(This is a sample cover image for this issue. The actual cover is not yet available at this time.

This article appeared in a journal published by Elsevier. The attached

copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elseviers archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
 Author's personal copy

Ecological Engineering 52 (2013) 112118

Contents lists available at SciVerse ScienceDirect

Ecological Engineering
journal homepage: www.elsevier.com/locate/ecoleng

Water-retention additives increase plant available water in green roof substrates

Claire Farrell , Xing Qi. Ang 1 , John P. Rayner
Melbourne School of Land and Environment, The University of Melbourne, Burnley Campus, 500 Yarra Boulevard, Richmond, Victoria 3121, Australia

a r t i c l e i n f o a b s t r a c t

Article history: Green roofs in hot and dry climates are frequently exposed to drought due to shallow depth and low
Received 17 September 2012 water holding capacity (WHC) of substrates (growing media). Water-retention additives have the poten-
Received in revised form tial to increase substrate water availability leading to greater plant growth and survival, expanding the
11 December 2012
range of plant species suitable for green roofs. We determined whether two different water-retention
Accepted 15 December 2012
additives (silicate granules and hydrogel) increased substrate WHC, plant available water (PAW), days
until permanent wilting and growth of winter wheat (Triticum aestivum L.) and white lupin (Lupinus albus
Keywords:
L.) grown in green roof substrates. Two substrates were compared; one based on scoria, the other based
Super absorbent polymer
Hydrophilic polymer
on crushed terracotta roof-tiles. Without additives both substrates had similar WHC (4043%) but PAW
Polyacrylamide was lower in scoria than in roof-tile due to greater air-lled porosity (AFP) and lower bulk density in
Sanoplant scoria. Silicates increased WHC in both scoria and roof-tile substrates, but hydrogel only improved scoria
Soil conditioner WHC. The effects of additives on PAW differed between plant species, with additives increasing PAW in
Substrate amendment both substrates for wheat but only in roof-tile for lupins. PAW was not signicantly different between
hydrogel and silicates. Although hydrogels increased substrate WHC and PAW there was no increase in
time until wilting in either substrate. Silicates on the other hand, increased time until wilting and total
and root biomass for wheat growing in scoria and for lupins in both substrates. While there was a species
effect on the efciency of water-retention additives which requires further study, the addition of silicates
is more likely to enhance species capacity to tolerate periods of low water availability on green roofs.
2013 Elsevier B.V. All rights reserved.

1. Introduction
In cities, green or vegetated roofs have the potential to mit-
igate the urban heat island effect (Bass and Baskaran, 2003),
improve building insulation and energy efciency (Sailor, 2008),
provide biodiversity habitat (Brenneisen, 2006) and manage urban
stormwater (VanWoert et al., 2005b; Berndtsson, 2010). Green
roofs generally consist of a shallow substrate (growing media) layer
planted with drought tolerant succulent species, usually Sedum
species (Snodgrass and Snodgrass, 2006; Oberndorfer et al., 2007).
However, as the cooling and stormwater functions of green roofs
require high plant water use, Sedum sp. are unlikely provide these
benets (Wolf and Lundholm, 2008; Nagase and Dunnett, 2010).
Expanding plant selection beyond succulent species on exten-
sive green roofs is challenging as water availability is often limited
due to the limited depth (Nagase and Dunnett, 2010) and high
porosity of the substrate; necessary to increase aeration and limit
weight loading. However, high air-lled porosity results in lower
Corresponding author. Tel.: +61 3 9250 6849; fax: +61 3 9250 6885.
E-mail address: c.farrell@unimelb.edu.au (C. Farrell).
1
Present address: National Parks Board, 1 Cluny Road, Singapore 259569,
Singapore.
water retention (Spomer, 1980) and can cause periodic droughts
and rapid uctuations in moisture availability. Increasing substrate
volume and depth increases water holding capacity (WHC), but also
increases structural weight loading, limiting its application. Water-
retention additives have the potential to increase WHC of green
roof substrates without signicantly adding to the weight of green
roofs.
Two widely available water-retention additives are hydrophilic
polymers, or hydrogels, and silicate-based granules. Hydrogels, also
described as Super Absorbent Polymers, are cross-linked crystalline
forms of insoluble polyacrylamide gel that absorb and store water
up to 500 times their own weight when saturated (Johnson, 1984).
Silicate-based granules such as Sanoplant contain natural silicate-
based stone powders (SiO2 ), carbon compounds and cellulose that
attach to the surface of soil particles, increasing the surface area
available for nutrient and water adsorption (Sanoway Australia,
2006). Hydrogels have been widely used for agricultural, horticul-
tural or forestry applications in drought affected climates, where
they can reduce irrigation frequency (Gehring and Lewis, 1980;
Baasiri et al., 1986; Sivapalan, 2006), improve tree establishment
(Httermann et al., 1999; Al-Humaid and Moftah, 2007), reduce
transplant shock of tree seedlings (Apostol et al., 2009) and enhance
root and shoot growth of trees (Httermann et al., 1999; Shi et al.,
2010) crops (Dorraji et al., 2010) and ornamental plants (Wang and

0925-8574/$ see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.ecoleng.2012.12.098
 Author's personal copy

