Model-Based Estimation of Amazonian Forests Recovery Time after Drought and Fire Events

Abstract

In recent decades, droughts, deforestation and wildfires have become recurring phenomena that have heavily affected both human activities and natural ecosystems in Amazonia. The time needed for an ecosystem to recover from carbon losses is a crucial metric to evaluate disturbance impacts on forests. However, little is known about the impacts of these disturbances, alone and synergistically, on forest recovery time and the resulting spatiotemporal patterns at the regional scale. In this study, we combined the 3-PG forest growth model, remote sensing and field derived equations, to map the Amazonia-wide (3 km of spatial resolution) impact and recovery time of aboveground biomass (AGB) after drought, fire and a combination of logging and fire. Our results indicate that AGB decreases by 4%, 19% and 46% in forests affected by drought, fire and logging + fire, respectively, with an average AGB recovery time of 27 years for drought, 44 years for burned and 63 years for logged + burned areas and with maximum values reaching 184 years in areas of high fire intensity. Our findings provide two major insights in the spatial and temporal patterns of drought and wildfire in the Amazon: (1) the recovery time of the forests takes longer in the southeastern part of the basin, and, (2) as droughts and wildfires become more frequent-since the intervals between the disturbances are getting shorter than the rate of forest regeneration-the long lasting damage they cause potentially results in a permanent and increasing carbon losses from these fragile ecosystems.

Full text

Article
Model-Based Estimation of Amazonian Forests Recovery Time
after Drought and Fire Events
Bruno L. De Faria 1, 2, * , Gina Marano 3, Camille Piponiot 4, Carlos A. Silva 5, Vinícius de L. Dantas 6,
Ludmila Rattis 7,8, Andre R. Rech 1and Alessio Collalti 9, 10


Citation: De Faria, B.L.; Marano, G.;
Piponiot, C.; Silva, C.A.; Dantas, V.d.L.;
Rattis, L.; Rech, A.R.; Collalti, A. Model-
Based Estimation of Amazonian Forests
Recovery Time after Drought and Fire
Events. Forests 2021,12, 8.
https://dx.doi.org/10.3390/f12010008
Received: 24 September 2020
Accepted: 21 December 2020
Published: 23 December 2020
Publisher’s Note: MDPI stays neu-
tral with regard to jurisdictional claims
in published maps and institutional
affiliations.
Copyright: © 2020 by the authors. Li-
censeeMDPI, Basel, Switzerland. This
articleis an open accessarticle distributed
under the terms and conditions of the
Creative CommonsAttribution (CCBY)
license(https://creativecommons.org/
licenses/by/4.0/).
1Programa de Pós-Graduação em Ciência Florestal, Universidade Federal Vales do Jequitinhonha e Mucuri
Campus JK, Diamantina 39100-000, MG, Brazil; andre.rech@ufvjm.edu.br
2Federal Institute of Technology North of Minas Gerais (IFNMG), Diamantina 39100-000, MG, Brazil
3Department of Agriculture, University of Napoli Federico II, 80055 Portici (Naples), Italy;
gina.marano@unina.it
4Smithsonian Tropical Research Institute, 03092 Panamá, Panama; PiponiotC@si.edu
5School of Forest Resources and Conservation, University of Florida, Gainesville, FL 32611, USA;
c.silva@ufl.edu
6Institute of Geography, Federal University of Uberlandia (UFU), Av. João Naves de Ávila 2121,
Uberlandia 38400-902, Minas Gerais, Brazil; viniciusdantas@ufu.br
7Woods Hole Research Center, Falmouth, MA 02540, USA; lrattis@woodwellclimate.org
8Instituto de Pesquisa Ambiental da Amazônia, Canarana 78640-000, MT, Brazil
9Institute for Agriculture and Forestry Systems in the Mediterranean, National Research Council of Italy,
06128 Perugia, Italy; alessio.collalti@cnr.it
10 Department of Innovation in Biological, Agro-food and Forest Systems, University of Tuscia,
01100 Viterbo, Italy
*Correspondence: blfaria@gmail.com
Abstract:
In recent decades, droughts, deforestation and wildfires have become recurring phenomena
that have heavily affected both human activities and natural ecosystems in Amazonia. The time
needed for an ecosystem to recover from carbon losses is a crucial metric to evaluate disturbance
impacts on forests. However, little is known about the impacts of these disturbances, alone and
synergistically, on forest recovery time and the resulting spatiotemporal patterns at the regional
scale. In this study, we combined the 3-PG forest growth model, remote sensing and field derived
equations, to map the Amazonia-wide (3 km of spatial resolution) impact and recovery time of
aboveground biomass (AGB) after drought, fire and a combination of logging and fire. Our results
indicate that AGB decreases by 4%, 19% and 46% in forests affected by drought, fire and logging +
fire, respectively, with an average AGB recovery time of 27 years for drought, 44 years for burned and
63 years for logged + burned areas and with maximum values reaching 184 years in areas of high fire
intensity. Our findings provide two major insights in the spatial and temporal patterns of drought
and wildfire in the Amazon: (1) the recovery time of the forests takes longer in the southeastern part
of the basin, and, (2) as droughts and wildfires become more frequent—since the intervals between
the disturbances are getting shorter than the rate of forest regeneration—the long lasting damage they
cause potentially results in a permanent and increasing carbon losses from these fragile ecosystems.
Keywords: Amazon; recovery time; aboveground biomass; climate change; 3-PG; fire; logging
1. Introduction
Natural disturbances have a key role in forest ecosystem dynamics [
1
], yet global
changes in climate and land-uses have intensified disturbances rates in several biomes
with important consequences on the ecosystems resilience [
2
]. Events like droughts and
wildfires are becoming widespread phenomena in vast areas of the globe, potentially
affecting the ecosystem services they provide [
3
,
4
] even in humid biomes with high rainfall
rates, such as Amazonia [
5
–
7
]. Housing more than half of the world’s remaining rainforest
Forests 2021,12, 8. https://dx.doi.org/10.3390/f12010008 https://www.mdpi.com/journal/forests

Forests 2021,12, 8 2 of 17
areas, Amazonian forests account for considerable carbon storage in living biomass and
soils, estimated at around 150–200 Pg [
8
,
9
]. In addition, the region represents one of the
most important biodiversity hotspots of the planet [
10
,
11
]. Amazonian forests are under
considerable pressure due to the increased frequency and intensity of disturbances in moist
tropical regions [
12
]. Forest fires and large-scale drought events are both directly dependent
on climate [
13
] and their effects are expected to become more severe with climate change
effects (i.e., mostly warming and reduction in precipitation). In combination with human
activities, such as selective logging and other land-use changes, increasing fire and drought
severity are expected to cause significant forest losses [14].
The Amazon Basin’s historical baseline of disturbances has been heavily altered in the
last 20 years as a result of anthropogenic activities, increasing the rates of deforestation,
drought and wildfire and their impacts [
15
]. In the early 2000s, logging activities affected
ca. 10,000–20,000 km
2
year
−1
of tropical forests in the Brazilian Amazon and it is estimated
that understory fires destroyed ca. 85,000 km
2
of standing forests in the period 1999–
2010 [
16
,
17
]. Moreover, recent studies have shown that Amazonian forests are becoming
more exposed to droughts [
18
,
19
], including extreme drought events that would not be
expected to take place more than once in a century (e.g., the three devastating droughts of
2005, 2010 and 2016; [
20
,
21
]). Altogether, droughts, wildfires and logging activities increase
the susceptibility of forests to successive burning by increasing ignition rates, wind speed,
creating drier microclimatic conditions near the soil surface and promoting exotic grass
invasion. The effect of fire in forest ecosystems contrasts with that observed at larger spatial
scales (i.e., global scale) and in fire-prone regions in which anthropogenic influences often
reduce fire spread [
22
]. Therefore, the increasing risk of wildfires is an additional driver of
change in the Amazon region [23].
Forest degradation due to more frequent and intense disturbances in the Amazon [
24
,
25
]
results in long-term reduction in carbon stocks [
26
] with potential release of the C stored
in Amazonian forests. The degree of degradation of the forest C stocks depends on four
major factors: (1) the type of disturbance (e.g., logging, droughts and wildfires); (2) intensity
(i.e., percentage of C loss); (3) the time return interval (i.e., years from one event to the
next one) [
25
,
27
,
28
]; and (4) disturbance synergisms (i.e., the interacting effects between
disturbances).
Several studies have analyzed forest recovery after disturbances at either broad or
at multiple scales disturbances [
29
,
30
], but few of them have been conducted in tropical
forests and specifically in the Amazon Basin. When conducted, these studies are usually
limited in temporal scale (usually <20 years) [
25
,
31
,
32
] and focus on the effects of a single
disturbance and in relatively small areas [
33
–
35
]. There is a lack of studies looking at
recovery beyond 30–40 years. As a result, we still have a limited understanding on forest
aboveground biomass (AGB) resilience to disturbance in Amazonian forests (i.e., how
much time does it take for the forest to return to its pre-disturbance status), especially at
the regional scale and taking interacting effects of multiple disturbance into consideration.
One straightforward way of addressing the consequences of disturbance in forest
AGB is by integrating geospatial techniques with remote sensing and process-based forest
growth models [
36
,
37
]. Specifically, remote sensing and GIS technologies allow the assess-
ment of forest AGB at broad scales [
38
] whereas process-based forest growth models can
provide insights on the mechanisms and processes involved in forest recovery and their
relationship with spatiotemporal climate (including human)-induced scenarios. Models
can help in assessing the recovery time of vegetation using climatic variables to predict
vegetation productivity and its spatial variability [
38
]. At a regional scale, net primary
productivity (NPP) is often used as an indicator of inherent plant growth potential [
39
].
Several studies have indeed assumed a strong relationship between productivity and
biomass [
40
] with the first one being a function of the second. Indeed, the targeted pa-
rameter AGB is also influenced by climate, water availability and soil fertility [
39
–
41
]. In
this study, we assessed the recovery time (i.e., the time necessary for a forest to recover its
pre-disturbance AGB levels) of Brazilian Amazon forests AGB from drought, fire and a

