ORIGINAL RESEARCH

published: 21 January 2020

doi: 10.3389/fpls.2019.01715

A Simple Method for Simulating

Drought Effects on Plants
Renée M. Marchin 1*, Alessandro Ossola 2, Michelle R. Leishman 2 and David S. Ellsworth 1
1 Hawkesbury Institute for the Environment, Western Sydney University, Penrith, NSW, Australia, 2 Department of Biological

Sciences, Macquarie University, North Ryde, NSW, Australia

Drought is expected to increase in frequency and severity in many regions in the future, so
it is important to improve our understanding of how drought affects plant functional traits
and ecological interactions. Imposing experimental water deficits is key to gaining this
understanding, but has been hindered by logistic difficulties in maintaining consistently low
water availability for plants. Here, we describe a simple method for applying soil water
deficits to potted plants in glasshouse experiments. We modified an existing method (the
“Snow and Tingey system”) in order to apply a gradual, moderate water deficit to 50 plant
species of different life forms (grasses, vines, shrubs, trees). The method requires less
maintenance and manual handling compared to other water deficit methods, so it can be
Edited by:
used for extended periods of time and is relatively inexpensive to implement. With only a
Randy D. Allen,
Oklahoma State University, few modifications, it is possible to easily establish and maintain soil water deficits of
United States differing intensity and duration, as well as to incorporate interacting stress factors. We
Reviewed by: tested this method by measuring physiological responses to an applied water deficit in a
Alejandra Navarro,
Council for Agricultural and
subset of 11 tree/shrub species with a wide range of drought tolerances and water-use
Economics Research, Italy strategies. For this subgroup of species, stomatal conductance was 2–17 times lower in
Mauro Guida Santos,
droughted plants than controls, although only half of the species (5 out of 11) experienced
Federal University of Pernambuco,
Brazil midday leaf water potentials that exceeded their turgor loss (i.e., wilting) point. Leaf
*Correspondence: temperatures were up to 8°C higher in droughted plants than controls, indicating that
Renée M. Marchin droughted plants are at greater risk of thermal damage, relative to unstressed plants. The
r.prokopavicius@
westernsydney.edu.au
largest leaf temperature differences (between droughted and well-watered plants) were in
species with high rates of water loss. Rapid osmotic adjustment was observed in leaves of
Specialty section: five species when drought stress was combined with an experimental heatwave. These
This article was submitted to
results highlight the potential value of further ecological and physiological experiments
Plant Abiotic Stress,
a section of the journal utilizing this simple water deficit method to study plant responses to drought stress.
Frontiers in Plant Science
Keywords: glasshouse experiment, moderate drought stress, plant drought tolerance, plant ecophysiology, soil
Received: 06 September 2019 water content, water deficit
Accepted: 05 December 2019
Published: 21 January 2020
Citation:
Marchin RM, Ossola A, Leishman MR
INTRODUCTION
and Ellsworth DS (2020) A Simple
Method for Simulating Drought
Drought has been described as the most damaging climate hazard facing our global population
Effects on Plants. (Kogan, 1997). It is expected to increase in frequency and severity in many regions in the future as a
Front. Plant Sci. 10:1715. result of decreased precipitation and increased evaporation due to global climate change (IPCC,
doi: 10.3389/fpls.2019.01715 2018; Naumann et al., 2018; Dey et al., 2019). The spatial extent and duration of recent droughts,

Frontiers in Plant Science | www.frontiersin.org 1 January 2020 | Volume 10 | Article 1715

Marchin et al. A Simple Drought Method

such as the Millennium Drought (1997–2009) in southern forms, growth rates, sizes, and leaf areas, and (3) it is simple to
Australia and the California drought (2011–2017) in the USA, establish water deficits of varying intensity, duration, and
are without precedent within at least the last 400 years (Griffin pulsation (i.e., repeated drying and wetting cycles; Figure 2).
and Anchukaitis, 2014; Freund et al., 2017). Approximately two- This method is especially useful for establishing gradual water
thirds of the global population will be affected by increasing deficits lasting for weeks/months and incorporating interacting
drought (Naumann et al., 2018), which threatens food security stress factors. Previous experiments used pure sand as the potting
(FAO, 2018), forest health (Allen et al., 2010; Choat et al., 2018), media (Fernández and Reynolds, 2000; Maseda and Fernández,
and even the global beer supply (Xie et al., 2018). One of the key 2016), which has likely limited the use of this method to date.
challenges for plant science is to improve our understanding of Here, we revisit the “Snow and Tingey system” and adapt it to
how drought affects plant ecology and plant functional traits, as simulate a moderate drought stress for a diverse set of plants with
this will impact agricultural productivity as well as vegetation different life forms (grasses, vines, shrubs, trees), extending its
management. One invaluable and long-used method for use beyond sand-based media to allow the use of more complex
examining plant drought responses is the experimental horticultural media and soils.
application of controlled water deficits in the glasshouse. Drought usually takes months or years to develop in natural
Methods for applying soil water deficits in pot studies date ecosystems (Zargar et al., 2011). Although pot studies cannot
back at least 50 years, but there has not been consensus on a best- recreate the complex heterogeneous factors present in field
practice method (Munns et al., 2010). The most basic method for environments, it is good practice to establish a gradual
generating soil water deficits is passive pot-drying by intensification of water deficit over at least several weeks
withholding irrigation, but this method risks fast drying rates (Snow and Tingey, 1985). Partial stomatal closure is one of the
that do not adequately mimic natural soil water deficits (Poorter earliest responses to water deficit and can occur rapidly (i.e.,
et al., 2012). Many early studies added osmotically-active within days), but acclimation responses – such as osmotic
substances (e.g., polyethylene glycol, PEG) to soil (Zur, 1966), adjustment – require longer stress events (Harb et al., 2010;
but PEG can limit oxygen diffusion to roots (Mexal et al., 1975) Blum, 2017). Osmotic adjustment is the accumulation of solutes
and interfere with ion uptake (Yeo and Flowers, 1984). Various in plant leaves under water deficit and is a strategy for
other methods have been used to decrease water availability to maintaining turgor at low leaf water potentials (Hsiao et al.,
plants, including reduction of water pressure inside microporous 1976; Morgan, 1984; Chen and Jiang, 2010). Plant species vary in
tubes (Steinberg and Henninger, 1997) or attachment of a their ability to osmotically adjust (Bartlett et al., 2014), and the
vacuum pump to pots (Bunce and Nasyrov, 2012), both of adaptive process of osmotic adjustment requires time, at least 6–
which require an additional apparatus and complex logistics to 15 days of water deficit in crop varieties with the capacity for
implement. The most commonly used method for applying soil high osmotic adjustment (Blum et al., 1980; Molinari et al., 2004;
water deficits is to air-dry and regularly weigh individual pots, Blum, 2017). Other drought responses, such as wilting and loss of
adding precise amounts of water in order to balance water loss stem conductivity, require more severe drought stress (Bartlett
from transpiration and establish the target soil water content et al., 2016). Damage to the light-harvesting capacity of leaf
(e.g., Earl, 2003). This method successfully simulates drought photosynthesis occurs only after severe loss of hydraulic function
stress for plants of different sizes, but requires a complex and under extreme dehydration (Trueba et al., 2019). In this way, it is
expensive automated computer system. While the same method important to carefully consider the desired level of drought
can be achieved manually, it is a laborious and time-consuming response before selecting an experimental water deficit
task, particularly for large experiments with hundreds of plants. treatment (see Figure 2).
Large experimental designs usually require the use of several We designed glasshouse experiments to screen large numbers
glasshouses and/or facilities, thus increasing the need for a of plant species (>50 species) in order to identify drought-
simple yet accurate method to simulate drought. tolerant species for urban plantings in Australian cities, though
One method that has received relatively little attention was the drought method we describe can be used in a wide variety of
first proposed by Haan and Barfield (1971) and later described by contexts. Drought is common in Australia (Nicholls et al., 1997;
Snow and Tingey (1985), where solid columns of low water Freund et al., 2017) and has been associated with tree decline in
permeability are used to separate the root zone from a water urban areas (Nitschke et al., 2017). Unprecedented extreme
table. The original method utilized complex float chambers to temperatures are also predicted for many parts of the world,
establish a more constant water stress (Snow and Tingey, 1985), including Australia, within the next 10–30 years (Perkins and
relative to methods involving repeated cycles of rewatering, and Alexander, 2013; Lewis et al., 2017; BoM, 2019), so we tested the
has since been modified several times (Wookey et al., 1991; efficacy of this method by applying an experimental heatwave in
Fernández and Reynolds, 2000). A simple and inexpensive the fifth week of water deficit to better understand how plants
modification uses commercial floral foam as the means for will cope with combined drought and heat stress. The overall
establishing soil water deficit of potted plants (Fernández and goals of this work were to: (1) refine a simple water deficit
Reynolds, 2000; Figure 1). Some other advantages of this version method to simulate drought stress on potted plants, (2) evaluate
of the ‘Snow and Tingey system’ are: (1) very little maintenance its effectiveness in generating some common ecophysiological
is required, (2) it can be used to apply water deficits responses to water deficits, and (3) demonstrate that this method
simultaneously to diverse plant species with different growth can be used to investigate interactive effects of drought and other

