Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
IMPLICATIONS OF URBANIZATION
RELATED LAND USE CHANGE ON THE
CARBON AND NITROGEN CYCLE
FROM SUBTROPICAL SOILS
Lona van Delden
Dipl.-Ing. agr.
Submitted in fulfilment of the requirements for the degree of
Doctor of Philosophy
Institute for Future Environments
School of Earth, Environmental and Biological Sciences
Science and Engineering Faculty
Queensland University of Technology
2017
Reducing nitrous oxide emissions while supporting subtropical cereal production in Oxisols
3
5
KEYWORDS
Urbanization, land use change, native forest, grazed pasture, turf grass, carbon,
nitrogen, soil-atmosphere gas exchange, greenhouse gas, nitrous oxide, methane,
carbon sequestration, subtropical, high-frequency greenhouse gas measurements,
climate change, Chromosols.
7
TABLE OF CONTENTS
Keywords ................................................................................................................................................ 5
Table of Contents .................................................................................................................................... 7
List of Figures ....................................................................................................................................... 11
List of Tables ........................................................................................................................................ 13
List of Abbreviations ............................................................................................................................. 15
Publications Incorporated into the Thesis ............................................................................................. 17
Statement of Original Authorship ......................................................................................................... 18
Acknowledgements ............................................................................................................................... 19
CHAPTER 1: INTRODUCTION ..................................................................................................... 21
1.1 Background ................................................................................................................................ 21
1.2 Research problem....................................................................................................................... 23
1.3 Research aim and objectives ...................................................................................................... 25
1.4 Method & Outcome ................................................................................................................... 27
1.5 Significance ............................................................................................................................... 28
CHAPTER 2: LITERATURE REVIEW ......................................................................................... 31
2.1 Land use change associated with urbanization .......................................................................... 31 2.1.1 Urbanization background ................................................................................................ 31 2.1.2 Land use change impact on the environment .................................................................. 34
2.2 Land use and climate change implications ................................................................................. 37 2.2.1 Climate change dynamics ............................................................................................... 37 2.2.2 Feedback effects ............................................................................................................. 40
2.3 Biogeochemical C and N cycling ............................................................................................... 44 2.3.1 Soil C and N ................................................................................................................... 44 2.3.2 Soil-atmosphere C and N exchange ................................................................................ 52
2.4 Summary & implications ........................................................................................................... 59
CHAPTER 3: RESEARCH DESIGN ............................................................................................... 65
3.1 Site description .......................................................................................................................... 65
3.2 Materials and Methods ............................................................................................................... 67 3.2.1 Experimental design ....................................................................................................... 67 3.2.2 GHG gas flux system ...................................................................................................... 68 3.2.3 Soil survey ...................................................................................................................... 68 3.2.3.1 Soil sampling ................................................................................................................. 69 3.2.3.2 C fractionation ............................................................................................................... 69 3.2.4 Environmental parameters .............................................................................................. 69 3.2.5 Data management and statistical analysis ....................................................................... 70
3.3 Thesis outline ............................................................................................................................. 71
CHAPTER 4: ESTABLISHING TURF GRASS INCREASES SOIL GREENHOUSE GAS
EMISSIONS IN PERI-URBAN ENVIRONMENTS (PAPER 1) .................................................... 77
4.1 Abstract ...................................................................................................................................... 77
4.2 Introduction ................................................................................................................................ 78
8
4.3 Materials and Methods ............................................................................................................... 80 4.3.1 Site description ............................................................................................................... 80 4.3.2 Experimental design ....................................................................................................... 81 4.3.3 CH4 and N2O flux measurements ................................................................................... 82 4.3.4 Auxiliary measurements ................................................................................................. 82 4.3.5 Flux calculations and statistical analysis ........................................................................ 83
4.4 Results ....................................................................................................................................... 84 4.4.1 Site description ............................................................................................................... 84 4.4.2 CH4 and N2O flux measurements ................................................................................... 85 4.4.3 Global warming potential ............................................................................................... 86
4.5 Discussion .................................................................................................................................. 89 4.5.1 CH4 and N2O flux measurements ................................................................................... 89 4.5.2 Global warming potential ............................................................................................... 90 4.5.3 Conclusion ...................................................................................................................... 92
CHAPTER 5: URBANIZATION-RELATED LAND USE CHANGE FROM FOREST AND
PASTURE INTO TURF GRASS MODIFIES SOIL NITROGEN CYCLING AND INCREASES
N2O EMISSIONS (PAPER 2) ............................................................................................................ 95
5.1 Abstract ...................................................................................................................................... 95
5.2 Introduction................................................................................................................................ 96
5.3 Materials and Methods ............................................................................................................... 98 5.3.1 Site description ............................................................................................................... 98 5.3.2 Experimental design ....................................................................................................... 99 5.3.3 N2O flux measurements ................................................................................................ 100 5.3.4 Auxiliary measurements ............................................................................................... 100 5.3.5 Flux calculations and statistical analysis ...................................................................... 101
5.4 Results ..................................................................................................................................... 102 5.4.1 Site characteristics ........................................................................................................ 102 5.4.2 Environmental parameters ............................................................................................ 103 5.4.3 Temporal variability of mineral N ................................................................................ 104 5.4.4 Temporal variability of N2O fluxes .............................................................................. 106 5.4.5 Environmental parameters influencing N2O fluxes ...................................................... 109
5.5 Discussion ................................................................................................................................ 110 5.5.1 Mineral N ..................................................................................................................... 111 5.5.2 N2O fluxes .................................................................................................................... 113 5.5.3 Effect of land use change associated with urbanization ............................................... 115
5.6 Conclusions.............................................................................................................................. 117
5.7 Acknowledgements .................................................................................................................. 117
CHAPTER 6: SOIL N2O AND CH4 FLUXES FROM URBANIZATION RELATED LAND
USE CHANGE; FROM EUCALYPTUS FOREST AND PASTURE TO URBAN LAWN
(PAPER 3) 121
6.1 Abstract .................................................................................................................................... 121
6.2 Introduction.............................................................................................................................. 122
6.3 Materials and Methods ............................................................................................................. 125 6.3.1 Site description ............................................................................................................. 125 6.3.2 Experimental design ..................................................................................................... 125 6.3.3 GHG flux measurements .............................................................................................. 126 6.3.4 Auxiliary measurements ............................................................................................... 126 6.3.5 Flux calculations and statistical analysis ...................................................................... 127
6.4 Results ..................................................................................................................................... 128 6.4.1 Environmental and soil parameters .............................................................................. 128 6.4.2 N2O fluxes .................................................................................................................... 130 6.4.3 CH4 fluxes .................................................................................................................... 131 6.4.4 Influence of environmental parameters on N2O and CH4 fluxes .................................. 132
9
6.4.5 Non-CO2 global warming potential .............................................................................. 136
6.5 Discussion ................................................................................................................................ 138 6.5.1 N2O fluxes .................................................................................................................... 138 6.5.2 CH4 fluxes ..................................................................................................................... 140 6.5.3 Inter-annual drivers of GHG fluxes .............................................................................. 141 6.5.4 Influence of land use change on GWP .......................................................................... 142 6.5.5 Outlook ......................................................................................................................... 143 6.5.6 Acknowledgements....................................................................................................... 144
CHAPTER 7: LAND USE CHANGE IMPLICATIONS ON THE SOIL C SEQUESTRATION
POTENTIAL OF PERI-URBAN ENVIRONMENTS (PAPER 4) ............................................... 145
7.1 Abstract .................................................................................................................................... 145
7.2 Introduction .............................................................................................................................. 146
7.3 Material and Methods .............................................................................................................. 149 7.3.1 Site description ............................................................................................................. 149 7.3.2 Experimental design ..................................................................................................... 149 7.3.3 Sampling ....................................................................................................................... 150 7.3.4 Sample preparation and analysis ................................................................................... 151 7.3.5 C fractionation .............................................................................................................. 151 7.3.6 Statistical analysis ......................................................................................................... 152
7.4 Results ...................................................................................................................................... 153 7.4.1 Environmental conditions ............................................................................................. 153 7.4.2 Carbon .......................................................................................................................... 154 7.4.3 Nitrogen ........................................................................................................................ 154 7.4.4 Environmental influence on C fractions ....................................................................... 155
7.5 Discussion ................................................................................................................................ 158 7.5.1 Soil C sequestration potential ....................................................................................... 158 7.5.2 Environmental influence on C and N cycling ............................................................... 160
7.6 Conclusion ............................................................................................................................... 162
7.7 Acknowledgements .................................................................................................................. 162
CHAPTER 8: DISCUSSION AND CONCLUSIONS ................................................................... 164
8.1 Environmental parameters ....................................................................................................... 164
8.2 Objective 1 ............................................................................................................................... 166
8.3 Objective 2 ............................................................................................................................... 167
8.4 Objective 3 ............................................................................................................................... 169
8.5 Objective 4 ............................................................................................................................... 171
8.6 Outlook .................................................................................................................................... 173
8.7 Conclusions .............................................................................................................................. 178
BIBLIOGRAPHY ............................................................................................................................. 181
APPENDIX 203
11
LIST OF FIGURES
Figure 1-1 Hypothesized multiple time-scale response of the Global Warming Potential of
newly established turf grass associated with urbanization processes when compared
to forest and pasture. ............................................................................................................ 28
Figure 2-1 Population of the world for the years 1950-2100, according to several projections of
the population increase based on of medium, high, low and constant human fertility
by the United Nations (2013). .............................................................................................. 32
Figure 2-2 Global amount of people living in urban and rural environments from 1950 to 2050
(United Nations 2008). ......................................................................................................... 33
Figure 2-3 IPCC 2013: Time series of temperature change relative to 1986–2005 averaged over
land grid points over the globe in December to February calculated from a variety of
Representative Concentration Pathways (RCPs) from the radiative forcing (+2.6,
+4.5, +6.0, and +8.5 W m-2
, respectively) of greenhouse gas concentration in the
atmosphere (Stocker et al. 2013). ......................................................................................... 38
Figure 2-4 Atmospheric concentrations of the three main long-lived greenhouse gases over the
last 2000 years. Increases since about 1750 are attributed to human activities in the
industrial era (Cubasch et al. 2001). ..................................................................................... 39
Figure 2-5 Heavy rainfall across Australia with over 300 mm d-1
in Samford Valley, SEQ
(Highvale weather station, BOM (2015 in January 2013, Source: Commonwealth of
Australia (2013. .................................................................................................................... 40
Figure 2-6 Conceptual model of climate change and the role of land ecosystem-atmosphere
interactions (Betts 2007). ..................................................................................................... 41
Figure 2-7 Socio-ecological framework by Grimm (2008) identifying the drivers and
responders of climate change on a local, regional and global scale. .................................... 43
Figure 2-8 Conceptual model of links between net primary productivity, litter C,
decomposition, microbial trace gas fluxes and soil N availability and their main
driving parameters climate and soil moisture (Pastor and Post 1986; Groffman et al.
1995). ................................................................................................................................... 46
Figure 2-9 Soil biological processes of GHG (a) uptake into the soil and (b) emissions from the
soil into the atmosphere (Baldock et al. 2012). .................................................................... 55
Figure 2-10 ‘Hole-in-the-pipe’ model of the regulation of trace-gas production and
consumption by nitrification and denitrification (Bouwman 1998). .................................... 56
Figure 3-1 Location of Samford Valley near Brisbane in South East Queensland, Australia. .............. 66
Figure 3-2 Experimental site at SERF showing pasture, turf grass and fallow plots as well as
the adjunct forest. ................................................................................................................. 67
Figure 4-1 Daily average CH4 (A, B) and N2O (C, D) fluxes for each treatment with error bars
from the annual measurements (2009/2010) and the intensive sampling campaign
(2013), as well as SERF climate data (E, F) for all sampling periods. ................................. 88
Figure 5-1 - Annual soil NO3- (A) and NH4
+ (B) contents variations from forest, pasture, turf
grass and fallow averaged across replicates (n = 3) and summed for separate
analysed soil depths of 0-10 and 10-20 cm with the climatic conditions (C) for the
experimental year 2013/2014 as well as fertilization and irrigation indication for the
turf grass treatment. ............................................................................................................ 105
Figure 5-2 - Daily N2O flux averages (max 8 fluxes per day for 3 replicates each) with standard
errors (n =3) over the experimental year 2013/2014 for forest (A), pasture (B), turf
grass (C) and fallow (D) with the treatment specific water filled pore space (WFPS). ...... 107
12
Figure 5-3 - Cumulative daily N2O fluxes (n = 3) for forest, pasture, turf grass and fallow with
rainfall for the experimental year 2013/2014. .................................................................... 108
Figure 5-4 – Linear relationship of log transformed N2O emissions with mineral N content
within 20 cm soil depth for each replicate of forest, pasture, turf grass and fallow
land use during the establishment phase (A) and the rest of the year (B), with the
coefficient of determination R2. ......................................................................................... 110
Figure 6-1 – Daily minimum and maximum temperatures and rainfall for the two years from
June 2013 until June 2015 for the experimental site .......................................................... 128
Figure 6-2 Two years of N2O (a) and CH4 (b) fluxes from the dry sclerophyll forest soil with
supporting environmental parameters (c) mean daily temperature and water filled
pore space (WFPS). ............................................................................................................ 133
Figure 6-3 Two years of N2O (a) and CH4 (b) fluxes from the agricultural pasture soil with
supporting environmental parameters (c) mean daily temperature and water filled
pore space (WFPS). ............................................................................................................ 134
Figure 6-4 Two years of N2O (a) and CH4 (b) fluxes from the turf grass soil with supporting
environmental parameters (c) mean daily temperature, water filled pore space
(WFPS) and fertilization events (↓).................................................................................... 135
Figure 6-5 Combined global warming potential from CO2-equivalents of N2O and CH4 soil-
atmosphere gas exchange for the peri-urban land uses forest, pasture and turf grass
for the first and the second experimental year separately as well as the inter-annual
average. .............................................................................................................................. 137
Figure 7-1 Selected private and public sites in Samford Valley, Queensland, Australia, of the
land use types forest (D2, JMP2, R2, SERF2, MR, BPP), pasture (D1, JMP1, URR,
KR, CSIRO, Dy), and turf grass lawn (A, MRDR, ELP, SPS, R1, SERF1). ..................... 153
Figure 7-2 Soil organic C average of in the form of active C (CA), slow C (CS) and resistant C
(CR) per land use type with standard error for 0-10 cm soil depth (A) and 10-20 cm
soil depth (B) ...................................................................................................................... 155
Figure 8-1 Hypothesized multiple time scale scheme corrected for the long-term response .............. 174
Figure A 1 Percentage of the population in urban areas, 2007, 2025 and 2050 (United Nations
2008). ................................................................................................................................. 203
Figure A 2 Major cities of Australia (Commonwealth of Australia 2013).......................................... 204
Figure A 3 Population distribution of selected countries; Source: Ellis in Commonwealth of
Australia (2013. .................................................................................................................. 204
Figure A 4 Population growth rates of OECD countries, 2000–10; Source: OECD 2012 in
Commonwealth of Australia (2013. ................................................................................... 205
Figure A 5 Principal global carbon pools (Lal 2004b). ....................................................................... 206
Figure A 6 Temperature change forecast for Australia from Appendix I in Stocker et al. (2013. ...... 206
Figure A 7 Australian Supersite Network (ASN) locations. Samford Ecological Research
Facility (SERF) is located at South East Queensland (SEQ) and is the only peri-
urban supersite in Australia. ............................................................................................... 207
Figure A 8 Global distribution of Planosols aka Chromosols by FAO/UNESCO (1998. ................... 207
Figure A 9 Typical soil profile of a Brown Chromosol defined by the Australian Soil
Classification (CSIRO 1996; Isbell 2002).......................................................................... 208
Figure A 10 Representative Australian soil types with their SOC content (Baldock et al. 2012). ...... 208
Figure A 11 Core site plot plan with automatic chambers organised in 3 measurement sets.............. 209
Figure A 12 ARIMA modeled confidence interval for CH4 and N2O fluxes over the
experimental timeframe from June 2013 to June 2015 for the forest, pasture and turf
grass (lawn) land use. ......................................................................................................... 210
13
LIST OF TABLES
Table 2-1 ‘Urban’ area and population compared to official metropolitan area of major cities
(Spencer 2015) ..................................................................................................................... 34
Table 2-2 Major GHG concentrations currently and historically (Stocker et al. 2013)......................... 54
Table 2-3 Approximate representation of the main literature in study related research topics per
climate zone in percent (%) and the overall global attention the topic has received ............ 62
Table 4-1 – SERF site description. ........................................................................................................ 84
Table 4-2 Average and cumulative fluxes of CH4 and N2O with standard error for each
treatment together with their significance, as well as calculated global warming
potential (GWP) for the intensive sampling campaign (80 days) in 2013. ........................... 85
Table 5-1 – SERF site characteristics .................................................................................................. 103
Table 5-2 - Seasonal and cumulative rain, number of rain events and seasonal and annual
averages of minimum and maximum Temperatures of the experimental year ................... 103
Table 5-3 - Annual mineral N averages as NH4+-N and NO3
--N in 0-20 cm soil depth, WFPS
and daily maximum and average N2O fluxes from all treatments with their
cumulative annual fluxes over the experimental year with their standard error. ................ 106
Table 5-4 - Spearman’s rho correlation coefficient between N2O fluxes and mineral N, WFPS
and temperature for each treatment. ................................................................................... 109
Table 6-1 – Site characteristics ........................................................................................................... 129
Table 6-2 Annual rainfall, number of heavy rain events and annual average minimum and
maximum temperatures for the experimental years 2013 and 2014. .................................. 129
Table 6-3 – Daily and annual N2O and CH4 flux averages, the non-CO2 global warming
potential (GWP) and water filled pore space (WFPS) for all three land uses with
their standard error and indication for significant differences between land uses .............. 130
Table 6-4 ARIMA time series coefficient for N2O and CH4 fluxes in g ha-1
d-1
and water filled
pore space (WFPS) for all three land uses and temperature as well as mineral N
(NO3- and NH4
+) and the factor of turf grass establishment impact on N2O and CH4
emissions ............................................................................................................................ 132
Table 7-1 Site characteristics for the topsoil (0-10 cm) averaged per land use type with
standard error ..................................................................................................................... 154
Table 7-2 Soil C contents averaged per land use type with standard errors in total (CT) and the
three C fractions of active (CA), slow (CS) and resistant (CR); total N (NT), mineral N
(Nmin) and soil C/N ratio ..................................................................................................... 155
Table 7-3 Spearman’s rho correlations of the active (CA), slow (CS) and resistant (CR) C
fractions with each other and their soil parameters total C (CT) and N (NT), mineral
N (Nmin), pH, electric conductivity (EC) and clay content for the upper 10 cm topsoil ..... 156
Table 7-4 Site parameters for all 18 sampling sites in Samford Valley based on 4 replicated
field subsamples and 3 laboratory replicates per value ...................................................... 157
15
LIST OF ABBREVIATIONS
C Carbon
CA Active carbon
CR Resistant carbon
CS Slow carbon
CT Total carbon
C:N Carbon-to-Nitrogen ratio
CO2 Carbon dioxide
EF Emission Factor
FAO Food and Agriculture Organization
GHG Greenhouse Gas
GWP Global Warming Potential
IPCC Intergovernmental Panel on Climate Change
LUC Land Use Change
N Nitrogen
N2 Dinitrogen
NH3 Ammonia
NH4+ Ammonium
N2O Nitrous oxide
NO Nitric oxide
NO2- Nitrite
NO3- Nitrate
NT Total nitrogen
O2 Oxygen
OECD Organisation for Economic Co-operation and Development
SERF Samford Ecological Research Facility
SEQ South East Queensland
SOC Soil Organic Carbon
SOM Soil Organic Matter
TERN Terrestrial Ecosystem Research Network
WFPS Water Filled Pore Space
17
PUBLICATIONS INCORPORATED INTO
THE THESIS
van Delden, L., E. Larsen, D. W. Rowlings, C. Scheer and P. R. Grace. 2016.
"Establishing turf grass increases soil greenhouse gas emissions in peri-
urban environments." Urban Ecosystems, Volume 19, Issue 2, pp 749–762.
van Delden, L., D. W. Rowlings, C. Scheer and P. R. Grace. 2016. " Urbanization-
related land use change from forest and pasture into turf grass modifies soil
nitrogen cycling and increases N2O emissions." Biogeosciences, Volume 13,
Issue 21, pp 6095-6106.
van Delden, L., D. W. Rowlings, C. Scheer, D. De Rosa and P. R. Grace. "Soil N2O
and CH4 fluxes from urbanization related land use change; from Eucalyptus
forest and pasture to urban lawn" submitted to Global Change Biology.
van Delden, L., D. W. Rowlings, C. Scheer and P. R. Grace. "Land use change
implications on the soil C sequestration potential of peri-urban
environments" in preparation.
18
STATEMENT OF ORIGINAL AUTHORSHIP
The work contained in this thesis has not been previously submitted to meet
requirements for an award at this or any other higher education institution. To the
best of my knowledge and belief, the thesis contains no material previously
published or written by another person except where due reference is made.
Signature: QUT Verified Signagure
Date: July 2017
19
ACKNOWLEDGEMENTS
I would like to thank my supervisors Dr David Rowlings, Dr Clemens Scheer,
Professor Peter R. Grace for providing me with the opportunity to study such an
interesting topic. I thank them for their encouragement, guidance and assistance
throughout the term of this research project.
I would also like to thank the following organisations for their significant
contributions towards my PhD:
The Queensland University of Technology and the Institute for Future
Environments for providing the scholarship;
The Samford Ecological Research Facility and Terrestrial Ecosystem
Research Network for the field site and materials;
The Moreton Bay Regional Council for providing detailed information and
access to the public sampling sites in Samford Valley;
The Central Analytical Research Facility for the use of high quality
laboratory equipment, technical support and data analysis.
I am extremely grateful to Marcus Yates, Karyn Gonano, and the whole HEEM
aka M4RL team for their professional, technical, physical and often mental support
during this journey.
A very special thank-you to my family and friends for their understanding,
support and encouragement. A special special-thank-you to my mother Lila for all
the professional resilience coaching, my sister Maike for the open ear and the best
reason to come home and my Omi Reni for the unlimited believe in me. This thesis is
dedicated to my partner Dan, who I would not have met without this PhD and who
was at my side every step of the way – FIGJAM.
Introduction
21
Chapter 1: Introduction
1.1 Background
Urban populations worldwide have not only exceeded rural populations but are
also predicted to account for most future population growth (United Nations 2014).
Increasing population densities and urban sprawl are causing rapid land use change
from natural ecosystems and commercially focused agriculture in rural areas into
smaller, residential properties. While over half the soils in build-up urban
environments are sealed (Scalenghe and Marsan 2009), the transitional stage of peri-
urban landscapes are associated with soil disturbance during construction processes
and increasingly the extensive establishment of turf grass as urban lawn for golf
courses, sports grounds, parks and residential properties (IPCC 2006). Significantly,
these land use changes influence ecosystem dynamics potentially causing substantial
nutrient losses in form of highly potent greenhouse gases (GHGs) from soils into the
atmosphere, where their radiative forcing accelerates climate change (IPCC 2013).
The majority of future global demographic growth is projected to take place in
tropical and subtropical regions of Africa, South America and Asia (UNFPA 2011).
These tropical and subtropical regions will also play a significant role in achieving
global food security in the future (FAO and ITPS 2015), which implies future land
use changes both from and into agricultural as well as residential land use. Data from
temperate zones identifies turf grass as contributing to climate change to a
comparable degree as intensive agriculture on an area basis (Kaye et al. 2004; Durán
et al. 2013). However, while agricultural soils emit approximately 70 % of the global
nitrous oxide (N2O) emissions (Baggs 2011), a GHG 298 times more potent than
carbon dioxide (CO2) (IPCC 2013). Emissions from urban areas such as turf grass
are currently not even included in the inventories as reliable data are lacking global
distribution.
Soil microbial activity, driving GHG emissions, is higher in the tropics compared
to temperate zones due to the consistently warm and moist environmental conditions,
resulting in higher ecosystem productivity and C and N turnover. These microbial
22
favourable climate conditions make even native forests an important N2O source in
the tropics (Breuer et al. 2000; Kiese and Butterbach-Bahl 2002; Werner et al. 2007).
Based on this increased ecosystem productivity from temperate to tropical, it could
be assumed that GHG emissions from subtropical peri-urban environments and
native land use range in their intensity between temperate and tropical emissions.
Despite covering wide areas in Africa, South America, Asia and Australia, the
subtropical climatic zone represents an often-neglected area of research despite the
potential of contributing significantly to future climate change. The subtropical
climatic zone covers 3.26 M ha in Australia alone (AGO 2010), though large
uncertainties still exist around C and N cycling in many subtropical land uses.
Therefore, this research analyses the effect of land use change from native forest and
grazed pasture, representative for a main rural land uses in the area, into turf grass, to
evaluate the effect of urbanization on biogeochemical nutrient cycling in the
subtropical climate.
Changes in global climate are driven in part by the radiative forcing of the three
major GHG’s CO2, N2O and methane (CH4) in the atmosphere (IPCC 2007, 2013).
The biogeochemical carbon (C) and nitrogen (N) cycles play an essential role in
global climate change mitigation by immobilizing C from and minimizing NOx
losses into the atmosphere by increasing soil organic matter (SOM) (Lal 2004b), i.e.
C sequestration. Soils contain over three times more C than either the atmosphere or
living vegetation, which makes them the largest terrestrial C pool (Schlesinger 1990;
1995).
Nitrous oxide is produced principally by microorganisms during nitrification and
denitrification processes from mineral N (ammonium and nitrate) in the soil,
representing a N loss to the ecosystem as well as contributing to climate change
when emitted to the atmosphere (Butterbach-Bahl et al. 2013). Atmospheric CH4
uptake into the soil occurs via microbial consumption by methanotrophic bacteria for
an energy source. This is the largest natural sink of CH4 and is highly sensitive to
physical alterations of soil conditions and diffusivity. Soils can change to a CH4
source when methanogenic activity dominates in saturated soil moisture conditions
(Groffman and Pouyat 2009). Soil CH4 flux can generally be considered the net-
result of simultaneous occurring production and consumption processes in the soil
(Butterbach-Bahl and Papen 2002). Soils represent a major source of GHGs, and the
Introduction
23
magnitude of emissions is heavily influenced by anthropogenic land use practices
such as fertilization, irrigation and physical disturbance such as tillage. These
practices are all part of productive turf grass management and therefore these
ecosystems are potential GHG sources, increasing the Global Warming Potential
(GWP) of the landscape.
The fragmented distribution of land use types within peri-urban environments
makes the quantification of GHGs and subsequent estimations of their GWP
difficult. Collectively, turf grass lawn occupies over 15 M ha in the USA alone, three
times more than any other irrigated crop in the country (Milesi et al. 2005).
Additionally, the rapid biogeochemical changes during land use change as well as
high soil heterogeneity creates intensive GHG hotspots or moments which make an
accurate quantification and process understanding especially difficult. High-
frequency GHG flux measurements are therefore needed for accurate daily, annual
and inter-annual estimations.
Urbanization is currently neglected in modelled IPCC climate scenarios, mainly
due to limited data on C and N cycling in peri-urban environments (IPCC 2006,
2013), and only a few studies have examined the effect of land use changes
associated with urbanization on biogeochemical cycling (Grimm 2008; Betts 2007).
This highlights the need to quantify changes in the GWP of peri-urban environments.
1.2 Research problem
This research will determine the non-CO2 GWP of the land uses forest, pasture
and turf grass by quantifying inter-annual soil-atmosphere N2O and CH4 fluxes as
well as the long-term C sequestration potential. It will determine the impact that key
environmental parameters have on soil C and N cycling and GHGs in these adjacent
land uses in a subtropical climate. How do environmental parameters such as (i)
climate; (ii) soil type and initial nutrient status before land use change (young or
highly weathered mature soils and texture); and (iii) land use history (native or
agricultural) impact biogeochemical C and N cycling in peri-urban environments (iv)
over time?
24
(i) Influence of the climate
The humid subtropical climate of South East Queensland (SEQ), which is
characterised by extreme annual and inter-annual variations in rainfall, includes
intense rainfall events and rapid changes in soil moisture. Combined with high year-
round soil temperatures, soil conditions become favourable for microbial activity and
rapid biogeochemical cycling suggesting a potential for both increased soil GHG
emissions (Rowlings et al. 2012) as well as enhanced nutrient cycling (Xu et al.
2013).
(ii) Soil type and initial C and N status before land use change
Soil organic C has a significant influence on denitrification processes producing
N2O emissions (Fageria 2012), and is together with N and water the main factors
limiting soil fertility and plant growth (Marschner 2007, 2012). Plant biomass
production subsequently drives soil organic matter (SOM) accumulation in form of C
sequestration in the soil. It is this interaction of the C and N biogeochemical cycles,
which needs to be evaluated for the C sequestration potential in peri-urban soils.
(iii) Land use history
Land use change in peri-urban environments can have positive or negative
consequences, depending on their land use history. Some negative consequences of
land use change from natural ecosystems to agriculture include a loss in soil quality
(structure and nutrient losses) and quantity (erosion), increased GHG emissions, and
reduced potential for soil C sequestration (Livesley et al. 2009; Grover et al. 2012).
Natural ecosystems, for example, are estimated to sequester 3.55 Pg CO2-e y-1
into
both soil and plant biomass (Dalal and Allen 2008) and therefore play a significant
role in mitigating climate change. The sequestration potential of ecosystems
however, can be reduced substantially when disturbed during land use change and
construction processes such as plant cover removal and soil cultivation. On the other
hand, intensively managed turf grass systems can increase C sequestration when
changed from seasonal cropping or extensively used grasslands in rural areas
(Golubiewski 2006; Raciti et al. 2011a; Brown et al. 2012).
(iv) C and N dynamics over time
Land use change need to be evaluated within several time scales as
biogeochemical cycling may affect soil-atmosphere GHG dynamics differently over
Introduction
25
time. Therefore, immediate, short, medium and long term C and N cycling responses
to land use change will be estimated. The immediate term reflects the change of input
such as fertilizer use, i.e. mineralization and soil disturbance. The short term
response is less affected by the soil disturbance and reflects the seasonal dynamics of
the newly established land use and increased immobilisation. The medium term
response takes inter-annual climate variations into account and the long term
response evaluates the new mineralization and immobilization equilibrium by
identifying the C and N storage capacity via the C sequestration of the new land use.
This C sequestration potential of peri-urban environments is based on the increased
ecosystem productivity due to fertilizer and irrigation practices of the turf grass
management. These intensified management practices however, might as well
increase the soil-atmosphere GHG exchange of CH4 and especially N2O, which then
limits the positive effect of C sequestration on the climate. Therefore, the C
sequestration potential of peri-urban environments needs to be evaluated in
combination with soil-atmosphere GHG exchange to estimate an accurate long-term
GWP. This study examines native forest and grazed pasture for comparison to the
establishment of a residential turf grass across multiple time scales to identify
alterations of C and N cycling after urbanization related land use change. Therefore,
an inter-annual non-CO2 GWP based on high frequency CH4 and N2O measurements
was complemented with the long-term soil C sequestration potential of peri-urban
turf grass compared to forest and pasture.
1.3 Research aim and objectives
Land use change associated with urbanization can impact biogeochemical nutrient
cycling in the transitioning environment. This research aims to identify how land use
change from native forest and grazed pasture to a peri-urban environment alters C
and N cycling and develops over time. These changes in C and N cycling need to be
determined for multiple time scales, as immediate and long-term ecosystem
responses can differ substantially. It is hypothesized that land use change associated
with urbanization significantly alters nutrient cycling and increases the non-CO2
GWP.
26
To evaluate this hypothesis, the following objectives will be addressed through
experimental design (Chapter 3).
Objective 1 – Evaluate the immediate ecosystem GHG exchange response to land
use change into peri-urban turf grass.
Hypothesis 1: Turf grass establishment increases soil N2O emissions and reduces
CH4 uptake within the first 90 days following land-use change from well-established
land uses such as native forest and grazed pasture due to soil disturbance and
increased inputs of N fertilizer and irrigation practices.
Objective 2 – Evaluate the annual ecosystem N cycling response after land use
change into turf grass to account for seasonal variation of the potent GHG N2O.
Hypothesis 2: Land use change associated with urbanization increases annual
ecosystem N losses in the form of N2O due to increased fertilizer use in turf grass
systems and plant cover removal during construction processes.
Objective 3 – Evaluate the current non-CO2 GWP of peri-urban environments in
subtropical Australia.
Hypothesis 3: Peri-urban land use significantly increases the non-CO2 GWP
compared to native forest by increasing N2O emissions and reducing CH4 uptake
with inter-annual significance due to differences in annual environmental conditions.
Objective 4 – Evaluate the longer-term effect of land use change on C and N cycling
by identifying the soil C sequestration potential in peri-urban environments.
Hypothesis 4: Peri-urban turf grass establishment significantly affects C and N
cycling in the long-term by increasing the soil’s C sequestration potential compared
to pasture and forest.
Introduction
27
1.4 Method & Outcome
The research plan has four major components, each representing one of the core
objectives above. Each objective highlights the impact of land use change on C and
N cycling over time such as immediate, annual and inter-annual N2O and CH4 fluxes
and long term C and N dynamics in form of a C sequestration potential. This
approach identifies the development of soil-atmosphere gas exchange dynamics over
time after land use change to estimate a non-CO2 GWP of a subtropical peri-urban
ecosystem as well as the long term ecosystem response in form of C sequestration.
Objectives 1-3 evaluated N2O and CH4 soil-atmosphere gas exchange dynamics
continuously over two years using high temporal frequency measurements. Objective
4 identified the long-term effect on the C and N cycle by estimating the C
sequestration of the main peri-urban land use types native forest, grazed pasture and
private and public turf grass within a soil survey across the Samford Valley.
Figure 1-1 illustrates the hypothesized ecosystem response to land use change into
peri-urban turf grass over multiple points in time. These changes are highlighted by
the developing non-CO2 GWP compared to the pasture it was converted from. The C
sequestration potential, which would influence the long-term GWP by storing GHG
related C and N in the soil, of the turf grass lawn compared to forest and pasture
could therefore balance the GHG emissions hypothesized in the long term.
28
Figure 1-1 Hypothesized multiple time-scale response of the Global Warming
Potential of newly established turf grass associated with urbanization processes when
compared to forest and pasture.
The outcome of this research is the first inter-annual non-CO2 GWP estimate for a
subtropical peri-urban environment using soil-atmosphere N2O and CH4 flux
measurements from a native forest, grazed pasture and newly established residential
turf grass system. This high temporal frequency GHG dataset not only provides the
foundation for continuous research on future urbanization processes in the Samford
Valley, but can be used to improve global GHG budget estimations and modelled
future climate scenarios. Furthermore, the soil survey highlights the long-term C
sequestration potential from these common peri-urban land use types and estimates
the C and N stock from one of the most widespread soil types in Australia.
1.5 Significance
Land use change associated with urbanization involves substantial changes to
ecosystems worldwide. Topsoil displacement and mixing during construction
processes in peri-urban environments can result in substantial nutrient losses and
increased emissions of GHG. High quality GHG flux measurements are needed to
calibrate process models for (1) climate change scenarios from current conditions
and, more importantly, (2) to identify the most efficient GHG mitigation and C
Introduction
29
sequestration strategies. Ecosystem response in form of soil-atmosphere GHG
exchange can substantially change over the immediate to inter-annual and can give
significantly different process model outputs when calibrated with a limited amount
of data.
This research demonstrated that soil-atmosphere gas exchange dynamics in a
subtropical peri-urban environment quickly stabilise and reach a new equilibrium
after land use change. In addition, the comprehensive data set developed here
improves our understanding of the climates influence on nutrient cycling in land use
change affected ecosystems. The subtropical of forest, pasture and turf grass soils
indicates a tight N cycle with a close coupling of soil N turnover and plant uptake,
which minimized losses and results in significantly less GHG emissions than
temperate ecosystems undergoing similar land use change. Based on this study, peri-
urban turf grass systems in subtropical environments become comparable in long-
term soil C sequestration to forest and pasture land use. The adjusted nutrient cycling
in subtropical turf grass systems might therefore not offset the GHG emissions
resulting from the fertilization, irrigation and mowing practices as it was suggested
by temperate turf grass systems. Reducing N fertilizer inputs in these subtropical
peri-urban environments could be a promising strategy to reduce the GWP of turf
grass, especially since the climate supports an efficiently tight N cycle. Overall, this
research outcome encourages further consideration about global climate change
mitigation strategies by identifying subtropical peri-urban environments as
substantial C and N pool with minor GHG emissions.
Literature Review
31
Chapter 2: Literature Review
Land use change is increasingly altering ecosystem functionality worldwide. Due
to the large size, intensity, and global distribution of these changing environments,
management strategies need to be adapted to mitigate nutrient losses and sustain
ecosystem productivity in the future. However, the biodiversity and heterogeneity of
ecosystems demands that these strategies must be modified according to specific
environmental conditions and a better understanding of biogeochemical cycling and
drivers. The first step to creating these complex strategies is analysing the current
knowledge base and identifying knowledge gaps. Current research suggests
significant changes in biogeochemical cycling of ecosystems as a result of land use
change. How these changes in biogeochemical processes influence soil-atmosphere
gas exchange and nutrient sequestration in ecosystems is only beginning to be
understood. It is hypothesized that land use change associated with urbanization will
affect the C and N cycle and alter the soil-atmosphere GHG exchange and future
climate. This literature review analyses current and predicted urbanization processes
driving land use change, the effect on the biogeochemical C and N cycle as well as
the impact of climate change on the environment.
2.1 Land use change associated with urbanization
This section will analyse current information available on historical and predicted
urbanization based on population increase and migration, and how these dynamics
drive land use change.
2.1.1 Urbanization background
The world population of 7.4 billion is projected to increase by almost one billion
people within the next twelve years under a medium population growth rate, reaching
9.6 billion in 2050 and 10.9 billion by 2100 (Figure 2-1, United Nations (2013)).
Historically, the world population has lived in rural environments, close to
agricultural ecosystems. Urbanization has increased significantly over the last 50
32
years due to extensive population growth and migration from rural to urban
environments (United Nations 2014). Urban populations worldwide now exceed
rural populations and will account for all future population growth (Figure 2-2,
United Nations (2008)). This global urbanization process is becoming increasingly
important in terms of climate change and ecosystem productivity worldwide (Hutyra
et al. 2011).
Figure 2-1 Population of the world for the years 1950-2100, according to several
projections of the population increase based on of medium, high, low and constant
human fertility by the United Nations (2013).
Urban areas currently occupy up to 2.4 % of the terrestrial land surface and house
approximately 50% of the total population (Potere and Schneider 2007), and are
forecast to increase rapidly worldwide (Figure A 1, United Nations (2008)). For
example, urban ecosystems within or adjacent to cities in the USA cover 25% of the
total terrestrial land surface, and over 50 % of regional areas are affected by
urbanization (Kaye et al. 2004). In fact, Australia has one of the highest global
urbanization rates where greater than 90 % of the population are expected to be
living in urban areas by 2050 (United Nations 2014). Urban centres have 1,000
residents or more per 2.5 km2 (ABS 2012), with 18 major cities in Australia having
more than 100,000 residents (Commonwealth of Australia 2013) (Figure A 2). In
2011, 77 % of Australia’s population lived in these 18 major cities, increasing to
Literature Review
33
88 % of the population if peri-urban environments surrounding those major cities are
included (Commonwealth of Australia 2013).
Figure 2-2 Global amount of people living in urban and rural environments from
1950 to 2050 (United Nations 2008).
Half of Australia’s urban population is distributed between the two largest cities,
Sydney and Melbourne (Figure A 3), while Brisbane presents the most extensive
urban sprawl of all Australian cities (Commonwealth of Australia 2013). Between
2011 and 2012 the population of Australia’s capital cities grew by 1.8 % per year,
faster than the national average of 1.2 % and the second highest growth rate within
OECD countries (Figure A 4). Brisbane currently has a population growth rate of
1.7 % per year with a population density of approximately 140 people per km2 (ABS
2015), while only 6 % of the Brisbane population lives within the urban centre
(Table 2-1) (Spencer 2015). This proportion highlights that the vast majority of
Brisbane’s population is living in less dense populated environments with more
diverse land uses such as turf grass lawns, gardens and pastures and less sealed soils
compared to urban centres. To upscale the urban sprawl with the annual population
increase results in an expansion of the urban and peri-urban area by 276 km2 y
-1 if
the population density does not change. However, any increase in urban area is at the
expense of natural environments or agricultural land and the subsequent changes in C
and N cycling in these ecosystems will ultimately affect global climate.
34
Table 2-1 ‘Urban’ area and population compared to official metropolitan area of
major cities (Spencer 2015)
City Metropolitan area as defined in by census boundaries
‘Urban’ area: census areas with ≥ 4 people ha
-1
‘Urban’ area as a proportion of metropolitan area
Total area (ha)
Population Area (ha) Population Area Population
Brisbane 1,582,593 2,066,660 101,835 1,830,157 6 % 89 %
Melbourne 999,052 4,000,315 171,456 3,798,882 17 % 95 %
London 157,104 8,173,941 126,075 8,122,564 80 % 99 %
2.1.2 Land use change impact on the environment
The process of urbanization includes significant land use changes with
deforestation and the conversion of native grasslands and commercially focused
agriculture into smaller residential properties and partial soil sealing (IPCC 2006).
Clearing natural vegetation has the strongest impact on the environment affecting
local ecosystem health as well as water quality and nutrient cycling (Maraseni et al.
2012). Perennial ecosystems are estimated to sequester 3.55 Pg CO2-e y-1
(Pg =
petagram = one billion tons) within soil and plant biomass (Dalal and Allen 2008)
and therefore play a significant role in mitigating climate change. During urban
expansion, the native or agricultural vegetation is removed before conversion to
residential space. Additionally, changes in the environment through dwelling
construction, hard infrastructure, foreign ornamental plants in gardens and parks and
invasive weeds have increased the level of ecological dysfunction of the landscape
(MacLeod and Kearney 2007).
The consequences of land use change from natural ecosystems to agriculture
include a loss in soil quality (structure and nutrient losses) and quantity (erosion),
increase GHG emissions, and reduce soil potential for C sequestration (Livesley et
al. 2009; Grover et al. 2012). For example, soil erosion can degrade environmental
productivity and is estimated to contribute about 1.14 Pg C y-1
to annual C losses
through erosion-induced processes (Lal 2004a). On the other hand, soils that have
converted from agricultural into residential use have the potential to improve critical
ecosystem services such as stormwater treatment and storage, sinks for atmospheric
N as well as C sequestration (Golubiewski 2006; Raciti et al. 2011a) by covering the
fertile topsoil with perennial plants which prevents erosion.
Literature Review
35
Grassland, as perennial land use, cover about 40 % of the terrestrial ice-free
surface worldwide (AGO 2010), storing approximately one third of the global
terrestrial C pool (IPCC 2007). This data however, is not partitioned into native,
agricultural (grasslands such as introduced or improved pastures), and residential use
(such as turf grass lawns) and therefore gives no clear definition of land use
management. It has been shown that managed peri-urban grasslands alter
biogeochemical C and N cycling substantially in temperate climates, significantly
increasing GHG emissions (Kaye et al. 2004; Conant et al. 2005) as well as C
sequestration (Golubiewski 2006; Raciti et al. 2011a). Given the large areas
worldwide undergoing these changes in land use, the resulting ecosystem
productivity and fertility will effect climate change and the global food security in
the long term.
The major land use change during transition from rural to peri-urban and urban
environments is the inclusion of turf grass, often in combination with extensive
construction processes (Kaye et al. 2005; Milesi et al. 2005; Pouyat et al. 2009). For
example, the conversion of rural grasslands to urban use introduces fertilization and
irrigation practices, which can increase emissions of CO2 and N2O and decrease net
sinks of CH4 (Kaye et al. 2004; Lorenz and Lal 2009). As the dominant peri-urban
vegetation and land use, turf grass is extensively established for residential
backyards, public parks, and golf courses (Milesi et al. 2005). Mown grass lawns and
golf courses, for example, are greater CO2 sources than the natural vegetation that
they replaced (Koerner and Klopatek 2002; 2010). This greater CO2 source is due to
increasing plant and soil respiration from increased ecosystem productivity as well as
CO2 emissions from the management practices themselves such as fossil fuel burning
during mowing, fertilizer production and distribution. Additionally, urbanization
significantly increased N2O emissions compared to native landscapes in the arid
regions of Arizona, USA, primarily due to the expansion of fertilized and irrigated
lawns (Hall et al. 2008).
