American Journal of Botany 97(12): 19261936. 2010.

INTRINSIC AND EXTRINSIC HYDRAULIC FACTORS IN VARYING

SIZES OF TWO AMAZONIAN PALM SPECIES (IRIARTEA DELTOIDEA
AND MAURITIA FLEXUOSA) DIFFERING IN DEVELOPMENT
AND GROWING ENVIRONMENT1

Heidi J. Renninger2 and Nathan Phillips

Department of Geography and Environment, Boston University, 675 Commonwealth Ave., Boston, Massachusetts 02215 USA
Premise of the study: This study seeks to determine how hydraulic factors vary with ontogeny and whether they begin to limit
further height growth in palms. Palms are an attractive group for physiological research because their columnar trunks and
simple leaf habit allow key intrinsic and extrinsic hydraulic variables to be estimated more easily than in complex arborescent
dicotyledons.
Methods: We measured various biometric and physiological factors including sap flux, leaf areas, turnover rates, and internode
lengths in two Amazonian rainforest species: terra firme Iriartea deltoidea and swamp-adapted Mauritia flexuosa. These two
palm species differ markedly in edaphic conditions, leaf type (pinnately compound vs. palmate), and bole development, making physiological comparisons between them important as well.
Key results: The species exhibited differing patterns in height growth rate along boles, which appear to relate to their differences
in bole development. Growth rates ultimately slowed at the tops of tall palms in both species. We also found a high degree of
convergence in total leaf area with height in both species even though they exhibited contrasting patterns in both live frond
number and leaf area per frond with height. Sap flux density from leaves was constant with height but four times greater in M.
flexuosa than in I. deltoidea.
Conclusions: Although height growth rates slow considerably in tall palms, neither species shows evidence that hydraulic factors
become limiting because they are able to support much greater leaf areas with similar sap flux densities as shorter palms.
Key words: Arecaceae; bole water storage; height growth rate; hydraulic limitation; leaf turnover; sap flux; vascular
anatomy

Hydraulic factors of trees differ considerably across ontogeny, and extrinsic factors such as tree size can begin to limit
intrinsic water transport and photosynthetic properties. This is
the basis behind the hydraulic limitation hypothesis, which
states that the hydraulic supply system of trees becomes challenged as they grow taller, and this increased hydraulic resistance ultimately limits maximum heights in trees (Ryan and
Yoder, 1997). Hydraulic limitation to maximum height in trees
has been studied in many complex gymnosperm (Hubbard et al.,
1999; Ryan et al., 2000; McDowell et al., 2002; Koch et al.,
2004) and angiosperm species (Schfer et al., 2000; Barnard
and Ryan, 2003; Phillips et al., 2003a) but has yet to be studied
in relatively simple tree systems like palms. The intrinsic properties of palms such as lack of branching and simple crown
form allow for extrinsic properties such as growth rates, sap
fluxes, and leaf areas to be estimated much more easily than in
dicotyledons (Phillips et al., 2008, 2009). Even though the intrin1

Manuscript received 11 January 2010; revision accepted 1 October 2010.

The authors thank the staff of the Universidad San Francisco de Quito
and Tiputini Biodiversity Station (TBS), especially D. Romo, C. de Romo,
J. Guerra, D. Mosquera, and R. Torres, for logistical support and field site
access. They also thank the labor staff at TBS, especially A. Papa, S.
Shiguango, J. C. Rodriguez, and R. Papa, for climbing the Mauritia
flexuosa palms, and they thank S. Shiguango for his assistance in installing
the solar panels on the canopy tower. Funding was provided by a grant
from the National Science Foundation (IOB #0517521).
2 Author for correspondence (e-mail: heidiren@bu.edu)
doi:10.3732/ajb.1000015

sic vasculature of palms differs considerably from that of dicotyledonous trees, because the same physical laws that govern
the movement of water through all plants apply to palms, this
research can shed light on the features and trade-offs that may
be universal and those that may differ between arborescent
monocots and dicots. Therefore, one of the aims of this study
deals with comparing palms of varying height to determine how
hydraulic factors differ or remain constant.
In addition to comparing palms across ontogeny, we are also
interested in comparing intrinsic and extrinsic hydraulic properties between Iriartea deltoidea (Ruiz & Pav.) and Mauritia
flexuosa (L.), two prominent palm species that, while growing
fairly closely to one another, differ in several ways. Whereas
both species are found primarily in the lowland rainforests of
western Amazonia, M. flexuosa grows mainly in permanently
flooded soils, where it forms nearly monospecific stands (Rull,
1998). Iriartea deltoidea, on the other hand, grows on terra
firme and varsea soils, where it has been found to be the most
common tree species in several locations in western Amazonia
(Pitman et al., 2001; Montufar and Pintaud, 2006). Besides differing in edaphic growing environment, I. deltoidea possesses
pinnately compound leaves, whereas M. flexuosa has costapalmate leaves (Uhl and Dransfield, 1987). Finally, I. deltoidea
and M. flexuosa differ greatly in their development. The trunk
of M. flexuosa increases in diameter underground until it
reaches its final diameter, at which point vertical growth occurs. Waterhouse and Quinn (1978) refer to this as type A development. Type B development, on the other hand, is exhibited
by I. deltoidea, in that all palms past the seedling stage exhibit
aboveground trunks that increase in diameter substantially over

American Journal of Botany 97(12): 19261936, 2010; http://www.amjbot.org/ 2010 Botanical Society of America

1926

December 2010]

