CSIRO PUBLISHING
www.publish.csiro.au/journals/ajb
Australian Journal of Botany, 2006, 54, 181–192
Impacts of tree plantations on groundwater in south-eastern Australia
Richard G. BenyonA,C , S. TheiveyanathanB and Tanya M. DoodyA
A Ensis,
PO Box 946, Mount Gambier, SA 5290, Australia.
PO Box E4008, Kingston, ACT 2604, Australia.
C Corresponding author. Email: [email protected]
B Ensis,
Abstract. In some regions dependent on groundwater, such as the lower south-east of South Australia in the Green
Triangle, deep-rooted, woody vegetation might have undesirable hydrological impacts by competing for finite,
good-quality groundwater resources. In other regions, such as the Riverina in south-central New South Wales,
where rising watertables and associated salinisation is threatening the viability of agriculture, woody vegetation
might have beneficial hydrological impacts. In response to a growing need to better understand the impacts of tree
plantations on groundwater, annual evapotranspiration and transpiration were measured at 21 plantation sites in
the Green Triangle and the Riverina. Sources of tree water uptake from rainfall and groundwater were determined
by measurements of evapotranspiration and soil water over periods of 2–5 years. In the Green Triangle, under a
combination of permeable soil over groundwater of low salinity (<2000 mg L−1 ) at 6-m depth or less, in a highly
transmissive aquifer, annual evapotranspiration at eight research sites in Pinus radiata D.Don and Eucalyptus
globulus Labill. plantations averaged 1090 mm year−1 (range 847–1343 mm year−1 ), compared with mean annual
precipitation of 630 mm year−1 . These plantation sites used groundwater at a mean annual rate of 435 mm year−1
(range 108–670 mm year−1 ). At eight other plantation sites that had greater depth to the watertable or a rootimpeding layer, annual evapotranspiration was equal to, or slightly less than, annual rainfall (mean 623 mm year−1 ,
range 540–795 mm year−1 ). In the Riverina, where groundwater was always present within 3 m of the surface,
Eucalyptus grandis Hill ex Maiden trees at three sites with medium or heavy clay, alkaline, sodic, saline subsoils
used little or no groundwater, whereas E. grandis and Corymbia maculata (Hook.) K.D.Hill and L.A.S.Johnson
trees at a site with a neutral sandy soil and groundwater of low salinity used 380 and 730 mm year−1 of groundwater
(respectively 41 and 53% of total annual evapotranspiration). We conclude that commonly grown Eucalyptus species
and P. radiata are able to use groundwater under a combination of light- or medium-textured soil and shallow depth
to a low-salinity watertable.
Introduction
Australia is a dry continent with limited water resources.
Sustainable management of these resources requires a
sound understanding of the availability of both surface and
groundwater and how this is influenced by climate, soil and
vegetation cover. In contrast, in some parts of the country, an
increase in the amount of water in the landscape because of
irrigation and clearing of deep-rooted perennial vegetation
has caused salinisation of soils, threatening the health of
remnant natural vegetation and aquatic ecosystems and the
viability of agriculture.
Hydrological studies from across the world demonstrate
that woody vegetation, including tree plantations, commonly
uses more water annually than grasses and non-irrigated
agricultural crops (Hibbert 1967; Bosch and Hewlett 1982).
It is generally accepted that forests evapotranspire more
water than shallow-rooted vegetation because they have
lower albedo, intercept more rainfall (particularly conifers),
© CSIRO 2006
possess deeper root systems and have taller, rougher canopies
(Holmes and Sinclair 1986; Zhang et al. 1999).
In Australia, the area under industrial-plantation forestry
has increased by ∼6000 km2 in the past decade. As a
proportion of the total area of agricultural land, this is a small
change, but in a few individual catchments, new plantation
forestry represents a significant change in land use. This has
stimulated interest and debate over the potential for both
adverse and beneficial impacts on water resources. In areas
threatened by salinity, there is strong interest in determining
the effectiveness of strategically locating trees to de-water
landscapes where saline watertables have been rising. These
issues have been summarised in Nambiar and Brown (2001)
and O’Loughlin and Nambiar (2001).
Trees can reduce stream-flow and groundwater recharge,
relative to shallower-rooted grasses and agricultural
crops, and in areas underlain by shallow watertables, the
deeper root systems of trees than herbaceous vegetation
10.1071/BT05046
0067-1924/06/020181
182
Australian Journal of Botany
might enable them to access groundwater (Knight 1999).
Several studies have quantified groundwater use by woody
perennials, including natural riparian vegetation in various
environments (Thorburn and Walker 1994; O’Grady et al.
2002, 2006), natural Savanna (O’Grady et al. 1999; Kelly
et al. 2002) and small woodlots established over shallow
groundwater (Thorburn 1999; Cramer et al. 1999; Morris
and Collopy 1999; Vertessy et al. 2000). In some cases,
natural vegetation, including eucalypts, has been shown to be
reliant almost solely on groundwater. For example, Thorburn
et al. (1993) observed that among various sites on the flood
plain of the Murray River in South Australia, groundwater
contributed between 40 and 100% of tree water use. To our
knowledge, however, quantification of groundwater uptake
by large-scale tree plantations established in locations
with shallow depth-to-groundwater has not previously
been undertaken.
In the Green Triangle region of south-eastern South
Australia and south-western Victoria and the Riverina in
south-central New South Wales, tree plantations have been
established over shallow watertables in recent years. In
both regions, impacts of tree plantations on groundwater
need to be understood and quantified to enable assessment
of potential beneficial or adverse effects on groundwater
supplies, remnant natural vegetation and groundwaterdependent ecosystems.
In the western half of the Green Triangle, mainly in
south-eastern South Australia, centred on Mount Gambier
(37.8◦ S, 140.8◦ E), 95% of available groundwater resources
occur in a transmissive, unconfined or semi-confined aquifer
which is commonly <20 m below ground level and in some
places a few metres from the surface. Generally, there is
little surface run-off and few permanent natural surface
streams and lakes. Most of the rainfall not returned to the
atmosphere via evapotranspiration drains vertically until it
enters the unconfined aquifer as recharge (Hopton et al.
2001; South East Catchment Water Management Board
2003). Pinus radiata D.Don has been grown for timber
production in this region for more than a century and
currently occupies ∼1000 km2 . In addition, a new industry
to supply eucalypt chips for export has been established
since the late 1990s, based on Eucalyptus globulus Labill
which now occupies ∼350 km2 in south-eastern South
Australia. Plantation forestry provides the raw material for
nearly 30% of regional economic activity and 25% of
regional employment (Green Triangle Regional Plantation
Committee 2001).
Prior to European settlement in the 19th century, the region
had a mixture of forest, woodland and wetland vegetation. In
the past 150 years, 93% of this has been cleared and replaced
with pasture, agricultural crops and plantation forests. Early
studies in south-eastern South Australia indicated lower
groundwater recharge under P. radiata plantations than with
grassland (Holmes and Colville 1970a, 1970b; Allison and
R. G. Benyon et al.
Hughes 1972; Colville and Holmes 1972). These studies
found that, once the canopy of a plantation has fully
closed, there is little, if any, groundwater recharge under
these plantations. However, none of these early studies
indicates whether deep-rooted vegetation, such as trees,
obtains water from groundwater (Benyon 2002). Sustainable
management of the region’s water resources requires accurate
assessment of water use by the main land uses, including
plantation forestry.
The Riverina, centred around Deniliquin in southcentral New South Wales (35.5◦ S, 145.0◦ E) is underlain
by the Murray Basin, made up of various layers of
sediment containing relatively thin aquifers (0.5–6 m thick)
and groundwater of varying salinity from almost fresh
(<1000 mg L−1 ) to highly saline (>10 000 mg L−1 ). Soils
are generally loam at the surface, underlain by subsoils
of various textures from coarse sand to heavy clay many
metres deep. Groundwater levels over the past few decades
have risen owing to accession from irrigation channels and
flood-irrigated crops, which occupy much of the region.