C. Farrell et al. / Ecological Engineering 52 (2013) 112118 113

Boogher, 1987). Although silicate-based water-retention additives Table 1

Composition of the two green roof substrates.
are widely used for golf course water management and horti-
cultural crops in arid regions (Sanoway Australia, 2006), there is Green roof substrate Components (% by volume)
little published scientic literature evaluating their use. However, Scoria-based mix 60% scoria (8 mm)
Hosseini et al. (2012) found silicates increased olive seedling root 20% scoria (7 mm aggregate)
biomass by 50% and halved irrigation frequency in sandy soils. 20% coir (horticulture grade)
The effectiveness of water-retention additives, in particular Roof-tile based mix 80% crushed terracotta roof tiles (7 mm)
20% coir (horticulture grade)
hydrogels, for increasing plant growth and plant available water
(PAW) is dependent on both substrate and additive properties,
with differences between plant species. Hydrogels can increase
roof substrates at manufacturers recommended rates (silicates
the WHC of soil/substrates, however an increase in WHC does not
10 g L1 and hydrogel 1 g L1 ). The silicate additive contains 57%
always relate to an increase in PAW (Ingram and Yeager, 1987;
SiO2 (Hosseini et al., 2012). Substrates without additives were also
Nguyen et al., 2009). Plant available water is described as the
prepared as a control. Slow release fertilizer (16 N:1.3P:9.1 K; 34
amount of water held in the soil/substrate at tensions between eld
month release) was mixed uniformly into all substrates at a rate
capacity and permanent wilting point (PWP) (1.5 MPa) (Slatyer,
of 3 g L1 . Additives and fertilizer were added to substrates at a
1957). Generally, hydrogels have been found to be more effective
volume rate to ensure the same quantity was applied in both sub-
at increasing plant available water in coarse textured sandy soils,
strates.
with signicantly reduced available water in loam and clay soils
Levels of pH and electrical conductivity (1:5 weight-to-volume
(Koupai et al., 2008; Agaba et al., 2010). Hydrogels have also been
extract of air-dried substrate to distilled water) were determined
found to have contrasting effects on plant growth and yield (Akhter
from samples taken from moist pots that were left in glasshouse
et al., 2004; Geesing and Schmidhalter, 2004).
conditions for one week. The substrates were tested for air-lled
Hydrogels have recently been evaluated to determine if they
porosity, water-holding capacity and bulk density using Australian
improve plant growth on green roofs (Oschmann et al., 2007;
Standard Methods AS 3743-2003 (Standards Australia, 2003) for
Sutton, 2008; Olszewski et al., 2010). Olszewski et al. (2010) found
potting mixes. A known volume of the mix was immersed in water
that hydrogel added at 2.2 kg m3 to expanded slate-based sub-
to soak and drained for three repeated cycles to ensure a full satu-
strate increased initial shoot growth and cover of two Sedum spp.
ration and left for a nal draining for 30 min before being removed.
However, in the same experiment, the addition of 20% compost to
The volume of the drained mix was measured and expressed as
the expanded slate-based substrate increased shoot growth by an
a percentage of the initial total volume of mix to give air-lled
extra 37%, compared with the addition of hydrogel (Olszewski et al.,
porosity. The mix was then oven-dried at 105 C for one week and
2010). Sutton (2008) and Oschmann et al. (2007) found hydrogel
the total water holding capacity was calculated as the difference
increased plant growth of grasses and forbs in green roof substrates.
between the wet and dry masses.
Although cover and growth is a key indicator of green roof success,
Moisture release curves of scoria and roof-tile substrates with-
none of these studies determined whether water-retention addi-
out additives were determined using the lter paper method
tives increased PAW. Given most previous research has been under
(Greacen et al., 1989). The lter paper method has been widely used
agricultural or containerized production conditions and the effec-
for measuring soil water potential in agronomic research (Hamblin,
tiveness of water-retention additives for improving plant available
1981; Greacen et al., 1989) and has also been used in glasshouse
water varies in different soil/substrates (Akhter et al., 2004), more
studies (Doube et al., 1996). This method is useful for green roof
research is necessary to better understand their application in a
substrates, which are often gravelly and heterogeneous, as it uses
green roof context.
larger sample sizes than methods using pressure plates or psy-
This study evaluated the use of hydrogel and silicate-based
chrometers (Hamblin, 1981). Substrates were oven dried at 105 C
water-retention additives to increase plant available water in
until constant weight was achieved. Distilled water was added at
green roof substrates. The study determined the effects of water-
5% increments of water-holding capacity to 200 g samples (20 sam-
retention additives on: (1) the physical properties of substrates and
ples) of each substrate so that water-holding capacity ranged from
(2) plant available water, days to permanent wilting and biomass
5% to 100%. The substrate samples were then placed in glass jars
of two indicator plant species with contrasting root morphologies
with a lter paper (Whatman No. 42) inserted between half the vol-
and drought tolerance.
ume in each jar to ensure direct contact between the lter paper
and substrate. The jars were then placed in an airtight, constant
temperature container to equilibrate for one week, after which the
2. Materials and methods
gravimetric water content of the lter paper (w) was determined.
The corresponding matric suction for each sample was calculated
2.1. Substrate preparation and analysis
using equations (Greacen et al., 1989) below, and the values plotted
against the percentage water holding capacity of each substrate by
Two green roof substrates were prepared for the experiment;
weight to give moisture release curves for each substrate.
one substrate mix was based on scoria, the other based on crushed
terracotta roof-tiles. Both substrates were developed according to
the green roof FLL guidelines (FLL, 2008) as part of a larger research For w < 0.453 g g1 : matric suction = exp (12.2717.93w)
program into green roofs for Australian conditions. The substrate For w > 0.453 g g1 : matric suction = exp (5.553.11w)
compositions are listed in Table 1. The two substrates differ in
particle size with the scoria containing 20% 7 mm aggregate in 2.2. Drought experiment
addition to the <8 mm fraction which includes nes; whereas, the
roof-tile substrate does not contain additional aggregate and is The effects of water retention additives on PAW, days to perma-
made up entirely of <7 mm fraction including nes (Table 1). Two nent wilting and biomass were evaluated in a glasshouse drought
