Research

Baobab trees (Adansonia) in Madagascar use stored water

Blackwell Publishing Ltd

to flush new leaves but not to support stomatal opening

before the rainy season
Saharah Moon Chapotin1, Juvet H. Razanameharizaka2 and N. Michele Holbrook1
1
Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA 02138, USA; 2Département de Biologie et Ecologie Végétales,
Université d’Antananarivo, Antananarivo (101), Madagascar

Summary

Author for correspondence: • Baobab trees (Adansonia, Bombacaceae) are widely thought to store water in
Saharah Moon Chapotin their stems for use when water availability is low. We tested this hypothesis by
Tel: +1 617 496 3580 assessing the role of stored water during the dry season in three baobab species in
Fax: +1 617 496 5854
Email: chapotin@post.harvard.edu
Madagascar.
• In the dry season, leaves are present only during and after leaf flush. We quantified
Received: 11 August 2005
the relative contributions of stem and soil water during this period through measures
Accepted: 10 October 2005
of stem water content, sap flow and stomatal conductance.
• Rates of sap flow at the base of the trunk were near zero, indicating that leaf flushing
was almost entirely dependent on stem water. Stem water content declined by up
to 12% during this period, yet stomatal conductance and branch sap flow rates
remained very low.
• Stem water reserves were used to support new leaf growth and cuticular transpi-
ration, but not to support stomatal opening before the rainy season. Stomatal opening
coincided with the onset of sap flow at the base of the trunk and occurred only after
significant rainfall.

Key words: Adansonia, Bombacaceae, leaf flushing, phenology, sap flow, stomatal
conductance, tropical dry forest, water storage.

New Phytologist (2006) 169: 549–559

© The Authors (2005). Journal compilation © New Phytologist (2005)

doi: 10.1111/j.1469-8137.2005.01618.x

to maximize daily stomatal opening during the rainy season;

Introduction
Stem-succulent trees are prominent in arid tropical ecosystems
around the world, and baobab trees (Adansonia L.; Bombacaceae)
are among the best-known examples. Because of the high
water content and large volume of their stems and branches,
and their occurrence in seasonally dry environments, it has
long been assumed that baobab trees depend on stored water
during periods of low water availability, and replenish the
stored water during periods of rainfall or reduced water demand
(Newton, 1974; Owen, 1974; Wickens, 1983; Baum, 1996;
Sorg & Rohner, 1996). The actual role of water storage in
baobab trees, however, had not been examined until recently,
and had simply been assumed from their physiognomy. In a
related study we found no evidence that stored water was used
instead, baobab trees appear to prioritize conduit safety and
turgor maintenance (Chapotin et al., 2006). The goal of the
present study was to understand the role of stem water during
the dry season. Baobab trees, like all stem-succulent trees,
are leafless during most of the dry season, and new leaves
are flushed only towards the end of the dry season. As this
phenological pattern considerably shortens their growing
season, the ability to use stored water for early leaf flushing
could be advantageous.
Stem water storage plays a physiological role in many
different plants, and contributes to their ability to survive in a
diverse range of ecosystems. Arborescent palms can use water
stored in the stem to buffer xylem potentials and support
leaves when water availability is limited (Holbrook & Sinclair,