Renninger and PhillipsHydraulic factors in two Amazonian palm species

an extended period of time (Waterhouse and Quinn, 1978). Ontogenetic differences with regard to trunk diameter formation
could lead to significant differences in vertical height growth
rates and intrinsic bole water storage capacity (Holbrook and
Sinclair, 1992a; Fisher et al., 1996; Meinzer et al., 2004).
The objective of this study was to compare intrinsic (vascular
and stomatal anatomy, bole water storage, etc.) and extrinsic
(height growth rates, sap flux, leaf turnover, and leaf areas)
hydraulic properties across ontogeny in I. deltoidea and
M. flexuosa to determine whether height growth rates become
hydraulically limited in the tallest individuals. Additionally, because these two palm species differ in growing environment,
intrinsic leaf structure, and aboveground growth development,
comparison of hydraulic properties between them was also of
interest. Not only will the information gained from this study
help to better understand these two palm species across ontogeny, but these data can be used to better understand palms as a
plant group, one that has been shown to be both economically
and environmentally important (Plotkin and Balick, 1984;
Kahn, 1988; Kahn, 1991; Salm et al., 2005; Tomlinson, 2006;
Walther et al., 2007).
MATERIALS AND METHODS
Site descriptionThis research was performed at Tiputini Biodiversity
Station (036S, 7627W), a 650-ha research facility located within Yasun
National Park in eastern Ecuador. The site receives approximately 2860 mm of
rainfall annually and has an average temperature of 25.5C (Maca, 2004). Research was conducted in May and June of 2006 (the wet season) and January,
February, and March from 2007 to 2009, which is the driest part of the year in
this region of the Amazon rainforest, receiving about 107 mm of rain per month
(data courtesy of Dr. J. Guerra, Tiputini Biodiversity Station). The I. deltoidea
palms chosen for the study grew in the terra firme forest where the canopy
reaches about 30 m in height, with a canopy cover of about 80% (estimated
from overhead photographs). Tall palms were dominant, unshaded trees. Medium-sized trees were codominant and shaded on their sides. Short palms were
understory, shaded trees. Mauritia flexuosa palms were found in a nearby, permanently inundated swamp with about 65% canopy cover; the area therefore
was more open than the rest of the terra firme forest. Mauritia flexuosa palms
were the dominant tree species, with Astrocaryum sp. palms also present at all
stages of development.
Growth rates and leaf turnover ratesBecause both I. deltoidea and
M. flexuosa have distinct leaf scar nodes, height growth rates could be determined by measuring the distance between these nodes and calculating leaf turnover rates across palms of varying heights. All measurements were performed
in February and March 2008. In small palms (1 m tall), internode distances
were determined with a measuring tape. In all other palms, internode distances
were measured with the use of digital photographs of the palm boles. Successive pictures of an individual bole were taken, then stitched together with
image software (ArcSoft PhotoStudio 5.5, Fremont, California, USA), and the
internode distances were measured with image-analysis software (Image J,
Scion Image, Frederick, Maryland, USA). The height of the palm to the base of
the live crown as well as the horizontal distance between the photographer and
the palm were measured with a TruPulse 200 hypsometer (Laser Technology,
Centennial, Colorado, USA). On a nearby canopy tower, flagging tape was tied
every 0.5 m from the base to 30 m. Standing the same distance from the tower
as from the palms, digital photographs were taken of this tower scale, and
these were used to correct for the angle at which the pictures were taken as well
as to scale the measurements in meters. Internode distances also were measured
opportunistically on a 20-m-tall I. deltoidea treefall, and these measurements
fell within the standard error bars of the distances calculated from the bole
photographs.
Because internode distances represent the height growth between the
production of successive leaves, leaf production rates are important for the
interpretation of internode distances as height growth rates. Therefore, leaf
turnover rates were determined for various-sized I. deltoidea and M. flexuosa
palms, with the assumption that turnover rates and production rates are approxi-

1927

mately equal. For I. deltoidea, the method for determining leaf turnover rates
was similar to that of Tomlinson (1963). In February and March 2008, four
small palms (about 1 m tall), two medium-sized palms (12 and 14 m tall), and
one tall palm (24 m tall) were marked by hammering an aluminum tag on the
most recently produced leaf scar node. Medium palms were accessed by using
a palm-climbing apparatus (www.nif.org.in/bd/node/125), and the tall palm
was accessed by using ropes. This tall I. deltoidea was located next to a canopy
tower built around a large Ceiba pentandra ((L.) Gaertn.). A rope was secured
to this C. pentandra, passed through the crown of the I. deltoidea palm, and
climbed with the use of ascenders. For two additional medium-sized palms
(6 and 11 m) and three additional tall palms (24 to 26 m tall), digital pictures of
the crowns were taken from a canopy tower, and the location of the photographer
was tagged. In February 2009, digital pictures were taken from the same location and were compared with those of the previous year to determine the number of leaves lost and newly made. Additionally, the leaf node scars above the
tagged scar were counted on the aluminum-tagged palms. Because the leaves of
M. flexuosa palms do not drop once they desiccate and this species is not found
growing near the canopy towers, another methodology was required. Therefore,
in February and March 2008, two juvenile palms that lacked an aboveground
trunk (0 m tall), two medium-sized palms (6 and 7 m tall), and two tall palms
(18 and 24 m tall) were marked for leaf turnover with flagging tape tied around
all green leaves. The following February (2009), these palms were revisited,
and the number of flagged and dead leaves, flagged and living leaves, and unflagged and living leaves were counted.
Sap flux and Mauritia flexuosa bole water storage estimationSap
flux was measured in the boles of three small I. deltoidea palms (about 1 m
tall), four medium I. deltoidea palms (between 6 and 14 m tall), and three tall
I. deltoidea palms (between 24 and 26 m tall) during May and June 2006
(wet season) and January and February 2007 (dry season) by using 1-cm-long
Granier-style heat dissipation sensors (Granier, 1987). One-cm-long sensors
were used because of the difficulty of installing sensors into I. deltoidea boles,
and sap fluxes were scaled up to the entire cross-sectional area of the bole (see
next paragraph). For M. flexuosa, sap flux was measured in the boles and/or
petioles of two juvenile palms (no aboveground trunk), two medium palms (6
to 7 m tall), and two large palms (18 and 22 m tall) during January and February
2007, February and March 2008, and January and February 2009 with the use
of 2-cm-long Granier sensors that were also scaled up to the entire cross-sectional bole area. To install sensors in the petioles of medium and tall M. flexuosa, palms were climbed by passing a rope through the crown of the intended
palm and securing it at both ends. First, heavy-gauge nylon wire was shot
through the crown with use of a bow and arrow, then heavier climbing rope was
passed through the crown. The rope then was climbed by using single rope ascenders. The Granier sensor methodology (Granier, 1987) involves measuring
the amount of heat dissipated by water flow around a heated sensor relative to a
reference sensor, both of which are radially inserted into the palm. When water
flows past the heated sensor, it dissipates some of the heat produced; the amount
of dissipation is related to the rate of sap flux. After sensor installation, aluminum insulating wrap was stapled around sensors to shield them from temperature variation caused by sun flecks. The heat dissipation sensors were attached
to a Campbell CR-10 data logger and AM-32 multiplexer (Campbell Scientific, Logan, Utah, USA), which collected data every 30 min for I. deltoidea and
every 2 min for M. flexuosa. The program, Baseliner version 2.4.2 (C-H20
Ecology Group, Duke University, Durham, North Carolina, USA), was used to
convert the millivolt signal from the data logger into sap fluxes (g m2 sec1)
by using the Granier equation (Granier, 1987). We performed a calibration of
the Granier-style sensors in the bole of a small I. deltoidea palm and found
the Granier equation was within the 95% confidence limits of our calibration
(Renninger et al., 2010). Solar panels located on the top of a canopy tower
provided power for the data logger and sensor heating in the I. deltoidea site.
For the M. flexuosa site, a solar panel located in the understory provided some
of the power for the system; however, batteries also needed to be shuttled back
and forth to the field camp generator to be recharged and to supply the site with
power.
To calculate daily sap flux values for I. deltoidea, we used vascular bundle
densities in the central and peripheral bole region reported in Rich (1987) and
assumed that metaxylem vessel diameters were more or less constant across the
bole radius. According to Rich (1987), vascular bundle densities in the central
core are approximately half that of the outer periphery. We used our own observations of fallen I. deltoidea palms to estimate the proportion of the bole area
considered central, which yielded the following equation: y = 0.8013e0.036x,
where x = palm height, and y = proportion of diameter at breast height (dbh)
occupied by the central region. These data were then used to scale our sap flux

