A Multilevel Assessment of Sediment Bioremediation with Bioturbating Macrofauna

A multilevel assessment of sediment bioremediation with bioturbating macrofauna

Sebastian Vadillo Gonzalez

A thesis submitted in fulfillment of the requirement for the degree of Doctor of Philosophy School of Biological, Earth and Environmental Sciences Faculty of Science

Supervisors: Prof. Emma Johnston, Dr. Katherine Dafforn and Assoc Prof. Paul E. Gribben

JANUARY 17, 2020

Thesis/Dissertation Sheet

Surname/Family Name : Vadillo Gonzalez Given Name/s : Sebastian Abbreviation for degree as give in the : PhD University calendar Faculty : Science School : Biological, Earth and Environmental Sciences A multilevel approach on the bioremediation of eutrophic sediments using Thesis Title : benthic macrofauna

Abstract 350 words maximum: (PLEASE TYPE)

Estuaries and coastal intertidal environments supply important ecosystem services and resources. Within these systems, sediments play a crucial role in processing organic inputs and providing nutrients for local food webs. Benthic macrofauna and sediment microbial communities participate actively in these processes and can interact to influence overall ecosystem function. Through different bioturbation processes, macrofauna can affect biogeochemical cycling in sediments and influence microbial communities. However, anthropogenic derived eutrophication has increased globally in recent years and has been found to negatively impact these systems. Excess inputs of anthropogenic contaminants can disrupt key sediment processes involving macrofauna and microbial communities. Current options for sediment remediation are costly and may result in further impacts on communities. An alternative method for sediment bioremediation involves the application of macrofaunal bioturbators to stimulate microbial processes including contaminant removal. However, to confidently apply such an alternative requires clarity on the effect of macrofauna on contaminants, the contaminant concentrations at which macrofauna will actively affect the sediments, the potential effects of bioturbators on microbial communities, and whether these applications can be applied in situ at large scales. Here I used a systematic review and meta- analysis to analyse current knowledge on the effect of bioturbating macrofauna on contaminants and quantify how this interaction changes depending on relevant biotic and abiotic variables. Secondly, I used an experimental approach with sediment mesocosms to further explore the contaminant thresholds for effective remediation by macrofauna bioturbation and interactions with microbial communities in highly enriched sediments. Finally, a large-scale field experiment was done to explore the influence in situ nitrogen enrichment on microbial communities in the sediment and within macrofauna burrows under variable environmental contexts. The combined results here demonstrate that the effect of macrofauna bioturbation on stimulating bacterial communities in organically enriched sediments is highly taxa driven and additional factors such as macrofauna intraspecific variations in body size, temperature, animal density, magnitude of contamination and site specific organic matter content need to be considered for potential bioremediation and coastal management plans.

Declaration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents a non-exclusive licence to archive and to make available (including to members of the public) my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known. I acknowledge that I retain all intellectual property rights which subsist in my thesis or dissertation, such as copyright and patent rights, subject to applicable law. I also retain the right to use all or part of my thesis or dissertation in future works (such as articles or books).

15/01/2020 ……….……………………...…….…

Signature Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years can be made when submitting the final copies of your thesis to the UNSW Library. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research. i

ORIGINALITY STATEMENT

‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’

Signed ……………………………………………......

Date ……………………………………………...... 15/01/2020

Inclusion of publications statement

UNSW is supportive of candidates publishing their research results during their candidature as detailed in the UNSW Thesis Examination Procedure.

Publications can be used in their thesis in lieu of a Chapter if: • The candidate contributed greater than 50% of the content in the publication and is the “primary author”, ie. the candidate was responsible primarily for the planning, execution and preparation of the work for publication • The candidate has approval to include the publication in their thesis in lieu of a Chapter from their supervisor and Postgraduate Coordinator. • The publication is not subject to any obligations or contractual agreements with a third party that would constrain its inclusion in the thesis

Please indicate whether this thesis contains published material or not: This thesis contains no publications, either published or submitted for ☐ publication (if this box is checked, you may delete all the material on page 2) Some of the work described in this thesis has been published and it has been documented in the relevant Chapters with acknowledgement