www.newphytologist.org 549
550 Research
1992), as can large, transpiring conifers (Waring & Running,
1978; Waring et al., 1979) and Andean rosette species
(Goldstein et al., 1984), while some tropical forest trees use
stored water on a daily basis (Goldstein et al., 1998). Fewer
studies have examined the role of water storage in deciduous
trees (Nilsen et al., 1990; Borchert, 1994a, 1994b; Machado
& Tyree, 1994; Stratton et al., 2000; Cedeño, 2001). Many
deciduous, dry forest trees have large, swollen stems with a
high water content (Holbrook et al., 1995), and the propor-
tion of species with this growth form increases in drier areas
(Medina, 1995). Contrary to expectation from their high
water-storage capacity, stem-succulent trees shed their leaves
earlier than co-occurring, nonsucculent trees, often before or
at the onset of the dry season, a pattern suggestive of a ‘drought-
avoidance’ strategy (Owen, 1974; Fenner, 1980; Holbrook,
1995; Borchert & Rivera, 2001; Lobo et al., 2003). Additionally,
during short periods of drought, stem succulents may close
their stomata to prevent transpiration and turgor loss (Nilsen
et al., 1990).
Baobab trees, like many stem succulents, flush new leaves
before the rainy season. (Daubenmire, 1972; Frankie et al.,
1974; Holbrook et al., 1995; Baum, 1996; Borchert &
Rivera, 2001; Rivera et al., 2002). The end of the dry season
is when soil water availability is lowest, and as baobab trees are
thought to be shallow-rooted (Fenner, 1980; Guy, 1971), any
growth during this time should be dependent on stored water.
Our major objective in this study was to learn to what extent
stored water is relied on during the dry season and for leaf
flushing. A second goal was to learn whether the use of stored
water allows baobab trees to extend their growing season,
either by initiating stomatal opening before the onset of the
rainy season, or simply by having the leaves present when the
rainy season begins.
Materials and Methods
Field site
This study was conducted at the Kirindy Forestry Reserve,
one of the largest remaining tracts of tropical dry forest in
western Madagascar (longitude 44°49′ E, latitude 20°27′ S)
(Rakotonirina, 1996). The majority of precipitation occurs
during the rainy season (December–March) and rainfall is
absent during the main part of the dry season (April–October).
Mean annual rainfall is 770 mm, but can be highly variable
and ranges from 300 to 1400 mm yr−1 (Ganzhorn et al., 1990).
Soils are sandy with low water retention, except near seasonal
watercourses where increased clay content confers a higher
water-retaining capacity (Sorg & Rohner, 1996).
Study species
The eight species in the genus Adansonia, commonly referred
to as baobabs, have a disjunct distribution across Africa,
New Phytologist (2006) 169: 549– 559 www.newphytologist.org © The Authors (2005). Journal compilation © New Phytologist (2005)
Madagascar and Australia (Armstrong, 1983). The six Malagasy
species, which probably form a monophyletic group (Baum,
1998), are restricted to the drier regions in the north, west and
south of the island, where they are found in deciduous forests,
spiny forests, scrublands and savannas. Baobabs range in
height from 5 to 30 m, and all have huge, swollen, cylindrical
or tapered trunks (Baum, 1995). Rooting depth is generally
shallow, and roots can access a large area around the tree, with
individual roots traveling along the surface for up to 70 m
(Guy, 1971; Fenner, 1980). Their low-density wood has large
vessels, few fibers, and abundant parenchyma throughout the
stem (Fisher, 1981; Newton, 1974).
The three species of baobab examined in this study were
Adansonia rubrostipa Jum. & H. Perrier; Adansonia za Baill.;
and Adansonia grandidieri Baill. (the latter was included only
for measures of water content and phenology). Adansonia
rubrostipa and A. za are found over a large range of habitat
types in western and southern Madagascar, and exhibit a wide
variety in overall morphology, while A. grandidieri has a
narrower range and a more consistent morphology (Baum,
1996). Study trees ranged from 1 to 3.5 m in diameter and
15–25 m in height. Additional trees smaller than 1 m in
diameter were included for phenological observations. Access
to the crown of the tree for installation of sap flow probes and
stomatal conductance measurements was through use of the
single-rope technique (Laman, 1995).
Phenology
Vegetative phenology was assessed during several seasons
between September 2001 and January 2003. Leaf bud break
and crown flush during the 2001 and 2003 leaf-flushing
periods, and extent of leaf drop during the 2002 leaf-shedding
period, were monitored on a weekly basis. Observations were
made on 45 trees of A. rubrostipa, 28 A. za and nine A.
grandidieri.
Soil water content
Soil cores were taken from a depth of 10–30 cm below the
leaf-litter layer at four representative sites in the Kirindy
Forest to quantify soil water content during the dry-to-rainy-
season transition. The soil cores were placed in sealed plastic
bags in the field, then brought to the laboratory, weighed,
dried at 75°C and reweighed. Water content was calculated as
a percentage of dry weight. Three cores were taken per site
every 7–14 d from September to December 2003.
Stem water content
Cores 12 mm in diameter and 50 cm in length were extracted
from the stem, between 0.5 and 1.5 m from the ground, using
an increment borer. Three sections 5 cm in length and centered
at 2.5, 15 and 30 cm beneath the bark were immediately
 Research 551

sealed into preweighed plastic bags. Fresh weight and volume Table 1 Dimensions of main stem and small branches at the location
were determined before drying the cores at 75°C to a constant of probe insertion for each tree on which sap flow measurements
weight (24 h). Water content was calculated as percentage of were made
fresh volume (Domec & Gartner, 2002). Cores were extracted
Parameter AR1 AR2 AR3 AZ1 AZ2 AZ3
from five trees each of A. rubrostipa, A. za and A. grandidieri
four times between October 2001 and February 2002, then Diameter at base (cm) 127 147 110 119 128 112
again in January, February and March 2003. Two more Diameter at top (cm) 80 92 116 82 94 87
measurements were made in September and December 2003. Diameter of branch 1 (cm) 10 11 8 11 10 11
Five additional trees of A. rubrostipa and A. za were included Diameter of branch 2 (cm) 13 11 8 9 11 13
Height to main branches (m) 13 12 11 13 13 17
to increase the sample size for these last two measurements.
Stem diameter at breast height was measured monthly with a Adansonia rubrostipa (AR); Adansonia za (AZ).
dbh tape to assess stem shrinkage and swelling in response to Sap flow measurements on AR1, AR2, AZ1 and AZ2 were made
uptake or use of stem water. Permanent guides were used to during the leaf-flushing period; measurements on AR3 and AZ3 during
ensure repeated measurements with the dbh tape were made the transitional period between leaf-flushing period and rainy season.
at the same height each time. Total water content in the
stems of several representative trees at different times of year is known. In the absence of accurate sapwood depth estima-
was estimated by taking the diameter of the tree at several tions, we focus on the relative timing and overall pattern of
points along the stem and the height at the point of branch sap flow at the different points in the tree, and we present our
divergence, and using the above stem water content values. data as sap flux densities.

