-8 mg L-1)

Variable (LWc 2-8, HWc 0-3 mg L-1)

Temperature High (29-32°C) High (29-32°C) High (29-32°C)

Variable (LW 29-34, HW 27-32 °C)d

Inorg. sediment load High Low Low

Sediment type Fine alluvial silt Sand Sand

Sediment fertility High Low Low

Herb. macrophytes Absent-Sparse Absent-Sparse Absent-Sparse

Igapó
Floodplain forest Várzea (high-productivity) Igapó (low-productivity)
(intermidiate-productivity)

estuary, associated mangroves, and the ocean near-neutral pH, and the relatively high concen-
(McClain and Naiman 2008; Filizola and Guyot tration of dissolved solids is reflected in the elec-
2009; Encalada et al. 2019). Andean-derived large tric conductivity, which varies between 40–300
sediment loads control downstream channel ero- μS/cm (McClain and Naiman 2008, Bogota-Greg-
sion and width, bed elevations, and the availabil- ory et al. 2020). White-water rivers are sur-
ity of riparian habitats and vegetation. These, in rounded by diverse várzea floodplain forests and
turn, influence the connectivity between river extensive floating meadow wetlands (Wittmann
channels and floodplains, and therefore spatial et al. 2011, see 4.2.2.2. below).
patterns of inundation and floodplain productiv-
ity (Constantine et al. 2014; Forsberg et al. 2017). Clear-water rivers (such as the Tapajós and
White-water rivers are turbid, with water trans- Xingu Rivers) have their upper catchments in the
parency ranging between 20 and 60 cm, because cerrado region of central Brazil and drain the an-
the high sediment loads contain suspended clay cient Brazilian shield, which has been strongly
particles from drained soil and completely de- eroded over millennia (Sioli 1984). The pH of
graded plant material. White-water rivers have clear-water rivers varies from acidic to neutral,

14
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

depending on the soil, and the water hardly car- besse 2012). Depending on fluvial processes, two
ries any suspended and dissolved solids (Sioli other groups are recognized: 1) lagoons formed
1984). The transparency of their greenish waters by the lateral displacement of the channel, in
is high (100–300 cm), electrical conductivity stretches of abandoned channels and meanders
ranges between 5–40 μS cm, and pH varies be- (lagoons or swamps depending on the degree of
tween 5.5–8 in large rivers (Bogota-Gregory et al. sedimentation), and lagoons that join islands to
2020). the floodplain; and 2) lakes generated by geo-
graphical features such as those built by vertical
Black-water rivers have their origin in lowlands, accretion processes in the main channel and by
are translucent, high in dissolved organic car- floods in the alluvial plain (e.g., square lagoons
bon, and low in nutrients. Rivers such as the Ne- also influenced by tectonics in SW Amazon), or
gro in Brazil and Vaupés and Apaporis in Colom- by deltas of alluvial plains, with dikes and
bia drain the Precambrian Guayana shield, char- blocked valleys (e.g., ria lakes).
acterized by large areas of white sands (podzols).
Water transparency ranges between 60–400 cm, In meandering rivers such as those found in the
with low quantities of suspended matter but high Amazon Basin, sediment deposits rich in clay
amounts of humic acids (rich in dissolved or- form within floodplains. These clay deposits slow
ganic carbon (DOC) from the incomplete degra- water flow and thus help to decrease the migra-
dation of forest plant material), which give the tion rates of the channel – up and down streams
water a brownish-reddish color. The pH values – affecting bank erodibility on a large scale (10–
are in the range of 3.5–6 and electrical conduc- 50 km) and sinuosity by 30% (Schwendel et al.
tivity varies between 5–20 μS/cm (Bogota-Greg- 2015). The grain size of clay-rich sediment de-
ory et al. 2020). Clear and black-water rivers are posits is similar to that of deposits near the outlet
surrounded by another type of flooded forest, of a meandering lake (1.5–3.0 μm) and form clay
igapó (See 4.2.2.2. below for a detailed descrip- plugs (Gautier et al. 2010). The abandoned mean-
tion of Amazonian floodplain wetlands). ders of rivers are known as oxbow lakes that may
or may not recover the sinuosity of the river.
Nevertheless, many rivers and streams do not However, while stagnant waters remain, aquatic
easily fit into these three categories and are con- submerged plant communities rapidly colonize
sidered as “mixed waters”. Greater variability in floodplain lakes, including species such as Victo-
water biochemistry results from the influence of ria amazonica, Lemna spp., Nymphaea gardneriana,
lower-order tributaries with different biogeo- and Eichhornia spp., among others. Oxbow lakes
chemical water properties that vary seasonally of black-water rivers are typically free of aquatic
depending on flooding levels and connectivity. plant communities due to their low nutrient lev-
els.
Amazonian lakes are the result of fluvial pro-
cesses in depressions or flooded valleys. Four Few areas within the lowland Amazon are more
main categories are distinguished: 1) lagoons in than 100 m above the river, where water comes
ancient lands not directly related to river sys- to the surface in the form of a dense network of
tems (e.g., the Hill of Six Lakes in the northern small streams. Most stream fauna depends on
Amazon), 2) lakes in river valleys and quaternary energy inputs from the surrounding forest (e.g,
sediments (not related to geographical features: insects and plant material) and much of the ter-
e.g., Pará and Rondonia states), 3) lakes gener- restrial flora and fauna also depend on resources
ated by river processes (e.g., the Boa Vista For- from streams. Intricate connections between
mation in the northern Amazon), and 4) "lakes" aquatic and terrestrial ecosystems continue as
of wetlands (a mosaic of lakes with a large diver- the streams coalesce to form larger rivers. In
sity in origin, shape, and functioning) (Latru- general, small streams are considered part of the

15
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

terra firme forest ecosystem and harbor great km2, 10%), and subjected to predictable, long-
aquatic biodiversity (Arbelaez et al. 2008). How- lasting, annual flood pulses (Junk et al. 2011; also
ever, as they form larger rivers, the forest canopy see 4.3.2. below). These forests are flooded due to
is no longer continuous, instead, the floodplain their low topographic location and poorly
areas around rivers support extensive forests drained soils. Flooding may last up to six months
(see 4.2.2.2. below), and the terrestrial and and water levels may fluctuate up to 10 m be-
aquatic ecosystems become more distinct (see tween the dry and flood seasons (Schöngart and
3.2. below). Junk 2007). The timing, duration, and magnitude
is variable across the basin. Such temporal and
4.2.2.2. Freshwater Wetlands spatial variation is mostly driven by air circula-
tion patterns and headwater precipitation modu-
There are several definitions of wetlands, but a lated by the Intertropical Convergence Zone and
broad and simple definition is proposed by Junk topography (Siddiqui et al. 2021). Although these
et al. (2011, 2014), which states that “wetlands are forests are flooded annually, different floristic
ecosystems at the interface between aquatic and zones are distinguished, which are influenced by
terrestrial environments with biota adapted for the input of sediments and nutrients in river wa-
life in water or in water-saturated soils.” Recent ters, flood regimes, and hydro-geomorphic dy-
large-scale mapping efforts have identified nu- namics (Prance 1979, Wittmann 2010).
merous wetlands dominated by vegetation, in
different sub-basins of the entire Amazon Basin. Floodplain forests along white-water rivers are
If we consider small riparian wetlands and wa- known as várzea in Brazil (or rebalse in Colombia)
terlogged savannas and grasslands, the esti- and represent the most extensive type of flooded
mated area covered by wetlands extends to 2.3 forest in South America, covering approximately
million km2 or 30% of the basin (Junk et al. 2011). 0.46 million km2 of the Amazon Basin (Junk and
Wetlands are divided into two main groups: 1) Wittmann, 2017). Amazonian white-water river
those with relatively stable water levels (e.g., floodplain forests contain around 1,000 species
Mauritia flexuosa palm swamps), and 2) those with of trees, making them the most diverse flood-
oscillating water levels (e.g., floodplain forests, plain forests in the world (Ferreira and Prance
mangroves). Some of these wetlands are forest- 1998; Wittmann et al. 2002, 2006). A significant
dominated and broadly distributed. In contrast, number of tree species are almost entirely re-
others are emblematic as they represent specific stricted to the floodplain (~40% of the most com-
regions within the basin, such as savanna eco- mon central Amazonian várzea tree species),
systems in the Llanos de Moxos, located in the while only ~31% of tree species in várzea are
Madeira basin of Bolivia; Bananal savannas of shared with terra firme forest (Wittmann et al.
Brazil which are seasonally inundated grass- 2011). Due to the seasonal influx of nutrients car-
lands, sedgelands, and open woodlands among ried by white-water rivers, floodplain forests are
many others (Castello et al. 2012, Figure 4.1). In eutrophic and highly productive (Junk and Pie-
the Upper Negro river basin, the Amazonas Sa- dade 1993), but their flora and fauna diversity is
vannahs Refuge and parts of the Imeri Refuge are less than that of terra firme forest (Patton et al.
considered centers of endemism for floodplain 2000; Haugaasen and Peres 2005a, b). This is be-
tree species, such as Mauritia carana, Ocotea es- cause of the selective pressure imposed by pro-
meraldana, and Vitex calothyrsa (Junk et al. 2010). longed annual floods. Due to its high productiv-
All of these wetlands are vital to support local ity, várzeas have been important centers of hu-
communities' livelihoods. man colonization which have intensified in the
last thirty years (Piedade et al. 2010). Data on the
Floodplain Forest Seasonally flooded forests are productivity of Amazonian aquatic ecosystems
second in area to terra firme forests (0.76 million are relatively few, but those available show that

16
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

remarkably high values are locally possible. This 2010; Junk et al. 2011). Due to the lack of soil nu-
is likely due to the combination of abundant nu- trients, tree abundance and biomass in igapó for-
trient and water supply, insolation, and macro- ests is much lower than in várzea and terra firme
phytes adapted to rapidly occupy the water-at- forests (Ferreira 1997, Junk et al. 2015, Wittman
mosphere interface when conditions permit (Ta- and Junk 2017). Montero et al. (2014) recorded
ble 4.2). The floodplain forests of Brazil, Peru, 6,126 trees with 243 species, 136 genera, and 48
and Ecuador are characterized by the presence families in 10 hectares along the middle Rio Ne-
of families such as Fabaceae, Moraceae, Are ca- gro. Most species found in igapó also occur in
ceae, Lecythidaceae and Annonaceae (Nebel et al. other ecosystems, such as terra firme and várzea
2001) and the flooded period may vary from 1 or forests, savanna, swamps, or white-sand forests
2 months to 6 months. In varzeas of the central (Junk et al. 2015). Among herbs, 55 species have
Amazon, characteristic tree species include Ceiba been documented, belonging to 20 families
pentandra, Hura crepitans, Nectandra amazonum, (Lopes et al. 2008); most of the species were found
and Cecropia spp. (Worbes 1997). These species with an exclusively terrestrial habit in the igapó
represent the early sequence forest species, have and belong to two main families: Cyperaceae
low wood density, and make up the successional (45% of the total) and Poaceae (7.3%) (Piedade et
process which is governed by hydrological sea- al. 2010).
sonality. Tree density (at 10 cm dbh) in várzea
varies along successional stages and flood-gradi- In general, comparison between terra firme,
ent position (i.e., high and low varzeas), being in várzea and igapó forests shows differences in tree
average 400–500 individuals ha−1 and with high- richness (Figure 4.4) and structural trends in the
est values occurring in early-secondary stages number of individuals. In general, terra firme for-
(800–1,000 individuals ha−1) (Wittmann et al. est shows greater density and richness of large
2011). trees (diameter at breast height ≥ 10 cm), fol-
lowed by várzea and igapó forests.
There are also floodplain forests along black-wa-
ter rivers (Junk et al. 2011), called igapó in Brazil. Permanently Flooded Swamps Permanently flooded
The igapó forests are seasonally flooded by black or waterlogged areas (swamps) occupy a small
(or clear) water rivers, for up to 9 m in depth, and area compared to other ecosystems in the Ama-
cover around 302,000 km2 (Melack and Hess, zon (80,000 km2, 1%). The extensive palm

Table 4.2 Net primary production (NPP, dry weight) for the most important populations and communities of aquatic herbaceous
plants in central Amazon várzea. NPP was measured under different methods and assumed to have a monthly loss between 10 and
25% of the biomass (Source: Piedade et al. 2010).

