Journal of Hydrology 237 (2000) 40–57
www.elsevier.com/locate/jhydrol
Gross rainfall and its partitioning into throughfall, stemflow and
evaporation of intercepted water in four forest ecosystems in
western Amazonia
C. Tobón Marin a,b,*, W. Bouten a, J. Sevink a
a
Fysisch Geografisch Bodemkundig Laboratorium, University of Amsterdam, Nieuwe Achtergracht 166, 1018 WV Amsterdam, The Netherlands
b
The Tropenbos Foundation, Colombia
Received 29 July 1999; revised 4 May 2000; accepted 4 July 2000
Abstract
The partitioning of gross rainfall into throughfall, stemflow and evaporation of intercepted rainfall was studied in four forest
ecosystems in the Middle Caquetá, Colombian Amazonia. Data on climate was collected automatically on an hourly basis
during a five-year period. Weekly measurements of rainfall, throughfall and stemflow were carried out during a period of two
years, while daily measurements, on an event basis, were carried out during two subsequent years. Throughfall, stemflow and
evaporation in each forest were checked for correlations with gross rainfall characteristics, canopy gap fraction, tree crown area
and bark texture. Canopy gap fraction differed between forests, ranging from 9% on the flood plain to 17% on the Tertiary
sedimentary plain. Rainfall was rather evenly distributed over the year, with one dry period from December to February. 92% of
the rain fell in single showers of less than 30 mm and most of the storms (56%) fell in less than one hour, during the afternoon or
early night. Throughfall ranged from 82 to 87% of gross rainfall in the forests studied and varied with gross rainfall in all
forests. It depended on the amounts and characteristics of rainfall, but differences in throughfall among forests, when comparing
similar rainfall events, clearly indicated that throughfall also depends on forest structure. Stemflow contributed little to net
precipitation (on average 1.1% of gross rainfall in all forests) and showed a power relation with gross rainfall. Correlations
between stemflow per tree, projected crown area and bark texture were very poor as indicated by the low coefficients of
determination. Evaporation during rainfall events exhibited a linear relation with rainfall duration and the ratio of evaporation
over gross rainfall increased with forest cover (1-gap fraction) in the forests studied. The structure of the forests seemed to vary
considerably and given its influence on rainfall partitioning it may explain both differences and similarities between results
from this study and those from most other studies within Amazonia. 䉷 2000 Elsevier Science B.V. All rights reserved.
Keywords: Amazonia; Tropical rain forest; Throughfall; Stemflow; Evaporation; Forest structure
1. Introduction
The Amazonian rain forest seems to play an important role in the regulation of regional and global
* Corresponding author. Tel.: ⫹31-205257442; fax: ⫹31205257431.
E-mail address: [email protected] (C. Tobón Marin).
climate (Salati and Vose, 1984). Using a modelling
approach, Eltahir and Bras (1993) concluded that the
atmosphere in the Amazon basin is an open system
and that the net input of atmospheric moisture into the
basin is about 32%, about 68% of the gross input
leaving the basin. They also found that the recycling
ratio of the Amazon basin is about 25–35%. This ratio
differs from those found by others, which were based
0022-1694/00/$ - see front matter 䉷 2000 Elsevier Science B.V. All rights reserved.
PII: S0022-1694(00 )00 301-2
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
on the erroneous hypothesis that the atmosphere over
the Amazon basin is a closed system (Molion, 1975;
Lettau et al., 1979; Salati et al., 1979). Eltahir and
Bras (1993) furthermore concluded that deforestation
of Amazonia would increase the surface temperature
and decrease the heating of the upper troposphere,
which would result in a reduction of precipitation.
The foregoing illustrates that recent concern about
tropical rain forest deforestation focuses on its impact
on climate at regional and global scale (e.g. The
Anglo Brazilian Amazonia Climate observation
study ABRACOS, Gash et al., 1996; the Large
Scale Biosphere–Atmosphere experiment in Amazonia, Nobre et al., 1996; Eltahir and Bras, 1993).
Current research concentrates on global circulation
models and on those parameters and fluxes that play
a role in the global climate (e.g. Gash et al., 1996).
Nevertheless, it is often argued that additional work is
needed to explain and improve model predictions for
impacts of deforestation (Bruijnzeel, 1990; Nobre et
al., 1991). Evidently, local hydrological studies on
undisturbed mature rain forests would provide base
level information on initial conditions that might
allow for an evaluation of the presumed influence of
deforestation on regional and global climate. This
paper concerns such local study.
In the Middle Caquetá (Colombian Amazonia), the
structure and species composition of the forest vary
considerably between the different landscape units
(Duivenvoorden and Lips, 1995; Londoño, 1993).
This causes equally large variations in temporal and
spatial patterns of water fluxes in these units, leading
to local differences in water and nutrient stocks in the
various forest compartments (Vitousek and Denslow,
1986; Tanner, 1985). Thus, the partitioning of rainfall
into throughfall and stemflow leads not only to a more
diffuse input of water into the forest floor, but also to
local concentration around the base of tree stems,
which is known to induce spatial variability in soil
properties and soil moisture conditions (Waidi et al.,
1992). In tropical forests, the abundance of epiphytes,
climbers and aerial roots renders this partitioning
much more complicated than in temperate forests
(Longman and Jenı́k, 1990).
It is well understood that in forested areas generally
total evaporation is larger than in areas with shorter
vegetation (e.g. grass) mainly due to the larger interception by the forest canopy (Bosch and Hewlett,
41
1982), which has been related to the large aerodynamic conductance of forest (Stewart, 1977). Additionally, there has been an increasing awareness that
evaporation of intercepted rainfall has to be investigated separate from transpiration, especially in very
humid areas (Hutjes et al., 1990; Shuttleworth and
Calder, 1979). Furthermore, most studies on forest
interception showed that this interception is closely
related to gross rainfall amounts and characteristics.
However, the influence of forest structure on
interception is poorly known.
Most rainfall interception and water balance studies
in Amazonia were executed in Brazil (Ubarana, 1996;
Leopoldo et al., 1995; Lesak, 1993; Lloyd and
Marques, 1988; Shuttleworth, 1988) and only a few
in other parts of Amazonia (Hölscher et al., 1997;
Jetten, 1996; Wright et al., 1992; Jordan, 1978).
Thus, in Colombian Amazonia, which represents
one of the most humid areas within the basin, very
little attention has been paid to the hydrology of forest
ecosystems and to the effects of forest structure on
water dynamics.
