You are on page 1of 9

Published August 28, 2015

Journal of Environmental Quality

Special Section
The Urban Forest and Ecosystem Services

Soil Carbon and Carbon/Nitrogen Ratio Change under Tree Canopy,


Tall Grass, and Turf Grass Areas of Urban Green Space
S. J. Livesley,* A. Ossola, C. G. Threlfall, A. K. Hahs, and N. S. G. Williams
Abstract
Soils in urban green spaces are an important carbon (C) store,
but urban soils with a high carbon to nitrogen (C/N) ratio can
also buffer N eutrophication from fertilizer use or atmospheric
deposition. The influence of vegetation management practices
on soil C cycling and C/N ratios in urban green spaces is largely
unknown. In 2013, we collected replicate (n = 3) soil samples
from tree canopy, tall grass, and short turf grass areas (n = 3)
at four random plot locations (n = 4) established in 13 golf
courses (n = 13). At each sample point, soil was separated into
0- to 0.1-, 0.1- to 0.2-, and 0.2- to 0.3-m depths (total n = 1404).
Linear mixed models investigated the relationships between
soil properties, vegetation attributes, and green space age. Tree
canopy soil was less compacted (1.07 g cm3) than grassy areas
(1.32 g cm3). Similarly, tree canopy soil had mean C/N ratios of
17.2, as compared with between 14.2 and 15.3 in grassy areas.
Soil properties in tree canopy areas were best explained by tree
basal area and understory vegetation volume. Soil C/N increased
with increasing understory vegetation, and the difference in soil
C/N between tree canopy and short turf grass areas increased
over time. The soil properties in tree canopy areas of urban green
space mean they can increasingly buffer the localized use of N
fertilizers and atmospheric N deposition. Managers of urban
green spaces concerned about N pollution of groundwater and
waterways could consider planting trees in suitable topographic
locations and promoting understory vegetation and surface litter
accumulation.

Core Ideas
We sampled soil under tree canopy, tall grass, and short grass in
13 urban greenspaces.
Tree canopy soil had significantly greater soil C density and C/N
ratios.
Soil C/N ratios increased significantly with increasing green
space age.
Soil C and C/N were best modeled by tree basal area and understory vegetation.
Tree canopy areas can be managed to buffer nitrogen eutrophication.

Copyright 2015 American Society of Agronomy, Crop Science Society of America,


and Soil Science Society of America. 5585 Guilford Rd., Madison, WI 53711 USA.
All rights reserved.
J. Environ. Qual.
doi:10.2134/jeq2015.03.0121
Freely available online through the author-supported open-access option.
Received 3 Mar. 2015.
Accepted 21 May 2015.
*Corresponding author (sjlive@unimelb.edu.au).

he global population is becoming progressively


urbanized, with close to 3 billion additional people
expected to live in urban areas by 2050 (Seto et al.,
2013). The vegetation and soil that persist in the urban landscape
are increasingly under threat from development associated with
continued urbanization while being under pressure to provide
ecosystems services, such as microclimate cooling, biogeochemical cycling, pollutant mitigation, and community health and
well-being (Grimm et al., 2008). Urban centers are hot spots for
CO2 emissions and other industrial and transport-related pollution (Grimm et al., 2008; Kaye et al., 2006). The ability of urban
centers to mitigate their own C emissions through sequestration
in vegetation and soil is small (Hutyra et al., 2014; Livesley et al.,
2010; Pataki et al., 2011). However, an increase in soil C stores
in these green spaces can help provide other highly valued ecosystem services such as infiltration and storage of stormwater,
immobilization of nutrient or industrial pollutants, and support
for productive vegetation and food systems (Groffman et al.,
2009; Pataki et al., 2011; Setl et al., 2014).
The soil supporting urban trees and other vegetation is likely
to represent a greater C store than the aboveground vegetation
(Davies et al., 2011; Edmondson et al., 2014; Golubiewski, 2006),
as is the case in most other ecosystems (Field and Raupach, 2004).
The degree to which an urban green space (or any ecosystem) is
able to sequester C is determined by the climate (temperature
and rainfall), the soil type (e.g., percentage clay), management
practices, vegetation type, and productivity (Foley et al., 2003;
Pouyat et al., 2006). Predicting a net loss or gain in green space
soil C following urbanization is difficult because of the complex
management and environmental gradients in urban landscapes
(Ossola and Livesley, 2015; Pouyat et al., 2006). Management of
urban landscapes often leads to an increase in plant productivity
through irrigation and fertilizer additions but can also disrupt C
and nutrient cycling processes (e.g., turf is mown and removed,
turf thatch is scarified, tree leaf and branch litter is collected and
removed, or organic mulch is added to excess). Changes in urban
green space soil C can relate strongly to the number of years
since those urban green spaces were established and, therefore,
the duration of increased (or decreased) management intensity
(Golubiewski, 2006). However, the variation in soil C among
S.J. Livesley, A. Ossola, C.G. Threlfall, and N.S.G. Williams, School of Ecosystem
and Forest Sciences, Faculty of Science, The Univ. of Melbourne, 500 Yarra Blvd.,
Richmond, Victoria 3121, Australia; A.K. Hahs, Australian Research Centre for Urban
Ecology, Royal Botanic Gardens Melbourne, c/o School of BioSciences, The Univ. of
Melbourne, VIC 3010 Australia. Assigned to Associate Editor Gregory McPherson.
Abbreviations: LMM, linear mixed models.

