376 A. Anandacoomaraswamy et al. Agricultural and Forest Meteorology 103 2000 375–386
High transpiration rates from extensive tea canopies cause significant soil water deficits which are respon-
sible for decreased leaf expansion rates Monteith and Elston, 1983; Squire, 1990; Stephens and Carr, 1993.
Even when the soil is wet, the excessive transpiration rates resulting from higher levels of irradiance and sat-
uration deficits around mid-day could cause transient water deficits within the plant Kramer, 1988; Smith
et al., 1994. Transpiration is closely linked to photo- synthesis which is the primary physiological process
responsible for growth of young leaves which form the economic yield of tea. De Wit 1958 showed that plant
biomass production is directly proportional to transpi- ration. The proportionality constant has been termed
transpiration efficiency Tanner and Sinclair, 1983 or dry matter:water ratio Monteith, 1986; Squire, 1990.
The saturation vapour pressure deficit D of the air exerts a significant influence on transpiration effi-
ciency T
E
of a crop by controlling the water vapour pressure gradient between the leaf sub-stomatal cham-
ber and the outside air Jones, 1992. For a given level of leaf conductance, a greater leaf-air vapour
pressure deficit causes a higher transpiration rate and consequently decreases T
E
. It has been shown that the product between transpiration efficiency T
E
and D is a constant for a given crop species Bierhuizen and
Slatyer, 1965; Monteith, 1986. On the other hand, D could exert a direct effect on stomata of many plant
species by decreasing stomatal conductance with in- creasing D Jarvis and Morrison, 1981. Because of
the varying degree of transpiration control by stom- atal conductance depending on the degree of coupling
between canopy and air Jarvis and McNaughton, 1986, the overall effect of D on transpiration and T
E
may vary with canopy characteristics and atmospheric conditions.
Despite the above complications, for most practical purposes, an estimation of T
E
and transpiration would enable the prediction of tea yields in a given environ-
ment. The main objectives of this experiment were to measure transpiration from a mature tea canopy and to
determine its controlling factors and the relationship with yield as estimated by T
E
. Tea is considered a shade-loving plant Squire and
Callander, 1981 and is normally grown as a mixture with trees which provide a natural shade. As solar
irradiance is the primary source of energy for tran- spiration, shading could reduce the transpiration rate
of tea growing under shade. Hence, the effect of ar- tificial shade, which simulated the different levels of
natural shade, on transpiration of tea was investigated in this study.
There have been several previous studies on water use of tea Dagg, 1970; Willat, 1971, 1973; Carr, 1974,
1985; Cooper, 1979; Callander and Woodhead, 1981; Stephens and Carr, 1991. However, in all these stud-
ies, evapotranspiration which includes both transpira- tion and soil evaporation, had been measured. These
measurements of water use include a variation com- ponent which does not directly contribute to the in-
ternal functioning of the plant as soil evaporation can vary depending on the degree of ground cover and top
0–5 cm depth soil water availability Ritchie, 1972. Therefore, a measurement of transpiration from tea
canopies is needed to predict tea yields accurately. In the present experiment, the heat pulse technique was
used to measure the transpiration from tea plants.
2. Materials and methods
The experiment was carried out at Talawakelle, St. Coomb’s estate of the Tea Research Institute of Sri
Lanka latitude, 6
◦
4
′
N; longitude, 80
◦
40
′
; altitude, 1380 m a.s.l.. The soil is a fine, mixed thermic Trop-
udult Panabokke, 1996. The field capacity was 44 vv and the permanent wilting point of 26 vv.
These were determined using the pressure plate ap- paratus. Field capacity and permanent wilting point
were defined as the soil water contents S at matric potentials of −0.01 and −1.5 MPa, respectively. The
soil available water content S
a
was defined as the difference between S at field capacity and permanent
wilting point. The soil profile contains about 200 mm of available water in its top 1 m.