Sap flow measurements Stomatal conductance

Rates of sap flow were measured using the thermal dissipation
technique (Granier, 1985), with probes similar in design to
those described by James et al. (2002). Unheated reference
probes revealed large underlying temperature gradients in the
stem and branches, which led to significant errors when the
traditional Granier calibration was applied to continuously
heated probes. The probes were therefore heated discontinuously,
following Do & Rocheteau (2002a, 2002b), and rates of sap
flow were calculated by taking the difference between heated
and unheated measurements and then applying their calibration
formulas. Probes were inserted to a depth of 1 cm beneath
the bark, and were placed at the base of the trunk (0.5–1.0 m
above ground) to measure sap flow from the roots to the stem;
at the top of the trunk (within 1 m of the point of main
branch divergence) to measure sap flow from the main stem
to the branches; and on distal branches (8–13 cm in diameter)
to measure sap flow from the branches to the leaves. Two
probes were installed at each location, and calculated sap flow
rates from probes at the same location were averaged.
Dimensions of the stem and branches at the locations of probe
insertion are presented in Table 1. Sap flow rates were measured
in two trees of each species during each measurement time
(leaf-flushing period and transitional period).
Our measurements indicated that the conductive sapwood
in some places was probably < 1 cm in depth, which we
estimated by placing probes at different depths to measure
presence or absence of sap flow at that point. As probe tips
were themselves 1 cm in length, it was not possible to resolve the
actual conductive sapwood depth. Clearwater et al. (1999)
discuss the implications of using sap flow probes longer or
shorter than conductive sapwood depth and the various
corrections that may be applied when the exact sapwood depth
Stomatal conductance of A. rubrostipa and A. za trees was
measured with a Li-Cor 1600 steady-state porometer.
Measurements were taken every 30–60 min from dawn to
dusk to quantify the diurnal pattern of stomatal conductance
during the 2002 rainy season ( January and February) in three
trees of A. rubrostipa and one A. za. During the 2003 dry
season, representative morning time-courses were made on
newly flushed leaves from one tree of each species. Ten to 15
leaves per tree were measured at each point in time, and the
average value calculated. Further measurements were taken in
two separate years (2001 and 2003) during the dry season to
compare levels of stomatal conductance before and after the
rainy season began. These measurements were made between
09:00 and 10:00 h on three trees of each species, using 12–15
leaves per tree. Measured leaves were distributed, to the extent
possible given the inherent difficulties with access, throughout
the outer crown of the tree. Air saturation deficit (ASD) at
each measurement time was calculated from the relative
humidity and temperature of the air as measured by the
porometer, according to the formula:
ASD = es − ea
where es is the saturated water vapor pressure and ea is the
ambient water vapor pressure of the air.
Results
Phenology
Leaf break was initiated in early October and leaf shedding in
early March. All trees had a full crown of mature leaves by

© The Authors (2005). Journal compilation © New Phytologist (2005) www.newphytologist.org New Phytologist (2006) 169: 549– 559
552 Research
mid-December, and were completely leafless by mid-April
(Fig. 1).
Soil water content
Soil water content varied between the different sites at the
Kirindy Forest, but remained constantly low within each site
from September to December, and did not begin to increase
until after significant rainfall had occurred. During the leaf-
flushing period, water content values ranged from 3.8 ± 0.1
to 8.8 ± 0.4 at the sites measured (percentage of dry weight,
mean ± SE). Note that the last measurements (28 December
2003) were made at the beginning of the rainy season and do
not reflect the maximal water content values that are expected
once the surface soil water has replenished fully.
Stem water content
In A. rubrostipa, stem volumetric water content did not change
significantly during the dry season ( March–September), but
decreases in stem water content of 10.4 and 12.2% were
measured during the leaf-flushing period (September–
December) in two separate years (Fig. 2). There was a slight
decrease in stem water content in A. za and A. grandidieri, but
the difference was not significant. All species experienced a
very small decrease in diameter during the dry season, but
A. za, and to a lesser extent A. grandidieri, shrank considerably
during the leaf-flushing period compared with the dry season
(Fig. 3). The total volume of water lost was calculated for
representative A. rubrostipa and A. za trees. Despite the fact
that the majority of water lost was through shrinkage for A. za
New Phytologist (2006) 169: 549– 559 www.newphytologist.org © The Authors (2005). Journal compilation © New Phytologist (2005)
Fig. 1 (a) Percentage of study trees of each
species with a full crown of leaves in two
separate seasons (Adansonia rubrostipa,
n = 45; Adansonia za, n = 28; Adansonia
grandidieri, n = 9). (b) Daily rainfall during the
same periods.
and through water extraction with little size change for A.
rubrostipa, the total volume of water for similarly sized trees of
each species was approximately the same (1000 l). Measurements
of total stem volume were not available for A. grandidieri,
which generally has a larger diameter, a greater height, and less
taper than A. rubrostipa. Stem shrinkage in A. grandidieri
would correspond on average to a four- to fivefold greater
volume of water loss for each meter of stem height than in
A. rubrostipa.
Sap flow rates
During the leaf-flushing period, sap flow rates at the bottom
of the stem were low, indicating little to no water entry from
the roots and soil into the stem during this time. The probes
at the top of the stem had the highest sap flow rates of the
three positions, and those in the small branches had moderate
rates. Positive sap flow rates through the upper stem and
branches while rates at the base of the tree were close to zero
confirmed that water reserves in the stem were used to support
leaf flushing (Fig. 4). Water-use rates in the branches and the
top of the stem increased as leaf flushing progressed, then
gradually decreased after the leaves were expanded. Occasional
rainfall events after the leaves had expanded caused sap flow
rates at all points in the tree (base, top and branches) to
increase, although sap flow rates decreased quickly as soil
water was depleted (Fig. 5). Sap flow rates were highest in the
morning, when ASD of the air was lowest, and declined to
lower values over the course of the day. After the onset of the
rainy season, coincident with greater soil water availability,
sap flow rates at all points in the tree reached higher values and
 Research 553