urban green spaces may also relate to their vegetation characteristics, as is clearly evident for C stored in aboveground biomass
(Davies et al., 2011).
Urban soils mediate the balance between many ecosystem
services and disservices. Urban landscapes, and the soils that
support them, can accumulate heavy metals and other nutrient
chemicals through widespread atmospheric deposition or more
site-specific point and corridor pollution events. The source of
these heavy metals and nutrients may be local industry, vehicles
and transport corridors, and excess fertilizer use in public and
private green space (Carey et al., 2012; Milesi et al., 2005). If
the capacity of urban soils to buffer or retain these pollutants
is exceeded, the pollutants can percolate into the groundwater
or waterways with negative impacts on riparian ecology, local
human health, and food production (De Kimpe and Morel, 2000;
Groffman et al., 2009; Law et al., 2004; Valtanen et al., 2014).
The C/N ratio of a soil is a good indicator of its buffering capacity to retain nutrient pollutants such as nitrate, ammonium, and
phosphate (Dise et al., 1998). Soils with greater C/N ratio will
also have slower N ammonification and nitrification rates and
a greater ability to immobilize additional N inputs (Groffman
et al., 2006), although this capacity can vary according to soil
type and rainfall regime (Gundersen et al., 1998; Gurmesa et al.,
2013). Soil C/N ratios can change through (i) N decrease (plant
uptake), (ii) N increase (fertilizer addition), or (iii) a change in
the C quality of the organic matter inputs (higher or lower C/N
ratio) (Brady and Weil, 2008). Urban green space previously
dominated by grass or annual crops into which woody vegetation is then planted will receive organic matter of a high C/N
ratio and so will invariably develop soils of a greater C/N ratio
over time (Berthrong et al., 2009). It is also expected that trees
will take up more N and store it within their biomass rather than
the soil organic matter (Murty et al., 2002).
Urban green spaces take up a large area of most western cities,
and as they are often primarily designed for active and passive
recreation, they are dominated by intensely managed turf grass
(Milesi et al., 2005) interspersed with areas of tree or shrub cover.
The C storage of urban soils, especially under turf grass systems,
has been investigated in many previous studies (Edmondson et
al., 2012; Golubiewski, 2006; Raciti et al., 2011). However, there
have been few studies of soil C and soil C/N ratios in intensively
and extensively managed grass systems in comparison with that
in tree dominated urban green space. A golf course provides an
opportunity to investigate changes in soil properties under tree
canopy, tall grass, and short turf grass that have been consistently managed since the time it was established. By studying the
impact of these three different vegetation types on soil properties
within one site, the confounding effects of differing soil type and
differing green space age may be reduced.
This study uses a network of 13 urban green spaces on sandy
soils throughout southeast Melbourne, Australia to answer three
research questions:
1. Do urban green space soils under tree canopy have a greater
C/N ratio than soils under grass?
2. Is the variation in soil properties under tree canopy more
strongly related to green space age or vegetation attributes?
3. Does soil under grass store greater C than soil under tree
canopy in these green spaces?

Materials and Methods


Experimental Design
Melbourne is Australias second largest city with ~4 million
inhabitants. Greater Melbourne spans four bioregions, but this
study was restricted to the Gippsland Plains bioregion (sandy
soils supporting grassy woodland or heathland vegetation communities) to minimize variability in soil properties and vegetation types before urbanization. We randomly selected 13 golf
courses located in suburbs of different development age (1890s
to 2000s) (Fig. 1), as they provide access to green spaces with
detailed land-use and management history and are areas that
have been continuously managed since establishment. The
decade each golf course was established was determined by
examining historical aerial imagery and consulting municipal
land release and construction records. Within each golf course
we established four 20 by 30 m plots (total n = 52) in out-ofplay areas under trees using GIS analysis of digitized aerial photographs (ArcMap Version. 10.2.1, ESRI). These tree canopy
plots provided access to adjacent tall grass rough and short turf
grass fairway, but formal plots were not established in these grass
dominated areas.
The proportion of each golf course that was tree canopy,
tall grass, and short turf grass was manually digitized in GIS.
Similarly, the proportion of each tree canopy plot that was tree
canopy and grass was also determined through GIS analysis. This
enabled more accurate scaling up of soil C to the whole-of-greenspace level.

Vegetation Attributes and Structural Complexity


Within each tree canopy plot, we identified the species of
each tree that was >8 cm in diameter and measured tree stem
diameter at 1.3 m (diameter at breast height). From this, we were
able to calculate tree basal area (square meter per hectare) and

Fig. 1. Location of the 13 large green spaces (golf courses) sampled in


the southeastern suburbs of Melbourne.
Journal of Environmental Quality

tree stem density (trees per hectare). We also measured understory vegetation structure, recording vegetation forms that intercepted a 2.0-m-tall pole at 0-, 5-, 10-, 15-, 20-, 25-, and 30-m
point locations along four parallel 30-m-long transects (e.g., four
transects at seven locations equals 28 points per plot). Vegetation
was recorded if it intercepted the pole at one of five height intervals: 0.0 to 0.2, 0.2 to 0.5, 0.5 to 1.0, 1.0 to 2.0, and >2.0 m. The
sum of vegetation intercepts in each height interval was used to
calculate vegetation volume:
VVEGHX = (PNIHX/PTHX) VSHX
where VVEGHX is estimated vegetation volume occupying a specific height interval, PNIHX is the actual number of times vegetation intercepted the pole for that height interval, PTHX is the
number of pole point locations surveyed (usually 28 for a 600 m2
plot), and VSHX is the total volume for that height interval based
on the area of the plot multiplied by the height of the sampled
height band interval. To estimate total understory vegetation
volume, tree intercepts were discarded and then the VVEGHX for
each height interval were summed together.

