rspb.royalsocietypublishing.org
An alternative water transport system
in land plants
M. Biddick, I. Hutton and K. C. Burns
School of Biological Sciences, Victoria University of Wellington, PO Box 600, Wellington, New Zealand
Research
Downloaded from https://royalsocietypublishing.org/ on 23 February 2024
Cite this article: Biddick M, Hutton I, Burns
KC. 2018 An alternative water transport system
in land plants. Proc. R. Soc. B 285: 20180995.
http://dx.doi.org/10.1098/rspb.2018.0995
Received: 2 May 2018
Accepted: 9 July 2018
Subject Category:
Evolution
Subject Areas:
ecology, evolution, plant science
Keywords:
Pandanus forsteri (Pandanaceae), plant water
relations, rainfall interception, stemflow
Author for correspondence:
M. Biddick
e-mail: matt.biddick@vuw.ac.nz
MB, 0000-0002-1196-5698; KCB, 0000-0002-4938-2877
The evolution of vascular tissue is a key innovation enabling plants to
inhabit terrestrial environments. Here, we demonstrate extra-vascular
water transport in a giant, prop-rooted monocot from Lord Howe Island.
Pandanus forsteri (Pandanaceae) produces gutter-like leaves that capture
rainwater, which is then couriered along a network of channels to the tips
of aerial roots, where it is stored by absorptive tissue. This passive mechanism of water acquisition, transport and storage is critical to the growth of
aerial prop roots that cannot yet attain water via vascular conduction. This
species therefore sheds light on the elaborate means by which plants have
evolved to attain water.
1. Introduction
The evolution of vascular tissue has enabled hundreds of thousands of land
plants to colonize most of terrestrial Earth [1,2]. Vascular land plants can conduct, transport and store soil water—adaptations paramount to terrestrial life.
Water is transported from roots to leaves in xylem as a result of both, transpiration at the leaves, as well as the cohesive and adhesive properties of water
molecules [3–5]. However, vascular conduction is contingent on soil moisture,
which often occurs patchily in the environment. Further, some plants lack
access to soil altogether (e.g. epiphytes); a phenomenon that has driven the
evolution of alternative means of water acquisition.
Rainfall interception is a complementary and unappreciated water source
for land plants. Intercepted rainwater collects at the stem, where it passively
descends to the roots via gravity (i.e. stemflow). Stemflow has become an
increasingly recognized phenomenon affecting forest hydrology (reviewed in
[6,7]), including soil moisture patterns [8–10], chemistry [11,12], erosion
[13,14] and understory species composition [15,16]. However, few studies
have document traits that increase rainwater harvesting or its physiological
benefits (although see [17]).
Pandanus forsteri (Pandanaceae) is a giant, prop-rooted monocot that is endemic to Lord Howe Island (figure 1a). New prop roots sometimes take years to
reach the ground, during which time they cannot conduct soil water. Here, we
investigate morphological traits in P. forsteri that appear to supply prop roots
with rainwater before they reach the ground. First, we test whether the gutterlike morphology of its leaves increases the amount of rainwater they capture
and channel to the trunk. We then test whether the grooved morphology of its
prop roots directs the flow of rainwater to aerial root tips. Next, we quantify
the water retention capacity of dead epidermal tissue (velamen radicum)
surrounding the tips of aerial roots. Finally, in a long-term growth experiment,
we disable each of these traits to establish whether they interact jointly to form
a water transport system that facilitates the growth of aerial roots.
2. Material and methods
(a) Study site and species
Lord Howe Island (figure 1c) is a small (less than 15 km2), subtropical island located
600 km off the east coast of Australia (318330 S, 1598050 E). The island is the remnant
& 2018 The Author(s) Published by the Royal Society. All rights reserved.
(a)
(b)
2
Proc. R. Soc. B 285: 20180995
Downloaded from https://royalsocietypublishing.org/ on 23 February 2024
rspb.royalsocietypublishing.org
(c)
Figure 1. (a) Pandanus forsteri (Pandanaceae) is a giant, endemic screwpine that supports itself with enormous prop roots. (b) Intercepted rainwater running down
the trunk and a steeply inclined aerial prop root. (c) Lord Howe Island is a UNESCO listed World Heritage Site located 600 km off the east coast of Australia.
of a shield volcano that erupted approximately 6.9 million years
ago and is now dominated by tropical rainforest [18]. Pandanus
forsteri is an endemic and unusually large species of screwpine
that produces a central axis with limited radial growth. For support, it produces sequentially larger prop roots through
ontogeny (figure 1a).