Fig. 2 Stem water content at three depths in

the tree before and after the dry season (left),
and before and after the leaf-flushing period
(right). (a) Adansonia rubrostipa (AR);
(b) Adansonia za (AZ); (c) Adansonia
grandidieri (AG). Values are mean of five trees
for all species during the dry season and AG
during the leaf-flushing period, and for 10
trees of AR and AZ during the leaf-flushing
period ± 1 SE. An asterisk between two data
points indicates a significant difference over
that time period (paired t-test, P < 0.05).
Values from before the dry season are
maximum values in stem water content
obtained at the end of the rainy season.

generally remained high throughout the day (0.5–1 l dm−2 h−1

at base; 2–3 l dm−2 h−1 at top; Chapotin, 2005).

Stomatal conductance
Stomatal conductance during the leaf-flushing period was close
to zero and remained low until after the rainy season began
(Fig. 6). Some trees exhibited a small amount of stomatal
opening shortly after sunrise, although the degree of stomatal
opening was lower than during the rainy season by over an
order of magnitude, but stomata closed as the morning progressed
and air-saturation deficit increased (Fig. 7). However, not all
trees exhibited early morning stomatal opening, and in some
cases values of stomatal conductance were zero even during
the morning after leaves were fully flushed (data not shown).
Fig. 3 Diameter decrease at breast height during the dry season and
leaf-flushing period for Adansonia rubrostipa (AR); Adansonia za (AZ);
Discussion Adansonia grandidieri (AG). Values are mean diameter decrease ± 1
SE. A paired t-test (P < 0.05) indicated significant differences in stem
Phenological patterns in African tropical dry forests vary diameter before and after each time period for AR and AZ during the
considerably from site to site. Leaf flushing in certain forests dry season, and for all species during the leaf-flushing period.

© The Authors (2005). Journal compilation © New Phytologist (2005) www.newphytologist.org New Phytologist (2006) 169: 549– 559
554 Research

Fig. 4 Sap flow rates in (a) two Adansonia rubrostipa (AR) trees; (b) two Adansonia za (AZ) trees during the leaf-flushing period. Leaves are
fully flushed by the times indicated by arrows. Each line is the average of values from two individual probes in similar locations (except one probe
only for the top probe on the first AZ tree; data are not available for any top probe on the second AZ tree).

New Phytologist (2006) 169: 549– 559 www.newphytologist.org © The Authors (2005). Journal compilation © New Phytologist (2005)
 Research 555

Fig. 5 Sap flow rates in (a) one Adansonia rubrostipa (AR) tree; (b) one Adansonia za (AZ) tree during the transitional period between dry and
rainy seasons. Leaves were fully flushed by the time measurements were initiated. Sap flow rates responded quickly to small rainfall events
(arrows) ranging from 1.5 to 8.5 mm total precipitation. Rainfall amount (mm) given next to each arrow. Each line is the average of values from
two individual probes in similar locations (except one probe only for top of AZ).