Soil Sampling and Analysis


Soil was sampled to a depth of 0.3 m at three random locations in each tree canopy plot using a drop hammer and stainless steel core sampler (AMS Inc.; 50-mm i.d., 300-mm length).
Soil was not sampled within 1.0 m of a tree stem. For each plot,
the three soil samples were combined for the following three
depths: 0.0 to 0.1, 0.1 to 0.2, and 0.2 to 0.3 m. Soil was sampled
in the same manner from the two other vegetation types adjacent
to the tree canopy plot: tall grass (rough) and short turf grass
(fairway). Soil was not sampled within 1.0 m of the transition
from tall grass to short turf grass. In total, 1404 individual soil
samples were collected (three depths at three random points of
three vegetation types at four locations per golf course on 13 golf
courses). Additional surface soil samples (0.00.1 m) were also
taken in the tree canopy plot and adjacent tall grass rough using
larger stainless steel rings (74-mm i.d., 100-mm length) to more
accurately measure surface soil bulk density as the drop hammer
noticeably compacted the topsoil. We were not given permission
to collect additional soil bulk density measures from the fairways, so values from the tall grass rough were applied to short
turf grass (0.00.1 m only) as they experienced similar vehicle
and pedestrian traffic.
All soil samples were air-dried and then weighed. A weighed
subsample (~10 g to two decimal places) was oven-dried at
105C for 48 h to estimate gravimetric moisture content and soil
dry bulk density according to the volume sampled. Remaining
air-dried soil was passed through a 2.0-mm sieve to separate
stones and coarse organic matter, and subsamples from the three
random point locations within a vegetation type were bulked
and then fine ground (Retsch ball mill) before C and N concentration analysis on a TrueMac elemental analyzer (Leco Corp.).

Data Analysis and Presentation


The relationship between green space age and vegetation
attributes (understory vegetation volume, tree basal area, and
tree stem density) and soil properties (bulk density, percentage
C, C density, and C/N) are presented in figures using untransJournal of Environmental Quality

formed data at the plot level (n = 52). Furthermore, for each soil
property, the difference between tree canopy and short turf grass
areas were estimated through subtraction and presented in figures using untransformed data at the plot level (n = 52). Linear
regressions were performed on lognormal transformed soil and
vegetation parameters as a function of green space age.
A linear mixed model (LMM) was used to determine significant differences in soil properties (bulk density, percentage C,
percentage N, and C/N) among the three vegetation types (tree
canopy, tall grass, and short turf grass) A Chi-squared test with
Bonferroni correction was used as post-hoc test. Similarly, the
relationships between soil properties in the tree canopy areas and
vegetation attributes (understory vegetation volume, tree basal
area, and tree stem density) and green space age were investigated
using a series of LMMs, with golf course as a random effect. Soil
and vegetation parameters were lognormal transformed before
modeling. Models were ranked using the Akaike information
coefficient.
The level of significance was set to p 0.05 for all the statistical
analyses, which were performed using R (R Development
Core Team, 2012) and Genstat software (Vers. 14.0; VSN
International, 2011).
Soil C stored in the upper 0.3 m was summed for each vegetation type on a milligram-per-hectare basis and according to the
GIS digitized extent of the three vegetation types in each urban
green space scaled up to the soil C stored at the whole of green
space.

Results
Vegetation Attributes in Tree Canopy Areas
Tree basal area showed a marginally significant (p = 0.07)
increase in the initial decades after golf course establishment (Fig.
2B), increasing from <1.0 to >2.0 m2 ha1 tree canopy area after
approximately 100 yr. There was, however, large variation among
the golf courses and among plots within golf courses. There was
no significant relationship between either tree stem density or
understory vegetation volume and green space age (Fig. 2A,C).

Soil Properties
Soil bulk densities were significantly smaller, and soil C concentrations significantly greater (p 0.05) in the upper 0.2 m
of soil in the tree canopy areas than the tall grass and short turf
grass areas (Table 1). Similarly, soil N concentrations were significantly (p 0.05) greater in the tree canopy areas but only in
the upper 0.1 m. Soil C/N ratios were significantly (p 0.05) different among all three vegetation types: tree canopy > tall grass >
short turf grass (Table 1).
Soil bulk density in tree canopy areas did decrease with
increasing green space age, but this was not significant (Fig. 3A).
Soil bulk densities were approximately 0.25 g cm3 less in the tree
canopy areas than short turf grass, and there was no age effect
to this difference (Fig. 3E). There was a significant increase in
soil C concentration, C density, and C/N ratio in tree canopy
areas (00.1 m) with increasing green space age (Fig. 3BD),
markedly so for soil C/N (p 0.01). The difference in soil C/N
between tree canopy and short turf grass areas increased significantly (p 0.05) over time (Fig. 3H). The difference in soil percentage C concentrations also increased over time between tree

In tree canopy areas, the LMMs that best predicted variation


in soil properties were based on either understory vegetation
volume or tree basal area (Table 2). No models based on, or
including, green space age showed strong predictive capacity.

Soil Carbon Stocks


At the per-hectare level, there was greater C stored in soil
under tree canopy than under fairway turf grass in all but three
of the golf courses (Table 3), and two of those were <10 yr old.
Once soil C densities were scaled up to the individual golf course
based on the total extent occupied by each vegetation type, it
was apparent that at a whole-of-green-space level, tree canopy
areas stored more soil C than turf grass fairways in the majority
of golf courses. The five exceptions were the three youngest golf
courses, where tree canopy cover at a whole-of-green-space level
was low (<10 ha and <15% of green area) and two older golf
courses where tree canopy was <26% of the total green space area
(Table 3).

Discussion
Variation in Soil Properties among Urban
Green Space Types
Soil properties were significantly different between the tree
canopy areas and the tall grass or short turf grass areas. Soil bulk
density was significantly smaller, and soil percentage C and C/N
ratio were significantly greater under tree canopy than tall grass or
short turf grass areas in the upper 0.2 m. This high level of spatial
heterogeneity in soil properties is typical of urban landscapes
and has been observed in other urban soil studies (Ossola and
Livesley, 2015; Pouyat et al., 2006).

Fig. 2. (A) Understory vegetation volume, (B) tree basal area, and (C)
tree stem density in tree canopy areas as a function of green space
age. Tree basal area and stem density are presented on a per-hectare
of tree canopy area basis.

Soil Bulk Density


Within a given soil type, the differences in soil bulk density
among urban green spaces will be due to a combination of (i)
soil biological activity, (ii) root density, and (iii) compaction
from pedestrian and vehicle traffic. Many urban soil studies do
not report soil bulk density as is often inferred from organic

canopy and short turf grass areas (Fig. 3F), but this age effect was
not significant.