different water-retentive additives: a polyacrylamide hydrophilic experiment using winter wheat (Triticum aestivum) and white lupin
polymer or hydrogel (SaturAid , Debco, Australia) and a silicon- (Lupinus albus). Wheat and lupin were chosen because of their abil-
based additive (Sanoplant , Sanoway) were added to the green ity to establish themselves within a short time period and their
 Author's personal copy
114 C. Farrell et al. / Ecological Engineering 52 (2013) 112118
lack of strong physiological adaptations to drought, unlike many
succulent green roof species. Determination of PWP is often done
on wheat in agricultural experiments with water-retention addi-
tives, enabling useful comparisons between applications (Akhter
et al., 2004; Geesing and Schmidhalter, 2004). The two species also
differ in drought tolerance and root morphologies, enabling the
effects of the additives on different plants species to be compared.
Wheat is less drought-tolerant than lupin (Brink and Belay, 2006)
and they have contrasting root morphologies; wheat has a dense
brous root system (Sayyad et al., 2010) and lupin has a branching
tap root (Bramley et al., 2010).
The experiment was a factorial design with two different sub-
strates, three additive treatments (none, hydrogel and silicates) and
two plant species. There were ten replicates for each treatment.
Pots were arranged in the glasshouse in a complete randomized
block design. Pots were 12 cm deep, had a total volume of 339 mL
and the mass of substrate in each pot was, on average, 333 g of
scoria or 429 g of crushed roof-tile mix. Although the two sub-
strate differed by mass, the volume was equivalent. This approach
was chosen as it most closely resembles green roof construction
where substrates are added by depth. Three seeds of either wheat
or lupin were sown per pot on 15 July 2010 and were thinned to
one seedling 2 weeks post germination. Plants were grown in a
controlled temperature glasshouse at the Burnley Campus, Uni-
versity of Melbourne, with mean daily temperature ranging from
14 C to 30 C, average relative humidity 59% and a daily photope-
riod of 10:14 h (L:D). Plants were irrigated to pot capacity with an
overhead sprinkler until well-establishment and root systems were
well-developed, 45 days post sowing.
On day 46, irrigation ceased and all plants were exposed to
drought conditions. To prevent water loss from evaporation, pots
were sealed at the container surface with two layers of transparent
polyethylene wrap, leaving 2 cm of air space above the substrate
surface for gaseous exchange. The glasshouse was also shaded (30%
reduction) to reduce transpiration and to ensure root temperatures
did not increase as a result of covering the pots with polyethylene
wrap. Plants were monitored daily for the rst sign of wilting: loss
of turgidity in the third lowest basal tiller of wheat and the sec-
ond lowest basal leaves of lupin. Once wilting was observed, the
whole pot was placed in a dark humid chamber overnight to ensure
wilting was permanent. If plants recovered turgor the next day,
they were transferred back into the glasshouse to resume growth
until wilting was again observed. This process was repeated until
no recovery was observed after an overnight period in the humid
chamber, thus permanent wilting was achieved (Veihmeyer and
Hendrickson, 1949). Pot weights were recorded daily to determine
substrate water contents and transpiration losses until permanent
wilting. At the permanent wilting point, plants were harvested,
separated into roots and shoots and oven dried at 80 C oven until
constant weight to determine dry weights.
Plant available water (%) was determined by calculating the
change in substrate water content from pot capacity (just prior to
drought commencing) to permanent wilting, divided by substrate
volume (339 mL) and then multiplied by 100%.
2.3. Statistical analyses
The effect of water retention additives on the physical and
chemical properties (water-holding capacity, air-lled porosity,
pH, EC and bulk density) of substrates were analysed using
two-way analysis of variance (ANOVA). Two-way ANOVA within
plant species (wheat or lupin) was also used to determine the
effects of water-retention additives and substrates on PAW, num-
ber of days taken to reach permanent wilting point and plant
biomass. Data were checked for normality and were transformed
100.000
10.000
Matric Suction (MPa)
1.000
0.100
0.010
0.001
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Substrate water content (% of total WHC)
Fig. 1. Moisture release curves for the two green roof substrates (crushed roof-tile
mix = dashed line; scoria mix = dotted line) without water-retention additives.
appropriately where necessary prior to analysis. All data presented
in gures and tables are non-transformed. Signicant differences
were determined by Tukeys post hoc test (P < 0.05). All data
analyses were performed in GenStat 14 (2011, VSN International
Ltd.).
3. Results
3.1. Substrate physical and chemical properties.
The air-lled porosity (AFP) and bulk density of the two
substrates with no water retention additives differed (Table 2).
However the two substrates had similar water holding capacities
(WHC), between 40 and 43%. The AFP of the scoria substrate was
16%, almost 3 times higher than the roof-tile AFP. The lower AFP
of the roof-tile substrate also corresponded with an increased bulk
density, compared with the scoria substrate.
When additives were added to substrates, there was a signi-
cant interaction between the effects of substrate and additives on
WHC (P < 0.001), AFP (P < 0.009) and bulk density (P < 0.001). Sili-
cates signicantly increased the WHC of both substrates, although
the effects were approximately 5% greater in scoria substrate when
compared with crushed roof-tile substrate. Hydrogel increased the
WHC of scoria relative to the control, but had no effect on the WHC
of roof-tile. Additives had no effect on AFP in the roof-tile substrate,
while silicates caused lower AFP in scoria. The addition of hydrogel
and silicates increased the bulk density of the roof-tile substrate
but had no effect on scoria.
Water retention additives had no effect on pH or salinity (EC1:5 )
of the two green roof substrates (data not shown). The mean
pH of both mixes ranged between 7.5 and 8.3 and salinity was
0.2 dS m1 .
3.2. Substrate water availability and the effect of water retention
additives on PAW and days until permanent wilting
Water availability, as shown by the moisture release curves,
differed in the two substrates without water-retention additives
(Fig. 1). Water was less available in scoria than in roof-tile growing
media when water content dropped below 30% WHC.
The effect of water-retention additives on plant available water
(PAW) differed between substrates and plant species (Table 3).
For wheat, PAW was signicantly lower in scoria than in roof-tile
(P < 0.001) but PAW was higher when hydrogel was added to both
substrates compared with control treatments (P = 0.013). There was
 Author's personal copy