Frontiers in Plant Science | www.frontiersin.org 2 January 2020 | Volume 10 | Article 1715

Marchin et al. A Simple Drought Method

FIGURE 1 | (A) Diagram of the simple water deficit method, adapted by permission from Springer Nature Customer Service Centre GmbH: Springer-Verlag
Oecologia (Potential growth and drought tolerance of eight desert grasses: lack of a trade-off?, Fernández and Reynolds), © 2000. Capillary irrigation is used to
control soil water content of potted plants (a, b), which are placed above a solid column of low water permeability (F, commercial porous foam) inside a plastic
container (C) filled with water (W). Intensity of the water deficit is controlled by the depth to water table (z). (B) Photo of drought tubs inside the MQ glasshouse.

environmental stresses, such as heatwaves. We show how the All seedlings were well-watered using drip irrigation for 6–15
gradual, moderate water deficit achieved with the “Snow and weeks to establish roots, allow formation of new leaves, and
Tingey system” affected plant physiological responses in a subset acclimate to glasshouse environmental conditions. During this
of 11 tree/shrub species. We also discuss the versatility of this time, seedlings were rotated within and between glasshouse bays
simple method for simulating drought effects on plants in order on a monthly basis to allow uniform solar irradiance for growth.
to highlight potential uses in other ecological experiments. At the beginning of the experiment, seedlings received one daily
watering of 1 L at 6:00. As seedlings grew, watering was increased
to a total of 1.5–4.5 L daily (delivered at 8:00, 13:00, and 17:00) to
keep all seedlings well-watered. The average glasshouse
MATERIALS AND METHODS temperature was 27°C to represent summer conditions in
southeastern Australia, with a diurnal range from 21 to 34°C
Plant Material and Experimental and the maximum temperature spanning 6 h at midday (12:00–
Conditions 18:00; Supplementary Figure S1). Glasshouse daytime relative
Two coordinated glasshouse experiments were conducted: one at humidity ranged from 40%–95%, which led to a range in daytime
the Hawkesbury Institute for the Environment (HIE; Western vapor pressure deficit (VPD) of 0.2–3.5 kPa. Daily maximum
Sydney University, Richmond, NSW, Australia) and the other at photosynthetically active radiation (PAR) was >2,000 µmol m-2
Macquarie University (MQ; North Ryde, NSW, Australia). A s-1 inside the HIE glasshouses, but only ~1600 µmol m-2 s-1
total of 50 plant species from 20 different families were selected, inside the MQ glasshouses.
including 25 tree, 11 shrub, 7 vine/groundcover, and 7
herbaceous species (Supplementary Table S1). Species
included both native Australian plants (42 species) and exotic Water Deficit Treatment
species (8 species) and occur in a range of different After the acclimation period, half of the plants (n = 5 plants per
environments, from semi-arid to rainforest ecosystems species) were exposed to a gradual, five-week water deficit using the
(Table 1). Twenty-four species were grown in two glasshouse method described by Snow and Tingey (1985) and modified by
bays at HIE from 1 November 2017 to 23 March 2018; 26 species Fernández and Reynolds (2000). Pure sand has been used to
were grown in two glasshouse bays at MQ from 26 January to 25 achieve fast equilibrium rates during drying (Fernández and
May 2018. Planting stock (n = 10 plants per species) was Reynolds, 2000), but it is not a preferred growth substrate for
obtained from commercial nurseries in Australia as tubestock, most plants. We used a native potting mix (as described above; bulk
140-mm, or 200-mm pot size, depending on species availability. density: 0.45 ± 0.004 g cm-3), instead of pure sand, to successfully
Seedlings were bare-rooted and transplanted into 6-L square pots grow a diverse set of 50 plant species (Supplementary Table S1). It
containing native potting mix (<30% sand/coir, >70% screened was not possible to drought all species in the same week due to a
composted pine bark; Australian Growing Solutions, Tyabb, limited number of glasshouses, so 5–6 species with similar growth
VIC, Australia), 38 g of controlled-release native plant fertilizer rates were batched and treated at the same time.
(Scotts Australia Osmocote Slow Release, Bella Vista, NSW, Before planting, four large 4.3-cm diameter circles were
Australia), and a 1.25-g tablet of systemic insecticide and drilled into the flat base of each drought pot and fitted with
fertilizer (Yates Confidor, Padstow, NSW, Australia). fine nylon mesh (20-µm, Allied Filter Fabrics, Berkeley Vale,

Frontiers in Plant Science | www.frontiersin.org 3 January 2020 | Volume 10 | Article 1715

Marchin et al. A Simple Drought Method

porous foam (Oasis IDEAL Floral Foam Maxlife brick; Smithers-

Oasis, Kent, OH, USA) with an adjustable water table. Pots were
able to be removed as needed for measurements, and the
constant water deficit was restored after pots were replaced
onto the commercial foam. However, it is essential to this
capillary irrigation method that: (1) there is adequate contact
area between the soil at the bottom of the pot and the foam
surface and (2) the pore size of the commercial foam is sufficient
for transporting water to the desired height by capillarity.
Depth to the water table was progressively increased every day
based on a predetermined schedule designed for the soil type, from
1 cm (on day 1) to 15 cm (on day 8) to 22 cm (on day 15), where it
was maintained for an additional three weeks (until day 35). Water
levels were checked daily and maintained within ±1 cm of the target
level, but this rate of change in the water table matched changes in
plant transpiration closely so that very little maintenance was
required over the first 15 days of drought. The soil volumetric
water content (VWC) of each drought pot was measured weekly
using a 20-cm water content probe (CS658 HydroSenseII,
Campbell Scientific Inc., Logan, UT, USA) for the first three
weeks of water deficit, but every 3–4 days thereafter to ensure the
final target intensity was reached. The target drought intensity was a
soil VWC of 7.5 ± 2.5%, which is below the permanent wilting point
for this soil (14%, Supplementary Figure S3). Soil VWC was
always measured in the morning (8:00–10:00). Pots exceeding the
upper limit of our target intensity on day 22 (>10%) were bench-
dried to achieve the target intensity, then replaced in the drought
tub. This only occurred in a limited number of pots containing
small plants with low transpiration rates (e.g., Cryptocarya
laevigata, Lophostemon confertus, Myoporum parvifolium). If soil
VWC surpassed the lower limit of our target intensity (<5%), 300
ml water was added to the soil surface of the pot to maintain the
target soil VWC. This was necessary for pots containing large plants
with high transpiration rates (e.g., Hakea laurina, Murraya
paniculata, Stenotaphrum secundatum). Thus, additional
maintenance was required during the final two weeks of water
deficit (days 25–35), although the total time investment was still
considerably less than for some other water deficit methods.
All control pots (n = 5 plants per species) were maintained at
field capacity (soil VWC ~35%, ranging from 25%–45%) for the
duration of the experiment. To determine if drought tubs were
necessary for watering control plants, we first tested if soil VWC
of control pots differed between our two watering methods:
capillary irrigation versus drip irrigation. We drilled four 4.3-
FIGURE 2 | Potential drought scenarios, i.e., experimental water deficit
cm diameter circles into the flat base of 14 control pots, fitted
treatments, that can be achieved using the modified “Snow & Tingey system”:
(A) intensity, from mild to severe, which can be achieved using different
with fine nylon mesh (before potting), and placed pots inside
depths to water table (z in Figure 1: mild, z ≤10 cm; severe, z ≥25 cm), (B) 100-L plastic containers on top of a 23-cm column of
duration, from weeks to months, and (C) pulsation, i.e., the use of repeated commercial foam with a constant 1–4 cm depth to the water
drying and wetting cycles, which can be achieved through repeated lowering table. The soil VWC of each control pot was measured weekly
and raising of the depth to water table (z in Figure 1).
using a 20-cm soil water content probe (CS658 HydroSenseII,
Campbell Scientific Inc.). For a subset of four pots, we
NSW) to allow exchange of air and water but prevent root continuously measured soil VWC using soil water content
passage (Supplementary Figure S2). Utility taps were installed probes (ThetaProbe type ML2X, Delta-T, Cambridge, UK) at a
into the base of 100-L plastic containers (hereafter referred to as 10-cm depth to compare: (1) the two methods for watering
drought tubs) to allow for water drainage. Pots were placed control pots, and (2) soil VWC differences between a control and
inside drought tubs on top of a 23-cm column of commercial drought pot. Output from these soil water content probes was