Table 1. Soil dry bulk density, C and N concentrations, and C/N ratio at three depth intervals in urban green space supporting tree canopy, tall grass
rough, and short turf grass fairway. Significant differences for a soil property among the three vegetation types is shown at p 0.05 using a, b, and c.
Depth
Soil bulk density (g cm3)
Canopy
Rough
Fairway
Soil C concentration (%)
Canopy
Rough
Fairway
Soil N concentration (%)
Canopy
Rough
Fairway
Soil C/N ratio
Canopy
Rough
Fairway

00.1 m

0.10.2 m

0.20.3 m

Mean

SE

Mean

SE

Mean

SE

1.16a
1.37b

0.04
0.03

1.44a
1.60b
1.56b

0.05
0.03
0.04

1.49a
1.49a
1.57b

0.05
0.03
0.03

5.10a
3.26b
2.85b

0.61
0.27
0.26

2.38a
1.68b
1.67b

0.29
0.22
0.29

2.12
2.52
1.76

0.26
0.34
0.25

0.30a
0.21b
0.20b

0.03
0.02
0.02

0.13
0.10
0.11

0.02
0.02
0.02

0.13
0.14
0.11

0.02
0.02
0.02

18.92a
15.91b
14.68c

1.12
0.56
0.40

25.52
18.33
18.03

6.37
1.31
1.79

19.59
14.73
16.96

1.27
4.29
0.85

Journal of Environmental Quality

Fig. 3. (A) Soil bulk density, (B) C concentration, (C) C density, and (D) C/N ratio in tree canopy areas (00.1 m depth) as a function of green space
age. Similarly, the differences in soil (E) bulk density, (F) C concentration, (G) C density, and (H) C/N ratio between tree canopy and short turf grass
areas as a function of green space age.

matter content or pedotransfer functions based on texture (Post


and Kwon, 2000; Pouyat et al., 2006). However, throughout the
urban green spaces of Leicester, UK, Edmondson et al. (2014)
measured no significant difference in soil bulk density between
tree canopy areas and park grassland areas. More in line with our
study, Golubiewski (2006) noted a significant decrease in soil
bulk density at older green spaces compared with younger and
suggested that this indicated that appropriate management, such
as parkland naturalization (Millward et al., 2011) and time, can
ameliorate compaction from construction or the previous land
use. Soil bulk density in urban green spaces can increase over time
because of vehicle and pedestrian traffic or can decrease over time
once drivers of compaction are removed (Millward et al., 2011).
The vegetation present in an urban green space has a central role in
Journal of Environmental Quality

ameliorating the effects of compaction, as woody vegetation can


act as a physical barrier preventing further compaction and act
as a key biological driver ameliorating previous compaction,
thorough root growth, and turnover. Future studies of soil
bulk density in urban green spaces should strive to account for
variation arising from (i) changing soil type, (ii) compaction from
vehicle and pedestrian traffic, and (iii) understory vegetation and
surface litter. Failing to account for these factors could result in
significant differences in soil bulk density being masked by the
high coefficients of variation.

Soil Carbon and Nitrogen Concentration


Both soil C and N concentration were greater under tree
canopy than tall or short grass. Greater soil C concentration

Table 2. Linear mixed models predicting tree canopy soil (0 0.1 m) bulk density (g cm3), C density (kg m2), N density (kg m2), and C/N ratio. Fixed
effect variables were individuals and pairs of years since establishment (Age), tree basal area (BA), tree stem density (TrD), and understory vegetation volume (Vol). Model suitability is indicated by the Akaike information coefficient (AIC), the best being indicated with bold values and second
and third indicated with italicized values. Akaike weighting (AIC wt) represents the probability of each model being the best performing one (1.00
high; 0.00 low).
Bulk density
AIC
AIC wt
2.84
0.01
2.21
0.18
4.42
0.54
2.62
0.23
10.13
0.00
8.90
0.00
4.46
0.01
2.39
0.02
4.46
0.01
2.60
0.02

Fixed effects
Age
Vol
BA
TrD
Age + Vol
Age + BA
Age + TrD
Vol + BA
Vol + TrD
BA + TrD

C density
AIC
218.4
212.2
210.6
213.0
219.3
218.71
220.14
212.6
214.0
212.7

N density
AIC wt
0.01
0.16
0.34
0.11
0.00
0.01
0.00
0.13
0.06
0.12

under tree canopy is likely due to greater organic matter input


from a lack of disturbance, or less disturbance, which will allow
leaf and bark litter to accumulate and decompose. The increase
in soil N concentration is more surprising as it has been proposed
that tree growth or afforestation will likely lead to a decrease
in soil N, as N is preferentially taken up and stored within
the woody vegetation biomass rather than the soil as the trees
develop (Berthrong et al., 2009).

Soil Carbon/Nitrogen Ratio


Soil C/N ratios are important because of their role in soil N
dynamics, microbial N immobilization, and capacity to buffer
eutrophication. Through our first research question, this study
set out to determine whether soils under tree canopies have
greater C/N ratios than soils under grass. It is clear that the
overall C/N ratio in top soil (00.1 m) under tree canopy was
greater than that under either of the two grass systems, as average
soil C/N was 14.7 under short turf grass and 18.9 under tree
canopy, a difference of 28%. This contrasts with Edmondson
et al. (2014) who reported no difference in soil C/N ratio
between tree canopy areas and turf grass systems in the urban
landscape of Leicester, UK. However, in a global meta-analysis
of afforestation impacts on soil, (Berthrong et al., 2009) noted