C. Farrell et al. / Ecological Engineering 52 (2013) 112118 115

Table 2
Effect of water retention additives on physical properties (water holding capacity (WHC), air-lled porosity (AFP) and bulk density) of two green roof substrates (scoria mix
and crushed roof-tile mix). None refers to no additives, silicates refers to silicate based granules added at a rate of 10 g L1 and hydrogel refers to hydrophyllic polyacrylamide
gel added at a rate of 1 g L1 . Mean SE (n = 3) (2-way ANOVA; interaction P-value shown).

Scoria Roof-tile

None Hydrogel Silicates None Hydrogel Silicates P-value

WHC (%) 40.12a (0.22) 44.86c (0.11) 49.56d (0.55) 43.54a (0.76) 41.94ab (0.51) 45.35c (0.08) <0.001
AFP (%) 15.94c (1.13) 13.19c (0.46) 9.58b (0.39) 5.41a (0.14) 4.75a (0.32) 3.49a (0.40) 0.009
Bulk density (kg L1 ) 0.83a (0.003) 0.83a (0.003) 0.82a (0.003) 1.16b (0.01) 1.21c (0.01) 1.20c (0.00) <0.001

Different letters denote signicant differences between additives and substrate.

Table 3
Effect of water-retention additives on mean (SE) plant available water (PAW) and days until permanent wilting point of wheat and lupin growing in two different green
roof substrates (scoria-mix and crushed roof-tile mix). None refers to substrates with no additives, silicates refers to silicate based granules added at a rate of 10 gL1 and
hydrogel refers to hydrophyllic polyacrylamide gel added at a rate of 1 gL1 .

Scoria Roof-tile P-values

None Hydrogel Silicates None Hydrogel Silicates Additive Substrate Interaction

PAW (%) Wheat 21.67 Aa(0.9) 27.49 Ba(1.68) 27.5 ABa(1.99) 29.15 Ab(1.0) 30.23 Bb(2.68) 34.75 ABb(2.29) 0.013 <0.001 0.364
Lupin 21.43 a (0.6) 23.18 a (1.54) 23.61 a (0.83) 27.88 b(0.98) 33.13 c (1.04) 35.10 c (0.69) <0.001 <0.001 <0.042
Days to PWP Wheat 8.4 a (0.4) 10.0 a (0.4) 9.6 a (0.5) 11.8 b (0.3) 11.9 b (0.9) 12.1 b (0.7) 0.139 <0.001 0.312
Lupin 8.35 Aa (0.6) 9.6 ABa (1.0) 10.4 Ba (1.0) 9.9 Ab (0.7) 11.1 ABb (0.6) 13.0 Bb (1.2) 0.006 0.007 0.759

Uppercase letters denote differences between additives and lower case letters denote signicant differences between substrates within plant species (2-way ANOVA; n = 12).
Where there is a signicant interaction, lowercase letters denote signicant differences between all means within species.

no signicant difference in PAW between the two additives. For
lupin, there was an interactive effect of substrate and additives
on permanent wilting point (PWP) (P 0.042), with signicantly
higher PAW in the roof-tile substrate with either hydrogel or sili-
cates added when compared with scoria. PAW was not signicantly
different between additives in either substrate.
There was no signicant effect of additives on the number of
days taken to reach permanent wilting point after the onset of
drought for wheat in either substrate (Table 3, P = 0.139). How-
ever, wheat plants grown in roof-tile substrate maintained turgor
approximately 2 days longer than in scoria (P < 0.001). For lupins,
the addition of silicates increased the number of days to reach
PWP by 3 days in roof-tile and 2 days in scoria, compared with
substrates without additives (P = 0.006). There was no signicant
difference between days to PWP of lupins growing in substrates
with- or without-hydrogel. The number of days it took lupins to
reach PWP was up to 3 days longer in roof-tile substrate compared
with scoria (P = 0.007).
3.3. Effect of water retention additives and substrate on plant
biomass
The addition of water-retention additives increased the total
and root biomass of both wheat and lupin but had no signicant
effect on shoot biomass (Table 4). In wheat grown in scoria
with added silicates, total biomass was at least 2-fold greater
(interaction P = 0.029) and root biomass was at least 78% greater
(interaction P = 0.022) than in any other substrate or additive treat-
ment. In lupin, both total and root biomass were greater in scoria
(total P = 0.017; root P = 0.006) and the addition of silicates to
either substrate increased total biomass by at least 40% (P = 0.017)
and root biomass by 90% (P = 0.006) relative to hydrogel or the
control. Root:shoot ratios (Table 4) reected the results for root
biomass, with 4-fold higher root:shoot ratios for wheat grow-
ing in scoria with added silicates (interaction P = 0.008). In lupin,
root:shoot ratios were greater in scoria substrate compared with
roof-tile (P = 0.003) and where substrates included added silicates
(P < 0.001).
4. Discussion
This study shows that water-retention additives can be used to
increase the water holding capacity of green roof substrates but
they behave differently in different substrates. Silicates increased
the water holding capacity of both scoria and roof-tile substrates,
while hydrogel only increased the WHC of scoria. The effect of
hydrogel on WHC was most likely related to substrate particle size,
with increased WHC in coarse-textured scoria versus no increase

Table 4
Effect of water-retention additives (no additives, hydrogel and silicates) and green roof substrates on plant biomass (root, shoot and total) and root:shoot ratio of winter
wheat and lupin.