Despite arguments that urban and peri-urban areas are too small to contribute
important biogeochemical fluxes on global scales (Kaye et al. 2004), urban lawns in
the USA currently cover over 160,000 km2, three times larger than any other
irrigated crop (Milesi et al. 2005; Groffman and Pouyat 2009). Florida, for example,
produces over 42,000 ha of commercial turf grass on its humid subtropical sandy
36
soils per year with a total economic impact estimated at $703 million USD
(Satterthwaite et al. 2009). No detailed estimates about the current turf grass cover in
Australia’s urban and peri-urban environments do currently exist, but annual turf
grass sales range between 4,918 ha and 17,320 ha over the last 10 years (ABS 2012;
Turf Australia 2012). The approximate gross value production of Australia’s turf
industry is $240 million AUD per annum, with over 40 % being produced by tropical
and sub-tropical Queensland suppliers (ABS 2012; Turf Australia 2012). The rapid
growth of the turf grass industry worldwide highlights the need for detailed
information to accurately predict trends within the turf grass industry to improve
economic and environmental benefits.
The importance of land use change from native land to urban areas has recently
become the focus of global change research (Pataki et al. 2007). The expansion of
urban areas has a significant influence on SOC storage by introducing turf grass
cover and intensive management (Golubiewski 2006; Kaye et al. 2005; Milesi et al.
2005). Despite the fact that some of these ecosystems have nearly the same
fertilization and irrigation inputs as agricultural land use, urban and peri-urban areas
are neglected so far in global IPCC climate change forecast scenarios (IPCC 2006,
2013).
Urban areas are still the least understood of all ecosystems when it comes to
climate change interactions (Durán et al. 2013), mainly because of the difficulties to
define urban land use and examine residential properties and the wide variety of
plant cover. With population growth and urbanization continuing to place pressure
on the natural environment, there is growing recognition of the need to manage urban
and peri-urban ecosystems to ensure they remain liveable. Therefore, researchers,
policy-makers, planners and resource managers are increasing their attentions to
greening the built environment in many major cities in Australia (Commonwealth of
Australia 2013). However, whether this greening, especially using intensive turf
grass, accelerates or mitigates climate change remains unknown.
About one tenth of Australia’s net GHG emissions of 525,202 Pg CO2-e y-1
(AGEIS 2015) is a result of land use and management change (Hatfield-Dodds et al.
2015). Agricultural soils emit approximately 70 % of the global nitrous N2O
emissions (Baggs 2011), a GHG 298 times more potent than CO2 (IPCC 2013). Data
from temperate zones identifies turf grass can contribute to climate change via GHG
Literature Review
37
emissions to a comparable degree as intensive agriculture on a unit area basis (Kaye
et al. 2004; Durán et al. 2013). However, there is potential to significantly reduce
those GHG emissions from turf grassland use by modifying current management
strategies. For example, modelled scenarios in the USA identified a substantial long-
term benefit for GHG reduction from turf grass systems with the simple management
practice of reducing fertilizer input and leaving mowed grass clipping behind instead
of removing them from the ecosystem (Zhang et al. 2013a; 2013b). To calibrate
process models to develop improved management strategies, new datasets from
public and private peri-urban environments are needed from various climates and soil
types to accurately predict local and global climate change impacts, which are often
neglected due to limited data available (Betts 2007; IPCC 2006, 2013).
2.2 Land use and climate change implications
As land use change can substantially alter biogeochemical processes, an
examination of the resulting GHG fluxes is required (Tratalos et al. 2007; Grimm et
al. 2000). The biogeochemical and physical effects of land use change due to
urbanization are often neglected when modelling future scenarios of climate change
(Betts 2007). Even the highly weathered soils of Australia have the potential for
substantial C sequestration to balance GHG emissions from management practices
(Grace and Basso 2012). However, in their ancient and fragile condition, climatic
changes are predicted to affect the net GHG sink and source behaviour of Australian
soils in the long-term (Baldock et al. 2012). Therefore, accurate baseline estimations
are needed to develop effective policies to offset GHG emissions (Viscarra Rossel et
al. 2014). This section will analyse current information available on GHG driven
climate change on the socio-ecological dynamic in peri-urban environments by
discussing the quantitative and qualitative impact of land use change associated with
urbanization and feedback effects from the changing global climate.
2.2.1 Climate change dynamics
Since the 1990s, IPCC reports are the most recognised achievements of all
international working groups on climate change science. In short, they identify all
climate system components in all timescales (past, present, future) and state
38
observations and process understanding on biogeochemical cycling. They claim to
have all the information included on natural and anthropogenic drivers of climate
change available to date. This all-embracing information is then used to model future
climate change scenarios worldwide. These results predict an increase in temperature
worldwide of about 2 °C by 2050 for most forecast scenarios (Figure 2-3).
Figure 2-3 IPCC 2013: Time series of temperature change relative to 1986–2005
averaged over land grid points over the globe in December to February calculated
from a variety of Representative Concentration Pathways (RCPs) from the radiative
forcing (+2.6, +4.5, +6.0, and +8.5 W m-2
, respectively) of greenhouse gas
concentration in the atmosphere (Stocker et al. 2013).
The rising global temperature is driven by the radiative forcing of various GHGs
in the atmosphere. The current radiative forcing of the three major GHGs is 1.46
w/m2 for CO2, 0.5 w/m
2 for CH4 and 0.15 w/m
2 for N2O (Lal 2004b). Historically,
the concentration increase of all GHGs combined (Figure 2-4) has already increased
the average global surface temperature by 0.6 jC since the late 19th century and is
currently warming with a rate of 0.17 jC/decade (IPCC 2001). This global
temperature increase will lead to higher water evapotranspiration into the atmosphere
and consequently higher rainfall (Stocker et al. 2013). However, due to higher
evapotranspiration of plants responding to heat stress, Australian’s climate
conditions are likely to become overall drier even with higher annual rainfall
(Baldock et al. 2012).
Literature Review
39
Figure 2-4 Atmospheric concentrations of the three main long-lived greenhouse
gases over the last 2000 years. Increases since about 1750 are attributed to human
activities in the industrial era (Cubasch et al. 2001).
Worldwide, IPCC models predict an accelerated weather cycle with more frequent
and extremer weather events such as heavy rain events, storms and floods in some
regions while other regions will experience longer droughts. The land-surface
precipitation will continue to increase at the rate of 0.5 – 1 % per decade in much of
the northern hemisphere, but decrease in subtropical areas at the rate of 0.3 % per
decade (IPCC 2007).
The 2100 projection suggests significant temperature changes for Australia
(Figure A 6), leading to an overall increase in droughts as well as floods after heavy
rain events. The subtropical climate in SEQ is frequently subjected to extreme
weather events, which are expected to intensify with future climate change. For
example, January is the most rain intense month in SEQ on long-term average (BOM
2015), receiving extreme localized rainfall with 536 mm, nearly half of the annual
average, in just one week in Samford Valley in 2013 (Figure 2-5), while receiving
less than 100 mm over the full month in 2014.
40
Figure 2-5 Heavy rainfall across Australia with over 300 mm d-1 in Samford Valley,
SEQ (Highvale weather station, BOM (2015 in January 2013, Source:
Commonwealth of Australia (2013.
It is hypothesized that future climate change will affect local ecosystems
differently according to their region and environment (Lal 2004b). Growing seasons
will extend and vegetation zones will shift into northern regions and boreal forest
will increase their productivity. These northern regions in temperate climates are
currently a strong net C sink but could become a net C source with increasing
temperatures worldwide (Schlesinger 1995; Lal 2004b).
2.2.2 Feedback effects
The potential increase in GHG due to rising temperatures may accelerate global
warming and climate change. These feedback effects are incorporated into IPCC
projections. How intense these feedback effects are depends on the ecosystem and
the anthropogenic response to the changing environmental conditions. A conceptual
model of climate change and the role of ecosystem-atmosphere interactions was
visualized by Betts (2007) (Figure 2-6).
Literature Review
41
Figure 2-6 Conceptual model of climate change and the role of land ecosystem-
atmosphere interactions (Betts 2007).
To date, the most well understood environmental response and feedback effect of
urbanization is the localized temperature increase, i.e. heat island (Oke 1982). The
replacement of vegetation with sealed surfaces of materials with high thermal
conductivity, high heat storage capacity and low albedo (reflectivity) enables cities to
influence local and global climates. This is where cities and their suburbs have
significantly warmer air and surface temperatures than rural areas (Commonwealth
of Australia 2013). Consequently, this temperature increase can enhance soil-
atmosphere gas exchange in urban green spaces as well as O3 concentrations.
Furthermore, urbanization has been observed to significantly affect precipitation by
producing large quantities of condensation and increasing the variability of local
heavy rain events (Oke 1982; Groffman et al. 1995).
Ecosystem vulnerability
The anthropogenic induced changes in global temperatures can be quantified as
the GWP of GHG emissions from ecosystems; however, the quality of the changed
ecosystem is much harder to estimate. A definition for human-environment
interactions by Turner et al. (2005) is: ‘Vulnerability is the degree to which a system,
subsystem, or system component is likely to experience harm due to exposure to a
hazard, either a perturbation or stress/stressor’’. With an emphasis on climate
change, the Intergovernmental Panel on Climate Change (IPCC) defined ecosystem
vulnerability as ‘the degree to which a system is susceptible to, or unable to cope
with, adverse effects of climate change, including climate variability and extremes’.
42
Overall, ecosystem vulnerability can be seen as a quantifiable factor contributing to
the risk of harm from human-induced climate change (Raupach et al. 2011). There is
increasing interest in estimating the anthropogenic impact on a socio-ecological basis
to identify not only the quantity but also the quality of our changed urban and peri-
urban environment to keep our cities liveable (Commonwealth of Australia 2013).
Furthermore, this socio-ecological relationship affects the environment as well as
anthropogenic activities to cope with these changing environments.
Link of local climate to global change
There is increasing evidence that areas undergoing urbanization significantly
influence their local climate (Betts 2007). With these areas constantly expanding
worldwide it is most likely that these locally changing climates combined will have
an increasing impact on global climate change. For example, CO2 concentrations
around cities can exceed 500 ppm (Pataki et al. 2007), where the global average is
currently 400 ppm (NOAA 2016). Byrne (2007) summarized key conclusions from
research associated with urbanization. These conclusions are; (1) habitat structure
provides a unifying theme for multivariate research about urban soil ecology; (2)
heterogeneous urban habitat structures influence soil ecological variables in different
ways; (3) more research is needed to understand relationships among sociological
variables, habitat structure patterns and urban soil ecology. Furthermore, this urban
soil ecology is strongly interlinked with the biogeochemistry of the C and N cycle,
which is increasingly driven by anthropogenic and environmental changes, i.e. land
use change. The relationships between these variables, drivers and changes are
illustrated in the framework by Grimm (2008) (Figure 2-7). This framework
highlights how local, regional and global environmental changes are interlinked and
together affect global climate change. However, we are currently only beginning to
understand how the process of urbanization influences both ecosystem dynamics in
their biogeochemical cycling and contributes to global climate change (Kaye et al.
2004; Byrne 2007; Lorenz and Lal 2009; Durán et al. 2013).
Literature Review
43
Figure 2-7 Socio-ecological framework by Grimm (2008) identifying the drivers and
responders of climate change on a local, regional and global scale.
While the trend of urbanization is global, the impact is not evenly distributed. For
example, 13 M ha are deforested yearly, but almost exclusively in tropical regions
(Canadell and Raupach 2008). Tropical deforestation as the major land use change
released annually about 1.5 Pg C to the atmosphere, accounting for almost 20 % of
anthropogenic GHG emissions in the 1990s (Gullison et al. 2007). On the other hand,
terrestrial ecosystems remove nearly 3 Pg of anthropogenic C every year through net
growth, absorbing about 30 % of all CO2 emissions from fossil fuel burning and net
deforestation (Canadell and Raupach 2008). If not deforested, the area in tropical
regions would account annually for 65 % of the total C offset (IPCC 2007).
Terrestrial C offset is the potential of C sequestration from the atmosphere into soils
and biomass (Conant et al. 2011). IPCC (2007) estimated an economic potential C
offset of 0.12 Pg C y−1
reachable by 2030 if we would start pricing our GHG
emissions with U.S. $20 per ton of CO2-e. These strategies could be included into
international policies for climate change mitigation, but because of the fragmentary
estimations and baseline data (Viscarra Rossel et al. 2014), it is still controversial to
make GHG budgeting compulsory worldwide.
44
Economic impact of land use change into peri-urban environments
Turf grass is the most highly managed land use of peri-urban environments and
shows the potential for GHG emissions comparable to intensive agriculture (Kaye et
al. 2004; Durán et al. 2013). However, some management practices show potential to
reduce the high GWP found in intensive agriculture (Trlica and Brown 2013). An
increase in C sequestration can reduce the GWP as well as the economic cost for
farmers by improving soil fertility (Grace et al. 2010; Grace and Basso 2012). To
encourage C sequestration practices, the in Carbon Farming Initiative (CFI),
promoted carbon credits for the land-based sectors in Australia, with emphasis on
emissions reporting, trading and C foot-printing (Cowie et al. 2012; Grace and Basso
2012). Currently, there is limited information on C footprints or GHG emission
budgets available for natural and none for peri-urban environments in Australia.
2.3 Biogeochemical C and N cycling
This section will analyse current information available on the biogeochemical C
and N cycle including the terrestrial C and N pool and soil-atmosphere gas exchange
by discussing the driving environmental parameters.
2.3.1 Soil C and N
The main components of organic material via plant growth, C and N, can be
stored as SOM in the soil after littering. While C and N can also be added to the soil
via organic or mineral fertilization, the availability, storage and accumulation of
SOM in the long-term depends on various environmental factors, which can be
greatly influenced by anthropogenic management. Soil C sequestration is one of the
main strategies to mitigate climate change by storing CO2 from the atmosphere into
SOM as well as achieving global food security by increasing the soils capacity to
store essential nutrients for plant growth and land use productivity (Lal 2004a). As
much as soils have the potential to take up anthropogenic produced CO2 and CH4
from the atmosphere, soils can also contribute to climate change by releasing GHGs
from the current pool of C and N ~ 2500 Pg C (Lal 2004b) and 190 Pg N
respectively (Baldock et al. 2012).
Literature Review
45
Organic C and N accumulation in soils is driven by processes of plant production
and decomposition, which are greatly influenced by biotic and abiotic parameters
(Morris et al. 2010). Anthropogenic management, such as fertilization and soil
disturbance, can alter these parameters substantially. However, few assessments to
date evaluate how disturbance, land-use history, and age of residential soils influence
C and N pools and fluxes (Raciti et al. 2011a). Urban green spaces have been
associated with substantially greater soil C and N accumulation than native forests
(Raciti et al. 2011a) and grasslands (Golubiewski 2006) in the temperate climate of
the USA. This increased C and N accumulation within urban green spaces suggests
that the global C and N pool could be underestimated due to the exclusion of urban
land cover despite the increase of sealed surfaces within the environment. Increased
C and N accumulation in urban and peri-urban soils can be explained by the higher
plant productivity with increasing management practices, such as fertilization and
irrigation, and consequently higher SOM production. The strong correlation of SOM
to its C and N content makes it a strong index of soil quality (Fageria 2012).
Identifying an ecosystems productivity and capacity to sequester C and N is therefore
a crucial factor to determine the long-term GWP of natural and managed
environments.
Soil C
The main component of SOM is soil organic C (SOC), making up about 58 % of
SOM (Lal 2004b; Cotrufo et al. 2011). The majority of the global soil C pool of
1550 Pg is in the form of SOC (Lal 2004a), and this is estimated to increase through
soil C sequestration by approximately 24 kg C ha-1
y-1
on average and over 100 kg C
ha-1
y-1
in forests (Schlesinger 1990). With global carbon stocks three times higher
than the atmosphere (Schlesinger 1990; Lal 2004b) (Figure A 5), soil management
has an enormous potential to influence global climate. Based on the fact that the C
cycle happens to be slow in its changes, long-term investigations are indispensable to
understand the C sequestration mechanisms (Grace and Basso 2012). These slow
processes complicate the determination of the ecosystems response to land use
change as they differ temporally and spatially due to the environmental
heterogeneity.
Currently, estimates of SOC are largely unavailable or uncertain for large areas
globally. However, Viscarra Rossel et al. (2014) attempted to create a baseline from
46
the scattered data available of current SOC levels in Australia, averaging 29.7 t ha-1
in 30 cm topsoil. With Australia occupying 5.2 % of the global terrestrial area, these
identified SOC contents add up to 25 Pg SOC in the topsoil, which represents
approximately 3.5 % of the global terrestrial C pool.
Soil N
The C and N cycle is driven by several environmental parameters including
climate, soil physical and chemical conditions and microbial activity. The link
between these two cycles is expressed through the influence of N availability from
litter quality and quantity, which effects microbial activity and drives C dynamics
(Figure 2-8) (Pastor and Post 1986; Groffman et al. 1995).
Figure 2-8 Conceptual model of links between net primary productivity, litter C,
decomposition, microbial trace gas fluxes and soil N availability and their main
driving parameters climate and soil moisture (Pastor and Post 1986; Groffman et al.
1995).
Soil N can be stored in organic form in the long term but is mostly available for
plant uptake in the mineral N form of ammonium (NH4+) and nitrate (NO3
-)
(Marschner 2012). Nitrogen availability depends on the microbial decomposition of
SOM and surface litter inputs and is enhanced in warm and moist environments that
favour soil microorganism growth. Changes in soil temperature and moisture
conditions due to modifications in plant cover or physical disturbance during land
use change can have a substantial impact on the N availability in soils (Groffman et
al. 1995). Other environmental parameters such as soil texture, pH and cation
Literature Review
47
exchange capacity (CEC) also significantly influence N availability in soils by
potentially immobilizing and fixing N in clay aggregates or transforming the stable
NH4+
into the mobile forms of NO3- (Blume et al. 2015). The highly mobile NO3
- can
be leached out of the soil profile and pollute groundwater and waterways. Nitrate can
cause eutrophication, an extensive algae growth which consequently results in
oxygen limitation and finally makes water bodies unable to support life
(Vollenweider 1970). Therefore, managing N availability from litter as well as
mitigating possible N losses throughout the nutrient cycle is one of the major factors
supporting soil and plant productivity, which eventually drives C dynamics.
C sequestration
Soil C sequestration implies the removal of atmospheric CO2 by plants and
storage of fixed C as SOM. This SOM production is highly dependent on the N
availability in the ecosystem, as N is the primary plant growth-limiting nutrient
(Marschner 2007, 2012; Neal et al. 2013). Most accumulation of SOM happens in
surface soils; C content increase in subsoils is mostly caused by increasing bulk
density (Golubiewski 2006; Bolstad and Vose 2005). This strategy to increase SOM
density in the topsoil and improve the distribution through depth, stabilizes C and N
by encapsulating it within stable microaggregates so that it is protected from
microbial processes as recalcitrant C with long turnover time (Lal 2004b). The
balance from accumulated C and N within SOM and the release via soil-atmosphere
gas exchange determines an ecosystem a C sink or source. The anthropogenic
influence, however, can transform the terrestrial C pool from a sink to a large source
by losing more C than input into cultivated and managed land use.
C Fractionation
The process of C sequestration is mediated by microorganisms and the transfer of
C from easily decomposable into recalcitrant forms, with the three major conceptual
fractions defined as active, slow and resistant C (Parton et al. 1987; Parton 1996).
Carbon fractionation can provide more information on the longevity of SOM in the
soil and its sensitivity to land-use change. Skjemstad et al. (2004) and Baldock et al.
(2012) introduced a combination of physical and chemical properties to allocate
SOM to the following three fractions:
48
(1) particulate organic carbon (active C) associated with particles > 50 mm
(excluding charcoal carbon); easily oxidated by soil microbes
(2) humus organic carbon (slow C) associated with particles < 50 mm (excluding
charcoal carbon); no direct oxidation by microbes but chemical processes such as
acid or base
(3) recalcitrant organic carbon (resistant C), found in the < 2 mm soil in poly-
aromatic chemical structure, consistent with the structure of charcoal, and resistant to
microbial and chemical processes.
Once transferred into the resistant fraction, SOM is recalcitrant in the global C
and N cycle in the long term and is not actively participating in the soil-atmosphere
GHG exchange compared to the slow and active C fractions (Conant et al. 2004; Paul
et al. 2008b). This transfer of SOM into the resistant fraction is mainly due to the
stabilization into macroaggregate-occluded microaggregates, which are formed by
clay colloids (Denef et al. 2007). Therefore, SOM has to be separated into their
physical and chemical properties to identify the transitional fractions within the C
and N cycle (Six et al. 2000).
Introducing fertilization and irrigation practices within peri-urban environments
has the potential to increase SOM through higher plant productivity and therefore
support C sequestration. On the other hand, excessive fertilizer use can increase N
losses when the nutrient holding capacity of the soil is exceeded. For example,
negatively charged clay minerals can hold NH4+
while sand relies on SOM to
improve its very low nutrient holding capacity (Blume et al. 2015). Plant roots and
microorganisms can interact with the soil and make nutrients available through soil
chemistry changes (Marschner 2007; Miller et al. 2007). The literature suggests that
management practices do not significantly change N cycling in peri-urban
environments by sustaining a tight coupling of N mineralization and immobilization
in the long-term due to an increased microbial C and N use efficiency (Shi et al.
2006). Others identified a clear change from nitrification to denitrification dominated
N cycling (Raciti et al. 2011b), which then suggests additional N losses in form of
other gaseous emissions such as N2, NO and NOX (Del Grosso et al. 2000). These
ecosystem N losses combined need to be compensated for to keep urban green spaces
and peri-urban land use highly efficient in SOM production. However, soil type is
Literature Review
49
one of the key factors determining if management practices such as fertilization and
irrigation results an accumulation of SOM or increased GHG emissions.
Climate
Generally, the SOC pool varies widely among ecosystems, being larger in cool
and moist environments compared to warm and dry regions (Lal 2004b).
Approximately 40% of global SOC is stored in subtropical and tropical ecosystems
(Lal 2004b) but it is being rapidly lost due to continuous deforestation (Richards et
al. 2007). It has been identified that land use change, such as deforestation and
degradation, in tropical climates can reduce SOC up to 75 %, which is then released
into the atmosphere (Lal 2004a), while some suggest a SOC reduction of even up to
97 % using IPCC developed C accounting methods (Ogle et al. 2004). However,
only 15 % of the C and N related studies on land use change are from tropical
(Veldkamp 1994; Paul et al. 2008b; Paul et al. 2008a) and even less from subtropical
regions (Conant et al. 2001).
Soil and land use history
Older soils, such as in Australia, are considered to have a lower C sequestration
potential than younger soils (< 3,000 years) and could become overall net C sources
with increasing global warming (Schlesinger 1990). Long term C accumulation in
3,000 to 10,000 year old soils determined from chronosequence studies varies from
2 kg C ha-1
y-1
in polar deserts to >100 kg C ha-1
y-1
in temperate forests, with an
average C sequestration of 24 ± 7 kg C ha-1
y-1
for tropical rain forests (Schlesinger
1990). Once, the maximum C storage capacity of the soil is reached, i.e. C saturation,
C contents level out by increasing C losses such as soil respiration (Six et al. 2002).
Native SOC levels reflect the balance of C inputs and C losses under native
conditions (i.e. productivity, moisture and temperature regimes), but do not
necessarily represent an upper limit in soil C stocks of that particular ecosystem.
Based on the C fractionation scheme described earlier, recent research suggests that
every fraction has its own C saturation limit (Mitchell et al. 2016), which makes
SOM management strategies a long-term investment. This means C contents in
intensively managed agricultural and pastoral ecosystems can exceed those under
native conditions (Six et al. 2002; Stewart et al. 2007; 2008). Therefore, some studies
suggests that local, regional, and global SOC pools are most likely underestimated by
50
neglecting urban and peri-urban ecosystems (Kaye et al. 2005; Pouyat et al. 2006;
Golubiewski 2006) despite the partly sealed soils.Forest
To date, most Australian land use change studies are focusing on deforestation
and afforestation (Forster 2006; Canadell and Raupach 2008; Newham et al. 2011;
Raupach et al. 2011). Afforestation of marginal agricultural soils or degraded soils
has a great potential of C sequestration but largely depends on climate, soil type, tree
species and nutrient management (Veldkamp 1994; Lal 2004a, 2004b). Some
evidence from temperate climates suggests that C and N contents in afforested soils
will not reach previous levels again once deforested (Raciti et al. 2011a).
Additionally, Paul et al. (2002) reviewed afforestation investigations across Australia
and concluded that soil C contents decrease with time, while Richards et al. (2007
predict afforested areas with native species as a C sequestration strategy in SEQ. In
particular afforestation from degraded agricultural land in SEQ, investigated by
Maraseni et al. (2012), can increase soil C content, which suggest recovery
capabilities of Australian soils in that region. Additionally, a fire regeneration
chronosequence study in temperate Australia identified a decrease in forest soil CH4
uptake without fires or harvest reducing the organic C levels sporadically (Fest et al.
2015b). However, the main environmental drivers deciding about an increase or
decrease in soil C contents after afforestation are climate, previous land use and the
type of forest established (Paul et al. 2002).
Grasslands
Land use change from forest to pasture can increase the C and N content in about
70 % of the global studies (Conant et al. 2001). This increase in SOM, however,
might be only from the initial decomposition of tree roots in the soil and will
substantially decrease in the long term of up to 22 t SOC ha-1
after 25 years
(Veldkamp 1994). Cerri et al. (2004) determined that the soil C content and the
distribution across the physical fractions changed significantly after conversion from
forest to pasture, using a modified partial dispersion method developed by Six et al.
(2002. This influence on the physical fractions reveals the impact of land use change
on the long-term C storage potential in the soil by altering the physical soil
conditions, which build the microaggregates for SOM sequestration. However, if
these physical changes in soil conditions result in an increase or decrease of the
overall soil C content depends strongly upon the soil texture and the availability of
Literature Review
51
SOM in the active, slow or resistant form (Neill and Davidson 2000; Cerri et al.
2004; 2007), which is therefore highly variable for every ecosystem. Therefore, the
land use history and land use age at the time of sampling are crucial components in
long-term C and N cycling research. Overall, some evidence suggests that
management methods can improve the C sequestration of pasture soils ranging
worldwide from 0.11 to 3.04 t C ha-1
y-1
with a mean of 0.54 t C ha-1
y-1
(Conant et
al. 2001) and approximately 17 % in tropical regions (Ogle et al. 2004).
Peri-urban areas
Urbanization related land use change studies are rare and mostly from temperate
climates of the USA (Golubiewski 2006; Qian et al. 2010; Raciti et al. 2011a). The
few studies on peri-urban environments suggest that because of the increased
management in peri-urban environments, grasslands can accumulate even more SOC
compared to rural environments (Pouyat et al. 2002; Golubiewski 2006; Raciti et al.
2011a). Additionally, Raciti et al. (2011a) argues that because of generally high SOC
contents in forests soils close to saturation, the C uptake under grassland soils such as
pasture and turf grass is higher. Especially when degraded soils, such as under
agricultural use, are converted into perennial land use, such as peri-urban forests, can
enhance the SOC pool (Lal 2004b). The management, especially fertilization, can
increase C sequestration due to higher plant productivity (Conant et al. 2001;
Golubiewski 2006). On the other hand, Selhorst and Lal (2011) hypothesize that to
intensive management might offset the positive impact of the increased C
sequestration by increasing GHG emissions. Turf grass lawn is intensely managed
with N fertilizer additions, irrigation and frequent mowing to ensure high
productivity. This intense management suggests a potential increase in both C
sequestration and GHG emissions and therefore needs to be quantified to identify
peri-urban environments as an overall C and N sink or source. Australian research
recently increases on well-established turf grass systems as part of peri-urban
environments, while focusing on water and fertilization strategies to optimize turf
grass productivity while minimizing nutrient losses (Barton and Colmer 2006;
Barton et al. 2009a; Barton and Colmer 2011).
For example, Kong et al. (2014) investigated the C sequestration limit of turf
grass soils in the humid subtropical climate of Hong Kong, identifying a C
sequestration capacity ranging from 13 to 49 t C ha-1
in 15 cm topsoil with the
52
potential to be offset by CO2 emissions in 5 to 24 years of land use age. This
identified subtropical C sequestration potential is slightly lower compared to
temperate turf grass systems ranging from 21 to 96 t C ha-1
in 15 cm topsoil of the
USA (Selhorst and Lal 2013) and from 25 to 64 t C ha-1
in 10 cm topsoil of South
Australia (Livesley et al. 2010).
Land use change data from deforestation has been investigated for various
climates whilst studies on urbanization are still rare. The few studies available on
peri-urban turf grass land use mostly nutrient cycling in temperate environments
(Kong et al. 2014), with limited information on abiotic parameters such as soil
texture. However, estimates of the SOC pool in different Australia land uses is still
incomplete (Viscarra Rossel et al. 2014) and most likely underestimated by
neglecting residential turf grass systems. Overall, research on the C and N pool and
nutrient cycling driving parameters in peri-urban environments received too little
attention and gives therefore still limited information on the C and N sink and source
behaviour of urbanization effected soils.
2.3.2 Soil-atmosphere C and N exchange
To identify ecosystems as a net C and N sink or source, the potential for
sequestration as well as losses needs to be determined and quantified. Besides
nutrient losses through physical displacement with the soil, such as erosion and
construction processes, substantial amounts of C and N can be lost from the soil via
leaching and gaseous losses. These losses imply economic costs to the agricultural
sector as well as residential communities to keep land use highly productive while
mitigating negative consequences to peri-urban environments such as eutrophication
and global warming from emitted GHGs such as CO2, CH4 and N2O (IPCC 2014).
Soils represent a major source of these GHGs, with the magnitude of emissions
greatly influenced by anthropogenic practices such as fertilization, irrigation and
physical disturbance. In agricultural environments, various environmental parameters
have been identified driving these GHG fluxes such as soil moisture dynamics, soil
texture, nutrient input and substrate availability (Rowlings et al. 2012), factors
greatly modified by land use change. Fertilization and irrigation increases N2O
emissions and decreases potential CH4 uptake in the soil by increasing substrate and
limiting oxygen availability (Rowlings et al. 2013; Scheer et al. 2008).
Literature Review
53
Current GHG concentrations in the atmosphere are about 1.4, 2.5, and 1.2 times
the concentrations before industrial times for CO2, CH4, and N2O respectively (Table
2-2) and contribute ~ 60 %, 20 %, and 6 % respectively to global warming (Dalal
and Allen 2008). These GHGs increased over the last century and are influencing the
climate in the long-term (IPCC 2007), with the balance between exchanges of these
gases constitutes the net GWP of an ecosystem. The largest proportion of the
increase in atmospheric CO2 is from human activities, like fossil fuel use
(production, distribution and consumption) and land use (particularly land use
change and agricultural activity) (Dalal and Allen 2008). Increase in CH4 emissions
is mainly from fossil fuel use and agriculture, especially in wetlands and N2O
emissions are mainly from fertilizer use and waste management in agriculture. Soils
alone account for approximately 70 % of all N2O emissions (Baggs 2011).
Commonly used crop production practices generate CO2 and N2O and decrease the
soil sink for atmospheric CH4 (Mosier et al. 2005).
The few studies examining GHG emissions in urban environments have focused
mainly on CO2 exchange, while CH4 and N2O have often been neglected (Tratalos et
al. 2007; Lorenz and Lal 2009; Ng et al. 2014). However, potential terrestrial CO2
uptake can be offset by only minor increases in CH4 and N2O emissions (Tian et al.
2014). For example, tropical rainforest soils indicate the potential to reduce their
GHG sink strength by emitting considerable amounts of N2O, globally averaging
1.2 kg N2O-N ha-1
y-1
with up to 32 g N2O-N ha-1
d-1
measured from Australian
rainforest soils (Werner et al. 2007). With these substantial N2O emissions tropical
forest increase their GWP to -0.03 t CO2-e ha-1
y-1
while temperate and boreal forests
range between -0.9 and -1.18 t CO2-e ha-1
y-1
(Dalal and Allen 2008). The strong
radiative forcing of CH4 and N2O results in a GWP of 34 and 298 respectively when
converted to their CO2-equivalents (CO2-e) (Myhre et al. 2013). According to the
IPCC (2001) the atmosphere can contain an annual GHG concentration of 8.4 Pg
CO2-e y-1
without affecting the climate. However, anthropogenic sources of CH4 and
N2O alone already total 7.7 Pg CO2-e y-1
(Robertson and Grace 2004). Despite CO2
being the greatest driver of global warming, CH4 and N2O alone already take up over
90 % of the atmospheric GHG threshold. This major opportunity for mitigation
strategies based on CH4 and N2O alone highlights the need to identify the driving
environmental parameters and underlying soil-atmosphere flux processes.
54
Table 2-2 Major GHG concentrations currently and historically (Stocker et al. 2013)
Year CO2 (ppm) CH4 (ppb) N2O (ppb)
PI* 278 ± 2 722 ± 25 270 ± 7
2011 390.5 ± 0.3 1803 ± 4 324 ± 1
Notes: abundances are mole fraction of dry air for the lower, well-mixed atmosphere (ppm =
micromoles per mole, ppb = nanomoles per mole). Values refer to single-year average. Pre-industrial
(pi*, taken to be 1750).
GHG flux processes
Biogeochemical C and N cycling processes control the soil-atmosphere gas
exchange of the three major GHGs CO2, CH4 and N2O as illustrated in Figure 2-9
(Baldock et al. 2012). Carbon dioxide assimilation through photosynthesis and
vegetation biomass increases with increasing precipitation. However, soil respiration
can exceed CO2 assimilation under certain environmental conditions such as low soil
water content. Aerobic microbial decomposition is optimal up to 60% soil water
filled pore space (WFPS); above this anaerobic respiration dominates (Dalal and
Allen 2008). Atmospheric CH4 uptake into the soil occurs via microbial consumption
by methanotrophic bacteria for an energy source and is the largest natural sink of
CH4. This process is highly sensitive to alterations of physical soil conditions and
diffusivity, which can change soils to a CH4 source when methanogenic activity
dominates in saturated soil moisture conditions (Groffman and Pouyat 2009). The
CH4 flux can generally be considered the net-result of simultaneous occurring
production and consumption processes in the soil (Butterbach-Bahl and Papen 2002).
Nitrous oxide is produced principally by microorganisms during nitrification and
denitrification processes from mineral N in the soil (Butterbach-Bahl et al. 2013).
Literature Review
55
Figure 2-9 Soil biological processes of GHG (a) uptake into the soil and (b)
emissions from the soil into the atmosphere (Baldock et al. 2012).
Nitrification and denitrification are closely linked to soil moisture and substrate
availability and well as the proportion of mineral N in the soil. Besides N2O,
substantial N amount can also be lost as N2 and NO gases (Bouwman 1998). To date
the processes of nitrification and denitrification which drive N gas production in the
soil are, however, not fully understood but schematically outlined in Figure 2-10.
56
Figure 2-10 ‘Hole-in-the-pipe’ model of the regulation of trace-gas production and
consumption by nitrification and denitrification (Bouwman 1998).
Environmental parameters driving GHG fluxes
These GHG fluxes are closely related and the intensity driven by environmental
parameters, such as the climate, soil moisture, and C and N availability (Groffman et
al. 1995). The annual amount of rainfall (Groffman et al. 2009) and daily
temperatures (Butterbach-Bahl and Kiese 2005; Fest et al. 2009) are the main
parameters regulating GHG fluxes in temperate climates, while being less significant
to subtropical GHG fluxes (Rowlings et al. 2015). Physical soil properties such as
soil texture and bulk density can be more influential by indirectly regulating soil
moisture, porosity and oxygen availability and directly by impairing root growth and
therefore plant productivity. The combination of chemical soil properties, such as
pH, electric conductivity (EC) indicating soil salinity and CEC indication soil
fertility by nutrient and water holding capacity, indirectly effecting GHG fluxes by
influencing C and N availability and soil moisture. All these biological, physical and
chemical parameters combined interact and create a unique environment for soil-
atmosphere gas exchange, which differs for every ecosystem.
For example, clay soils with low pH can occlude N and make it unavailable for
soil microbes and plants. Slow drainage together with extensive mineral fertilizer use
in clay soils can result in salinization of the soil as well as increase N2O emissions
Literature Review
57
via increased denitrification. The most commonly occurring clay in the highly
weathered soils of Australia is kaolinite with a low CEC of approximately 10 meq+
100g-1
, while clays such as illite and smectite have CECs ranging from 25 to 100
meq+ 100g
-1 (Moore et al. 1998). Soil GHG emissions are generally strongly
correlated to soil texture, which means with increasing clay content increases the
potential of water logging conditions and therefore CH4 and N2O emissions in
particular, but become less correlated to texture with increasing sand content (Grover
et al. 2012; Raciti et al. 2011a; Rowlings et al. 2012).
Commonly sandy soils have very low CECs (< 10 meq+ 100g
-1), resulting in very
low nutrient and water holding capacity which, again, reduces microbial and plant
productivity. Therefore, some evidence suggests that sandy soils maybe more
affected by land use age than heavier textural soils as the age related SOM
accumulation is crucial for soil fertility and is more important in sand because of
lower fertility and therefore less resilience to disturbance (Golubiewski 2006).
Increasing SOM, especially in sandy soils, can increase the fertility of the ecosystem
by substantially increasing the CEC ranging from 250 to 400 meq+ 100g
-1 SOM
(Moore et al. 1998). However, fluxes of N2O are reported to be low in nutrient-poor,
acid soils with low CEC (Castaldi et al. 2006). Generally, the soil CEC can indicate
the fertility and productivity of an ecosystem. Soil fertility decreases with decreasing
pH, which can be induced by anthropogenic practices such as acidifying N fertilizer,
N leaching and by land clearing (McKenzie et al. 2004).
Temperature influences all biogeochemical reactions as well as soil microbe
populations and is therefore an important parameter effecting gas exchange in the
soil. For example, the literature review by Davidson and Janssens (2006) gives the
clear assumption that CO2 production of soils are temperature dependant almost
entirely from root respiration and microbial decomposition. Research from temperate
grasslands determined a shift from CO2 sinks to become neutral or even sources with
the introduction of N fertilizer use (Leahy et al. 2004). This suggests an even higher
CO2 productivity of tropical and subtropical regions as well as an increase over time
along the global temperature rise predicted by IPCC (2014).
58
Land use effect on GHG fluxes
Little research has been conducted on the effect of land use change associated
with urbanization on the C and N cycles, especially the effects in combination, and
the impact on net soil GHG flux. The few studies examining GHG emissions as
indicators for C and N cycle alterations in urban areas focused mainly on net primary
production (CO2 exchange) (Pouyat et al. 2006; Tratalos et al. 2007; Lorenz and Lal
2009). Generally, higher CO2 fluxes have been observed in peri-urban soils
compared to native environments (Pataki et al. 2007). Additionally, the conversion to
residential grasslands in the USA can increase N2O emissions and result in a weaker
net CH4 sink because of fertilization and irrigation practices. These changes in non-
CO2 GHG flux have the potential to offset the climate change mitigating soil C
sequestration (Kaye et al. 2004; Conant et al. 2005; Lorenz and Lal 2009; Wang et
al. 2014).
While some research suggests that an increase in SOM through intensive
management practices of urban environments, increases the soil’s capacity to oxidize
CH4 (Lal 2004a), others determined that residential lawns with relatively high SOM
appear to have almost no capacity for CH4 uptake (Groffman and Pouyat 2009).
Furthermore, Groffman and Pouyat (2009) observed that the same land use, such as
native forest, decreases its soil CH4 uptake potential with decreasing distance to
urban areas. These soil-atmosphere CH4 exchange dynamics and their relation to
urbanized ecosystems is not well understood and therefore demands further research
to establish accurate upscaling techniques for soil CH4 uptake within the same land
use type.
However, SOM is also an excellent predictor of the amount of total N in the soil
and has significant influence on denitrification in the soil producing N2O emissions
(Fageria 2012). Additionally, management practices have been identified to change a
CH4 sink to a source during wet seasons, particularly in pastures (Castaldi et al.
2006). Research from temperate zones identified a high C sequestration potential due
to the increased turf grass productivity, which reducing global warming effect can be
offset by N2O emissions from the high fertilizer N demand and all management
practices combined (Conant et al. 2005; Lorenz and Lal 2009; Wang et al. 2014).
Low soil C contents, such as found in some of the highly weathered soils of Australia
(Livesley et al. 2009), can limit N2O emissions but accelerate emissions with
Literature Review
59
increasing C content, offsetting the advantages of C sequestration in the long term
(Zhang et al. 2013a). The active C content can be an indicator of microbial activity
and therefore GHG production in the soil, which regulates net GWP (Mosier et al.
2005; Robertson and Grace 2004).
Measurement methods
To date, most GHG emission studies are based on manually sampled chamber
measurements over short time periods, this can substantially over or underestimate
annual fluxes (Scheer et al. 2013). Nitrous oxide emissions in particular vary
spatially and temporally (Rowlings et al. 2012; 2015) and can be easily over- or
underestimated with short-term and low frequency measurements. Mosier et al.
(2005) investigated a correlation of rain events and N2O emissions in agricultural
soils of Colorado, showing a delay of days between N2O production in the soil and
the release. An increased frequency in gas sampling is recommended during periods
of high gaseous emissions to ensure an accurate estimate over the growing season or
year.
Continuous high-frequency measurements are now considered an important tool
to ensure accurate annual GHG estimations and identification of the main driving
environmental parameters to create efficient mitigation strategies. Special emphasis
is needed for ecosystems in the highly variable weather conditions of the subtropics
and affected by changing environmental conditions such as urbanization.
2.4 Summary & implications
This literature review analysed current information available on urbanization
processes driving land use change, the effect on the biogeochemical C and N cycle as
well as the impact of climate change on the environment.
Currently, over half of our world population is living in urban and peri-urban
environments, these areas are predicted to account for all future population growth
(United Nations 2014). Further population increase and rural to urban migration
causes extensive land use changes around cities. By 2050, over 90 % of the
Australian population is predicted to live within such urban and peri-urban
environments which are currently in native or agricultural use (Commonwealth of
60
Australia 2013). This extensive urban sprawl proceeds with soil disturbance during
construction processes and increasingly the establishment of turf grass for residential
backyards, public parks and sportsgrounds, and golf courses (IPCC 2006). For
example, urban and peri-urban environments within or adjacent to cities in the USA
cover already 25 % of the terrestrial surface and is highly managed with fertilization
and irrigation resulting in substantial GHG emissions comparable to intensive
agriculture (Kaye et al. 2004).
Land use change from forest to pasture can increase the C and N content in about
70 % of the global studies (Conant et al. 2001). This increase in SOM, however,
might be only from the initial decomposition of tree roots in the soil and will
substantially decrease in the long term (Veldkamp 1994), and is hypothesized to not
reach native levels after afforestation (Raciti et al. 2011a). Peri-urban turf grass
systems in temperate climates, however, suggest a substantial C sequestration
potential by increased plant productivity based on fertilization and irrigation
management practices (Golubiewski 2006; Raciti et al. 2011a; Selhorst and Lal
2011).
Carbon sequestration dynamics is of great interest but the analysis is far from
uniformity (Paul 2006; Poeplau et al. 2016). For the purpose of identifying C
sequestration in soils, the microbial availability and therefore long-term storage
potential of SOM has been identified through fractionation schemes based on
physical and chemical SOM properties (Skjemstad et al. 2004; Baldock et al. 2012).
The three following fractions simplify SOM turnover into temporal context of active
(years), slow (decades) and resistant (centuries) C. The production of SOM depends
strongly on the N availability in the ecosystem, as N is the primary plant growth
limiting nutrient (Marschner 2007, 2012; Neal et al. 2013). The availability of N in
the soil depends strongly on microbial activity, which is subsequently driven by
environmental parameters such as soil temperature, moisture, texture, chemistry, and
the quantity and quality of decomposable plant material. These C and N cycle
driving environmental factors are significantly affected by changes in land use and
management such as fertilization, irrigation and physical disturbance.
There is increasing evidence that these anthropogenic induced land use changes
and management practices increase C and N losses from peri-urban ecosystems in
form of the main GHGs CO2, CH4, and N2O (Pouyat et al. 2002; Kaye et al. 2004;
Literature Review
61
2005; Groffman et al. 2009; Livesley et al. 2010; Selhorst and Lal 2013; Tian et al.
2014; Zhang et al. 2013a). Therefore, C sequestration needs to be quantified relative
to the ecosystem’s GHG emissions and land use management practices (Pouyat et al.
2009; Pouyat et al. 2006). This combination of estimating C and N sinks and sources
together with identifying the main driving environmental parameters, established the
full GWP of peri-urban environments.