can be entirely opportunistic with respect to water availability
(Lieberman, 1982), while in others, such as the Kirindy Forest,
many trees flush new leaves before the rainy season (Sorg &
Rohner, 1996). The leaf-flushing period in baobab trees at
Kirindy occurs towards the end of the dry season, and a full
crown of mature leaves is often maintained for several weeks
before the onset of the rainy season. The end of the dry season
is generally when soil water content is at its lowest and
evaporative demand at its highest, and we therefore hypothesized
that the water required for early leaf flushing and transpiration
in baobab trees derives from stored reserves in the tree.
Furthermore, relying on stored water might allow baobab trees
to maintain stomatal opening even when soil water is unavailable,
thereby increasing the length of the growing season.
Stem water usage
Stem water content was high in all species (71–75% by
volume in the outer sapwood), and generally decreased with
increasing depth into the stem. When expressed as a percentage
of dry weight, water-content values in baobab trees (500–
800%) are much higher than published values for other
tree species (30–300%, Tsoumis, 1991) as well as other
stem-succulent trees (120–410%, Borchert, 1994a); although
it should be noted that these values are high also because
the solid fraction in baobab stems is quite low (Chapotin,
2005).
The measured sap flow patterns were consistent with
the hypothesis that the water necessary for leaf flushing was
derived almost entirely from the stem and branches, and not
from the soil. The presence of sap flow at the top of the stem
indicated substantial amounts of water moving from the stem
into the branches, and moderate flow rates in smaller branches
reflect water moving into the leaves to support leaf expansion
and cuticular water loss. Increasing water use during leaf
flushing and decreasing water use thereafter probably reflect
changing leaf area and cuticular properties of the leaves during
maturation (Fig. 4).

© The Authors (2005). Journal compilation © New Phytologist (2005) www.newphytologist.org New Phytologist (2006) 169: 549– 559
556 Research
Fig. 6 Leaf stomatal conductance in Adansonia rubrostipa and
Adansonia za during leaf flushing (November 2003); after leaves are
fully mature (December 2003); just after the rainy season has begun
(January 2004); and in the middle of the rainy season (January/
February 2002). All measurements were made between 09:00 and
10:00 h. Values are mean of three trees per species ± 1 SE (except one
tree only for A. za in the rainy season of 2002), with five to 15 leaves
measured per tree. Significant differences between months within a
species are indicated by different letters (t-test, P < 0.05). Difference
between the values marked by a′ and b′ had only a tendency towards
significance (P = 0.06).
Adansonia rubrostipa and A. za differed in the degree to
which stem diameter shrank with water withdrawal, although
both species lost a comparable volume of water during this
period. In A. rubrostipa the uniformly distributed fibers, vessels
and parenchyma cells appeared to maintain the volume of the
wood as air replaced withdrawn water. In A. za, however, the
grouping of the parenchyma cells into tangential bands,
alternating with bands comprised primarily of vessels and
fibers, allowed the stem to shrink when water was withdrawn
New Phytologist (2006) 169: 549– 559 www.newphytologist.org © The Authors (2005). Journal compilation © New Phytologist (2005)
without an increase in the air fraction. These results under-
score the importance of not relying solely on water content or
diameter changes when evaluating stem water loss, as even
closely related species can differ greatly in wood composition
and the extent of shrinkage with water loss.
Stem water does not support stomatal opening
There was no evidence that stem water reserves were being
used to support stomatal opening, as stomatal conductance
was consistently low until after the rainy season began and soil
water was replenished (Fig. 6). This was similar to the pattern
observed in another dry forest deciduous tree, Enterolobium
cyclocarpum, where electron-transfer rates in newly flushed
leaves were low until after the rainy season began (Brodribb
et al., 2002). Comparing the relationship between stomatal
opening and ASD during the dry season with that during the
rainy season demonstrates that greater stomatal closure is not
simply a factor of increased ASD during the dry season, and
thus stomata must also be independently regulated by another
factor (Fig. 8). Measured leaf-water potentials before the rainy
season reached a minimum of c. −1.0 MPa during the
afternoon, which was near the turgor-loss point for leaves
(unpublished data for A. rubrostipa). We propose that limited
water availability in the upper soil layers during this time
restricts stomatal opening and, as a consequence, photosynthetic
activity.
Whole-tree water use increased rapidly after isolated rain-
fall events (Fig. 5). This increase in water flow was noticeable
at all points measured in the tree, but particularly so in the
small branches as a result of increases in transpirational water
loss from the leaves, and at the base of the stem, indicating the
seasonal onset of water uptake by the roots. The rapidity with
which water uptake followed individual rainfall events is
significant in that it indicates that fine roots were present by
the time leaves were flushed or produced quickly upon soil
Fig. 7 Stomatal conductance during early
morning on newly flushed leaves for one
Adansonia rubrostipa and one Adansonia za
tree in December 2003. Values are mean of
10 –15 leaves per tree ± 1 SE. Air saturation
deficits (ASD) during each measurement period
are also shown.