Scoria Roof-tile P-values

None Hydrogel Silicates None Hydrogel Silicates Additive Substrate Interaction

Root (g) Wheat 0.91a (0.13) 0.88a (0.13) 1.23b (0.27) 1.82 a(0.22) 1.35a (0.18) 3.25a (0.53) <0.001 <0.001 0.022
Lupin 0.37Aa (0.03) 0.48Aa (0.06) 0.92Ab (0.11) 0.79 Ba(0.23) 0.73Ba (0.16) 1.4Bb (0.25) <0.001 0.006 0.765
Shoot (g) Wheat 0.69 (0.05) 0.81(0.07) 0.65 (0.09) 0.78(0.06) 0.76 (0.04) 0.77 (0.04) 0.514 0.291 0.320
Lupin 0.54 (0.05) 0.63 (0.04) 0.57 (0.04) 0.58(0.05) 0.58 (0.06) 0.58 (0.04) 0.554 0.965 0.613
Total (g) Wheat 1.60a (0.17) 1.69a (0.19) 1.88b (0.35) 2.60 a(0.28) 2.11a (0.19) 4.03a (0.56) 0.004 <0.001 0.029
Lupin 0.90Aa (0.07) 1.11Aa (0.09) 1.49Ab (0.13) 1.36 Ba(0.26) 1.31Ba (0.22) 1.99Bb (0.28) 0.005 0.017 0.709
Root:shoot (g) Wheat 1.31b (0.14) 1.05ab (0.09) 1.71c (0.19) 2.28 ab(0.17) 1.83a (0.31) 4.14ab (0.52) <0.001 <0.001 0.008
Lupin 0.69Aa (0.05) 0.76Aa (0.09) 1.66Ab (0.23) 1.31Ba (0.32) 1.13Ba (0.17) 2.31Bb (0.29) <0.001 0.003 0.776