Forests 2021,12, 8 3 of 17
combination of logging and fire disturbances, using a dynamic forest carbon model that
simulates vegetation recovery time as a function of climate scenarios and geospatial data.
With the present study, we aim to investigate the recovery time of AGB in the Amazon
forests when subject to a disturbance caused by: (1) an extreme drought, (2) a catastrophic
fire and (3) a combination of logging and fire disturbances by integrating the existing
knowledge [24,42–44] within our modeling framework.
2. Materials and Methods
We used a spatially implicit forest productivity model based on the net primary pro-
ductivity of the 3-PG model (Figure 1) (see Section 2.1 The Model) to estimate forest recovery
time (here defined as the time necessary for a forest to recover at its pre-disturbance AGB
levels). Analysis of AGB recovery was carried out for the Brazilian Amazon biome, which
encompasses about 3.5 million km
2
located between 15
◦
S–5
◦
N and 40
◦
W–80
◦
W. The
region consists of one of the largest preserved forests in the world that has been experienc-
ing strong human disturbances in recent times, especially in “the arch of deforestation”
(Figure 2).
Forests 2020, 11, x FOR PEER REVIEW 3 of 18
productivity (NPP) is often used as an indicator of inherent plant growth potential [39].
Several studies have indeed assumed a strong relationship between productivity and bi-
omass [40] with the first one being a function of the second. Indeed, the targeted parame-
ter AGB is also influenced by climate, water availability and soil fertility [39–41]. In this
study, we assessed the recovery time (i.e., the time necessary for a forest to recover its pre-
disturbance AGB levels) of Brazilian Amazon forests AGB from drought, fire and a com-
bination of logging and fire disturbances, using a dynamic forest carbon model that sim-
ulates vegetation recovery time as a function of climate scenarios and geospatial data.
With the present study, we aim to investigate the recovery time of AGB in the Amazon
forests when subject to a disturbance caused by: (1) an extreme drought, (2) a catastrophic
fire and (3) a combination of logging and fire disturbances by integrating the existing
knowledge [24,42–44] within our modeling framework.
2. Materials and Methods
We used a spatially implicit forest productivity model based on the net primary
productivity of the 3-PG model (Figure 1) (see Section 2.1 The Model) to estimate forest
recovery time (here defined as the time necessary for a forest to recover at its pre-disturb-
ance AGB levels). Analysis of AGB recovery was carried out for the Brazilian Amazon
biome, which encompasses about 3.5 million km2 located between 15° S–5° N and 40° W–
80° W. The region consists of one of the largest preserved forests in the world that has
been experiencing strong human disturbances in recent times, especially in “the arch of
deforestation” (Figure 2).
Figure 1. Proof–of–concept vegetation recovery time simulations as a function of climate variables
(i.e., soil-plant available water (fSW), photosynthetically active radiation (PAR), vapor pressure
deficit (fVPD), and air temperature (fTemp), see The Model for description). Aboveground biomass
(AGB) losses resulting from drought stress and fire are a function of the maximum climatological
water deficit (MCWD, see The Model for description).
Figure 1.
Proof–of–concept vegetation recovery time simulations as a function of climate variables (i.e., soil-plant available
water (fSW), photosynthetically active radiation (PAR), vapor pressure deficit (fVPD), and air temperature (fTemp), see The
Model for description). Aboveground biomass (AGB) losses resulting from drought stress and fire are a function of the
maximum climatological water deficit (MCWD, see The Model for description).

Forests 2021,12, 8 4 of 17
Forests 2020, 11, x FOR PEER REVIEW 4 of 18
Figure 2. Study area: Amazonian forest in Brazil. Amazon biome extent (gray area). Forest loss
map (yellow-red) has been displayed according to [45,46] (Global Forest Change dataset in Google
Earth Engine). Red pixels identify areas of where tree cover loss has been detected.
2.1. The Model
In this study, recovery time dynamics are simulated using the 3-PG model (Physio-
logical Principles in Predicting Growth; [47]), as embedded and parameterized into the
CARLUG model by [48], driven by four monthly climatic variables: photosynthetically
active radiation (PAR, mol PAR m−2 month−1), vapor pressure deficit (VPD, KPa), precipi-
tation (mm month−1) and air temperature (°C), respectively. The 3-PG model was used to
estimate gross and net primary productivity (GPP and NPP, both in g C m−2 month−1) as
follows:
   
(1)
where Y is the carbon use efficiency (i.e., the fraction of GPP not used to support auto-
trophic respiration, known as CUE [49–51]). GPP is computed as:
      
(2)
where αx is the maximum quantum canopy efficiency (i.e., the maximum capacity in con-
verting light into photosynthates without environmental or other functional limitations,
in mol C mol PAR−1 m−2 month−1), modifiers comprise environmental limitations to maxi-
mum photosynthetic rate (temperature, fTEMP; soil water, fSW; and vapor pressure defi-
cit, fVPD), with values ranging from zero (complete limitation) to one (no limitation). For
an in-depth description of modifiers algorithms see also [48,52]. The last two terms in
Equation (2) reflect the incident PAR effectively absorbed by the canopies (i.e., APAR)
based on their leaf area index (LAI, m2 m−2) and the leaf light extinction coefficient (k,
unitless) as in Beer’s Law [53].
Each month, the model assumes that leaf, wood, and root carbon pools increase by
an overall amount equal to the NPP, which are, respectively, allocated proportionally in
their three pools as in the standard 3-PG carbon partitioning-allocation scheme [54]. The
Figure 2.
Study area: Amazonian forest in Brazil. Amazon biome extent (gray area). Forest loss map (yellow-red) has been
displayed according to [
45
,
46
] (Global Forest Change dataset in Google Earth Engine). Red pixels identify areas of where
tree cover loss has been detected.
2.1. The Model
In this study, recovery time dynamics are simulated using the 3-PG model (Physio-
logical Principles in Predicting Growth; [
47
]), as embedded and parameterized into the
CARLUG model by [
48
], driven by four monthly climatic variables: photosynthetically
active radiation (PAR, mol PAR m
−2
month
−1
), vapor pressure deficit (VPD, KPa), precipi-
tation (mm month
−1
) and air temperature (
◦
C), respectively. The 3-PG model was used to
estimate gross and net primary productivity (GPP and NPP, both in g C m
−2
month
−1
) as
follows:
NPP =GPP ×Y(1)
where Y is the carbon use efficiency (i.e., the fraction of GPP not used to support autotrophic
respiration, known as CUE [49–51]). GPP is computed as:
GPP =αx×modi f iers ×PAR ×1−e−k×LAI(2)
where
αx
is the maximum quantum canopy efficiency (i.e., the maximum capacity in
converting light into photosynthates without environmental or other functional limitations,
in mol C mol PAR
−1
m
−2
month
−1
), modifiers comprise environmental limitations to
maximum photosynthetic rate (temperature, fTEMP; soil water, fSW; and vapor pressure
deficit, fVPD), with values ranging from zero (complete limitation) to one (no limitation).
For an in-depth description of modifiers algorithms see also [
48
,
52
]. The last two terms
in Equation (2) reflect the incident PAR effectively absorbed by the canopies (i.e., APAR)
based on their leaf area index (LAI, m
2
m
−2
) and the leaf light extinction coefficient (k,
unitless) as in Beer’s Law [53].

Forests 2021,12, 8 5 of 17
Each month, the model assumes that leaf, wood, and root carbon pools increase by
an overall amount equal to the NPP, which are, respectively, allocated proportionally in
their three pools as in the standard 3-PG carbon partitioning-allocation scheme [
54
]. The
partitioning of NPP is the outcome of the climate and soil conditions interacting with
vegetation through a series of differential equations that describe the flow of C within the
tree compartments [
48
]. Therefore, the model predicts the distribution of forest biomass
from carbon stocks, but in order to obtain biomass we converted C to biomass assuming
that one ton of biomass contains 0.5 tons of C [
55
]. We assume that the re-equilibration of
forest carbon after disturbances (i.e., steady state undisturbed conditions) is when the AGB
growth and the decay rates stabilize. We also estimated the average time to recover 90% of
old-growth forests’ carbon levels. The 90% threshold has often been used in similar studies
(e.g., [
56
]) and can thus more easily be compared to previous results; the 100% threshold
corresponds to a full recovery of carbon stocks, but it may take significantly longer.
The study conducted by [
48
] uses the recalibrated 3-PG model parameters for the
Amazonian forests (the overall parameters description and their values are shown in
Supporting Information, see Table S1). The 3-PG calculates NPP as a constant fraction of
GPP, using an NPP/GPP ratio (Y = 0.47) based on empirical evidence [
47
]. For Brazilian
Amazon forests other studies suggest Y to be closer to 0.3 [
57
] while others report much
higher values at some tropical sites, even including Amazonian ones (i.e., Y > 0.5; [
51
]).
However, the issue of whether Y is a constant value, its actual value, even including its
top-down limits, is a much-debated issue as described in [51,58].
An overall 3-PG model parameter sensitivity analysis has been performed already by
a number of authors (e.g., [
59
]) showing how the 3-PG model is mostly sensitive to stem
allometric parameters (i.e., those used to obtain from trees structure the tree biomass), ratios
for biomass partitioning and allocation, maximum canopy conductance, turnover time of
wood, and maximum canopy quantum efficiency. For an in-depth 3-PG model parameter
sensitivity analysis we refer to the works of [
48
,
59
] and this will be not considered and
discussed further here. In addition, we used the pan-tropical biomass map generated by
Avitabile et al. [
60
] as reference (pre-impact) levels to initialize the model and combining it
with two comprehensive recent estimates of carbon density (i.e., estimations of [
55
,
61
] and
covering a wide 250–500 Mg ha−1range (Figure S1).
2.2. Estimating Drought, Fire and Logging Impacts on AGB Stocks
The loss of AGB due to drought events was modeled as a function of the MCWD
(Maximum Climatological Water Deficit index, representing the maximum climatological
water deficit reached in the year), a common index used to measure the cumulative water
stress in Amazonia (e.g., [
42
,
62
,
63
]). The MCWD reflects the intensity and length of the
dry season, when evapotranspiration exceeds precipitation (i.e., negative balance). A
measure of water deficit related to tree mortality in Amazonian forests that is denoted as in
Lewis et al. [42], that is:
∆AGB =0.378 −0.052 ×∆MCWD (3)
We estimated the MCWD anomalies (namely,
∆
MCWD) for the year 2010 by first
estimating the mean MCWD for the baseline period from 1998 to 2015, without considering
both the years 2005 and 2010. The
∆
MCWD have been shown to be strong predictors
of drought-associated tree mortality in the Amazon [
62
]. Specifically, a monthly water
deficit was calculated as the difference between precipitation and evapotranspiration (with
ground measurements estimated at 100 mm per month [
63
,
64
], i.e., evapotranspiration is
fixed at 100 mm month
−1
). As a result, we assume that the forest is in water deficit when
monthly precipitation falls below 100 mm. MCWD was calculated as the sum of sequential
monthly water deficits, where more negative MCWD values indicate higher drought stress.
We quantified the MCWD for the year of 2010 using the product 3B43 of TRMM (Tropical
Rainfall Measuring Mission at 0.25
◦
grid-resolution), and then, the average of carbon losses