☒ (if this box is checked, you may delete all the material on page 2)

This thesis has publications (either published or submitted for publication) ☐ incorporated into it in lieu of a chapter and the details are presented below

CANDIDATE’S DECLARATION I declare that: • I have complied with the UNSW Thesis Examination Procedure • where I have used a publication in lieu of a Chapter, the listed publication(s) below meet(s) the requirements to be included in the thesis. Candidate’s Name Signature Date (dd/mm/yy) Sebastian Vadillo Gonzalez 15/01/2020

iii

‘I hereby grant the University of New South Wales or its agents a non- exclusive licence to archive and to make available (including to members of the public) my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known. I acknowledge that I retain all intellectual property rights which subsist in my thesis or dissertation, such as copyright and patent rights, subject to applicable law. I also retain the right to use all or part of my thesis or dissertation in future works (such as articles or books).’

‘For any substantial portions of copyright material used in this thesis, written permission for use has been obtained, or the copyright material is removed from the final public version of the thesis.’

Signed ……………………………………………......

Date ……10/03/2021………………………………………......

AUTHENTICITY STATEMENT ‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis.’

Signed ……………………………………………......

Date …10/03/2021…………………………………………......

ACKNOWLEDGMENTS

From the moment I came to the UNSW, “contrast” has been the word to describe every day, every feeling and every lesson learned during this path called Ph.D. Everything began as a contrast between what I knew back in Mexico City to what had to be known in a place like Sydney. The food to be had every day, the transport that had to be taken and of course the new skills I had to learn. Everything was different and I had to step up from everything I knew before. In this journey filled with contrasts, I want to thank the people which in my first days in Australia guided me and provided the first taste of a new life. My supervisors of course receive my most sincere acknowledgments: the sweet and always trustworthy Katie Dafforn; the busy and truly amazing Emma Johnston; and the patient and always supportive Paul Gribben. Through their support, I was able to get a scholarship from The Bushell Foundation through the Sydney Institute of Marine Science and have ongoing financial stability through the Australian Research Council Future Fellowship Scheme. Without their help during the beginning, the massive experiment in port Stephens, the fieldwork in New Zealand, the data processing and analysis, and finally the long writing and review process; I wouldn’t have been able to survive or reached this point. I am deeply grateful for everything they did for and with me.

In the AMEE lab and the current marine floor in E26, many people that are still here struggling on their journey or were once my examples to follow, also deserve recognition. Be for help in analysis, as volunteers, counselors, or moral support; the following people deserve my acknowledgment for this work and it’s a shame I cannot account for everyone. From the AMEE lab, I want to thank my guru and predecessor Simone Birrer, my fellow latin American mentors Mariana Mayer Pinto and Ana Bugnot (former AMEE), the jiu-jitsu master Mark Browne and all other friends that have been in the lab during all these years. Too many friends to be mentioned here have helped me through my Ph.D. however I would like to mention a few who have acted like important volunteers and have supported this project. Special thanks to Rosie Steinberg for helping me always and for her efforts in collecting 360 cockles in one single day; Miguel Mazkiaran Ramirez (my soul brother and Mexican housemate) who helped in the same v ordeal; Wayne O’Connor my unofficial supervisor in Port Stephens, Giulia Filippini my bioinformatic confident and a great friend; Stephanie J. Connor for her unique but super cool support in the lands of the Kiwis; and finally to Jiawei Xu who opened the world of who I really was even if not directly related to my academic life. Also, I want to extend special acknowledgments to everybody involved in the multi-university project from the Sustainable seas initiative in New Zealand: ‘Tipping points in ecosystem structure, function and services’ which I was very happy to collaborate in. Direct people that need my sincere thanks include Simon Thrush, Teri O’ Meara, Kaiwen Yang, Stefano Schenone and Boyd Taylor for permitting me to have my base in the beautiful Leigh lab in Goat Island. In addition, I thank Rebecca Gladstone-Gallagher and her team for all her support in Waikato and for supporting me with the data for Chapter 4. My sincere respect and gratitude to everybody involved in this large scale project involved at every level of analysis, data processing, and experimental setupAnd of course, finally a special mention needs to be done to the 60 Sydney cockles who gave their life during experiments in this thesis. Ph.D. was one of the most difficult things I have done in my life beyond the academic. It taught me valuable lessons in every aspect of my life and I’m proud to say I have walked this path and endured. Definitely one of the most important experiences I have had but also the beginning of the rest of my life.