Fig. 8 Stomatal conductance as a function of
air saturation deficit (ASD) in Adansonia
rubrostipa trees during the dry and rainy
seasons. All measurements were made
between 07:00 and 15:30 h. Each point is the
mean of five to 15 leaves from one tree ± 1 SE.
Three to five trees per curve.
hydration, as they were in another dry forest site (Kummerow
et al., 1990). These increased rates of water use were not sus-
tained, however, and rates of sap flow in the branches and
stem declined gradually over the several days following the
rainfall event, in response to declining soil water availability.
Sap flow rates then remained low until the next rainfall event.
It was not until after the onset of the rainy season, when rainfall
events were frequent enough to prevent soil drying, that sap
flow and stomatal conductance reached their maximal values
and remained consistently high (Chapotin et al., 2006).
Limited use of stem water despite early flushing
Early leaf flushing in baobab trees does not appear to allow
stomata to open before the rainy season when soil water
availability is still low. This raises two main questions: why
does leaf flushing in baobab trees precede the onset of the
rainy season; and why is stored water not drawn upon more
extensively to maximize carbon assimilation? The first question
is pertinent because the use of stored water for leaf flushing
appears to necessitate some investment in water-storage
tissue. Furthermore, daily temperatures reach their yearly
maximum during this time, and the inability to reduce leaf
temperatures through evaporative cooling exposes the leaves
to potentially damaging temperatures. Nontranspiring canopy
leaves frequently reached 40°C, a temperature that induces
heat-stress responses in many plants ( Jones, 1992).
It has been suggested that a physiological advantage of early
leaf flushing is that leaves are ready to begin photosynthesiz-
ing when the rainy season begins (Rivera et al., 2002). Having
the leaves present by the end of the dry season allowed baobab
trees to open their stomata as soon as soil water became avail-
able, and eliminated the time lag that would otherwise have
existed between the onset of the rainy season and the onset of
photosynthetic activity. Additionally, early leaf flushing allows
baobab trees to take advantage of scattered or early rainfall
events, although they may be of limited duration or intensity.
© The Authors (2005). Journal compilation © New Phytologist (2005) www.newphytologist.org
Research 557
Even very small rainfall events occurring before the full onset
of the rainy season were accompanied by immediate increases
in sap flow, not only at the base of the tree, indicating water
uptake from the soil, but also in the branches, indicating
transpirational water loss caused by stomatal opening. In an
environment where the rainy season is short and its length can
vary by several months, maximizing photosynthesis in this
way would be advantageous.
With regard to the second question, as to why more exten-
sive use of stored water does not occur, results from two
related studies indicate that stored water may be unavailable
because of tissue water relations, transport limitations and
biomechanical considerations. Baobab stem wood is highly
vulnerable to cavitation and has a turgor-loss point near actual
field water potentials (Chapotin et al., 2006). In addition, the
conductive portion of the sapwood is restricted to the 1–2 cm
just beneath the bark, so that radial transport of water in the
stem to the transpiration stream is through a relatively high
resistance pathway, and is likely to occur very slowly. This rate
of water movement may be sufficient for the needs of growing
tissues, but is unlikely to meet the demand of actively tran-
spiring leaves. Furthermore, the wood is soft and weak, and
comprised primarily of living tissues which may exhibit
decreased mechanical stability under excessive water withdrawal
(Niklas, 1988; Chapotin, 2005).
Water use in baobab trees before the onset of the rainy season
appears surprisingly conservative, given the vast amounts of
water in the trunk. However, the constraints described above,
and the highly unpredictable climate to which baobab trees
are subject, may require that use of stored water be limited in
this way. The timing of the rainy season can vary by several
months, and the total amount of precipitation in any given
rainy season is similarly variable (Fenner, 1980). By restricting
water use to leaf flushing rather than stomatal opening, the
baobab tree is able to maintain leaf turgor until the onset
of the rainy season, and to produce a new crop of leaves the
following year if current-year precipitation levels were not
New Phytologist (2006) 169: 549– 559
558 Research
sufficient to replenish the water content in the wood. The
results of this study appear to contradict general ideas about
water use in baobab trees by indicating a limited role for
stored water. To our knowledge there are no similar studies on
stem-succulent trees that focus specifically on water use dur-
ing the period between leaf flushing and the onset of the rainy
season. Stem-succulent trees occur in many distantly related
plant taxa, yet exhibit striking convergence in physiog-
nomy, ecology and habitat. Although this study focuses on
just one genus, the results suggest that water-use patterns in
stem-succulent trees may need further study.
Acknowledgements
This project was made possible through funding by the
Sinclair Kennedy Traveling Fellowship, the Arnold Arboretum
and the Department of Organismic and Evolutionary Biology
at Harvard University, the Garden Club of America, and
the National Science Foundation. We are thankful to Lalao
Andriamahefarivo and other Missouri Botanical Garden staff
