Nitrogen Cycling Processes and Microbial Communities In

NITROGEN CYCLING PROCESSES AND MICROBIAL COMMUNITIES IN

RECONSTRUCTED OIL-SANDS SOILS

Jacynthe Masse

B.Sc., Université de Montréal, 2009

M.Sc., Université de Montréal, 2011

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF

THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES

(Forestry)

THE UNIVERSITY OF BRITISH COLUMBIA

(Vancouver)

July 2016

Covering 140,200 square km, the Athabasca Oil Sands deposit in Alberta is one of the largest single oil deposits in the world. Following surface mining, companies are required to restore soil- like profiles that can support the previous land capabilities. The overall objective of this thesis was to measure, compare and understand processes underlying nitrogen cycling rates and microbial communities in 20- to 30- year-old reconstructed oil-sands soils and in natural boreal- forest soils. The use of 15N tracer methods in combination with massively parallel sequencing techniques of the 16S and ITS genes identified key dissimilarities between reconstructed and

+ natural boreal-forest soils. In reconstructed soils, NH4 was mainly cycled through the

+ recalcitrant organic-N pool. In natural soils, NH4 was produced from the recalcitrant organic-N pool, but predominantly consumed in the labile organic-N pool, suggesting greater prominence of microbial N-cycling activity in the natural soils compared to the reconstructed soils.

- Reconstructed soils also produced more NO3 than they immobilized it resulting in net nitrification rates. Prokaryotic and fungal β-diversity, but not α-diversity, differed between reconstructed and natural forest soils. Microorganisms associated with a copiotrophic lifestyle were more abundant in reconstructed soils, whereas microorganisms associated with an oligotrophic lifestyle were more abundant in natural forest soils. Vegetation cover was the main factor influencing prokaryotic and fungal α-diversity in reconstructed and natural forest soils.

Nitrogen deposition, pH, soil nutrient content and plant cover influenced prokaryotic and fungal

β-diversity. The results of this thesis deepen our understanding of the distinct pedological environments of oil-sands reconstructed soils and highlighted the importance of above- and below-ground interactions in reconstructed and natural ecosystems. ii

Preface

This study stemmed from an NSERC Collaborative Research and Development grant (PI Sylvie

Quideau). As such, Sylvie Quideau, Cindy Prescott and Sue Grayston have identified key research questions and helped in the development of the general research design.

Together with Jeff Anderson, Jacynthe Masse selected the specific research sites. Sampling design, soil description and soil collection was done by Jacynthe Masse with the help of Jeff

Anderson, Meghan Laidlaw and research assistants (Hannah Mei and Eli Rechtschaffen).

Vegetation survey on all sites was done solely by Jeff Anderson. Meghan Laidlaw was the main researcher to carry the microbial biomass measurements.

Laboratory work was done by Jacynthe Masse with the help of research assistants. Sébastien

Renaut suggested using the massively parallel sequencing techniques for microbial community analyses. Statistical work, data analyses and discussion of the results were done by Jacynthe

Masse.

Chapters were written by Jacynthe Masse with manuscript edits provided by Cindy Prescott and

Sue Grayston.

Chapter 2 has been accepted for publication in Geoderma.

Masse J., Prescott, C.E., Müller, C., Grayston, S.J. (accepted Geoderma) “Gross nitrogen transformation rates differ in reconstructed oil-sand soils from natural boreal forest soils as revealed using a 15N tracing method”

iii

• Jacynthe Masse planned the experimental design, did the soil description, soil sampling

and laboratory analyses. She also performed the statistical work, data analyses and wrote

the manuscript

• Cindy Prescott helped with the experimental design and provided manuscript edits.

• Christoph Müller designed the original 15N tracing model and used the data provided by

J.Masse to model the gross rates of N transformation. He also provided manuscript edits

• Sue Grayston helped with the experimental design and provided manuscript edits.

Chapter 2 will be submitted: Masse J., Prescott, C.E., Renaut S., Terrat, Y., Grayston S.J.

“Vegetation cover and nitrogen deposition as drivers of α- and β- prokaryotic diversity in reconstructed oil-sand soils and in natural boreal forest soils”

• Jacynthe Masse planned the experimental design, did the soil description, soil sampling

and laboratory analyses. She also performed the statistical work, data analyses and wrote

the manuscript

• Cindy Prescott helped with the experimental design and provided manuscript edits.

• Sébastien Renaut suggested the use of massively parallel techniques to analyze the

microbial community and provided help with the bioinformatics work and with the

manuscript edits

• Yves Terrat provided helped with bioinformatics work and manuscript edits

• Sue Grayston helped with the experimental design and provided manuscript edits.

Table of Contents

Abstract ...... ii

Preface ...... iii

Table of Contents ...... v

List of Tables ...... xi

List of Figures ...... xii

Acknowledgements ...... xvii

Dedication ...... xix

Chapter 1: Introduction ...... 1

1.1 The oil sands in Canada ...... 1

1.1.1 The resource ...... 2

1.1.2 The history of the oil sands development in Alberta ...... 2

1.1.3 The challenge of ecological restoration in the oil sands ...... 5

1.2 Nitrogen cycle ...... 6

1.2.1 Nitrogen fixation ...... 8

1.2.2 Depolymerization ...... 11

1.2.3 Mineralization ...... 12

1.2.4 Nitrification ...... 13

1.2.5 Denitrification ...... 15

1.2.6 Other nitrogen transformations in soil ...... 17

1.3 Measuring nitrogen cycling processes ...... 19

1.3.1 Measuring pools of N ...... 19 v

1.3.2 Measuring net fluxes of N ...... 20

1.3.3 Measuring gross fluxes of N using analytical methods ...... 20

1.3.4 Measuring gross fluxes of N using analytical methods ...... 22

1.3.5 Comparison between analytical and numerical methods to measure gross fluxes of N ...... 24

1.4 The microbial role in the nitrogen cycle: does diversity matter? ...... 25

1.5 Impacts of tree species on soil processes and microbial communities ...... 27

1.6 Impacts of forest fire on the nitrogen cycle and soil microbial communities ...... 29

1.7 Restoration ecology ...... 31

1.8 Reclamation of soil nitrogen cycling processes in the Athabasca Region ...... 32

1.9 Objectives and hypotheses ...... 34

Chapter 2: Gross nitrogen transformation rates differ in reconstructed oil-sands soils and natural boreal-forest soils as revealed using a 15N tracing method ...... 37

2.1 Introduction ...... 37

2.2 Materials and methods ...... 41

2.2.1 Study area ...... 41

2.2.2 Study sites ...... 43

2.2.3 Laboratory analyses ...... 45

2.2.4 Quantification of gross rates of N-transformation ...... 46

2.2.5 Calculations and statistics ...... 48

2.3 Results ...... 54

2.3.1 Soil biological, physical and chemical characteristics ...... 54

2.3.2 Soil nitrogen content ...... 55

2.3.3 Gross N transformation rates ...... 57

2.3.4 Net ammonification and nitrification rates ...... 62

2.3.5 Turnover times ...... 63

2.4 Discussion ...... 64

2.4.1 Distinct ammonification cycling processes in reconstructed and natural forest soils ...... 66

2.4.2 Higher net nitrification rates in reconstructed soils ...... 68

2.4.3 Higher carbon content as a cause of higher N-transformation rates in reconstructed soils ...... 68

2.4.4 Explanations for high nitrification rates in two reconstructed soils ...... 71

2.4.5 Limitations of the 15N tracing model to measure gross rates of N-transformation ...... 72

2.4.6 Revegetation treatment had no impact on gross N transformation rates in reconstructed soils ...... 73

2.5 Conclusions ...... 74

Chapter 3: Vegetation cover and nitrogen deposition as drivers of α- and β- prokaryotic diversity in reconstructed oil-sand soils and in natural boreal-forest soils ...... 75

3.1 Introduction ...... 75

3.2 Materials and methods ...... 79

3.2.1 Study area ...... 79

3.2.2 Study sites ...... 80

3.2.3 Laboratory analyses ...... 82

3.2.4 Bioinformatics analyses ...... 83

3.2.5 Calculations and statistics ...... 84

3.3 Results ...... 87

3.3.1 Soil chemical, physical and biological characteristics ...... 87

3.3.2 Soil microbial community structure ...... 87

3.3.3 α-diversity ...... 92

3.3.4 β-diversity ...... 93

3.3.5 Relationships between environmental variables and prokaryotic communities ...... 103

3.4 Discussion: ...... 107

3.4.1 α- diversity ...... 107

3.4.2 Bacterial β- diversity ...... 107

3.4.3 Archaeal β- diversity ...... 109

vii

3.4.4 Influence of vegetation on α- and β- diversity ...... 112

3.4.5 Influence of soil characteristics on α- and β- diversity ...... 115

3.4.6 Microbial community structure in reconstructed soils planted with grass ...... 115

3.4.7 Effect of disturbances, other than forest-fire, on β- diversity in natural soils ...... 117

3.5 Conclusion ...... 117

Chapter 4: Nutrient availability and vegetation cover as drivers of fungal α- and β- diversity in reconstructed oil-sands soils and natural boreal-forest soils ...... 119

4.1 Introduction ...... 119

4.2 Materials and methods ...... 125

4.2.1 Study area ...... 125

4.2.2 Study sites ...... 127

4.2.3 Laboratory analyses ...... 128

4.2.4 Bioinformatics analyses ...... 130

4.2.5 Calculations and statistics ...... 130

4.3 Results ...... 133

4.3.1 Soil chemical, physical and biological characteristics ...... 133

4.3.2 Soil fungal community structure ...... 135

4.3.3 α-diversity ...... 139

4.3.4 β-diversity ...... 140

4.3.5 Relationships between environmental variables and fungal communities ...... 146

4.4 Discussion ...... 148

4.4.1 Fungal biomass and number of fungal OTUs ...... 148

4.4.2 Fungal α-diversity ...... 151

4.4.3 Fungal β-diversity ...... 152

4.4.4 Fungal communities in reconstructed soils planted with grasses ...... 157

4.4.5 Litter-degrading capabilities of fungi ...... 157

viii

4.5 Conclusion ...... 160

Chapter 5: Conclusion ...... 162

5.1 Main findings ...... 168

5.1.1 Role of nitrogen deposition in shaping microbial communities and nitrogen cycling in

reconstructed soils ...... 168

5.1.2 Nitrate cycle and its fate in reconstructed soils ...... 169

5.1.3 Alpha-diversity and soil functions ...... 170

5.1.4 Role of vegetation on nitrogen cycle processes and communities ...... 171

5.2 Contributions ...... 172

5.2.1 Practical contributions ...... 172

5.2.2 Fundamental contributions ...... 173

5.3 Future research avenues ...... 174

References ...... 177

Appendices ...... 210

Appendix A Physical, chemical and biological characteristics of the studied soils ...... 210

Appendix B : Vegetation description ...... 212

B.1 Tree cover at each study site (source : Anderson, 2014) ...... 212

B.2 Shrubs cover at each study site (source : Anderson. 2014) ...... 213

B.3 Grass cover at each study site (source: Anderson. 2014) ...... 215

B.4 Forbs cover at each study site (source: Anderson. 2014) ...... 216

B.5 Lichen and bryophyte cover at each study site (source : Anderson. 2014) ...... 219

Appendix C Gross N-transformation rates presented in different units ...... 220

C.1 Gross N-transformation rates in the studied soils express in mg N m-2 day-1 ...... 220

-1 C.2 Gross N-transformation rates in the studied soils express in mg N kgdrysoil day ...... 221

Appendix D F values, p values of ANOVA on prokaryotic communities ...... 222 ix

D.1 F values, p values and signification of ANOVA-1way on bacterial phyla (p<0.001 : ***;

p<0.01 : **; p<0.05 : *; p<0.1 : .) ...... 222

D.2 F values, p values and signification of ANOVA 1 way on bacterial classes (p<0.001 : ***;

p<0.01 : **; p<0.05 : *; p<0.1 :.) ...... 223

Appendix E F values. p values and signification of ANOVA 1-way on fungi classes (*** : p<0.001; ** :p<0.01; * : p<0.05; . :p<0.1) ...... 224

List of Tables

Table 1.1 Main forms of nitrogen in soil and their oxidation states (Robertson & Groffman,

2007) ...... 6

Table 2.1 Description of parameters (rates) used in the conceptual N-cycle model presented in figure 2.1 ...... 52

Table 2.2 Selected physical, chemical and biological characteristics of each of the study soils .. 54

Table 2.3 Gross ammonification, NH4+ consumption, nitrification and NO3- consumption rates calculated using Hart et al. (1994) equations ...... 57

Table 2.4 Gross N-transformation rates in the AOSR (mg N kg-1 day-1) ...... 65

Table 3.1 Selected physical, chemical and biological characteristics of the studied soil ...... 88

Table 3.2 Number of sequences, OTU and coverage in the studied soils ...... 89

Table 3.3 Coefficients and signifiance of the multiple linear regression explaining α-diversity in the studied soils (*** : p<0.01; ** p<0.05; * : p<0.1) ...... 93

Table 3.4 Phyla (bold) and classes (italics) that are more abundant in either reconstructed soils, natural or in grassland soils...... 100

Table 4.1 Selected physical. chemical and biological characteristics of the studied soils ...... 134

Table 4.2 Number of sequences, OTU and coverage in soils in the studied soils ...... 135

Table 4.3 Coefficients and signifiance of the multiple linear regression explaining α-diversity in the studied soils (*** : p<0.001; . : p<0.1) ...... 140

Table 4.4 Phyla (bold) and classes (italics) that are more abundant in either reconstructed soils, natural soils or grassland soils ...... 143

Table 5.1 Main findings and processes identified in the thesis ...... 164

List of Figures

Figure 1.2 A process nitrogen cycle drawn by Löhnis in 1913 (Clark, 1981) ...... 7

Figure 2.1 Conceptual 15N tracing model to analyze N transformation in soil (Müller et al.

2004). For explanation on transformation rates and parameters see table 1 ...... 51

Figure 2.2 Soil total nitrogen (A) and inorganic nitrogen (B) content in reconstructed soils planted with coniferous trees, deciduous trees and grass, and in natural forest soils. Lower-case characters indicate significant differences among treatments ...... 55

Figure 2.3 Linear regression between total carbon content (%) and total nitrogen content (%) in coniferous (C); deciduous (D); grassland soils (G) and natural forest soils (N) ...... 56

Figure 2.4 Soil C:N ratio in reconstructed soils planted with coniferous trees, deciduous trees, grass and in natural forest soils. Lower-case characters indicate significant differences among treatments ...... 56

-2 -1 Figure 2.5 Soil gross ammonium transformation rates (mg NH4-N m day ) in coniferous, deciduous, grassland and naturally disturbed sites. Lower-case characters indicate significant differences among treatments. Note the scale difference between plot B and E ...... 59

-2 -1 Figure 2.6 Soil gross nitrate transformation rates (mg NO3-N m day ) in reconstructed soils planted with coniferous trees, deciduous trees and grass, and in natural forest soils. Lower-case characters indicate significant differences among treatments ...... 60

Figure 2.7 Soil ammonification and nitrification net rates in reconstructed soils planted with coniferous trees, deciduous trees, grass, and in natural forest soils. The red line represents the limit between net production (above the line) and net assimilation (below the line) ...... 62

+ - Figure 2.8 NH4 (A) and NO3 (B) turnover times in reconstructed soils planted with coniferous trees, deciduous trees, grasses and natural forest soils ...... 63 xii

Figure 2.9 N cycle with measured (average) gross rates of N transformation in reconstructed (a) and natural soil (b). The thickness of the arrows represents the importance of that rate within the soil N-cycle ...... 64

-1 -1 Figure 2.10 Soil gross ammonium transformation rates (mg NH4-N kg OM day ) in coniferous, deciduous, grassland and naturally disturbed sites. Lower-case characters indicate significant differences among treatments ...... 70

Figure 2.11 Pictures of reconstructed-coniferous sites C3 (A) and C4 (B). Patches of peat material are pointed on the C4 soil profile ...... 71

Figure 3.1 Rarefaction curves for reconstructed soils planted with coniferous trees (C1 to C5), reconstructed soils planted with deciduous trees (D1 to D5), reconstructed soils planted with grasses (G1 to G5) and in natural forest soils (N1 to N5) ...... 90

Figure 3.2 Proportion of bacterial phyla (A) and classes (B) in reconstructed soils planted using coniferous, deciduous and grasses species and in natural forest soils ...... 91

Figure 3.3 Inverse Simpson index in reconstructed soils planted using coniferous, deciduous and grasses and in natural forest soils ...... 92

Figure 3.4 Principal Coordinate Analysis (scaling 1) showing ordination of bacterial phyla in reconstructed soils planted with coniferous (C) trees, deciduous (D) trees or grasses (G) and in natural (N) forest soils. Phyla enriched in reconstructed soils are in blue; phyla enriched in natural forst soils are in green; (p<0.001: ***; p<0.01: **; p<0.05: *; p<0.1:.) ...... 94

Figure 3.5 Principal Coordinate Analysis (scaling 1) showing ordination of bacterial classes in reconstructed soils planted with coniferous (C) trees, deciduous (D) trees or grasses (G) and in natural (N) forest soils. Classes enriched in reconstructed soils are in blue; classes enriched in natural forest soils are in green; (p<0.001: ***; p<0.01: **; p<0.05: *; p<0.1:.) ...... 94 xiii

Figure 3.6 Z-scores of bacterial phyla in reconstructed soils planted with coniferous trees, deciduous trees or grasses and in natural forest soils. The first rectangle from the bottom groups phyla that are more abundant in natural forest soils; the second rectangle groups phyla that are more abundant in reconstructed soils; the third rectangle groups phyla that are more abundant in reconstructed soils planted with grasses (p<0.001: ***; p<0.01: **; p<0.05: *; p<0.1: .) ...... 96

Figure 3.7 Z-scores of bacterial classes belonging to phyla more abundant in reconstructed soils

(p<0.001: ***; p<0.01: **; p<0.05: *; p<0.1:.) ...... 97

Figure 3.8 Z-scores of bacterial classes belonging to phyla more abudnant in natural forest soils

(p<0.001: ***; p<0.01: **; p<0.05: *; p<0.1:.) ...... 98

Figure 3.9 Z-scores of bacterial classes belonging to phyla more abundant in reconstructed soils planted with grasses (p<0.001: ***; p<0.01: **; p<0.05: *; p<0.1: .) ...... 99

Figure 3.10 Proportion of bacteria and archaea classes in the studied soils. Statistical differences among bacteria are identified with lower-case characters (p>0.1) and statistical differences among archaea are identified with upper-case character (p<0.001) ...... 101

Figure 3.11 Proportion of archaeal phyla (top) and classes (bottom) in reconstructed soils planted using coniferous, deciduous and grass species and in natural forest soils ...... 102

Figure 3.12 Canonical redundancy analysis (scaling 3) (A) and partition of variation (B) showing relations between site and soil characteristics and bacterial phyla in reconstructed soils planted with coniferous trees (C1 to C5), reconstructed soils planted with deciduous trees (D1 to D5), reconstructed soils planted with grasses (G1 to G5) and in natural forest soils (N1 to N5).

(p<0.001: ***; p<0.01: **) ...... 104

Figure 3.13 Canonical redundancy analysis (scaling 3) (A) and partition of variation (B) showing relations between site and soil characteristics and bacterial classes in reconstructed soils planted xiv

with coniferous trees (C1 to C5), reconstructed soils planted with deciduous trees (D1 to D5), reconstructed soils planted with grasses (G1 to G5) and in natural forest soils (N1 to N5).

(p<0.001: ***; p<0.01: **) ...... 106

Figure 4.1 Rarefaction curves for reconstructed soils planted with coniferous trees (C1 to C5), reconstructed soils planted with deciduous trees (D1 to D5), reconstructed soils planted with grasses (G1 to G5) and in natural forest soils (N1 to N5) ...... 136

Figure 4.2 Number of fungal-phyla OTUs identified (A) and their relative abundance (B) in reconstructed soils planted with coniferous trees, deciduous trees or grasses and in natural forest soils ...... 137

Figure 4.3 Number of fungal-classes OTUs identified (A) and their relative abundance (B) in reconstructed soils planted with coniferous trees, deciduous trees or grasses species and in natural forest soils ...... 138

Figure 4.4 Inverse Simpson index in reconstructed soils planted with coniferous trees, deciduous trees or grasses and in natural forest soils ...... 139

Figure 4.5 Principal Coordinate Analysis (scaling 1) showing fungal classes in reconstructed soils planted with coniferous trees (C1 to C5), deciduous trees (D1 to D5), grasses (G1 to G5) and in natural boreal-forest (N1 to N5) soils. Classes more abundant in reconstructed soils planted to trees are in blue; classes more abundant in reconstructed soils planted to grass are in green, and classes more abundant in naturally disturbed soils are in red (p<0.01: **; p<0.05: *; p<0.1:.) ...... 140

Figure 4.7 Z-scores of fungal phyla in reconstructed soils planted with coniferous trees, deciduous trees or grasses and in natural forest soils. The first rectangle from the bottom groups phyla that are most abundant in reconstructed soils planted with trees; the second rectangle groups phyla that are most abundant in reconstructed soils planted with grasses; the third rectangle groups phyla that are more abundant in natural soils (p<0.001: ***; p<0.01: **; p<0.05: *) ...... 142

Figure 4.8 Z-scores of fungal classes belonging to phyla that were more abundant in reconstructed soils. The first rectangle from the bottom groups classes that are most abundant in reconstructed soils planted with trees; the second rectangle groups classes that are more abundant in reconstructed soils planted with grasses; the third rectangle groups classes that are most abundant in natural soils (p<0.001: ***; p<0.01: **) ...... 144

Figure 4.9 Z-scores of fungal classes belonging to phyla that were more abundant in natural soils. The first rectangle from the bottom represents the class that is most abundant in natural soils; the second rectangle represents the class that is most abundant in reconstructed soils

(p<0.001: ***; p<0.1: .) ...... 145

Figure 4.10 Canonical redundancy analysis (scaling 2) showing relations among vegetation, soil characteristics and fungal phyla in reconstructed soils planted with coniferous trees (C1 to C5), deciduous trees (D1 to D5) or grasses (G1 to G5) and in natural forest soils (N1 to N5).

(p<0.001: ***; p<0.01: **; p<0.05: *) ...... 147

Figure 4.11 Canonical redundancy analysis (scaling 2) showing relations among sites, soil characteristics and fungal classes in reconstructed soils planted with coniferous trees (C1 to C5), deciduous trees (D1 to D5) or grasses (G1 to G5) and in natural forest soils (N1 to N5).

(p<0.001: ***; p<0.05: *) ...... 148 xvi

Acknowledgements

This is with an intangible mix feeling of happiness and sadness that I’m finding myself at the end of this PhD journey. It has sometimes been ecstatic, sometimes depressing, but never boring. I enjoyed every step of the adventure and I have a lot of people to thank for contributing to this feeling.

Foremost, I want to thank my supervisor Sue Grayston. Thank you for believing enough in me when I was just fresh out of the master degree. Thank you for your constant support

(scientifically, emotionally and financially) all the way through the PhD. Heartfelt thank you to

Cindy Prescott who played a role thousands-of-light-years beyond what her official role of committee member would have dictated. Thanks for always pushing me, making me a better

English writer (damn English :-) and most importantly a better scientist. Your role was fundamental to the evolution of this PhD. Thank you to Sylvie Quideau for having put this project together and for your constant support all the way through the program. I enjoy our discussions in conferences and I sincerely hope we will be able to collaborate in the future.

When I came in this program, I was secretly happy to have an all-woman committee. I am now grateful to walk in the steps you have created for us.

Je ne pourrais jamais remercier assez mes parents. C’est votre doctorat à vous aussi. Encore une fois merci d’avoir cru en l’éducation, de m’avoir toujours soutenu même à 3000 km de distance et même si je n’aime toujours pas, à trente ans, me faire dire d’aller travailler plus, vous aviez raison. Pascal, j’ai été parfois l’enfer durant cette aventure. Merci de l’avoir fait avec moi. Merci d’avoir été là à toutes les étapes et d’avoir vécu avec moi les rires comme les larmes. Je t’aime et xvii

espère vivre encore des milliers de tempêtes avec toi. Merci aussi à tous les membres de nos familles qui nous ont épaulés de toutes les façons possibles pendant ces années. Aux amis, merci d’avoir partagé angoisses, joies et surtout bières avec moi! To Jeff, Vicki, Calum, Megan and

Susannah thank you for being our Vancouver-family.

This project will never have been possible without the help of people on the field, in the laboratory and in the office. Jeff and Hannah, that first trip in Alberta was challenging for me. I want to thank you not only for your help and the fruitful discussions we had on the project, but also for making this field trip a life-changing moment. Meghan and Eli you were the perfect partners for my second round in the oil sands. I enjoyed all the discussion we had together and especially the late-night in the lab (that’s mainly a joke). Meghan, you were an inspiration that kept me working, you are a great scientist and a great person. To all the other people who helped me in the lab (Hannah, Emily, Megan and Zack), I would never have made it without you. Thank you to Alice and Kate for having taking care of us in the lab. Thank you also to Carolyn, David and Tim for the discussions in the office.

This study was funded by a Natural Science and Engineering Research Council of Canada

(NSERC) Collaborative Research and Development grant (PI Sylvie Quideau). Personal funding was provided by NSERC PGS-D scholarship, FQRNT scholarship, UBC FYF, UBC Forestry

Strategic Recruitment Fellowship, UBC Forestry Edward W. Bassett Memorial Scholarship and

UBC Forestry Mary, David Macaree Fellowship and the Agricultural Institute of Canada

Foundation. Thank you for supporting students. Merci!

xviii

Dedication

À Marie, Yves et Pascal

xix

Chapter 1: Introduction

1.1 The oil sands in Canada

Covering 140,200 square km, the Alberta oil sands reserve is divided in three regions of northern Alberta (Peace River, Athabasca, and Cold Lake deposits) (Figure 1.1). In 2013, the proven oil reserve of Alberta was 170 billion barrels. This represents the third reserve of oil worldwide after Saudi Arabia and Venezuela. The Albertan oil reserve counts for 98% of the total oil reserve of Canada and approximately 13% of the total global oil reserve. Most of this oil is in the unconventional heavy form, which is usually referred to as oil sands or bituminous sands (Government of Alberta, 2013a).

Figure 1.1 Oil sands region in northern Alberta, Canada (Source: Government of Alberta, 2012).

1 1.1.1 The resource

Athabasca oil sand is a mixture of sands, water, clay and hydrocarbons that is thick at ordinary temperatures and contains high carbon and sulphur contents compared to traditional crude oil

(Ferguson, 1984; Government of Alberta, 2013a). As such, it is not recoverable using traditional methods of extraction such as wells. Oil sands extraction can be done via surface mining if the deposit is less than 75 meters deep or through in-situ recovery using a hot-water extraction method, developed in the first half of the 20th century by Karl Clark from the

Alberta Research Council (Ferguson, 1984; Government of Alberta, 2013a). The bitumen extracted from the Alberta’s oil sands area requires either carbon removal, via a process called coking, or hydrogen addition, through hydro-cracking. For all these reasons, the oil sands are also called heavy crude or synthetic oil and pose unique environmental challenges (Ferguson,

1984).

1.1.2 The history of the oil sands development in Alberta

First Nations in the Athabasca region historically used the oil sands in Alberta to waterproof their canoes (Ferguson, 1984; Government of Alberta, 2016a). When Canada acquired the

Hudson’s Bay Company, in 1870, the government commissioned the Geological Survey of

Canada to inventory of Canada’s geophysical resources done by the. John Macoun (in 1875) and Robert Bell (in 1882) led the first two surveys in the Athabasca region. They noted the presence of the bitumen along the river shore in the region, and believed that the bitumen would be extractable through conventional wells. The first well dug by Alfred Von

Hammerstein, at the end of the 19th century, reached 1,500 feet (457.2 meters), but was inconclusive. Drillings continued for eight years in the area with a total bill of $38,000 paid by

2 the Canadian Geological Survey. Drilling resumed a few years later supported this time by private initiatives. Altogether approximately 24 wells were dug in the Athabasca region between 1894 and 1910, all unsuccessful (Ferguson, 1984; Syncrude Canada Ltd, 2013).

Between 1912 and 1914, the Dominium of Canada acknowledged the importance of oil for its development and wanted to get away from the extreme dependence of Canada on foreign supplies. Therefore, the government made of the Athabasca region a federal reserve in which private development was invited with the Government maintaining the right to expropriate lands and equipment (Ferguson, 1984). In 1913, Sidney Ells, an employee of the federal

Department of Mines sent out samples for laboratory analysis; the oil sands were then assessed as a road-paving material (Syncrude Canada Ltd, 2013).

In 1921, the Research Council of Alberta was founded and Dr. Karl A. Clark, a chemist, was appointed as the first full-time employee of the institution. His mission was to find a way to utilize the bituminous sands. At that time, he was certain that the future of the oil sands lay in devising a way to separate the oil from the sand. He focused on a method of separation and recovery using hot water, a method still in use in the oil sands industry. Because of his work,

Dr. Clark is often cited as the father of the oil sands industry in Alberta (Sheppard, 2005).

The entrepreneur R.C. Fitzsimmons came to Fort McMurray 10 years later hoping to extract oil using Dr. Clark’s method. He formed the International Bitumen Company and built a plant at Bitumount, 80 kilometres north of what is now the town of Fort McMurray. This project

3 indirectly led to the first commercial project in the oil sands, now owned by Suncor

(Sheppard, 2005; Syncrude Canada Ltd, 2013).

After twenty years of both private and public research on the exploitation of the oil sands,

S.M. Blair, an engineer commissioned by the provincial government of Alberta, produced a technical and economic survey of the feasibility of commercial exploitation of the oil sands in northern Alberta (Sheppard, 2005). The report was released in December 1950. It concluded that the production of oil from the bituminous sands could be a profitable venture. It established that the total cost of production would be $3.10 per barrel and that the potential revenue at that time was $3.50 per barrel. It also estimated that the region had “few square miles that each contain 200 million barrels of bitumen and possibly an appreciable number that contain 100 million barrels” (Blair, 1950). This report marked the beginning of commercial development of the oil sands.

Seventeen years after that report, in 1967, the first commercial production of oil sands began.

In 1964, the Syncrude consortium (a joint government and private venture) was formed. In

1969 Syncrude received government permission and started to construct a plant. Production began in 1978; a hundred years after the first survey by the Geological Survey of Canada

(Sheppard, 2005; Syncrude Canada Ltd, 2013).

Thirty-six years later, in 2014, the total production of bitumen in the region reached 2.3 million barrels per day (Government of Alberta, 2016b). Approximately 121,500 workers were employed in businesses evolving either in oil sands, conventional oil and gas and mining

4 in Alberta. Royalties collected from the industry during the fiscal year 2013-2014 were of $5.2 billion, which represented about 13% of the estimated provincial revenues for the year 2013

(Government of Alberta, 2016b; Horner, 2013). Based on 2011 estimates, it is believed that the capacity of the oil sands could peak at 7.1 million barrels a day by the end of 2027 with cumulative royalties of $350 billion (CERI, 2012). However, in early 2016, the price of oil dropped to around US$36 per barrel after a US$106 per barrel maximum in 2011 (CNBC,

2016). At this price, it is believed that mining operations are loosing about CDN $3 per barrel

(Williams, 2016). This worldwide drop in the price of oil has profoundly affected the oil sands industry and the province of Alberta.

1.1.3 The challenge of ecological restoration in the oil sands

Overall, the total minable area in the region spans over 4,800 km2 of land. In December 2013,

813 km2 have been disturbed in the oil sands; a 100 km2 increase since 2011 (Government of

Alberta, 2015, 2013b). Before operating the mines, companies are required to design a closing plan that includes their reclamation strategies. According to the Conservation and Reclamation

Regulation of Alberta, companies are required to restore soils that can achieve equivalent land capability which is described as “the ability of the land to support various land uses after conservation and reclamation [that] is similar to the ability that existed prior to activity being conducted on the land, but that the individual land uses will not necessarily be identical”

(Government of Alberta, 1993; Powter et al., 2012). Mine operators must also provide a reclamation security bond as a guarantee that reclamation work will take place. As of August

31st 2014, the government had approximately $1 billion in reclamation securities from oil

5 sands companies (Government of Alberta, 2013b). Because of the size of the area to reclaimed, the Canadian Energy Research Institute believes that the reclamation of the tailings ponds and other disturbed environments could become a large liability to companies (CERI,

2012). Critical to the long-term sustainability of the reclaimed landscapes is the re- establishment of biogeochemical cycling of nutrients between the reconstructed soils and plants. Among these biogeochemical cycles, the nitrogen cycle is of prime importance because of its impacts on site productivity.

1.2 Nitrogen cycle

Nitrogen is a key element in soil ecosystems. As an essential constituent of proteins, N is a vital element for all living organisms. While deficiencies in nitrogen can induce poor crop productivity and lead to famine, excesses can lead to contamination of groundwater by nitrates and emission of the potent greenhouse gas, nitrous oxide (N2O). Consequently, management of nitrogen in soil is ecologically, financially and environmentally critical (Brady and Weil,

2002). Nitrogen occurs in many forms, few of which are bioavailable (Table 1.1.) (Robertson and Groffman, 2007).

Table 1.1 Main forms of nitrogen in soil and their oxidation states (Robertson & Groffman, 2007)

Name Chemical formula Oxidation state - Nitrate NO3 +5 Nitrogen dioxide (g) NO2 +4 - Nitrite NO2 +3 Nitric oxide (g) NO +2 Nitrous oxide (g) N2O +1 Dinitrogen (g) N2 0 Ammonia NH3 -3 + Ammonium NH4 -3 Organic N RNH3 -3 Gases (g) occur both free in the soil atmosphere and dissolved in soil water

6 Transformations among these forms are systematic and mediated by microbes. In 1913, Felix

Löhnis was one of the first researchers to design a conceptual model of the nitrogen cycle

(Clark, 1981). His model formalized the idea that N species were converted from one form to another in a stepwise fashion (Figure 1.2).

Figure 1.2 A process nitrogen cycle drawn by Löhnis in 1913 (Clark, 1981)

Since then, numerous models have been developed, each one bringing mostly cosmetic, rather than substantive changes; the major pathways remaining N-fixation, mineralization, nitrification and denitrification (Clark, 1981). However, in nearly a hundred years of research a few paradigms about the nitrogen cycle have been challenged and new processes have been discovered. For example, it was thought that the limiting step in the nitrogen cycle was mineralization ⎯ the microbial transformation from the previously thought unavailable organic

+ form of N to the available inorganic form of N, ammonium (NH4 ). Schimel and Bennett

(2004) suggested that the limiting transformation was rather the depolymerization of soil organic N, as in N-poor environment plants are able to take up organic nitrogen (Näsholm et al., 2009; Paungfoo-Lonhienne et al., 2010; Schimel and Bennett, 2004). Consequently, nitrogen models can now address differences in N-poor (dominated by organic-N) and N-rich

7 (dominated by inorganic N forms) ecosystems. Other processes such as nitrifier- denitrification, dissimilatory nitrate reduction to ammonium (DNRA), the anaerobic oxidation of ammonium (annamox) and the complete oxidation of ammonium to nitrate (comammox) were later identified.

The model I use in this thesis was designed by Prescott 2012 and covers the majority of these concepts (Figure 1.3)

Figure 1.3: Nitrogen cycle (Prescott, pers. comm., 2012)

1.2.1 Nitrogen fixation

Nitrogen fixation is the main pathway by which new nitrogen enters terrestrial ecosystems.

Gaseous N2 composes 78% of the Earth’s atmosphere. However, the strong hydrogen triple-

8 bond between the two nitrogen atoms makes N2 virtually non-available for living organisms except for N-fixing organisms. The reactive N pool of the soil consists of the nitrogen that is

+ - bound to carbon (organic N), hydrogen (NH3, NH4 ) or oxygen (NOx, NO3 , N2O) atoms. In order to pass from the inactive to the reactive pool, the inert N2 needs to be transformed through a process called “fixation”. This process can be achieved through three major pathways: energy discharge (lightning), biological fixation and industrial fixation (McNeill and Unkovich, 2007).

Lightning is not considered to be a major pathway for N-fixation. It is estimated that less than10 Tg N are fixed through this process per year (Galloway et al., 1995). Biological fixation, on the other hand, is historically perceived as the dominant process by which N enters the reactive pool (Robertson and Groffman, 2007). The reduction from N2 to NH3 is catalyzed by the enzyme nitrogenase. About 90 genera of specialised prokaryotes, carrying the gene nifH, can produce this enzyme (McNeill and Unkovich, 2007). These microorganisms exist either as free-living organisms (e.g., Cyanobacteria) or in a symbiotic association with plants in nodules of legumes (e.g., Rhizobium can be the symbiont of

Trifolium) or of non-legumes (e.g., the actinomycetes Frankia with Alnus) (McNeill and

Unkovich, 2007). Nitrogenase is inhibited by O2 and the reduction of N2 to NH3 requires a lot of energy; by associating with higher plant, the microorganisms get both protection against O2 and access to energy. Finally, the ammonia (NH3) resulting from N-fixation is combined with organic acids to form amino acids and, ultimately, proteins. These organic forms of nitrogen

+ can be either taken up by plants or depolymerised and transformed in inorganic N (NH4 ) by

9 microorganisms through mineralization. Accumulation of ammonium in soil inhibits N- fixation (Brady and Weil, 2002).

Preindustrial biological fixation of N in terrestrial ecosystems was estimated to be between 90 and 130 Tg N/year. With the industrialisation of modern societies, industrial fixation processes outpaced biological fixation and became the dominant pathway of N-fixation.

Industrial fixation includes fabrication and the use of fertilizers, enhanced cultivation of N- fixing legumes and combustion of fossil fuels (Galloway et al., 1995). The invention of synthetic agricultural fertilizers in the 20th century was at the root of the Green Revolution in agriculture. The development Haber-Bosch process permitted the industrial fixation of N2 in a relatively cost-effective manner and helped to boost agricultural production to face the challenges of a fast-growing human population. The Haber-Bosch N-fixation process starts by generating H2 with natural gas and water, and then combines the H2 with N2 from the atmosphere at very high temperature and pressure to produce NH3 in the presence of a catalyst

(McNeill and Unkovich, 2007). In 1995, the production of fertilizers was believed to fix around 80Tg N/year (Galloway et al., 1995). Combustion of fossil fuel releases into the atmosphere previously fixed N that was stored in long-term reservoirs. The high temperatures of the combustion of the fuel also fix a small amount of present-day atmospheric N2. This process can emit more than 20 Tg/year of fixed nitrogen. Finally, agricultural practices that specifically select crops with N-fixing capabilities add about 32 to 53 Tg N / year to terrestrial ecosystems (Galloway et al., 1995).

10 In summary, preindustrial N-fixation processes were able to fixed around 100 to 150 Tg N / year. In the last century, human activities caused the release of approximately 140 Tg N/year of newly fixed nitrogen. Hence, human activities have approximately doubled the amount of fixed N in terrestrial ecosystems. Consequently, humans have become the major player in N- fixation processes (Galloway et al., 1995; Vitousek et al., 1997).

1.2.2 Depolymerization

The N2 fixed by bacteria is converted to ammonia (NH3), which is then assimilated into organic matter. Soil organic matter is the biggest N pool in soil. Previously, it was assumed that higher plants could only take up inorganic forms of nitrogen, so the soil organic matter needed to be decomposed and the N mineralized by microorganisms before it became available to plants (McNeill and Unkovich, 2007). The thinking was then that nitrogen mineralization regulated the overall N cycle. However, recent studies have shown that plants can take up, to various degrees, amino-acids, peptides, proteins and even microbes (Näsholm et al., 2009; Paungfoo-Lonhienne et al., 2010), making depolymerisation – the breakdown of insoluble organic N into amino-acids and other SON (proteins, peptides) – the new rate- regulating step of the N cycle (Schimel & Bennett, 2004).

Depolymerization is mainly achieved through the process of proteolysis, which is the breakdown of proteins into simpler compounds. Extracellular enzymes, mainly proteases, secreted by free-living microbes, mycorrhizal fungi and plant roots, carry out proteolysis. This process is still not yet fully understood, but pH, soil organic-matter pools and protein concentrations seems to influence the proteolytic activity. Low pH increases proteolytic

11 activity and amino-acid turnover, while high concentrations of amino acids slow the process

(Näsholm et al., 2009).

The ecological significance of organic N in plant nutrition remains uncertain. All plants tested to date can take up amino acids (Lipson and Näsholm, 2001), and this pathway could explain the discrepancy between measured rates of production of inorganic N forms and measured

+ annual rates of plant N uptake. NH4 is absorbed by roots at the highest rates, followed by

- amino acids, and finally NO3 (Näsholm et al., 2009). If the rate of absorption is a proxy for plant preference, that would mean that even though plants can take up SON, they

+ preferentially absorb NH4 . Organic N uptake is probably more important in N-poor

+ - ecosystems, where concentrations of NH4 and NO3 are low (Sauheitl et al., 2009; Schimel &

+ Bennett, 2004). In such N-poor systems, it is likely that most of the NH4 mineralized is consumed by microorganisms, leaving plants with only SON to meet their N needs.

1.2.3 Mineralization

The transformation of organic N into inorganic forms of nitrogen is called mineralization.

Extracellular enzymes produced by a variety of soil microorganisms mediate this process.

These organisms are decomposing soil organic matter (SOM) as a source of energy and carbon to support their growth. By doing so, they also mineralize the organic forms of N into

+ + NH4 . If the SOM is N-rich, microorganisms’ needs can be met and the excess NH4 is available for plants; this process results in net N mineralization. On the contrary, if microorganisms’ needs for N are not fulfilled, they immobilize the transformed N in their tissues, causing net N immobilization (Robertson and Groffman, 2007). The N is eventually

12 returned to the SON pool through microbial turnover. This route through which N is recycled is called Mineralization-Immobilization Turnover (MIT) (McNeill and Unkovich, 2007).

Whether a system generates net mineralization or immobilization is mostly influenced by the

C:N ratio of the SOM and of the microorganisms themselves, as well as by the labile carbon available to microorganisms. Generally, a C:N ratio of SOM lower than 25:1 results in net mineralization and a C:N ratio higher than 25:1 results in immobilization (Paul and Juma,

1981). The specific N needs of each organism can also influence the N net mineralization or immobilization rates and, this is determined by the C:N ratio of their tissues. For instance, fungi have wider C:N ratios in their tissues than bacteria, and so can grow more efficiently on

+ low-N substrate and immobilize less NH4 than bacteria (Robertson and Groffman, 2007).

Finally, the availability of labile C reduces net mineralization, mainly because microorganisms decompose less SOM if they have access to an easy source of carbon; by doing so they mineralize less N and increase the potential for N immobilization (Prescott and McDonald,

1994).

1.2.4 Nitrification

Nitrification is the microbial oxidation of reduced forms of nitrogen into less reduced forms,

- - principally nitrite (NO2 ) and nitrate (NO3 ) (Robertson and Groffman, 2007). Equation 1 describes this three-step process.

+ - - NH4 ↔ NH3 → NH2OH → NO2 → NO3 (eq. 1)

13 The first step is the oxidation of ammonia to hydroxylamine carried out by the membrane- bound ammonia mono-oxygenase enzyme. This enzyme is produced only by a select group of organisms, called ammonia-oxidizing bacteria (AOB) (e.g. Nitrososcoccus, Nitrosospira,

Nitrosomonas, Nitrosovibrio and Nitrosolobus). These organisms are obligate aerobes and have the gene amoA in their genomes. Some archaeal species called ammonia-oxidizing archaea (AOA) also carry the gene amoA, making them able to oxidize ammonia. AOA have been seen to outnumber their bacterial counterparts in some soils (Leininger et al., 2006).

Within the second step of nitrification, hydroxylamine is further oxidized to nitrite by hydroxylamine oxidoreductase enzymes. In most soil nitrite does not accumulate as it is quickly oxidized to nitrate (step 3) by nitrite-oxidizing bacteria (e.g. Nitrobacter, Nitrospira,

Nitrococcus and Nitrospina) that produce the nitrite oxidoreductase enzyme (Teske et al.,

- 1994). The end-product, NO3 , will then be assimilated by organisms or be further reduced through denitrification pathways. Interestingly, ammonia-oxidizers also appear to be able to

- produce gaseous NO via NO2 reduction. The nitrite acts as an electron acceptor when O2 levels are low in soil. This process is call nitrifier denitrification (Wrage et al., 2001).

Autotrophic bacteria or archaea gain up to 440kJ of energy per mole of NH3 oxidized when

- NO3 is the end product. Up to 80% of the energy produced during nitrification will be respired via the Calvin cycle. This energy will be used to fix CO2 that helps to meet their C needs (Robertson and Groffman, 2007). Even though autotrophic nitrification remains the main pathway by which ammonium is oxidized to nitrate, heterotrophic microbes (bacteria and fungi) can also nitrify. In the latter case, no energy is gained from the transformation. The

14 heterotrophic microbes use the same kind of extracellular enzymes as autotrophic nitrifiers.

However, it appears that these enzymes can oxidize a number of different substrates; therefore, it might be possible that ammonia oxidation is only a secondary process carried out by these enzymes. Heterotrophic nitrification may be important in environments where autotrophic nitrifiers are chemically inhibited by low pH, such as some forest soils (Hayatsu et al., 2008).

Nitrification rates are controlled mainly by the supply of ammonium and by the level of oxygen in the soil as, to date, all known nitrifiers are obligate aerobes. It was believed that both low pH and low temperature inhibit nitrification, but it was recently found that nitrifiers can adapt to a range of temperatures and that some nitrification occurs even in very acidic forest soils through heterotrophic and autotrophic nitrification (Robertson and Groffman,

2007). Wertz et al. (2012) found that autotrophic nitrification dominated in forest soils located in interior British-Columbia and was responsible for 55-97% of total nitrification (Wertz et al.,

2012).