(b) Leaf channels
Leaves have deep grooves that appear to capture and channel
rainwater (figures 2 and 3a). To characterize leaf channelling,
we measured the length, width and channel depth of a single
leaf from each of 73 individuals using a pair of digital callipers.
Measurements were taken at a randomly chosen distance from
the point of attachment (random number generated between
zero and total leaf length). Channel area was then calculated as
the area of a triangle. To test whether relative channel area
increases with total catchment area along the length of the leaf,
we used linear regression to test whether the ratio of channel
area to leaf width increases with proximity to the trunk (n ¼ 73).
To test whether leaf channels capture and transport rainwater, a single leaf from each of 15 individuals was chosen
at random. Leaves were placed into a collection container fixed
at 308 (mean leaf angle measured in the field) and a mechanical
sprinkler system was then used to sprinkle 430 ml of water (the
maximum volume of sprinkler apparatus) onto a standardized
length of leaf (50 cm). Water was collected at the end of the
leaves and weighed using a digital scale (i.e. 1 g ¼ 1 ml). This
procedure was then repeated for each leaf after disabling the
channel with a layer of adhesive tape that was fitted over
the channel along the length of the lamina. We then ran a
linear mixed effects model of water captured against channel
area, with treatment as a fixed factor with two levels (channels
present and channels removed).
(c) Root channels
Roots often have conspicuous longitudinal grooves that appear
to channel the flow of rainwater from the trunk to the tips of
aerial roots (figure 3b). To characterize how root channels are
positioned along prop roots, one root from each of 83 individuals
channel area
leaf width
150
2.5
2.0
100
50
50
Downloaded from https://royalsocietypublishing.org/ on 23 February 2024
100
unk (
cm)
leaf width
30
50
distan
ce to
tr
1.5
3. Results
1.0
Leaf morphology is increasingly more gutter-like towards the
point of attachment to the trunk (figure 2). Experimentally
disabled leaf channels capture less water than control leaves
(figure 3a, d.f. ¼ 27, T ¼ 25.250, p , 0.01). Leaves with channels present captured a mean of 426.1 ml of water (+4.5 s.d.,
range ¼ 415.7, 430), whereas leaves with channels disabled
captured a mean of 328.1 ml (+71.3 s.d., range ¼ 144.8, 403.5).
The radial position of root channels varied with root
inclination (figure 4). Weakly inclined roots (I , 608) primarily
produce channels on their upper surface (cardinal mean ¼
8.118), whereas strongly inclined roots (I . 608) primarily
produce channels on their underside (cardinal mean ¼
177.28). Channels do not form a uniform distribution
around a central tendency (T ¼ 0.481, p , 0.01) and the cardinal means of the two groups (less than 608 versus greater
than 608) were significantly different (T ¼ 0.406, p , 0.01).
Experimental manipulations of root channels demonstrated
that channel presence and root inclination interact to determine the amount of water transported to the tips of aerial
roots (figure 3b; d.f. ¼ 26, T ¼ 3.452, p , 0.01).
The velamen radicum surrounding aerial root tips
retained more water than regular root epidermis (figure 3b;
d.f. ¼ 9.141, T ¼ 7.478, p , 0.01). Root epidermal tissue held
1.5-times its dry mass when soaked (+ 0.30 s.d., range ¼ 1.18,
2.20). Whereas velamen radicum tissue held 9.6-times its dry
mass when soaked (+ 3.33 s.d., range ¼ 4.19, 13.34).
The long-term growth experiment illustrated that both
stemflow diversion and velamen radicum removal affected
root growth. Roots with stemflow diverted grew less longitudinally than control roots (figure 5; d.f. ¼ 13.893, T ¼ 2.475,
p ¼ 0.026). Roots with the velamen radicum removed
grew less radially relative to control roots (d.f. ¼ 11.892,
T ¼ 4.520, p , 0.01).
0.5
Figure 2. Strap leaves in P. forsteri are increasingly more gutter-like towards
the point of attachment to the trunk. That is, leaf channels take up a larger
portion of leaf lamina closer to the trunk.
was randomly chosen for measurement. The orientation of
longitudinal grooves on the root surface and root inclination
from the ground was measured using an electronic protractor.
To test whether the position of root channels varied with root
inclination, we separated the data into two distinct categories:
less than 608 and greater than 608. We then ran Rayleigh’s test
of uniformity and calculated the respective mean cardinal
biases. A Watson’s two-sample test of homogeneity was then
used to test for a difference between means.