PUBLICATIONS AND PRESENTATIONS ARISING FROM THIS THESIS

Publications

1. Vadillo Gonzalez, S., Johnston, E., Gribben, P.E. and Dafforn, K., (2019). The application of bioturbators for aquatic bioremediation: Review and meta-analysis. Environmental pollution. 250, pp. 426-436.

Conference presentations

1. Oral presentation “The application of bioturbation to sediment bioremediation: a review and meta-analysis” on the 3-6th September 2018 Estuarine and Coastal Sciences Association (ECSA 57). Changing estuaries coasts and shelf systems. Diverse threats and opportunities held at Perth, Western Australia, Australia.

2. Oral presentation “Analyzing the potential of bioturbation for sediment bioremediation: Identifying knowledge gaps and mechanisms for its optimization” on the 20-22th May 2019 3rd Australia and New Zealand Marine Biotechnology Society Conference held at the University of New South Wales, Sydney, New South Wales, Australia.

3. Oral presentation “Application of macrofauna bioturbation for contaminant bioremediation in sediments: a review and meta-analysis” on the 7-10th July 2019 Society of Environmental Toxicology and Chemistry Australasia 2019 held at Darwin, Northern Territory, Australia.

4. Digital poster presentation “Ontogenetic differences of the Sydney cockle’s survivorship, mobility and impact on microbial communities in eutrophic sediments” on the 7-10th July 2019 Society of Environmental Toxicology and Chemistry Australasia 2019 held at Darwin, Northern Territory, Australia.

5. Oral presentation “Sediment bioremediation using benthic macrofauna: knowledge gaps and application in coastal environments” on the 30th October to 1st November 2019, 28th Annual NSW Coastal Conference held at Terrigal, New South Wales, Australia.

6. Oral presentation “Investigating the microbial communities in sediment bioturbator burrows and links to nutrient cycling” on the 25th to 28th November 2019 Environmental DNA Conference and workshop held at the University of Otago, Dunedin, New Zealand.

vii GENERAL ABSTRACT

Estuaries and coastal intertidal environments supply important ecosystem services and resources. Within these systems, sediments play a crucial role in processing organic inputs and providing nutrients for local food webs. Benthic macrofauna and sediment microbial communities participate actively in these processes and can interact to influence overall ecosystem function. Through different bioturbation processes, macrofauna can affect biogeochemical cycling in sediments and influence microbial communities. However, anthropogenic derived eutrophication has increased globally in recent years and has been found to negatively impact these systems. Excess inputs of anthropogenic contaminants can disrupt key sediment processes involving macrofauna and microbial communities. Current options for sediment remediation are costly and may result in further impacts on communities. An alternative method for sediment bioremediation involves the application of macrofaunal bioturbators to stimulate microbial processes including contaminant removal. However, to confidently apply such an alternative requires clarity on the effect of macrofauna on contaminants, the contaminant concentrations at which macrofauna will actively affect the sediments, the potential effects of bioturbators on microbial communities, and whether these applications can be applied in situ at large scales. Here I used a systematic review and meta-analysis to analyse current knowledge on the effect of bioturbating macrofauna on contaminants and quantify how this interaction changes depending on relevant biotic and abiotic variables. Secondly, I used an experimental approach with sediment mesocosms to further explore the contaminant thresholds for effective remediation by macrofauna bioturbation and interactions with microbial communities in highly enriched sediments. Finally, a large-scale field experiment was done to explore the influence in situ nitrogen enrichment on microbial communities in the sediment and within macrofauna burrows under variable environmental contexts. The combined results here demonstrate that the effect of macrofauna bioturbation on stimulating bacterial communities in organically enriched sediments is highly taxa driven and additional factors such as macrofauna intraspecific variations in body size, temperature, animal density, magnitude of contamination and site specific organic matter content need to be considered for potential bioremediation and coastal management plans.