1928

[Vol. 97

American Journal of Botany

values measured in the bole periphery to the entire cross-sectional area of the
I. deltoidea palms. For M. flexuosa, we did not have access to vascular bundle
density data; therefore, we assumed that flux was more or less constant throughout the cross section. This assumption is supported by Roupsard et al. (2006),
who found a constant pattern of sap flux throughout the stem of coconut (Cocos
nucifera L.) palms up to a 12-cm radius; Sellami and Sifaoui (2003), who found
that sap flow at 3-cm depth and 6-cm depth did not differ significantly in date
palms (Phoenix dactylifera L.); and Renninger et al. (2009), who found that sap
flux at 2 cm and 4 cm did not differ significantly in Mexican fan palms (Washingtonia robusta H.Wendl.). For M. flexuosa, we multiplied individual petiole
sap flux by the number of fronds to get an estimate of whole palm sap flux. Sap
fluxes scaled up from individual bole sensors matched well with sap fluxes
scaled up from individual fronds (regressions of sap flux vs. palm height had
neither slopes nor y-intercepts that differed significantly), and this lends confidence to the assumptions that were made in scaling total sap flux from the
single sensor values.
For M. flexuosa, because concurrent measurements of sap flux were taken in
the base of the bole as well as in the petiole, daily usage of stored bole water
could be examined. In palms that are more reliant on stored water in the stem,
sap flux in the lower bole tends to lag behind sap flux in the petiole (i.e., initiation of morning sap flux, peak midday sap flux, and nighttime decline will occur
later). Cross-correlation analysis of the sap flux time series data were performed
in the 6-, 7-, 18-, and 22-m-tall palms. The time lag corresponding to the maximum degree of correlation between the petiole sap flux signal and the bole sap
flux signal, therefore, represented the approximate amount of stored bole water
used daily by the palm. Comparison of diurnal sap flux signals from boles and
petioles also allowed for estimates of daily stored water use according to procedures and assumptions described in Phillips et al. (2003b).
Leaf anatomical propertiesLeaflets were collected from I. deltoidea and
M. flexuosa palms both opportunistically and by climbing the trees. For I. deltoidea, leaflets were collected from medium-sized palms (6 to 14 m tall) by
climbing the bole with the use of a palm-climbing apparatus (www.nif.org.in/
bd/node/125) to the base of the live crown and by using a pole saw to cut leaflets down. Large I. deltoidea palms could not be accessed this way. Therefore,
leaflets were collected opportunistically when they, or the entire frond, had
fallen from a tall I. deltoidea. For M. flexuosa, leaflets were collected from
palms when the petiole sap flux sensors were installed. For other palms, dead
leaflet (no longer green) material was gathered by either climbing the boles to
the dead fronds or by collecting from fronds that had recently fallen to the
ground. These fronds retained all their microscopic anatomical features even
though they were no longer green. Once leaflets were obtained, they were returned to the laboratory, where they were hand sectioned with a razor blade,
taking thin sections of epidermal tissue from the abaxial side of the leaf. Mauritia flexuosa posed some difficultly, as stomata were concentrated on the main
parallel ribs. Therefore, hand sectioning was concentrated in these regions. Sections then were stained with a solution of 1% toluidine blue O and mounted on
slides with Permount (Fisher Scientific, Pittsburgh, Pennsylvania, USA).
The slides were viewed at 200 magnification with a compound light
microscope (Leica CME, Bannockburn, Illinois, USA), and photographs were
taken with a digital camera (Olympus SP-550 UZ, Center Valley, Pennsylvania, USA). These photographs were imported into image-analysis software
(Image J, Scion Image) for measurement. Leaf epidermal cell areas were measured by tracing around the perimeter of approximately 300 to 500 cells per
palm distributed across 20 to 30 photographs. Stomatal densities were determined by counting the number of stomata within a field of view and calculating
the area of that view field. Approximately 20 to 30 different view fields were
used per palm. Guard cell lengths were calculated by measuring the distance
between the two points where the guard cells met. Approximately 50 and 150
stomata were measured per palm to calculate average guard cell length. Total
stomatal pore area index (SPI) was then calculated as (stomatal density guard
cell length2) (Sack et al., 2003).
Iriartea deltoidea bole and stilt root anatomical propertiesBole material
was collected from two I. deltoidea palms (14 and 20 m tall) that had been
pushed over in a storm. Material was collected from midheight and from the
base of the live crown. The bole sections were split into four quadrants, and
samples were taken from both the inner (center) and outer (peripheral) bole region of each of the four quadrants. Additionally, five small I. deltoidea palms
(1 to 5 m tall) were harvested and split in half, and inner and outer bole material
from each half was collected as well as were samples from the stilt roots. Stilt
roots also were collected from five medium-sized I. deltoidea palms (10 to 15 m
tall) and from five large I. deltoidea palms (20 to 25 m tall). Samples were

hand sectioned with a razorblade on the transverse (cross-sectional) plane.

Sections were stained with a solution of 1% toluidine blue O and mounted
on slides with Permount. The slides were viewed at 40 magnification with a
compound light microscope (Leica CME, Bannockburn, Illinois, USA), and
photographs were taken with a digital camera (Olympus SP-550 UZ). These
photographs were imported into image-analysis software for measurement
of vessel diameters and vascular bundle densities. Approximately 50 to 100
metaxylem vessels were measured for each palm height/location category. Vascular bundle densities were determined by counting the number of bundles
within a field of view and calculating the area of that view field. Approximately 10 to 20 different fields of view were used per palm. These same fields
of view were then used to calculate Hagen-Poiseuille specific conductivities
(kHP) with the following equation:

k HP

r4
8
As

where r is the radius of metaxylem vessels, is the viscosity of water, and As

is the cross-sectional area of the field of view, with the summation over all
metaxylem vessels in the field of view (Zimmermann, 1983).
To quantify the capacity for water storage in both the inner and outer regions
of the bole in I. deltoidea, proportions of the cross sections that were occupied
by parenchyma tissue were calculated. Parenchyma tissues in boles of Sabal
palmetto have been found to have very high specific capacitance (0.34 to 1.16
MPa1 in Holbrook and Sinclair, 1992b). Therefore, estimation of parenchyma
proportions should allow for comparison of capacities for bole water storage in
I. deltoidea palms of differing heights. For this calculation, only parenchyma
cells that were outside the vascular bundles were included. Approximately 20
to 30 fields of view were used for each height/location category. To begin,
vascular bundles and any area not to be included in the field of view were blackened out with image-analysis software (Image J, Scion Image). Then, we used
the thresholding technique to select and measure the area of parenchyma. This
was compared with the total area being evaluated to determine the percent area
occupied by parenchyma capable of storing water within the bole.
Statistical analysesWe calculated r2 and P values for all regressions with
SigmaPlot 2000 version 6.1 (SPSS Chicago, Illinois, USA).