AIC
28.4
38.1
35.7
35.8
23.4
21.0
21.2
30.1
30.6
28.3

C/N
AIC wt
0.00
0.60
0.18
0.19
0.00
0.00
0.00
0.01
0.01
0.00

AIC
239.0
232.9
235.6
235.2
237.9
241.2
240.6
233.3
234.1
237.1

AIC wt
0.01
0.27
0.07
0.09
0.02
0.00
0.01
0.22
0.15
0.03

that C/N increased by between 6 and 12%, which is more in line


with our observations.
An increase in soil C/N ratio can be due to an increase in
plant N uptake and subsequent long-term storage in vegetation
biomass or an increase in the C/N ratio of the litter returned to
the soil by woody vegetation (Brady and Weil, 2008). Near our
study sites, in five local nature reserves with remnant vegetation,
the soil C/N ratio is 20.2 1.5 (data not presented), so it is quite
possible that soil C/N in tree canopy areas of these green spaces
will continue to increase. All but one of the green spaces in this
study were established on land that was previously agricultural or
horticultural, so they were likely to have similar soil conditions
before urbanization. The differences in soil C/N ratio between
the tree and grass systems are most likely related to an increase in
the C/N ratio over time because of long-term litter supply rather
than increased plant N uptake (Brady and Weil, 2008).
The divergence in soil C/N between tree canopy areas and
short turf grass areas could be influenced by N inputs through
intensive fertilizer management of the fairways (Carey et al.,
2012; Milesi et al., 2005) and subsequent cycling of organic
matter with a lower C/N. Either way, the greater soil C/N under
tree canopy provides a greater capacity to absorb sporadic, or
cumulative, N eutrophication, as soil C/N is one of the major
controls of soil N availability and leaching (Brady and Weil,

Table 3. Soil C density (00.3 m) under the three vegetation types and scaled up to the whole-of-green-space according to area (ha) covered.
Age

Tree
canopy

yr
5
7
9
37
46
49
63
74
84
87
98
99
109

Mg ha1
113.9
104.7
127.9
67.7
59.8
72.2
82.6
58.4
55.7
80.0
90.3
66.5
157.9
84.0
62.8
103.3
69.4
59.9
171.9
135.9
118.6
107.0
84.6
79.3
109.2
75.0
90.0
137.5
137.3
147.5
136.4
103.6
88.3
120.7
101.6
99.4
91.3
73.1
49.8

Tall grass
rough

Short grass
fairway

Tree
canopy

Tall grass
rough

Short grass
fairway

ha
7.5
48.6
18.7
9.5
33.2
20.6
2.0
30.1
18.4
21.8
17.2
11.7
19.7
26.5
8.0
26.7
27.3
19.6
15.6
19.3
17.5
14.9
17.5
14.1
10.4
18.8
18.5
9.9
15.7
12.9
14.4
20.0
18.6
39.2
1.2
4.9
10.2
10.0
9.2

Tree
canopy

Tall grass
rough

Short grass
fairway

Green
space

Mg
858
5090
2394
8342
641
1987
1484
4112
165
1759
1026
2950
1748
1558
776
4082
3107
2224
501
5832
2756
1894
1175
5825
2684
2626
2078
7388
1589
1480
1118
4187
1136
1410
1669
4215
1366
2158
1906
5429
1964
2076
1640
5680
4731
124
489
5344
932
731
460
2123
Journal of Environmental Quality

2008; Dungait et al., 2012; Groffman et al., 2009). Microbes


are better able to immobilize N in their biomass as soil C/N
increases, and, as a consequence, N mineralization rates decrease
leading to lower plant available N and greater capacity to buffer
N eutrophication with the combined uptake capacity of the
microbial biomass and tree root systems (Berthrong et al., 2009;
Brady and Weil, 2008; Groffman et al., 2009). Urban green
spaces with tree canopy and an increased understory vegetation
volume have been also demonstrated to have higher water infiltration rates than less complex vegetation (Ossola et al., 2015),
which, coupled with the increased soil C/N ratios, might determine a higher capacity to absorb N loads from runoff and atmospheric deposition.

Soil Organic Carbon Density


The total amount of C stored within a soil generally correlates
positively with mean annual precipitation and soil clay content
and negatively with mean annual temperature. Whereas, the vertical distribution of soil C is often influenced by the characteristics of the long-term vegetation system ( Jobbgy and Jackson,
2000). The influence of vegetation is mediated by differences
in shoot/root biomass allocation, productivity, phenology, and
rooting pattern. The soil C density profiles (00.3 m) presented
in our study showed some differences according to vegetation
type, ranging from 5.0 to 14.8 kg C m2 under tall grass rough
and from 6.8 to 17.2 kg C m2 under tree canopy. However,
despite the slight increase in soil C density with increasing green
space age, the large variation among sites meant that differences
among vegetation types were not significant (p 0.05). Such
variation is common in studies of urban soils (Golubiewski,
2006; Bae and Ryu, 2015; Edmondson, 2012). For example,
Bae and Ryu (2015) measured considerable variation in soil
C among different urban green space types in a single park in
Seoul, South Korea, from 1.58 kg C m2 under bare soil, 3.74 kg
C m2 under short turf grass up to 10.19 kg C m2 under mixed
tree canopy. In a synthesis study, Pouyat et al. (2006) assumed
that green spaces in all cities and towns throughout the United
States stored either 14.4 kg C m2 (n = 3) in residential short turf
grass soil or 7.1 kg m2 in nonresidential green space soil (n = 2).
Comparing soil C density values between cities, and even among
different soil types within a city, may not be very informative, as
the comparison will be greatly influenced by inherent differences
in pedology, soil texture and climate, and previous vegetation
system before urbanization (Ossola and Livesley, 2015; Pouyat et
al., 2006). However, it is clear that the soils in these urban green
spaces of a relatively young city (Melbourne) represent large C
stores comparable to those reported for other major cities of the
world. Regardless, the large variation in soil properties among
urban green spaces can be partly ascribed to differences in managed vegetation systems. This suggests that it would be wise to
adequately stratify for key drivers of soil C variation when sampling to estimate urban C stocks because, even in urban landscapes, soil represents the dominant C store.