Maximum NPP Time for production

Population/Community
(t.ha-1) (months)
Monospecific stands of Echinochloa polystachya
100 12
(Kunth) Hitchock1
Monospecific stands of Paspalum fasciculatum
70 7.7
Willd.2
Mixed populations dominated by Hymenachne
48 9.5
amplexicaulis (Ruudge) Nees2
Monospecific stands of Paspalum repens P.J.
33 4
Bergius2
Monospecific stands of Oryza perennis Moench2 27 4
Mixed populations dominated by Oryza peren-
17.5 5
nis Moench2

17
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

formations of Mauritia flexuosa, Oenocarpus ba- flooded regions. Such ecosystem shifts favor
taua, and Euterpe oleracea (Arecaceae) are very grasslands and deteriorated aquatic communi-
characteristic of swamps of the Amazon. Their ties, as was demonstrated in the Pantanal which
distribution is azonal as they are found from the is considered a hyper-seasonal savanna (Nunes
lowland plain to the Andean foothills, up to 500 da Cunha and Junk 2004).
m of altitude, always associated with highly stag-
nant black-waters (Moraes R et al. 2020), such as Mangroves Mangroves occupy relatively small ar-
in permanent wet depressions within the sa- eas in a narrow littoral belt towards the Atlantic
vanna landscape (Mauritia flexuosa) (Junk et al. Ocean and in the Amazon estuary. Mangroves are
2010). There are also permanent swamp areas subject to flooding by salt water or brackish wa-
with rooted plants in channels or depressions ter and have only a few tree species, generally
within the alluvial plain, characterized by herba- uniform in structure, not exceeding 10 m in
ceous species including Cyperus giganteum, Thalia height. The dominant mangrove species (in or-
geniculata, Pontederia spp., Eichornia spp., among der of abundance) are Rhizophora mangle (com-
others (Pires and Prance 1985; Beck and Moraes mon names are mangue verdadeiro in Brazil, red
R 1997). mangrove elsewhere), Avicennia nitida, and La-
guncularia racemosa (Pires and Prance 1985, Junk
Flooded Savanna The seasonally flooded savannas et al 2010). Brazilian mangroves occur mostly
of the alluvial plain cover an area of ca. 200,000 along the coasts of Amapá, Pará, and Maranhão
km2 (Pires and Prance 1985) and represent 6% of states and cover an area of about 14,000 km2 (IC-
flooded plant communities (Meirelles 2006). MBio 2018). The largest mangrove area extends
They occur in the northern (Roraima and Ru- southward from Belém and measures at least
pununi) and southern (Beni savanna) Amazon, 7,000 km2 (FAO 2007; Menezes et al. 2008). Little
along the cerrado belts in Brazil and the Guianas, is known about the wetlands along the coastline
and have strong climatic seasonality (several dry north of Belém. For Guyana, Huber et al. (1995)
moths) (Junk et al. 2011). estimated that there are about 900 km2 of coastal
mangroves. In areas with very strong freshwater
Flooding is mainly influenced by rainfall and the influence near the Atlantic coast, várzea forests
overflow of rivers during 3-5 months of the year, may replace mangroves.
but in a matter of hours, the flooding percolates
and the landscape returns to its natural state 4.3. Ecosystem Functioning
without permanent water, except in lower places
and in depressions linked to rivers. On alluvial 4.3.1. Primary productivity, nutrients, forest
plains of white-water rivers, Poaceae species dynamics and decomposition
predominate (32% of the total), followed by Cy-
peraceae (20%) (Junk and Piedade 1993), and 4.3.1.1. Terrestrial ecosystems
their contribution to net primary production
(NPP) make them the most important aquatic In the Amazon, climatic factors exert the greatest
herbaceous plant community (Piedade et al. influence on gross production (GPP) in terrestrial
2010). ecosystems, but a wide range of other factors re-
lated to soil, forest disturbance, and species com-
Flooded savannas and grasslands are very fragile position are also influential in determining how
ecosystems. Savannization processes are being captured carbon is allocated and how long it is
generated by the reduction of floodplain forests stored in tree woody biomass and other ecosys-
due to various dynamics, such as deforestation tem compartments. Thus, bottom-up studies of
and fires driven by severe droughts in minimally the carbon budget and its seasonal variation us-
ing intensive measurements in plots of the GEM

18
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

(Global Ecosystems Monitoring) network (Malhi the productivity of the forest (e.g., Sobral et al.
et al. 2021) show variation in GPP between sites 2017), and it is possible that the pre-Colombian
from around 33 to 38 Mg C ha-1 yr-1 for more hu- extinction of Amazonian megafauna has im-
mid forests (in the west and north) to lower val- pacted productivity negatively by slowing the nu-
ues of 25 to 30 Mg C ha-1 yr-1 in drier forests of the trient transfer from richer floodplains to hinter-
Brazilian Shield and central Amazon (Malhi et al land terra firme forests, a function which the orig-
2015). However, carbon-use efficiency (CUE), de- inal large herbivores would have performed
fined as the fraction of fixed carbon that is used (Doughty et al. 2016).
to produce plant matter, i.e. NPP divided by GPP,
appears to be lower (0.3 – 0.4) in wetter sites than What does all this mean for forest dynamics, bio-
in more seasonal sites (0.4 – 0.5). Overall, the de- mass, and carbon storage? Inventory plots show
cline in GPP in the drier sites is compensated by that differences in above-ground biomass track
shifts in CUE and in allocation, so that in these more closely to underlying edaphic factors than
studies there is often no clear decline in tree to climate factors.
woody growth toward more seasonal parts of the
Amazon. Compensatory shifts in CUE and alloca- Mortality rates vary greatly across the Amazon,
tion unrelated to climate thereby may effectively being higher in the western and southern re-
decouple spatial variation in GPP, NPP, and gions, around 2.2-2.8% per year, than in the
woody growth. northern and eastern central regions where 1.1 -
1.5% is typical (Phillips et al. 2004, Marimon et al.
Less intensive but more extensive measure- 2014, Esquivel et al. 2020). Fast turnover forests
ments of woody growth and tree mortality (Box often correspond to where soils are relatively
4.1), combined with species composition and soil rich chemically but offer poor structural support
measurements, help confirm the role of non-cli- physically. Associated with these high rates of
matic factors in affecting how carbon is allocated stand-level tree mortality is the prevalence of
in Amazonian ecosystems. In the widespread species with ‘live-fast-die-young’ life-history
RAINFOR forest inventories, above-ground strategies that tend to favor growth over survi-
woody production is more closely linked to vorship, with lower wood density storing less car-
edaphic factors, such as phosphorus concentra- bon (Baker et al. 2004, ter Steege et al. 2006,
tions, than to climate (e.g., Quesada et al. 2012). Honorio Coronado et al. 2009, Patiño et al. 2009).
Other non-climate factors play a role too. Thus,
the high tree mortality rates of some Amazonian Remarkably, basal-area weighted wood density
forests as a result of wind-disturbance (e.g., Es- in the slow-turnover forests of the northeast Am-
quivel Muelbert et al. 2020), while the poor phys- azon is up to 50% greater than in fast-turnover
ical structure and shallow rooting depths of forests in the south and west (Phillips et al. 2019).
many western Amazonian soils (Quesada et al. In sum, three decades of careful observation in
2012), ensure that more forest here is naturally permanent plots shows that spatial variation in
in early to intermediate successional states. Amazonian biomass carbon stocks and dynam-
These tend to produce wood faster and may have ics are driven more by soil conditions than cli-
greater carbon use efficiencies (Rödig et al. 2018). mate, and more by spatial variation in mortality
Additionally, the nature of the species present than productivity. These findings run counter to
makes a difference too; where tree phylogenetic the dominant paradigm in ecosystem vegetation
diversity is greatest, forests have greater levels of models which has emphasized the role of climate
woody productivity, even accounting for and processes of carbon production (GPP, NPP,
covarying climate and edaphic factors (de Souza tree growth), rather than its turnover and loss (es-
et al. 2019). There is also evidence that animals pecially mortality), and which often ignore the
may increase nutrient cycling and subsequently physical constraints and floristic compositional

19
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

BOX 4.1 How much does the longevity of Amazonian species vary?

Tree age has generally been inferred based on trunk diameter growth rates (growth rings) (Figure
B4.1A)., mortality (Condit et al. 1995, Shöngart et al. 2015), or radiocarbon dating (14C) (Chambers 1989,
Vieira et al. 2005). The maximum longevity values based on demographic studies were inferred in 93
species of canopy trees in the rain forest in the Central Amazon, considering the influence of the life
cycle, such as wood density, growth form, mortality rate, rate of recruitment, trunk diameter, increase
in growth, and population density. Maximum longevity ranged from 48 years for the pioneer tree
Pourouma bicolor (Cecropiaceae) to 981 years for the canopy tree Pouteria manaosensis (Sapotaceae), with
an overall average of 336 + 196 years (Laurance et al. 2004). These estimates of tree maximum age
coincided with the analyses of the average mortality rates; the longevity of the tree was positively cor-
related with the density of the wood, the maximum diameter of the stem, and the population density,
while it was negatively related to annual mortality, recruitment, and growth rates; pioneer species had
much shorter longevity than climax trees (Laurance et al. 2004).

Tree age data provide important information for conservation and sustainable forest management.
Emergent old-age trees in the central Amazon, for instance, represent a key component of the forest’s
carbon budget, as around 50% of the aboveground biomass is retained in less than the 10% of the larg-
est trees (Chambers et al. 1989). The time required for a tree to achieve a certain diameter varies with

Figure B4.1 (A) Stem disk, and Tree rings of Bertholletia excelsa Bonpl. (Lecythidaceae) from a plantation tree in Manaus. Tree
rings are defined by an alternating pattern of fiber (dark tissue) and parenchyma (light tissue) (Shongart et al. 2015, ©Wiley). (B)
Bertholletia excelsa achieves 50 meters’ height tree in terra firme forests and 400 years of age (© WWF-Brazil / Clóvis Miranda).

