TYPE Original Research

PUBLISHED 23 September 2022

DOI 10.3389/frwa.2022.886558

Dry season rainfall as a source of

OPEN ACCESS transpired water in a seasonal,
evergreen forest in the western
EDITED BY
Matthias Sprenger,
Berkeley Lab (DOE), United States

REVIEWED BY
Laura Andrea Benegas Negri,
Amazon region inferred by
Centro Agronomico Tropical de
Investigacion y Ensenanza Catie,
Costa Rica
water stable isotopes
Jia Hu,
University of Arizona, United States
Laura De Simone Borma1*, Wilian Carlo Demetrio1 ,
*CORRESPONDENCE
Laura De Simone Borma Ranieli Dos Anjos De Souza2,3 , Anne Verhoef4 , Alberto Webler5
laura.borma@inpe.br
and Renata Gonçalves Aguiar5
SPECIALTY SECTION
1
This article was submitted to Division of Impact, Adaptation and Vulnerability (DIIAV) of the National Institute for Space Research
Water and Critical Zone, (INPE), Sao Jose dos Campos, São Paulo, Brazil, 2 Division of Earth Observation (DIOTG) of the
a section of the journal National Institute for Space Research (INPE), Sao Jose dos Campos, São Paulo, Brazil, 3 Space
Frontiers in Water Research Group (GREES) of the Federal Institute of Education, Science and Technology of Rondônia
(IFRO), Colorado do Oeste, Brazil, 4 Department of Geography and Environmental Science, The
RECEIVED 28 February 2022
University of Reading, Reading, United Kingdom, 5 Department of Environmental Engineering,
ACCEPTED 16 August 2022
Federal University of Rondonia (UNIR), Ji-Paraná, Brazil
PUBLISHED 23 September 2022

CITATION
Borma LDS, Demetrio WC, Souza
RDAD, Verhoef A, Webler A and The present work aimed to investigate the potential sources of water for plants
Aguiar RG (2022) Dry season rainfall as in an area of evergreen forest located in western Amazonia (Rebio Jaru). We
a source of transpired water in a
used a natural abundance of water isotopes—δ2 H and δ18 O—to trace the main
seasonal, evergreen forest in the
western Amazon region inferred by source of water to plants at the beginning of the dry period (May 2016) and at
water stable isotopes. the end of the dry period/transition to the wet period (October 2016) following
Front. Water 4:886558.
doi: 10.3389/frwa.2022.886558
a severe El Niño drought (ENSO 2015/16). Soil samples were collected in a soil
profile up to 4 m depth. Plant samples from 18 trees (14 species) were collected
COPYRIGHT
© 2022 Borma, Demetrio, Souza, in May and in October 2016. Rainwater and river water samples were collected
Verhoef, Webler and Aguiar. This is an between September 2015 and February 2017. We found that, at the end of the
open-access article distributed under
the terms of the Creative Commons
dry period/transition to the wet period (i.e., October 2016), the average plant
Attribution License (CC BY). The use, xylem signal was more enriched (δ2 H: −20.0 ± 8.1‰; δ18 O: −1.13 ± 1.88‰)
distribution or reproduction in other than in May 2016 (δ2 H: −36.7 ± 5.6‰; δ18 O: −3.50 ± 1.30‰), the onset of the
forums is permitted, provided the
original author(s) and the copyright dry period. The averaged isotopic soil signal in May 2016 (δ2 H: −35.4 ± 5.90‰;
owner(s) are credited and that the δ18 O: −5.19 ± 0.70‰) is slightly more depleted than in October (δ2 H: −27.6 ±
original publication in this journal is
cited, in accordance with accepted
13.8‰; δ18 O: −4.35 ± 1.73‰) and, in general, more depleted than the xylem
academic practice. No use, distribution signal. In the dual isotope space, the xylem signal at the beginning of the dry
or reproduction is permitted which period follows the rainfall signal of the wet period, while the xylem signal at
does not comply with these terms.
the end of the dry period/transition to the wet period follows the signal of the
dry season rainfall, suggesting that plants mostly transpire recent rainwater.
Contrary to what was expected, we did not find evidence in the xylem signal
of the water stored in the soil pores, which suggests that to meet to the water
demands of the dry period, plants do not use the water from past periods stored
in the soil layers.

KEYWORDS

Amazon forest, plant transpiration, root water uptake, droughts, water isotopes

Frontiers in Water 01 frontiersin.org

Borma et al. 10.3389/frwa.2022.886558

Introduction El Niño drought (May 2016 to October 2016) (Souza et al., 2022).
It suggests that the sampled trees have mechanisms to deal with
An improved understanding of how the Amazon forest severe droughts, without a significant impact on the canopy
behaves under seasonal and prolonged droughts is crucial to structure and, consequently, on ET rates and photosynthesis
help predict the future of the Amazon forest when it faces a (Souza et al., 2022). However, questions still arise about the
drier and warmer climate, as is projected to occur as a result potential sources of water that maintain elevated ET rates at this
of local and global anthropogenic drivers (Seneviratne et al., site and the Amazon forest in general.
2013; Berg et al., 2016). Observations at several Amazonian The natural abundance of heavy water isotopes (2 H and
experimental sites indicate dry-season evapotranspiration rates 18 O) has been used to infer the potential water sources for