Rising saline watertables pose a threat to the viability of
irrigated agriculture in some locations. Small farm-forestry
plantations, mainly of Eucalyptus grandis Hill ex Maiden and
Corymbia maculata (Hook.) K.D.Hill & L.A.S.Johnson, have
been established over the past 10 years. The question here is
whether these woodlots have potential to use groundwater
in some locations and help arrest the rise in watertables
(Polglase et al. 2002).
Although the land and water-resource management issues
differ between the two regions, the fundamental research
questions are the same, including the following:
(1) do tree plantations use all the rainfall (is there any
groundwater recharge), and
(2) can the trees access shallow groundwater; is there net
aquifer discharge through direct uptake by tree roots; if
so, how much groundwater can tree plantations use and
what site factors determine the amount of groundwater
uptake?
Over the past 10 years in the Green Triangle and
south-central New South Wales, water balance and tree
water-use studies have been used to quantify plantation
water use in research plots of P. radiata, E. globulus,
E. grandis and C. maculata. These studies provide key
information on plantation water use for these major
plantation species and how water use varies with rainfall,
potential evapotranspiration, depth-to-groundwater, soil
conditions and easily measurable plantation characteristics
such as stem growth rates and leaf area index. This paper
uses the results from these studies to identify site factors
influencing the impacts of plantations on groundwater in
the Green Triangle and the Riverina and to assess whether
plantation growth rates can be used as an indicator of
evapotranspiration.
Plantations and groundwater
Australian Journal of Botany
Materials and methods
Quantifying groundwater recharge and uptake
Various techniques have been used to quantify groundwater use by plants
in previous studies. Some have used measurements of changes in depthto-watertable in comparison with measured evapotranspiration rates to
indicate groundwater uptake (Dugas et al. 1990; Farrington et al. 1990).
This technique does not work at locations where aquifer transmissivity
is high relative to evapotranspiration rates. Aquifer transmissivity in
south-eastern South Australia is high (F. Stadter, pers. comm.). Others
have used measurements of concentrations of stable isotopes of O and
H of groundwater, soil water and plant water to identify sources of water
used by plants (Thorburn et al. 1993). However, in south-eastern South
Australia, the isotope concentrations in deep soil water and groundwater
are not sufficiently different to permit the use of this technique (McEwen
and Leaney 1996).
At locations where there is negligible surface water flow or
subsurface lateral flow within the vadose zone, groundwater uptake can
be estimated by comparing measured evapotranspiration with rainfall
183
inputs and changes in root-zone soil water. A statistically significantly
higher evapotranspiration rate than the sum of rainfall and any reduction
in soil water over a period, allows quantification of groundwater uptake.
We used this method, i.e. components of the water balance were
measured in experimental plots ranging in size from 20 × 20 m to
30 × 30 m, at various sites in the Green Triangle and the Riverina.
Sources and amounts of water used by the trees were estimated by
comparing evapotranspiration with rainfall and changes in soil water.
In the Riverina, the plantations were occasionally irrigated. Additional
water supplied by irrigation was accounted for. For each site, tree water
use and some or all of the other components of evapotranspiration were
measured.
At all sites, measurements of water use began at or after canopy
closure and continued for between 2 and 5 years. Sites described
in Table 1 include E. globulus (labelled ‘EGl’), P. radiata (‘PR’),
E. grandis (‘EGr’) and C. maculata (‘CM’). Plantation ages at
commencement of measurement ranged from 3 to 6 years in the
eucalypts and from 4 to 27 years in P. radiata. All sites had little or
no understorey.
Table 1. Characteristics of the research sites
Column 1 indicates the site number, as referred to in the text, and the species, as defined in the text. Column 2 indicates the plantation age, to the
nearest year, at the commencement and conclusion of measurements. Column 3 indicates the median depth to the watertable over the study period
and the watertable depth range (in parentheses). Sites highlighted in bold used some groundwater (see results). n.m. indicates not measured
Site/
species
Age range
(years)
Median watertable
depth (m)A
1. EGl
4–6
1.7 (0.4–2.7)
Green Triangle sites
Extensive flats
2. EGl
6–9
1.7 (0.6–3.2)
Extensive flats
3. EGl
6–9
1.9 (0.8–3.4)
Extensive flats
4. EGl
5–10
3.0 (2.7–3.3)
5. EGl
4–7
3.2 (2.4–3.9)
6. EGl
7. EGl
8. EGl
9. EGl
10. PR
11. PR
12. PR
13. PR
4–7
4–8
4–6
4–7
4–7
4–9
27–31
14–18
10.3 (9.8–11.0)
4.4 (3.7–4.7)
15.9 (15.4–16)
7.8 (7.3–8.3)
3.9 (3.5–4.4)
6.0 (5.4–6.4)
4.4 (4.1–5.2)
8.9 (8.7–9.3)
Lower slope hollow
among dunes
Lower slope hollow
among dunes
Top of dune
Lower mid-slope
Mid-slope
Mid-slope
Lower slope
Lower slope
Mid-slope
Mid-slope
14. PR
15. PR
16. PR
20–24
25–29
21–25
8.6 (8.5–8.9)
23.0 (22.9–23.3)
n.m.
Mid-slope
Lower slope
Mid-slope
17. EGr
18. EGr
19. EGr
20. EGr
3–5
3–5
3–5
3–5
2.8 (2.5–3.0)
2.9 (2.8–3.0)
2.9 (2.8–3.0)
2.7 (2.5–2.8)
21. CM
3–5
3.1 (2.8–3.5)
A
Slope position
Riverina sites
Extensive flats
Extensive flats
Extensive flats
Extensive flats –
prior stream
Extensive flats –
prior stream
Soil description
Stocking
(trees ha−1 )
Sandy to 0.4 m, medium
sandy clay to 2.5 m
Sandy to 0.9 m, sandy clay to 1.7 m,
medium to heavy clay to 2.5 m,
sandy clay to 4 m
Sandy to 1.2 m, sandy light
to medium clay to 4 m
Sand with calcified sand at 1 to
1.5 m, sand to 3 m
Sand and loamy sand to 1.6 m,
light to medium sandy clay to 4 m
Sand to 2.5 m, sandy light clay to 3 m
Sand to 0.4 m, sandy clay to >3 m
Sand to 1.2 m, sandy clay to >3 m
Medium to heavy clay to >5 m
Sand 0.6 m, sandy clay to >3 m
Sand to 1.5 m, sandy clay to >3 m
Sand to 1.2 m, cemented hardpan at 1.2 m
Sand to 0.6 m, light to
medium sandy clay to >5 m
Sand to 2 m, sandy clay to >3 m
Sand to 1.0 m, sandy clay to >3 m
Sand to 0.7 m, light clay to 1.5 m
1175
700
1200
818
1175
1200
1125
1025
925
1275
1200
374
912
432
374
353
Loam to 0.5 m, heavy clay to >3 m
Heavy clay to >3 m
Loam to 0.5 m, medium clay to >3 m
Light sand to >3 m
∼1333
∼1333
∼1333
∼1333
Light sand to >3 m
∼1333
The variation in depth-to-groundwater at each site was largely seasonal. Over the periods of measurement there was little overall change.
184
Australian Journal of Botany
More details of the research projects from which data for this
paper were derived have been reported in Benyon and Doody
(2004) for the Green Triangle (available via the internet from
www.ffp.csiro.au/downloads/PlantationWaterUse.pdf ) and Polglase
et al. (2002) for the Riverina (available at www.rirdc.gov.au/reports/
AFT/02-146.pdf ).