Different letters denote signicant differences between biomass means within species (2-way ANOVA; n = 12). Where there is a signicant interaction, lowercase letters
denote signicant differences between all means within species. Where there is no signicant interaction uppercase letters show signicant differences between substrate
and lowercase between additives.
 Author's personal copy
116 C. Farrell et al. / Ecological Engineering 52 (2013) 112118
in ne-textured roof-tile substrate. This is consistent with research
in soils, where hydrogel increases WHC of sand to a greater degree
than in clay soils, as ner textured soils limit the expansion and
therefore water absorption of hydrogel polymers (Abedi-Koupai
et al., 2008; Mamedov et al., 2011). The increase in WHC in scoria
with added hydrogel was not associated with a change in poros-
ity or bulk density. However, although hydrogel increased bulk
density in roof-tile substrate, WHC and porosity were unchanged.
Generally, hydrogels are thought to increase permeability of media
and reduce bulk density due to swelling when hydrated (Baasiri
et al., 1986; Han et al., 2010). In our study, the absence of an
affect of hydrogels on porosity in green roof substrates may have
been due to differences in application rate or hydrogel polymer
size. Olszewski et al. (2010) found that porosity only increased in
expanded shale green roof substrate when hydrogel was applied at
a rate of 2.2 kg m3 , more than double the application rate in our
study. However, Fonteno and Bilderback (1993) found application
rate had no affect on porosity of either pine bark or sand-pine bark
media.
Increased WHC of substrates with added silicates was associated
with a change in bulk density and porosity. Other silicate-based soil
amendments such as zeolite and diatomaceous earth have been
shown to increase water holding capacity by increasing internal
surface area for water retention and reducing drainage (Githinji
et al., 2011). It is likely that the silicate additive used in our study
behaved similarly, as increased WHC was associated with a change
in bulk density. In the roof-tile substrate, silicates increased bulk
density but bulk density was reduced in scoria. The measured
reduction in bulk density in our study could have be due to sili-
cates lling internal pore spaces as scoria has high internal porosity
(Wallach et al., 1992).
Although additives increased substrate WHC, the additional
stored water was not always plant available. This is consistent
with other studies where hydrogel increased WHC but not PAW in
Orthosiphon aristatus (Nguyen et al., 2009) and Lingustrum japon-
icum (Ingram and Yeager, 1987). In our study silicates did not
increase PAW in scoria despite increasing WHC. Hydrogel on the
other hand increased both WHC and PAW in scoria, however the
effects differed between plant species with increased PAW for
wheat but not lupin. In roof-tile substrate both additives increased
PAW for wheat and lupin, despite hydrogel having no effect on WHC
of roof-tile substrate. These differences in WHC and PAW between
substrates are likely to have been exacerbated by the inherent dif-
ferences in water availability between the two substrates. Despite
similar WHC between the two substrates, water is less available
at higher substrate water contents in scoria than for roof-tile sub-
strate, due to the high internal porosity of scoria (Wallach et al.,
1992).
Increased PAW for wheat in hydrogel amended scoria and in
roof-tile substrate amended with either additive did not increase
time to wilting. Jobin et al. (2004) also found that increased PAW
in pine bark substrate did not increase the time until wilting in
Petunia hybrida. For hydrogel-amended substrates it is possible
that the additional stored water was restricted to roots which
aggregated around or penetrated hydrogel polymers. Roots were
observed in the hydrogel polymers in our study and root penetra-
tion of hydrated polymers has been suggested increase PAW and
time to wilting in lettuce and barley seedlings (Woodhouse and
Johnson, 1991) and in Populus popularis cuttings (Shi et al., 2010).
Restricted access to stored water would not have signicantly
improved overall plant water status and time until permanent wil-
ting. In contrast, silicates increased the number of days until wilting
of lupins by two days in scoria and three days in roof-tile substrate.
Silicates have been shown to improve PAW by lling mesopores
and reducing water loss through drainage (Hosseini et al., 2012),
consequently in our study it is likely that improved water availabil-
ity would have been throughout the pot and not localized around
roots as found for hydrogel (Woodhouse and Johnson, 1991). Dif-
ferences in the response of wheat and lupins to additives are also
likely due to differences in root morphology. As wheat has dense
brous roots and had a greater root biomass than lupin it would
have exhausted the available water faster than lupin which has a
coarse, branching taproot root system. This may also explain why
hydrogel did not increase the number of days until wilting in lupin
in roof-tile substrate, as fewer roots would result in reduced contact
with hydrogel polymers in the substrate.
Silicates increased plant growth (total and root biomass) in the
two green roof substrates, whereas hydrogel had no effect on plant
biomass. Both additives had no effect on either wheat or lupin
shoot growth. The effects of hydrogel on plant growth and yield
are often contrasting, even within the same species and substrate.
For example, in wheat grown in sandy loam with 0.3% hydrogel,
Akhter et al. (2004) reported increased growth whereas Geesing
and Schmidhalter (2004) reported no growth effect. We applied
hydrogel at the manufacturers recommended rate of 0.1% and this
may have been too low to affect growth. While other studies have
shown 0.1% additions of hydrogel increased plant growth of let-
tuce and barley (Woodhouse and Johnson, 1991) and cucumber
(Al-Harbi et al., 1999), rates less than 0.3% have been shown to
have no effect on growth of Agrostis stolonifera (Agaba et al., 2011),
Orthosiphon aristatus (Nguyen et al., 2009) and Conocarpus erectus
(Al-Humaid and Moftah, 2007). In a green roof context, Oschmann
et al. (2007) also found that 0.3% hydrogel application increased
cover of grasses and forbs to a greater extent than 0.1% hydro-
gel. However, while increased application rates generally increase
plant growth, the relationship is not linear and over-application
can reduce yield (Geesing and Schmidhalter, 2004). Furthermore,
the benets of increased growth with higher application rates need
to be evaluated against increased application costs.
In addition to application rate, the longevity and effectiveness