(ET) and forest productivity comparable to, or even exceeding, plant transpiration (Ehleringer and Dawson, 1992) and, more
wet season values (Shuttleworth, 1988; Restrepo-Coupe et al., recently, to track the movement of water in the vadose zone
2013; Wu et al., 2017). Some studies also suggest that Amazon (Sprenger et al., 2016). To investigate the potential water sources
forests are resilient to extreme droughts (Saleska et al., 2016), which sustain elevated EVI and the ET rates in the Rebio Jaru
thus contradicting current perceptions and simulation results site, during the dry period of the 2015/2016 El Niño drought,
from most Earth system models, which show a decrease in ET we used the natural abundance of stable isotopes of water—2 H
and productivity during the dry season and drought periods, as and 18 O. The use of water isotopes to trace plant water sources
a result of limited water availability (Christoffersen et al., 2014). is usually less common in tropical humid regions such as the
In the Amazon, the deep root water uptake (deep-RWU) Amazon forest than in arid regions (Sprenger et al., 2016; Sohel
has been widely identified to explain the forest resilience to et al., 2021). This is because, in these environments, the soil
droughts (Nepstad et al., 1994; Jipp et al., 1998; Bruno et al., isotopic signal is usually diluted by frequent infiltration of rain
2006) and the maintenance of relatively high ET values during waters with different isotopic signatures; making the distinction
seasonal dry periods (Shuttleworth, 1998). It has been proposed between plant water sources stemming from different layers in
that the larger trees, which have the highest transpiration rates, the soil profile difficult (Sprenger et al., 2016). Here, we had the
would be mainly responsible for the higher ET rates (Kunert advantage that samples were taken during the dry period when
et al., 2017). However, challenging deep-RWU as a mechanism rainfall amount and rainwater infiltration in the soil was strongly
of resilience to droughts, some studies have shown that the larger reduced. We combined information on soil physical properties,
trees, which potentially have the deeper root systems, presented soil moisture, natural abundance of water isotopes at the wet-
higher mortality rates during experimental (Nepstad et al., 2007) to-dry (WTD) transition period (May 2016) and at the dry-to-
and natural extreme droughts (Phillips et al., 2010). In fact, in wet (DTW) transition period (October 2016), and leaf water
experiments where the wet season rainfall was suppressed by potential at midday (ψmd ) to test the hypothesis that, to attend
plastic panels placed above the canopy, the larger trees (diameter the atmospheric and plant water demands of the dry period,
at breast height, DBH > 10 cm) and the lianas presented the plants take up water from the deeper soil layer (i.e., >1 m depth).
higher mortality rates after 2 years of rainfall exclusion (Nepstad We discuss the obtained results in terms of potential sources of
et al., 2007). water for plants, and the role of the water stress in the stem in
Severe droughts are part of the natural climate pattern of the the enrichment of the plant xylem observed in the peak of the
Amazon (Marengo, 2004) and their occurence can be used to dry period.
increase our understanding of the Amazon forest’s response to
droughts. Long-term meteorological measurements suggest that
the 2015/2016 El Niño drought was the warmest observed to Materials and methods
date period in the Amazon basin (Jiménez-Muñoz et al., 2016).
In the Biological Reserve of Jaru (Rebio Jaru—RJA site), a very Study area
seasonal experimental site located in the southwestern Amazon,
Souza et al. (2022) investigated the forest response to the The study site is located at Jaru Biological Reserve (Rebio
2015/2016 El Niño drought by measuring enhanced vegetation Jaru), 10◦ 11′ 11.4′′ S; 61◦ 52′ 29.9′′ W, a national conservation area
indices (EVI), green chromatic coordinate (Gcc) vegetation in the Rondônia state, Amazonas, at situated around 290 km
indices, and plant physiological traits, such as the potential at from Porto Velho, the capital. The Rebio Jaru has a total area
which plants lose 50% of the hydraulic conductance (P50 ), turgor of 353,163 ha (IBAMA, 2007). Inside the reserve, flux tower
loss point (πTLP), and hydraulic safety margin (HSM), which was installed, owned by the LBA (Large Biosphere-Atmosphere
characterizes plant resistance to droughts (Sperry et al., 2002). Experiment in Amazonia) and operating since 1999 (Webler
They found that the tropical forest at this experimental site is et al., 2007). At around 80 m far from the LBA flux tower was
formed by species with low to moderate resilience to droughts. installed, in 2016, a permanent plot composed of 25 subplots of
However, only a small reduction in the EVI (from 0.52 to 0.48; 20 × 25 m (Da Silva, 2021). Our soil and plant samples were
dimensionless) was found during the dry period of the 2015/16 collected from one of these subplots.

Frontiers in Water 02 frontiersin.org

Borma et al. 10.3389/frwa.2022.886558

Rebio Jaru Reserve contains a forest characterized by a dense occurred between May and September (2016) (Figure 1A).
evergreen forest. The canopy average height is 20 ± 6.7 m, The driest month in our sampling period was July 2016,
with emergent trees of about 44 m. Based on NOAA’s 34-year when no rainfall was registered. The lowest soil moisture
historical series, the site’s average annual rainfall is 1,923 mm, conditions along the first meter of the soil profile were
with a dry period of 5 months typically occurring between May recorded in September 2015 and September 2016 (Figure 1B).
and September (Sombroek, 2001). In terms of rainfall pattern, In the present work, isotopic measurements in soil and xylem
this site is considered one of the most seasonal LBA sites, with water were performed in May and October 2016, which
a longer dry period than the other LBA sites (Restrepo-Coupe are considered, respectively, as the WTD and the DTW
et al., 2013). transition periods.

