IMPLICATIONS OF URBANIZATION RELATED LAND USE CHANGE ON THE CARBON AND NITROGEN CYCLE ... Delden... · 2017-07-12 · IMPLICATIONS OF URBANIZATION RELATED LAND USE CHANGE ON THE CARBON

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.


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.


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.


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


Peter R. Grace Aided experimental design and data analysis, and


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


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.


4.3.1 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


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).


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.


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


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.


85


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


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.


89

4.5 Discussion


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


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










and



requirements.


Chapter 5: Urbanization-related land use change from forest and pasture into turf

grass modifies soil nitrogen cycling and increases N2O emissions (Paper 2)





Signature

05/07/2017







Chapter 5 (Paper 2) has been published in Biogeosciences in November 2016, Volume 13,

Issue 21, pp 6095-6106.

94



authorship.


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


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.




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


99

classified as brown Chromosol according to the Australian soil classification (Isbell

2002) and Planosol according to the World Reference Base (WRB 2015).


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.


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


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.


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

.


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.


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.


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.


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


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.


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


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).


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.

http://www.n2o.net.au/knb/metacat/vandelden.3.3/html

118










and



requirements.


Chapter 6: Soil N2O and CH4 fluxes from urbanization related land use change: From

forest to pasture and turf grass (Paper 3)





Signature

05/07/2017





Daniele de Rosa Aided data analysis, and reviewed the manuscript.



Chapter 5 (Paper 3) has been submitted to Global Change Biology.

119



authorship.


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


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


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.


125




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).


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).


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).


127


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.


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


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


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


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


Te

mp

era

ture

[C

°]

5

10

15

20

25

30

WF

PS

[%

]

0

20

40

60

80

100TemperatureWFPS

a

b

c


from the agricultural pasture soil with supporting environmental parameters (c) mean

daily temperature and water filled pore space (WFPS).


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


Te

mp

era

ture

[C

°]

5

10

15

20

25

30

WF

PS

[%

]

0

20

40

60

80

100TemperatureWFPS

a

b

c


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.


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


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).


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


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.

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


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.


149

7.3 Material and Methods


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.


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


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.


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.


155

0-10 cm

SO

C [

t C

ha

-1]

0

10

20

30

40

CA

CS

CR

10-20 cm


A B


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

http://www.supersites.net.au/knb

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


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


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


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.

http://www.sciencedirect.com/science/article/pii/S1877343511000893

http://dx.doi.org/10.1016/j.cosust.2011.08.011

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://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.

http://dx.doi.org/10.2134/agronj2008.0078


http://dx.doi.org/10.1016/j.agwat.2005.07.011

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://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.



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.

http://dx.doi.org/10.1111/gcb.12238


http://dx.doi.org/10.1016/j.agrformet.2008.09.007



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://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.


http://dx.doi.org/10.1016/j.jenvman.2015.01.008

http://dx.doi.org/10.1038/nature16065

http://www.nature.com/nature/journal/v527/n7576/abs/nature16065.html#supplementary-information

http://www.nature.com/nature/journal/v527/n7576/abs/nature16065.html#supplementary-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.

http://www.ipcc-nggip.iges.or.jp/public/2006gl/pdf/4_Volume4/V4_08_Ch8_Settlements.pdf

http://www.ipcc-nggip.iges.or.jp/public/2006gl/pdf/4_Volume4/V4_08_Ch8_Settlements.pdf

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

http://dx.doi.org/10.1007/978-3-540-68027-7_6

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://dx.doi.org/10.1016/j.envpol.2014.06.009.


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.

https://dl.sciencesocieties.org/publications/jeq/abstracts/40/6/1932

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://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.


http://dx.doi.org/10.1016/j.landurbplan.2008.10.011

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://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.


http://dx.doi.org/10.1016/j.soilbio.2005.05.008

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

http://dx.doi.org/10.1021/es3033007

https://www.turfaustralia.com.au/

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."

http://www.biogeosciences-discuss.net/bg-2016-216/

https://www.soils.org/publications/sssaj/abstracts/58/1/175

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.

http://dx.doi.org/10.1029/1998GB900019

http://dx.doi.org/10.1111/gcb.12569

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.



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.

Documents

IMPLICATIONS OF URBANIZATION RELATED LAND USE CHANGE ON THE CARBON AND NITROGEN CYCLE ... Delden... · 2017-07-12 · IMPLICATIONS OF URBANIZATION RELATED LAND USE CHANGE ON THE CARBON