in Antananarivo for acquiring research permits and providing
logistical support, and to the German Primate Center (DPZ)
for permission to work at the Kirindy Research Station.
Michael Burns provided invaluable technical expertise and
Randol Villalobos assisted with equipment assembly. Rodolphe
Rasoloarivony, Heather Patt and Mario Ramohavelo provided
field assistance. We also thank three reviewers for their helpful
comments and suggestions.
References
Armstrong P. 1983. The disjunct distribution of the genus Adansonia L.
National Geographic Journal of India 29: 142–163.
Baum DA. 1995. A systematic revision of Adansonia (Bombacaceae). Annals
of the Missouri Botanical Garden 82: 440 – 470.
Baum DA. 1996. The ecology and conservation of the baobabs of
Madagascar. Primate Report 44: 311–328.
Baum DA. 1998. Biogeography and floral evolution of baobabs (Adansonia,
Bombacaceae) as inferred from multiple data sets. Systematic Biology 47:
181–207.
Borchert R. 1994a. Soil and stem water storage determine phenology and
distribution of tropical dry forest trees. Ecology 75: 1437–1449.
Borchert R. 1994b. Water status and development of tropical trees during
seasonal drought. Trees 8: 115 –125.
Borchert R, Rivera G. 2001. Photoperiodic control of seasonal development
and dormancy in tropical stem-succulent trees. Tree Physiology 21:
213–221.
Brodribb TJ, Holbrook NM, Gutiérrez MV. 2002. Hydraulic and
photosynthetic co-ordination in seasonally dry tropical forest trees. Plant,
Cell & Environment 25: 1435 –1444.
Cedeño LF. 2001. Almacenamiento de recursos y relaciones hídricas durante
el desarrollo ontogenético de Cochlospermum vitifolium, una especie
pionera del Bosque Seco de Costa Rica. Licenciada Thesis, Universidad de
Costa Rica.
Chapotin S. 2005. Why are baobab trees so fat? Water relations and
biomechanics in the genus Adansonia L. PhD Thesis, Harvard University,
Harvard, MA, USA.
Chapotin S, Razanameharizaka JH, Holbrook NM. 2006. Water relations
of baobab trees (Adansonia L.) during the rainy season: does stem water
New Phytologist (2006) 169: 549– 559 www.newphytologist.org © The Authors (2005). Journal compilation © New Phytologist (2005)
buffer daily water deficits? Plant, Cell & Environment. (In press.)
doi:10.1111/j.1365-3040.2005.01456.x
Clearwater MJ, Meinzer FC, Andrade JL, Goldstein G, Holbrook NM.
1999. Potential errors in measurement of nonuniform sap flow using heat
dissipation probes. Tree Physiology 19: 681–687.
Daubenmire R. 1972. Phenology and other characteristics of tropical
semi-deciduous forest in north-western Costa Rica. Journal of Ecology 60:
147–170.
Do F, Rocheteau A. 2002a. Influence of natural temperature gradients
on measurements of xylem sap flow with thermal dissipation probes.
1. Field observations and possible remedies. Tree Physiology 22:
641–648.
Do F, Rocheteau A. 2002b. Influence of natural temperature gradients on
measurements of xylem sap flow with thermal dissipation probes. 2.
Advantages and calibration of a noncontinuous heating system.
Tree Physiology 22: 649–654.
Domec J-C, Gartner BL. 2002. How do water transport and water storage
differ in coniferous earlywood and latewood? Journal of Experimental
Botany 53: 2369–2379.
Fenner M. 1980. Some measurements on the water relations of baobab trees.
Biotropica 12: 205–209.
Fisher JB. 1981. Wound healing by exposed secondary xylem in Adansonia
(Bombacaceae). IAWA Bulletin 2: 193–199.
Frankie GW, Baker HG, Opler PA. 1974. Comparative phenological studies
of trees in tropical wet and dry forests in the lowlands of Costa Rica.
Journal of Ecology 62: 881–919.
Ganzhorn JU, Ganzhorn AW, Abraham J-P, Andriamanarivo L,
Ramananjatovo A. 1990. The impact of selective logging on forest
structure and tenrec populations in western Madagascar. Oecologia 84:
126–133.
Goldstein G, Meinzer FC, Monasterio M. 1984. The role of capacitance
in the water balance of Andean giant rosette species. Plant, Cell &
Environment 7: 179–186.
Goldstein G, Andrade JL, Meinzer FC, Holbrook NM, Cavelier J,
Jackson P, Celis A. 1998. Stem water storage and diurnal patterns of
water use in tropical forest canopy trees. Plant, Cell & Environment 21:
397–406.
Granier A. 1985. Une nouvelle méthode pour la mesure de flux de sève brute
dans le tronc des arbres. Annales des Sciences Forestières 42: 193–200.
Guy GL. 1971. The baobabs: Adansonia spp. (Bombacaceae). Journal of the
Botanical Society of South Africa 57: 30–37.
Holbrook NM. 1995. Stem water storage. In: Gartner BL, ed. Plant Stems:
Physiology and Functional Morphology. San Diego, CA, USA: Academic
Press, 151–174.
Holbrook NM, Sinclair TR. 1992. Water balance in the arborescent palm,
Sabal palmetto. II. Transpiration and stem water storage. Plant, Cell &
Environment 15: 401–409.
Holbrook NM, Whitbeck JL, Mooney HA. 1995. Drought responses of
neotropical dry forest trees. In: Bullock SH, Mooney HA, Medina E,
eds. Tropical Dry Forests. Cambridge, UK: Cambridge University Press,
243–276.
James SA, Clearwater MJ, Meinzer FC, Goldstein G. 2002. Heat
dissipation sensors of variable length for the measurement of sap flow in
trees with deep sapwood. Tree Physiology 22: 277–283.
Jones HG. 1992. Plants and Microclimate, 2nd edn. Cambridge, UK:
Cambridge University Press.
Kummerow J, Castillanos J, Maas M, Larigauderie A. 1990. Production of
fine roots and the seasonality of their growth in a Mexican deciduous dry
forest. Vegetatio 90: 73–80.
Laman TG. 1995. Safety recommendations for climbing rain forest trees
with ‘Single Rope Technique’. Biotropica 27: 406–409.
Lieberman D. 1982. Seasonality and phenology in a dry tropical forest in
Ghana. Journal of Ecology 70: 791–806.
Lobo JA, Quesada M, Stoner KE, Fuchs EJ, Herrerías-Diego Y, Rojas J,
Saborío G. 2003. Factors affecting phenological patterns of
 Research 559