Nitrate is more mobile in soil than ammonium and could be lost through leaching (possibly causing eutrophication of surrounding water bodies) or through denitrification; therefore, careful management practices are needed.

1.2.5 Denitrification

Denitrification is the reduction of soil nitrate to the N gases NO, N2O and N2 by microorganisms (Robertson and Groffman, 2007). Among the gases released through

15 denitrification, N2O is the most studied, as it is a greenhouse gas with a global warming potential 300 times higher than CO2 (EPA, 2013). Organisms that carry the genes for denitrification are widespread with over 25 genera and 125 species of either bacteria, archaea or fungi that can participate in the various steps of denitrification; all obtain a certain amount

- of energy by using NO3 as a terminal electron acceptor during respiration. However, this process gives them less energy than using O2 as a terminal electron; for this reason denitrifiers undertake denitrification only when O2 is unavailable, making denitrification an anaerobic process (Brady & Weil, 2002; Jones et al., 2008).

- Equation 2 shows the general pathway of conversion NO3 into N2. Each step is mediated by extracellular enzymes (nitrate reductase (nar), nitrite reductase (nir), nitric oxide reductase

(nor) and nitrous oxide reductase (nos)) secreted by organisms; among which nirK, nirS and nosZ are the most studied because of their control over critical parts of the denitrification process.

- - 2NO3 → 2NO2 → 2NO → N2O → N2 (eq. 2)

Denitrification is influenced by the concentrations of O2 in the soil, and by the amount of

- moisture, soil C content and NO3 available (Robertson and Groffman, 2007). In acidic soils, denitrification is mostly carried by fungi (Chen et al., 2014). In a silver birch (Betula pendula) plantation on drained organic forest soils, Rütting et al. (2013) measured a strong negative correlation between pH and fungal:bacteria PLFA ratios in soils with high N2O emissions.

Moreover, in-situ studies have shown that fungi denitrifiers could be responsible to more than

16 50% of denitrification in grassland soils (Crenshaw et al., 2008; Laughlin and Stevens, 2002), in arable peat soils (Yanai et al., 2007) and in forest plantation soils (Chen et al., 2014).

1.2.6 Other nitrogen transformations in soil

Although the nitrogen cycle has been studied for over 100 years, new transformation pathways are still being discovered.

DNRA: Dissimilatory nitrate reduction to ammonium

DNRA is the anaerobic reduction of nitrate to nitrite and then to ammonium. It is the anaerobic reverse pathway of the nitrification (Robertson and Groffman, 2007). It has been observed in tropical forest soils (Silver et al., 2001) and in paddy soils (Yin et al., 2002). Like denitrification, this process allows for respiration to occur in the absence of O2 (Robertson and

- + Groffman, 2007). Since DNRA allows for the direct transformation NO3 to NH4 , so from a form that can easily be lost from ecosystems through leaching and denitrification to a much more stable form, it can have an important ecological role. However, because of its anaerobic nature, the DNRA pathway is not expected to be of great ecological importance in non- flooded soils.

Annamox: Anaerobic ammonium oxidation:

The annamox reaction was discovered in 1995 in a pilot plant treating wastewater (Mulder et al., 1995). It coverts ammonium and nitrite directly to N2 gas under anoxic conditions and, therefore, leads to the loss of ammonium in the environment. Our current knowledge of annamox bacteria is limited to three genera (Brocadia, Scalindua and Kuenenia) in the

17 phylum Planctomycetes. Although it was assumed that annamox bacteria were restricted to marine environments, it is becoming clear that they are more widely distributed, and might play an important role in both artificial and natural environments (Hayatsu et al., 2008).

Abiotic ammonium fixation

Compared to nitrate, which is easily leached out of soil, ammonium is more stable in soil.

+ Because it is a positively charged ion, NH4 is attracted to the negatively-charged surfaces of

+ the soils such as organo-mineral complexes and clays. NH4 is then adsorbed to the organic matter or clay particle and is available for plant or microbe uptake while being relatively

+ protected from leaching. This NH4 pool is called a pool of exchangeable ammonium (Brady

+ and Weil, 2002). Because of its size, NH4 can also become entrapped in 2:1 clay minerals

+ (illite and vermiculite) rendering NH4 non-exchangeable (Nômmik, 1981; Trehan, 1996).

Trehan (1996) showed that up to 31% of the ammonium applied to a soil was fixed by clay and 6-13% became incorporated in the organic matter structure of the soil. This fixation renders ammonium virtually unavailable to microorganisms and plants. However, it was also shown that some higher plants are able to utilize the nitrogen in NH4-fixing clay soils; these

+ soils therefore act as storage facilities for NH4 (Nômmik, 1981).

Aerobic denitrification

For a long time, it was thought that denitrification was limited to anaerobic zones of the soil, but this idea was challenged by Ritchie and Nicholas (1972), and confirmed 2001 by Wrage et al (2001). These authors showed that some N2O can be produced during nitrification, which is an aerobic process (Wrage et al., 2001).

Comammox: Complete oxidation of ammonium to nitrate

It was long held that the oxidation of ammonium to nitrate was a two-step process carried out by two different groups of organisms ⎯ i.e. specific bacteria or archaea oxidized ammonia to nitrite and others oxidized nitrite to nitrate. However, it was recently discovered that three cultivated bacteria and one uncultivated bacterium could perform the complete oxidation of ammonium to nitrate without the help of any other organisms. This process through which these organisms were able to oxidize ammonium to nitrate was called comammox. Because, comammox-bacteria can compete with ammonia-oxidizer for ammonium resources, this finding suggests that the degree of microbial competition for ammonium has been underestimated (Santoro, 2016)

1.3 Measuring nitrogen cycling processes

1.3.1 Measuring pools of N

Conceptually, the nitrogen cycle has two main components: pools (organic N, inorganic N and

N gases) and fluxes (e.g. mineralization, nitrification, denitrification). Soil N pools can be measured with a panoply of methods. Soil total N concentration can be measured using either wet digestion method (Kjeldahl method) or dry combustion (e.g. Dumas method). The wet digestion method involves the dissolution of the organic and inorganic nitrogen in acid and their subsequent measurements. With the dry combustion method, nitrogen is initially oxidized and the gases pass through a reduction furnace which reduces NOx to N2; the N2 concentration is then measured (Rutherford et al. 2008). The inorganic pool of N is usually

+ - measured after extracting NH4 and NO3 in a salt (KCl or K2SO4) solution. Li et al. (2012)

19 reviewed the influence of various methodological parameters (e.g. the extracting solution and its concentration, extraction time, shaking time and filter used) on the quality of the extractions. The organic N pool is usually calculated as the difference between the total N pool and the inorganic N pools. Gaseous forms of nitrogen are most commonly measured using field chambers (Hutchinson and Rochette, 2003; Mosier and Klemedtsson, 1994).

1.3.2 Measuring net fluxes of N

Nitrogen fluxes in soil are estimated in three main ways. In the field, “ion-exchange resin” are

+ - installed in the soil for a fixed period of time during which the NH4 and NO3 accumulates on the resins. The resins are removed and extracted in KCl or K2SO4 solution, and the amount of inorganic N extracted is considered to equal the net mineral net mineralization or nitrification rates during the incubation period (Binkley et al. , 1986). Net N mineralization and nitrification rates can also be estimated during incubation of soil samples in the field or in the

+ - laboratory. Concentrations of NH4 and NO3 are measured before and after the soil is

+ - incubated for a period of time, often 28 days. The change in NH4 and NO3 concentrations during the incubation period yields estimates of net ammonification and net nitrification rates respectively (Drury et al., 2008; Hart, Stark et al. , 1994).

1.3.3 Measuring gross fluxes of N using analytical methods

The first two methods estimate net rates of transformation (i.e. gross rate minus consumption) and not gross rates of transformation, these rates do not reflect the amount of nitrogen that is being transformed by microorganisms. To measure gross rates of N transformation, scientists have developed tracers-based method such as the 15N pool-dilution method. The main idea behind this method is to enrich of the “receiving” pool of nitrogen with a stable isotope (15N)

20 and follow the dilution of this isotope as new 14N is produced and incorporated into the

receiving N-pool. Since only the receiving pool is enriched, there are no priming effects on the

measured transformation rates that could arise from adding new nitrogen to the system.

- 15 Basically, to measured gross rates of nitrification, the soil NO3 pool is enriched with NO3

15 14 15 (usually using K NO3 or NH4 NO3). A subsample of the soil is extracted immediately after

- 15/14 - the enrichment and both the concentration of NO3 the NO3 ratio are measured.

The remainder of the soil is incubated for 24h or 48h. sub-sampled and the same parameters

14 + 14 - are measured. During incubation, the NH4 is transformed to NO3 , which reduces the ratio

15 14 15 of NO3/ NO3 and diluting the soil N content (Davidson and Hart, 1991). In their

pioneering work, Kirkham and Bartholomew (1954; 1955) developed a set of simple linear

equations to estimate gross rates of transformation using the 15N pool-dilution method (Figure

1.4) (Kirkham & Bartholomew, 1954; 1955).

Figure 1.4: N-cycle conceptual model for 15N pool-dilution method developed by Kirkham & Bartholomew

(1954)

Kirkham and Bartholomew equations are still being used; however, these equations heavily rely on some assumptions that have been challenged (Schimel, 1996). The main criticism of the Kirkham and Bartholomew model stems from the fact that rates of transformation of nitrogen are not at steady-state, but follow a non-linear function that changes through time depending on various factors such as size of the departing and receiving pools of nitrogen

(Myrold and Tiedje, 1986). The nitrogen cycle also has many more pools and transformation rates than Kirkham and Bartholomew recognized in 1954. Analytical solutions, such as those used by Kirkham and Bartholomew, have limited usefulness considering these concepts.

1.3.4 Measuring gross fluxes of N using analytical methods

Numerical solutions that allow the simultaneous optimization of various rates of transformation began to appear in the late 1980’s, among them the model FLUAZ (Figure 1.5)

(Mary et al., 1998). Müller et al. (2004) developed another model to use with a15N-tracing method which offers the possibility to estimate more parameters than possible with FLUAZ.

The Müller et al. model uses a Markov Chain Monte Carlo method to optimize the fit of the data to a conceptual model chosen by the user. The method allows for the determination of the probability density functions of a whole set of parameters, which are, in this case, the transformation rates. This method is one of the few that can estimate Michaelis-Menten kinetics ⎯ i.e. reaction models that best represent enzymatic reactions in which the transformation rate is link to the substrate pool size (Figure 1.6)

22 Figure 1.5: N-cycle conceptual model for 15N pool-dilution methods developed by Mary et al (1998).

(m=mineralization; ia: ammonium immobilization; n=nitrification; in: nitrate immobilization;

v=volatilization; d: denitrification; r=remineralization; s=N mineralization coming from plant residue

decomposition; j=direct assimilation of organic-N by microorganisms; h=N humification)

Figure 1.6: N-cycle conceptual model for 15N pool dilution method developed by Müller et al. (2004).

+ - (Nlab=labile soil organic-N; Nrec=recalcitrant soil organic-N; NH4 = ammonium; NO3 = nitrate;

+ + - - + NH4 ads=adsorbed NH4 ; NO3 sto = stored NO3; MNlab=mineralization of Nlab; INH4=immobilization of NH4 in

- - Nlab; MNrec: mineralization of Nrec; INO3=immobilization of NO3 in Nrec; ONrec=oxidation of Nrec; DNO3 =DNRA;

+ + + - ONH4=oxidation of NH4 ; RNH4a=release of NH4 from adsorbed NH4 ; RNO3s=release of NO3 from stored NO3-)

23 There are two main differences between FLUAZ and the Müller et al. model. FLUAZ assumes the rates of transformation within a single time interval to be constant, but rates can change from one time interval to the other; whereas the Müller et al. model assumes a single transformation rates for the entire experiment. This aspect of the Müller et al. model has been criticised (Luxhøi et al., 2005), but as Müller et al. (2005) explains, it is also peculiar that microbes work at a certain rate during a experimentally-set time interval, but use a different rate for a second time interval without any changes in the experimental settings. The rapid

+ immobilization and remineralization of NH4 at the beginning of the incubation is also not treated in same way. FLUAZ assumed that this process is driven by microbes, whereas the

Müller et al. assumes that it could be driven by microbes and by abiotic fixation, and so it splits this process into two rates of transformation and two pools and optimizes it.

1.3.5 Comparison between analytical and numerical methods to measure gross fluxes of N

There are some pitfalls in using either the analytical or the numerical method to measure gross fluxes of N. The analytical solutions can only assume the zero-order kinetics (steady-state systems). Yet it is conceivable that transformation rates will change depending on substrate pool size or on the enzymatic reaction rates (i.e. Michaelis-Menten kinetics). Also the analytical solutions calculate the transformation rates successively starting with the gross rates of mineralization. In doing so, the calculations errors are transferred to the last rate calculated, therefore decreasing its robustness (Murphy et al., 2003). By simultaneously calculating N- transformation rates using different kinetics, the numerical solution resolves these issues.

However, numerical solutions could be susceptible to over-determination if one wants to

24 calculate a highly complex system with many N-transformation rates. In order to avoid over- determination, the number of N-transformation fluxes which can be simultaneously optimized should not exceed the number of independent variables provided, otherwise the optimization may result in biased results. It is, therefore, recommended to keep the model as simple as possible and to carefully increased its complexity only if needed (Luxhøi et al., 2005).

Moreover, 15N additions, even to the receiving pools, can increase microbial activity, alter rates of N-transformations and lead to experimental artefacts. Therefore, a good practice would be to limit the addition of 15N to 5-25% of the initial pool size (Luxhøi et al., 2005). As

Myrold and Tiedje (1986) pointed out, the choice of a model and of its assumptions has the potential to generate more profound errors of estimation than does the use of the incorrect kinetic order or even the perturbations associated by the introduction of tracers into the system. Therefore, the choice of model to measure soil gross rates of nitrogen transformation should be made with great care. Measuring N-transformation rates using both the analytical and the numerical solution could be an interesting solution to increase confidence in the results while measuring more complex rates using different kinetics.

1.4 The microbial role in the nitrogen cycle: does diversity matter?

The importance of microbial diversity to ecological function depends on the ecological process in question and the numbers of organisms capable of carrying out that process

(Griffiths et al., 2001; Schimel et al., 2005; Wardle, 2002). If an ecological process is ubiquitous in the population, the process occurs regardless of the diversity of the microbial population. However, if only specific organisms carry out the process, the probability of these specific organisms being present increases with the diversity of the population. Therefore,

25 higher microbial diversity increases the chance that an organism that performs that specific function will be present.

The discrepancy in response of soil functioning to microbial diversity was supported in a dilution experiment in which soil processes behaved differently to a decrease in microbial

- diversity; NO3 accumulation, respiratory growth didn’t change with changes in biodiversity, but substrate induced respiration showed a gradual increase with increasing dilution, potential nitrification only declined at the highest dilution treatment, while long-term decomposition of barely straw only differed between the least diverse and the more diverse treatment (Griffiths et al., 2001).

Schimel et al. (2005) developed a theoretical background in which a process could be narrow, broad or aggregated. Narrow processes are those for which only a phylogenetically limited number of organisms carry out a specific function (e.g. N-fixation), making these processes sensitive to microbial community composition. Martin et al. (1999) found a positive correlation between denitrification and substrate microbial richness, evenness and functional diversity making denitrification rates sensitive to soil biodiversity. Broad, or universal, biochemical processes (e.g. the metabolism of glucose) are widely distributed across living organisms, making diversity less important. Processes are aggregated if a single process is in fact an aggregate of multiple distinct physiological processes (Schimel et al., 2005). Within the traditional view of the nitrogen cycle, only the N-fixation and nitrification pathways were believed to be driven by a narrow set of organisms. However, Schimel et al (2005) suggested that perhaps, mineralization may not be as broad as it has been considered; rather is an

26 aggregate process comprising first, the depolymerization of macromolecular organic material into simpler compounds, followed by the mineralization of these new compounds. Hence, the depolymerization step might be biologically narrow, if the requisite extracellular enzymes are synthesized by a limited set of organisms.

1.5 Impacts of tree species on soil processes and microbial communities

According to Jenny (1941) theory on soil formation, a soil is formed under the action of climate, organisms (mainly vegetation and humans), relief (topography), parent material and time. In order to be included, a soil-forming variable has to play a significant role in soil formation and be able to vary independently to the other variables included in the model.

There was some debate as to which extent vegetation should be a soil-forming factor. On one hand, Jenny cited Joffe (1936) who wrote in his book “Pedology”: “without plants, no soil can form”. He opposed him Robinson (1935 – “Soils of Great Britain”) who argued that the vegetation was not an independent variable since it is itself closely governed by situation, soil and climate. Bridging the two arguments, Jenny claimed that vegetation was indeed a soil- forming factor if it is considered for its biotic effect as a group and not as a set independent species which distribution is, indeed, dependant on other soil forming factors.

Jenny further stated that in order to effectively compared the effect two types of vegetation (or tree species) have on soil characteristics, all the other variables having effects on soil characteristics (i.e. climate, topography, parent material and time) must be kept constant. Such settings are now called common-garden designs – i.e. experimental designs in which all species grow side-by-side on the same parent material and de facto within the same climate.

27 Using only published common-garden studies, Augusto et al. (2015) identified distinct soil processes under evergreen gymnosperms (mainly coniferous tree species) and deciduous angiosperm forests. However, while 20 ecological functions were analysed, only 8 were considered consensual. Among them precipitation interception, soil acidity, accumulation of forest floor mass and weathering of soil minerals were higher under evergreen gymnosperms stands. Water throughfall, water seepage, decomposition of aboveground necromass and possibly decomposition of belowground necromass, accumulation of organic matter in mineral soils, nitrogen mineralization and nitrification rates were higher under deciduous angiosperm forests. It was hypothesized that the lower rates of nitrogen mineralization in evergreen gymnosperms were linked to the higher C:N ratio and lower pH of the litter which slowed down microbial decomposition process (Schimel and Bennett, 2004; Ste-Marie and Paré,

1999). Also, some compounds (polyphenols and terpenes) in coniferous litter could inhibit litter-decomposing microbial communities (Augusto et al., 2015). Finally, the nature of the mycorrhizal relationships could also explain differences in decomposition processes, N- transformation and C mineralization rates in the rhizosphere of trees. Cornelissen et al. (2001) noted that the litter of trees associated with arbuscular mycorrhizal and of ecto/arbuscular mycorrhizae species was more decomposable than the litter of trees associated with ericoid or ectomycorrhizal fungi in lowland and upland forests in the UK. In contrast, Phillips and Fahey

(2006) noted that N mineralization, nitrification and C mineralization rates were higher in the rhizosphere of ectomycorrhizal trees in a monospecific tree plantation of upstate New-York

(USA). Moreover, the presence of soil fauna under the nutrient and base cations-rich litter of deciduous forest can alter the process of litter decomposition, ammonification and nitrification

28 + - rates and increase soil NH4 and NO3 content of soils under deciduous stands (Laganière et al., 2009; Melillo, 1981; Nordin et al., 2001).

Tree species also affect the structure and abundance of soil microbial communities through the chemical nature of litter and exudates, as well as mycorrhizal fungal associates (Wardle, 2002;

Grayston & Prescott, 2005; Prescott & Grayston, 2013). Using PLFA analyses, Weand et al.

(2010) found a lower abundance of Gram-negative bacteria and higher abundance of fungi and

Actinomycetes as well as a higher fungal:bacteria ratio in forest floor material under western hemlock (Tsuga heterophylla) stands than under deciduous stands. This slightly contrasts with the higher abundance of Gram-positive bacteria and Actinomycetes, but lower fungal biomass and fungi:bacteria ratio found under western red cedar (Thuja plicata) stands compared to deciduous stands (Grayston and Prescott, 2005).

It must be noted that forest processes depend on a myriad of interactions among environmental factors that interplay at global and local scales and are, as such, a complex system that are difficult (possibly impossible) to translate into general trends (Augusto et al.,

2015; Prescott and Grayston, 2013; Prescott and Vesterdal, 2013). Hence, tree specie effects on soil characteristics are context-dependant and simplistic statements should be treated with caution before being apply locally.

1.6 Impacts of forest fire on the nitrogen cycle and soil microbial communities

Wildfires, the main natural disturbance in the boreal forest, influence the nitrogen cycle in many ways. Fire causes nutrient losses through oxidation of various forms of nitrogen

29 compounds into gaseous N2. Post-fire soil erosion due to the temporary reduction in vegetation also contributes to nutrient depletion. A boost in nutrient availability typically follows fires as a result of reduced competition among plants and the release of elements from organic matter in the form of ash (Binkley & Fisher, 2013). The addition of ash increases soil pH, and ash is also believed to be enriched in ammonium and depleted in nitrate, contributing to higher nitrification rates after fire (Binkley & Fisher, 2013; Raison et al., 2009). By reducing canopy cover, fires also alter soil temperature and moisture regimes, with attendant influences on soil microbial communities and rates of soil processes. The charcoal resulting from wildfires may also affect microbial communities and N cycling. Ball et al. (2010) found that addition of charcoal enhanced nitrification and abundance of ammonia-oxidizing bacteria abundance, possibly by absorbing allelochemicals, such as monoterpenes that inhibit ammonia mono-oxygenase activity and/or by creating local microsites with higher pH due to the high alkalinity of charcoal (Ball et al., 2010). Lysis of microbial cells and death of plants roots during fire will cause changes in microbial communities (Hart et al., 2005); increased nitrification rates may be partially attributable to reduced uptake of nitrate by microbial communities. Switzer et al. (2012) measured reductions in the abundance of Gram-positive bacteria, Gram-negative bacteria, actinomycetes, total bacteria, arbuscular mycorrhizae and fungi in forest floors after prescribed fire, which were associated with an increase in soil pH.

Microbial communities had not returned to their pre-fire status one year after fire. Yeager et al

(2005) also observed reductions in soil microbial biomass after fire and a shift in the structure of ammonia-oxidizing organisms. Pre-fire sites were co-dominated by cluster 1/2/4 of

Nitrosospira spp. In contrast, post-fire soils were dominated by a single cluster, the 3A cluster of Nitrosospira spp, which has been shown to respond favourably to environmental

30 + perturbations and higher concentration of soil NH4 (Yeager et al., 2005). In a fire- chronosequence, Sun et al (2005) found that Ascomycota-saprophytic fungi were more abundant in recently burned forest, probably because of the higher availability of easily decomposable material; with time since fire, Basidiomycota-mycorrhizal species increased in abundance and became the dominant fungal species in these soils. The post-fire survival of mycorrhizal species seems also to be influenced by their ability to survive without their hosts.

This ability has been suggested to be higher for arbuscular mycorrhizae than for ectomycorrhizae (Amaranthus and Perry, 1987; Hart et al., 2005). Finally, nitrogen-fixing organisms may also be more abundant after a fire, but only if the severity of the fire is high enough to trigger vegetation re-initiation (Yeager et al., 2005).

Depending on the intensity of the fire, these effects of fire may vanish after few seasons

(Raison et al., 2009; Binkley & Fisher, 2013), but can last more than 14 years (Ball et al.,

2010).

1.7 Restoration ecology

Ecological restoration is the process of assisting the recovery of an ecosystem that has been degraded, damaged, or destroyed (Society for Ecological Restoration 2004). The goal of ecological restoration has been to return the ecosystem to its previous ‘natural’ or ‘pristine’ condition; i.e. its historical baseline. Reclamation, on the other hand, does not aim to return the ecosystem to a historical state. Reclamation has been used to describe practices which restore ecological function to drastically disturbed sites or landscapes. In this sense, reclaiming something means rescuing it from an undesirable state in order to stabilise a

31 landscape while increasing its utility and economical value (Harris et al., 2005; Higgs, 2003).

Restoring soil ecological functions such as decomposition and nutrient cycling are critical to ecosystem reclamation and rely heavily on restoration of the soil physical and chemical properties and the soil biological community, as well as interactions between the above- and belowground ecosystem components.

1.8 Reclamation of soil nitrogen cycling processes in the Athabasca Region

Alberta’s first regulation on reclamation was voted in 1963 and was called the Surface

Reclamation Act. Since then, the laws have evolved and reclamation prescriptions are now encompassed in the Environment Protection and Enhancement Act (1993). This act states that all disturbed land should be returned to pre-disturbance land capabilities such that “… the ability of the land to support various land uses after conservation and reclamation is similar to the ability that existed prior to an activity being conducted on the land, but the individual land uses will not necessarily by identical” (Conservation and Reclamation Regulation - Alberta,

1993). Since ‘pre-disturbance’ land capability does not require the return to an historical condition, the objective is best considered as reclamation rather than restoration (Powter et al.,

2012).

Research on nitrogen cycling in reclaimed soils of the Athabasca oil sands region has so far portrayed a situation different from the natural boreal forest soils of northern Alberta. Firstly,

NOx emissions in the vicinity of the oil-sands mining sites have increased with the growth of the oil-sands industry (Golder, 2003). In 2011, atmospheric N deposition rates on oils sands

-1 -1 + reclaimed soils were approximately 15 kg N ha y and mainly in the NH4 form (Davis et al.,

32 2015; Hemsley, 2012). This contrasts with the 3 kg N ha-1y-1 typically deposited on natural boreal forest soils of northern Alberta (Davis et al., 2015). Secondly, the majority of soil

- available N in reconstructed soils is in the inorganic NO3 form; whereas, most soil available N

+ in the boreal forest is in organic and NH4 forms (Hemsley, 2012). MacKenzie and Quideau

(2012) also found higher nitrogen mineralization rates in the reclaimed material used for soil reconstruction than in the boreal forest soils, suggesting that the rates of both ammonification and nitrification rates are higher in the reconstructed soils. A simultaneous analysis of the entire nitrogen cycle, including an assessment of gross rates of ammonification and

+ - nitrification, as well as NH4 and NO3 immobilization is still needed.

Differences in soil microbial communities have also been reported between reconstructed and natural boreal forest soils. Using denaturing gradient gel electrophoresis (DGGE) and phospholipid fatty acid (PLFA) profiling, reduced abundance and distinct microbial community structure were found in reconstructed oil-sands soils ranging in age from 5-35 years compared with natural boreal-forest soils (Dimitriu et al., 2010; Hahn and Quideau,

2013)While DGGE and PLFA do not allow for in-depth (species-level) characterization of microbial communities, it was possible to detect a general increase in Gram-negative bacteria and of fungal phyla Zygomycota and Ascomycota in reconstructed soils compared to natural soils (Dimitriu et al., 2010). Dimitriu and Grayston (2010) showed that abiotic variables, such as pH and soil moisture, rather than plant cover, were the main drivers of community structure in both reconstructed and natural soils. Differences among reclaimed sites have also been reported, Sorenson et al. (2011) found that reconstructed soils under coniferous stands were associated with greater abundance of fungi, whereas deciduous stands had greater abundance

33 of bacteria. Vegetation influenced soil microbial communities only when canopy cover was above 30%; below this level, soil microbial communities were mainly influenced by characteristics of the material used during reclamation. Dimitriu et al. (2010) have also found that reclamation material characteristics had greater influence than time-since-reclamation in reconstructed soils between 5- to 30-year-old further underlining the importance of the reclamation material in the restoration effort.

These differences in nutrient availability and microbial communities between reconstructed soils and natural forest soils indicate that it is unlikely that these reconstructed forest soils will exactly mirror the natural boreal-forest soils (Chazdon, 2008; Hobbs et al., 2006). Thus, novel soil ecosystems will most probably result from the reconstruction and reclamation efforts. In this context, re-establishing ecological function rather than replicating structural qualities of the previous soil ecosystem is the most reasonable goal of reclamation and will assist in ensuring the long-term sustainability of reclaimed boreal forest landscapes (Quideau et al.,

2013).

1.9 Objectives and hypotheses

In this study I investigate differences in nitrogen cycling processes and communities between reconstructed soils and natural boreal-forest soils in order to assess if reconstructed soils can provide the same ecological functions as natural soils and therefore achieve previous land capabilities. I also evaluate if the vegetation planted during soil reconstruction influences nitrogen cycling processes and microbial communities. An in-depth characterization of microbial communities and of nitrogen transformation rates as well as their interactions with

34 soil abiotic and biotic characteristics and above-ground vegetation in mature reconstructed soils will deepen our understanding of these ecosystems and will allow us to delve into their ecological competencies. Oil sands extraction is going to be Canada’s biggest environmental legacy; our ability to recreate functional soils is fundamental in restoring these lands for future generations.

I compare fifteen 20-30-year-old reconstructed soils planted with coniferous trees (5 sites), deciduous trees (5 sites) and grass species (5 sites) with five natural boreal-forest soils in the vicinity of the mining operations1, Specifically, I :

(1) Assess if soil processes underlying the N cycle in reconstructed oil-sands soils are similar to those of natural boreal-forest soils. I hypothesize that a) N content and gross rates of N transformations will be higher in reconstructed soils than natural soils; b) gross ammonification and nitrification rates will be highest in reconstructed soils planted with deciduous trees;

(2) Assess if soil prokaryotic diversity and structure (i.e. α- and β- diversity) in reconstructed oil-sands soils are similar to those in natural boreal-forest soils. Relationships between prokaryotic α- and β- diversity and above- and below-ground characteristics will also be determined. I hypothesise that a) α-diversity will be higher in natural soils; b) β-diversity will differ between reconstructed and natural soils; and c) pH and soil moisture will be the main drivers of prokaryotic α- and β- diversity.

1 Soil characteristics measured for this project and vegetation assessment are available in Appendices A and B 35

(3) Assess if soil fungal diversity and structure in reconstructed oil-sands soils are similar to those in natural boreal-forest soils. Relationships between fungal α- and β- diversity and above- and below-ground characteristics will also be determined. I hypothesise that a) α-diversity will be higher in natural soils; b) β-diversity will differ between reconstructed and natural soils; c) Basidiomycota fungal species will be more abundant in natural soils, Ascomycota and Zygomycota fungal species will be more abundant in reconstructed soils, and Glomeromycota will be more abundant in reconstructed soils planted with grass; d) reconstructed soils planted with coniferous trees will have greater fungal abundance compared to reconstructed soils planted with deciduous trees or grass and e) pH and nutrient content of the soils will be the main drivers of fungal α- and β- diversity.

36 Chapter 2: Gross nitrogen transformation rates differ in reconstructed oil- sands soils and natural boreal-forest soils as revealed using a 15N tracing method2

2.1 Introduction

The Athabasca Oil Sands deposit, located in the boreal forests of northern Alberta, is part of the largest single oil deposit in the world, with proven reserves of 166 billion barrels of bitumen, and covering 142,200 km2 (Government of Alberta, 2016). Most (80%) of the bituminous sands can be extracted using in situ recovery methods, but 20% of the resource is shallow and can be recovered through open-pit mining (Government of Alberta, 2012). To date, about 895 km2 of land has been disturbed by oil-sands mining activity (Government of

Alberta, 2016). According to the Conservation and Reclamation Regulation of Alberta, following surface mining, companies are required to restore soils that can achieve equivalent land capability which is described as “the ability of the land to support various land uses after conservation and reclamation [that] is similar to the ability that existed prior to activity being conducted on the land, but that the individual land uses will not necessarily be identical”

(Government of Alberta, 1993; Powter et al., 2012). After soil reconstruction, the area is re- vegetated. When reclamation in the area began in the 1980’s, revegetation predominantly focused on erosion control, and used both native and introduced grasses and shrubs. However, more recent revegetation practices use native tree species (conifers i.e. jack pine, white and

2This chapter has been submitted (accepted Geoderma): Masse J., Prescott, C.E., Müller, C., Grayston, S.J. “Gross nitrogen transformation rates differ in reconstruced oil-sands soils from natural boreal forest soils as revealed using a 15N tracing method”. 37 black spruce; deciduous i.e. aspen) and understory shrubs (e.g. blueberry and willow), to re- establish a boreal-forest plant community. Differences in soil organic matter composition, soil available N and microbial communities between reconstructed oil sands soils and natural boreal-forest soils of northern Alberta have been previously identified (Dimitriu et al., 2010;

Hahn and Quideau, 2013; Hemsley, 2012; Turcotte et al., 2009). As such, it is unlikely that these reconstructed forest soils will exactly mirror pre-existing boreal-forest soils (Chazdon,

2008; Hobbs et al., 2006). Thus, novel soil ecosystems will most probably rise from the reconstruction efforts. Aiming at re-establishing soil functions, and chiefly nutrient cycling, rather than simply trying to replicate structural qualities of previous soil ecosystem is key to ensure the long-term sustainability of reclaimed boreal forest landscapes (Quideau et al.,

2013)

Nitrogen (N) is an essential nutrient for plant growth and metabolic activities, being a key element in amino acids, enzymes, proteins and nucleic acids (Binkley and Fisher, 2013; Lupi et al., 2013). In forest ecosystems, N availability influences photosynthetic rates, tree growth, root size, root structure and root distribution. In soil, N is present in many forms and transformations of N from one form to another are mediated by soil microorganisms

+ - (Robertson and Groffman, 2007). Inorganic forms of N – ammonium (NH4 ) and nitrate (NO3

) – rarely comprise more than 1% of the total N pool, and were previously assumed to be the only plant-available forms of N. However, it is now recognized that plants can take up simple organic N compounds, such as amino acids and peptides (Näsholm et al., 2009; Paungfoo-

Lonhienne et al., 2010; Inselsbacher and Näsholm, 2012).

38 Canadian boreal forest ecosystems are generally N-poor environments, and N availability is the primary limitation to plant productivity (Vitousek and Howarth, 1990; Matson, et al.,

+ 2002). In Alberta, gross rates of ammonification (mineralization of NH4 ) in mature upland

+ -1 -1 + forest soils ranged from 3.75 to 164 mg NH4 kg soil day and gross rates of NH4 immobilization were similar, resulting in low net ammonification rates (Carmosini et al.,

+ 2002; Cheng et al., 2013). Gross rates of ammonification and NH4 immobilization were lower

-1 -1 (14 to 19 mg NH4 kg soil day ) in boreal forest soils in Ontario, but net ammonification was still close to zero (Westbrook and Devito, 2004). Gross rates of nitrification in the Canadian

- -1 -1 boreal forest are highly variable, but generally close to 0 mg NO3 kg soil day (Carmosini et al., 2002; Cheng et al., 2013; Westbrook and Devito, 2004). Stark and Hart (1997) measured high gross nitrification rates (25 mgN m-2 day-1in a ponderosa pine sites in New Mexico during summer to >300 mg N m-2 day-1 in a Douglas-fir site during spring) in eleven

- undisturbed forest ecosystems of New Mexico and Oregon. Gross rates of NO3 consumption were similar to gross nitrification rates resulting in nill net nitrification. Through their effects on litter chemistry, tree species can also influence N-transformation rates in boreal forest soils.

Studies comparing N-cycling in boreal forest soils under deciduous stands (composed mainly of trembling aspen (Populus tremuloides), balsam poplar (Populus baslamifera) and paper birch (Betula papyrifera)) and coniferous stands (composed mainly of black spruce (Picea mariana), white spruce (Picea glauca), white cedar (Thuja occidentalis) and jack pine (Pinus banksiana)) found higher pH, higher base-cation contents and lower C:N ratio in forest floors of deciduous stands compared to forest floors of coniferous stands (Jerabkova et al., 2006;

+ Paré and Bergeron, 1996; Ste-Marie and Paré, 1999). Higher N availability, NH4 pools, net rates of mineralization and nitrification, and nitrate accumulation were also measured in forest

39 floors under deciduous stands compared to coniferous stands. The higher N content of deciduous litter positively affects net N mineralization rates, while the high C:N ratio and low pH typical of coniferous litter seems to reduce net mineralization and nitrification rates

(Jerabkova et al., 2006; Paré and Bergeron, 1996; Ste-Marie and Paré, 1999).

Fire is the primary natural disturbance in the boreal forest and both nitrogen mineralization and nitrification rates increase following fire (Ball et al., 2010; Binkley and Fisher, 2013;

Raison et al. , 2009; Yeager et al., 2005). Depending on the intensity of the fire, these effects may vanish after few seasons (Raison et al., 2009; Binkley & Fisher, 2013), but can last more than 14 years (Ball et al., 2010). Historically low, throughfall N-deposition in the boreal forest of the Athabasca Oil Sands Region (AOSR) are higher near mining sites and decreased to background level with distance (Fenn et al., 2015; Proemse et al., 2013). Ammonium deposition measured in the close vicinity of the mining sites (<3km) varied between of 14.7

+ -1 -1 and 19.6 kg NH4 ha year ; whereas nitrate depositions were lower, ranging between 2.1 and

- -1 -1 6.7 kg NO3 ha year (Fenn et al., 2015; Hemsley, 2012). These depositions were reduced to

+ -1 -1 - -1 -1 0.81 kg NH4 ha year and 0.27 kg NO3 ha year 120 km away from the mining sites

(Fenn et al., 2015).

In order to evaluate if the reconstructed soils of the AOSR were able to re-established key soil functions such as nutrient cycling, the objective of this study was to assess if soil N- transformation rates in oil-sands soils that were reconstructed 20-30 years previously are similar to those of natural boreal-forest soils that were subject to wildfire disturbance at approximately the same time. In a previous study of N-cycling in reconstructed oil-sands soils,

40 McMillan et al. (2007) detected no differences in gross ammonification rates between a 5- year-old reconstructed soil and a natural aspen-forest soil despite the reconstructed soil having higher microbial biomass and higher N content. In contrast to McMillan et al. (2007), we examined soils at least 20 years after reconstruction, at which point previous studies in the oils sands region (Rowland et al., 2009) and elsewhere (Frouz et al., 2001, 2008; Šourková et al.,

2005) indicate that reconstructed soils have stabilized to some extent with respect to vegetation cover and composition, forest-floor development, nutrient-cycling processes, soil carbon, and soil faunal communities. We evaluated gross rates of N transformation, as this allowed us to separately examine N production and consumption rates in order to better understand N cycling processes in the reconstructed soils. We also assessed the influence of the vegetation treatments (conifers, deciduous trees and grasses) used in reclamation on N- cycling rates. We hypothesize that a) N content and gross rates of N transformations will be higher in reconstructed soils than natural soils; b) gross ammonification and nitrification rates will be highest in reconstructed soils planted deciduous trees.

2.2 Materials and methods

2.2.1 Study area

The study area was situated in the Athabasca Oil Sands Region (AOSR) in northern Alberta,

Canada (56°39’N, 111°13’W, altitude: 369m). Short warm summers and long cold winters characterize the climate. The mean annual temperature is 1°C, ranges from -17.4°C in January to 17.1°C in July. Mean annual precipitation is 418.6 mm, of which 316.3 mm occurs as rainfall during the growing season (Environment Canada, 2015). Medium- to fine-textured

Gray Luvisols and Dystric Brunisols underlie landscapes shaped by the impact of Pleistocene

41 ice activity, deglaciation and post-glacial modifications in upland areas. Organic soils are found under wetland areas (Natural Regions Committee, 2006). This region falls within the central mixedwood region of the Canadian boreal forest. Dominant tree canopy species in upland landscapes are trembling aspen (Populus temuloides Michx), white spruce (Picea glauca (Moench) Voss) and Jack pine (Pinus banksiana Lamb) (Natural Regions Committee,

2006). Fire is the major natural disturbance in these forests (Thomson, 1979).

Oil-sands mining activities involve the removal of surface soil materials followed by the removal of 40 m of overburden material (approximate regional average) to expose the oil-sand ore body. Salvaged soil materials are preferably used for reclamation of an area within the footprint of the mines that is ready for reclamation, or are stockpiled for later use. The overburden is used for berm, dyke wall or road construction, or deposited in a dedicated disposal area to create large-scale overburden landform units. The oil-sand ore is transported to the extraction and upgrading facility. Oil-sands soil reconstruction involves a number of cover designs, depending on the landform substrate being reclaimed. There are two main cover designs: one that uses only cover-soil and the other consisting of a combination of cover-soil and subsoil. Cover-soil and subsoil materials are salvaged from surface soils within the mine-development footprint. Only sites at which cover-soil had been placed on top of overburden material were used in this study. The cover-soil materials used consisted of surface peat mixed with mineral soils material having a loam or coarser texture and is hereafter referred to as ‘peat-mineral mix’. In the studied soils, the depth of the peat-mineral mix ranged from 12 cm to more than 100 cm. Early revegetation objectives in the AOSR were to establish native or introduced grass and shrub species to control erosion; however, oil-sands

42 operators are now required to use native trees and species with the intention of promoting the re-establishment of a boreal forest community. During the period when the sites used in this study were reclaimed (20-30 years ago), 250 to 350 kg ha-1 of varying proportions of N:P:K fertilizer was typically applied in the first year of re-vegetation. Some oil-sands operators also applied fertilizer annually for four additional years.

2.2.2 Study sites

Nine reconstructed sites and three natural forest sites were studied in the AOSR; all sites were reconstructed (or naturally fire-disturbed) 20 to 30 years previously. The 12 reconstructed sites were previously studied by Sorenson et al. (2011). Among them, three sites (D1, D2 and D3) were planted with deciduous species (mostly trembling aspen), three (C1, C2 and C3) with coniferous species (mostly white spruce), and three (G1, G2 and G3) with grasses (fescue, slender wheatgrass and alfalfa). For reference sites, we selected three sites (N1, N2 and N3) outside of the mining footprint, at which the stands were of similar age-since-disturbance as the reconstructed sites. The disturbance at these sites was a natural stand-replacing wildfire, which boreal forests in this region experience every 250 years or less (Bergeron et al., 2001;

Binkley and Fisher, 2013). As these sites would be at approximately the same stage in the post-disturbance trajectory as the reconstructed sites, they were considered to provide a more realistic reference for the reconstructed soils than would soils from older natural forests. Soils at the natural sites were classified as Brunisolic Gray Luvisol (soils N1 and N2) and Eluviated

Eutric Brunisol (soil N3). They are located approximately 40 to 150 km south of Fort

McMurray (Alberta, Canada).

43 At each site, one 10-m2 plot was sub-divided into 10 1-m2, from which 7 subplots were randomly selected. A 30-cm-deep soil pit was dug and carefully described in each sub-plot.

The litter and forest floor layers were removed and the top 0 – 15 cm of the mineral part of the soil (i.e. A and top of B horizons) was sampled (~ 1 kg) at each sub-plot using a plastic trowel and the 7 subplot samples were pooled to produce one sample per plot (nsample = 12). A sub- sample (~5 g) of each soil was immediately put on ice for microbial analyses. The remaining samples were put in a cooler and kept cold (4°C) until soil chemical and physical analyses.

Removal of any surface organic material prior to sampling enabled us to directly compare the mineral substrate on which the organic layers develop, and so determine how similar the reconstructed soils are to natural soils. If we assume that the similar soil development factors – the action of climate, organisms, relief through time – are at play in reconstructed and natural boreal forest soils with few natural variations such as parent material, types of organisms, etc., the peat-mineral mix is the analogue of the mineral part of the natural soils on top of which a forest floor develops. Also, characteristics of the forest floor layers in reconstructed and natural soils in the AOSR have been described in previous research (Rowland et al., 2009).

Vegetation is described in (Anderson, 2014) (Appendices B). Tree density was not significantly different among reconstructed soils planted with deciduous species, coniferous species and naturally disturbed sites. Grassland sites were devoid of trees.

44 2.2.3 Laboratory analyses

2.2.3.1 Soil characteristics

To estimate soil bulk density at each site a soil core was collected from one of the sub-plots.

The weight of the soil core was determined after drying at 105°C. The final bulk density was calculated by dividing the dry weight of the soil sample by the volume of the core (Blake and

Hartge, 1986). Soil pH in water was determined at a 1:10 ratio (soil:water) using a UB-10 pH meter (Denver instruments, Bohemia, USA) (Thomas, 1996). Soil samples (<2mm) were send to the laboratoire de pédologie de l’Université de Montréal for cation exchange capacity

(CEC) measuring which was calculated as the sum of Ca, Mg, K, Na, Al, Fe and Mn that could be exchanged by BaCl2 (Hendershot and Duquette, 1986). Soil total carbon and nitrogen content were measured with an Elementar Vario El Cube Elemental Analyzer (Elementar,

Hanau, Germany) using 15 mg of soil sieved (< 0.5 mm) to insured homogenization

(Rutherford et al., 2008; Skjemstad and Baldock, 2008). Soil particle size of sand (60 µm to 2 mm) and clay (< 2 µm) were determined in the Environmental Geoscience Research Centre at

Trent University using laser diffraction with a Partica LA-950 (Horiba Ltd., Kyoto, Japan)

(Konert and Vandenberghe, 1997). Analyses were completed in triplicate after homogenizing the sample. Microbial biomass carbon was measured on the day of the sampling using the chloroform fumigation extraction (CFE) technique (Tate et al., 1988). Ten g sub-samples were weighed; the first sub-sample was immediately extracted in 0.5M K2SO4, shaken for 1 hour and filtered through 1.2-µm glass-fibre filter (Fisher scientific, Waltham, Massachusetts,

USA). The second sub-sample was put in a desiccator in which the air was replaced with 10 ml of ethanol-free chloroform. The chloroform was safely evacuated after 24 h and the carbon in the soil sample was extracted using the same technique as for the first sub-sample. The

45 samples were kept cold until analysed at the University of Alberta. Total organic carbon

(TOC) from each sub-sample was measured by oxidative combustion and infrared analysis using a TOC-TN autoanalyzer (Shimadzu, Kyoto, Japan). Microbial biomass C was calculated as the difference in TOC from the fumigated sub-sample and the unfumigated sub-sample. A calibration factor (kEC) is typically applied to extractable MBC to estimate total biomass C

(Sparling et al., 1990). No constant was used in this study because reconstructed soils differ from natural soils, particularly due to the peat amendment, and thus applying a kEC that is based on natural soils would probably produce inaccurate results. Therefore, extractable microbial biomass C, as opposed to total microbial biomass C, was compared among sites, and is referred to as microbial biomass C in this study. Results are presented in grams of C per kg of soil dry weight. Unless otherwise mentioned, a quarter of all samples (n=3) were randomly selected and analysed in duplicate and controls (AgroMAT AG-2 soil standard, SCP Science,

Baie d’Urfé, Canada) were added.