Samples collection Soil physical properties and plant traits

Disturbed and undisturbed soil samples were collected The measured soil physical properties comprise: (i) soil
from a soil profile up to 4 m deep (at 0.05, 0.1, 0.15, 0.2, texture, determined according to the Bouyoucos method
0.4, 0.8, 1.2, 1.6, 2, 3, and 4 m). Undisturbed soil samples (Bouyoucos, 1927; Gee and Bauder, 1986); (ii) soil dry bulk
were used to determine soil water retention curves (SWRCs), density, ρd , determined from the ratio of the mass of oven-
while disturbed soil samples were used to determine the soil dried soil (at 105◦ C for 48 h) and the volume of the soil core
physical properties and for isotopic analysis. Plant samples samples; (iii) soil-water retention curves (SWRCs), obtained
for isotopic analyses were collected from 18 individuals using a tension table (for soil suctions of 0, 1, 2, 4, 6,
from 14 different plant species located in the study area 10, 30, 50, 100, and 500 kPa) and pressure chamber (for
(Supplementary Table 1) (Souza et al., 2022). For isotopic 1,500 kPa), and (iv) total porosity (ϕ), determined by 1 –
analysis, soil and plant samples were collected in two periods— ρd /ρs , with the density of soil particles, ρs , taken as 2.65
May 2016 and October 2016. To collect soil samples in these g cm−3 .
two different periods, we used two different pits <5 m apart. Soil cation exchange capacity (CEC) was obtained by
To avoid evaporation and fractionation, for each tree, we the sum of exchangeable contents of Ca, Mg, Al (extracted
used suberized twig segments (i.e., 1–2 cm diameter), from by KCl 1 mol L−1 ), and K (Mehlich−1 ) following standard
which we removed the bark before storing the sample to methodologies (Teixeira et al., 2017). Soil organic matter was
avoid mixing between the phloem and xylem water. Plant obtained using the Walkley–Black method (Walkley and Black,
and soil samples were quickly sealed in vials, tightly wrapped 1934).
with parafilm, kept refrigerated in the field, and frozen in The SWRCs for each soil layer were fitted using the van
the laboratory until water extraction. Rainwater samples were Genuchten (1980) equation
collected during rainfall events between September 2015 and
October 2016. Seven water samples were collected in the θs − θr
θ = θr +   
n 1−1/n (1)
Machado River (March and June), 1.2 km away from the EC 1 + α h
flux tower. Rainfall samples were not mixed, so the data
presented here refer to event-specific values and not to averaged where θ is the volumetric water content (cm3 cm−3 ), and θs
monthly values. and θr are the saturated and residual water content (cm3 cm−3 ),
respectively; h is the soil suction (hPa), and α (hPa−1 ) and n are
empirical parameters. The van Genuchten parameters required
Environmental conditions in Equation (1) (θs , θr , α, n) were found by fitting 9 h-θ data-
pairs (one set for each sampled layer) with the excel solver (non-
To characterize the environmental conditions during the linear least-squares fitting) to find the best adjustment between
experiment, we used precipitation (P) and soil moisture (θ) the observed and the estimated soil moisture. Parameters were
data obtained from the EC flux tower between September 2015 calculated for each layer, with r2 -values >0.97 for all layers
and October 2016. Precipitation amounts were automatically (Supplementary Table 2).
measured by a rain gauge (Environmental Measurements Total plant-available water (TAW), in cm3 cm−3 , was
Ltd.—Arg 100). Soil moisture content was measured using calculated as the difference between the volumetric water
Timing Domain Reflectometers (TDR), installed at 0.2, 0.4, content at the field capacity (θFC ) and the water content at the
0.6, 0.8, and 1 m depths. These data were collected and permanent wilting point (θPWP ), with FC taken at a suction, h,
treated by the LBA office at Rebio Jaru. Evapotranspiration of 33 kPa and PWP at a suction of 1,500 kPa (1.5 MPa). From the
data were obtained from MODIS (Muler, 2018). The dry fitted SWRCs (Supplementary Figure 1) for each soil depth, we
period, that is, months where P < 100 mm (Sombroek, 2001), derived four classes of pore sizes: macropores (Ma, equivalent

Frontiers in Water 03 frontiersin.org

Borma et al. 10.3389/frwa.2022.886558

FIGURE 1
Environmental conditions in the study area from September 2015 to December 2016. (A) Averaged monthly precipitation (P, mm; blue bars) and
ET rates (mm; green bars). (B) Daily precipitation rates (mm), and (C) soil water content (θ, cm3 cm−3 ) at the upper meter of the soil profile. ET
data from MODIS (MOD16A2; Souza et al., 2022); monthly and daily P, and θ obtained from the LBA data from the Rebio Jaru EC flux tower.

pore diameter, φ, > 300 µm), mesopores (Me, 300 < φ < pores” (Tr), while φ between 50 and 0.2 µm (i.e., Mi + Cr; h >
50 µm), micropores (Mi, 50 < φ < 0.2 µm), and cryptopores 6 kPa) refers to water “storage pores” (St pores). To compare the
(Cr, φ < 0.2 µm) (Teixeira et al., 2017). soil’s physical properties along the 4 m soil profile, we computed
According to the capillarity equation (Washburn, 1921; TAW and pore size distributions as a percentage of the total
Rowell, 1994): porosity (ϕ).
Plant traits such as tree height (H, m) and circumference at
h = 0.15/r (2) breast height (CBH, cm) were measured with a tape measure.
Circumference at breast height was transformed into DBH (cm)
where h is the suction in (Pa) and r is the pore radius in m, by dividing CBH by π . The leaf water potential at midday (ψmd ,
for macroporosity, h < 1 kPa; mesoporosity 1 < h < 6 kPa; MPa), which is an indicator of plant water stress, was measured
microporosity 6 < h < 1,500 kPa, and for Cr, h > 1,500 kPa. with a Scholander camera (PMS Instruments Co., Albany, NY,
According to Rowell (1994), soil pores diameters (φ) of 50 µm or USA). These measurements were taken in October 2016, the
larger (i.e., Ma+Me; h < 6 kPa) corresponds to the “transmission DTW transition period.

Frontiers in Water 04 frontiersin.org

Borma et al. 10.3389/frwa.2022.886558

TABLE 1 Physical and physic-chemical soil properties along the upper 4 m of the soil profile.

Soil depth (m) Soil texture Chemical properties

Clay Silt Sand OM CEC ρd ϕ
% % cmolc dm−3 g cm−3 cm3 cm−3

0.2 60.3 6.5 33.3 1.61 1.64 1.46 0.480

0.4 60.9 5.5 33.6 1.21 1.49 1.30 0.531
0.8 61.5 5.3 33.2 0.81 1.58 1.42 0.496
1.2 63.1 4.7 32.2 0.81 1.11 1.65 0.451
1.6 60.9 6.0 33.1 0.67 1.20 1.54 0.480
2.0 66.3 7.3 26.4 0.54 0.80 1.47 0.470
2.4 70.3 9.2 20.6 0.13 0.84 1.51 0.480
3.0 61.7 5.6 32.7 0.00 1.20 1.54 0.459
4.0 61.6 7.5 30.9 0.00 1.10 1.62 0.400

Soil physical properties comprise soil texture, dry bulk density (ρd ; gcm−3 ), and total porosity (ϕ (cm3 cm−3 ). Soil physic-chemical properties comprise organic matter content (OM; %),
cation exchange capacity (CEC, cmolc dm−3 ).