Green Triangle sites
The Green Triangle region has a Mediterranean climate, with cool wet
winters and warm to hot, dry summers. The mean daily maximum
temperature at Mount Gambier Aerodrome ranges from 13.1◦ C in July
(mid-winter) to 25.1◦ C in January, with minima in these months of
5.1◦ C and 11.0◦ C, respectively. Mean precipitation is 711 mm year−1 , of
which 485 mm (68%) occurs between May and October. Aerial potential
evapotranspiration is ∼980 mm year−1 (Wang et al. 2001).
The western half of the region is underlain by karst or sedimentary
geology, largely laid down under shallow seas, or as stranded coastlines
as sea levels have risen and fallen over the past 40 million years.
These areas have sandy surface soils, of various depths (a few
decimetres to several metres), overlying sandy clay. According to
McKenzie et al. (2000), hydraulic conductivity of the A horizons of
such soils is high (>100 mm h−1 ). The underlying sedimentary geology
is generally highly porous. Rainfall not used by the vegetation tends
to drain vertically rather than runoff. Consequently, there are few
natural surface streams or lakes, except within a few kilometres of
the coast.
Most of the tree plantations have been established in the southern
half of the region where rainfall is highest (mean precipitation ranging
from ∼650 to 850 mm year−1 ). P. radiata has been grown in the region
for timber for nearly 120 years, currently on a 28–37-year rotation for
a mix of wood-based products. E. globulus has only been grown on a
commercial scale since the late 1990s. Rotation lengths are anticipated
to be 10–12 years, mainly for pulpwood.
Between 1994 and 2004, research plots were established at seven
P. radiata sites and nine E. globulus sites. All but one of these
was located over karst or sedimentary geology. One E. globulus site,
located in south-western Victoria, had a heavy clay soil derived from
basalt.
Five P. radiata sites were established between 1994 and 1997. Two
additional P. radiata sites were established in 2000. Plot sizes were
chosen largely on the basis of current stocking density, resulting in
plots varying in size from 20 × 20 m to 30 × 30 m, containing between
30 and 50 trees each.
The nine E. globulus sites represent the range of depth-to-watertable
commonly encountered at sites used for this species in the karst
geological areas of the region. Plots at all sites were nominally
20 × 20 m.
Riverina sites
The Riverina has mild winters and hot summers, with 406 mm year−1
rainfall distributed uniformly through the year. Mean daily maxima
range from 14.4◦ C in July to 32.5◦ C in January and minima 3.1◦ C
in July to 15.5◦ C in January. Aerial potential evapotranspiration is
∼1200 mm year−1 (Wang et al. 2001). The region is composed of
extensive flood plains from the Murray River. Soils vary widely from
sandy prior-streams to deep clays.
Water use at these sites was measured as part of a research project
investigating plantation productivity in relation to water availability. Five
sites were chosen in 1997 to include the range of soil types in which
farm-forestry plantations were being established. Irrigation water was
available, but because of the landowners’ decisions to minimise use of
limited irrigation water, four sites were irrigated only once or twice
each year.
R. G. Benyon et al.
Determination of net water balances
At each site, for each period of several weeks between soil water
measurements, net water balance (deep drainage or water uptake from
groundwater) was estimated to be
Qwt = P − (I + E + T ) − (Sc − Sp )
(1)
where Qwt = net water balance: either drainage (a positive value) or
water uptake (a negative value) below the maximum depth of soil water
measurement; P = gross total precipitation for the period, measured in a
rain gauge in the open nearby, plus net irrigation in the case of Riverina
sites; I = rainfall interception losses, estimated from measurements of
throughfall; E = evaporative losses from the soil surface measured with
mini lysimeters; T = transpiration determined by direct measurement
with sap flow sensors; Sc = the current volumetric water content of the
root zone measured with a neutron moisture meter; Sp = the previous
volumetric water content of the root zone.
The sum of I, E and T makes up the total evapotranspiration. Thus,
for a given period, the net water balance is equal to net water applied
(i.e. rain plus irrigation) from which evapotranspiration and any change
in soil water is subtracted. A positive net water balance can imply either
deep drainage, net surface runoff or subsurface lateral flow out of the
research plot, or a combination of these, whereas a negative net water
balance can imply net lateral movement of water into the research plot,
either from higher up-slope as overland flow or subsurface lateral flow,
or as lateral flow through a watertable.
Climate
In the Green Triangle, for each site, rainfall in a rain gauge in an open area
(located within 2 km for all sites) was recorded every time the site was
visited (nominally every 2–4 weeks). In the Riverina, meteorological
data (rainfall, pan evaporation, wet- and dry-bulb temperature, solar
radiation and wind speed and direction) were recorded hourly with
an automatic weather station (Starlog, UNIDATA, Australia) at
one site.
Soil water
Soil water was measured at the end of each measurement period to
maximum depths of 3 m at most sites and 6 m at three others (Sites 6,
9 and 14) with a neutron probe. At all sites with <3-m depth-togroundwater, this was to the watertable or to the capillary fringe above
the watertable. Measurements were collected every 2–4 weeks in five
randomly located access holes at each site at depths of 0.075, 0.15, 0.3 m
and then every 0.3 m to the maximum depth of measurement. For each
site, a calibration curve was derived by using soil samples collected
during installation of the access tubes.
Interception
At 10 sites in the Green Triangle (Sites 1–3, 5–7, 9, 11, 13,
14 in Table 1) throughfall was measured by using eight randomly
located collection troughs. These were 90-degree, V-shaped pieces
of aluminium, each 1.20 m long by 0.14 m wide, draining into a
collection drum. The volume of water collected in each drum was
measured whenever soil water was measured. Stem flow was not
measured, but estimated assuming 10% of rainfall for P. radiata
(Langford and O’Shaughnessy 1977) and 2% of rainfall for eucalypts
(Vertessy et al. 2000). Interception loss was calculated as rainfall
minus throughfall (including the 10 or 2% allowance for stem flow).
For the six other Green Triangle sites, interception was estimated
on the basis of the mean percentage interception determined for
that species.
For the Riverina sites, interception was estimated on the basis of
seasonal percentages previously measured for E. grandis in that climate
(Myers et al. 1996).
Plantations and groundwater
Soil evaporation
In the Green Triangle, soil evaporation was measured at Sites 1–3, 5–7,
9, 11, 13 and 14, with five to nine mini-lysimeters per site. Each lysimeter
contained a soil column collected in situ 0.27 m deep. Net rainfall beside
each lysimeter was measured with a standard rain gauge. Drainage
through the lysimeter was collected and measured. Evaporation from
the soil column for each measurement period was estimated on the basis
of rainfall in, drainage out and the change in weight over the period.
Measurements were collected each time soil water was measured. The
lysimeter surface was kept free of litter fall.
Because the lysimeters excluded live roots, rates of soil water
depletion might be lower than in the surrounding soil. This would, at
times, result in the soil water content inside the lysimeters being higher
than in the corresponding depth of soil outside the lysimeters, causing
an over-estimation of soil evaporation. Examination of the rate of soil
evaporation from the lysimeters, in relation to soil water content over
2–3 years of measurements, indicated that for water content lower than
∼50% of the maximum water content, there was a linear relationship
between the evaporation rate and soil water content. For water content
greater than 50% of the maximum water content there was no significant
correlation between the rate of evaporation and soil water content. In
these conditions, soil evaporation proceeded at a rate determined by
weather rather than soil water content. Thus, for periods when soil
water measurements indicated that soil water content in the top 0.3 m
of soil outside the lysimeters was greater than 50% of the maximum
water content, observed soil evaporation was used in the water-balance
equation, on the assumption that the rate of soil evaporation would be
the same inside and outside the lysimeter because it was unconstrained
by water availability. For periods when soil water content in the top 0.3 m
of soil outside the lysimeters was lower than 50% of the maximum, soil
evaporation for the period was estimated on the basis of the average soil
water for the period and a linear regression derived from the lysimeter
data, relating daily evaporation rate to soil water content in the top 0.3 m
of soil. For the remaining six Green Triangle sites, soil evaporation was
estimated on the basis of the average rates observed at the sites where
it was measured.