To test whether root channels increase the flow of rainwater
to aerial root tips, we randomly chose a single root from each of
15 trees. We precipitated 430 ml of water onto the trunk –root
junction and recorded the amount water channelled to a silicone
interception collar and collection container fitted 10 cm from the
root tip. The channel was then disabled using a layer of adhesive
tape that was fitted over the channel along with length of the root
and the procedure repeated. A linear model was then conducted
using treatment as a fixed factor with two levels (channel
present, channel disabled) and inclination as a covariate.
(d) Root tips
Root tips are surrounded by a velamen radicum consisting
of dead epidermal tissue that appears to absorb rainwater
(figure 3c). To quantify its water retention capacity, we removed
a section of the velamen radicum and adjacent epidermal tissue
from a single root from each of 10 individuals. Samples were
dried using a low heat convection oven to a constant weight,
immersed in water for 20 s, strained for 10 s to remove excess
water and then weighed. The water retention capacity of each
tissue type was then calculated by dividing wet mass by dry
mass. A paired t-test was used to test for a difference in the
water holding capacity of each tissue type.
(e) Root growth
To test whether channelled rainwater facilitates the growth of
aerial roots, a single root was randomly chosen from each of 30
trees. Root channels in 10 trees were disabled by a silicone interception collar fitted 10 cm above the trunk – root junction. The
velamen radicum in 10 trees was surgically removed without
damaging the root meristem. The remaining 10 trees were treated
as controls. All roots were then ring-marked 10 cm from the tip
as a point of reference to measure future growth.
After 22 months (September 2015 – July 2017), two dependent
variables were measured. Longitudinal root growth was
measured as the distance from the ring mark to the end of the
4. Discussion
Pandanus forsteri is unusual among screwpines in that adults
often grow to over 15 m and new prop roots can remain aerial
for years before they ground. During this time, roots are
unable to access and conduct soil water. Selection appears
to have mitigated this constraint by favouring the evolution
of a passive water transport system that does not rely on
soil water. Gutter-like leaves intercept and transport rainwater to the trunk, which is then couriered along a network
of channels directly to aerial root tips, where it is stored by
absorptive tissue.
Other plants are known to intercept rainfall with specialized leaves. Tank bromeliads capture rainwater in cup-like
structures formed by coalescing leaf axils, which enables
them to mitigate drought conditions associated with an
epiphytic existence [20,21]. Tropical palms accumulate rainwater and nutrients in their large, upward-facing fronds,
Proc. R. Soc. B 285: 20180995
40
(mm)
channel
2
)
area (cm
3.0
3
rspb.royalsocietypublishing.org
200
root. Radial root growth was measured by dividing post-treatment
root diameter (10 cm below ring-mark) by pre-treatment root
diameter. Separate Welsh unequal variance t-tests were then
used to test for differences in long-term longitudinal and radial
root growth between treatments. All statistical analyses were performed in R environment [19]. Radial analyses were conducted
using the ‘circular’ and ‘CircStats’ packages.
(a)
4
rspb.royalsocietypublishing.org
water captured (ml)
500
400
300
Proc. R. Soc. B 285: 20180995
200
channels present
channels disabled
Downloaded from https://royalsocietypublishing.org/ on 23 February 2024
(b)
450
water captured (ml)
400
350
300
250
channels present
channels disabled
200
40
60
50
root inclination (°)
70
water retention capacity
(c)
12
10
8
6
4
2
root epidermis
velamen radicum
Figure 3. (a) Gutter-like leaves intercept and transport rainwater to the trunk. Leaves with these channels disabled capture less water. (b) Root channels direct
descending rainwater to aerial root tips. When disabled, roots transport less water to the root tip, particularly at shallower root inclinations. (c) Dead epidermal
tissue (velamen radicum) that stores water at growing aerial root tips. Dried samples of velamen radicum retain markedly more water (greater wet mass : dry mass
ratio) than regular root epidermis when immersed in water and strained.
which then descend the trunk and creates an area of high soil
moisture and nutrients at their base [17,22 –24]. Though the
acquisition of supplementary water transcends leaf traits.