viii

ix TABLE OF CONTENTS

List of figures xvi

List of Tables xxvii

Introduction 1

Bioturbation, microbial communities, and ecosystem 2 function

The Global Issue of Eutrophication 4

Macrofauna bioremediation: a new alternative for 5 reducing eutrophication in aquatic systems

Thesis outline 7

1. Application of bioturbators for aquatic bioremediation: 11 review and meta-analysis

1.1. Abstract 11

1.2. Introduction 12

1.3. Methods 14

1.3.1. Search methodology 14

1.3.2. Qualitative analysis 15

1.3.3. Design criteria for meta-analysis 15

1.3.4. Meta-analyses of the data 16

1.4. Results 17

1.4.1. Qualitative assessment of contaminant types in 17 sediments tested with bioremediation

1.4.2. Meta-analysis of bioturbator effects on selected 18 contaminants

1.4.3. Meta-regression analysis of effects of moderator 20 variables

1.4.4. Publication bias 24

1.5. Discussion 24

1.5.1. Bioturbators promote nutrient release and oxygen uptake 27

1.5.2. Bioturbator effects on metals is limited to one taxonomic 28 group

1.5.3. Further research on bioturbation effect on PAHs fate in 29 sediments

1.5.4. Factors that modulate the effect of bioturbators 29

1.5.5. Future research directions 32

1.6. Conclusion 35

2. The influence of bioturbator intraspecific body size variation in sediment bioremediation: effects on 37 survivorship and mobility of a sediment bioremediator in organically enriched sediments

2.1. Abstract 37

2.2. Introduction 38

2.3. Methods 41

2.3.1. Study species and collection 41

2.3.2. Sediment collection and mesocosm enrichment setup 42

2.3.3. Mesocosm setup and experimental design 43

2.3.4. Quantifying mortality, mobility and changes in organic matter 45 content xi 2.3.5. Statistical analysis 46

2.4. Results 48

2.4.1. Lethal effects of enrichment vary between body size 48 combinations

2.4.2. Cockle mobility decreased by enrichment in large cockles 48

2.4.3. Effect of different cockle size treatments on sediment organic 53 matter content decrease

2.5. Discussion 53

2.5.1. Cockle mortality in enriched sediments is lower in large 54 organisms

2.5.2. Associational susceptibility between small and large 55 phenotypes in enriched sediments

2.5.3. Intraspecific variation in lateral mobility in enriched sediments 57

2.5.4. Large cockle’s lateral movement increases in higher 57 intraspecific diversity

2.5.5. Organic matter loss was not dependant on the presence of the 58 Sydney cockle

2.6. Conclusion 60

3. Organic enrichment reduces microbial diversity and changes community structure, but these effects are not 63 mitigated by the Sydney cockle (Anadara trapezia)