Isotopic analysis dual isotopic space. We used Pearson’s correlation to test the
correlation between plant traits such as DBH, H, ψmd , soil
The isotopic ratios of H (δ2 H) and O (δ18 O) (Coplen, physical properties (i.e., TAW and soil porosity), and δ18 O
2011) of the collected and extracted waters were analyzed and δ12 H.
by laser absorption spectroscopy using a Picarro Li2130. We
used two different standards—PLRM1 (δ2 H: 16.9‰ and δ18 O:
1.65‰) and PLRM2 (δ2 H: −123.1‰ and δ18 O: −16.52‰)— Results
and one Quality Control (δ2 H: 46‰ and δ18 O: 7.25‰),
Soil physical properties
previously calibrated with the V-SMOW (Vienna Standard
Mean Ocean Water).
Our results point to a very clayey soil, with clay content
Deuterium excess (d-excess) was calculated according to
>60% along the entire 4 m profile, and sand content of ∼30%,
(Dansgaard et al., 1964):
except at 2 and 2.4 m, where sand content decreases to values
dexcess = δ 2 H + 8δ 18 O (3) between 20% and 30% (Table 1). The soil presents a relatively
high bulk density (ρd ) > 1.3 g cm−3 , with a distinctly denser
Soil and plant water samples were extracted using the layer (ρd > 1.65 g cm−3 ) occurring at 1.2 m depth (Table 1).
cryogenic distillation method. To prevent the influence of Total soil porosity (8) is high relatively, with the lowest value
organic contamination in the isotopic results, we immersed (0.4 cm3 cm−3 ) occurring at 4 m depth. The soil porosity is
activated charcoal for 48 h in the extracted soil and plant water mostly (>50%) formed by Cr pores (i.e., pores with φ < 0.2 µm;
samples. Besides that, the Picarro CRDS L2130-i instrument Table 2) where water is retained at suctions >1.5 MPa (Rowell,
used has a built-in software—Post Process Chemcorrect—that 1994). Consequently, this soil presents a high water retention
flags any sample with organic contamination. If samples flag capacity, with θPWP > 0.3 cm3 cm−3 for most of the soil layers
once, they are re-analyzed; and if they flag two times, isotopic and St pores >70% along the entire 4-m soil profile. The highest
analysis is done in the isotope ratio mass spectrometry (IRMS). values of θPWP and St pores occur between 1.6 and 3 m depths
During our study, no sample was required to be re-analyzed or (Figure 2; Table 2). Van Genuchten parameters for each layer are
transferred to the IRMS. presented in Supplementary Table 2.

Data analysis Isotopic analysis

In the dual-isotope space (δ2 H × δ18 O), we used analysis During our sampling period, the isotopic signal of rainfall
of covariance (ANCOVA) to identify differences among the water ranged between −82.2 and 36.7‰ for δ2 H and between
slopes of Global Meteorological Water Line (GMWL), Local −10.1 and 12.2‰ for δ18 O (Figures 3A,B; Table 3). δ2 Hxylem
Meteorological Water Line (LMWL), River Water Line (RWL), ranged from −46.6 to −27.3‰ and δ18 Oxylem ranged from −5.0
Plant Water Line (PWL), and Soil Water Line (SWL) on the to −0.2‰ in May, while in October, xylem water presented

Frontiers in Water 05 frontiersin.org

Borma et al. 10.3389/frwa.2022.886558

TABLE 2 Soil properties, expressed in terms of water retention properties and soil porosity for the upper 4 m of the soil profile.

Soil depth (m) Water retention properties Soil pore types Rowell
θFC θPWP TAW TAW Ma Me Mi Cr St Tr
cm3 cm−3 cm3 cm−3 % % %

0.2 0.364 0.322 0.0423 8.8 5.8 10.0 17.1 67.1 84.2 15.8
0.4 0.342 0.288 0.0540 10.2 17.5 10.5 17.7 54.3 72.0 28.1
0.8 0.346 0.301 0.0456 9.2 10.7 11.3 17.3 60.8 78.1 22.0
1.2 0.255 0.227 0.0285 6.3 21.7 14.0 14.0 50.4 64.3 35.7
1.6 0.383 0.347 0.0363 7.6 4.8 8.3 14.6 72.3 86.9 13.1
2.0 0.381 0.347 0.0340 7.2 6.8 6.8 12.6 73.8 86.4 13.6
2.4 0.408 0.376 0.0313 6.5 4.8 5.6 11.3 78.4 89.6 10.4
3.0 0.383 0.347 0.0359 7.8 7.0 6.5 10.9 73.8 84.7 13.5
4.0 0.312 0.252 0.0603 15.1 4.3 8.5 24.3 63.0 87.3 12.8

θFC is the field capacity, θPWP is the plant wilting point, both derived from the adjusted SWRC. TAW is the total available water to plants (i.e., θFC – θPWP ) expressed in cm3 cm−3 and in
a percentage (%) of the total porosity (ϕ) (Table 1). Soil porosity is expressed in terms of macropores (Ma), mesopores (Me), micropores (Mi), and cryptopores (Cr), and also in terms of
transmission (Tr) and storage (St) pores, according to Rowell (1994).

FIGURE 2
Soil properties expressed in terms of soil porosity, water content (θ), and plant water availability) for the upper 4 m of the soil profile. The yellow
area is the storage (St) zone, which comprises the Cr (orange solid line) and Mi (pink solid line) pore sizes, and the blue area is the transmission
(Tr) zone, which comprises the Me (in green) and Ma pores. Total porosity (ϕ; black dotted line) comprises the sum of Cr, Mi, Me, and Ma pores.
In terms of water availability conditions, θFC is the field capacity, and coincides with Mi pores (pink solid line), θPWP is the plant wilting point and
coincides with the Cr pores (orange solid line), and TAW is the total available water to plants (i.e., θFC – θPWP ; yellow dashed area). It is also shown
soil moisture conditions (cm3 cm−3 ) at the upper meter of the soil profile, in May 2016 (dashed line in green) and in October 2016 (dashed line in
orange) based on EC flux tower data from LBA. Data used are presented in Table 2.

more enriched values for both water isotopes, with δ2 H ranging −10.8‰ for δ2 H; and from −6.2 to −4.2‰ and −6.7 to −2.2
from −35.9 to −9.6 ‰ and δ18 O from −3.7 to 2.4‰. Similar ‰ for δ18 O in May and October, respectively (Table 3). Analysis
behavior was observed for soil water signals with enriched values based on single isotopic data (δ2 H and δ18 O; Figures 3A,B,
in the dry season, ranging from −42.9 to −27.5‰ and −46.3 to respectively; Table 3) showed dry season rainfall (measured in

Frontiers in Water 06 frontiersin.org

Borma et al. 10.3389/frwa.2022.886558

FIGURE 3
Boxplot charts for δ2 H (A) and δ18 O (B) signals for the rainfall (all periods, wet and dry periods, and throughfall) (in bright blue), xylem water (in
May 2016 and October 2016) for all sampled trees (in green), soil water at various depths (in yellow), and river water (in blue–green) samples. It is
also shown the averaged values of all xylem water signals (in green), and for all soil water signals (in brown) for May 2016 and October 2016. May
2016 is considered the wet-to-dry (WTD) transition period, and October 2016 is the dry-to-wet (DTW) transition period. The boxplots show
median (black line), 1st and 3rd quartiles (box), max/min observations (upper and lower lines), and the outliers (black points), when present.