Plants also capitalize on intercepted rainwater through
the production of adventitious roots. Some tropical trees, for
instance, produce adventitious roots in the presence of strong
stemflow, exploiting its nutrient content before it reaches the
soil where it can be leeched by competing understory species
[13,25]. Desiccation resistant Velloziaceae similarly use adventitious roots in their ‘pseudostems’ of dead leaf bases to absorb
rainwater incidentally trapped inside [26–28]. Most curiously,
Amazonian trees have been shown to produce apogeotropic
100
50
C
90°
Downloaded from https://royalsocietypublishing.org/ on 23 February 2024
270
90
270
0
90
180
180
I < 60°
60° > I > 70°
270
90
I > 70°
Figure 4. Pandanus forsteri produces aerial prop-roots that differ in their
inclination relative to the ground. Prop-roots bare deep channels on their
surfaces. The orientation of root channels radially is closely associated with
root inclination (I ). Shallowly inclined roots (I , 608) primarily have channels
located on their upper surface (i.e. 2708 , x , 908, where 08 ¼ upper root
surface oriented skyward), whereas steeply inclined roots (I . 608) tend to
have channels on their underside (908 , x , 2708). Near-vertical roots
seldom have channels, whereas shallowly inclined roots often have several.
roots that grow vertically up the trunks of neighbouring trees,
thereby securing first access to nutrient-rich stemflow [29].
Results from this study demonstrate that mechanisms of
rainwater harvesting can be much more sophisticated than
previously thought.
Pandanus forsteri produces root channels that orient the flow
of intercepted rainwater to the tips of aerial roots. Channels on
the upper root surface of weakly inclined roots reduce the
amount of stemflow that is lost to gravity by preventing runoff. Similarly, channels on the underside of strongly inclined
roots reduce run-off by increasing the surface area contacted
by stemflow. Near-vertical roots generally lack channels,
presumably because water passively descends regardless.
Root channels maximize the volume of stemflow that
reaches root tips primarily by modulating its course. Prior
studies have shown the influence of hydrophilic –hydrophobic
properties of plant surfaces on water dynamics [30–32]. While
the water-repellent nature of Pandanus leaves presumably
facilitates the movement of intercepted rainwater to the
trunk, the degree to which such properties modulate stemflow
dynamics in P. forsteri is not yet known.
The production of root channels entails a physiological
cost. Root channels are formations of missing vascular
tissue and therefore reduce the cross-sectional area of prop
roots. As the primary function of roots is to conduct soil
water and minerals, and to provide structural support, root
channels necessarily reduce both. The results presented
SD
VR
5
AB
B
0.9
0.8
C
SD
VR
Figure 5. The effects of stemflow diversion (SD) and velamen radicum
removal (VR) on root growth after 22-months compared to control roots (C).
Groups sharing common letters do not differ significantly, while groups
that do not share letters are statistically distinguishable. Outliers are denoted
by closed circles.
here, however, suggest the cost of their production is
outweighed by their benefit to growth while aerial.
Many epiphytes are aerially rooted and therefore prone to
drought-induced cavitation and embolism that can render
vascular tissue dysfunctional [33–35]. Some epiphytic orchids
avoid drought by absorbing and storing incident rainfall in a
sheath layer of dead cells that surrounds root tips (‘velamen
radicum’, [36,37]). Pandanus forsteri is not epiphytic. However,
it does produce aerial roots that are enclosed in a type of
velamen; supporting the recent discovery that the velamen
radicum is not an exclusively epiphytic adaptation [38].
Experimentally removing the velamen radicum affected
root growth differently than diverting stemflow. Diverting
the flow of rainwater to aerial roots constrained their longitudinal growth. Removing their absorptive velamen radicum
constrained their growth radially. Why the velamen radicum
and stemflow affect root growth in different ways is unknown.
Pandanus forsteri has evolved a simple water transport
system that captures, transports and stores rainwater, facilitating the growth of its aerial prop roots that cannot yet
attain water from the soil. This system is relatively simple
compared to vascular conduction in that it is comprised of
organs that have evolved to perform other functions, but
have been modified to fulfil the additional role of rainwater
harvesting. Nevertheless, P. forsteri illustrates how plants
have evolved innovative and alternative mechanisms of
water acquisition. Better understanding such adaptations
might prove important to predicting the outcomes of a
changing climate for species already under water constraints.
Data accessibility. All data are deposited at: http://dx.doi.org/10.5061/
dryad.2dp8877 [39].
Authors’ contributions. The study was conceived together by M.B. and
K.C.B. All experimental and morphometric data were obtained by
M.B. The long-term growth experiment was supervised by I.H. for
its 22-month duration. M.B. wrote the manuscript with significant
inputs from K.C.B. and I.H.
Competing interests. We declare we have no competing interests.
Funding. We received no funding for this study.
References
1.