3.1. Abstract 63

3.2. Introduction 64

3.3. Methods 68

3.3.1. Study species and collection 68

3.3.2. Mesocosm experimental enrichment setup 69

xii

3.3.3. Sample collection, DNA extraction and sequencing 70

3.3.4. Bioinformatics sequence analysis 71

3.3.5. Statistical analysis 72

3.4. Results 74

3.4.1. Lower microbial α-diversity indices in enriched sediments 74

3.4.2. Bacterial composition differed between natural and enriched 77 conditions

3.4.3. Archaeal composition differed between natural and enriched 86 conditions

3.5. Discussion 91

3.5.1. The Sydney cockle had no influence on sediment microbial 92 communities

3.5.2. Sediment enrichment decreased bacterial α-diversity and 93 changed community composition

3.5.3. Sediment enrichment reduces archaeal α-diversity and 96 changes community composition

3.6. Conclusion 97

4. Linking bacterial community shifts to nitrogen enrichment and sediment characteristics in macrofauna 99 burrows

4.1. Abstract 99

4.2. Introduction 100

4.3. Methods 102

4.3.1. Site selection, treatment preparation and sample collection 102

4.3.2. DNA extraction, amplification and sequencing 104

4.3.3. Bioinformatic sequence analysis 105 xiii 4.3.4. Statistical analysis 107

4.4. Results 110

4.4.1. Changes in alpha diversity are driven by sediment position 110

4.4.2. Community composition dissimilarity 113

4.4.3. Phylum composition analysis and specific phyla sensitivity to 113 environmental covariates

4.4.4. Community structure dissimilarities explained by selected 118 genera

4.5. Discussion 124

4.5.1. Macrofauna may create low oxygen bacterial microniches in 125 burrows

4.5.2. Organic matter as an important factor in burrow microniches 126

4.5.3. The role of sediment physical characteristics in shaping 129 bacterial microniches in burrows

4.5.4. Alternative explanations for the null effect of in situ nitrogen 130 enrichment on bacterial communities

4.6. Conclusion 131

5. General discussion 133

5.1.Macrofauna sediment bioremediation: advantages 135 and drawbacks

5.1.1.Spatial variability and site-specific factors 135

5.1.2. Importance of taxa selection: inter- and intraspecific species 137 differences in bioremediation

5.1.3. Animal density and inter and intraspecific associational 139 interactions

5.2. Final remarks and future directions 140

xiv

6. References 145

Appendix 1: Application of bioturbators for aquatic bioremediation: review and meta-analysis 195

Appendix 2: The influence of bioturbator intraspecific body size variation in sediment bioremediation: effects on survivorship and mobility of a sediment bioremediator 231 in organically enriched sediments

Appendix 3: Organic enrichment reduces microbial diversity and changes community structure, but these effects are not mitigated by the Sydney cockle (Anadara 241 trapezia)

Appendix 4: Linking bacterial community shifts to nitrogen enrichment and sediment characteristics in 255 macrofauna burrows

xv LIST OF FIGURES

1. Application of bioturbators for aquatic bioremediation: review and meta-analysis

Fig.1.1. Frequency diagram of main aquatic systems where the 19 effect of bioturbation on contaminant fate were studied.

Fig. 1.2. Frequency diagram of the main taxonomic groups where 19 the effect of bioturbation on contaminant fate were studied.

Fig. 1.3. Overall effect size of the release of evaluated responses from sediments to the overlying water and sediment oxygen 21 uptake (SOU) in the presence of bioturbators.

Fig.1.4. Overall effect size of the release of evaluated contaminants from sediments to the overlying water and sediment oxygen 22 uptake in the presence of bioturbators and between different taxonomic groups.

Fig.1.5. Overall effect size of the evaluated responses in the presence of bioturbators and between different aquatic 23 systems.

Fig. 1.6. Scatter plot matrix showing the meta-regression between the obtained effect size (SMD) of each response variable and 25 moderator variables obtained from the studies

Fig. 1.7. Funnel plots assessing publication bias on evaluated metanalysis: A) metal release, B) Ammonia fluxes, C) 26 Phosphorous fluxes, D) Sediment Oxygen Uptake (SOU).

xvi

2. The influence of bioturbator intraspecific body size variation in sediment bioremediation: effects on survivorship and mobility of a sediment bioremediator in organically enriched sediments

Fig.2.1. Overall cockle mortality (%) between different cockle size treatments exposed to 16 days in enriched sediments only (14% OM) (A). In addition, cockle mortality (%) of only small (B), medium (C) and large cockles (D). Data are presented as 51 mean % ± SE. In comparisons where significant differences occurred, treatments with a common letter do not differ significantly. Significant differences are shown with different letters. Abbreviations: S = Small, M= Medium and L= Large.

Fig.2.2. Cockle mean lateral mobility (cm) evaluated as the total distance travelled during a 16 day exposure to natural (8% OM) and enriched (14% OM) sediments between all sizes (A,B) and analysed separately with only small (C, sediment type data separated between body size combinations due to the presence of an interaction), medium (D,E) and large cockles (F,G) included. All data are expressed as mean ± SE. 52 Models to calculate these differences included biomass loss as an offset covariate to correct by high mortality in enriched treatments. Different letters represent significant differences between body size combinations within each sediment type. Asterisks indicate differences in cockle lateral movement between natural and enriched sediments. Abbreviations: S= Small, M= Medium and L= Large.