This paper concerns a study designed to address
this lack of knowledge by measuring rainfall and its
partitioning after entering the canopy in four undisturbed rain forests in the Middle Caquetá, Colombian
Amazonia. It focuses on the analysis of long-term
hydrological measurements of rainfall, throughfall,
stemflow, the resultant evaporation and the related
structure of these forests.
2. The study area
The study area is in Peña Roja (Nonuya Indian
community) near Araracuara, Middle Caquetá,
Colombia, (0⬚ 37 0 and 1⬚ 24 0 S, 72⬚ 23 0 and 70⬚ 43 0
W; Fig. 1). Climate is classified as equatorial superhumid Afi (Köppen, 1936). The research sites are
permanent undisturbed forest plots, used by the
Tropenbos Foundation for its research. They lie
approximately 200–250 metres above sea level and
form a sequence from the lower terrace of the River
Caquetá to the Tertiary sedimentary plain. Based on
data from the manual meteorological station at Araracuara (IDEAM), average annual rainfall in the area is
about 3100, April being the wettest month and
January the driest.
42
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
Fig. 1. Location of the research sites in the various land units in the Middle Caquetá, Colombian Amazonia.
Colombian Amazonia comprises 403,000 km 2 and
the major part of this area is covered by mature rain
forests, classified as ombrophilous tropical forest
(Duivenvoorden, 1995). The research plots are
located in the four main land units in the area: the
Tertiary sedimentary plain, the upland terraces of
the River Caquetá (high and low terraces) and the
flood plain. The vegetation is very rich in species
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
(Duivenvoorden and Lips, 1995; Londoño, 1993) and
is typical for undisturbed mature forests in the western
part of the Amazon basin. The canopy reaches about
25–30 m above the forest floor with some emergent
trees reaching up to 45 m in the rarely inundated flood
plain. There are three to four canopy layers, but large
tree crowns in the upper canopy form the bulk of the
vegetation. Lower canopies contribute far less to the
forest cover. Small palms reaching a height of 2–4 m
constitute the lowest layer. The land units differ in the
total standing biomass, species diversity and tree
density (Duivenvoorden and Lips, 1995; Londoño,
1993). Other differences between plots pertain to the
structure of the forest canopy (canopy cover) and the
contribution of epiphytes, climbers and aerial roots. A
more detailed description and vegetation classification of the research sites is given by Duivenvoorden
and Lips (1995); and Londoño (1993).
3. Materials and methods
The areas for the present study were selected as
being representative for the natural vegetation in
the main land units from this part of the Amazon
basin. Three subplots were selected in the Tertiary
sedimentary plain (SP) and two subplots in the
high terrace (HT), the low terrace (LT) and the
flood plain (FP), respectively, to measure gross
rainfall above the forest canopy, throughfall and
stemflow (Fig. 1). In 1992, approximately 3 km
from the plots, in an open area of about 20 hectare
(within an Indian community village), an
automatic weather station (AWS) was installed
to measure gross rainfall, temperature, air
humidity, incoming radiation, wind speed, wind
direction and Class A pan evaporation.
Parameters were measured and recorded every
30 s using a datalogger (CR10 Campbell Scientific
Instruments), which additionally recorded means
or totals every 20 min.
Rainfall in the open area was measured by a tipping
bucket raingauge with a resolution of 0.2 mm providing information on the number and duration of
showers and on total rainfall. Gross rainfall in each
plot was measured in two ways: (1) automatic
measurements with one tipping bucket installed in
the top of an emergent tree crown after clearing all
43
branches; (2) manual measurements with two raingauges per subplot suspended from cords attached to
two emergent trees in small gaps within the forest.
Throughfall was measured in the same rainfall
subplots using 20 collectors per subplot, randomly
located in an area of 50 by 20 m (1000 m 2). Evaporation from the collectors was avoided by using an
internal plastic tube running from the funnels to the
bottom of the collectors. To allow for direct correlation, all funnels for gross rainfall above the forest
canopy and for throughfall had an orifice of
298.6 cm 2. Throughfall and forest rainfall collectors
were calibrated against standard raingauges in the
open.
Because of the large variability in throughfall due
to the forest structure (Jetten, 1996; Ford and Deans,
1978) many readings are needed to study forest interception. However, when using average values,
moving the collectors has a positive effect by reducing
the standard error of estimations (Lloyd and Marques,
1988). Therefore, each month (after five measurements) collectors were randomly relocated within
all subplots during the period of ‘single event
measurements’, i.e. between 1996–1997.
Stemflow was measured for 15 randomly selected
trees in each subplot. Collars, constructed from 8 mm
thick black polyethylene plastic, were sealed to the
stems in an upward spiral pattern and the water
diverted into bottle gauges on the forest floor. The
opening of each collar extended only about 2–3 cm
from the trunk to avoid drips from the branches or
leaves being collected by the collars. The amount of
water, which drops in a diffuse pattern around roughbarked trees rather than adhering to and flowing down
the trunks, was considered throughfall. For practical
reasons, only trees with diameter larger than 10 cm
were selected for stemflow measurements. Where
palm trees were present in the subplots, depending
on their frequency of occurrence one or two of
palms per subplot were randomly selected. In the SP
and HT plots, two palms were selected, in the LT four
palms in total and in the FP three palms. Stemflow
measurements were rotated once by installing collars
around stems of new trees within the same subplots.
Horizontally or downwards inclined branches of
trees may not direct intercepted rainfall to the centre
of the tree to be drained as stemflow. Therefore, the
flat area of the tree crown was mapped by means of
44
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
vertical projections from the edge of outstanding
upward branches to the forest floor. At least six
projection lines were drawn for each tree. The
crown area was found by integrating measured areas
of each triangle. Stemflow is expressed as millilitre of
water coming in over the horizontal area of the tree
crown and flowing down along the tree trunk. Thus,
each tree was considered as a single catchment area.
Estimates for stemflow at subplot level were obtained
by multiplying the average stemflow per tree of a
specific diameter by the total number of trees with
approximately the same diameter within the subplot.
Data on sampled trees with a diameter of about 10 cm
were used to extrapolate the information to trees with
smaller diameter. Average stemflow per subplot was
related to gross rainfall measured in the same subplot
to determine the percentage of rainfall coming to the
forest floor as stemflow.