Frontiers in Plant Science | www.frontiersin.org 4 January 2020 | Volume 10 | Article 1715

Marchin et al. A Simple Drought Method

TABLE 1 | Eleven tree/shrub species selected for detailed leaf-level physiological measurements, ranked from low to high drought tolerance using mean water potential
(± SE) at turgor loss point (ptlp, MPa) estimated from osmometer measurements of po (n = 16–20 leaves per species). Species abbreviation (used in Figures 6–8) and
natural occurrence are also provided, as well as minimum midday water potential (Ymid, MPa) during the heatwave and mean wood density (± SE, g cm-3; n = 4–5
plants per species).

Species Species Code Natural Occurrence ptlp (MPa) Minimum HW Ymid (MPa) Wood density (g cm-3)

Magnolia grandiflora L. Magr warm temperate forest -1.71 ± 0.02 -1.85 0.46 ± 0.01
Castanospermum australe Caau tropical rainforest -1.71 ± 0.03 -1.73 0.50 ± 0.02
A.Cunn. & C.Fraser
Lophostemon confertus (R.Br.) Loco sclerophyll forest, rainforest -1.71 ± 0.03 -2.65 0.42 ± 0.01
Peter G.Wilson & J.T.Waterh.
Syzygium floribundum F.Muell. Syfl riverine rainforest -1.78 ± 0.04 -2.00 0.56 ± 0.04
Hymenosporum flavum F.Muell. Hyfl sclerophyll forest, rainforest -1.94 ± 0.03 -3.06 0.59 ± 0.01
Toechima erythrocarpum Toer tropical rainforest -1.99 ± 0.06 -2.45 0.64 ± 0.03
(F.Muell.) Radlk.
Harpullia pendula Planch. ex Hape subtropical to tropical rainforest -2.05 ± 0.05 -2.65 0.66 ± 0.03
F.Muell.
Hakea laurina R.Br. Hala semi-arid mallee plain -2.08 ± 0.03 -2.56 0.44 ± 0.02
Hakea salicifolia (Vent.) B.L.Burtt Hasa sclerophyll forest -2.12 ± 0.04 -3.85 0.49 ± 0.004
Grevillea baileyana McGill. Grba tropical rainforest -2.21 ± 0.02 -2.98 0.53 ± 0.01
Buckinghamia celsissima F.Muell. Buce tropical rainforest -2.30 ± 0.02 -2.75 0.55 ± 0.01

For additional species information (horticultural variety, family, growth form, origin), see Supplementary Table S1.

converted into soil VWC using generalised settings for organic The po of fully-expanded, fully-hydrated leaves was also
soil, but further normalization was required to match point measured independently using an osmometer (WP4C Dewpoint
measurements of soil VWC. A constant offset (10%–20%) was PotentiaMeter, Decagon Devices, Pullman, WA, USA) for these 29
applied to the data from each probe to achieve consistency across species (n = 5–10 leaves per species), following the method
all soil VWC measurements. described by Bartlett et al. (2012). Briefly, leaves and/or stems
In the final week of water deficit (days 29–35), we applied an were collected and rehydrated overnight for 12 h using the standing
experimental heatwave to the droughted plants to better rehydration method (Arndt et al., 2015). Rehydration may cause
understand how plants cope with combined stress factors. We solute leakage into the apoplast and underestimation of osmotic
exposed droughted plants to a 7-day heatwave by moving pots potential in some plant species (Kubiske and Abrams, 1991; Arndt
into another glasshouse chamber maintained at higher et al., 2015), but was necessary to ensure fully-hydrated leaves (i.e.,
temperatures than the growth regime. The average heatwave Yleaf ≥ –0.3 MPa) were used for comparison of control and
temperature was 35°C, with a diurnal range of 30°C–41°C and droughted plants. The midrib was removed from leaves before
the maximum temperature spanning 2 h at midday (12:00–14:00; 40-mm diameter leaf discs or leaf pieces were quickly cut from each
Supplementary Figure S1). plant. Leaf pieces were wrapped in foil and frozen in liquid N2 for
2 min, then equilibrated for 10 min inside a sealed, humidified
Pressure-Volume Curves and Leaf plastic bag. Leaf pieces were punctured repeatedly with sharp-tipped
Osmotic Potential forceps immediately before measurement using the osmometer.
Pressure-volume curves were measured for well-watered plants of Measurements were recorded for 20–30 min, until equilibrium as
29 species (n = 1–10 leaves per species) to allow estimation of the indicated by <0.01 MPa change over 2 min. Osmometer
water potential at wilting point (or turgor loss point, ptlp), which is measurements of po were used to estimate species ptlp using the
strongly related to plant drought tolerance (Engelbrecht et al., 2000; following equation: ptlp = 0.832po – 0.631 (Bartlett et al., 2012).
Baltzer et al., 2008). Logistical constraints prevented measurement
of all 50 species. The pressure-volume curves were measured using Plant Physiological Responses to Water
a pressure chamber (Model 1505D, PMS Instrument Company, Deficit Treatment
Albany, OR, USA) following the bench-drying method (Tyree and Eleven of the 50 species, all broadleaf evergreen trees/shrubs
Hammel, 1972; Schulte and Hinckley, 1985). Leaves were (Table 1), were selected to assess the impact of the soil water
rehydrated overnight for 12 h using the standing rehydration deficit on plant function. We collected a set of leaf-level
method (Arndt et al., 2015) to ensure leaves were fully hydrated. physiological measurements: stomatal conductance (gs), leaf water
Leaf fresh weight (g) and leaf water potential (Yleaf, MPa) were potential (Ypre, Ymid), leaf temperature (Tleaf), and leaf osmotic
measured periodically as leaves dried under ambient laboratory potential (po). Control and drought plants (n = 4–5 plants per
conditions. Leaf dry mass was measured after oven-drying at 70°C treatment) were measured under the target soil VWC (34%–41%
for 72 h. We determined ptlp, osmotic potential at full turgor (po), vs. 5%–11%, respectively) during the fourth week of water deficit.
relative water content at turgor loss point (RWCtlp), and bulk Stomatal conductance was measured on sunny days at midday
modulus of elasticity (e) following standard methods (Turner, (11:00–14:00) on three fully-expanded leaves per plant using a
1981) using a pressure-volume curve analysis routine developed porometer (AP-4, Delta-T, Cambridge, UK). The same leaves (n =
by Kevin Tu (available at: http://landflux.org/Tools.php, accessed 1 1–2 leaves per plant) were subsequently removed for measurement
November 2019) and based on Schulte and Hinckley (1985). of Ymid with a pressure chamber (Model 1505D, PMS Instruments).