Frontiers in Water 07 frontiersin.org

Borma et al. 10.3389/frwa.2022.886558

TABLE 3 Minimum (Min), maximum (Max) and average (Mean) values (including standard deviation, ± std) for all the isotopic data measured, where
N is the number of samples.

Sample type N δ2 H δ18 O

Min Max Mean ± std Min Max Mean ± std

Rainfall 40 −82.2 36.7 −26.7 ± 33.0 −10.1 12.2 −2.87 ± 4.46

Rainfall wet 30 −82.2 −2.8 −41.5 ± 23.0 −10.1 1.5 −4.65 ± 3.04
Rainfall dry 10 0.4 36.7 17.7 ± 9.6 −1.4 12.2 2.49 ± 3.75
Throughfall 20 −20.0 14.4 −8.2 ± 13.1 −3.6 1.4 −1.43 ± 1.64
Plant xylem May 2016 18 −46.6 −27.3 −36.7 ± 5.6 −5.0 −0.2 −3.50 ± 1.31
Plant xylem Oct 2016 18 −35.9 −9.58 −20.0 ± 8.1 −3.7 2.4 −1.13 ± 1.88
Soil May 2016 8 −42.9 −27.4 −35.4 ± 5.9 −6.2 −4.2 −5.19 ± 0.70
Soil Oct 2016 11 −46.3 −10.8 −27.6 ± 13.8 −6.7 −2.2 −4.35 ± 1.73

May, June 2016, and October 2016), and throughfall isotopic (Supplementary Table 5), which means that an enriched isotopic
water signals being more enriched than the wet season rainfall signal, in October 2016, shows a better correlation with Cr pores
(measured in March, April, and December 2016, and January than with other pore sizes or soil physical properties.
and February 2017), for both δ2 H and δ18 O. In May 2016, Correlation analysis between plant traits and isotopic
δ2 Hxylem was in the range of δ2 Hsoil , however, δ18 Oxylem signal showed no significant correlation between isotopic
was more enriched than δ18 Osoil (Table 3). In October 2016, signal (δ2 H and δ18 O) and plant height (H) or DBH
δ2 Hxylem plotted in the range of the more depleted δ2 Hsoil , but, (Supplementary Table 4). However, we found a significant
as in May 2016, δ18 Oxylem is outside the range of δ18 Osoil . negative correlation (r-Pearson of −0.5 and −0.7), respectively,
In the dual isotope space (Figure 4), SWLwet plotted along for δ2 H and δ18 O with ψmd , both measured in October 2015
the GMWL and LMWL in the wet period (LMWLwet , p > (Supplementary Table 4). It suggests that the higher (i.e., more
0.05; Supplementary Table 3), while PWLwet , measured in May negative) the plant water potential, the more enriched the
2016, followed the LMWLwet (1slope = 0.24; p = 0.930) isotopic signal.
and the throughfall water line (1slope = 0.34; p = 0.853;
Supplementary Table 3). In October 2016, PWLdry better agreed
with the LMWLdry (1slope = 1.28; p = 0.37), and deviated
Discussion
from the LWMLwet , in response to the higher enrichment of Potential sources of water for plants
the δ18 Oxylem in relation to the δ2 Hxylem . However, some trees
presented an isotopic signal that followed the LWMLwet . These Our isotopic results show that the wet season rainfall was
trees were marked in bold, in Supplementary Table 3. isotopically more depleted than the dry season rainfall, a pattern
that can be attributed to the rainfall amount effect (Araguás-
Araguás et al., 2000). In the dual isotope space, the agreement
Isotopic signal and correlation analysis between GMWL and LMWLwet , GMWL and throughfall, and
The variability of the isotopic signal along the soil profile GMWL and SWLOct (Supplementary Table 3) suggests that the
showed that the isotopically enriched layers were found in the rainfall of past events would be mainly responsible for the soil
upper 0.15 m. The soil water became more depleted with depth, water recharge. Considering the hypothesis that to meet the
approximating to the Machado river water at around 3 m depth, plant-atmospheric water demands of the dry period, trees uptake
and even more depleted than the river water at 4 m depth the more depleted water from the past wet periods, stored in
(Figures 3A,B). the deeper soil layers; we would expect a more depleted xylem
Correlation coefficients between soil physical properties signal as the dry season progresses. However, contrary to this,
[soil texture, pore sizes (Ma, Me, Mi, and Cr pores), θFC and we found xylem water more enriched in October 2016, end of
θPWP , and TAW] presented negative r-Pearson (−0.82 and the dry period/transition to wet period, than in May 2016, the
−0.76 for δ2 H and δ18 O, respectively) for TAW and Mi, in beginning of the dry period. Besides, instead of following the
May 2016 (Supplementary Table 5). It means that the larger isotopically depleted soil signal of the deeper soil layers, the
the TAW and Mi, the lower (i.e., the most depleted) the xylem signal is isotopically more enriched and shows a better
isotopic signal. In October 2016, the DTW transition period, agreement with the recent rainfall. This pattern can be observed
the larger positive correlations were found for the percentage in dual isotope space (Figure 4) and in the ANCOVA analysis
of Cr pores, respectively, 0.43 and 0.35 for δ2 H and δ18 O (Supplementary Table 3). These results suggest that instead of

Frontiers in Water 08 frontiersin.org

Borma et al. 10.3389/frwa.2022.886558

FIGURE 4
Dual isotope plot of the global meteorological water line (GMWL), local meterological water line for the wet (LMWLwet) and the dry periods
(LMWLdry, taken in May, June and October 2016), and throughfall; soil water line that comprise the isotopic soil signal at different soil depths in
May 2016 (SWLMay16) and in October 2016 (SWLOct16); plant water line that comprises the isotopic signal of the xylem in May 2016
(PWLMay16) and in October 2016 (PWLOct16). May 2016 corresponds to the wet-to-dry (WTD) transition period, and October 2016
corresponds to the dry-to-wet (DTW) transition period.