Forests 2021,12, 8 6 of 17
for each pixel using Equation (3). The 2010 drought is one of the most intense and spatially
extensive drought events ever recorded in the Brazilian Amazon [42].
Effects of wildfire were estimated by using the CARLUC-Fire model [
44
]. This model
specifically accounts for the effects of fire by estimating forest carbon losses after a fire
event as a function of its intensity (FI). FI is defined as the energy released per unit length
of fire-line (kWm
−2
), which is a key factor in estimating how vegetation responds to fire
events. The relationship between fire intensity and fire-induced biomass losses was derived
from a large-scale fire experiment in southeast Amazonia [
24
,
44
] (Equation (4)). Based on
this experiment, AGB losses were calculated as a function of FI as follows:
AGBlosses =1
1+e(2.45−0.002373×FI)(4)
We limit our fire analysis to areas that burned between 2003 and 2016 [
65
] using
information at 500 m resolution from the Moderate Resolution Imaging Spectroradiometer
(MODIS) Collection 6 MCD64A1 burned area product over the period 2003–2016.
As a substantial proportion of fires occurred in areas likely to have been previously
logged, we accounted for this effect in the estimation of the initial AGB by incorporating
an additional loss in fire effects of 40% in burned areas that were also cleared. We assumed
this based on findings of Berenguer et al. [
43
] that an average forest under selective
logging stores about 40% less carbon. Logged areas were defined using data from the
annual Landsat-based Project for Monitoring Amazonian Deforestation (PRODES, http:
//www.obt.inpe.br/prodes). Because edge effects from logging have been shown to affect
forests up to 2–3 km from the border [
66
], we include forests located within 3 km from a
deforested pixel, as a selective logging influence zone and they were defined using data
from PRODES with cumulative deforestation up to 2017.
2.3. Experimental Runs
We ran the 3-PG model at 3 km
×
3 km spatial resolution under mean monthly
climate conditions for the 1980–2009 period, to estimate the forest recovery time for both
drought, fire and logging + fire impacts (includes loss from logging and losses from fire).
Climate input variables used to calculate the climatic means consisted of monthly series
of temperature and mean vapor pressure deficit from the Climate Research Unit (CRU
TS; [
67
]), while PAR was obtained from the GOES–9 satellite product [
68
]. In each pixel,
AGB recovery was assessed by simulating AGB dynamics with the model after an AGB
loss corresponding to disturbance impact.
2.4. Assessing Model Results
Light detection and ranging (Lidar) remote sensing is widely used for monitoring
forest structure and biomass dynamics [
69
,
70
] in many forest ecosystems [
71
]. For instance,
airborne lidar (ALS) technologies help quantify changes in canopy structure, carbon stocks
and recovery time at the local-to-regional scale under different types of forest degradation
(e.g., [25,72,73]).
In the present study, we compare our modeled recovery time from fire in logged
areas with airborne lidar-derived aboveground carbon density (ACD) recovery estimates
in forest stands (2891.45 Ha) located in Feliz Natal (Mato Grosso, Brasil) that were logged
and burned once. For computing the recovery time of ACD from lidar, we applied a
model developed by Rappaport et al. [
25
] that used multiple linear regression to model the
recovery time of ACD (Kg C m
−2
) in degraded forest stands based on degradation type. In
their study, the model was calibrated using a chronosequence of ACD maps derived from
lidar and degradation history data (from 2013 to 2018) across degraded forests stands [
25
].
The model is presented in Equation (5) and shows adjusted R
2
of 0.89. Herein, we chose to

Forests 2021,12, 8 7 of 17
compare our results with those provided in Rappaport et al. [
25
] due to lack of available
field data on the time scale addressed here to assess recovery time.
RT = 62.259 + 11.395 ×log(t) −10.268 ×CF1 (5)
where RT refers to recovery time, t refers to time (years) and CF1 refers to degradation
history, once-burned stands.
2.5. Disturbance Return Interval
In order to inquire whether global changes could determine an increase in future
drought and fire frequency we projected the areal extent and spatial patterns of future
drought and fire impacts up to the year 2100 in order to understand whether global changes
could determine an increase in future drought and fire frequency in the study area. We
analyzed both future precipitations (based on Representative Concentration Pathways,
i.e., RCP 8.5—representing unmitigated climate change scenario) and a land use changes
scenario (based on Aguiar et al. [
74
]) with a decrease in the extension and level of protection
of the areas and increases in deforestation rates from 2014 to 2020 and continuing until
2100.
We built drought scenarios (2040–2070 and 2071–2100) using precipitation (related
with water stress, MCWD) from the ensemble of 35 climate models participating in the
Coupled Model Intercomparison Project phase 5 (CMIP-5, [
75
]). In detail, we derived the
forcing from the mean monthly simulated precipitation anomalies first averaged for all 35
models and then bias corrections with Tropical Rainfall Measuring Mission (TRMM data
product 3B43 [
63
]). To investigate frequency of future Amazonian droughts we assumed
severe drought condition when MCWD anomalies (subtraction between future projections
and the historical average) is <
−
40 mm (threshold derived by Phillips et al. [
62
]), below this
threshold water stress is assumed to induce losses in AGB. We also used maps of predicted
change in fire recurrence in response to global changes obtained from Fonseca et al. [
76
]
based on future land-use change data by Aguiar et al. [
74
]. The fire scenarios (2040–2070
and 2071–2100) developed by Fonseca et al. [
76
] combine the effects of future land-use and
climate change on fire relative probability in the Brazilian Amazon in the best-case and
worst-case scenarios. We assume fire relative probability to equal fire relative frequency
and then determine the mean fire return interval as the inverse of fire relative frequency.
3. Results
Results show that disturbances have substantially affected biomass in Brazilian Ama-
zonia. In the locations affected by drought, fire and logging + fire, AGB decreased by 4%,
19% and 46%, respectively (Figure 3). Our results suggest that during the 2010 drought,
about 1.5 million km
2
of the Brazilian Amazon lost a considerable amount of AGB (we
considered losses
≥
10% of the initial AGB). Fire could also produce substantial losses
in above-ground carbon affecting 550,000 km
2
especially in southern Brazilian Amazon.
Approximately 150,000 km
2
of the burned forest patches were located within 3 km from a
logged forest.
Average AGB recovery time was 27 years for drought-impacted, 44 years for burned,
and 63 years for logged + burned areas (includes loss from logging and loss from fire).
Recovery time from drought revealed a northwest-to-southeast gradient in the study area
(Figure 4a). Roughly 20% of these drought-affected areas, corresponding to ca. 364,000 km
2
,
were estimated to recover in the first 10 years, with maximum values reaching 90 years in
parts of southeastern Brazilian Amazonia (Figure 4). Forest fires were widespread across
the “arch of deforestation” (the region in southern and eastern Amazonia where the rates
of deforestation are higher) during the period 2003–2016 (Figure 4b). The longest recovery
times during this period were concentrated along the eastern and southwestern extent of
Amazon forests in Brazil, where the maximum was about 150 years after fire disturbance.
Subsequent wildfires events (i.e., multiple fires in the same location) accounted for 10% of
all forest fires during the period 2003–2016, delaying forest recovery times within these

Forests 2021,12, 8 8 of 17
areas (Figure 4b). The longest recovery times were found in logged-and-burned forests
with maximum values reaching 184 years (Figure 4c,c1). These results consider a recovery
of the carbon stock corresponding to 100% (i.e., recovery time ~184 years) (see The Model)
resulting in a difference of about 122 years in logged and burned forest which would be
much faster if we would consider a recovery threshold of 90% (i.e., recovery time ~62 years)
(Figure S2).
Forests 2020, 11, x FOR PEER REVIEW 7 of 18
2.5. Disturbance Return Interval
In order to inquire whether global changes could determine an increase in future
drought and fire frequency we projected the areal extent and spatial patterns of future
drought and fire impacts up to the year 2100 in order to understand whether global
changes could determine an increase in future drought and fire frequency in the study
area. We analyzed both future precipitations (based on Representative Concentration
Pathways, i.e., RCP 8.5—representing unmitigated climate change scenario) and a land
use changes scenario (based on Aguiar et al. [74]) with a decrease in the extension and
level of protection of the areas and increases in deforestation rates from 2014 to 2020 and
continuing until 2100.
We built drought scenarios (2040–2070 and 2071–2100) using precipitation (related
with water stress, MCWD) from the ensemble of 35 climate models participating in the
Coupled Model Intercomparison Project phase 5 (CMIP-5, [75]). In detail, we derived the
forcing from the mean monthly simulated precipitation anomalies first averaged for all 35
models and then bias corrections with Tropical Rainfall Measuring Mission (TRMM data
product 3B43 [63]). To investigate frequency of future Amazonian droughts we assumed
severe drought condition when MCWD anomalies (subtraction between future projec-
tions and the historical average) is <−40 mm (threshold derived by Phillips et al. [62]),
below this threshold water stress is assumed to induce losses in AGB. We also used maps
of predicted change in fire recurrence in response to global changes obtained from Fon-
seca et al. [76] based on future land-use change data by Aguiar et al. [74]. The fire scenarios
(2040–2070 and 2071–2100) developed by Fonseca et al. [76] combine the effects of future
land‐use and climate change on fire relative probability in the Brazilian Amazon in the
best-case and worst-case scenarios. We assume fire relative probability to equal fire rela-
tive frequency and then determine the mean fire return interval as the inverse of fire rel-
ative frequency.
3. Results
Results show that disturbances have substantially affected biomass in Brazilian Ama-
zonia. In the locations affected by drought, fire and logging + fire, AGB decreased by 4%,
19% and 46%, respectively (Figure 3). Our results suggest that during the 2010 drought,
about 1.5 million km2 of the Brazilian Amazon lost a considerable amount of AGB (we
considered losses ≥10% of the initial AGB). Fire could also produce substantial losses in
above-ground carbon affecting 550,000 km2 especially in southern Brazilian Amazon. Ap-
proximately 150,000 km2 of the burned forest patches were located within 3 km from a
logged forest.
Figure 3. Biomass density plots describing patterns before and after drought (a), fire (b) and log-
ging + fire (c) impacts. Only areas that burned between 2003 and 2016 are considered and, for (c),
only burned areas up to 3 km from logging areas. Recovery is defined as 100% of pre-disturbance
AGB.
Figure 3.
Biomass density plots describing patterns before and after drought (
a
), fire (
b
) and logging + fire (
c
) impacts. Only
areas that burned between 2003 and 2016 are considered and, for (
c
), only burned areas up to 3 km from logging areas.
Recovery is defined as 100% of pre-disturbance AGB.
Forests 2020, 11, x FOR PEER REVIEW 8 of 18
Average AGB recovery time was 27 years for drought-impacted, 44 years for burned,
and 63 years for logged + burned areas (includes loss from logging and loss from fire).
Recovery time from drought revealed a northwest-to-southeast gradient in the study area
(Figure 4a). Roughly 20% of these drought-affected areas, corresponding to ca. 364,000
km2, were estimated to recover in the first 10 years, with maximum values reaching 90
years in parts of southeastern Brazilian Amazonia (Figure 4). Forest fires were widespread
across the “arch of deforestation” (the region in southern and eastern Amazonia where
the rates of deforestation are higher) during the period 2003–2016 (Figure 4b). The longest
recovery times during this period were concentrated along the eastern and southwestern
extent of Amazon forests in Brazil, where the maximum was about 150 years after fire
disturbance. Subsequent wildfires events (i.e., multiple fires in the same location) ac-
counted for 10% of all forest fires during the period 2003–2016, delaying forest recovery
times within these areas (Figure 4b). The longest recovery times were found in logged-
and-burned forests with maximum values reaching 184 years (Figure 4c,c1). These results
consider a recovery of the carbon stock corresponding to 100% (i.e., recovery time ~184
years) (see The Model) resulting in a difference of about 122 years in logged and burned
forest which would be much faster if we would consider a recovery threshold of 90% (i.e.,
recovery time ~62 years) (Figure S2).
Figure 4. Aboveground recovery time (in years) for 2010 drought (a), fire areas that burned be-
tween 2003 and 2016 (b) and in areas that were both burned and logged (c). Histogram plots sum-
marize AGB recovery pixels distributions (in years), for drought (a1), fire (b1) and logging + fire
(c1).
We compared our results with a lidar-derived model of recovery time in stands that
were logged and burned once (Figure 5a). Our estimations show smaller AGB decreases
in comparison with lidar-based estimates of carbon losses from fire (loss of AGB of 46%
vs. 55%). However, recovery rates were shown to be strongly correlated (Figure 5b).
Figure 4.
Aboveground recovery time (in years) for 2010 drought (
a
), fire areas that burned between 2003 and 2016 (
b
) and
in areas that were both burned and logged (
c
). Histogram plots summarize AGB recovery pixels distributions (in years), for
drought (a1), fire (b1) and logging + fire (c1).
We compared our results with a lidar-derived model of recovery time in stands that
were logged and burned once (Figure 5a). Our estimations show smaller AGB decreases in
comparison with lidar-based estimates of carbon losses from fire (loss of AGB of 46% vs.
55%). However, recovery rates were shown to be strongly correlated (Figure 5b).
Increases in the extent and frequency of drought and fire (Figure 6) suggest that these
future disturbances could undermine the full forest recovery. Our results suggest that by
2070 the area affected by drought will increase approximately three-fold (Figure 6—top
panel). Moreover, from the middle to the end of the century, the mean fire return intervals
(FRI) was projected to decrease from 10 to 8 years and the median FRI to decrease from 8 to
6 years from the 2040–2070 period to the 2070–2100 periods, respectively, in a worst case