In the Riverina, soil evaporation was similarly determined from the
measurements collected from mini-lysimeters. Measurements of daily
soil evaporation over drying cycles of several days, on several occasions,
were used to determine relationships between surface soil water content
and daily evaporation rate.
Transpiration
Transpiration at all sites was determined with sapflow sensors employing
the heat-pulse technique (Models SF100 and SF300, Greenspan
Technology, Warwick, Qld). Sap velocities were recorded every 30 min
in six or eight sample trees per site and transpiration each day was
estimated as the product of observed mean sap velocity and plot sap
conducting-wood area.
To select the sample trees, the stem diameters of all trees in each
plot were measured and stem basal area over bark calculated. Plot total
basal area was also calculated and each plot was divided into two or
three classes of tree size. Each class contained an equal total basal area.
In each plot, either four trees (if only two size classes were used) or two
trees (if three size classes were used) for sap-flow measurement were
randomly selected from each tree size class.
Estimates of wound size around the holes drilled for installation of
the sensors were based on previous measurements (3.0 mm for eucalypts
and 2.4 mm for radiata pine). Wood and water volume fractions were
determined on several occasions from 5-mm-diameter cores. Within
plots, these did not change significantly with time. Heat-pulse velocities
were converted to sap velocities after Swanson and Whitfield (1981)
and Edwards and Warwick (1984). To account for radial variation, the
Australian Journal of Botany
185
sapwood area was divided into two (SF100 sensors) or four (SF300
sensors) concentric rings of equal area, and one sensor (SF300) or two
sensors (SF100) located at a random depth and azimuth within each ring,
after Benyon (1999). Zero flows were identified by the method described
by Benyon (1999). Mean sap velocity for a tree was calculated as the
average of the two or four sample points within the tree.
In eucalypts, sapwood thickness was determined every 3–4 months
with the sap-flow sensors after Hatton et al. (1995) and Benyon
(1999). Young P. radiata trees (<15 years old) contained little or no
heartwood. In older P. radiata trees, sapwood thickness was determined
by examining the colour change in three 12-mm-diameter wood cores
per sample tree. Tree sapwood area at the sap-flow measurement height
was determined on the basis of measured stem diameter, bark thickness
and sapwood thickness. For trees in which sap velocity and sapwood
thickness were not measured, sapwood area was estimated from a
linear regression with tree basal area over bark. Plot sapwood area
was calculated by summing the estimated sapwood areas of all trees
in the plot. Daily transpiration was estimated as the product of the 24-h
mean sap velocity (0600 to 0600 hours) of the sample trees and plot
sapwood area.
Statistical analysis
Statistical confidence limits were estimated for annual interception,
annual transpiration and annual soil evaporation for each site. For
interception, confidence limits were based on the variation in mean
throughfall between collection troughs. For soil evaporation, they were
based on the variation in mean evaporation between lysimeters and for
transpiration they were based on the variation in mean sap velocity
between sample trees.
Evapotranspiration is the sum of three independently measured
variables. Assuming sampling errors in the measurements of one
variable were independent of the sampling errors in the measurements
of the other variables, a combined 95% confidence interval for
evapotranspiration was initially estimated as the sum of the 62%
confidence intervals for the three variables, on the basis that there
is only a 5% probability the true mean of all three variables would
fall either above or below the 62% confidence intervals at the same
time (i.e. 0.38 × 0.38 × 0.38 = 0.05). However, because the confidence
intervals for interception and evaporation were small, relative to those
for transpiration, it was found that the 95% confidence interval for
evapotranspiration estimated by this method was narrower than the
95% confidence interval for transpiration alone. Thus, we have assumed
that the 95% confidence interval for evapotranspiration is equal to that
for transpiration.
Leaf area index
In the Green Triangle, measurements of leaf area and sapwood area at
breast height from individual P. radiata trees (Teskey and Sheriff 1996)
and E. globulus trees (D. N. Fife, CSIRO, unpubl. data) were used to
derive linear relationships between tree leaf area and tree sapwood area
for these two species. These relationships were applied to subsequent
estimates of tree sapwood area to estimate tree leaf area and site leaf
area index (LAI) two to four times per year during the study periods.
Results
Annual evapotranspiration
Annual evapotranspiration varied widely among sites and,
overall, was not well correlated with rainfall (Green Triangle
sites) or with rainfall plus net irrigation (Riverina sites)
(Fig. 1). In about half the sites (Sites 1–5, 7, 10, 11, 19–21 in
Tables 1, 2, annual evapotranspiration significantly exceeded
186
Australian Journal of Botany
Water use (mm year–1)
1500
1200
R. G. Benyon et al.
E. globulus
P. radiata
E. grandis
C. maculata
900
600
300
300
450
600
750
900
Rainfall or rainfall + net irrigation (mm year–1)
Fig. 1. Relationship between annual evapotranspiration (water use)
and rainfall (Pinus radiata and Eucalyptus globulus) or rainfall + net
irrigation (E. grandis and Corymbia maculata). Error bars indicate the
95% statistical confidence limits for water use. The solid line shows the
one-to-one relationship (water used = water applied).
annual rainfall. The range in annual evapotranspiration
among sites (from 415 to 1343 mm year−1 ) was substantially
higher than the range in annual water applied (from 362 to
784 mm year−1 ). For all but one of the sites, annual water
applied was in the range 489–784 mm year−1 . Excluding the
one P. radiata site with very low rainfall (Site 10), the range in
evapotranspiration among sites (928 mm year−1 ) was more
than three times greater than the range in water applied
(295 mm year−1 ).
The one-to-one line (water used = water applied) is also
shown on Fig. 1. For about half the sites (Sites 6, 8, 9, 12–18),
the one-to-one line passed through or slightly above the
95% statistical confidence interval for annual water use.
However, for the remaining sites, the one-to-one line was
well below the 95% confidence intervals for mean annual
evapotranspiration, indicating another source of water was
available to the trees in addition to rainfall and net irrigation
(Sites 1–5, 7, 10, 11, 19–21).
Sources of water used by the trees
Table 2 shows for each site, the mean annual water
applied, best available estimate of the theoretical potential
evapotranspiration, actual evapotranspiration and the net
water balance, as determined from Eqn 1. For the Green
Triangle sites, the use of either point or areal potential
evapotranspiration, as defined by Wang et al. (2001), was
Table 2. Annual water balance for each site
Column 1 lists site numbers as in Table 1. Column 2 indicates species as defined in the text. Column 3 indicates mean annual water
applied (P in Eqn 1). Column 4 gives the best available estimate of potential evapotranspiration, as either point potential (p) or areal
potential (a) for the Green Triangle and estimated with the Penman–Monteith equation for the Riverina. Observed evapotranspiration
is the sum of interception (I in Eqn 1), soil evaporation (E in Eqn 1) and transpiration (T in Eqn 1). Sites highlighted in bold
used significant quantities of groundwater
Site
Species
Rain
(+net irr)
(mm year−1 )
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
EGl
EGl
EGl
EGl
EGl
EGl
EGl
EGl
EGl
PR
PR
PR
PR
PR
PR
PR
662
641
675
545
669
701
732
489
658
362
747
650
678
667
724
600
17
18
19
20
21
EGr
EGr
EGr
EGr
CM
784
534
633
664
619
Potential
evapotranspiration
(mm year−1 )
Observed
evapotranspiration
(mm year−1 )
Green Triangle sites
1250 p
970 a
1230 p
1180 p
980 a
980 a
1230 p
1400 p
1150 p
1340 p
1230 p
960 a
970 a
980 a
970 a
970 a
Net change in
soil water
(mm year−1 )
Net water
balance
(mm year−1 )
1059
847
1158
1193
904
713
1167
488
604
1074
1343
560
648
795
635
540
16
–99
–43
–12
–9
–10
–9
–3
0
–41
–35
0
0
–19
108
0
–413
–107
–440
–636
–226
–2
–426
4
54
–671
–561
90
30
–109
–19
60
704
415
712
928
1277
80
119
37
113
75
0
0
–116
–377
–733
Riverina sites
1124
1015
1124
1220
1280
Plantations and groundwater
187
Median depth-to-groundwater (m)
0
5
10
15
20
25
400
200
Net water balance (mm year–1)
a subjective determination based on plantation size and
proximity of the site to an exposed plantation edge.