Frontiers in Plant Science | www.frontiersin.org 5 January 2020 | Volume 10 | Article 1715

Marchin et al. A Simple Drought Method

Measurements of gs and Ymid were repeated on two different moderate water deficit (target soil VWC: 7.5 ± 2.5%) over a period
days and averaged for each plant. Leaves (n = 1–2 leaves per of several weeks (Figure 3). Our method was successfully
plant) for measurement of Ypre were collected at 5:00. All leaves implemented at two glasshouse locations for a diverse set of 50
for water potential measurement were stored inside a sealed, plant species, including grasses, vines, shrubs, and trees
humidified plastic bag and kept cool and dark until (Supplementary Table S1). The largest plants with the highest
measurement, which was completed within 3 h of collection. transpiration rates reached a soil VWC of 10% in about 15 days,
Leaf temperature was measured on sunny days (9:00–14:00) on while the smallest plants required 28 days (Figure 3). There was a
three fully-developed leaves per plant using an infrared significant difference in soil VWC between the control and drought
thermometer (Agri-Therm III Model 6110L, Everest Interscience, treatments (t48 = 33.28, p < 0.001) at maximum drought intensity in
Inc., Chino Hills, OR, USA) held at a distance of about 10 cm from the fourth week of water deficit (Figure 4D). Mean soil VWC was
the leaf surface. Thermal emissivity was set to 0.92, a representative higher in the control treatment, relative to the drought treatment,
value for individual plant leaves (Jones, 2004). for all 50 study species (32%–42% vs. 3%–13%, respectively).
The po was measured in both the fourth and fifth weeks (on Pots that were watered via drip irrigation had larger diurnal
the fifth heatwave day) of drought using an osmometer (WP4C variation in soil VWC, relative to pots watered via capillary
Dewpoint PotentiaMeter, Decagon Devices) following the irrigation inside drought tubs (Figures 4A, C). Either method,
Bartlett et al. (2012) method described above. Leaves (n = 1–2 drip irrigation or capillary irrigation, was capable of maintaining
leaves per plant) were also collected on the fifth heatwave day for soil VWC of control pots within the target range (25%–45%),
measurement of Ymid with a pressure chamber (Model 1505D, although soil VWC was significantly higher in pots watered via drip
PMS Instruments). After the end of the heatwave, all plants irrigation (t26 = 2.84, p = 0.009; Figure 4B). Mean soil VWC of 14
(including drought plants) were well-watered then monitored for drip-irrigated control pots ranged from 32%–40%, while mean soil
survival after a two-week recovery period. VWC of 14 capillary-irrigated control pots (of the same species)
Wood density was determined for the 11 focal species (n = 4–5 ranged from 30%–39%.
plants per species) after the completion of the experiment. A 5-cm
stem segment was split to remove the pith and bark before Plant Physiological Responses to Water
determining fresh volume using the water displacement method. Deficit Treatment
The wood sample was then dried to constant mass at 70°C Leaf osmotic potential was tightly correlated with ptlp across 29
and weighed. species/varieties (r2 = 0.82, p < 0.001; Figure 5, Supplementary

Statistical Analyses
Differences in soil VWC between control and drought pots were
analyzed in the fourth week of drought by using a Student’s t test.
The overall effect of drought on gs, Tleaf, Ypre, Ymid, and po was
determined by using full‐factorial, mixed‐model analyses of
variance (ANOVAs) with species and treatment as the main
effects; species was analyzed as a random effect, and treatment
was analyzed as a fixed effect. When there was a significant
species × treatment interaction (p ≤ 0.05), individual species’
responses were analyzed using Student’s t tests. Separate
paired-sample Student’s t tests were used for each species to
analyze differences in po between the fourth and fifth week of
drought; values were paired by plant. We used ordinary least
squares regression to correlate po with ptlp, then used analysis of
covariance (ANCOVA) to test if the slope and intercept of our
relationship differed from a previously published relationship by
Bartlett et al. (2012). All data were tested for normality with the
Shapiro and Wilk’s test; gs, Tleaf, Ypre, and po measurements were
ln-transformed to achieve normality. All statistical analyses were
completed using R Statistical Software 3.5.1 (R Core Team,
2018). Means were considered significantly different at p ≤ 0.05.

RESULTS FIGURE 3 | Differences in soil volumetric water content (VWC, %) between

control (blue) and drought (coral) pots for two different plant species over the
Achieving Experimental Plant Water 5-week experimental drought period. The fastest-drying species (Lomandra
longifolia, circles) and slowest-drying species (Tristaniopsis laurina, triangles)
Deficits reached the target soil VWC of 10% in 17 and 28 days, respectively, in the
The experimental treatment gradually reduced soil volumetric HIE glasshouse experiment. Values are means of 5 plants per treatment, and
water content of potted plants from field capacity (~35%) to a error bars indicate SE. The target drought intensity is shaded gray.

Frontiers in Plant Science | www.frontiersin.org 6 January 2020 | Volume 10 | Article 1715

Marchin et al. A Simple Drought Method

FIGURE 4 | (A) Changes in soil volumetric water content (VWC, %) over a 3-week period in the austral summer for one control pot watered via drip irrigation on the
bench and one control pot watered via capillary irrigation (i.e., in plastic tub with water level kept constantly at 1-cm depth). Circles indicate point measurements of
soil VWC. (B) Mean soil VWC of 14 control pots watered via drip irrigation and 14 control pots (of the same species) watered via capillary irrigation over a four-week
period; treatments are significantly different (t26 = 2.842, p = 0.009). (C) Changes in soil VWC over the first four weeks of drought for one control pot (watered via
drip irrigation) and one drought pot (watered via capillary irrigation). From 29 January to 2 February, both pots were well-watered using drip irrigation. Circles indicate
point measurements of soil VWC. (D) Species mean soil VWC of control and drought treatments for all 50 species during the fourth week of drought; treatments are
significantly different (t48 = 33.275, p < 0.001). Points are means of 4–5 plants per treatment.

Table S2). This relationship was compared to the relationship 80.177, p < 0.001), for 9 of the 11 species (Figure 6B). For the
published by Bartlett et al. (2012) to verify its use for estimating species with the highest gs (Magnolia grandiflora: 457 mmol m-2
ptlp, and thus drought tolerance, of 11 tree/shrub species used for s-1), Tleaf was an average of 8°C higher under simulated drought.
detailed physiological measurements. The slopes did not differ The two species for which Tleaf did not vary with the water deficit
within the overlapping range of po (‒0.9 to ‒2.3 MPa; F1,48 = treatment had low rates of gs (Lophostemon confertus: 151 mmol
3.181, p = 0.081), although the intercepts were significantly m-2 s-1; T. erythrocarpum: 126 mmol m-2 s-1).
different (F1,49 = 6.910, p = 0.011). The ptlp ranged from –1.71 Reductions in Yleaf in response to the water deficit treatment
to –2.30 MPa and wood density ranged from 0.42‒0.66 g cm-3 for were less consistent across species (Figure 7). Droughted plants
these 11 species (Table 1), whose natural occurrences span the had significantly lower Ypre than control plants (F1,10 = 69.26, p <
semi-arid plains to rainforest habitats within Australia. The four 0.001) for 9 of the 11 species, but all species maintained mean
species in the Proteaceae family had the lowest ptlp (<–2.0 MPa), Ypre >–1.5 MPa (Figure 7A). Only six species had significantly
indicating high drought tolerance. lower Ymid in droughted plants, relative to controls (F1,10 =
Droughted plants of all 11 species had significantly lower 33.43, p < 0.001; Figure 7B). While control plants of all species
rates of gs, relative to control plants (F1,10 = 619.29, p < 0.001) in maintained Ymid at or above their turgor loss point, droughted
the fourth week of water deficit (16–62 vs. 126–457 mmol m-2 s-1, plants of five species had Ymid that exceeded their turgor loss
respectively; Figure 6A). For species with high rates of water loss, point (Figure 7B). Severe wilting was observed in L. confertus
gs was 6–17 times higher in control plants than in droughted and Hymenosporum flavum, but wilting was less obvious for
plants. In the species with the lowest rate of water loss (Toechima three other species with Ymid below their turgor loss point
erythrocarpum), g s of control plants was double that of (T. erythrocarpum, Harpullia pendula, Hakea salicifolia).
droughted plants. Leaf temperature was also significantly There was no evidence of osmotic adjustment in any species
higher for droughted plants, relative to control plants (F1,10 = in the fourth week of water deficit (F1,10 = 1.22, p = 0.272). Three

Frontiers in Plant Science | www.frontiersin.org 7 January 2020 | Volume 10 | Article 1715