using the water stored in the soil layers, to meet the plant- (Nepstad et al., 2007; Phillips et al., 2010; Esquivel-Muelbert
atmospheric water demands of the dry period, trees are using et al., 2020). These results can be of great concern considering
a more recent water, which is the rainfall dropped in the past the perspectives of increase in length and severity of the dry
recent events. In May 2016, beginning of the dry period, trees season rainfall, as a result of local and global climate changes
used water from the past wet season rainfall while in October (Wright et al., 2017; Gatti et al., 2021). However, it is important
2016, the peak of the dry period/transition to the wet period, to note that the Amazon forest comprises a very huge and
trees used the more enriched water from the dry period. diverse ecosystem, and more research is needed to generalize the
These results agree with the results found by Miguez-Macho behavior found here to all the Amazon forest environments, and
and Fan (2021), who found that in a global survey of studies for more severe dry periods.
based on isotope data, 70% of plants used water from the
recent rainfall, while 18% use water from the past rainfall events
stored in the soil layers. The plant dependence on the recent
rainfall events challenges the deep-RWU as a mechanism of Uncertainties in the results and another
forest resilience to droughts. The fact that instead of relying on potential explanation for xylem isotopic
the water from past events stored in the soil, the maintenance enrichment
of transpiration rates relies on a more recent rainwater, could
explain the high mortality rates of the larger trees during Despite statistical analysis (ANCOVA, methods)
severe droughts that have been registered in the Amazon forest indicating the existence of significant similarity between

Frontiers in Water 09 frontiersin.org

Borma et al. 10.3389/frwa.2022.886558

PWLOct and LMWLdry slopes (1slope = −0.11; p = It is also important to note that we did not find an overlap
0.873; Supplementary Table 3), visually it is possible between the soil water signal and xylem signal for any of
to observe a displacement in both lines so that the both sampled periods (i.e., in May and in October 2016).
PWLOct plots below the LMWLdry (Figure 4). This Very fine resolution soil moisture measurements performed
displacement is a result of the pronounced enrichment in central Amazon show a very wet layer developed in the
of the δ18 O in relation to the δ2 H (Figures 3A,B), a upper 5 cm of the soil layer during droughts (Negrón-Juárez
pattern which can be also observed in the d-excess et al., 2020). Further investigation is needed to consider
(Supplementary Figure 2). if there is a missing source of water for soil layers that
The relationship between δ2 H and δ18 O, expressed in terms were not sampled and could also explain the enriched xylem
of the d-excess (Equation 3), has been used as an indicator isotopic signal.
of evaporative enrichment (Sprenger et al., 2016). Some works
have considered d-excess <-10‰ as an indicator of evaporative
enrichment of the rainfall water (Kern et al., 2020). Despite Conclusion
tree waters being sampled, stored, and analyzed following the
same procedure, in October 2016—peak of the dry period— Knowledge of how rainforests cope with droughts can be
we found a pronounced enrichment in the δ18 O (i.e., d-excess leveraged to assess how a changing climate may impact these
<-10‰) for 7 of the 18 sampled trees; Swartzia ingifolia, ecosystems. Deep (>1 m) soil water, stored during the previous
Dipterix odorata, Dipterix magnifica, Copaifera multijuga, wet season, has been regarded as playing a critical role in the
Cariniana decandra, Astronium lecontei, and Anamalocalyx maintenance of relatively high ET rates in Amazonia. However,
uleanus (Supplementary Figure 3). Four of these trees presented taller and larger trees, which potentially have deep roots, are
moderate resistance to drought (Dipterix odorata, Cariniana the most vulnerable to severe droughts, casting doubt on this
decandra, Astronium lecontei, and Anamalocalyx uleanus) and hypothesis. Using naturally abundant stable water isotopes
one (Swartzia ingifolia) presented low resistance to drought, together with soil physical data and ecophysiological plant
according to Souza et al. (2022). For the other species, resistance traits—H, DBH, and ψmd –we found that recently infiltrated
to drought (using πTLP as a proxy) was not determined (Souza water is crucial for sustaining plant transpiration during the
et al., 2022). However, other species that also presented moderate dry periods. These results challenge deep-RWU theory and
resistance to drought according to Souza et al. (2022) did the use, by the plants, of the water from past rainfall events
not present a pronounced δ18 O enrichment as the dry season stored in the soil layer. In other words, it means that the
progresses. This is the case for Sagotia brachysepala, Pouteria dependence of the Amazon forest on the dry-season rainfall
durlandi, and Licania sprucei (Supplementary Figure 3). Based could make the Amazon forest more vulnerable to droughts
on this, we did not find a relationship between the pronounced than previously thought, in particular, with the prospect of
enrichment in the xylem isotopic signal and the plant resistance the dry-season lengthening and projected decreases in the dry
to drought. We also did not find any correlation between season rainfall. When using the natural abundance of water
enriched xylem isotopic signal in October 2016 and tree H or isotopes as tracers of the source of transpired water, caution is
DBH (Supplementary Table 4). However, the strong correlation needed for processes that may cause enrichment in the xylem
between the enriched signal and the ψmd , both measured signal, or the existence of missing sources not sampled. In our
in October 2016 (Supplementary Table 4) suggests that an case, the pronounced enrichment of the xylem water in the
enriched xylem signal is more pronounced in trees submitted dry period suggests that plants are using the more enriched
to higher water stress (ψmd > 2 MPa). According to Martín- rainwater of the dry period. However, it is also possible that
Gómez et al. (2017), under low water availability and high some trees may have branches where the stem water flow ceased
evaporative demand, stem water loss via evaporation can due to water stress at the end of the dry period. The reduction
create significant isotopic enrichment of stem water (Martín- of the stem flow may cause isotopic enrichment and biases
Gómez et al., 2017). That is because, when leaf transpiration in the identification of the source of the water transpired by
is limited, it reduces the input of non-enriched fresh water, plants. However, it is important to note that, in the study
allowing for cumulative evaporative enrichment. However, site, the maintenance of the elevated ET rates and the EVI
in our study, given the maintenance of relatively high ET during the dry period do not point to huge water stress in the
rates and EVI along the dry period (i.e., between May sampled trees.
and October 2016)—as presented in Souza et al. (2022)—we
conclude that the isotopic enrichment of the xylem signal is
more a result of the dry season rainwater uptake than of a Data availability statement
limitation of the transpiration rates and stem water flow as
the dry season progresses, as proposed to Martín-Gómez et al. The raw data supporting the conclusions of this article will
(2017). be made available by the authors, without undue reservation.