20
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

BOX 4.1 continued

radial growth rates, with the cambial activity being influenced by abiotic site conditions and precipita-
tion that limits water in the dry season (Worbes 1999). Bertholletia excelsa (Lecythidaceae), a tree of 50
m height, may have 400 years and a diameter of 150 cm (Figure B4.1). As growth is higher under favor-
able light conditions (e.g., under canopy gaps), a tree of 10 cm diameter can have an age varying from
13 to 50 years (Shöngart et al. 2015). The flood-tolerant tree Calophyllum brasiliense (Calophyllaceae) may
achieve a maximum age of 490 years in a black-water floodplain. Under permanently waterlogged con-
ditions the longevity is reduced to 72 and 134 years. As consequence, for achieving the 50 cm diameter-
cutting limit based on forest management norms in the Brazilian Amazon, C. brasiliense would spend 70
years in white-river floodplains, but a remarkable 400 years in black-water floodplains (Rosa et al.
2017), suggesting habitat-specific Growth-Oriented Logging is needed to ensure species conservation
(Schöngart 2008).

The relation between radial growth rates and precipitation in the Amazon floodplain allows an esti-
mate of the effect of climate variability induced by the El Niño phenomenon with forest dynamics. Low
precipitation events influenced by El Niño (see Chapter 22) are related to increased growth periods in
the long-living (143 to 289 years old) hardwood species Piranhea trifoliata Baill. (Picrodendraceae). Un-
like in terra firme forests, the influence of drought on growth rates in floodplain trees may increase
carbon absorption, partially compensating the carbon emitted from terra firme forests under El Niño
periods (Shongart et al. 2004). Efforts to determine the age and growth rate of tropical trees under
flooded and non-flooded conditions, and the influence of climate and soil conditions on growing pat-
terns, are essential to guide wise use and long-term preservation (Vetter and Botosso 1989, Shöngart
et al. 2008).

factors which largely determine Amazonian for- et al. (2016) found that Neotropical primate abun-
est biomass. dance and diversity are largely controlled by fruit
production, and with much greater biomass and
The key effects of soils on Amazonian ecosystem diversity in the western Amazon than in the Gui-
function extend also to animals and their im- ana and Brazilian Shields. Such effects are likely
portant functions, including herbivory and seed to extend to many other animal groups, as we
dispersal. Travelers from the west to the east of have known for more than a third-of-a-century
the Amazon are often struck by the remarkably that production of flowers and fruits in the neo-
low level of insect activity, which can make field- tropics is closely tied to soil nutrient status (Gen-
work much more comfortable. This likely reflects try and Emmons 1987).
fundamental controls of cations and other nutri-
ents on the metabolism of animal consumers Finally, we note that climate nevertheless does
(e.g., Kaspari et al. 2009) as well as plant produc- impact rates of woody production, and clearly
ers (e.g., Lloyd et al. 2015). In the white sand for- has consequences for forest carbon storage and
ests of the Amazon, the interaction of impover- biodiversity. Both worldwide and in the Amazon,
ished soils and herbivory can select for invest- woody production is suppressed in the most ex-
ment in defense by the plants, while in forest for- treme seasonal tropical forest climates with high
mations with clay soils species are instead fa- maximum temperatures and high seasonal water
vored that commit resources to rapid growth deficits (Sullivan et al. 2020). This means that
(Fine et al. 2006). Large animals too respond to some Amazonian forests are already at the cli-
bottom-up soil controls; for example, Stevenson matic limits capable of sustaining productive

21
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

forest ecosystems. As a consequence, in some of and hence on forest plant and faunal community
the tropical forests which have warmed and dynamics. In elevation gradients at the Andes-
dried most, the long-term carbon sink of a ma- Amazon interface in Peru, temperature is the
ture forest appears to have recently weakened variable that best explains variations in litter de-
(Hubau et al. 2020). In the Amazon we also know composition rates (Salinas et al. 2011). Pinto et al.
from long-term RAINFOR plots that forest com- 2018 indicate that, as an effect of global change,
position is being affected by recent droughts, increases in temperature and dry season dura-
with the mortality of wet-affiliated genera in- tion are anticipated for the southern Amazon Ba-
creasing in places where the dry season has in- sin and the Pantanal (Gatti, et al. 2014; Junk
tensified most (Esquivel Muelbert et al. 2019). 2013), so these are likely to induce changes in de-
However, not all Amazonian forests appear to be composition rates and patterns. Also, the physio-
so impacted, with large areas with shallow water logical, morphological, and biochemical charac-
tables in the central and western Amazon poten- teristics of Amazonian tree species (their func-
tially effectively immunized against drought via tional traits) play an important role in their de-
local water supplies, in some cases even seeing composition. Species type has a large influence
an increase in growth and carbon stocks during on the decomposition rate (k) (Hättenschwiler et
recent drought (Sousa et al. 2020). Key areas of al. 2011), most probably through its influence on
scientific uncertainty include the extent to which wood density and leaf quality and morphology.
recent climate change has actually caused the For example, the influence of leaf anatomy is
slowdown in the intact Amazonian biomass car- manifested primarily through spongy paren-
bon sink (Brienen et al. 2015), and whether it chyma thickness, which strongly influences the
might soon go into reverse, with the remaining moisture-holding capacity of the leaf material,
intact Amazonian forests becoming a net carbon which in turn largely explains the observed mois-
source under further warming, as some have ture content in the leaves.
predicted (e.g., Hubau et al. 2020, Sullivan et al.
2020). Alternatively, forests may prove more cli- 4.3.1.2. Freshwater ecosystems
mate-change resistant than expected, especially
if the shallow water tables, wetter climates, and As with terrestrial ecosystems, the functions of
rich biodiversity of many Amazonian forests, in aquatic ecosystems comprise biochemical activ-
the west especially, help prevent large regions of ities of productivity (plants and algae), decompo-
the Amazon from becoming a net carbon source. sition of dead organic matter, and processes re-
Critical, of course, to the fate of the intact forest lated to the flow of energy and nutrient recycling
sink will be whether the forests themselves sur- (Morris 2010). These functions affect and are af-
vive. A recent analysis shows that for parts of the fected by interactions between living organisms
eastern Amazon carbon losses from deforesta- and consecutively sustain biodiversity and hu-
tion and degradation already exceed the sink in man well-being. However, unlike terrestrial eco-
remaining forest lands (Gatti et al. 2021). systems, the flow of water makes aquatic ecosys-
tems highly dynamic in both space and time.
To complete our picture of forest dynamics, we This is due to changing physical conditions and
need to understand the decomposition of dead biotic components along stream and river chan-
organic material as a fundamental biogeochemi- nels, from the headwaters to downstream conflu-
cal process, both through its role in the forest ence with other rivers or the sea, and the influ-
carbon (C) cycle and, perhaps more importantly, ence of precipitation on streamflow.
through its role in the recycling of nutrients to
soil and plant communities. Any changes in de- The flow of energy and nutrient recycling are
composition processes will have profound im- prime examples of the dynamic nature of aquatic
pacts on the rate and pattern of nutrient cycling, ecosystems, and the Amazon is no exception.

22
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Headwater and forest streams are shaded by veg- forests represent the most carbon-dense ecosys-
etation, inhibiting algae growth, a key energy tem (1,391 ± 710 Mg C ha−1) in the Amazon
producer in aquatic ecosystems. Instead, ripar- (Draper et al. 2014). Our knowledge here is in-
ian vegetation subsidizes aquatic food webs that complete, and peat may extend up to nine meters
are dominated by shredder invertebrates and de- deep (Householder et al. 2012). Recent intensive
composer bacteria that help recycle nutrients fieldwork revealed 61% more area of peatland
(Vannote et al. 1980). Nutrients travel down- pole forest in north Peru than initially thought
stream in a spiral-like pattern and, as the width (Coronado et al. 2021, Honorio et al. 2021). Be-
of the river channel expands downstream, algae cause peatland ecosystems function as carbon
growth is no longer limited by shading (Vannote sinks, they play a crucial role in maintaining the
et al. 1980). The lack of dissolved nutrients limits natural balance of the carbon cycle, modulating
algae production in nutrient-poor rivers such as global warming. Recent models estimate that
Amazonian clear-water and black-water rivers, 38% of Amazonian wetlands form peat deposits;
while acidity and low light penetration in dark- however, the lack of climate data needed to build
stained water further limits productivity in hydrological models hinders quantification of
black-water rivers. In turbid white-water rivers, the true extent of peatland ecosystems within the
light penetration also is a limiting factor to algae Amazon basin, and thus the importance of the re-
growth (Moreira-Turcq et al. 2003; Dustan 2009). gion in the global greenhouse gas budgets (Gum-
By connecting rivers with floodplain habitats, the bricht et al. 2017).
flood pulse provides a mechanism to compensate
for limited in-situ algal productivity by replen- 4.3.2. The Flood Pulse and Aquatic-Terrestrial
ishing nutrients during the annual flood (Junk Transition Zone
and Wantzen 2003, see 4.3.2 below).
Variation in water flow and depth is driven by re-
Some wetlands contribute to carbon storage at a gional and local precipitation patterns, which,
global scale due to the extensive and deep accu- coupled with variations in stream order, latitude,
mulation of below-ground peat deposits. Peat is a and elevation across the enormous Amazon Ba-
type of soil with a top layer composed of at least sin, create distinctive flow regimes (Goulding et
50% decomposed or semi-decomposed organic al. 2003, Siddiqui et al. 2021). In a recent classifi-
material (i.e., 29% carbon content), extending at cation, Siddiqui and collaborators (2021) identi-
least 30 cm deep (Gumbricht et al. 2017). Several fied 6-7 flow regimes based on a combination of
factors are essential in determining the location hydrological characteristics that include the tim-
of peatland ecosystems, including high rainfall, ing of the wet season, the magnitude of change in
frequent flooding, low drought and fire fre- streamflow, and the number of times streamflow
quency, and a low-lying topography that creates changes from rising to falling within a year. The
waterlogging and anoxic conditions for peat ac- timing of maximum flow, for instance, changes
cumulation (Draper et al. 2014). Peatland ecosys- spatially across the Amazon basin, with maxi-
tems also are influenced by different types of wa- mum flooding occurring in February-March in
ters, with a gradient of nutrient content. They can the southern tributaries and June-July in the
be nutrient-poor ombrotrophic bogs if they are northern tributaries. The magnitude of change in
dominated by atmospheric water, or they can be streamflow increases in lower elevation areas,
nutrient-rich swamps that are influenced by riv- while at the same time, the frequency is reduced
ers (Lähteenoja and Page 2011). For example, in to a single large flood episode. Rainfall in the
the Pastaza-Marañon foreland basin located in headwaters of large Andean rivers causes a flood
the western Amazon in Peru, an area of 35,600 ± pulse that travels downstream and leads to a pre-
2,133 km2 contains 3.14 (0.44–8.15) Pg C below dictable annual hydrological cycle with distinct
palm swamps. At the same time, peatland pole water-level periods (rising, flood, falling, and dry)