Mature field-grown tea plants of the clone TRI2025 were used for measurements. Tea plants were grown
with shade trees of Grevillea robusta spaced at 12 m×12 m. Experimental plots were 8 m×6 m and
the tea plants were at a spacing of 1.2 m×0.6 m. The canopy spread of a Grevillea tree was around 5–6 m
in diameter whereas the canopy diameter of a tea plant was 1–1.2 m. However, the tea plants formed
a continuous, smooth canopy at 0.8–1 m height as they were planted at a closer spacing. In contrast,
the widely-planted Grevillea formed a discontinuous
A. Anandacoomaraswamy et al. Agricultural and Forest Meteorology 103 2000 375–386 377
canopy at a height of around 12–15 m. Excavation studies have shown that clonal tea has a maximum
rooting depth of around 0.9–1 m, but more than 90 of the roots are located within the top 0.6 m of the
soil profile Anandacoomaraswamy, unpublished. In contrast, Grevillea has a deeper root system with the
maximum rooting depth exceeding 3 m. The lateral spread of the tea root system is over an area of 1.2 m
in diameter. In each plot, 10 plants with round stem bases and a stem diameter of approximately 55 mm
were selected for transpiration measurement.
2.1. Measurement of transpiration Transpiration was measured by the heat pulse tech-
nique as described by Cohen et al. 1981. It is based on measuring the time required for a heat pulse applied
at a given point of the stem to be transferred to a given point downstream. This gives a measurement of the
rate of sap flow in the xylem. The flux density of sap is then calculated using flow geometry, cross-sectional
area and flow velocity Swanson, 1994. The measur- ing instrumentation consisted of a line heating needle
resistance, 38 , a thin rod 30 mm in length con- taining three thermocouples along its length at 5, 15
and 25 mm from the tip Thermalogic, USA and a data logger for recording the thermocouple output sig-
nals. The heating needles and the thermocouples were inserted in to the stem base of each of the selected
plants. Before insertion, the bark was trimmed away and holes were drilled using a drill guide template.
The thermocouple rod was inserted so that the three temperature sensors were placed at radial distances of
5, 15 and 25 mm beneath the outer surface of the stem. Diameter of the stem bases after trimming away the
bark was 50 mm. The inserted thermocouples were at a distance of 5 mm downstream of the heating nee-
dle. After installation of the heater and the thermocou- ples, the stems were covered with insulating material
to prevent radial heat loss to the outside. The heater was powered by a 12 V battery. The heater and ther-
mocouples were connected to a 21X Campbell data logger Campbell Scientific Inc.. The heat pulse was
switched on and off every second by a control port on the data logger and the duration of the heat pulse was
controlled by the data logger through a programmed counter as described in Ishida et al. 1991. The data
logger was programmed to record the rise in temper- ature of the thermocouples every second and the data
were averaged and stored at 30 min time intervals. The single-sensor heat pulse method used in the present
study is appropriate for relatively high sap speeds whereas the up- and down-stream dual-sensor method
is appropriate for slower sap speeds Swanson, 1994.
2.2. Calculation of volumetric water flux The volumetric water flux, J m
3
h
− 1
, was the sum of fluxes in different radial rings of the stem as mea-
sured from thermocouples inserted at different radial depths:
J = J
1
A
1
+ J
2
A
2
+ J
3
A
3
1 where A
1
, A
2
and A
3
are the respective stem ring areas from which flux densities J
1
, J
2
and J
3
are measured. The respective stem ring areas were calculated as
A
1
= π
[R
2 x
− R
x
− 0.5
2
], A
2
= π
[R
x
− 0.5
2
− R
x
− 1.5
2
], A
3
= π
[R
x
− 1.5
2
− R
x
− 2.5
2
] 2
where R
x
is the xylem radius as obtained by measur- ing the circumference C of the stem after removing
the bark R
x
= C2π . The flux density J
i
through a given area A
i
was calculated following the procedure described by Cohen et al. 1981. This is based on the
equation of Marshall 1958 to describe the temper- ature rise detected at a distance r from a line heater
after a time t following a heat pulse
T = H
4πρcκt exp
− x −
Vt
2
+ y
2
4κt 3
where H is the heat output per unit length of the heater, ρ
, c and κ the density, specific heat and thermal dif- fusivity of wet wood, respectively, x and y distances
related to the distance r between the line heater and the thermocouples as r=x
2
+ y
2 12
. V is the convec- tive velocity of the heat pulse.