The radiative forcing of GHGs in the atmosphere drive local and global climate
change (Myhre et al. 2013). These climate changes towards more extreme weather
events and rising temperatures will increase environmental pressure and
consequently increase GHG emissions even further as feedback effects (IPCC 2014).
Especially ecosystems under transition are particularly vulnerable to changing
climate conditions and therefore need to be examined for their C and N sink or
source potential on a multiple time scale as ecosystems response to changing
environmental conditions slowly and over long term. Arid regions seem to be the
most vulnerable ecosystems undergoing urbanization (Koerner and Klopatek 2002;
2010; Hall et al. 2008) and indicate with the predicted future drought increase in
Australia (IPCC 2014) a substantial socio-ecological impact on peri-urban
environments.
Conclusion
C sequestration could be the major strategy for climate change mitigation if these
gains are not offset by N2O and CH4 emissions from anthropogenic activities. In
addition, the tropical and subtropical climatic zone will play a significant role in
achieving global food security in the future (FAO and ITPS 2015). Australia has
great potential to reduce GHG emissions, which is currently four times the global
average (Hatfield-Dodds et al. 2015), by limiting management intensities and
improve C sequestration strategies.
62
Gaps
Based on the gaps that have been identified in the literature (Table 2-3), the
following four main points will be addressed by this research.
Table 2-3 Approximate representation of the main literature in study related research
topics per climate zone in percent (%) and the overall global attention the topic has
received
Climate Global
Topic Temperate Subtropical Tropical
LUC* involving forest 45 10 45 high
LUC* involving agriculture 70 10 20 high
LUC* involving turf grass 98 1 1 low
C & N cycle 75 5 20 high
GHG emissions 70 10 20 medium
Non-CO2 GWP 85 5 10 low
C sequestration 70 10 20 medium
Multiple time scale 50 0 50 low
* Land Use Change (LUC)
1. High-frequency GHG baseline data for native, agricultural and residential land
use for future predictions and climate change mitigation scenarios:
Most studies to date use infrequent, short term or measurements up to one year
(Groffman et al. 2009; Koerner and Klopatek 2010; Page et al. 2011; Fest et al.
2015a). With these low-frequency measurements, the high intra- and inter-annual
temporal variability of GHG fluxes (Scheer et al. 2014a; Rowlings et al. 2015) can
easily cause an over- or underestimation of annual GWPs. Improving the global
baseline information for peri-urban environments in transition will ensure process
models and predictions are more precise and mitigation strategies based on these
predictions more efficient (IPCC 2013; Tian et al. 2014; Henderson et al. 2015)
IPCC (2013.
2. C and N cycling in turf grass systems and driving environmental parameters:
To date, C and N studies on Australian land use change are mostly on
deforestation, afforestation and reforestation (Forster 2006; Canadell and Raupach
2008; Newham et al. 2011; Raupach et al. 2011). Data from urban environments and
related land use change into turf grass systems are still rare and nearly exclusively
from temperate USA (Golubiewski 2006; Qian et al. 2010; Raciti et al. 2011a) or
Literature Review
63
other temperate environments (Vellinga et al. 2004; Tratalos et al. 2007; Livesley et
al. 2010; Schaufler et al. 2010). The current state of subtropical research on turf grass
systems, gives limited information about environmental parameters such as soil
texture, which makes a generalization and global comparison difficult (Kong et al.
2014)
3. Subtropical climate:
Generally, limited information is available on C and N cycling affected by land
use change in the tropics (Veldkamp 1994; Paul et al. 2008b; Paul et al. 2008a) and
even less is known about the subtropics (Xu et al. 2013; Kong et al. 2014).
4. Multiple time-scales research:
Ecosystem responses can change substantially from medium to long-term time
scales and can give significantly different data model outputs when calibrated with
limited time representative measurements. Chronosequence studies, such as from
Veldkamp (1994) and Fest et al. (2015b), could be a strategy to evaluate long-term
effects on ecosystems by taking multiple time scales after land use change into
account to identify the temporal response variation (van Lent et al. 2015).
All studies agree that land use changes associated with urbanization has received
little attention with respect to soil C sequestration and GHG emissions. Information
on changes in GWP in peri-urban environments with multiple land uses is critical to
assess the potential impacts on global climate change.
Research Design
65
Chapter 3: Research Design
The land use transition from rural to urban environments includes intensive
construction processes, which involve plant cover removal, topsoil displacement and
mixing with the subsoil, as well as extended periods of bare soil (i.e. fallow land).
The resulting peri-urban environments include a combination of land use types such
as forest, pasture and turf grass. This research will use a native forest for the baseline
estimation before land use change and grazed pasture as representative of the main
rural land use. However, peri-urban environments vary widely in their land use types
and also include highly disturbed or secondary forest, grazed and ungrazed pasture
and turf grass systems with high or low management intensity. This wide variety of
land use types and their management practices, along with limited accessibility for
research on private and public areas, restricts continuous GHG estimations and
nutrient cycling observations. This research used a combination of soil-atmosphere
gas measurements and the C sequestration potential of a peri-urban environment to
identify the immediate, inter-annual and long-term (> 10 years) ecosystem response
to land use change associated with urbanization. Two years of high frequency GHG
measurements generated the immediate and inter-annual ecosystem response to land
use change and a soil survey focusing on C sequestration evaluated the long-term
effect of urbanization on the C and N cycle.
3.1 Site description
Research was conducted in the Samford Valley, 20 km from Brisbane in South
East Queensland (SEQ), Australia (Figure 3-1). Brisbane currently has a population
growth rate of 1.7 % per year and is considered the most biologically diverse city in
Australia with the most extensive area of urban sprawl (ABARES 2010;
Commonwealth of Australia 2013). The Samford Valley covers an area of
approximately 82 km2 and is surrounded by mountains to the north, west and south.
Mostly cleared in the early 1900s, the valley was significantly developed in the
1960s for dairy and beef cattle as well as intensive agriculture including banana and
66
pineapple. Since the early 1990s, population density has increased and almost
doubled from 1996 – 2006 causing substantial land use change from predominately
rural to residential properties (Moreton Bay Regional Council 2011). The first three
research objectives were evaluated by experiments at the Samford Ecological
Research Facility (SERF), the only peri-urban Supersite within the Terrestrial
Ecosystem Research Network (TERN) in Australia (Figure A 7). To evaluate the
fourth objective, a soil survey was conducted across public parks and sportsgrounds
of the Valley, with access provided by the Moreton Bay Regional Council, as well as
private residential volunteers from the Samford community.
Figure 3-1 Location of Samford Valley near Brisbane in South East Queensland,
Australia.
The granite parent material of the Samford Valley has given rise to mostly
Chromosols and Kurosols type soils based on the Australian soil classification (Isbell
2002) and Planosols according to the World Reference Base (WRB 2015). These are
characterized by a strong texture contrast between the A and B horizon (Figure A 8)
and are amongst the most widespread soil types currently in agricultural use in
Australia (Figure A 9). The distribution of SOC through depth of the most common
Australian soils can be found in the Figure A 10. Construction processes associated
with urbanization has resulted in the random mixing of soil horizons under many turf
grass areas across the Valley, creating a wide range of soil chemical and physical
properties.
South East Queensland is influenced by a humid subtropical climate with seasonal
summer rain. The long term mean annual precipitation at SERF is 1110 mm with a
mean annual minimum and maximum temperature of 13 °C and 25.6 °C, respectively
(BOM 2015). Over the two years of this research the inter-annual difference in
rainfall was 430 mm, ranging from 740 in 2013/14 to 1170 mm in 2014/15, with
Research Design
67
temperatures 1.6°C on average higher in year two of the GHG flux measurements.
These differences illustrate the extreme annual and inter-annual rainfall variations of
the humid subtropical Australia, which are dominated by heavy rain events and
therefore rapid changes in soil moisture.
3.2 Materials and Methods
3.2.1 Experimental design
The turf grass lawn and fallow treatments were established in a randomised plot
design within a well-established grazed pasture with three replicated plots per land
use, each 2 m by 10 m and separated by 0.5 m buffer zone (Figure 3-2, Figure A 11),
adjacent to the native forest. The remnant forest at SERF is classified as a Dry
sclerophyll eucalypt forest and is the typical native forest of the sandy soils in this
region. The Chloris gayana pasture with 10 % of white clover (Trifolium repens) had
been grazed extensively for the last 15 years. Livestock were excluded over the two
years of this research and the pasture was cut 11 times when a height of 20 cm was
reached, to simulate grazing.
Figure 3-2 Experimental site at SERF showing pasture, turf grass and fallow plots as
well as the adjunct forest.
68
The turf grass lawn was established in June 2013 according to the common
regional practice of removing the dense pasture sward and surface roots to expose the
topsoil. The topsoil was then rotary hoed twice to a depth of 15 cm and fertilized
with 50 kg N ha-1
before Blue Couch (Digitaria didactyla) turf rolls laid over the top.
Over the two years the turf grass was fertilized an additional 4 times with 50 kg N ha-
1 surface applied (26.10.13, 6.3.14, 28.9.14, 8.1.15) with Prolific Blue AN fertilizer
(8 % ammonium, 4 % nitrate, 5.2 % phosphorus, 14.1 % potassium, 1.2 %
magnesium) and irrigated immediately. The total 250 kg of N fertilizer ha-1
over the
two years was less than the local industry practice recommendation (300 kg N ha-1
y-
1) and was chosen to represent average application rates for private, public and
industrial use (Moreton Bay Regional Council 2011). The turf grass was irrigated
during both the establishment phase and periods of extended dry. After each of the
11 occasions of mowing, the turf grass clippings were removed and weighed for their
nutrient content.
The fallow treatment simulated the impact of construction processes associated
with urbanization by removing the pasture sward with approximately 5 cm of topsoil.
The remaining topsoil was then rotary hoed twice to a depth of 15 cm. The fallow
soil was kept free from plant cover over the two experimental years with a non-
selective herbicide (Bi-Active 360g/L Glyphosate) and a broad leaf herbicide
(Double Time, 340g/l MCPA + 80g/l Dicambra).
3.2.2 GHG gas flux system
Two years of high frequency N2O and CH4 measurements identified inter-annual
flux dynamics after land use change from native forest to a grazed pasture and peri-
urban turf grass lawn. Gas flux sampling was based on the static chamber technique
using an automated sampling system as detailed by Scheer et al. (2014b). The
pneumatically operated 50 cm x 50 cm x 15 cm high static chambers were secured to
stainless steel bases, permanently inserted 10 cm into the ground. Detailed
information about the measurements cycle can be found in Chapter 4 to 6.
3.2.3 Soil survey
This is a brief description of the soil sampling and C fractionation procedure.
Additional details can be found in Chapter 7.
Research Design
69
3.2.3.1 Soil sampling
Intact soil cores were taken from each sampling site to 1 m depth with a hydraulic
soil auger to determine the soil type. The soil core was divided into horizons
according to Isbell (2002) with each bagged separately. Three replicated soil samples
mixed from four subsamples were taken for 0-10 cm and 10-20 cm soil depth from
each sampling site with a hand auger.
3.2.3.2 C fractionation
The C fractionation studied here is based on the scheme established by Skjemstad
et al. (2004) and Baldock et al. (2012), dividing SOC into an active (CA), slow (CS)
and resistant fraction (CR). The CA fraction drives the C cycle’s fastest turnover
(years) in the form of soil respiration and microbial available C. The CS fraction has
a slow turnover (decades) and the transitional stage to the CR fraction which turnover
is approaching stagnation (centuries) and is considered resistant to C cycling in the
soil. This research analysed the CA and CR fractions without the physical separation,
as the main soil texture of the area of interest is sand (particle size 0.2 - 2 mm)
without major amounts of micro-aggregates, and then calculated the CS fraction from
the total soil C (CT) according to equation 1.
CS = CT – (CA + CR) (Equation 1)
3.2.4 Environmental parameters
Soil samples were taken for site characterization to 1 m depth, air dried and sieved
to 2 mm. Particle size analysis for soil texture as well as bulk density (BD), pH and
electrical conductivity (EC) analyses were undertaken as established by Carter and
Gregorich (2007). Soil moisture and temperature for each land use type were
collected using a TDR probe (HydroSense CD 620 CSA) and a PT100 probe (IMKO
Germany). Soil moisture was then converted with the treatment specific BD to water-
filled pore space (WFPS). Total C and N content of soil and plant material was
determined by dry combustion (CNS-2000, LECO Corporation, St. Joseph, MI,
USA) from ground samples. Regular soil samples were taken fortnightly to support
gas flux measurements from all replicated land use type plots over both experimental
years and divided into 2 depths (0-10 cm, 10-20 cm). NH4+ and NO3
- were extracted
70
from the soil using a 1:5 KCl solution with 20 g of fresh soil. The extract was
analysed for NH4+ and NO3
- with an AQ2+ discrete analyser (SEAL Analytical WI,
USA).
3.2.5 Data management and statistical analysis
Soil-atmosphere GHG fluxes were calculated from the slope of the linear increase
or decrease of the 4 concentrations measured over the closure time and corrected for
chamber temperature and atmospheric pressure using the procedure outlined by
Scheer et al. (2014b). The coefficient of determination (r2) was calculated and used
as a quality check for fluxes above the detection limit to assure linearity of the gas
concentration increase. Flux rates were discarded if r2 was < 0.85 for N2O and < 0.95
for CH4. Daily fluxes from the automated chambers were calculated by averaging
sub-daily measurements for each chamber over the 24 hour period. The detection
limit determined for the gas chromatograph is ± 1 g N2O and CH4 ha-1
d-1
. Gaps in
the dataset were filled by linear interpolation across missing days. The non-CO2
GWP was calculated from the CO2-equivalents (CO2-e) for N2O and CH4 of 298 and
34 respectively (IPCC 2014).
Statistical analyses for cumulated annual N2O and CH4 fluxes, non-CO2 GWP and
annual WFPS averages were undertaken using SPSS Statistics 21.0 (IBM Corp.,
Armonk, NY) and differences between land use types were assumed to be significant
when the significance value (p) was < 0.05. Non-normal distribution meant all data
were log-transformed for ANOVA analysis using the Ryan-Einot-Gabriel-Welch Q
(REGWQ) as post-hoc test. A Spearman’s rho correlation analysis was used to
examine relationships between gas fluxes and environmental parameters such as soil
chemistry, soil moisture and temperature. The effect of land use types on C
sequestration was identified by an ANCOVA using the correlated environmental
parameters clay, total C and N content as covariates. An autoregressive integrated
moving average (ARIMA) model (Box and Pierce 1970) was used in R studio to
determine autocorrelation between successive daily N2O and CH4 averages which
includes covariate effects between measurements. The ARIMA coefficient was
interpreted as the expected difference between current and lagged values for a
covariate unit increase.
Research Design
71
3.3 Thesis outline
This research investigating C and N cycling in peri-urban environments before
and after land use change consists of four results chapters represented by
publications (Chapters 4 to 7). The following four publications with the hypotheses
and main outcomes are briefly outlined within the context of the overall objectives of
this research.
The first paper (Chapter 4) addresses Objective 1 and uses manual and automated
gas sampling methods to investigate annual GHG fluxes from rural land use types
and the immediate ecosystem response when turf grass is established.
The main outcomes of this paper were:
1. Turf grass establishment significantly increases the non-CO2 GWP compared
to native forest and grazed pasture by:
a. reduced CH4 uptake for up to a month after establishment and
b. high N2O emissions due to fertilizer use and irrigation, for up to 2
month after establishment of the turf grass.
2. Manual gas sampling detected no N2O emissions over the full year of low
frequency measurements
Paper 1 was published as:
van Delden, L., E. Larsen, D. Rowlings, C. Scheer and P. Grace. 2016a.
"Establishing turf grass increases soil greenhouse gas emissions in peri-urban
environments." Urban Ecosystems: 1-14.
The second paper (Chapter 5) addresses Objective 2 and compares land use types
and corresponding N cycling over a one year period. High-frequency automated flux
measurements were used for the annual N2O emissions to identify the main driving
environmental parameters effecting N cycling in peri-urban ecosystems.
The main outcomes of this paper were:
1. Native forest soil is a low N2O emitter because of an efficient N cycle
72
2. Pasture soil is comparable to the forest in N cycling and N2O emissions
3. Turf grass establishment and fallow land increases mineral N turnover and
losses
4. Fallow land significantly increases N2O emissions and NO3- leaching potential
Paper 2 was published as:
van Delden, L., D. W. Rowlings, C. Scheer and P. R. Grace. 2016b. "Urbanization
related land use change from forest and pasture into turf grass modifies soil nitrogen
cycling and increases N2O emissions.." Biogeosciences 2016: 1-23.
The third paper (Chapter 6) addresses Objective 3 and complements the papers 1
and 2 by analysing non-CO2 GHG fluxes influenced by land use type. It identified
the inter-annual variations of subtropical land use change in peri-urban
environments.
The main outcomes of this paper were:
1. Land use change associated with urbanization significantly increases the non-
CO2 GWP mainly through the establishment of turf grass, while decreasing
the non-CO2 GWP significantly from year 1 to year 2, becoming comparable
with grazed pasture
2. Native forest and grazed pasture show little intra- and inter-annual variation in
N2O emissions
3. Pasture soil becomes temporarily a CH4 source with high WFPS, but annually
represents a CH4 sink
Paper 3 was submitted as:
van Delden, L., D. W. Rowlings, C. Scheer, D. De Rosa, P. R Grace. “Soil N2O
and CH4 fluxes from urbanization related land use change; from Eucalyptus forest
and pasture to urban lawn.” Global Change Biology
Research Design
73
The fourth paper (Chapter 7) addresses Objective 4 and evaluated the long-term
influence of land use change on the C and N cycle by identifying the C sequestration
potential of peri-urban ecosystems after more than a decade of establishment.
The main outcomes of this paper were:
1. Turf grass land use did not significantly increase soil C sequestration or the
total C and N content in the subtropical peri-urban Samford Valley
2. Turf grass land use increased the slow C fraction compared to pasture but also
decreased resistant C fraction compared to forest
3. Introduced clay content from construction processes during land use change
may be more dominant in affecting soil C sequestration over land use type
4. Subtropical peri-urban topsoils can exceed the native C and N conditions
Paper 4 is in preparation and will be submitted as:
van Delden, L., D. W. Rowlings, C. Scheer, P. R Grace. “Long-term implications
of land use change associated with urbanization on the terrestrial C and N cycle.”
This thesis concludes with Chapter 8, which synthesis the outcomes from all four
publications and discusses the overall research findings, as well as identifying the
practical implications to for future research.
75
Statement of Contribution of Co-Authors for Thesis by Published Paper
The authors listed below have certified* that:
1. they meet the criteria for authorship in that they have participated in the conception,
execution, or interpretation, of at least that part of the publication in their field of expertise;
2. they take public responsibility for their part of the publication, except for the responsible
author who accepts overall responsibility for the publication;
3. there are no other authors of the publication according to these criteria;
4. potential conflicts of interest have been disclosed to (a) granting bodies, (b) the editor or
publisher of journals or other publications, and (c) the head of the responsible academic unit,
and
5. they agree to the use of the publication in the student’s thesis and its publication on the
Australasian Research Online database consistent with any limitations set by publisher
requirements.
In the case of this chapter:
Chapter 4: Establishing turf grass increases soil greenhouse gas emissions in peri-
urban environments (Paper 1)
Contributor Statement of contribution*
Lona van Delden Performed experimental design, conducted
fieldwork and laboratory analyses, data analysis,
and wrote the manuscript.
Signature
05/07/2017
Eloise Larsen Aided experimental design, contributed field work,
and reviewed the manuscript.
David W. Rowlings Aided experimental design and data analysis, and
reviewed the manuscript.
Clemens Scheer Aided experimental design and data analysis, and
reviewed the manuscript.
Peter R. Grace Aided experimental design and data analysis, and
reviewed the manuscript.
Chapter 4 (Paper 1) has been published in Urban Ecosystems in June 2016, Volume 19, Issue
2, pp 749–762.
76
Principal Supervisor Confirmation
I have sighted email or other correspondence from all Co-authors confirming their certifying
authorship.
David W. Rowlings 30/11/2016
Name Signature Date
Establishing turf grass increases soil greenhouse gas emissions in peri-urban environments (Paper 1)
77
Chapter 4: Establishing turf grass increases
soil greenhouse gas emissions in
peri-urban environments
(Paper 1)
4.1 Abstract
Urbanization is becoming increasingly important in terms of climate change and
ecosystem functionality worldwide. We are only beginning to understand how the
processes of urbanization influence ecosystem dynamics and how peri-urban
environments contribute to climate change. Brisbane in South East Queensland
(SEQ) currently has the most extensive urban sprawl of all Australian cities. This
leads to substantial land use changes in urban and peri-urban environments and the
subsequent gaseous emissions from soils are to date neglected for IPCC climate
change estimations. This research examines how land use change effects methane
(CH4) and nitrous oxide (N2O) fluxes from peri-urban soils and consequently
influences the Global Warming Potential (GWP) of rural ecosystems in agricultural
use undergoing urbanization. Therefore, manual and fully automated static chamber
measurements determined soil gas fluxes over a full year and an intensive sampling
campaign of 80 days after land use change. Turf grass, as the major peri-urban land
cover, increased the GWP by 415 kg CO2-e ha-1
over the first 80 days after
conversion from a well-established pasture. This results principally from increased
daily average N2O emissions of 0.5 g N2O ha-1
d-1
from the pasture to 18.3 g N2O ha-
1 d
-1 from the turf grass due to fertilizer application during conversion. Compared to
the native dry sclerophyll eucalypt forest, turf grass establishment increases the GWP
by another 30 kg CO2-e ha-1
. The results presented in this study clearly indicate the
substantial impact of urbanization on soil-atmosphere gas exchange in form of non-
CO2 greenhouse gas emissions particularly after turf grass establishment.
78
4.2 Introduction
Urban populations worldwide now exceed rural populations and will account for
all future population growth (United Nations 2008). Australia has one of the highest
proportions of urban versus rural population and is expected to exceed 90 % of the
population living in urban areas by 2050 (United Nations 2014). Globally, this
urbanization becomes increasingly important in terms of climate change and
ecosystem functionality (Hutyra et al. 2011).
Australia’s urban, peri-urban and rural residential land use accounted for 2.5
million ha in 2005/06, with Brisbane City in South East Queensland (SEQ) covering
1,300 km2 and it is considered the most biologically diverse of Australia’s capital
cities (ABARES 2010; Commonwealth of Australia 2013). In June 2014 the
population of Brisbane was 2.27 million people, this is nearly half of Queensland's
population and is an increase of 1.7 % between 2013 and 2014 (ABS 2015). This
results in the most extensive urban sprawl of all Australian cities with substantial
land use changes associated with deforestation and the conversion of commercially
focused agriculture into smaller residential properties (Commonwealth of Australia
2013). Clearing natural vegetation has the strongest impact on the environment by
the removal of biomass, which influences water quality as well as nutrient cycling.
For example, land use change from intact biomes to agricultural use can lead to a
loss in soil quality (structure and nutrient losses) and quantity (erosion), increase
greenhouse gas (GHG) emissions, and reduce soil potential for carbon sequestration
(Grover et al. 2012; Livesley et al. 2009). Given the large areas worldwide
undergoing these land use changes, the implications for ecosystem functionality and
health are significant.
The changes in global climates are driven by the radiative forcing of various
greenhouse gases (IPCC 2007). Soils represent a major source of these GHG’s, with
the magnitude of emissions greatly influenced by anthropogenic practices. Produced
by biogeochemical processes, GHG emissions are strongly influenced by land use
type and management, soil moisture, and soil type. Especially irrigation accelerates
the microbial C and N turnover and leads to conditions that promote elevated
emissions from soils (Scheer et al. 2008). Fluxes of the three major greenhouse gases
carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O) depend on soil
Establishing turf grass increases soil greenhouse gas emissions in peri-urban environments (Paper 1)
79
moisture dynamics, soil texture, nutrient input and substrate availability (Rowlings et
al. 2012); factors that are greatly modified by land use change. While CH4 when
taken up from the atmosphere by soil microbes reduces the radiative forcing, land
uses can also become CH4 sources in certain circumstances and contribute to global
warming (Groffman and Pouyat 2009). Nitrous oxide, however, is produced in
nitrification and denitrification processes and emitted into the atmosphere where its
increases the radiative forcing (Butterbach-Bahl et al. 2013).
Urban development involves the construction and sealing of soils for commercial,
industrial and residential use (WRB 2015), and urbanization research suggests that
once influenced by humans, ecosystem dynamics are no longer dominated by natural
factors (Kaye et al. 2006). Urban soils, however, can also improve critical ecosystem
services by providing stormwater treatment, acting as a sink for atmospheric nitrogen
(N), and sequestering carbon (C) (Raciti et al. 2011a; Golubiewski 2006). How these
processes of urbanization influence ecosystem dynamics in biogeochemical cycling
and contribute to climate change is only beginning to be understood (Kaye et al.
2004; Byrne 2007; Lorenz and Lal 2009).
This transition of rural to urban and peri-urban environments is characterized by
an extensive construction process and the establishment of turf grass in residential
backyards, public parks, and golf courses (Kaye et al. 2005; Milesi et al. 2005;
Pouyat et al. 2009). To date, research has focussed on the biogeochemistry of turf
grass in urban and peri-urban environments of the temperate zones but little is known
for subtropical and tropical regions (Golubiewski 2006; Grimm et al. 2008; Lorenz
and Lal 2009; Pouyat et al. 2009; Raciti et al. 2011a; Barton and Colmer 2011).
Biochemical processes like decomposition of organic matter, accumulation and
losses of nutrients like C and N, nitrification and denitrification and even C
sequestration in soil are known to be affected by urbanization. Changes in
management specific to public and private residential properties such as fertilization,
irrigation, and frequent mowing substantially modify the ecological biogeochemistry
of ecosystems and need to be quantified for different land uses and climates.
Research on GHG fluxes within changed land cover are still quite rare and mostly
limited to temperate zones (Groffman and Pouyat 2009; Kaye et al. 2004) or arid
regions (Hall et al. 2008; Koerner and Klopatek 2010). The few studies examining
80
GHG emissions as indicators for C and N cycle alteration in urban areas focused
mainly on net primary production, i.e. CO2 exchange, while CH4 and N2O have often
been neglected (Tratalos et al. 2007; Lorenz and Lal 2009; Ng et al. 2015). Most
recent data modelling approaches on soil C and N cycling under turf grass suggest
that management is crucial for the impact of urban and peri-urban land use on the
soil-atmosphere gas exchange in temperate climates (Zhang et al. 2013a; Gu et al.
2015). However, little is known about subtropical and tropical climates affecting
soil-atmosphere gas exchange during land use change.
Estimating the effect of urbanization on ecosystems, as well as the contribution to
climate change, requires identifying peri-urban land uses as either a net GHG sink or
source and can be achieved by determining N2O and CH4 fluxes from intensified
peri-urban land use. To estimate the true impact of land-use change on climate, the
combined global warming potential (GWP) of GHG emissions need to be evaluated.
Most GHG emission studies to date are based on manual measurements over short
time periods which can substantially over or underestimate annual fluxes (Rowlings
et al. 2015). Precise methods are therefore necessary to quantify GHG emissions to
determine the impact of urban peri-urban environments on the climate, especially for
subtropical and tropical climates with the highest urbanization rates.
This research quantified the impact of peri-urban land use change on the soils
GWP in SEQ using a combination of manual GHG measurements over a full year,
together with an intensive sampling campaign of high frequency measurements
immediately following further land use intensification.
4.3 Materials and Methods
4.3.1 Site description
The study was conducted at the Samford Ecological Research Facility (SERF) in
the Samford Valley, 20 km from Brisbane in SEQ. The Samford Valley covers an
area of approximately 82 km2 and is surrounded by mountains to the north, west and
south. Mostly cleared in the early 1900s, the valley was significantly developed in
the 1960s for dairy and beef cattle as well as intensive agriculture including banana
and pineapple. Since the early 1990s, population density has increased and almost
Establishing turf grass increases soil greenhouse gas emissions in peri-urban environments (Paper 1)
81
doubled from 1996 – 2006 causing substantial land use change from predominately
rural to residential properties (Moreton Bay Regional Council 2011). South East
Queensland is influenced by a humid subtropical climate with seasonal summer rain.
The long term mean annual precipitation is 1110 mm with a mean annual minimum
and maximum temperature of 13 °C and 25.6 °C, respectively (BOM 2015). The
overall shallow soil profile of the experimental site is covering the granite floor of
the valley. Characterised by a strong texture contrast between A and B horizon
classifies the experimental site as Chromosols according to the Australian soil
classification (Isbell 2002) or Planosols according to the World Reference Base and
is defined as poor soils (WRB 2015).
4.3.2 Experimental design
A combination of manually sampled and automated closed static chambers in
conjunction with gas chromatography was used to quantify GHG fluxes from four
peri-urban land cover treatments. One year of GHG data were collected from March
2009 to February 2010 to compare native forest and grazed pasture and to determine
any major seasonal variations in fluxes. Native forest at SERF (Dry Sclerophyll
Eucalypt Forest), was analysed as a baseline for historical land use change and
pasture (Chloris gayana and Setaria sphacelata) represented rurally developed areas.
An intensive, 80 days, sampling campaign followed the transformation of the pasture
into bare soil (fallow) and the establishment of Blue Couch (Digitaria didactyla) turf
grass lawn. The fallow treatment was established to simulate the impact of
transitional processes such as construction work and plant cover replacement.
For the yearlong baseline observation, four representative plots each for forest and
pasture were selected adjacent to each other ensuring that slope, aspect and soils
were identical. The additional treatments for the intensive campaign (fallow and
lawn) were installed on the well-established extensively grazed pasture, within 50 m
of the dry sclerophyll forest. Three replicated plots per treatment, 2 m wide by 10 m
long were established in a randomised plot design in June 2013. The pasture and 10
cm of topsoil and roots was removed from the fallow and lawn treatments and the
plots rotary hoed twice to a depth of 15 cm. Turf grass was planted with 50 kg N ha-1
of Prolific Blue AN fertilizer (12.0 % nitrogen, 5.2 % phosphorus, 14.1 % potassium,
82
1.2 % magnesium), half of the local construction industry practice recommend for
the area. Irrigation was applied to the lawn only.
4.3.3 CH4 and N2O flux measurements
The year-long baseline observation (2009/2010) used manually sampled
chambers for biweekly measurements. The PVC chambers had a headspace of
230 cm3, and were permanently inserted into the soil to 10 cm depth and replicated
four times per land cover. Closure was achieved using a gas tight lid containing a
rubber septum as a sampling port. Four gas samples were taken over one hour of
closure (0, 20, 40, 60 min) with a double-ended syringe to extract a 12 ml headspace
sample into an evacuated glass vial (Exetainer; Labco, High Wycombe,
Buckinghamshire, UK). The gas samples were analysed with a gas chromatograph
(Shimadzu GC-2014) in laboratory facilities at Queensland University of
Technology.
The intensive campaign (Jun – Aug 2013) employed a high resolution fully
automated GHG measurement system as detailed by (Scheer et al. 2014b). The
pneumatically operated 50 cm x 50 cm static chambers were secured to stainless steel
bases permanently inserted 10 cm into the ground. The chambers had a headspace of
15 cm height and were connected to an in situ sampling system and gas
chromatograph (SRI GC8610, Torrance, CA, USA); equipped with 63Ni Electron
Capture Detector (ECD) for N2O and a Flame Ionization Detector (FID) for CH4.
One replicate chamber from each of the four treatments closed for one hour and four
gas concentrations from each chamber were measured at 15 minute intervals,
followed by a known calibration standard (4.0 ppm CH4, 0.5 ppm N2O, Air Liquide,
Houston, TX, USA). This process was repeated two more times for the remaining
two replicates, building a full cycle of three hours, with eight fluxes per day for 12
chambers.
4.3.4 Auxiliary measurements
Soil moisture and temperature for each treatment were collected using a TDR
probe (HydroSense CD 620 CSA) and a PT100 probe (IMKO Germany). Soil
samples were taken for site characteristics with a hydraulic auger to one meter depth.
For analysis preparation the samples were air dried and sieved to 2 mm particle size.
Establishing turf grass increases soil greenhouse gas emissions in peri-urban environments (Paper 1)
83
Particle Size Analysis for sand, silt, and clay content as well as pH and electrical
conductivity (EC) analysis and bulk density (BD) were done according to (Carter and
Gregorich 2007). Total C and N content was determined by dry combustion (CNS-
2000, LECO Corporation, St. Joseph, MI, USA) from ground soil samples.
4.3.5 Flux calculations and statistical analysis
Fluxes were calculated from the slope of the linear increase or decrease over the 4
concentrations measured over the closure time as well as corrected for chamber
temperature and atmospheric pressure similar to the procedure outlined by (Scheer et
al. 2014b). Pearson’s correlation coefficient (r2) for the linear regression was
calculated and used as a quality check (linearity of the concentration increase) for the
measurement. Flux rates were discarded if r2 was < 0.95 for CH4 and < 0.85 for N2O
fluxes (Scheer et al. 2013). Daily fluxes from the automated chambers were
calculated by averaging sub-daily measurements for each chamber over the 24 hour
period before averaging across replicates. Gaps in the dataset were filled by linear
interpolation across missing days. The GWP was calculated using the CO2-
equivalents (CO2-e) of 25 and 298 for CH4 and N2O, respectively (IPCC 2007).
Statistical analysis was undertaken using SPSS Statistics 21.0 (IBM Corp.,
Armonk, NY). Non-normal distribution meant all data were log-transformed for
ANOVA analysis using Games-Howell as the post-hoc test. A Spearman’s rho
correlation analysis was used to examine relationships between gas fluxes, soil
moisture and temperature. The significance value (p) is shown for each analysis, as
well as the correlation coefficient (r) with its significance level (p < 0.05*, p <
0.01**).
84
4.4 Results
4.4.1 Site description
Specific site parameters for the SERF site description can be found in Table 4-1.
During the baseline observation in 2009/2010 the annual rainfall of 1490 mm
exceeded the long term annual mean for the site. Because of the relatively dry winter
season with only 40 mm of rain in 2013, the turf grass lawn had to be irrigated after
planting with additional 12 mm over the 80 days to ensure root establishment. The
mean annual minimum and maximum temperatures for the year-long baseline
observation 2009/2010 were 16.6 °C and 27 °C respectively; and 13 °C and 20.6 °C
respectively for the intensive sampling campaign 2013.
Table 4-1 – SERF site description.
Parameters
Longitude 152° 52' 37.3" E
Latitude 27° 23' 22.211" S
Altitude 60 m
Slope 2°
Mean annual min temp. 13 °C
Mean annual max temp. 25.6 °C
Mean annual rain 1110 mm*
Soil profile Horizons Depths
(cm)
Sand
(%)
Silt
(%)
Clay
(%)
BD
(g cm-3)
pH EC
(μS)
Total
C
(%)
Total
N
(%)
A1 0 – 17 70 24 6 1.4 5.4 46 1.5 0.12
A2 17 – 45 74 18 8 1.6 6.0 10 0.9 0.07
B2 45 – 92 9 18 73 1.8 6.1 31 0.4 0.03
B3 92 – 110 62 16 22 1.7 6.2 39 0.5 0.04
*Commonwealth Bureau of Meteorology, Australian Government (BOM 2015)
The site parameters of the soil profile indicate a low fertility as already suggested
by the WRB (2015). The high sand content in the surface soil with underlying clay
horizon combined with the moderate slope suggests the potential for nutrient losses
via leaching and run off in the surface soil, while preventing excess water logging.
As N2O and CH4 is frequently associated with extended periods of high soil water
content, it is likely the emissions from this study are at the lower range of gas fluxes
that can be expected from natural and peri-urban environments in this climate.
Establishing turf grass increases soil greenhouse gas emissions in peri-urban environments (Paper 1)
85
4.4.2 CH4 and N2O flux measurements
Both the annual and intensive campaign results determined that native forest is a
sustained soil CH4 sink due to consistent CH4 uptake (Table 4-2, Figure 4-1A, B)
with fluxes ranging from -4.8 to -0.9 g CH4 ha-1
d-1
in 2009/2010 and from -12.7 to -
6.9 g CH4 ha-1
d-1
in 2013. The pasture, turf grass and fallow all took up CH4 during
drier months. The wetter period at the start of the campaign saw CH4 uptake
restricted to turf grass and the fallow treatment. Pasture, on the other hand, became a
soil CH4 source after rainfall events with fluxes ranging from -3.8 to
10.6 g CH4 ha-1
d-1
in 2009/2010 and from -8.3 to 12.5 g CH4 ha-1
d-1
in 2013.
Cumulative fluxes over the 80 day sampling campaign illustrate the strength of
native forest soil CH4 sink with an uptake 7 times stronger than pasture. Turf grass,
following conversion, increased CH4 uptake by 3 times compared to the pasture with
fluxes ranging from -12.1 to -0.4 g CH4 ha-1
d-1
, which is comparable to the native
forest. Daily averages and cumulative CH4 fluxes in both data sets show highly
significant differences in all treatments (p = 0.000 – 0.013). Methane uptake in the
native forest is significantly stronger than all the peri-urban land cover in this study
(p = 0.002, both data sets). All treatments, however, increased their CH4 uptake with
decreasing soil moisture despite any soil disturbance through the fallow and turf
grass establishment.
Table 4-2 Average and cumulative fluxes of CH4 and N2O with standard error for
each treatment together with their significance, as well as calculated global warming
potential (GWP) for the intensive sampling campaign (80 days) in 2013.
CH4 flux
avg
(2009/2010) [g ha-1 d-1]
CH4 flux
avg
(2013) [g ha-1 d-1]
CH4 flux
total
(2013) [g ha-1 80 d-1]
N2O flux
avg
(2009/2010) [g ha-1 d-1]
N2O flux
avg
(2013) [g ha-1 d-1]
N2O flux
total
(2013) [g ha-1 80 d-1]
GWP
[kg CO2-e ha-1 80 d-1]
Forest -3.2a
(± 0.6) -10.1
a
(± 0.2) -807.8
a
(± 29.4) 0.5
a
(± 0.3) 0.03
a
(± 0.02) 2.3
a
(± 4.6) -19.5
a
Pasture -0.2b
(± 0.7) -1.1
b
(± 0.4) -90.1
b
(± 12.9) 0.7
a
(± 0.3) 0.5
b
(± 0.04) 41.6
a
(± 7.9) 10.1
b
Turf
grass
-5.0c
(± 0.4) -396.5
c
(± 41.8) 18.3
c
(± 2.3) 1,460.1
b
(± 269.6) 425.2
c
Fallow -3.1d
(± 0.2) -248.5
d
(± 9.22) 1.6
ab
(± 0.2) 123.2
a
(± 35.0) 30.5
b
± () Standard error from n = 22 (2009/2010), n = 80 (average 2013), n = 3 (total 2013) abcd Significant differences between treatments for each column (p < 0.05), same letters state no statistically significant difference
86
Annual soil N2O measurements identified no significant differences between
forest and pasture, in contrast to the intensive campaign where pasture emitted
significantly more N2O (Table 4-2, Figure 4-1C, D). The native forest had
significantly lower daily N2O emissions compared to all other treatments ranging
from -0.7 to 1.6 g N2O ha-1
d-1
in 2009/2010 and -0.3 to 0.4 g N2O ha-1
d-1
in 2013.
Pasture showed a weak significant difference to the daily averages of forest with
fluxes ranging from -0.4 to 2.7 g N2O ha-1
d-1
in 2009/2010 and -0.1 to 1.3
g N2O ha-1
d-1
in 2013 (p = 0.044). The fallow treatment was not significantly
different to the forest and pasture with fluxes ranging from -0.1 to 10.2 g N2O ha-1
d-1
(p = 0.056). Turf grass fluxes ranged from -0.2 to 83 g N2O ha-1
d-1
, and were
significantly greater than any other treatment (p < 0.01) due to the application of
nitrogen-based fertilizer during planting with emissions responding rapidly to rainfall
and irrigation. After short pulses of highly elevated N2O emissions during the first
month after fertilizer application and lawn establishment, the turf grass lawn’s N2O
fluxes decreased consistently with decreasing soil moisture until the end of the
sampling campaign. The standard error indicates that N2O emissions are variable and
episodic and therefore more difficult to estimate, demonstrating that more intensive
monitoring is necessary to accurately capture N2O flux dynamics as demonstrated
from the intensive sampling campaign.
4.4.3 Global warming potential
To compare cumulative CH4 and N2O fluxes together for the GWP of each land
cover of the intensive sampling campaign, all fluxes were converted to their CO2-
equivalents. Turf grass had the highest GWP (p < 0.01), caused by elevated N2O
emissions from N fertilization. In contrast the consistent CH4 uptake in the forest,
together with negligible N2O emissions, showed a negative GWP and therefore
reduced the radiative forcing in the atmosphere. Overall, pasture and fallow were not
significantly different from each other but both slightly increased the GWP compared
to forest. Pasture, however, still had a substantially lower GWP than turf grass. In
this study, the conversion of pasture to fallow and turf grass resulted in a GWP
increase of 20.4 and 415.1 kg CO2-e ha-1
respectively for the first 80 days after
conversion. Had turf grass been established on forest soil, the GWP increase would
have been 444.7 kg CO2-e ha-1
80 d-1
. These results confirm that area estimations of
Establishing turf grass increases soil greenhouse gas emissions in peri-urban environments (Paper 1)
87
land cover, as well as their land use history need to be considered for urban and peri-
urban GWPs.
Soil water content in the annual 2009/2010 measurements had only a small and
insignificant effect on CH4 and N2O fluxes in native forest (r = 0.12, r = 0.15)
(Figure 4-1E). In the pasture, however, water content significantly affected CH4 but
not N2O fluxes (r = 0.54*, r = 0.32) which means CH4 emissions in the pasture
significantly increased with increasing soil moisture. Temperature had no significant
effect on emissions in the annual measurement. During the 2013 intensive sampling
campaign, soil water content had a significant effect on CH4 fluxes in all treatments
(r = 0.72** – 0.87**), while temperature was only weakly correlated to CH4 in the
turf grass and fallow (r = -0.28*, r = -0.23*) (Figure 1F). Nitrous oxide fluxes were
significantly influenced by water content (r = 0.67** - 0.84**) in all treatments
except pasture (r = 0.16), while no temperature effect was observed. In both data sets
neither CH4 nor N2O fluxes in forest and pasture were affected by temperature.
Significant differences in daily CH4 fluxes showed that the results from the
shorter but more intensive sampling campaign were consistent with the annual
measurements. A comparison of CH4 fluxes of the same season (dry and cold) from
2009/2010 and 2013 show that the differences between native forest and pasture
moved from weakly related (p = 0.052) in 2009/2010 to highly significant
(p = 0.000) in 2013. The statistical evidence for N2O fluxes also changed from an
insignificant difference between treatments in 2009/2010 (p = 0.066) to significantly
different in 2013 (p = 0.012). This change may be due to a lower average water
content (19 % in 2009/2010 and 26 % in 2013), but also the larger data set from the
intensive sampling campaign.
88
Climate
2009/2010
Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar
Rain
[m
m]
0
20
40
60
80
100
120
140
CH4
ForestPastureTurf grassFallow
CH4
CH
4 [
g h
a-1
d-1
]
-15
-10
-5
0
5
10
15
20
N2ON2O
N2O
[g h
a-1
d-1
]
0
10
20
30
40
50
60
Climate
2013
17/06/13 1/07/13 15/07/13 29/07/13 12/08/13 26/08/13
0
10
20
30
40
50
60
Tem
pera
ture
[°C
]
Wate
r conte
nt
[%]
RainIrrigation turf grassAir temperatureSoil water content
C D
A B
E F
Figure 4-1 Daily average CH4 (A, B) and N2O (C, D) fluxes for each treatment with
error bars from the annual measurements (2009/2010) and the intensive sampling
campaign (2013), as well as SERF climate data (E, F) for all sampling periods.
Establishing turf grass increases soil greenhouse gas emissions in peri-urban environments (Paper 1)
89
4.5 Discussion
4.5.1 CH4 and N2O flux measurements
The native forest in this study emitted less N2O fluxes than those reported from
other tropical and subtropical forests around Australia, which range from
2 to 32 g N2O ha-1
d-1
(Breuer et al. 2000; Kiese and Butterbach-Bahl 2002;
Rowlings et al. 2012), which can possibly be explained by the particular forest type
at SERF as well as soil fertility and texture. The lower CH4 uptake from the forest
soil in 2009/2010 compared to 2013 might be explained by the higher annual rainfall
than the long term mean for the experimental site. Reported CH4 and N2O fluxes
from other dry sclerophyll eucalypt forests in Australia are limited to temperate
climates only and range from -0.9 to -16.4 g CH4 ha-1
d-1
and < 0.5 g N2O ha-1
d-1
(Fest et al. 2009; Fest et al. 2015a; Livesley et al. 2009). While the CH4 fluxes
determined here fit well within the range of reported fluxes for this forest type, the
N2O fluxes on the other hand, are the higher end of reported values. Differences
between the temperate zones and the humid subtropical climate with in SEQ are the
heavy rains during the hot summer season. These sometimes extreme weather events
increase N2O emissions due to the strong correlation to the water content of the soil
identified by the intensive sampling campaign. The insignificant correlation analysis
between water content and N2O in the annual 2009/2010 measurements show the
necessity of high frequency measurements as research suggests time differences in
the production of N2O reacting on water content and release even by days (Mosier et
al. 1998), which makes it unlikely for manual sampling to catch a representative flux
after rain events.