Fig.2.3. Sediment organic matter percent decrease (% ± SE) during a 16-day exposure to A) natural (8% OM) and B) enriched (14% OM) sediments. Asterisks indicate differences of organic 54 matter breakdown in cockle body size treatments between natural and enriched sediments. Abbreviations: S = Small, M= Medium and L= Large.

xvii

3. Organic enrichment reduces microbial diversity and changes community structure, but these effects are not mitigated by the Sydney cockle (Anadara trapezia)

Fig. 3.1. α-diversity indices for A) Bacterial zOTU richness, B) archaeal zOTU richness, C) bacterial zOTU diversity, D) archaeal zOTU diversity, E) bacterial zOTU evenness and F) 76 archaeal zOTU evenness between natural (8% OM) and enriched (14%OM) treatments. Data is expressed as mean ± SD. Significant differences between sediment types within each index is shown with an asterisk.

Fig. 3.2. Metric multidimensional scaling ordination (PCA, dimensions= 2) where, coloured outlines denote discrete 79 enrichment groups and labels indicate cockle treatment categories for each point (NC= No cockles, S=Small, M= Medium, L=Large and SML= Small + Medium + Large).

Fig. 3.3. Relative abundances (%) of bacterial phyla in natural and enriched sediments (all phyla with smaller than 5 % of 80 abundance where grouped in ‘Others’, see Appendix 3, Table C.2 for all bacterial phyla registered).

Fig. 3.4. Relative abundance of identified genera with >1% contribution to overall community dissimilarity between natural and enriched sediments. All genera with higher 85 relative abundance in enriched or natural sediments where found to be significant (p<0.001). All relative abundance data (%) is expressed as means ± SE.

Fig. 3.5. Archaea composition analysis between natural and enriched sediment treatments .Metric multidimensional scaling ordination (PCA, dimensions= 2, stress= 0.05) where, 87 coloured outlines denote discrete enrichment groups and labels indicate cockle treatment categories for each point (NC= No cockles, S=Small, M= Medium, L=Large and SML= Small + Medium + Large)..

xviii

Fig. 3.6. Archaea composition analysis between natural and 88 enriched sediment treatments. Relative abundances (%) of archaea phyla in natural and enriched conditions.

Fig. 3.7. Relative abundance of identified archaea classes that contributed to overall community dissimilarity between natural and enriched sediments. Significant increase in class 90 relative abundance between enrichment treatments is marked with an Asterix (8). All relative abundance data (%) is expressed as means ± SE.

4. Linking bacterial community shifts to background nutrients and organic enrichment in intertidal sediments

Fig. 4.1. A) Map of New Zealand from which a close up (B) of the North Island and selected sites are marked with red triangles. C) Expanded map of 3 sites in the Northland region (Whangarei: Onerahai (WGR-O), Parua bay (WGR-P) and Takahiwai (WGR-T)) and 4 sites in the Auckland region 106 (Whangateau (WTA) and Mahurangi: Lagoon bay (MAH-L) and Mandaley Bay (MAH-M). D) Expanded map of 3 sites in Waikato region (Raglan (RAG) and Whitianga: lower section (WHI-L) and upper section (WHI-U)) and 1 site in the Bay of Plenty region (Tauranga: Tuapiro (TAU-T)). Specific GPS location can be seen in Appendix 4, Table D1.

Fig.4.2. Differences in alpha diversity indices of bacterial communities sampled from two nitrogen treatments (Control and High; richness: A, Shannon diversity: C and Pielou Evenness: E) and in two different sediment positions 112 (Surface and burrows: richness: B, Shannon diversity: D and Pielou Evenness: F). Data is shown as mean ± SE. Asterisks (“*”) indicate significant (p < 0.05) lower alpha diversity values between factor levels.

Fig. 4.3. Bacterial composition multivariate analysis (Bray-Curtis dissimilarity index) between A) nitrogen treatments (control 115 and high) and B) sediment position (burrows and surface). Non-parametric multidimensional scaling ordination xix (stress= 0.15, k=2) where polygons and points represent grouping levels of fixed factors. Vectors represent fitted environmental covariates scaled by their correlation with the community composition from the grouping factor centroid. Here stronger predictors have longer arrows.