Relationships between Soil Properties, Vegetation


Attributes, and Green Space Age
We found that soil properties under tree canopy were better
predicted by vegetation attributes than green space age. Both
Journal of Environmental Quality

tree basal area and understory vegetation volume best predicted


soil property variation, and this is ecologically meaningful, as
these two parameters reflect green space structure management
practices.
Figure 3H suggests that over several decades there is a large
divergence in soil C/N ratio between tree canopy and short
turf grass green space. All vegetation types sites we sampled are
altered from what occurred there under the previous land use;
if one site accumulated C but lost N from having trees planted
on former pasture while the other lost C and had N added, then
it is quite possible to achieve the large differences we observed
(Berthrong et al., 2009). In systems that are greatly changed
through anthropogenic intervention, such as urban green spaces,
time has been shown to affect most soil properties. Golubiewski
(2006) studied soil C change in urban green space along a
decadal chronosequence after urbanization and found that soil
C density recovered and surpassed that of the surrounding native
grasslands in the second decade after establishment. For lawns
established on previously forested soils in Virginia, USA, there
was a positive relationship between the time since development
and surface (00.1 m) soil C concentration but a negative relationship at 0.2 to 0.3 m (p = 0.03), perhaps as soil C in the previous ecosystem was mineralized (Campbell et al., 2014). Bae and
Ryu (2015) observed a large 2.5-fold increase in soil C over 10
yr in a new and intensively managed urban park in Seoul. Soils
under new residential turf grass in North America were estimated to have the potential to accumulate C at 0.082kgm2 yr1
(Raciti et al., 2011).
The changes in soil C and N properties following land-use
change are influenced by soil pedology and texture (percentage
clay) but are generally superseded by the greater influence
of anthropogenic management activities (Neill et al., 1997),
and this has unsurprisingly been demonstrated in urbanized
landscapes (Golubiewski, 2006).
Our study recommends that managing and maintaining areas
of tree canopy within an urban green space, especially when combined with dense understory vegetation, can provide a means to
purposefully increase soil C density and soil C/N ratios. This
makes sense when leaf-litter fall, accumulation, decomposition,
and cycling are considered, as these processes will be more complete in areas where mowing and leaf removal (sweeping or blowing) are prevented. However, there are multiple benefits from
promoting and maintaining a dense understory beneath tree canopies: it acts as vegetation barrier (i) preventing soil compaction
from pedestrian and vehicle traffic; (ii) preventing surface soil
litter removal thereby promoting surface organic matter accumulation, decomposition, and input to the soil; and iii) increasing fine root density and turnover. Root turnover is one of the
most important mechanisms by which organic matter enters the
soil system and can greatly determine soil C stocks and profile
distribution ( Jobbgy and Jackson, 2000).

Urban Green Space Soil Carbon Density, Tree Planting,


and Landscape Management
Turf grass systems dominate urban green space; in the United
States, they cover the largest land area of any irrigated crop
(Milesi et al., 2005) and can represent the largest proportion of
the urban C stock (Pouyat et al., 2006). Similarly, in our study,

grass areas (tall and short combined) dominated green space


C stocks, representing on average 65.7%. However, somewhat
surprisingly, the soil under tall grass rough areas was a greater C
store overall than soil under short turf grass, and this was probably because the large areas of tall grass rough in some of the
younger green spaces as the tree canopy areas were still establishing. Our third research question asked whether soil under grass
stores more C than soil under trees. This study determined that
the answer to this depends on whether this is on an equal-area
basis or a whole-of-green-space basis. On an equal-area basis, tree
canopy areas store more soil C than grass green space. However,
as these green spaces are (as are most urban green spaces are)
spatially dominated by short turf grass, this vegetation type is
proportionally a greater contributor to the overall soil C stock
of these urban green spaces. It must be recognized that in our
C stock estimates we have not included belowground tree root
biomass, which is often assumed to be 25% of that aboveground
for Eucalyptus spp. trees (Snowdon et al., 2000). Nor have we
accounted for C stored in the soil profile below 0.3 m, as the soil
C stock could be doubled by including the subsoil from 0.3- to
1.0-m depth ( Jobbgy and Jackson, 2000). Both of these aspects
would increase the soil C stock estimates. Vegetation systems
that allocate biomass to root growth deep in the soil profile
(>0.5 m) will often have greater soil C concentrations at these
depths than vegetation systems that are more superficial in their
root biomass allocation ( Jobbgy and Jackson, 2000). As such,
establishing tree canopy on land that was previously occupied by
shallow rooting annuals and grasses may have led to an increase
in soil C below the depth investigated in this study (up to 0.3
m). Therefore, the estimates of soil C increase under tree canopy
presented in this study are likely to be conservative.
Our study demonstrates that higher soil C concentrations
can be found in areas under tree canopy than under grass within
large urban green spaces, such as the golf courses investigated
here. Moreover, these tree canopy areas significantly increase
their C/N ratio over time, which is likely to enhance the capacity of these urban green spaces to buffer N eutrophication. From
an urban C stock perspective, it is clear that tree canopy areas
store greater soil C per unit of area than tall grass area or short
turf grass system. This suggests that the conclusions of several
published studies of C stored in soil and vegetation in residential
and nonresidential urban green space may be too simplistic. For
example, studies in Leicester, UK, found that soil C was greatest
under trees in residential green space as opposed to under trees
in public parks, and it was suggested that this reflected the use of
compost, mulches, and organic fertilizers by private residentialspace managers (Davies et al., 2011; Edmondson et al., 2014).
While there may be some support for this, the managed act of
simply promoting greater root development, litter accumulation,
and decomposition in tree canopy areas should not be underestimated. Furthermore, it would be wrong to assume that planting trees in nonresidential green space, such as public parks and
golf courses, will not provide significant soil C benefits, as well as
other ecosystem service benefits, to the urban landscape.
The capacity of any urban green space to accumulate C or N,
or to alter the soil C/N ratio, will be determined by (i) the length
of time they are managed as a green space after urbanization, and
(ii) the structural composition of vegetation habitat, such as the
presence of trees, an understory, and surface litter.