Stemflow amounts from each rainfall event (ml)
and from individual trees were correlated with the
measured projected area of the central crown of the
tree concerned and with some tree characteristics such
as trunk surface area and bark texture. Consequently,
the periphery of the trunks or surface area was determined by measuring the tree trunk circumference at
breast height and the height of the tree trunk to the first
set of divergent branches. Trunk area was determined
by considering the tree trunk as being rectangular and
applying a correction factor to account for the conical
form of tree. Bark texture was classified according to
the roughness of the bark in a range from smooth to
very fibrous. Per plot, 40 hemispherical photographs
were taken to estimate the gap fraction of the forest
canopy in the subplots where throughfall collectors
were installed. The photographic camera was installed
horizontally at the same height as the throughfall
collectors. The black and white photographs were
digitised with a scanner and analysed with
the Hemiphot program (Ter Stegee, 1994), which
calculates the gap fraction of a forest and roughly
estimates the LAI from the forest cover. Although
the photographs were taken under covered sky with
diffuse sun light, they were corrected for the reflected
light from leaves, branches and trunks. It is assumed
that the mean fraction of white pixels in the
photographs gives the best estimate of the gap
fraction.
Most methods for the determination of the water
storage capacity of the forest canopy (C), which
have been presented thusfar (Klaassen et al., 1998),
mainly differ in the way of accounting for the drainage
before the forest is completely saturated and the
gradual saturation of forest layers with continuously
proceeding evaporation. In the present study, C is the
intercept of the regression of the estimated evaporation versus gross rainfall. Therefore, only single, highintensity rainfall events of short duration (less than
one hour) were used. Late afternoon and night rainfalls were preferred, representing rainfalls under
conditions of a low moisture deficit. In total, 30–40
events were selected for each ecosystem. Free
throughfall is considered as an important parameter,
among others, in studies on nutrient cycling, since it
may represent the fraction of throughfall which is not
involved in washing out dry deposition, exudates and
released nutrients from leaves, branches and trunks.
Free throughfall in the forests may be taken as the gap
fraction (ps). Therefore, it was estimated for each
forest from the set of digitised and scanned black
and white photographs taken in each subplot and
from the regression coefficient of throughfall versus
gross rainfall (pt), using data from small storms,
which were insufficient to saturate the forest canopy
(Gash and Morton, 1978). About 14 small storms were
selected for each forest to determine the value of pt.
Manual measurements of rainfall, throughfall and
stemflow were carried out on weekly basis from
December 1993 to February 1996, without moving
collectors. During these years, some measurements
were carried out on event basis during periods with
limited or no rainfall (mostly in the dry season). Daily
measurements were performed from February 1996 to
August 1997. Readings were made early in the morning (around 08:00 local time) or after the storm. Personal observation in the plots showed that the upper
canopy dried out within 6 h after rainfall, if this
event occurred during the daytime or part of it.
Therefore, the criteria to separate the single events
was that only those events where no rainfall occurred
in the previous 6 h were considered. In those cases
where two or more events occurred during the night
time, since evaporation during the night is expected to
be low, they were considered as a single event. Most
analysed single events in this study are events,
preceded by a dry period of more than 10 h.
Evaporation (E), which is taken here as the total
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
amount of water intercepted and evaporated from the
forest canopy, is calculated on event basis from the
difference between gross rainfall (Pg) and net rainfall
(throughfall Th plus stemflow Sf) of the single rainfall
events. Cumulative evaporation during the rainfall
events (Ew) is calculated from the difference between
gross rainfall and net precipitation plus the water
stored in the forest canopy after rainfall ceased (S),
deduced from the estimated evaporation of single
events minus the storage capacity (C)
Ew ˆ Pg ⫺ …Th ⫹ Sf ⫹ S†
…1†
To avoid negative values in Eq. 1, although
they may reflect those events where the forest
canopy does not reach saturation, the value for S
was inferred from the single rainfall events where
E was larger than C.
Statistical analyses were made for the entire
collected data. However, for the assessment of the
effect of rainfall sizes on net rainfall and resulting
evaporation, measurements of single rainfall events
are required. Therefore, throughfall percentages relative to gross rainfall and evaporation were calculated
from these single events. Moreover, to determine the
effect of forest structure on throughfall percentage and
evaporation, we used only data from single rainfall
events collected during the period of weekly
measurements when collectors were not relocated.
4. Results
4.1. Forest structure
Average values of measured variables of forest
structure in each forest ecosystem studied are
presented in Table 1. The largest mapped crown
2
45
2
area was 62.2 m and the smallest 2.4 m , both in
the high-terrace plot. For the trunk surface area, the
correction factor of 0.5 was applied to trees with
diameter larger than 0.1 m. The apparent inconsistency in the flood plain data, of a large tree crown
area in combination with the lowest tree trunk surface
area, is mainly due to the abundance of small trees
with a diameter of about 0.1 m. The data on the estimated gap fraction of the canopy and the free throughfall coefficient show that considerable differences
exist between forests, the largest values being
observed in the sedimentary plain. Moreover, the
leaf area index appeared to decrease from the sedimentary plain to the flood plain of the River Caquetá.
The estimated LAI values are within the range of
values found for similar forests in Brazilian Amazonia
(Roberts et al., 1996; Klinge et al., 1975). The
decrease in gap fraction and increase in LAI towards
the flood plain is in line with the results from Duivenvoorden and Lips (1995), who observed that litter fall
in the flood plain of the River Caquetá is highest when
compared to the other ecosystems in the area.