2.2.4 Quantification of gross rates of N-transformation

For each site (n=12) a 15N tracing study was carried out following the procedure of Müller et al. (2004; 2007). Briefly, 120g of wet soil (sieved at < 5 mm) were placed in six jars per sample (total number of jars=72). The jars were then sealed with a parafilm punctured with seven holes and incubated overnight, in the dark, at room temperature. The following day, three jars per site were 15N-labelled using 4 ml solution containing 60 at. % excess of

15 14 15 NH4 NO3. The remaining three jars were N-labelled using a 4 ml solution containing 60

14 15 15 + 15 - at. % excess of NH4 NO3 (njars=72). The added NH4 and NO3 in the injected solution

+ - contained less than 2% of the original soil NH4 or NO3 content to avoid any “priming

46 effects” on the microbial communities. Labelled N was injected in 1ml intervals four times over one minute using a plastic syringe, gently homogenized to ensure uniformity of the 15N application within the soil sample and the parafilm seal was replaced. At times 3h, 24h, 48h and 288h after labelling, inorganic N was extracted from each jar using 10 g of soil and 100 ml of 2M KCl (1:10 ratiow:v). Samples were shaken for 1 hour on a mechanical shaker and filtered through a ø 12.5 cm fiberglass G6 microfilter (Fisher Scientific, Waltham,

Massachusetts, USA). Part of each extract (10 ml) was sent to the University of British

Columbia Environmental Engineering Department to be analysed using a Lachat Flow

Injection Analyzer (FIA) QuickChem 8000 (Lachat Instruments, Loveland, Colorado, USA)

+ - for colorimetric determination of total NH4 and NO3 concentrations. The remainder of each extract (90 ml) was stored in specimen cups (Straplex Scientific inc., Etobicoke, Ontario,

15 15 - - Canada) to be used for the diffusion of NH4 and NO3 following the International Atomic

Energy Agency protocol (IAEA, 2001). An acidified (2.5 M KHSO4) glass-fibre disc

(Whatman glass-microfiber filters grade 934-H cut to a ø 1 cm) sealed in Teflon

(polytetrafluoroethylene) tape is added to each cup. Then 0.2 g of magnesium oxide (MgO) was added and the cups were immediately sealed. Samples were shaken on a horizontal shaker

+ at 100 rpm. MgO causes the pH to increase, which slowly liberates NH4 from the solution in a gaseous ammonia (NH3) form that is collected on the filter disc. After 72 h, the disks were removed from the Teflon tape and dried in a dessicator along with a beaker of HCl (17.5%) used to trap atmospheric NH3. The specimen cups were then left open overnight in a dark room to liberate any remaining trace of NH3. Another acidified disc and 0.2 g of Devarda’s alloy were added into the cups, which were immediately sealed. The Devarda’s alloy reduces

- + the NO3 to NH4 , which in turn is liberated from the solution as NH3 and is collected on the

47 acidified disc. Once again, the cups were put on a horizontal shaken for 72 h and disks were collected and put in a dessicator overnight. Once dried, disks were packaged in tin cups and sent to the Stable Isotope Facility of the Department of Soil Science at the University of

Saskatchewan where the 15N ratio was measured using Costech ECS4010 elemental analyzer

(Costech Analytical, Ventura, CA, USA), coupled to a Thermo Scientific mass spectrometer with a Conflo IV interface (Thermo Scientific, Bremen, Germany) (IAEA, 2001). At each extraction time, a subsample (2 g) of soil from each jar was also collected in order to measure soil moisture content. This was used to transform data on a dry weight basis. The soil water content was constant during the experiment.

2.2.5 Calculations and statistics

2.2.5.1 Gross rates of N transformation – analytical solution

We first used Hart et al. (1994) equations as described in Drury et al. (2008) to calculate gross

+ rates of ammonification (Equation 2.1), NH4 consumption (Equation 2.2), nitrification

- (Equation 2.3) and of NO3 consumption (Equation 2.4). These equations are all based on the pioneer work of Kirkham and Bartholomew (1954).

+ + + + GAR = (([NH4 ]0 - [NH4 ]t)/t *(log(APEA0/APEAt)/log[NH4 ]0/[NH4 ]t)) (eq. 2.1)

+ + Ca = GAR - (([NH4 ]t – [NH4 ]0)/t (eq. 2.2)

- - - - GNR = (([NO3 ]0 - NO3 ]t)/t *(log(APEN0/APENt)/log NO3 ]0/[ NO3 ]t)) (eq. 2.3)

- - Cn = GAR - (([NO3 ]t – [NO3 ]0)/t (eq. 2.4)

Where

+ -1 -1 GAR = gross ammonification rate (mg of NH4 -N kg dry soil day ).

48 + + -1 [NH4 ]0 = Total NH4 concentration (mg N kg dry soil ) at time 0.

+ + -1 [NH4 ]t = Total NH4 concentration (mg N kg dry soil ) at time t.

15 + APEA0 = atom% N excess of NH4 pool at time 0.

15 + APEAt = atom% N excess of NH4 pool at time t.

+ + -1 -1 Ca = NH4 consumption rate (mg of NH4 -N kg dry soil day ).

- -1 -1 GNR = gross nitrification rate (mg of NO3 -N kg dry soil day ).

- - -1 [NO3 ]0 = Total NO3 concentration (mg N kg dry soil ) at time 0.

- - -1 [NO3 ]t = Total NO3 concentration (mg N kg dry soil ) at time t.

15 - APEA0 = atom% N excess of NO3 pool at time 0.

15 - APEAt = atom% N excess of NO3 pool at time t.

- - -1 -1 Cn = NO3 consumption rate (mg of NO3 -N kg dry soil day ). t = 2 days (48 hours of incubation time).

Gross rates of N-transformation (GR) were transformed in mg N m-2 day-1 using Equation 2.5.

GR (mg N m-2 day-1) = GR (mg N kg dry soil-1 day-1) * bulk density (kg dry soil m-3) * 0.12

(m)

(eq. 2.5)

This unit was chosen because soil bulk density was significantly different in reconstructed and natural soils. Accounting for these differences by expressing values as content (m-2) rather than concentration (kg-1) allows us to make effective comparisons among the treatments. We used a depth of 0.12 m in the equation to reflect the average depth of sampling.

2.2.5.2 Gross rates of N transformations – numerical solution

To estimate rates of additional N-transformation, we used the 15N tracing model developed by

Müller et al (2007). This model uses Markov Chain Monte Carlo method to optimize the fit of the data to a conceptual model chosen by the user. The method allows for the determination of the probability density functions of a whole set of parameters, which are, in this case, the transformation rates. In total, seven gross N transformation rates were estimated, following the procedure outlined by Müller et al. (2007). This method is also one of the few that can estimate Michaelis-Menten kinetics ⎯ i.e. reaction models that best represent enzymatic reaction in which the transformation rate is link to the substrate pool size. The conceptual model chosen for this study (Figure 2.1) comprises four pools of nitrogen and seven rates of

N-transformation. The names and descriptions of the pools and rates are based on the current understanding of the N-cycle (Müller et al., 2007, 2004; Schimel and Bennett, 2004). The organic-nitrogen (SON) pool was calculated as the difference between soil total nitrogen (TN)

+ - content and soil inorganic nitrogen content (SON= TN – ([NH4 ] + [NO3 ])). SON was further divided in two pools: the more active labile organic-nitrogen pool was calculated as 1% SON and represent the microbial biomass pool (which undergoes quick N transformations) and possibly N-monomers; and the more passive recalcitrant organic-N pool was calculated as

99% SON and comprises more complex forms of organic nitrogen. The other two pools are

+ - + - + the inorganic pools: NH4 and NO3 . The initial pools (SON, NH4 and NO3 ) as well as NH4 ,

- 15/14 NO3 concentration and N ratio at each extraction time were provided. This method is not a simulation model, but a 15N-tracing model with the purpose of quantifying simultaneous gross

N transformation rates. Although the model was developed using mainly grassland soils, it is

50 adaptable to a variety of soils. The strategy is to start with a simple model containing the most essential N-transformation processes and then add additional realistic N pools and N transformation rates depending on the study soils (Müller et al., 2007, 2004). During optimisation several model versions (differing in kinetics, number of N pools and N- transformation rates) are tested and only the simplest one, which still adequately represents the observations, is retained (as evaluated by Akaike Information Criteria AIC). Non-significant transformation rates (with probability density function towards zero) were not considered for downstream analyses. The transformation rates retained are illustrated in Figure 2.1 and described in Table 2.1. To account for differences in bulk density, gross rates of N- transformation are expressed in mg N m-2 day-1 (Appendix C.1) using Equation 2.5. Gross

-1 rates of N-transformation expressed in mg N kgdrysoil day are available in Appendix C.2.

Labile organic Recalcitrant organic nitrogen nitrogen

INH4-Nrec I NH4-Nlab N Nrec INO3-Nrec ANlab ANrec

+ - NH4 NO3 NNH4

Figure 2.1 Conceptual 15N tracing model to analyze N transformation in soil (Müller et al. 2004).

For explanation on transformation rates and parameters see table 1

Table 2.1 Description of parameters (rates) used in the conceptual N-cycle model presented in figure 2.1 Parameters (rates) Description kinetica ANlab Ammonification of labile soil organic N 1 INH4-Nlab Immobilization of NH4 in labile soil organic N 0 INH4-Nrec Immobilization of NH4 in recalcitrant soil organic N 0 ANrec Ammonification of recalcitrant soil organic N 0 + 3- NNH4 Nitrification (oxidation) of NH4 into NO 1 - NNrec Nitrification (oxidation) of recalcitrant organic N into NO3 0 INO3-Nrec Immobilization of NO3 to recalcitrant soil organic N 0 a 0 = zero-order kinetics, 1 = first-order kinetics

2.2.5.3 Net rates of ammonification and nitrification

Net ammonification and net nitrification rates for each soil were calculated using Equation 2.6 and 2.7, respectively.

+ Net ammonification rate = Total ammonification rate – total NH4 immobilization rate (eq. 2.6)

+ Net nitrification rate = Total nitrification rate – total NO3 immobilization rate (eq. 2.7)

Where

Total ammonification rate = Ammonification rate from labile organic N + ammonification rate from recalcitrant organic-N

+ + Total NH4 immobilization rate = immobilization rate NH4 into labile organic-N +

+ immobilization rate of NH4 into recalcitrant organic-N.

+ Total nitrification rate = Nitrification rate of recalcitrant organic-N + nitrification rate of NH4

- - total NO3 immobilization rates = immobilization rate of NO3 into recalcitrant organic-N

52 2.2.5.4 Turnover times

+ + NH4 and NO3 turnover time (or mean residence times) were calculated with Equations 2.8 and 2.9. They represent the amount of time an atom of N stays in a pool before being transformed. Mean and standard errors of the turnover rates for each treatment are presented.

+ + NH4 turnover time = Total ammonification / NH4 pool (eq. 2.8)

- - NO3 turnover time = Total nitrification / NO3 pool (eq. 2.9)

Where

Total ammonification rate = Ammonification rate from labile organic N + ammonification rate from recalcitrant organic-N

+ Total nitrification rate = Nitrification rate of recalcitrant organic-N + nitrification rate of NH4

Differences among treatment types (reconstructed with deciduous, coniferous and grasses species and naturally disturbed sites) were assessed using one-way ANOVA with permutation testing (Legendre, 2007). When significant differences were observed, post-hoc tests were performed using the Kruskall-Wallis test from the R package pgirmess (Giraudoux, 2014).

Differences between calculation techniques of gross N-transformation rates were assessed using Wilcoxon signed-rank test. Significance level was set at α=0.1. We accepted an α of 0.1 because it represents a satisfying balance between the detection of statistical differences with a low sample size (n=12) and the probability of being wrong when rejecting the null hypothesis

(type 1 error) (Wasserstein and Lazar, 2016) Statistical analyses were done using the software

R (R Core Team, 2014).

53 2.3 Results

2.3.1 Soil biological, physical and chemical characteristics

Key plot characteristics influencing N-cycling rates, such as site age (F(3,16)=1,49, p=0.255) and soil texture (F(3,16)=7.20, p=0.003), were similar among reconstructed planted with deciduous trees, coniferous trees and grass and natural forest soils. Soil pH values ranged from

6.04 to 7.11 and were highest at the grassland sites and lowest at the natural sites (F(3,16)=8.08,

-6 p=0.002). Cation exchange capacity (CEC) (F(3,16)=23.39, p=4.31x10 ) and microbial biomass

C (F(3,16)=3.44, p=0.042) were higher in reconstructed soils than in natural soils. Bulk density was lower in the reconstructed soils than in the natural soils (F(3,16)=9.94, p=0.0006) (Table

2.2).

Table 2.2 Selected physical, chemical and biological characteristics of each of the study soils

CEC Microbial biomass carbon Bulk density Site pH -1 -1 -3 (cmol+ kg soil ) (gC kgsoil ) (g cm ) Coniferous (mean±sd) 6.5±0.2ab 37.7±11.0ab 1.31±0.37ab 0.92±0.25a C1 6.4 31.3 1.15 1.06 C2 6.4 30.6 1.60 0.78 C3 6.6 56.9 1.27 0.57 C4 6.9 36.5 1.73 0.98 C5 6.5 33.2 0.79 1.19 Deciduous (mean±sd) 6.4±0.3ab 50.5±14.7a 1.53±0.84a 0.78±0.25a D1 6.5 68.8 1.53 0.75 D2 5.8 47.2 2.84 0.36 D3 6.5 34.9 0.58 1.05 D4 6.6 39.1 1.64 0.95 D5 6.5 62.4 1.08 0.81 Grassland (mean±sd) 7.0±0.3a 52.3±6.4a 0.93±0.44ab 0.99±0.17ab G1 6.5 57.8 0.98 0.73 G2 7.2 59.1 1.35 1.13 G3 7.1 50.5 1.03 1.15 G4 7.1 43.3 0.17 1.05 G5 7.0 50.8 1.10 0.91 Natural (mean±sd) 6.1±0.3b 2.6±1.2b 0.53±0.29b 1.46±0.20a N1 6.6 1.3 0.93 1.69 N2 6.1 3.1 0.24 1.29 N3 5.8 3.6 0.35 1.32 N4 5.9 3.7 0.41 1.65 N5 6.1 1.5 0.73 1.32

54 2.3.2 Soil nitrogen content

Total N content in the first 15 cm of mineral soil (or peat-mineral mix) ranged between 50 and

600 g N m-2 (Figure 2.2A). More than 95% of this total N was in the form of organic N

(Figure 2.2A). Reconstructed soils had significantly higher total N content than natural soils

(F(3,16)=7.20, p=0.003). Grassland soils had higher total N content followed by coniferous and

+ - deciduous soils. NH4 (F(3,16)=0.66, p=0.591) and NO3 (F(3,16)=1.36, p=0.365) content were

-2 + similar among all treatments and ranged from 0.75 to about 0.8 g m (Figure 2.2B). NH4 was

- the dominant inorganic N form in coniferous soils; NO3 was dominant in deciduous and

+ - grassland soils, and there were similar proportions of NH4 and NO3 in natural soils.

600 a 1.00 Organic nitrogen A NH4-N B p<0.01 p>0.1 Inorganic nitrogen NO3-N

0.75 a

)) 400 -2

0.50 Nitrogen content (g m (g content Nitrogen

200

0.25

0 0

Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

55 Total C and total N concentrations were strongly positively correlated (Figure 2.3). C:N ratios

of peat-mineral mix at reconstructed sites were between 15 and 25, while mineral soils at

natural sites had C: N ratios of around 10 (Figure 2.4).

0.8 %N = 0.049 %C + 0.012 R2 = 0.8531; p<0.001

0.6

0.4 Total N (%) N Total

0.2

0.0

0510 15 Total C (%)

Figure 2.3 Linear regression between total carbon content (%) and total nitrogen content (%) in coniferous (C); deciduous (D); grassland soils (G) and natural forest soils (N)

25 a ab p<0.001 a

C:N ratio C:N 15

Coniferous Deciduous Grassland Natural

56 2.3.3 Gross N transformation rates

2.3.3.1 Gross N-transformation rates – analytical solution

+ Gross rates of ammonification (F(3,8)=3.07;p=0.091) and NH4 consumption

(F(3,8)=4.04;p=0.05) were higher in reconstructed soils than in natural boreal-forest soils (Table

- 2.3). Gross rates of nitrification (F(3,8)=1.88;p=0.211) and of NO3 consumption

(F(3,8)=0.09;p=0.962) did not differ among treatment (Table 2.3).

Table 2.3 Gross ammonification, NH4+ consumption, nitrification and NO3- consumption rates calculated using Hart et al. (1994) equations + - Gross Gross NH4 Gross Gross NO3 ammonification consumption nitrification consumption ------mg N m-2 day-1 ------

Coniferous 286.7±154.4a 303.8±175.4a 553.7±473.3a 277.2.0±520.8a (mean±sd) C1 145.1 127.7 62.8 77.4 C2 263.7 305.3 591.2 -114.2 C3 451.3 478.5 1007.1 868.4 Deciduous 302.2±94.8a 359.5±170.8a 170.8±269.2a 112.2±69.6a (mean±sd) D1 341.0 397.5 481.7 190.6 D2 371.5 508.1 13.3 57.5 D3 194.1 173.0 17.5 88.5 Grassland 176.4±82.4ab 135.2±54.7ab 1156.4±1162.4a 129.4±1031.2a (mean±sd) G1 242.9 189.6 715.0 491.8 G2 202.2 135.6 2474.8 930.4 G3 84.2 80.3 279.4 -1034.1 Natural 80.8±48.7b 45.6±14.3a 14.3±12.5a 32.2±62.5a (mean±sd) N1 46.5 32.0 0.0 0.0 N2 136.6 60.5 19.9 104.2 N3 59.4 44.3 23.0 -7.6

2.3.3.2 Gross N-transformation rates – numerical solution

2.3.3.2.1 Gross ammonification rates

+ In the conceptual model used for this study, ammonification (mineralization of NH4 ) could originate from two organic-N pools: the labile organic-N pool and the recalcitrant organic-N

+ pool. In both reconstructed and natural forest soils, more NH4 was transformed from

-2 -1 recalcitrant organic-N pool (≈ 350 mg NH4-N m day ) (Figure 2.5A) than from the labile

-2 -1 organic-N pool (≈ 0.5 mg NH4-N m / day ) (Figure 2.5B).

Reconstructed soils ammonified (F(3,16)=3.87, p=0.055) and immobilized (F(3,16)=3.14,

+ p=0.087) significantly more NH4 from the recalcitrant organic-N pool than did natural soils.

No significant differences in these process rates were observed among the different vegetation types within the reconstructed sites. (Figure 2.5A and D).

Ammonification rates from the labile organic-N pool were significantly lower in reconstructed

-2 -1 -2 -1 soils (≈ 0.1 mg NH4-N m day ) than in the natural soils (≈ 0.6 mg NH4-N m day )

-5 (F(3,16)=38.24, p=4.33x10 ) (Figure 2.5B). There were no significant differences among any

+ type of soils for rates of immobilization of NH4 into the labile organic-N pool (F(3,16)=2.23,

-2 - p=0.163). Ammonification from the labile organic-N pool was low (≈ 0.1 mg NH4-N m day

1 + -2 -1 ) compared to immobilization of NH4 into this pool (≈ 60 mg NH4-N m day ) in all soils

(Figure 2.5 B and E).

58 a a A a D 400 400 a a into 4

300 300 a

200 200 recalcitrant organic N organic recalcitrant

100 NH of Immobilization 100 b p < 0.1 p < 0.1 b Ammonification of recalcitrant organic N organic recalcitrant of Ammonification Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

120 B a E

0.6 a

into into 90 4 b

0.4 a 60 a labile organic N organic labile 0.2 p < 0.001 30 a NH of Immobilization a a Ammonification of labile organic N organic labile of Ammonification 0.0 0 Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

a a a a C F

400 400 a a

300 300 immobilization 4 200 200 Total ammonification Total Total NH Total 100 100 b p < 0.1 b p < 0.1

Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

59 + Total ammonification and NH4 immobilization represent the joint contribution from (and to)

the recalcitrant and the labile organic-N pools. Reconstructed soils ammonified (F(3,16)=3.86,

+ p=0.056) and immobilized (F(3,16)=3.23, p=0.082) more NH4 per square meter than natural

soils. Within the reconstructed soils, there were no differences in rates of ammonification and

+ NH4 immobilization among vegetation treatments (Figure 2.5 C and F). Total rates of

+ ammonification (W=56; p=0.377) and of NH4 consumption (W=59; p=0.478) were similar

using either the analytical or the numerical solution.

2.3.3.2.2 Gross nitrification rates

In the conceptual model used for this study, nitrification could originate from the oxidation of

+ recalcitrant organic N (heterotrophic nitrification) or the oxidation of NH4 (autotrophic

nitrification).

A B

2000 2000

1000 1000 Nitrification of ammonium of Nitrification 0 0

Nitrification of recalcitrant organic N organic recalcitrant of Nitrification Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

C D into - 2000 3 2000

1000 1000 Total nitrification Total recalcitrant organic N organic recalcitrant Immobilization of NO of Immobilization 0 0 Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

-2 -1 Nitrification from the oxidation of recalcitrant organic N (≈ 1000 mg NO3-N m day ) was

+ -2 -1 much higher than from the oxidation of NH4 (≈ 50 mg NO3-N m day ) in all soils (Figure

2.6 A, B).

Gross rates of nitrification from the recalcitrant organic pool (F(3,16)=1.12, p=0.395) and from

+ the NH4 pool (F(3,16)=0.90, p=0.484) did not differ significantly between reconstructed and natural soils, or among vegetation treatments (Figure 2.6A and 2.6B). In coniferous and grassland soils, total gross nitrification rates from the recalcitrant organic-N pool (1000 mg

-2 -1 NO3-N m day ) (Figure 2.6C) were higher than ammonification rates from the same pool (≈

-2 -1 400 mg NH4-N m day ) (Figure 2.5A). These high rates were driven by “extreme values” which were measured in soils that had either a large amount of peat material (one of the coniferous sites) or a lot of macrofaunal activity (one of the grassland sites) compared to the other sites. When we removed data from these sites, gross nitrification rates were around 400

-2 -1 mg NO3-N m day at all sites (data not shown), which is similar to the ammonification rates.

- Immobilization of NO3 in the model only entered the recalcitrant organic-N pool, because dissimilatory nitrate reduction to ammonium (DNRA) was thought to be negligible in these

- non-flooded soils. Gross rates of NO3 immobilization did not differ between reconstructed and natural forest soils or among vegetation treatments (F(3,16)=0.93, p=0.467) (Figure 2.6D).

- Total rates of nitrification (W=65; p=0.712) and of NO3 consumption (W=67; p=0.799) were similar using either the analytical or the numerical solution. Once again, the two sites with unusually high rates of nitrification of recalcitrant organic-N increased the immobilization

61 rates in coniferous and grassland soils. When we removed the contributions of these two sites,

average gross rates of NO3-N immobilization into recalcitrant organic-N rates were around

-2 -1 200 mg NO3-N m day in reconstructed and natural soils (data not shown).

2.3.4 Net ammonification and nitrification rates

Net ammonification rates did not differ between reconstructed and natural soils, or among

vegetation treatments (F(3,16)=0.72, p=0.569) (Figure 2.7A). Net ammonification rates were not

significantly different from zero in both reconstructed (p=0.570) and natural soils (p=0.500)

suggesting that production and consumption of ammonium were equivalent in all cases. Net

nitrification rates were not different between reconstructed and natural soils (F(3,16)=1.89,

p=0.569) (Figure 2.7B). However, net nitrification rates were significantly greater than zero in

reconstructed soils (p=0.002), but not in natural soils (p=0.75) meaning there was net

production of nitrate in reconstructed soils, but not in natural soils.

200 200 A B

150 150 ) ) -1 -1 day day -2 -2 100 100 -N m -N 3 -N m -N 4 (mg NO (mg Net nitrification rates nitrification Net (mg NH (mg 50 50 Net ammonification rates ammonification Net

0 0

Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

+ Ammonium turnover times varied from 0.09 (2.16h) to 2.39 days. NH4 turnover times did not

significantly differ among reconstructed deciduous, coniferous and grassland sites and natural

sites (F(3,16)=1.29, p=0.341) although natural sites had systematically shorter turnover times

than reconstructed soils (Figure 2.8a).

3 3

2 2 (days) (days) turnover time turnover turnover time turnover + - 4 3 NH NO

1 1

0 0 Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

+ - Figure 2.8 NH4 (A) and NO3 (B) turnover times in reconstructed soils planted with coniferous trees, deciduous trees, grasses and natural forest soils Nitrate turnover times were more variable among sites, ranging from 0.03 (<1h) to 15 days.

Although no significant differences among all treatments were identified (F(3,16)=1.28,

- p=0.345) (Figure 2.8b), deciduous and natural soils had consistently shorter mean NO3

- turnover times. One grassland site had an extremely long NO3 residence time compared to the

other sites.

63 2.4 Discussion

The relative rates of the various N-cycling processes in reconstructed and natural soils are summarized in Figure 2.9.

a - Reconstructed soils

Labile organic Recalcitrant organic nitrogen nitrogen INH4-Nrec 269.7 mg N m-2 day-1

INH4-Nlab NNrec 69.9 mgN m-2 day-1 238.9 mgN m-2 day-1 INO3-Nrec 138.85mgN m-2 day-1 ANlab A -2 -1 Nrec 0.03 mg N m day 345.3 mg N m-2 day-1

+ - NH4 NO3 NNH4 15.05 mg N m-2 day-1

b - Natural soils

Labile organic Recalcitrant organic nitrogen nitrogen INH4-Nrec 20.35 mg N m-2 day-1

I NH4-Nlab NNrec 35.02 mg N m-2 day-1 18.72 mg N m-2 day-1 INO3-Nrec -2 -1 ANlab 17.60 mg N m day 0.64 mg N m-2 day-1 ANrec 49.05 mg N m-2 day-1

+ - NH4 NO3 NNH4 0.07 mg N m-2 day-1

64 Gross N transformation rates measured in this study fall within previously published rates with

+ the exception of gross NH4 immobilization rates, which were lower in the current study

+ (Table 2.4). Gross ammonification and NH4 immobilization rates in natural soils studied by

McMillan et al. (2007) were higher than those measured in the current study. Mc Millan

(2007) measured gross N-transformation rates in forest floor soil horizons which have a higher

C and N content than mineral horizons and, therefore, likely to have higher N-transformation

rates.

Table 2.4 Gross N-transformation rates in the AOSR (mg N kg-1 day-1)

McMillan et al. Cheng et al. Current study1 (2007)2 (2011)3 Reconstructed soils Gross ammonification 1.6 to 3.7 1.5 to 3.5 - + Gross NH4 immobilization 1.15 to 4.34 - - Gross nitrification 0.23 to 3.204 0.7 to 0.9 - - 4 Gross NO3 immobilization 0.02 to 2.8 - - Natural soils Gross ammonification 0.2 to 0.28 2.8 to 4.6 0.25 to 0.75 + Gross NH4 immobilization 0.18 to 0.34 - 1.2 to 1.5 Gross nitrification 0.003 to 0.15 0.25 to 0.55 0.025 to 0.075 - Gross NO3 immobilization 0.001 to 0.21 0.05 to 0.2 1. Ranges are calculated using the minimum and maximum values obtained for each transformation rates calculated using the numerical solution 2. McMillan et al (2007) sampled the first 7 cm of peat mineral mixes in reconstructed soils and the first 7 cm of forest floor (horizons LFH) in the natural soils (situated 4 km away from the reconstructed soils). Ranges are estimated from the mean and standard errors presented in the paper 3. Cheng et al. (2011) only sampled in natural soils situated in the southern edge of the AOSR. They sampled the first 20 cm of mineral soil layer. Ranges are estimated from the mean and standard deviation presented in the paper 4. Ranges are calculated without the outliers values (from sites C3 and G2)

65 2.4.1 Distinct ammonification cycling processes in reconstructed and natural forest soils

Rates of most N-cycling processes were higher in reconstructed oil-sands soils than in natural boreal-forest soils. The main differences in N-cycling processes between reconstructed and

+ natural soils are 1) NH4 is predominantly cycle through the recalcitrant organic-N pool in the

+ reconstructed soils; whereas, in the natural soils, NH4 is mainly being mineralize from

+ recalcitrant organic-N pool, but then this NH4 is being primarily immobilize in the labile

- organic-N pool. 2) More NO3 is being cycle in the reconstructed soils compare to the natural

- soils and a net production of NO3 was measured in the reconstructed soils.

+ In the reconstructed soils, NH4 was mainly cycled through the recalcitrant organic-N pool,

+ with only a small amount of the mineralized NH4 being produced and consumed from the

+ labile organic N pool. In the natural soil, NH4 was also mainly mineralized from the recalcitrant organic-N pool, but it was predominantly consumed into the labile organic-N pool.

+ Also, mineralization of NH4 from the labile organic-N pool was 20-fold higher in natural soils compared to reconstructed soils. This could indicate greater prominence of microbial N cycling activity in the natural soils compared to the reconstructed soils, despite the natural soils having lower microbial biomass and lower N content. Quicker turnover rates (of both

+ - NH4 and NO3 ) in natural soils also support this hypothesis. Even if this theory was not tested

+ in this study, the high rate of immobilization of NH4 into recalcitrant organic matter in

+ reconstructed soils may also indicate that chemical fixation of NH4 into the residual peat is a prominent process in the reconstructed soils (McNevin et al., 1999; Cho et al., 2014;).

66 In reconstructed soils, gross ammonification rates from the recalcitrant organic-N pools were on average 11,510 times higher than ammonification from the labile organic-N pool.

However, in natural soils, gross ammonification rates were only 76 times higher from the recalcitrant organic-N pools compared to the labile organic-N pool. Part of the higher gross ammonification rates from the recalcitrant organic-N pool can be explained by the differences in the initial pool sizes; the labile organic-N pool was set at 1% of the SON pool in the model, the rest of the organic-N being the recalcitrant organic-N. Higher transformation rates are, therefore, expected from the recalcitrant organic-N pool. With gross ammonification rates being 76 times higher from the recalcitrant organic-N pool than from the labile organic-N pool, gross ammonification rates in natural soils fall within expected ranges. However, in reconstructed soils, the discrepancy between gross ammonification rates from the recalcitrant organic-N pool and from the labile organic-N pool is larger than anticipated. This could be due to lower microbial turnover rates in the reconstructed soils compared to the natural soils, which induce a much lower gross ammonification rates from labile organic-N pools than from

+ recalcitrant organic-N pools in the reconstructed soils. It could also be that NH4 is simply adsorbed on the negatively charged organo-mineral complexes in the reconstructed soils and it is being exchanged back and forth between the recalcitrant organic-N pool and the NH4+ pool creating an “adsorption-desorption loop” between these two N pools. Such an adsorption-

+ desorption loop could have induced (an artificially) greater gross ammonification and NH4 immobilization rates into the recalcitrant organic-N pool in reconstructed soils. This process can be further reinforced by the greater ammonium deposition on reconstructed soils compared to natural soils.

67 2.4.2 Higher net nitrification rates in reconstructed soils

- + Although there was some production of NO3 from the NH4 pool (autotrophic nitrification) in reconstructed soils, this pathway was negligible in natural soils suggesting that mineralization

- of NO3 in both reconstructed and natural soils is not primarily an autotrophic process. Mean

- values of production and consumption of NO3 were higher in reconstructed soil than in

- natural soils. More NO3 was also produced than was consumed in reconstructed soils, resulting in positive net nitrification rates. The higher soil N content, ammonification rates and microbial biomass in the reconstructed soils could compared to the natural soils explain positive net nitrification rates in the reconstructed soils. The very low autotrophic nitrification rates measured in both reconstructed and natural soils suggest that the higher net nitrification rates in the reconstructed soils can be predominantly attributed to heterotrophic nitrification.

The positive net nitrification rates in the reconstructed soils can have ecological impacts as

- - NO3 is an anion and most of the exchange sites in soils are negatively charged. NO3 is therefore less stable in soils and could be further lost by leaching or gas emission (Robertson and Groffman, 2007).

2.4.3 Higher carbon content as a cause of higher N-transformation rates in reconstructed soils

The higher net nitrification in reconstructed soils, along with the higher N content, higher gross rates of ammonification from the recalcitrant organic pool, and higher gross rates of

+ NH4 immobilization into the recalcitrant organic N pool, indicate much greater availability of

+ - N in reconstructed soils. Given that measured rates of NH4 and NO3 deposition in the vicinity of the oil-sands mining can be ten-fold higher that those 120 km from the mining sites

68 (Fenn et al., 2015), one could hypothesise that the higher N transformation rates in the reconstructed soils are attributable to the higher N deposition in the area. However, when we calculate our rates of N transformations on an organic-matter-content basis (mg NH4-N kg

OM-1 day-1), the rates become comparable between reconstructed and natural soils (Figure

2.10). Thus the high gross rates of N transformations (per m2) in reconstructed soils can be explained by the high organic-matter content (hence higher N content) of the reconstructed soil (resulting from the incorporation of peat during reclamation) and not due to the higher nitrogen deposition in the area. The strong-positive correlation observed between soil C and N contents support this conclusion.

69 a a A B

30 30

a a

in recalcitrant organic N organic recalcitrant in 20 a 4

20 a a

10 a Mineralization of recalcitrant organic N organic recalcitrant of Mineralization

Coniferous Deciduous Grassland Natural NH of Immobilization Coniferous Deciduous Grassland Natural

0.3 25 C D

0.2 b 15 in labile organic N organic labile in 4

10 0.1 a a a p < 0.001 5 a

Mineralization of labile organic N organic labile of Mineralization a ab ab

0.0 NH of Immobilization

Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

40 E F

a 30 30

immobilization a a a 4 a 20 a 20 a Total mineralization Total Total NH Total

10 Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

70 2.4.4 Explanations for high nitrification rates in two reconstructed soils

- Rates of nitrification of recalcitrant organic N and immobilization of NO3 into recalcitrant organic N rates in reconstructed soils were augmented by two extreme values from one coniferous (C4) and one grassland site (G3). Site C4 had the highest peat content (Figure

+ - 2.11); it also had higher pH, root content, C:N ratio and concentrations of NH4 and NO3 than the other coniferous reconstructed soils (data available in Appendix A). Nitrification is

+ enhanced under alkaline conditions and high concentrations of NH4 , which could result in the

- higher rates of NO3 transformation in the C4 soil (Robertson and Groffman, 2007).

Figure 2.11 Pictures of reconstructed-coniferous sites C3 (A) and C4 (B). Patches of peat material are pointed on the C4 soil profile

71 Site G3 had the highest coarse-root, clay and fine-sand content (Appendix A). The higher coarse-root content and the soil texture could have improved aeration in the soil and thereby increased nitrification rates (Gebauer et al., 1996; Corre et al., 2002; Li et al., 2003; Tan et al.,

2006). Evidence of a particularly high level of soil faunal activity (i.e. faeces and direct observation of worms and ants) was also observed during field sampling at this site. Activity of soil fauna, especially of earthworms, could lead to better aeration and therefore higher nitrification rates (Olsson et al., 2012).

2.4.5 Limitations of the 15N tracing model to measure gross rates of N-transformation

Like any model, both the analytical and numerical solutions to measure N-transformation rates have their limitations. Seven N-transformation rates were simultaneously optimized using the numerical solution while only four pools and two 15N enrichment level were provided for each extraction time. This ratio still respects one of the condition of the model which state that “the number of N-transformation rates which can be simultaneously optimized should not exceed the number of independent variables, otherwise the system is overdetermined and optimisation may result in erroneous results” (Müller et al., 2004). By supplying to the model 8

+ - 15 + 15 - 15 independent variables ([NH4 ], [NO3 ], [SON], NH4 and NO3 for both N labelling incubation experiments at each extraction time, it was possible simulate a maximum of 8 rates.

During optimisation, several model versions (differing in kinetics, number of N pools and N- transformation rates) were tested and only the simplest one, which still adequately represented the observations, was retained (as evaluated by Akaike Information Criteria AIC). The finding

+ - that estimated rates of total ammonification, NH4 consumption, nitrification and NO3 consumption rates were similar using both the simpler analytical and the numerical solutions

72 further increased the level of confidence in the N-transformation rates generated by the

- numerical solution. The model also assumes that NO3 can only be immobilized into the

+ recalcitrant organic-N pool while NH4 could be immobilized in both labile and recalcitrant organic-N pools. During development of the model (Müller et al., 2004) it was assumed that

+ - bacteria preferentially use NH4 as their N source and that fungi may prefer NO3 as a N

+ - source. However, several studies have shown that both NH4 and NO3 as well as organic forms of nitrogen may be consumed by both bacteria and fungi depending on a panoply of factors such as acidity and soil nutrient regime (Geisseler et al., 2010; Schimel and Bennett,

2004; Tahovská et al., 2013). This knowledge could be implemented to future 15N-tracing models using this optimization technique.

2.4.6 Revegetation treatment had no impact on gross N transformation rates in reconstructed soils

Contrary to our hypotheses, N-cycling processes in reconstructed soils did not significantly differ among different vegetation treatments. Afforestation studies have shown that vegetation effects on various mineral soil properties (pH, humic acid characteristics, carbon and nitrogen contents) become measurable 20 to 50 years after afforestation, when litter inputs rates have been restored and organic matter is being redistributed through the soil profile (Compton et al., 2007; Abakumov et al., 2010; Bárcena et al., 2014). Schrijver et al. (2012) did not detect a significant tree-species effect on mineral soil properties (at a depth 10-20 cm in the mineral soil) until 35 years after planting. The reconstructed oil-sands soils that we studied were planted less than 30 years ago, but some differences in C content in the upper 10 cm of mineral soil have been observed between these vegetation treatments (Anderson, 2014). These

73 differences may eventually lead to variation in rates of soil N-cycling processes among vegetation types, but at this point in time it appears that the effect of the residual peat is overwhelming any influence of vegetation on N-cycling rates in reconstructed soils.

2.5 Conclusions

Relative to natural soils, the reconstructed soils had higher rates of most N transformations, reflecting their greater total N content due to the incorporation of peat during reclamation.

+ Mineralized NH4 was mainly re-immobilized in the recalcitrant organic-N pool in

+ reconstructed soil; NH4 transformations through the labile organic-N pool were less pronounced in reconstructed soils than natural soils, indicating reduced prominence of microbial N-cycling in the reconstructed soils. Gross nitrification rates were similar in reconstructed and natural soils, but net nitrification rates were higher in reconstructed soils, and appeared to be heterotrophic in nature. The higher net nitrification rates in reconstructed soils indicate a surplus of N relative to microbial requirements, which may be attributable to

+ - higher rates of NH4 and NO3 deposition that have been measured in the vicinity of the oil- sands mines.

74 Chapter 3: Vegetation cover and nitrogen deposition as drivers of α- and β- prokaryotic diversity in reconstructed oil-sand soils and in natural boreal- forest soils3

3.1 Introduction

The Athabasca Oil Sands deposit, located in the boreal forests of northern Alberta, is part of the largest single oil deposit in the world, with proven reserves of 166 billion barrels of bitumen, and covering 142,200 km2 (Government of Alberta, 2016b). Most (80%) of the bituminous sands can be extracted using in situ recovery methods, but 20% of the resource is shallow and can be recovered through open-pit mining (Government of Alberta, 2012). To date, about 895 km2 of land has been disturbed by oil-sands mining activity (Government of

Alberta, 2016b). Following surface mining, companies are required to restore soils that can achieve equivalent land capability (Government of Alberta, 1993; Powter et al., 2012). After soil reconstruction, the area is re-vegetated. When reclamation in the area began in the 1980’s, revegetation predominantly focused on erosion control, and used both native and introduced grasses and shrubs. However, more recent revegetation practices use native tree species such as jack pine, white and black spruce and aspen and understory shrubs such as blueberry and willow to re-establish a boreal-forest plant community. It is unlikely that these reconstructed forest soils will exactly mirror pre-existing boreal-forest soils (Chazdon, 2008; Hobbs et al.,

2006), so novel soil ecosystems will probably arise from the reconstruction efforts. Re-

3 This chapter will be submitted: Masse J., Prescott, C.E., Renaut S., Terrat, Y., Grayston S.J. Vegetation cover and nitrogen deposition as drivers of α- and β- prokaryotic diversity in reconstructed oil-sand soils and in natural boreal forest soils. 75 establishing soil functions, chiefly nutrient cycling, rather than trying to replicate the structural qualities of the previous soil ecosystem is necessary to ensure the long-term sustainability of reclaimed boreal forest landscapes (Quideau et al., 2013).

From decomposition to nutrient cycling and greenhouse gases regulation, soil microbial communities are responsible for a plethora of functions in soils. While some broad, or universal, biochemical processes are widely distributed across living organisms (e.g. the metabolism of glucose), others are carried out by a phylogenetically limited number of organisms (e.g. N-fixation or nitrification). Consequently, for narrow processes, high microbial diversity can increase the probability that these functions will be retained in the event that a species is affected by disturbances (Schimel et al., 2005).

The main phyla of bacteria found in boreal forest soils are Proteobacteria, followed by

Bacteriodetes, Acidobacteria, Actinobacteria, Firmicutes, Planctomycetes and

Verrucomicrobia (Roesch et al., 2007; Sun et al., 2014). Environmental factors such as pH, moisture, base cation abundance, as well as the quality of available carbon, influence soil microbial communities (Fierer & Jackson, 2006; Hansel et al., 2008; Brockett et al., 2012;

Ferrenberg et al., 2013). For example, soils with near-neutral pH tend to have the highest microbial diversity and richness (Fierer and Jackson, 2006). Tree species affect the structure and abundance of soil microbial communities through influences on the chemical nature of litter and exudates, as well as mycorrhizal fungal associates (Wardle, 2002; Grayston &

Prescott, 2005; Prescott & Grayston, 2013). For example, in a common-garden experiment with contrasting levels of tree diversity Thoms et al., (2010) found different microbial

76 communities at different diversity levels. However, the differentiation was related to the presence of specific tree species (Tilia and Acer), both of which produce base-rich litter; this appeared to exert the greatest influence on microbial communities.

The main disturbance in boreal forest ecosystems is wildfire, which can cause shifts in soil microbial communities (Ball et al., 2010; Hart et al., 2005; Switzer et al., 2012). Changes in soil microbial communities have been be linked to both direct and indirect effects of fire.

Lysis of microbial cells and death of plants roots during fire causes changes in microbial communities; microorganisms especially sensitive to high temperature (i.e. Nitrobacter spp. and fungi) are more prone to lysis during fire (Hart et al., 2005). Post-fire changes in soil characteristics (i.e. increased pH and temperature, reduced moisture) modify the structure of soil microbial communities. The increased soil pH 10 days after a prescribed fire in a

Douglas-fir forest in British Columbia (Canada) was thought to be the main factor responsible for the reduced abundance of Gram-positive bacteria, Gram-negative bacteria, Actinomycetes and arbuscular mycorrhizae, as well as the total bacterial biomass in soils (Switzer et al.,

2012). Ball et al. (2010) reported that forest fires increased charcoal content, gross nitrification rates and abundance of ammonia-oxidizing bacteria. Moreover, changes in the structure of bacterial community were still measurable more than 14 years after fire.

Previous studies contrasting reconstructed oil-sands soils ranging in age from 5–35 years and natural boreal-forest soil, using denaturing gradient gel electrophoresis (DGGE) and phospholipid fatty acid (PLFA) profiling, showed reduced abundance and differences in microbial community structure (Dimitriu et al., 2010; Hahn and Quideau, 2013). While DGGE

77 and PLFA do not allow for in-depth (species-level) characterization of microbial communities, it was possible to detect a general increase in Gram-negative bacteria in reconstructed soils compared to natural soils (Dimitriu et al., 2010). Dimitriu and Grayston (2010) showed that abiotic variables, such as pH and soil moisture, rather than plant cover, were the main drivers of community structure in both reconstructed and natural soils. Most previous studies in the oil sands compared microbial communities in the top 15 cm of soils regardless of the presence or absence of a forest floor. It could be predicted that samples containing forest floor material, with high carbon and nutrient content would have de facto greater microbial biomass.

However, Dimitriu and Grayston (2010) found no differences in microbial community richness and structural diversity between samples containing exclusively mineral soil or exclusively forest-floor material.

Since microorganisms have a pivotal role in soil functioning, the objective of this study was to assess if soil prokaryotic diversity and structure (i.e. α- and β- diversity) in oil-sands soils reconstructed 20–30 years previously were similar to those found in natural boreal-forest soils subjected to wildfire disturbance at approximately the same time. Specifically, we evaluated prokaryotic diversity and community structure in the top 12 cm of mineral soils of 20–30- year-old reconstructed and fire-disturbed boreal forest soils using massively parallel sequencing of 16S rRNA genes. We also assessed the influence of three vegetation treatments

(coniferous trees, deciduous trees and grasses) used during reclamation as well as the influence of plant cover, soil chemical and physical characteristics on prokaryotic α- and β- diversity. We hypothesised that 1) α-diversity will be higher in natural soils; 2) β-diversity

78 will differ between reconstructed and natural soils; and 3) pH and soil moisture will be the main drivers of prokaryotic α- and β- diversity.

3.2 Materials and methods

3.2.1 Study area

The study area was situated in the Athabasca Oil Sands Region (AOSR) in northern Alberta,

Canada (56°39’N, 111°13’W, altitude: 369 m). Short warm summers and long cold winters characterize the climate. The mean annual temperature is 1°C; mean monthly temperatures range from -17.4°C in January to 17.1°C in July. Mean annual precipitation is 418.6 mm, of which 316.3 mm occurs as rainfall during the growing season (Environment Canada, 2015).