Forests 2021,12, 8 9 of 17
land use change scenario (Figure 6bottom panel). However, in a more optimistic scenario
the area subject to high fire frequency would be smaller (Figure 6middle panel).
Forests 2020, 11, x FOR PEER REVIEW 9 of 18
Figure 5. Airborne light detection and ranging (lidar) data were sampled (red line) in Feliz Natal,
within the Xingu basin (light green), Brazilian state of Mato Grosso (a). The forest growth model
(3-PG green line) shows the relationship between aboveground biomass (%) and recovery time in
years. We compared it with a lidar-derived model of recovery time in stands that were logged and
burned once (CF1 refers to once-burned) [25] (orange line) (b). A sample of vertical profile of a
recovering forest which was degraded by fire and selective logging (c). The discrete return lidar
data used for creating the transect figure were acquired in 2018 with a point density of 22.98
points m−2 covering an area of 2891.25 ha in Feliz Natal, Mato Grosso, Brazil [25], as part of the
Sustainable Landscapes Brazil project program (data available from:
https://www.paisagenslidar.cnptia.embrapa.br/webgis/; details of airborne lidar (ALS) data acqui-
sitions are presented in the supplementary material, Table S2).
Increases in the extent and frequency of drought and fire (Figure 6) suggest that these
future disturbances could undermine the full forest recovery. Our results suggest that by
2070 the area affected by drought will increase approximately three-fold (Figure 6—top
panel). Moreover, from the middle to the end of the century, the mean fire return intervals
(FRI) was projected to decrease from 10 to 8 years and the median FRI to decrease from 8
to 6 years from the 2040–2070 period to the 2070–2100 periods, respectively, in a worst
case land use change scenario (Figure 6 bottom panel). However, in a more optimistic
scenario the area subject to high fire frequency would be smaller (Figure 6 middle panel).
Figure 5.
Airborne light detection and ranging (lidar) data were sampled (red line) in Feliz Natal, within the Xingu
basin (light green), Brazilian state of Mato Grosso (
a
). The forest growth model (3-PG green line) shows the relationship
between aboveground biomass (%) and recovery time in years. We compared it with a lidar-derived model of recovery
time in stands that were logged and burned once (CF1 refers to once-burned) [
25
] (orange line) (
b
). A sample of vertical
profile of a recovering forest which was degraded by fire and selective logging (
c
). The discrete return lidar data used for
creating the transect figure were acquired in 2018 with a point density of 22.98 points m
−2
covering an area of 2891.25 ha in
Feliz Natal, Mato Grosso, Brazil [
25
], as part of the Sustainable Landscapes Brazil project program (data available from:
https://www.paisagenslidar.cnptia.embrapa.br/webgis/; details of airborne lidar (ALS) data acquisitions are presented in
the supplementary material, Table S2).
Forests 2020, 11, x FOR PEER REVIEW 10 of 18
Figure 6. Projected changes in droughts (as maximum climatological water deficit anomalies,
ΔMCWD) (upper panel) and fire return interval based on an optimistic land use scenario (mid
panel) and in the unmitigated scenarios with the worst‐case land‐use scenario (bottom panel).
4. Discussion
In the present study, we explored the AGB changes after drought, fire and a combi-
nation of logging and fire disturbances and the time needed for complete recovery as a
function of both climatic conditions and AGB in the Brazilian Amazon forest, using a
modeling-based approach. Our results suggest that fire is a much greater threat than
drought for the forest resilience, especially if logging occurs. These results highlight the
key threat imposed by fire to Amazon forests. The intensity of the disturbance event is
strongly related to both the amount of AGB lost and the recovery time of the forest. The
biomass recovery rates estimates reported here are consistent with those from Poorter et
al. [56] that showed AGB of Neotropical second growth forest took a median time of 66
years to recover to 90% of previous growth values after multiple disturbances events, in-
cluding land use changes. On the other hand, recent evidence [77] suggests that recovery
time might take at least 150 years until secondary forests (re)gain carbon levels similar to
primary forests, after drought disturbances thus indicating that these biomes have recov-
ery rates that are much lower than previously suggested.
Our results also suggest that by the end of the century, especially after 2070, the Bra-
zilian Amazon will be affected by more frequent droughts with the southern area being
more vulnerable since it will need a longer time to recover after these events. Thus, climate
change will greatly increase the threat imposed to the forest, potentially jeopardizing for-
est resilience. The interplay between longer forest recovery times and more frequent
droughts has been previously evidenced in the Amazonia, where longer recovery times
have been documented [78]. Moreover, if on the one hand the extreme droughts of 2005,
2010 and 2016 have prevented the full recovery of the forests, on the other, drought effects
on forest canopy carbon fixation capacity could potentially persist for several years during
recovery processes [78], leading to forest degradation and changes in forest species com-
position, and evidence suggests that taller tree species have significantly higher mortality
than small tree species, when subject to drought [79,80].
Figure 6.
Projected changes in droughts (as maximum climatological water deficit anomalies,
∆
MCWD) (
upper panel
)
and fire return interval based on an optimistic land use scenario (
mid panel
) and in the unmitigated scenarios with the
worst-case land-use scenario (bottom panel).

Forests 2021,12, 8 10 of 17
4. Discussion
In the present study, we explored the AGB changes after drought, fire and a com-
bination of logging and fire disturbances and the time needed for complete recovery as
a function of both climatic conditions and AGB in the Brazilian Amazon forest, using
a modeling-based approach. Our results suggest that fire is a much greater threat than
drought for the forest resilience, especially if logging occurs. These results highlight the key
threat imposed by fire to Amazon forests. The intensity of the disturbance event is strongly
related to both the amount of AGB lost and the recovery time of the forest. The biomass
recovery rates estimates reported here are consistent with those from Poorter et al. [
56
] that
showed AGB of Neotropical second growth forest took a median time of 66 years to recover
to 90% of previous growth values after multiple disturbances events, including land use
changes. On the other hand, recent evidence [
77
] suggests that recovery time might take
at least 150 years until secondary forests (re)gain carbon levels similar to primary forests,
after drought disturbances thus indicating that these biomes have recovery rates that are
much lower than previously suggested.
Our results also suggest that by the end of the century, especially after 2070, the
Brazilian Amazon will be affected by more frequent droughts with the southern area being
more vulnerable since it will need a longer time to recover after these events. Thus, climate
change will greatly increase the threat imposed to the forest, potentially jeopardizing forest
resilience. The interplay between longer forest recovery times and more frequent droughts
has been previously evidenced in the Amazonia, where longer recovery times have been
documented [
78
]. Moreover, if on the one hand the extreme droughts of 2005, 2010 and
2016 have prevented the full recovery of the forests, on the other, drought effects on forest
canopy carbon fixation capacity could potentially persist for several years during recovery
processes [
78
], leading to forest degradation and changes in forest species composition,
and evidence suggests that taller tree species have significantly higher mortality than small
tree species, when subject to drought [79,80].
Our findings also confirm that the land carbon sink in the Brazilian Amazon will
be strongly impacted by a regime of a chronic state of incomplete recovery [
78
], with
adverse consequences also on the GPP due to shifts in precipitation patterns caused by
anthropogenic emissions [
81
–
83
]. Indeed, across Amazon forests, GPP is modeled to
decrease linearly with increasing seasonal water deficit [
82
]. Longer and more intense
dry seasons have been forecasted, together with an increased frequency and severity of
drought events [
84
–
86
] and future Amazon droughts are expected to become even more
frequent [
87
,
88
]. Our projections suggest about one extreme drought per decade (drought
return interval ranging from 4 to 16 years depending on the scenario of climate change).
If drought frequency increases, Amazon forest, both as species composition and regional
carbon sink, will be affected, which will thereby have an impact on global carbon cycling and
contribute further to climate change [
62
,
80
,
89
,
90
]. Previous studies have shown increased
fire occurrence and tree mortality during and after Amazon droughts [6,89,91–93]. If these
events continue to increase in frequency, large parts of the Amazon could potentially shift
from rainforest vegetation to a fire-maintained degraded forest and may promote the
persistence of degraded forests with a savanna-like structure [
94
,
95
]. This change in forest
type, structure and ecology would most likely reduce both the forest sink capacity and
even its biodiversity and ecosystem services [
94
]. The net increase in areas that are more
susceptible to wildfires, induced by either drought events increase, or potentially intensified
by climate change, could lead to significant biomass losses [9,96].
Human pressures play a crucial role in fire ignitions, wildfires could break out also
in non-dry years as in 2019, when more than 69,000 km
2
burnt despite the absence of
anomalous drought [
97
]. As droughts and wildfires are expected to become more frequent,
the time of occurrence between these disturbances may even get shorter than forest recovery
time, determining permanently damaged ecosystems and widespread degradation [
95
].
Although forest growth models are powerful tools that can be applied in simulating the
C dynamics in forests [
98
,
99
], our results are subject to some uncertainty and a number