Sources of water used by the trees were determined
by comparing rainfall (or rainfall + net irrigation),
evapotranspiration and changes in soil water, by using
Eqn 1. In both regions, rainfall is highest and potential
evapotranspiration is lowest during the winter months.
In the Green Triangle, from June to September, water
use of plantations is lower than rainfall, and thus soil
water accumulates. Soil-moisture profiles indicated water
movement through the soil to greater than 3-m depth at
most sites in winters with average or above-average rainfall
(data not shown). Water stored in the soil during winter was
transpired over the following spring and summer, so that over
periods of ∼12 months, changes in soil water were relatively
small. The plantations in the Riverina were flood-irrigated
1–8 times each summer to supplement rainfall, but not in
sufficient amounts to match potential evapotranspiration.
None of the plantations substantially depleted stores of
soil water over the 2–5 years of measurements. At all sites
except Sites 9 and 12, the plantations used all of the rainfall
available during the study period (Fig. 1).
The right-hand column in Table 2 shows the net annual
water balance for each site as estimated from Eqn 1.
A negative value indicates the site used more water than
could be accounted for from water applied and net change in
soil water, whereas a positive value indicates deep drainage
or runoff. Eleven of the sites (E. globulus Sites 1–5 and 7,
P. radiata Sites 10 and 11, E. grandis Sites 19 and 20 and
C. maculata Site 21) used additional water which could not
be accounted for by depletion in soil water in the upper
3 or 6 m of the soil profile. We conclude this additional
water was taken up from the groundwater by the trees.
At 10 of the 11 sites where groundwater uptake occurred
(Sites 1–5, 7, 10 and 19–21), the maximum depth of soil
water measurement was sufficient to account for all soil
water between the soil surface and capillary fringe above
the watertable. At P. radiata Site 11, there was ∼2 m of soil
above the watertable in which soil water was not measured.
However, on the basis of maximum and minimum soil water
over a 5-year period in the top 3 m of soil at this site, no
more than 5% of the unaccounted for water use could have
been derived from this zone through depletion of stored
soil water.
Figure 2 illustrates the relationship between net water
balance and median depth-to-groundwater. Groundwater
uptake, indicated by a significant negative net water balance,
was related to depth-to-groundwater. Of six sites with 7.5 m
or greater depth-to-groundwater (E globulus Sites 6, 8 and 9
and P. radiata Sites 13–15), none used significantly more
water than received from rainfall and all water use was
accounted for in the long term by rainfall (Table 2, Fig. 2).
Of 14 sites with groundwater within 6 m of ground level, we
conclude that at 11 sites, the trees used significant amounts
of groundwater (E. globulus Sites 1–5 and 7, P. radiata
Australian Journal of Botany
0
–200
–400
–600
E. globulus
P. radiata
E. grandis
C. maculata
–800
–1000
Fig. 2. Relationship between depth-to-groundwater and net water
balance for closed-canopy Pinus radiata and Eucalyptus globulus sites
in the Green Triangle and E. grandis and Corymbia maculata sites in
the Riverina. Error bars indicate the 95% statistical confidence limits.
Sites 10 and 11, E. grandis Sites 19 and 20 and C. maculata
Site 21), accounting for between 13 and 72% of annual
water use. These sites had sandy or light or medium clay
soils (Table 1). Of three sites where the trees were not
using groundwater despite shallow depth-to-groundwater,
two (E. grandis Sites 17 and 18) had heavy, sodic, clay
subsoils, whereas one (P. radiata Site 12) had a cemented
hardpan layer at 1.2-m depth. The latter used significantly
less water than received from rainfall, implying net
groundwater recharge.
Mean transpiration at the sites where the trees
used groundwater was 767 mm year−1 , compared with
358 mm year−1 for the sites where no groundwater use was
detected, a difference of 409 mm year−1 . Thus, the difference
in mean annual water use between sites using and not using
groundwater was due to a difference in transpiration.
Relationship between leaf area index and water use
One of the aims of the Green Triangle studies was to
determine how accurately water use of existing plantations
could be estimated on the basis of plantation inventory
data or variables measurable by remote sensing, such as
leaf area index (LAI). For the Green Triangle dataset, the
relationship between annual transpiration and leaf area index
(Fig. 3) appeared to differ between sites using and not using
groundwater. The three sites with the lowest LAI (Sites 8,
9 and 12) did not use groundwater and had low water
use, whereas the three sites with the highest LAI (Sites 4,
10 and 11) used groundwater and had high annual wateruse rates. However, of sites with LAI between 3.2 and
188
Australian Journal of Botany
R. G. Benyon et al.
Transpiration (mm year–1)
1000
800
600
400
using groundwater
not using groundwater
200
0
0
1
2
3
4
5
6
7
8
LAI (m2 m–2)
Fig. 3. Relationship between annual transpiration and leaf area index
(LAI) for the Green Triangle sites. Filled circles indicate sites that
used some groundwater. The grey box indicates plots having mean LAI
between 3.2 and 4.2. Two lines of best fit are shown.
4.2, five accessed groundwater (Sites 1–3, 5, 7) whereas
four did not (Sites 6, 13–15). Mean annual transpiration
was 350 mm year−1 higher for the five sites which accessed
groundwater. The relationship between transpiration and LAI
was not influenced by species (data not shown).
Discussion
Groundwater recharge under plantations
All but 2 of the 21 sites in the Green Triangle and the
Riverina, listed in Table 1, used all of the available rainfall
during the periods of measurement. In both regions, mean
annual rainfall is substantially lower than annual potential
evapotranspiration (Table 2), indicating that these are waterlimited environments. Thus, in all cases there is potential for
tree plantations to use more water than available from rainfall.
Once the canopy of a plantation has closed, deep-drainage and
groundwater recharge is likely to occur if root-zone waterholding capacity is exceeded during periods when cumulative
rainfall is greater than cumulative evapotranspiration. This
is possible in the Green Triangle where, from May to
August, mean total rainfall exceeds mean total potential
evapotranspiration by ∼200 mm.
The method used to estimate net water balances in this
study cannot determine whether the proportion of rainfall
which is unaccounted for, went to groundwater recharge,
surface runoff or net subsurface lateral flows. A statistically
significant positive net water balance was observed at Sites 9
and 12. Site 12, in south-eastern South Australia, has a
sandy surface soil and gently sloping topography, with a
hardpan present at 1.2-m depth. Hydraulic conductivity of
the surface soil, although not measured, was probably high,
and therefore surface runoff from the site, which was in a
mid-slope position, was unlikely. It is possible that excess
rainfall drained vertically to the hardpan layer, where it
may have moved laterally further down-slope, rather than
recharging the groundwater. Site 9, in south-western Victoria,
was almost flat, with a heavy clay soil. Periodic measurements
of soil water down to 6-m depth indicated that several wetting
fronts moved slowly through the soil down to >6 m between
August and November during a wet winter, suggesting
the likelihood of deep drainage reaching the watertable at
7.5-m depth.
No studies in south-eastern Australia have measured
evapotranspiration in the period of the rotation between
harvesting of one tree crop and establishment of the canopy
of the next. In south-eastern South Australia, Mitchell and
Correll (1987) observed that the soil water profile fully
recharged in the 1-year fallow period between the first
and second rotation P. radiata plantation, indicating the
possibility of some groundwater recharge during this period.