4.2. Rainfall characteristics
Gross rainfall above the forest canopy did not vary
considerably within plots. On average, the differences
in the amounts of gross rainfall between subplots in
the SP were 5.5% (^5.6) with n ˆ 3; whereas in the
HT these were 3.7% (^3.5), in the LT gauges 3.7%
(^3.0) and in the FP 3.1% (^2.7), all with n ˆ 2:
Rainfall distribution differs between plots when
examining separate storms, although annual totals
are rather similar. During the measurement period
(1992–1997), the mean annual rainfall at the AWS
was about 3400 mm y ⫺1 and the average period with
rainfall was 616 h y ⫺1. In total, 1584 rainfall events
Table 1
Forest structure characteristics of four forest ecosystems in the Middle Caquetá, Colombian Amazonia
Parameter
Sedimentary plain
High terrace
Low terrace
Flood plain
Crown area (m 2)
Stem trunk area (m 2)
Gap fraction (%)
Free throughfall coefficient, pt
LAI
Forest storage capacity (mm)
9.5 (^7.3)
7.2 (^4.6)
16.8 (^2.4)
0.59
4.4 (^0.7)
1.16
17.8 (^14.0)
8.2 (^5.2)
15.4 (^3.4)
0.52
4.9 (^0.8)
1.28
9.8 (^7.2)
8.4 (^5.4)
11.7 (^1.5)
0.49
5.6 (^0.6)
1.32
12.1 (^8.3)
5.4 (^4.6)
8.2 (^1.5)
0.27
6.6 (^0.4)
1.55
46
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
Fig. 2. Rainfall characteristics at the research site in the Middle Caquetá, Colombian Amazonia.(August 1992 to August 1997)
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
Fig. 2. (continued)
47
48
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
Table 2
Throughfall percentages of daily gross rainfall for 5 storm classes, in four forests in the Middle Caquetá, Colombian Amazonia. (SP)
sedimentary plain, (HT) high terrace, (LT) low terrace, (FP) flood plain, (std) standard deviation of the means and (n) number of events
Rainfall ranges
Throughfall %
SP
HT
LT
FP
(mm)
%
std
n
%
std
n
%
std
n
%
std
n
⬍5
5–20
20–40
40–80
⬎ 80
Total
58.7
81.4
88.9
90.6
92.8
87.2
11.4
6.3
2.8
2.1
–
2.4
41
78
39
19
1
178
56.2
80.5
87.9
90.0
92.2
86.7
12.1
5.6
2.6
1.9
0.8
2.4
34
68
36
19
3
160
52.3
79.8
87.7
88.8
92.4
85.8
9.6
6.8
3.4
2.5
1.0
1.2
32
71
41
17
2
163
47.4
74.5
83.0
84.6
88.5
81.9
13.2
6.9
2.5
3.1
1.2
1.0
27
57
32
19
5
140
were recorded at the AWS in Peña Roja, with storms
ranging from 0.2 to 161.6 mm and lasting between
20 min and 13 h. During the total period, 37% of the
incident rain fell in single showers of less than 2 mm
and 92.3% of these showers contributed with less than
30 mm. Rainfall intensity, calculated for the total
period with some rainfall, averaged 5.46 mm h ⫺1
with a maximum of 78.16 mm h ⫺1 (Fig. 2a). Most
showers (63%) fell during the afternoon and at night
(Fig. 2b) and 56% of these storms fell in less than 1 h
(Fig. 2c). Monthly rainfall distribution during the fiveyear period shows that there was a slightly drier
period from December to February (Fig. 2d). Comparing our data on five years rainfall with data from
earlier years in the Middle Caquetá (Duivenvoorden
and Lips, 1995), rainfall characteristics appear to be
similar to the long-term average.
4.3. Throughfall
The variability of throughfall within a subplot was
large, with the smallest variation in the FP forest,
although differences in average values between
subplots were small. The average coefficient of
variation (CV) of individual gauges in each plot was
0.285 (^0.10) in the SP, 0.306 (^0.07) in the HT,
0.279 (^0.09) in the LT and 0.225 (^0.08) in the
FP forest. The CV of the mean throughfall in each
subplot was 0.062 (^0.058) in the SP, 0.043
(^0.05) in the HT, 0.046 (^0.04) in the LT and
0.047 (^0.04) in the FP forest. As a general trend,
for small rainfall events the value of the standard
deviation (std) of throughfall (expressed as a percen-
tage of mean throughfall) over gross rainfall varied
more than for major events. Furthermore, for some
individual throughfall gauges values exceeded gross
rainfall (e.g. 29% of the individual gauge values in the
SP were larger than gross rainfall, whereas in the HT
this was 30%, in the LT 27% and in the FP forest
21%), but the average (of 60 and 40 gauges) was
always lower than gross rainfall.
Throughfall was calculated as a percentage of gross
rainfall for five different rainfall sizes and from the
totals of the measured daily gross rainfall and
throughfall during the study. Throughfall ranged
from zero, with events below 2 mm, to 95% in storms
larger than 100 mm, but mean throughfall varied from
50 to 93% depending on gross rainfall amounts and
the type of forest (Table 2). The calculated value of
total throughfall relative to total gross rainfall ranged
from 82 to 87% in the four forests.
Although empirical regression equations provide
only site-specific information, they may indicate a
trend especially if explained variance is high. Therefore,
Table 3
Regression parameters of throughfall versus gross rainfall in four
different forest ecosystems in Colombian Amazonia. (se) standard
error of regression coefficient (Note: The equation for linear form is
T ˆ a ⫹ bPg ; where T is throughfall amount and Pg is gross rainfall
(mm))
Landscape unit
a
b
se
R2
n
Sedimentary plain
High terrace
Low terrace
Flood plain
⫺1.02
⫺1.02
⫺1.07
⫺1.48
0.926
0.918
0.906
0.887
0.003
0.003
0.004
0.003
0.99
0.99
0.99
0.99
102
97
97
84
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
49
Fig. 3. Average throughfall and its standard deviation (std) against gross rainfall in a forest ecosystem (Sedimentary plain) in the Middle
Caquetá, Colombian Amazonia.
regressions of throughfall versus gross rainfall
were computed from single storms for each forest
(Table 3). Average throughfall per plot was highly
correlated with gross rainfall in all forests (Fig. 3).
ANOVA analysis showed that the ratio of mean
throughfall over gross rainfall in the FP forest was
significantly different from the other forests (at 95%
level).
4.4. Stemflow
Large differences were observed in the amount of
stemflow of individual trees and among subplots. In
general, however, the contribution of stemflow to net
rainfall was very low. The average CV in each plot
was 0.295 (^0.12) in the SP, 0.207 (^0.12) in the HT,
0.323 (^0.20) in the LT and 0.303 (0.26) in the FP
Fig. 4. Average stemflow and its standard deviation (std) against gross rainfall in a forest ecosystem (Sedimentary plain) in the Middle Caquetá,
Colombian Amazonia.
50
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
Table 4
Summary statistics for regressions of daily stemflow against gross
rainfall, in four forest ecosystems in the Colombian Amazonia
(Note: The power form is Ps ˆ c…Pdg † where c and d are the regression coefficients for stemflow (mm))
Landscape unit
Regression coefficient
c
d
se
R2
n
Sedimentary plain
High terrace
Low terrace
Flood plain
0.0015
0.0020
0.0029
0.0031
0.049
0.038
0.035
0.050
0.92
0.94
0.95
0.91
86
92
87
73
1.53
1.467
1.423
1.325
forest. The percentage of stemflow in all plots varied
from 0.2 to 3.2% of gross rainfall. The total average
percentage of stemflow relative to gross rainfall was
0.85% (^0.46) in the SP, 0.94 (^0.51) in the HT, 1.45
(^0.88) in the LT and 1.12 (^0.56) in the FP forest.