23
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

and long-lasting flooding (4-15 m in depth and performing annual lateral migrations into the
weeks to months in duration) in floodplains of white-water floodplain habitats that largely con-
lowland rivers (≤ 500 m). This flood pulse drives tribute to their productivity (Junk et al. 1984; Bay-
multiple physical, biological, and ecological pro- ley and Petrere 1989; Bayley 1995; Goulding et al.
cesses in the Amazon Basin, from sediment 1996, 2019; Isaac et al. 2016). In floodplain lakes
transport to fish migration. In addition, the flood connected to white-water rivers, the lack of cur-
pulse drastically transforms the landscape of low- rents allows sediment settling and greater water
land rivers by creating an aquatic-terrestrial transparency, facilitating phytoplankton growth
transition zone (ATTZ) that allows the movement and fueling a zooplankton-based food web. Thus,
of nutrients and organisms between river chan- floodplain lakes play a key role as nurseries and
nels and floodplain habitats (Junk and Wantzen feeding grounds to juvenile fish of commercial
2003). value (Oliveira 2006). The current consensus
among researchers is that a mixture of carbon
Interactions between terrestrial and aquatic generated in seasonally available floodplain hab-
components are among the most important pro- itats by algae, forest vegetation, and aquatic
cesses of Amazonian ecosystems. Floodplain plants plays a pivotal role in subsidizing aquatic
wetlands controlled by the seasonal flood pulse of food webs and commercial fisheries across the
white-water rivers are probably the best-docu- Amazon (Benedito-Cecilio et al. 2000, Santos et al.
mented examples of the importance of ATTZ in 2017, Correa and Winemiller 2018).
the Amazon basin (Junk 1984). These Amazonian
floodplains, which are among the most produc- Massive annual fish migrations transfer a small
tive natural systems on Earth, originate from the portion of Andean-derived energy and nutrients
accumulation of large sediment loads drifting from the white-water floodplains to the nutrient-
from the Andes, fueled by their associated nutri- poor black- or clear-water tributaries (see details
ents (Junk 1984; Melack and Forsberg 2001; below). Another perfect illustration of the inti-
McClain and Naiman 2008). Complex floodplain mate ecological interactions between the aquatic
macrophyte and forest communities have and terrestrial systems is the ancient mutually
adapted to these seasonal sediment fluxes and beneficial co-evolution and co-adaptation be-
year-round lateral exchanges between the main tween trees and fishes in Amazonian floodplains.
channel of rivers and their floodplains. Most tree species produce fruit during the high-
water season when fish invade the flooded forest
Terrestrial primary production, organic matter, (Ferreira et al. 2010; Hawes and Peres 2016).
and nutrients captured when floodwaters invade Hundreds of fish species have evolved frugivory
the floodplains decompose or are consumed by habits and may have been the first vertebrate
organisms become the basis of the aquatic food seed dispersers in the Amazon (Goulding 1980;
chain (Junk 1984; Melack and Forsberg 2001). Correa and Winemiller 2014; Correa et al. 2015a).
Part of this productivity goes back to the river’s They eat fruits falling in the water from flood-
main stem through the many organisms that plain trees and disperse their seeds over long
move between the floodplains and the river, in- distances, improving their germination and
cluding large numbers of fishes during massive thereby contributing to the maintenance of the
annual migrations (Goulding 1980, 1993). Flood- flooded forest (Goulding 1980; Kubitzki and Zi-
plains play crucial roles as feeding grounds and burski 1994; Waldhoff et al. 1996; Correa et al.
nursery areas for many fishes (Lima and Araujo- 2015a, b). In addition to fruits, fish also consume
Lima 2004; Castello et al. 2015, 2019). For in- copious amounts of invertebrates that undergo
stance, most commercially important fishes sup- vertical migrations toward the forest canopy dur-
porting large fisheries in the Amazon basin are ing the flood season. The consumption of leaf-
detrivore, herbivore, and omnivore species eating insects and carnivorous invertebrates

24
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

that, in turn, predate upon leaf-eating insects the total flora range from 15,000 – 55,000. As in
creates an indirect feeding link between fish and other tropical areas, Fabaceae (the bean family)
trees. Whether directly or indirectly, flooded for- is the most abundant and species-rich of all
ests provide a critical terrestrial subsidy to river- woody plant groups. South America and the Am-
ine fishes, particularly in nutrient-poor black- or azon are also renowned for their great abun-
clear-water rivers (Correa and Winemiller 2018). dance and richness of palms.

The flood pulse influences multiple aspects of The Amazon region holds the largest tropical
fish reproductive strategies, including fecundity wetland system on Earth, home to 15% of all
(number of eggs), age at first reproduction, num- known fish species (see Chapter 3, Jézéquel et al.
ber of reproductive episodes per year, and paren- 2020).
tal care (Tedesco et al. 2008). As a result, changes
in water levels affect fish species differently, and Its rivers are classified as white-water (rivers car-
fishing yields can lag 2-3 years. The flood pulse rying sediments from the Andes); clear-water
also affects the movement patterns of terrestrial (draining the two shield areas); and black-water
animals between floodplain and adjacent terra (draining the white sand areas). The water type
firme forests. During the flood period, abundant determines the forest type along the rivers, with
fruits attract frugivorous monkeys to floodplain igapó forest growing in sediment-poor clear and
forests, while kingfishers track fish movement to black-waters floodplains, and várzea floodplain
the interior of flooded forests. During the dry pe- forests in white, sediment-rich waters. The phys-
riod, seedling germination drives the movement ical-chemical characteristics of the different wa-
of terrestrial animals to floodplain forests, while ter types, particularly electrical conductivity and
hummingbirds take advantage of the synchro- turbidity are major factors shaping fish commu-
nicity in flower production (Haugaasen and nities in rivers and associated floodplains. The
Peres 2007, Beja et al. 2009). Moreover, flooding flood pulse causes marked periods of floods and
enhances habitat heterogeneity in floodplain for- droughts, which drive physical, biological, and
ests, which influences the formation of unique ecological processes, from sediment transport to
bird, bat, and amphibian communities not found fish migration, and together with the elevational
in adjacent terra firme forests (Beja et al. 2009, Pe- gradients in the floodplain are factors that favor
reira et al. 2009, Ramalho et al. 2018). the maintenance of various plant communities.
The white-water wetlands are probably the best-
4.4. Conclusions documented examples of the importance of the
aquatic-terrestrial transition zone and among
The Amazon biogeographical region covers ~7 the most productive systems on the planet.
million km2, 5.79 million km2 of which are low-
land tropical rainforests. We have shown that as Variation in gross primary productivity between
well as climate, soil has a powerful influence on forest sites ranges from 33 to 38 Mg C ha-1 yr-1 for
species richness and composition and on forest more humid forests (in the west and north) to
function. Based primarily on the geological age of lower values of 25 to 30 Mg C ha-1 yr-1 in drier for-
parent material and soil nutrients, the Amazon ests of the Brazilian Shield and central Amazon.
can broadly be divided into six regions (Figure It is also partly driven by soil characteristics. Cli-
4.3). mate nevertheless also impacts the rate of wood
production, and the prevailing baseline climate
The total species richness of the Amazon is still has consequences for forest carbon storage and
actively debated. A well-supported estimate for biodiversity. Both worldwide and in the Amazon,
trees (diameter >10 cm) is 16,000, of which wood production is suppressed in the most ex-
~10,000 have been collected there. Estimates of treme seasonal tropical forest climates with high

25
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

maximum temperatures and high seasonal wa- Allan E, Manning P, Alt F, et al. 2015. Land use intensification
alters ecosystem multifunctionality via loss of biodiversity
ter deficits. This means that some Amazonian
and changes to functional composition. Ecol Lett 18: 834–
forests are already at the climatic limits capable 43.
of sustaining productive tropical forest ecosys- Álvarez Alonso J, Metz MR, and Fine PVA. 2013. Habitat Spe-
tems. Further heating or drying in the Amazon cialization by Birds in Western Amazonian White-sand
Forests. Biotropica 45: 365–72.
risks pushing its trees beyond critical physiolog-
Amaral IL do, Matos FDA, and Lima J. 2000. Composição flo-
ical thresholds. rística e parâmetros estruturais de um hectare de floresta
densa de terra firme no rio Uatumã, Amazônia, Brasil. Acta
4.5. Recommendations Amaz 30: 377.
Amazon Tree Diversity Network. 2021. Amazon Tree Diver-
sity Networkhttp://atdn.myspecies.info. Viewed
● Document ecological networks and their im- Arbeláez F, Duivenvoorden JF, and Maldonado-Ocampo JA.
plications for maintaining these ecosystems in 2008. Geological differentiation explains diversity and
the long-term to understand the truly composition of fish communities in upland streams in the
southern Amazon of Colombia. J Trop Ecol 24: 505–15.
astounding ecological and evolutionary rela-
Arroyo-Kalin M. 2010. The Amazonian formative: crop do-
tionships among species and ecosystems. mestication and anthropogenic soils. Diversity 2: 473–504.
● Conservation initiatives must protect not only Arroyo-Kalin M. 2012. Slash-burn-and-churn: Landscape
forests, but also all the animal and plant spe- history and crop cultivation in pre-Columbian Amazonia.
Quat Int 249: 4–18.
cies within them to guarantee ecological func-
Avissar R and Werth D. 2005. Global hydroclimatological tel-
tioning. Large individual areas of forests, sa- econnections resulting from tropical deforestation. J Hy-
vannas, and aquatic ecosystems need to be drometeorol 6: 134–45.
protected to establish large-scale, landscape- Baker TR, Phillips OL, Malhi Y, et al. 2004. Variation in wood
density determines spatial patterns inAmazonian forest
level conservation initiatives, maintain core
biomass. Glob Chang Biol 10: 545–62.
areas, and provide security for the survival of Bala G, Caldeira K, Wickett M, et al. 2007. Combined climate
wide-ranging species, migratory species, in- and carbon-cycle effects of large-scale deforestation. Proc
numerable less abundant species, species Natl Acad Sci 104: 6550–5.
Bayley PB. 1995. Understanding Large River: Floodplain Eco-
with patchy distributions, and the full range of
systems. Bioscience 45: 153–8.
functional traits. Bayley PB and Petrere Jr M. 1989. Amazon fisheries: assess-
● Maintain connectivity of ecosystems and land- ment methods, current status and management options.
scapes from the Andes to the Amazon region, Can Spec Publ Fish Aquat Sci Spec Can des Sci halieutiques Aquat
1989.
as well as the interaction between terrestrial
Beck SG, Moraes R M, Davis SD, et al. 1997. Llanos de Mojos
and aquatic environments. This is vital to en- Region, Bolivia. Centers plant Divers a Guid Strateg their Con-
sure that Amazonian diversity and processes serv 3: 421–5.
can be sustained. Beja P, Santos CD, Santana J, et al. 2010. Seasonal patterns of
spatial variation in understory bird assemblages across a
● Ensure large, connected areas spanning cli-
mosaic of flooded and unflooded Amazonian forests. Bio-
mate gradients are protected. Connecting Am- divers Conserv 19: 129–52.
azonian and Andean forests is especially es- Benedito-Cecilio E, Araujo-lima CARM, Forsberg BR, et al.
sential for ensuring that forest functions are 2000. Carbon sources of Amazonian fisheries. Fish Manag
Ecol 7: 305–15.
maintained in a changing climate, and that
Bentos TV, Nascimento HEM, Anjos Vizcarra M dos, and Wil-
Amazonian species have some prospect of be- liamson GB. 2017. Effects of lightgaps and topography on
ing able to track climate change in space. Amazon secondary forest: Changes in species richness
and community composition. For Ecol Manage 396: 124–31.
Berner RA and Rao J-L. 1994. Phosphorus in sediments of the
5. References
Amazon River and estuary: Implications for the global flux
of phosphorus to the sea. Geochim Cosmochim Acta 58:
Ab’Saber AN. 1988. O Pantanal Mato-Grossense e a teoria dos 2333–9.
refúgios e redutos. Rev Bras Geogr 50: 57. Bodmer R, Mayor P, Antunez M, et al. 2018. Major shifts in Am-
Adeney JM, Christensen NL, Vicentini A, and Cohn-Haft M. azon wildlife populations from recent intensification of
2016. White-sand Ecosystems in Amazonia. Biotropica 48: floods and drought. Conserv Biol 32: 333–44.
7–23.