Frontiers in Water 10 frontiersin.org

Borma et al. 10.3389/frwa.2022.886558

Author contributions Conflict of interest

LB, WD, and AV wrote the paper and performed soil physics The authors declare that the research was conducted in the
and isotopic analysis. RS performed work experiments, plant absence of any commercial or financial relationships that could
traits, and remote sensing analysis. AW and RA performed field be construed as a potential conflict of interest.
experiments and maintenance of the LBA experimental site at
Rebio Jaru. All authors contributed to the article and approved
the submitted version. Publisher’s note
All claims expressed in this article are solely those of the
Funding authors and do not necessarily represent those of their affiliated
organizations, or those of the publisher, the editors and the
This research was supported by the FAPESP reviewers. Any product that may be evaluated in this article, or
grant (2013/50531-2). claim that may be made by its manufacturer, is not guaranteed
or endorsed by the publisher.

Acknowledgments
Supplementary material
We thank the Large Biosphere-Atmosphere (LBA)
Program, coordinated by the National Institute for Amazon The Supplementary Material for this article can be
Research (INPA), for data availability, logistical support, and found online at: https://www.frontiersin.org/articles/10.3389/
infrastructure during field campaigns. frwa.2022.886558/full#supplementary-material

References
Araguás-Araguás, L., Froehlich, K., and Rozanski, K. (2000). IBAMA (2007). Avaliação Ecológica Rápida para Revisão do Plano de Manejo da
Deuterium and oxygen-18 isotope composition of precipitation Reserva Biológica do Jaru, Brasilia, DF.
and atmospheric moisture. Hydrol. Process. 14, 1341–1355.
Jiménez-Muñoz, J. C., Mattar, C., Barichivich, J., Santamaría-Artigas, A.,
doi: 10.1002/1099-1085(20000615)14:8<1341::AID-HYP983>3.0.CO;2-Z
Takahashi, K., Malhi, Y., et al. (2016). Record-breaking warming and extreme
Berg, A., Findell, K., Lintner, B., Giannini, A., Seneviratne, S. I., Van Den Hurk, drought in the Amazon rainforest during the course of El Niño 2015–2016. Sci.
B., et al. (2016). Land-atmosphere feedbacks amplify aridity increase over land Rep. 6:33130. doi: 10.1038/srep33130
under global warming. Nature Clim. Change 6, 869–874. doi: 10.1038/nclimate3029
Jipp, P. H., Nepstad, D. C., Cassel, D. K., and Reis De Carvalho, C.
Bouyoucos, G. J. (1927). A rapid method for mechanical analysis of soils. Science (1998). Deep soil moisture storage and transpiration in forests and pastures of
65, 549–551. doi: 10.1126/science.65.1692.549 seasonally-dry Amazonia. Clim. Change 39, 395–412. doi: 10.1023/A:10053089
30871
Bruno, R. D., da Rocha, H. R., de Freitas, H. C., Goulden, M. L., and Miller, S. D.
(2006). Soil moisture dynamics in an eastern Amazonian tropical forest. Hydrol. Kern, Z., Hatvani, I., Czuppon, G., Fórizs, I., Erdélyi, D., Kanduč, T., et al.
Process. 20, 2477–2489. doi: 10.1002/hyp.6211 (2020). Isotopic ‘altitude’ and ‘continental’ effects in modern precipitation across
the adriatic–Pannonian region. Water 12:1797. doi: 10.3390/w12061797
Christoffersen, B. O., Restrepo-Coupe, N., Arain, M. A., Baker, I. T., Cestaro,
B. P., Ciais, P., et al. (2014). Mechanisms of water supply and vegetation demand Kunert, N., Aparecido, L. M. T., Wolff, S., Higuchi, N., dos Santos, J., de
govern the seasonality and magnitude of evapotranspiration in Amazonia and Araujo, A. C., et al. (2017). A revised hydrological model for the Central Amazon:
Cerrado. Agric. For. Meteorol. 191, 33–50. doi: 10.1016/j.agrformet.2014.02.008 the importance of emergent canopy trees in the forest water budget. Agric. For.
Coplen, T. B. (2011). Guidelines and recommended terms for expression of Meteorol. 239, 47–57. doi: 10.1016/j.agrformet.2017.03.002
stable-isotope-ratio and gas-ratio measurement results. Rapid Commun. Mass Marengo, J. A. (2004). Interdecadal variability and trends of rainfall across the
Spectrom. 25, 2538–2560. doi: 10.1002/rcm.5129 Amazon basin. Theor. Appl. Climatol. 78, 79–96. doi: 10.1007/s00704-004-0045-8
Da Silva, J. O. (2021). Distribuição Espacial dos atributos físicos do solo em uma Martín-Gómez, P., Serrano, L., and Ferrio, J. P. (2017). Short-term dynamics
Reserva Biológica no bioma Amazônia. Ji-Parana. of evaporative enrichment of xylem water in woody stems: implications for
Dansgaard, W., Daansgard, W., and Dansgaard, W. (1964). Stable isotopes in ecohydrology. Tree Physiol. 37, 511–522. doi: 10.1093/treephys/tpw115
precipitation. Tellus. 6n, 436–468. doi: 10.3402/tellusa.v16i4.8993 Miguez-Macho, G., and Fan, Y. (2021). Spatiotemporal origin of soil water taken
Ehleringer, J. R., and Dawson, T. E. (1992). Water uptake by plants: up by vegetation. Nature 598, 624–628. doi: 10.1038/s41586-021-03958-6
perspectives from stable isotope composition. Plant Cell Environ. 15, 1073–1082. Muler, R. A. S. (2018). Analysis of Seasonal and Interannual Response of Forest
doi: 10.1111/j.1365-3040.1992.tb01657.x Dynamics using MODIS/MAIAC Sensor Data and in situ Measurements in a Forest
Esquivel-Muelbert, A., Phillips, O. L., Brienen, R. J. W., Fauset, S., Sullivan, M. J. Fragment in the southwest Amazon. INPE.
P., Baker, T. R., et al. (2020). Tree mode of death and mortality risk factors across Negrón-Juárez, R., Ferreira, S. J. F., Mota, M. C., Faybishenko, B., Monteiro, M.
Amazon forests. Nat. Commun. doi: 10.1038/s41467-020-18996-3 T. F., Candido, L. A., et al. (2020). Calibration, measurement and characterization
Gatti, L. V., Basso, L. S., Miller, J. B., Gloor, M., Domingues, L. G., Cassol, H. L. of deep soil moisture dynamics in a Central Amazonian tropical forest. Vadose
G., et al. (2021). Amazonia as a carbon source linked to deforestation and climate Zone J. 19:e20070. doi: 10.1002/vzj2.20070
change. Nature 595, 388–393. doi: 10.1038/s41586-021-03629-6 Nepstad, D. C., De Carvalho, C. R., Davidson, E. A., Jipp, P. H., Lefebvre, P.
Gee, G. W., and Bauder, J. W. (1986). “Methods of soil analysis. Part I,” in A., Negreiros, G. H., et al. (1994). The role of deep roots in the hydrological
Particle-size Analysis, ed A. Klute (Madison, WI: American Society of Agronomy and carbon cycles of Amazonian forests and pastures. Nature 372, 666–669.
and Soil Science Society of America), 383–411. doi: 10.2136/sssabookser5.1.2ed.c15 doi: 10.1038/372666a0