RESULTS
Internode lengths and leaf turnover ratesIn all I. deltoidea
height categories, from the base of the bole to the top, internode
lengths initially increased, reached a maximum that was sustained to differing degrees, then decreased toward smaller and
smaller values (Fig. 1A). For I. deltoidea, leaf turnover rates
decreased as palms got taller (r2 = 0.57, P = 0.033) (Fig. 2A).
For M. flexuosa, small and medium-sized palms exhibited constant internode lengths along their boles after an initial increase
in internode length (Fig. 1B). In large M. flexuosa, from the
base of the bole to the top, internode lengths initially increased,
remained constant over an extended length of the bole, then
decreased sharply to very low levels (Fig. 1B). However, unlike
I. deltoidea, in M. flexuosa, leaf turnover rates increased nonlinearly (fitted with a power function) with palm height (r2 = 0.997,
P = 0.0002) (Fig. 2A).
Leaf propertiesIriartea deltoidea and M. flexuosa exhibited
opposing but complimentary relations between the number of
live fronds per palm and individual frond leaf area with height
(Fig. 2B, C). In I. deltoidea, individual frond leaf areas
increased linearly with height (r2 = 0.97, P < 0.0001) (Fig. 2C),
whereas the number of live fronds per palm increased nonlinearly (fitted with a power function), with smaller rates of increase in taller palms (r2 = 0.76, P = 0.0005) (Fig. 2B). However,

December 2010]

Renninger and PhillipsHydraulic factors in two Amazonian palm species

1929

Fig. 1. Internode lengths (m) along the boles of (A) Iriartea deltoidea
and (B) Mauritia flexuosa. Points with standard error bars represent the
mean and variance of four or five individuals.

in M. flexuosa, the individual frond leaf area increased nonlinearly (fitted with a power function), with smaller rates of increase
in taller palms (r2 = 0.80, P < 0.0001) (Fig. 2C), whereas the
number of live fronds per palm increased linearly with height
(r2 = 0.83, P < 0.0001) (Fig. 2B). Putting these two variables
together, both I. deltoidea and M. flexuosa showed a similar
linear increase in total leaf area with height (r2 = 0.91, P <
0.0001) (Fig. 2D), with neither the slopes nor y-intercepts of the
individual species linear regressions differing significantly at
= 0.05.
The leaves of I. deltoidea palms had significantly larger
epidermal cells than leaves of M. flexuosa (Fig. 3A). Leaf
epidermal cell sizes decreased with height in both species, with
I. deltoidea exhibiting a nonlinear, quadratic relation (r2 = 0.74,
P = 0.018) and M. flexuosa exhibiting a linear one (r2 = 0.93,
P = 0.0021) (Fig. 3A). Stomatal densities and guard cell lengths
showed opposing, complimentary relations with height in
I. deltoidea and M. flexuosa (Fig. 3B, C). In I. deltoidea, stomatal
densities increased nonlinearly (fitted with a power function)
with height (r2 = 0.85, P = 0.0005) (Fig. 3B), whereas guard
cell lengths decreased nonlinearly and exponentially with height
(r2 = 0.9, P = 0.0012) (Fig. 3C). On the other hand, in M. flexuosa, stomatal densities decreased nonlinearly and exponentially
with height (r2 = 0.68, P = 0.0063) (Fig. 3B), whereas guard
cell lengths increased nonlinearly (fitted with a quadratic function) with height (r2 = 0.94, P = 0.0038) (Fig. 3C). Stomatal

Fig. 2. Relation between palm height (m) and (A) number of fronds
lost per year for Iriartea deltoidea (y = 0.062x + 2.56) and Mauritia flexuosa (y = 1.34x0.44), (B) number of live fronds in I. deltoidea (y = 3.83x0.17)
and M. flexuosa (y = 0.32x + 3.62), (C) individual frond leaf area (m2) in
I. deltoidea (y = 0.40x + 0.64) and M. flexuosa (y = 3.25x0.19), and (D) total
palm leaf area (m2) in I. deltoidea and M. flexuosa, where a single regression (y = 2.53x + 5.95) fits the data for both species.

pore area index was constant for leaves from I. deltoidea palms
of different heights (Fig. 3, inset). However for M. flexuosa, SPI
decreased nonlinearly and exponentially in leaves from taller
palms (r2 = 0.90, P = 0.001) (Fig. 3, inset).
Sap flux For both I. deltoidea and M. flexuosa, no relation
existed between bole cross-sectional area and sap flux per unit
area of outer bole material (kg m2 d1; data not shown).
However, when sap flux data were scaled up to the tree level
(kg d1), both I. deltoidea and M. flexuosa showed greater sap
flux rates in taller palms than in shorter ones (Fig. 4A, B). As
well, for I. deltoidea, sap flux measured in the boles of palms
showed the same relation with height during the wet season
(MayJune) and the following dry season (JanuaryFebruary),
as the slopes and y-intercepts were not significantly different at
= 0.05, and a single line was fitted to the data (r2 = 0.79, P <
0.0001) (Fig. 4A). In M. flexuosa, sap flux measured in the base
of the bole showed the same relation with palm height as sap

1930

American Journal of Botany

Fig. 3. The relation between palm height (m) and (A) leaf epidermal
cell sizes (m2) in Iriartea deltoidea (y = 77 + 21x 1.3x2) and Mauritia
flexuosa (y = 7.2x + 350), (B) stomatal density (mm2) in I. deltoidea (y =
52x0.47) and M. flexuosa (y = 37e0.076x), and (C) guard cell length (m) in
I. deltoidea (y = 29.4e0.028x) and M. flexuosa (y = 14.7 + 0.001x + 0.007x2).
Guard cell lengths and stomatal densities were used to calculate stomatal
pore area indices, which are plotted vs. palm height (m) (inset) for I. deltoidea and M. flexuosa (y = 0.081e0.066x).

flux measured in a petiole and scaled up to the whole palm

level, as the slopes and y-intercepts did not significantly differ
at = 0.05, and a single line was fitted to the data (r2 = 0.80,
P < 0.0001) (Fig. 4B). To determine whether sap flux density
from the leaves also differed in palms of differing heights, daily
sap flux (kg d1) was divided by the total leaf area of the palm
being measured. Because total leaf area also increased with
palm height (Fig. 2D), no relation existed between daily sap
flux density from the leaves and palm height for either I. deltoidea (P = 0.1983) or M. flexuosa (P = 0.1583) (Fig. 4A, B,
inset graphs). Sap flux density from leaves was approximately
four times higher in M. flexuosa palms than in I. deltoidea palms.
Bole water storage Cross-correlation analysis between
petiole and lower bole sap flux in M. flexuosa revealed differences in bole water storage in palms of different heights. For a
6- and 7-m-tall M. flexuosa palm, an average lag of 14.4 min
(SE = 3.1) corresponded to the highest correlation between petiole and bole sap flux over a 10-d period. This value represents
about 11.8% (SE = 1.0%) of the total daily water use for these
palms. For an 18-m-tall M. flexuosa palm, an average lag of
20.5 min (SE = 5.3) corresponded to the highest correlation

[Vol. 97

Fig. 4. The relation between palm height (m) and (A) Iriartea deltoidea sap flux (kg d1) measured in the boles during the wet season and
the subsequent dry season. Sap flux increased linearly in taller palms (y =
0.27x + 0.41); however, no relation exists when sap flux is presented on a
per leaf area basis (inset). (B) Mauritia flexuosa sap flux (kg d1) measured in the boles and in the petioles, where sap flux was scaled up by
multiplying by the number of fronds. Sap flux increased linearly in taller
palms (y = 3.95x + 1.82); however, no relation exists when sap flux is presented on a per leaf area basis (inset). Error bars represent the standard
error in sap flux measured over consecutive days.