Medium- to fine-textured Gray Luvisols and Dystric Brunisols underlie landscapes shaped by the impact of Pleistocene ice activity, deglaciation and post-glacial modifications in upland areas. Organic soils are found under wetland areas (Natural Regions Committee, 2006). The

AOSR falls within the central mixedwood region of the Canadian boreal forest. Dominant tree canopy species in upland landscapes are trembling aspen (Populus temuloides Michx), white spruce (Picea glauca (Moench) Voss) and jack pine (Pinus banksiana Lamb) (Natural

Regions Committee, 2006). Fire is the major natural disturbance in these forests (Thomson,

1979).

79 for berm, dyke-wall or road construction, or deposited in a dedicated disposal area to create large-scale overburden landform units. The oil-sand ore is transported to the extraction and upgrading facility. Oil-sands soil reconstruction involves a number of cover designs, depending on the landform substrate being reclaimed. There are two main cover designs: one that uses only cover-soil and the other consisting of a combination of cover-soil and subsoil.

Cover-soil and subsoil materials are salvaged from surface soils within the mine-development footprint. Only sites at which cover-soil had been placed on top of overburden material were used in this study. The cover-soil materials used consisted of surface peat mixed with mineral soils material having a loam or coarser texture, and is hereafter referred to as ‘peat-mineral mix’. In the studied soils, the depth of the peat-mineral mix ranged from 20 cm to more than

100 cm. Early revegetation objectives in the AOSR were to establish native or introduced grass and shrub species to control erosion; however, oil-sands operators are now required to use native trees and species with the intention of promoting the re-establishment of a boreal forest community. During the period when the sites used in this study were reclaimed (20–30 years ago), 250 to 350 kg ha-1 of varying proportions of N:P:K fertilizer was typically applied in the first year of re-vegetation. Some oil-sands operators also applied fertilizer annually for four additional years.

3.2.2 Study sites

Fifteen reconstructed sites and five natural forest sites were studied in the AOSR; all sites were reconstructed (or naturally fire-disturbed) 20 to 30 years previously. The 15 reconstructed sites were previously studied by Sorenson et al. (2011). Five sites were planted with deciduous species (mostly trembling aspen), five with coniferous species (mostly white

80 spruce), and five with grasses (fescue, slender wheatgrass and alfalfa). The natural sites have similar time-since-disturbance (fire) and similar soil texture as the reconstructed soils. Natural forest soils were classified as Gleyed Eluviated Eutric Brunisol (soil N1), Eluviated Eutric

Brunisol (N5) or Brunisolic Gray Luvisol (soils N2, N3, N4). The natural sites are located approximately 40 to 150 km south of Fort McMurray (Alberta, Canada).

At each site, one 10-m2 plot was sub-divided into ten 1-m2 subplots, from which 7 subplots were randomly selected. A 30-cm-deep soil pit was dug and carefully described in each sub- plot. The litter and forest floor layers were removed and the top 0–15 cm of the mineral part of the soil was sampled (~ 1 kg) at each sub-plot using a plastic trowel and the 7 subplot samples were pooled to produce one sample per plot (nsample = 20). A sub-sample (~5 g) of each sample was immediately put on ice for microbial analyses. The remaining samples were put in a cooler and kept cold until arrival at the laboratory for soil chemical and physical analyses.

Removal of any surface organic material prior to sampling enabled us to directly compare the mineral substrate on which the organic layers develop, and so determine how similar the reconstructed soils are to natural soils. If we assume that the similar soil development factors – the action of climate, organisms, relief through time – are at play in reconstructed and natural boreal forest soils with few natural variation such as parent material, types of organisms, etc., the peat-mineral mix is the analogue of the mineral part of the natural soils on top of which a forest floor develops. Also, characteristics of the forest floor layers in reconstructed and natural soils in the AOSR have been described in previous research (Rowland et al., 2009).

81 Vegetation is described by Anderson (2014). (Appendix B). Tree density was not significantly different among reconstructed soils planted with deciduous species, coniferous species and naturally disturbed sites. Grassland sites were devoid of trees.

3.2.3 Laboratory analyses

3.2.3.1 Soil characteristics

Soil moisture and bulk density at each site were determined using soil cores from one of the subplots (Blake and Hartge, 1986). Soil pH in water was determined in a 1:10 (soil:water) solution using a UB-10 pH meter (Denver instruments, Bohemia, USA) (Thomas, 1996). Soil total carbon and nitrogen concentrations were measured with an Elementar Vario El Cube

Elemental Analyzer (Elementar, Hanau, Germany) using 15 mg of soil sieved (< 0.5 mm) to assure homogenization (Rutherford et al., 2008; Skjemstad and Baldock, 2008). C:N ratio was calculated by dividing total C content by total N content. Soil particle size of sand (60 µm to 2 mm) and clay (< 2 µm) were determined in the Environmental Geoscience Research Centre at

Trent University using laser diffraction with a Partica LA-950 (Horiba Ltd., Kyoto, Japan).

Analyses were completed in triplicate after homogenizing the sample (Konert and

Vandenberghe, 1997). Microbial biomass carbon was measured on the day of the sampling using the chloroform fumigation extraction (CFE) technique (modified from Tate et al., 1988).

Unless otherwise mentioned, 25% of the samples were analyzed in duplicate and controls

(AgroMAT AG-2 soil standard, SCP Science, Baie d’Urfé, Canada) were added.

Nitrogen deposition, as a result of oil sands mining activity, at each site was estimated using data from Davis et al. (2015).

3.2.3.2 DNA extraction and amplification

Soil DNA was extracted in triplicate using PowerSoil ® DNA isolation kit (MO BIO,

Carlsbad, CA, USA) with 0.25g of soil. Triplicates were pooled together and sent to McGill

University and Génome Québec Innovation Centre (Montréal, Canada) for sequencing. The

V4 region of the 16S rRNA gene was amplified with the 515F (5’-

GTGCCAGCMGCCGCGGTAA-3’) and 806R (5’-GGACTACVSGGGTATCTAAT-3’) primers with expected length of 250 base pairs (bp) using the following program: 94°C for 3 minutes, followed by 35 cycles of 95°C for 45 seconds, 50°C for 60 seconds and 72°C for 90 seconds and a final 10 minutes at 72°C. This primer pair is used for phylogenetic studies to target specifically bacterial and archaeal species (Caporaso, 2011; Caporaso et al., 2012). PCR products (n=20) were barcoded, pooled and sequenced using Illumina MiSeq sequencer with

250 base pairs on the forward and the reverse reads (Illumina, San Diego, CA, USA).

3.2.4 Bioinformatics analyses

A total of 736,710 paired-end Illumina sequences were obtained from Génome Québec from the 20 samples (36,836±8,981 sequences per sample). They were analyzed using Mothur

MiSeq SOP pipeline (Kozich et al., 2013). Briefly, forward and reverse sequences were merged. Non-ambiguous sequences shorter than 500 bp were kept and aligned using SILVA, a manually curated ribosomal RNA sequence database (Quast et al., 2013). Sequences were further screened for chimeras using chimera.uchime script (Edgar et al., 2011). Singletons, chloroplasts and mitochondria sequences were removed from the dataset using Mothur’s remove.lineage function. At the end of the screening process, 115,629 sequences were kept

83 and clustered in 778 operational taxonomic units (OTU) with a 97% similarity using the dist.seqs and cluster (using the average neighbor algorithm) functions in Mothur. We assigned the taxonomy of each OTU using the SILVA database (Quast et al., 2013). In order to make comparisons among samples, we subsampled sequences from each sample in our dataset to the lowest number of sequences found in one sample and so restricted the number of sequences in each sample to 769. Because they generated fewer unclassified sequences, we analyzed the data using the taxonomic level of phyla and classes.

3.2.5 Calculations and statistics

3.2.5.1 Good’s coverage

Good’s coverage (C) is an indicator of the depth of sequencing. It estimates the proportion of total species that are represented in a sample using equation 1.

' C = 1 − ( (eq. 1) )

Where, n1 is the number of OTU’s that have been sampled once and N is the total number of individuals in the sample (Kozich et al., 2013).

3.2.5.2 α-diversity

The microbial diversity at each site (α-diversity) was calculated using the inverse Simpson diversity index (Simpson, 1949). The Simpson index measures the probability that two sequences taken at random from the dataset represent the same OTU. The lower the index is, the lower the probability of getting two identical OTU from the dataset, and the higher the α-

84 diversity. Taking the inverse of the Simpson index makes it more intuitive index as the α- diversity is higher when the index is higher. Mean and standard errors of the inverse Simpson index for each treatment are presented. Differences among treatment types (reconstructed using deciduous, coniferous and grass species and natural forest soils) were assessed using one-way ANOVA with permutations testing (Legendre, 2007). A multiple linear regression model using forward selection was applied to find variables that significantly contribute to α- diversity in our samples.

3.2.5.3 β-diversity

β-diversity is the differentiation of microbial composition among habitats. It was assessed using different methods. A principal coordinate analysis (PCoA) using a Bray-Curtis dissimilarity matrix on Hellinger-transformed community data was done to visually assess if research sites harboured different microbial communities (Borcard et al., 2011). Hellinger- transformation of the community data provides an appropriate matrix of species abundances that eliminates the bias that species-abundance matrices generate by emphasizing the weight of rare species in ordinations (Legendre and Gallagher, 2001). PCoA was followed by an analysis of molecular variance (AMOVA) – a nonparametric analysis of variance – to test if the centers of the clouds representing treatments are more separated than the variation among samples of the same treatment (Kozich et al., 2013).

We also evaluated β-diversity using one-way ANOVA with permutations testing on all phyla and classes (Legendre, 2007) and by representing the data using the Z-scores (eq. 2) .

85 +,+ Z = (eq .2) -

Where x is abundance of the phylum or class in the sample, x is the mean abundance of the phylum or class using all samples; s is the standard deviation of the abundance of the phylum or class using all samples. Represented are the Z-scores means of each phylum or class for each treatment.

When significant differences were observed, post-hoc Kruskall-Wallis tests were performed using the R package pgirmess (Giraudoux, 2014).

3.2.5.4 Relationships between environmental variables and communities

A canonical redundancy analysis (RDA) using Hellinger-transformed microbial data was performed to identify environmental variables – i.e. soil physical, chemical and biological characteristics and vegetation cover – that significantly explain the distribution of microbial communities in the studied soils (Legendre and Legendre, 2012). A complete list of environmental variables tested in the RDAs (and their values) is available in Appendix A and

B. Significant variables were selected using a forward selection algorithm – packfor’s forward.sel function (Dray et al., 2011). The RDA model, axis and explanatory variables were tested using permutations testing (Borcard et al., 2011). RDAs are presented using scaling 3.

A partition of the variation was also performed to assess how much of the variation was explained by the soil and the vegetation characteristics (Borcard et al., 2011).

Statistical analyses used the software R (R Core Team, 2014).

86 3.3 Results

3.3.1 Soil chemical, physical and biological characteristics

Microbial biomass carbon was significantly greater (F(3,16) = 4.613; p=0.0165) in reconstructed soils planted with deciduous trees than in natural forest soils (Table 3.1). Soil nitrogen concentration (F(3,16) = 7.183; p=0.0029), carbon concentration (F(3,16) = 8.649; p=0.0012),

-6 C:N ratio (F(3,16) = 25.457; p=2.48x10 ) as well as nitrogen deposition rates (F(3,16) = 52.184;

-8 p=1.75x10 ), were all substantially higher in reconstructed soils. Soil pH (F(3,16) = 8.085; p=0.0017) and soil moisture (F(3,16) = 2.608; p=0.0874) were higher in reconstructed soils under grasses than in natural soils. Clay content of soil was similar at all sites (F(3,16) = 1.323; p=0.3016)) (Table 3.1).

3.3.2 Soil microbial community structure

Reconstructed soils planted with coniferous species had fewer prokaryotic sequences than natural forest soils (F(3,16) = 6.518; p=0.0043). Once sequences were cleaned and aligned, numbers of sequences were similar among all soils (F(3,16) = 2.017; p=0.1521), although reconstructed soils were still depauperate relative to natural soils. The number of OTUs identified per site (F(3,16) = 0.508; p=0.6824) were similar among reconstructed and natural forest soils. Good’s coverage was similar among treatments (F(3,16) = 0.337; p=0.7987) and ranged from 97.8 to 100% (Table 3.2).

87 Table 3.1 Selected physical, chemical and biological characteristics of the studied soil

Microbial Soil Total Clay Total carbon biomass Nitrogen Site pH moisture nitrogen C:N ratio content (%) carbon deposition (%) (%) (%) (kg N ha-1 y-1) (g C kgsoil-1) Coniferous (mean±sd) 6.5±0.2ab 20.2±12.1ab 0.35±0.22ab 7.6±4.7ab 21.9±2.5a 1.31±0.37ab 2.61±0.52a 12.5±2.5a C1 6.4 11.6 0.19 3.8 20.6 1.15 3.33 12.5 C2 6.4 18.5 0.42 7.8 18.7 1.60 2.33 15.0 C3 6.6 39.9 0.71 15.5 21.9 1.27 2.26 10.0 C4 6.9 21.3 0.19 4.6 24.8 1.73 2.97 10.0 C5 6.5 9.6 0.27 6.4 23.7 0.79 2.15 15.0 Deciduous (mean±sd) 6.4±0.3ab 22.9±6.3ab 0.41±0.14a 7.6±2.2a 19.0±2.9ab 1.53±0.84a 1.87±0.90a 14.0±1.4a D1 6.5 22.3 0.46 7.2 15.5 1.53 1.95 15.0 D2 5.8 27.1 0.62 11.4 18.6 2.84 1.04 12.5 D3 6.5 16.1 0.23 5.4 23.4 0.58 2.40 12.5 D4 6.6 31.3 0.38 6.9 18.0 1.64 0.94 15.0 D5 6.5 17.8 0.36 7.2 19.7 1.08 3.06 15.0 Grassland (mean±sd) 7.0±0.3a 25.3±8.1a 0.39±0.13a 7.5±1.8a 19.8±2.1a 0.93±0.44ab 1.59±0.50a 12.5±0.0ab G1 6.5 25.2 0.37 7.2 19.4 0.98 1.27 12.5 G2 7.2 35.9 0.61 10.4 17.0 1.35 2.08 12.5 G3 7.1 26.9 0.37 8.0 21.5 1.03 2.18 12.5 G4 7.1 13.3 0.26 5.8 22.1 0.17 1.27 12.5 G5 7.0 25.2 0.33 6.3 18.8 1.10 1.14 12.5 Natural (mean±sd) 6.1±0.3b 12.1±1.4b 0.04±0.008b 0.39±0.13b 10.1±1.8b 0.53±0.29b 3.43±2.69a 3.0±0.5b N1 6.6 10.8 0.028 0.31 11.1 0.93 1.20 3.3 N2 6.1 12.8 0.037 0.36 9.7 0.24 5.62 3.5 N3 5.8 10.5 0.047 0.53 11.2 0.35 3.35 2.5 N4 5.9 13.5 0.044 0.49 11.2 0.41 6.58 2.5 N5 6.1 13.1 0.030 0.21 7.1 0.73 0.38 3.3

88 Table 3.2 Number of sequences, OTU and coverage in the studied soils

Number of Number of raw sequences after Number of Good's Site sequences cleaning and OTU coverage alignment Coniferous (mean±sd) 27,855±9,930a 2,603±1429a 167±68a 99.3±0.9a C1 28,066 1,555 73 100.0 C2 31,026 3,842 222 99.8 C3 10,802 769 120 97.8 C4 34,838 4,069 229 99.8 C5 34,547 2,782 192 99.4 Deciduous (mean±sd) 34,646±2,068ab 3,220±965a 185±39a 99.7±0.1a D1 34,711 3,734 217 99.7 D2 31,425 3,471 177 99.5 D3 35,663 4,038 205 99.9 D4 37,021 1,568 119 99.7 D5 34,410 3,289 205 99.8 Grassland (mean±sd) 39,197±5,817ab 3,129±392a 203±23a 99.4±0.6a G1 30,851 2,992 173 99.9 G2 35,865 2,913 202 99.4 G3 41,313 3,659 195 99.6 G4 42,490 3,396 208 99.9 G5 45,464 2,683 238 98.5 Natural (mean±sd) 45,644±5976b 4,701±2215a 170±63a 99.5±0.5a N1 39,957 2,227 181 98.8 N2 40,293 6,734 230 99.9 N3 44,740 7,302 230 99.9 N4 53,823 3,251 111 99.4 N5 49,405 3,990 97 99.7

89 Rarefaction curves for each sample reached an asymptote at approximately 1,500 sequences

(Figure 3.1). As such, both Good’s coverage and the rarefaction curves suggest that the depth

of sequencing was adequate.

250

200

C1# C2# C3# 150 C4# C5# D1# D2# D3# D4#

OTUs Number of 100 D5# G1# G2# G3# G4# G5# 50 N1# N2# N3# N4# N5#

0 0 1000 2000 3000 4000 5000 6000 7000 Number of sequences

In reconstructed soils the most abundant phyla were Proteobacteria > Actinobacteria =

Acidobacteria > Chloroflexi > Bacteriorides > Planctomycetes. In the natural soils, the most

abundant phyla were Proteobacteria > Acidobacteria > Planctomycetes > Actinobacteria >

Chloroflexi > Verrumicrobia (Figure 3.2).

90 A B

1.00 1.00

0.75 0.75

0.50 0.50 Relative abundance Relative Relative abundance abundance Relative

0.25 0.25

0.00 0.00 Coniferous Deciduous Grassland Naturally disturbed Coniferous Deciduous Grassland Naturally disturbed

Bacterial classes: Bacteria phyla: Acidimicrobiia Caldilineae Gammaproteobacteria Phycisphaerae Acidobacteria Chloroflexi Acidobacteria Chloroflexia Gemmatimonadetes Planctomycetacia Actinobacteria Cyanobacteria Proteobacteria Actinobacteria Chthonomonadetes Holophagae Spartobacteria Armatimonadetes Elusmicrobia TM6 Alphaproteobacteria Cytophagia Spingobacteriia BH 180 139 Firmicutes Verrucomicrobia Ktedonobacteria Deltaproteobacteria Thermoleophilia Bacterioridetes Gemmatimonadetes WD272 Anaerolineae Melainabacteria Candidate division OD1 Nitrospirae Unclassified Bacilli Elusmicrobia Nitrospira Thermomicrobia Candidate division WS3 Planctinomycetes Betaproteobacteria Flavobacteriia Opitutae Verrucomicrobiae

Figure 3.2 Proportion of bacterial phyla (A) and classes (B) in reconstructed soils planted using coniferous, deciduous and grasses species and in natural forest soils

91 3.3.3 α-diversity

α-diversity indices in reconstructed and natural soils were not significantly different (F(3,16) =

2.016; p=0.1523), although reconstructed soils had higher α-diversity (Figure 3.3).

30 Inverse Simpson Inverse Index

Coniferous Deciduous Grassland Natural

Figure 3.3 Inverse Simpson index in reconstructed soils planted using coniferous, deciduous and grasses and in natural forest soils

C:N ratio, abundance of Pinus banksiana, Medicago sativa, Salix bebbiana and Vaccinium myrtriolloides significantly correlated with α-diversity. All variables were positively correlated with α-diversity except for Pinus banksiana, which was negatively correlated with

α-diversity (Table 3.3).

Table 3.3 Coefficients and signifiance of the multiple linear regression explaining α-diversity in the studied soils (*** : p<0.01; ** p<0.05; * : p<0.1) Regression model : α-diversity = β1 C:N ratio – β2 Pinus banksiana + β3 Medicago sativa + β4 Salix bebbina + β5 Vaccinium myrtilloides + β0+ εi

Variable Coefficients (βi) t value p-value C:N ratio 1.25 1.80 0.0938 * Pinus banksiana -0.48 -3.17 0.0067 *** Medicago sativa 0.72 3.11 0.0076 *** Salix bebbiana 4.15 2.49 0.0266 ** Vaccinium myrtilloides 2.08 2.57 0.0224 ** Intercept 11.47 0.86 0.4064 F-statistic : 9.83 on 5 and 14 degrees of freedom; Adj. R2 :0.699; p-value : 0.0003393

3.3.4 β-diversity

Microbial community structure differed between reconstructed and natural soils (F(3,16) =

7.358; p<0.001) (Figure 3.2, Figure 3.4 and Figure 3.5). Within reconstructed soils, those planted with grasses harboured different communities than those planted with either deciduous

(F(8,9) = 4.763; p=0.005) or coniferous trees (F(8,9) = 1.915; p=0.06) (Figure 3.4 and Figure 3.5)

G3 G1

G5 Candidate division WS3 BHI80 139 D1 G4 N4 TM6 C2 C5 Chloroflexi . Bacteroidetes** Thaumarchaeota . Gemmatimonadetes Archaea*** G2 Euryarchaeota unclassified . N1 D5 C4 Actinobacteria* Bacteria* Planctomycetes . Proteobacteria*** Acidobacteria*** Nitrospirae Verrucomicrobia .

Dimension 2 Dimension Cyanobacteria* WD272 *** N5 N2 0.02 0.00 0.02 0.04 0.06 Firmicutes . Elusimicrobia . − N3 Armatimonadetes D4 C1 D3 Candidate division OD1 0.04 D2 − C3 0.06

−

−0.05 0.00 0.05 0.10 Dimension 1

C1 D4 0.10 N4

Melainabacteria . G2 D3 Flavobacteriia C2 Chthonomonadetes Cytophagia* N2 N3 Deltaproteobacteria . Acidimicrobiia D2 Thermoleophilia* Elusimicrobia . Planctomycetacia Actinobacteria Anaerolineae* JG37.AG.4*** Acidobacteria*** Alphaproteobacteria*** Verrucomicrobiae* Spartobacteria** Gammaproteobacteria* Thermomicrobia Ktedonobacteria*** unclassified** Nitrospira Caldilineae** Phycisphaerae* Sphingobacteriia* D1 C4 Holophagae Betaproteobacteria C3Chloroflexia** Gemmatimonadetes Dimension 2 Dimension Opitutae Bacilli . G5 Verrucomicrobia Incertae Sedis G4 G3 G1 C5 0.05 0.00 0.05

− D5 N1 0.10

− N5

−0.3 −0.2 −0.1 0.0 0.1 Dimension 1

94 Members of the phyla Proteobacteria, Bacteroidetes and Actinobacteria were more abundant in reconstructed soils than in natural soils, while Acidobacteria, Cyanobacteria, Elusmicrobia,

Firmicutes, Plantomycetes, Verrucomicrobia and WD272 were less abundant in reconstructed soils. Finally, phylum Chloroflexi was more abundant in reconstructed soils planted with grasses compared to reconstructed soils planted to trees and natural soils (Figure 3.6; F and p values are presented in Appendix D.1).

Within phyla that were relatively more abundant in reconstructed soils, the bacterial classes

Thermoleophilia, Cytophagia, Sphingobacteria, α-proteobacteria and γ-proteobacteria were still more abundant in reconstructed soils, but the class δ-proteobacteria was more abundant in natural soils (Figure 3.7; Table 3.4; F and p values are presented in Appendix D.2).

Within phyla that were more abundant in natural soils, bacterial classes Acidobacteria,

Melainabacteria, Elusmicrobia, Bacilli, Phycisphaerae and Spartobacteria were more abundant in natural soils but the class Verrumicrobiae was more abundant in reconstructed soils (Figure

3.8; Table 3.4; F and p values are presented in Appendix D.2). Classes Melainabacteria and

Elusimicrobia were absent from reconstructed soils and class Bacilli was absent from grassland soils (Figure 3.8; Table 3.4).

Within the phylum Chloroflexi, which was more abundant in grassland soils, the classes

Anaerolinea, Caldilineae and Chlroflexia were more abundant in grassland soils, but the class

Ktedonobactria was more abundant in natural soils (Figure 3.9; Table 3.4; F and p values are presented in Appendix D.2).

95 Coniferous Deciduous Grassland Natural

TM6

Nitrospirae

Gemmatimonadetes

Candidate_division_WS3

Candidate_division_OD1

BHI80_139

Armatimonadetes

. Chloroflexi

*** Proteobacteria

** Bacteroidetes

* Actinobacteria

. unclassified

*** WD272

. Verrucomicrobia

. Planctomycetes

. Firmicutes

. Elusimicrobia

* Cyanobacteria

*** Acidobacteria

−2 −1012 −2 −1012 −2 −1012 −2 −1012 Z scores

96 Coniferous Deciduous Grassland Natural

Betaproteobacteria

. Deltaproteobacteria

* Gammaproteobacteria Proteobacteria

*** Alphaproteobacteria

−2 −10 12 −2 −10 12 −2 −10 12 −2 −10 1 2

Coniferous Deciduous Grassland Natural

Flavobacteriia

* Sphingobacteriia Bacteroidetes * Cytophagia

−2 −10 12 −2 −10 12 −2 −10 12 −2 −10 1 2

Coniferous Deciduous Grassland Natural

TakashiAC.B11

MB.A2.108

Actinobacteria

Acidimicrobiia Actinobacteria

* Thermoleophilia

−2 −10 12 −2 −10 12 −2 −10 12 −2 −10 1 2 Z scores

Figure 3.7 Z-scores of bacterial classes belonging to phyla more abundant in reconstructed soils (p<0.001: ***; p<0.01: **; p<0.05: *; p<0.1:.)

97 Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

S.BQ2.57_soil_group Planctomycetacia

Opitutae OM190 OPB35_soil_group BD7.11 * Verrucomicrobiae Planctomycetes Verrucomicrobia Phycisphaerae *** Spartobacteria *

−2 −1012−2 −10 12−2 −10 12−2 −10 1 2 −2 −10 12−2 −10 12−2 −10 12−2 −10 1 2

Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

. Bacilli . Elusimicrobia Firmicutes Elusimicrobia

−2 −1012−2 −10 12−2 −10 12−2 −10 1 2 −2 −1012−2 −10 12−2 −10 12−2 −10 1 2

Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

Holophagae

. Melainabacteria

***

Acidobacteria Acidobacteria Cyanobacteria

−2 −1012−2 −10 12−2 −10 12−2 −10 1 2 −2 −1012−2 −10 12−2 −10 12−2 −10 1 2 Z scores Z scores

Figure 3.8 Z-scores of bacterial classes belonging to phyla more abudnant in natural forest soils (p<0.001: ***; p<0.01: **; p<0.05: *; p<0.1:.)

98 Coniferous Deciduous Grassland Natural

Thermomicrobia

TK10

S085

Gitt.GS.136

*** Ktedonobacteria

*** JG37.AG.4 Chloroflexi

* JG30.KF.CM66

. KD4.96

** Chloroflexia

** Caldilineae

* Anaerolineae

−2 −1012 −2 −1012 −2 −10 12 −2 −1012 Z scores

Figure 3.9 Z-scores of bacterial classes belonging to phyla more abundant in reconstructed soils planted with grasses (p<0.001: ***; p<0.01: **; p<0.05:

*; p<0.1: .)

Table 3.4 Phyla (bold) and classes (italics) that are more abundant in either reconstructed soils, natural or in grassland soils. Enriched in reconstructed Enriched in natural soils Enriched in grassland soils soils Acidobacteria • Acidobacteria

Actinobacteria Cyanobacteriaθ • Thermoleophilia • Melainabacteriaθ

Bacterioridetes Elusmicrobiaθ • Cytophagia • Elusmicrobiaθ • Sphingobacteriia γ Chloroflexi Firmicutes ρ γ • Anaerolineae Proteobacteria • Bacilli σ • Caldilineae • Alphaproteobacteria • Chloroflexia • Gammaproteobacteria Planctomycetes • KD4 96 • Phycisphaerae Verrucomicrobiae Verrumicrobia JG30.KF.CM66 • Spartobacteri JG37.AG.4 • Verrucomicrobiaeδ Ktedonobacteria WD272θ

Deltaproteobacteria θ: Absent from reconstructed soils; γ: Absent from grassland soils; ρ: Absent from natural soils and reconstructed soils planted with coniferous species; σ: Absent from natural soils and reconstructed soils planted with deciduous species

100 There were significantly less archaea in reconstructed soils planted with trees compared to reconstructed soils planted with grass and natural forest soils (F(3,16) = 13.810; p=0.00001) (Figure 3.10)

1.00 A A

0.99 B

0.98

Bacteria Archaea

0.97 Proportion of bacteria and archaea and bacteria of Proportion

0.96

a a a a 0.95 Coniferous Deciduous Grassland Natural

The archaeal phylum Euryarchaeota was found only at one natural site N1 (Figure 3.11).

The phylum Thaumarchaeota was found in both reconstructed soils and in natural forest soils, but Thaumarchaeota were significantly more abundant in reconstructed soil planted with grass and in natural soils (F(3,16) = 8.90; p=0.00105). Within the phylum

Thaumarchaeota, the class Soil Crenarchaeotic Group (also known as the 1.1b group) was more abundant in reconstructed soils planted with grasses than in natural soils (F(3,16) =

3.870; p=0.029). In contrast, the archaeal class Terrestrial Group (also known as the 1.1c

101 group) of the phylum Thaumarchaeota was only present in natural soils and, therefore, was significantly more abundant in these soils than in reconstructed soils (F(3,16) =

14.231; p=8x10-5) (Figure 3.11).

1.00

0.75

Archaeal phyla: Thaumarchaeota 0.50 Euryarchaeota the abundance of archaea archaea of abundance the

Proportion of archaeal phyla relative to phyla relative archaeal of Proportion 0.25

0.00 Coniferous Deciduous Grassland Natural

1.00

0.75

Archaeal classes Soil crenarchaeotic group Thermoplasmata Terrestrial group 0.50 the abundance of archaeal of abundance the

0.25 Proportion of archaeal classes relative to to relative classes archaeal of Proportion

0.00 Coniferous Deciduous Grssland Natural

Figure 3.11 Proportion of archaeal phyla (top) and classes (bottom) in reconstructed soils planted using coniferous, deciduous and grass species and in natural forest soils

102 3.3.5 Relationships between environmental variables and prokaryotic communities

Canonical redundancy analysis indicated that distribution of phyla in the studied soils can be significantly explained by the estimated nitrogen deposition, clay content and the abundance of the mosses Lycopodium obscurum and Polytrichum juniperinum and the woodland perennial, Trientalis borealis (F(5,14) = 4.495, p = 0.001, Figure 3.12A).

Vegetation was positively correlated with the microbial communities that were abundant in the natural soils, whereas nitrogen deposition and soil clay content were positively correlated with the microbial communities that were more abundant in reconstructed soils

(Figure 3.12A). The variables included in the model explained 52.8% of the variation in microbial communities at a phylum level (Figure 3.12B). Alone, the abundance of

Lycopodium obscurum, Polytrichum juniperinum and Trientalis borealis explained

18.6% of the variation (F(3,14) = 5.551; p = 0.001); nitrogen deposition and soil clay content explained 12.7% of the same variation (F(2,14) = 3.150; p = 0.001). Jointly these variables explained 21.5% of the variation (F(5,14)=5.250; p=0.001) (Figure 3.12B).

103

1.5 N3 A Trientalis borealis B

Elusimicrobia Armatimonadetes N2 Firmicutes Soil characteristics Vegetation Nitrospirae - Nitrogen deposition - Trientalis borealis C3 Planctomycetes D3 C1 Bacteria - Polytrichum juniperinum Gemmatimonadetes - Clay content N5 WD272 D2 Acidobacteria Candidate division OD1 - Lycopodium obscurum Polytrichum N1 D4 Cyanobacteria unclassified EuryarchaeotaD5 D1 C4Proteobacteria 0 Candidate division WS3 Bacteroidetes juniperinum Verrucomicrobia G4C5G2 Archaea G5Chloroflexi Actinobacteria Thaumarchaeota C2 G3 12.7% *** 21.5% 18.6% *** G1 Nitrogen deposition 0.5 0.0 0.5 1.0 − Clay content 1.0 −

BHI80 139 Lycopodium TM6

1.5 obscurum − Canonical axis 2 (10.42% of the variance explained **) explained variance the of (10.42% 2 axis Canonical 1

− Residuals = 47.2% N4 2.0 − −3 −2 −10 12 Canonical axis 1 (37.12% of the variance explained ***)

1.5 N3 A Trientalis borealis B

BHI80 139 Lycopodium TM6

1.5 obscurum − Canonical axis 2 (10.42% of the variance explained **) explained variance the of (10.42% 2 axis Canonical 1

− Residuals = 47.2% N4 2.0 − −3 −2 −10 12 Canonical axis 1 (37.12% of the variance explained ***)

Figure 3.12 Canonical redundancy analysis (scaling 3) (A) and partition of variation (B) showing

relations between site and soil characteristics and bacterial phyla in reconstructed soils planted with

coniferous trees (C1 to C5), reconstructed soils planted with deciduous trees (D1 to D5), reconstructed

soils planted with grasses (G1 to G5) and in natural forest soils (N1 to N5). (p<0.001: ***; p<0.01: **)

104

Slightly different variables explained the distribution of classes of microbial communities in the studied soils. Forest-floor depth, pH, and abundance of trees and shrubs, Pinus banksiana (< 5m), and Virburnum edule further added to the already contributing nitrogen deposition, clay content and abundance of Polytrichum juniperinum variables

(F(7,12) = 2.790; p=0.001) (Figure 3.13A). Vegetation data were positively correlated with the soil microbial classes abundant in natural soils, while soil chemical variables were positively correlated with the microbial classes abundant in reconstructed soils (Figure

3.13A). These variables explained 44.6% of the variation in microbial communities at the class level (Figure 3.13B). Alone, the abundance of Pinus banksiana (< 5m), Polytrichum juniperinum and Virburnum edule explained 6.2% of the variation (F(3,12) = 1.563; p =

0.03); nitrogen deposition, soil clay content, soil pH and forest floor depth explained 9% of the same variation (F(4,12) = 1.650; p=0.007). Jointly these variables explained 29.4% of the variation in prokaryotic microbial communities at the class level (F(7,12) = 3.186; p

= 0.001) (Figure 3.13B).

105

A B N4

Clay content Soil characteristics Vegetation S.BQ2.57 soil group - Nitrogen deposition - Pinus banksiana < 5m BD7.11 12 - Forest floor depth - Polytrichum juniperinum Actinobacteria Pinus banksiana Melainabacteria - Clay content - Viburnium edule D4 <5 m G2 - pH C2Acidimicrobiia Sphingobacteriia C1 Deltaproteobacteria Nitrogen deposition ThermoleophiliaD3 Phycisphaerae Polytrichum 9% ** 29.4% 6.2% * Alphaproteobacteria ChloroflexiaG3 G1Flavobacteriia Spartobacteria Cytophagia Opitutae Planctomycetacia Verrucomicrobiae Caldilineae terrestrial group juniperinum

Anaerolineae 0 pH ThermomicrobiaG5C4D1 JG37.AG.4 Gammaproteobacteria G4C5 Acidobacteria Ktedonobacteria Betaproteobacteria D5D2Chthonomonadetes unclassified Thermoplasmata Elusimicrobia N2 Soil Crenarchaeotic Group.SCG. Holophagae N3 Gemmatimonadetes NitrospiraVerrucomicrobia Incertae Sedis N1 C3 Forest Floor Depth

10 Bacilli − Viburnum edule Canonical axis 2 (13.79% of variation explained) ** explained) variation of (13.79% 2 axis Canonical N5 Residuals = 55.4%

−3 −2 −10 123 Canonical axis 1 (58.17% of variation explained)***

A B N4

Clay content Soil characteristics Vegetation S.BQ2.57 soil group BD7.11 - Nitrogen deposition - Pinus banksiana < 5m 12 - Forest floor depth - Polytrichum juniperinum Actinobacteria Pinus banksiana Melainabacteria - Clay content - Viburnium edule D4 <5 m G2 - pH C2Acidimicrobiia Sphingobacteriia C1 Deltaproteobacteria Nitrogen deposition ThermoleophiliaD3 Phycisphaerae Polytrichum 9% ** 29.4% 6.2% * Alphaproteobacteria ChloroflexiaG3 G1Flavobacteriia Spartobacteria Cytophagia Opitutae Planctomycetacia Verrucomicrobiae Caldilineae terrestrial group juniperinum

10 Bacilli − Viburnum edule Canonical axis 2 (13.79% of variation explained) ** explained) variation of (13.79% 2 axis Canonical N5 Residuals = 55.4%

−3 −2 −10 123 Canonical axis 1 (58.17% of variation explained)*** Figure 3.13 Canonical redundancy analysis (scaling 3) (A) and partition of variation (B) showing relations between site and soil characteristics and bacterial classes in reconstructed soils planted with coniferous trees (C1 to C5), reconstructed soils planted with deciduous trees (D1 to D5), reconstructed soils planted with grasses (G1 to G5) and in natural forest soils (N1 to N5). (p<0.001: ***; p<0.01: **)

106 3.4 Discussion:

3.4.1 α- diversity

Contrary to our hypothesis, α-diversity was not significantly different between reconstructed and natural soils, although the average inverse Simpson index was systematically higher in reconstructed soils. Comparing only data from mineral soils,

Dimitriu and Grayston (2010) also found similar α-diversity indices for both natural, undisturbed boreal forest and reconstructed soils of a broader range of ages (3 to 29 years). It might be that the variety of materials used for soil reconstruction in the AOSR

(peat and stockpiled mineral horizons) once mixed together, created a heterogeneous medium that offers a variety of niches to the microbial communities.

3.4.2 Bacterial β- diversity

As predicted, bacterial communities differed between reconstructed and natural soils.

Phyla Actinobacteria, Bacteroidetes and Proteobacteria were more abundant in reconstructed than in natural soils. These bacteria can be categorized as copiotrophic, i.e. bacteria that thrive in nutrient-rich environments and are able to rapidly use a resource when available, but unable to survive in a nutritionally deprived environment. They contrast with the oligotrophic bacteria which are slow-growing and better adapted to nutrient-poor environments (Eilers et al., 2010; Fierer et al., 2007; Koch, 2001). Phyla

Acidobacteria, Cyanobacteria, Elusmicrobia, Firmicutes, Planctomycetes, Verrumicrobia and WD272 were impoverished in reconstructed soils and are predominantly recognized as oligotrophic bacteria. Even if this “copiotrophic-oligotrophic” spectrum is a simplification of reality, members of each category still tend to react in these ways to

107 nutrient availability. For example, copiotrophic bacteria and especially Actinobacteria,

Bacteroidetes, as well as Proteobacteria (particularly α-, β- and γ- proteobacteria), are more abundant when N (Fierer et al., 2012; Freedman and Zak, 2014; Li et al., 2014) or labile C (Eilers et al., 2010; Fierer et al., 2007; Leff et al., 2012) are added to soil via fertilization. In contrast, the abundance of Acidobacteria has been shown to decrease with

N fertilization or deposition (Fierer et al., 2012; Leff et al., 2012), as well as with increased soil C content (Nemergut et al., 2010) or C mineralization rates (Fierer et al.,

2007). It could be that the higher C and N content of reconstructed soils selects for copiotrophic bacteria. However, only estimated N deposition was positively correlated to

Actinobacteria, Acidobacteria, Bacteroidetes and Proteobacteria abundances. This suggests that the most important factor shaping microbial communities in these soils is not the total N or C content of the soils, but the availability of a new and readily available source of N through atmospheric deposition. While copiotrophic bacteria are more abundant in reconstructed soils, they do not possess the metabolic pathways to degrade more recalcitrant carbon, which may promote OM accumulation in these soils (Fontaine et al., 2004; Freedman and Zak, 2014; Leff et al., 2012). The ability to degrade lignin compounds in copiotrophs is still under debate, as there is evidence that Actinobacteria and γ-proteobacteria may be able to secrete enzymes that degrade lignin compounds under certain circumstances (i.e. alkaline or neutral pH) (McCarthy and Williams, 1992;

McCarthy, 1987; Tian et al., 2014; Tuomela et al., 2000). The finding that reconstructed soils select for copiotrophic bacteria with limited capabilities to degrade complex organic-matter might somehow contradict conclusion for the Chapter 2 namely that most of the N was cycle thought recalcitrant organic-N in reconstructed soils. Per contra, it

108 could also reinforce the hypothesis that N is simply exchange through a “adsorption- desorption loop” in reconstructed soils. Abiotic exchange of N would be artificially considered as a mineralization-immobilization process within the 15N pool dilution method used in chapter 2. Moreover, Dimitriu et al. (2010) noted a decrease in activity of phenol oxidase, a key enzyme in lignin decomposition, and in products of lignin degradation, in reconstructed soils of the AOSR. This suggests that the microbial communities most abundant in reconstructed soils may be less capable of degrading recalcitrant OM relative to communities in natural soils. Turcotte et al. (2009) found that most of the OM in the reconstructed soils of the AOSR is composed of undecomposed peat, while Laidlaw (2015) found that about 33% of the C in reconstructed soils was old and biochemically resistant (probably originating from fibric or hemic residual peat), compared to only 10% in natural soils. The higher C:N ratio in the reconstructed soils in this study supports these findings. In contrast, some Acidobacteria have been shown to have the capacity to degrade, utilize and biosynthesize diverse structure of polysaccharide structures (i.e. hemicellulose, cellulose, pectin, starch, chitin, lignin, DNA and urea) (Leff et al., 2012; Li et al., 2014; Rawat et al., 2012). This capacity to utilize complex source of

OM might give Acidobacteria an advantage over copiotrophs in natural soils, hence their greater abundance in natural soils.

3.4.3 Archaeal β- diversity

In the studied soils, archaea represented only 0.6% of the OTUs identified. Similarly, using pyrosequencing techniques, Roesch et al. (2007) found that only 0.009% of the

109 total sequences extracted from a boreal-forest soil of northern Ontario (Canada) were archaeal species, compared with 4 to 12% in agricultural soils.

There were significantly fewer archaea in reconstructed soils planted with trees than in reconstructed soils planted with grass and in natural forest soils. Within the archaea identified, 3% were from the phylum Euryarchaeota (class Thermoplasmata), which were only identified in site N1. Not much is known about the class Thermoplasmata. Cultured species of this class are aerobic, heterotrophic and thermoacidophile and have optimum growth at pH 0.7 to 3 and at a temperature of about 50°C (Angelov and Liebl, 2006). Site

N1 was one of the more recently burned sites (1995) and therefore these archaeal

Thermoplasmata may be remnant from extreme post-fire conditions.

The remainder of the archaea identified (97%) belonged to the phylum Thaumarchaeota.

Reconstructed soils planted with grass harboured more organisms from the

Thaumarchaeota class Soil Crenarchaeotic Group (also known as group 1.1b) than reconstructed soils planted with trees. However, the Soil Crenarchaeotic Group only comprised 9% of the Thaumarchaeota sequences in the natural soils. The rest of the sequences identified in the natural soils (91%) belonged to the Terrestrial Group (also known as group 1.1c). All known members of the Soil Crenarchaeotic Group are autotrophic ammonia-oxidizing archaea (AOA) (Lin et al., 2015; Weber et al., 2015).

Nicol et al. (2008) demonstrated that the abundance of the bacterial amoA gene (one of the genes responsible for ammonia oxidation) decreased under acidic conditions, while the archaeal amoA gene abundance increased in acidic conditions. These results

110 demonstrate the potential importance of archaea (and especially of Thermoarchaeota) for the nitrogen cycle in acidic soils.

The Terrestrial Group of archaea, restricted to natural soils, seems to prefer even lower pH than organisms from the Soil Crenarchaeotic Group (He et al., 2012; Lehtovirta et al.,

2009; Oton et al., 2016). Organisms from the Terrestrial Group composed 29% of all

16SrRNA gene sequences in acidic soils of a mixed deciduous forest in Germany and were not detected in soils with pH above 4.5 (Kemnitz et al., 2007; Nicol et al., 2005).

The canonical ordination of prokaryotic classes supports the influence of pH on the abundance of Thaumarchaeota classes in the current study, as the abundance of the Soil

Crenarchaeotic Group was positively correlated with pH while the abundance of the

Terrestrial Group was negatively correlated with pH. Weber et al. (2015) found that, compared to the other members of the phylum Thaumarchaeota, organisms from the

Terrestrial Group were not able to grow autotrophically (gaining energy from the oxidation of ammonia to nitrite) and therefore showed no evidence of ammonia oxidation during growth.

These results indicate that reconstructed soils planted with grass promote ammonia- oxidizing archaea of the class Soil Crenarcheotic Group. The lower pH in natural soils seems to limit the growth of ammonia-oxidizing archaeal species, therefore promoting archaeal species from the Terrestrial Group which seem unable to grow autotrophically by oxidizing ammonia. The higher prevalence of AOA in reconstructed soils planted with

+ - grass could induce conversion of ammonium (NH4 ) to nitrate (NO3 ) in these soils.

111 Because nitrate is an anion and most of the exchange sites in soils are negatively charged,

- NO3 is less stable in soil and so could be further lost by leaching or gas emission

(Robertson and Groffman, 2007). The greater abundance of AOA in reconstructed soils planted with grass may, therefore, have a significant effect on the nitrogen cycle in these soils. However, the greater abundance of AOA in the reconstructed soils planted with

+ grass did not translate into higher gross rates of NH4 oxidation in the studied soils (see

Figure 2.6). This suggests that the AOA community might not be actively contributing to

+ - the conversion of NH4 to NO3 in the studied soils.

3.4.4 Influence of vegetation on α- and β- diversity

Vegetation, more than soil characteristics, was related to bacterial α- and β- diversity in all the studied soils. α-diversity was correlated with the abundance of alfalfa (Medicago sativa), beaked willow (Salix bebbiana), velvetleaf blueberry (Vaccinium myrtriolloides) and jack pine (Pinus banksiana). A diverse plant community can create a high level of heterogeneity in root-exudate patterns capable of supporting a diverse microbial community (Ding et al., 2013; Kowalchuk et al., 2002). The rhizosphere of alfalfa has been shown to harbour more bacteria than other grass species, such as rye (Miethling et al., 2000). Through its root exudates, alfalfa could, therefore promote α-diversity in reconstructed grassland soils. The fact that alfalfa was present only in the reconstructed soils that were re-vegetated using grasses and especially in site G4, which had the highest

α-diversity among all the studied soils (data not shown), supports this hypothesis. Alfalfa

+ also commonly host N-fixing bacteria. This association can increase increasing soil NH4 content, N-cycling rates and the presence of ammonia-oxidizers in grassland soils.