More research is needed if recharge during this period of each
rotation is to be quantified and the net water balance over a
full rotation estimated accurately.
Groundwater uptake by plantations
Significant groundwater uptake occurred at 11 of the study
sites, providing between 13 and 72% of annual transpiration.
In both the Green Triangle and the Riverina, the highest rate
of net groundwater uptake was ∼700 mm year−1 . This is
the highest absolute annual rate of groundwater uptake by
trees that we are aware of. Thorburn (1999) collated results
from 20 diverse field experiments on trees, crops, shrubs
and pasture, from which estimates of groundwater uptake
had been derived. These ranged from 10 to 590 mm year−1 .
At all sites, depth-to-watertable was <5 m; however,
groundwater was moderately saline (>3000 mg L−1 ). In
several studies in northern Victoria and south-eastern
Queensland, groundwater uptake of 100–279 mm year−1 by
E. camaldulensis and E. grandis woodlots was estimated
(Cramer et al. 1999; Morris and Collopy 1999; Vertessy et al.
2000). These sites were on shallow, slightly or moderately
saline groundwater, in soils of generally low permeability.
Under such conditions, annual evapotranspiration rates
were substantially lower than potential evapotranspiration,
and groundwater uptake may have been restricted by low
soil hydraulic conductivity, low leaf area index or salt
accumulation in the root zone.
In the Green Triangle and the Riverina, whether the
trees accessed groundwater was the most significant factor
influencing water use, accounting for 67% of the total
among-site variation in mean annual evapotranspiration.
Accessibility of groundwater was largely a function of depthto-the-watertable and soil type.
For the eight sites in the Green Triangle that used some
groundwater, the among-site range in water use was almost
500 mm year−1 . For these eight sites, the site and stand
variables we measured were only weakly correlated with
Plantations and groundwater
annual water use and groundwater uptake. There was no
significant correlation between annual water use and rainfall
(r2 = 0.01), median depth-to-the-watertable (r2 = 0.13)
or LAI (r2 = 0.14). There was, however, a significant
(r2 = 0.51, P < 0.05) correlation between theoretical
potential evapotranspiration (as defined by Wang et al.
2001) and actual evapotranspiration for these sites. This
was largely related to whether we considered the site fitted
the Wang et al. (2001) definition of ‘areal’ or ‘point’ for
determining potential evapotranspiration. For two sites
located in the middle of large plantations (Sites 2 and 5),
areal potential evapotranspiration was within or slightly
above the 95% confidence limits for observed annual
evapotranspiration. The remaining six sites were located
either in small plantations (<1 km2 , Sites 1 and 4), or
within 100 m of a plantation edge (Sites 3, 7, 10, 11). For
these, the theoretical point potential evapotranspiration was
within the 90% statistical confidence limits for observed
annual evapotranspiration. The y-intercept term for a
regression relating observed evapotranspiration to potential
evapotranspiration was not significantly different from 0,
and the slope coefficient was not statistically significantly
different from 1. Within the group of six sites considered
representative of point potential evapotranspiration, there
was a significant difference in annual water use between
the site with lowest water use (Site 1) and the site with
the highest water use (Site 11), but there were no other
statistically significant differences. Thus, for the eight Green
Triangle sites at which the trees accessed groundwater,
the among-site variation in annual water use was partly a
function of potential evapotranspiration. It is also possible
that aquifer transmissivity or soil hydraulic conductivity in
the root zone restricted the rate of water movement to the tree
root zone at some sites, but not others, or that the distribution
of roots between capillary fringe and surface soil varied
among sites.
In the Riverina, two sites (Sites 17 and 18) where there
was no detectable groundwater uptake by E. grandis,
despite groundwater present at 3-m depth, had heavy,
sodic, alkaline, saline clay subsoils below ∼0.5-m depth.
Groundwater at these sites was also saline. At Site 19,
where groundwater uptake was estimated to be small, the
texture of the clay subsoil was lighter than at Sites 17 and
18. The remaining Riverina sites (Sites 20 and 21), where
E. grandis and C. maculata both used significant quantities
of groundwater, had sandy soil and low groundwater salinity.
In the Riverina, soil type had the greatest influence on
plantation water use and groundwater uptake. The effect
of soil type, however, was confounded with the effect
of groundwater salinity, as those sites with heavy clay
subsoils also had saline groundwater. These effects could be
separated if additional sites with clay soils and low-salinity
groundwater, or sandy soils and saline ground water,
were studied.
Australian Journal of Botany
189
We conclude that the four species can take up groundwater
at sites with a combination of shallow depth-to-groundwater
with high aquifer transmissivity, low groundwater salinity,
light-textured soil with no root-impeding layers, and
annual potential evapotranspiration substantially exceeding
annual rainfall. When all these conditions are present,
plantation annual water use approaches theoretical potential
evapotranspiration at most sites, once the canopy has closed.
We have suggested that groundwater uptake in the Green
Triangle is partly a function of depth-to-the watertable. For
trees to use groundwater, their roots must have penetrated
to the depth of the capillary fringe above the watertable.
Tree roots soon after planting are very shallow (<0.3 m).
Clearly, it must take time for the root system to develop
as individual trees grow in size and for roots to extend
deep enough to access groundwater. Roots might continue
to penetrate to greater depths as the trees age. This raises the
question of whether there is a maximum depth for significant
groundwater uptake. The results from the Green Triangle
suggest this depth is somewhere between 6 and 8 m. Whether
this is an indication of species maximum rooting depth, or
resistance to water movement through the soil–root–tree–
atmosphere pathway is not known.
At some sites, trees which currently are not using
groundwater might eventually do so. However, all sites where
trees used groundwater were doing so when measurement
began at 3–5 years of age. At five eucalypt sites in the Green
Triangle (Sites 5, 8 and 9) and the Riverina (Sites 17 and 18)
trees were not using groundwater when measurements began
at 3 or 4 years of age and still were not doing so by 5–7 years
of age. In P. radiata in the Green Triangle, trees at the two
youngest sites of this species (Sites 10 and 11) were already
using groundwater when measurements began at 4 years of
age, one from a watertable at 6-m depth. The five P. radiata
sites which used no significant groundwater (Sites 12–16)
were considerably older (age range 14–32 years). Site 12,
with a watertable at 4.5 m, but a root-impeding layer at 1.2-m
depth, used no groundwater. These trees were 28–31 years
old during the period of measurement. Presumably, their
roots would have penetrated to the watertable by then if they
were going to. The trees at Sites 13 and 14, with watertables
at ∼8.5-m depth, were not using groundwater by the time
measurements ceased at Ages 18 and 24 years. Thus,
there was no evidence that accessibility of groundwater
increased with tree age after canopy closure. We are not
aware of many studies reporting groundwater uptake from
deeper than ∼5 m. Most studies of planted trees determined
groundwater uptake from watertables only a few metres
below the ground surface (Cramer et al. 1999; Morris
and Collopy 1999; Thorburn 1999; Vertessy et al. 2000).
However, O’Grady et al. (2006) inferred groundwater uptake
by some trees in remnant riparian communities from up to
10-m depth. Species rooting depth, depth-to-groundwater
and the availability of other water sources are all likely to
190
Australian Journal of Botany
R. G. Benyon et al.
be important in determining groundwater use by vegetation
(O’Grady et al. 2006).
the same site conditions to have different patterns of water
uptake, even within the one genus.
Species differences in water use and groundwater uptake
Implications for groundwater management
Although our study was not designed to provide valid
statistical comparisons of water use between species,
the results do provide some indication of whether
evapotranspiration differed between species. At most sites,
only a single species was present, but one plantation in the
Riverina included plots of E. grandis and C. maculata.