Marchin et al. A Simple Drought Method

(Maseda and Fernández, 2016). Capillary irrigation is used to

control the soil water content of potted plants, which are placed
above a solid column of low water permeability with an
adjustable water table (Figure 1, Supplementary Figure S2).
In these previous experiments, plants were grown in pure sand,
which has a relatively high bulk density and low porosity. The
use of pure sand can restrict root growth and has likely limited
the use of this water deficit method in other studies. Instead of
pure sand, we used a potting mix with low bulk density (0.45 ±
0.004 g cm-3), which is ideal for root growth as it allows
movement of air, water, and nutrients through the soil. The
higher water-holding capacity of the potting mix, relative to pure
sand, resulted in different rates of drying for species with large
differences in plant size and hence transpiration rate (Marchin,
unpublished; Figure 3). Although a difference in drying rate may
not be appropriate for all studies, it was acceptable for our
primary aim, which was to establish a constant water deficit
treatment that could be contrasted to the well-watered control
treatment in a large number of species and replicates.
There are several advantages of our simple water deficit
method that are ideal for ecological studies. Perhaps most
FIGURE 5 | Relationship between leaf osmotic potential (po, MPa) and the
importantly, it requires minimal maintenance throughout the
water potential at leaf turgor loss point (ptlp, MPa) for 29 plant species (n =
1–10 leaves per species; r2 = 0.82, p < 0.001). Leaf osmotic potential was experiment and is relatively inexpensive to implement. Further,
measured with an osmometer, and ptlp was calculated from pressure-volume the method does not require electricity, wiring, or application of
curves. The dashed line is the relationship published by Bartlett et al. (2012), chemicals, and as such, it can be applied inside growth chambers
which includes 30 woody species. The slope does not differ between the two if necessary. We have established that pure sand is not a vital
relationships within the overlapping range of po (F1,48 = 3.181, p = 0.081).
requirement, suggesting that any potting media or soil type could
be adapted for use with this method. We have also shown that
species (M. grandiflora, H. flavum, T. erythrocarpum) showed control plants can be watered via drip irrigation, rather than by
the opposite response, however, by having significantly higher po capillary irrigation inside plastic tubs (Figure 4B), which further
in control plants, relative to drought plants (Figure 8A). cuts the materials cost and space requirements. With only a few
Droughted plants were exposed to heatwave temperatures in manipulations of the adjustable water table, it is possible to
the fifth week of drought, which resulted in significant osmotic establish a broad range of soil water deficits of differing intensity,
adjustment (F1,10 = 4.73, p = 0.032) for 5 of the 11 species (Figure duration, and pulsation (Figure 2). For example, mild water
8B). For these species, mean po decreased by –0.15 to –0.54 MPa deficits could be established by using a lower water table depth
in just one week. The magnitude of this adjustment surpassed (e.g., ≤10 cm), whereas severe water deficits require a greater
mean po for control plants of M. grandiflora, H. flavum, and depth (e.g., ≥25 cm; Figure 2A). It is straightforward to simulate
T. erythrocarpum in the fourth week of water deficit. Interestingly, gradual droughts of long duration (weeks or months; Figure 2B),
rapid osmotic adjustment was observed for the species with the since daily maintenance consists of simply tracking and adjusting
lowest Ymid during the heatwave (‒3.9 MPa, H. salicifolia, water depths inside plastic drought tubs. Repeated wetting and
Table 1). Plants of several species showed high levels of leaf drying cycles (Figure 2C) can also be simulated simply by raising
desiccation (>50%; H. salicifolia, L. confertus, T. erythrocarpum) as and lowering the water table following the desired schedule. We
a result of the combination of drought and heat stress, leading to have also shown that it is possible to incorporate interacting
the death of four H. salicifolia plants and one T. erythrocarpum stress factors, such as a heatwave, into the experimental design.
plant in the following weeks (data not shown). Multi-factorial experiments are needed in ecological and
physiological investigations, as the importance of determining
DISCUSSION the interaction of environmental factors is increasingly
recognized as critical for our understanding of the impacts of
We modified an experimental method first introduced almost 50 global change (Dermody, 2006).
years ago (Haan and Barfield, 1971) to apply a gradual, moderate It is important to consider the desired plant physiological
water deficit (Figure 3) to a diverse set of 50 plant species, responses before selecting an experimental water deficit
including grasses, vines, shrubs, and trees (Supplementary Table treatment, although responses to the same treatment can often
S1). The original method was previously modified to eliminate differ among species with different water-use strategies
the need for complex float chambers, but has rarely been used, (Figures 7 and 8). The simulated drought treatment described
and only applied to small numbers of species such as desert here resulted in a universal decrease in gs (Figure 6A) across a
grasses (Fernández and Reynolds, 2000) and Eucalyptus trees broad range of species with different drought tolerances

Frontiers in Plant Science | www.frontiersin.org 8 January 2020 | Volume 10 | Article 1715

Marchin et al. A Simple Drought Method

FIGURE 6 | Differences in (A) rates of stomatal conductance (gs, mmol m-2 s-1) and (B) leaf temperature (Tleaf, °C) between control and drought plants of 11
selected tree/shrub species. Measurements were completed during the fourth week of drought, when species mean soil volumetric water content (VWC) was
34%–41% in control pots and 5%–11% in drought pots. Species are ordered from low to high drought tolerance according to ptlp. Species are denoted according
to abbreviations in Table 1. Values are means of 4–5 plants, and error bars indicate SE (unidirectional SE are presented for clarity). Asterisks denote significant
differences between treatments: *p < 0.05; **p < 0.01; ***p < 0.001.

(Table 1). Stomatal conductance of droughted plants was 2–17 during the moderate water deficit until after heatwave
times less than that of control plants (Figure 6A). Therefore, this temperatures were applied. Therefore, our moderate water
same treatment could feasibly be applied to any study aiming to deficit can be easily applied to bring plants very close to their
examine plant responses resulting from differences in physiological limits while ensuring low mortality rates
transpiration rates. For example, studies aimed at (if desired).
understanding plant-insect interactions during drought could
use this same moderate water deficit treatment. Our Simulating Drought and Heat
experimental treatment did not, however, decrease Ymid for all Stress Interactions
study species (Figure 7B). The most drought-tolerant species, The experimental water deficit treatment led to some coherent
Grevillea baileyana and Buckinghamia celsissima, maintained plant physiological responses. All species within a subset of 11
Ymid at or above their ptlp in control and droughted plants, at evergreen trees/shrubs partially closed their stomata in response
least before the heatwave. If drought-related changes in Yleaf are to simulated drought (Figure 6A), resulting in higher Tleaf for
an important experimental aim, the method can be easily droughted plants (Figure 6B). Stomatal conductance and
adjusted to impose a greater intensity and/or longer duration transpiration result in evaporative cooling of leaves (Nobel,
water deficit than was used in these experiments. For example, 1974; Farquhar and Sharkey, 1982), so partial stomatal closure
studies aimed at understanding the physiological processes that under water deficit can result in higher Tleaf (Medina and Gilbert,
occur during drought mortality would require a longer and/or 2015). Transpiration alone can cool leaves by at least 2°C–3°C
more severe water deficit. While some species wilted and/or (Lin et al., 2017) and up to 8°C for species with high transpiration
dropped leaves during our simulated drought, no plant died rates (e.g., M. grandiflora, Figure 6). About half of the species (5

Frontiers in Plant Science | www.frontiersin.org 9 January 2020 | Volume 10 | Article 1715

Marchin et al. A Simple Drought Method

FIGURE 7 | Differences in (A) predawn leaf water potential (Ypre, MPa) and (B) midday leaf water potential (Ymid, MPa) between control and drought plants of 11
tree/shrub species. The gray points in panel (B) indicate species mean water potential at leaf turgor loss point (ptlp). Measurements were completed during the fourth
week of drought, when species mean soil volumetric water content (VWC) was 34%–41% in control pots and 5%–11% in drought pots. Species are ordered from
low to high drought tolerance according to ptlp. Species are denoted according to abbreviations in Table 1. Values are means of 4–5 plants, and error bars indicate
SE (unidirectional SE are presented for clarity). Asterisks as in Figure 6.