Native grasslands from the temperate zones clearly show more CH4 uptake
compared to urban lawns with a daily average of -17.2 and -6.2 g CH4 ha-1
d-1
respectively (Kaye et al. 2004). But the N2O emissions from these urban lawns were
10 times higher than from the native grassland of the temperate zones. The pasture at
SERF showed far less daily CH4 uptake with temporary source behaviour which is
reported for other Australian pastures with comparable soils (Livesley et al. 2009).
The turf grass soil acted as a CH4 sink immediately, despite the disruption to the soil
structure during establishment, this might be due to the intact soil-root structure that
90
came with the turf grass. The effect on N2O emissions was comparable to the
temperate zones, with the urban lawn establishment increasing N2O emissions
significantly due to the management practices. Reported GWP from non-CO2 GHG
fluxes from soils suggest that turf grass lawns can increase the GWP even
comparable to intensive agriculture (Kaye et al. 2004). This reflects the turf grass
N2O emissions from SERF, being 2.4 times higher than average daily values reported
from subtropical Australian intensive pastures (Scheer et al. 2011). Grasslands in
Australia cover an area of approximately 450 million ha and some 40 % of the
terrestrial ice-free surface and are therefore the principal land cover (AGO 2010),
though this figure is not divided into agricultural and residential use. Considering the
significant extent of pastures and grassland worldwide, more data are needed from
different climates and soil types, as well as an internationally uniform categorization
of native, agricultural and residential grassland.
4.5.2 Global warming potential
The N2O fluxes from the fallow soil were not significantly different from the
forest and pasture fluxes at SERF, therefore the increase in GWP must be due to the
turf grass management during establishment rather than soil disturbance from
construction processes. Research from the temperate zones determined the GWP
from well-established lawns of 147 kg CO2-e ha-1
y-1
is higher compared to rural
forests by estimating the decrease in CH4 uptake only (Groffman and Pouyat 2009).
Using the same approach with the decrease in CH4 uptake from forest to turf grass of
0.4 kg CH4 ha-1
after the 80 days of lawn establishment showed a GWP increase of
10.3 kg CO2-e ha-1
When the cumulative N2O emission difference of 1.5 kg N2O ha-1
between forest and turf grass is included in the calculation, the GWP increases about
40 times. This highlights the importance to include N2O emissions into GWP
calculations, as turf grass varies widely in management, e. g. in temperate zones the
reported turf grass was fertilized with up to 200 kg N ha-1
y-1
. The study by Groffman
and Pouyat (2009 also observed a significant decrease in CH4 uptake from rural to
urban forests which indicates that ecosystem productivity is highly influenced by
changing environments during urbanization processes. As native vegetation, the dry
sclerophyll eucalypt forest at SERF might therefore be expected to decrease its CH4
uptake with advancing urbanization in the Samford Valley due to the changing
Establishing turf grass increases soil greenhouse gas emissions in peri-urban environments (Paper 1)
91
environment. Long term monitoring of those unique remnant vegetation sites within
peri-urban environments becomes therefore more important than one-time
estimations.
Half the world’s 7.2 billion population currently occupies 2.4 % of the terrestrial
land surface in urban areas, which are constantly expanding (Potere and Schneider
2007; United Nations 2013). In 2014 the population density of Brisbane was
approximately 140 people per km2 and calculated from the current population
increase this results in an annual urban sprawl of 276 km2 y
-1 (ABS 2015). Despite
arguments that urban and peri-urban areas are too small to contribute important
biogeochemical fluxes on global scales (Kaye et al. 2004), urban lawns in the US
already cover 160,000 km2, a proportion 3 times larger than any other irrigated crop
(Milesi et al. 2005; Groffman and Pouyat 2009). Calculating a general annual GWP
estimate from daily non-CO2 GHG fluxes from the SERF turf grass soil results in 1.9
t CO2-e ha-1
y-1
and exceeds reported values from irrigated lawns in temperate
Australia 1.6 times (Livesley et al. 2010). This coarse annual GWP is based on daily
averages from the first 80 days after turf grass establishment and can be expected to
decrease over time. The climate and continuously high management necessary in the
subtropics, however, might results in less decrease than expected from temperate
climates. Considering turf grass suppliers recommend fertilizing three times a year
with irrigation throughout the year, turf grass will have a strong impact on the GWP
of peri-urban environments. Detailed estimates about the current turf grass cover in
Australia’s urban and peri-urban environments does currently not exist, but annual
turf grass sales range between 4,918 ha and 17,320 ha over the last 10 years (ABS
2012; Turf Australia 2012). The approximate gross value production of Australia’s
turf industry is $ 240 million AUD per annum, while over 40 % is produced by
tropical and sub-tropical Queensland suppliers (ABS 2012; Turf Australia 2012). The
rapid growth of the turf grass industry worldwide as shown on the example of the
extensive turf grass cover in the US, detailed information is needed to accurately
predict trends within the turf grass industry to improve economic and environmental
benefits.
Constantly expanding urban and peri-urban areas worldwide indicate it is most
likely that locally changing climates combined will have an increasing impact on
92
global climate change. Chromosols, similar to these found at SERF, are the most
widespread soil type in agricultural use in Australia (Isbell 2002) and therefore the
most likely to be effected by current and future urbanization. Therefore a range of
soils and different climates need to be studied for an accurate global assessment of
urbanization effects on GHG fluxes and nutrient cycling. In particular, the complex
mechanisms of C and N cycling, gas fluxes and the potential carbon sequestration of
peri-urban soils should be of focus of further research to improve the estimation of
land use change effects due to urbanization. Urbanization effects are currently
neglected in modelled climate scenarios within official IPCC calculations, as little
data exist on urban and peri-urban ecosystems (Betts 2007; Stocker et al. 2013),
highlighting the urgency to research urban and peri-urban ecosystems in the future.
4.5.3 Conclusion
This study distinguishes that turf grass lawn establishment in peri-urban
environments such as Samford in SEQ, Australia, significantly increases soil GHG
emissions. The environmental conditions examined here are representative for wide
areas in Australia and highlight the need for optimised management strategies for
peri-urban environments after land use change. Intensely managed land cover like
turf grass will result in highly elevated N2O emissions due to N fertilizer use and
irrigation, as well as being accelerated by the subtropical climate. This unique data
set is the baseline for long term research on peri-urban environments in the humid
subtropics. The Global Warming Potential of land use change, as determined in this
study, needs to be included in future climate scenarios models to estimate the full
impact of urbanization on climate change and ecosystem health.
93
Statement of Contribution of Co-Authors for Thesis by Published Paper
The authors listed below have certified* that:
1. they meet the criteria for authorship in that they have participated in the conception,
execution, or interpretation, of at least that part of the publication in their field of expertise;
2. they take public responsibility for their part of the publication, except for the responsible
author who accepts overall responsibility for the publication;
3. there are no other authors of the publication according to these criteria;
4. potential conflicts of interest have been disclosed to (a) granting bodies, (b) the editor or
publisher of journals or other publications, and (c) the head of the responsible academic unit,
and
5. they agree to the use of the publication in the student’s thesis and its publication on the
Australasian Research Online database consistent with any limitations set by publisher
requirements.
In the case of this chapter:
Chapter 5: Urbanization-related land use change from forest and pasture into turf
grass modifies soil nitrogen cycling and increases N2O emissions (Paper 2)
Contributor Statement of contribution*
Lona van Delden Performed experimental design, conducted
fieldwork and laboratory analyses, data analysis,
and wrote the manuscript.
Signature
05/07/2017
David W. Rowlings Aided experimental design and data analysis, and
reviewed the manuscript.
Clemens Scheer Aided experimental design and data analysis, and
reviewed the manuscript.
Peter R. Grace Aided experimental design and data analysis, and
reviewed the manuscript.
Chapter 5 (Paper 2) has been published in Biogeosciences in November 2016, Volume 13,
Issue 21, pp 6095-6106.
94
Principal Supervisor Confirmation
I have sighted email or other correspondence from all Co-authors confirming their certifying
authorship.
David W. Rowlings 30/11/2016
Name Signature Date
Urbanization related land use change modifies soil nitrogen (Paper 2)
95
Chapter 5: Urbanization-related land use
change from forest and pasture
into turf grass modifies soil
nitrogen cycling and increases
N2O emissions
(Paper 2)
5.1 Abstract
Urbanization is becoming increasingly important in terms of climate change and
ecosystem functionality worldwide. We are only beginning to understand how the
processes of urbanization influence ecosystem dynamics, making peri-urban
environments more vulnerable to nutrient losses. Brisbane in South East Queensland
has the most extensive urban sprawl of all Australian cities. This research estimates
the environmental impact of land use change associated with urbanization by
examining soil nitrogen (N) turnover and subsequent nitrous oxide (N2O) emissions
with a fully automated system that measured emissions on a sub-daily basis. There
was no significant difference in soil N2O emissions between a native dry sclerophyll
eucalypt forest and an extensively grazed pasture, wherefrom only low annual
emissions were observed amounting to 0.1 and. 0.2 kg N2O ha-1
y-1
, respectively. The
establishment of a fertilized turf grass lawn increased soil N2O emissions by 18 fold
(1.8 kg N2O ha-1
y-1
) with highest emission occurring in the first 2 month after
establishment. Once established, the turf grass lawn presented relatively low N2O
emissions after fertilization and rain events for the rest of the year. Soil moisture was
significantly higher and mineralised N accumulated in fallow land, resulting in
highest N2O emissions (2.8 kg N2O ha-1
y-1
) and significant nitrate (NO3-) losses of
up to 63 kg N ha-1
from a single rain event due to plant cover removal. The study
concludes that urbanization processes into peri-urban ecosystems can greatly modify
N cycling and increase the potential for losses in form of N2O and NO3-.
96
5.2 Introduction
Global urbanization processes are becoming increasingly important in terms of
global warming and ecosystem functionality. Urban populations worldwide have not
only exceeded rural populations but are also predicted to account for all future
population growth (United Nations 2008). Urban sprawl and increasing population
densities are causing severe land use changes from intact biomes and commercially
focused agriculture into smaller residential properties with introduced species. This
transition from rural to semi-rural, i.e. peri-urban, and urban environments is
increasingly associated with development and construction processes and the
extensive establishment of turf grass for residential backyards, public parks and
sportsgrounds, and golf courses (IPCC 2006). How these urbanization processes
influence ecosystem dynamics in biogeochemical cycling, and therefore contribute to
ecosystem vulnerability and global warming is only beginning to be understood.
The consequences of land use change from native to agriculture have been
identified by several studies, including a loss in soil quality (structure and nutrient
losses) and quantity (erosion), increase greenhouse gas (GHG) emissions, and
reduced potential for soil carbon (C) sequestration (Livesley et al. 2009; Grover et al.
2012). On the other hand, changing soils from agricultural to residential use in
temperate climates has shown the potential to improve critical ecosystem services by
(i) providing stormwater treatment, (ii) acting as a sink for atmospheric nitrogen (N)
and (iii) sequestering C (Golubiewski 2006; Raciti et al. 2011a).
Studies on the impact of those land use changes on climate change are few but
suggest that urbanization will change the biogeochemical cycling of C and N and
associated nutrient turnover (Grimm et al. 2008). These biogeochemical alterations
induced by land use change interact with urban effects (Betts 2007); such as creating
heat islands through vegetation replacement and surface sealing but also increased
local carbon dioxide concentration of over 500 ppm around cities compared to 390
ppm in natural environments (Pataki et al. 2007; IPCC 2013). These changing local
climatic conditions and their feedback effects onto natural ecosystems make peri-
urban environments more vulnerable to nutrient losses and potential sources of GHG
emissions. With these peri-urban areas expanding worldwide it is most likely that
these changing local climates will increasingly have an impact on global climate
Urbanization related land use change modifies soil nitrogen (Paper 2)
97
change, making an examination of GHG emissions from peri-urban land uses all the
more urgent.
Nitrous oxide (N2O) along with carbon dioxide (CO2) and methane (CH4) is one
of the major greenhouse gases with a global warming potential (GWP) nearly 300
times that of CO2 (IPCC 2013). Nitrous oxide is produced principally by
microorganisms during nitrification and denitrification processes from mineral N
(NH4+ and NO3
-) in the soil. The production of N2O is influenced by a number of soil
parameters including substrate availability, temperature, and availability of oxygen,
which is dependent on water content and texture of the soil (Rowlings et al. 2015).
With the predicted climatic changes, Australia’s ancient and fragile soils will most
likely be affected in their balance between GHG gas emissions and consumptions
(Baldock et al. 2012). Management practices such as fertilization and irrigation
enhance N2O production in the soil by increasing the mineral N content and limiting
the oxygen availability (Scheer et al. 2008; Rowlings et al. 2013). Turf grass is the
most highly managed land use of peri-urban environments in terms of fertilization,
irrigation and frequent mowing, which therefore has a high potential for N2O
emissions.
Research on urban and peri-urban areas in temperate zones suggests that changes
in biogeochemical cycling due to urbanization will substantially affect the global
climate comparable to agriculture, with those areas and their intensive management
expanding rapidly worldwide (Milesi et al. 2005; Groffman et al. 2009). More than
half the world’s 7.2 billion population currently occupies 2.4 % of the global
terrestrial land surface in urban areas (Potere and Schneider 2007; United Nations
2013). While peri-urban environments are often considered too small to be of
consequence, the rapid growth of the turf grass industry as highlighted in the USA
covers over 160,000 km2 occupied with turf grass lawn, three times more than any
other irrigated crop in the country (Milesi et al. 2005). In Australia about 60 % of all
anthropogenic N2O emissions come from cropped and grazed soils and the first GHG
estimations from turf grass establishment support the emission intensity reported
from temperate zones (AGO 2010; van Delden et al. 2016a). The use of turf grass is
consistently growing in Australia with up to 17,320 ha in turf grass sales and an
approximate gross value production of $240 million AUD per annum (ABS 2012;
98
Turf Australia 2012). Detailed estimates of turf grass cover, however, currently do
not exist for the Australian continent and other subtropical regions like South-east
Asia, China, India or Mexico. Urbanization is currently neglected in modelled IPCC
climate scenarios, mainly due to limited data on C and N processes in urban and peri-
urban environments (IPCC 2006, 2013).
Therefore, this study aims to identify the impact of those land use changes
associated with urbanization on annual N2O emissions and their driving parameters
in subtropical peri-urban environments. Following a short-term (80 days) GHG
sampling campaign focussing on lawn establishment (van Delden et al. 2016a), a
fully automated closed static chamber system was used to continuously monitor N2O
fluxes together with soil biogeochemical processes over a full year to determine the
seasonal impact of construction work and conversion from extensively grazed
pasture to turf grass lawn. This study’s high-resolution flux measurements and
supporting soil N mineralisation illustrate the vulnerability of ecosystems to
urbanization processes and the potential impact on N cycling and N2O emissions.
5.3 Materials and Methods
5.3.1 Site description
The study was conducted at the Samford Ecological Research Facility (SERF) in
the Samford Valley, 20 km from Brisbane in South-East Queensland, Australia. The
Samford Valley covers an area of approximately 82 km2 and is surrounded by
mountains to the north, west and south. Mostly cleared in the early 1900s, the valley
was developed in the 1960s for dairy and beef cattle as well as intensive agriculture
including banana and pineapple. Samford’s population density has increased rapidly,
almost doubling from 1996 – 2006, causing land use change from predominately
rural to residential properties (Moreton Bay Regional Council 2011). As a result,
SERF contains the last remnant forest of the valley floor. The valley is influenced by
a humid subtropical climate with seasonal summer rain. The long term mean annual
precipitation is 1110 mm with mean annual minimum and maximum temperatures of
13 °C and 25.6 °C respectively (BOM 2015). The soil at the experimental site is
characterised by a strong texture contrast between the A and B horizon and is
Urbanization related land use change modifies soil nitrogen (Paper 2)
99
classified as brown Chromosol according to the Australian soil classification (Isbell
2002) and Planosol according to the World Reference Base (WRB 2015).
5.3.2 Experimental design
This study examines the impact of land use change from a native forest to well-
established pasture, turf grass lawn and fallow soil without plant cover using the
same sampling campaign setup as van Delden et al. (2016a). Each land use treatment
included 3 replicated plots, 2 m wide by 10 m long and separated by 0.5 m of pasture
as a buffer zone. The turf grass lawn and fallow treatments were established within
the well-established pasture to create a randomised plot design, 50 m from the native
forest. The SERF native forest (Dry sclerophyll eucalypt forest) is a baseline for
historical land use and was unmanaged. The well-established Chloris gayana pasture
represents rural development in the area and has been extensively grazed for the last
15 years. Livestock, however, was excluded over the course of the study and the
pasture grass was slashed 5 times during the study to ensure it did not exceed the
maximum height of the GHG measurement chamber.
The turf grass lawn was established from the well-established pasture by
removing 5 cm of topsoil with grass roots. The soil was rotary hoed twice to a depth
of 15 cm and Blue Couch (Digitaria didactyla) was planted with 50 kg N ha-1
fertilization (13.6.2013) to aid in establishment. Over the experimental year the turf
grass lawn was fertilized twice (26.10.13, 6.3.2014) with 50 kg N ha-1
and irrigated,
in all 150 kg N ha-1
y-1
of Prolific Blue AN fertilizer (12.0 % nitrogen, 5.2 %
phosphorus, 14.1 % potassium, 1.2 % magnesium) with two-thirds of the N content
by mass in the ammonium form. The turf grass lawn was irrigated to a total of
30 mm during drier months as well as after fertilization. The turf grass was mowed
with the clippings removed as soon as the grass grew to the maximum chamber
height, once in spring, twice in summer and twice in autumn, and kept free of weeds
manually at all times. Fertilization rates were based on half the local industry
practices recommendation. Infrequent mowing represents the normal management
for residential properties in this region and is normally in response to increased
growth in the wetter and warmer summer months.
100
The fallow treatment simulated the impact of transitional processes such as
construction work and plant cover replacement. In the fallow treatment, the grass
cover was removed and the bare soil was rotary hoed twice to a depth of 15 cm. The
fallow treatment was kept free from plant cover over the full experimental year with
a non-selective herbicide (Bi-Active 360g/L Glyphosate) and a broad leaf herbicide
(Double Time, 340g/l MCPA + 80g/l Dicambra). During the experimental year,
high-resolution sub-daily N2O flux measurements were combined with mineral N
analysis and site-specific climate and soil moisture measurements.
5.3.3 N2O flux measurements
Nitrous oxide fluxes were determined from mid-June 2013 to mid-June 2014
using an automated sampling system as detailed by Scheer et al. (2014b), extending
the turf grass establishment phase documented by van Delden et al. (2016a) into a
full measurement year. The pneumatically operated 50 cm x 50 cm x 15 cm high,
clear acrylic glass chambers were secured to stainless steel bases, permanently
inserted 10 cm into the ground. The chambers were moved each week between two
bases per treatment plot, to minimize the influence of the chamber microclimate,
while measurements were analysed continuously. The chambers were connected to
an automated sampling system and an in-situ gas chromatograph (SRI GC8610,
Torrance, CA, USA) equipped with 63Ni Electron Capture Detector (ECD) for N2O.
One replicate chamber from each of the four treatments was closed for one hour, and
four headspace gas concentrations measured at 15 minute intervals, followed a
known calibration standard (0.5 ppm N2O, Air Liquide, Houston, TX, USA). This
process was repeated for the remaining two replicate chambers over a full cycle of
three hours, allowing eight flux measurements to be calculated per day, for each of
the 12 chambers.
5.3.4 Auxiliary measurements
Soil samples were taken fortnightly from all replicated treatment plots over the
experimental year and divided into 2 depths (0-10 cm, 10-20 cm). NH4+ and NO3
-
were extracted from the soil using a 1:5 KCl solution with 20 g of fresh soil with
additional soil moisture determination at 105°C to identify the dry soil weight for the
mineral N calculation as described by Carter and Gregorich (2007. The extract was
Urbanization related land use change modifies soil nitrogen (Paper 2)
101
analysed for NH4+ and NO3
- with an AQ2+ discrete analyser (SEAL Analytical WI,
USA). The net mineralisation rate was determined from differences in mineral
content between sampling dates (Hart et al. 1994). Soil moisture and temperature for
each treatment were collected using a TDR probe (HydroSense CD 620 CSA) and a
PT100 probe (IMKO Germany). Soil moisture was then converted with the treatment
specific bulk density (BD) to water-filled pore space (WFPS). Soil samples were
taken for site characterization with a hydraulic soil corer to 1 m depth, air dried and
sieved to 2 mm. Particle size analysis for soil texture as well as BD, pH and electrical
conductivity (EC) analysis were undertaken according to Carter and Gregorich
(2007. The cation exchange capacity (CEC) was determined based on Rayment and
Higginson (1992. Total C and N content of air dried soil and plant material was
determined by dry combustion (CNS-2000, LECO Corporation, St. Joseph, MI,
USA) from ground samples.
5.3.5 Flux calculations and statistical analysis
Fluxes were calculated from the slope of the linear increase or decrease of the 4
concentrations measured over the closure time and corrected for chamber
temperature and atmospheric pressure, using the procedure outlined by Knowles and
Singh (2003) and Scheer et al. (2014b). The linear regression coefficient (r2) was
calculated and used as a quality check for fluxes above the detection limit to assure
linearity of the gas concentration increase. Flux rates were discarded if r2 was < 0.85
for N2O fluxes (Scheer et al. 2013). Daily fluxes from the automated chambers were
calculated by averaging sub-daily measurements for each chamber over the 24 hour
period. The detection limit determined for the gas sampling system is ± 1.2 g N2O ha-
1 d
-1. Gaps in the dataset were filled by linear interpolation across missing days.
Statistical analysis was undertaken using SPSS Statistics 21.0 (IBM Corp.,
Armonk, NY). Non-normal distribution meant all cumulative data were log-
transformed for ANOVA analysis using Games-Howell as the post-hoc test. Daily
N2O flux differences between treatments were interpreted by plotting 95 %
confidence intervals using R studio. A significant difference of p < 0.05 between
treatments was assumed in case the confidence intervals of all treatments were not
overlapping. A Spearman’s rho correlation analysis was used to examine
102
relationships between gas fluxes, soil chemistry, soil moisture and temperature. The
significance value (p) is shown for each analysis, as well as the correlation
coefficient (r) with its significance level (p < 0.05*, p < 0.01**).
5.4 Results
5.4.1 Site characteristics
The site received 740 mm of rain during the experimental year, substantially less
than the long-term average (Table 5-1). Wet season rainfall was delayed compared to
the historic average, with less than half the rainfall in summer (December to
February) compared to autumn (March to May) (Table 5-2). Substantial out of
season rain also fell in the spring with over 200 mm in November alone. Rainfall was
highly episodic, with the highest daily rain event of 108.8 mm in March 2014. The
mean annual minimum and maximum temperatures for the experimental year were
16.7 °C and 27.1 °C respectively, and light ground frost occurred twice in August.
The turf grass and fallow treatment were established within the pasture and therefore
share the same soil profile with its characteristics, except for bulk density in the A1
horizon, which changed after the turf grass establishment from1.4 to 1.2 g cm-3
. The
CEC of the sandy topsoil is very low, and slightly higher in the A1 compared to the
A2 horizon due to the higher soil organic matter as indicated by the total C and N
content. Nutrient removal in turf grass clippings added up to 1.8 t C ha-1
y-1
and 30
kg N ha-1
y-1
lost from the system during the experiment year. The turf’s biomass
production was approximately 6.3 kg C ha-1
d-1
and 0.13 kg N ha-1
d-1
in dry matter
but varied widely depending on fertilization and available water and increased up to
10.4 kg C ha-1
d-1
.
Urbanization related land use change modifies soil nitrogen (Paper 2)
103
Table 5-1 – SERF site characteristics
Parameters
Longitude 152° 52' 37.3" E Latitude 27° 23' 22.211" S Altitude 60 m Slope 2° Mean annual min temp. 13 °C* Mean annual max temp. 25.6 °C* Mean annual rain 1110 mm*
Soil profile Horizon**
Depth (cm)
Sand (%)
Silt (%)
Clay (%)
BD (g cm
-3)
pH EC (μS)
CEC (meq+/100g)
Total C (%)
Total N (%)
Pasture A1 0 – 17 70 24 6 1.4 5.4 46 4.0 1.5 0.12 A2 17 – 45 74 18 8 1.6 6.0 10 0.9 0.9 0.07 B2 45 – 92 9 18 73 1.8 6.1 31 11.8 0.4 0.03
Forest A1 0 - 20 75 18 7 1.4 5.5 29 2.8 1.8 0.14 A2 20 - 47 78 15 7 1.5 5.6 30 0.9 1.1 0.08 B2 47 - 70 44 39 17 1.7 5.6 30 11.8 0.2 0.02
*Long term means by Commonwealth Bureau of Meteorology, Australian Government (BOM)
**According to the Australian soil classification
Table 5-2 - Seasonal and cumulative rain, number of rain events and seasonal and
annual averages of minimum and maximum Temperatures of the experimental year
Sum Rain (mm)
Number of rain events*
Avg Temperature (°C)
Min Max
Winter 51.2 0 11.5 22.6 Spring 248.2 5 16.7 28.2 Summer 137.2 3 20.7 30.2 Autumn 303.2 3 17.6 27.5
739.8 11 16.7 27.1
5.4.2 Environmental parameters
The lowest WFPS during the experimental year was 13 % in the forest, with the
highest occurring in the pasture, which briefly reached saturation in March 2014
(Figure 5-3). In all treatments, the lowest WFPS occurred in spring and summer with
an average of 33 and 32 % respectively together with the highest average maximum
daily temperatures of 28 and 30 °C. While the highest seasonal WFPS for all
treatments occurred in winter, the maximum WFPS occurred during autumn after the
heavy rain in March 2014. The forest had significantly lower WFPS throughout the
experimental year than all other treatments (p < 0.01, Table 5-3), while the fallow
had significantly higher WFPS (p < 0.01). No significant difference in WFPS was
observed between pasture and turf grass (p > 0.05) although during spring, summer
and autumn turf grass had lower minimum and maximum values than the pasture.
104
Fallow had significantly higher and forest significantly lower WFPS than pasture and
turf grass (p < 0.01) throughout the experimental year.
5.4.3 Temporal variability of mineral N
Averaged over the experimental year the fallow treatment had the highest NH4+
and NO3- content across 20 cm soil profile, followed by turf grass, pasture and forest
(Table 5-3). These differences in mineral N were significant for all treatments (p <
0.01) except between pasture and forest (p > 0.05). The 0-10 cm depth had higher
average mineral N, NH4+
and NO3- than the 10 – 20 cm depth for all treatments with
significant differences between all treatments (p < 0.01). In 10 – 20 cm soil depth
only the fallow had significantly higher mineral N, NH4+
and NO3-
contents (p <
0.01). Soil NH4+ showed relatively little temporal variation and remained
consistently above 3 kg NH4+ ha
-1 while NO3
- decreased substantially after rain
events and fell below detection limit several times in all treatments but the fallow
(Figure 5-1).
Total mineral N in the forest ranged from 8 to 40.1 kg N ha-1
20 cm-1
throughout
the year with marginally higher mineral N content from 0-10 cm than 10-20 cm with
9.7 and 8 kg N ha-1
respectively. Total mineral N in the pasture ranged from 5.1 to
42.4 kg N ha-1
20 cm-1
, with a comparable distribution in depth than the 0-10 cm and
10-20 cm forest soil with 10.7 and 7.8 kg N ha-1
respectively. Total mineral N in the
turf grass soil ranged from 9.1 to 127.6 kg N ha-1
20 cm-1
. The turf grass had twice as
much mineral N in 0-10 cm than 10-20 cm depths with 20.7 and 10.1 kg N ha-1
respectively. A short-term increase in both NH4+ and NO3
- content in the soil was
evident after fertilization in June, October and March, which decreased to the
background levels after approximately one month. Total mineral N contents in the
fallow soil ranged from 19.7 to 160.7 kg N ha-1
20 cm-1
with about 2/3 of the mineral
N being located in the upper 10 cm. All main changes in the fallow’s total mineral N
content were caused by variations in NO3-
rather than NH4+. The NO3
- content
increased in the fallow until the major rain event in March when it dropped from
95.5 to 32.8 kg N ha-1
20 cm-1
. From the linear increase in mineral N content within
the upper 10 cm between January and March 2014 a soil mineralization rate of 0.6 kg
N ha-1
d-1
was estimated.
Urbanization related land use change modifies soil nitrogen (Paper 2)
105
NO3
NO
3
- [k
g N
ha
-1]
0
20
40
60
80
100
120
NH4
NH
4
+ [
kg
N h
a-1
]
0
20
40
60
80
100
120
ForestPastureTurf grassFallow
Climate
Jun
13
Jul 1
3
Aug
13
Sep
13
Oct 1
3
Nov
13
Dec
13
Jan
14
Feb 1
4
Mar
14
Apr
14
May
14
Jun
14
Jul 1
4
Ra
in [
mm
]Ir
rig
ati
on
[m
m]
0
20
40
60
80
100
120
Te
mp
era
ture
[°C
]
0
10
20
30
40
50
Rain Irrigation (turf grass)Min temperatureMax temperature
Fertilization (turf grass)
A
B
C
Figure 5-1 - Annual soil NO3
- (A) and NH4
+ (B) contents variations from forest,
pasture, turf grass and fallow averaged across replicates (n = 3) and summed for
separate analysed soil depths of 0-10 and 10-20 cm with the climatic conditions (C)
for the experimental year 2013/2014 as well as fertilization and irrigation indication
for the turf grass treatment.
106
Table 5-3 - Annual mineral N averages as NH4+-N and NO3
--N in 0-20 cm soil
depth, WFPS and daily maximum and average N2O fluxes from all treatments with
their cumulative annual fluxes over the experimental year with their standard error.
NH4+-N
(kg ha-1
) NO3
--N
(kg ha-1
) WFPS
(%) Max daily flux (g N2O ha
-1 d
-
1)
Avg daily flux (g N2O ha
-1 d
-1)
Annual flux (kg N2O ha
-1 y
-1)
Forest 13.7a ± 1.2 3.9
a ± 0.6 23
a 8.1 0.4
a ± 0.1 0.1
a ± 0.03
Pasture 17.4b ± 1.4 1.1
b ± 0.3 42
b 18.3 0.6
a ± 0.1 0.2
a ± 0.2
Turf grass 21.9bc
±2.4 8.9c ± 2.5 43
b 83.0 4.9
b ± 0.6 1.8
b ± 0.3
Fallow 26.0c ± 1.9 35.2
d ± 5.6 55
c 123.8 7.7
b ± 1.0 2.8
b ± 1.0
abcd different letters indicate significant differences between treatments based on p <0.05
5.4.4 Temporal variability of N2O fluxes
Daily N2O fluxes across all treatments ranged from extended periods close to zero
to over 123 g N2O ha-1
d-1
from the fallow with the highest WFPS after heavy rain
events (Figure 5-2). The N2O fluxes from turf grass were more often significantly
different on daily basis than any other treatment with 76 days (21 %) of the
experimental year, 80 % of this difference occurred in the first two months after
establishment. This was followed by the forest with 65 days (18 %), fallow with 58
days (16 %) and pasture with 29 days (8 %). Daily N2O fluxes from the forest soil
showed no substantial temporal variation throughout the experimental year, with
minor emission peaks up to 8.1 g N2O ha-1
d-1
after large rain events (> 60 mm) in
November and March. From September until October one of the two bases in one
pasture replicate emitted substantially more N2O than the other replicates, however,
the exact cause of this is unknown. Without these spatially variable emissions, the
annual flux would have been about 40 % lower and therefore comparable to the
forest N loss of 0.09 kg N ha-1
y-1
. During the initial emission peak between June and
August the daily average N2O flux from the turf grass was 24 g N2O ha-1
d-1
,
reaching a maximum of 83 g N2O ha-1
d-1
. Excluding this initial emission peak, daily
N2O fluxes from the turf grass averaged 1.2 g N2O ha-1
d-1
. The highest annual N2O
flux was measured in the fallow from only 3 large peaks over 19, 10, and 44
consecutive days after rain events which together accounted for 85 % of the total N
losses. Over a third of the significantly high daily N2O fluxes in the fallow occurred
from the heavy rain event in March 2014.
Urbanization related land use change modifies soil nitrogen (Paper 2)
107
Forest
N2O
[g
ha
-1 d
-1]
0
50
100
150
200
WF
PS
[%
]
0
20
40
60
80
Turf grass
N2O
[g
ha
-1 d
-1]
0
50
100
150
200
WF
PS
[%
]
0
20
40
60
80
Pasture
N2O
[g
ha
-1 d
-1]
0
50
100
150
200
WF
PS
[%
]
0
20
40
60
80
Fallow
Jun
13
Jul 1
3
Aug 1
3
Sep 1
3
Oct 1
3
Nov 1
3
Dec
13
Jan
14
Feb 1
4
Mar
14
Apr 1
4
May
14
Jun
14
Jul 1
4
N2O
[g
ha
-1 d
-1]
0
50
100
150
200
WF
PS
[%
]
0
20
40
60
80
WFPS A
B
C
D
Figure 5-2 - Daily N2O flux averages (max 8 fluxes per day for 3 replicates each)
with standard errors (n =3) over the experimental year 2013/2014 for forest (A),
pasture (B), turf grass (C) and fallow (D) with the treatment specific water filled pore
space (WFPS).
108
Annual N2O losses were highest in the fallow and turf grass treatments totalling
1.78 and 1.15 kg N ha-1
y-1
respectively compared to the pasture and forest losses of
0.15 and 0.09 kg N ha-1
y-1
(p < 0.01, Table 5-3). About 80 % of the annual N2O
losses in the turf occurred in the first 8 weeks after establishment (Figure 5-3).
Mineral N fertilizer input of 150 kg N ha-1
y-1
and the yearly N2O-N losses from the
turf grass lawn corrected for background emissions (zero N fertilization) from the
pasture resulted in an emission factor (EF) of 0.7 % (Kroeze et al. 1997).
Jun
13
Jul 1
3
Aug
13
Sep
13
Oct
13
Nov
13
Dec
13
Jan
14
Feb 1
4
Mar
14
Apr
14
May
14
Jun
14
Jul 1
4
Ra
in [
mm
]
0
20
40
60
80
100
120
Cu
mu
lati
ve
N2O
flu
xe
s [
kg
N2O
ha
-1 y
-1]
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Rain
Forest
Pasture
Turf grass
Fallow
Figure 5-3 - Cumulative daily N2O fluxes (n = 3) for forest, pasture, turf grass and
fallow with rainfall for the experimental year 2013/2014.
Urbanization related land use change modifies soil nitrogen (Paper 2)
109
5.4.5 Environmental parameters influencing N2O fluxes
Mineral N contents in the forest and fallow soils were not significantly correlated
to N2O fluxes on a daily basis (Table 5-4). However, the linear regression shown in
Figure 5-4 identified a clear increase of N2O emissions with increasing annual
mineral N contents for all treatments during the establishment phase as well as
during the rest of the year. This relationship is supported by the substantial N2O
emissions peaks from the fallow and simultaneous decrease in NO3-
after the two
biggest rain events in November 2013 and March 2014, with WFPS above 70 %. The
separate linear regression for all land uses with plant cover, i.e. forest, pasture, and
turf grass, identified an even stronger relationship of mineral N and N2O. Forest and
turf grass N2O fluxes were strongly and fluxes from the pasture and fallow
moderately correlated to their WFPS. Temperature was moderate negatively
correlated to N2O fluxes as well as mineral N for pasture and turf grass. In the fallow
temperature strongly affected mineral N contents but not N2O fluxes. Mineral N in
the fallow soil was strongly negative correlated to its WFPS mostly because of the
strong negative correlation of NO3- with WFPS with r = -0.56**.
Table 5-4 - Spearman’s rho correlation coefficient between N2O fluxes and
mineral N, WFPS and temperature for each treatment.
N2O Mineral N
Mineral N WFPS Temperature WFPS Temperature
Forest -0.39 0.61** 0.40** 0.07 -0.17 Pasture 0.49** 0.30** -0.46** 0.39** -0.41** Turf grass 0.47** 0.78** -0.45** 0.47** -0.50** Fallow -0.0 0.43** 0.20** -0.61** 0.72**
** correlation coefficient significant with p < 0.01
110
Mineral N [kg ha-1
]
5 10 15 20 25 30 35 40
Lo
g (
N2O
flu
x)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0Forest
Pasture
Turf grass
Fallow
Linear regression
Mineral N [kg ha-1
]
5 10 15 20 25 30 35 40
y = 0.0878x-0.2743
R2 = 0.96
A B
y = 0.0464x-0.4306
R2 = 0.85
Figure 5-4 – Linear relationship of log transformed N2O emissions with mineral N
content within 20 cm soil depth for each replicate of forest, pasture, turf grass and
fallow land use during the establishment phase (A) and the rest of the year (B), with
the coefficient of determination R2.
5.5 Discussion
This study combines the first high frequency estimates of subtropical N2O fluxes
and annual mineral N cycling from dry sclerophyll forests, unfertilized pastures and
turf grass lawns, the most common land uses associated with urban and peri-urban
environments. The lack of high frequency field measurements in urban and peri-
urban environments makes accurate assumptions and mitigation strategies difficult.
Conventional gas sampling methods most likely result in an over- or underestimation
of emissions, as the production and release of N2O can differ in time (Mosier et al.
1998). This research gap together with the strong temporal variability of subtropical
heavy rain events underlines the importance of automated high frequency
measurements to capture representative soil-atmosphere gas exchange. The
subtropical climatic zone represents an often neglected area of research, despite the
subtropics covering 3.26 M ha in Australia alone, as well as large areas on the North
and South American continent, Africa and Asia. Differences between other climates
and the humid subtropics are the heavy summer rains leading to high soil moisture
and temperatures favourable for high soil microbial activity. Therefore this study
identified mineral N content and WFPS as the main parameters driving N2O
Urbanization related land use change modifies soil nitrogen (Paper 2)
111
production in the soil while studies from temperate zones report temperature as the
main driver (Butterbach-Bahl and Kiese 2005; Fest et al. 2009).
5.5.1 Mineral N
Mineralised N in form of NH4+
and NO3- determines the production and loss of N
via N2O and depends on climatic parameters like temperature as well as substrate and
oxygen availability. Nutrient mineralisation is often faster in sandy soils but the rapid
infiltration and low nutrient holding capacity of the A horizon of the Chromosol
decreases the highly mobile NO3- content substantially after heavy rain events. This
amount of NO3-
is not only lost for plant uptake but can pollute groundwater and
open waterways resulting in eutrophication. In this study mineral N contents from the
forest and pasture treatments where driven by annual variation in temperature and
moisture contents whereas turf grass lawn and fallow where dominated by
management. The negative correlation of temperature in the pasture and turf grass
can be most likely explained by the higher plant productivity during the warmer
summer and spring resulting in higher plant NO3- and water uptake with increasing
temperature which then reduces soil moisture conditions.
Soil mineral N in the SERF forest was generally low and dominated by NH4+, and
while less seasonally variable throughout the year than NO3-, still responded to
rainfall. The overall mineral N reported from temperate eucalypt forest soils was
double the annual SERF average of 17.6 kg N ha-1
with up to 38.1 kg N ha-1
(Fest et
al. 2009; Livesley et al. 2009; Fest et al. 2015a). However, the greater NO3-
proportion in the sandy SERF soil of 3.9 kg N ha-1
compared to 0.8 kg N ha-1
of
temperate sandy forest soils (Livesley et al. 2009) indicates a higher mineral N
availability in the subtropics. Average NO3-
contents from other dry sclerophyll
forest are even lower with 0.02 kg N ha
-1 (Fest et al. 2015a). The higher N
availability is most likely due to faster soil organic matter turnover in the subtropical
climate with higher temperatures in combination with the main annual rainfall. While
in temperate summers it is mostly dry during the high temperatures which limits
microbial activity, the humid summers in SEQ not only accelerate N turnover but
also water and N uptake by plant, which therefore reduces potential N losses.
Subtropical rainforests, on the other hand, present with 97.7 kg N ha-1
up to 6 times
112
higher mineral N contents than the SERF soil, suggesting a lower N turnover
associated with the low net primary productivity (NPP) of the dry sclerophyll forests
(Rowlings et al. 2012). Overall NH4+:NO3
- ratios from Australian forests indicate
higher NO3- availability in subtropical forest soils (3-4) compared to temperate zones
(28-125) (Livesley et al. 2009; Rowlings et al. 2012; Fest et al. 2015a). These
differences in N availability suggest that N cycling in forest soils is mainly regulated
by the climate as opposed to soil type and NPP.
Ammonium was the dominant mineral N form in the SERF pasture, similar to the
forest and in line with other subtropical pastures in Australia (Rowlings et al. 2015).
The SERF soil reflects the overall minor annual variability of NH4+ compared to
NO3- across most climates in Australia. The overall mineral N content at the SERF
pasture soil was at the lower end of the reported values from both temperate and
subtropical pastures which is most likely explained by the lower clay contents at the
site which fixes NH4+
and higher N inputs by legumes (Livesley et al. 2009;
Rowlings et al. 2015). For example, in other extensively used subtropical pastures
NH4+ annual values did not drop below 55 kg N ha
-1, which is three times higher than
the SERF annual NH4+ average (Rowlings et al. 2015). While NH4
+ at SERF is
comparable to temperate Australian pastures, NO3- in the SERF pasture soil is at the
lower end (Livesley et al. 2009). This indicates an efficient system from tied up N in
organic material to the plant uptake of NO3-
, which supports the hypothesis of an
efficient N cycle within well-established land use.
Annual mineral N variations in the SERF turf grass were mainly controlled by the
fertilization events but rapidly fell back to background levels after each application.
The fertilizer mineral N peak was particularly emphasized after the first application,
where soil NO3- was more than double than after subsequent fertilization events. This
is possibly due to the undeveloped root system and therefore less N uptake as well as
additional plant available N in the added turf grass rolls. These NO3- peaks together
with irrigation, which is particularly needed during turf grass establishment, implies
a high N leaching potential in other Australian sandy soils of up to 80 kg N ha-1
y-1
(Barton et al. 2006). With the high potential of heavy rain events in the subtropics,
fertilizer rates and timing needs to be considered carefully to avoid excessive N
losses in form of NO3- displacement.
Urbanization related land use change modifies soil nitrogen (Paper 2)
113
The fallow soil had the highest WFPS content throughout the year due to plant
cover removal and therefore no further water uptake by the roots, creating favourable
condition for soil mineralisation and losses (Robertson and Groffman 2007). The
moist conditions together with the temperatures of the warmer season resulted in
accelerated N turnover and without the N uptake by plants substantial amounts of
NO3- accumulated in the soil. Despite the fact that mineral N in the fallow soil never
dropped back to zero, substantial amounts of NO3 were lost from the topsoil after
heavy rain events, not only as N2O emissions but also through NO3-
displacement
into deeper soil layers. The low CEC of the sandy topsoil highlights the minor
nutrient holding capacity of this peri-urban environment. These potential N losses
after heavy rain events demonstrates the significant impact of plant cover removal
and soil disturbance in peri-urban ecosystems.
5.5.2 N2O fluxes
The study illustrates that land use change associated with urbanization can
significantly alter soil N turnover resulting in elevated soil N2O emissions and
increased N losses from the soil. During the experimental year of this study, autumn
was the wettest season and therefore had the highest N2O emissions from all
treatments but with different intensity from the different land use systems. Soil N2O
emissions were significantly different between the investigated land use systems with
the temporal variations in daily N2O fluxes and primarily controlled by WFPS.
However, the linear increase of N2O emissions with increasing NO3- content in the
soil may be the result of higher denitrification than nitrification rates in the SERF
soil. The high surface sand content of the Chromosol, combined with the moderate
slope, prevents excessive water logging over long periods of time, which limits N2O
gaseous losses from denitrification in saturated soil conditions.
The daily N2O average of 0.4 g N2O ha-1
d-1
from this study’s subtropical dry
sclerophyll forest is lower than the averages of < 1.2 g N2O ha-1
d-1
reported from
temperate Australian dry sclerophyll forests (Fest et al. 2009; Livesley et al. 2009).