Forests 2021,12, 8 11 of 17
of caveats [
100
,
101
]. In this study, we modeled vegetation recovery time as a function
of climate only. This approach does not account for regional variation in growth rates
depending on soils types (due to their inner physico-chemical properties such as water
retention or local-scale variation based on prior land use [
92
,
93
]) growth rates are also
known to vary significantly by species [
43
]. In addition to the mechanisms mentioned
above, CO
2
fertilization of Amazonian vegetation and nitrogen deposition could play
an important, but yet often neglected, role in forest regeneration [
102
]. It has also been
suggested that atmospheric CO
2
generally stimulates plant growth with increased rates
in photosynthetic activity and indirectly through increased water-use efficiency [
103
], but
not in all cases [
104
]. As CO
2
accumulates in the atmosphere, Amazonian trees may
also accumulate more biomass resulting in denser canopies and faster growth [
105
]. But
an increased atmospheric CO
2
concentration necessarily implies an increase in mean air
temperature which is in turn speculated to increase plants’ respiration and should result in a
levelled-off forest carbon use efficiency [
83
]. Recent studies indicate that the ability of intact
tropical forests to remove carbon from the atmosphere may be already saturating [
9
,
106
]
while others indicate for tropical species higher thermal acclimation capacities to buffer
C–losses by respiration [
51
], thus, calling for more studies on the possible consequences
of warming and increased atmospheric CO
2
concentration on forest dynamics. However,
in the Amazon phosphorus is an important limiting nutrient over large parts and its low
availability may limit positive CO2fertilization effects.
Future Possibilities for Model Improvement
Lidar-derived 3D-point cloud and biomass products can be used to enhance models’
representation of complex and heterogeneous forest ecosystems, such as those found in
Amazonia [
107
], and therefore can be used as input or to initialize vegetation models [
108
].
For instance, Longo et al. [
109
] have used lidar to obtain initial conditions for an ecosystem
model that requires an initial state for forest structure. Their method to derive the vertical
structure of the canopy from high-resolution airborne lidar successfully characterized the
diversity of forest structure variability caused by human-induced forest degradation (such
as logging and fire).
This new approach has strong implications on modeling recovery time and the suc-
cessional trajectories of the Amazonian disturbed forest because it does not require any
assumption on the successional stage of the forest, but only the vertical distribution of
returns. Moreover, it could be adapted to space-borne lidar data, including NASA’s Global
Ecosystem Dynamics Investigation (GEDI, [
110
]). Fusion of GEDI and optical data [
111
]
will further expand the spatial extent of available lidar data and potentially provide tools
capable of mapping drought, fire and logging impacts helping models to assess recovery
time. Moreover, integration of GEDI with either optical or radar [
112
] wall-to-wall data
could allow large-scale characterization of forest ecosystems structure providing accurate
measurements of biomass stock that could be used for assessing recovery time via repeated
measurements.
5. Conclusions
This study shows how forest growth models can be used as tools for complementing
field-based studies on recovery time by investigating the spatial and temporal dynamics
and processes of forest recovery. Indeed, our biomass recovery map illustrates both spatial
and climatic variability in carbon sequestration potential due to forest re-growth. By
mapping potential for biomass recovery across Amazonia, policy makers could focus their
efforts on specific areas that require special protection and need to be preserved. Moreover,
such recovery maps could also help by identifying areas with higher carbon sequestration
potential thus supporting policies and concrete actions to mitigate forest degradation in
areas where biomass resilience is under increasing stress (such as southeastern Amazonia).
The capability and timing of forest recovery after drought, fire and logging are urgent and
hot topics for applied research calling upon conservation and policy actions in Amazonia.

Forests 2021,12, 8 12 of 17
Future changes in fire regimes could push some Amazonian regions into a permanently
drier climate regime and weaken the resilience of the region to possible large-scale drought–
fire interactions driven by climate change. We are far from an integrated view of forest
recovery processes, yet the results presented in this study may provide some new insights
about forest recovery time after disturbances. The consequences that an extreme climatic
event, such as a drought, may cause in the forest can result in a net loss of ecosystem
services compromising these ecosystems dynamics in the long term. As a major result of
projected increases in fire and drought frequency and intensity in the region, Amazonian
forest resilience appears, in the medium and long term, to be severely jeopardized.
Supplementary Materials:
The following are available online at https://www.mdpi.com/1999-4
907/12/1/8/s1. Figure S1. Pre-disturbance reference biomass map [
44
]. Figure S2. The ABG
dynamic as reproduced by the forest growth model (3-PG green line) showing the relationship
between aboveground biomass (%) and recovery time in years to reach recovery threshold. Red
dotted line 90% threshold and black dotted line 100% threshold. Table S1: Parameters description
and their values used in 3-PG model (modified from Hirsch et al., [
48
]). Table S2: Details of ALS data
acquisitions.
Author Contributions:
Conceptualization, B.L.D.F.; methodology, B.L.D.F., A.C. and C.P.; software,
B.L.D.F. and C.A.S.; draft preparation, G.M., A.C., B.L.D.F., C.P., L.R., V.d.L.D. and A.R.R.; writing—
review and editing, G.M., A.C., C.P., L.R., C.A.S., V.d.L.D., B.L.D.F. and A.R.R.; visualization, B.L.D.F.,
C.A.S. and G.M.; supervision, G.M., A.C. and A.R.R. All authors have read and agreed to the
published version of the manuscript.
Funding: This research received no external funding.
Acknowledgments:
B.L.D.F. would like to thank the IFNMG and CNPq for financial support. He
would like to especially thank the IFNMG campus Pirapora for giving him the opportunity to pursue
a PhD. A.R.R. and B.L.D.F. thanks the support received from PPGCF/UFVJM.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Seidl, R.; Fernandes, P.M.; Fonseca, T.F.; Gillet, F.; Jönsson, A.M.; Merganiˇcová, K.; Netherer, S.; Arpaci, A.; Bontemps, J.D.;
Bugmann, H.; et al. Modelling natural disturbances in forest ecosystems: A review. Ecol. Model. 2011,222, 903–924. [CrossRef]
2.
D’Andrea, E.; Rezaie, N.; Prislan, P.; Griˇcar, J.; Collalti, A.; Muhr, J.; Matteucci, G. Frost and drought: Effects of extreme weather
events on stem carbon dynamics in a Mediterranean beech forest. Plant Cell Environ. 2020,43, 2365–2379. [CrossRef] [PubMed]
3.
Pyne, S. The Ecology of Fire. Available online: https://www.nature.com/scitable/knowledge/library/the-ecology- of-fire-1325
9892/ (accessed on 10 June 2020).
4.
Noce, S.; Collalti, A.; Valentini, R.; Santini, M. Hot spot maps of forest presence in the Mediterranean basin. IForest
2016
,9,
766–774. [CrossRef]
5. Mishra, A.K.; Singh, V.P. A review of drought concepts. J. Hydrol. 2010,391, 202–216. [CrossRef]
6.
Brando, P.M.; Nepstad, D.C.; Davidson, E.A.; Trumbore, S.E.; Ray, D.; Camargo, P. Drought effects on litterfall, wood production
and belowground carbon cycling in an Amazon forest: Results of a throughfall reduction experiment. Philos. Trans. R. Soc. B
Biol. Sci. 2008,1498, 1839–1848. [CrossRef]
7.
Leskinen, P.; Cardellini, G.; González García, S.; Hurmekoski, E.; Sathre, R.; Seppälä, J.; Smyth, C.E.; Stern, T.; Verkerk, H.
Substitution effects of wood-based products in climate change mitigation. From Sci. Policy 2018,7, 28.
8.
Mitchard, E.T.A.; Feldpausch, T.R.; Brienen, R.J.W.; Lopez-Gonzalez, G.; Monteagudo, A.; Baker, T.R.; Lewis, S.L.; Lloyd, J.;
Quesada, C.A.; Gloor, M.; et al. Markedly divergent estimates of Amazon forest carbon density from ground plots and satellites.
Glob. Ecol. Biogeogr. 2014,23, 935–946. [CrossRef]
9. Brienen, R.J.W.; Phillips, O.L.; Feldpausch, T.R.; Gloor, E.; Baker, T.R.; Lloyd, J.; Lopez-Gonzalez, G.; Monteagudo-Mendoza, A.;
Malhi, Y.; Lewis, S.L.; et al. Long-term decline of the Amazon carbon sink. Nature 2015,519, 344–348. [CrossRef]
10.
Saatchi, S.; Houghton, R.A.; Dos Santos Alvalá, R.C.; Soares, J.V.; Yu, Y. Distribution of aboveground live biomass in the Amazon
basin. Glob. Chang. Biol. 2007,13, 816–837. [CrossRef]
11.
Gardner, T.A.; Barlow, J.; Chazdon, R.; Ewers, R.M.; Harvey, C.A.; Peres, C.A.; Sodhi, N.S. Prospects for tropical forest biodiversity
in a human-modified world. Ecol. Lett. 2009,12, 561–582. [CrossRef] [PubMed]
12.
Lewis, S.L.; Edwards, D.P.; Galbraith, D. Increasing human dominance of tropical forests. Science
2015
,349, 827–832. [CrossRef]
13.
Seidl, R.; Schelhaas, M.J.; Lexer, M.J. Unraveling the drivers of intensifying forest disturbance regimes in Europe. Glob. Chang. Biol.
2011,17, 2842–2852. [CrossRef]