out of 11) experienced declines in Ymid that exceeded their ptlp for osmotic adjustment can acclimate to drought stress within 6–
(Figure 7B), despite partial stomatal closure (Figure 6A). 15 days (Blum et al., 1980; Molinari et al., 2004; Blum, 2017), but
Stomata respond to variations in leaf (or guard cell) water our results suggest osmotic adjustment requires a longer
potential (Peak and Mott, 2011; Buckley, 2016) and close response time (i.e., months) in horticultural and wild species.
before reaching thresholds for xylem cavitation (Bartlett et al., Interestingly, rapid osmotic adjustment was observed for five
2016; Martin-StPaul et al., 2017), thus limiting tissue damage species when water deficit was combined with heat stress (Figure
under water deficit. The intensity and duration of our 8B). This osmotic adjustment occurred in about one week,
experimental treatment led to wilting and/or leaf loss for some, between the fourth and fifth weeks of water deficit, after 5 days
but not all, study species (Marchin, unpublished), indicating the of heatwave temperatures. The importance of po for high thermal
simulated drought stress was moderate overall. tolerance was first noted for plants growing in extreme
Osmotic adjustment was not observed for any species in the environments, such as the deserts of Western Australia
fourth week of water deficit (Figure 8A). Some plant species may (Hellmuth, 1971) and Death Valley, USA (Seemann et al.,
not be capable of osmotic adjustment, but it is a widely prevalent 1986). Increases in cell sugar concentrations and the resulting
response to drought, with only 15% of measured species showing decrease in po may provide a mechanism for rapid temperature
no seasonal adjustment of po (Bartlett et al., 2014). There are two acclimation in water-stressed leaves (Santarius, 1973; Huve et al.,
mechanisms for osmotic adjustment of plant leaves: (1) 2006). In our glasshouse experiments, decreases in po were
accumulation of organic solutes (e.g., glycine betaine, proline, observed for trees/shrubs that naturally occur in diverse
sugars) and (2) increasing inorganic ion concentrations (e.g., K+, environments, including temperate forests, tropical rainforests,
Ca2+; Chen and Jiang, 2010). Crop varieties with a high capacity and sclerophyll forests. These results suggest rapid osmotic

Frontiers in Plant Science | www.frontiersin.org 10 January 2020 | Volume 10 | Article 1715

Marchin et al. A Simple Drought Method

FIGURE 8 | (A) Differences in leaf osmotic potential (po, MPa) between control and drought plants of 11 tree/shrub species. Measurements were completed during
the fourth week of drought, when species mean soil volumetric water content (VWC) was 34%–41% in control pots and 5%–11% in drought pots. (B) Differences in
po of plants between the fourth and fifth week of drought; in week 5, drought plants were exposed to heatwave (HW) temperatures. Species mean soil VWC was
3%–11% in both weeks. Species are ordered from low to high drought tolerance according to ptlp. Species are denoted according to abbreviations in Table 1.
Values are means of 3–5 plants, and error bars indicate SE (unidirectional SE are presented for clarity). Asterisks as in Figure 6.

adjustment may be a widespread mechanism for plant tolerance CONCLUSIONS

of the combination of drought and heat stress. Despite the
decrease in po, however, four H. salicifolia plants and one We have described a simple method for simulating drought
T. erythrocarpum plant died as a result of drought and heat effects on plants in glasshouse experiments. Our approach can be
stress. Hakea salicifolia had relatively low wood density and the easily applied to investigate drought responses: (1) of large
lowest Ymid during the heatwave (Table 1), so it is possible that numbers of species, provenances, genotypes, etc., (2) to
xylem embolism occurred and restricted access to water even different intensities of a constant stress (mild, moderate,
after the drought and heatwave ended. Hydraulic failure is a key severe), (3) to gradual water deficits of long duration (i.e.,
mechanism leading to tree mortality during drought (Hoffmann months), and (4) combined with interacting abiotic or biotic
et al., 2011; Anderegg et al., 2016), but further research is factors (e.g., heatwaves, atmospheric CO 2 concentration,
required to determine why species such as H. salicifolia may be nutrient level, mycorrhizal symbiosis, insect pest or pathogen
particularly vulnerable. The combination of heat and drought presence, etc.). One limitation to the method is that it resulted in
stress can quickly kill large swathes of trees (Allen et al., 2010; different rates of drying for species with large differences in leaf
Williams et al., 2012; Choat et al., 2018), but our current area and transpiration rate, which may not be appropriate for all
understanding of the physiological mechanisms preceding studies. It is also important to note that the method was tested in
drought-induced tree mortality is incomplete. Our results a highly-controlled glasshouse, and variations or fluctuations of
indicate that droughted plants are at greater risk of thermal soil drying profiles are to be expected when attempting to use this
damage (Figures 6B and 7B) and provide valuable insights into method in less-controlled environments, where, for instance,
how plants cope with multiple stresses (Figure 8). changes in PAR or VPD can affect plant evapotranspiration and

Frontiers in Plant Science | www.frontiersin.org 11 January 2020 | Volume 10 | Article 1715

Marchin et al. A Simple Drought Method

thus drying rates. As demonstrated here, this method can be used FUNDING
to rank species according to drought tolerance and elucidate species’
differences in physiological mechanisms for coping with drought This is a contribution from the Which Plant Where project,
and heat stress. The most drought-tolerant species in our study was which is funded by the Green Cities Fund, as part of the Hort
Buckinghamia celsissima (ptlp = ‒2.3 MPa), based on the ptlp Frontiers Strategic Partnership Initiative developed by Hort
ranking, whereas Castanospermum australe (ptlp = ‒1.7 MPa) is Innovation Australia, with co-investment from Macquarie
sensitive to drought and should be avoided for urban plantings in University, Western Sydney University, and the NSW
drought-prone cities with frequent water limitations or imposed Department of Planning, Industry and Environment, and
restrictions. The most vulnerable species to drought in combination funds from the Australian Government.
with heat stress was Hakea salicifolia (ptlp = ‒2.1 MPa), indicating
that the interactive effects of heat and drought stress are complex
and cannot easily be predicted based on measurement of ptlp alone. ACKNOWLEDGMENTS
We recommend this simple water deficit method for the study of
plant drought responses in a range of ecological contexts, The authors would like to thank Matthew Alfonzetti, David
particularly those involving other abiotic or biotic effects. Coleman, Ellen Curtis, Elise Dando, Yanisa Olaranont, Camille
Parise, and Nikki Sperinck for their technical assistance during
glasshouse experiments, plus Craig Barton, Andrew Gherlenda,
DATA AVAILABILITY STATEMENT and Muhammad Masood for assistance with collecting
glasshouse environmental data and running glasshouse
The raw data supporting the conclusions of this article will be
operations. Leigh Staas is kindly acknowledged for help with
made available by the authors, without undue reservation, to any
compiling species lists, while Hendrik Poorter and the two
qualified researcher.
reviewers are thanked for their helpful comments on drafts of
this manuscript.
AUTHOR CONTRIBUTIONS
ML and DE conceived the ideas and the overall study design. RM SUPPLEMENTARY MATERIAL
and AO designed the methodology and collected the data. RM
analysed the data. RM led the writing of the manuscript. All The Supplementary Material for this article can be found online
authors contributed critically to the drafts and gave final at: https://www.frontiersin.org/articles/10.3389/fpls.
approval for publication. 2019.01715/full#supplementary-material