In the Green Triangle, there was no obvious difference
in mean annual transpiration or evapotranspiration between
P. radiata and E. globulus. Both species used some
groundwater at locations where the watertable was within 6 m
of the ground surface and where there were no distinct rootimpeding layers. Annual interception loss, as a percentage
of rainfall, was higher in P. radiata; however, this was
balanced by lower soil evaporation. In both species, the total
of interception loss and soil evaporation averaged ∼46% of
annual rainfall, with similar variation between P. radiata and
E. globulus sites (Table 3).
Sites 20 and 21 in the Riverina provided some evidence of
a species difference in water use. Here, Polglase et al. (2002)
observed groundwater uptake of ∼350 mm year−1 higher in
C. maculata than in an adjacent stand of E. grandis. Although
the two species were not grown together in a statistically valid
experimental design, the similarity of soil types under the two
sites lead Falkiner et al. (2006) to speculate the difference was
related to species, not soil. A study of fine-root distributions
under each species in this plantation revealed a higher rootlength density in the capillary fringe above the watertable
in C. maculata than in E. grandis (Falkiner et al. 2006).
Differences between species in the depths from which soil
water or groundwater is accessed have been reported in mixed
stands of native riparian vegetation (O’Grady et al. 2006) and
mixed-eucalypt tree belts (White et al. 2000), indicating it
is not uncommon for different tree species growing under
Less than 20 km2 of eucalypt plantations, in woodlots
generally <1 km2 , have been established in the Riverina,
some in soil types where root growth to the watertable is
restricted. In this region, increased groundwater recharge
associated with irrigation has resulted in rising saline
watertables, posing a significant risk to the sustainability
of agriculture. Establishment of tree plantations in strategic
locations might be one way to ameliorate this risk.
Polglase et al. (2002) suggested that establishment of
∼100 km2 of plantations on sandy soils over prior streams,
where trees would be able to take up groundwater of
low salinity, would be sufficient to balance the excess
groundwater recharge from irrigation in the region. These
farm-forestry plantations might also have other economic
and environmental benefits, such as wood production, carbon
sequestration and enhancement of biodiversity.
In the Green Triangle, plantations have been grown
on a commercial scale for many decades and support an
important wood-processing industry in the region. In southeastern South Australia, groundwater management plans
already account for the lower rates of groundwater recharge
under plantations than for pastures. However, no account
has been taken of the possibility of groundwater uptake
by trees in locations with shallow depth-to-groundwater
and light-textured soils. Since the late 1990s, ∼350 km2 of
E. globulus plantations have been established on ex-pasture
in an area to the west and north-west of Penola where
depth-to-groundwater is generally shallow. The potential
for these plantations to influence groundwater levels is
currently being examined. More information is also being
collected on the dependence of wetlands in this area on
groundwater. Groundwater management plans in some parts
of the region will be revised in the next few years to
Table 3. Interception, soil evaporation and transpiration
Components of evapotranspiration for sites in the Green Triangle where interception (I ), soil evaporation (E)
and transpiration (T ) were all measured. Column 1 lists site numbers as in Table 1. Column 2 indicates species
as defined in the text. The column labelled %I + %E shows the sum of I and E as a percentage of annual rainfall
Site
Species
Rain
(mm year−1 )
I
(mm year−1 )
E
(mm year−1 )
%I + %E
T
(mm year−1 )
1
2
3
5
6
7
9
11
13
14
EGl
EGl
EGl
EGl
EGl
EGl
EGl
PR
PR
PR
662
641
675
669
701
732
658
747
678
667
116
78
182
86
130
185
100
336
150
134
127
181
251
161
181
241
231
108
103
128
37
40
64
37
44
58
50
59
37
39
816
588
725
656
402
741
274
899
395
533
Plantations and groundwater
more-accurately account for the effects of various land uses
on groundwater resources. Ultimately, policies for naturalresource management need to include the broadest possible
analysis of all the economic, social and environmental
benefits of alternative land uses.
Acknowledgments
Thanks go to technical staff, including Mark Tunningley,
Michele Michael, Ian Craig, Dean Tompkins, Leroy Stewart,
David Gritton and Kevin Nickolls. Thanks go also to
Quinton Clasen and Melissa Cann who conducted soil
classifications at some of the Green Triangle sites. Site
access in the Green Triangle was provided by ForestrySA,
Auspine, Green Triangle Forest Products, Woakwine Pastoral
Co. Pty Ltd, Balak Park Pty Ltd, WA Plantation Resources,
GPFL, Martin Harvey and Tony Buffon. Site access in
the Riverina was provided by Bruce Moore, Geoff and
Leanne Small, Bill and Elwin Hermiston and Ray Baxter.
Funding for the Green Triangle sites was provided by
CSIRO, the South Australian Department of Water, Land
and Biodiversity Conservation, the South East Catchment
Water Management Board, ForestrySA, the Natural Heritage
Trust, Auspine, Green Triangle Forest Products, the Forest
and Wood Products Research and Development Corporation,
the Glenelg Hopkins Catchment Management Authority,
the Green Triangle Regional Plantation Committee and the
District Council of Grant. Funding for the Riverina sites
was provided by CSIRO, Agriculture, Fisheries and ForestryAustralia (AFFA), and the Natural Heritage Trust in
Collaboration with State Forests of NSW. David Sheriff and
Bob Teskey initiated the water-use studies in radiata pine
in the Green Triangle. Sadanandan Nambiar, Clive Carlyle,
Brian Myers, Don White and Barrie May provided valuable
advice and support at various times. Michael Battaglia and
Jenny Carter reviewed an earlier draft of the manuscript and
two anonymous reviewers provided valuable comments on
the final draft.
References
Allison GB, Hughes MW (1972) Comparison of recharge to
groundwater under pasture and forest using environmental tritium.
Journal of Hydrology 17, 81–95.
doi: 10.1016/0022-1694(72)90067-4
Benyon RG (1999) Nighttime water use in an irrigated Eucalyptus
grandis plantation. Tree Physiology 19, 853–859.
Benyon RG (2002) ‘Water use by tree plantations in the Green Triangle:
a review of current knowledge.’ (Glenelg Hopkins Catchment
Management Authority: Hamilton, Vic.; and CSIRO Forestry and
Forest Products: Mt Gambier, SA)
Benyon RG, Doody TM (2004) Water use by tree plantations in south
east South Australia. Technical Report No. 148. CSIRO Forestry and
Forest Products, Mount Gambier, SA.
Bosch JM, Hewlett JD (1982) A review of catchment experiments
to determine the effects of vegetation changes on water yield and
evaporation. Journal of Hydrology 55, 3–23.
doi: 10.1016/0022-1694(82)90117-2
Australian Journal of Botany
191
Colville JS, Holmes JW (1972) Watertable fluctuations under forest
and pasture in a karstic region of southern Australia. Journal of
Hydrology 17, 61–80. doi: 10.1016/0022-1694(72)90066-2
Cramer V, Fraser G, Thorburn P (1999) Tree water use and root
zone salt accumulation in saline farm forest plots in south-east
Queensland: 1 Tree water use. In ‘Agroforestry over shallow
watertables: the impact of salinity on sustainability’. Water and
Salinity Issues in Agroforestry No. 4. Publication No. 99/36.
(Ed. PJ Thorburn) pp. 44–45. (Rural Industries Research and
Development Corporation: Canberra)
Dugas WA, Meyer WS, Barris HD, Fleetwood RJ (1990) Effects of soil
type on soybean crop water use in weighing lysimeters. II. Root
growth, soil water extraction and water-table conditions. Irrigation
Science 11, 77–81. doi: 10.1007/BF00188443
Edwards WRN, Warwick NWM (1984) Transpiration from a kiwifruit
vine as estimated by the heat pulse technique and the Penman–
Monteith equation. New Zealand Journal of Agricultural Research
27, 537–543.
Falkiner RA, Nambiar EKS, Polglase PJ, Theiveyanathan S, Stewart LG
(2006) Root distribution of Eucalyptus grandis and Corymbia
maculata in degraded saline soils of south-eastern Australia.