over time of these two water-retention additives was not assessed
in our study as we were interested in plant available water dur-
ing initial plant establishment. While both manufacturers state
that their additive life-spans are ve and ten years, respectively,
this has not been evaluated under green roof conditions. Al-Harbi
et al. (1999) found the effects of hydrogel on substrate water avail-
ability reduced linearly at a rate of 1015% per year, while Jobin
et al. (2004) found no benet to PAW after 9 weeks in potted
Petunia hybrida. Under eld conditions, the longevity and per-
formance of additives has been shown to be reduced by high
temperatures (Baasiri et al., 1986), repeated drying and wetting
cycles (Geesing and Schmidhalter, 2004; Bai et al., 2010) and micro-
bial degradation (Holliman, 2005). On green roofs, where substrate
temperatures are often extreme and water availability uctuates
dramatically between rainfall events (Oberndorfer et al., 2007), the
longevity of water-retention additives is likely to be limited and
further research is needed to determine their effectiveness beyond
initial plant establishment.
Water-retention additives, particularly silicates which
increased growth and time until wilting could prove useful
to increase the range of plant species suitable for green roofs. It is
unlikely that these additives will have any benet to existing green
roof succulents as Farrell et al. (2012) found that several Sedum
spp. could survive more than 100 days without water in scoria
and roof-tile green roof substrates. However, water-retention
additives could expand plant selection to include less drought
tolerant species that use more water, increasing the stormwater
management and cooling potential of green roofs. Increased
substrate WHC with additives could also increase the volume
of stormwater retained (VanWoert et al., 2005a; Villarreal and
 Author's personal copy
C. Farrell et al. / Ecological Engineering 52 (2013) 112118
Bengtsson, 2005; Mentens et al., 2006) and increasing cooling
due to greater latent heat conversion from evaporation (Carter
and Rasmussen, 2006). These new applications of water-retention
additives in green roof substrates would be useful avenues for
future research.
5. Conclusions
The benets of water-retention additives on plant available
water, drought stress and plant growth in green roof substrates
is clearly substrate and plant species specic. At recommended
application rates, silicates performed better than hydrogel in these
two green roof substrates. Although hydrogels increased substrate
WHC and PAW there was no increase in time until wilting in either
substrate. Silicates on the other hand, increased time until wil-
ting and total and root biomass for wheat growing in scoria and
for lupins in either substrate. Therefore silicates are more likely to
assist in plant establishment on green roofs. However, as high heat
loads from solar radiation and exposure on green roofs may affect
additive performance and longevity further research is needed to
evaluate their long term effectiveness.
Acknowledgements
We thank the Burnley nursery staff Nick Osborne and Sascha
Andrusiak for technical assistance during the experiment. We
also thank Nick Williams and Chris Szota for useful feedback on
this manuscript. This research was funded by The University of
Melbourne and the Australian Research Council Linkage Grant
LP0990704 supported by the Victoria Department of Sustainability
and Environment, Melbourne Water, City of Melbourne, and The
Committee for Melbourne.
References
Abedi-Koupai, J., Sohrab, F., Swarbrick, G., 2008. Evaluation of hydrogel application
on soil water retention characteristics. J. Plant Nutr. 31, 317331.
Agaba, H., Baguma Orikiriza, L.J., Osoto Esegu, J.F., Obua, J., Kabasa, J.D., Httermann,
A., 2010. Effects of hydrogel amendment to different soils on plant available
water and survival of trees under drought conditions. Clean-Soil Air Water 38,
328335.
Agaba, H., Orikiriza, L.J.B., Obua, J., Kabasa, J.D., Worbes, M., Httermann, A., 2011.
Hydrogel amendment to sandy soil reduces irrigation frequency and improves
the biomass of Agrostis stolonifera. Agric. Sci. 2, 544550.
Akhter, J., Mahmood, K., Malik, K., Mardan, A., Ahmad, M., Iqbal, M., 2004. Effects
of hydrogel amendment on water storage of sandy loam and loam soils and
seedling growth of barley, wheat and chickpea. Plant Soil Environ. 50, 463469.
Al-Harbi, A., Al-Omran, A., Shalaby, A., Choudhary, M., 1999. Efcacy of a hydrophilic
polymer declines with time in greenhouse experiments. HortScience 34,
223224.
Al-Humaid, A.I., Moftah, A.E., 2007. Effects of hydrophilic polymer on the survival
of buttonwood seedlings grown under drought stress. J. Plant Nutr. 30, 5366.
Apostol, K.G., Jacobs, D.F., Dumroese, R.K., 2009. Root desiccation and drought stress
responses of bareroot Quercus rubra seedlings treated with a hydrophilic poly-
mer root dip. Plant Soil 315, 229240.
Baasiri, M., Ryan, J., Mucheik, M., Harik, S., 1986. Soil application of a hydrophilic con-
ditioner in relation to moisture, irrigation frequency and crop growth. Commun.
Soil Sci. Plant Anal. 17, 573589.
Bai, W., Zhang, H., Liu, B., Wu, Y., Song, J., 2010. Effects of super-absorbent polymers
on the physical and chemical properties of soil following different wetting and
drying cycles. Soil Use Manag. 26, 253260.
Bass, B., Baskaran, B., 2003. Evaluating rooftop and vertical gardens as an adapta-
tion strategy for urban areas. Institute for Research and Construction, National
Research Council, Canada, Ottawa, p. 111.
Berndtsson, J.C., 2010. Green roof performance towards management of runoff water
quantity and quality: a review. Ecol. Eng. 36, 351360.
Bramley, H., Turner, N.C., Turner, D.W., Tyerman, S.D., 2010. The contrasting inu-
ence of short-term hypoxia on the hydraulic properties of cells and roots of
wheat and lupin. Funct. Plant Biol. 37, 183193.
Brenneisen, S., 2006. Space for urban wildlife: designing green roofs as habitats in
Switzerland. Urban Habitat. 4, 2736.
Brink, M., Belay, G. (Eds), 2006. Plant Resources of Tropical Africa. 1. Cereals and
Pulses. PROTA Foundation, Wageningen.
117
Carter, T.L., Rasmussen, T.C., 2006. Hydrologic behaviour of vegetated roofs. J. Am.
Water Resour. Assoc. 42, 12611274.
Dorraji, S.S., Golchin, A., Ahmadi, S., 2010. The effects of hydrophilic polymer and
soil salinity on corn growth in sandy and loamy soils. Clean-Soil Air Water 38,
584591.
Doube, B.M., Williams, P.M.L., Willmott, P.J., 1996. The inuence of two species of
earthworm (Aporrectodea trapezoides and Aporrectoedea rosea) on the growth
of wheat, barley and faba beans in three soil types in the greenhouse. Soil Biol.
Biochem. 29, 503509.