REFERENCES Blum, A., Sinmena, B., and Ziv, O. (1980). An evaluation of seed and seedling
drought tolerance screening tests in wheat. Euphytica 29 (3), 727–736. doi:
Allen, C. D., Macalady, A. K., Chenchouni, H., Bachelet, D., McDowell, N., 10.1007/bf00023219
Vennetier, M., et al. (2010). A global overview of drought and heat-induced Blum, A. (2017). Osmotic adjustment is a prime drought stress adaptive engine in
tree mortality reveals emerging climate change risks for forests. For. Ecol. support of plant production. Plant Cell Environ. 40 (1), 4–10. doi: 10.1111/
Manage. 259 (4), 660–684. doi: 10.1016/j.foreco.2009.09.001 pce.12800
Anderegg, W. R. L., Klein, T., Bartlett, M., Sack, L., Pellegrini, A. F. A., Choat, B., BoM (2019). Widespread heatwaves during December 2018 and January 2019
et al. (2016). Meta-analysis reveals that hydraulic traits explain cross-species (Bureau of Meteorology Special Climate Statement 68, Commonwealth of
patterns of drought-induced tree mortality across the globe. Proc. Natl. Acad. Australia).
Sci. U.S.A. 113 (18), 5024–5029. doi: 10.1073/pnas.1525678113 Buckley, T. N. (2016). Stomatal responses to humidity: has the ‘black box’ finally
Arndt, S. K., Irawan, A., and Sanders, G. J. (2015). Apoplastic water fraction and been opened? Plant Cell Environ. 39 (3), 482–484. doi: 10.1111/pce.12651
rehydration techniques introduce significant errors in measurements of Bunce, J. A., and Nasyrov, M. (2012). A new method of applying a controlled soil
relative water content and osmotic potential in plant leaves. Physiol. Plant. water stress, and its effect on the growth of cotton and soybean seedlings at
155 (4), 355–368. doi: 10.1111/ppl.12380 ambient and elevated carbon dioxide. Environ. Exp. Bot. 77, 165–169. doi:
Baltzer, J. L., Davies, S. J., Bunyavejchewin, S., and Noor, N. S. M. (2008). The role 10.1016/j.envexpbot.2011.11.015
of desiccation tolerance in determining tree species distributions along the Chen, H., and Jiang, J. G. (2010). Osmotic adjustment and plant adaptation to
Malay-Thai Peninsula. Funct. Ecol. 22 (2), 221–231. doi: 10.1111/j.1365- environmental changes related to drought and salinity. Environ. Rev. 18, 309–
2435.2007.01374.x 319. doi: 10.1139/a10-014
Bartlett, M. K., Scoffoni, C., Ardy, R., Zhang, Y., Sun, S. W., Cao, K. F., et al. (2012). Choat, B., Brodribb, T. J., Brodersen, C. R., Duursma, R. A., Lopez, R., and Medlyn,
Rapid determination of comparative drought tolerance traits: using an B. E. (2018). Triggers of tree mortality under drought. Nature 558 (7711), 531–
osmometer to predict turgor loss point. Methods Ecol. Evol. 3 (5), 880–888. 539. doi: 10.1038/s41586-018-0240-x
doi: 10.1111/j.2041-210X.2012.00230.x Dermody, O. (2006). Mucking through multifactor experiments; design and
Bartlett, M. K., Zhang, Y., Kreidler, N., Sun, S. W., Ardy, R., Cao, K. F., et al. analysis of multifactor studies in global change research. New Phytol. 172
(2014). Global analysis of plasticity in turgor loss point, a key drought tolerance (4), 598–600. doi: 10.1111/j.1469-8137.2006.01921.x
trait. Ecol. Lett. 17 (12), 1580–1590. doi: 10.1111/ele.12374 Dey, R., Lewis, S. C., Arblaster, J. M., and Abram, N. J. (2019). A review of past and
Bartlett, M. K., Klein, T., Jansen, S., Choat, B., and Sack, L. (2016). The correlations projected changes in Australia’s rainfall. Wiley Interdiscip. Rev.: Climate
and sequence of plant stomatal, hydraulic, and wilting responses to drought. Change 10 (3), e577. doi: 10.1002/wcc.577
Proc. Natl. Acad. Sci. U.S.A. 113 (46), 13098–13103. doi: 10.1073/ Earl, H. J. (2003). A precise gravimetric method for simulating drought stress in
pnas.1604088113 pot experiments. Crop Sci. 43 (5), 1868–1873. doi: 10.2135/cropsci2003.1868

Frontiers in Plant Science | www.frontiersin.org 12 January 2020 | Volume 10 | Article 1715