Conclusions
Urban green space soil under tree canopies had significantly
lower bulk density but significantly greater soil C and N concentrations and C/N ratios than soil under tall grass rough or short
grass turf. Soil C/N ratios in tree canopy areas were related to
green space age, whereas surface soil C and bulk density related
better to understory vegetation volume and tree stem density.
Overall, tree canopy areas contribute the most soil C to urban
soils on a per-unit-area basis because of their greater soil C density (00.3 m depth). However, at the whole-of-green-space
scale, the contribution of short turf grass areas dominates the soil
C store because of the larger proportional area they occupy. The
greater soil C/N ratios of tree canopy areas mean they are better
able to buffer soil and water N pollution from poor fertilizer
management or cumulative atmospheric N deposition. They are
therefore providing an important ecosystem service more effectively than other green space types. The planting and placement
of trees within urban green spaces should consider management
options that also allow understory vegetation and litter to accumulate where possible.

Acknowledgments
We would like to thank Stephen Stewart and Lee Wilson for helping
with field work and the Australian Golf Course Superintendents
Association, in particular John Neylan and John Geary, for cofunding
the study and arranging site access and infrastructure support. The study
was funded through the ARC Linkage program (ARC LP 110100686),
with contributions from the Australian Golf Course Superintendents
Association and Royal Botanical Gardens Melbourne. AKH would like
to acknowledge support from the Baker Foundation. Soil analysis was
performed by Najib Ahmady at the Creswick campus laboratory of the
University of Melbourne.

References
Bae, J., and Y. Ryu. 2015. Land use and land cover changes explain spatial
and temporal variations of the soil organic carbon stocks in a constructed urban park. Landsc. Urban Plan. 136:5767. doi:10.1016/j.
landurbplan.2014.11.015
Berthrong, S.T., E.G. Jobbgy, and R.B. Jackson. 2009. A global meta-analysis
of soil exchangeable cations, pH, carbon, and nitrogen with afforestation.
Ecol. Appl. 19:22282241. doi:10.1890/08-1730.1
Brady, N.C., and R.R. Weil. 2008. The nature and properties of soils. Pearson
Prentice Hall, Upper Saddle River, NJ.
Campbell, C.D., J.R. Seiler, E.P. Wiseman, B.D. Strahm, and J.F. Munsell. 2014.
Soil carbon dynamics in residential lawns converted from Appalachian
mixed oak stands. Forests 5:425438. doi:10.3390/f5030425
Carey, R.O., G.J. Hochmuth, C.J. Martinez, T.H. Boyer, V.D. Nair, M.D. Dukes,
et al. 2012. A review of turfgrass fertiliser management practices: Implications for urban water quality. Hortic. Tech. 22:280291.
Davies, Z.G., J.L. Edmondson, A. Heinemeyer, J.R. Leake, and K.J. Gaston. 2011. Mapping an urban ecosystem service: Quantifying aboveground carbon storage at a city-wide scale. J. Appl. Ecol. 48:11251134.
doi:10.1111/j.1365-2664.2011.02021.x
De Kimpe, C.R., and J.L. Morel. 2000. Urban soil management: A growing concern. Soil Sci. 165:3140. doi:10.1097/00010694-200001000-00005
Dise, N.B., E. Matzner, and M. Forsius. 1998. Evaluation of organic horizon C:N
ratio as an indicator of nitrate leaching in conifer forests across Europe.
Environ. Pollut. 102:453456. doi:10.1016/S0269-7491(98)80068-7
Dungait, J.A.J., L.M. Cardenas, M.S.A. Blackwell, L. Wu, P.J.A. Withers, D.R.
Chadwick, et al. 2012. Advances in the understanding of nutrient dynamics and management in UK agriculture. Sci. Total Environ. 434:3950.
doi:10.1016/j.scitotenv.2012.04.029
Edmondson, J.L., Z.G. Davies, N. McHugh, K.J. Gaston, and J.R. Leake. 2012.
Organic carbon hidden in urban ecosystems. Scientific Reports 2:963.
doi:10.1038/srep00963
Edmondson, J.L., O.S. OSullivan, R. Inger, J. Potter, N. McHugh, K.J. Gaston, et
al. 2014. Urban tree effects on soil organic carbon. PLoS ONE 9:e101872.
doi:10.1371/journal.pone.0101872
Journal of Environmental Quality