112 Diverse root-exudate production in the rhizosphere could also result in the increase α- diversity of bacteria in the soils revegetated with beaked willow and velvetleaf blueberry.

Both plants can develop a symbiotic relation with fungi: ectomycorrhizae for the beaked willow and ericoid mycorrhizae for the velvetleaf blueberry (Bell et al., 2015, 2014;

Kourtev et al., 2002). Willows promote specific ectomycorrhizal fungi, which, once established, increase bacterial diversity in the soil (Bell et al., 2015). Willows also produce a small, but highly labile pool of organic carbon that can support abundant copiotrophic proteobacteria (Männistö et al., 2013). Similarly, following formation of ericoid mycorrhizae on velvetleaf blueberry, a diverse set of bacteria, including biofilm bacteria, mycorrhizal helper bacteria and plant-growth-promoting rhizobacteria can colonize and proliferate around the roots, increasing soil bacterial diversity (Robertson et al., 2007). These results indicate that specific vegetation, through effects on either root exudates or symbiotic mycorrhizal relationships, or both, can create soil microenvironments which favour diverse microbial communities – even in reconstructed soils. The only type of vegetation that was associated with a reduction in bacterial α- diversity was jack pine. This may be related to jack pine being an early successional species in the boreal forest; jack pine was predominant in the most recently burned natural stands (site N1 – burned in 1995 and site N4 – burned in 1981). Pine also tends to dominate on the driest and most nutrient-poor sites, which may also contribute to the low bacterial α-diversity in soils under the natural jack-pine stands.

Vegetation also correlated with the structure of microbial communities within the natural soils. Phyla Cyanobacteria, Planctomycetes and Verrucomicrobia were positively

113 correlated with the presence of Polytrichum juniperinum. Cyanobacteria are known for their association with bryophytes (DeLuca et al., 2002; Gundale et al., 2013; Rousk et al.,

2013; Zackrisson et al., 2004). While moss provides a stable living-environment for the free-living Cyanobacteria, these bacteria fix nitrogen and can preferentially leak N to the moss (DeLuca et al., 2002; Rousk et al., 2014). These moss–Cyanobacteria associations can replenish the soil N-pool, and therefore are an important mechanism in N-limited environments, such as boreal forests. Nitrogen fixation has been shown to increase in boreal forests with time since fire because of moss re-establishment, and with it, the increased presence of Cyanobacteria (Zackrisson et al., 2004). On the other hand, in areas of high N deposition (near roads in Sweden), N-fixation rates decreased, as did the abundance of mosses and Cyanobacteria (Ackermann et al., 2012; Gundale et al., 2013).

The positive correlation between Cyanobacteria and the abundance of Polytrhichum juniperium and their negative correlations with N deposition in this study support this mechanism. Verrucomicrobia and Planctomycetes were also positively correlated with the abundance of Polytrichum juniperinum. Perhaps these bacteria also colonize mosses.

It has been shown that Planctomycetes are capable of anaerobic ammonium oxidation to dinitrogen with nitrate as an electron acceptor (Hayatsu et al., 2008; Mulder et al., 1995).

It could be hypothesized that the high N concentrations in the moss stimulate colonization by annamox-capable bacterial species. Perhaps the moss–bacteria association is contributing to the nitrogen cycle beyond nitrogen fixation. These results indicate that vegetation, more than soil characteristics, drive bacterial α- and β-diversity, highlighting the importance of above- and belowground relationships in shaping soil microbial communities.

114

3.4.5 Influence of soil characteristics on α- and β- diversity

While vegetation was positively correlated with microbial community structure (β- diversity) in natural soils, pH and N deposition were positively correlated with the most abundant microbial communities in reconstructed soils. Abundance of Actinobacteria,

Bacteroidetes and Proteobacteria were positively correlated with pH. On a continental scale, soil pH has been shown to drive bacterial community structure; Acidobacteria preferred acidic environments, Actinobacteria and Bacteroidetes preferred alkaline environments, and Proteobacteria have an optimum pH of 6 (Lauber et al., 2009). Soil pH was generally slightly more alkaline in reconstructed soils, which may have favoured the growth of Actinobacteria, Bacteroidetes and Proteobacteria in these soils. The precedence of soil characteristics over tree cover in shaping microbial communities in reconstructed soils is not unexpected. Schrijver et al. (2012) did not detect a significant tree-species effect on soil properties (at a depth 10–20 cm in mineral soil) until 35 years after reforestation of agricultural land in northern Belgium. The reconstructed oil-sands soils that we studied were planted less than 30 years ago. At this time, it appears that the residual peat used for soil reconstruction are still overwhelming any influence of arboreal vegetation on community structure in the mineral portion of the reconstructed soils.

3.4.6 Microbial community structure in reconstructed soils planted with grass

Within reconstructed soils, sites planted with grasses had more bacteria from classes

Anaerolineae, Caldilineae and Chloroflexia of the phylum Chloroflexi than sites planted with either deciduous or coniferous trees. Members of the phyla Chloroflexi are slow-

115 growing, heterotrophic bacteria ubiquitous in natural ecosystems (Yamada and

Sekiguchi, 2009). Both Anaerolinae and Caldilineae are anaerobic (or facultative aerobic for Caldilineae) classes of Chloroflexi that can be found at depth of 140 m in both marine and terrestrial environments (Breuker et al., 2011; Yamada and Sekiguchi, 2009). Phyla

Chloroflexi and grassland soils were positively correlated with soil clay content in the redundancy analyses. Soil bulk density was not significantly higher in grassland than other soils (data not shown), although soil compaction was noted during field sampling at grassland sites. Laidlaw (2015) showed that these grassland soils had the highest average proportion of microaggregates of any of the sites. These results suggest that the interaction between roots, clay content, compaction and aggregation might increase the number of anaerobic micro-environments in grassland soils compared to the other reconstructed soils, which would favour anoxic classes of Chloroflexi (Anaerolineae and

Caldilineae).

As mention in section 4.3, archaeal species of the Thaumarchaeota class Soil

Crenarchaeotic Group were also more abundant in reconstructed soils planted with grass.

The micro-environments of reconstructed grassland soils could select for these ammonia- oxidizing species, hence favouring nitrification in these soils and potentially increasing losses of nitrogen through leaching or gas emission. Higher total gross nitrification and net nitrification rates were measured in the grassland soils (see Figures 2.6 and 2.7).

However, autotrophic nitrification was not higher in the reconstructed soils planted with grass. This suggests that the AOA community might not be actively contributing to

+ - conversion of NH4 to NO3 in the studied soils.

116

3.4.7 Effect of disturbances, other than forest-fire, on β- diversity in natural soils

Natural site N1 had a different prokaryotic microbial community than the other natural soils, and the community was more similar to reconstructed soils. Soil N1 burned in 1995 and was visited in 2012. However, between summer 2012 and summer 2013, the young forest stand was clear-cut to permit the establishment of a small access road. During the summer of 2013, soil at this site was collected from an undisturbed forest stand approximately 10 metres away from the newly constructed road. As such, the establishment of the access road at site N1 may have modified soil conditions, such as water and air movement, due to compaction, and therefore, changed the structure of microbial communities. The finding that site N1 did not tightly group with the other reconstructed sites in the multivariate analyses supports the hypothesis that these altered environmental conditions are responsible for the distinct microbial community in N1.

3.5 Conclusion

Using next-generation sequencing, we were able to distinguish the prokaryotic communities in reconstructed and natural forest soils in the AOSR. Contrary to our hypothesis, α-diversity did not differ in reconstructed and natural soils and was actually higher in reconstructed soils. Vegetation cover, especially Pinus banksiana, Medicago staiva, Salix bebbina and Vaccinium mytrilloides, was the main factor influencing α- diversity. As predicted, β-diversity differed between reconstructed and natural forest soils. Within reconstructed soils, those planted with grasses harboured a distinct microbial community than those planted with trees. Copiotrophic bacteria

117 (Actinobacteria, Bacterioridetes and Proteobacteria) were more abundant in reconstructed soils, whereas oligotrophic bacteria (Acidobacteria, Cyanobacteria, Elusmicrbia,

Firmicutes, Planctomycetes and Verrumicrobia) were more abundant in natural forest soils. Ammonia-oxidizing archaea and anaerobic bacteria (from classes Anaerolineae,

Caldilineae and Chloroflexia) were more abundant in reconstructed soils planted with grasses. Nitrogen deposition, pH, clay content and plant cover were the main variables influencing the structure of the communities. Taken together, these results indicated that key soil and site characteristics, such as N deposition, pH and plant cover engendered distinct prokaryotic communities in the reconstructed and natural forest soils.

118 Chapter 4: Nutrient availability and vegetation cover as drivers of fungal α- and β- diversity in reconstructed oil-sands soils and natural boreal-forest soils

4.1 Introduction

The Athabasca Oil Sands deposit, located in the boreal forest of northern Alberta, is part of the largest single oil deposit in the world, with proven reserves of 166 billion barrels of bitumen, and covering 142,200 km2 (Government of Alberta, 2016b). Most (80%) of the bituminous sands can be extracted using in situ recovery methods, but 20% of the resource is shallow and can be recovered through open-pit mining (Government of

Alberta, 2012). To date, about 895 km2 of land has been disturbed by oil-sands mining activity (Government of Alberta, 2016b). Following surface mining, companies are required to restore soils that can achieve equivalent land capability (Government of

Alberta, 1993; Powter et al., 2012). After soil reconstruction, the area is re-vegetated.

When reclamation in the area began in the 1980’s, revegetation predominantly focused on erosion control, and used both native and introduced grasses and shrubs. However, more recent revegetation practices use native tree species such as jack pine, white and black spruce and aspen and understory shrubs such as blueberry and willow to re- establish a boreal-forest plant community. It is unlikely that these reconstructed forest soils will exactly mirror pre-existing boreal-forest soils (Chazdon, 2008; Hobbs et al.,

2006), so novel soil ecosystems will probably arise from the reconstruction efforts. Re- establishing soil functions, chiefly nutrient cycling, rather than trying to replicate the

119 structural qualities of the previous soil ecosystem is necessary to ensure the long-term sustainability of reclaimed boreal forest landscapes (Quideau et al., 2013).

Over 100,000 species of fungi have been described globally to date (Tedersoo et al.,

2014), but current estimates are that there are up to to 5.1 millions of fungal species in existence (Blackwell, 2011). In acidic, nutrient-poor environments fungal biomass can surpass the belowground biomass of all other soil-organisms combined, except for plant roots (Brady and Weil, 2002; Thorn and Lynch, 2007). Fungi are keystone organisms for soil functioning. With their ability to penetrate solid soil particles, plant-material or woody debris, fungi are the main decomposers of organic matter in forest soils

(Hättenschwiler et al., 2005; Sinsabaugh, 2010; Thorn and Lynch, 2007). The ability of saprotrophic fungi to degrade complex organic matter, mainly through the use of phenol oxidase and peroxidase extracellular enzymes, regulates soil nutrient availability

(Lonsdale et al., 2008; Sinsabaugh, 2010). The majority of the saprophytic fungi also belong to the phyla Basidiomycota and Ascomycota, but are not limited to them as some species of Chytridiomycota and Zygomycota are also able to degrade organic matter

(Thorn and Lynch, 2007). In addition to their ability to degrade organic matter, many fungi can also form symbiotic relationships with plant roots (Brady and Weil, 2002;

Hättenschwiler et al., 2005; Thorn and Lynch, 2007). In this symbiotic relationship, the plant supply the fungi in recently photosynthesised C and, in return, the fungi uses its mycelia to explore vast volume of soil and gain access to nutrients (N and P) and water that are, then, transfer to the plant (Binkley and Fisher, 2013). In boreal forests, these mycorrhizal fungi can be responsible for the majority of the nitrogen acquired by plants

120 annually (Van Der Heijden et al., 2008). There are two main classes of mycorrhizae that are important for plants. The ectomycorrhizae are characterized by fungal hyphae that do not penetrate inside of the root cells and usually form a compact mantle around the roots.

The arbuscular mycorrhizae penetrate inside the root cell and usually do not form a fungal mantle around the roots (Binkley and Fisher, 2013). The vast majority of ectomycorrhizal fungi belongs to the phyla Basidiomycota and some to Ascomycota, while Glomeromycota form the majority of arbuscular mycorrhizae relationships

(Binkley and Fisher, 2013; Thorn and Lynch, 2007). Fungi also play a major role in humus formation and aggregate stability adding the role of soil architecture to their capabilities (Brady and Weil, 2002).

In a portrait of global fungal biogeography, Tedersoo et al. (2014) found that the

Basidiomycota fungal class Agaricomycetes (mainly ectomycorrhizal and saprotrophic) largely dominated boreal-forest ecosystems followed by the Ascomycota classes

Leotiomycetes (diverse functions including saprophytic, orchid mycorrhizal and pathogenic) and Eurotiomycetes (diverse functions including mycorrhizal and parasitic).

Within mycorrhizal fungi, ectomycorrhizae are mostly restricted to trees and dominate in temperate, boreal and alpine forests, while arbuscular mycorrhizae form symbiotic relationships with many herbaceous species, woody tropical plants and some deciduous tree species (Acer, Alnus, Fraxinus and Populus) (Binkley and Fisher, 2013; Thorn and

Lynch, 2007). In a study of boreal forest soils across a fertility gradient of 25 Swedish old-growth forests sites, fungal abundance was influenced by soil pH, C:N ratio and

+ NH4 content (Sterkenburg et al., 2015). Ascomycetes, especially from class

121 Leotiomycetes, Chaetothyriales and Archaerhizomycetes, dominated less fertile soils, while Basidiomycetes were more abundant in more fertile soils (Sterkenburg et al.,

2015). Because trees should not rely as heavily on symbiotic relationships in high nutrient environments as in low-nutrient environments, it was believed that mycorrhizal development may be reduced or even prevented in high N-environments (Binkley and

Fisher, 2013). However, Kranabetter et al. (2015) showed that ectomycorrhizal genera such as Tomentella were more abundant in N-rich environments, while the genera

Cortinarius was more abundant in N-poor environments. Fungal functional traits, such as

+ NH4 uptake rate was hypothesized to define those ectomycorrhizal environmental niches. Fungal species also seem to be vertically segregated in the soil profile as a shift from saprotrophic fungi to mycorrhizal fungi has been observed from the top of the forest floor to the mineral part of a boreal-forest soil profile (Clemmensen et al., 2015; Lindahl et al., 2007). Because of their ability to degrade more complex organic matter, saprotrophic fungi (especially white-rot fungi) might have a competitive advantage over mycorrhizal species in undecomposed-fresh litter, therefore explaining the higher abundance of saprotrophic fungi in forest-floor material (Lindahl et al., 2007).

Fire is the major disturbance in boreal ecosystems and can cause shifts in microbial communities (Ball et al., 2010; Hart et al., 2005; Switzer et al., 2012). Fungal biomass has been shown to decrease, while fungal richness and diversity were noted to be higher in recently burned soils (Holden et al., 2013; Kennedy and Egger, 2010; Sun et al., 2015).

It was hypothesised that changes in soil organic inputs due to fire (i.e. higher abundance and diversity of woody debris versus leaf-litter) boost the diversity of saprotrophic fungi,

122 thereby increasing general fungal diversity in recently burned soils (Kennedy and Egger,

2010). Fungal community structure is also changed by wildfire (Holden et al., 2013;

Kennedy and Egger, 2010; Sun et al., 2015). Changes in microbial communities can be prompted by lysis of microbial cells and death of plants roots directly due to fire (Hart et al., 2005; Kennedy and Egger, 2010). The post-fire survival of mycorrhizal species can also be influenced by changes in above-ground vegetation, as mycorrhizal species are plant-host dependent. Because ectomycorrhizae are more limited in the number of possible hosts than are arbuscular mycorrhizae, fire was hypothesised to reduce the abundance of ectomycorrhizal species more than the abundance of arbuscular mycorrhizae (Hart et al., 2005). Supporting that hypothesis, Holden et al. (2013) noted a decreased in Basidiomycota in recently burned soils in Alaskan boreal soils, which was attributed to a decline in ectomycorrhizal species in the recently burned soils. In a fire- chronosequence in the northern boreal subarctic coniferous forest of Finland, Sun et al

(2005) found that Ascomycota-saprotrophic fungi were more abundant in recently burned forests, whereas the abundance of Basidiomycota-mycorrhizal species increased with time since fire. Genes encoding enzymes for hemicellulose degradation were also more abundant in the younger soils, suggesting that the higher availability of easily degradable organic matter in recently burned soils favour its colonization by saprotrophic fungi.

Previous studies of microbial communities in reconstructed oil-sands soils ranging in age from 5-35 years, using denaturing gradient gel electrophoresis (DGGE) and phospholipid fatty acid (PLFA) profiling, showed reduced abundance and different in microbial community structure in reconstructed soils compared to natural soils (Dimitriu et al.,

123 2010; Hahn and Quideau, 2013). DGGE and PLFA do not allow for in-depth (species- level) characterization of microbial communities, however, it was possible to detect that fungal communities in natural soils in the AOSR were dominated by Basidiomycota species, while reclaimed soils were dominated by Zygomycota and Ascomycota species

(Dimitriu et al., 2010). In a study of thirty-two 16- to 33-year-old reclaimed soils in the

AOSR, Sorenson et al. (2011) found that reclaimed soils planted with coniferous trees had higher fungal abundance than soils reclaimed sols planted with deciduous trees.

Sorenson et al. (2011) also found that vegetation affected microbial community composition only when canopy cover was above 30%. Below this level, soil microbial communities were mainly influenced by characteristics of the materials used during reclamation.

Since fungi play an essential role in soil functioning and especially in decomposition and nutrient cycling, the objective of this study was to assess if soil fungal diversity and structure (i.e. α- and β- diversity) in oil-sands soils reconstructed 20-30 years previously were similar to those found in natural boreal-forest soils that had been subject to wildfire disturbance at approximately the same time. Specifically, we evaluated fungal diversity and community structure in the top 12 cm of mineral soils of 20- to 30-year-old reconstructed and fire-disturbed boreal soils using massively parallel sequencing of ITS 2 region. We also assessed the influence of the vegetation treatments (coniferous trees, deciduous trees and grasses) used in soil reclamation as well as the influence of plant cover, soil chemical and physical characteristics on fungal α- and β- diversity. We hypothesised that a) α-diversity will be higher in natural soils; b) β-diversity will differ

124 between reconstructed and natural soils; c) Basidiomycota fungal species will be more abundant in natural soils, Ascomycota and Zygomycota fungal species will be more abundant in reconstructed soils, and Glomeromycota will be more abundant in reconstructed soils planted with grass; d) reconstructed soils planted with coniferous trees will have greater fungal abundance compared to reconstructed soils planted with deciduous trees or grass and e) pH and nutrient content of the soils will be the main drivers of fungal α- and β- diversity.

4.2 Materials and methods

4.2.1 Study area

The study area was situated in the Athabasca Oil Sands Region (AOSR) in northern

Alberta, Canada (56°39’N, 111°13’W, altitude: 369m). Short warm summers and long cold winters characterize the climate. The mean annual temperature is 1°C, ranges from -

17.4°C in January to 17.1°C in July. Mean annual precipitation is 418.6 mm, of which

316.3 mm occurs as rainfall during the growing season (Environment Canada, 2015).

Regions Committee, 2006). This region falls within the central mixedwood region of the

Canadian boreal forest. Dominant tree canopy species in upland landscapes are trembling aspen (Populus temuloides Michx), white spruce (Picea glauca (Moench) Voss) and jack pine (Pinus banksiana Lamb) (Natural Regions Committee, 2006). Fire is the major natural disturbance in these forests (Thomson, 1979).

125

Oil-sands mining activities involve the removal of surface soil materials followed by the removal of 40 m of overburden material (approximate regional average) to expose the oil-sand ore body. Salvaged soil materials are preferably used for reclamation of an area within the footprint of the mines that is ready for reclamation, or are stockpiled for later use. The overburden is used for berm, dyke wall or road construction, or deposited in a dedicated disposal area to create large-scale overburden landform units. The oil-sand ore is transported to the extraction and upgrading facility. Oil-sands soil reconstruction involves a number of cover designs, depending on the landform substrate being reclaimed. There are two main cover designs: one uses only cover-soil and the other one consists of a combination of cover-soil and of subsoil. Cover-soil and subsoil materials are salvaged from surface soils within the mine-development footprint. Only sites at which cover-soil had been placed on top of overburden material were used in this study.

The cover-soil materials used consisted of surface peat mixed with mineral soils material having a loam or coarser texture and is hereafter referred to as ‘peat-mineral mix’. In the studied soils, the depth of the peat-mineral mix ranged from 10 cm to more than 100 cm.

Early objectives for re-vegetation objectives in the AOSR were to establish native or introduced grass and shrub species to control erosion; however, oil-sands operators are now required to use native tree and species with the objective of re-establishing a boreal forest community. During the period when the sites used in this study were reclaimed

(20-30 years ago), 250 to 350 kg ha-1 of varying proportions of N:P:K fertilizer was typically applied in the first year of re-vegetation.

126 4.2.2 Study sites

Fifteen reconstructed sites and five natural forest sites were studied in the AOSR; all sites had been reconstructed (or naturally fire-disturbed by wildfire) 20 to 30 years previously.

The 15 reconstructed sites were previously studied by Sorenson et al. (2011). Five sites had been planted with deciduous species (mostly trembling aspen), five with coniferous species (mostly white spruce), and five with grasses (fescue, slender wheatgrass and alfalfa). The natural sites have similar time-since-disturbance (fire) and similar soil texture as the reconstructed soils. The natural soils were classified as Gleyed Eluviated

Eutric Brunisol (soil N1), Eluviated Eutric Brunisol (N5) or Brunisolic Gray Luvisol

(soils N2, N3, N4). The natural forest sites are located approximately 40 to 150 km south of the town of Fort McMurray (Alberta, Canada).

At each site, one 10-m2 plot was sub-divided into 10 1-m2 and 7 subplots were randomly selected. Soil horizons were described in a 30-cm-deep soil pit in each sub-plot. Any litter or forest floor layers were removed and the top 0–15 cm of the mineral soil (or peat- mineral mix) was sampled (~ 1 kg) at each sub-plot using a plastic trowel, and the 7 subplot samples were pooled to produce one sample per plot (nsample = 20). A sub-sample

(~5 g) of each soil was immediately put on ice for soil microbial analyses. The remaining samples were put in a cooler and kept cold until arrival at the laboratory for soil chemical and physical analyses. Removal of any surface organic material prior to sampling enabled us to directly compare the mineral substrate on which the organic layers develop, and so determine how similar the reconstructed soils are to natural soils. If we assume that the similar soil development factors – the action of climate, organisms, relief through time –

127 are at play in reconstructed and natural boreal forest soils with few natural variation such as parent material, types of organisms, etc., the peat-mineral mix is the analogue of the mineral part of the natural soils on top of which a forest floor develops. Characteristics of the forest floor layers in reconstructed and natural soils in the AOSR have been described in previous research (Rowland et al., 2009).

Vegetation is described in Anderson (2014) (Supplemental material). Tree density was not significantly different among reconstructed soils planted with deciduous species, coniferous species and naturally disturbed sites. Grassland sites were devoid of trees.

4.2.3 Laboratory analyses

4.2.3.1 Soil characteristics

Soil moisture at each site was determined using soil cores from one of the subplots.

Samples were weighed, dried overnight at 105°C and weighed again to determine the volumetric water content (Blake and Hartge, 1986). Soil pH in water was determined in a

1:10 (soil:water) solution using a UB-10 pH meter (Denver instruments, Bohemia, USA)

(Thomas, 1996). Cation exchange capacity (CEC) was calculated as the sum of Ca, Mg,

K, Na, Al, Fe and Mn that could be exchanged by BaCl2 and measured using an Argilent

240 Atomic Absorption Spectrometer (Argilent, Santa Clara, CA, USA) (Hendershot and

Duquette, 1986). Soil total carbon concentrations were measured with an Elementar

Vario El Cube Elemental Analyzer (Elementar, Hanau, Germany) using 15 mg of grinded and sieved (< 0.5 mm) soil to assured homogenization (Rutherford et al., 2008;

Skjemstad and Baldock, 2008). Analyses were completed in triplicate after homogenizing

128 - the sample (Konert and Vandenberghe, 1997). Nitrate (NO3 ) was extracted using 10 g of soil and 100 ml of 2M KCl (1:10 ratiow:v). Samples were shaken for 1 hour on a mechanical shaker and filtered through a ø 12.5 cm fiberglass G6 microfilter (Fisher

Scientific, Loughborough). Extracts were sent to the University of British Columbia

Environmental Engineering Department to be analysed using a Lachat Flow Injection

Analyzer (FIA) QuickChem 8000 (Loveland, Colorado) for colorimetric determination of

- total NO3 concentrations. Microbial biomass carbon was measured on the day of the sampling using the chloroform fumigation extraction (CFE) technique (modified from

Tate et al., 1988). TOC in both fumigated and unfumigated samples were determined by oxidative combustion and infra-red analysis using a TOC-TN autoanalyzer (Shimadzu,

Kyoto, Japan). Unless otherwise mentioned, 25% of the samples were analyzed in duplicate and controls (AgroMAT AG-2 soil standard, SCP Science, Baie d’Urfé,

Canada) were added.

4.2.3.2 DNA extraction and amplification

Soil DNA was extracted in triplicate using PowerSoil ® DNA isolation kit (MO BIO,

Carlsbad, CA, USA) with 0.25g of soil. Triplicates were pooled together and sent to

Génome Québec (Montréal, Canada) for sequencing. The Internal Transcribed Spacer region ITS 2 were amplified using the ITS 1F (5’- CTTGGTCATTTAGAGGAAGTAA -

3’) and ITS4 (5’- TCCTCCGCTTATTGATATGC -3’) primers with expected length of

300 base pairs (bp) using the following program: 95°C for 15 minutes, followed by 34 cycles of 95°C for 1 minute, 51°C for 1 minute and 72°C for 1 minute and a final 7 minutes at 72°C (Větrovský and Baldrian, 2013). This primer pair is used for

129 phylogenetic studies to target specifically fungi species and allow for a good assignment quality due to its length (300 bp). PCR products (n=20) were barcoded, pooled and sequenced using Illumina MiSeq sequencer with 250 base pairs on the forward and the reverse reads (Illumina, San Diego, CA, USA).

4.2.4 Bioinformatics analyses

Sequences (752,587) obtained from Genome Quebec were treated and analyzed using

Mothur MiSeq SOP pipeline (Kozich et al., 2013). Briefly, forward and reverse sequences were merged. Non-ambiguous sequences shorter than 500 bp were kept.

Sequences were further screened for chimeras using chimera.uchime script (Edgar et al.,

2011). Singletons, chloroplasts and mitochondria sequences were removed from the dataset. At the end of the screening process, 530,811 sequences were kept and clustered in 1293 operational taxonomic units (OTU). We assigned the taxonomy of each OTU using the UNITE database. To facilitate comparison among samples, we subsampled sequences from each sample in our dataset to the lowest number of sequences found in a single sample, which was 6819 sequences. Because they generated fewer unclassified sequences, we analyzed the data using the taxonomic level of phyla and classes.

4.2.5 Calculations and statistics

4.2.5.1 Good’s coverage

Good’s coverage (C) is an indicator of the depth of sequencing. It estimates the proportion of total species that are represented in a sample using equation 1.

130 % C = 1 − & (eq. 1) '

Where, n1 is the number of OTU’s that have been sampled once and N is the total number of individuals in the sample (Kozich et al., 2013).

4.2.5.2 α-diversity

The fungal diversity in each site (α-diversity), was calculated using the inverse Simpson diversity index (Simpson, 1949). The Simpson index measures the probability that two sequences taken at random from the dataset represent the same OTU. The lower the index, the lower the probability of getting two identical OTU from the dataset, and the higher the α-diversity. Taking the inverse of the Simpson index transforms the index to a more intuitive index for which the α -diversity is higher when the index is higher. Mean and standard errors of the inverse Simpson index for each treatment are presented.

Differences among treatment types (reconstructed with deciduous, coniferous and grass species and natural sites) were assessed using one-way ANOVA with permutations testing (Legendre, 2007). A multiple linear regression model using forward selection was applied to find variables that significantly contribute to α-diversity in our samples.

4.2.5.3 β-diversity

β-diversity is the differentiation of fungal composition among habitats. It was assessed using four methods. A Principal coordinate analysis (PCoA) using a Bray-Curtis dissimilarity matrix on community data allowed us to visually assess if soils at the 20 sites harboured different fungal communities (Borcard et al., 2011). PCoAs were

131 followed by an analysis of molecular variance (AMOVA) – a nonparametric analysis of variance – to test if the centers of the clouds representing a treatment were more separated than the variation among samples of the same treatment itself; in other words it test if the genetic diversity within two (or more) communities was greater than their pooled genetic diversity (Kozich et al., 2013; Schloss, 2008). Weighted UniFrac distances were measured and tested for significant differences among groups (Lozupone et al., 2007). Briefly, the un-weighted UniFrac method measures the phylogenetic distance among communities in a phylogenetic tree “as the fraction of the branch length of the tree that leads to descendants from either one environment or the other, but not both” (Lozupone and Knight, 2005). The weighted version of the UniFrac uses the same philosophy, but takes into consideration the abundance of each taxa belonging to a lineage (Lozupone et al., 2007). Finally, we also evaluated β-diversity using one-way

ANOVA with permutations testing on all phyla and classes (Legendre, 2007) and by representing the data using the Z-scores (eq. 2) .

)*) Z = (eq .2) +

Wallis tests were performed using the R package pgirmess (Giraudoux, 2014).

132 4.2.5.4 Relationships between environmental variables and communities

A canonical redundancy analysis (RDA) was performed to identify variables that significantly explain the distribution of fungal communities in the studied soils (Legendre and Legendre, 2012). Significant variables were selected using a forward selection algorithm – packfor’s forward.sel function (Dray et al., 2011). The RDA model, axis and explanatory variables were tested using permutations testing (Borcard et al., 2011). RDAs are presented using scaling 2 in which the angles in the biplot between the fungal phyla or classes (the response variables) and the soil or vegetation characteristics (the explanatory variables), and between the response variables themselves or explanatory variables themselves, reflect their correlation (Borcard et al., 2011). Statistical analyses were done using the software R (R Core Team, 2014).

4.3 Results

4.3.1 Soil chemical, physical and biological characteristics

-6 Soil CEC (F(3,16) = 28.07; p=1.30x10 ) and soil total carbon (F(3,16) = 12.05; p=0.0002) were significantly greater in the reconstructed soils planted with deciduous trees and grasses than in the natural forest soils. Microbial biomass carbon was significantly greater (F(3,16) = 4.613; p=0.0165) in the reconstructed soils planted with deciduous trees than in the natural soils. Soil pH (F(3,16) = 8.085; p=0.0017) and soil moisture (F(3,16) =

2.608; p=0.0874) were higher in reconstructed soils under grasses than in natural soils.

- NO3 concentration did not significantly differ between reconstructed and natural soils

(F(3,16) = 0.934; p=0.45) (Table 4.1).

133

Table 4.1 Selected physical. chemical and biological characteristics of the studied soils

Soil Nitrate Microbial biomass Canopy CEC Total carbon Site pH moisture -1 concentration carbon cover (cmol+ kg soil ) (%) -1 -1 (%) (mg NO3-N kg dry soil ) (g C kgsoil ) (%) Coniferous (mean±sd) 6.5±0.2ab 20.2±12.1ab 37.7±11.0ab 7.6±4.7ab 3.1±1.8a 1.31±0.37ab 60.8±14.6ab C1 6.4 11.6 31.3 3.83 1.59 1.15 58.4 C2 6.4 18.5 30.6 7.84 6.10 1.60 72.6 C3 6.6 39.9 56.9 15.51 2.94 1.27 70.1 C4 6.9 21.3 36.5 4.61 2.45 1.73 66.2 C5 6.5 9.6 33.2 6.35 2.18 0.79 36.5 Deciduous (mean±sd) 6.4±0.3ab 22.9±6.3ab 50.5±14.7a 7.6±2.2a 3.8±3.5a 1.53±0.84a 52.5±17.3a D1 6.5 22.3 68.8 7.22 9.73 1.53 72.2 D2 5.8 27.1 47.2 11.42 4.24 2.84 61.6 D3 6.5 16.1 34.9 5.43 2.65 0.58 53.6 D4 6.6 31.3 39.1 6.92 1.30 1.64 25.7 D5 6.5 17.8 62.5 7.16 1.26 1.08 49.3 Grassland (mean±sd) 7.0±0.3a 25.3±8.1a 52.3±6.4a 7.5±1.8a 4.4±4.1a 0.93±0.44ab 0.0±0.0b G1 6.5 25.2 57.8 7.16 2.22 0.98 0.0 G2 7.2 35.9 59.1 10.39 1.31 1.35 0.0 G3 7.1 26.9 50.5 7.99 1.39 1.03 0.0 G4 7.1 13.3 43.3 5.85 10.66 0.17 0.0 G5 7.0 25.2 50.8 6.27 6.28 1.10 0.0 Natural (mean±sd) 6.1±0.3b 12.1±1.4b 2.6±1.2b 0.39±0.13b 1.5±0.5a 0.53±0.29a 42.7±19.1ab N1 6.6 10.8 1.3 0.31 1.82 0.93 33.0 N2 6.1 12.8 3.1 0.36 0.81 0.24 42.7 N3 5.8 10.5 3.6 0.53 2.15 0.35 69.6 N4 5.9 13.5 3.7 0.49 1.38 0.41 49.7 N5 6.1 13.1 1.5 0.21 1.59 0.73 18.5

134 4.3.2 Soil fungal community structure

The number of sequences extracted (F(3,16) = 1,68; p=0.2114) and the number of sequences kept after screening (F(3,16) = 2.8452; p=0.07) were similar for the reconstructed soils and the natural forest soils. After combining these sequences in OTUs, reconstructed soils planted with deciduous and coniferous trees had more OTUs than natural soils (F(3,16) = 6.829; p=0.004). Good’s coverage ranged from 99.2 to 99.9% and was significantly lower in reconstructed soils planted with deciduous trees than in natural soils (F(3,16) = 4.45; p=0.02) (Table 4.2).

Table 4.2 Number of sequences, OTU and coverage in soils in the studied soils

Number of Number of Number of Good's Site sequences sequences left OTUs coverage (%) extracted after screening (p=0.004) (p=0.02) Coniferous (mean±sd) 23,609±12,000a 14,798±7,654a 215±39a 99.6±0.28ab C1 12,568 6,819 218 99.3694 C2 40,386 2,3914 258 99.862 C3 11,324 7,071 237 99.2505 C4 27,221 19,846 209 99.7178 C5 26,548 16,338 155 99.8164 Deciduous (mean±sd) 20,832±7,829a 12,138±4,198a 240±58a 99.5±0.22a D1 29,658 16,538 320 99.74 D2 17,633 11,676 172 99.5546 D3 28,816 16,316 260 99.7487 D4 14,605 8,598 252 99.244 D5 13,449 7,564 197 99.3654 Grassland (mean±sd) 32,311±5,275a 21,780±4,840a 202±79ab 99.8±0.05ab G1 30,762 18,447 203 99.9078 G2 38,226 27,641 171 99.8372 G3 30,158 24,750 92 99.8384 G4 36,984 22,518 305 99.8357 G5 25,424 15,546 241 99.762 Natural (mean±sd) 29,327±9,661a 22,614±9,383a 90±44b 99.9±0.07b N1 24126 21,504 55 99.8558 N2 15768 8,444 93 99.775 N3 30266 23,033 143 99.9001 N4 37063 25,604 119 99.9258 N5 39413 34,488 38 99.9623

135 Both coverage values and rarefaction curves (Figure 4.1) suggested that the depth of sequencing was adequate. Rarefaction curves for each samples reached an asymptote at approximately 5,000 sequences.

C1 300 C2 C3 C4 250 C5 C1 D2 200 D3 D4 D5

150 G1 G2

Number OTUs of G3 G4 100 G5 N1 N2 50 N3 N4 N5 0 0 5000 10000 15000 20000 25000 30000 35000 Number of sequences Figure 4.1 Rarefaction curves for reconstructed soils planted with coniferous trees (C1 to C5), reconstructed soils planted with deciduous trees (D1 to D5), reconstructed soils planted with grasses (G1 to G5) and in natural forest soils (N1 to N5)

There were significantly more fungi in the reconstructed soils planted with coniferous and with deciduous trees than in the natural forest soils (F(3,16) = 6.83; p=0.0036) (Figure

4.2A). In both reconstructed and natural soils the most abundant phyla were Ascomycota

> Basidiomycota > Chytridiomycota > Glomeromycota > Rozellomycota (Figure 4.2B).

Fungal class distribution in each treatment is shown in Figure 4.3.

136 250 a 1.00 A p = 0.003564 B a ab 200 0.75

150

0.50 OTUs

100 b

0.25 50 OTU's proportionnal to the abundance of fungi fungi of abundance the to proportionnal OTU's

0 0.00 Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

Fungi phyla:

Ascomycota Basidiomycota Chytridiomycota Glomeromycota Rozellomycota Zygomycota unclassified

Figure 4.2 Number of fungal-phyla OTUs identified (A) and their relative abundance (B) in reconstructed soils planted with coniferous trees, deciduous trees or grasses and in natural forest soils

137 200 a 1.00 A p = 0.003564 B a

150 0.75

OTUs 100 0.50

50 0.25 OTU's proportional to the abudnance the to proportional OTU's fungi of

0 0.00 Coniferous Deciduous Grassland Natural Coniferous Deciduous Grassland Natural

Fungi classes

Archaeorhizomycetes Ascomycota class Incertae sedis Dothideomycetes Eurotiomycetes Leotiomycetes Orbiliomycetes

Pezizomycetes Saccharomycetes Sordariomycetes Agaricomycetes Agaricostilbomycetes Microbotryomycetes

Pucciniomycetes Tremellomycetes Tritirachiomycetes Wallemiomycetes Chytridiomycetes Glomeromycetes Zygomycota class Incertae sedis

138

4.3.3 α-diversity

α-diversity was higher in the reconstructed soils planted with deciduous and coniferous trees than in the reconstructed soils planted with grass or in the natural forest soils (F(3,16)

= 3.39; p=0.044) (Figure 4.4).

a F = 3.39; p = 0.044 20 a (3,16)

10 Inverse Simpson Inverse Index

b b

Coniferous Deciduous Grassland Natural

Figure 4.4 Inverse Simpson index in reconstructed soils planted with coniferous trees, deciduous

trees or grasses and in natural forest soils

Abundance of Populus balsamifera, Betula payrifera, Picea glauca, Alnus virdis spp. crispa and Salix sp. as well as the soil C:N ratio significantly correlated with α-diversity.

All variables were positively correlated with α-diversity except for Betula papyrifera which was negatively correlated with α-diversity (Table 4.3).

139 Table 4.3 Coefficients and signifiance of the multiple linear regression explaining α-diversity in the studied soils (*** : p<0.001; . : p<0.1) Regression model : α-diversity = β1 Populus balsamifera – β2 Betula payrifera + β3 Picea glauca + β4 Alnus virdis spp. crispa + β5 Salix sp. + β6 C :N ratio+ β0+ εi

Variable Coefficients (βi) t value p-value Populus balsamifera 0.30 6.47 2.10x10-5 *** Betula papyrifera -7.78 -7.91 2.52x10-6 *** Picea glauca 0.22 6.68 1.51x10-5 *** Alnus virdis spp. crispa 5.46 7.83 2.82x10-6 *** Salix sp. 2.80 4.69 0.000424 *** C :N ratio 0.26 2.01 0.0657 . Intercept -1.07 -0.47 0.644 F-statistic : 35.41 on 6 and 13 degrees of freedom; Adj. R2 :0.9157; p-value : 2.54x10-7

4.3.4 β-diversity

Soil fungal community structure differed between the reconstructed soils and the natural

forest soils (p<0.05) (Figure 4.5). Fungal community structure also differed between

reconstructed soils planted with deciduous trees, but not coniferous trees, and

reconstructed soils planted with grasses (p<0.05) (Figure 4.5).

0.4

N2 0.2 N3 N4

Archaeorhizomycetes **

Agaricomycetes ** C3 C1 G1 C2 C4 D5 Pezizomycetes * N5 0.0 D1 D4Eurotiomycetes *** WallemiomycetesD3 . Axis 2 Axis G5 D2 Sordariomycetes ** G4Glomeromycetes *** Chytridiomycetes ** C5

G2 N1

-0.4 G3

-0.2

-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 Axis 1 Figure 4.5 Principal Coordinate Analysis (scaling 1) showing fungal classes in reconstructed soils planted with coniferous trees (C1 to C5), deciduous trees (D1 to D5), grasses (G1 to G5) and in natural boreal-forest (N1 to N5) soils. Classes more abundant in reconstructed soils planted to trees are in blue; classes more abundant in reconstructed soils planted to grass are in green, and 140 classes more abundant in naturally disturbed soils are in red (p<0.01: **; p<0.05: *; p<0.1:.) Phylogenetically, the soil fungi communities differed between reconstructed soils and natural soils (p<0.001), as well as between reconstructed soils planted with trees and reconstructed soils planted with grasses (p<0.001) (Figure 4.6).

Tree scale: 0.01

N5 N3 N4 N2 N1 G3 G2 D1 D5 G1 G5 G4 C5 D3 C2 C1 D1 C4 D4 C3

Figure 4.6 Dendrogram showing the dissimilarity of the structure of the fungal community among reconstructed soils planted with coniferous trees (C1 to C5), deciduous trees (D1 to D5), grasses (G1 to G5) and in natural forest soils (N1 to N5) Relative abundance (abundance of each phyla divided by the total abundance of fungi in each treatment) of Ascomycota was greater in reconstructed soils planted with coniferous trees and grass than in natural forest soils (F(3,16) = 4.75; p=0.015). Abundance of

Chytridiomycota was greater in the reconstructed soils planted with grass than in the natural forest soils (F(3,16) = 10.95; p=0.0003). Members of the phylum Basidiomycota were less abundant in reconstructed soils than in natural soils (F(3,16) = 7.56; p=0.002)

(Figure 4.7; Table 4.4). The phylum Glomeromycota was more abundant in reconstructed soils planted with grass species than in reconstructed soils planted with trees and natural soils (F(3,16) = 7.56; p=0.002) (Figure 4.7; Table 4.4). Reclaimed soils also harboured more unclassified sequences than natural soils (F(3,16) = 4.13; p=0.024) (Figure 4.7).

141 Coniferous Deciduous Grassland Natural

Zygomycota

Rozellomycota

Basidiomycota ***

Glomeromycota **

Unclassified *

Chytridiomycota *

Ascomycota *

-2 -1 0 1 2 -2 -1 0 1 2 -2 -1 0 1 2 -2 -1 0 12

142 Within the phyla that were relatively more abundant in reconstructed soils (Ascomycota and Chytridiomycota), the fungal classes Eurotiomycetes, Sordariomycetes and

Chytridiomycetes were more abundant in reconstructed soils, while the classes

Archaeorhizomycetes, Leotiomycetes and Saccharomycetes were more abundant in natural soils (Figure 4.8; Table 4.4; F and p values are presented in Appendix E).

Glomeromycetes – the only identified class of the phyla Glomeromycota – was more abundant in reconstructed soils planted with grasses (Figure 4.8; Table 4.4; F and p values are presented in Appendix E). Within the phylum that was more abundant in natural soils (Basidiomycota), only the class Agaricomycetes was significantly more abundant in natural soils, whereas the class Wallemiomycetes was more abundant in reconstructed soils (Figure 4.9; Table 4.4; F and p values are presented in Appendix E).

Classes Chytridiomycetes, Wallemiomycetes and Glomoromycetes were not found in natural soils whereas class Archaeorhizomycetes was not found in reconstructed soils planted with coniferous or grass species (Table 4.4).

Table 4.4 Phyla (bold) and classes (italics) that are more abundant in either reconstructed soils, natural soils or grassland soils Reconstructed soils planted with Reconstructed soils planted Natural forest soils trees with grass Basidiomycota Ascomycota • Agaricomycetes • Eurotiomycetes • Sordariomycetes

Glomeromycota Archaeorhizomycetes σ Chytridiomycota θ • Glomeromycetesθ Leotiomycetes • Chytridiomycetes θ Saccharomycetes

Wallemiomycetesθ θ σ Absent from natural soils; Absent from reconstructed soils planted with coniferous and grass species;

143 Coniferous Deciduous Grassland Natural

Pezizomycetes

Orbiliomycetes

Dothideomycetes

Saccharomycetes **

Leotiomycetes ***

Archaeorhizomycetes **

Glomeromycetes ***

Chytridiomycetes **

Sordariomycetes **

Eurotiomycetes ***

-2 -1 0 1 2 -2 -1 0 1 2 -2 -1 0 1 2 -2 -1 0 12

Tritirachiomycetes

Tremellomycetes

Pucciniomycetes

Microbotryomycetes

Agaricostilbomycetes

Wallemiomycetes .