This might be explained by the overall low total C and N and the below average
rainfall during the experimental year. Considering the positive correlation of N2O
emissions and NO3-
content in the soil, it was expected that the higher NO3-
114
availability in the SERF forest compared to the temperate dry sclerophyll forest also
causes higher N2O emissions. The low WFPS which was > 40 % for most of the year
inhibited denitrification processes and caused therefore lower N2O emissions
compared to the temperate zones as well as increased NO3- uptake during the humid
subtropical summer. This efficient N cycling together with the low NPP of the dry
sclerophyll forest and low clay content at SERF causes also lower N2O losses
compared to subtropical rainforests (Rowlings et al. 2012). This study supports the
general hypothesis that forest soils are minor contributors to the global N2O budget,
although other N2O emission studies of Australian forest soils provide only a limited
comparison of temporal N2O variability due to infrequent or short-term
measurements (Fest et al. 2009; Page et al. 2011; Fest et al. 2015a).
The annual N2O emissions from the SERF pasture are comparable to other
reported extensive pastures across Australia (1-2 kg N ha-1
y-1
) but substantially
lower than unfertilized pasture in the northern hemisphere (Dalal et al. 2003). Annual
emissions from other studies on subtropical Australian pastures reported to be up to
3.4 kg N2O ha-1
y-1
and highly inter-annual variable depending on rainfall (Rowlings
et al. 2015). This exceeded the annual N2O emissions at SERF by nearly 17 times,
which may have been limited by the dry year and high sand content.
The first N2O emission peak after the turf grass’s establishment caused the
majority of the annual N2O emissions and was not repeated after two additional
fertilization events. This initial N2O peak can be explained by the underdeveloped
root system and consequently a reduced NO3- uptake by the turf grass, which together
with the irrigation stimulated nitrification and denitrification and consequential N2O
emissions. The high N demand from the highly productive turf grass later on results
in the immediate uptake of mineral N and therefore minor N2O emissions. The
annual N2O emissions from the SERF’s turf grass are more than double the N2O
emissions from extensive Australian pastures reported in the literature (Dalal et al.
2003). The SERF turf grass lawn emitted about 3.3 times more N2O on daily average
during the experimental year than native pasture from the temperate zones, but only
half of the reported values for urban turf grass in the USA which were comparable to
intensive agriculture (Kaye et al. 2004). However, compared to Australian
intensively managed pastures, N2O emissions from the SERF turf grass were 50 %
lower (Scheer et al. 2011). Differences between reported values and the SERF turf
Urbanization related land use change modifies soil nitrogen (Paper 2)
115
grass are most likely explained by differences in texture and the total N content in the
SERF soil being nearly 4 times lower. Reported EFs from temperate pastures also
vary substantially between experimental years due to differences in received rainfall
(Jones et al. 2005). It could be therefore expected that the SERF’s turf grass EF will
increase in wetter years. However, in subtropical systems is has been proven that the
total amount of annual rainfall received is not as decisive for annual N2O emissions
as rainfall patterns and intensities (Rowlings et al. 2015). These differences between
temperate and subtropical N cycling make short-term N2O flux measurements
difficult to compare and need further investigation in the global subtropics.
Significantly higher NO3- contents occurred 3 months after plant cover removal in
the fallow soil but only during the warm and wet summer season substantial N2O
emissions were observed. The two significant N2O emission peaks from the fallow
were most likely caused by denitrification processes from the accumulated NO3- and
soil moisture conditions after major rain events. These N2O emission peaks mirror
the NO3- decrease from the soil after those rain events but cannot completely account
for it, suggesting that most NO3-
was leached below 20 cm soil depths or lost via
other gases such as N2. All other treatments, including the fertilized turf grass,
prevented potential N2O production in the soil by rapid NO3-
uptake from plants.
Therefore, plant cover removal makes ecosystems undergoing land use change most
vulnerable to substantial N losses under humid subtropical climate conditions.
5.5.3 Effect of land use change associated with urbanization
This study determined that urbanization related land use change results in an
accumulation of NO3-
in fallow topsoil and elevated N2O emissions, mainly after
heavy rain events. The results presented here verify that subtropical N2O emissions
positively correlate to mineral N content in the soil and therefore indicate that land
use change increases N2O emissions from the soil, especially after plant cover
removal and establishment of fertilized turf grass lawn. The annual variation in daily
N2O fluxes confirm that despite soil moisture as the strongest climatic parameter
influencing N2O emissions, the individual land use is the main influence on the soil-
atmosphere gas exchange. Extended periods of fallow soil in particular should be
avoided during urbanization processes, as bare soil is highly vulnerable to N losses
116
due to plant cover removal. Turf grass lawn, as a fertilized and highly managed land
cover, leads to significantly changed soil conditions compared to the forest and
pasture land use types. However, this turf grass lawn in the subtropical climate of
SEQ has lower emissions against expectations based on the high emission findings
from temperate zones (Kaye et al. 2004; Tratalos et al. 2007; Grimm et al. 2008).
Substantial N2O emissions were only observed within the first 2 months after turf
grass establishment, while over the remaining 10 months only minor fluxes occurred
even after further fertilization events. While N2O emissions from the turf grass were
reduced substantially over time, emissions from the fallow increased with time due to
more available NO3-. Therefore, the N2O emissions of well-established turf grass
lawns need to be considered separately to their production and establishment phase
as well as potential N losses from fallow land targeted for the entire duration of land
use change, which should be kept as short as possible (Barton et al. 2006; van Delden
et al. 2016a).
Research from temperate zones suggests a C sequestration potential from the
higher productivity of turf grass lawns (Golubiewski 2006; Lorenz and Lal 2009;
Raciti et al. 2011a). Others argue that the positive effect of C sequestration however
can easily be offset by the high N demand together with irrigation, resulting in
increased N2O emissions and overall nutrient losses caused by management practices
like mowing and clipping removal (Conant et al. 2005; Wang et al. 2014). Australian
ecosystems with highly weathered soils, however, are generally low in nutrient
stocks and often limited in their C sequestration potential (Livesley et al. 2009). The
SERF turf grass, however, presented relatively low N2O emissions when excluding
the establishment phase, which implies the potential to balance emissions with C
sequestration. A full life cycle assessment needs to determine if turf grass lawn in the
subtropics is increasing or decreasing the GWP of peri-urban environments by
balancing C sequestration and GHG emissions, not only from the soil but also
through the production, distribution and use of fertilizer, fuel and chemicals (Selhorst
and Lal 2011).
Urbanization related land use change modifies soil nitrogen (Paper 2)
117
5.6 Conclusions
This study provides evidence that land use change associated with urbanization
accelerates N turnover and increase N2O emissions from soils by presenting the first
high temporal frequency dataset on peri-urban soils in the subtropics for a full year
after land use change. These findings demonstrate that GHG emissions from peri-
urban areas should be included into future IPCC climate change scenarios and rural
to urban land development guidelines need to be established for GHG emission
mitigation. Three main factors need to be considered to target N2O losses from soils
during land use change associated with urbanization: (i) previous land use, (ii)
duration of development process, and (iii) new land use purpose that it is being
changed into, i.e. public or private. The dry sclerophyll forest in this study supports
the general hypothesis that forest soils are low N2O emitters, contrary to expectation
that the humid subtropical summer conditions would increase emissions compared to
temperate forest soils. The accumulation of NO3- in fallow soil increases the potential
for N2O emissions and may amplify considering future predictions of rising
temperatures and more frequent heavy rain events worldwide. Increased fertilizer
application may be required to compensate for these N losses after land use change
to keep land uses, such as turf grass, highly productive while altering N cycling in
peri-urban environments. The outcomes of this study highlight the substantial NO3-
accumulation in soils during land use change, which consequently increases N2O
emissions and should be accounted for in global climate forecasts as urbanization
processes are predicted to increase worldwide with increasing population growth.
5.7 Acknowledgements
This study was undertaken at the Samford Ecological Research Facility (SERF)
one of the Supersites in the Terrestrial Ecosystem Research Network (TERN). The
study was supported by the Institute for Future Environments (IFE) of the
Queensland University of Technology (QUT). The data set “Greenhouse gas
emissions from peri-urban land use at SERF, SEQ. 2013-2015” can be found online
at the N2O network under http://www.N2O.net.au/knb/metacat/vandelden.3.3/html.
118
Statement of Contribution of Co-Authors for Thesis by Published Paper
The authors listed below have certified* that:
1. they meet the criteria for authorship in that they have participated in the conception,
execution, or interpretation, of at least that part of the publication in their field of expertise;
2. they take public responsibility for their part of the publication, except for the responsible
author who accepts overall responsibility for the publication;
3. there are no other authors of the publication according to these criteria;
4. potential conflicts of interest have been disclosed to (a) granting bodies, (b) the editor or
publisher of journals or other publications, and (c) the head of the responsible academic unit,
and
5. they agree to the use of the publication in the student’s thesis and its publication on the
Australasian Research Online database consistent with any limitations set by publisher
requirements.
In the case of this chapter:
Chapter 6: Soil N2O and CH4 fluxes from urbanization related land use change: From
forest to pasture and turf grass (Paper 3)
Contributor Statement of contribution*
Lona van Delden Performed experimental design, conducted
fieldwork and laboratory analyses, data analysis,
and wrote the manuscript.
Signature
05/07/2017
David W. Rowlings Aided experimental design and data analysis, and
reviewed the manuscript.
Clemens Scheer Aided experimental design and data analysis, and
reviewed the manuscript.
Daniele de Rosa Aided data analysis, and reviewed the manuscript.
Peter R. Grace Aided experimental design and data analysis, and
reviewed the manuscript.
Chapter 5 (Paper 3) has been submitted to Global Change Biology.
119
Principal Supervisor Confirmation
I have sighted email or other correspondence from all Co-authors confirming their certifying
authorship.
David W. Rowlings 30/11/2016
Name Signature Date
Soil N2O and CH4 fluxes from urbanization related land use change (Paper 3)
121
Chapter 6: Soil N2O and CH4 fluxes from
urbanization related land use
change; from Eucalyptus forest
and pasture to urban lawn
(Paper 3)
6.1 Abstract
Increasing population densities and urban sprawl are causing rapid land use
change from natural and agricultural ecosystems into smaller, urban residential
properties, altering biogeochemical C and N cycles. However the impact of
urbanization on the soil-atmosphere exchange is largely unknown. This study
quantified the soil–atmosphere exchange of N2O and CH4 in three land uses
representing typical land use intensification from a native forest to a well-established
pasture and a fertilized turf grass lawn in the subtropical peri-urban region of
Brisbane, Australia. Fluxes were measured continuously over two years using a high
resolution automated chamber system to account for short-term and inter-annual
variability. The fertilised turf grass had the highest temporal variation in N2O
emissions, dominated by extremely high fluxes immediately following
establishment, while only small fluxes occurred in the forest and pasture (0.08 – 0.15
kg N2O-N ha-1
y-1
). Apart from the high N2O emissions in the turf grass during the
establishment phase, there was little inter-annual variability in fluxes across all land
uses, despite substantial rainfall variations between years. The high aeration of the
sandy topsoil limited N2O emissions while promoting substantial CH4 uptake with all
land uses being net CH4 sinks. Native forest was consistently the strongest CH4 sink
(-2.9 kg CH4-C ha-1
y-1
), while the pasture became a short-term CH4 source after
heavy rainfall when the soil reached saturation. On a two years average, land use
change from native forest to turf grass increased the non-CO2 GWP by 329 kg CO2-e
ha-1
y-1
, turning it from a net GHG sink into a source. The study highlights that
urbanization can substantially alter soil atmosphere exchange by increasing bulk
122
density and inorganic N availability. However on well drained tropical soils, the long
term non-CO2 GWP of turf grass was comparably low compared to results reported
from temperate climates.
6.2 Introduction
Urban populations worldwide have not only exceeded rural populations but are
also predicted to account for the majority future population growth (United Nations,
2014). Increasing population densities and urban sprawl are causing rapid land use
change from natural ecosystems and commercially focused agriculture into smaller,
urban residential properties. This transition from rural to urban environments, i.e.
peri-urban, is associated with construction processes and increasingly the extensive
establishment of intensively managed turf grass for residential backyards, public
parks and sportsgrounds and golf courses (IPCC, 2006). The impact this ecosystem
change has on biogeochemical processes associated with soil-atmosphere gas
exchange and the effects on global climate are only beginning to be understood
(Betts, 2007; Grimm et al., 2008).
The consequences of land use change from natural ecosystems to agriculture have
been identified by several studies and include a loss in soil quality (structure and
nutrient losses) and quantity (erosion), increased greenhouse gas (GHG) emissions,
and a reduced soil potential for carbon sequestration (Livesley et al., 2009; Grover et
al., 2012). Natural ecosystems are estimated to sequester 3.55 Gt CO2-e y-1
from the
atmosphere into biomass and any reduction in this sink will have a significant impact
on climate change (Dalal and Allen, 2008). On the other hand, urban soils have been
shown to increase carbon (C) sequestration over their existing agricultural land uses,
as well as providing critical ecosystem services such as being a sink for atmospheric
nitrogen (N), and stormwater treatment and storage (Golubiewski, 2006; Raciti et al.,
2011a).
Soils represent a major source of the trace GHGs methane (CH4) and nitrous
oxide (N2O) driven by the natural biogeochemical cycling of C and N and greatly
modified by anthropogenic practices such as fertilization, irrigation and physical
disturbance. The few studies examining GHG emissions in peri-urban environments
Soil N2O and CH4 fluxes from urbanization related land use change (Paper 3)
123
have focused mainly on CO2 exchange, while N2O and CH4 have often been
neglected (Tratalos et al., 2007; Lorenz and Lal, 2009; Ng et al., 2014). However,
potential terrestrial CO2 uptake can be offset by relatively small increases in N2O
and CH4 emissions (Tian et al., 2014). Nitrous oxide is 298 times more potent than
CO2 as a GHG (Myhre et al., 2013) and is produced principally by microorganisms
during nitrification and denitrification processes in the soil. Atmospheric CH4 uptake
into the soil occurs via microbial consumption by methanotrophic bacteria for an
energy source and is the largest natural sink of CH4. This process is highly sensitive
to alterations of physical soil conditions and diffusivity, which can change soils to a
CH4 source when methanogenic activity dominates under saturated soil moisture
conditions (Groffman and Pouyat, 2009). The CH4 flux can generally be considered
the net result of simultaneously occurring production and consumption processes in
the soil (Butterbach-Bahl and Papen, 2002).
Emissions of these GHG’s are driven by environmental parameters such as soil
moisture dynamics, which is also a function of soil bulk density, nutrient input and
substrate availability; factors that are greatly modified by land use change (Verchot
et al., 1999; Werner et al., 2006; Yashiro et al., 2008; Rowlings et al., 2012b).
Fertilization and irrigation enhances GHG production in the soil by increasing N
substrate and limiting oxygen availability (Scheer et al., 2008; Rowlings et al.,
2013). Turf grass is the most highly managed land use of peri-urban environments in
terms of fertilization, irrigation and physical disturbances associated with
establishment and frequent mowing and emissions comparable to intensive
agriculture (Kaye et al., 2004; Durán et al., 2013). While peri-urban environments
are often neglected due to their fragmented distribution, collectively peri-urban turf
grass occupies over 15 Mha in the USA alone, three times more than any other
irrigated crop in the country (Milesi et al., 2005). Subtropical Florida for example,
produces over 42,000 ha of new commercial turf grass per year with a total economic
impact estimated at $703 M USD (Satterthwaite et al., 2009).
Humid subtropical climatic zones such as South-East Queensland (SEQ),
Australia, are dominated by large episodic rain events during the summer and early
autumn. Combined with high year-round soil temperatures, this creates soil
conditions favourable for high GHG emissions with large annual and inter-annual
124
variability (Rowlings et al., 2015). Brisbane in SEQ currently has a population
growth rate of 1.7 % per year with one of the most extensive areas of urban sprawl in
Australia (ABARES, 2010; Commonwealth of Australia, 2013). Recent research
identified the potential for high N2O emissions and modified N cycling from turf
establishment due to higher management intensity compared to extensively used
rural pasture and native forest (van Delden et al., 2016b). However, the contribution
of urban land use change to changes in soil-atmosphere GHG exchange has yet to be
estimated due to limited data. Therefore, this study established the first non-CO2
global warming potential (GWP) from an Australian subtropical peri-urban
environment. High frequency N2O and CH4 measurements using automated
chambers identified flux dynamics following land use change from native forest to a
well-established grazed pasture and residential turf grass lawn. Previous research on
this subtropical peri-urban environment identified a significant but short-lived
increase in N2O emissions and reduction in CH4 uptake up to 2 months after turf
grass establishment (van Delden et al., 2016a; van Delden et al., 2016b). However,
substantial inter-annual N2O flux variations from subtropical pastures suggest that
the climate conditions can significantly
However, substantial inter-annual climate variations can account for significant
differences in annual GHG estimations of subtropical pastures, especially for highly
temporal variable N2O emissions (Rowlings et al., 2015). Therefore can be expected
that the non-CO2 GWP of a peri-urban environment with fertilized turf grass lawn
will vary significantly between years with substantial climate variations. This study
was conducted over two consecutive years to account for the significant inter-annual
climate variations of the humid subtropics to evaluate the temporal variability of soil-
atmosphere GHG exchange. It was hypothesized that soil-atmosphere N2O and CH4
flux dynamics are mainly driven by the substantial inter-annual climate variations
and result in significantly higher emissions than during the turf grass establishment
phase.
Soil N2O and CH4 fluxes from urbanization related land use change (Paper 3)
125
6.3 Materials and Methods
6.3.1 Site description
The study was conducted at the Samford Ecological Research Facility (SERF) in
the Samford Valley, 20 km from Brisbane, Australia. The Samford Valley was
intensively cleared of native vegetation in the early 1900s, and developed in the
1960s for dairy and beef cattle as well as intensive agriculture including banana and
pineapple. Samford’s population density has increased rapidly, almost doubling from
1996 – 2006 as land uses changed from predominantly rural to peri-urban (Moreton
Bay Regional Council, 2011). The region is influenced by a humid subtropical
climate with seasonal summer (December to February) rain. The long term mean
annual precipitation is 1110 mm with mean annual minimum and maximum
temperatures of 13 °C and 25.6 °C respectively (BOM, 2015). The soil at the
experimental site is characterised by a strong texture contrast between the A and B
horizon and is classified as brown a Chromosol according to the Australian soil
classification (Isbell, 2002) and a Planosol according to the World Reference Base
(WRB, 2015).
6.3.2 Experimental design
The turf grass lawn was established in June 2013 following the typical practice of
the region by removing the dense pasture sward and surface roots (0-2 cm) to expose
the topsoil which was then rotary hoed twice to a depth of 15 cm. Fertilizer was
applied at the rate of 50 kg N ha-1
(Prolific Blue AN fertilizer: 8 % ammonium, 4 %
nitrate, 5.2 % phosphorus, 14.1 % potassium, 1.2 % magnesium) immediately prior
to the placement of Blue Couch (Digitaria didactyla) turf rolls and irrigated with 10
mm. The same fertilizer was surface applied at the rate of 50 kg N ha-1
followed by
immediate irrigation an additional four times (26.10.13, 6.3.14, 28.9.14, 8.1.15). This
totalled 250 kg of N fertilizer ha-1
over the two year study, less than the local
industry practice recommendation (300 kg N ha-1
y-1
) but representative of private
and public turf grass use in the region (Moreton Bay Regional Council, 2011).
Additional irrigation was applied during very dry periods to ensure turf survival. The
turf grass was mowed 11 times with the clippings removed and weighed as soon as
126
the grass reached 20 cm height and kept manually free of weeds at all times. This
infrequent mowing represented the average management for residential properties in
this region.
6.3.3 GHG flux measurements
High resolution N2O and CH4 measurements were collected from mid-June 2013
to mid-June 2015 using an automated sampling system as detailed by Scheer et al.
(2014) and van Delden et al. (2016a). The pneumatically operated 50 cm x 50 cm x
15 cm high clear acrylic glass chambers were secured to stainless steel bases,
permanently inserted 10 cm into the ground. The chambers were moved weekly
between two bases per plot to minimize the influence of the chamber microclimate
on plant growth. The chambers were connected to an automated sampling system and
an in-situ gas chromatograph (SRI 8610C, Torrance, CA, USA) equipped with 63
Ni
Electron Capture Detector (ECD) for N2O and a Flame Ionization Detector (FID) for
CH4. One replicate chamber from each of the four treatments was closed for one
hour, and four gas concentrations measured at 15-minute intervals followed by a
known calibration standard (0.5 ppm N2O, 4.0 ppm CH4, Air Liquide, Houston, TX,
USA). This process was repeated for the remaining two replicates over a full cycle of
three hours, allowing a maximum of eight flux measurements per day for each of the
12 chambers. The sampling system was also equipped with a nondispersive infra-red
CO2 analyser for continuous measurements which was used to detect chamber leaks
(LI-820; LI-COR, Lincoln Nebraska, USA).
6.3.4 Auxiliary measurements
Soil samples were taken for site characterization to 1 m depth, air dried and sieved
to 2 mm. Particle size analysis for soil texture as well as bulk density (BD), pH and
electrical conductivity (EC) analysis were undertaken according to Carter and
Gregorich (2007). Soil moisture and temperature for each treatment were collected
using a TDR probe (HydroSense CD 620 CSA) and a PT100 probe (IMKO
Germany) respectively. Soil moisture was then converted with the treatment specific
BD to water-filled pore space (WFPS) according to Haney and Haney (2010). Total
C and N content of soil and air dried turf grass clippings were determined by dry
combustion (CNS-2000, LECO Corporation, St. Joseph, MI, USA).
Soil N2O and CH4 fluxes from urbanization related land use change (Paper 3)
127
6.3.5 Flux calculations and statistical analysis
Fluxes were calculated from the slope of the linear increase or decrease of the four
concentrations measured over the closure time and corrected for chamber
temperature and atmospheric pressure using the procedure outlined by Scheer et al.
(2014). The coefficient of determination (r2) was calculated and used as a quality
check for fluxes above the detection limit to assure linearity of the gas concentration
increase. Fluxes were discarded if the r2 was < 0.85 for N2O, < 0.95 for CH4 and <
0.98 for CO2 fluxes (Scheer et al., 2013). Daily fluxes from the automated chambers
were calculated by averaging sub-daily measurements for each chamber over the 24
hour period. The detection limit determined for the gas chromatograph was ± 1 g
N2O and CH4 ha-1
d-1
. Gaps in the dataset were filled by linear interpolation across
missing days. The non-CO2 GWP was calculated from the CO2-equivalents (CO2-e)
for N2O and CH4 of 298 and 34 respectively (Myhre et al., 2013). The emission
factor (EF) for the fertilized turf grass lawn was calculated and corrected for
background emissions (zero N fertilization from the pasture) according to Kroeze et
al. (1997).
Statistical analyses for cumulated annual N2O and CH4 fluxes, non-CO2 GWP and
annual WFPS averages were undertaken using SPSS Statistics 21.0 (IBM Corp.,
Armonk, NY). Non-normal distribution meant all data were log-transformed for
ANOVA using the Ryan-Einot-Gabriel-Welch Q (REGWQ) as post-hoc test. The
value (p) is shown for significance between treatments. An autoregressive integrated
moving average (ARIMA) model (Box and Pierce, 1970) was used in R studio to
determine autocorrelation between successive daily N2O and CH4 averages which
includes covariate effects between measurements. A significant difference in daily
fluxes between treatments of p < 0.05 was assumed when the 95 % confidence
intervals were not overlapping. The ARIMA coefficient was interpreted as the
expected difference between current and lagged values for a covariate unit increase.
128
6.4 Results
6.4.1 Environmental and soil parameters
The experimental site received 740 mm rain in year one, two-thirds of the long-
term mean annual precipitation and 430 mm less than the 1170 mm that fell in year
two (Table 6-1, Figure 6-1). Rainfall was highly episodic for both years, with over
70 % of the annual rainfall occurring in only 17 days in year one and nearly 80 % in
19 days in year two. Three times as many heavy rain events (> 50 mm day-1
)
occurred in year two compared to year one, when WFPS exceeded 60 % (Table 6-2).
The largest rain events during the experiment occurred over 2-4 successive days in
March 2014 (204 mm), February 2015 (234 mm), and April 2015 (206 mm). Overall,
the highest 24-hour rain event was 169 mm in April 2015. The annual minimum and
maximum temperatures for the experiment ranged from 3 to 39 °C, with year two on
average 1.6 °C warmer than year one (p < 0.05).
Jun 13 Oct 13 Feb 14 Jun 14 Oct 14 Feb 15 Jun 15
Te
mp
era
ture
[C
°]
0
10
20
30
40
50
Ra
infa
ll [
mm
]
0
50
100
150
200
Daily minimum temperatureDaily maximum temperatureRain
Figure 6-1 – Daily minimum and maximum temperatures and rainfall for the two
years from June 2013 until June 2015 for the experimental site
The bulk density of the A1 horizon in the turf changed from 1.4 g cm-3
prior to
land use change to 1.2 g cm-3
one year after the turf grass establishment. Nutrient
removal in turf grass clippings totalled 2.9 t C and 54 kg N ha-1
over the two years
with no significant difference between years. This averaged 6.3 kg C ha-1
d-1
and
0.13 kg N ha-1
d-1
and varied widely depending on fertilization and available water
with highest removal of up to 20.8 kg C ha-1
d-1
and 0.56 kg N ha-1
d-1
in January
2015.
Soil N2O and CH4 fluxes from urbanization related land use change (Paper 3)
129
Table 6-1 – Site characteristics
Parameters
Longitude 152° 52' 37.3" E Latitude 27° 23' 22.211" S Altitude 60 m Slope 2° Mean annual min temp. 13 °C* Mean annual max temp. 25.6 °C* Mean annual precipitation 1110 mm*
Soil profile
Horizon **
Depth (cm)
Sand (%)
Silt (%)
Clay (%)
BD (g cm
-3)
pH EC (μS)
CEC (meq+/100g)
Total C (%)
Total N (%)
Forest A1 0 - 20 75 18 7 1.4 (1.3)*** 5.5 29 3 1.8 0.14
A2 20 - 47 78 15 7 1.5 5.6 30 1 1.1 0.08
B2 47 - 70 41 7 52 1.7 5.6 30 12 0.2 0.02
Pasture &Turf grass
A1 0 – 17 70 24 6 1.4 (1.4)*** 5.4 46 4 1.5 0.12
A2 17 – 45 74 18 8 1.6 6.0 10 1 0.9 0.07
B2 45 – 92 9 18 73 1.8 6.1 31 12 0.4 0.03
*Long term means by Commonwealth Bureau of Meteorology, Australian Government (BOM) **According to the Australian soil classification
***BD in 0-10 cm soil depth
Table 6-2 Annual rainfall, number of heavy rain events and annual average
minimum and maximum temperatures for the experimental years 2013 and 2014.
Sum Rain (mm)
Number of heavy rain events*
Temperature (°C)
Min Max Avg
Year 1 740 2 16.7 27.1 19.7 Year 2 1170 6 16.5 26.7 21.3
* Heavy rain event if sum > 50 mm per day resulting in >60 % WFPS
The lowest WFPS recorded during the experiment was 9 % in the forest in early
December 2014 (Figure 6-2c), with the highest occurring in the pasture (Figure 6-3c)
which briefly reached saturation in late March 2014. The forest had significantly
lower WFPS than pasture and turf grass (p < 0.01, Table 6-3) over the two years,
while no significant difference was observed between pasture and turf grass (p >
0.05). However, the increase in WFPS following the heavy rain events was lower in
the turf grass (Figure 6-4c) which showed only two thirds of the amplitude of
response compared to the WFPS increase in the pasture. All land uses showed
significantly higher WFPS on average in year two compared to year one (p < 0.05),
which reflects the differences in rainfall. The mean seasonal WFPS was higher in
autumn and winter in year one but higher in summer and autumn in the year two.
WFPS responded quickly to rainfall with highest values across all land uses after
heavy rain events regardless the season.
Soil NH4+
in the top 10 cm ranged from 2 to 36 kg N ha-1
and while NO3- ranged
from zero to 31 kg NO3--N ha
-1 and varied little between years. The turf grass soil
130
had on average the highest annual total inorganic N concentrations (NO3- + NH4
+)
with an average of 31 kg N ha-1
, while forest and pasture had significantly less with
18 and 19 kg N ha-1
respectively. Inorganic N was dominated by the NH4+ form, with
NO3- only accounting for 22% and 6% for the forest and pasture respectively, and
29% of the turf. While concentrations in the turf grass where highest after
fertilization events, the seasonal dynamic in the forest and pasture was highest at the
end of winter and lowest at the end of summer. Annual dynamics of inorganic N can
be found in van Delden et al. (2016b).
Table 6-3 – Daily and annual N2O and CH4 flux averages, the non-CO2 global
warming potential (GWP) and water filled pore space (WFPS) for all three land uses
with their standard error and indication for significant differences between land uses
WFPS (%)
Avg N2O-N (g ha
-1 d
-1)
Annual N2O-N (kg ha
-1 y
-1)
Avg CH4-C (g ha
-1 d
-1)
Annual CH4-C (kg ha
-1 y
-1)
GWP (kg CO2-e ha
-1 y
-1)
Year 1
Forest 23a 0.2
a ± 0.06 0.09
a ± 0.02 -8.1
a ± 0.38 -2.9
a ± 0.14 -93.1
a ± 15.4
Pasture 42b 0.4
a ± 0.29 0.15
a ± 0.11 -2.1
b ± 0.59 -0.8
b ± 0.22 34.0
a ± 44.6
Turf grass 43b 3.2
b ± 0.56 1.15
b ± 0.20 -5.2
c ± 0.14 -1.9
c ± 0.05 451.5
b ± 95.4
Year 2
Forest 27a 0.2
a ± 0.07 0.08
a ± 0.03 -6.8
a ± 0.31 -2.5
a ± 0.11 -74.1
a ± 13.7
Pasture 47b 0.4
a ± 0.17 0.14
a ± 0.06 -2.2
b ± 0.79 -0.8
b ± 0.29 -9.9
ab ± 15.0
Turf grass 44b 0.6
a ± 0.18 0.21
a ± 0.07 -3.6
b ± 0.18 -1.3
b ± 0.07 38.6
b ± 28.3
abcd Different letters indicate significant differences between treatments per column based on p <0.05
6.4.2 N2O fluxes
Daily N2O fluxes ranged from below the detection limit in all land uses to a
maximum of 9 g N2O-N ha-1
d-1
in the forest (Figure 6-2a), 38 g N2O-N ha-1
d-1
in the
pasture (Figure 6-3a) and 73 g N2O-N ha-1
d-1
in the turf grass (Figure 6-4a). The turf
had the highest temporal variation in emissions, dominated by extremely high fluxes
immediately following establishment, which then decreased to levels comparable
with the pasture after 2 months (Figure A 12). This resulted in over 18 % of days in
year one having significantly higher fluxes from the turf grass than forest and pasture
compared to just 4% in year two. Daily forest and pasture N2O fluxes on the other
hand showed no substantial temporal variation throughout the two years. The
significant differences in daily fluxes between experimental years is supported by the
annual N losses, which were 8 fold higher from the turf grass compared to the forest
and pasture in year one, but comparable emissions in year two (p > 0.05, Table 6-3).
Over the full two years of the experiment however the turf still lost significantly
more N (1.36 kg N2O-N ha-1
2y-1
) than the forest and pasture (p < 0.01), which were
not significantly different from each other (p > 0.05). Forest and pasture emitted
Soil N2O and CH4 fluxes from urbanization related land use change (Paper 3)
131
about half of their annual N2O emissions when WFPS was greater than 60 %,
compared to 80 % from the turf. Mineral N fertilizer input and the yearly N2O-N
losses from the turf grass resulted in an EF of 0.7 % and 0.07 % for the year one and
two respectively.
6.4.3 CH4 fluxes
Daily CH4 flux averages across all treatments ranged from -11.1 to 23.1 g CH4-C
ha-1
d-1
, with lowest values measured from the forest and highest from the pasture
(Table 6-3). The forest soil acted as a constant sink for CH4 throughout the
experiment (Figure 6-2b). In year two, CH4 uptake decreased to almost zero for short
periods, following a series of large (>100 mm) rain events in February and May 2015
where WFPS exceeded 80%. By comparison, CH4 uptake in the pasture was close to
zero for 152 days over the two years, with a total of 309 g CH4-C ha-1
of emissions
over 58 days. This occurred mostly in autumn 2014 and 2015 when WFPS exceeded
85% (Figure 6-3b). Overall, the pasture was an annual CH4 sink for both
experimental years. No CH4 uptake was observed from the turf grass soil on 56 days
over the two years (Figure 6-4b). Methane uptake decreased in all land uses when
WFPS was extremely low, following long periods without major rainfall, and when
very high, following heavy rain events. Mowing and slashing of the grasslands
pasture and turf grass did not significantly influence WFPS and therefore soil CH4
fluxes. Time series analysis identified the forest having significantly stronger daily
CH4 uptake than the pasture and turf on 63 % of days over the two-year experiment
(Figure A 12). The pasture soil had on average 42 % of the two years significantly
less CH4 uptake than the forest and turf grass, mainly during the CH4 emission
phases. The turf grass was only 4 % of the time significantly different in daily CH4
uptake to the other land uses.
The annual CH4 flux from the forest averaged -2.7 kg CH4-C ha-1
y-1
over the two
years, with 16 % higher uptake recorded in year two. This was significantly higher
than the pasture and turf grass (p < 0.01) where an average annual flux of -0.8 and -
1.6 kg CH4-C ha-1
y-1
respectively was recorded. Annual uptake was higher in the turf
grass in year one, decreasing over 30 % in year two. Pasture and turf grass were not
significantly different from each other in year two (p > 0.05), although a series of
132
localised CH4 emissions of up 47.2 g CH4-C ha-1
d-1
from one chamber frame over
86 days resulted in significantly less CH4 uptake than the turf grass in year one (p <
0.05).
6.4.4 Influence of environmental parameters on N2O and CH4
fluxes
The time series analysis of all land uses identified a small but highly significant
effect of WFPS that increased N2O emissions by 0.02 – 0.05 g N2O-N ha-1
d-1
for
every 1 % WFPS increase while CH4 uptake decreased by 0.05 – 0.07 g CH4-C ha-1
d-1
(Table 6-4). This makes WFPS the dominant climatic driver of N2O and CH4
fluxes over temperature in all three land uses. Despite this clear trend of decreasing
CH4 uptake in all land use types with extremely high and low WFPS, no clear
correlation could be determined for this dynamic (r2 = 0.33). There was a positive but
non-significant effect of soil NO3- on N2O in the pasture and the forest, however the
ARMIA time-series analysis predicted increasing N2O emission concurrent with
decreasing NO3- concentrations in the turf, with a 1 g N2O-N ha
-1 d
-1 emission
increase for every -0.3 kg decrease in NO3-. This same analysis also confirmed the
significant impact the turf grass establishment phase, increasing N2O emissions 12
fold. Soil mineral N content had a contrasting effect on N2O emissions in the turf
grass with a time series predicted 1 g N2O-N ha-1
d-1
emission increase with a
decrease of -0.3 kg NO3- ha
-1 but increase of 0.1 kg NH4
+ ha
-1 (Table 6-4).
Table 6-4 ARIMA time series coefficient for N2O and CH4 fluxes in g ha-1
d-1
and
water filled pore space (WFPS) for all three land uses and temperature as well as
mineral N (NO3- and NH4
+) and the factor of turf grass establishment impact on N2O
and CH4 emissions
WFPS (%)
Temperature (°C)
NO3-
(kg N ha-1
) NH4
+
(kg N ha-1
) Establishment factor
N2O 12***
Forest 0.03*** 0.02*** 0.05 -0.01 Pasture 0.02*** 0.01 0.3 -0.01 Turf grass 0.05*** 0.01*** -0.3*** 0.1*
CH4 0.7
Forest 0.06*** -0.04* -0.14 0.02 Pasture 0.05*** 0.09 -0.37 0.24 Turf grass 0.05*** -0.01 0.02 0.02 * ARIMA coefficient significance with p < 0.05
*** ARIMA coefficient significance with p < 0.001
Soil N2O and CH4 fluxes from urbanization related land use change (Paper 3)
133
N2O
-N [
g h
a-1
d-1
]
0
2
4
6
8
10
12
Forest
CH
4-C
[g
ha
-1 d
-1]
-16
-12
-8
-4
0
Jun 13 Oct 13 Feb 14 Jun 14 Oct 14 Feb 15 Jun 15
Te
mp
era
ture
[C
°]
5
10
15
20
25
30
WF
PS
[%
]
0
20
40
60
80
100
TemperatureWFPS
a
b
c
Figure 6-2 Two years (June 2013 to June 2015) of N2O (a) and CH4 (b) fluxes
from the dry sclerophyll forest soil with supporting environmental parameters (c)
mean daily temperature and water filled pore space (WFPS).
134
N2O
-N [
g h
a-1
d-1
]
0
10
20
30
40Pasture
CH
4-C
[g
ha
-1 d
-1]
-10
0
10
20
30
40
50
Jun 13 Oct 13 Feb 14 Jun 14 Oct 14 Feb 15 Jun 15
Te
mp
era
ture
[C
°]
5
10
15
20
25
30
WF
PS
[%
]
0
20
40
60
80
100TemperatureWFPS
a
b
c
Figure 6-3 Two years (June 2013 to June 2015) of N2O (a) and CH4 (b) fluxes
from the agricultural pasture soil with supporting environmental parameters (c) mean
daily temperature and water filled pore space (WFPS).
Soil N2O and CH4 fluxes from urbanization related land use change (Paper 3)
135
N2O
-N [
g h
a-1
d-1
]
0
20
40
60
80
100 Turf grassC
H4-C
[g
ha
-1 d
-1]
-15
-10
-5
0
5
Jun 13 Oct 13 Feb 14 Jun 14 Oct 14 Feb 15 Jun 15
Te
mp
era
ture
[C
°]
5
10
15
20
25
30
WF
PS
[%
]
0
20
40
60
80
100TemperatureWFPS
a
b
c
Figure 6-4 Two years (June 2013 to June 2015) of N2O (a) and CH4 (b) fluxes
from the turf grass soil with supporting environmental parameters (c) mean daily
temperature, water filled pore space (WFPS) and fertilization events (↓).
136
6.4.5 Non-CO2 global warming potential
The non-CO2 GWP of the forest averaged -83.6 ± 15 kg CO2-e ha-1
y-1
, with
minor inter-annual variation between the two years (Table 6-3). The pasture on the
other hand, changed from a GHG source in year one to a sink in year two due to a
slight decrease in N2O emissions of 0.01 kg N2O-N ha-1
y-1
, averaging a minor
source over two years of 12.1 ± 23 kg CO2-e ha-1
y-1
. The non-CO2 GWP of the turf
was 12 times higher in year one compared to year two (p < 0.05), decreasing from
over 450 to less than 40 kg CO2-e ha-1
y-1
, resulting in a two-year average of 245.1 ±
53.2 kg CO2-e ha-1
y-1
. Despite this the turf was still the largest non-CO2 GWP
source in year two, though was not significantly different to the pasture (p > 0.05).
On a two years average, land use change from native forest to pasture increased
the non-CO2 GWP by 96 kg CO2-e ha-1
y-1
, though this was not statistically
significant (p > 0.05). Peri-urban turf grass lawn however, increased the non-CO2
GWP by an average of 329 kg CO2-e ha-1
y-1
and 233 kg CO2-e ha-1
y-1
compared to
the native forest and rural pasture, respectively (p < 0.01, Figure 6-5). Most of this
was associated with the 2-month turf grass establishment phase, which increased
N2O emissions 12 fold and reduced CH4 uptake by 30% compared to the forest.
Inter-annual changes in N2O emissions accounted for approximately 90 % of the
non-CO2 GWP.
Soil N2O and CH4 fluxes from urbanization related land use change (Paper 3)
137
GWP
Forest Pasture Turf grass
GWP
CO
2-e
[kg
ha
-1 y
-1]
-200
0
200
400
600
800
N2O year 1
N2O year 2
N2O avg
CH4 year 1
CH4 year 2
CH4 avg
GWP
Figure 6-5 Annual and the inter-annual averaged N2O and CH4 fluxes converted
to CO2-e (kg ha-1
y-1
) for the forest, pasture and turf grass land uses. For all land uses
N2O was an annual emission source (above the line) and CH4 an annual sink (below
the line). Combined global warming potential (GWP) was calculated by summing the
N2O and CH4 in CO2-e.
138
6.5 Discussion
The soil GHG flux measurements presented here provides some of the first
estimates of non-CO2 GWP for peri-urban environments. Overall, the high
infiltration rate of the sandy topsoil limited the occurrence of saturated soil
conditions, reducing the potential for both N2O and CH4 emissions while favouring
CH4 uptake, though the clay-textured subsoil potentially created saturated conditions
favourable for CH4 production and denitrification at depth. The relatively low soil
organic matter content and across all land use types and high NO3- leaching capacity
of the soil may further limit N2O production by soil microbes due to limited available
C and N substrates (Giles et al., 2012), though there is potential for substantial losses
in the turf shortly after establishment.
6.5.1 N2O fluxes
Daily N2O emissions across all land use types in this study were correlated with
WFPS rather than temperature, while N2O emissions from temperate forests are
commonly reported to be more affected by temperature (Butterbach-Bahl and Kiese,
2005; Fest et al., 2009).
The dry sclerophyll forest at SERF was only a minor source of N2O and had the
lowest daily and inter-annual flux variations of the three land uses. Daily N2O
emissions from the SERF forest were (0.2 g N2O-N ha-1
d-1
) slightly lower than
temperate dry sclerophyll forests with similar soil type (0.8 g N2O-N ha-1
d-1
, Fest et
al. (2015a) and substantially less than subtropical rainforests emissions where 1.3 g
N2O-N ha-1
d-1
has been reported (Rowlings et al., 2012b). Annual N2O emissions
from the SERF forest were more than 5 and 13 times lower than subtropical
(Rowlings et al., 2012b) and tropical (Werner et al., 2007) rainforests respectively.
The overall lower N2O emissions from the sclerophyll forest were in part limited by
the low available NO3- content (0.0 – 3.9 kg N ha
-1) typical of these sandy eucalypt
forest soils (Fest et al., 2009; Livesley et al., 2009) compared to the > 20 kg N ha-1
reported from rainforests (Rowlings et al., 2012a). These findings support the
general hypothesis that these dry forests with C:N ratios > 20 are minor contributors
to the global N2O budget (Page et al., 2011; Fest et al., 2015a) due to the limited
Soil N2O and CH4 fluxes from urbanization related land use change (Paper 3)
139
supply of N substrates for nitrification and denitrification processes (Fest et al.,
2009).
Overall the annual N2O losses from the subtropical pasture were low (0.4 g N2O-
N ha-1
day-1
) and comparable to the forest. Emissions were 3-12 times lower than
other extensively grazed tropical and subtropical pastures where N2O emissions
between 1.2 – 4.8 g N2O-N ha-1
day-1
(Neill et al., 2005; Grover et al., 2012;
Rowlings et al., 2015) have been reported, and well below the 4.3 – 9.3 g N from
legume-based temperate pastures (Eckard et al., 2003; Livesley et al., 2009). The
lack of substantial N inputs into tropical pastures either as fertilizer N or through
legume fixation together with the low soil organic N content typically limits
substrate availability and therefore potential N2O emissions in these systems. The
overall lower N2O emissions from the SERF pasture can also be attributed to the soil
only being at or near saturation for only short periods after large rain events due to
the rapid drainage of the sandy topsoil and high pasture evapotranspiration.
However, the occurrence of short-term CH4 production suggests extended water
logging within microsites for some periods at least, and suggests that the highly
dynamic but substantial emission peaks recorded (up to 38 g N2O-N ha-1
day-1
)
following the large rain events may have been substrate as well as water limited.
Interestingly though, highest N2O peaks did not correlate with highest CH4
emissions, indicating either substantial NO3- leaching (van Delden et al., 2016b) or
complete denitrification to N2.
Emissions of N2O from the turf grass were dominated by the initial six weeks of
the experiment when N from the applied basal fertilizer was in excess of root
demand from the newly established turf (van Delden et al., 2016a). This most likely
occurred within microsites at the interface of the freshly laid turf roll, full of labile C
as a result of fine root death from the harvest and transport processes, and the
granular fertiliser applied to the existing soil surface. Despite the application of an
additional 200 kg ha-1
of top-dressed N fertilizer emissions continued to decline over
the remainder of the experiment averaging just 0.6 g N ha-1
day-1
over the second
year as the turf root system rapidly developed, increasing N uptake and reducing
both available inorganic N and soil moisture (van Delden et al., 2016b). As such,
annual N2O emissions from the SERF turf grass were up to 15 times lower than
140
temperate peri-urban environments where up to 3 kg N2O-N ha-1
y-1
has been
reported (Groffman et al., 2009), mostly due to higher fertilizer application rates and
less well drained soils. Overall, the low N2O emissions from these well drained soils,
when compared to temperate zones, suggest the tightly coupled N cycle limits excess
N in the profile and subsequent losses to the environment (Xu et al., 2013).