26
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Bogotá-Gregory JD, Lima FCT, Correa SB, et al. 2020. Biogeo- Constantine JA, Dunne T, Ahmed J, et al. 2014. Sediment sup-
chemical water type influences community composition, ply as a driver of river meandering and floodplain evolu-
species richness, and biomass in megadiverse Amazonian tion in the Amazon Basin. Nat Geosci 7: 899–903.
fish assemblages. Sci Rep 10: 1–15. Coronado EHN, Hastie A, Reyna J, et al. 2021. Intensive field
Brando PM, Paolucci L, Ummenhofer CC, et al. 2019. sampling increases the known extent of carbon-rich Am-
Droughts, Wildfires, and Forest Carbon Cycling: A Pan- azonian peatland pole forests. Environ Res Lett 16: 74048.
tropical Synthesis. Annu Rev Earth Planet Sci 47: 555–81. Correa SB, Araujo JK, Penha JMF, et al. 2015. Overfishing dis-
Brienen RJW, Phillips OL, Feldpausch TR, et al. 2015. Long- rupts an ancient mutualism between frugivorous fishes
term decline of the Amazon carbon sink. Nature 519: 344– and plants in Neotropical wetlands. Biol Conserv 191: 159–
8. 67.
Bush MB, Hanselman JA, and Hooghiemstra H. 2007. Andean Correa SB, Costa-Pereira R, Fleming T, et al. 2015. Neotropical
montane forests and climate change. In: Tropical rainfor- fish-fruit interactions: eco-evolutionary dynamics and
est responses to climatic change. Springer. conservation. Biol Rev 90: 1263–78.
Camargo PB, Salomão R de P, Trumbore S, and Martinelli LA. Correa SB, Oliveira PC de, Cunha C da, et al. 2018. Water and
1994. How old are large Brazil-nut trees (Bertholletia ex- fish select for fleshy fruits in tropical wetland forests. Bio-
celsa) in the Amazon? Sci Agric 51: 389–91. tropica 50: 312–8.
Cardinale BJ, Duffy JE, Gonzalez A, et al. 2012. Biodiversity Correa SB and Winemiller K. 2018. Terrestrial–aquatic
loss and its impact on humanity. Nature 486: 59–67. trophic linkages support fish production in a tropical oli-
Cardoso D, Särkinen T, Alexander S, et al. 2017. Amazon plant gotrophic river. Oecologia 186: 1069–78.
diversity revealed by a taxonomically verified species list. Correa SB and Winemiller KO. 2014. Niche partitioning
Proc Natl Acad Sci 114: 10695–700. among frugivorous fishes in response to fluctuating re-
Castello L, Bayley PB, Fabré NN, and Batista VS. 2019. Flood- sources in the Amazonian floodplain forest. Ecology 95:
ing effects on abundance of an exploited, long-lived fish 210–24.
population in river-floodplains of the Amazon. Rev Fish Biol Crowther TW, Glick HB, Covey KR, et al. 2015. Mapping tree
Fish 29: 487–500. density at a global scale. Nature 525: 201–5.
Castello L, Isaac VJ, and Thapa R. 2015. Flood pulse effects on Daily GC. 1997. Nature’s services: societal dependence on
multispecies fishery yields in the Lower Amazon. R Soc natural ecosystems. In: Yale University Press.
open Sci 2: 150299. D’Almeida C, Vörösmarty CJ, Hurtt GC, et al. 2007. The effects
Castello L and Macedo MN. 2016. Large-scale degradation of of deforestation on the hydrological cycle in Amazonia: a
Amazonian freshwater ecosystems. Glob Chang Biol 22: review on scale and resolution. Int J Climatol 27: 633–47.
990–1007. Souza FC de, Dexter KG, Phillips OL, et al. 2019. Evolutionary
Castello L, McGrath DG, Hess LL, et al. 2012. The vulnerability diversity is associated with wood productivity in Amazo-
of Amazon freshwater ecosystems. Conserv Lett 6: 217–29. nian forests. Nat Ecol Evol 3: 1754–61.
Chambers JQ, Higuchi N, and Schimel JP. 1998. Ancient trees Aguila-Pasquel J del, Doughty CE, Metcalfe DB, et al. 2014. The
in Amazonia. Nature 391: 135–6. seasonal cycle of productivity, metabolism and carbon dy-
Chase EM and Sayles FL. 1980. Phosphorus in suspended sed- namics in a wet aseasonal forest in north-west Amazonia
iments of the Amazon River. Estuar Coast Mar Sci 11: 383– (Iquitos, Peru). Plant Ecol \& Divers 7: 71–83.
91. Dirzo R and Raven PH. 2003. Global state of biodiversity and
Chave J, Navarrete D, Almeida S, et al. 2010. Regional and sea- loss. Annu Rev Environ Resour 28: 137–67.
sonal patterns of litterfall in tropical South America. Bioge- Dirzo R, Young HS, Galetti M, et al. 2014. Defaunation in the
osciences 7: 43–55. Anthropocene. Science 345: 401–6.
Churchill SP, Balslev Forero HE, and Luteyn JL (Eds). 1995. Doughty CE, Roman J, Faurby S, et al. 2016. Global nutrient
Biodiversity and Conservation of Neotropical Montane transport in a world of giants. Proc Natl Acad Sci 113: 868–
Forests. New York: New York Botanical Garden. 73.
Clement C, Cristo-Araújo M De, Coppens D’Eeckenbrugge G, Draper FC, Costa FRC, Arellano G, et al. 2021. Amazon tree
et al. 2010. Origin and Domestication of Native Amazonian dominance across forest strata. Nat Ecol Evol 5: 757–67.
Crops. Diversity 2: 72–106. Draper FC, Roucoux KH, Lawson IT, et al. 2014. The distribu-
Coe MT, Costa MH, and Howard EA. 2008. Simulating the sur- tion and amount of carbon in the largest peatland complex
face waters of the Amazon River basin: impacts of new in Amazonia. Environ Res Lett 9: 124017.
river geomorphic and flow parameterizations. Hydrol Pro- Dustan P. 2009. Terrestrial limitation of Amazon River
cess An Int J 22: 2542–53. productivity: Why the Amazon River is not green. Evol Ecol
Comer PJ, Hak JC, Josse C, and Smyth R. 2020. Long-term loss Res 11: 421–32.
in extent and current protection of terrestrial ecosystem Eden MJ, Bray W, Herrera L, and McEwan C. 1984. Terra preta
diversity in the temperate and tropical Americas (SP Al- soils and their archaeological context in the Caquetá basin
drich, Ed). PLoS One 15: e0234960. of southeast Colombia. Am Antiq: 125–40.
Condit R, Hubbell SP, and Foster RB. 1995. Mortality rates of Encalada AC, Flecker AS, Poff NL, et al. 2019. A global perspec-
205 neotropical tree and shrub species and the impact of tive on tropical montane rivers. Science 365: 1124–9.
a severe drought. Ecol Monogr 65: 419–39. Endo W, Peres CA, and Haugaasen T. 2016. Flood pulse

27
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

dynamics affects exploitation of both aquatic and terres- 1–6.