Differences are mainly due to the higher contribution
of tree palms to the total stemflow per plot. For
palms, high-capacity collectors (more than 35 l)
were required to measure the incoming water. In
subplots with abundant palms, these palms produced
about 43% of total stemflow.
Upon rainfall, in all forests stemflow increased
very gradually until a threshold of about 25 mm
gross rainfall is reached (Fig. 4). However, values
tend to scatter with increasing rainfall. The relationship between measured stemflow and gross
rainfall could be described with a power function
(Table 4). Some rainfall events smaller than 3 mm
did not produce stemflow in most plots, which
explains the lower number of events (n) reported
for stemflow regressions. For rainfall events with
an intensity ⬎5 mm h ⫺1, stemflow showed no clear
relationship with tree trunk area or bark texture
…R2 ˆ 0:3†: Nevertheless, there seems to be an
inverse relationship between crown area and the
amount of collected stemflow for each tree. We
also observed that lower parts of tree trunks with
fibrous bark texture were slowly wetted during
long storm events, which points to high water
storage.
4.5. Evaporation
Evaporation of intercepted water by the forest
Fig. 5. Evaporation against gross rainfall in the sedimentary plain forest ecosystem, Middle Caquetá, Colombian Amazonia.
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
51
Table 5
Statistics of evaporation from wet forest canopy in single rainfall events in four forest ecosystems in the Middle Caquetá, Colombian Amazonia.
Equations are of the form Ew ˆ e ⫹ ft ; (e in mm and f in mm h ⫺1) (Note: e and f are the regression coefficients for the linear function of
evaporation loss during rainfall events)
Forest
SP
HT
LT
FP
Total
rainfall
(mm)
3273.8
3293.0
3158.4
3120.9
Total
throughfall
(mm)
2853.7
2854.8
2711.7
2555.0
Total
stemf.
(mm)
32.4
36.2
38.6
30.5
time
(h)
557.0
464.2
487.8
472.3
Evap. during
rainfall (Ew)
mm
190.1
207.4
201.1
320.0
canopy is calculated by subtracting the measured
daily throughfall and stemflow from gross rainfall.
Furthermore, it is related to gross rainfall characteristics and forest system parameters. Following net
throughfall trends, the percentage of evaporation relative to gross rainfall varied from 6 to 100% in all
forests, depending on rainfall size. Mean evaporation,
expressed as percentage of total gross rainfall, also
differed between the forests: 11.84 (^2.4) in the SP,
12.24 (^1.2) in the HT, 12.92 (^1.1) in the LT and
17.15 (^0.96) in the FP forest. For small storms (less
than 2 mm), evaporation values were very close to
those of gross rainfall. For heavy showers, however,
the relative value of evaporation became smaller
(Fig. 5).
An assessment of the cumulative evaporation
during rainfall (Ew) was calculated with Eq. 1 for
the single rainfall events in each forest. Additionally,
Ew was related to rainfall duration. The average
evaporation rate during the rainfall varied from 0.34
to 0.68 mm h ⫺1 among the forests (Table 5) and Ew
exhibited a linear relation with rainfall duration for all
forests (Fig. 6). Negative values correspond to rainfall
events of short duration and low intensity, indicating
that the forest canopy did not reach saturated conditions during such events.
Though climatic conditions are similar in the
forests studied, there is a clear difference in amounts
of evaporation when comparing similar rainfall
events. This implies that amounts of evaporation
from the wet forest canopy did not only depend
upon gross rainfall and climate conditions. Differences in evaporation between these close-by forests
may be related to differences in their structure. For
Evap. rate
from wet
canopy
mm h ⫺1
0.342
0.447
0.412
0.677
Ew versus rainfall
duration. Regression
coeffficient
e
f
⫺0.424
⫺0.263
⫺0.351
⫺0.366
0.46
0.52
0.52
0.78
R2
n
178
0.86
0.75
0.82
0.88
178
160
163
140
that reason, the ratio of evaporation over gross rainfall
was plotted against the mean forest cover (1-gap
fraction) established for each forest. Fig. 7 indicates
that there is an increase of evaporation from the wet
forest canopy with increasing canopy cover. This
figure also shows an inflexion in the curve, i.e. a steeping from the low terrace to the flood plain, which
indicates that also other forest structural parameters
(e.g. leaf surface characteristics) affected canopy
interception.
5. Discussion
Storage capacity values of the forests studied on the
whole resemble the values found by Ubarana (1996)
in the reserves Vale do Rio Doce and Jaru Duke in
Brazil, which were based on linear regressions of
throughfall against gross rainfall. However, Ubarana
concluded that this method results in an overestimation of evaporative losses. Since our estimates of the
canopy storage capacity were based on specific events
for which it was assumed that evaporation was negligible, it might be that we somewhat overestimated the
storage capacity by neglecting evaporation during the
selected events. The storage capacity of the flood
plain forest is higher than the values commonly
reported in Amazonian rain forest studies. This can
be explained by the fact that most of the latter
studies were executed in so called “terra firme” forests
(non-flooded ecosystem). As we studied a broader
range of ecosystems, differences between parameters
values reported here and those reported in other
studies therefore should be interpreted in terms of
52
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
differences in forest structure between the ecosystems
studied.
The free throughfall coefficients of our forests,
derived from the data on selected small storms (pt),
conform to the values found by Jetten (1996).
However, these values which range from 0.27 to
0.59 clearly differ from those derived from the
scanned photographs (ps) and their application
resulted in an overestimation of net rainfall rates in
all forests. According to Ubarana (1996) this may be
explained by the waxy nature of the tree leaf surfaces
causing the drops to splash off, which thus contribute
to throughfall before the canopy is saturated. Additionally, our values were based on the results for
small showers producing throughfall and this may
have influenced the estimation of the free throughfall
because of the low frequency of such events (for the
research sites only 7–10 events were registered
during the total period). Photographs taken from
the forest canopy under non-direct sunlight and
with covered sky can easily be analysed, the white
pixels reflecting the non-covered part. This estimate
may provide better results, especially if large
numbers of photographs have been taken and
analysed.
The range in our values for throughfall and stemflow, expressed as percentage of gross rainfall, is
similar to the range in values reported in earlier
studies on rainfall partitioning in Amazonian rain
forests (Table 6). This most probably also explains
why such variability exist in the latter values, i.e. it
is probably largely due to differences in rainfall characteristics and forest structure between the forests
studied. That coefficients of determination for our
regressions are significantly higher than most values
presented in the literature can be explained by the size
of our data set, which is much larger than in earlier
studies. Our results also indicate that the partitioning
of rainfall depends, among others, on the size of the
rainfall event. Moreover, it is clear from the relation
between throughfall and storm size (Table 2) that the
high CV of throughfall is the result of the large variability in rainfall classes.