Agaricomycetes ***

-2 -1 0 1 2 -2 -1 0 1 2 -2 -1 0 1 2 -2 -1 0 12 Figure 4.9 Z-scores of fungal classes belonging to phyla that were more abundant in natural soils. The first rectangle from the bottom represents the class that is most abundant in natural soils; the second rectangle represents the class that is most abundant in reconstructed soils (p<0.001: ***; p<0.1: .) 145 4.3.5 Relationships between environmental variables and fungal communities

The distribution of fungal phyla in the studied soils was significantly influenced by soil

- NO3 concentrations, % canopy cover and the abundance of the shrubs Cornus canadensis, Alnus virdis spp. crispa and the herb Astragalus (F(5,14) = 8.8751, p = 0.001)

- (Figure 4.10). Soil NO3 concentration and the abundance of Cornus canadensis were strongly correlated with axis 1 and explained 57.25% of the variation in fungal phyla

(p=0.001), while canopy cover and the abundance of Astragalus as well as Alnus virdis spp. crispa were correlated with axis 2 which explained 13.09% of the variation in fungal phyla (p=0.001). As a result, these variables significantly explained 76.02% of the variation of the distribution of fungi at the phylum level. All the explanatory variables were positively correlated with the abundance of fungi, except Cornus canadensis abundance, which was negatively correlated with the abundance of fungi. The abundance of Glomeromycota was positively influenced by the abundance of Astragalus; Phyla

- Chytridiomycota, Zygomycota and Ascomycota were positively related to soil NO3 content and Basidiomycota abundance increased with increasing abundance of Alnus virdis spp. crispa and % canopy cover.

146 2 G1 G5 1

1.5 Astragalus *

1 G3 Glomeromycota G2 N1

0.50 G4 Chytridiomycota N5 - NO3 concentration *** C2 Zygomycota Cornus canadensis *** 0 Ascomycota C5 N2 Rozellomycota C3 N4 D2 D1 D5 -0.5 C1 C4 Basidiomycota D3 Alnus crispa ** N3 -1-2 Canonical axis 2 (13.09% of the variance explained) *** variance the (13.09% 2 of axis Canonical Canopy cover * D4 -1 -2 -1 0 12 Canonical axis 1 (57.25% of the variance explained) ***

At the fungal class level, the effect of vegetation on the distribution of fungi disappeared,

- leaving soil NO3 concentrations, CEC, soil moisture and canopy cover to significantly explain the distribution of fungi in the studied soils (F(4,15) = 3.5708, p = 0.001) (Figure

- 4.11). These variables explained 48.78% of the distribution of fungal classes. Soil NO3 concentrations, CEC, soil moisture and canopy cover were positively correlated with the majority of the fungal classes, except for the class Tritirachiomycetes and

- Archaorhizomycetes, which were negatively correlated with NO3 concentration, CEC and soil moisture, and classes Glomeromycetes and Saccharomycetes which were negatively correlated with canopy cover (Figure 4.11).

147

G5 21

G4 N1 1 G2

G3 Glomeromycetes - NO3 concentration * Moisture * Saccharomycetes N5 CEC *** Chytridiomycetes Sordariomycetes Pucciniomycetes Tremellomycetes Dothideomycetes Wallemiomycetes 0 0 MicrobotryomycetesD2Tr itirachiomycetes Eurotiomycetes Leotiomycetes Archaeorhizomycetes N4 Agaricostilbomycetes D1 Orbiliomycetes N2 C3 C1 D5 Pezizomycetes Agaricomycetes

C5 D4 -1 N3 Canonical axis 2 (8.59% of the variance explained) ** explained) variance the (8.59% 2 of axis Canonical -1 C4 C2 Canopy cover * D3

-3 -2 -1 0 12 Canonical axis 1 (35.56% of the variance explained) ***

4.4 Discussion

4.4.1 Fungal biomass and number of fungal OTUs

Greater total microbial biomass and fungal OTUs were measured in the reconstructed

soils than in natural forest soils. This result was unexpected since previous studies in the

AOSR using denaturing gradient gel electrophoresis (DGGE) and phospholipid fatty acid

148 (PLFA) profiling showed reduced microbial and fungal biomass in reconstructed oil- sands soils ranging in age from 5-35 years compared to both forest-floor and mineral horizons of natural boreal-forest soils (Dimitriu and Grayston, 2010; Dimitriu et al.,

2010; Hahn and Quideau, 2013). This apparent disparity might be due to the fact that previous studies mainly compared the top of the reconstructed soils with the forest floor of natural forest soils, which may exhibit a higher microbial biomass due to their higher

C content. However, Dimitriu and Grayston (2010) did not noted any differences of microbial community structure and number of OTU identified between forest floor and mineral horizons of natural boreal-forest soils. It might be that the use of the more sensitive sequencing techniques in the present study detected more species than do PLFA or DGGE techniques. Moreover, Högberg et al. (2007) measured increasing fungal biomass with increasing C:N ratio in a Fennoscandian boreal forest, and the reconstructed soils had significantly higher C:N ratio than natural soils. It is probable that plants depend more on symbiotic relationships with fungi in less-decomposed litter with high C:N ratio, such as in the reconstructed soils of the present study (Thorn and Lynch, 2007). Soil N content has also been observed to either increase or decrease fungal biomass (Boberg et al., 2008; Frey et al., 2014; Högberg et al., 2007; Sterkenburg et al., 2015). Sterkenburg et al. (2015) have reported that fungal biomass in humus of boreal forest soils of central

Sweden was three times higher in a N-rich litter compared to in a N-poor litter. Addition of ammonium to a Pinus sylvestris litter mixed with the saprophytic fungi Mycena epipterygia resulted in an 32% increase in respiration, 31% increase in needle-litter-mass- loss and increased mycelial production in a laboratory study (Boberg et al., 2008). On the other hand, Frey et al. (2014) noted a 86% reduction in fungal biomass in soils (forest

149 floor and mineral horizons) with high N deposition compared to soils not exposed to N deposition in a temperate forest of Massachusetts (USA). The differing concentrations of

+ - soil N enrichment and the type of inorganic-N – NH4 vs NO3 – being added to the soil could be the cause of these apparent conflicting results. In cases where soils were not subject to N deposition or artificial N addition through fertilization, increased fungal biomass was noticed with increasing soil N content. However, in studies in which soils were subjected to chronic N deposition, sometimes as high as 150 kg N ha-1 year-1, a decline in fungal biomass was observed with increasing N content. A threshold effect may exist, below which higher soil N content induces a higher fungal biomass, but above which adding more N is detrimental to fungal growth. In the current study, N deposition

-1 -1 + -1 -1 has been estimated to be around 15 kg N ha y (14.7 to 19.6 kg NH4 ha year and 2.1

- -1 -1 to 6.7 kg NO3 ha year ) on the reconstructed soils (within the mining footprint)

-1 -1 + -1 -1 - -1 -1 compared to 3 kg N ha y (0.81 kg NH4 ha year and 0.27 kg NO3 ha year ) on

+ natural forest soils (Davis et al., 2015; Fenn et al., 2015; Hemsley, 2012). NH4 deposition around the mining sites are approximately 21 times higher than the

+ - background deposition of NH4 in the area and NO3 deposition, 16 times higher. The level of N enrichment in reconstructed soils, in an otherwise N-limited environment such as in the boreal ecosystem, might be beneficial to fungal growth. The positive correlation

- between NO3 concentrations and fungi in the ordination plot supports the hypothesis that

N enrichment positively influenced fungal biomass in soils.

150 4.4.2 Fungal α-diversity

Contrary to our hypothesis, fungal biodiversity (α-diversity) was higher in reconstructed soils planted with trees than in reconstructed soils planted with grasses and in natural forest soils. Some fungal classes (Chytridiomycetes, Wallemiomycetes and

Glomeromycetes) were only found in reconstructed soils, whereas all fungal classes identified in natural soils were also identified in reconstructed soils. This discrepancy could increase α-diversity in the reconstructed soils compared to the natural forest soils.

The variety of materials used for soil reconstruction in the AOSR (peat and stockpiled mineral horizons) may, once mixed together, create a heterogeneous medium that provides a variety of niches and substrates for fungal communities. However, if the diversity of the material used for soil reconstruction was the main factor explaining fungal diversity, α-diversity would have been also high in reconstructed soils planted with grasses, which was not the case. The phyla Glomeromycota (arbuscular mycorrhizae) was dominant in reconstructed soils planted with grasses as this phyla mainly forms symbiotic relationships with herbaceous plants (Thorn and Lynch, 2007).

Fungi that form arbuscular mycorrhizae are less diverse than those forming ectomycorrhizae; the higher abundance of these fungi in reconstructed soils planted with grasses could explain the lower fungal α-diversity in these soils (Johnson et al., 2005).

Vegetation, and especially the abundance of Populus balsamifera, Betula papyrifera,

Picea glauca, Alnus virdis spp. crispa and Salix sp. was a key variable influencing fungal

α-diversity in the studied soils. The abundance of all tree species positively influenced fungal α-diversity, except for Betula papyrifera which was negatively related to fungal α- diversity. The effect of tree species on fungal diversity can be exert via host specificity,

151 or via leaf litter characteristics. Ectomycorrhizal diversity was positively related to overstory host tree diversity in 12 boreal mixed-wood stands in north-western Québec

(Kernaghan et al., 2003). In an old-growth mixed forest in southeastern Estonia, Bahram et al. (2011) found that a single individual tree (Populus tremula) individual may host at least 200 species of ectomyccorhizae. High host-specificity was also found with alder species (Alnus virdis spp. crispa) which had distinct fungal assemblages compared to black spruce (Picea mariana), trembling aspen (Populus tremuloides) and paper birch

(Betula papyrifera) in Alaska (Bent et al., 2011). Kernaghan and Patriquin (2015) also found that Picea glauca roots supported higher fungal diversity than Abies Balsamea and

Betula papyrifera roots in two boreal soils in Eastern Canada. However, using structural equation modeling techniques, Tedersoo et al. (2014) demonstrated that the correlation between global plant diversity and global fungal diversity was largely explained by their covariance with edaphic conditions.

4.4.3 Fungal β-diversity

Fungal community structure (β-diversity) differed between reconstructed soils and natural forest soils. Fungal classes Eurotiomycetes and Sordariomycetes (phyla

Ascomycota), class Chytridiomycetes (Chytridiomycota) and class Wallemiomycetes

(Basicomycota) were more abundant in reconstructed soils than in natural forest soils. On the other hand, fungal classes Agaricomycetes, Archaeorhizomycetes, Leotiomycetes and

Saccharomycetes were less abundant in reconstructed soils than in natural forest soils.

152 4.4.3.1 Fungal community in reconstructed soils

Species belonging to fungal classes that were more abundant in reconstructed soils have been recognized to be either copiotrophic (Fan et al., 2012), N2O producers (Mothapo et al., 2015), pathogenic (Adams, 1991) or extremophiles (Nguyen et al., 2013). In a long- term fertilization experiment in China, the abundance of the class Eurotiomycetes was shown to increase with increasing N fertilization (Fan et al., 2012). Fan et al. (2012) suggested that members of the Eurotiomycetes, especially Penicillium sp. are copiotrophic organisms – organisms that thrive in nutrient-rich environments – that are able to outcompete other fungal species when nutrient availability is high. In the AOSR, atmospheric N deposition on reconstructed soils has been estimated to be around 15 kg N

-1 -1 + ha y and mainly in NH4 form (Davis et al., 2015; Hemsley, 2012). Moreover, total C

- content, NO3 concentrations and CEC were higher in reconstructed soils. It could be that the nutrient-rich environments in the reconstructed soils favour copiotrophic species from

- the Ascomycota phylum. The positive correlation between CEC, soil NO3 concentration and Ascomycota classes in the ordinations supports this hypothesis.

- Generally higher than in the natural forest soils, NO3 concentrations in reconstructed soils could also favour N2O-producing fungi of the class Eurotiomycetes and

Sordariomycetes. Ninety percent of the known N2O-producing fungi belong to the phylum Ascomycota, among them 46% belong to the class Sordariomycetes and 24% to the class Eurotiomycetes. Fusarium sp. and Trichoderma sp., which form the majority of

Sordariomycetes in the studied soils, dominate N2O-producing fungi (Mothapo et al.,

- 2015). The positive correlation between soil NO3 concentration, Ascomycota and

153 Sordariomycetes observed in the ordinations supports this hypothesis. Because of higher abundance of Sordariomycetes and Eurotiomycetes in the reconstructed soils, it is

- possible that some of this NO3 is being transformed into N2O by denitrifying fungi in these soils.

The vast majority (78%) of the identified Chytridiomycota in the studied soils belonged to the species Olpidium sp. Olpidium is an obligate plant parasite (Powell, 1993). The fungus itself has not been recognized to be directly harmful to plants, but the fungus can transmit virus that can lead to plant diseases (Adams, 1991). Apart from their parasitic roles, Chytridiomycota can also decompose a variety of organic material including chitin, keratin, cellulose and even pollen (Powell, 1993). Even if Chytridiomycota are environmentally ubiquitous, they have only been identified in reconstructed soils in this study and were more abundant in reconstructed soils planted with grasses. At this stage, it is impossible to know if they are actively transmitting viruses to plants in the reconstructed soils, but it is possible that they are diminishing plant resistance against disease.

Geminibasidium sp. was the only species of the Basidiomycota class Wallemiomycetes identified in the studied soils and was only present in the reconstructed soils.

Geminibasidium are soil inhabitant fungi that are xero-tolerant and heat-resistant

(Nguyen et al., 2013). Perhaps because of their extremophile characteristics,

Geminibasidium fungi have preferentially survived stockpiling and the soil reconstruction process, hence their greater abundance in the reconstructed soils.

154

4.4.3.2 Fungal community structure in natural boreal-forest soils

Species from classes more abundant in natural forest soils are either oligotrophic (Rosling et al., 2011; Sterkenburg et al., 2015), ectomycorrhizal or saprophytic (Edwards and Zak,

2010; Riley et al., 2014; Thorn and Lynch, 2007), and also included a leaf-colonizing yeast (Fonseca and Inacio, 2006; Kachalkin and Yurkov, 2012; Sláviková et al., 2007).

Natural forest soils contained more sequences from oligotrophic organisms – slow- growing fungi that compete effectively in nutrient-poor environments – of the

Ascomycetes classes Leotiomycetes and Archaerhyizomycetes. This pattern was also observed within boreal soils from 25 forest sites representing a fertility gradient in central

Sweden in which Leotiomycetes and Archaerhizomycetes classes increased in abundance in sites with low-soil fertility (Sterkenburg et al., 2015). Archaeorhizomycetes are slow- growing fungi that have been relatively recently described by Rosling et al. (2011).

Contrary to most classes of fungi, Leotiomycetes increased in diversity towards the pole, where nitrogen deficiency and soil acidity are greater (Tedersoo et al., 2014). With their oligotrophic growth strategy, the Leotiomycetes and Archaerhizomycetes seems to have a competitive advantage over the copiotrophic fungi in nutrient-poor and acidic environments hence their greater abundance in the natural forest soils of the present

- study. The negative correlation between soil NO3 concentration, CEC and

Archaeorhizomycetes as well as Leotiomycetes in the ordination plot supports this hypothesis.

155 With over 22,000 taxa, the Agaricomycetes are the largest class of Basidiomycota (Riley et al., 2014; Thorn and Lynch, 2007). They comprise most of the wood- and leaf- decomposers as well as ectomycorrhizal species and, therefore, provide key ecosystem functions (Edwards and Zak, 2010; Lindahl et al., 2007; Riley et al., 2014; Roy-Bolduc et al., 2015; Thorn and Lynch, 2007). These fungi were more abundant in the natural soils than in the reconstructed soils. Using DGGE and PLFA, Dimitriu et al (2010) also found that ectomycorrhizal Basidiomycota dominated natural forest soils, whilst Ascomycota sequences dominated reconstructed soils in the AOSR. Edwards and Zak (2010) showed that certain species of Agaricomycetes were exclusive to forest-floor horizons

(Sistotrema, Mycena and Clitocybe); while some species (Cortinarius, Russula,

Thelephora, Piloderma, Trechispora, Collybia/Gymnopus and Clavariaceae) were exclusive to mineral soils in upland forest ecosystems in northwestern Lower Michigan.

They attributed this stratification to a shift in fungi from saprotrophic lifestyles to mycorrhizal lifestyle in the mineral soils in which fine roots thrive. Lindahl et al. (2007) observed the same shift of fungal communities between forest floor and mineral-soil horizons in a Pinus sylvestris forest of central Sweden. This stratification is observable in the present study as species Cortinarius and Clavariacea dominated Agaricomycetes species in the mineral horizons of the studied soils. The strong positive correlation between the abundance of Agaricomycetes and canopy cover further supports the mycorrhizal nature of the fungi present in the natural forest soils.

Yeast are normally well adapted for aqueous environments hence their positive correlation with moisture within the studied soils. Yeast can also be found in soils in

156 which they can decompose plant tissues (Thorn and Lynch, 2007). The Ascomycota

Saccharomycetes have also been found on plant leaves (Fonseca and Inacio, 2006;

Sláviková et al., 2007) and on sphagnum (Kachalkin and Yurkov, 2012). Their higher abundance in the mineral horizons of the natural forest soils in this study might be related to the presence of forest-floor material and sphagnum on top of these mineral soil horizons.

4.4.4 Fungal communities in reconstructed soils planted with grasses

Fungal communities in reconstructed soils planted with grasses were distinct from fungal communities in reconstructed soils planted with trees and were also more similar to the fungal community of natural forest soils. The higher abundance of arbuscular mycorrhizae (of phyla Glomeromycota), which forms mycorrhizae mostly with herbaceous species, in reconstructed soils planted with grasses probably drives the differences in fungal communities between these soils and the reconstructed soils planted with trees. The positive correlation of Astragalus with Glomeromycota in the ordination plot supports this hypothesis. It could also be hypothesized that the Glomeromycota outcompete other fungi “endemic” to reconstructed soils such as the Chytridiomycetes and Wallemiomycetes, placing the fungal community inhabiting reconstructed soils planted to grasses closer to fungal communities found in natural boreal-forest soils.

4.4.5 Litter-degrading capabilities of fungi

Basidiomycetes are generally assumed to be better litter decomposers than Ascomycota as most of them possess both cellulosic and lignin decomposing capabilities, whereas the

157 Ascomycota are mainly restricted to cellulose degradation (Boberg et al., 2011; Osono and Takeda, 2002). However, among the Ascomycota, the order Xylariales from the class

Sordariomycetes have been shown to possess lignin-decomposing capabilities (Osono and Takeda, 2002). In a cool, temperate deciduous forest near the city of Kyoto (Japan),

Osono and Takeda (2001) found that beech leaf litter decomposed in two phases ⎯ the immobilization phase followed by the mobilization phase. During the immobilization phase, the C:N ratio of the litter varied between 55-26, favouring the immobilization of nutrients (i.e. N and P) in recalcitrant organic matter. The high availability of holocellulose material during this phase lead to rapid fungal growth which was dominated by the Ascomycota. Decomposition of lignin was then slow and carried only by Xylariales species, which preferably decomposed holocellulose material over lignin.

By decomposing holocellulose material, the authors suggested that Xylariales changed the chemistry of the litter, decreasing the C:N ratio and creating an environment that favours Basidiomycota. The decomposition process then entered in the second phase, the mobilization phase during which nutrients were mobilized from recalcitrant material and lignin via the Basidiomycota activity (Osono and Takeda, 2001).

The reconstructed soils, being mainly colonized by Ascomycota, may have lower decomposition capability for recalcitrant material and lignin compounds, which would keep the reconstructed soils in an immobilization phase. Dimitriu et al. (2010) noted a lower phenol-oxidase activity (a key enzyme in lignin decomposition), and fewer products of lignin degradation in reconstructed soils compared with natural forest soils of the AOSR. This suggested that microbial communities most abundant in reconstructed

158 soils might be limited in their capacity to degrade organic matter. The high ammonification rates from the recalcitrant organic-N pool observed in reconstructed soils

(Chapter 2) could be due to an “adsorption-desorption” process rather that a

“mineralization-immobilization” process. In contrast, the fungal community of natural forest soils is dominated by Basidiomycota, especially of the class Agaricomycetes.

These fungi are able to degrade more complex organic matter and lignin materials and, therefore, place the natural soils in the mobilization phase of decomposition. The lower phenol-oxidase activity in reconstructed soils could also be attributable to the communities colonizing reconstructed soils being adapted to peat material with low content of lignin material. With soil development and greater inputs of lignin material from decomposing litter, the fungal communities in reconstructed soils could potentially shift towards communities with greater ability to degrade lignin materials. The reconstructed soils could also eventually move to the mobilization phase of decomposition as Ascomycota decompose holocellulose materials, thereby changing soil conditions and allowing colonization by Basidiomycota. However, the high N deposition and high nutrient content of the reconstructed soils could continuously favour copiotrophic fungi of the phylum Ascomycota, therefore, impeding the decomposition of recalcitrant material in these soils. Sinsabaugh (2010) noted that in northern temperate forests, phenol oxidase expression by ectomycorrhizae declined with higher soil N content. This reduced phenol oxidase activity may, in turn, increase the concentration of refractory carbon and reduce rates of decomposition. It would be interesting to revisit the reclaimed sites in few years to assess if the reconstructed soils of the AOSR have proceeded to the mobilization phase of decomposition or if the high N-deposition of the

159 area is maintaining fungal communities with potentially lower competency to degrade recalcitrant organic carbon.

4.5 Conclusion

The use of massively parallel sequencing techniques to analyse fungal species in reconstructed oil-sand soils and in natural boreal-forest soils provided a comprehensive assemblage of fungal species colonizing these soils. Understanding key differences in fungal communities inhabiting reconstructed and natural soils allowed us to delve into their ecological capabilities. Contrary to our hypothesis, fungal OTUs and α-diversity was higher in reconstructed soils planted with trees than in reconstructed soils planted with grass or natural forest soils. Vegetation cover and especially the abundance of

Populus balsamifera, Betula payrifera, Picea glauca, Alnus virdis spp. crispa and Salix sp, possibly via their host-specificity mechanisms or via leaf-litter characteristics, were the best predictors of α-diversity in the studied soils. β-diversity differed among all treatments except reconstructed soils planted with deciduous or coniferous tree species.

Ectomycorrhizal- and saptrotrophic- Agaricomycetes species, together with oligotrophic- archaeorhizomycetes and leotiomycetes species, were relatively less abundant in reconstructed soils than in natural forest soils. Putative denitrifiers fungal species of the classes eurotimocytes and sordiomycetes and the plant-parasitic class chytridiomycetes were more abundant in reconstructed soils planted with coniferous and deciduous species than in natural forest soils. Constructed soils planted with grass had more arbuscular mycorrhizal species of the phylum Glomeromycota than any other studied soils. Soil

- NO3 content and CEC were the main drivers explaining differences between fungal

160 communities in reconstructed and natural soils, whereas plant species and canopy cover influenced differences among reconstructed soils planted with trees and reconstructed soils planted with grass. Taken together these results indicate that key soil and site

- characteristics such as NO3 content, CEC and vegetation cover resulted in contrasting fungal communities in the reconstructed and natural forest soils.

161 Chapter 5: Conclusion

The overall objective of this thesis was to measure, compare and understand processes underlying nitrogen cycling rates and microbial communities in reconstructed oil-sands soils and in natural boreal-forest soils. The use of 15N tracer methods in combination with massively parallel sequencing techniques allowed me to successfully reach the objectives of this study and, more importantly, to better understand the pedological environments of the reconstructed soils and their dissimilarities with natural boreal-forest soils.

It was already known that oil-sands reconstructed soils were structurally different than natural boreal-forest soils of northern Alberta. Turcotte et al. (2009) showed that organic- matter composition was different in reconstructed soils compared to natural forest soils.

Furthermore, atmospheric N deposition rates on oils-sands-reconstructed soils were approximately three times higher than on natural boreal-forest soils of northern Alberta

(Davis et al., 2015; Hemsley, 2012). Also Hemsley (2012) have shown that the majority

- of the soil available N in the forest floor of reconstructed soils is in the inorganic NO3

+ form; whereas, most soil available N in the boreal forest is in organic and NH4 forms ().

MacKenzie and Quideau (2012) also found higher rates of nitrogen mineralization in the reclaimed material used for soil reconstruction than in the boreal forest soils, suggesting that the rates of both ammonification and nitrification are higher in the reconstructed soils. Differences in soil microbial communities have also been reported between reconstructed and natural boreal-forest soils. Using denaturing gradient gel electrophoresis (DGGE) and phospholipid fatty acid (PLFA) profiling, reduced abundance and distinct microbial community structure were found in reconstructed oil-

162 sands soils ranging in age from 5-35 years compared with natural boreal-forest soils

(Dimitriu et al., 2010; Hahn and Quideau, 2013). These differences in nutrient availability and microbial communities between reconstructed soils and natural forest soils indicate that it is unlikely that these reconstructed forest soils will exactly mirror natural boreal-forest soils (Chazdon, 2008; Hobbs et al., 2006). Thus, novel soil ecosystems will probably result from the reconstruction and reclamation efforts. In this context, re-establishing ecological function, rather than replicating structural qualities of the previous soil ecosystem, is the most reasonable goal of reclamation and will assist in ensuring the long-term sustainability of reclaimed boreal-forest landscapes (Quideau et al., 2013).

The contributions of this thesis to the field of ecosystem restoration lie in providing a simultaneous analysis of the soil nitrogen cycle, including an assessment of gross rates of

+ - ammonification and nitrification, as well as of NH4 and NO3 immobilization and of microbial communities carrying out these functions, which deepens our understanding of soil functioning in both reconstructed and natural forest soils. A fundamental understanding of how soils function in reconstructed soils is necessary to guide the long- term sustainability of these reclaimed boreal-forest landscapes. Fundamentally, the thesis also sheds light on the interactions between above- and below-ground communities and their impacts on soil nitrogen cycle in contrasting environments. Table 5.1 summarizes the main findings of this research project.

163 Table 5.1 Main findings and processes identified in the thesis Objective 1. Assess if soil N-transformation rates in oil-sands soils that were reconstructed 20-30 years previously are similar to those of natural boreal-forest soils that were subject to wildfire disturbance at approximately the same time. Main findings Hypotheses Processes + - In the reconstructed soils, NH4 was - Reconstructed soils mineralized and mainly cycled through the recalcitrant + organic-N pool immobilized more NH4 from and to + the recalcitrant organic-N pool - In the natural forest soils, NH4 was mainly cycled through the labile - Natural forest soils mineralized more H1.1: N content and gross rates of N NH + from the labile organic-N pool organic-N pool, suggesting greater 4 transformations will be higher in prominence of microbial N-cycling - Reconstructed and natural forest soils reconstructed soils than natural soils. activity in the natural soils despite immobilized NH + into the labile 4 Verdict: True, except for gross rates of having a lower microbial biomass than organic-N at the same rate mineralization from labile organic N-pool, the reconstructed soils. Shorter - No differences in NO - transformation 3 which was higher in natural forest soils turnover rates in natural forest soils rates were found between reconstructed than in reconstructed soils support this prominence of microbial and natural forest soils N-cycling activity in the natural soils - Net nitrification rates were positive in - Positive nitrification rates in reconstructed soils, but null in natural H1.2: Gross ammonification and reconstructed soils suggest that forest soils nitrification rates will be highest in microorganisms are limited by other - Autotrophic nitrification was negligible reconstructed soils planted to deciduous nutrients than N in these soils. in both reconstructed and natural forest - Net production of NO - (net soils. trees 3 nitrification) rates in reconstructed soils - Turnover rates of both NH + and NO - Verdict: Not true. Vegetation treatment did 4 3 not have any impacts on gross rates of N- has ecological implications because were longer in reconstructed soils than NO - is less stable in soils in natural forest soils transformation in reconstructed soils. 3 - Characteristics of the material used - Within the reconstructed soils, during soil reconstruction are still vegetation treatments did not influence overpowering any influences of N-transformation rates vegetation on N-cycling rates in reconstructed soils

164 Table 5.1. cont’ Objective 2: Assess if soil prokaryotic diversity and structure (i.e. α- and β- diversity) in oil-sands soils reconstructed 20–30 years previously were similar to those found in natural boreal-forest soils. Relationships between prokaryotic α- and β- diversity and above- and below-ground characteristics will also be determined. Main findings Hypotheses Processes

- α-diversity did not differ in - Copiotrophic bacteria were more abundant H2.1: α-diversity will be higher in natural reconstructed and natural soils and was in reconstructed soils soils actually higher in reconstructed soils - Oligotrophic bacteria were more abundant Verdict: False. α-diversity did not differ in natural soils - The abundance of Pinus banksiana, significantly between reconstructed and - This copiotrophic–oligotrophic shift Medicago staiva, Salix bebbina and natural soils, but was generally higher in between reconstructed and natural forest Vaccinium mytrilloides were the main reconstructed soils soils was mainly related to N deposition factors influencing α-diversity rates, suggesting that the availability of a - β-diversity differed between new and readily available source of N reconstructed and natural forest soils through atmospheric deposition was the H2.2.: β-diversity will differ between - Actinobacteria, Bacterioridetes and most important factor shaping microbial reconstructed and natural soils communities in the reconstructed soils. Proteobacteria were more abundant in Verdict: True. reconstructed soils - Copiotrophic bacteria have lower ability to degrade recalcitrant organic matter. Their - Acidobacteria, Cyanobacteria, higher abundance in reconstructed soils can Elusmicrbia, Firmicutes, H2.3.: pH and soil moisture will be the main therefore strongly influence carbon and Planctomycetes and Verrumicrobia drivers of prokaryotic α- and β- diversity nitrogen cycles and further drive were more abundant in natural forest Verdict: Mainly false. differences between reconstructed and soils - α-diversity was largely explained by natural soils. - Thaumarchaeota and anaerobic- the abundance of Pinus banksiana, - Grassland soils favoured anoxic bacteria Chloroflexi classes were more Medicago staiva, Salix bebbina and and ammonia-oxidizing archaeal prokaryotic species, highlighting the abundant in reconstructed soils planted Vaccinium mytrilloides with grass than any other soils. uniqueness of their soil environments. The - β-diversity differences between - Nitrogen deposition and pH were the higher abundance of AOA in the grassland reconstructed and natural forest soils soils may also explain the higher main factors influencing the were mainly influenced by N nitrification rates in these soils and may distribution of microbial community in deposition, pH and vegetation cover. lead to greater losses of N through leaching reconstructed and natural soils or gaseous emissions.

165 Table 5.1 cont’ Objective 3: To assess if soil fungal diversity and structure (i.e. α- and β- diversity) in oil-sands soils reconstructed 20-30 years previously were similar to those in natural boreal-forest soils. Relationships between fungal α- and β- diversity and above- and below- ground characteristics will also be determined Main findings Hypotheses Processes H3.1: α-diversity will be higher in natural - α-diversity was higher in reconstructed - Fungal α-diversity was mainly soils; soils planted with trees than in influenced by vegetation (via host- Verdict: False. α-diversity was higher in reconstructed soils planted with grass specificity or leaf-litter characteristics), reconstructed soils planted with trees or natural forest soils highlighting the importance of above-

- the abundance of Populus balsamifera, ground diversity for below-ground H3.2: β-diversity will differ between Betula payrifera, Picea glauca, Alnus diversity in both reconstructed and reconstructed and natural soils; virdis spp. crispa and Salix sp were the natural forest soils Verdict: True main factors influencing α-diversity - As with bacterial community, a shift - β-diversity differed among all from copiotrophic to oligotrophic H3.3 Basidiomycota fungal species will be treatments except reconstructed soils fungal communities was observed in more abundant in natural soils, Ascomycota planted with deciduous or coniferous reconstructed and natural forest soils and Zygomycota fungal species will be more tree species abundant in reconstructed soils, and - Fungal communities in reconstructed - Eurotimocytes and sordiomycetes and Glomeromycota will be more abundant in soils have limited organic-matter chytridiomycetes were more abundant reconstructed soils planted with grass communities and therefore rely on a in reconstructed soils Verdict: Partially true constant supply of readily available C - Agaricomycetes, archaeorhizomycetes and nutrients to meet their nutritional - The phylum Basidiomycota was more and leotiomycetes were more abundant needs as highlighted by their strong abundant in natural soils, but driven by in natural forest soils correlation with soil NO - content and the class Agaricomycetes. Some 3 CEC; this may lead to accumulation of - Glomeromycota were more abundant in Ascomycota (Leotiomycetes and reconstructed soils planted with grass recalcitrant C in reconstructed soils. - archaeorhizomycetes) were more - NO3 content and CEC were the main abundant in natural soils - The presence of putative denitrifier- fungal communities, coupled with the drivers of differences between fungal - Only the phylum Ascomycota was higher content of NO - in reconstructed communities in reconstructed and more abundant in reconstructed soils. 3 soils, indicates that there might be natural soils, whereas plant species and - Glomeromycota were more abundant in emission of N O from these soils, canopy cover influenced differences reconstructed soils planted with grasses 2

166 among reconstructed soils planted with H3.4. Reconstructed soils planted with which would increase nitrogen trees and reconstructed soils planted coniferous trees will have greater fungal deposition in the area. with grass abundance compared to reconstructed soils planted with deciduous trees or grass Verdict: Partially true. More fungal OTUs were identified in reconstructed soils planted with coniferous trees than in reconstructed soils planted with grass, but not compared to reconstructed soils planted with deciduous trees

H3.5. pH and nutrient content of the soils will be the main drivers of fungal α- and β- diversity - Verdict: Partially true. NO3 content and CEC were the main drivers of differences between fungal communities in reconstructed and natural soils

167 5.1 Main findings

5.1.1 Role of nitrogen deposition in shaping microbial communities and nitrogen cycling in reconstructed soils

The results presented in this thesis highlight the key role of nitrogen deposition in shaping microbial communities and nitrogen cycling rates in the reconstructed and natural forest soils as pointed by the ordination plots. In the AOSR, nitrogen deposition represents a constant source of readily available N that seems to favour the growth of copiotrophic bacterial and fungal species

(Proteobacteria, Actinobacteria, Bacterioridetes and Eurotiomycetes). However, according to the literature, these microorganisms are limited in their capabilities to degrade complex organic matter, which could limit decomposition processes in the reconstructed soils and favour C accumulation. The high ammonification rates from the recalcitrant organic-N pool observed in reconstructed soils could then be mostly due to an “adsorption-desorption” process rather that a

“mineralization-immobilization” process. Ammonification rates from the labile organic-N are also lower in reconstructed soils than in natural forest soils although reconstructed soils had greater microbial biomass. This further suggests that microbes colonizing reconstructed soils have a low competency to degrade organic matter. As forest floors develop on the reconstructed soils, and roots die and decompose, it is likely that more complex organic matter will be incorporated into the mineral soils. It could be that this new organic matter will trigger changes the composition of the microbial community in reconstructed soils and favour community that are able to degrade this complex organic matter. However, if nitrogen deposition remains high in the AORS, it might also continuously favour microorganisms with reduced abilities to degrade

168 more complex organic matter in reconstructed soils, thus favouring organic matter accumulation and diminishing soil functioning in reconstructed soils.

5.1.2 Nitrate cycle and its fate in reconstructed soils

- Because soil exchange sites are negatively charged, NO3 is more mobile in soils and could be lost from the system by denitrification or leaching processes (Thorn and Lynch, 2007). These losses could have detrimental impacts on environments surrounding the reclaimed landscapes.

Among the gases released through denitrification, N2O is a greenhouse gas with a global

- warming potential 300 times higher than CO2 (EPA, 2013). NO3 can also cause eutrophication of surrounding bodies of water with resultant impacts on water quality and survival of aquatic

- species. Mean NO3 content of the reconstructed soils were 2 to 3 times higher than NO3-

- -2 -1 content of the natural boreal-forest soils. Measured at an average rate of 100 mg NO3 m day , net nitrification rates were positive and higher than in natural soils. Moreover, higher abundances of AOA archaea and of denitrifying fungi were found in the reconstructed soils than in the natural soils. Taken together these results indicate that the nitrogen cycle in reconstructed soils is in the “N-rich environment” part of the conceptual model used in this thesis (Figure 1.3). The higher N content of reconstructed soils coupled with the high N deposition in the area could favour AOA archaeal species, which will, in turn, increase nitrification rates. The higher abundance of denitrifier fungi in reconstructed soils is concerning as these organisms can enhance greenhouse gas emissions in the AOSR and increase nitrogen deposition, thereby creating a positive feedback loop between soil N content, microorganisms and gaseous

- emissions, as well as NO3 leaching from the reconstructed soils.

169 The higher rates of nitrification in the reconstructed soils might also indicate that the microbial communities colonizing these soils are limited by available carbon or other nutrients, rather than

N, which would cause them to keep decomposing the organic matter even if their N-needs are fulfilled, further stimulating release of inorganic N.

5.1.3 Alpha-diversity and soil functions

Contrary to our hypotheses, α-diversity was usually higher in reconstructed soils than in natural forest soils. In the Introduction, I presented a model in which higher microbial diversity will increase the chances that biologically-narrow processes will continue. The results of the thesis showed that even if α-diversity is high, it does not ensure that key soil functions will be retained.

As such, the identity of microbial species colonizing the soils is a better predictor of soil functioning. For example, fungal α-diversity was higher in reconstructed soils than in natural forest soils, but the fungal species driving this higher biodiversity were possible parasites, N2O producers or extremophiles that have a limited impact on fundamental processes such as decomposition. In contrast, the natural forest soils had a lower α-diversity, but the main fungal species found in these soils was the Agaricomycetes which are saprotrophic and ectomycorrhizal, and so have fundamental roles in soil functioning. These results indicate that higher soil biodiversity (or α-diversity) does not necessarily translate into greater soil functioning capabilities. Soil biodiversity can indeed increase the probability finding keystone species in a soil, but characterization of the soil microbial community proved to be more meaningful in identifying and understanding soil functions. A further step would be to characterize the active soil communities through RNA or proteomics studies to assess which microbial community are actively carrying out functions.

170

5.1.4 Role of vegetation on nitrogen cycle processes and communities

The influence of vegetation on soil nitrogen cycling and communities was assessed in this research in two main ways. First, I assessed if vegetation treatment had an impact on nitrogen cycling rates and communities in reconstructed soils by comparing reconstructed soils planted with deciduous trees, coniferous trees or grasses. Second, I used the vegetation survey of

Anderson (2014) on the same sites to link above-ground plant cover with soil processes.

Contrary to my hypothesis, vegetation treatment did not have a significant influence on nitrogen cycling rates in reconstructed soils. This result is consistent with afforestation studies which have shown that vegetation effects on various mineral soil properties (pH, humic acid characteristics, carbon and nitrogen contents) become measurable 20 to 50 years after afforestation, when litter inputs rates have been restored and organic matter is being redistributed through the soil profile

(Compton et al., 2007; Abakumov et al., 2010; Bárcena et al., 2014). The reconstructed oil-sands soils that I studied were planted less than 30 years previously, but some differences in C content in the upper 10 cm of mineral soil have been observed among these vegetation treatments

(Anderson, 2014). These differences may eventually lead to variation in rates of soil N-cycling processes among vegetation types, but at this point in time it appears that the effect of the residual peat is overwhelming any influence of vegetation on N-cycling rates in reconstructed soils. However, vegetation treatment did have a significant influence on microbial communities, as reconstructed soils planted with grasses had different prokaryotic and fungal communities than natural forest soils. Reconstructed soils planted with grasses had higher abundances of anoxic bacteria, ammonia-oxidizing archaea and Glomeromycota, highlighting the uniqueness of

171 the below-ground environment in these reconstructed soils. The higher abundance of

Glomeromycota can be easily explained by the fact that arbuscular mycorrhizae (which are mainly from the fungal phylum Glomeromycota) form the majority of their symbiotic relationships with herbaceous species, hence their higher abundance in grassland soils (Thorn and Lynch, 2007). However, the higher abundance of ammonia-oxidizing archaea can lead to higher nitrification rates in the grassland soils, with potential loss of N from the soil system.

Vegetation cover was the main factor explaining α-diversity in the studied soils, emphasizing the importance of above- and below-ground interactions in natural and reconstructed ecosystems. It was also an important factor explaining the fine distribution of microorganisms among soils (β- diversity) as highlighted in the ordination plots. It was not possible to distinguish between coniferous or deciduous species effects on soil microbial communities, but this was not one of the objectives of the study. However, this research indicates a possible important role of understorey plant species in influencing microbial communities, as they were the main type of vegetation to have an impact on either α- or β- diversity.

5.2 Contributions

5.2.1 Practical contributions

The results presented in this thesis further confirm that reconstructed soils are a novel soil ecosystem distinct from the natural boreal-forest soils of northern Alberta. It was already known that nitrogen cycling rates and microbial communities differed in the two environments. The methods used in this thesis allowed for a more fundamental understanding of key differences in reconstructed and natural soils. By being able to estimate meaningful gross rates of

172 transformations, it was determined that in the reconstructed soils nitrogen cycles mainly through recalcitrant organic-N with limited microbial N-cycling activity despite higher microbial biomass. The use of massively parallel sequencing techniques also expanded our understanding of how microbial communities differ between reconstructed oil-sands soils and natural boreal- forest soils.

The findings of this thesis can be used to effectively manage the reclaimed ecosystems and aid in the development of soil reconstruction strategies that might reduce differences between reconstructed and natural forest soils and thereby assist in re-creating soils with pre-disturbance land capabilities.

5.2.2 Fundamental contributions

The objectives of this thesis were mainly practical. However, the thesis also contributes to fundamental knowledge in the science of forest ecology. The influence of nitrogen deposition on specific microbial communities has been previously studied both in laboratory settings and in the field, with contrasting results, possibly due to a threshold effect. The results of the present study positively contribute to this body of knowledge since it has detected a shift from copiotrophic to oligotrophic communities over a relatively short gradient of both distance and nitrogen deposition rates. Interestingly, this shift was associated with distinct soil functions in copiotrophic and oligotrophic environments.

The results of this thesis also highlighted the importance of above-ground and below-ground interactions in forest ecosystems, including reconstructed ecosystems. The understory species

173 explained more of the variation in either prokaryotic or fungal species distribution than any other type of vegetation, highlighting their important role in both natural and reconstructed environments.

5.3 Future research avenues

Building on the results of this thesis, I suggest looking further into active soil processes at play in reconstructed soils, to delve deeper into their ecological competencies. The use of proteomics techniques would allow for a characterization of active processes taking place in the reconstructed soils and test the conclusions about soil functioning identified in this study drawn on the structure of the communities.

- The cycling and fate of NO3 is also worth investigating further, as high nitrification rates and

- greater abundance of microbial communities contributing to the NO3 cycle have been identified in the reconstructed soils. This could have detrimental effects on ecosystems surrounding reclaimed landscapes and should be investigated further.

As an element fundamental to ecosystem health, nitrogen has been accordingly studied in the oil- sands reconstructed soils. However, high rates of nitrification coupled with low rates of turnover of inorganic-N in reconstructed soils might indicate that communities colonizing these soils are limited by either available-C or nutrients other than N. Investigating the factors limiting microbial growth in the reconstructed soils might be helpful to ensure long-term sustainability of these ecosystems.

174 Fundamentally, one of the peculiar results of this research was the identification of a link between the bryophyte Polytrichum juniperinum, and Planctomycetes – an annamox-capable bacterial phyla. It was already known that Cyanobacteria were preferably colonizing bryophyte species, thereby making bryophytes hot-spots for N-fixation. The positive correlation between

Planctomycetes and Polytrichum juniperinum in the present study might suggest that the moss– bacteria association is contributing to the nitrogen cycle beyond nitrogen fixation – an intriguing possibility worth investigating further.

In the discussion of the third chapter, it became apparent that knowledge about functions carried by individual fungal species is substantially less than our knowledge of functions carried by bacterial species. In forest ecosystems, fungi play a central role in ecosystem functioning and are arguably more important than bacteria, yet we know less about their individual abilities than we do for bacteria. There also seems to be a dichotomist vision of fungal functions, divided between mycorrhizal and saprotrophic species. It is possible that single fungal species carry genes enabling it to perform both functions (like ammonia-oxidizing bacteria), and to perform other tasks. New cultivation and whole-genome sequencing techniques might help to fill gaps in our understanding of fungal species and their ecological functions in forest soils.

As a whole this thesis made a contribution to both practical and fundamental science. The results deepen our understanding of the functioning of oil-sands reconstructed soils and identify areas that need further investigation. This thesis also contributed to the current body of knowledge of how forest soils function and how interactions between vegetation, soil physical and chemical characteristics and microbial communities shape soil functions.

175

The oil sands are Canada’s environmental legacy. I hope that the research effort, to which I am contributing through this thesis, will ensure that this legacy is one of successful restoration of self-sustaining ecosystems that provide ecological services at least equal to those of the boreal forest prior to the exploitation of the oil sands.

176 References

Abakumov, E., Trubetskoj, O., Demin, D., Celi, L., Cerli, C., Trubetskaya, O., 2010. Humic acid

characteristics in podzol soil chronosequence. Chemistry and Ecology 26, 59–66.

doi:10.1080/02757540.2010.497758

Ackermann, K., Zackrisson, O., Rousk, J., Jones, D.L., DeLuca, T.H., 2012. N2 Fixation in

Feather Mosses is a Sensitive Indicator of N Deposition in Boreal Forests. Ecosystems 15,

986–998. doi:10.1007/s10021-012-9562-y

Adams, M.J., 1991. Transmission of plant viruses by fungi. Annals of Applied Biology 118,

479–492.

Amaranthus, M.P., Perry, D.A., 1987. Effect of soil transfert on ectomycorrhiza formation and

the survival and growth of conifer seedlings on old, nonreforested clear-cuts. Canadian

Journal of Forest Research 17, 944–950. doi:10.1139/x87-147

Anderson, J., 2014. Organic matter accumulation in reclaimed soils beneath different vegetation

types in the athabasca oil sands. University of British Columbia. doi:10.1007/s13398-014-

0173-7.2

Angelov, A., Liebl, W., 2006. Insights into extreme thermoacidophily based on genome analysis

of Picrophilus torridus and other thermoacidophilic archaea. Journal of Biotechnology 126,

3–10. doi:10.1016/j.jbiotec.2006.02.017

Augusto, L., De Schrijver, A., Vesterdal, L., Smolander, A., Prescott, C., Ranger, J., 2015.