trial prey by Amazonian floodplain settlements. Biol Con- Gatti L V., Basso LS, Miller JB, et al. 2021. Amazonia as a car-
serv 201: 129–36. bon source linked to deforestation and climate change.
Espinoza Villar JC, Ronchail J, Guyot JL, et al. 2009. Spatio- Nature 595: 388–93.
temporal rainfall variability in the Amazon basin coun- Gautier E, Brunstein D, Vauchel P, et al. 2007. Temporal rela-
tries (Brazil, Peru, Bolivia, Colombia, and Ecuador). Int J tions between meander deformation, water discharge and
Climatol 29: 1574–94. sediment fluxes in the floodplain of the Rio Beni (Bolivian
Esquivel‐Muelbert A, Baker TR, Dexter KG, et al. 2019. Com- Amazonia). Earth Surf Process Landforms 32: 230–48.
positional response of Amazon forests to climate change. Gentry AH. 1988. Tree species richness of upper Amazonian
Glob Chang Biol 25: 39–56. forests. Proc Natl Acad Sci 85: 156–9.
Esquivel-Muelbert A, Baker TR, Dexter KG, et al. 2016. Sea- Gentry AH and Emmons LH. 1987. Geographical Variation in
sonal drought limits tree species across the Neotropics. Fertility, Phenology, and Composition of the Understory of
Ecography (Cop) 40: 618–29. Neotropical Forests. Biotropica 19: 216–27.
Esquivel-Muelbert A, Phillips OL, Brienen RJW, et al. 2020. Goulding M. 1993. Flooded forests of the Amazon. Sci Am 268:
Tree mode of death and mortality risk factors across Am- 114–20.
azon forests. Nat Commun 11: 5515. Goulding M. 1980. The fishes and the forest: explorations in
Faber-Langendoen D, Keeler-Wolf T, Meidinger D, et al. 2016. Amazonian natural history. Univ of California Press.
Classification and description of world formation types. Goulding M, Barthem R, and Ferreira EJG. 2003. The Smith-
FAO. 2007. Mangroves of South America 1980–2005: country sonian atlas of the Amazon. Smithsonian books, Washing-
reports. Rome: Food and Agriculture Organization of the ton, DC, USA.
United Nations. Goulding M, Smith NJH, and Mahar DJ. 1996. Floods of for-
Feldpausch TR, Banin L, Phillips OL, et al. 2011. Height-diam- tune: ecology and economy along the Amazon. New Yourk,
eter allometry of tropical forest trees. Biogeosciences 8: Columbia University Press.
1081–106. Goulding M, Venticinque E, Ribeiro ML de B, et al. 2019. Eco-
Ferreira CS, Piedade MTF, Wittmann A de O, and Franco AC. system-based management of Amazon fisheries and wet-
2010. Plant reproduction in the Central Amazonian flood- lands. Fish Fish 20: 138–58.
plains: challenges and adaptations. AoB Plants 2010. Gumbricht T, Roman‐Cuesta RM, Verchot L, et al. 2017. An ex-
Ferreira L V and Prance GT. 1998. Structure and species rich- pert system model for mapping tropical wetlands and
ness of low-diversity floodplain forest on the Rio Tapajós, peatlands reveals South America as the largest contribu-
Eastern Amazonia, Brazil. Biodivers \& Conserv 7: 585–96. tor. Glob Chang Biol 23: 3581–99.
Ferreira LV. 1997. Effects of the duration of flooding on spe- Hansen MC, Potapov P V, Moore R, et al. 2013. High-Resolu-
cies richness and floristic composition in three hectares tion Global Maps of 21st-Century Forest Cover Change. Sci-
in the Jaú National Park in floodplain forests in central ence 342: 850–3.
Amazonia. Biodivers Conserv 6: 1353–63. Harley CDG. 2003. Abiotic stress and herbivory interact to set
Field CB, Lobell DB, Peters HA, and Chiariello NR. 2007. Feed- range limits across a two-dimensional stress gradient.
backs of terrestrial ecosystems to climate change. Annu Ecology 84: 1477–88.
Rev Environ Resour 32: 1–29. Hättenschwiler S, Coq S, Barantal S, and Handa IT. 2011. Leaf
Figueiredo FOG, Zuquim G, Tuomisto H, et al. 2017. Beyond traits and decomposition in tropical rainforests: revisiting
climate control on species range: The importance of soil some commonly held views and towards a new hypothe-
data to predict distribution of Amazonian plant species. J sis. New Phytol 189: 950–65.
Biogeogr 45: 190–200. Haugaasen T and Peres CA. 2005a. Mammal assemblage
Filizola N and Guyot JL. 2009. Suspended sediment yields in structure in Amazonian flooded and unflooded forests. J
the Amazon basin: an assessment using the Brazilian na- Trop Ecol: 133–45.
tional data set. Hydrol Process 23: 3207–15. Haugaasen T and Peres CA. 2007. Vertebrate responses to
Fine PVA, Miller ZJ, Mesones I, et al. 2006. The growth--de- fruit production in Amazonian flooded and unflooded for-
fense trade-off and habitat specialization by plants in Am- ests. Biodivers Conserv 16: 4165–90.
azonian forests. Ecology 87: S150--S162. Haugaasen T and Peres CA. 2005b. Primate assemblage
Foley JA, Asner GP, Costa MH, et al. 2007. Amazonia revealed: structure in amazonian flooded and unflooded forests. Am
Forest degradation and loss of ecosystem goods and ser- J Primatol 67: 243–58.
vices in the Amazon Basin. Front Ecol Environ 5: 25–32. Haugaasen T and Peres CA. 2006. Floristic, edaphic and
Forsberg BR, Melack JM, Dunne T, et al. 2017. The potential structural characteristics of flooded and unflooded forests
impact of new Andean dams on Amazon fluvial ecosys- in the lower Rio Purús region of central Amazonia, Brazil.
tems. PLoS One 12: e0182254. Acta Amaz 36: 25–35.
Forsberg BR, Araújo-Lima CARM, Martinelli LA, et al. 1993. Hawes JE and Peres CA. 2016. Patterns of plant phenology in
Autotrophic carbon sources for fish of the central Ama- Amazonian seasonally flooded and unflooded forests. Bio-
zon. Ecology 74: 643–52. tropica 48: 465–75.
Galetti M and Dirzo R. 2013. Ecological and evolutionary con- Hess LL, Melack JM, Affonso AG, et al. 2015. Wetlands of the
sequences of living in a defaunated world. Biol Conserv 163: Lowland Amazon Basin: Extent, Vegetative Cover, and

28
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Dual-season Inundated Area as Mapped with JERS-1 Syn- Amazon floodplain near Manaus: Species diversity and
thetic Aperture Radar. Wetlands 35: 745–56. adaptations to the flood pulse. Amaz Limnol Oecologia Reg
Hoch G and Körner C. 2005. Growth, demography and carbon Syst Fluminis Amaz 12: 467–84.
relations of Polylepis trees at the world’s highest treeline. Junk WJ, Piedade MTF, Schöngart J, et al. 2011. A classifica-
Funct Ecol 19: 941–51. tion of major naturally-occurring Amazonian lowland
Hofhansl F, Chacón-Madrigal E, Fuchslueger L, et al. 2020. wetlands. Wetlands 31: 623–40.
Climatic and edaphic controls over tropical forest diver- Junk WJ and Wantzen KM. 2003. The flood pulse concept:
sity and vegetation carbon storage. Sci Rep 10: 1–11. new aspects, approaches and applications - an update. In
Honorio Coronado EN, Baker TR, Phillips OL, et al. 2009. R. L. Welcomme, & T. Petr (Eds.), Proceedings of the Sec-
Multi-scale comparisons of tree composition in Amazo- ond International Symposium on the Management of
nian terra firme forests. Biogeosciences 6: 2719–31. Large Rivers for Fisheries (pp. 117-149). Bangkok: Food
Hooper DU, Adair EC, Cardinale BJ, et al. 2012. A global syn- and Agriculture Organization and Mekong River Commis-
thesis reveals biodiversity loss as a major driver of ecosys- sion, FAO Regional Office for Asia and the Pacific.
tem change. Nature 486: 105–8. Junk WJ, Wittmann F, Schöngart J, and Piedade MTF. 2015. A
Householder JE, Janovec JP, Tobler MW, et al. 2012. Peatlands classification of the major habitats of Amazonian black-
of the Madre de Dios River of Peru: Distribution, Geomor- water river floodplains and a comparison with their white-
phology, and Habitat Diversity. Wetlands 32: 359–68. water counterparts. Wetl Ecol Manag 23: 677–93.
Hubau W, Lewis SL, Phillips OL, et al. 2020. Asynchronous Kalliola R, Puhakka M, and Danjoy W. 1993. Amazonia peru-
carbon sink saturation in African and Amazonian tropical ana: vegetación húmeda tropical en el llano subandino.
forests. Nature 579: 80–7. Kaspari M, Yanoviak SP, Dudley R, et al. 2009. Sodium short-
Humboldt A Von and Bonpland A. 1805. Essai sur la age as a constraint on the carbon cycle in an inland tropi-
géographie des plantes: Accompagné d’un tableau phy- cal rainforest. Proc Natl Acad Sci 106: 19405–9.
sique des régions équinoxiales. Paris: Chez Levrault, Keddy PA, Fraser LH, Solomeshch AI, et al. 2009. Wet and
Schoell et Compagnie. Wonderful: The World’s Largest Wetlands Are Conserva-
IBGE. 1992. Manual técnico da vegetação Brasileira. Rio de tion Priorities. Bioscience 59: 39–51.
Janeiro: Instituto Brasileiro de Geografia e Estatística. Kricher JC. 1997. A neotropical companion: an introduction
ICMBio. 2018. Atlas dos Manguezais do Brasil. Instituto Chico to the animals, plants, and ecosystems of the New World
Mendes de Conservação da Biodiversidade. – Brasília: In- tropics. Princeton University Press.
stituto Chico Mendes de Conservação da Biodiversidade. Kubitzki K and Ziburski A. 1994. Seed dispersal in flood plain
Isaac VJ, Castello L, Santos PRB, and Ruffino ML. 2016. Sea- forests of Amazonia. Biotropica: 30–43.
sonal and interannual dynamics of river-floodplain multi- Kurten EL. 2013. Cascading effects of contemporaneous de-
species fisheries in relation to flood pulses in the Lower faunation on tropical forest communities. Biol Conserv 163:
Amazon. Fish Res 183: 352–9. 22–32.
Jézéquel C, Tedesco PA, Bigorne R, et al. 2020. A database of Lähteenoja O and Page S. 2011. High diversity of tropical
freshwater fish species of the Amazon Basin. Sci data 7: 1– peatland ecosystem types in the Pastaza-Marañón basin,
9. Peruvian Amazonia. J Geophys Res Biogeosciences 116.
Josse C, Cuesta F, Navarro G, et al. 2009. Ecosistemas de los Lähteenoja O, Ruokolainen K, Schulman L, and Oinonen M.
Andes del norte y centro. Bolivia, Colombia, Ecuador, Perú 2009. Amazonian peatlands: an ignored C sink and poten-
y Venezuela. tial source. Glob Chang Biol 15: 2311–20.
Junk WJ, Piedade MTF, Lourival R, et al. 2014. Brazilian wet- Lasso CA, Agudelo-Córdoba E, Jiménez-Segura LF, et al. 2011.
lands: their definition, delineation, and classification for Catálogo de los recursos pesqueros continentales de Co-
research, sustainable management, and protection in. lombia: memoria técnica y explicativa, resumen ejecutivo.
Aquat Conserv Mar Freshw Ecosyst 24: 5–22. Instituto de Investigación de Recursos Biológicos Alexan-
Junk WJ. 1970. Investigations on the ecology and production- der von Humboldt.
biology of the" floating meadows"(Paspalo-Echinochloe- Latrubesse E. 2012. Lakes and Reservoirs. In: Amazon lakes.
tum) on the middle Amazon. Part 1: The floating vegeta- Springer Verlag.
tion and its ecology. Amaz Limnol Oecologia Reg Syst Fluminis Laurance WF. 1998. A crisis in the making: responses of Am-
Amaz 2: 449–95. azonian forests to land use and climate change. Trends Ecol
Junk WJ. 2013. The central Amazon floodplain: ecology of a \& Evol 13: 411–5.
pulsing system. Springer Science \& Business Media. Laurance WF, Camargo JLC, Fearnside PM, et al. 2018. An Am-
Junk WJ. 1984. Ecology of the várzea, floodplain of Amazo- azonian rainforest and its fragments as a laboratory of
nian whitewater rivers. In: The Amazon. Springer. global change. Biol Rev 93: 223–47.
Junk WJ and Piedade MTF. 2010. An Introduction to South Laurance WF and Williamson GB. 2001. Positive Feedbacks
American Wetland Forests: Distribution, Definitions and among Forest Fragmentation, Drought, and Climate
General Characterization. In: Junk W, Piedade M, Witt- Change in the Amazon. Conserv Biol 15: 1529–35.
mann F, et al. (Eds). Amazonian Floodplain Forests. Eco- Laurance WF, Lovejoy TE, Vasconcelos HL, et al. 2002. Eco-
logical Studies (Analysis and Synthesis). Springer,. system decay of Amazonian forest fragments: a 22-year
Junk WJ and Piedade MTF. 1993. Herbaceous plants of the investigation. Conserv Biol 16: 605–18.