Raven JA 1977 The evolution of vascular land plants
in relation to supracellular transport processes. In
Advances in botanical research, pp. 153 –219.
London, UK: Elsevier.
2.
Lucas WJ et al. 2013 The plant vascular system:
evolution, development and functions.
J. Integ. Plant Biol. 55, 294–388. (doi:10.1111/jipb.
12041)
3.
Tyree MT, Zimmermann MH 2002 Hydraulic
architecture of whole plants and plant performance.
In Xylem structure and the ascent of sap,
pp. 175–214. Berlin, Germany: Springer.
Proc. R. Soc. B 285: 20180995
0
AB
post/pre-treatment root diameter
longitudinal root growth (cm)
B
A
A
rspb.royalsocietypublishing.org
0°
root
inclination
(I)
1.0
150
4.
6.
Downloaded from https://royalsocietypublishing.org/ on 23 February 2024
8.
9.
10.
11.
12.
13.
14.
15.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
desiccation-tolerant plant Vellozia flavicans in the
savannas of central Brazil. J. Trop. Ecol. 21, 585–
588. (doi:10.1017/S0266467405002658)
Sanford RL. 1987 Apogeotropic roots in an Amazon
rain forest. Science 235, 1062 –1064. (doi:10.1126/
science.235.4792.1062)
Levia D, Herwitz S. 2005 Interspecific variation of
bark water storage capacity of three deciduous tree
species in relation to stemflow yield and solute flux
to forest soils. Catena 64, 117 –137. (doi:10.1016/j.
catena.2005.08.001)
Holder CD. 2007 Leaf water repellency of species in
Guatemala and Colorado (USA) and its significance
to forest hydrology studies. J. Hydrol. 336,
147–154. (doi:10.1016/j.jhydrol.2006.12.018)
Rosado BH, Holder CD. 2013 The significance of leaf
water repellency in ecohydrological research: a review.
Ecohydrology 6, 150–161. (doi:10.1002/eco.1340)
Pickard WF. 1981 The ascent of sap in plants. Prog.
Biophys. Mol. Biol. 37, 181 –229. (doi:10.1016/
0079-6107(82)90023-2)
Tyree MT, Sperry JS. 1989 Vulnerability of xylem to
cavitation and embolism. Annu. Rev. Plant Biol.
40, 19–36. (doi:10.1146/annurev.pp.40.060189.
000315)
Venturas MD, Sperry JS, Hacke UG. 2017 Plant
xylem hydraulics: what we understand, current
research, and future challenges. J. Integr. Plant Biol.
59, 356 –389. (doi:10.1111/jipb.12534)
Capesius I, Barthlott W. 1975 IsotopenMarkierungen and Raster-elektronenmikroskopische
Untersuchungen des Velameya radicum der
Orchideen: Isotope Labelling and Scanning Electron
Microscope Studies of the Velamen radicum of
Orchideen. Z. Pflanzenphysiol. 75, 436–448.
(doi:10.1016/S0044-328X(75)80138-3)
Zotz G, Winkler U. 2013 Aerial roots of epiphytic
orchids: the velamen radicum and its role in water
and nutrient uptake. Oecologia 171, 733– 741.
(doi:10.1007/s00442-012-2575-6)
Zotz G, Schickenberg N, Albach D. 2017 The
velamen radicum is common among terrestrial
monocotyledons. Ann. Bot. 120, 625–632. (doi:10.
1093/aob/mcx097)
Biddick M, Hutton I, Burns KC. 2018 Data from: An
alternative water transport system in land plants.
Dyrad Digital Repository. (doi:10.5061/10.5061/
dryad.2dp8877)
6
Proc. R. Soc. B 285: 20180995
7.
16.
soil chemistry and vegetation patterns. Can. J. For.
Res. 21, 917–924. (doi:10.1139/x91-127)
Awasthi O, Sharma E, Palni L. 1995 Stemflow: a source
of nutrients in some naturally growing epiphytic
orchids of the Sikkim Himalaya. Ann. Bot. 75, 5–11.
(doi:10.1016/S0305-7364(05)80003-4)
Edwards PJ, Fleischer Dogley F, Kaiser Bunbury CN.
2015 The nutrient economy of Lodoicea maldivica, a
monodominant palm producing the world’s largest
seed. New Phytol. 206, 990–999. (doi:10.1111/
nph.13272)
McDougall I, Embleton B, Stone D. 1981 Origin and
evolution of Lord Howe Island, southwest Pacific
Ocean. J. Geol. Soc. Aust. 28, 155 –176. (doi:10.