Forests 2021,12, 8 13 of 17
14.
Malhi, Y.; Aragão, L.E.O.C.; Galbraith, D.; Huntingford, C.; Fisher, R.; Zelazowski, P.; Sitch, S.; McSweeney, C.; Meir, P. Exploring
the likelihood and mechanism of a climate-change-induced dieback of the Amazon rainforest. Proc. Natl. Acad. Sci. USA
2009
,
106, 20610–20615. [CrossRef] [PubMed]
15.
Exbrayat, J.F.; Liu, Y.Y.; Williams, M. Impact of deforestation and climate on the Amazon Basin’s above-ground biomass during.
Sci. Rep. 2017,7, 1–7. [CrossRef] [PubMed]
16.
Morton, D.C.; Le Page, Y.; DeFries, R.; Collatz, G.J.; Hurtt, G.C. Understorey fire frequency and the fate of burned forests in
southern Amazonia. Philos. Trans. R. Soc. B Biol. Sci. 2013,368, 20120163. [CrossRef] [PubMed]
17.
Asner, G.P.; Knapp, D.E.; Broadbent, E.N.; Oliveira, P.J.C.; Keller, M.; Silva, J.N. Ecology: Selective logging in the Brazilian
Amazon. Science 2005,310, 480–482. [CrossRef] [PubMed]
18.
Stocker, B.D.; Zscheischler, J.; Keenan, T.F.; Prentice, I.C.; Seneviratne, S.I.; Peñuelas, J. Drought impacts on terrestrial primary
production underestimated by satellite monitoring. Nat. Geosci. 2019,12, 264–270. [CrossRef]
19.
Brodribb, T.J.; Powers, J.; Cochard, H.; Choat, B. Hanging by a thread? Forests and drought. Science
2020
,368, 261–266. [CrossRef]
[PubMed]
20.
Jiménez-Muñoz, J.C.; Mattar, C.; Barichivich, J.; Santamaría-Artigas, A.; Takahashi, K.; Malhi, Y.; Sobrino, J.A.; Schrier, G. Van Der
Record-breaking warming and extreme drought in the Amazon rainforest during the course of El Niño 2015–2016. Sci. Rep.
2016
,
6, 33130. [CrossRef] [PubMed]
21.
Marengo, J.A.; Espinoza, J.C. Extreme seasonal droughts and floods in Amazonia: Causes, trends and impacts. Int. J. Climatol.
2016,36, 1033–1050. [CrossRef]
22.
Nepstad, D.C.; Stickler, C.M.; Soares-Filho, B.; Merry, F. Interactions among Amazon land use, forests and climate: Prospects for a
near-term forest tipping point. Philos. Trans. R. Soc. B Biol. Sci. 2008,363, 1737–1746. [CrossRef] [PubMed]
23.
Marengo, J.A.; Souza, C.M.; Thonicke, K.; Burton, C.; Halladay, K.; Betts, R.A.; Alves, L.M.; Soares, W.R. Changes in Climate and
Land Use Over the Amazon Region: Current and Future Variability and Trends. Front. Earth Sci. 2018,6, 228. [CrossRef]
24.
Brando, P.M.; Balch, J.K.; Nepstad, D.C.; Morton, D.C.; Putz, F.E.; Coe, M.T.; Silvério, D.; Macedo, M.N.; Davidson, E.A.; Nóbrega,
C.C.; et al. Abrupt increases in Amazonian tree mortality due to drought-fire interactions. Proc. Natl. Acad. Sci. USA
2014
,111,
6347–6352. [CrossRef]
25.
Rappaport, D.I.; Morton, D.C.; Longo, M.; Keller, M.; Dubayah, R.; Dos-Santos, M.N. Quantifying long-term changes in carbon
stocks and forest structure from Amazon forest degradation. Environ. Res. Lett. 2018,13, 065013. [CrossRef]
26.
Walker, X.J.; Baltzer, J.L.; Cumming, S.G.; Day, N.J.; Ebert, C.; Goetz, S.; Johnstone, J.F.; Potter, S.; Rogers, B.M.; Schuur, E.A.G.; et al.
Increasing wildfires threaten historic carbon sink of boreal forest soils. Nature 2019,572, 520–523. [CrossRef] [PubMed]
27.
Laurance, W.F.; Nascimento, H.E.M.; Laurance, S.G.; Andrade, A.; Ribeiro, J.E.L.S.; Giraldo, J.P.; Lovejoy, T.E.; Condit, R.; Chave,
J.; Harms, K.E.; et al. Rapid decay of tree-community composition in Amazonian forest fragments. Proc. Natl. Acad. Sci. USA
2006,103, 19010–19014. [CrossRef]
28.
Barlow, J.; Gardner, T.A.; Lees, A.C.; Parry, L.; Peres, C.A. How pristine are tropical forests? An ecological perspective on the
pre-Columbian human footprint in Amazonia and implications for contemporary conservation. Biol. Conserv.
2012
,151, 45–49.
[CrossRef]
29.
Pickell, P.D.; Hermosilla, T.; Frazier, R.J.; Coops, N.C.; Wulder, M.A. Forest recovery trends derived from Landsat time series for
North American boreal forests. Int. J. Remote Sens. 2016,37, 138–149. [CrossRef]
30.
White, J.C.; Wulder, M.A.; Hermosilla, T.; Coops, N.C.; Hobart, G.W. A nationwide annual characterization of 25 years of forest
disturbance and recovery for Canada using Landsat time series. Remote Sens. Environ. 2017,194, 303–321. [CrossRef]
31.
Andrade, R.B.; Balch, J.K.; Parsons, A.L.; Armenteras, D.; Roman-Cuesta, R.M.; Bulkan, J. Scenarios in tropical forest degradation:
Carbon stock trajectories for REDD+. Carbon Balance Manag. 2017,12, 1–7. [CrossRef]
32.
Sato, L.Y.; Gomes, V.C.F.; Shimabukuro, Y.E.; Keller, M.; Arai, E.; dos-Santos, M.N.; Brown, I.F. Post-fire changes in forest biomass
retrieved by airborne LiDAR in Amazonia. Remote Sens. 2016,8, 839. [CrossRef]
33.
Barlow, J.; Peres, C.A.; Lagan, B.O.; Haugaasen, T. Large tree mortality and the decline of forest biomass following Amazonian
wildfires. Ecol. Lett. 2003,6, 6–8. [CrossRef]
34.
Balch, J.K.; Nepstad, D.C.; Curran, L.M.; Brando, P.M.; Portela, O.; Guilherme, P.; Reuning-Scherer, J.D.; de Carvalho, O. Size,
species, and fire behavior predict tree and liana mortality from experimental burns in the Brazilian Amazon. For. Ecol. Manag.
2011,261, 68–77. [CrossRef]
35.
Feldpausch, T.R.; Jirka, S.; Passos, C.A.M.; Jasper, F.; Riha, S.J. When big trees fall: Damage and carbon export by reduced impact
logging in southern Amazonia. For. Ecol. Manag. 2005,219, 199–215. [CrossRef]
36.
Marano, G.; Langella, G.; Basile, A.; Cona, F.; Michele, C.D.; Manna, P.; Teobaldelli, M.; Saracino, A.; Terribile, F. A geospatial
decision support system tool for supporting integrated forest knowledge at the landscape scale. Forests
2019
,10, 690. [CrossRef]
37.
Vacchiano, G.; Magnani, F.; Collalti, A. Modeling Italian forests: State of the art and future challenges. IForest
2012
,5, 113–120.
[CrossRef]
38.
Kumar, L.; Sinha, P.; Taylor, S.; Alqurashi, A.F. Review of the use of remote sensing for biomass estimation to support renewable
energy generation. J. Appl. Remote Sens. 2015,9, 097696. [CrossRef]
39.
Keeling, H.; Phillips, O. The global relationship between forest productivity and biomass. Glob. Ecol. Biogeogr.
2007
,16, 618–631.
[CrossRef]