Farrell, C., Mitchell, R.E., Szota, C., Rayner, J.P., Williams, N.S.G., 2012. Green roofs
for hot and dry climates: interacting effects of plant water use, succulence and
substrate. Ecol. Eng. 49, 270276.
FLL, 2008. Guidelines for the planning, construction and maintenance of green roof-
ing green roong guidleine. Forschungsgesellschaft Landschaftsentwicklung
Landschaftsbau.
Fonteno, W., Bilderback, T., 1993. Impact of hydrogel on physical properties of
coarse-structured horticultural substrates. J. Am. Soc. Hortic. Sci. 118, 217222.
Geesing, D., Schmidhalter, U., 2004. Inuence of sodium polyacrylate on the water-
holding capacity of three different soils and effects on growth of wheat. Soil Use
Manag. 20, 207209.
Gehring, J., Lewis III, A., 1980. Effect of hydrogel on wilting and moisture stress of
bedding plants. J. Am. Soc. Hortic. Sci. 105, 511513.
Githinji, L.J.M., Dane, J.H., Walker, R.H., 2011. Physical and hydraulic properties of
inorganic amendments and modeling their effects on water movement in sand-
based root zones. Irrigat. Sci. 29, 6577.
Greacen, E.L., Walker, G.R., Cook, P.G., 1989. Procedure for lter paper method
of measuring soil water suction. In: Division of Soils Report No. 108. CSIRO,
Australia.
Hamblin, A.P., 1981. Filter-paper method for routine measurement of eld water
potential. J. Hydrol. 53, 355360.
Han, Y.G., Yang, P.L., Luo, Y.P., Ren, S.M., Zhang, L.X., Xu, L., 2010. Porosity change
model for watered super absorbent polymer-treated soil. Environ. Earth Sci. 61,
11971205.
Holliman, P.J., 2005. Model and eld studies of the degradation of cross-linked poly-
acrylamide gels used during the revegetation of slate waste. Sci. Total Environ.
336, 13.
Hosseini, M.R., Haile, A.M., McClain, M.E., 2012. Improving agricultural production
under water scarcity in Fars province, Iran. Agro Environ.
Httermann, A., Zommorodi, M., Reise, K., 1999. Addition of hydrogels to soil for
prolonging the survival of Pinus halepensis seedlings subjected to drought. Soil
Till. Res. 50, 295304.
Ingram, D.L., Yeager, T.H., 1987. Effects of irrigation frequency and a water-absorbing
polymer amendment on Ligustrum growth and moisture retention by a con-
tainer medium. J. Environ. Hort. 5, 1219.
Jobin, P., Caron, J., Bernier, P.Y., Dansereau, B., 2004. Impact of two hydrophilic
acrylic-based polymers on the physical properties of three substrates and the
growth of petunia hybrida Brilliant Pink. J. Am. Soc. Hortic. Sci. 129, 449457.
Johnson, M.S., 1984. The effects of gel-forming polyacrylamides on moisture storage
in sandy soils. J. Sci. Food Agric. 35, 11961200.
Koupai, J.A., Eslamian, S.S., Kazemi, J.A., 2008. Enhancing the available water con-
tent in unsaturated soil zone using hydrogel, to improve plant growth indices.
Ecohydrol. Hydrobiol. 8, 6775.
Mamedov, A., Shi, J., Levy, G., Yu, J., Yan, Y., Shainberg, I., 2011. Superabsorbents
and Semiarid Soil Properties Affecting Water Absorption. Soil Sci. Soc. Am. J. 75,
23052313.
Mentens, J., Raes, D., Hermy, M., 2006. Green roofs as a tool for solving the rainwater
runoff problem in the urbanized 21st century? Landsc. Urban Plan. 77, 217226.
Nagase, A., Dunnett, N., 2010. Drought tolerance in different vegetation types for
extensive green roofs: Effects of watering and diversity. Landsc. Urban Plan. 97,
318327.
Nguyen, T.T., Joyce, D.C., Dinh, S.Q., 2009. Effects of articial amendments in pot-
ting media on Orthosiphon aristatus growth and development. Sci. Hortic. 123,
129136.
Oberndorfer, E., Lundholm, J., Bass, B., Coffman, R.R., Doshi, H., Dunnett, N., Gaf-
n, S., Kohler, M., Liu, K.K.Y., Rowe, B., 2007. Green roofs as urban ecosystems:
Ecological structures, functions, and services. Bioscience 57, 823833.
Olszewski, M.W., Holmes, M.H., Young, C.A., 2010. Assessment of physical proper-
ties and stonecrop growth in green roof substrates amended with compost and
hydrogel. HortTechnology 20, 438444.
Oschmann, C., Kobayashi, N., Perkuhn, C., Grneberg, H., Wissemeier, A., 2007. Study
to Expand the Range of Wild Plants for Extensive Roof Greening Systems using
Superabsorbent Polymers (SAP). Acta Hort. (ISHS) 813, 421426.
Sailor, D.J., 2008. A green roof model for building energy simulation programs. Energ.
Build. 40, 14661478.
Sanoway Australia, 2006. SANOPLANT. Vol. 2010, Australia. Available at:
http://www.sanoway.com.au/sanoplant.overview.ews (accessed 2.11.10).
Sayyad, G., Afyuni, M., Mousavi, S.-F., Abbaspour, K.C., Richards, B.K., Schulin,
R., 2010. Transport of Cd, Cu, Pb and Zn in a calcareous soil under
wheat and safower cultivation a column study. Geoderma 154,
311320.
Shi, Y., Li, J., Shao, J., Deng, S., Wang, R., Li, N., Sun, J., Zhang, H., Zhu, H., Zhang, Y., 2010.
Effects of Stockosorb and Luquasorb polymers on salt and drought tolerance of
Populus popularis. Sci. Hortic. 124, 268273.
Sivapalan, S., 2006. Benets of treating a sandy soil with a cross-linked type poly-
acrylamide. Aust. J. Exp. Agric. 46, 579584.
 Author's personal copy
118 C. Farrell et al. / Ecological Engineering 52 (2013) 112118
Slatyer, R., 1957. The signicance of the permanent wilting percentage in studies of
plant and soil water relations. Bot. Rev. 23, 585636.
Snodgrass, E.C., Snodgrass, L.L., 2006. Green Roof Plants: A Resource and Planting
Guide. Timber Press, Portland, OR, p. 203.
Spomer, L.A., 1980. Container soil water relations: Production, maintenance, and
transplating. J. Arboricult. 6, 315320.
Standards Australia, 2003. AS 3743-2003: Australlian Standards for potting mixes.
Sutton, R.K., 2008. Media Modications for Native Plant Assemblages on Green Roofs.
VanWoert, N.D., Rowe, B., Andresen, J.A., Rugh, C.L., Fernandez, R.T., Xiao, L., 2005a.
Green roof stormwater retention: effects of roof surface, slope, and media depth.
J. Environ. Qual. 34, 10361044.
VanWoert, N.D., Rowe, D.B., Andresen, J.A., Rugh, C.L., Xiao, L., 2005b. Watering
regime and green roof substrate design affect Sedum plant growth. HortScience
40, 659664.
Veihmeyer, F.J., Hendrickson, A.H., 1949. Methods of measuring eld capacity and
permanent wilting percentage of soils. Soil Sci. 68, 7594.
Villarreal, E.L., Bengtsson, L., 2005. Response of a Sedum green-roof to individual
rain events. Ecol. Eng. 25, 17.
Wallach, R., Da Silva, F., Chen, Y., 1992. Hydraulic characteristics of tuff (scoria) used
as a container medium. J. Am. Soc. Hortic. Sci. 117, 415421.
Wang, Y.T., Boogher, C.A., 1987. Effect of a medium-incorporated hydrogel on plant
growth and water use of two foliage species. J. Environ. Hortic. 5, 125127.
Wolf, D., Lundholm, J.T., 2008. Water uptake in green roof microcosms: Effects of
plant species and water availability. Ecol. Eng. 33, 179186.
Woodhouse, J., Johnson, M., 1991. Effect of superabsorbent polymers
on survival and growth of crop seedlings. Agric. Water Manage. 20,
6370.