Fig. 4.4. Relative abundance (%) of selected bacterial phyla (relative abundance >1%) between A) nitrogen treatments (Control and high) and B) sediment position (Burrows and surface). 116 Relative abundance data is shown as mean ± SE. Significant differences between fixed factor levels are denoted with an asterisk “*” and detailed in Table 4.3 and and multiple pairwise comparisons are given in Appendix 4, Table D6.

Fig. 4.5. Relative abundance (%) of selected identified and unclassified genera (>1% contribution to community structure dissimilarity) between A) nitrogen treatments (Control and high) and B) sediment position (Burrows and 121 surface). Relative abundance data is shown as mean ± SE. Significant differences between fixed factor levels are marked with an asterisk “*”, and detailed in Appendix 4 Table D13 and multiple pairwise comparisons are given in Appendix 4, Table D14.

A. Appendix 1

Fig. A.1. Percentage of the main contaminants found in peer 195 reviewed scientific articles where contaminant-bioturbation interaction is analysed.

Fig.A.2. Flow diagram of the specific search, selection and article 196 exclusion for metal systematic analysis

Fig.A.3. Flow diagram of the specific search, selection and article 197 exclusion for organic matter/ nutrients systematic analysis

Fig.A.4. Flow diagram of the specific search, selection and article 198 exclusion for Polycyclic Aromatic Hydrocarbons (PAH) systematic analysis

Fig. A.5. Forest plot of the effect size (Standard Mean Difference, SMD) in the selected studies where the interaction between 199 bioturbators and metal release to overlying water was evaluated.

Fig. A.6. Forest plot of the effect size (Standard Mean Difference, SMD) in the selected studies where the interaction between 200 bioturbators and ammonia fluxes (NH4+) in sediment to the overlying water was evaluated.

Fig. A.7. Forest plot of the effect size (Standard Mean Difference, SMD) in the selected studies where the interaction between 201 bioturbators and phosphorous fluxes in sediment to the overlying water was evaluated

Fig. A.8. Forest plot of the effect size (Standard Mean Difference, SMD) in the selected studies where the interaction between 202 bioturbators and sediment oxygen uptake (SOU) was evaluated.

Fig.A.9. Main countries found in the systematic review where the 203 effect of bioturbation and metals, organic matter/nutrients and PAHs was evaluated

Attached publication of Chapter 1 results acknowledged throughout 220 the thesis as Vadillo-Gonzalez et al. 2019.

B. Appendix 2

Fig.B.1. Summary of the experimental design developed for the main experiment for two enrichment treatments (natural or 231 enriched sediments) crossed with 8 cockle single and mixed body size treatments. Total mesocosm number = 70).

xxi Fig. B2. Image processing graphic protocol to describe calculation of cockle mobility using the Image J 1.x. A) A scale in pixels is measured with the photograph using the known diameter of the mesocosm (i.e. 25 cm). As distance change slightly with every daily photo an average of this distance in pixels was 233 calculated. B)After establishing the scale, a straight line from the center of each individual cockle was measured from Day 0 to its new position next day. This was done for each cockle for the 16 days of experiment. From this data, accumulated mobility per cockle in each mesocosm was obtained in both natural and enriched treatments.

Fig. B.3. Pearson correlation of A) cockle mortality (%) and B) cockle lateral movement (cm/16 days) and the percentage of organic matter decrease in sediments in natural and 234 sediments exposed to 16 days of enrichment. Coloured area around regression line correspond to calculated confidence intervals.

Fig.B.4. General Mixed model diagnostics (QQ plots and residual vs fitted plots) to determine differences in percent cockle 235 mortality between body size combinations in two sediment enrichment types.

Fig.B5. General Mixed model diagnostics (QQ plots and residual vs fitted plots) to determine differences in cockle lateral 236 movement between cockle body size combinations in two sediment enrichment types.