29
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Laurance WF, Nascimento HEM, Laurance SG, et al. 2004. In- Distribution and Extent of Wetlands in the Amazon Basin.
ferred longevity of Amazonian rainforest trees based on a In: Junk W, Piedade M, Wittmann F, et al. (Eds). Amazonian
long-term demographic study. For Ecol Manage 190: 131– Floodplain Forests. Ecological Studies (Analysis and Syn-
43. thesis). Springer.
Lehmann J, Kern DC, Glaser B, and Woods WI. 2007. Amazo- Melack JM and Forsberg BR. 2001. Biogeochemistry of Ama-
nian dark earths: origin properties management. zon floodplain. Biogeochem Amaz Basin; Oxford Univ Press
Lima ÁC and Araujo-Lima CARM. 2004. The distributions of New York, NY, USA: 235.
larval and juvenile fishes in Amazonian rivers of different Montero JC, Piedade MTF, and Wittmann F. 2014. Floristic
nutrient status. Freshw Biol 49: 787–800. variation across 600 km of inundation forests (Igapó)
Lindeman JC and Moolenaar SP. 1959. Preliminary survey of along the Negro River, Central Amazonia. Hydrobiologia
the vegetation types of northern Suriname. Meded van het 729: 229–46.
Bot Museum en Herb van Rijksuniv te Utr 159: 1–45. Moraes R M, Hurtado R, and Mejía K. Mauritia flexuosa un sím-
Lloyd J, Domingues TF, Schrodt F, et al. 2015. Edaphic, struc- bolo de las palmas útiles sudamericanas. Palmeras y usos
tural and physiological contrasts across Amazon Basin Especies Boliv y la región: 71.
forest–savanna ecotones suggest a role for potassium as a Moreira-Turcq P, Seyler P, Guyot JL, and Etcheber H. 2003.
key modulator of tropical woody vegetation structure and Exportation of organic carbon from the Amazon River and
function. Biogeosciences 12: 6529–71. its main tributaries. Hydrol Process 17: 1329–44.
Lombardo U, Iriarte J, Hilbert L, et al. 2020. Early Holocene Morris RJ. 2010. Anthropogenic impacts on tropical forest bi-
crop cultivation and landscape modification in Amazonia. odiversity: a network structure and ecosystem function-
Nature 581: 190–3. ing perspective. Philos Trans R Soc B Biol Sci 365: 3709–18.
Lu D, Moran E, and Mausel P. 2002. Linking Amazonian sec- Mostacedo B, Balcazar J, and Montero JC. 2006. Tipos de
ondary succession forest growth to soil properties. L bosque, diversidad y composición flor{\’\i}stica en la Ama-
Degrad \& Dev 13: 331–43. zonia sudoeste de Bolivia. Ecol en Boliv 41: 99–116.
Luther DA, Cooper WJ, Wolfe JD, et al. 2020. Tropical forest Naeem S. 1998. Species redundancy and ecosystem reliabil-
fragmentation and isolation: Is community decay a ran- ity. Conserv Biol 12: 39–45.
dom process? Glob Ecol Conserv 23: e01168. Nepstad DC, Carvalho CR de, Davidson EA, et al. 1994. The role
Macía MJ and Svenning J-C. 2005. Oligarchic dominance in of deep roots in the hydrological and carbon cycles of Am-
western Amazonian plant communities. J Trop Ecol 21: azonian forests and pastures. Nature 372: 666–9.
613–26. Nepstad DC, Stickler CM, Filho BS-, and Merry F. 2008. Inter-
Madriñán S, Cortés AJ, and Richardson JE. 2013. Páramo is actions among Amazon land use, forests and climate: pro-
the world’s fastest evolving and coolest biodiversity spects for a near-term forest tipping point. Philos Trans R
hotspot. Front Genet 4: 192. Soc B Biol Sci 363: 1737–46.
Malhi Y, Doughty CE, Goldsmith GR, et al. 2015. The linkages Nunes da Cunha C and Junk WJ. 2004. Year-to-year changes
between photosynthesis, productivity, growth and bio- in water level drive the invasion of Vochysia divergens in
mass in lowland Amazonian forests. Glob Chang Biol 21: Pantanal grasslands. Appl Veg Sci 7: 103–10.
2283–95. Oliveira ACB, Martinelli LA, Moreira MZ, et al. 2006. Seasonal-
Malhi Y, Girardin C, Metcalfe DB, et al. 2021. The Global Eco- ity of energy sources of Colossoma macropomum in a
systems Monitoring network: Monitoring ecosystem floodplain lake in the Amazon - lake Camaleao, Amazonas,
productivity and carbon cycling across the tropics. Biol Brazil. Fish Manag Ecol 13: 135–42.
Conserv 253: 108889. Oliveira AA and Mori SA. 1999. A central Amazonian terra
Marengo JA, Nobre CA, Sampaio G, et al. 2011. Climate change firme forest. I. High tree species richness on poor soils. Bi-
in the Amazon Basin: Tipping points, changes in ex- odivers Conserv 8: 1219–44.
tremes, and impacts on natural and human systems. In: Parolin P, Simone O De, Haase K, et al. 2004. Central Amazo-
Tropical rainforest responses to climatic change. nian floodplain forests: tree adaptations in a pulsing sys-
Springer. tem. Bot Rev 70: 357–80.
Marengo JA and Espinoza JC. 2016. Extreme seasonal Parrens M, Bitar A Al, Frappart F, et al. 2019. High resolution
droughts and floods in Amazonia: causes, trends and im- mapping of inundation area in the Amazon basin from a
pacts. Int J Climatol 36: 1033–50. combination of L-band passive microwave, optical and ra-
Marimon BS, Marimon-Junior BH, Feldpausch TR, et al. 2014. dar datasets. Int J Appl Earth Obs Geoinf 81: 58–71.
Disequilibrium and hyperdynamic tree turnover at the Patiño S, Lloyd J, Paiva R, et al. 2009. Branch xylem density
forest–cerrado transition zone in southern Amazonia. variations across the Amazon Basin. Biogeosciences 6: 545–
Plant Ecol Divers 7: 281–92. 68.
McClain ME and Naiman RJ. 2008. Andean influences on the Patton JL, Silva MNF Da, and Malcolm JR. 2000. Mammals of
biogeochemistry and ecology of the Amazon River. Biosci- the Rio Juruá and the evolutionary and ecological diversi-
ence 58: 325–38. fication of Amazonia. Bull Am Museum Nat Hist 2000: 1–306.
Meirelles JM. 2006. O livro de ouro da Amazônia. Rio de Pereira MJR, Marques JT, Santana J, et al. 2009. Structuring of
Janeiro: Ediouro. Amazonian bat assemblages: the roles of flooding patterns
Melack JM and Hess LL. 2010. Remote Sensing of the and floodwater nutrient load. J Anim Ecol 78: 1163–71.

30
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Phillips OL. 1997. The changing ecology of tropical forests. Bi- environments? Biotropica 50: 338–45.
odivers Conserv 6: 291–311. Räsänen ME, Salo JS, and Kalliola RJ. 1987. Fluvial perturb-
Phillips OL, Baker TR, Arroyo L, et al. 2004. Pattern and pro- ance in the western Amazon basin: regulation by long-
cess in Amazon tree turnover, 1976–2001 (Y Malhi and OL term sub-Andean tectonics. Science 238: 1398–401.
Phillips, Eds). Philos Trans R Soc London Ser B Biol Sci 359: Reid W V, Mooney HA, Cropper A, et al. 2005. Ecosystems and
381–407. human well-being-Synthesis: A report of the Millennium
Phillips OL, Sullivan MJP, Baker TR, et al. 2019. Species Mat- Ecosystem Assessment. Island Press.
ter: Wood Density Influences Tropical Forest Biomass at Resende AF, Schöngart J, Streher AS, et al. 2019. Massive tree
Multiple Scales. Surv Geophys 40: 913–35. mortality from flood pulse disturbances in Amazonian
Phillips O, Miller JS, and Miller JS. 2002. Global patterns of floodplain forests: The collateral effects of hydropower
plant diversity: Alwyn H. Gentry’s forest transect data set. production. Sci Total Environ 659: 587–98.
Missouri Botanical Press. Ríos-Villamizar EA, Piedade MTF, Costa JG Da, et al. 2013.
Piedade MTF, Junk WJ, and Long SP. 1991. The productivity Chemistry of different Amazonian water types for river
of the C4 grass Echinochloa polystachya on the Amazon classification: a preliminary review. In: Brebbia C (Ed).
floodplain. Ecology 72: 1456–63. Water and Society II. Southampton: WIT Press.
Piedade MTF, Junk W, D’Ângelo SA, et al. 2010. Aquatic her- Rocha M, Assis RL de, Piedade MTF, et al. 2019. Thirty years
baceous plants of the Amazon floodplains: state of the art after Balbina Dam: Diversity and floristic composition of
and research needed. Acta Limnol Bras 22: 165–78. the downstream floodplain forest, Central Amazon, Brazil.
Pires J and Prance G. 1985. The vegetation types of the Brazil- Ecohydrology 12: e2144.
ian Amazon. Rödig E, Cuntz M, Rammig A, et al. 2018. The importance of
Pires THS, Borghezan EA, Machado VN, et al. 2018. Testing forest structure for carbon fluxes of the Amazon rainfor-
Wallace’s intuition: water type, reproductive isolation and est. Environ Res Lett 13: 054013.
divergence in an Amazonian fish. J Evol Biol 31: 882–92. Rosa SA, Barbosa ACMC, Junk WJ, et al. 2017. Growth models
Pitman NCA, Terborgh JW, Silman MR, et al. 2001. Dominance based on tree-ring data for the Neotropical tree species
and Distribution of Tree Species in Upper Amazonian Calophyllum brasiliense across different Brazilian wet-
Terra Firme Forests. Ecology 82: 2101. lands: implications for conservation and management.
Pouilly M, Beck S, Moraes R M, and Ibañez C. 2004. Diversidad Trees - Struct Funct 31: 729–42.
biológica en la llanura de inundación del Río Mamoré. Im- Rozendaal DMA, Bongers F, Aide TM, et al. 2019. Biodiversity
portancia ecológica de la dinámica fluvial. Bolivia: Centro recovery of Neotropical secondary forests. Sci Adv 5.
de Ecologia Simon I. Patiño. Salinas N, Malhi Y, Meir P, et al. 2011. The sensitivity of tropi-
Prada CM, Morris A, Andersen KM, et al. 2017. Soils and rain- cal leaf litter decomposition to temperature: results from
fall drive landscape-scale changes in the diversity and a large-scale leaf translocation experiment along an eleva-
functional composition of tree communities in premon- tion gradient in Peruvian forests. New Phytol 189: 967–77.
tane tropical forest. J Veg Sci 28: 859–70. Salo J, Kalliola R, Häkkinen I, et al. 1986. River dynamics and
Prance G. 1979. Notes on the vegetation of amazonia III. The the diversity of Amazon lowland forest. Nature 322: 254–8.
terminology of amazonian forest types subject to inunda- Sanchez-Vega I and Dillon MO. 2006. Jalcas- PARTE II Eco-
tion. Brittonia 31: 26–38. sistemas productivos en los Andes centrales. In: Moraes R
Quesada CA, Lloyd J, Anderson LO, et al. 2011. Soils of Amazo- M, Øllgaard B, Kvist L, et al. (Eds). Botánica Económica de
nia with particular reference to the RAINFOR sites. Biogeo- los Andes Centrales. Herbario Nacional de Bolivia, Univer-
sciences 8: 1415–40. sidad Mayor de San Andrés, Plural Editores, La Paz.
Quesada CACA, Phillips OLOL, Schwarz M, et al. 2012. Basin- Santos JA, Soriano L, Barroco A, et al. 2017. Stable isotopes in
wide variations in Amazon forest structure and function ecological studies with fish in the Brazilian Amazon
are mediated by both soils and climate. Biogeosciences 9: Isótopos estáveis em estudos ecológicos com peixes na
2203–46. Amazônia Brasileira.
Quesada CA and Lloyd J. 2016. Soil–Vegetation Interactions Schargel R. 2011. Una reseña de la geografía física de Vene-
in Amazonia. In: Nagy L, Artaxo P, Forsberg BR (Eds). In- zuela, con énfasis en los suelos. BioLlania Edic Esp 10: 11–
teractions Between Biosphere, Atmosphere and Human 26.
Land Use in the Amazon Basin. Springer. Schargel R and Marvez P. 2001. Características de los suelos
Quesada CA, Paz C, Oblitas Mendoza E, et al. 2020. Variations alrededor de san carlos de río negro, estado Amazonas,
in soil chemical and physical properties explain basin- Venezuela. BioLlania Edic Esp 7: 234–64.
wide Amazon forest soil carbon concentrations. SOIL 6: Schöngart J. 2008. Growth-Oriented Logging (GOL): A new
53–88. concept towards sustainable forest management in Cen-
Quintana C, Pennington RT, Ulloa CU, and Balslev H. 2017. Bi- tral Amazonian vá rzea floodplains. For Ecol Manage 256:
ogeographic Barriers in the Andes: Is the Amotape— 46–58.
Huancabamba Zone a Dispersal Barrier for Dry Forest Schöngart J, Gribel R, Ferreira da Fonseca-Junior S, and
Plants? Ann Missouri Bot Gard 102: 542–50. Haugaasen T. 2015. Age and Growth Patterns of Brazil Nut
Ramalho WP, Machado IF, and Vieira LJS. 2018. Do flood Trees ( Bertholletia excelsa Bonpl.) in Amazonia, Brazil. Bi-
pulses structure amphibian communities in floodplain otropica 47: 550–8.