Forests 2021,12, 8 14 of 17
40.
Rödig, E.; Cuntz, M.; Rammig, A.; Fischer, R.; Taubert, F.; Huth, A. The importance of forest structure for carbon fluxes of the
Amazon rainforest. Environ. Res. Lett. 2018,13, 054013. [CrossRef]
41.
Fauset, S.; Gloor, M.; Fyllas, N.M.; Phillips, O.L.; Asner, G.P.; Baker, T.R.; Patrick Bentley, L.; Brienen, R.J.W.; Christoffersen, B.O.;
del Aguila-Pasquel, J.; et al. Individual-Based Modeling of Amazon Forests Suggests That Climate Controls Productivity While
Traits Control Demography. Front. Earth Sci. 2019,7, 83. [CrossRef]
42.
Lewis, S.L.; Brando, P.M.; Phillips, O.L.; Van Der Heijden, G.M.F.; Nepstad, D. The 2010 Amazon drought. Science
2011
,331, 554.
[CrossRef] [PubMed]
43.
Berenguer, E.; Ferreira, J.; Gardner, T.A.; Aragão, L.E.O.C.; De Camargo, P.B.; Cerri, C.E.; Durigan, M.; De Oliveira, R.C.; Vieira,
I.C.G.; Barlow, J. A large-scale field assessment of carbon stocks in human-modified tropical forests. Glob. Chang. Biol.
2014
,20,
3713–3726. [CrossRef] [PubMed]
44.
De Faria, B.L.; Brando, P.M.; Macedo, M.N.; Panday, P.K.; Soares-Filho, B.S.; Coe, M.T. Current and future patterns of fire-induced
forest degradation in amazonia. Environ. Res. Lett. 2017,12, 095005. [CrossRef]
45.
Hansen, M.C.; Potapov, P.V.; Moore, R.; Hancher, M.; Turubanova, S.A.; Tyukavina, A.; Thau, D.; Stehman, S.V.; Goetz, S.J.;
Loveland, T.R.; et al. High-resolution global maps of 21st-century forest cover change. Science
2013
,342, 850–853. [CrossRef]
[PubMed]
46.
Wheeler, D.; Guzder-Williams, B.; Petersen, R.; Thau, D. Rapid MODIS-based detection of tree cover loss. Int. J. Appl. Earth
Obs. Geoinf. 2018,69, 78–87. [CrossRef]
47.
Landsberg, J.J.; Waring, R.H. A generalised model of forest productivity using simplified concepts of radiation-use efficiency,
carbon balance and partitioning. For. Ecol. Manag. 1997,95, 209–228. [CrossRef]
48.
Hirsch, A.I.; Little, W.S.; Houghton, R.A.; Scott, N.A.; White, J.D. The net carbon flux due to deforestation and forest re-growth in
the Brazilian Amazon: Analysis using a process-based model. Glob. Chang. Biol. 2004,10, 908–924. [CrossRef]
49. Coops, N.C.; Waring, R.H.; Landsberg, J.J. Assessing forest productivity in Australia and New Zealand using a physiologically-
based model driven with averaged monthly weather data and satellite-derived estimates of canopy photosynthetic capacity.
For. Ecol. Manag. 1998,104, 113–127. [CrossRef]
50.
Collalti, A.; Prentice, I.C. Is NPP proportional to GPP? Waring’s hypothesis 20 years on. Tree Physiol.
2019
,39, 1473–1483. [CrossRef]
51.
Collalti, A.; Ibrom, A.; Stockmarr, A.; Cescatti, A.; Alkama, R.; Fernández-Martínez, M.; Matteucci, G.; Sitch, S.; Friedlingstein, P.;
Ciais, P.; et al. Forest production efficiency increases with growth temperature. Nat. Commun. 2020. [CrossRef]
52.
Waring, R.H.; Landsberg, J.J.; Williams, M. Net primary production of forests: A constant fraction of gross primary production?
Tree Physiol. 1998,18, 129–134. [CrossRef]
53.
Collalti, A.; Perugini, L.; Santini, M.; Chiti, T.; Nolè, A.; Matteucci, G.; Valentini, R. A process-based model to simulate growth in
forests with complex structure: Evaluation and use of 3D-CMCC Forest Ecosystem Model in a deciduous forest in Central Italy.
Ecol. Model. 2014,272, 362–378. [CrossRef]
54.
Merganiˇcová, K.; Merganiˇc, J.; Lehtonen, A.; Vacchiano, G.; Sever, M.Z.O.; Augustynczik, A.L.D.; Grote, R.; Kyselová, I.; Mäkelä,
A.; Yousefpour, R.; et al. Forest carbon allocation modelling under climate change. Tree Physiol. 2019,39, 1937–1960. [CrossRef]
55.
Baccini, A.; Goetz, S.J.; Walker, W.S.; Laporte, N.T.; Sun, M.; Sulla-Menashe, D.; Hackler, J.; Beck, P.S.A.; Dubayah, R.; Friedl,
M.A.; et al. Estimated carbon dioxide emissions from tropical deforestation improved by carbon-density maps. Nat. Clim. Chang.
2012,2, 182–185. [CrossRef]
56.
Poorter, L.; Bongers, F.; Aide, T.M.; Almeyda Zambrano, A.M.; Balvanera, P.; Becknell, J.M.; Boukili, V.; Brancalion, P.H.S.;
Broadbent, E.N.; Chazdon, R.L.; et al. Biomass resilience of Neotropical secondary forests. Nature
2016
,530, 211–214. [CrossRef]
57.
Chambers, J.Q.; Tribuzy, E.S.; Toledo, L.C.; Crispim, B.F.; Higuchi, N.; Dos Santos, J.; Araújo, A.C.; Kruijt, B.; Nobre, A.D.;
Trumbore, S.E. Respiration from a tropical forest ecosystem: Partitioning of sources and low carbon use efficiency. Ecol. Appl.
2004,14, 72–88. [CrossRef]
58.
Collalti, A.; Marconi, S.; Ibrom, A.; Trotta, C.; Anav, A.; D’Andrea, E.; Matteucci, G.; Montagnani, L.; Gielen, B.; Mammarella, I. val-
idation of 3D-CMCC Forest Ecosystem Model (v.5.1) against eddy covariance data for ten European forest sites. Geosci. Model Dev.
2016,9, 479–504. [CrossRef]
59.
Almeida, A.C.; Landsberg, J.J.; Sands, P.J. Parameterisation of 3-PG model for fast-growing Eucalyptus grandis plantations.
For. Ecol. Manag. 2004,193, 179–195. [CrossRef]
60.
Avitabile, V.; Herold, M.; Heuvelink, G.B.M.; Lewis, S.L.; Phillips, O.L.; Asner, G.P.; Armston, J.; Ashton, P.S.; Banin, L.; Bayol, N.; et al.
An integrated pan-tropical biomass map using multiple reference datasets. Glob. Chang. Biol. 2016,22, 1406–1420. [CrossRef]
61.
Saatchi, S.S.; Harris, N.L.; Brown, S.; Lefsky, M.; Mitchard, E.T.A.; Salas, W.; Zutta, B.R.; Buermann, W.; Lewis, S.L.; Hagen,
S.; et al. Benchmark map of forest carbon stocks in tropical regions across three continents. Proc. Natl. Acad. Sci. USA
2011
,108,
9899–9904. [CrossRef]
62.
Phillips, O.L.; Aragão, L.E.O.C.; Lewis, S.L.; Fisher, J.B.; Lloyd, J.; López-González, G.; Malhi, Y.; Monteagudo, A.; Peacock, J.;
Quesada, C.A.; et al. Drought sensitivity of the amazon rainforest. Science 2009,323, 1344–1347. [CrossRef] [PubMed]
63.
Zemp, D.C.; Schleussner, C.F.; Barbosa, H.M.J.; Hirota, M.; Montade, V.; Sampaio, G.; Staal, A.; Wang-Erlandsson, L.; Rammig, A.
Self-amplified Amazon forest loss due to vegetation-atmosphere feedbacks. Nat. Commun. 2017,8, 1–10. [CrossRef]
64.
Aragão, L.E.O.C.; Malhi, Y.; Roman-Cuesta, R.M.; Saatchi, S.; Anderson, L.O.; Shimabukuro, Y.E. Spatial patterns and fire response
of recent Amazonian droughts. Geophys. Res. Lett. 2007,34. [CrossRef]

Forests 2021,12, 8 15 of 17
65.
Andela, N.; Morton, D.C.; Giglio, L.; Paugam, R.; Chen, Y.; Hantson, S.; Van Der Werf, G.R.; Anderson, J.T. The Global Fire Atlas
of individual fire size, duration, speed and direction. Earth Syst. Sci. Data 2019,11, 529–552. [CrossRef]
66.
Broadbent, E.N.; Asner, G.P.; Keller, M.; Knapp, D.E.; Oliveira, P.J.C.; Silva, J.N. Forest fragmentation and edge effects from
deforestation and selective logging in the Brazilian Amazon. Biol. Conserv. 2008,141, 1745–1757. [CrossRef]
67.
Harris, I.; Jones, P.D.; Osborn, T.J.; Lister, D.H. Updated high-resolution grids of monthly climatic observations-the CRU TS3.10
Dataset. Int. J. Climatol. 2014,34, 623–642. [CrossRef]
68.
Lee, H. Climate Algorithm Theoretical Basis Document (C-ATBD): Outgoing Longwave Radiation (OLR)-Daily. NOAA’s Climate Data
Record (CDR) Program, CDRP-ATBD-0526; Broadway: New York, NY, USA, 2014.
69.
Hunter, M.O.; Keller, M.; Victoria, D.; Morton, D.C. Tree height and tropical forest biomass estimation. Biogeosciences
2013
,10,
8385–8399. [CrossRef]
70.
Shao, G.; Stark, S.C.; de Almeida, D.R.A.; Smith, M.N. Towards high throughput assessment of canopy dynamics: The estimation
of leaf area structure in Amazonian forests with multitemporal multi-sensor airborne lidar. Remote Sens. Environ.
2019
,221, 1–13.
[CrossRef]
71.
Rex, F.E.; Silva, C.A.; Corte, A.P.D.; Klauberg, C.; Mohan, M.; Cardil, A.; da Silva, V.S.; de Almeida, D.R.A.; Garcia, M.;
Broadbent, E.N.; et al. Comparison of statistical modelling approaches for estimating tropical forest aboveground biomass
stock and reporting their changes in low-intensity logging areas using multi-temporal LiDAR data. Remote Sens.
2020
,12, 1498.
[CrossRef]
72.
Asner, G.P.; Powell, G.V.N.; Mascaro, J.; Knapp, D.E.; Clark, J.K.; Jacobson, J.; Kennedy-Bowdoin, T.; Balaji, A.; Paez-Acosta,
G.; Victoria, E.; et al. High-resolution forest carbon stocks and emissions in the Amazon. Proc. Natl. Acad. Sci. USA
2010
,107,
16738–16742. [CrossRef]
73.
Meyer, V.; Saatchi, S.; Ferraz, A.; Xu, L.; Duque, A.; García, M.; Chave, J. Forest degradation and biomass loss along the Chocó
region of Colombia. Carbon Balance Manag. 2019,14, 2. [CrossRef] [PubMed]
74.
Aguiar, A.P.D.; Vieira, I.C.G.; Assis, T.O.; Dalla-Nora, E.L.; Toledo, P.M.; Oliveira Santos-Junior, R.A.; Batistella, M.; Coelho, A.S.;
Savaget, E.K.; Aragão, L.E.O.C.; et al. Land use change emission scenarios: Anticipating a forest transition process in the Brazilian
Amazon. Glob. Chang. Biol. 2016,22, 1821–1840. [CrossRef] [PubMed]
75.
Taylor, K.E.; Stouffer, R.J.; Meehl, G.A. An overview of CMIP5 and the experiment design. Bull. Am. Meteorol. Soc.
2012
,93,
485–498. [CrossRef]
76.
Fonseca, M.G.; Alves, L.M.; Aguiar, A.P.D.; Arai, E.; Anderson, L.O.; Rosan, T.M.; Shimabukuro, Y.E.; de Aragão, L.E.O.E.C.
Effects of climate and land-use change scenarios on fire probability during the 21st century in the Brazilian Amazon. Glob. Chang.
Biol. 2019,25, 2931–2946. [CrossRef] [PubMed]
77.
Elias, F.; Ferreira, J.; Lennox, G.D.; Berenguer, E.; Ferreira, S.; Schwartz, G.; de Oliveira Melo, L.; Reis Júnior, D.N.; Nascimento,
R.O.; Ferreira, F.N.; et al. Assessing the growth and climate sensitivity of secondary forests in highly deforested Amazonian
landscapes. Ecology 2020,101, e02954. [CrossRef]
78.
Schwalm, C.R.; Anderegg, W.R.L.; Michalak, A.M.; Fisher, J.B.; Biondi, F.; Koch, G.; Litvak, M.; Ogle, K.; Shaw, J.D.; Wolf, A.; et al.
Global patterns of drought recovery. Nature 2017,548, 202–205. [CrossRef]
79.
Engelbrecht, B.M.J.; Comita, L.S.; Condit, R.; Kursar, T.A.; Tyree, M.T.; Turner, B.L.; Hubbell, S.P. Drought sensitivity shapes
species distribution patterns in tropical forests. Nature 2007,447, 80–82. [CrossRef]
80.
Saatchi, S.; Asefi-Najafabady, S.; Malhi, Y.; Aragão, L.E.O.C.; Anderson, L.O.; Myneni, R.B.; Nemani, R. Persistent effects of a
severe drought on Amazonian forest canopy. Proc. Natl. Acad. Sci. USA 2013,110, 565–570. [CrossRef]
81.
Malhi, Y.; Roberts, J.T.; Betts, R.A.; Killeen, T.J.; Li, W.; Nobre, C.A. Climate change, deforestation, and the fate of the Amazon.
Science 2008,319, 169–172. [CrossRef]
82.
Malhi, Y.; Doughty, C.E.; Goldsmith, G.R.; Metcalfe, D.B.; Girardin, C.A.J.; Marthews, T.R.; del Aguila-Pasquel, J.; Aragão,
L.E.O.C.; Araujo-Murakami, A.; Brando, P.; et al. The linkages between photosynthesis, productivity, growth and biomass in
lowland Amazonian forests. Glob. Chang. Biol. 2015,21, 2283–2295. [CrossRef]
83.
Collalti, A.; Trotta, C.; Keenan, T.F.; Ibrom, A.; Bond-Lamberty, B.; Grote, R.; Vicca, S.; Reyer, C.P.O.; Migliavacca, M.; Veroustraete,
F.; et al. Thinning Can Reduce Losses in Carbon Use Efficiency and Carbon Stocks in Managed Forests Under Warmer Climate.
J. Adv. Model. Earth Syst. 2018,10, 2427–2452. [CrossRef]
84.
Joetzjer, E.; Douville, H.; Delire, C.; Ciais, P. Present-day and future Amazonian precipitation in global climate models: CMIP5
versus CMIP3. Clim. Dyn. 2013,41, 2921–2936. [CrossRef]
85.
Boisier, J.P.; Ciais, P.; Ducharne, A.; Guimberteau, M. Projected strengthening of Amazonian dry season by constrained climate
model simulations. Nat. Clim. Chang. 2015,5, 656–660. [CrossRef]
86.
Duffy, P.B.; Brando, P.; Asner, G.P.; Field, C.B. Projections of future meteorological drought and wet periods in the Amazon.
Proc. Natl. Acad. Sci. USA 2015,112, 13172–13177. [CrossRef]
87.
Cai, W.; Borlace, S.; Lengaigne, M.; Van Rensch, P.; Collins, M.; Vecchi, G.; Timmermann, A.; Santoso, A.; Mcphaden, M.J.;
Wu, L.; et al. Increasing frequency of extreme El Niño events due to greenhouse warming. Nat. Clim. Chang.
2014
,4, 111–116.
[CrossRef]
88.
Lau, W.K.M.; Kim, K.-M. Robust Hadley Circulation changes and increasing global dryness due to CO
2
warming from CMIP5
model projections. Proc. Natl. Acad. Sci. USA 2015,112, 3630–3635. [CrossRef]