over a 9-d period, which represents about 12.2% (SE = 1.6%)

of the total daily water use for this palm. For a 22.5-m-tall
M. flexuosa palm, an average lag of 48.6 min (SE = 8.3) corresponded to the highest degree of correlation over an 8-d period.
This represents about 19.4% of the total daily water use for this
palm. Figure 5 shows the petiolebole sap flux correlation for a
representative day (Julian day 59), with no lag applied to the
data (left panels) and a lag in petiole sap flux corresponding to
the maximum r2 value (right panels) for a 6-, 18-, and 22.5-mtall palm. The lags corresponding to a maximum degree of correlation are higher than the mean lags for each palm, but they
show the same trend of increasing lags and, therefore, increasing reliance on stored bole water in taller palms.
For I. deltoidea, cross-correlation analysis between petioles
and boles could not be performed because the petioles of I. deltoidea palms could be accessed only in a relatively small proportion of individuals. Therefore, the percent area occupied by
parenchyma cells in the inner and outer bole was quantified to
determine possible reliance on stored water. Only parenchyma
cells external to the vascular bundles were quantified, since this

December 2010]

Renninger and PhillipsHydraulic factors in two Amazonian palm species

1931

Fig. 5. Bole water storage estimation for Mauritia flexuosa, in which left panels represent no lag between bole and petiole sap flux (g m2 s1), and
right panels incorporate a lag that maximizes the r2 of the regression. Top panels, (A) and (B), represent data from a 6-m-tall palm; middle panels, (C) and
(D), represent data from an 18-m-tall palm; and bottom panels, (E) and (F), represent data from a 22.5-m-tall palm. All data were collected on the same
day: 28 February 2008.

is a likely location for bole water storage. Parenchyma in the

inner bole region occupied 48% (SE = 2.6%) (Fig. 6) of the area
with large lacunae present (Fig. 6, inset). In the outer bole
region, parenchyma occupied 25% (SE =1.5%) (Fig. 6) of the
area with no lacunae present (Fig. 6, inset). Therefore, the inner
bole of I. deltoidea palms had approximately twice the water
storage capacity as the outer bole region. Additionally, the percent area occupied by parenchyma tissue remained relatively
constant across height in both the inner and outer bole (Fig. 6).
However, bole cross-sectional areas increased linearly (r2 =
0.92, P < 0.0001) as palms got taller, thus increasing the size of
the inner bole and outer bole region in taller palms.
Bole anatomical properties In the boles of I. deltoidea
palms, vessel diameters increased nonlinearly (fitted with a
power function) with increasing height above ground (r2 = 0.65,
P = 0.0002) (Fig. 7A). Vessels from the inner bole and the outer
bole showed statistically similar curves between vessel diameter and palm height. Unlike boles, vessel diameters in stilt roots

were constant across height classes and therefore were pooled

(Fig. 7A). Vascular bundle densities showed the opposite relation, as vessel diameter in the boles of I. deltoidea decreased
nonlinearly (fitted with a power function) in taller palms (Fig.
7B). Vascular bundle densities in the inner bole decreased more
sharply (r2 = 0.95, P < 0.0001) with increasing height than did
bundle densities in the outer bole (r2 = 0.85, P = 0.0012) (Fig.
7B). As with vessel diameters, vascular bundle densities in the
stilt roots were constant across height classes and were pooled
(Fig. 7B). Calculated Hagen-Poiseuille specific conductivities
increased nonlinearly (fitted with a power function) with palm
height in the outer bole (r2 = 0.74, P = 0.006) but showed no
significant relations with palm height in the inner bole (Fig.
7C). Stilt root conductivities remained constant across height
categories and were pooled (Fig. 7C). Vessel diameters varied
nonlinearly and exponentially with vascular bundle densities
(r2 = 0.63, P < 0.0001), with smaller vessels exhibiting greater
vascular bundle densities (Fig. 8). Likewise, the relation between vessel diameter and vascular bundle density from the

1932

American Journal of Botany

[Vol. 97

Fig. 6. Proportion of parenchyma on an area basis vs. height of sample above ground (m) in the outer bole (lower left picture, cross section, 40,
stained with toluidine blue O) and the inner bole (upper right picture, cross section, 40, stained with toluidine blue O) in Iriartea deltoidea palms. The
inset graph presents bole cross-sectional area measured at breast height (m2) vs. palm height (y = 0.0018x + 0.0008) in I. deltoidea. Bars = 1mm; arrows
point to parenchyma cells measured.

outer bole, inner bole, and stilt roots from I. deltoidea palms
ranging in height from 1 to 25 m all converged on a single line.
DISCUSSION
In both I. deltoidea and M. flexuosa, internode sizes were
significantly reduced at the tops of the tallest palms compared
with sizes at the midheight range. One interesting aspect of
palms is their height growth rates are tied to their leaf production rates. Therefore, although internode lengths are important
in determining height growth rates, leaf turnover rates also need
to be incorporated. According to Lugo and Rivera Batlle (1987),
studies that have not incorporated both pieces of information
may be limited in their interpretations of growth. Iriartea deltoidea and M. flexuosa, in addition to displaying differing patterns of internode lengths along their boles, also have differing
leaf turnover rates with height. In I. deltoidea, leaf turnover
rates slow down in taller palms, and this would tend to accentuate the height growth declines at the tops of tall palms based on
decreasing internode sizes. In M. flexuosa, leaf turnover rates
speed up, and internode lengths represent decreasing lengths of
time from the bottom of M. flexuosa palms to the top. This
would tend to decrease the growth rates indicated by the large
internode lengths at midheight and would suggest slightly faster
growth rates at the tops of tall M. flexuosa than are implied by
the internode lengths alone. But since internode sizes are very
small at the tops of tall palms, rates of height growth are still
reduced in them. Therefore, we found evidence for height
growth declines in the tallest palms of both I. deltoidea and

M. flexuosa, which were near the maximum reported heights

for each species (Henderson et al., 1995). Height growth declines also have been found at the tops of other tall palm species, including coconut palms (Cocos nucifera L.) (Friend and
Corley, 1994) and Prestoea montana ((Graham) Nicholson)
Lugo and Rivera Batlle (1987).
Daily sap fluxes increase as palms get taller, with I. deltoidea
and M. flexuosa having daily sap fluxes of about 10 kg d1 and
45 kg d1, respectively, which are at the lower end of the range
reported by Fisher et al. (2006) for tropical rainforest tree species. However, in both species, leaf areas increase significantly
as palms get taller, which contrasts with the pattern seen in the
subtropical palm species Washingtonia robusta, which exhibited both smaller and fewer leaves in taller palms relative to
shorter ones (Renninger et al., 2009). Therefore, because sap
flux densities from leaves remain relatively constant across
height in both I. deltoidea and M. flexuosa, and leaf areas increase with height, it appears that neither palm shows evidence
for hydraulic limitations in taller palms relative to shorter ones
(Ryan and Yoder, 1997). We did observe that leaf epidermal
cell sizes decreased with palm height in both I. deltoidea and M.
flexuosa. This may be due to height-related decreases in turgor
pressure that may limit cell expansion in leaves from taller
palms and therefore could be considered a hydraulic limitation (Woodruff et al., 2004; Meinzer et al., 2008). Although
these decreases in individual leaf cell sizes did not lead to decreases in overall leaf size, diffusive resistances within leaves
of taller palms may be increased (Niinemets, 2002; Thomas and
Winner, 2002), which could cause decreased gas exchange capacity in leaves from taller palms.