To define the total error (t.e.) of the mean throughfall as a percentage of gross rainfall, we applied the
proposed formula for random relocation of n gauges
Fig. 6. Evaporation from the wet forest canopy (Ew) in relation to rainfall duration, in a forest ecosystem (sedimentary plain) in the Middle
Caquetá, Colombian Amazonia.
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
53
Fig. 7. Fraction of evaporation from gross rainfall as a function of forest cover fraction (FC ˆ 1-gap fraction) in the four forest ecosystems.
Middle Caquetá, Colombian Amazonia.
by Lloyd and Marques (1988), although larger
diameter funnels were used in the current study
p
t:e: ˆ s:e:…1 ⫹ N=nm†
…2†
where s.e. is the standard error resulting from the
random relocations of collectors, expressed as the
best estimate of the standard error of mean throughfall
in each collector, under the assumption that the specific canopy structure is properly described by N
(number of grid points) and m (the relocation of
collectors). Based on the formula of Lloyd and
Marques (1988), the arrangement of 60 funnels with
23 relocations in the SP forest results in a total error in
measured throughfall of 3.5% of gross rainfall,
whereas in the other ecosystems with 40 gauges and
23 relocations the error is 3.8% due to variation in
canopy structure. These figures are lower than those
found by other authors, which can be explained by the
continuous relocation of our collectors. Nevertheless,
our values for t.e. are larger than those found by Lloyd
and Marques (1988).
Whether throughfall percentage depends on storm
size remains to be established, as clearly stated by
Lloyd and Marques (1988). Accordingly, we investigated the effect of storm size on the variability of
throughfall percentage by using only those single
events that were measured during the 20-months
period in which weekly data of throughfall and
gross rainfall were collected. During that period, the
same methodologies were used with the exception
that gauges were not relocated. Fig. 8, given as an
example, shows that the variation in the ratio of
throughfall from a single funnel over the average of
20 funnels tends to decrease as storm size increases.
This is a trend observed for most non-moving collectors but also for the relocated collectors, as stated
earlier in this paper, which suggests that storm size
also affect throughfall variability in our ecosystems.
We did not fully investigate the relation between
storm size and throughfall (expressed as percentage of
gross rainfall). Nevertheless, we conclude from our
results that when the method of relocation of collectors is used to estimate this throughfall, it is essential
54
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
Table 6
Partitioning of gross rainfall (percentages) in Amazonian rain forests
Location
Forest type
Throughfall %
Stemflow %
Evaporative loss %
Reference
Venezuela
Brazil
Brazil
Brazil
Colombia
Catinga
Rain forest
Rain forest
Rain forest
Rain forest
91
80.2
87–91
86–87
82–87
0.8–14
–
1.8
0.8–1.4
0.9–1.5
–
19.8
8.9(^3.6)
11.6–12.9(^5.9)
12–17
Herrera (1979)
Franken et al. (1992)
Lloyd and Marques (1988)
Ubarana (1996)
This study
that this relocation is preceded by sampling of a wide
range of storms sizes with a fixed set of collectors, in
order to assess the combined effect of site forest
structure and rainfall characteristics on throughfall.
Lastly, although litter fall in the ecosystems studied
exhibits some temporal dynamics (Duivenvoorden
and Lips, 1995), no relation was observed between
throughfall percentage and litter fall. Such relation
was reported for a Bornean rain forest by Burghouts
et al. (1988).
The variability of stemflow in mature tropical rain
forest has been attributed to the high species diversity
(Hutjes et al., 1990; Hertwitz, 1985) and this variability certainly is larger in tropical forests than in temperate forest plantations (Lloyd and Marques, 1988). In
the present study, this parameter was estimated for
different tree species with different diameter. Stemflow values from this study ranged from 0.9 to 1.5% of
gross rainfall, which is within the range of values
presented in other studies on similar forest types
(Table 6). Although the contribution of stemflow to
net rainfall was very low, it probably causes an important input of solutes to the forest floor, concentrated
around the base of trees. Results suggest that little
water was stored in excess of the storage capacity of
the stem elements, as indicated by the very small
stemflow quantities collected once rainfall has ceased
or during small storms. This can be explained by the
presence of some tree trunks with hydrophobic bark
(personal observations 1992–1997) and of bark with
fibrous texture. Upon rainfall, tree species with these
characteristics exhibited significant stemflow, even
without being completely wet. However, once rainfall
stopped, there was a sharp decline in stemflow.
We found static models to be capable of describing
rainfall partitioning for the forests studied. The applicability of these models is most probably restricted to
the area and conditions during the period of research.
Though the observed relationships may contribute
little to the explanation of the hydrological processes
at canopy level, the models nevertheless provide clear
indications for the extent to which this partitioning is
controlled by the parameters used. While linear functions produce better fits for correlation between
throughfall and gross rainfall, power functions
produce better fits for such correlation with stemflow,
in terms of the significance levels and standard deviation of residuals. The linear regression equations of
throughfall versus gross rainfall fit most points and
have a high coefficient of determination in all ecosystems. Nevertheless, their application to very small
storms (lower than 2 mm) results in negative values
for throughfall and they slightly underestimate
throughfall for very high-rainfall events, which illustrates the limitation of regressions, which fit a curve to
a set of data.
Throughfall and gross rainfall were highly
correlated in all forests, which is probably due to
the similarity of our forests with regard to relevant
system parameters. However, the correlation between
interception values and gross rainfall is less prominent, the coefficient of determination of the regression
being distinctly lower (R 2 between 0.66 and 0.83). In
other words, throughfall percentages can be predicted
with a high accuracy based on data on rainfall
amounts and characteristics, whereas for the prediction of interception other parameters, such as forest
structure, must be included.
Although our values are within the range of interception values reported in other studies from the
Amazon basin (Table 6), the values found for the FP
and LT forests are rather high compared to those
reported in earlier Amazonian studies (Lloyd and
Marques, 1988). This is in line with the higher canopy
storage capacity of our forests as compared to those
described in these other studies (e.g. Lloyd and
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
55
Fig. 8. The ratio of single gauge throughfall over average throughfall from 20 gauges, as a function of gross rainfall size.