Bombacaceous trees in seasonal forests in Costa Rica and Mexico.
American Journal of Botany 90: 1054 –1063.
Machado J-L, Tyree MT. 1994. Patterns of hydraulic architecture and water
relations of two tropical canopy trees with contrasting leaf phenologies:
Ochroma pyramidale and Pseudobombax septenatum. Tree Physiology 14:
219–240.
Medina E. 1995. Diversity of life forms of higher plants in neotropical dry
forests. In: Bullock SH, Mooney HA, Medina E, eds. Seasonally Dry
Tropical Forests. Cambridge, UK: Cambridge University Press, 221–242.
Newton L. 1974. Is the baobab tree succulent? Cactus and Succulents Journal
36: 57–58.
Niklas KJ. 1988. Dependency of the tensile modulus on transverse
dimensions, water potential, and cell number of pith parenchyma.
American Journal of Botany 75: 1286 –1292.
Nilsen ET, Sharifi MR, Rundel PW, Forseth IN, Ehleringer JR. 1990.
Water relations of stem succulent trees in north-central Baja California.
Oecologia 82: 299–303.
Owen J. 1974. A contribution to the ecology of the African baobab
(Adansonia digitata L.). Savanna 3: 1–12.
Rakotonirina. 1996. Composition and structure of a dry forest on sandy soils
near Morondava. Primate Report 46: 81–88.
Rivera G et al. 2002. Increasing day-length induces spring flushing of tropical
dry forest trees in the absence of rain. Trees 16: 445–456.
Sorg JP, Rohner U. 1996. Climate and tree phenology of the dry deciduous
forest of the Kirindy Forest. Primate Report 46: 57–80.
Stratton L, Goldstein G, Meinzer FC. 2000. Stem water storage capacity and
efficiency of water transport: their functional significance in a Hawaiian
dry forest. Plant, Cell & Environment 23: 99–106.
Tsoumis G. 1991. Science and Technology of Wood. New York, USA:
Chapman & Hall.
Waring RH, Running SW. 1978. Sapwood water storage: its contribution
to transpiration and effect upon water conductance through the stems of
old-growth Douglas-fir. Plant, Cell & Environment 1: 131–140.
Waring RH, Whitehead D, Jarvis PG. 1979. The contribution of
stored water to transpiration in Scots pine. Plant, Cell & Environment
2: 309–317.
Wickens GE. 1983. The baobab – Africa’s upside-down tree. Kew Bulletin
37: 173–209.

About New Phytologist

• New Phytologist is owned by a non-profit-making charitable trust dedicated to the promotion of plant science, facilitating projects
from symposia to open access for our Tansley reviews. Complete information is available at www.newphytologist.org.

• Regular papers, Letters, Research reviews, Rapid reports and both Modelling/Theory and Methods papers are encouraged.
We are committed to rapid processing, from online submission through to publication ‘as-ready’ via OnlineEarly – the 2004 average
submission to decision time was just 30 days. Online-only colour is free, and we provide 25 offprints as well as a PDF for each
article.

• For online summaries and ToC alerts, go to the website and click on ‘Journal online’. You can take out a personal subscription to
the journal for a fraction of the institutional price. Rates start at £125 in Europe/$232 in the USA & Canada for the online edition
(click on ‘Subscribe’ at the website).

• If you have any questions, do get in touch with Central Office (newphytol@lancaster.ac.uk; tel +44 1524 594691) or, for a local
contact in North America, the US Office (newphytol@ornl.gov; tel +1 865 576 5261).

© The Authors (2005). Journal compilation © New Phytologist (2005) www.newphytologist.org New Phytologist (2006) 169: 549– 559