December 2010]

Renninger and PhillipsHydraulic factors in two Amazonian palm species

1933

Fig. 8. Vascular bundle density (mm2) vs. metaxylem vessel diameter (m) for samples from Iriartea deltoidea palms collected from the
outer bole and inner bole at various heights, and from stilt roots of palms
of varying height. The following equation was fitted to the data: y = 68.5
e 0.023x.

Fig. 7. Vascular conduit sizes and distributions in the outer bole, inner
bole, and stilt roots from Iriartea deltoidea palms of various heights.
(A) Metaxylem vessel diameters (m) vs. sample height above ground (m)
for outer and inner bole (y = 71x0.43) and pooled stilt roots, (B) vascular
bundle density (mm2) vs. palm height (m) for outer bole (y = 17 x0.97),
inner bole (y = 27 x1.6), and pooled stilt roots, and (C) calculated HagenPoiseuille specific conductivity (kg m1 s1 MPa1) vs. palm height for
outer bole (y = 6.4x1.2), inner bole, and pooled stilt roots.

If hydraulic factors are not limiting height growth rates in tall

palms, then other constraints need to be evaluated. Homeier et al.
(2002) attribute decreases in height growth at the tops of tall
I. deltoidea palms growing in Costa Rica to higher energy requirements needed for reproduction. Mechanical constraints
also may play a central role in the maximum heights palms can
achieve. In many cases (type A design), final bole diameter in
palms is achieved long before lengthening of the bole occurs
(Niklas, 1992). Furthermore, Niklas (1993) found that estimates
of stem height composed of sclerenchyma matched closely with
mean heights seen in palms, whereas the same was not true for
woody stems. However, Gale and Barfod (1999) found that
most I. deltoidea palms (with a type B design) either died while
standing or were snapped off by other treefalls. This would
suggest a biological, rather than mechanical, mechanism as the
primary driver of height growth reductions in I. deltoidea. Increased respiration costs may be another possibility because
palms accumulate parenchyma tissue as they grow taller. Both
Breure (1988) and Henson (2004) found that maintenance res-

piration per unit biomass decreased with age in oil palms (Elaeis
guineensis Jacq.), but total maintenance respiration increased
because of the increased accumulation of biomass.
Changes in conduit properties with height may provide a clue
as to why taller palms do not show evidence for hydraulic limitation. In I. deltoidea boles, the metaxylem vessel diameters
increase nonlinearly from the lower to the upper portions of the
boles. This is opposite of the pattern predicted by the West et al.
(1999) model. However, the West et al. (1999) model refers to
branched tree systems that exhibit secondary growth, whereby
new conduits added at the base are increasingly larger, and
conduits in new terminal branches are increasingly smaller.
Because palms cannot add new vessels at the base, the taper of
their vascular system cannot readjust itself as they get taller.
Nevertheless, calculated Hagen-Poiseuille specific conductivities increase toward the top of the bole because of increases
in vessel diameter. Therefore, height growth increments add
smaller increments of hydraulic resistance from the soil to the
leaves as palms get taller. This may allow tall palms to overcome hydraulic limitation. Larger parenchyma cells as well as
larger bole diameters may allow taller I. deltoidea palms to
have greater water storage capacity, which also would tend to
mitigate any increased instantaneous resistance in water flow
due to an increased path length resistance (Goldstein et al.,
1998; Phillips et al., 2003b). Likewise, the increase in lacunae
allow for large changes in parenchyma cell water content
without incurring additional pressure forces within the bole
(Holbrook and Sinclair, 1992b). To evaluate the reliance of
M. flexuosa palms on bole water storage, we performed time lag
analysis of sap flux measured in the lower bole and petioles and
found that daily sap flux in taller palms was more reliant on
stored bole water than it was in shorter palms. This may seem
counterintuitive considering M. flexuosa grows in inundated
swamps. However, reliance on stored water, again, would decrease the resistance of water flow to the leaves compared with
pulling water from the inundation zone.
The differences between I. deltoidea and M. flexuosa in terms
of edaphic growing conditions, leaf type, and bole development

1934

[Vol. 97

American Journal of Botany

have led to interesting physiological differences as well. One

interesting difference is the contrasting patterns of internode
length along the boles in these two species. Mauritia flexuosa
exhibited fairly constant internode distances with significant
decreases only at the top of the tallest palms, whereas all palms
of I. deltoidea exhibited increasing internode lengths until
reaching a maximum in approximately the midtrunk region,
followed by further decreases in internode length. Differences
in the pattern of internode length with height may be related
to their differing growth types. Mauritia flexuosa does not increase in diameter once it creates an aboveground trunk,
whereas I. deltoidea has a bole that increases substantially in
diameter after it has been formed. Therefore, internode sizes
appear to be relatively fixed in M. flexuosa. However in I. deltoidea, it seems plausible that if stems can increase in girth after
formation, internodes may be able to lengthen as well. Currently, this idea is speculative, but if internodes did not lengthen
over time, this would mean the small palms we observe today
have drastically smaller vertical growth patterns than the older
palms would have had at the same age range. This remains a
possibility, but the relative decadal stability of climate and vegetation in this area argues against such a drastic difference in
the conditions for growth experienced by the young vs. old
palms. Waterhouse and Quinn (1978) argue that it is anatomically impossible for internodes to lengthen after leaf fall because metaxylem vessels are fully differentiated at that point.
However, it also is known that vascular bundles follow a helical
pattern through the bole (Zimmermann and Tomlinson, 1965;
Zimmermann and Tomlinson, 1972). Therefore, vascular bundles could lengthen in a similar way that a spring lengthens
(without the addition of new material). Secondary stem lengthening in I. deltoidea has yet to be validated but is a novel explanation for the differences in internode lengths observed for the
various palm age classes.
Iriartea deltoidea and M. flexuosa also differ in leaf type,
which could be related to the differing patterns of leaf turnover,
leaf size, and frond counts with height. Leaf turnover rates slow
as I. deltoidea palms get taller, which makes sense considering
taller palms have significantly larger fronds than shorter palms.
Rich (1986) found that leaf production rates increased with
height in I. deltoidea, but only palms up to 13 m tall were studied. In M. flexuosa, leaf turnover rates speed up in taller palms;
slower leaf turnover rates in shorter palms are related to the increased stem tissue increments that are made by shorter palms
relative to the tops of taller palms. Additionally, Lugo and
Rivera Batlle (1987) found that Prestoea montana exhibited
higher rates of leaf production in dominant palms, and de
Steven et al. (1987) found that palms with larger crowns produce more leaves per year than those with fewer leaves.
One other interesting finding from this study is the level of
overlap in the leaf area between I. deltoidea and M. flexuosa
across all ranges of height, with a single line fitting data from
both species. Not only do I. deltoidea and M. flexuosa have
different leaf types (pinnately compound vs. palmate), but they
also exhibit contrasting strategies for increasing leaf area with
height. Iriartea deltoidea shows significant increases in individual frond leaf area in taller palms while maintaining a more
or less constant number of fronds per palm (Rich et al., 1995).
In M. flexuosa, individual frond leaf areas remain more or less
constant, while the number of live fronds increases significantly
in taller palms. The ability to dissipate heat may set an upper
limit to the size M. flexuosa leaves can achieve (Parkhurst and
Loucks, 1972), whereas it may not affect the pinnate leaflets of