The temperature rise T reaches a maximum at time t
m
after the heat pulse. At this point when dTdt=0, Eq. 3 gives an expression for the convective velocity
of the heat pulse as V =
r
2
− 4κt
m 12
t
m
4
378 A. Anandacoomaraswamy et al. Agricultural and Forest Meteorology 103 2000 375–386
When t
m0
is the time required to maximize T at V=0, κ can be given as
κ = r
2
4t
m
5 By combining Eqs. 4 and 5, V can be computed
as V =
r 1 − t
m
t
m0 12
t
m
6 Here, t
m0
was determined by temperature rise data collected at night during which the sap flow rates were
assumed to be zero. J
i
can be given as: J
i
= V
ρc ρ
l
c
l
7 where ρc is the volumetric specific heat of tea wood
3.2×10
6
J m
− 3
K
− 1
as determined previously by Anandacoomaraswamy, unpublished and ρ
l
c
l
is the volumetric specific heat of water 4.18 J m
− 3
K
− 1
. Total flux of water was then calculated from Eq. 1.
Transpiration rates of three trees of G. robusta grow- ing in the same field were measured simultaneously
with tea using the same method.
The transpiration measurements obtained from the heat pulse method, were of the same order of magni-
tude as the evapotranspiration values estimated by the soil water balance method during the corresponding
period of the previous year Anandacoomaraswamy, unpublished. As the tea canopy in the measured fields,
covered the soil completely, soil evaporation could be assumed to be negligible. Hence, the above evapotran-
spiration measurements could be safely assumed to reflect the transpiration levels.
2.3. Experimental treatments To examine the factors influencing the transpiration
rate of tea, three treatments were imposed on the ex- perimental plots at different times. To determine the
influence of irradiance, four different levels of shad- ing 25, 60, 80 and 85 were imposed on the plots
by covering them with varying layers of shade net- ting. The whole plot area of 8 m×6 m, which included
about 66 tea plants, was covered. A control plot was kept unshaded. One week after the imposition
of shading treatments, transpiration was measured simultaneously over a period of 3 days at different
shade levels and the control. During these measure- ments, all plots were irrigated to eliminate water
stress.
To examine the influence of soil water availabil- ity, some of the plots were irrigated to field capacity
on the day prior to the commencement of the tran- spiration measurements. Soil water content at a depth
of 15 cm was measured gravimetrically on a weekly basis.
To examine the effect of a change in canopy re- flectance on transpiration, a 5 solution of Kaolin
was sprayed on the canopy of some of the plots. Canopy temperature in the Kaolin-sprayed and control
plots was measured using an infra-red thermometer HORIBA, IT-330 simultaneously with the transpi-
ration measurements.
Continuous measurements of incident solar radia- tion, air temperature, relative humidity and wind speed
at 1.2 m above the soil surface were made by Kipp so- larimeter, thermometer, hygrometer and anemometer,
respectively, in an automatic weather station Camp- bell Scientific Inc. located in the experimental field.
Measurements were recorded at 5 min time intervals, integrated or averaged over 30 min intervals and stored
in a data logger. Saturation vapour pressure deficit of the air D was computed from relative humidity
measurements as the difference between the satura- tion vapour pressure at air temperature and the actual
vapour pressure.
2.4. Tea leaf yield Leaf yield of tea was measured by plucking all
the new shoots emerging above the canopy surface at 7-day intervals. Dry weight of leaf yield was measured
after oven-drying at 105
◦
C to a constant weight usu- ally over a period of 12 h. Transpiration efficiency T
E
for leaf yield in terms of made tea was estimated by a linear regression between harvested leaf dry weight
and transpiration. Value of the product between T
E
and D was estimated by a linear regression between estimated total plant dry weight and the ratio between
transpiration and D. Total dry weight was estimated as the ratio between the measured leaf dry weight and a
harvest index of 0.10 Magambo and Cannell, 1981. All measurements were made between 1 January and
14 February 1997.
A. Anandacoomaraswamy et al. Agricultural and Forest Meteorology 103 2000 375–386 379
Table 1 Meteorological conditions during the experimental period from 1 January to 13 February 1999
a
Day of Mean solar irradiance
Mean maximum Mean minimum
Mean vapour pressure the year
MJ m
− 2
per day temperature
◦
C temperature
◦
C deficit kPa
0–7 22.26
23.8 9.9
0.52 8–14
20.51 24.7
12.0 0.62
15–21 23.63
24.4 9.7
0.81 22–28
24.11 24.5
7.7 0.88
29–35 22.91
22.9 10.3
0.82 36–42
21.37 21.4
8.6 0.85
a
Means of temperatures and vapour pressure deficits were measured at 1.2 m above soil surface and are 24 h mean values. No rainfall occurred during the experimental period.
3. Results