31
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Schöngart J and Junk WJ. 2007. Forecasting the flood-pulse Ter Steege H, Mota de Oliveira S, Pitman NCA, et al. 2019b. To-
in Central Amazonia by ENSO-indices. J Hydrol 335: 124– wards a dynamic list of Amazonian tree species. Sci Rep 9:
32. 3501.
Schöngart J, Junk WJ, Piedade MTF, et al. 2004. Teleconnec- Ter Steege H, Pitman NCA, Sabatier D, et al. 2013. Hyperdom-
tion between tree growth in the Amazonian floodplains inance in the Amazonian tree flora. Science 342.
and the El Niño-Southern Oscillation effect. Glob Chang Biol Ter Steege H, Pitman NCA, Killeen TJ, et al. 2015. Estimating
10: 683–92. the global conservation status of more than 15,000 Ama-
Schwendel AC, Nicholas AP, Aalto RE, et al. 2015. Interaction zonian tree species. Sci Adv 1: e1500936.
between meander dynamics and floodplain heterogeneity Ter Steege H, Pitman NCA, Phillips OL, et al. 2006. Continen-
in a large tropical sand-bed river: the Rio Beni, Bolivian tal-scale patterns of canopy tree composition and function
Amazon. Earth Surf Process Landforms 40: 2026–40. across Amazonia. Nature 443: 444–7.
Siddiqui SF, Zapata‐Rios X, Torres‐Paguay S, et al. 2021. Clas- Ter Steege H, Pitman N, Sabatier D, et al. 2003. A spatial model
sifying flow regimes of the Amazon basin. Aquat Conserv of tree α diversity and tree density for the Amazon. Bio-
Mar Freshw Ecosyst 31: 1005–28. divers \& Conserv 12: 2255–77.
Silva CV de J, Santos JR dos, Galvao LS, et al. 2016. Floristic Ter Steege H, Prado PI, Lima RAF de, et al. 2020. Biased-cor-
and structure of an Amazonian primary forest and a rected richness estimates for the Amazonian tree flora. Sci
chronosequence of secondary succession. Acta Amaz 46: Rep 10: 10130.
133–50. Ter-Steege H, Sabatier D, Castellanos H, et al. 2000. An analy-
Sioli H. 1984. The Amazon and its main affluents: Hydrogra- sis of the floristic composition and diversity of Amazonian
phy, morphology of the river courses, and river types. In: forests including those of the Guiana Shield. J Trop Ecol:
Sioli H (Ed). The Amazon. Monographiae Biologicae, vol 801–28.
56. Springer. Thorbjarnarson J and Silveira R Da. 2000. Secrets of the
Sobral M, Silvius KM, Overman H, et al. 2017. Mammal diver- Flooded Forest. Nat Hist 109: 70–9.
sity influences the carbon cycle through trophic interac- Tuomisto H, Ruokolainen K, Kalliola R, et al. 1995. Dissecting
tions in the Amazon. Nat Ecol Evol 1: 1670–6. amazonian biodiversity. Science 269: 63–6.
Sousa TR, Schietti J, Coelho de Souza F, et al. 2020. Palms and Tuomisto H, Doninck J Van, Ruokolainen K, et al. 2019. Dis-
trees resist extreme drought in Amazon forests with shal- covering floristic and geoecological gradients across Ama-
low water tables (C McMichael, Ed). J Ecol 108: 2070–82. zonia. J Biogeogr 46: 1734–48.
Stefanelli-Silva G, Zuanon J, and Pires T. 2019. Revisiting Am- Ulloa Ulloa C and Jørgensen PM. 2018. From Humboldt’s
azonian water types: experimental evidence highlights cross-section of the Chimborazo to the map of the plants
the importance of forest stream hydrochemistry in shap- of the Americas: Making checklists. Taxon 67: 1059–61.
ing adaptation in a fish species. Hydrobiologia 830: 151–60. Vannote RL, Minshall GW, Cummins KW, et al. 1980. The river
Stevenson PR. 2016. Neotropical primate communities: Ef- continuum concept. Can J Fish Aquat Sci 37: 130–7.
fects of disturbance, resource production and forest type Venticinque E, Forsberg B, Barthem R, et al. 2016. An explicit
heterogeneity. Am J Primatol 78: 391–401. GIS-based river basin framework for aquatic ecosystem
Sullivan MJP, Talbot J, Lewis SL, et al. 2017. Diversity and car- conservation in the Amazonhttps://knb.ecoinformat-
bon storage across the tropical forest biome. Sci Rep 7: ics.org/view/doi%3A10.5063%2FF1BG2KX8#snapp_com
39102. puting.6.1. Viewed
Sullivan MJP, Lewis SL, Affum-Baffoe K, et al. 2020. Long- Vetter RE and Botosso PC. 1989. Remarks on Age and Growth
term thermal sensitivity of earth’s tropical forests. Science Rate Determination of Amazonian Trees. IAWA J 10: 133–
368: 869–74. 45.
Swap R, Garstang M, Greco S, et al. 1992. Saharan dust in the Vieira S, Trumbore S, Camargo PB, et al. 2005. Slow growth
Amazon Basin. Tellus B 44: 133–49. rates of Amazonian trees: Consequences for carbon cy-
Tedesco PA, Hugueny B, Oberdorff T, et al. 2008. River hydro- cling. Proc Natl Acad Sci U S A 102: 18502–7.
logical seasonality influences life history strategies of Vormisto J, Svenning J-C, Hall P, and Balslev H. 2004. Diver-
tropical riverine fishes. Oecologia 156: 691–702. sity and dominance in palm (Arecaceae) communities in
Tedesco PA, Beauchard O, Bigorne R, et al. 2017. A global da- terra firme forests in the western Amazon basin. J Ecol 92:
tabase on freshwater fish species occurrence in drainage 577–88.
basins. Sci Data 4: 170141. Waldhoff D, Saint-Paul U, and Furch B. 1996. Value of fruits
Terborgh J, Foster RB, and Percy N V. 1996. Tropical tree and seeds from the floodplain forests of central Amazonia
communities: a test of the nonequilibrium hypothesis. as food resource for fish. Ecotropica 2: 143–56.
Ecology 77: 561–7. Winemiller KO, McIntyre PB, Castello L, et al. 2016. Balancing
Ter Steege H, Pitman NCA, Sabatier D, et al. 2013. Hyperdom- hydropower and biodiversity in the Amazon, Congo, and
inance in the Amazonian Tree Flora. Science 342: Mekong. Science 351: 128–9.
1243092–1243092. Wittmann F, Anhuf D, and Junk WJ. 2002. Tree species distri-
Ter Steege H, Henkel TW, Helal N, et al. 2019a. Rarity of bution and community structure of central Amazonian
monodominance in hyperdiverse Amazonian forests. Sci várzea forests by remote-sensing techniques. J Trop Ecol:
Rep 9: 13822. 805–20.

32
Science Panel for the Amazon
Chapter 4: Biodiversity and Ecological Functioning in the Amazon

Wittmann F and Junk WJ. 2017. Amazon river basin. The

Wetland Book II. In: Finlayson C, Milton G, Prentice R, Da-
vidson N (Eds). Distribution, Description and Conserva-
tion. Dordrecht: Springer Netherlands.
Wittmann F, Schöngart J, and Junk WJ. 2010. Phytogeogra-
phy, Species Diversity, Community Structure and Dynam-
ics of Central Amazonian Floodplain Forests. In: Amazo-
nian Floodplain Forests. Springer, Dordrecht.
Wittmann F, Schöngart J, Montero JC, et al. 2006. Tree species
composition and diversity gradients in white-water for-
ests across the Amazon Basin. J Biogeogr 33: 1334–47.
Wittmann H, Blanckenburg F von, Maurice L, et al. 2011. Sed-
iment production and delivery in the Amazon River basin
quantified by in situ--produced cosmogenic nuclides and
recent river loads. Bulletin 123: 934–50.
Wittmann H, Blanckenburg F von, Maurice L, et al. 2011. Sed-
iment production and delivery in the Amazon River basin
quantified by in situ--produced cosmogenic nuclides and
recent river loads. Bulletin 123: 934–50.
Worbes M. 1997. The forest ecosystem of the floodplains. In:
The central Amazon floodplain. Springer.
Young K, León B, Jørgensen P, and Ulloa UC. 2007. Tropical
and subtropical landscapes of the Andes. In: Veblen T,
Young K, Orme A (Eds). The Physical Geography of South
America. Oxford University Press.
Zemp DC, Schleussner C-F, Barbosa HMJ, et al. 2017. Self-am-
plified Amazon forest loss due to vegetation-atmosphere
feedbacks. Nat Commun 8: 1–10.

33
Science Panel for the Amazon
CONTACT INFORMATION

SPA Technical-Scientific Secretariat New York

475 Riverside Drive, Suite 530
New York NY 10115
USA
+1 (212) 870-3920
[email protected]

SPA Technical-Scientific Secretariat South America

Av. Ironman Victor Garrido, 623
São José dos Campos – São Paulo
Brazil
[email protected]

WEBSITE theamazonwewant.org
INSTAGRAM @theamazonwewant
TWITTER @theamazonwewant