Influences of evergreen gymnosperm and deciduous angiosperm tree species on the

functioning of temperate and boreal forests. Biological Reviews 90, 444–466.

doi:10.1111/brv.12119

177 Bahram, M., Polme, S., Koljalg, U., Tedersoo, L., 2011. A single European aspen (Populus

tremula) tree individual may potentially harbour dozens of Cenococcum geophilum ITS

genotypes and hundreds of species of ectomycorrhizal fungi. FEMS Microbiology Ecology

75, 313–320. doi:10.1111/j.1574-6941.2010.01000.x

Ball, P.N., MacKenzie, M.D., DeLuca, T.H., Montana, W.E.H., 2010. Wildfire and charcoal

enhance nitrification and ammonium-oxidizing bacterial abundance in dry montane forest

soils. Journal of Environment Quality 39, 1243. doi:10.2134/jeq2009.0082

Bárcena, T.G., Kiaer, L.P., Vesterdal, L., Stefánsdóttir, H.M., Gundersen, P., Sigurdsson, B.D.,

2014. Soil carbon stock change following afforestation in Northern Europe: a meta-analysis.

Global Change Biology 20, 2393–2405. doi:10.1111/gcb.12576

Bell, T.H., Cloutier-Hurteau, B., Al-Otaibi, F., Turmel, M.-C., Yergeau, E., Courchesne, F., St-

Arnaud, M., 2015. Early rhizosphere microbiome composition is related to the growth and

Zn uptake of willows introduced to a former landfill. Environmental Microbiology 17, n/a–

n/a. doi:10.1111/1462-2920.12900

Bell, T.H., El-Din Hassan, S., Lauron-Moreau, A., Al-Otaibi, F., Hijri, M., Yergeau, E., St-

Arnaud, M., 2014. Linkage between bacterial and fungal rhizosphere communities in

hydrocarbon-contaminated soils is related to plant phylogeny. Isme J 8, 331–343.

doi:10.1038/ismej.2013.149

Bent, E., Kiekel, P., Brenton, R., Taylor, D.L., 2011. Root-associated ectomycorrhizal fungi

shared by various boreal forest seedlings naturally regenerating after a fire in interior

Alaska and correlation of different fungi with host growth responses. Applied and

Environmental Microbiology 77, 3351–3359. doi:10.1128/AEM.02575-10

Bergeron, Y., Gauthier, S., Kafka, V., Lefort, P., Lesieur, D., 2001. Natural fire frequency for the

178 eastern Canadian boreal forest: consequences for sustainable forestry. Canadian Journal of

Forest Research 31, 384–391. doi:10.1139/x00-178

Binkley, D., Aber, J., Pastor, J., Nadelhoffer, K., 1986. Nitrogen availability in some Wisconsin

forests: comparisons of resin bags and on-site incubations. Biology and Fertility of Soils 2,

77–82. doi:10.1007/BF00257583

Binkley, D., Fisher, R.F., 2013. Ecology and Management of Forest Soils, Fourth Edition. West

Sussex.

Blackwell, M., 2011. The fungi: 1, 2, 3 ... 5.1 million species? American Journal of Botany 98,

426–438. doi:10.3732/ajb.1000298

Blair, S.M., 1950. Report on the Alberta Bituminous Sands, Alberta Se. ed. Toronto.

Blake, G.R., Hartge, K.H., 1986. Bulk density, in: Klute, A. (Ed.), Method of Soil Analysis. Part

1: Physical and Mineralogical Methods. ASA-SSSA, Madison, pp. 363–375.

Boberg, J., Finlay, R.D., Stenlid, J., Näsholm, T., Lindahl, B.D., 2008. Glucose and ammonium

additions affect needle decomposition and carbon allocation by the litter degrading fungus

Mycena epipterygia. Soil Biology and Biochemistry 40, 995–999.

doi:10.1016/j.soilbio.2007.11.005

Boberg, J.B., Ihrmark, K., Lindahl, B.D., 2011. Decomposing capacity of fungi commonly

detected in Pinus sylvestris needle litter. Fungal Ecology 4, 110–114.

doi:10.1016/j.funeco.2010.09.002

Borcard, D., Gillet, F., Legendre, P., 2011. Numerical Ecology with R. Springer, New York.

Brady, N.C., Weil, R.R., 2002. The Nature and Properties of Soils - Thirteenth Edition, 13th ed.

Pearson Education, Upper Saddle River.

Breuker, A., Köweker, G., Blazejak, A., Schippers, A., 2011. The deep biosphere in terrestrial

179 sediments in the Chesapeake Bay area, Virginia, USA. Frontiers in Microbiology 2, 1–13.

doi:10.3389/fmicb.2011.00156

Brockett, B.F.T., Prescott, C.E., Grayston, S.J., 2012. Soil moisture is the major factor

influencing microbial community structure and enzyme activities across seven

biogeoclimatic zones in western Canada. Soil Biology and Biochemistry 44, 9–20.

doi:10.1016/j.soilbio.2011.09.003

Caporaso, G., 2011. Earth microbiome project. 16S rRNA amplification protocol [WWW

Document]. URL http://www.earthmicrobiome.org/emp-standard-protocols/16s/ (accessed

11.29.15).

Caporaso, J.G., Lauber, C.L., Walters, W. a, Berg-Lyons, D., Huntley, J., Fierer, N., Owens,

S.M., Betley, J., Fraser, L., Bauer, M., Gormley, N., Gilbert, J. a, Smith, G., Knight, R.,

2012. Ultra-high-throughput microbial community analysis on the Illumina HiSeq and

MiSeq platforms. The ISME Journal 6, 1621–1624. doi:10.1038/ismej.2012.8

Carmosini, N., Devito, K.J., Prepas, E.E., 2002. Gross nitrogen transformations in harvested and

mature aspen-conifer mixed forest soils from the Boreal Plain. Soil Biology and

Biochemistry 34, 1949–1951.

CERI, 2012. Canadian Oil Sands Supply and Cost (2011-2045). Calgary.

Chazdon, R.L., 2008. Beyond Deforestation : Restoring Degraded Lands. Communities 1458,

1458–1460. doi:10.1126/science.1155365

Chen, H., Mothapo, N. V., Shi, W., 2014. The significant contribution of fungi to soil N2O

production across diverse ecosystems. Applied Soil Ecology 73, 70–77.

doi:10.1016/j.apsoil.2013.08.011

Cheng, Y., Wang, J., Mary, B., Zhang, J. bo, Cai, Z. cong, Chang, S.X., 2013. Soil pH has

180 contrasting effects on gross and net nitrogen mineralizations in adjacent forest and grassland

soils in central Alberta, Canada. Soil Biology and Biochemistry 57, 848–857.

doi:10.1016/j.soilbio.2012.08.021

Cho, Y., Kim, S., Park, H., Komarneni, S., Hong, Y., 2014. Removal of inorganic pollutants in

rainwater by a peat-derived porous material. Journal of Porous Materials 21, 387–394.

doi:10.1007/s10934-014-9784-9

Clark, F.E., 1981. The Nitrogen Cycle, Viewed With Poetic Licence, in: Clark, F.E., Rosswall,

T. (Eds.), Terrestrial Nitrogen Cycles. Processes, Ecosystem Strategies and Management

Impacts. Ecological Bulletins, Stockholm, pp. 13–24.

Clemmensen, K.E., Finlay, R.D., Dahlberg, A., Stenlid, J., Wardle, D. a, Lindahl, B.D., 2015.

Carbon sequestration is related to mycorrhizal fungal community shifts during long term

succession in boreal forests. New Phytologist 1525–1536.

CNBC, 2016. West Texas Intermediate crude oil. Market exchange prices [WWW Document].

URL http://data.cnbc.com/quotes/%40CL.1/tab/1 (accessed 3.5.16).

Compton, J.E., Hooker, T.D., Perakis, S.S., 2007. Ecosystem N Distribution and δ15N during a

Century of Forest Regrowth after Agricultural Abandonment. Ecosystems 10, 1197–1208.

doi:10.1007/s10021-007-9087-y

Cornelissen, J., Aerts, R., Cerabolini, B., Werger, M., van der Heijden, M., 2001. Carbon cycling

traits of plant species are linked with mycorrhizal strategy. Oecologia 129, 611–9.

doi:10.1007/s004420100752

Corre, M.D., Schnabel, R.R., Stout, W.L., 2002. Spatial and seasonal variation of gross nitrogen

transformations and microbial biomass in a northeastern US grassland. Soil Biology and

Biochemistry 34, 445–457. doi:10.1016/S0038-0717(01)00198-5

181 Crenshaw, C.L., Lauber, C., Sinsabaugh, R.L., Stavely, L.K., 2008. Fungal control of nitrous

oxide production in semiarid grassland. Biogeochemistry 87, 17–27. doi:10.1007/s10533-

007-9165-4

Davidson, E., Hart, S., 1991. Measuring gross nitrogen mineralization, and nitrification by 15 N

isotopic pool dilution in intact soil cores. Journal of Soil … 42, 335–349.

Davidson, E.A., Hart, S.C., Firestone, M.K., 1992. Internal Cycling of Nitrate in Soils of a

Mature Coniferous Forest. Ecology 73, 1148–1156.

Davis, M., Bajwa, K., Person, R., 2015. Predicted spatial variations of sulphur and nitrogen

compound concentrations and deposition in the AOSR, in: Clair, T.A., Percy, K.E. (Eds.),

Assessing Forest Health in the Alberta Oil Sands Region. Wood Buffalo Environmental

association, Fort McMurray, pp. 51–63.

DeLuca, T.H., Zackrisson, O., Nilsson, M.-C., Seastedt, A., 2002. Quantifying nitrogen-fixation

in feather moss carpets of boreal forests. Nature 419, 917–920. doi:10.1038/nature01136.1.

Dimitriu, P. a., Grayston, S.J., 2010. Relationship between soil properties and patterns of

bacterial β-diversity across reclaimed and natural boreal forest soils. Microbial Ecology 59,

563–573. doi:10.1007/s00248-009-9590-0

Dimitriu, P. a., Prescott, C.E., Quideau, S. a., Grayston, S.J., 2010. Impact of reclamation of

surface-mined boreal forest soils on microbial community composition and function. Soil

Biology and Biochemistry 42, 2289–2297. doi:10.1016/j.soilbio.2010.09.001

Ding, G.-C., Piceno, Y.M., Heuer, H., Weinert, N., Dohrmann, A.B., Carrillo, A., Andersen,

G.L., Castellanos, T., Tebbe, C.C., Smalla, K., 2013. Changes of soil bacterial diversity as a

consequence of agricultural land use in a semi-arid ecosystem. PLoS ONE 8, e59497.

doi:10.1371/journal.pone.0059497

182 Dray, S., Legendre, P., Blanchet, G., 2011. Packfor: Forward selection with permutation (Canoco

p.46). R package version 0.0-8.

Drury, C.F., Hart, S.C., Yang, X.M., 2008a. Nitrification Techniques for Soils, in: Carter, M.R.,

Gregorich, E.G. (Eds.), Soil Sampling and Methods of Analysis. Second Edition. Canadian

Society of Soil Science, Boca Raton, pp. 495–513.

Drury, C.F., Hart, S.C., Yang, X.M., 2008b. Nitrification Techniques for Soils, in: M.R., C.,

E.G., G. (Eds.), Soil Sampling and Methods of Analysis. Second Edition. Canadian Society

of Soil Science, Boca Raton, pp. 495–513.

Edgar, R.C., Haas, B.J., Clemente, J.C., Quince, C., Knight, R., 2011. UCHIME improves

sensitivity and speed of chimera detection. Bioinformatics 27, 2194–2200.

doi:10.1093/bioinformatics/btr381

Edwards, I.P., Zak, D.R., 2010. Phylogenetic similarity and structure of Agaricomycotina

communities across a forested landscape. Molecular Ecology 19, 1469–1482.

doi:10.1111/j.1365-294X.2010.04566.x

Eilers, K.G., Lauber, C.L., Knight, R., Fierer, N., 2010. Shifts in bacterial community structure

associated with inputs of low molecular weight carbon compounds to soil. Soil Biology and

Biochemistry 42, 896–903. doi:10.1016/j.soilbio.2010.02.003

Environment Canada, 2015. Canadian climate normals 1821-2000 - Fort McMurray [WWW

Document]. URL

http://climate.weather.gc.ca/climate_normals/results_1981_2010_e.html?stnID=2519&lang

=f&StationName=Fort&SearchType=Contains&stnNameSubmit=go&dCode=1

EPA, 2013. Overview of Greenhouse Gases [WWW Document]. URL

http://epa.gov/climatechange/ghgemissions/gases/n2o.html

183 Fan, F., Li, Z., Wakelin, S.A., Yu, W., Liang, Y., 2012. Mineral fertilizer alters cellulolytic

community structure and suppresses soil cellobiohydrolase activity in a long-term

fertilization experiment. Soil Biology and Biochemistry 55, 70–77.

doi:10.1016/j.soilbio.2012.06.008

Fenn, M.E., Bytnerowicz, A., Schilling, S.L., Ross, C.S., 2015. Atmospheric deposition of

nitrogen, sulfur and base cations in jack pine stands in the Athabasca Oil Sands Region,

Alberta, Canada. Environmental Pollution 196, 497–510. doi:10.1016/j.envpol.2014.08.023

Ferguson, B.G., 1984. Athabasca Oil Sands : Northern Resource Exploration : 1875-1951.

Calgary.

Ferrenberg, S., O’Neill, S.P., Knelman, J.E., Todd, B., Duggan, S., Bradley, D., Robinson, T.,

Schmidt, S.K., Townsend, A.R., Williams, M.W., Cleveland, C.C., Melbourne, B. a, Jiang,

L., Nemergut, D.R., 2013. Changes in assembly processes in soil bacterial communities

following a wildfire disturbance. The ISME Journal 7, 1102–1111.

doi:10.1038/ismej.2013.11

Fierer, N., Bradford, M. a., Jackson, R.B., 2007. Toward an ecological classification of soil

bacteria. Ecology 88, 1354–1364. doi:10.1890/05-1839

Fierer, N., Jackson, R.B., 2006. The diversity and biogeography of soil bacterial communities.

Proceedings of the National Academy of Sciences of the United States of America 103,

626–631. doi:10.1073/pnas.0507535103

Fierer, N., Lauber, C.L., Ramirez, K.S., Zaneveld, J., Bradford, M.A., Knight, R., 2012.

Comparative metagenomic, phylogenetic and physiological analyses of soil microbial

communities across nitrogen gradients. The ISME Journal 6, 1007–1017.

doi:10.1038/ismej.2011.159

184 Fonseca, A., Inacio, J., 2006. Phylloplane yeasts, in: Rosa, C., Peter, G. (Eds.), Biodiversity and

Ecophysiology of Yeasts. The Yeast Handbook. Springer-Verlag, New York, pp. 263–303.

Fontaine, S., Bardoux, G., Abbadie, L., Mariotti, A., 2004. Carbon input to soil may decrease

soil carbon content. Ecology Letters 7, 314–320. doi:10.1111/j.1461-0248.2004.00579.x

Freedman, Z., Zak, D.R., 2014. Atmospheric N deposition increases bacterial laccase-like

multicopper oxidases: implications for organic matter decay. Applied and Environmental

Microbiology 80, 4460–4468. doi:10.1128/AEM.01224-14

Frey, S.D., Ollinger, S., Nadelhoffer, K., Bowden, R., Brzostek, E., Burton, A., Caldwell, B.A.,

Crow, S., Goodale, C.L., Grandy, A.S., Finzi, A., Kramer, M.G., Lajtha, K., LeMoine, J.,

Martin, M., McDowell, W.H., Minocha, R., Sadowsky, J.J., Templer, P.H., Wickings, K.,

2014. Chronic nitrogen additions suppress decomposition and sequester soil carbon in

temperate forests. Biogeochemistry 121, 305–316. doi:10.1007/s10533-014-0004-0

Frouz, J., Keplin, B., Pižl, V., Tajovský, K., Starý, J., Lukešová, A., Nováková, A., Balı́k, V.,

Háněl, L., Materna, J., Düker, C., Chalupský, J., Rusek, J., Heinkele, T., 2001. Soil biota

and upper soil layer development in two contrasting post-mining chronosequences.

Ecological Engineering 17, 275–284. doi:http://dx.doi.org/10.1016/S0925-8574(00)00144-0

Frouz, J., Prach, K., Pižl, V., Háněl, L., Starý, J., Tajovský, K., Materna, J., Balík, V., Kalčík, J.,

Řehounková, K., 2008. Interactions between soil development, vegetation and soil fauna

during spontaneous succession in post mining sites. European Journal of Soil Biology 44,

109–121. doi:10.1016/j.ejsobi.2007.09.002

Galloway, N., Schlesinger, W.H., Levy, H., Michaels, A., Schnoor, J.L., 1995. Nitrogen

fixation : Anthropogenic enhancement-environmental response. Gobal Biogeochemical

Cycles 9, 235–252.

185 Gebauer, R.L.E., Tenhunen, J.D., Reynolds, J.F., 1996. Soil aeration in relation to soil physical

properties, nitrogen availability, and root characteristics within an arctic watershed. Plant

and Soil 178, 37–48. doi:10.1007/BF00011161

Geisseler, D., Horwath, W.R., Joergensen, R.G., Ludwig, B., 2010. Pathways of nitrogen

utilization by soil microorganisms - A review. Soil Biology and Biochemistry 42, 2058–

2067. doi:10.1016/j.soilbio.2010.08.021

Giraudoux, P., 2014. pgirmess: Data analysis in ecology.

Golder, 2003. Evaluation of historic and future acid deposition effects on soils in the Athabasca

oil sands region. Appendix I. Air emissions for the evaluation of historic and future acid

deposition effects on soils in the Athabasca Oil Sands Region. Calgary.

Government of Alberta, 2016a. Alberta oil sands industry. Quarterly update.

Government of Alberta, 2016b. Alberta Energy: Facts and Statistics [WWW Document]. URL

http://www.energy.alberta.ca/oilsands/791.asp (accessed 3.5.16).

Government of Alberta, 2015. The Oil Sands - Reclamation.

Government of Alberta, 2013a. Oil Sands - The Resource.

Government of Alberta, 2013b. Oil Sands. Reclamation 1–2.

Government of Alberta, 2012. Oil Sands. The Resource 1–2.

Government of Alberta, 1993. Conservation and Reclamation Regulation (AR 115-1993).

Queen’s printer, Edmonton.

Grayston, S.J., Prescott, C.E., 2005. Microbial communities in forest floors under four tree

species in coastal British Columbia. Soil Biology and Biochemistry 37, 1157–1167.

doi:10.1016/j.soilbio.2004.11.014

Griffiths, B., Ritz, K., Wheatley, R., Kuan, H., Boag, B., Christensen, S., Ekelund, F., Sørensen,

186 S., Muller, S., Bloem, J., 2001. An examination of the biodiversity–ecosystem function

relationship in arable soil microbial communities. Soil Biology and Biochemistry 33, 1713–

1722. doi:10.1016/S0038-0717(01)00094-3

Gundale, M.J., Bach, L.H., Nordin, A., 2013. The impact of simulated chronic nitrogen

deposition on the biomass and N 2 -fixation activity of two boreal feather moss –

cyanobacteria associations. Biology Letters 9, 3–7. doi:10.1098/rsbl.2013.0797

Hahn, A.S., Quideau, S. a., 2013. Long-term effects of organic amendments on the recovery of

plant and soil microbial communities following disturbance in the Canadian boreal forest.

Plant and Soil 363, 331–344. doi:10.1007/s11104-012-1306-4

Hansel, C.M., Fendorf, S., Jardine, P.M., Francis, C. a., 2008. Changes in bacterial and archaeal

community structure and functional diversity along a geochemically variable soil profile.

Applied and Environmental Microbiology 74, 1620–1633. doi:10.1128/AEM.01787-07

Harris, J. a, Grogan, P., Hobbs, R.J., 2005. Restoration ecology and the role of soil biodiversity,

in: Bardgett, R.D., Usher, M.B., Hopkins, D.W. (Eds.), Biological Diversity and Function in

Soils. Cambridge University Press, Cambridge, pp. 319–342.

Hart, S.C., DeLuca, T.H., Newman, G.S., MacKenzie, M.D., Boyle, S.I., 2005. Post-fire

vegetative dynamics as drivers of microbial community structure and function in forest

soils. Forest Ecology and Management 220, 166–184. doi:10.1016/j.foreco.2005.08.012

Hart, S.C., DeLuca, T.H., Newman, G.S., MacKenzie, M.D., Boyle, S.I., 2005. Post-fire

vegetative dynamics as drivers of microbial community structure and function in forest

soils. Forest Ecology and Management 220, 166–184. doi:10.1016/j.foreco.2005.08.012

Hart, S.C., Stark, J.M., Davidson, E.A., Fireston, M.K., 1994. Nitrogen Mineralization,

Immobilization, and Nitrification, in: Weaver, R.W., Angle, J.S., Bottomley, P.S. (Eds.),

187 Methods of Soil Analysis. Part 2 - Microbiological and Biochemical Properties. Soil

Science Society of America, Madison, pp. 985–1018.

Hartmann, M., Howes, C.G., VanInsberghe, D., Yu, H., Bachar, D., Christen, R., Henrik

Nilsson, R., Hallam, S.J., Mohn, W.W., 2012. Significant and persistent impact of timber

harvesting on soil microbial communities in Northern coniferous forests. The ISME Journal

6, 2199–2218. doi:10.1038/ismej.2012.84

Hartmann, M., Niklaus, P. a, Zimmermann, S., Schmutz, S., Kremer, J., Abarenkov, K., Lüscher,

P., Widmer, F., Frey, B., 2014. Resistance and resilience of the forest soil microbiome to

logging-associated compaction. The ISME Journal 8, 226–44. doi:10.1038/ismej.2013.141

Hättenschwiler, S., Tiunov, A., Scheu, S., 2005. Biodiversity and litter deomposition in

terrestrial ecosystems. Annual Review of Ecology and Evolution 36, 191–218.

doi:10.1146/annurev.ecolsys.36.112904.151932

Hayatsu, M., Tago, K., Saito, M., 2008. Various players in the nitrogen cycle: Diversity and

functions of the microorganisms involved in nitrification and denitrification. Soil Science

and Plant Nutrition 54, 33–45. doi:10.1111/j.1747-0765.2007.00195.x

He, J.Z., Hu, H.W., Zhang, L.M., 2012. Current insights into the autotrophic thaumarchaeal

ammonia oxidation in acidic soils. Soil Biology and Biochemistry 55, 146–154.

doi:10.1016/j.soilbio.2012.06.006

Hemsley, T.L., 2012. Ecological Response of Atmospheric Nitrogen Deposition on

Reconstructed Soils in the Athabasca Oil Sands Region. University of Alberta.

Hendershot, W.H., Duquette, M., 1986. A simple barium chloride method for determining cation

exchange capacity and exchangeable cations. Soil Science Society of America Journal 50,

605–608.

188 Higgs, E., 2003. Nature by Design, First Edit. ed. MIT Press, Cambridge.

Hobbs, R.J., Arico, S., Aronson, J., Baron, J.S., Cramer, V.A., Epstein, P.R., Ewel, J.J., Klink,

C.A., Lugo, A.E., Norton, D., Ojima, D., Richardson, D.M., 2006. Novel ecosystems :

theoretical and management aspects of the new ecological world order. Global Ecology and

Biogeography 15, 1–7. doi:10.1111/j.1466-822x.2006.00212.x

Högberg, M.N., Högberg, P., Myrold, D.D., 2007. Is microbial community composition in boreal

forest soils determined by pH, C-to-N ratio, the trees, or all three? Oecologia 150, 590–601.

doi:10.1007/s00442-006-0562-5

Holden, S.R., Gutierrez, A., Treseder, K.K., 2013. Changes in Soil Fungal Communities,

Extracellular Enzyme Activities, and Litter Decomposition Across a Fire Chronosequence

in Alaskan Boreal Forests. Ecosystems 16, 34–46. doi:10.1007/s10021-012-9594-3

Horner, D., 2013. Budget 2013. Responsible change. Edmonton.

Hutchinson, G., Rochette, P., 2003. Non-flow-through steady-state chambers for measuring soil

respiration. Soil Science Society of … 67, 166–180.

IAEA, 2001. Use of isotope and radiation methods in soil and water management and crop

nutrition. Vienna.

Inselsbacher, E., Näsholm, T., 2012. The below-ground perspective of forest plants: Soil

provides mainly organic nitrogen for plants and mycorrhizal fungi. New Phytologist 195,

329–334. doi:10.1111/j.1469-8137.2012.04169.x

Jenny, H., 1941. Factors of soil formation. McGraw-Hill, New York.

Jerabkova, L., Prescott, C.E., Kishchuk, B.E., 2006. Effect of variable-retention harvesting on

soil nitrogen availability in boreal mixedwood forests. Canadian Journal of Forest Research

36, 3029–3038. doi:10.1139/x06-175

189 Johnson, D., IJdo, M., Genney, D.R., Anderson, I.C., Alexander, I.J., 2005. How do plants

regulate the function, community structure, and diversity of mycorrhizal fungi? Journal of

Experimental Botany 56, 1751–1760. doi:10.1093/jxb/eri192

Jones, C.M., Stres, B., Rosenquist, M., Hallin, S., 2008. Phylogenetic analysis of nitrite, nitric

oxide, and nitrous oxide respiratory enzymes reveal a complex evolutionary history for

denitrification. Molecular Biology and Evolution 25, 1955–66. doi:10.1093/molbev/msn146

Kachalkin, A. V., Yurkov, A.M., 2012. Yeast communities in Sphagnum phyllosphere along the

temperature-moisture ecocline in the boreal forest-swamp ecosystem and description of

Candida sphagnicola sp. nov. Antonie van Leeuwenhoek, International Journal of General

and Molecular Microbiology 102, 29–43. doi:10.1007/s10482-012-9710-6

Kemnitz, D., Kolb, S., Conrad, R., 2007. High abundance of Crenarchaeota in a temperate acidic

forest soil. FEMS Microbiology Ecology 60, 442–448. doi:10.1111/j.1574-

6941.2007.00310.x

Kennedy, N., Egger, K.N., 2010. Impact of wildfire intensity and logging on fungal and

nitrogen-cycling bacterial communities in British Columbia forest soils. Forest Ecology and

Management 260, 787–794. doi:10.1016/j.foreco.2010.05.037

Kernaghan, G., Patriquin, G., 2015. Diversity and host preference of fungi co-inhabiting

Cenococcum mycorrhizae. Fungal Ecology 17, 84–95. doi:10.1016/j.funeco.2015.05.001

Kernaghan, G., Widden, P., Bergeron, Y., Legare, S., Pare, D., 2003. Biotic and abiotic factors

affecting ectomycorrhizal diversity in boreal mixed-woods. Oikos 102, 497–504.

doi:10.1034/j.1600-0706.2003.12415.x

Kirkham, D., Bartholomew, W. V., 1955. Equations for Following Nutrient Transformations in

Soil, Utilizing Tracer Data: II. Soil Science Society of America Journal 19, 189.

190 doi:10.2136/sssaj1955.03615995001900020020x

Kirkham, D., Bartholomew, W., 1954. Equations for following nutrient transformations in soil,

utilizing tracer data. Soil Science Society of … 1, 33–34.

Koch, A.L., 2001. Oligotrophs versus copiotrophs. BioEssays 23, 657–661.

doi:10.1002/bies.1091

Konert, M., Vandenberghe, J., 1997. Comparison of laser grain size analysis with pipette and

sieve analysis: a solution for the underestimation of the clay fraction. Sedimentology 523–

535.

Kourtev, P.S., Ehrenfeld, J.G., Häggblom, M., 2002. Exotic plant species alter the microbial

community structure and function in the soil. Ecology 83, 3152–3166. doi:10.1890/0012-

9658(2002)083[3152:EPSATM]2.0.CO;2

Kowalchuk, G.A., Buma, D.S., de Boer, W., Klinkhamer, P.G.L., van Veen, J.A., 2002. Effects

of above-ground plant species composition and diversity on the diversity of soil-borne

microorganisms. International Journal of General and Molecular Microbiology 81, 509–

520. doi:10.1023/A:1020565523615

Kozich, J.J., Westcott, S.L., Baxter, N.T., Highlander, S.K., Schloss, P.D., 2013. Development of

a dual-index sequencing strategy and curation pipeline for analyzing amplicon sequence

data on the MiSeq Illumina sequencing platform. Applied and Environmental Microbiology

79, 5112–5120. doi:10.1128/AEM.01043-13

Kranabetter, J.M., Hawkins, B.J., Jones, M.D., Robbins, S., Dyer, T., Li, T., 2015. Species

turnover (β-diversity) in ectomycorrhizal fungi linked to NH4+ uptake capacity. Molecular

Ecology 24, 5992–6005. doi:10.1111/mec.13435

Laganière, J., Paré, D., Bradley, R.L., 2009. Linking the abundance of aspen with soil faunal

191 communities and rates of belowground processes within single stands of mixed aspen-black

spruce. Applied Soil Ecology 41, 19–28. doi:10.1016/j.apsoil.2008.08.005

Laidlaw, M., 2015. Soil carbon stabilization under three reclaimed vegetation types in the

Alberta oil sands. University of British Columbia.

Lauber, C.L., Hamady, M., Knight, R., Fierer, N., 2009. Pyrosequencing-based assessment of

soil pH as a predictor of soil bacterial community structure at the continental scale. Applied

and Environmental Microbiology 75, 5111–5120. doi:10.1128/AEM.00335-09

Laughlin, R.J., Stevens, R.J., 2002. Evidence for fungal dominance of denitrification and

codenitrification in a grassland soil. Soil Science Society of America Journal 66, 1540–

1548. doi:10.2136/sssaj2002.1540

Leff, J.W., Nemergut, D.R., Grandy, A.S., O’Neill, S.P., Wickings, K., Townsend, A.R.,

Cleveland, C.C., 2012. The effects of soil bacterial community structure on decomposition

in a tropical rain forest. Ecosystems 15, 284–298. doi:10.1007/s10021-011-9510-2

Legendre, P., 2007. anova.1way [WWW Document]. URL

http://adn.biol.umontreal.ca/~numericalecology/FonctionsR/

Legendre, P., Gallagher, E.D., 2001. Ecologically meaningful transformations for ordination of

species data. Oecologia 129, 271–280. doi:10.1007/s004420100716

Legendre, P., Legendre, L., 2012. Numerical Ecology, Third engl. ed. Elsevier, Amsterdam.

Lehtovirta, L.E., Prosser, J.I., Nicol, G.W., 2009. Soil pH regulates the abundance and diversity

of Group 1.1c Crenarchaeota. FEMS Microbiology Ecology 70, 367–376.

doi:10.1111/j.1574-6941.2009.00748.x

Leininger, S., Urich, T., Schloter, M., Schwark, L., Qi, J., Nicol, G.W., Prosser, J.I., Schuster,

S.C., Schleper, C., 2006. Archaea predominate among ammonia-oxidizing prokaryotes in

192 soils. Nature 442, 806–9. doi:10.1038/nature04983

Li, C., Yan, K., Tang, L., Jia, Z., Li, Y., 2014. Change in deep soil microbial communities due to

long-term fertilization. Soil Biology and Biochemistry 75, 264–272.

doi:10.1016/j.soilbio.2014.04.023

Li, K., Zhao, Y., Yuan, X., Zhao, H., Wang, Z., Li, S., Malhi, S.S., 2012. Comparison of Factors

Affecting Soil Nitrate Nitrogen and Ammonium Nitrogen Extraction. Communications in

Soil Science and Plant Analysis 43, 571–588. doi:10.1080/00103624.2012.639108

Li, Q., Allen, H.L., Wilson, C. a, 2003. Nitrogen mineralization dynamics following the

establishment of a loblolly pine plantation. Canadian Journal of Forest Research 33, 364–

374. doi:10.1139/x02-184

Lin, X., Handley, K.M., Gilbert, J.A., Kostka, J.E., 2015. Metabolic potential of fatty acid

oxidation and anaerobic respiration by abundant members of Thaumarchaeota and

Thermoplasmata in deep anoxic peat. The ISME Journal 9, 2740–2744.

doi:10.1038/ismej.2015.77

Lindahl, B.D., Ihrmark, K., Boberg, J., Trumbore, S.E., Högberg, P., Stenlid, J., Finlay, R.D.,

2007. Spatial separation of litter decomposition and mycorrhizal nitrogen uptake in a boreal

forest. New Phytologist 611–620. doi:10.1111/j.1469-8137.2006.01936.x

Lipson, D., Näsholm, T., 2001. The Unexpected Versatility of Plants : Organic Nitrogen Use and

Availability in Terrestrial Ecosystems. Oecologia 128, 305–316.

doi:10.1007/S004420100693

Lonsdale, D., Pautasso, M., Holdenrieder, O., 2008. Wood-decaying fungi in the forest:

Conservation needs and management options. European Journal of Forest Research 127, 1–

22. doi:10.1007/s10342-007-0182-6

193 Lozupone, C., Knight, R., 2005. UniFrac : a New Phylogenetic Method for Comparing Microbial

Communities. Applied and Environmental Microbiology 71, 8228–8235.

doi:10.1128/AEM.71.12.8228

Lozupone, C.A., Hamady, M., Kelley, S.T., Knight, R., 2007. Quantitative and qualitative β-

diversity measures lead to different insights into factors that structure microbial

communities. Applied and Environmental Microbiology 73, 1576–1585.

doi:10.1128/AEM.01996-06

Lupi, C., Morin, H., Deslauriers, A., Rossi, S., Houle, D., 2013. Role of Soil Nitrogen for the

Conifers of the Boreal Forest : A Critical Review. International Journal of Plant and Soil

Science 2, 155–189.

Luxhøi, J., Mary, B., Jensen, L.S., 2005. Comments to “A 15N tracing model to analyse N

transformations in old grassland soil.” Soil Biology and Biochemistry 37, 1003–1005.

doi:10.1016/j.soilbio.2004.11.001

MacKenzie, M., Quideau, S., 2012. Laboratory-based nitrogen mineralization and

biogeochemistry of two soils used in oil sands reclamation. Canadian Journal of Soil

Science. doi:10.4141/CJSS2010-070

Männistö, M.K., Kurhela, E., Tiirola, M., Häggblom, M.M., 2013. Acidobacteria dominate the

active bacterial communities of Arctic tundra with widely divergent winter-time snow

accumulation and soil temperatures. FEMS Microbiology Ecology 84, 47–59.

doi:10.1111/1574-6941.12035

Martin, T.L., Trevors, J.T., Kaushik, N.K., 1999. Soil microbial diversity , community structure

and denitrification in a temperate riparian zone. Biodiversity and Conservation 8, 1057–

1078.

194 Mary, B., Recous, S., Robin, D., 1998. A model for calculating nitrogen fluxes in soil using 15N

tracing. Soil Biology and Biochemistry 30, 1963–1979.

Matson, P., Lohse, K. a, Hall, S.J., 2002. The globalization of nitrogen deposition: consequences

for terrestrial ecosystems. Ambio 31, 113–119. doi:10.1639/0044-

7447(2002)031[0113:TGONDC]2.0.CO;2

McCarthy, A.J., 1987. Lignocellulose-degrading actinomycetes. FEMS Microbiology Reviews

46, 145–163. doi:10.1016/0378-1097(87)90061-9

McCarthy, A.J., Williams, S.T., 1992. Actinomycetes as agents of biodegradation in the

environment - A review. Gene 115, 189–192. doi:10.1016/0378-1119(92)90558-7

McMillan, R., Quideau, S. a, MacKenzie, M.D., Biryukova, O., 2007. Nitrogen mineralization

and microbial activity in oil sands reclaimed boreal forest soils. Journal of Environmental

Quality 36, 1470–8. doi:10.2134/jeq2006.0530

McNeill, A., Unkovich, M., 2007. The Nitrogen Cycle in Terrestrial Ecosystems, in: Marschner,

P., Rengel, Z. (Eds.), Nutrient Cycling in Terestrial Ecosystems. Berlin.

McNevin, D., Barford, J., Hage, J., 1999. Adsorption and biological degradation of ammonium

and sulfide on peat. Water Research 33, 1449–1459. doi:10.1016/S0043-1354(98)00345-5

Melillo, J.M., 1981. Nitrogen cycling in deciduous forests, in: Clark, F.E., Rosswall T. (Eds.),

Terrestrial Nitrogen Cycles. Processes, Ecosystem Strategies and Management Impacts.

Ecological Bulletins, Stockholm, pp. 427–442.

Miethling, R., Wieland, G., Backhaus, H., Tebbe, C.C., 2000. Variation of Microbial

Rhizosphere Communities in Response to Crop Species, Soil Origin, and Inoculation with

Sinorhizobium meliloti L33. Microbial Ecology 40, 43–56. doi:10.1007/s002480000021

Mosier, A.R., Klemedtsson, L., 1994. Measuring Denitrification on the Field, in: Weaver, R.W.,

195 Angle, J.S., Bottomley, P.S. (Eds.), Methods of Soil Analysis. Part 2 - Microbiological and

Biochemical Properties. Soil Science Society of America, Madison, pp. 1047–1065.

Mothapo, N., Chen, H., Cubeta, M.A., Grossman, J.M., Fuller, F., Shi, W., 2015. Phylogenetic,

taxonomic and functional diversity of fungal denitrifiers and associated N2O production

efficacy. Soil Biology and Biochemistry 83, 160–175. doi:10.1016/j.soilbio.2015.02.001

Mulder, A., Graaf, A.A., Robertson, L.A., Kuenen, J.G., 1995. Anaerobic ammonium oxidation

discovered in a denitrifying fluidized bed reactor. FEMS Microbiology … 16, 177–184.

doi:10.1111/j.1574-6941.1995.tb00281.x

Müller, C., Rütting, T., Kattge, J., Laughlin, R.J., Stevens, R.J., 2007. Estimation of parameters

in complex 15N tracing models by Monte Carlo sampling. Soil Biology and Biochemistry

39, 715–726. doi:10.1016/j.soilbio.2006.09.021

Müller, C., Stevens, R.J., Laughlin, R.J., 2005. Response to comments by Luxhøi et al. to the

paper “A 15N tracing model to analyse N transformations….” Soil Biology and

Biochemistry 37, 1007–1008. doi:10.1016/j.soilbio.2005.01.024

Müller, C., Stevens, R.J., Laughlin, R.J., 2004. A 15N tracing model to analyse N

transformations in old grassland soil. Soil Biology and Biochemistry 36, 619–632.

doi:10.1016/j.soilbio.2003.12.006

Murphy, D.V., Recous, S., Stockdale, E.A., Fillery, I.R.P., Jensen, L.S., Hatch, D.J., Goulding,

K.W.T., 2003. Gross nitrogen fluxes in soil: theory, measurement and application of 15N

pool diluiton techniques. Advances in Agronomy 79, 69–118. doi:10.1016/S0065-

2113(02)79002-0

Myrold, D.D., Tiedje, J.M., 1986. Simultaneous estimation of several nitrogen cycle rates using

15 N: Theory and application. Soil Biology and Biochemistry 18, 559–568.

196 Näsholm, T., Kielland, K., Ganeteg, U., 2009. Uptake of organic nitrogen by plants. The New

Phytologist 182, 31–48. doi:10.1111/j.1469-8137.2008.02751.x

Natural Regions Committee, 2006. Natural Regions and Subregions of Alberta. Government of

Canada, Edmonton.

Nemergut, D.R., Cleveland, C.C., Wieder, W.R., Washenberger, C.L., Townsend, A.R., 2010.

Plot-scale manipulations of organic matter inputs to soils correlate with shifts in microbial

community composition in a lowland tropical rain forest. Soil Biology and Biochemistry 42,

2153–2160. doi:10.1016/j.soilbio.2010.08.011

Nguyen, H.D.T., Nickerson, N.L., Seifert, K. a, 2013. Basidioascus and Geminibasidium: a new

lineage of heat-resistant and xerotolerant basidiomycetes. Mycologia 105, 1231–50.

doi:10.3852/12-351

Nicol, G.W., Leininger, S., Schleper, C., Prosser, J.I., 2008. The influence of soil pH on the

diversity, abundance and transcriptional activity of ammonia oxidizing archaea and bacteria.

Environmental Microbiology 10, 2966–2978. doi:10.1111/j.1462-2920.2008.01701.x

Nicol, G.W., Tscherko, D., Embley, T.M., Prosser, J.I., 2005. Primary succession of soil

Crenarchaeota across a receding glacier foreland. Environmental Microbiology 7, 337–347.

doi:10.1111/j.1462-2920.2004.00698.x

Nômmik, H., 1981. Fixation and biological availability of ammonium on soil clay minerals, in:

Clark, F.E., Rosswall, T. (Eds.), Terrestrial Nitrogen Cycles. Processes, Ecosystem

Strategies and Management Impacts. Stockholm, pp. 273–279.

Nordin, A., Högberg, P., Näsholm, T., 2001. Soil nitrogen form and plant nitrogen uptake along

a boreal forest productivity gradient. Oecologia 129, 125–132. doi:10.1007/s004420100698

Olsson, B. a., Hansson, K., Persson, T., Beuker, E., Helmisaari, H.S., 2012. Heterotrophic

197 respiration and nitrogen mineralisation in soils of Norway spruce, Scots pine and silver

birch stands in contrasting climates. Forest Ecology and Management 269, 197–205.

doi:10.1016/j.foreco.2011.12.031

Osono, T., Takeda, H., 2002. Comparison of litter decomposing ability among diverse fungi in a

cool temperate deciduous forest in Japan. Mycologia 94, 421–427. doi:10.2307/3761776

Osono, T., Takeda, H., 2001. Organic chemical and nutrient dynamics in decomposing beech

leaf litter in relation to fungal ingrowth and succession during 3-year decomposition

processes in a cool temperate deciduous forest in Japan. Ecological Research 16, 649–670.

doi:10.1046/j.1440-1703.2001.00426.x

Oton, E.V., Quince, C., Nicol, G.W., Prosser, J.I., Gubry-Rangin, C., 2016. Phylogenetic

congruence and ecological coherence in terrestrial Thaumarchaeota. The ISME Journal 10,

85–96. doi:10.1038/ismej.2015.101

Paré, D., Bergeron, Y., 1996. Effect of colonizing tree species on soil nutrient availability in a

clay soil of the boreal mixedwood. Canadian Journal of Forest Research 26, 1022–1031.

doi:10.1017/CBO9781107415324.004

Paul, E.A., Juma, N.G., 1981. Mineralization and immobilization of soil nitrogen by

microorganisms, in: Clark, F.E., Rosswall, T. (Eds.), Terrestrial Nitrogen Cycles. Processes,

Ecosystem Strategies and Management Impacts. Ecological Bulletins, Stockholm, pp. 179–

195.

Paungfoo-Lonhienne, C., Rentsch, D., Robatzek, S., Webb, R.I., Sagulenko, E., Näsholm, T.,

Schmidt, S., Lonhienne, T.G. a, 2010. Turning the table: plants consume microbes as a

source of nutrients. PloS One 5, e11915. doi:10.1371/journal.pone.0011915

Phillips, R.P., Fahey, T.J., 2006. Tree species and mycorrhizal associations influence the

198 magnitude of rhizosphere effects. Ecology 87, 1302–1313. doi:10.1890/0012-

9658(2006)87[1302:TSAMAI]2.0.CO;2

Powell, M.J., 1993. Looking at mycology with a Janus face: a glimpse at Chytridiomycetes

active in the environment. Mycologia 85, 1–20.

Powter, C., Chymko, N., Dinwoodie, G., Howat, D., Janz, A., Puhlmann, R., Richens, T.,

Watson, D., Sinton, H., Ball, K., Patterson, B., Brocke, L., Dyer, R., 2012. Regulatory

history of Alberta’ s industrial land conservation and reclamation program. Canadian

Journal of Soil Science 92, 39–51. doi:10.4141/CJSS2010-033

Prescott, C.E., Grayston, S.J., 2013. Tree species influence on microbial communities in litter

and soil: Current knowledge and research needs. Forest Ecology and Management 309, 19–

27. doi:10.1016/j.foreco.2013.02.034

Prescott, C.E., McDonald, M.A., 1994. Effects of carbon and lime additions on mineralization of

C and N in humus from cutovers of western red cedar - western hemlock forests on northern

Vancouver Island. Canadian Journal of Forest Resources 24, 2432–2438.

Prescott, C.E., Vesterdal, L., 2013. Tree species effects on soils in temperate and boreal forests:

Emerging themes and research needs. Forest Ecology and Management 309, 1–3.

doi:10.1016/j.foreco.2013.06.042

Proemse, B.C., Mayer, B., Fenn, M.E., Ross, C.S., 2013. A multi-isotope approach for

estimating industrial contributions to atmospheric nitrogen deposition in the Athabasca oil

sands region in Alberta, Canada. Environmental Pollution 182, 80–91.

doi:10.1016/j.envpol.2013.07.004

Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T., Yarza, P., Peplies, J., Glöckner, F.O.,

2013. The SILVA ribosomal RNA gene database project: Improved data processing and

199 web-based tools. Nucleic Acids Research 41, 590–596. doi:10.1093/nar/gks1219

Quideau, S.A., Swallow, M.J.B., Prescott, C.E., Grayston, S.J., Oh, S.W., 2013. Comparing soil

biogeochemical processes in novel and natural boreal forest ecosystems. Biogeosciences 10,

5651–5661. doi:10.5194/bg-10-5651-2013

R Core Team, 2014. R: A language and environement for statistical computing.