Field, C.B., and M.R. Raupach. 2004. The global carbon cycle. Island Press,
Washington, DC.
Foley, J.A., M.H. Costa, C. Delire, N. Ramankutty, and P. Snyder. 2003. Green
surprise? How terrestrial ecosystems could affect earths climate. Front.
Ecol. Environ 1:3844.
Golubiewski, N.E. 2006. Urbanization increases grassland carbon pools: Effects of landscaping in Colorados front range. Ecol. Appl. 16:555571.
doi:10.1890/1051-0761(2006)016[0555:UIGCPE]2.0.CO;2
Grimm, N.B., S.H. Faeth, N.E. Golubiewski, C.L. Redman, J.G. Wu, X.M. Bai,
et al. 2008. Global change and the ecology of cities. Science 319:756760.
doi:10.1126/science.1150195
Groffman, P.M., R.V. Pouyat, M.L. Cadenasso, W.C. Zipperer, K. Szlavecz,
I.D. Yesilonis, et al. 2006. Land use context and natural soil controls on
plant community composition and soil nitrogen and carbon dynamics in
urban and rural forests. For. Ecol. Manage. 236:177192. doi:10.1016/j.
foreco.2006.09.002
Groffman, P.M., C.O. Williams, R.V. Pouyat, L.E. Band, and I.D. Yesilonis. 2009.
Nitrate leaching and nitrous oxide flux in urban forests and grasslands. J.
Environ. Qual. 38:18481860. doi:10.2134/jeq2008.0521
Gundersen, P., I. Callesen, and W. De Vries. 1998. Nitrate leaching in forest ecosystems is related to forest floor C/N ratios. Environ. Pollut. 102:403407.
doi:10.1016/S0269-7491(98)80060-2
Gurmesa, G.A., I.K. Schmidt, P. Gundersen, and L. Vesterdal. 2013. Soil carbon accumulation and nitrogen retention traits of four tree species grown
in common gardens. For. Ecol. Manage. 309:4757. doi:10.1016/j.
foreco.2013.02.015
Hutyra, L.R., R. Duren, K.R. Gurney, N. Grimm, E.A. Kort, E. Larson, et al.
2014. Urbanization and the carbon cycle: Current capabilities and research
outlook from the natural sciences perspective. Earths Future 2:473495.
doi:10.1002/2014EF000255
Jobbgy, E.G., and R.B. Jackson. 2000. The vertical distribution of soil organic
carbon and its relation to climate and vegetation. Ecol. Appl. 10:423436.
doi:10.1890/1051-0761(2000)010[0423:TVDOSO]2.0.CO;2
Kaye, J.P., P.M. Groffman, N.B. Grimm, L.A. Baker, and R.V. Pouyat. 2006.
A distinct urban biogeochemistry? Trends Ecol. Evol. 21:192199.
doi:10.1016/j.tree.2005.12.006
Law, N., L. Band, and M. Grove. 2004. Nitrogen input from residential lawn care
practices in suburban watersheds in Baltimore County, MD. J. Environ.
Plann. Manage. 47:737755. doi:10.1080/0964056042000274452
Livesley, S.J., B.J. Dougherty, A.J. Smith, D. Navaud, L.J. Wylie, and S.K. Arndt.
2010. Soil-atmosphere exchange of carbon dioxide, methane and nitrous
oxide in urban garden systems: Impact of irrigation, fertiliser and mulch.
Urban Ecosyst. 13:273293. doi:10.1007/s11252-009-0119-6
Milesi, C., S.W. Running, C.D. Elvidge, J.B. Dietz, B.T. Tuttle, and R.R. Nemani. 2005. Mapping and modeling the biogeochemical cycling of turf
grasses in the United States. Environ. Manage. 36:426438. doi:10.1007/
s00267-004-0316-2
Millward, A.A., K. Paudel, and S.E. Briggs. 2011. Naturalization as a strategy
for improving soil physical characteristics in a forested urban park. Urban
Ecosyst. 14:261278. doi:10.1007/s11252-010-0153-4

Journal of Environmental Quality

Murty, D., M.U.F. Kirschbaum, R.E. McMurtrie, and H. McGilvray. 2002.


Does conversion of forest to agricultural land change soil carbon and
nitrogen? A review of the literature. Glob. Change Biol. 8:105123.
doi:10.1046/j.1354-1013.2001.00459.x
Neill, C., J.M. Melillo, P.A. Steudler, C.C. Cerri, J.F.L. De Moraes, M.C. Piccolo,
et al. 1997. Soil carbon and nitrogen stocks following forest clearing for
pasture in the southwestern Brazilian Amazon. Ecol. Appl. 7:12161225.
doi:10.1890/1051-0761(1997)007[1216:SCANSF]2.0.CO;2
Ossola, A., A.K. Hahs, and S.J. Livesley. 2015. Habitat complexity influences
fine scale hydrological processes and the incidence of stormwater runoff in
managed urban ecosystems. J. Environ. Manage 159:110. doi:10.1016/j.
jenvman.2015.05.002
Ossola, A., and S.J. Livesley. 2015. Drivers of soil heterogeneity in the urban
landscape. In: R.A. Francis, editor, Urban landscape ecology: Science,
policy and practice. Routledge, New York.
Pataki, D.E., M.M. Carreiro, J. Cherrier, N.E. Grulke, V. Jennings, S. Pincetl, et
al. 2011. Coupling biogeochemical cycles in urban environments: Ecosystem services, green solutions, and misconceptions. Front. Ecol. Environ
9:2736. doi:10.1890/090220
Post, W.M., and K.C. Kwon. 2000. Soil carbon sequestration and landuse change: Processes and potential. Glob. Change Biol. 6:317327.
doi:10.1046/j.1365-2486.2000.00308.x
Pouyat, R.V., I.D. Yesilonis, and D.J. Nowak. 2006. Carbon storage by urban
soils in the United States. J. Environ. Qual. 35:15661575. doi:10.2134/
jeq2005.0215
Raciti, S., P. Groffman, J. Jenkins, R. Pouyat, T. Fahey, S.A. Pickett, et al. 2011. accumulation of carbon and nitrogen in residential soils with different landuse histories. Ecosystems 14:287297. doi:10.1007/s10021-010-9409-3
R Development Core Team. 2012. R: A language and environment for statistical
computing. R Foundation for Stat. Comput., Vienna, Austria.
Setl, H., R.D. Bardgett, K. Birkhofer, M. Brady, L. Byrne, P.C. de Ruiter, et al.
2014. Urban and agricultural soils: Conflicts and trade-offs in the optimization of ecosystem services. Urban Ecosyst. 17:239253. doi:10.1007/
s11252-013-0311-6
Seto, K.C., S. Parnell, and T. Elmqvist. 2013. A global outlook on urbanization.
In: T. Elmqvist, M. Fragkias, J. Goodness, B. Gneralp, P.J. Marcotullio,
R.I. McDonald, S. Parnell, M. Schewenius, M. Sendstad, K.C. Seto, and
C. Wilkinson, editors, Urbanization, biodiversity and ecosystem services:
Challenges and opportunities. Springer, the Netherlands. p. 112.
Snowdon, P., D. Eamus, P. Gibbons, P. Khanna, H. Keith, J. Raison, et al. 2000.
Synthesis of allometrics, review of root biomass and design of future woody
biomass sampling strategies. NCAS Technical Series. Australian Greenhouse Office, Canberra, Australia.
Valtanen, M., N. Sillanp, and H. Setl. 2014. The effects of urbanization on
runoff pollutant concentrations, loadings and their seasonal patterns under
cold climate. Water, Air, Soil Pollut. 225:1977.
VSN International. 2011. GenStat for Windows. 14th ed. VSN International,
Hemel Hempstead, UK.