Marques, 1988). Thus, even though climatic conditions for our forests were similar, forest interception
varied, due to differences in amounts and characteristics of the rainfall and in forest structure. The relative proportion of evaporation from our forests was
also higher than the values reported in these earlier
studies. This can be attributed to the higher gross rainfall in our forests, at least when compared with the
mean annual value of 2500 mm reported from central
Amazonia (Leopoldo et al., 1987).
In many studies on rainfall interception, it is
concluded that leaf surfaces determine the interception storage capacity of woody plants (Hertwitz, 1985;
Gash, 1979; Singh, 1977; Rutter et al., 1975). Studies
in the research plots, using destructive methods and
derived regression equations for leaf biomass estimation (Overman et al., 1990; Alvarez, 1993), showed
that leaf biomass is higher in the flood plain forest
(9.5 tonnes/ha) than in the other forests from which
a higher leaf surface area can be inferred. Therefore,
the relatively high interception by the flood plain
forest may be explained by its higher leaf biomass.
Cumulative evaporation during rainfall events (Ew)
from our forests strongly depended on rainfall dura-
tion. Although not evaluated in this study, it might
also depend on specific climatic conditions (e.g.
wind speed). Additionally, a distinct relationship
seemed to exist with the forest cover fraction
(Fig. 7). Although it should be realised that the
number of forests studied is small and the relationship
is rather uncertain, it may serve for the estimation
of evaporation by a forest for which measurements
are not available. Provided that climatic conditions
are similar, such estimations mainly rely upon an
adequate estimation of the gap fraction or LAI.
6. Conclusions
Of the gross rainfall of about 3400 mm y ⫺1, most
fell in small showers during the afternoon and at night.
The overall average rainfall intensity was about
5 mm h ⫺1. These rainfall characteristics largely
explain the partitioning of rainfall into throughfall,
stemflow and ensuing evaporation in the forests
studied.
Water fluxes in the forest canopy of four forest
ecosystems in western Amazonia have been quantified
56
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
as a percentage of gross rainfall. Amounts of net
precipitation reaching the forest floor and evaporation
from the wet forest canopy varied for the forests
studied: the SP forest had the highest percentage of
throughfall relative to gross rainfall and the FP forest
the lowest. The observed differences in throughfall,
stemflow and evaporation can partly be attributed to
differences in forest structure (gap fraction). Their
range is similar to the overall range in these parameters as published in earlier studies from the
Amazon basin, implying that the latter variability
may very well be connected with differences in forest
structure.
Results from the forests studied provide some
insight into the rates of evaporation from a wet forest
canopy and strengthen the understanding of the
contribution of forests to atmospheric moisture. The
mean evaporation rate from a wet forest canopy
during rainfall events in the Middle Caquetá, (Colombian Amazonia) was estimated at 0.47 mm h ⫺1 and it
increased with increasing forest cover. Moreover, this
study of throughfall, stemflow and evaporation in a
range of forests demonstrates the relevance of forest
structure for the evaporation of rainfall intercepted by
the forest canopy and for the net precipitation reaching the forest floor. The results show that within the
scope of this research forest structure can be
adequately characterised by the gap fraction and
LAI. These structural characteristics together with
the rainfall amount and rainfall duration are the
main parameters determining rainfall partitioning in
the western Amazonian rain forests.
Acknowledgements
We are grateful to Dr John Gash from the UK Institute of Hydrology and to Dr Sampurno Bruijnzeel
from the Free University, Amsterdam for their
suggestions and corrections of earlier drafts of this
paper. This work, forming part of a larger research
project on water and nutrient cycling in Colombian
Amazonia, was supported by the Tropenbos Foundation in Colombia and the Netherlands and by the
Colombian Institute for Science and Technology
“Colciencias”. Collected data (four years data on
gross rainfall, throughfall, stemflow and evaporative
loss from western Amazonia) is available on request
to the Tropenbos Foundation, E-mail: [email protected]
References
Alvarez, D.E., 1993. Composición florı́stica, diversidad, estructura
y biomasa de un bosque inundable, en la Amazonia Colombiana. Ir dissertation, Universidad de Antioquia, Colombia.
Bosch, J.M., Hewlett, J.D., 1982. A review of catchment experiments to determine the effect of vegetation changes on water
yield and evapotranspiration. J. Hydrol. 55, 3–23.
Bruijnzeel, L.A., 1990. Hydrology of moist tropical forests and
effects of conversion: a state of knowledge review. UNESCO,
Paris.
Burghouts, T.B.A., van Straalen, M.N., Bruijnzeel, L.A., 1998.
Spatial heterogeneity of elements and litter turnover in a
borneam rain forest. J. Trop. Ecol. 14, 477–506.
Duivenvoorden, J.F., 1995. Tree species composition and rain
forest-environment relationships in the Middle Caquetá area,
Colombia, NW Amazonia. Vegetatio 120, 1–113.
Duivenvoorden, J.M., Lips, J.M., 1995. A land-ecological study of
soils, vegetation and plant diversity in Colombian Amazonia.
PhD thesis, Landscape and Environmental Research Group,
Faculty of Environmental sciences, University of Amsterdam
Tropenbos Foundation, Wageningen.
Eltahir, E.A., Bras, R.L., 1993. On the response of the tropical
atmosphere to large-scale deforestation. Q.J.R. Meteorol. Soc.
119, 779–793.
Ford, E.D., Deans, J.D., 1978. The effects of canopy structure on
stemflow, throughfall and interception loss in a young Sitka
spruce plantation. J. Appl. Ecol. 15, 905–917.
Franken, W., Leopoldo, P.R., Matsui, E., Ribeiro, M. 1992. Estudo
da Interceptaçâo da água de chuva em cobertura florestal
amazônica do tipo terra firme. Acta Amazonica, 12, 327–331.
Gash, J.H.C., Nobre, C.A., Roberts, J.M., Victoria, R.L. (Eds.),
1996. Amazonian Deforestation and Climate Wiley, Chichester,
UK (611pp.).
Gash, J.H.C., 1979. An analytical model of rainfall interception by
forests. Q.J.R. Meteorol. Soc. 105, 43–55.
Gash, J.H.C., Morton, A.J., 1978. An application of the Rutter
model to the estimation of the interception loss from Thetford
forest. J. Hydrol. 38, 49–58.
Herrera, R.A., 1979. Nutrient distribution and cycling in an Amazon
caatinga forest on Spodosols in southern Venezuela,
Dissertation, University of Reading.
Hertwitz, S.R., 1985. Interception storage capacities of tropical rain
forest canopy trees. J. Hydrol. 77, 237–252.
Hölscher, D., de A Sa, T., Bastos, T., Denich, M., Fölster, H., 1997.