Raison, Rj., Khanna, P., Jacobsen, K., Romanya, J., Serrasolses, I., 2009. Effects of Fire on

Forest Nutrient Cycles, in: Fire Effects on Soils and Restoration Strategies. Science

Publishers, pp. 225–256. doi:doi:10.1201/9781439843338-c8

Rawat, S.R., Männistö, M.K., Bromberg, Y., Häggblom, M.M., 2012. Comparative genomic and

physiological analysis provides insights into the role of Acidobacteria in organic carbon

utilization in Arctic tundra soils. FEMS Microbiology Ecology 82, 341–355.

doi:10.1111/j.1574-6941.2012.01381.x

Riley, R., Salamov, A.A., Brown, D.W., Nagy, L.G., Floudas, D., Held, B.W., Levasseur, A.,

Lombard, V., Morin, E., Otillar, R., Lindquist, E.A., Sun, H., LaButti, K.M., Schmutz, J.,

Jabbour, D., Luo, H., Baker, S.E., Pisabarro, A.G., Walton, J.D., Blanchette, R.A.,

Henrissat, B., Martin, F., Cullen, D., Hibbett, D.S., Grigoriev, I. V., 2014. Extensive

sampling of basidiomycete genomes demonstrates inadequacy of the white-rot/brown-rot

paradigm for wood decay fungi. Proceedings of the National Academy of Sciences of the

United States of America 111, 9923–9928. doi:Doi 10.1073/Pnas.1400592111

Ritchie, G.A.F., Nicholas, D.J., 1972. Identification of the sources of nitrous oxide produced by

oxidative and reductive processes in Nitrosomonas europaea. The Biochemical Journal 126,

1181–1191.

Robertson, G., Groffman, P.M., 2007. Nitrogen Transformations, in: Paul, E.A. (Ed.), Soil

200 Microbiology, Ecology and Biochemistry. Elsevier, Burlington, pp. 341–364.

Robertson, S.J., McGill, W.B., Massicotte, H.B., Rutherford, P.M., 2007. Petroleum

hydrocarbon contamination in boreal forest soils: a mycorrhizal ecosystems perspective.

Biological Reviews 82, 213–240. doi:10.1111/j.1469-185X.2007.00012.x

Roesch, L.F.W., Fulthorpe, R.R., Riva, A., Casella, G., Hadwin, A.K.M., Kent, A.D., Daroub,

S.H., Camargo, F. a O., Farmerie, W.G., Triplett, E.W., 2007. Pyrosequencing enumerates

and contrasts soil microbial diversity. The ISME Journal 1, 283–290.

doi:10.1038/ismej.2007.53

Rosling, A., Cox, F., Cruz-martinez, K., Ihrmark, K., Menkis, A., James, T.Y., 2011.

Archaeorhizomycetes : Unearthing an. Science 333, 876–879. doi:10.1126/science.1206958

Rousk, K., Jones, D.L., DeLuca, T.H., 2014. Moss-nitrogen input to boreal forest soils: Tracking

15N in a field experiment. Soil Biology and Biochemistry 72, 100–104.

doi:10.1016/j.soilbio.2014.01.031

Rousk, K., Jones, D.L., Deluca, T.H., 2013. Moss-cyanobacteria associations as biogenic sources

of nitrogen in boreal forest ecosystems. Frontiers in Microbiology 4, 150.

doi:10.3389/fmicb.2013.00150

Rowland, S.M., Prescott, C.E., Grayston, S.J., Quideau, S.A., Bradfield, G.E., 2009. Recreating a

functioning forest soil in reclaimed oil sands in northern alberta: an approach for measuring

success in ecological restoration. Journal of Environmental Quality 38, 1580–90.

doi:10.2134/jeq2008.0317

Roy-Bolduc, A., Laliberté, E., Hijri, M., 2015. High richness of ectomycorrhizal fungi and low

host specificity in a coastal sand dune ecosystem revealed by network analysis. Ecology and

Evolution 349–362. doi:10.1002/ece3.1881

201 Rutherford, P.M., McGill, W.B., Arocena, J.M., Figueiredo, C.T., 2008. Total Nitrogen, in:

Carter, M.R., Gregorich, E.G. (Eds.), Soil Sampling and Methods of Analysis. Second

Edition. Canadian Society of Soil Science, Boca Raton, pp. 239–250.

Rütting, T., Huygens, D., Boeckx, P., Staelens, J., Klemedtsson, L., 2013. Increased fungal

dominance in N2O emission hotspots along a natural pH gradient in organic forest soil.

Biology and Fertility of Soils 49, 715–721. doi:10.1007/s00374-012-0762-6

Santoro, B.A.E., 2016. The do-it-all nitrifier. Science 351, 342–343.

doi:10.1126/science.aad9839

Sauheitl, L., Glaser, B., Weigelt, a., 2009. Uptake of intact amino acids by plants depends on

soil amino acid concentrations. Environmental and Experimental Botany 66, 145–152.

doi:10.1016/j.envexpbot.2009.03.009

Schimel, J., 1996. Assumptions and errors in the 15NH4 pool dilution technique for measuring

mineralization and immobilization. Soil Biology and Biochemistry 28, 827–828.

Schimel, J.P., Bennett, J., 2004. Nitrogen Mineralization: Challenges of a changing paradigm.

Ecology 85, 591–602.

Schimel, J.P., Bennett, J., Fierer, N., 2005. Microbial community compostion and soil nitrogen

cycling: is there really a connection, in: Bardgett, R.D., Usher, M.B., Hopkins, D.W. (Eds.),

Biological Diversity and Function in Soils. Ecological Reviews, Cambridge, p. 411.

Schloss, P.D., 2008. Evaluating different approaches that test whether microbial communities

have the same structure. The ISME Journal 2, 265–275. doi:10.1038/ismej.2008.5

Schrijver, A., Frenne, P., Staelens, J., Verstraeten, G., Muys, B., Vesterdal, L., Wuyts, K., Nevel,

L., Schelfhout, S., Neve, S., Verheyen, K., 2012. Tree species traits cause divergence in soil

acidification during four decades of postagricultural forest development. Global Change

202 Biology 18, 1127–1140. doi:10.1111/j.1365-2486.2011.02572.x

Sheppard, M.C., 2005. Athabasca Oil Sands. From Laboratory to Production. The Letters of Karl

A. Clark, 1950-66, Geoscience. ed. Sherwood Park.

Silver, W.L., Herman, D.J., Firestone, M.K., Ecology, S., Sep, N., 2001. Dissimilatory nitrate

reduction to ammonium in upland tropical forest soils. Ecology 82, 2410–2416.

doi:10.2307/2679925

Simpson, E.H., 1949. Measurement of diversity. Nature 163, 688–688.

Sinsabaugh, R.L., 2010. Phenol oxidase, peroxidase and organic matter dynamics of soil. Soil

Biology and Biochemistry 42, 391–404. doi:10.1016/j.soilbio.2009.10.014

Skjemstad, J.O., Baldock, J.A., 2008. Total and organic carbon, in: Carter, M.R., Gregorich, E.G.

(Eds.), Soil Sampling and Methods of Analysis. Second Edition. Canadian Society of Soil

Science, Boca Raton, pp. 225–237.

Sláviková, E., Vadkertiová, R., Vránová, D., 2007. Yeasts colonizing the leaf surfaces. Journal

of Basic Microbiology 47, 344–350. doi:10.1002/jobm.200710310

Sorenson, P.T., Quideau, S.A., MacKenzie, M.D., Landhäusser, S.M., Oh, S.W., 2011. Forest

floor development and biochemical properties in reconstructed boreal forest soils. Applied

Soil Ecology 49, 139–147. doi:10.1016/j.apsoil.2011.06.006

Šourková, M., Frouz, J., Šantrùčková, H., 2005. Accumulation of carbon, nitrogen and

phosphorus during soil formation on alder spoil heaps after brown-coal mining, near

Sokolov (Czech Republic). Geoderma 124, 203–214. doi:10.1016/j.geoderma.2004.05.001

Sparling, G.P., Feltham, C.W., Reynolds, J., West, A.W., Singleton, P., 1990. Estimation of Soil

Microbial C by a Fumigation Extraction Method - Use on Soils of High Organic-Matter

Content, and a Reassessment of the Kec-Factor. Soil Biology & Biochemistry 22, 301–307.

203 doi:Doi 10.1016/0038-0717(90)90104-8

Stark, J.M., Hart, S.C., 1997. High rates of nitrification and nitrate turnover in undisturbed

coniferous forests. Nature 385, 61–64.

Ste-Marie, C., Paré, D., 1999. Soil, pH and N availability effects on net nitrification in the forest

floors of a range of boreal forest stands. Soil Biology and Biochemistry 31, 1579–1589.

Sterkenburg, E., Bahr, A., Brandström Durling, M., Clemmensen, K.E., Lindahl, B.D., 2015.

Changes in fungal communities along a boreal forest soil fertility gradient. The New

Phytologist 1145–1158. doi:10.1111/nph.13426

Sun, H., Santalahti, M., Pumpanen, J., K??ster, K., Berninger, F., Raffaello, T., Jumpponen, A.,

Asiegbu, F.O., Heinonsalo, J., 2015. Fungal community shifts in structure and function

across a boreal forest fire chronosequence. Applied and Environmental Microbiology 81,

7869–7880. doi:10.1128/AEM.02063-15

Sun, H., Terhonen, E., Kasanen, R., Asiegbu, F.O., 2014. Diversity and community structure of

primary wood-inhabiting bacteria in boreal forest. Geomicrobiology Journal 31, 315–324.

doi:10.1080/01490451.2013.827763

Switzer, J.M., Hope, G.D., Grayston, S.J., Prescott, C.E., 2012. Changes in soil chemical and

biological properties after thinning and prescribed fire for ecosystem restoration in a Rocky

Mountain Douglas-fir forest. Forest Ecology and Management 275, 1–13.

doi:10.1016/j.foreco.2012.02.025

Syncrude Canada Ltd, 2013. Unlocking the potential of the oil sands. Oil sands history [WWW

Document]. URL http://www.syncrude.ca/users/folder.asp?FolderID=5657

Tahovská, K., Kaňa, J., Bárta, J., Oulehle, F., Richter, A., Šantrůčková, H., 2013. Microbial N

immobilization is of great importance in acidified mountain spruce forest soils. Soil Biology

204 and Biochemistry 59, 58–71. doi:10.1016/j.soilbio.2012.12.015

Tan, X., Chang, S.X., Kabzems, R., 2006. Effects of tree harvesting, forest floor removal, and

compaction on soil microbial biomass, microbial respiration, and N availability in a boreal

aspen forest in British Columbia. Soil Biology and Biochemistry 38, 1734–1744.

doi:10.1016/j.soilbio.2005.11.029

Tate, K.R., Ross, D.J., Feltham, C.W., 1988. A direct extraction method to estimate soil

microbial C: Effects of experimental variables and some different calibration procedures.

Soil Biology and Biochemistry 20, 329–335.

Tedersoo, L., Bahram, M., Polme, S., Koljalg, U., Yorou, N.S., Wijesundera, R., Villarreal Ruiz,

L., Vasco-Palacios, A.M., Thu, P.Q., Suija, A., Smith, M.E., Sharp, C., Saluveer, E., Saitta,

A., Rosas, M., Riit, T., Ratkowsky, D., Pritsch, K., Poldmaa, K., Piepenbring, M., Phosri,

C., Peterson, M., Parts, K., Paertel, K., Otsing, E., Nouhra, E., Njouonkou, A.L., Nilsson,

R.H., Morgado, L.N., Mayor, J., May, T.W., Majuakim, L., Lodge, D.J., Lee, S.S., Larsson,

K.-H., Kohout, P., Hosaka, K., Hiiesalu, I., Henkel, T.W., Harend, H., Guo, L., Greslebin,

A., Grelet, G., Geml, J., Gates, G., Dunstan, W., Dunk, C., Drenkhan, R., Dearnaley, J., De

Kesel, A., Dang, T., Chen, X., Buegger, F., Brearley, F.Q., Bonito, G., Anslan, S., Abell, S.,

Abarenkov, K., 2014. Global diversity and geography of soil fungi. Science 346, 1079–

1088. doi:10.1126/science.aaa1185

Teske, A., Alm, E., Regan, J.M., Toze, S., Rittmann, B.E., Stahl, D.A., 1994. Evolutionary

Relationships among Ammonia- and Nitrite-Oxidizing Bacteria. Journal of Bacteriology

176, 6623–6630.

Thomas, G.W., 1996. Soil pH and soil acidity, in: Sparks, D.L. (Ed.), Methods of Soil Analysis

Part3. Chemical Methods. ASA-SSSA, Madison, pp. 475–490.

205 Thoms, C., Gattinger, A., Jacob, M., Thomas, F.M., Gleixner, G., 2010. Direct and indirect

effects of tree diversity drive soil microbial diversity in temperate deciduous forest. Soil

Biology and Biochemistry 42, 1558–1565. doi:10.1016/j.soilbio.2010.05.030

Thomson, M.D., 1979. Ecological habitat mapping of the AOSERP study area (supplement):

phase 1. Edmonton.

Thorn, R.G., Lynch, M.D.J., 2007. Fungi and eukayotic algae, in: Paul, E.A. (Ed.), Soil

Microbiology, Ecology and Biochemistry. Elsevier, Burlington, pp. 145–162.

Tian, J.-H., Pourcher, A.-M., Bouchez, T., Gelhaye, E., Peu, P., 2014. Occurrence of lignin

degradation genotypes and phenotypes among prokaryotes. Applied Microbiology and

Biotechnology 98, 9527–44. doi:10.1007/s00253-014-6142-4

Trehan, S.P., 1996. Immobilisation of 15NH4 in three soils by chemical and biological

processes. Soil Biology and Biochemistry 28, 1021–1027.

Tuomela, M., Vikman, M., Hatakka, a., Itävaara, M., 2000. Biodegradation of lignin in a

compost environment: A review. Bioresource Technology 72, 169–183. doi:10.1016/S0960-

8524(99)00104-2

Turcotte, I., Quideau, S. a., Oh, S.-W., 2009. Organic matter quality in reclaimed boreal forest

soils following oil sands mining. Organic Geochemistry 40, 510–519.

doi:10.1016/j.orggeochem.2009.01.003

Van Der Heijden, M.G.A., Bardgett, R.D., Van Straalen, N.M., 2008. The unseen majority: Soil

microbes as drivers of plant diversity and productivity in terrestrial ecosystems. Ecology

Letters 11, 296–310. doi:10.1111/j.1461-0248.2007.01139.x

Větrovský, T., Baldrian, P., 2013. Analysis of soil fungal communities by amplicon

pyrosequencing: Current approaches to data analysis and the introduction of the pipeline

206 SEED. Biology and Fertility of Soils 49, 1027–1037. doi:10.1007/s00374-013-0801-y

Vitousek, P.M., Aber, J.D., Howarth, R.W., Likens, G.E., Matson, P.A., Schindler, D.W.,

Schlesinger, W.H., Tilman, D.G., 1997. Human alteration of the global nitrogen cycle:

Sources and consequences. Ecological Applications 7, 737–750. doi:10.2307/2269431

Vitousek, P.M., Howarth, R.W., 1990. Nitrogen limitation on land and in the sea : How can it

occur ? Biogeochemistry 13, 87–115.

Wardle, D.A., 2002. Communities and Ecosystems. Linking the aboveground and belowground

components. Princeton University Press, Princeton.

Wasserstein, R.L., Lazar, N.A., 2016. The ASA’s statement on p-values: context, process, and

purpose. The American Statistician 1305, 00–00. doi:10.1080/00031305.2016.1154108

Weand, M.P., Arthur, M.A., Lovett, G.M., McCulley, R.L., Weathers, K.C., 2010. Effects of tree

species and N additions on forest floor microbial communities and extracellular enzyme

activities. Soil Biology and Biochemistry 42, 2161–2173. doi:10.1016/j.soilbio.2010.08.012

Weber, E.B., Lehtovirta-Morley, L.E., Prosser, J.I., Gubry-Rangin, C., 2015. Ammonia oxidation

is not required for growth of group 1.1c soil Thaumarchaeota. FEMS Microbiology Ecology

91, 1–7. doi:10.1093/femsec/fiv001

Wertz, S., Leigh, A.K.K., Grayston, S.J., 2012. Effects of long-term fertilization of forest soils

on potential nitrification and on the abundance and community structure of ammonia

oxidizers and nitrite oxidizers. FEMS Microbiology Ecology 79, 142–54.

doi:10.1111/j.1574-6941.2011.01204.x

Westbrook, C.J., Devito, K.J., 2004. Gross nitrogen transformations in soils from uncut and cut

boreal upland and peatland coniferous forest stands. Biochemistry 63, 33–50.

Williams, N., 2016. At $22, three quarters of oilsands production is underwater and losing up to

207 $3 on every barrel [WWW Document]. The National Post. URL

http://business.financialpost.com/news/energy/at-22-three-quarters-of-oilsands-producers-

are-underwater-and-losing-3-on-every-barrel?__lsa=27ef-9b86 (accessed 3.5.16).

Wrage, N., Velthof, G.L., van Beusichem, M.L., Oenema, O., 2001. Role of nitrifier

denitrification in the production of nitrous oxide. Soil Biology and Biochemistry 33, 1723–

1732.

Yamada, T., Sekiguchi, Y., 2009. Cultivation of uncultured chloroflexi subphyla: Significance

and ecophysiology of formerly uncultured chloroflexi “subphylum I” with natural and

biotechnological relevance. Microbes and Environments / JSME 24, 205–216.

doi:10.1264/jsme2.ME09151S

Yanai, Y., Toyota, K., Morishita, T., Takakai, F., Hatano, R., Limin, S.H., Darung, U., Dohong,

S., 2007. Fungal N2O production in an arable peat soil in Central Kalimantan, Indonesia.

Soil Science and Plant Nutrition 53, 806–811. doi:10.1111/j.1747-0765.2007.00201.x

Yeager, C.M., Northup, D.E., Grow, C.C., Barns, M., Kuske, C.R., Barns, S.M., 2005. Changes

in Nitrogen-Fixing and Ammonia-Oxidizing Bacterial Communities in Soil of a Mixed

Conifer Forest after Wildfire Changes in Nitrogen-Fixing and Ammonia-Oxidizing

Bacterial Communities in Soil of a Mixed Conifer Forest after Wildfire. Applied and

Environmental Microbiology 71, 2713–2722. doi:10.1128/AEM.71.5.2713

Yin, S.X., Chen, D., Chen, L.M., Edis, R., 2002. Dissimilatory nitrate reduction to ammonium

and responsible microorganisms in two Chinese and Australian paddy soils. Soil Biology

and Biochemistry 34, 1131–1137. doi:10.1016/S0038-0717(02)00049-4

Zackrisson, O., DeLuca, T.H., Nilsson, M.C., Sellstedt, a., Berglund, L.M., 2004. Nitrogen

fixation increases with successional age in boreal forests. Ecology 85, 3327–3334.

208 doi:10.1890/04-0461

209 Appendices

Appendix A Physical, chemical and biological characteristics of the studied soils

Site Forest Bulk Field Plot Vegetation age Floor pH CEC Ca Mg K Na Fe Al Mn density moisture (2013) Depth years cm g cm-3 % ------cmol(+)/kg ------C1 Coniferous 23 3.34 1.052 11.56 6.36 34.93 29.91 4.72 0.23 0.06 < BDL < BDL < BDL C2 Coniferous 22 1.79 0.947 18.49 6.39 39.10 33.02 5.70 0.18 0.18 < BDL < BDL 0.01 C3 Coniferous 29 3.36 0.809 39.93 6.63 62.45 50.58 10.66 0.38 0.80 < BDL < BDL 0.02 C4 Coniferous 28 1.77 0.784 21.30 6.87 30.56 28.64 1.68 0.24 < BDL < BDL < BDL < BDL C5 Coniferous 21 2.08 1.193 9.57 6.47 33.19 29.40 3.57 0.22 < BDL < BDL < BDL 0.01 D1 Deciduous 21 4.60 0.749 22.30 6.49 68.76 62.14 6.04 0.35 0.23 < BDL < BDL 0.01 D2 Deciduous 29 2.58 0.365 27.06 5.76 47.22 43.02 3.33 0.21 0.51 0.10 < BDL 0.05 D3 Deciduous 28 3.20 1.056 16.13 6.53 31.28 26.52 4.30 0.39 0.06 < BDL < BDL 0.01 D4 Deciduous 29 5.53 0.567 31.27 6.57 56.92 49.00 6.49 0.47 0.93 < BDL < BDL 0.01 D5 Deciduous 21 6.33 0.980 17.81 6.5 36.49 31.86 4.27 0.35 < BDL < BDL < BDL < BDL G1 Grassland 41 1.93 0.733 25.17 6.45 57.78 53.61 3.84 0.33 < BDL < BDL < BDL < BDL G2 Grassland 28 2.70 1.130 35.94 7.2 59.14 54.24 4.70 0.20 < BDL < BDL < BDL 0.01 G3 Grassland 28 4.63 1.153 26.91 7.1 50.52 44.18 6.11 0.23 < BDL < BDL < BDL < BDL G4 Grassland 35 5.37 1.053 13.27 7.13 43.27 39.90 2.93 0.16 0.28 < BDL < BDL < BDL G5 Grassland 30 3.00 0.907 25.23 7.04 50.79 45.72 4.93 0.14 < BDL < BDL < BDL < BDL N1 Natural 18 2.40 1.694 10.80 6.57 1.30 0.56 0.14 0.08 < BDL 0.03 0.49 < BDL N2 Natural 41 10.73 1.294 12.76 6.13 3.13 1.18 0.36 0.15 < BDL 0.12 1.29 0.02 N3 Natural 32 7.47 1.321 10.54 5.75 3.58 1.16 0.32 0.12 < BDL 0.05 1.80 0.12 N4 Natural 32 5.90 1.651 13.47 5.93 3.69 0.68 0.22 0.13 < BDL 0.16 2.50 0.01 N5 Natural 18 16.67 1.322 13.05 6.07 1.46 0.51 0.15 0.06 0.12 0.06 0.55 0.00

210 Appendix A : cont’

Inorganic Inorganic Nitrogen Total Root Microbial Plot Total N Total C C:N + - Clay Silt Sand Texture NH4 NO3 deposition content biomass C

-1 mg N-NH4 kg mg N-NO3 kg kgN ha -1 ug g dry % % -1 -1 T ha % % % soil soil year-1 soil-1 C1 0.19 3.83 20.6 1.46 2.65 12.5 4.09 1151.7 3.33 33.62 63.05 Sandy loam C2 0.42 7.84 18.7 1.35 1.30 15.0 8.34 1596.5 2.33 27.99 69.69 Sandy loam C3 0.71 15.51 21.9 2.49 1.26 10.0 5.05 1274.5 2.26 43.87 53.87 Sandy loam C4 0.19 4.61 24.8 11.74 6.10 10.0 8.20 1732.6 2.97 31.41 65.62 Sandy loam C5 0.27 6.35 23.7 2.04 2.18 15.0 5.22 787.8 2.15 38.20 59.38 Sandy loam D1 0.47 7.22 15.5 2.25 9.73 15.0 5.91 1526.6 1.95 34.57 63.48 Sandy loam D2 0.62 11.42 18.6 13.74 4.24 12.5 10.52 2841.2 1.04 26.21 72.76 Loamy sand D3 0.23 5.43 23.4 1.50 1.59 12.5 5.24 576.3 2.40 28.83 68.78 Sandy loam D4 0.38 6.92 18.0 1.76 2.94 15.0 5.20 1644.5 0.94 25.83 73.12 Loamy sand D5 0.36 7.16 19.7 2.67 2.45 15.0 3.96 1077.3 3.06 28.73 68.21 Sandy loam G1 0.37 7.16 19.4 1.39 2.22 12.5 3.24 975.5 1.30 22.88 74.46 Loamy sand G2 0.61 10.39 17.0 1.45 1.31 12.5 2.35 1351.8 2.08 42.02 55.91 Sandy loam G3 0.37 7.99 21.5 3.80 1.39 12.5 2.53 1029.7 2.18 39.15 58.67 Sandy loam G4 0.26 5.85 22.1 1.40 10.66 12.5 2.33 174.8 1.27 22.89 75.56 Loamy sand G5 0.33 6.27 18.8 2.59 6.28 12.5 2.52 1098.6 1.14 22.75 76.11 Loamy sand N1 0.03 0.31 11.1 1.88 1.82 3.3 4.51 926.6 1.20 26.69 72.05 Loamy sand N2 0.04 0.36 9.7 1.15 0.81 3.5 6.78 235.5 5.62 63.80 30.59 Silt loam N3 0.05 0.53 11.2 3.10 2.15 2.5 4.92 348.0 3.87 54.41 41.64 Silt loam N4 0.04 0.49 11.2 1.91 1.38 2.5 3.47 413.8 6.58 87.67 5.75 Silt N5 0.03 0.21 7.1 0.83 1.59 3.3 2.74 728.8 0.38 9.34 87.61 Sand

211 Appendix B : Vegetation description

B.1 Tree cover at each study site (source : Anderson, 2014)

Canopy Tree Populus Populus Betula Total Pinus Total Picea Total site Treatment Total tree cover density tremuloides balsamifera papyrifera Deciduous banksiana Pine glauca Spruce % ------cover (%) ------

C1 Coniferous 53.64 1900 71 0 4 0 4 0 0 65 66 C2 Coniferous 25.70 2400 65 0 0 0 0 0 0 65 65 C3 Coniferous 49.25 1300 50 3 0 0 3 0 0 45 47 C4 Coniferous 72.61 2700 84.1 0.1 0 4 4.1 0 0 80 80 C5 Coniferous 36.54 900 65.1 0 0 0 0.1 0 0 0 65 D1 Deciduous 72.19 1000 60.1 0 60 0 60.1 0 0 0 0 D2 Deciduous 61.56 3700 77 70 0 0 70 6 6 0 1 D3 Deciduous 58.43 1200 50 45 0 0 50 0 0 0 0 D4 Deciduous 70.11 3800 55 55 0 0 55 0 0 0 0 D5 Deciduous 66.20 2425 17.5 15 0 1.5 16.5 0 0 1 1 G1 Grassland 0.00 0 0 0 0 0 0 0 0 0 0 G2 Grassland 0.00 0 9 0 0 0 3 0 6 0 0 G3 Grassland 0.00 0 0.1 0 0 0 0.1 0 0 0 0 G4 Grassland 0.00 0 0 0 0 0 0 0 0 0 0 G5 Grassland 0.00 0 0 0 0 0 0 0 0 0 0 N1 Natural 33.02 0 85.1 4 0 0 4.1 75 75 6 6 N2 Natural 42.67 3900 41 10 0 0 14 8 9 12 18 N3 Natural 69.64 5500 82 65 0 0 75 1 1 0 6 N4 Natural 49.68 4900 54 0 0 0 1 40 42 1 11 N5 Natural 18.50 4900 55.5 19.75 0 0 19.75 31 31 4.75 4.75

212 B.2 Shrubs cover at each study site (source : Anderson. 2014)

Pinus Alnus Cornus Betula Picea Pinus contorta Populus viridis Amelanchier Arctostaphylos Caragana sericea Elaeagnus Kalmia Rhododendron Prunus site Treatment papyrifera glauca banksiana var. balsamifera ssp. alnifolia uvaursi species ssp. commutata polifolia groenlandicum pensylvanica < 5m <5m <5m latifolia <5m crispa sericea < 5 m ------cover (%) ------

C1 Coniferous 0 0 0 0 0 0 0 0 0 0 1 0 0 1 C2 Coniferous 0.1 0 0 0 0 0 0 0 0 0 0 0 0 0 C3 Coniferous 0 0 0 0 0 0.1 0.1 0 0 0 2 0 0 0 C4 Coniferous 2 0 0 0 0 1 0 0 0 0 0 0 0 0 C5 Coniferous 0 0 0 0 0 0 0 0 0 0 65 0 0 0 D1 Deciduous 0 0 0 0 0 0 0 0 0 0 0 0 0 0 D2 Deciduous 0 0.1 0.1 0 1 0.1 0 0 0 0 1 0 0 0 D3 Deciduous 0 0 0 0 0 1 0 0 0 0 0 0 0 0 D4 Deciduous 4 0 0 0 0 0 0 0 0 4 0 0 0 0 D5 Deciduous 0.025 0.025 0 0.25 0.275 0 0 0 1 0.25 0 0 0 1.275 G1 Grassland 0 0 0 0 0 0 0 0 0 0 0 0 0 0 G2 Grassland 0 0 0 0 0 0.1 0 0 0 0 0 0 6 0 G3 Grassland 0 0 0 0 0 0 0 0 0 0 0 0 0 0 G4 Grassland 0 0 0 0 0 0 0 0 0 0 0 0 0 0 G5 Grassland 0 0 0 0 0 0.025 0 0 0 0 0 0 1.5 0 N1 Natural 0 0 0 0.1 0 0 0 0 5 0 0 0 0 0 N2 Natural 0 0 0 0 0 0 0 2 40 0 6 1 0 0 N3 Natural 0 0 0 0 0 0 0 0 0 0 6 0 0 0 N4 Natural 0 0 0 1 0 0 0 0 1 0 10 2 0 0 N5 Natural 0 0 0 0.275 0 0 0 0.5 11.5 0 5.5 0 0 0

213 Appendix B2. Cont’ Populus Ribes Rosa Rubus Salix Salix Salix Salix Shepherdia Vaccinium Vaccinium Viburnum Total site tremuloides Salix sp. oxyacanthoides acicularis idaeus bebbiana discolor exigua scouleriana canadensis myrtilloides vitis-idaea edule Shrub <5m ------cover (%) ------

C1 0 0 4 4 0 0 0 0 0 0 0 0 0 8 C2 0 0 0 0 0 0 0 4 0 0 0 0 0 4.1 C3 0 0 4 0 1 1 0.1 0 0 0 2 0 0 8.3 C4 0 1 0.1 0 1 0 0 0 0 0 0 0 0 5.1 C5 0.1 0 0 0 0 0 0 0 0 0 0 0 0 0 D1 0.1 0 0 7 0 0 0 0.1 0 0 0 0 0 7.1 D2 0 0 3 2 2 0 0 0 0 0 0.1 0 0 8.4 D3 5 0 0.1 0 0 0 0 3 0 1 0.1 0 0 5.2 D4 0 0 4 12 6 0 0 0 8 0 0 0 0 38 D5 0 1.77 5.25 2 0 0 0.77 2 0.25 0 0.07 0 0 15.22 G1 0 0 0.1 0 0 2 0 0 0 0 0 0 0 2.1 G2 3 0 0 0 0 0.1 0 0 0 0 0 0 0 0.2 G3 0.1 0 4 0 0 0 0 0 0 0 0 0 0 4 G4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 G5 0.77 0 1.02 0 0 0.52 0 0 0 0 0 0 0 3.85 N1 0 0 0 0 0 0 0 0 0 0 8 12 0 25 N2 4 0 0 0 0 0 0 2 0 0 15 10 0 69 N3 10 0 2 0 0 0 0 0 0 0 1 4 12 19 N4 0 0 0 0 0 0 0 0 0 0 5 3 0 9 N5 3.5 0 0.5 0 0 0 0 0.5 0 0 7.25 0 0 29.52

214 B.3 Grass cover at each study site (source: Anderson. 2014)

Petasites Elymus Elymus lance- Pascopyrum Bromus Calamagrosti Festuca Oryzopsis Phleum Poa Total site Treatment Carex sp. frigidus var. trachycaulus olatus ssp. smithii inermis s canadensis sp. asperifolia pratense pratensis Grass palmatus ------cover (%) 1------

C1 Coniferous 0 0 0 0 4 0 0 0 0 0 0 4 C2 Coniferous 0 0 0.1 0 0 0.1 0 0 0 0 0.1 0.3 C3 Coniferous 0 0 0 0 0 0 0.1 0.1 0 0 0 0.2 C4 Coniferous 0 0 0 0 0 0 0 0 0 0 0.1 0.1 C5 Coniferous 0 0 0 0 0 0 0 0 0 0 0.1 0.1 D1 Deciduous 1 0 0 0 10 1 0 0 0 0 1 13 D2 Deciduous 2 0 0 0 0 0 0 0 0 0 0.1 2.1 D3 Deciduous 0 0 0 0 0 0 60 0 0 0 0 60 D4 Deciduous 0 0 0 0 8 0 0 0 0 0 0 8 D5 Deciduous 0 0 0.075 4.5 0.25 0 0.025 0 0.275 0 1.525 6.65 G1 Grassland 80 0 0 0 0 0 0 0 0 0 1 81 G2 Grassland 0 0 0 0 0 0 80 0 0 0 0 80 G3 Grassland 0 10 0 5 0 0 70 0 1 0 10 96 G4 Grassland 0 0 0 35 0 0 7 0 0 0 2 44 G5 Grassland 20 2.5 0 10 0 0 39.25 0 0.25 0 3.25 75.25 N1 Natural 0 0 0 0 0 0 0 0 0 0 0 0 N2 Natural 0 0 0 0 0 0 0 0 0 0 0 0 N3 Natural 0 0 0 0 0 0 0 0 0 0 2 2 N4 Natural 0 0 0 0 0 0 0 0 0 0 0 0 N5 Natural 0 0 0 0 0 0 0 0 0 0 0.5 0.5

215 B.4 Forbs cover at each study site (source: Anderson. 2014)

Achillea Symphyotrichum Eurybia Astragalus Castilleja Comandra Cornus Chamerion Equisetum Fragaria Galium site Treatment millefolium ciliolatum conspicua sp. miniata umbellata canadensis angustifolium sylvaticum virginiana boreale ------cover (%) ------

C1 Coniferous 0.1 0.1 0 0 0 0 0 0 0 2 1 C2 Coniferous 0 0 0 0 0 0 0 0 0 0.1 0 C3 Coniferous 0.1 0.1 0 0 0.1 0 0 0 0 1 0 C4 Coniferous 0 0.1 0 0 0 0 0 0 0 1 0 C5 Coniferous 0 0 0 0 0 0 0 1 0 0.1 0 D1 Deciduous 0.1 0.1 0 0 0 0 0 2 0.1 0.1 0 D2 Deciduous 0.1 0.1 0.1 0 0 0 0 0.1 0 0 0 D3 Deciduous 0 0 0 0 0 0 0 0 0 0 0 D4 Deciduous 0.1 0.1 0 0 0 0 0 0 0 0 1 D5 Deciduous 0.75 0.025 0 0 0 0 0.525 0.025 15 0.25 0 G1 Grassland 0 0 0 0.1 0 0 0 0.1 0 0 0.1 G2 Grassland 0 0.1 0 0 0 0 0 2 0 0 0 G3 Grassland 0.1 0 0 0 0 0 0 0 0 0 0 G4 Grassland 0 0 0 0 0 0 0 0 0 0 0 G5 Grassland 0.025 0.025 0 0.025 0 0 0 0.525 0 0 0.025 N1 Natural 0 0 0 0 0 0 4 0 0 0 0 N2 Natural 0 0 0 0 0 0 2 0 0 0 0 N3 Natural 0 0 0 0 0 0 2 0 0 0 0 N4 Natural 0 0 0 0 0 0 1 0 0 0 0 N5 Natural 0 0 0 0 0 0 2.25 0 0 0 0

216 Appendix B.4. Cont’ Petasites Geocaulon Lathyrus Lilium Linnaea Lonicera Maianthemum Medicago Melilotus Mertensia Sibbaldiopsis Pyrola Site frigidus var. lividum ochroleucus philadelphicum borealis dioica canadense sativa sp. paniculata tridentata asarifolia palmatus ------cover (%) ------C1 0 0 0 0 0 0 0 0 0 0 0 0 C2 0 0 0 0 0 0 0 30 0 0 0 0 C3 0 0 0 0 0 0 0 0 0 0 0 2 C4 0 0 0 0 0 0 0 0 0 0 0 0 C5 0 0 0 0 0 0 4 0.1 0 0 0 0 D1 0 0 0 0 0 0 0 0 0 0 0 0.1 D2 0 0 0 0 0.1 0 0 0 0 0 0 4 D3 0 0 0 0 0 0 0 0 0 0 0 2 D4 0 0 0.1 0 0 0 0 0 0 0 0 0 D5 0 0 0 0.025 0 0 0 0 0 0 0 0.025 G1 0 0.1 0 0 0 0 1 6 0.1 0 0 0 G2 0 0 0 0 0 0 2 0.1 0 0 0 0 G3 0 0 0 0 0 0 0 0 0 0 0 0 G4 0 0 0 0 0 0 45 0 0 0 0 0 G5 0 0.025 0 0 0 0 12 1.525 0.025 0 0 0 N1 1 0 0 1 0 0 0 0 0 0 0 0 N2 0.1 0 2 2 0 0 0 0 0 0 0 0 N3 0 0 0.1 4 0 1 0 0 0 0 0 2 N4 0 0 0 2 0 0 0 0 0 0 0 0 N5 0.275 0 0 2.25 0 0.25 0 0 0 0 0 0.5

217 Appendix B.4. Cont’ Solidago Solidago Sonchus Stellaria Taraxacum Trientalis Trifolium Vicia site Trifolium sp. Total Forb canadensis sp. arvensis longifolia officinale borealis repens americana ------cover (%) ------

C1 0 0 0 0 5 0 0 0 0.1 8.3 C2 0 0 0.1 0 1 0 0 0 0 31.2 C3 0 0 0 0 0.1 0 0.1 0 0 3.5 C4 0 0 0 0 1 0 0 0 0 2.1 C5 0 0 0 0 5 0 0 0 0 10.2 D1 0 0 0 0.1 4 0 0 0 0 6.6 D2 0.1 0 0 0 8 0 0 0 0 12.6 D3 0 0.1 0 0 0 0 0 0.1 0 2.2 D4 0 0 0 0 3 0 0 0 0 4.3 D5 0.025 0 0.025 3.75 0 0 0.025 0 1 5.925 G1 0 0 2 0 8 0 0 0 1 18.5 G2 0 0 0.1 0 1 0 0.1 0 0 5.4 G3 0 0 2 0 0.1 0 1 0 0 3.2 G4 0 0 0.1 0 1 0 0 0 0 46.1 G5 0 0 1.05 0 2.525 0 0.275 0 0.25 16.75 N1 0 0 0 0 0 0 0 0 0 6 N2 0 0 0 0 0 0.1 0 0 0 6.2 N3 0 0 0 0 0 1 0 0 0 10.1 N4 0 0 0 0 0 0 0 0 0 3 N5 0 0 0 0 0 0.275 0 0 0 3.55

218 B.5 Lichen and bryophyte cover at each study site (source : Anderson. 2014)

Pleuurozium Peltigera Ptilium crista- Polytrichum Ceratodon Hylocomium site Treatment schreberi aphthosa castrensis juniperinum purpureus splendens ------cover (%) ------

C1 Coniferous 15 0 0 0 0 0 C2 Coniferous 0 0 0 0 45 0 C3 Coniferous 0 0 0 0 35 0 C4 Coniferous 0 0 0 0 0 0 C5 Coniferous 5 2 0 0 70 0 D1 Deciduous 0 0 0 0 0 0 D2 Deciduous 0 0 0 0 0 0 D3 Deciduous 3 0 0 0 0 0 D4 Deciduous 1 0 0 0 0 0 D5 Deciduous 0 0 0 0 0 0 G1 Grassland 0 0 0 0 0 0 G2 Grassland 0 0 0 0 15 0 G3 Grassland 0 0 0 0 2 0 G4 Grassland 0 0 0 0 0 0 G5 Grassland 0 0 0 0 4.25 0 N1 Natural 0 3 0 0 0 0 N2 Natural 45 1 25 0 0 15 N3 Natural 6 0 1 2 0 1 N4 Natural 8 0 0 1 0 1 N5 Natural 14.75 1 6.5 0 0 0

219 Appendix C Gross N-transformation rates presented in different units

C.1 Gross N-transformation rates in the studied soils express in mg N m-2 day-1

Ammonification Immobilization NH + Ammonification Immobilization Oxidation from Immobilization NO - 4 3 Oxidation ID from recalcitrant in recalcitrant from labile NH + in labile recalcitrant in recalcitrant 4 of NH + organic-N organic-N organic-N organic-N organic-N organic-N 4 C1 260.7 220.3 0.001 26.5 19.1 3.1 17.9 C2 524.1 448.7 0.020 78.0 309.9 100.2 11.6 C3 379.1 388.0 0.019 3.3 2070.6 1998.9 3.6 D1 251.7 146.9 0.002 21.8 16.1 6.2 95.9 D2 290.7 151.8 0.023 141.0 186.3 38.1 1.7 D3 380.7 362.1 0.034 31.7 322.3 297.3 0.5 G1 584.5 467.5 0.100 122.5 540.4 480.9 1.3 G2 287.0 178.9 0.018 110.7 3823.1 3709.0 2.7 G3 149.2 63.1 0.104 93.4 278.0 46.2 0.1 N1 47.5 25.1 0.795 28.3 0.5 0.3 0.0 N2 46.5 27.6 0.656 18.0 25.7 51.1 0.0 N3 53.1 8.4 0.472 58.8 30.0 1.4 0.2

220 -1 C.2 Gross N-transformation rates in the studied soils express in mg N kgdrysoil day

Ammonification Immobilization NH + Ammonification Immobilization Oxidation from Immobilization 4 Oxidation of ID from recalcitrant in recalcitrant from labile NH + in labile recalcitrant NO - in recalcitrant 4 3 NH + organic-N organic-N organic-N organic-N organic-N organic-N 4 C1 1.65 1.40 0.0000 0.17 0.12 0.02 0.113 C2 4.32 3.70 0.0002 0.64 2.56 0.83 0.096 C3 3.23 3.30 0.0002 0.03 17.62 17.01 0.031 D1 2.24 1.31 0.0000 0.19 0.14 0.06 0.854 D2 3.42 1.79 0.0003 1.66 2.19 0.45 0.020 D3 2.59 2.46 0.0002 0.22 2.19 2.02 0.004 G1 3.45 2.76 0.0006 0.72 3.19 2.84 0.008 G2 1.66 1.03 0.0001 0.64 22.12 21.46 0.016 G3 1.10 0.46 0.0008 0.69 2.04 0.34 0.001 N1 0.24 0.13 0.0041 0.15 0.00 0.00 0.000 N2 0.19 0.11 0.0026 0.07 0.10 0.21 0.000 N3 0.27 0.04 0.0024 0.30 0.15 0.01 0.001

221

Appendix D F values, p values of ANOVA on prokaryotic communities

D.1 F values, p values and signification of ANOVA-1way on bacterial phyla (p<0.001 : ***; p<0.01 : **; p<0.05 : *; p<0.1 : .)

Bacterial phyla F value p value Level of signification

Acidobacteria 9.445 0.001 *** Actinobacteria 3.685 0.034 * Armatimonadetes 1.000 0.418 Not significative BHI80.139 1.000 0.418 Not significative Bacteroidetes 8.806 0.001 ** Candidate_division_OD1 1.000 0.418 Not significative Candidate_division_WS3 2.447 0.101 Not significative Chloroflexi 3.072 0.058 . Cyanobacteria 4.269 0.021 * Elusimicrobia 2.667 0.083 . Firmicutes 2.858 0.070 . Gemmatimonadetes 0.717 0.556 Not significative Nitrospirae 0.907 0.460 Not significative Planctomycetes 4.549 0.017 * Proteobacteria 14.381 0.000 *** TM6 1.000 0.418 Not significative Verrucomicrobia 2.997 0.062 . WD272 6.598 0.004 ** Unclassified 2.861 0.070 .

222 D.2 F values, p values and signification of ANOVA 1 way on bacterial classes (p<0.001 : ***; p<0.01 : **; p<0.05 : *; p<0.1 :.)

Bacterial classes F value p value Level of signification Acidobacteria 9.570 0.001 *** Acidimicrobiia 0.599 0.625 Not significative Actinobacteria 2.042 0.148 Not significative Alphaproteobacteria 10.367 0.000 *** Anaerolineae 3.782 0.032 * Bacilli 2.858 0.070 . Betaproteobacteria 1.932 0.165 Not significative Caldilineae 7.569 0.002 ** Chloroflexia 6.283 0.005 ** Chthonomonadetes 1.000 0.418 Not significative Cytophagia 3.878 0.029 * Deltaproteobacteria 2.844 0.071 . Elusimicrobia 2.667 0.083 . Flavobacteriia 1.000 0.418 Not significative Gammaproteobacteria 4.806 0.014 * Gemmatimonadetes 0.717 0.556 Not significative Gitt.GS.136 1.340 0.296 Not significative Holophagae 1.921 0.167 Not significative JG30.KF.CM66 3.265 0.049 * JG37.AG.4 10.209 0.001 *** KD4.96 2.948 0.064 . Ktedonobacteria 13.168 0.000 *** MB.A2.108 1.248 0.325 Not significative Melainabacteria 2.664 0.083 . Nitrospira 0.907 0.460 Not significative Opitutae 1.357 0.292 Not significative Phycisphaerae 5.137 0.011 * Planctomycetacia 1.680 0.211 Not significative S.BQ2.57 soil group 1.000 0.418 Not significative S085 1.404 0.278 Not significative Spartobacteria 6.815 0.004 ** Sphingobacteriia 4.858 0.014 * TakashiAC.B11 0.614 0.616 Not significative Thermoleophilia 3.456 0.042 * Thermomicrobia 1.490 0.255 Not significative TK10 1.220 0.335 Not significative unclassified 6.695 0.004 ** Verrucomicrobia Incertae Sedis 1.001 0.418 Not significative Verrucomicrobiae 3.969 0.027 *

223 Appendix E F values. p values and signification of ANOVA 1-way on fungi classes (*** : p<0.001; ** :p<0.01; * : p<0.05; . :p<0.1)

Fungi classes F value p value Level of signification Agaricomycetes 11.74 0.000 *** Agaricostilbomycetes 0.52 0.678 Not significative Archaeorhizomycetes 6.14 0.006 ** Chytridiomycetes 3.54 0.039 * Dothideomycetes 1.22 0.336 Not significative Eurotiomycetes 12.74 0.000 *** Glomeromycetes 7.56 0.002 ** Leotiomycetes 16.84 0.000 *** Microbotryomycetes 0.61 0.616 Not significative Orbiliomycetes 1.50 0.252 Not significative Pezizomycetes 0.87 0.476 Not significative Pucciniomycetes 1.12 0.371 Not significative Saccharomycetes 5.19 0.011 * Sordariomycetes 8.94 0.001 ** Tremellomycetes 1.11 0.373 Not significative Tritirachiomycetes 1.00 0.418 Not significative Wallemiomycetes 2.73 0.078 .

224