Agroforestry Systems, in press.
Farrington P, Watson GD, Bartle GA, Greenwood EAN (1990)
Evaporation from dampland vegetation on a groundwater mound.
Journal of Hydrology 115, 65–75.
doi: 10.1016/0022-1694(90)90198-7
Green Triangle Regional Plantation Committee (2001) ‘South east forest
industry development strategy.’ (GTRPC: Mount Gambier, SA)
Hatton TJ, Moore SJ, Reece PH (1995) Estimating stand transpiration
in a Eucalyptus populnea woodland with the heat pulse method:
measurement errors and sampling strategies. Tree Physiology 15,
219–227.
Hibbert AR (1967) Forest treatment effects on water yield. In
‘International symposium on forest hydrology’. (Eds WE Sopper,
HW Lull) pp. 527–543. (Pergamon: Oxford)
Holmes JW, Colville JS (1970a) Grassland hydrology in a karstic region
of southern Australia. Journal of Hydrology 10, 38–58.
doi: 10.1016/0022-1694(70)90053-3
Holmes JW, Colville JS (1970b) Forest hydrology in a karstic region of
southern Australia. Journal of Hydrology 10, 59–74.
doi: 10.1016/0022-1694(70)90054-5
Holmes JW, Sinclair JA (1986) Water yield from some afforested
catchments in Victoria. In ‘Hydrology and water resources
symposium, Griffith University, Brisbane, 25–27 November 1986’.
pp. 214–218. (The Institute of Engineers Australia: Barton, ACT)
Hopton H, Schmidt L, Stadter F, Dunkley C (2001) Forestry, land
use change and water management: a Green Triangle perspective.
In ‘Plantations, farm forestry and water. Water and Salinity Issues
in Agroforestry No. 7’. Publication No. 01/20. (Eds ESK Nambiar,
AG Brown) pp. 36–43. (Rural Industries Research and Development
Corporation: Canberra)
Kelly G, Hutley LB, Eamus D, Jolly P (2002) Role of savanna
vegetation in groundwater dynamics in a wet–dry tropical climate.
In ‘Proceedings, balancing the groundwater budget, Darwin,
12–17 May 2002’. [CD-ROM.] (International Association of
Hydrogeologists: Darwin)
Knight JH (1999) Root distribution and water uptake patterns in
eucalypts and other species. In ‘The way trees use water. Water
and Salinity Issues in Agroforestry No. 5’. Publication No. 99137.
(Ed. JJ Landsberg) pp. 55–85. (Rural Industries Research and
Development Corporation: Canberra)
Langford KJ, O’Shaughnessy PJ (1977) A study of canopy
interception in native forests and conifer plantations. Melbourne
and Metropolitan Board of Works, Report No. MMBW-W-0007,
Melbourne.
192
Australian Journal of Botany
R. G. Benyon et al.
McEwen K, Leaney F (1996) Using the natural abundance composition
of stable isotopes in sapwater, groundwater and soilwater to estimate
water use by pine plantations near Mt Gambier, South Australia.
Consultancy Report No. 96-34, CSIRO Division of Water Resources,
Adelaide.
McKenzie NJ, Jacquier DW, Ashton LJ, Cresswell HP (2000) Estimation
of soil properties using the Atlas of Australian Soils. Technical
Report 11/00, CSIRO Land and Water, Canberra.
Mitchell BA, Correll RL (1987) The soil water regime in a young radiata
pine plantation in southeastern Australia. New Forests 1, 273–289.
doi: 10.1007/BF00031738
Morris JD, Collopy JJ (1999) Water use and salt accumulation by
Eucalyptus camaldulensis and Casuarina cunninghamiana on a site
with shallow saline groundwater. Agricultural Water Management
39, 205–227. doi: 10.1016/S0378-3774(98)00079-1
Myers BJ, Theiveyanathan S, O’Brien ND, Bond WJ (1996) Growth
and water use of Eucalyptus grandis and Pinus radiata plantations
irrigated with effluent. Tree Physiology 16, 211–219.
Nambiar EKS, Brown AG (Eds) (2001) Plantations, farm forestry
and water. Proceedings of a national workshop, 20–21 July 2000,
Melbourne. Water and Salinity Issues in Agroforestry No. 7.
Publication No. 01/20. (Rural Industries Research and Development
Corporation: Canberra)
O’Grady AP, Eamus D, Hutley LB (1999) Transpiration increases in the
dry season: patterns of tree water use in the eucalypt open forests of
Northern Australia. Tree Physiology 19, 591–597.
O’Grady AP, Cook P, Lamontagne S, Eamus D (2002) Water use
by riparian vegetation along the Daly River in the Northern
Territory. In ‘Proceedings, balancing the groundwater budget’,
Darwin, 12–17 May 2002. [CD-ROM.] (International Association
of Hydrogeologists: Darwin)
O’Grady AP, Cook PG, Howe P, Werren G (2006) Groundwater use by
dominant tree species in tropical remnant vegetation communities.
Australian Journal of Botany 54, 155–171.
O’Loughlin E, Nambiar EKS (2001) Plantations, farm forestry and
water. A discussion paper. Water and Salinity Issues in Agroforestry
No. 8. Publication No. 01/137. (Rural Industries Research and
Development Corporation: Canberra)
Polglase PJ, Theiveyanathan S, Benyon RG, Falkiner RA (2002)
‘Irrigation management and groundwater uptake in young tree
plantations growing over high watertables.’ (Rural Industries
Research and Development Corporation: Canberra)
South East Catchment Water Management Board (2003) ‘South East
Catchment Water Management Plan 2003–2008.’ (SECWMB:
Mount Gambier, SA)
Swanson RH, Whitfield DWA (1981) A numerical analysis of heat pulse
velocity theory and practice. Journal of Experimental Botany 32,
221–239.
Teskey RO, Sheriff DW (1996) Water use by Pinus radiata trees in a
plantation. Tree Physiology 16, 273–279.
Thorburn PJ (1999) The limits to evaporation from shallow, saline
water tables—the big picture? In ‘Agroforestry over shallow
watertables: the impact of salinity on sustainability. Water and
Salinity Issues in Agroforestry No. 4’. Publication No. 99/36.
(Ed. PJ Thorburn) pp. 18–20. (Rural Industries Research and
Development Corporation: Canberra)
Thorburn PJ, Walker GR (1994) Variations in stream water uptake by
Eucalyptus camaldulensis with different access to stream water.
Oecologia 100, 293–301. doi: 10.1007/BF00316957
Thorburn PJ, Hatton TJ, Walker GR (1993) Combining measurements of
transpiration and stable isotopes of water to determine groundwater
discharge from forests. Journal of Hydrology 150, 563–587.
doi: 10.1016/0022-1694(93)90126-T
Vertessy R, Connell L, Morris J, Silberstein R, Heuperman A, Feikema P,
Mann L, Komarzynski M, Collopy J, Stackpole D (2000) Sustainable
hardwood production in shallow watertable areas. Water and Salinity
Issues in Agroforestry No. 6. Publication No. 00/163. (Rural
Industries Research and Development Corporation: Canberra)
Wang QJ, Chiew FHS, McConachy FLN, James R, de Hoedt GC,
Wright WJ (2001) ‘Climatic atlas of Australia: evapotranspiration.’
(Bureau of Meteorology: Melbourne)
White DA, Turner NC, Galbraith JH (2000) Leaf water relations
and stomatal behaviour of four allopatric Eucalyptus species
in Mediterranean south-western Australia. Tree Physiology 20,
1157–1165.
Zhang L, Dawes WR, Walker GR (1999) ‘Predicting the effect
of vegetation changes on catchment average water balance.’
(Co-operative Research Centre for Catchment Hydrology:
Canberra)
Manuscript received 8 March 2005, accepted 27 September 2005
http://www.publish.csiro.au/journals/ajb