Evaporation from young secondary vegetation in eastern
Amazonia. J. Hydrol. 193, 293–305.
Hutjes, R.W.A., Wierda, A., Veen, A.W.L., 1990. Rainfall
interception in the Tai forest, Ivory Coast: application of two
simulation models to a humid tropical system. J. Hydrol. 114,
259–275.
Jetten, V.G., 1996. Interception of tropical rain forest: performance
of a canopy water balance model. Hydrol. Process 10, 671–685.
C. Tobón Marin et al. / Journal of Hydrology 237 (2000) 40–57
Jordan, C.F., 1978. Stemflow and nutrient transfer in a tropical rain
forest. Oikos 31, 257–263.
Klaassen, W., Bosveld, F.C., de Water, E., 1998. Water storage and
evaporation as constituents of rainfall interception. J. Hydrol.
212–213, 35–50.
Klinge, H., Rodriguez, W.A, Brunig, E.F., Fittkau, E.J., 1975.
Biomass and structure in a central Amazonian Rain Forest. In:
Golley, F.B., Medina, E. (Eds.). Tropical Ecological Systems.
Trends in Terrestrial and Aquatic Research, Springer, Berlin,
pp. 115–122.
Köppen, W., 1936. Das geographische System der Klimate.
Handbuch der Klimatologie. Verlag Gebrüder Borntraeger,
Berlin (44pp.).
Leopoldo, P.R., Franken, W., Villa Nova, N., 1995. Real evaporation and transpiration through a tropical rain forest in central
Amazonia as estimated by the water balance method. For. Ecol.
Manag. 73, 185–195.
Leopoldo, P.R., Franken, W., Salati, E., Ribeiro, M.N., 1987.
Towards a water balance in central Amazonia region.
Experentia 43 Birkhäuser Verlag, CH-4010 Basel/Switzerland.
Lesak, L.F.W., 1993. Water balance and hydrologic characteristics
of a rain forest catchment in the Central Amazon Basin. Water
Resour. Res. 29 (3), 759–777.
Lettau, H., Lettau, K., Molion, L.C.B., 1979. Amazonian’s Hydrology cycle and the role of atmospheric recycling in assessing
deforestation effects. Mon. Weather Rev. 107, 227–238.
Londoño, V.C., 1993. Análisis estructural de dos bosques asociados
a unidades fisiográficas contrastantes, en la region de
Araracuara (Amazonia Colombiana). Ir Dissertation,
Universidad Nacional de Colombia.
Longman, K.A., Jenı́k, J., 1990. Tropical Forest and its
Environment. Longman, New York.
Lloyd, C.R., Marques, F.A.de.O., 1988. Spatial variability of
throughfall and stemflow measurements in Amazonian rain
forests. Agri. For. Meteorol. 42, 63–73.
Molion, L.C.B., 1975. A climatonomic study of the energy and
moisture fluxes of the Amazon’s basin with considerations of
deforestation effects. PhD thesis, University of Wisconsin,
Madison.
Nobre, C.A., Sellers, P., Shukla, J., 1991. Amazonia deforestation
and regional climate change. J. Climate 4, 957–988.
Nobre, C.A., Coauthors, 1996. The Large-scale Biosphere-Atmosphere Experiment in Amazonia (LBA). Concise experiment
plan; LBA Science Planning Group, Staring Center-DLO, The
Netherlands.
Overman, J.P.M., Saldarriaga, J.G., Duivenvoorden, J.F., 1990.
57
Estimación de la biomasa aérea en el bosque del medio Caquetá,
Colombia. Colombia Amazonica 4 (2), 135–147.
Roberts, J.M., Cabral, O.M.R., da Costa, J.P., McWilliam, A.L.C.,
de A. SÁ, T.D., 1996. An overview of the leaf area index and
physiological measurements during ABRACOS. Amazonian
Deforestation and Climate. Wiley, Chichester, UK (611pp.).
Rutter, A.J., Morton, A.J., Robins, P.C., 1975. A predictive model of
rainfall interception in forests. II Generalisation of the model
and comparison with observations in some coniferous and
hardwood stands. J. Appl. Ecol. 12, 367–380.
Salati, E., Vose, P.B., 1984. Amazon Basin: a system in
equilibrium. Science 225, 129–138.
Salati, E., Dall’olio, A., Matsui, E., Gat, J.R., 1979. Recycling of
water in the Amazon Basin: an isotopic study. Water Resources
Res. 15, 67–71.
Shuttleworth, W.J., Calder, I.R., 1979. Has the Priestley–Taylor
equation any relevance to forest evaporation? J. Appl. Meteorol.
18, 639–646.
Shuttleworth, W.J. 1988. Evaporation from Amazonian rain forest.
Proc. R. Soc., London, B233, 321–346.
Singh, B., 1977. The effect of rainfall characteristics and postwetting synoptic characteristics on evaporation rates from a
wetted hardwood canopy. Climatol. Bull. 21, 12–33.
Stewart, J.B., 1977. Evaporation from the wet canopy of a pine
forest. Water Resour. Res. 13, 915–921.
Tanner, E.V.J., 1985. Jamaican Montane forests: Nutrient capital
and coast growth. J. Ecol. 75, 553–568.
Ter Stegee, H., (1994). HEMIPHOT: a programme to analyse vegetation indices, light and light quality from hemispherical photographs. Tropenbos Documents 3. The Tropenbos Foundation,
Wageningen, The Netherlands.
Ubarana, V.N., 1996. Observations and modelling of rainfall
interception at two experimental sites in Amazonia. In: Gash,
J.H.C., Nobre, C.A., Roberts, J.M., Victoria, R.L. (Eds.).
Amazonian Deforestation and Climate, Wiley, Chichester, UK.
Vitousek, P.M., Denslow, J.S., 1986. Nitrogen and phosphorous
availability in treefall gaps of a lowland tropical rain forest.
J. Ecol. 74, 1157–1178.
Waidi, S., Wong, W.M., Douglas, I., 1992. Throughfall, stemflow,
overland flow and throughflow in the Ulu Segama rain forest,
Sabah Malaysia. Phil. Trans. R. Soc. London. B. 335, 389–395.
Wright, I.R., Gash, J.H., da Rocha, H.R., Shuttleworth, W.J., Nobre,
C.A., Maitelli, G.T., Zamparoni, C.A., Carvalho, P., 1992. Dry
season micrometeorology of central Amazonian ranchland.
Q.J.R. Meteorol. Soc. 118, 1083–1099.