6.5.2 CH4 fluxes
All land uses were net CH4 sinks over both years. While the annual CH4 uptake
decreased in the forest and turf grass in year two due to the higher annual rainfall,
uptake in the pasture remained constant over both years. Soil CH4 uptake decreased
in all land use types when WFPS increased or decreased to very high and low levels.
CH4 uptake dynamics are driven by decreased methanotrophic and/or increased
methanogenic activity when soil oxygen or moisture reaches extreme levels (Smith et
al., 2000; Kaye et al., 2004). However the dry sclerophyll forest was the strongest
CH4 sink, despite having significantly lower WFPS compared to the pasture and turf
grass with extensive periods where WFPS was very low. This strong CH4 uptake
even with extremely low soil moisture may indicate less simultaneous occurring CH4
production in the forest soil due the exposed mineral soil without thick grass thatch.
The upper 10 cm of these sandy forest topsoils have the highest CH4 uptake potential
(Butterbach-Bahl and Papen, 2002; Fest et al., 2015b), compared to the dense root
zone with fine roots in the pasture and turf grass soils. The pasture was the only land
use, which emitted significant amounts of CH4 after heavy rain events, possible due
to accumulation of water and C substrates at the top of the clay subsoil, while the turf
grass limited the water infiltration into the subsoil by increases plant uptake due to
higher root density.
Daily CH4 uptake of the SERF dry sclerophyll forest was (-7.5 g CH4-C ha-1
d-1
)
comparable to temperate dry sclerophyll forest soils (≈ -9 g CH4-C ha-1
d-1
, Fest et al.
(2015a)) and 25 % higher than native forests in Western Australia (-5 g CH4-C ha-1
d-1
(Livesley et al., 2009). Annual CH4 uptake of the SERF forest however, was with
-2.7 kg CH4-C ha-1
y-1
averaged over the two years, 27 % less than subtropical
rainforest soils where up to -3.7 kg CH4-C ha-1
y-1
, which was mostly due to higher
methanothropic activity based on higher soil organic matter contents (Rowlings et
al., 2012b).
Soil N2O and CH4 fluxes from urbanization related land use change (Paper 3)
141
The higher topsoil BD and consequently WFPS in the pasture compared to forest
and turf grass saturated the soil and changed the pasture from a CH4 sink to a source
for short periods of time. These WFPS driven sink to source dynamics are in line
with observations from a similar texture-contrast soil under clover-grass pasture in
Western Australian (Livesley et al., 2009) where fluxes switched between a minor
sink (-0.7 g CH4-C ha-1
d-1
) to a minor source (1.2 g CH4-C ha-1
d-1
) following
rainfall. However, uptake from the SERF pasture was 3 times higher (-2.1 g CH4-C
ha-1
d-1
) than in Western Australia. On the other hand, the maximum CH4 emissions
in the SERF pasture were also 25 times higher than in Western Australia, which is
comparable to 28 g CH4-C ha-1
d-1
from tropical pastures (Grover et al., 2012). This
suggests an important impact of climate conditions on CH4 uptake with comparable
soil and land use type.
The CH4 uptake in turf grass soil was depleted when soil moisture conditions
were extremely high and extremely low but never became a clear CH4 source like the
pasture. This consistent CH4 uptake in the SERF turf grass represents with -1.6 kg
CH4-C ha-1
y-1
, within the -3 kg CH4-C ha-1
y-1
(Kaye et al., 2004) to negligible range
of CH4 uptake (Groffman and Pouyat, 2009) reported from other turf grass systems.
6.5.3 Inter-annual drivers of GHG fluxes
Despite 400 mm more rainfall and 3 times as many rain events in the second year
of the study, there was no corresponding increase in the average WFPS which was
comparable across years within each land use type. The consistent inter-annual soil
moisture conditions are most likely due to the limited water-holding capacity of the
soil and the higher temperatures and therefore evapotranspiration in year two
offsetting the additional rainfall. As WFPS was identified as the main abiotic driver
of the GHG fluxes, only minor inter-annual variations in fluxes occurred between the
experimental years. This finding is in contrast to the few other multi-year GHG
datasets where annual estimates of N2O have varied by up to 84%. Rowlings et al.
2015 for instance, reported a 46% increase in emissions with a halving of summer
rainfall from a humid subtropical pasture on an alluvial loam due to potential losses
of N2 as opposed to N2O. Alternatively, in temperate zones, however, clear
correlations can be found between higher rainfall and increased GHG emissions
142
(Groffman et al., 2009). In this study although WPFS was the core controller of the
timing of N2O emission events, the magnitude of losses were restricted by limited
substrate availability.
6.5.4 Influence of land use change on GWP
The sandy Chromosol at the study site had a higher BD in the first 10 cm of the
grazed pasture compared to the native forest, most likely a result of compaction from
cattle and historic management activities. The turf grass soil had a lower topsoil BD
due to the soil cultivation during establishment in combination with the higher root
density of the turf grass. These BD alterations illustrate how land use change
processes can alter soil physical conditions, significantly influencing GHG
measurements by the changed diffusivity of the soil (Veldkamp, 1994).
The high CH4 uptake in the subtropical dry sclerophyll forest soil at SERF
resulted in a net non-CO2 GWP sink of -0.08 t CO2-e ha-1
y-1
. With increasing
urbanization of the surrounding environments however, the strength of this GHG
sink might be reduced in the future as some research suggests that CH4 uptake
decreases from forests soils along a rural to urban gradient (Groffman and Pouyat,
2009). This reduced sink of forest soils is the result of changes in atmospheric GHG
concentrations, e.g. higher CO2 levels within urban environments (Pataki et al.,
2007). Land use change from native forest to rural pasture increased the non-CO2
GWP from a GHG sink to a weak source by reducing CH4 uptake rather than
increasing N2O emissions. This was mainly caused by the nearly doubled average
WFPS in the pasture soil, the main abiotic environmental parameter influencing
GHG soil-atmosphere gas exchange in this study. The introduction of turf grass with
the associated additional fertilizer and irrigation inputs, not only decreased CH4
uptake but also significantly increased N2O emissions by 12 times the annual
average, resulting in 0.42 t CO2-e ha-1
in only two months after establishment.
The inter-annual flux variation observed in the turf were a result of an increased
emission peak following the turf grass establishment rather than climate variations.
This is supported by the minor interannual variation observed in the forest and
pasture. This substantial emission peak however, did not reoccur following
subsequent fertilization events suggesting subtropical turf lawns can potentially be
managed to have little environmental impact when well-established. This outcome
Soil N2O and CH4 fluxes from urbanization related land use change (Paper 3)
143
contradicts findings from temperate turf grass systems, where GHG emissions have
been reported comparable to intensive agriculture (Kaye et al., 2004). While
balancing the intensive GHG emissions from fertilizer input, some research however
suggests a strong C sequestration potential of turf grass soils (Conant et al., 2005;
Lorenz and Lal, 2009; Wang et al., 2014). The sclerophyll forest had nearly half the
annual WFPS average than the pasture and turf grass despite receiving the same
amount of rainfall, indicating a higher water uptake by the dry sclerophyll eucalypt
trees. Although the average WFPS did not differ significantly between pasture and
turf grass, short-term WFPS peaks after heavy rain events were lower in the turf
grass soil. This buffering effect is most likely the result of higher evapotranspiration
due to the higher root density of the turf grass (Barton et al., 2009). Temperature
showed only a minor effect on the GHG fluxes measured over the two years, with
stronger significance for N2O than CH4 fluxes.
The results from this study may indicate a great potential to reduce management
inputs in form of fertilization and irrigation shortly after turf grass establishment in
subtropical peri-urban environments while maintaining and may even contribute to
climate change mitigation with C sequestration in the long-term.
6.5.5 Outlook
One tenth of the current Australian GHG inventory of approximately 525,202 Gt
CO2-e (AGEIS, 2015) is accounted for by emissions due to land use change and
management (Hatfield-Dodds et al., 2015). With 17,320 ha of turf grass being
established in Australia annually and nearly half distributed in tropical and
subtropical Queensland (ABS, 2012; Turf Australia, 2012), the non-CO2 GHG
emissions for the establishment alone could be estimated from the non-CO2 GWP
presented here of up to 7,326 t CO2-e y-1
. However, Australia has a great potential to
reduce national emissions from currently four times the global average (Hatfield-
Dodds et al., 2015), by optimizing fertilizer use efficiency which therefore reduces
fertilizer inputs and subsequent GHG emissions while supporting C sequestration
strategies into the biomass of natural and peri-urban ecosystems.
144
6.5.6 Acknowledgements
This study was undertaken at the Samford Ecological Research Facility (SERF)
one of the Supersites in the Terrestrial Ecosystem Research Network (TERN). The
study was supported by the Central Analytical Research Facility (CARF) operated by
the Institute for Future Environments (IFE) of the Queensland University of
Technology (QUT). The data set “Greenhouse gas emissions from peri-urban land
use at SERF, SEQ. 2013-2015” can be found online at the N2O network under
http://www.N2O.net.au/knb/metacat/vandelden.3.3/html.
Land use change implications on the soil C sequestration potential of peri-urban environments (Paper4)
145
Chapter 7: Land use change implications on
the soil C sequestration potential
of peri-urban environments
(Paper 4)
7.1 Abstract
Urban populations worldwide have exceeded rural populations and will account
for most future population growth. This populations growth and rural to urban
migration increasingly result in extensive urban sprawl, which causes rapid land use
change from native forests, rural pastures and commercially focused agriculture into
smaller, residential properties, i.e. peri-urban environments. Soil biogeochemical
carbon (C) and nitrogen (N) cycling has the potential to contribute to climate change
by emitting greenhouse gases (GHG) to the atmosphere. On the other hand, soils can
remove GHGs from the atmosphere by storing C and N in soil organic matter
(SOM), i.e. C sequestration. The C sequestration potential of peri-urban
environments is often neglected due to the fragmented distribution of these land
areas and partially sealed soils. This omission, however, potentially represents an
underestimate of the global terrestrial C pool as unsealed peri-urban soil has the
potential to increased C sequestration by increased ecosystem productivity due to
higher management inputs such as fertilization and irrigation. This study identified
the long-term effect of land use change on the soil C and N pools in a peri-urban
environment to estimate the C sequestration potential of subtropical turf grass
systems when compared to forest and pasture. It was hypothesized that due to the
intensive fertilization and irrigation management practices of turf grass systems the
higher ecosystem productivity would result in a higher soil C sequestration than in
the forest and pasture land use. A soil survey was conducted from 18 sites in the
dominant land uses of Samford Valley, Australia, an area of rapid peri-urbanization.
The predominately sandy soils were analysed for total C, which was fractionated
according to turnover velocity into the active, slow and resistant soil C pools. Total
soil C and N varied widely across sampling sites from 17.3 to 46.6 t C ha-1
and 1.0 to
146
3.4 t N ha-1
respectively, with the widest C and N ranges in pasture soils. The turf
grassland use however, showed no significant difference in total C when compared to
forest and pasture. Overall, the slow soil C pool was dominant across all land use
types with 1.1 % on average, which suggests soil C storage in the long term. This
study proves that peri-urban environments can contribute substantially to the global
terrestrial C pool but the intense management of turf grass does not improve C
sequestration in the subtropical Samford Valley.
7.2 Introduction
Urban populations worldwide have not only exceeded rural populations but are
also predicted to account for all future population growth (United Nations 2014).
This populations growth and rural to urban migration increasingly result in extensive
urban sprawl, which causes rapid land use change from native forests, rural pastures
and commercially focused agriculture into smaller, residential properties, i.e. peri-
urban environments. This transition from rural to urban environments, i.e. peri-urban,
is associated with construction processes and increasingly the extensive
establishment of turf grass for residential backyards, public parks and sportsgrounds,
and golf courses (IPCC 2006). How these land use changes influence ecosystem
dynamics in terrestrial biogeochemical cycling is only beginning to be understood.
The biogeochemical carbon (C) and nitrogen (N) cycles play important roles in
climate change mitigation by immobilizing C from the atmosphere in vegetation and
soil organic matter (SOM), i.e. C sequestration, and therefore reduce the radiative
forcing of greenhouse gases (GHG) in the atmosphere (IPCC 2013). Soil organic
matter comprises 58 % of soil organic carbon (SOC) on average (Jain et al. 1997) as
well as a combination of macro- and micronutrients, and tends to increase as CEC
increases. The majority of the terrestrial 2500 Pg soil C pool is beside inorganic C
with 1550 Pg mainly in SOC form (Lal 2004a), and is estimated to increase through
natural C sequestration by approximately 24 kg C ha-1
y-1
on average, with over 100
kg C ha-1
y-1
in boreal or temperate forests (Schlesinger 1990). Older soils, such as
found in Australia, however are considered to have a lower C sequestration potential
than younger soils (< 3,000 years) and could even become overall net C sources with
increasing global warming (Schlesinger 1990).
Land use change implications on the soil C sequestration potential of peri-urban environments (Paper 4)
147
Carbon sequestration represents an increase in the total soil content (Lal 2004b),
while the turnover velocity is crucial for assessing the long-term storage potential of
the sequestered C. Partitioning of soil C and N into active, slow and resistant
fractions is therefore needed to evaluate the long-term impact of the land use on C
sequestration. The active C fraction drives the C cycle’s fastest turnover (weeks-
months) in the form of soil respiration and microbial available C. The slow C
fraction has a slow turnover (decades) and is the transitional stage to the resistant C
fraction, which turnover is approaching a steady state (centuries) in the soil. The
transformation from easily microbial available C into microbial resistant SOM is
strongly correlated to establishment of macro-and micro-aggregates of the sand, silt
and clay texture of the soil.
The consequences of land use change from native to crop cultivation have been
identified by several studies. These include a loss in soil quality (structure and
nutrient losses) and quantity (erosion), increased GHG emissions, and reduced
potential for soil C sequestration (Grover et al. 2012; Livesley et al. 2009). On the
other hand, changing soils from seasonally cultivated to perennial land use such as
residential ecosystems has shown the potential in temperate climates to improve
critical ecosystem services by (i) providing stormwater treatment, (ii) acting as a sink
for atmospheric N and (iii) sequestering C (Lal 2004b; Golubiewski 2006; Raciti et
al. 2011a).
Managed land use types within peri-urban environments influence the
biogeochemical C and N cycling with frequent fertilization, irrigation and biomass
removal through mowing. While management practices such as fertilization and
irrigation increases biomass production and subsequently increases SOM, the same
practices potentially increase gaseous losses such as CO2 respiration and N2O
emissions (Conant et al. 2005; Lorenz and Lal 2009; Wang et al. 2014), which is a
GHG nearly 300 times more potent than CO2 (IPCC 2013). Considering the
substantial C sequestration potential of managed peri-urban environments
(Golubiewski 2006; Raciti et al. 2011a), the global C pool may currently be
underestimated due to the exclusion of urban and peri-urban soils.
While the extent of peri-urban environments are difficult to quantify due to their
fragmented distribution, collectively peri-urban turf grass occupies over 15 Mha in
148
the USA alone, three times more than any other irrigated crop in the country (Milesi
et al. 2005). These managed peri-urban grasslands imply substantial alterations in
biogeochemical C and N cycling in temperate climates, resulting in significantly
increasing GHG emissions mostly due to intense fertilizer use (Kaye et al. 2004;
Conant et al. 2005). Despite the subtropical climate affects 3.26 Mha in Australia
alone (Department of Trade 1982), as well as large areas on the North and South
American continent, Africa and Asia, it remains unknown if urbanization related land
use change into turf grass systems affects biogeochemical C and N cycling in this
warm and humid climate.
The humid subtropical climatic zone of South East Queensland (SEQ), Australia,
is dominated by extreme annual and inter-annual variations in rainfall, with heavy
rain events and rapid soil moisture changes. Combined with year-round high soil
temperatures, these soil conditions are favourable for microbial activity. Brisbane in
SEQ currently has an annual population growth rate of 1.7 % and is considered the
most biologically diverse city in Australia with the most extensive area of urban
sprawl (ABARES 2010; Commonwealth of Australia 2013). Tropical and subtropical
ecosystems may have the potential for improved N cycling, minimizing N losses
within biogeochemical cycling in natural and managed ecosystems (Xu et al. 2013;
van Delden et al. 2016b). This tightly coupled N turnover with rapid plant uptake
may potentially support C sequestration in subtropical peri-urban environments by
increased biomass production.
This study quantifies the total C and N pool from a subtropical peri-urban
environment and identifies the C sequestration potential of turf grass when converted
from forest and pasture to identify the long-term implications of urbanization related
land use change on the C and N cycle. This study hypothesizes that the subtropical
climate favours soil C sequestration though high rates of primary production driven
by favourable temperatures throughout the year and humid summers. Land use
change from unmanaged forest and pasture into fertilized peri-urban turf grass
systems will therefore show an even higher C sequestration potential than indicated
from temperate climates.
Land use change implications on the soil C sequestration potential of peri-urban environments (Paper 4)
149
7.3 Material and Methods
7.3.1 Site description
This study was conducted in Samford Valley, 20 km from Brisbane, Australia
(152° 52' 37.3" E, 27° 23' 22.211" S). The Samford Valley was mainly cleared of
native vegetation in the early 1900s, and developed in the 1960s for dairy and beef
cattle as well as intensive agriculture including banana and pineapple. Samford’s
population density has increased rapidly, almost doubling from 1996 – 2006, causing
land use change from predominately rural to peri-urban residential properties
(Moreton Bay Regional Council 2011). The region is influenced by a humid
subtropical climate with seasonal summer rain (December to February),
approximately 1110 mm long term annual mean rainfall (BOM 2015). The mean
annual minimum and maximum temperatures are 13 °C and 25.6 °C respectively
(BOM 2015). The Samford Valley floor has granite as the parent material with soils
characterised as Chromosols and Kurosols based on the Australian soil classification
(Isbell 2002) and Planosols based on the World Reference Base (WRB 2015). These
soil types are characterized by a strong texture contrast between the A and B horizon.
However, construction processes associated with urbanization mean possible
intermediate layers can be found across the Samford Valley, ranging widely in soil
texture and properties.
7.3.2 Experimental design
This study compared the C and N pool of three well-established land use types in
the Samford Valley of native and secondary forest, grazed and ungrazed pasture, and
turf grass lawn under low and high management intensity. All land use types had
been established for at least one decade to ensure this analysis was of long-term land
use effects. Overall, 18 sites were sampled for soil and plant material in November
2014. Six sites were sampled per land use type, which included three sites each under
private and public management to establish a representative average for the Valley.
The sampling sites under private management include SERF1, SERF2, Dy, D1, D2,
A, R1, R2, and CSIRO; and under public management ELP, URR, MRDR, SPS, KR,
JMP1, JMP2 and MR (Figure 7-1).
150
The native forest type for the Samford Valley is a dry sclerophyll eucalypt forest,
which is typical for the region of sandy soils. The secondary forest included
Flindersia schottiana, Podocarrpus eletus and Grevillea robusta. The forest under
both private and public use is mostly unmanaged. Pasture species included Chloris
gayana, Setaria, and Nardus stricta with white clover (Trifolium repens). Pasture
under private and public use is infrequently mowed or grazed, mostly unfertilized
and not irrigated. The common practice to establish turf grass as lawn in this region
includes removal of the dense pasture sward and surface roots to expose the topsoil,
which is then mixed during construction processes of the peri-urban environment.
Mainly Blue Couch turf (Digitaria didactyla) and Buffalo turf (Bouteloua
dactyloides) are then laid with up to 100 kg N ha-1
fertilization applied to the root
layer. Turf suppliers in this region recommend fertilization at 300 kg N ha-1
y-1
for
regular use and up to 500 kg N ha-1
y-1
for high end users such as golf courses and
sports grounds.
Information on the particular management intensity for selected public sites was
provided by the Moreton Bay Regional Council and included fertilization, frequent
mowing and irrigation. Nine public sites were randomly selected form the Council’s
provided 44 public sites with forest, pasture and turf grass land use, wherefrom 35
were suitable for their land use age, representative size and accessibility. Access to
private residential sites was limited and selected by availability, provided by
residents of the Samford community and the Samford Ecological Research Facility
(SERF). All sites were located around Samford Village in the North-East of the
Valley, to meet main requirement for age of land use.
7.3.3 Sampling
Intact soil cores were taken from each sampling site to 1 m depth with a hydraulic
soil auger to determine the soil type according to Isbell (2002). Separate soil samples
were taken at 0-10 cm and 10-20 cm soil depths with a hand auger from four
subsamples, replicated three times across the sampling site. All soil samples were
kept refrigerated until analysed. Additionally, three replicated bulk density (BD)
samples were taken for the two sample depths according to Carter and Gregorich
(2007). Aboveground plant samples were taken from each site and air-dried, leave
Land use change implications on the soil C sequestration potential of peri-urban environments (Paper 4)
151
and stem material from the forest and grass and root material from the pasture and
turf grass.
7.3.4 Sample preparation and analysis
Soil from the intact cores were air-dried and sieved to 2 mm for particle size
analysis per horizon. The soil samples were extracted for mineral N (NH4+ and NO3
-)
using a 1:5 KCl (2 M) solution with 20 g of soil. The extract was analysed for NH4+
and NO3- with an AQ2+ discrete analyser (SEAL Analytical WI, USA). The
remaining soil samples were air-dried and sieved to 2 mm for further analysis.
Gravimetric water content (GWC), pH and electrical conductivity (EC) were
analysed according to Carter and Gregorich (2007). Total C (CT) and N (NT) content
of air-dried soil and plant samples were determined by dry combustion (CNS-2000,
LECO Corporation, St. Joseph, MI, USA) from ground samples. The laser sizing
technique (Mastersizer 3000, Malvern Instruments Ltd, UK) was used to determine
particle size texture for the upper horizon per sampling site and clay content
corrected based on Konert and Vandenberghe (1997). However, after comparison
with the pipette method (Carter and Gregorich 2007) for two soils with contrasting
low and high clay content, the correction was furthermore modified to <5 µm instead
of <8 µm grain size output for this particular laser analyser.
7.3.5 C fractionation
The C fractionation scheme used was based on a simplified version of the
CENTURY model pools (Parton 1996) based on the concepts of Skjemstad et al.
(2004) and Baldock et al. (2012). Soil organic C was partitioned into an active (CA),
slow (CS) and resistant fraction (CR). This research analysed the CA and CR fractions
without physical separation of the soil according to the particle size of sand, silt, and
clay content, as the soil texture of the Samford Valley is predominately sand (particle
size 0.2 - 2 mm) lacking micro-aggregation. The CS fraction was then calculated
from CT according to equation 1. All analyses were conducted on three laboratory
replicates, per field sample replicate.
CS = CT – (CA + CR) (Equation 1)
152
The CA fraction was analysed according to the protocol of Hbirkou et al. (2011),
based on the original method of Blair et al. (1995) and modified after Tirol-Padre and
Ladha (2004). Ground soil of 15 mg C equivalent was shaken for 24 h at 12 rpm on a
15 cm radius tumbler with 25 ml of 33 mM KMnO4 in centrifuge tubes covered with
aluminium foil to avoid UV interference. The tubes were than centrifuged for 5
minutes at approximately 5000 rpm and the supernatant diluted 1:25 with DI water.
The absorbance was measured with a split beam spectrophotometer at 565 nm,
calibrated with standards of 1.0, 1.4, 2.0 ml 33 mM KMnO4. Each mM KMnO4
consumed equals 0.75 mM or 9 mg of CA.
The CR fraction was analysed as established by Siregar et al. (2005) and modified
according to Thomsen et al. (2009. Briefly, 5 g of bulk soil was oxidized in
centrifuge tubes for 18 h with 45 ml of 6 % NaOCl, and adjusted to pH 8 with
concentrated HCl. The tubes were centrifuged at 1000 g for 15 min, decanted and the
supernatant discarded. The soil was then washed with deionised water, centrifuged
and decanted. The whole oxidation procedure was repeated three times and washed
twice at the end of the third oxidation procedure. The oxidized soil was then dried at
40 °C and ground for dry combustion analysis of CR.
7.3.6 Statistical analysis
Correlations of environmental parameters and the C sequestration potential were
identified with a Spearman’s rho correlation using SPSS Statistics 21.0 (IBM Corp.,
Armonk, NY). ANCOVA analysis with the main correlated environmental
parameters as covariance was used to identify differences between land use types,
when the significance value (p) was < 0.05.
Land use change implications on the soil C sequestration potential of peri-urban environments (Paper 4)
153
Figure 7-1 Selected private and public sites in Samford Valley, Queensland,
Australia, of the land use types forest (D2, JMP2, R2, SERF2, MR, BPP), pasture
(D1, JMP1, URR, KR, CSIRO, Dy), and turf grass lawn (A, MRDR, ELP, SPS, R1,
SERF1).
7.4 Results
7.4.1 Environmental conditions
During 2014, the Samford Valley received 677 mm of rain, below the long-term
average of 1110 mm (BOM 2015). The topsoil BD ranged from 1.1 to 1.6 g cm-3
across all sites with the lowest BD on average in the forest, and highest in the pasture
soil (Table 7-1). Bulk density in the 10-20 cm soil depth was on average 10 % higher
than the 0-10 soil depth. Soil pH and EC ranged from 6.4 to 7.2 and 38 to 147 μS
respectively, with the highest pH and EC values in the turf grass soil. Clay content
ranged from 7 to 18 % with the lowest and highest contents represented in all land
use types. Total plant C and N ranged from 38 to 50 % and from 0.7 to 3.1 %
respectively, resulting in a higher C/N ratio in the forest leaves over the pasture and
turf grass.
154
Table 7-1 Site characteristics for the topsoil (0-10 cm) averaged per land use type
with standard error
BD (g cm
-3)
pH EC (μS)
Clay (%)
Plant C (%)
Plant N (%)
Plant C/N
Forest 1.2 ± 0.05 6.6 ± 0.09 57 ± 5.9 12.0 ± 2.0 47 ± 0.7 1.0 ± 0.02 50
Pasture 1.3 ± 0.08 6.7 ± 0.07 60 ± 8.8 11.9 ± 1.6 41 ± 0.7 1.2 ± 0.1 23
Turf grass 1.4 ± 0.02 6.9 ± 0.12 86 ± 17.5 10.2 ± 1.2 41 ± 0.6 1.8 ± 0.4 34
7.4.2 Carbon
Total C ranged widely from 0.8 to 3.8 % and averaging 2.6 % in 10 cm topsoil
across all sites with the highest C content in pasture soils (Figure 7-2, Table 7-2).
The majority of the CT was in the CS fraction for forest and turf grass, while CA was
the dominant fraction in the pasture soil. The CA fraction ranged from 20 to 77 % of
the CT content across all sites with both the lowest and highest percentage in the
pasture soils. The CS fraction ranged from 5 to 57 % of the CT content across all sites
with the lowest percentage in the pasture, and highest in the turf grass land use. The
CR fraction ranged from 18 to 42 % of the CT content across all sites with the lowest
percentage in turf grass and highest in pasture land use. Overall, CT and the C
fractions were on average 22 % lower in the 10-20 cm soil depth compared to the 0-
10 cm depth.
7.4.3 Nitrogen
Total N ranged between sites from 0.1 to 0.3 %, averaging 0.2 % in 10 cm topsoil
with the highest and lowest contents in the pasture soils (Table 7-4). Mineral N in the
form of NO3- accounted for 1 % of the NT content of the fertilized turf grass land use
at a sports ground which was the highest on average across all land use types. This
compared to 0.1 % on average of the NT across the pasture land use. The NH4+
content ranges from 0.4 to 2 % of the NT content across all sites with the lowest
percentage on average in the forest and highest percentage in turf grass land use.
Overall, NT and mineral N were on average 24 % lower in the 10-20 cm soil depth
compared to the 0-10 cm depth.
Land use change implications on the soil C sequestration potential of peri-urban environments (Paper 4)
155
0-10 cm
SO
C [
t C
ha
-1]
0
10
20
30
40
CA
CS
CR
10-20 cm
Forest Pasture Turf grass
A B
Forest Pasture Turf grass
Figure 7-2 Soil organic C average of in the form of active C (CA), slow C (CS) and
resistant C (CR) per land use type with standard error for 0-10 cm soil depth (A) and
10-20 cm soil depth (B)
Table 7-2 Soil C contents averaged per land use type with standard errors in total
(CT) and the three C fractions of active (CA), slow (CS) and resistant (CR); total N
(NT), mineral N (Nmin) and soil C/N ratio
Depth (cm)
CT (t ha
-1)
CA (t ha
-1)
CS (t ha
-1)
CR (t ha
-1)
NT (t ha
-1)
Nmin (kg ha
-1)
Soil C/N
Forest 0-10 30.9 ± 4.9 8.9a ± 1.6 13.8
a ± 2.0 8.3
a ± 1.6 2.1 ± 0.3 26.5 ± 3.0 13
10-20 26.5 ± 4.3 7.2a ± 1.3 12.0
a ± 1.9 7.2
a ± 1.4 1.9 ± 0.3 23.0 ± 5.3 15
Pasture 0-10 37.4 ± 6.1 12.3a ± 1.3 13.5
ab ± 3.6 11.6
a ± 2.7 2.7 ± 0.4 24.9 ± 3.3 12
10-20 30.5 ± 6.0 8.6a ± 1.1 11.9
ab ± 2.9 10.0
a ± 2.7 2.4 ± 0.4 17.6 ± 1.1 12
Turf grass 0-10 30.4 ± 2.7 9.3a ± 1.5 14.5
ac ± 0.9 6.5
b ± 0.7 2.0 ± 0.3 35.1 ± 8.9 13
10-20 19.7 ± 3.1 5.4a ± 1.2 9.9
ac ± 1.5 4.4
b ± 0.7 1.6 ± 0.3 25.2 ± 5.4 12
abc Different letters indicate significant differences between land use types per column with p < 0.05
7.4.4 Environmental influence on C fractions
Total C and N content of the soils were significantly correlated to the mass of the
CA, CS, and CR fractions (Table 7-3). The CS and CR fractions were also correlated to
the clay content of the soil, while clay content showed no significant effects on the
CA fraction. The CA fraction, on the other hand, was significantly correlated to the
soil EC. Mineral N and pH had no correlation with any of the C fractions.
The total C and N content, texture, BD, pH and EC of sample site SERF2 (Table
7-4) is representative of the native forest of Samford Valley, and provides the
baseline for assessing land use change in response to agricultural development and
urbanization. Approximately 34 ha of native dry sclerophyll forest had been
surveyed previously at three different locations at Samford
(www.supersites.net.au/knb) in addition to the SERF2 site to with comparable total
156
C, N and clay contents in the topsoil. Overall, the land use type, did not significantly
influence the mass of the C fractions (p = 0.053). Comparing every land use type to
its original land use revealed some significance. Land use change from forest to turf
grass significantly decreased the CR fraction (p < 0.05) but not CA or CS in both soil
depths. Land use change from pasture to turf grass significantly increased the CS
fraction in both soil depths (p < 0.05) without affecting CA or CR (p > 0.05). Land
use change from forest to pasture did not significantly influence the CA, CS, or CR
fraction in both soil depths.
Table 7-3 Spearman’s rho correlations of the active (CA), slow (CS) and resistant
(CR) C fractions with each other and their soil parameters total C (CT) and N (NT),
mineral N (Nmin), pH, electric conductivity (EC) and clay content for the upper 10 cm
topsoil
CA CS CR CT NT Nmin pH EC Clay
CA - 0.48* 0.83** 0.87** 0.52* 0.21 0.08 0.67** 0.45
CS - 0.64** 0.76** 0.78** 0.03 -0.12 0.38 0.48*
CR - 0.96** 0.65** -0.01 -0.13 0.56* 0.61** * Correlation significant with p < 0.05 ** Correlation significant with p < 0.01
Land use change implications on the soil C sequestration potential of peri-urban environments (Paper4)
157
Table 7-4 Site parameters for all 18 sampling sites in Samford Valley based on 4 replicated field subsamples and 3 laboratory replicates per value
Land use Site ID Management Depth (cm)
CT (%)
CA (%)
CS (%)
CR (%)
NT (%)
Nmin (% of NT)
Soil C/N BD (g cm
-3)
pH EC (μS)
Clay (%)
Plant C (%)
Plant N (%)
Plant C/N
Forest SERF2 private 0-10 1.2 0.35 0.64 0.24 0.10 12.8 13 1.4 6.5 49 6.8 50.3 0.9 55 10-20 0.8 0.20 0.38 0.17 0.07 11.6 12 1.5 6.5 49 6.8
R2 private 0-10 1.8 0.39 0.93 0.44 0.16 11.6 11 1.4 6.4 51 7.3 46.7 1.0 46 10-20 1.5 0.29 0.78 0.39 0.13 10.6 11 1.5 6.4 51 7.3
D2 private 0-10 2.0 0.70 0.82 0.50 0.16 16.7 13 1.2 7.0 45 14.8 45.5 1.0 46 10-20 1.5 0.51 0.62 0.36 0.12 9.7 13 1.4 7.0 33 16.0
MR public 0-10 2.6 0.73 1.04 0.78 0.21 12.8 12 1.1 6.5 53 16.8 46.0 0.9 51 10-20 2.2 0.59 0.92 0.69 0.19 10.6 12 1.3 6.5 53 16.8
BP public 0-10 3.8 1.25 1.55 0.97 0.28 5.5 13 1.2 6.7 84 8.9 46.0 1.0 48 10-20 2.6 0.82 1.13 0.66 0.19 6.1 14 1.4 6.7 84 8.9
JMP2 public 0-10 3.8 0.94 1.73 1.14 0.27 10.9 14 1.2 6.6 62 17.7 47.9 0.9 52 10-20 3.0 0.71 1.36 0.89 0.10 34.3 28 1.3 6.6 62 17.7
Pasture Dy private 0-10 1.5 1.12 0.07 0.28 0.12 14.6 12 1.4 6.8 43 9.1 42.8 1.1 40 10-20 1.5 0.76 0.50 0.28 0.12 9.9 13 1.6 6.8 43 9.1
D1 private 0-10 1.6 0.55 0.59 0.40 0.13 19.7 12 1.3 7.0 38 10.5 41.1 1.1 36 10-20 1.0 0.33 0.34 0.32 0.08 16.2 12 1.5 7.1 26 10.5
CSIRO private 0-10 2.9 1.21 0.90 0.84 0.26 10.2 11 1.2 6.8 93 7.6 38.3 0.8 46 10-20 1.1 0.44 0.32 0.33 0.10 9.9 11 1.5 6.8 93 7.6
URR public 0-10 3.4 0.82 1.60 0.95 0.29 4.1 12 1.6 6.7 79 16.8 42.6 1.8 24 10-20 2.5 0.55 1.17 0.75 0.22 4.9 11 1.6 6.7 79 16.8
JMP1 public 0-10 4.2 0.81 1.60 1.74 0.30 3.3 14 1.2 6.6 53 11.0 41.4 1.1 37 10-20 3.9 0.74 1.63 1.54 0.28 3.8 14 1.3 6.6 53 11.0
K public 0-10 4.2 1.32 1.55 1.32 0.32 8.9 13 1.1 6.5 55 16.2 42.3 1.3 33 10-20 2.9 0.76 1.07 1.06 0.20 8.0 15 1.2 6.5 55 21.5
Turf grass SERF1 private 0-10 1.7 0.42 0.99 0.34 0.13 13.7 14 1.4 6.5 40 7.0 41.2 1.3 31 10-20 1.0 0.24 0.60 0.20 0.08 10.1 13 1.5 6.5 40 7.0
R1 private 0-10 1.7 0.40 0.91 0.37 0.14 9.8 12 1.4 6.7 52 6.7 39.6 0.7 59 10-20 1.1 0.20 0.62 0.27 0.08 30.5 13 1.5 6.7 52 6.7
A private 0-10 2.6 0.91 1.20 0.45 0.20 16.5 13 1.4 6.7 55 9.6 42.2 1.5 28 10-20 1.9 0.66 0.91 0.37 0.15 12.1 13 1.4 6.5 42 9.6
SPS public 0-10 2.0 0.51 1.05 0.44 0.18 30.3 11 1.3 7.1 147 14.2 42.4 2.9 15 10-20 1.6 0.35 0.83 0.43 0.15 18.5 11 1.5 7.1 147 23.2
MRDR public 0-10 2.4 0.87 0.88 0.64 0.18 6.5 13 1.4 7.2 109 11.8 38.9 1.2 32 10-20 0.6 0.19 0.21 0.15 0.05 14.6 11 1.5 7.4 117 22.0
ELP public 0-10 3.0 0.99 1.37 0.64 0.25 9.7 12 1.3 7.1 113 12.0 39.8 3.1 13 10-20 1.8 0.57 0.86 0.36 0.15 9.9 12 1.5 7.1 113 12.0
158
7.5 Discussion
Annually about 17,320 ha of turf grass is established in Australia with nearly half
distributed in Queensland (ABS 2012; Turf Australia 2012). These subtropical peri-
urban environments have the potential to reduce atmospheric CO2 levels by C
sequestration into biomass of natural and managed ecosystems (Hatfield-Dodds et al.
2015). Fertilized turf grass in the temperate USA showed higher C sequestration
rates than native land use (Golubiewski 2006), but varies widely in its management
intensity. However, due to the limited access for research in residential areas and
gaps in the known land use history (Golubiewski 2006; Raciti et al. 2011a), a wide
variety of peri-urban environments is still needed for generalized C sequestration
estimations.
7.5.1 Soil C sequestration potential
The averaged total C content across all land uses in Samford Valley of 58 t C ha-1
in 20 cm topsoil presents a substantial increase in soil C compared to the native
conditions of the dry sclerophyll forest soil (SERF2) with 29 t C ha-1
20 cm-1
. This
single site of remnant vegetation (SERF2) might not be entirely representative of the
soils now under secondary forest and the fact that soil spatial heterogeneity and
topography might account for some variations in clay content. The higher total C and
N contents of the secondary forests, pastures and turf grass lawn soils therefore
indicate a substantial increase in soil C and N since the Samford Valley was cleared
of native forest over a century ago. This increase would include charcoal, which
would have been deposited on burning of the original forest (Conant et al. 2001).
Additionally, increased net primary productivity (NPP) of the high C/N products in
the pasture would increase both total C and CR (McLauchlan et al. 2006), as is the
case for the Samford Valley pastures.
If we consider the collective soil C data from the native and secondary forest sites,
there was no difference in the total amount of soil C found per area basis when
compared to the pastures. This is consistent with Guo and Gifford (2002) who
reported no difference in soil C stocks between forest and converted pasture systems
in regions with less than 2000 mm of annual rainfall. Overall, the slightly higher total
Land use change implications on the soil C sequestration potential of peri-urban environments (Paper 4)
159
C contents in the pasture land use is due to more roots per area including through soil
depth, while the turf grass roots do not extend as far down the soil profile as the
pasture, hence less C below 0-10 soil depth.
Australian ecosystems with mostly highly weathered soils (Attiwill et al. 1996)
are generally expected to be limited in their C sequestration potential (Livesley et al.
2009). The change from native to peri-urban environments in the Samford Valley
included decades of agricultural practices associated with livestock production
across. The average CT content of over 2 % across all sites is indicative of increased
primary production from intensive agriculture across the Valley after clearing of the
original vegetation, with organic inputs from cattle and mineral fertilizer. The
outcome of this study in Samford Valley, however, supports the fact that intensive
agricultural management such as pasture and turf grass can result in soil C exceeding
native conditions (Six et al. 2002) through substantial C sequestration despite the soil
age (Grace and Basso 2012). This outcome of relatively high soil C reflects other
reported total C contents from subtropical peri-urban environments in Hong Kong
with up to 49 t C ha-1
15 cm-1
(Kong et al. 2014).
The relative masses of the C fractions varied widely across sampling sites, with no
significant influence of the type of land use. However, analysing the C fractions and
land use types separately with their particular land use history indicated some trends.
The conversion from forest to pasture did not affect any of the C fractions in their
absolute mass but change from forest into turf grass suggests a decrease in the stable
CR fraction. This decrease in CR can be the result of physical disturbance of the
topsoil during construction processes and turf grass establishment, mixing in the
lower CR content from the subsoil as the land use change was too recent to affect this
resistant form of SOM. Construction processes in the region generally include plant
cover removal, which includes some of the fertile topsoil (approximately 5 cm). This
removal of the upper CR rich topsoil could additionally lower the CR contents in the
turf grass topsoil. The higher CS content in the turf grass, on the other hand, is
indicative of a soil in transition receiving a low plant C/N input (Table 7-4). The
consistent proportion of active C relative to total C across all land uses and depths
(29 %) indicates no change in C inputs between land uses with the subtropical
climate supporting high primary production. The average plant C/N of 29 at the turf
160
grass sites highlights the effect of N fertilizer use and legumes in maintaining high
levels of productivity.
The lack of significance in terms of land use type on the absolute C fraction mass
could be due to one or a combination of the following scenarios. 1) Topsoil
displacement and mixing with the subsoil due to construction processes during
urbanization substantially alters the C fractions and disturbs C sequestration. 2) The
rate of soil C sequestration in these highly weathered soils is slow and the age of the
land use is not enough to detect significant changes. The first scenario is supported
by the reduced amount of CR fraction in the turf grass soil of this study, as most C
storage generally occurs in the topsoil (Golubiewski 2006) and therefore might have
been reduced solely or partly by displacement and mixing during establishment as
customary in the Valley (van Delden et al. 2016a). This increasing soil heterogeneity
caused by construction processes suggests the need for including deeper soil samples
(20-50 cm) in future sampling designs to confirm the insignificance of the land use
type on the C sequestration potential determined in this study. The second scenario is
supported by findings from another study in a humid subtropical climate identifying
a potential change from a C sink to a source of peri-urban environments after two
decades (Kong et al. 2014). The potential for peri-urban soils to become C sources
requires regular sampling for accurate C budgeting in the future. Increased accuracy
could also be achieved in future using a larger sample size of increasing land use age.
Our present result would lead to the conclusion that in subtropical regions, climate
has a more dominant effect on C cycling than land use type.
7.5.2 Environmental influence on C and N cycling
This study identified the environmental parameters, which correlate to C
sequestration such as total soil C and N content, the proportion of active and resistant
C and clay content. While the majority of C sequestration occurs in the topsoil based
on SOM accumulation, the subsoil has some potential for increased C content by
increasing BD (Golubiewski 2006; Bolstad and Vose 2005). Even the topsoil in this
study had higher BD in the pasture and turf grass land use due to the agricultural and
peri-urban management activities as sandy topsoils generally tend to higher BD
(Bowman et al. 2002). These higher BDs in peri-urban environments may imply
diffusivity alterations in the soil (Veldkamp 1994) and could therefore be expected to
Land use change implications on the soil C sequestration potential of peri-urban environments (Paper 4)
161
affect the CA fraction by altered respiration dynamics, which was, however, not
confirmed by the statistics. Additionally, high water infiltration rates and low
nutrient holding capacity of the sandy topsoils in the Samford Valley may impair N
accumulation even in the fertilized pasture and turf grass soils. The mineral N
content of the soil was not correlated with any of the C fractions, which is not
surprising considering it was a one-off static measurement of what is a very dynamic
soil process (i.e. mineralisation). On the other hand, highly efficient N cycling has
been reported in subtropics ecosystems (Xu et al. 2013; van Delden et al. 2016b),
minimizing the limiting effect of plant available N for SOM production.
Mineral fertilization in this peri-urban environment, however, is generally
associated with irrigation, which removes this limiting factor and may affects pH
and/or EC in the long-term. In these intensively managed systems with sandy
topsoils, subtropical rain events can increase the potential for NO3-
leaching when
treated with mineral fertilizer, which could result in substantial pollution of
groundwater and open waterways (Barton et al. 2006). The fertilization of these
sandy soils may support long-term increases in organic N in form of SOM rather
than frequent mineral N supply. This necessity of organic bound nitrogen for long-
term N accumulation can easily be improved by management practices such as
leaving clippings behind after mowing instead of removal as suggested from
modelled long-term turf grass management in temperate zones (Zhang et al. 2013b).