Forests 2021,12, 8 16 of 17
89.
Nepstad, D.C.; Tohver, I.M.; David, R.; Moutinho, P.; Cardinot, G. Mortality of large trees and lianas following experimental
drought in an amazon forest. Ecology 2007,88, 2259–2269. [CrossRef]
90.
Poulter, B.; Hattermann, F.; Hawkins, E.; Zaehle, S.; Sitch, S.; Restrepo-Coupe, N.; Heyder, U.; Cramer, W. Robust dynamics
of Amazon dieback to climate change with perturbed ecosystem model parameters. Glob. Chang. Biol.
2010
,16, 2476–2495.
[CrossRef]
91.
Nepstad, D.; Lefebvre, P.; Da Silva, U.L.; Tomasella, J.; Schlesinger, P.; Solórzano, L.; Moutinho, P.; Ray, D.; Benito, J.G. Amazon
drought and its implications for forest flammability and tree growth: A basin-wide analysis. Glob. Chang. Biol.
2004
,10, 704–717.
[CrossRef]
92.
Liu, J.; Vogelmann, J.E.; Zhu, Z.; Key, C.H.; Sleeter, B.M.; Price, D.T.; Chen, J.M.; Cochrane, M.A.; Eidenshink, J.C.; Howard,
S.M.; et al. Estimating California ecosystem carbon change using process model and land cover disturbance data: 1951–2000.
Ecol. Model. 2011,222, 2333–2341. [CrossRef]
93.
Doughty, C.E.; Metcalfe, D.B.; Girardin, C.A.J.; Amézquita, F.F.; Cabrera, D.G.; Huasco, W.H.; Silva-Espejo, J.E.; Araujo-Murakami,
A.; Da Costa, M.C.; Rocha, W.; et al. Drought impact on forest carbon dynamics and fluxes in Amazonia. Nature
2015
,519, 78–82.
[CrossRef] [PubMed]
94.
Yang, Y.; Saatchi, S.S.; Xu, L.; Yu, Y.; Choi, S.; Phillips, N.; Kennedy, R.; Keller, M.; Knyazikhin, Y.; Myneni, R.B. Post-drought
decline of the Amazon carbon sink. Nat. Commun. 2018,9, 1–9. [CrossRef] [PubMed]
95.
De Faria, B.L.; Staal, A.; Martin, P.A.; Panday, P.K.; Castanho, A.D.; Dantas, V.L. Climate change and deforestation boost post-fire
grass invasion of Amazonian forests. bioRxiv 2019. [CrossRef]
96.
Jolly, W.M.; Cochrane, M.A.; Freeborn, P.H.; Holden, Z.A.; Brown, T.J.; Williamson, G.J.; Bowman, D.M.J.S. Climate-induced
variations in global wildfire danger from 1979 to 2013. Nat. Commun. 2015,6, 1–11. [CrossRef] [PubMed]
97.
Cardil, A.; de-Miguel, S.; Silva, C.A.; Reich, P.B.; Calkin, D.E.; Brancalion, P.H.S.; Vibrans, A.C.; Gamarra, J.G.P.; Zhou, M.;
Pijanowski, B.C.; et al. Recent deforestation drove the spike in Amazonian fires. Environ. Res. Lett. 2020,15, 121003. [CrossRef]
98.
Vanderwel, M.C.; Coomes, D.A.; Purves, D.W. Quantifying variation in forest disturbance, and its effects on aboveground
biomass dynamics, across the eastern United States. Glob. Chang. Biol. 2013,19, 1504–1517. [CrossRef]
99.
Jin, W.; He, H.S.; Thompson, F.R. Are more complex physiological models of forest ecosystems better choices for plot and regional
predictions? Environ. Model. Softw. 2016,75, 1–14. [CrossRef]
100.
Collalti, A.; Thornton, P.E.; Cescatti, A.; Rita, A.; Borghetti, M.; Nolè, A.; Trotta, C.; Ciais, P.; Matteucci, G. The sensitivity of
the forest carbon budget shifts across processes along with stand development and climate change. Ecol. Appl.
2019
,29, 1–18.
[CrossRef]
101.
Collalti, A.; Tjoelker, M.G.; Hoch, G.; Mäkelä, A.; Guidolotti, G.; Heskel, M.; Petit, G.; Ryan, M.G.; Battipaglia, G.; Matteucci, G.; et al.
Plant respiration: Controlled by photosynthesis or biomass? Glob. Chang. Biol. 2020,26, 1739–1753. [CrossRef]
102.
Swann, A.L.S.; Hoffman, F.M.; Koven, C.D.; Randerson, J.T. Plant responses to increasing CO
2
reduce estimates of climate impacts
on drought severity. Proc. Natl. Acad. Sci. USA 2016,113, 10019–10024. [CrossRef]
103.
Holtum, J.A.M.; Winter, K. Elevated [CO
2
] and forest vegetation: More a water issue than a carbon issue? Funct. Plant Biol.
2010
,
37, 694–702. [CrossRef]
104.
Jiang, M.; Medlyn, B.E.; Drake, J.E.; Duursma, R.A.; Anderson, I.C.; Barton, C.V.M.; Boer, M.M.; Carrillo, Y.; Castañeda-Gómez, L.;
Collins, L.; et al. The fate of carbon in a mature forest under carbon dioxide enrichment. Nature
2020
,580, 227–231. [CrossRef]
[PubMed]
105.
Hofhansl, F.; Andersen, K.M.; Fleischer, K.; Fuchslueger, L.; Rammig, A.; Schaap, K.J.; Valverde-Barrantes, O.J.; Lapola, D.M.
Amazon forest ecosystem responses to elevated atmospheric CO
2
and alterations in nutrient availability: Filling the gaps with
model-experiment integration. Front. Earth Sci. 2016,4, 19. [CrossRef]
106.
Hubau, W.; Lewis, S.L.; Phillips, O.L.; Affum-Baffoe, K.; Beeckman, H.; Cuní-Sanchez, A.; Daniels, A.K.; Ewango, C.E.N.; Fauset,
S.; Mukinzi, J.M.; et al. Asynchronous carbon sink saturation in African and Amazonian tropical forests. Nature
2020
,579, 80–87.
[CrossRef]
107.
Longo, M.; Keller, M.; dos-Santos, M.N.; Leitold, V.; Pinagé, E.R.; Baccini, A.; Saatchi, S.; Nogueira, E.M.; Batistella, M.; Morton,
D.C. Aboveground biomass variability across intact and degraded forests in the Brazilian Amazon. Glob. Biogeochem. Cycles
2016
,
30, 1639–1660. [CrossRef]
108.
Longo, M.; Knox, R.G.; Medvigy, D.M.; Levine, N.M.; Dietze, M.C.; Kim, Y.; Swann, A.L.S.; Zhang, K.; Rollinson, C.R.; Bras,
R.L.; et al. The biophysics, ecology, and biogeochemistry of functionally diverse, vertically and horizontally heterogeneous
ecosystems: The Ecosystem Demography model, version 2.2-Part 1: Model description. Geosci. Model. Dev. Discuss.
2019
,12,
4309–4346. [CrossRef]
109.
Longo, M.; Saatchi, S.; Keller, M.; Bowman, K.; Ferraz, A.; Moorcroft, P.R.; Morton, D.C.; Bonal, D.; Brando, P.; Burban, B.; et al.
Impacts of Degradation on Water, Energy, and Carbon Cycling of the Amazon Tropical Forests. J. Geophys. Res. Biogeosci.
2020
,
125, e2020JG005677. [CrossRef]
110.
Dubayah, R.; Blair, J.B.; Goetz, S.; Fatoyinbo, L.; Hansen, M.; Healey, S.; Hofton, M.; Hurtt, G.; Kellner, J.; Luthcke, S.; et al. The
Global Ecosystem Dynamics Investigation: High-resolution laser ranging of the Earth’s forests and topography. Sci. Remote Sens.
2020,1, 100002. [CrossRef]

Forests 2021,12, 8 17 of 17
111.
Potapov, P.; Li, X.; Hernandez-Serna, A.; Tyukavina, A.; Hansen, M.C.; Kommareddy, A.; Pickens, A.; Turubanova, S.; Tang, H.;
Silva, C.E.; et al. Mapping global forest canopy height through integration of GEDI and Landsat data. Remote Sens. Environ.
2020
,
112165. [CrossRef]
112.
Duncanson, L.; Neuenschwander, A.; Hancock, S.; Thomas, N.; Fatoyinbo, T.; Simard, M.; Silva, C.A.; Armston, J.; Luthcke, S.B.;
Hofton, M.; et al. Biomass estimation from simulated GEDI, ICESat-2 and NISAR across environmental gradients in Sonoma
County, California. Remote Sens. Environ. 2020,242, 111779. [CrossRef]