1080/00167618108729154)
R Core Team. 2000 R language definition. Vienna,
Austria: R foundation for statistical computing.
Zotz G, Thomas V. 1999 How much water is in the
tank? Model calculations for two epiphytic
bromeliads. Ann. Bot. 83, 183–192. (doi:10.1006/
anbo.1998.0809)
Benzing DH. 2008 Vascular epiphytes: general
biology and related biota. Cambridge, UK:
Cambridge, University Press.
Raich JW. 1983 Understory palms as nutrient traps:
a hypothesis. Brenesia. San Jose 21, 119– 129.
Enright N. 1987 Stemflow as a nutrient source for
nikau palm (Rhopalostylis sapida) in a New Zealand
forest. Austral Ecol. 12, 17 –24. (doi:10.1111/j.14429993.1987.tb00923.x)
Germer S, Zimmermann A, Neill C, Krusche AV,
Elsenbeer H. 2012 Disproportionate single-species
contribution to canopy –soil nutrient flux in an
Amazonian rainforest. Forest Ecol. Manag. 267, 40 –
49. (doi:10.1016/j.foreco.2011.11.041)
Nadkarni NM. 1981 Canopy roots: convergent
evolution in rainforest nutrient cycles. Science 214,
1023 –1024. (doi:10.1126/science.214.4524.1023)
Porembski S, Barthlott W. 1995 On the occurrence
of a velamen radicum in Cyperaceae and
Velloziaceae. Nord. J. Bot. 15, 625–629. (doi:10.
1111/j.1756-1051.1995.tb02131.x)
Porembski S, Barthlott W. 2000 Granitic and gneissic
outcrops (inselbergs) as centers of diversity for
desiccation-tolerant vascular plants. Plant Ecol. 151,
19 –28. (doi:10.1023/A:1026565817218)
Oliveira RS, Dawson TE, Burgess SS. 2005 Evidence
for direct water absorption by the shoot of the
rspb.royalsocietypublishing.org
5.
Kramer P. 2012 Physiology of woody plants.
Amsterdam, the Netherlands: Elsevier.
Jones HG. 2013 Plants and microclimate: a
quantitative approach to environmental
plant physiology. Cambridge, UK: Cambridge
University Press.
Levia Jr DF, Frost EE. 2003 A review and evaluation
of stemflow literature in the hydrologic and
biogeochemical cycles of forested and agricultural
ecosystems. J. Hydrol. 274, 1–29. (doi:10.1016/
S0022-1694(02)00399-2)
Levia DF, Germer S. 2015 A review of stemflow
generation dynamics and stemflow–environment
interactions in forests and shrublands. Rev. Geophys.
53, 673–714. (doi:10.1002/2015RG000479)
Durocher M. 1990 Monitoring spatial variability of
forest interception. Hydrol. Processes 4, 215–229.
(doi:10.1002/hyp.3360040303)
Chang S-C, Matzner E. 2000 The effect of beech
stemflow on spatial patterns of soil solution
chemistry and seepage fluxes in a mixed beech/oak
stand. Hydrol. Process 14, 135 –144. (doi:10.1002/
(SICI)1099-1085(200001)14:1,135::AIDHYP915.3.0.CO;2-R)
Germer S. 2013 Development of near-surface
perched water tables during natural and artificial
stemflow generation by babassu palms. J. Hydrol.
507, 262–272. (doi:10.1016/j.jhydrol.2013.10.026)
Escudero A, Hernandez M, Del Arco J 1991 Spatial
patterns of soil composition around isolated trees.
In Developments in geochemistry, pp. 507 –517.
London, UK: Elsevier.
Matschonat G, Falkengren-Grerup U. 2000 Recovery
of soil pH, cation-exchange capacity and the
saturation of exchange sites from stemflow-induced
soil acidification in three Swedish beech (Fagus
sylvatica L.) forests. Scand. J. Forest R. 15, 39– 48.
(doi:10.1080/02827580050160457)
Herwitz SR. 1991 Aboveground adventitious roots
and stemflow chemistry of Ceratopetalum virchowii
in an Australian montane tropical rain forest.
Biotropica 23, 210–218. (doi:10.2307/2388197)
Cattan P, Bussière F, Nouvellon A. 2007 Evidence of
large rainfall partitioning patterns by banana and
impact on surface runoff generation. Hydrol.
Processes 21, 2196–2205. (doi:10.1002/hyp.6588)
Andersson T. 1991 Influence of stemflow and
throughfall from common oak (Quercus robur) on