I. deltoidea (Balding and Cunningham, 1976). The design of

I. deltoidea leaves also may be important during periods of high
wind, as it has been shown that pinnately compound leaves can
close in around the rachis, forming a cylinder that has low drag
in high winds (Vogel, 1989). The lowland topography of M. flexuosa swamps may make wind a much weaker factor. It is also
interesting to note that M. flexuosa is the only palmate species
in its subtribe; Horn et al. (2009) found palmate leaves evolved
twice in the palm family, with M. flexuosa representing one of
those occurrences. Mauritia flexuosa dominates the permanently flooded areas where it is found, and it can be speculated
as to whether the palmate leaf habit contributes to that.
Therefore, although I. deltoidea and M. flexuosa are both
palm species that are found in the western Amazonian rainforest,
they exhibited interesting differences in intrinsic and extrinsic
properties that were related, to varying degrees, to their differences in edaphic growing environment, leaf habit, and bole
development (type A or type B habit). The similarities and differences between these two palm species exhibit both convergence in solving similar physiological problems as well as the
plasticity seen within the palm family. We also were able to
compare these two palm species individually across a range of
heights to determine how growth rates, leaf area, sap flux rates,
and intrinsic anatomic properties changed as palms grew taller.
These findings have helped elucidate the various strategies
these palms use to move water ever-increasing distances to support expanding leaf areas as they get taller (increased reliance
on stored water, changes in conduit properties). Although
growth rates declined at the tops of tall palms, we found tall
palms were not more hydraulically limited than shorter palms.
Taller palms maintained the same sap flux densities from their
leaves as shorter palms while supporting significantly greater
leaf areas. Finally, we found a very interesting convergence in
the pattern of total leaf area with height in I. deltoidea and
M. flexuosa even though they exhibit very different leaf types.
LITERATURE CITED
Balding, F. R., and G. L. Cunningham. 1976. A comparison of heat
transfer characteristics of simple and pinnate leaf models. Botanical
Gazette 137: 6574.
Barnard, H. R., and M. G. Ryan. 2003. A test of the hydraulic limitation hypothesis in fast-growing Eucalyptus saligna. Plant, Cell &
Environment 26: 12351245.
Breure, C. J. 1988. The effect of palm age and planting density on the partitioning of assimilates in oil palm (Elaeis guineensis). Experimental
Agriculture 24: 5366.
De Steven, D., D. M. Windsor, F. E. Putz, and B. De Len. 1987.
Vegetative and reproductive phenologies of a palm assemblage in
Panama. Biotropica 19: 342356.
Fisher, J. B., J. N. Burch, and L. R. Noblick. 1996. Stem structure of
the Cuban belly palm (Gastrococos crispa). Principes 40: 125128.
Fisher, R. A., M. Williams, R. Lobo Do Vale, A. Lola da Costa, and
P. Meir. 2006. Evidence from Amazonian forest is consistent with
isohydric control of leaf water potential. Plant, Cell & Environment
29: 151165.
Friend, D., and R. H. V. Corley. 1994. Measuring coconut palm dry
matter production. Experimental Agriculture 30: 223235.
Gale, N., and A. S. Barfod. 1999. Canopy tree mode of death in a
western Ecuadorian rain forest. Journal of Tropical Ecology 15:
415436.
Goldstein, G., J. L. Andrade, F. C. Meinzer, N. M. Holbrook, J.
Cavelier, P. Jackson, and A. Celis. 1998. Stem water storage and
diurnal patterns of water use in tropical forest canopy trees. Plant,
Cell & Environment 21: 397406.

December 2010]

Renninger and PhillipsHydraulic factors in two Amazonian palm species

Granier, A. 1987. Evaluation of transpiration in a Douglas-fir stand by

means of sap flow measurements. Tree Physiology 3: 309320.
Henderson, A., G. Galeano, and R. Bernal. 1995. Field guide to the
palms of the Americas. Princeton University Press, Princeton, New
Jersey, USA.
Henson, I. E. 2004. Estimating maintenance respiration of oil palm. Oil
Palm Bulletin 48: 110.
Holbrook, N. M., and T. R. Sinclair. 1992a. Water balance in the
arborescent palm, Sabal palmetto. II. Transpiration and stem water
storage. Plant, Cell & Environment 15: 401409.
Holbrook, N. M., and T. R. Sinclair. 1992b. Water balance in the arborescent palm, Sabal palmetto. I. Stem structure, tissue water release
properties and leaf epidermal conductance. Plant, Cell & Environment
15: 393399.
Homeier, J., S. W. Breckle, H. Dalitz, C. Leyers, and R. Ortiz.
2002. Demography, spatial distribution, and growth of three arborescent palm species in a tropical premontane rain forest in Costa Rica.
Ecotropica 8: 239247.
Horn, J. W., J. B. Fisher, P. B. Tomlinson, C. E. Lewis, and K.
Laubengayer. 2009. Evolution of lamina anatomy in the palm family (Arecaceae). American Journal of Botany 96: 14621486.
Hubbard, R. M., B. J. Bond, and M. G. Ryan. 1999. Evidence that
hydraulic conductance limits photosynthesis in old Pinus ponderosa
trees. Tree Physiology 19: 165172.
Kahn, F. 1988. Ecology of economically important palm in Peruvian
Amazonia. Advances in Economic Botany 6: 4249.
Kahn, F. 1991. Palms as key swamp forest resources in Amazonia.
Forest Ecology and Management 38: 133142.
Koch, G. W., S. C. Sillett, G. M. Jennings, and S. D. Davis. 2004.
The limits to tree height. Nature 428: 851854.
Lugo, A. E., and C. T. Rivera Batlle. 1987. Leaf production, growth
rate, and age of the palm Prestoea montana in the Luquillo Experimental
Forest, Puerto Rico. Journal of Tropical Ecology 3: 151161.
Maca, M. J. 2004. Multiplicity in palm uses by the Huaorani of
Amazonian Ecuador. Botanical Journal of the Linnean Society 144:
149159.
McDowell, N. G., N. Phillips, C. Lunch, B. J. Bond, and M. G. Ryan.
2002. An investigation of hydraulic limitation and compensation in
large, old Douglas-fir trees. Tree Physiology 22: 763774.
Meinzer, F. C., B. J. Bond, and J. A. Karanian. 2008. Biophysical
constraints on leaf expansion in a tall conifer. Tree Physiology 28:
197206.
Meinzer, F. C., S. A. James, and G. Goldstein. 2004. Dynamics of
transpiration, sap flow and use of stored water in tropical forest canopy trees. Tree Physiology 24: 901909.
Montufar, R., and J. C. Pintaud. 2006. Variation in species composition, abundance and microhabitat preference among western
Amazonian terra firme palm communities. Botanical Journal of the
Linnean Society 151: 127140.
Niinemets, U. 2002. Stomatal conductance alone does not explain
the decline in foliar photosynthetic rates with increasing tree age
and size in Picea abies and Pinus sylvestris. Tree Physiology 22:
515535.
Niklas, K. J. 1992. Plant biomechanics: An engineering approach to
plant form and function. The University of Chicago Press, Chicago,
Illinois, USA.
Niklas, K. J. 1993. Influence of tissue densityspecific mechanical properties on the scaling of plant height. Annals of Botany 72: 173179.
Parkhurst, D. F., and O. L. Loucks. 1972. Optimal leaf size in relation
to environment. Journal of Ecology 60: 505537.
Phillips, N., B. J. Bond, N. G. McDowell, M. G. Ryan, and A.
Schauer. 2003a. Leaf area compounds height-related hydraulic
costs of water transport in Oregon white oak trees. Functional Ecology
17: 832840.
Phillips, N., T. N. Buckley, and D. T. Tissue. 2008. Capacity of old
trees to respond to environmental change. Journal of Integrative Plant
Biology 50: 13551364.
Phillips, N., M. G. Ryan, B. J. Bond, N. G. McDowell, T. M. Hinckley,
and J. Cermak. 2003b. Reliance on stored water increases with tree