Micro-aggregates include clay material and mostly stable SOM inaccessible to
microbes, which stores C in the soil in the long term (Denef et al. 2004; 2007). This
interaction of clay material and C sequestration is supported by the observation of a
strong correlation between clay content and the slow and resistant C fractions. The
sandy topsoils in the Samford Valley would support increased C sequestration when
clay material is imported during the construction processes, which was the case at
some of the sampled sites. However, this incorporated clay material might then
increasingly effect other aspects of the C and N cycle, for example fixing NH4+
in the
soil (Marschner 2012) or affect the GHG flux balance by changing soil moisture
dynamics (Fest et al. 2015b).
The neutral range in the soil pH across the Samford Valley should have minimal
impact on soil C and N cycling. Electric conductivity, on the other hand, had a
162
positive correlation on some of the C fractions, which could mean that a slight EC
increase may positively affect microbial activity and therefore SOM production in
these extremely low EC soils. As EC can limit microbial activity when extremely
high, extremely low EC can also limit soil fertility even with high CT contents based
on the low amount of plant growth supporting nutrient salts in the soil solution
(Schlesinger 1995). This correlation of EC and the CA fraction may suggest an
acceleration of the active C cycle such as respiration processes within fertilized and
irrigated systems. However, the more stable C fractions CS and CR are not affected
by EC as the salt is unlikely to reach critical levels in these well-drained and highly
weathered soils.
7.6 Conclusion
This study illustrates that land use change into peri-urban environments can
support C sequestration in subtropical sandy soils and exceed native conditions. Soils
of secondary forest, pasture and turf grass land use had on average the same long-
term C sequestration rates, regardless of the plant cover. Practices during land use
change such as topsoil displacement and soil disturbance for construction purposes
have a stronger long-term influence than the land use. Incorporated clay material
during construction can significantly affect C sequestration into more stable SOM
fractions of peri-urban environments. Overall, these biogeochemical data on
subtropical land use change associated with urbanization highlight the potential of
peri-urban environments to store substantial amounts of C and N in the soil in the
long-term. However, the higher management of turf grass systems does not result in
significantly higher C sequestration and can therefore not negate the higher
emissions resulting from fertilizer and irrigation practices in the long term.
7.7 Acknowledgements
The study was supported by the Central Analytical Research Facility (CARF)
operated by the Institute for Future Environments (IFE) of the Queensland University
of Technology (QUT). The support from the Moreton Bay Regional Council was
Land use change implications on the soil C sequestration potential of peri-urban environments (Paper 4)
163
greatly appreciated for providing generous information and access to the public
sampling sites. Special thanks for the private access of residential properties to the
Samford Valley community, Samford Valley Research Facility (SERF) and Marcus
Yates.
164
Chapter 8: Discussion and Conclusions
This research is the first to establish an inter-annual non-CO2 GWP using high
frequency N2O and CH4 soil-atmosphere gas fluxes for a peri-urban environment. It
was hypothesized that urbanization processes can increase non-CO2 GWPs by
significantly altering C and N cycling immediately after land use change, inter-
annually and in the long-term. The experimental outcome was analysed in chapters 4
to 7 within four individual publications and will now be discussed within the
research objectives developed in chapter 1.
8.1 Environmental parameters
Several environmental parameters that influenced biogeochemical C and N
cycling were identified by this research. These included annual rainfall, heavy rain
events, temperature, soil texture, BD, WFPS, pH, EC, total C and N content as well
as N in the mineral form.
Carbon sequestration in the sand dominated Samford Valley soils, as identified in
Chapter 7, was mainly influenced by the clay content, total C and N as well as the
proportion of active and resistant C. The clay content became more influential with
increasing C stability in the soil, which supports the general concept of C
sequestration into soil micro-aggregates structured by clay colloids (Denef et al.
2004; 2007). The uncorrelated impact of mineral N on C sequestration might be the
results of an efficient N cycle in the subtropics (Xu et al. 2013), minimizing the
limiting effect of plant available N for SOM production. Mineral fertilization in
urban lawns, however, often occurs in conjunction with irrigation and therefore
substantially increases management inputs, which may influence soil parameters
such as pH or EC in the long-term by affecting the salt content. Soil pH generally
affects N cycling in the soil, by influencing NH4+ availability and N fixation
(Marschner 2012), but was not correlated in this study, which is most likely the result
of the neutral pH range and minor differences between sites. Electrical conductivity,
on the other hand, had a positive effect on the active C fraction, which may increase
Discussion
165
microbial activity in these overall low EC soils with limited salt content. This
limitation in the highly weathered soils of tropics and subtropics can results in
relatively low fertility of the soil even with high total C contents (Schlesinger 1995).
The main parameters correlated to soil-atmosphere GHG exchange in the
subtropical peri-urban ecosystem experiment at SERF, as identified in Chapters 4 to
6, were soil WFPS and mineral N content. Rainfall (Groffman et al. 2009) and
temperature (Butterbach-Bahl and Kiese 2005; Fest et al. 2009), the main parameters
regulating GHG fluxes in temperate climates, were only minor or not at all correlated
to subtropical GHG fluxes. Heavy rains in the humid subtropical zone typically fall
when temperatures and plant biomass growth are highest during the summer
resulting in high evapotranspiration, a key control of soil moisture and subsequent
WFPS. Inter-annual rainfall variations have been identified to cause annual N2O flux
differences in subtropical grasslands (Rowlings et al. 2015), but could not be
confirmed by this research even with the substantial inter-annual rainfall variation of
430 mm. This can potentially be attributed to to differences in soil texture and
legume content of the grassland between the two studies, with the higher sand
content and lower total soil N limiting N2O emissions. Annual CH4 fluxes, on the
other hand, showed some inter-annual variations from the forest and turf grass land
use but not the pasture. The dry sclerophyll forest averaged nearly half the annual
WFPS than the pasture and turf grass despite receiving the same amount of rainfall,
clearly indicating higher water uptake by the forest. Short-term WFPS peaks after
heavy rain events were lower in the turf grass, while the average WFPS did not differ
significantly between pasture and turf grass, which indicates a buffering effect
resulting from higher evapotranspiration from the even turf grass root system (Barton
et al. 2009b) compared to the patchiness of the pasture cover.
The soil at the core research site SERF, a Chromosol, represents one of the most
widespread soil types in agricultural use in Australia (Isbell 2002), particularly
around the eastern and southern coastlines, which occupy Australia’s major cities
such as Brisbane, Sydney and Melbourne (Commonwealth of Australia 2013) and
therefore the most likely to be effected by current and future urban sprawl. The
generally low SOC content in Chromosols (Baldock et al. 2012) may limit N2O
production by soil microbes because of a deficient C energy source (Giles et al.
166
2012). Furthermore, the high sand content of the topsoil, combined with the
moderate slope, prevents excessive water logging for extended periods of time. This
limits N2O gaseous losses from denitrification, but indicates a high potential for N
leaching as indicated by the substantial decrease in NO3- of >60 kg N ha
-1 from the
fallow soil after a single heavy rain event. This research did not specifically quantify
the proportion of N lost via leaching verses total denitrification (N2+N2O) but
reported N losses via leaching from turf grass systems can account for approximately
>80 kg N ha-1
y-1
from sandy Australian soils (Barton et al. 2006), leading to the
pollution of groundwater and open waterways. The clay-textured subsoil of the
Chromosol, on the other hand, may reach water saturated conditions and therefore
potentially produce CH4 at depth as well as denitrification processes produce N2O.
The pasture soil had a higher topsoil BD than the forest, which is most likely due
to grazing activity as well as the pasture management with heavy machinery. The
turf grass soil, however, had a lower BD than the native forest topsoil, most likely
due to the soil cultivation during establishment in combination with the higher root
density of the turf grass. These BD alterations illustrate the physical impact of land
use change associated with urbanization has on soil structure, which can significantly
influence GHG measurements by the changed diffusivity of the soil (Veldkamp
1994). These combined environmental parameters may suggest that the GHG
emissions identified by this research range at the lower end of subtropical peri-urban
ecosystems.
8.2 Objective 1
Evaluate the immediate ecosystem response to land use change into peri-urban turf
grass in form of soil-atmosphere GHG exchange.
Paper 1 confirmed the hypothesis that turf grass establishment increases soil N2O
emissions and reduces CH4 uptake when changed from well-established land uses
such as native forest and grazed pasture. Turf grass, as the major peri-urban land
cover, increased the non-CO2 GWP by 415 kg CO2-e ha-1
over the first 80 days after
establishment from the converted pasture, measured by the high frequency automated
gas sampling system. Turf grass establishment increased the non-CO2 GWP by
another 30 kg CO2-e ha-1
when compared to the native dry sclerophyll eucalypt
Discussion
167
forest, solely due to strong CH4 uptake in the forest. Turf grass increased daily N2O
emissions from 0.3 g N2O-N ha-1
d-1
in the pasture to 11.6 g N2O-N ha-1
d-1
from the
turf grass due to fertilizer application during establishment phase. Calculating a
general annual non-CO2 GWP estimate from these 80 days averaged daily CH4 and
N2O fluxes from the turf grass, results in 1.9 t CO2-e ha-1
y-1
and exceeds reported
values from irrigated lawns in temperate Australia 1.6 times (Livesley et al. 2010).
The annual baseline for forest and pasture before turf grass establishment estimated
using manual gas sampling, measured moderate CH4 fluxes but did not detect any
N2O emissions throughout year. This might be explained by the time lag between
N2O production and release (Mosier et al. 1998), which highlights the high temporal
variability of emissions and difficulty for accurate long-term estimations. Together
with the strong temporal variability of subtropical heavy rain events underlines the
importance of automated high frequency measurements to capture representative
soil-atmosphere gas exchange.
8.3 Objective 2
Evaluate the annual ecosystem response to N cycling after land use change with the
implication on the potent GHG N2O.
Paper 2 confirmed the hypothesis that land use change associated with
urbanization increases ecosystem N losses in the form of N2O. This increase in the
highly potent GHG from land use change processes into peri-urban environments is a
major contributor to the non-CO2 GWP of peri-urban land use. Fallow land
associated with construction processes and turf grass significantly increased annual
N2O emissions by 30 and 19 times respectively compared to the native forest. The
grazed pasture, however, did not significantly differ to the forest. The SERF soil
reflects the overall minor annual variability of NH4+ and more dynamic nature of
NO3- across most climates in Australia (Livesley et al. 2009; Rowlings et al. 2012;
Fest et al. 2015a). Overall NH4+:NO3
- ratios from Australian forests indicate a higher
NO3- availability in subtropical forest soils (3-4, Rowlings et al. (2012)) compared to
temperate zones (28-125, Livesley et al. (2009; Fest et al. (2015a)).
168
This highlights that the climate is an important driver of N cycling while the
subtropical influence suggests a higher NO3-
availability for plant uptake with less
emissions were observed compared to temperate zones. These results indicate a
higher mineralization rate in the subtropics, which causes a higher mineral N content
for a single point in time, but low N2O losses because of a simultaneous high plant
uptake rate. These mineral N dynamics together with the low N2O losses observed
from this subtropical system indicates a rapid N cycling resulting in an constant flow
from tied up N in organic material to the plant uptake of NO3-, which supports the
hypothesis of an efficient N cycle within subtropical climates (Xu et al. 2013). This
efficiency implies a sufficient N availability for plant and SOM production on a
microscale but no excessive accumulation of the mobile NO3- in the topsoil, such as
highlighted from the fallow soil without the plant uptake. This N dynamic lack of
NO3-
accumulation during specific seasons or dry periods limits the potential N
losses in form of N2O emissions and NO3- leaching from heavy rain events.
The daily N2O flux of 0.2 g N2O-N ha-1
d-1
from the subtropical dry sclerophyll
forest in this study is slightly lower than the averages of < 0.8 g N2O-N ha-1
d-1
reported from temperate Australian dry sclerophyll forests with similar soil type
(Fest et al. 2009; Livesley et al. 2009) and substantially lower than subtropical
rainforests with 1.3 g N2O-N ha-1
d-1
(Rowlings et al. 2012) and tropical estimations
(Werner et al. 2007). These findings support the general hypothesis that these dry
forests with C:N ratios > 20 are minor contributors to the global N2O budget (Page et
al. 2011; Fest et al. 2015a) due to the limited supply of N substrates for nitrification
and denitrification processes (Fest et al. 2009).
The SERF pasture did not significantly alter N cycling in the form of mineral N
dynamics and N2O emissions when compared to the forest land use. The turf grass
establishment, however, significantly increased N2O emissions within the first 2
months after turf grass establishment about 12 times the annual intensity, while over
the remaining 10 months only minor fluxes occurred even after further fertilization
events. The establishment phase should therefore be considered separately when
calculating the non-CO2 GWP from new land uses. Emissions from the fallow on the
other hand increased over the 12 month period. The fallow soil accumulated NO3- in
the topsoil due to the lack of plant uptake which also created high soil water
conditions favourable for mineralisation (Robertson and Groffman 2007). With
Discussion
169
increasing NO3- contents in the soil, N2O emissions increased significantly from the
fallow soil with daily averages of up to 78 g N2O-N ha-1
d-1
after heavy rain events.
However, the amount of N lost via N2O emissions from the fallow soil cannot
account for the substantial decrease of NO3-
after those heavy rain events, which
suggests a substantial leaching potential considering the sandy texture of the topsoil.
8.4 Objective 3
Evaluate the non-CO2 GWP of peri-urban environments in subtropical Australia.
Paper 3 confirmed the hypothesis that peri-urban land use significantly increases
the non-CO2 GWP compared native forest by increasing N2O emissions and reducing
CH4 emissions. The pasture and turf grass land uses reduced soil CH4 uptake
compared to native forest but only turf grass increased N2O emissions due to the
application of fertilizer. Short, intermittent periods of CH4 emissions from the
pasture resulted in substantially less annual CH4 uptake compared to the forest, while
the non-CO2 GWP was not significantly different. Turf grass, however, increased the
non-CO2 GWP significantly by 329 and 233 kg CO2-e ha-1
y-1
compared to the forest
and pasture respectively. This highlights the dominant influence of the potent N2O
emissions, which were comparably low in the forest and pasture land use. While the
non-CO2 GWP of the forest showed little inter-annual variation, emissions from the
pasture changed from an overall GHG source in year one to a sink in year two with
increasing rainfall. However, the most significant inter-annual non-CO2 GWP
decrease occurred in the turf grass system due to 12 times lower N2O emissions
resulting in a 10 fold reduction in the emission factor from 0.7 % to only 0.07 % in
the second year of the study.
All land use types were annual CH4 sinks with strong soil CH4 uptake ranging
from 0.8 to -2.9 kg CH4-C ha-1
y-1
, in the increasing uptake order of pasture < turf
grass < forest. Soil CH4 uptake decreased in all land use types when WFPS reached
extremely high or low levels. These variations in CH4 uptake are driven by the
decreased methanotrophic and/or increased methanogenic activity when soil oxygen
or moisture reaches extreme levels (Smith et al. 2000; Kaye et al. 2004). The dry
sclerophyll forest had significantly lower WFPS compared to the pasture and turf
170
grass with extended periods of very low WFPS but was still the strongest CH4 sink.
This could be explained by the higher simultaneously occurring CH4 production from
the thick organic root layer in the pasture compared to the barely covered mineral
soil in the forest, which has the highest CH4 uptake potential (Butterbach-Bahl and
Papen 2002; Fest et al. 2015b). Annual CH4 uptake of the SERF forest however, was
with -2.7 kg CH4-C ha-1
y-1
averaged over the two years, 27 % less than subtropical
rainforest soils with up to -3.7 kg CH4-C ha-1
y-1
, which is mostly due to the
consistent year-round soil moisture conditions in the rainforest and therefore higher
methanothropic activity (Rowlings et al. 2012).
While the annual CH4 uptake decreased in the forest and turf grass soil by 14 %
and 41 % respectively from year one to year two with increasing rainfall, the pasture
soil CH4 uptake was constant. This inter-annual decrease in CH4 uptake with
increased rain suggests a stronger influence of climatic conditions in turf grass
systems than the pasture or forest. This variation could be due to a number of factors
such as fertilization and alteration of soil physical conditions that influence diffusion
(Groffman and Pouyat 2009). The CH4 uptake in turf grass soil depleted when soil
moisture conditions were extremely high and extremely low but never became clear
a CH4 source unlike the pasture. This consistent rate of daily CH4 uptake and annual
CH4 uptake by the SERF turf grass soil makes this peri-urban land use a solid CH4
sink with 1.6 kg CH4-C ha-1
y-1
averaged over two years, while reported temperate
turf grass systems range widely from CH4 uptake (Groffman and Pouyat 2009) to
-3 kg CH4-C ha-1
y-1
(Kaye et al. 2004).
The annual N2O emissions emitted from the forest and pasture soil were low both
years, while the turf grass soil emitted more than 5 times less annual N2O in year two
compared to year one. This high inter-annual N2O flux variation from the turf grass
soil was chiefly associated with the establishment phase of 80 days rather than the
climatic conditions, despite the higher rainfall in year two. This increased rainfall
had no significant effect on the inter-annual N2O flux variations in the forest and
pasture soil. The consistent inter-annual N2O emissions of 0.15 kg N2O-N ha-1
y-1
from the SERF pasture was much lower than N2O dynamics from other Australian
pastures where up to 3.4 kg N2O-N ha-1
y-1
was reported (Livesley et al. 2009), with
substantial inter-annual variation (Rowlings et al. 2015). These differences can be
Discussion
171
most likely explained by the strong influence of soil texture such as clay content as
well as legume content as the low legume content of the SERF pasture.
The establishment phase of the turf grass lawn clearly caused the significant
higher annual N2O emissions in year one compared to year two, most likely due to
the fertilizer input and irrigation while the root system was not fully established for
efficient mineral N uptake, which was then available for denitrification processes.
The 10 times lower EF in year two compared to year one indicates the rapid
establishment of the root system and increasing N uptake and reducing both the
available mineral N and soil moisture for N2O emissions. However, despite the
higher annual N2O emissions in year one cause by the establishment phase, the
emissions were still relatively low on an annual basis representing half of reported
emissions from temperate turf grass systems of up to 3 kg N2O-N ha-1
y-1
(Groffman
et al. 2009). This difference in N2O emissions between temperate zones and the
results presented here are based on higher fertilization rates used in the USA and the
faster mineral N uptake by turf grass in the warm subtropical climate.
8.5 Objective 4
Evaluate the long-term effect of land use change on C and N cycling by identifying
the soil C sequestration potential in peri-urban environments.
Paper 4 rejected the hypothesis that peri-urban turf grass significantly affects C
and N cycling in the long-term by increasing the soil’s C sequestration potential
compared to pasture and forest. In this study, the total C content of forest, pasture
and turf grass topsoils were examined for their active, slow, and resistant C fraction
after at least one decade of land use age. Total soil C and N varied widely across
sampling sites from 17.3 to 46.6 t C ha-1
and 1.0 to 3.4 t N ha-1
respectively with the
widest C and N range in pasture soils. All C fractions varied widely across sampling
sites, resulting in an overall insignificant effect of land use type on C sequestration
most likely due to the high variation due to inherent soil characteristics and
topography.
Analysing the fractions and land use types separately with their particulate land
use history indicates some trends. The change from forest to pasture did not affect
172
any of the C fractions but the change from forest into turf grass suggests a decrease
of the most stable CR fraction. This decrease in the most stable fraction is most likely
the result of physical disturbance of the topsoil during construction processes and
turf grass establishment, as the land use change was too recent to affect the resistant
SOM form. Construction processes in the region generally involve plant cover
removal and therefore some of the fertile surface soil. The change from pasture to
turf grass, however, increased the CS fraction, which might be the first indication of
potentially higher C sequestration in the form of resistant C after one decade of land
use age. The overall insignificant effect of land use on C fractions could be one or a
combination of the following scenarios. 1) Topsoil displacement and mixing with the
subsoil due to construction processes during urbanization substantially alters the C
fractions and disturbs C sequestration. 2) The rate of soil C sequestration in these
highly weathered soils is slow and the age of the land use is not enough to detect
significant changes.
The averaged total C content across all land uses in Samford Valley of 58 t C ha-1
in 20 cm topsoil represents a substantial increase in soil C compared to the native
conditions of the dry sclerophyll forest with 29 t C ha-1
20 cm-1
. This outcome
reflects other reported soil C contents from subtropical peri-urban environments in
Hong Kong with up to 49 t C ha-1
15 cm-1
(Kong et al. 2014). This outcome supports
the general hypothesis that even highly weathered soils, such as Australian
Chromosols (Attiwill et al. 1996), can sequester substantial amounts of C (Grace and
Basso 2012). On the other hand, some research suggests a limited C sequestration
potential by highly weathered soils (Livesley et al. 2009), resulting in increasing soil
C losses via CO2 respiration when soils reach their C saturation level (Six et al. 2002;
Stewart et al. 2007; 2008). This accelerated C cycling with increasing C content is
supported by this study’s strong correlation of the active C fraction to the total C
content. However, the wide range of total C measured in the Samford Valley, which
vary from 1.2 to 4.2 %, demonstrates that within peri-urban environments, soils can
exceed their native conditions (Six et al. 2002).
Paper 4 estimated that peri-urban environments have a substantial potential for C
sequestration and therefore reducing GHG in the atmosphere. Subtropical turf grass
soil, however, did not significantly sequester more C and N than pasture and forest
due to its higher productivity and management as it was suggested by temperate turf
Discussion
173
grass systems (Golubiewski 2006; Raciti et al. 2011a). However, further research is
required with greater replication and awareness of land use age to support the trends
identified by this study.
8.6 Outlook
This research identified land use change from forest and to a peri-urban
environment alters C and N cycling immediately after establishment but becomes
quickly comparable. In the long term, peri urban land uses such as forest, pasture and
turf grass can become comparable and even exceed native C and N pools. This long-
term potential for comparable C and N cycling in peri-urban environments is most
likely the result of a tight coupling of N turnover and plant uptake resulting in an
efficient nutrient cycling in subtropical soils. Therefore, these subtropical peri-urban
environments highlight the importance of natural and some managed ecosystems of
Australia to offset the current GHG inventory of approximately 525,202 Pg CO2-e y-1
by sequestering atmospheric GHG into biomass and soil (AGEIS 2015; Hatfield-
Dodds et al. 2015).
The subtropical dry sclerophyll forest soil at SERF is an important GHG sink,
sequestering 0.08 t CO2-e ha-1
y-1
. Tropical forest soils indicate the potential to
reduce this GHG sink by emitting considerable amounts of N2O, globally averaging
1.2 kg N2O-N ha-1
y-1
and up to 32 g N2O-N ha-1
d-1
from Australian rainforest soils
(Werner et al. 2007). These substantial N2O emissions from rainforest soils increase
the non-CO2 GWP to -0.03 t CO2-e ha-1
y-1
while temperate and boreal forests range
between -0.9 and -1.18 t CO2-e ha-1
y-1
(Dalal and Allen 2008). Additionally,
research on CH4 uptake from forests soils along a rural to urban gradient suggest a
decrease in soil CH4 uptake within increasing urbanization of the neighbouring
environment (Groffman and Pouyat 2009), which may limit the sink potential of the
SERF forest soil with increasing population density of Samford Valley. While the
current research identified comparable C and N cycling in forest and pasture soils
and a generally low non-CO2 GWP, turf grass establishment significantly alters
nutrient cycling in peri-urban environments and can increase the non-CO2 GWP.
174
Grasslands, which range from unmanaged native rangelands to intensively
agricultural used pastures, cover about 40 % of the terrestrial ice-free surface
worldwide with 450 M ha in Australia alone (AGO 2010), making it the principal
land use type with significant potential for climate change mitigation or acceleration,
depending on their management. However, uniform international classifications for
the management intensity of these grasslands are required to quantify a non-CO2
GWP for the entire peri-urban environment.
Of the hypothesized outcomes from all four initial project objectives, only the
long-term impact of the land use needs to be reconsidered (Figure 8-1). The
comparable GHG emissions from the unfertilized pasture soil to the fertilized turf
grass after the establishment phase together with the comparable C sequestration
strength The intense management of subtropical peri-urban turf grass systems with
fertilization and irrigation does not significantly increase either soil GHG emissions
after the establishment phase or the C sequestration when compared to grazed
pasture. However, a full lifecycle assessment has to determine how much GHG
emissions from fertilizer production and distribution and management practices
would add to the turf grass’s non-CO2 GWP to give the full extent of land use change
into peri-urban turf grass.
Figure 8-1 Hypothesized multiple time scale scheme corrected for the long-term
response
Discussion
175
More than half the world’s 7.4 billion population occupies about 2.4 % of the
global terrestrial land surface and growth is expected to expand mostly in urban areas
through population growth and rural to urban migration (United Nations 2013). This
population expansion will advance with extensive land use change and increasingly
turf grass establishment and management. While turf grass establishment
significantly changes nutrient cycling in soils and therefore soil-atmosphere GHG
exchange, green spaces in and around cities are important to keep our cities liveable
and enjoyable (Commonwealth of Australia 2013). However, the predicted climatic
changes towards more extreme weather events and rising temperatures (IPCC 2013),
will undoubtedly alter soil nutrient cycling and the subsequent balance between GHG
gas emissions and consumptions. The urbanization effects combined with the
changing global climate may increase the GWP of natural ecosystems and therefore
increase feedback effects on the climate (Betts 2007; Grimm 2008).
Additionally, the substantial climate variations make generalized annual flux
estimations difficult, especially with the predicted climate changes towards more
extreme and season-untypical heavy rain event, and therefore demands long-term
evaluations in future biogeochemical cycling. However, the substantial difference in
annual rainfall and temperature between both experimental years identified the well-
established land uses forest and pasture to vary little in their inter-annual N2O
emissions, which make up 90 % of the non-CO2 GWP. Methane uptake, on the other
hand, can be reduced with higher annual rainfall especially in form of heavy rain
events. The annual rainfall mainly occurs with high temperatures during summer and
therefore results in high evapotranspiration, which regulates WFPS more than the
amount of rain. This makes these GHG fluxes not directly correlated to rainfall but
significantly more effected by physical soil parameters such as soil texture, porosity
and BD. More information is therefore needed from different soil types and highly
managed turf grass systems to provide a holistic estimation on the peri-urban
contribution to climate change.
Sandy soils in intensely managed systems are highly prone to increased N
leaching from mineral fertilizers with the subtropical heavy rain events, resulting in
the substantial pollution of groundwater and open waterways (Barton et al. 2006).
Frequent mineral fertilization of sandy soils, therefore, might not support the long-
176
term ecosystem productivity for increased C sequestration as much as the increase in
organic input would do, especially because of the tightly coupled N cycling in the
subtropics as identified in this study. The necessity of organic bound nitrogen for
long-term N accumulation can easily be improved by management practices such as
leaving clippings behind after mowing as suggested from modelled long-term turf
grass management in temperate zones (Zhang et al. 2013b). Additionally, the mineral
fertilizer use in turf grass systems may increase denitrification (Raciti et al. 2011b)
and potential N losses in form of N2, which then need to be compensated for with
more fertilizer to ensure turf grass productivity.
Despite no reliable data currently being available for the extent of existing turf
grass Australia wide (pers. comm. with Turf Australia), the turf grass industry is
growing rapidly as shown by consistently increasing sales numbers. Nearly half of
the distributed turf grass in Australia is cultivated in tropical and subtropical
Queensland (ABS 2012). With about 17,320 ha of turf grass being established in
Australia annually (Turf Australia 2012), the GHG emissions for the establishment
alone could be estimated from the non-CO2 GWP presented here of up to 7,326 t
CO2-e y-1
. However, Australia has the potential to reduce national emissions, which
are currently four times the global average (Hatfield-Dodds et al. 2015), by
optimizing fertilizer use efficiency to reduce fertilizer inputs and subsequent GHG
emissions and by increasing C sequestration into biomass of natural and managed
ecosystems. The efficient nutrient cycling of the tropics and subtropics (Xu et al.
2013), as supported by this research in SEQ, limits C and N losses and increase GHG
uptake from the atmosphere into SOM. Therefore, peri-urban forests and grasslands
are currently most likely undervalued in their contribution to climate change
mitigation, especially because the majority of future global demographic growth is
projected to take place in tropical and subtropical regions (UNFPA 2011).
Based on this research and the importance of this subject as highlighted by the
literature, the following research topics can be recommended:
1. Ecosystem modeling using this baseline data set could identify the best
possible management practices for subtropical turf grass to make peri-urban
environments more sustainable and give recommendations for future land use
change processes.
Discussion
177
2. Establishing a full peri-urban life cycle assessment with a GWP including CO2
ecosystem response to land use change could quantify urbanization processes
for the national GHG inventory.
3. Upscaling of soil-atmosphere GHG fluxes within peri-urban environments
while taking sealed soils in proportion into account.
4. Mineralization experiments could support the tight N cycling in the subtropics
to review public and private fertilizer use recommendations for the turf grass
industry.
5. A stronger categorization of land use age and history together with an
increased sample size of the soils in Samford Valley could reveal a hidden
significance of land use change on the C and N cycle.
6. Research on the socio-ecological aspect of urbanization related land use
change could identify the driving anthropogenic decision patterns to predict
future urbanization developments more accurately and make sustainability
recommendations for urban and peri-urban greenspaces before they are
established.
Overall, this research highlights that urbanization related land use change might have
a more substantial impact on soils and the climate than accredited for, while the
subtropical climate might support quick adaptation through tight nutrient cycling.
However, more research could identify the most efficient management strategies for
these urbanization processes as they will undoubtedly increase in the future. This
century’s global challenge of achieving food security and mitigating climate change
while increasing economic growth will only succeed when we improve our current
economic and ecological management strategies to minimize input such as
fertilization with equivalent productivity.
178
8.7 Conclusions
This research establishes the first data set for subtropical C and N cycling of peri-
urban environments effected by land use change. High-frequency N2O and CH4 flux
measurements identified the low non-CO2 GWP baseline for the native dry
sclerophyll forest. Land use change into pasture increases this GWP mainly by
reducing CH4 uptake. Establishing turf grass systems during urbanization processes
did significantly increase N2O emissions, however, reducing to a pasture comparable
rate in only 2 months. This temporal development in combination with the
comparable soil C sequestration of the turf grass, forest and pasture highlights the
significant influence of the subtropical climate and site characteristics rather than
management practices such as fertilization and irrigation. These environmental
conditions result in an efficient N cycle with rapid turnover and plant uptake, and
consequently minor N losses in form of NO3- and N2O. These results countered the
hypothesis that high temperatures and moisture conditions, which favour microbial
activity, would result in substantial annual GHG emissions.
Subtropical land use change increases the non-CO2 GWP when converted from
native forest to pasture and turf grass. However, the turf grass presented here was
managed at an average industry rate and some high-end users, such as golf courses
and sports grounds, to identify a representative average for residential areas. This
relatively low non-CO2 GWP of the turf grass after two years may increase when
fertilized at the industrial maximum, when there is more mineral N available in the
soil than the plants can take up before denitrification or leaching occurs, i.e. over-
fertilization. The biogeochemical N cycling in the fallow soil, representative for local
construction processes, identified a substantial NO3-
accumulation after plant cover
removal. This accumulated NO3- decreased rapidly after heavy rain events and was
lost to the system. Such additional N losses in form of potential NO3- leaching need
to be quantified to optimize fertilizer use within private and public land use
management to compensate for the losses during land use change.
The native forest in this research proved to be very efficient in its N cycling as
shown by the higher NO3- availability in the soil compared to other forest systems but
with minor N2O losses. The low N2O emissions with the strongest CH4 uptake in the
Conclusion
179
forest soil illustrates the close coupling of microbial N cycling and plant growth in
the subtropical climate and support efficient nutrient cycling. The pasture soil
emitted minor amounts of N2O comparable to the forest but became a CH4 source
over short periods of time, which was decisive in changing the negative non-CO2
GWP of the forest to a positive non-CO2 GWP of the pasture. Especially changes in
BD effect the water infiltration and therefore aeration of the soil pores, which drives
CH4 production and uptake. Establishing turf grass during land use change
significantly effects the non-CO2 GWP by increasing annual N2O emissions with a
short-term but significant emission peak of two months after the establishment. Due
to the extensive area undergoing these urbanization processes worldwide, such
significant N2O emission peaks should be included into national and international
GHG inventories and support future IPCC climate change scenarios.
This data set of high-frequency soil-atmosphere GHG exchange is not only of
high quality but also highly accurate due to the fully automated continuous
measurements, which highlights that N2O emissions are generally short lived but
intense. This substantial temporal variability makes them easy to over- or
underestimate with conventional low intensity sampling methods over short periods
of time. To overcome the significant inter-annual climate variations of the humid
subtropics and identify the main drivers of soil-atmosphere GHG exchange, this
research was conducted over two consecutive years.
Optimized management strategies together with the efficient N cycling driven by
the humid subtropical climate can improve long-term environmental benefits while
keeping peri-urban land use such as turf grass highly productive. Based on the
biogeochemical data gathered by this research, three peri-urban land use
management factors can be suggested to policy makers for consideration to make
urbanization processes economically efficient and environmentally sustainable: (i)
previous land use, (ii) duration of development process, and (iii) purpose of the new
land use and productivity required, such as public or private, sports ground or park.
Overall, the peri-urban environment of this research illustrates the C sequestration
potential of subtropical turf grass to be long-term comparable to forest and pasture
and even exceed the native C and N pool within Samford Valley. These findings
highlight that ecosystem response to land use change needs to be evaluated on a
180
multiple time scale, as the impact on the C and N cycle can change substantially
from the immediate to the long-term. This research concludes that well-established
subtropical peri-urban environments have the potential to store substantial amounts
of C and N in the soil while emitting minor GHG emissions. Including these
transitioning environments into global C and N pool estimations can support C
sequestration strategies worldwide to mitigate climate change and improve soil
fertility to achieve food security.
Bibliography
181
Bibliography
ABARES. 2010. "Land use and land management information for Australia:
workplan of the Australian Collaborative Land Use and Management
Program (ACLUMP)", edited by Australian Bureau of Agricultural and
Resource Economics and Sciences. Canberra.
ABS. 2012. "7106.0 - Australian Farming in Brief, 2012", edited by Australian
Bureau of Statistics. Canberra.
ABS. 2015. "3218.0 - Regional Population Growth, Australia, 2013-14", edited by
Australian Bureau of Statistics. Canberra.
AGEIS. 2015. "National Greenhouse Gas Inventory": Australian Greenhouse
Emissions Information System, Australian Government Department of the
Environment.
AGO. 2010. "National Inventory Report 2008": National Greenhouse Account,
Australian Greenhouse Office, Commonwealth of Australia, Canberra.
Attiwill, PM, PJ Polglase, CJ Weston and MA Adams. 1996. "Nutrient cycling in
forests of south-eastern Australia." Nutrition of eucalypts. CSIRO,
Melbourne: 191-227.
Baggs, Elizabeth M. 2011. "Soil microbial sources of nitrous oxide: recent advances
in knowledge, emerging challenges and future direction." Current Opinion in
Environmental Sustainability 3 (5): 321-327.
http://www.sciencedirect.com/science/article/pii/S1877343511000893. doi:
http://dx.doi.org/10.1016/j.cosust.2011.08.011.
Baldock, J. A., I. Wheeler, N. McKenzie and A. McBrateny. 2012. "Soils and
climate change: potential impacts on carbon stocks and greenhouse gas
emissions, and future research for Australian agriculture." Crop & Pasture
Science 63 (3): 269-283. <Go to ISI>://WOS:000304489700009. doi:
10.1071/cp11170.
Barton, L, GGY Wan, RP Buck and TD Colmer. 2009a. "Does N fertiliser regime
influence N leaching and quality of different-aged turfgrass (Pennisetum
clandestinum) stands?" Plant and soil 316 (1-2): 81-96.
182
Barton, L, GGY Wan and TD Colmer. 2006. "Turfgrass (Cynodon dactylon L.) sod
production on sandy soils: II. Effects of irrigation and fertiliser regimes on N
leaching." Plant and soil 284 (1-2): 147-164.
Barton, L. and T. D. Colmer. 2011. "Granular wetting agents ameliorate water
repellency in turfgrass of contrasting soil organic matter content." Plant and
Soil 348 (1-2): 411-424. <Go to ISI>://WOS:000295587700029. doi:
10.1007/s11104-011-0765-3.
Barton, L., G. G. Y. Wan, R. P. Buck and T. D. Colmer. 2009b. "Nitrogen Increases
Evapotranspiration and Growth of a Warm-Season Turfgrass All rights
reserved. No part of this periodical may be reproduced or transmitted in any
form or by any means, electronic or mechanical, including photocopying,
recording, or any information storage and retrieval system, without
permission in writing from the publisher." Agronomy Journal 101 (1).
http://dx.doi.org/10.2134/agronj2008.0078. doi: 10.2134/agronj2008.0078.
Barton, Louise and Timothy D. Colmer. 2006. "Irrigation and fertiliser strategies for
minimising nitrogen leaching from turfgrass." Agricultural Water
Management 80 (1–3): 160-175.
http://www.sciencedirect.com/science/article/pii/S0378377405002969. doi:
http://dx.doi.org/10.1016/j.agwat.2005.07.011.
Betts, Richard. 2007. "Implications of land ecosystem‐atmosphere interactions for
strategies for climate change adaptation and mitigation." Tellus B 59 (3): 602-
615.
Blair, Graeme J, Rod DB Lefroy and Leanne Lisle. 1995. "Soil carbon fractions
based on their degree of oxidation, and the development of a carbon
management index for agricultural systems." Crop and Pasture Science 46
(7): 1459-1466.
Blume, Hans-Peter, Gerhard W Brümmer, Heiner Fleige, Rainer Horn, Ellen
Kandeler, Ingrid Kögel-Knabner, Ruben Kretzschmar, Karl Stahr and Berndt-
Michael Wilke. 2015. Scheffer/Schachtschabel soil science: Springer.
Bolstad, P. V. and J. M. Vose. 2005. "Forest and pasture carbon pools and soil
respiration in the southern Appalachian Mountains." Forest Science 51 (4):
372-383. <Go to ISI>://WOS:000231006800009.
BOM. 2015. Commonwealth Bureau of Meteorology, Australian Government.
Bouwman, AF. 1998. "Environmental science: Nitrogen oxides and tropical
agriculture." Nature 392 (6679): 866-867.
Bibliography
183
Bowman, GM, J Hutka, N McKenzie, K Coughlan and H Cresswell. 2002. "Particle
size analysis." Soil physical measurement and interpretation for land
evaluation: 224-239.
Box, George EP and David A Pierce. 1970. "Distribution of residual autocorrelations
in autoregressive-integrated moving average time series models." Journal of
the American statistical Association 65 (332): 1509-1526.
Breuer, Lutz, Hans Papen and Klaus Butterbach‐Bahl. 2000. "N2O emission from
tropical forest soils of Australia." Journal of Geophysical Research:
Atmospheres (1984–2012) 105 (D21): 26353-26367.
Brown, Sally, Eric Miltner and Craig Cogger. 2012. "Carbon sequestration potential
in urban soils." In Carbon Sequestration in Urban Ecosystems, 173-196:
Springer.
Butterbach-Bahl, K and H Papen. 2002. "Four years continuous record of CH4-
exchange between the atmosphere and untreated and limed soil of a N-
saturated spruce and beech forest ecosystem in Germany." Plant and Soil 240
(1): 77-90.
Butterbach-Bahl, Klaus, Elizabeth M Baggs, Michael Dannenmann, Ralf Kiese and
Sophie Zechmeister-Boltenstern. 2013. "Nitrous oxide emissions from soils:
how well do we understand the processes and their controls?" Philosophical
Transactions of the Royal Society of London B: Biological Sciences 368
(1621): 20130122.
Butterbach-Bahl, Klaus and Ralf Kiese. 2005. "Significance of forests as sources for
N2O and NO." In Tree species effects on soils: implications for global
change, 173-191: Springer.
Byrne, Loren B. 2007. "Habitat structure: a fundamental concept and framework for
urban soil ecology." Urban ecosystems 10 (3): 255-274.
Canadell, J. G. and M. R. Raupach. 2008. "Managing forests for climate change
mitigation." Science 320 (5882): 1456-1457. <Go to
ISI>://WOS:000256676400033. doi: 10.1126/science.1155458.
Carter, R. and E.G. Gregorich. 2007. Soil Sampling and Methods of Analysis, Second
Edition, edited by Canadian Society of Soil Science: Taylor & Francis.
184
Castaldi, S., A. Ermice and S. Strumia. 2006. "Fluxes of N2O and CH4 from soils of
savannas and seasonally-dry ecosystems." Journal of Biogeography 33 (3):
401-415. <Go to ISI>://WOS:000235256700003. doi: 10.1111/j.1365-
2699.2005.01447.x.
Cerri, C. E. P., M. Easter, K. Paustian, K. Killian, K. Coleman, M. Bernoux, P.
Falloon, et al. 2007. "Predicted soil organic carbon stocks and changes in the
Brazilian Amazon between 2000 and 2030." Agriculture Ecosystems &
Environment 122 (1): 58-72. <Go to ISI>://WOS:000246320300007. doi:
10.1016/j.agee.2007.01.008.
Cerri, C. E. P., K. Paustian, M. Bernoux, R. L. Victoria, J. M. Melillo and C. C.
Cerri. 2004. "Modeling changes in soil organic matter in Amazon forest to
pasture conversion with the Century model." Global Change Biology 10 (5):
815-832. <Go to ISI>://WOS:000221421600020. doi: 10.1111/j.1529-
8817.2003.00759.x.
Commonwealth of Australia. 2013. "State of Australian Cities 2013", edited by
Department for Infrastructure and Transport: Australian Government.
Conant, R. T., S. M. Ogle, E. A. Paul and K. Paustian. 2011. "Measuring and
monitoring soil organic carbon stocks in agricultural lands for climate
mitigation." Frontiers in Ecology and the Environment 9 (3): 169-173. <Go
to ISI>://WOS:000289377800018. doi: 10.1890/090153.
Conant, R. T., K. Paustian, S. J. Del Grosso and W. J. Parton. 2005. "Nitrogen pools
and fluxes in grassland soils sequestering carbon." Nutrient Cycling in
Agroecosystems 71 (3): 239-248. <Go to ISI>://WOS:000229627500003. doi:
10.1007/s10705-004-5085-z.
Conant, R. T., K. Paustian and E. T. Elliott. 2001. "Grassland management and
conversion into grassland: Effects on soil carbon." Ecological Applications
11 (2): 343-355. <Go to ISI>://WOS:000167876900003. doi:
10.2307/3060893.
Conant, R. T., J. Six and K. Paustian. 2004. "Land use effects on soil carbon
fractions in the southeastern United States. II. changes in soil carbon fractions
along a forest to pasture chronosequence." Biology and Fertility of Soils 40
(3): 194-200. <Go to ISI>://WOS:000223263900008. doi: 10.1007/s00374-
004-0754-2.
Cotrufo, M. F., R. T. Conant and K. Paustian. 2011. "Soil organic matter dynamics:
land use, management and global change." Plant and Soil 338 (1-2): 1-3. <Go
to ISI>://WOS:000285389300001. doi: 10.1007/s11104-010-0617-6.
Bibliography
185
Cowie, A., R. Eckard and S. Eady. 2012. "Greenhouse gas accounting for inventory,
emissions trading and life cycle assessment in the land-based sector: a
review." Crop & Pasture Science 63 (3): 284-296. <Go to
ISI>://WOS:000304489700010. doi: 10.1071/cp11188.
CSIRO. 1996. "Interactive Key to the Australian Soil Classification":
Commonwealth Scientific and Industrial Research Organisation.
Cubasch, U, GA Meehl, GJ Boer, RJ Stouffer, M Dix, A Noda, CA Senior, S Raper
and KS Yap. 2001. "Projections of future climate change.", in: JT Houghton,
Y. Ding, DJ Griggs, M. Noguer, PJ Van der Linden, X. Dai, K. Maskell, and
CA Johnson (eds.): Climate Change 2001: The Scientific Basis: Contribution
of Working Group I to the Third Assessment Report of the Intergovernmental
Panel: 526-582.
Dalal, R. C. and D. E. Allen. 2008. "Greenhouse gas fluxes from natural
ecosystems." Australian Journal of Botany 56 (5): 369-407. <Go to
ISI>://WOS:000257868700001. doi: 10.1071/bt07128.
Dalal, R. C., W. J. Wang, G. P. Robertson and W. J. Parton. 2003. "Nitrous oxide
emission from Australian agricultural lands and mitigation options: a review."
Australian Journal of Soil Research 41 (2): 165-195. <Go to
ISI>://WOS:000182064800001. doi: 10.1071/sr02064.
Davidson, E. A. and I. A. Janssens. 2006. "Temperature sensitivity of soil carbon
decomposition and feedbacks to climate change." Nature 440 (7081): 165-
173. <Go to ISI>://WOS:000235839500036. doi: 10.1038/nature04514.