Frontiers in Water 11 frontiersin.org

Borma et al. 10.3389/frwa.2022.886558

Nepstad, D. C., Tohver, I. M., David, R., Moutinho, P., and Cardinot, G. (2007). Souza, R. D. A., Moura, V., Paloschi, R. A., Aguiar, R. G., Webler, A. D.,
Mortality of large trees and lianas following experimental drought in an amazon Borma, L. D. S., et al. (2022). Assessing drought response in the Southwestern
forest. Ecology 88, 2259–2269. doi: 10.1890/06-1046.1 Amazon forest by remote sensing and in situ measurements. Remote Sens. 14:1733.
doi: 10.3390/rs14071733
Phillips, O. L., van der Heijden, G., Lewis, S. L., López-González, G., Aragão, L.
E. O. C., Lloyd, J., et al. (2010). Drought-mortality relationships for tropical forests. Sperry, J. S., Hacke, U. G., Oren, R., and Comstock, J. P. (2002). Water deficits
New Phytol. 187, 631–646. doi: 10.1111/j.1469-8137.2010.03359.x and hydraulic limits to leaf water supply. Plant, Cell Environ. 25, 251–263.
doi: 10.1046/j.0016-8025.2001.00799.x
Restrepo-Coupe, N., da Rocha, H. R., Hutyra, L. R., da Araujo, A. C.,
Borma, L. S., Christoffersen, B., et al. (2013). What drives the seasonality of Sprenger, M., Leistert, H., Gimbel, K., and Weiler, M. (2016). Illuminating
photosynthesis across the Amazon basin? A cross-site analysis of eddy flux hydrological processes at the soil-vegetation-atmosphere interface with water
tower measurements from the Brasil flux network. Agric For. Meteorol. 182–183, stable isotopes. Rev. Geophys. 54, 674–704. doi: 10.1002/2015RG000515
128–144. doi: 10.1016/j.agrformet.2013.04.031
Teixeira, P. C., Donagemma, G. K., Fontana, A., and Teixeira, W. G. (2017).
Rowell, D. C. (1994). Soil Science: Methods and Applications. London: Routledge. Manual de Métodos de Análise de Solo, 3rd Edn. Brasilia, DF: Embrapa.
368 pages.
van Genuchten, M. T. A. (1980). Closed-form equation for predicting the
Saleska, S. R., Wu, J., Guan, K., Araujo, A. C., Huete, A., Nobre, A. D., hydraulic conductivity of unsaturated soils. Soil Sci. Soc. Am. J. 44, 892–898.
et al. (2016). Dry-season greening of Amazon forests. Nature 531, E4–E5. doi: 10.2136/sssaj1980.03615995004400050002x
doi: 10.1038/nature16457
Walkley, A., and Black, I. A. (1934). An examination of the degtjareff method for
Seneviratne, S. I., Wilhelm, M., Stanelle, T., Van Den Hurk, B., Hagemann, determining soil organic matter, and a proposed modification of the chromic acid
S., Berg, A., et al. (2013). Impact of soil moisture-climate feedbacks on CMIP5 titration method. Soil Sci. 37, 29–38. doi: 10.1097/00010694-193401000-00003
projections: first results from the GLACE-CMIP5 experiment. Geophys. Res. Lett.
Washburn, E. W. (1921). The dynamics of capillary flow. Phys. Rev. 17:273.
40, 5212–5217. doi: 10.1002/grl.50956
doi: 10.1103/PhysRev.17.273
Shuttleworth, W. J. (1988). Macrohydrology - the new challenge for process
Webler, A. D., Aguiar, R. G., and Aguiar, L. J. G. (2007). Características da
hydrology. J. Hydrol. 100, 31–56. doi: 10.1016/0022-1694(88)90180-1
precipitação em área de floresta primária e área de pastagem no Estado de
Shuttleworth, W. J. (1998). Evaporation from Amazonian rainforest. Proc. R. Soc. Rondônia. Ciên. Nat. 2007, 55–58. doi: 10.5902/2179460X9755
B Biol. Sci. (1988). B223, 321–346. doi: 10.1098/rspb.1988.0024
Wright, J. S., Fu, R., Worden, J. R., Chakraborty, S., Clinton, N. E., Risi, C., et al.
Sohel, M. S. I., Grau, A. V., McDonnell, J. J., and Herbohn, J. (2021). (2017). Rainforest-initiated wet season onset over the southern Amazon. Proc. Natl.
Tropical forest water source patterns revealed by stable isotopes: a Acad. Sci. U.S.A. 114, 8481–8486. doi: 10.1073/pnas.1621516114
preliminary analysis of 46 neighboring species. For. Ecol. Manage. 494:119355.
Wu, J., Guan, K., Hayek, M., Restrepo-Coupe, N., Wiedemann, K. T., Xu, X.,
doi: 10.1016/j.foreco.2021.119355
et al. (2017). Partitioning controls on Amazon forest photosynthesis between
Sombroek, W. (2001). Spatial and temporal patterns of Amazon rainfall. Ambio. environmental and biotic factors at hourly to interannual timescales. Glob. Change
30, 388–396. doi: 10.1579/0044-7447-30.7.388 Biol. 23, 1240–1257. doi: 10.1111/gcb.13509

Frontiers in Water 12 frontiersin.org