1935

size in three species in the Pacific Northwest. Tree Physiology 23:

237245.
Phillips, N., F. G. Scholz, S. J. Bucci, G. Goldstein, and F. C.
Meinzer. 2009. Using branch and basal trunk sap flow measurements
to estimate whole-plant water capacitance: Comment on Burgess and
Dawson (2008). Plant and Soil 315: 315324.
Pitman, N. C. A., J. W. Terborgh, M. R. Silman, V. P. Nez, D. A.
Neill, C. E. Cern, W. A. Palacios, and M. Aulestia. 2001.
Dominance and distribution of tree species in upper Amazonian terra
firme forests. Ecology 82: 21012117.
Plotkin, M. J., and M. J. Balick. 1984. Medicinal uses of South
American palms. Journal of Ethnopharmacology 10: 157179.
Renninger, H. J., N. Phillips, and D. R. Hodel. 2009. Comparative
hydraulic and anatomic properties in palm trees (Washingtonia robusta) of varying heights: Implications for hydraulic limitation to
increased height growth. Trees: 23: 911921.
Renninger, H. J., N. Phillips, and G. D. Salvucci. 2010. Wet- vs. dryseason transpiration in an Amazonian rain forest palm Iriartea deltoidea. Biotropica 42: 470478.
Rich, P. M. 1986. Mechanical architecture of arborescent rain forest
palms. Principes 30: 117131.
Rich, P. M. 1987. Developmental anatomy of the stem of Welfia
georgii, Iriartea gigantea and other arborescent palms: Implications
for mechanical support. American Journal of Botany 74: 792
802.
Rich, P. M., N. M. Holbrook, and N. Luttinger. 1995. Leaf development and crown geometry of two Iriarteoid palms. American Journal
of Botany 82: 328336.
Roupsard, O., J. M. Bonnefond, M. Irvine, P. Berbigier, Y.
Nouvellon, J. Dauzat, S. Taga, et al. 2006. Partitioning energy
and evapo-transpiration above and below a tropical palm canopy.
Agricultural and Forest Meteorology 139: 252268.
Rull, V. 1998. Biogeographical and evolutionary considerations of
Mauritia (Arecaceae), based on palynological evidence. Review of
Palaeobotany and Palynology 100: 109122.
Ryan, M. G., B. J. Bond, B. E. Law, R. M. Hubbard, D. Woodruff,
E. Cienciala, and J. Kucera. 2000. Transpiration and whole-tree
conductance in ponderosa pine trees of differing heights. Oecologia
124: 553560.
Ryan, M. G., and B. J. Yoder. 1997. Hydraulic limits to tree height and
tree growth. Bioscience 47: 235242.
Sack, L., P. D. Cowan, N. Jaikumar, and N. M. Holbrook. 2003.
The hydrology of leaves: Co-ordination of structure and function in temperate woody species. Plant, Cell & Environment 26:
13431356.
Salm, R., E. Jalles-Filho, and C. Schuck-Paim. 2005. A model
for the importance of large arborescent palms in the dynamics of seasonally dry Amazonian forests. Biota Neotropica 5:
151156.
Schfer, K. V., R. Oren, and J. D. Tenhunen. 2000. The effect
of tree height on crown level stomatal conductance. Plant, Cell &
Environment 23: 365375.
Sellami, M. H., and M. S. Sifaoui. 2003. Estimating transpiration
in an intercropping system: Measuring sap flow inside the oasis.
Agricultural Water Management 59: 191204.
Thomas, S. C., and W. E. Winner. 2002. Photosynthetic differences
between saplings and adult trees: An integration of field results by
meta-analysis. Tree Physiology 22: 117127.
Tomlinson, P. B. 1963. Measuring growth rates in palms. Principes 7:
4044.
Tomlinson, P. B. 2006. The uniqueness of palms. Botanical Journal of
the Linnean Society 151: 514.
Uhl, N. W., and J. Dransfield. 1987. Genera palmarum. Allen Press,
Kansas, USA.
Vogel, S. 1989. Drag and reconfiguration of broad leaves in high winds.
Journal of Experimental Botany 40: 941948.
Walther, G. R., E. S. Gritti, S. Berger, T. Hickler, Z. Tang, and
M. T. Sykes. 2007. Palms tracking climate change. Global Ecology
and Biogeography 16: 801809.

1936

American Journal of Botany

Waterhouse, J. T., and C. J. Quinn. 1978. Growth patterns in the stem

of the palm Archontophoenix cunninghamiana. Botanical Journal of
the Linnean Society 77: 7393.
West, G. B., J. H. Brown, and B. J. Enquist. 1999. A general model
for the structure and allometry of plant vascular systems. Nature 400:
664667.
Woodruff, D. R., B. J. Bond, and F. C. Meinzer. 2004. Does turgor
limit growth in tall trees? Plant, Cell & Environment 27: 229236.

Zimmermann, M. H., and P. B. Tomlinson. 1965. Anatomy of the

palm Rhapis excelsa I. Mature vegetative axis. Journal of the Arnold
Arboretum. 46: 160180.
Zimmermann, M. H., and P. B. Tomlinson. 1972. The vascular system of monocotyledonous stems. Botanical Gazette 133:
141155.
Zimmermann, M. H. 1983. Xylem structure and the ascent of sap.
Springer-Verlag, New York City, New York, USA.