Del Grosso, S. J., W. J. Parton, A. R. Mosier, D. S. Ojima, A. E. Kulmala and S.
Phongpan. 2000. "General model for N2O and N-2 gas emissions from soils
due to dentrification." Global Biogeochemical Cycles 14 (4): 1045-1060. <Go
to ISI>://WOS:000166341000005.
Denef, K., L. Zotarelli, R. M. Boddey and J. Six. 2007. "Microaggregate-associated
carbon as a diagnostic fraction for management-induced changes in soil
organic carbon in two Oxisols." Soil Biology & Biochemistry 39 (5): 1165-
1172. <Go to ISI>://WOS:000245415900020. doi:
10.1016/j.soilbio.2006.12.024.
Denef, Karolien, Johan Six, Roel Merckx and Keith Paustian. 2004. "Carbon
sequestration in microaggregates of no-tillage soils with different clay
mineralogy." Soil Science Society of America Journal 68 (6): 1935-1944.
186
Department of Trade. 1982. "Australian farming systems: the subtropics ": Australian
Government Publishing Service, Canberra.
Durán, Jorge, Alexandra Rodríguez, Jennifer L. Morse and Peter M. Groffman. 2013.
"Winter climate change effects on soil C and N cycles in urban grasslands."
Global Change Biology: n/a-n/a. http://dx.doi.org/10.1111/gcb.12238. doi:
10.1111/gcb.12238.
Fageria, N. K. 2012. "Role of soil organic matter in maintaining sustainability of
cropping systems." Communications in Soil Science and Plant Analysis 43
(16): 2063-2113. <Go to ISI>://WOS:000307353600001. doi:
10.1080/00103624.2012.697234.
FAO and ITPS. 2015. Status of the World’s Soil Resources (SWSR) – Main Report.
Edited by Food and Agriculture Organization of the United Nations and
Intergovernmental Technical Panel on Soils. Rome, Italy.
FAO/UNESCO. 1998. "Soil Map of the World": Food and Argriculture Organization
of the United Nations and United Nations Educational, Scientific and Cultural
Organization.
Fest, Benedikt J., Stephen J. Livesley, Matthias Drösler, Eva van Gorsel and Stefan
K. Arndt. 2009. "Soil–atmosphere greenhouse gas exchange in a cool,
temperate Eucalyptus delegatensis forest in south-eastern Australia."
Agricultural and Forest Meteorology 149 (3–4): 393-406.
http://www.sciencedirect.com/science/article/pii/S0168192308002517. doi:
http://dx.doi.org/10.1016/j.agrformet.2008.09.007.
Fest, Benedikt J., Stephen J. Livesley, Joseph C. von Fischer and Stefan K. Arndt.
2015a. "Repeated fuel reduction burns have little long-term impact on soil
greenhouse gas exchange in a dry sclerophyll eucalypt forest." Agricultural
and Forest Meteorology 201: 17-25.
http://www.sciencedirect.com/science/article/pii/S0168192314002743. doi:
http://dx.doi.org/10.1016/j.agrformet.2014.11.006.
Fest, Benedikt, Tim Wardlaw, Stephen J Livesley, Thomas J Duff and Stefan K
Arndt. 2015b. "Changes in soil moisture drive soil methane uptake along a
fire regeneration chronosequence in a eucalypt forest landscape." Global
change biology 21 (11): 4250-4264.
Forster, Clive. 2006. "The challenge of change: Australian cities and urban planning
in the new millennium." Geographical research 44 (2): 173-182.
Bibliography
187
Giles, Madeline, Nicholas Morley, Elizabeth M Baggs and Tim J Daniell. 2012.
"Soil nitrate reducing processes-drivers, mechanisms for spatial variation, and
significance for nitrous oxide production." Frontiers in microbiology 3: 407.
Golubiewski, N. E. 2006. "Urbanization increases grassland carbon pools: Effects of
landscaping in Colorado's front range." Ecological Applications 16 (2): 555-
571. <Go to ISI>://WOS:000237052200011. doi: 10.1890/1051-
0761(2006)016[0555:uigcpe]2.0.co;2.
Grace, P. R., J. Antle, S. Ogle, K. Paustian and B. Basso. 2010. "Soil carbon
sequestration rates and associated economic costs for farming systems of
south-eastern Australia." Australian Journal of Soil Research 48 (8): 720-
729. <Go to ISI>://WOS:000284360600008. doi: 10.1071/sr10063.
Grace, P. R. and B. Basso. 2012. "Offsetting greenhouse gas emissions through
biological carbon sequestration in North Eastern Australia." Agricultural
Systems 105 (1): 1-6. <Go to ISI>://WOS:000299191900001. doi:
10.1016/j.agsy.2011.08.006.
Grimm, Nancy B, David Foster, Peter Groffman, J Morgan Grove, Charles S
Hopkinson, Knute J Nadelhoffer, Diane E Pataki and Debra PC Peters. 2008.
"The changing landscape: ecosystem responses to urbanization and pollution
across climatic and societal gradients." Frontiers in Ecology and the
Environment 6 (5): 264-272.
Grimm, Nancy B, J Grove Grove, Steward TA Pickett and Charles L Redman. 2000.
"Integrated Approaches to Long-Term Studies of Urban Ecological Systems
Urban ecological systems present multiple challenges to ecologists—
pervasive human impact and extreme heterogeneity of cities, and the need to
integrate social and ecological approaches, concepts, and theory." BioScience
50 (7): 571-584.
Grimm, Nancy B; Faeth, Stanley H; Golubiewski, Nancy E; Redman, Charles L; Wu,
Jianguo; Bai, Xuemei; Briggs, John M. 2008. "Global change and the ecology
of cities." science 319 (5864): 756-760.
Groffman, P. M. and R. V. Pouyat. 2009. "Methane Uptake in Urban Forests and
Lawns." Environmental Science & Technology 43 (14): 5229-5235. <Go to
ISI>://WOS:000268138000013. doi: 10.1021/es803720h.
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."
Journal of Environmental Quality 38 (5): 1848-1860. <Go to
ISI>://WOS:000269627400008. doi: 10.2134/jeq2008.0521.
188
Groffman, Peter M, Richard V Pouyat, Mark J McDonnell, Steward TA Pickett and
Wayne C Zipperer. 1995. "Carbon pools and trace gas fluxes in urban forest
soils." In Soil Management and Greenhouse Effect, 147-58.
Grover, S. P. P., S. J. Livesley, L. B. Hutley, H. Jamali, B. Fest, J. Beringer, K.
Butterbach-Bahl and S. K. Arndt. 2012. "Land use change and the impact on
greenhouse gas exchange in north Australian savanna soils." Biogeosciences
9 (1): 423-437. <Go to ISI>://WOS:000300229000029. doi: 10.5194/bg-9-
423-2012.
Gu, Chuanhui, John Crane Ii, George Hornberger and Amanda Carrico. 2015. "The
effects of household management practices on the global warming potential
of urban lawns." Journal of Environmental Management 151: 233-242.
http://www.sciencedirect.com/science/article/pii/S0301479715000092. doi:
http://dx.doi.org/10.1016/j.jenvman.2015.01.008.
Gullison, R. E., P. C. Frumhoff, J. G. Canadell, C. B. Field, D. C. Nepstad, K.
Hayhoe, R. Avissar, et al. 2007. "Tropical forests and climate policy."
Science 316 (5827): 985-986. <Go to ISI>://WOS:000246554000029. doi:
10.1126/science.1136163.
Guo, Lanbin B and RM Gifford. 2002. "Soil carbon stocks and land use change: a
meta analysis." Global change biology 8 (4): 345-360.
Hall, S. J., D. Huber and N. B. Grimm. 2008. "Soil N2O and NO emissions from an
arid, urban ecosystem." Journal of Geophysical Research-Biogeosciences
113 (G1). <Go to ISI>://WOS:000253531900001. doi:
10.1029/2007jg000523.
Haney, RL and EB Haney. 2010. "Simple and rapid laboratory method for rewetting
dry soil for incubations." Communications in soil science and plant analysis
41 (12): 1493-1501.
Hart, Stephen C, John M Stark, Eric A Davidson and Mary K Firestone. 1994.
"Nitrogen mineralization, immobilization, and nitrification." Methods of Soil
Analysis: Part 2—Microbiological and Biochemical Properties
(methodsofsoilan2): 985-1018.
Hatfield-Dodds, Steve, Heinz Schandl, Philip D. Adams, Timothy M. Baynes,
Thomas S. Brinsmead, Brett A. Bryan, Francis H. S. Chiew, et al. 2015.
"Australia is ‘free to choose’ economic growth and falling environmental
pressures." Nature 527 (7576): 49-53. http://dx.doi.org/10.1038/nature16065.
doi: 10.1038/nature16065
http://www.nature.com/nature/journal/v527/n7576/abs/nature16065.html#sup
plementary-information.
Bibliography
189
Hbirkou, C, C Martius, A Khamzina, JPA Lamers, G Welp and W Amelung. 2011.
"Reducing topsoil salinity and raising carbon stocks through afforestation in
Khorezm, Uzbekistan." Journal of Arid Environments 75 (2): 146-155.
Henderson, Benjamin B, Pierre J Gerber, Tom E Hilinski, Alessandra Falcucci,
Dennis S Ojima, Mirella Salvatore and Richard T Conant. 2015. "Greenhouse
gas mitigation potential of the world’s grazing lands: Modeling soil carbon
and nitrogen fluxes of mitigation practices." Agriculture, Ecosystems &
Environment 207: 91-100.
Hutyra, L. R., B. Yoon and M. Alberti. 2011. "Terrestrial carbon stocks across a
gradient of urbanization: a study of the Seattle, WA region." Global Change
Biology 17 (2): 783-797. <Go to ISI>://WOS:000285878000012. doi:
10.1111/j.1365-2486.2010.02238.x.
IPCC. 2001. "Climate change 2001: The scientific basis. Contribution of Working
Group I of the Intergovernmental Panel on Climate Change." Cambridge
University Press: Cambridge.
IPCC. 2006. Chapter 8 Settlements in Guidelines for National Greenhouse Gas
Inventories: Agriculture, Forestry and Other Land Use. Edited by Hector
Daniel Ginzo (Argentina) Jennifer C. Jenkins (USA), Stephen M. Ogle
(USA), Louis V. Verchot (ICRAF/USA), Mariko Handa (Japan), Atsushi
Tsunekawa (Japan). Vol. 4. http://www.ipcc-
nggip.iges.or.jp/public/2006gl/pdf/4_Volume4/V4_08_Ch8_Settlements.pdf.
IPCC. 2007. Climate Change 2007: The Physical Science Basis. Edited by S.
Solomon, D. Qin, M. Manning, M. Marquis, K. Averyt, M. M. B. Tignor, H.
L. Miller and Z. L. Chen, Climate Change 2007: The Physical Science Basis.
New York: Cambridge Univ Press. <Go to ISI>://WOS:000292238900014.
IPCC. 2013. Climate Change 2013: The Physical Science Basis. Contribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental
Panel on Climate Change. Edited by T.F. Stocker, D. Qin, G.-K. Plattner, M.
Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M.
Midgley (eds.). Cambridge University Press, Cambridge, United Kingdom
and New York, NY, USA, 1535 pp,. doi: 10.1017/CBO9781107415324.
IPCC. 2014. "Climate Change 2014: Synthesis Report. Contribution of Working
Groups I, II and III to the Fifth Assessment Report of the Intergovernmental
Panel on Climate Change", edited by R.K. Pachauri and L.A. Meyer (eds.)
Core Writing Team. IPCC, Geneva, Switzerland.
190
Isbell, Raymond. 2002. The Australian soil classification. Vol. 4: CSIRO publishing:
Collingwood.
Jain, Theresa B, Russell T Graham and David L Adams. 1997. "Carbon to organic
matter ratios for soils in Rocky Mountain coniferous forests." Soil Science
Society of America Journal 61 (4): 1190-1195.
Jones, S. K., R. M. Rees, U. M. Skiba and B. C. Ball. 2005. "Greenhouse gas
emissions from a managed grassland." Global and Planetary Change 47 (2-
4): 201-211. <Go to ISI>://WOS:000231753100013. doi:
10.1016/j.gloplacha.2004.10.011.
Kaye, J. P., I. C. Burke, A. R. Mosier and J. P. Guerschman. 2004. "Methane and
nitrous oxide fluxes from urban soils to the atmosphere." Ecological
Applications 14 (4): 975-981. <Go to ISI>://WOS:000223156600002. doi:
10.1890/03-5115.
Kaye, J. P., R. L. McCulley and I. C. Burke. 2005. "Carbon fluxes, nitrogen cycling,
and soil microbial communities in adjacent urban, native and agricultural
ecosystems." Global Change Biology 11 (4): 575-587. <Go to
ISI>://WOS:000228179500005. doi: 10.1111/j1365-2486.2005.00921.x.
Kaye, Jason P, Peter M Groffman, Nancy B Grimm, Lawrence A Baker and Richard
V Pouyat. 2006. "A distinct urban biogeochemistry?" Trends in Ecology &
Evolution 21 (4): 192-199.
Kiese, R. and K. Butterbach-Bahl. 2002. "N2O and CO2 emissions from three
different tropical forest sites in the wet tropics of Queensland, Australia." Soil
Biology & Biochemistry 34 (7): 975-987. <Go to
ISI>://WOS:000176977600009. doi: 10.1016/s0038-0717(02)00031-7.
Knowles, T. A. and B. Singh. 2003. "Carbon storage in cotton soils of northern New
South Wales." Australian Journal of Soil Research 41 (5): 889-903. <Go to
ISI>://WOS:000185266100006. doi: 10.1071/sr02023.
Koerner, B and J Klopatek. 2002. "Anthropogenic and natural CO< sub> 2</sub>
emission sources in an arid urban environment." Environmental Pollution
116: S45-S51.
Koerner, Brenda A. and Jeffrey M. Klopatek. 2010. "Carbon fluxes and nitrogen
availability along an urban-rural gradient in a desert landscape." Urban
Ecosystems 13 (1): 1-21. <Go to ISI>://BIOABS:BACD201000151374. doi:
10.1007/s11252-009-0105-z.
Bibliography
191
Konert, Martin and JEF Vandenberghe. 1997. "Comparison of laser grain size
analysis with pipette and sieve analysis: a solution for the underestimation of
the clay fraction." Sedimentology 44 (3): 523-535.
Kong, L., Z. J. Shi and L. M. Chu. 2014. "Carbon emission and sequestration of
urban turfgrass systems in Hong Kong." Science of the Total Environment
473: 132-138. <Go to ISI>://WOS:000331923900017. doi:
10.1016/j.scitotenv.2013.12.012.
Kroeze, C., A. Mosier, C. Nevison, O. Oenema, S. Seitzinger and O. van Cleemput.
1997. Revised 1996 IPCC guidelines for national greenhouse gas inventories:
Intergovernmental Panel on Climate Change.
Lal, R. 2004a. "Soil carbon sequestration impacts on global climate change and food
security." Science 304 (5677): 1623-1627. <Go to
ISI>://WOS:000221934300036. doi: 10.1126/science.1097396.
Lal, R. 2004b. "Soil carbon sequestration to mitigate climate change." Geoderma 123
(1-2): 1-22. <Go to ISI>://WOS:000224881800001. doi:
10.1016/j.geoderma.2004.01.032.
Leahy, P., G. Kiely and T. M. Scanlon. 2004. "Managed grasslands: A greenhouse
gas sink or source?" Geophysical Research Letters 31 (20). <Go to
ISI>://WOS:000224881700004. doi: 10.1029/2004gl021161.
Livesley, S. J., R. Kiese, P. Miehle, C. J. Weston, K. Butterbach-Bahl and S. K.
Arndt. 2009. "Soil-atmosphere exchange of greenhouse gases in a Eucalyptus
marginata woodland, a clover-grass pasture, and Pinus radiata and Eucalyptus
globulus plantations." Global Change Biology 15 (2): 425-440. <Go to
ISI>://WOS:000262510500011. doi: 10.1111/j.1365-2486.2008.01759.x.
Livesley, Stephen J, Ben J Dougherty, Alison J Smith, Damian Navaud, Luke J
Wylie and Stefan K Arndt. 2010. "Soil-atmosphere exchange of carbon
dioxide, methane and nitrous oxide in urban garden systems: impact of
irrigation, fertiliser and mulch." Urban ecosystems 13 (3): 273-293.
Lorenz, Klaus and Rattan Lal. 2009. "Biogeochemical C and N cycles in urban
soils." Environment international 35 (1): 1-8.
MacLeod, N. D. and F. C. Kearney. 2007. "Adoption of sustainable landscape design
practices on small holdings." Final Report for AG-SIP Project AG14, CSIRO
Sustainable Ecosystems.
192
Maraseni, T. N., G. Cockfield, T. Cadman, G. N. Chen and J. S. Qu. 2012.
"Enhancing the value of multiple use plantations: a case study from southeast
Queensland, Australia." Agroforestry Systems 86 (3): 451-462. <Go to
ISI>://WOS:000310952000015. doi: 10.1007/s10457-012-9506-8.
Marschner, Petra. 2007. "Plant-Microbe Interactions in the Rhizosphere and Nutrient
Cycling." In Nutrient Cycling in Terrestrial Ecosystems, edited by Petra
Marschner and Zdenko Rengel, 159-182: Springer Berlin Heidelberg.
http://dx.doi.org/10.1007/978-3-540-68027-7_6. doi: 10.1007/978-3-540-
68027-7_6.
Marschner, Petra. 2012. "Nutrient Availability in Soils." In Marschner's Mineral
Nutrition of Higher Plants, p. 315. doi: 0-12-384905-5, 978-0-12-384905-2.
McKenzie, Neil, David Jacquier, Ray Isbell and Katharine Brown. 2004. "Australian
Soils and Landscapes : An Illustrated Compendium". Melbourne: CSIRO
Publishing.
McLauchlan, Kendra K, Sarah E Hobbie and Wilfred M Post. 2006. "Conversion
from agriculture to grassland builds soil organic matter on decadal
timescales." Ecological applications 16 (1): 143-153.
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." Environmental Management 36 (3): 426-438. <Go to
ISI>://WOS:000231959000008. doi: 10.1007/s00267-004-0316-2.
Miller, Amy E, William D Bowman and Katharine Nash Suding. 2007. "Plant uptake
of inorganic and organic nitrogen: neighbor identity matters." Ecology 88 (7):
1832-1840.
Mitchell, Elaine, Clemens Scheer, David W Rowlings, Richard T Conant, M
Francesca Cotrufo, Lona van Delden and Peter R Grace. 2016. "The influence
of above-ground residue input and incorporation on GHG fluxes and stable
SOM formation in a sandy soil." Soil Biology and Biochemistry 101: 104-
113.
Moore, G, P Dolling, B Porter and L Leonard. 1998. "Soil acidity in Chapter 5.1 of
Soil Guide: a handbook for understanding and managing agricultural soils,
Moore ed." Bulletin 4343.
Moreton Bay Regional Council. 2011. "Community profile; Samford valley area."
http://profile.id.com.au/Default.aspx?id=311&pg=101&gid=560&type=enum
Bibliography
193
Morris, S. J., R. Conant, N. Mellor, E. Brewer and E. A. Paul. 2010. "Controls on
Soil Carbon Sequestration and Dynamics: Lessons from Land-use Change."
Journal of Nematology 42 (1): 78-83. <Go to ISI>://WOS:000284387600012.
Mosier, A. R., A. D. Halvorson, G. A. Peterson, G. P. Robertson and L. Sherrod.
2005. "Measurement of net global warming potential in three
agroecosystems." Nutrient Cycling in Agroecosystems 72 (1): 67-76. <Go to
ISI>://WOS:000232265900007. doi: 10.1007/s10705-004-7356-0.
Mosier, Arvin, Carolien Kroeze, Cindy Nevison, Oene Oenema, Sybil Seitzinger and
Oswald Van Cleemput. 1998. "Closing the global N2O budget: nitrous oxide
emissions through the agricultural nitrogen cycle." Nutrient cycling in
Agroecosystems 52 (2-3): 225-248.
Myhre, G. D. , D. Shindell, F. M. Bréon, W. Collins, J. Fuglestvedt, J. Huang, D.
Koch, et al. 2013. Anthropogenic and Natural Radiativ Forcing, Climate
Change 2013: The Physical Science Basis. Contribution of Working Group I
to the Fifth Assessment Report of the Intergovernmental Panel on Climate
Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J.
Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge
University Press, Cambridge, United Kingdom and New York, NY, USA.
Neal, J. S., S. M. Eldridge, W. J. Fulkerson, R. Lawrie and I. M. Barchia. 2013.
"Differences in soil carbon sequestration and soil nitrogen among forages
used by the dairy industry." Soil Biology & Biochemistry 57: 542-548. <Go to
ISI>://WOS:000317247100062. doi: 10.1016/j.soilbio.2012.09.019.
Neill, C. and E. A. Davidson. 2000. Soil carbon accumulation or loss following
deforestation for pasture in the Brazilian Amazon, edited by R. Lal, J. M.
Kimble and B. A. Stewart, Global Climate Change and Tropical Ecosystems.
Boca Raton: Crc Press-Taylor & Francis Group. <Go to
ISI>://WOS:000085521600010.
Newham, Michael J, Christine S Fellows and Fran Sheldon. 2011. "Functions of
riparian forest in urban catchments: a case study from sub-tropical Brisbane,
Australia." Urban Ecosystems 14 (2): 165-180.
Ng, B. J. L., L. R. Hutyra, H. Nguyen, A. R. Cobb, F. M. Kai, C. Harvey and L.
Gandois. 2015. "Carbon fluxes from an urban tropical grassland."
Environmental Pollution 203: 227-234.
http://www.sciencedirect.com/science/article/pii/S0269749114002413. doi:
http://dx.doi.org/10.1016/j.envpol.2014.06.009.
194
Ng, BJL, LR Hutyra, H Nguyen, AR Cobb, FM Kai, C Harvey and L Gandois. 2014.
"Carbon fluxes from an urban tropical grassland." Environmental Pollution.
NOAA, National Oceanic & Atmospheric Administration. 2016. "Trends in
Atmospheric Carbon Dioxide - Recent Global CO2": U.S. Department of
Commerce.
Ogle, S. M., R. T. Conant and K. Paustian. 2004. "Deriving grassland management
factors for a carbon accounting method developed by the intergovernmental
panel on climate change." Environmental Management 33 (4): 474-484. <Go
to ISI>://WOS:000222309300006. doi: 10.1007/s00267-003-9105-6.
Oke, Timothy R. 1982. "The energetic basis of the urban heat island." Quarterly
Journal of the Royal Meteorological Society 108 (455): 1-24.
Page, K. L., R. C. Dalal and R. J. Raison. 2011. "The impact of harvesting native
forests on vegetation and soil C stocks, and soil CO2, N2O and CH4 fluxes."
Australian Journal of Botany 59 (7): 653-668. <Go to
ISI>://WOS:000297263400006. doi: 10.1071/bt11207.
Parton, W. J., D. S. Schimel, C. V. Cole and D. S. Ojima. 1987. "Analysis of factors
controlling soil organic-matter levels in great-plains grasslands." Soil Science
Society of America Journal 51 (5): 1173-1179. <Go to
ISI>://WOS:A1987K564900015.
Parton, WJ. 1996. "The CENTURY model." In Evaluation of soil organic matter
models, 283-291: Springer.
Pastor, John and WM Post. 1986. "Influence of climate, soil moisture, and
succession on forest carbon and nitrogen cycles." Biogeochemistry 2 (1): 3-
27.
Pataki, DE, T Xu, YQ Luo and JR Ehleringer. 2007. "Inferring biogenic and
anthropogenic carbon dioxide sources across an urban to rural gradient."
Oecologia 152 (2): 307-322.
Paul, Eldor A. 2006. Soil microbiology, ecology and biochemistry: Academic press.
Paul, K. I., P. J. Polglase, J. G. Nyakuengama and P. K. Khanna. 2002. "Change in
soil carbon following afforestation." Forest Ecology and Management 168
(1-3): 241-257. <Go to ISI>://WOS:000177478900020. doi: 10.1016/s0378-
1127(01)00740-x.
Bibliography
195
Paul, S., H. Flessa, E. Veldkamp and M. Lopez-Ulloa. 2008a. "Stabilization of recent
soil carbon in the humid tropics following land use changes: evidence from
aggregate fractionation and stable isotope analyses." Biogeochemistry 87 (3):
247-263. <Go to ISI>://WOS:000254360400003. doi: 10.1007/s10533-008-
9182-y.
Paul, S., E. Veldkamp and H. Flessa. 2008b. "Soil organic carbon in density fractions
of tropical soils under forest - pasture - secondary forest land use changes."
European Journal of Soil Science 59 (2): 359-371. <Go to
ISI>://WOS:000253758500024. doi: 10.1111/j.1365-2389.2007.01010.x.
Poeplau, Christopher, Thomas Kätterer, Niki IW Leblans and Bjarni D Sigurdsson.
2016. "Sensitivity of soil carbon fractions and their specific stabilization
mechanisms to extreme soil warming in a subarctic grassland." Global
Change Biology.
Potere, David and Annemarie Schneider. 2007. "A critical look at representations of
urban areas in global maps." GeoJournal 69 (1-2): 55-80.
Pouyat, R, P Groffman, I Yesilonis and L Hernandez. 2002. "Soil carbon pools and
fluxes in urban ecosystems." Environmental pollution 116: S107-S118.
Pouyat, Richard V, Ian D Yesilonis and David J Nowak. 2006. "Carbon storage by
urban soils in the United States." Journal of environmental quality 35 (4):
1566-1575.
Pouyat, Richard V., Ian D. Yesilonis and Nancy E. Golubiewski. 2009. "A
comparison of soil organic carbon stocks between residential turf grass and
native soil." Urban Ecosystems 12 (1): 45-62. <Go to
ISI>://BIOABS:BACD200900147772. doi: 10.1007/s11252-008-0059-6.
Qian, Y. L., R. F. Follett and J. M. Kimble. 2010. "Soil Organic Carbon Input from
Urban Turfgrasses." Soil Science Society of America Journal 74 (2): 366-371.
<Go to ISI>://WOS:000275187300004. doi: 10.2136/sssaj2009.0075.
Raciti, S. M., P. M. Groffman, J. C. Jenkins, R. V. Pouyat, T. J. Fahey, S. T. A.
Pickett and M. L. Cadenasso. 2011a. "Accumulation of carbon and nitrogen
in residential soils with different land-use histories." Ecosystems 14 (2): 287-
297. <Go to ISI>://WOS:000288172300009. doi: 10.1007/s10021-010-9409-
3.
Raciti, Steve M., Amy J. Burgin, Peter M. Groffman, David N. Lewis and Timothy J.
Fahey. 2011b. "Denitrification in Suburban Lawn Soils." J. Environ. Qual. 40
(6): 1932-1940.
196
https://dl.sciencesocieties.org/publications/jeq/abstracts/40/6/1932. doi:
10.2134/jeq2011.0107.
Raupach, M. R., J. G. Canadell, P. Ciais, P. Friedlingstein, P. J. Rayner and C. M.
Trudinger. 2011. "The relationship between peak warming and cumulative
CO2 emissions, and its use to quantify vulnerabilities in the carbon-climate-
human system." Tellus Series B-Chemical and Physical Meteorology 63 (2):
145-164. <Go to ISI>://WOS:000288516400001. doi: 10.1111/j.1600-
0889.2010.00521.x.
Rayment, GE and Francis Ross Higginson. 1992. Australian laboratory handbook of
soil and water chemical methods: Inkata Press Pty Ltd.
Richards, A. E., R. C. Dalal and S. Schmidt. 2007. "Soil carbon turnover and
sequestration in native subtropical tree plantations." Soil Biology &
Biochemistry 39 (8): 2078-2090. <Go to ISI>://WOS:000247295800024. doi:
10.1016/j.soilbio.2007.03.012.
Robertson, G. Philip and Peter R. Grace. 2004. "Greenhouse gas fluxes in tropical
and temperate agriculture: The need for a full-cost accounting of global
warming potentials." Environment Development and Sustainability 6 (1-2):
51-63. <Go to ISI>://BIOABS:BACD200400335077. doi:
10.1023/B:ENVI.0000003629.32997.9e.
Robertson, G. Philip and P. Groffman. 2007. "Nitrogen transformations." Paul EA
CF, ed. Soil Microbiology, Ecology, and Biochemistry. Oxford, UK: Elsevier
Academic Press: 341-364.
Rowlings, D. W., P. R. Grace, R. Kiese and K. L. Weier. 2012. "Environmental
factors controlling temporal and spatial variability in the soil-atmosphere
exchange of CO2, CH4 and N2O from an Australian subtropical rainforest."
Global Change Biology 18 (2): 726-738. <Go to
ISI>://WOS:000299042500027. doi: 10.1111/j.1365-2486.2011.02563.x.
Rowlings, DW, PR Grace, C Scheer and R Kiese. 2013. "Influence of nitrogen
fertiliser application and timing on greenhouse gas emissions from a lychee
(Litchi chinensis) orchard in humid subtropical Australia." Agriculture,
ecosystems & environment 179: 168-178.
Rowlings, DW, PR Grace, C Scheer and S Liu. 2015. "Rainfall variability drives
interannual variation in N 2 O emissions from a humid, subtropical pasture."
Science of The Total Environment 512: 8-18.
Satterthwaite, LN, AW Hodges, JJ Haydu and JL Cisar. 2009. "iAn Agronomic and
Economic Profile of Florida’s Sod Industry in 2007.
Bibliography
197
Scalenghe, Riccardo and Franco Ajmone Marsan. 2009. "The anthropogenic sealing
of soils in urban areas." Landscape and Urban Planning 90 (1): 1-10.
http://www.sciencedirect.com/science/article/pii/S0169204608001710. doi:
http://dx.doi.org/10.1016/j.landurbplan.2008.10.011.
Schaufler, G., B. Kitzler, A. Schindlbacher, U. Skiba, M. A. Sutton and S.
Zechmeister-Boltenstern. 2010. "Greenhouse gas emissions from European
soils under different land use: effects of soil moisture and temperature."
European Journal of Soil Science 61 (5): 683-696. <Go to
ISI>://WOS:000282474200007. doi: 10.1111/1.1365-2389.2010.01277.x.
Scheer, C., S. J. Del Grosso, W. J. Parton, D. W. Rowlings and P. R. Grace. 2014a.
"Modeling nitrous oxide emissions from irrigated agriculture: testing
DayCent with high-frequency measurements." Ecological Applications 24
(3): 528-538. <Go to ISI>://WOS:000333242300009. doi: 10.1890/13-
0570.1.
Scheer, C., P. R. Grace, D. W. Rowlings, S. Kimber and L. Van Zwieten. 2011.
"Effect of biochar amendment on the soil-atmosphere exchange of
greenhouse gases from an intensive subtropical pasture in northern New
South Wales, Australia." Plant and Soil 345 (1-2): 47-58. <Go to
ISI>://WOS:000292999700005. doi: 10.1007/s11104-011-0759-1.
Scheer, C., P. R. Grace, D. W. Rowlings and J. Payero. 2013. "Soil N2O and CO2
emissions from cotton in Australia under varying irrigation management."
Nutrient Cycling in Agroecosystems 95 (1): 43-56. <Go to
ISI>://WOS:000314334300003. doi: 10.1007/s10705-012-9547-4.
Scheer, Clemens, David W Rowlings, Mary Firrel, Peter Deuter, Stephen Morris and
Peter R Grace. 2014b. "Impact of nitrification inhibitor (DMPP) on soil
nitrous oxide emissions from an intensive broccoli production system in sub-
tropical Australia." Soil Biology and Biochemistry 77: 243-251.
Scheer, Clemens, Reiner Wassmann, Kirsten Kienzler, Nazar Ibragimov and
Ruzimboy Eschanov. 2008. "Nitrous oxide emissions from fertilized,
irrigated cotton (< i> Gossypium hirsutum</i> L.) in the Aral Sea Basin,
Uzbekistan: Influence of nitrogen applications and irrigation practices." Soil
Biology and Biochemistry 40 (2): 290-301.
Schlesinger, W. H. 1990. "Evidence from chronosequence studies for a low carbon-
storage potential of soils." Nature 348 (6298): 232-234. <Go to
ISI>://WOS:A1990EH79600056. doi: 10.1038/348232a0.
198
Schlesinger, William H. 1995. "An overview of the carbon cycle." Soils and global
change 25.
Selhorst, A. L. and R. Lal. 2011. "Carbon budgeting in golf course soils of Central
Ohio." Urban Ecosystems 14 (4): 771-781. <Go to
ISI>://WOS:000297670700015. doi: 10.1007/s11252-011-0168-5.
Selhorst, Adam and Rattan Lal. 2013. "Net carbon sequestration potential and
emissions in home lawn turfgrasses of the United States." Environmental
management 51 (1): 198-208.
Shi, Wei, Huaiying Yao and Daniel Bowman. 2006. "Soil microbial biomass, activity
and nitrogen transformations in a turfgrass chronosequence." Soil Biology
and Biochemistry 38 (2): 311-319.
http://www.sciencedirect.com/science/article/pii/S0038071705002026. doi:
http://dx.doi.org/10.1016/j.soilbio.2005.05.008.
Siregar, A, M Kleber, R Mikutta and R Jahn. 2005. "Sodium hypochlorite oxidation
reduces soil organic matter concentrations without affecting inorganic soil
constituents." European Journal of Soil Science 56 (4): 481-490.
Six, J., R. T. Conant, E. A. Paul and K. Paustian. 2002. "Stabilization mechanisms of
soil organic matter: Implications for C-saturation of soils." Plant and Soil 241
(2): 155-176. <Go to ISI>://WOS:000176778900001. doi:
10.1023/a:1016125726789.
Six, JΑΕΤ, ET Elliott and K Paustian. 2000. "Soil macroaggregate turnover and
microaggregate formation: a mechanism for C sequestration under no-tillage
agriculture." Soil Biology and Biochemistry 32 (14): 2099-2103.
Skjemstad, JO, LR Spouncer, B Cowie and RS Swift. 2004. "Calibration of the
Rothamsted organic carbon turnover model (RothC ver. 26.3), using
measurable soil organic carbon pools." Soil Research 42 (1): 79-88.
Smith, KA, KE Dobbie, BC Ball, LR Bakken, BK Sitaula, S Hansen, R Brumme, W
Borken, Søren Christensen and Anders Priemé. 2000. "Oxidation of
atmospheric methane in Northern European soils, comparison with other
ecosystems, and uncertainties in the global terrestrial sink." Global Change
Biology 6 (7): 791-803.
Spencer, A., J. Gill, L. Schmahmann. 2015. "Urban or suburban? Examining the
density of Australian cities in a global context." Paper presented at the State
of Australian Cities Conference 2015.
Bibliography
199
Stewart, C. E., K. Paustian, R. T. Conant, A. F. Plante and J. Six. 2007. "Soil carbon
saturation: concept, evidence and evaluation." Biogeochemistry 86 (1): 19-31.
<Go to ISI>://WOS:000249402000002. doi: 10.1007/s10533-007-9140-0.
Stewart, C. E., A. F. Plante, K. Paustian, R. T. Conant and J. Six. 2008. "Soil carbon
saturation: Linking concept and measurable carbon pools." Soil Science
Society of America Journal 72 (2): 379-392. <Go to
ISI>://WOS:000254060200011. doi: 10.2136/sssaj2007.0104.
Stocker, Thomas F, Q Dahe, G.-K. Plattner, M Tignor, S.K. Allen, J. Boschung, A.
Nauels, Y. Xia, V. Bex and P.M. Midgley. 2013. "Climate Change 2013: The
Physical Science Basis." Working Group I Contribution to the Fifth
Assessment Report of the Intergovernmental Panel on Climate Change
(IPCC, 2013). Cambridge University Press, Cambridge, United Kingdom and
New York, NY, USA.
Thomsen, Ingrid K, Sander Bruun, Lars S Jensen and Bent T Christensen. 2009.
"Assessing soil carbon lability by near infrared spectroscopy and NaOCl
oxidation." Soil Biology and Biochemistry 41 (10): 2170-2177.
Tian, Hanqin, Guangsheng Chen, Chaoqun Lu, Xiaofeng Xu, Daniel J Hayes, Wei
Ren, Shufen Pan, Deborah N Huntzinger and Steven C Wofsy. 2014. "North
American terrestrial CO2 uptake largely offset by CH4 and N2O emissions:
toward a full accounting of the greenhouse gas budget." Climatic Change: 1-
14.
Tirol-Padre, A and JK Ladha. 2004. "Assessing the reliability of permanganate-
oxidizable carbon as an index of soil labile carbon." Soil Science Society of
America Journal 68 (3): 969-978.
Tratalos, Jamie, Richard A Fuller, Philip H Warren, Richard G Davies and Kevin J
Gaston. 2007. "Urban form, biodiversity potential and ecosystem services."
Landscape and urban planning 83 (4): 308-317.
Trlica, Andrew and Sally L. Brown. 2013. "Greenhouse gas emissions and the
interrelation of urban and forest sectors in reclaiming one hectare of land in
the Pacific Northwest." Environmental Science & Technology. Accessed
2013/06/02. http://dx.doi.org/10.1021/es3033007. doi: 10.1021/es3033007.
Turf Australia. 2012. "Newsletter 2012." https://www.turfaustralia.com.au/.
Turner, J., M. J. Lambert and D. W. Johnson. 2005. "Experience with patterns of
change in soil carbon resulting from forest plantation establishment in eastern
200
Australia." Forest Ecology and Management 220 (1-3): 259-269. <Go to
ISI>://WOS:000233365300018. doi: 10.1016/j.foreco.2005.08.025.
UNFPA. 2011. "State of the world population 2011", edited by United Nations
Population Fund. New York, New York, USA.
United Nations. 2008. "World urbanization prospects: The 2007 Revision", edited by
Economics and Social Affairs. United Nations (NY): United Nations
Population Division.
United Nations. 2013. "World Population Prospects: The 2012 Revision, Highlights
ans Advance Tables", edited by Economics and Social Affairs. United
Nations (NY): United Nations Population Division.
United Nations. 2014. "World Population Prospects: The 2014 Revision, Highlights",
edited by Economics and Social Affairs. United Nations (NY): United
Nations Population Division.
van Delden, L., D. W. Rowlings, C. Scheer and P. R. Grace. 2016b. "Land use
change associated with urbanization modifies soil nitrogen cycling and
increases N2O emissions." Biogeosciences Discuss. 2016: 1-23.
http://www.biogeosciences-discuss.net/bg-2016-216/. doi: 10.5194/bg-2016-
216.
van Delden, Lona, Eloise Larsen, David Rowlings, Clemens Scheer and Peter Grace.
2016a. "Establishing turf grass increases soil greenhouse gas emissions in
peri-urban environments." Urban Ecosystems: 1-14.
van Delden, Lona, David W Rowlings, Clemens Scheer and Peter R Grace. 2016.
"Urbanisation-related land use change from forest and pasture into turf grass
modifies soil nitrogen cycling and increases N 2 O emissions."
Biogeosciences 13 (21): 6095-6106.
van Lent, J, K Hergoualc'h and LV Verchot. 2015. "Reviews and syntheses: Soil
N2O and NO emissions from land use and land use change in the tropics and
subtropics: a meta-analysis." Biogeosciences 12 (15).
Veldkamp, E. 1994. "Organic Carbon Turnover in Three Tropical Soils under
Pasture after Deforestation." Soil Sci. Soc. Am. J. 58 (1): 175-180.
https://www.soils.org/publications/sssaj/abstracts/58/1/175. doi:
10.2136/sssaj1994.03615995005800010025x.
Vellinga, T. V., A. van den Pol-van Dasselaar and P. J. Kuikman. 2004. "The impact
of grassland ploughing on CO2 and N2O emissions in the Netherlands."
Bibliography
201
Nutrient Cycling in Agroecosystems 70 (1): 33-45. <Go to
ISI>://WOS:000225103200004. doi:
10.1023/B:FRES.0000045981.56547.db.
Verchot, Louis V., Eric A. Davidson, Henrique Cattânio, Ilse L. Ackerman, Heather
E. Erickson and Michael Keller. 1999. "Land use change and biogeochemical
controls of nitrogen oxide emissions from soils in eastern Amazonia." Global
Biogeochemical Cycles 13 (1): 31-46.
http://dx.doi.org/10.1029/1998GB900019. doi: 10.1029/1998GB900019.
Viscarra Rossel, Raphael A., Richard Webster, Elisabeth N. Bui and Jeff A. Baldock.
2014. "Baseline map of organic carbon in Australian soil to support national
carbon accounting and monitoring under climate change." Global Change
Biology 20 (9): 2953-2970. http://dx.doi.org/10.1111/gcb.12569. doi:
10.1111/gcb.12569.
Vollenweider, Richard A. 1970. Scientific fundamentals of the eutrophication of
lakes and flowing waters, with particular reference to nitrogen and
phosphorus as factors in eutrophication: OECD Paris.
Wang, Yi, Cong Tu, Chunyue Li, Lane Tredway, David Lee, Mark Snell, Xingchang
Zhang and Shuijin Hu. 2014. "Turfgrass Management Duration and
Intensities Influence Soil Microbial Dynamics and Carbon Sequestration."
International Journal of Agriculture and Biology 16 (1): 139-145.
Werner, C, K Butterbach‐Bahl, E Haas, Thomas Hickler and R Kiese. 2007. "A
global inventory of N2O emissions from tropical rainforest soils using a
detailed biogeochemical model." Global Biogeochemical Cycles 21 (3).
Werner, Christian, Xunhua Zheng, Janwei Tang, Baohua Xie, Chunyan Liu, Ralf
Kiese and Klaus Butterbach-Bahl. 2006. "N2O, CH4 and CO2 emissions from
seasonal tropical rainforests and a rubber plantation in Southwest China."
Plant and Soil 289 (1-2): 335-353.
WRB, IUSS Working Group. 2015. World Reference Base for Soil Resources 2014,
update 2015. Vol. World Soil Resources Reports No. 106., International soil
classification system for naming soils and creating legends for soil maps.:
FAO, Rome.
Xu, Yongbo, Zhihong Xu, Zucong Cai and Frédérique Reverchon. 2013. "Review of
denitrification in tropical and subtropical soils of terrestrial ecosystems."
Journal of Soils and Sediments 13 (4): 699-710.
202
Yashiro, Yuichiro, Wan Rashidah Kadir, Toshinori Okuda and Hiroshi Koizumi.
2008. "The effects of logging on soil greenhouse gas (CO2, CH4, N2O) flux in
a tropical rain forest, Peninsular Malaysia." Agricultural and Forest
Meteorology 148 (5): 799-806.
http://www.sciencedirect.com/science/article/pii/S0168192308000348. doi:
http://dx.doi.org/10.1016/j.agrformet.2008.01.010.
Zhang, Yao, Yaling Qian, Dale J Bremer and Jason P Kaye. 2013a. "Simulation of
Nitrous Oxide Emissions and Estimation of Global Warming Potential in
Turfgrass Systems Using the DAYCENT Model." Journal of environmental
quality 42 (4): 1100-1108.
Zhang, Yao, Yaling Qian, Brent Mecham and William J Parton. 2013b.
"Development of best turfgrass management practices using the DAYCENT
model." Agronomy Journal 105 (4): 1151-1159.
Appendix
203
Appendix
Figure A 1 Percentage of the population in urban areas, 2007, 2025 and 2050 (United
Nations 2008).
204
Figure A 2 Major cities of Australia (Commonwealth of Australia 2013).
Figure A 3 Population distribution of selected countries; Source: Ellis in
Commonwealth of Australia (2013.
Appendix
205
Figure A 4 Population growth rates of OECD countries, 2000–10; Source: OECD
2012 in Commonwealth of Australia (2013.
206
Figure A 5 Principal global carbon pools (Lal 2004b).
Figure A 6 Temperature change forecast for Australia from Appendix I in Stocker et
al. (2013.
Appendix
207
Figure A 7 Australian Supersite Network (ASN) locations. Samford Ecological
Research Facility (SERF) is located at South East Queensland (SEQ) and is the only
peri-urban supersite in Australia.
Figure A 8 Global distribution of Planosols aka Chromosols by FAO/UNESCO
(1998.
208
Figure A 9 Typical soil profile of a Brown Chromosol defined by the Australian Soil
Classification (CSIRO 1996; Isbell 2002).
Figure A 10 Representative Australian soil types with their SOC content (Baldock et
al. 2012).
Appendix
209
Figure A 11 Core site plot plan with automatic chambers organised in 3 measurement
sets.
210
Figure A 12 ARIMA modeled confidence interval for CH4 and N2O fluxes over the
experimental timeframe from June 2013 to June 2015 for the forest, pasture and turf
grass (lawn) land use.