Marchin et al. A Simple Drought Method

Engelbrecht, B. M. J., Velez, V., and Tyree, M. T. (2000). Hydraulic conductance of Medina, V., and Gilbert, M. E. (2015). Physiological trade-offs of stomatal closure
two co-occuring neotropical understory shrubs with different habitat under high evaporative gradients in field grown soybean. Funct. Plant Biol. 43
preferences. Ann. For. Sci. 57 (3), 201–208. doi: 10.1051/forest:2000111 (1), 40–51. doi: 10.1071/FP15304
FAO (2018). 2017 The impact of disasters and crises on agriculture and food security Mexal, J., Fisher, J. T., Osteryoung, J., and Reid, C. P. P. (1975). Oxygen availability
(Rome, Italy: Food and Agriculture Organization of the United Nations). in polyethylene glycol solutions and its implications in plant-water relations.
Farquhar, G. D., and Sharkey, T. D. (1982). Stomatal conductance and Plant Physiol. 55 (1), 20–24. doi: 10.1104/pp.55.1.20
photosynthesis. Ann. Rev. Plant Physiol. 33 (1), 317–345. doi: 10.1146/ Molinari, H. B. C., Marur, C. J., Bespalhok, J. C., Kobayashi, A. K., Pileggi, M.,
annurev.pp.33.060182.001533 Leite, R. P., et al. (2004). Osmotic adjustment in transgenic citrus rootstock
Fernández, R. J., and Reynolds, J. F. (2000). Potential growth and drought Carrizo citrange (Citrus sinensis Osb. x Poncirus trifoliata L. Raf.)
tolerance of eight desert grasses: lack of a trade-off? Oecologia 123 (1), 90– overproducing proline. Plant Sci. 167 (6), 1375–1381. doi: 10.1016/
98. doi: 10.1007/s004420050993 j.plantsci.2004.07.007
Freund, M., Henley, B. J., Karoly, D. J., Allen, K. J., and Baker, P. J. (2017). Morgan, J. M. (1984). Osmoregulation and water stress in higher plants. Annu.
Multi-century cool- and warm-season rainfall reconstructions for Australia’s Rev. Plant Physiol. Plant Mol. Biol. 35, 299–319. doi: 10.1146/
major climatic regions. Clim. Past 13 (12), 1751–1770. doi: 10.5194/cp-13- annurev.pp.35.060184.001503
1751-2017 Munns, R., James, R. A., Sirault, X. R. R., Furbank, R. T., and Jones, H. G. (2010).
Griffin, D., and Anchukaitis, K. J. (2014). How unusual is the 2012-2014 California New phenotyping methods for screening wheat and barley for beneficial
drought? Geophys. Res. Lett. 41 (24), 9017–9023. doi: 10.1002/2014gl062433 responses to water deficit. J. Exp. Bot. 61 (13), 3499–3507. doi: 10.1093/jxb/
Haan, C. T., and Barfield, B. J. (1971). Controlling the soil moisture environment erq199
of transpiring plants. Plant Soil 35 (1-3), 439–443. doi: 10.1007/bf01372675 Naumann, G., Alfieri, L., Wyser, K., Mentaschi, L., Betts, R. A., Carrao, H., et al.
Harb, A., Krishnan, A., Ambavaram, M. M. R., and Pereira, A. (2010). Molecular (2018). Global changes in drought conditions under different levels of
and physiological analysis of drought stress in Arabidopsis reveals early warming. Geophys. Res. Lett. 45 (7), 3285–3296. doi: 10.1002/2017gl076521
responses leading to acclimation in plant growth. Plant Physiol. 154 (3), Nicholls, N., Drosdowsky, W., and Lavery, B. (1997). Australian rainfall variability
1254–1271. doi: 10.1104/pp.110.161752 and change. Weather 52 (3), 66–72. doi: 10.1002/j.1477-8696.1997.tb06274.x
Hellmuth, E. O. (1971). Ecophysiological studies on plants in arid and semi-arid Nitschke, C. R., Nichols, S., Allen, K., Dobbs, C., Livesley, S. J., Baker, P. J., et al.
regions in Western Australia: V. Heat resistance limits of photosynthetic (2017). The influence of climate and drought on urban tree growth in southeast
organs of different seasons, their relation to water deficits and cell sap Australia and the implications for future growth under climate change.
properties and the regeneration ability. J. Ecol. 59 (2), 365. doi: 10.2307/ Landscape Urban Plan. 167, 275–287. doi: 10.1016/j.landurbplan.2017.06.012
2258318 Nobel, P. (1974). Introduction to biophysical plant physiology (San Francisco, CA:
Hoffmann, W. A., Marchin, R. M., Abit, P. P., and Lau, O. L. (2011). Hydraulic WH Freeman & Company).
failure and tree dieback are associated with high wood density in a temperate Peak, D., and Mott, K. A. (2011). A new, vapour-phase mechanism for stomatal
forest under extreme drought. Glob. Change Biol. 17, 2731–2742. doi: 10.1111/ responses to humidity and temperature. Plant Cell Environ. 34 (1), 162–178.
j.1365-2486.2011.02401.x doi: 10.1111/j.1365-3040.2010.02234.x
Hsiao, T. C., Acevedo, E., Fereres, E., and Henderson, D. W. (1976). Water stress, Perkins, S. E., and Alexander, L. V. (2013). On the measurement of heat waves.
growth, and osmotic adjustment. Philos. T. R. Soc B. 273 (927), 479–500. doi: J. Climate 26 (13), 4500–4517. doi: 10.1175/jcli-d-12-00383.1
10.1098/rstb.1976.0026 Poorter, H., Fiorani, F., Stitt, M., Schurr, U., Finck, A., Gibon, Y., et al. (2012). The
Huve, K., Bichele, I., Tobias, M., and Niinemets, U. (2006). Heat sensitivity of art of growing plants for experimental purposes: a practical guide for the plant
photosynthetic electron transport varies during the day due to changes in biologist. Funct. Plant Biol. 39 (11), 821–838. doi: 10.1071/FP12028
sugars and osmotic potential. Plant Cell Environ. 29 (2), 212–228. doi: 10.1111/ R Core Team (2018). R: A language and environment for statistical computing
j.1365-3040.2005.01414.x (Vienna, Austria: R Foundation for Statistical Computing).
IPCC (2018). “Global Warming of 1.5°C: Summary for Policymakers,” in Global Santarius, K. A. (1973). The protective effect of sugars on chloroplast membranes
warming of 1.5°C. An IPCC Special Report on the impacts of global warming of 1.5°C during temperature and water stress and its relationship to frost, desiccation
above pre-industrial levels and related global greenhouse gas emission pathways, in and heat resistance. Planta 113 (2), 105–114. doi: 10.1007/bf00388196
the context of strengthening the global response to the threat of climate change, Schulte, P. J., and Hinckley, T. M. (1985). A Comparison of Pressure-Volume
sustainable development, and efforts to eradicate poverty. Eds. V. Masson-Delmotte, Curve Data-Analysis Techniques. J. Exp. Bot. 36, 1590–1602. doi: 10.1093/jxb/
P. Zhai, H. O. Pörtner, D. Roberts, J. Skea, P. R. Shukla, A. Pirani, W. Moufouma- 36.10.1590
Okia, C. Péan, R. Pidcock, S. Connors, J. B. R. Matthews, Y. Chen, X. Zhou, M. I. Seemann, J. R., Downton, W. J. S., and Berry, J. A. (1986). Temperature and leaf
Gomis, E. Lonnoy, T. Maycock, M. Tignor and T. Waterfield, (Geneva, osmotic potential as factors in the acclimation of photosynthesis to high
Switzerland: World Meteorological Organization), pp. 32 temperature in desert plants. Plant Physiol. 80 (4), 926–930. doi: 10.1104/
Kogan, F. N. (1997). Global drought watch from space. B. Am. Meteorol. Soc 78 pp.80.4.926
(4), 621–636. doi: 10.1175/1520-0477(1997)078<0621:Gdwfs>2.0.Co;2 Snow, M. D., and Tingey, D. T. (1985). Evaluation of a system for the imposition of
Kubiske, M. E., and Abrams, M. D. (1991). Rehydration effects on pressure- plant water stress. Plant Physiol. 77 (3), 602–607. doi: 10.1104/pp.77.3.602
volume relationships in four temperate woody species: variability with site, Steinberg, S. L., and Henninger, D. L. (1997). Response of the water status of
time of season and drought conditions. Oecologia 85 (4), 537–542. doi: soybean to changes in soil water potentials controlled by the water pressure in
10.1007/bf00323766 microporous tubes. Plant Cell Environ. 20 (12), 1506–1516. doi: 10.1046/
Lewis, S. C., King, A. D., and Mitchell, D. M. (2017). Australia’s unprecedented j.1365-3040.1997.d01-46.x
future temperature extremes under Paris limits to warming. Geophys. Res. Lett. Trueba, S., Pan, R. H., Scoffoni, C., John, G. P., Davis, S. D., and Sack, L. (2019).
44 (19), 9947–9956. doi: 10.1002/2017gl074612 Thresholds for leaf damage due to dehydration: declines of hydraulic function,
Lin, H., Chen, Y. J., Zhang, H. L., Fu, P. L., and Fan, Z. X. (2017). Stronger cooling stomatal conductance and cellular integrity precede those for photochemistry.
effects of transpiration and leaf physical traits of plants from a hot dry habitat New Phytol. 223 (1), 134–149. doi: 10.1111/nph.15779
than from a hot wet habitat. Funct. Ecol. 31 (12), 2202–2211. doi: 10.1111/ Turner, N. C. (1981). Techniques and experimental approaches for the
1365-2435.12923 measurement of plant water status. Plant Soil 58 (1-3), 339–366. doi:
Martin-StPaul, N., Delzon, S., and Cochard, H. (2017). Plant resistance to drought 10.1007/bf02180062
depends on timely stomatal closure. Ecol. Lett. 20 (11), 1437–1447. doi: Tyree, M., and Hammel, H. (1972). The measurement of the turgor pressure and
10.1111/ele.12851 the water relations of plants by the pressure-bomb technique. J. Exp. Bot. 23,
Maseda, P. H., and Fernández, R. J. (2016). Growth potential limits drought 267–282. doi: 10.1093/jxb/23.1.267
morphological plasticity in seedlings from six Eucalyptus provenances. Tree Williams, A. P., Allen, C. D., Macalady, A. K., Griffin, D., Woodhouse, C. A.,
Physiol. 36 (2), 243–251. doi: 10.1093/treephys/tpv137 Meko, D. M., et al. (2012). Temperature as a potent driver of regional forest

Frontiers in Plant Science | www.frontiersin.org 13 January 2020 | Volume 10 | Article 1715

Marchin et al. A Simple Drought Method

drought stress and tree mortality. Nat. Clim. Change 3, 292. doi: 10.1038/ Zur, B. (1966). Osmotic control of the matric soil–water potential: I. Soil–water
nclimate1693 system. Soil Sci. 102, 394–398. doi: 10.1097/00010694-196612000-00007
Wookey, P. A., Atkinson, C. J., Mansfield, T. A., and Wilkinson, J. R. (1991).
Control of plant water deficits using the ‘Snow and Tingey system’ and their
influence on the water relations and growth of sunflower. J. Exp. Bot. 42 (238), Conflict of Interest: The authors declare that the research was conducted in the
589–595. doi: 10.1093/jxb/42.5.589 absence of any commercial or financial relationships that could be construed as a
Xie, W., Xiong, W., Pan, J., Ali, T., Cui, Q., Guan, D. B., et al. (2018). Decreases in potential conflict of interest.
global beer supply due to extreme drought and heat. Nat. Plants 4 (11), 964–
973. doi: 10.1038/s41477-018-0263-1 Copyright © 2020 Marchin, Ossola, Leishman and Ellsworth. This is an open-access
Yeo, A. R., and Flowers, T. J. (1984). Nonosmotic effects of polyethylene glycols article distributed under the terms of the Creative Commons Attribution License (CC
upon sodium transport and sodium-potassium selectivity by rice roots. Plant BY). The use, distribution or reproduction in other forums is permitted, provided the
Physiol. 75 (2), 298–303. doi: 10.1104/pp.75.2.298 original author(s) and the copyright owner(s) are credited and that the original
Zargar, A., Sadiq, R., Naser, B., and Khan, F. I. (2011). A review of drought indices. publication in this journal is cited, in accordance with accepted academic practice. No
Environ. Rev. 19, 333–349. doi: 10.1139/a11-013 use, distribution or reproduction is permitted which does not comply with these terms.

Frontiers in Plant Science | www.frontiersin.org 14 January 2020 | Volume 10 | Article 1715