Fig.B6. General Mixed model diagnostics (QQ plots and residual vs fitted plots) to determine differences in sediment organic 237 matter decrease between cockle body size combinations in two sediment enrichment types.

xxii

C. Appendix 3

Fig.C.1. Summary of the experimental design developed for the experiment for two enrichment treatments (natural and 241 enriched sediments) crossed with 3 monocultures (S, M and L), 1 mixed body size combination (SML) and 1 control group (NC, no cockles). Total mesocosm number = 28).

Fig. C.2. Rarefaction curves obtained from A) Bacterial and B) 242 Archaea zOTUs table of abundance. Fig.C3. General linear model diagnostics (residual plots and normality QQ plots) for bacterial alpha diversity indices compared with different sediment types (natural vs. 245 enriched), cockle body size treatments (i.e. No cockles, small, medium, large and small+medium+large) and interaction of both variables. Fig.C4. General linear model diagnostics (residual plots and normality QQ plots) for archaeal alpha diversity indices compared with different sediment types (natural vs. 246 enriched), cockle body size treatments (i.e. No cockles, small, medium, large and small+medium+large) and interaction of both variables. Fig.C5. General linear model diagnostics (residual plots and normality QQ plots) for main bacterial phyla compared with different sediment types (natural vs. enriched), cockle body 248 size treatments (i.e. No cockles, small, medium, large and small+medium+large) and interaction of both variables Fig.C6. General linear model diagnostics (residual plots and normality QQ plots) for identified genera (>1% contribution 251 to overall community dissimilarity) compared with natural and enriched sediments. Fig.C7. General linear model diagnostics (residual plots and normality QQ plots) for main archaeal phyla compared with different sediment types (natural vs. enriched), cockle body 253 size treatments (i.e. No cockles, small, medium, large and small+medium+large) and interaction of both variables. xxiii Fig.C8. General linear model diagnostics (residual plots and normality QQ plots) for identified archaeal classes that 254 contributed to overall community dissimilarity between natural and enriched sediments.

D. Appendix 4

Fig. D1. Difference in porewater ammonia concentration in sediments within control (no nitrogen addition) and high nitrogen addition (600g N/ m2). A previous statistical analysis (general linear mixed model) was done with 255 nitrogen treatment as fixed factor and plot as a random factor. Main effects inferences were done with a Wald Chi test (χ2=215.5, df=1 and p<0.001). Data is shown as mean ± SE. Asterisks (“*”) indicate significant (p < 0.05) lower porewater ammonia concentration.

Fig. D2. Arrangement of N enriched plots within a sampling site. Circles represent the area where the plots were placed and 256 coloured squares the different levels of enrichment within an experimental area. Diagram modified from Rebecca Gallagher’s personal notes.

Fig. D3. Rarefaction curves of bacterial community samples in A) macrofauna burrows and B) surface sediment. Blue lines 262 indicate curves from bacterial communities with natural levels of N (i.e. control) and red lines from communities with a high N enrichment treatment

Fig. D4. Model diagnostics for alpha diversity index analysis. For all indices, QQ normality plots, residual plots (Fitted vs observed residuals, and Fitted vs Studentized residuals) are shown. For models where plot was not found to be a 263 significant random factor and a GLM approach was done (A), residual vs leverage plots and Cook’s distance plots are also shown

xxiv

Fig. D5. Model diagnostics for the GLM of Phylum composition analysis. QQ normality plots, residual plots (fitted vs observed residuals, and fitted vs studentized residuals) are 264 shown. As plot was not found to be a significant random factor, a GLM approach was used and a residual vs leverage plot and Cook’s distance were included

Fig. D6. Model diagnostics for the GLM of SIMPER genera community dissimilarity. QQ normality plots, residual plots (fitted vs observed residuals, and fitted vs studentized 265 residuals) are shown. As plot was not found to be a significant random factor, a GLM approach was used and a residual vs leverage plot and Cook’s distance were included.

Fig. D7. Planned comparative Pearson correlations of selected covariates and significant fixed factors to determine the 268 effect of A) Organic matter content (%) and B) Mud content (%) on alpha diversity indices (Richness, Shannon diversity and Pielou index).

Fig. D8. Pearson correlations of the effect of organic matter content (%) on bacterial relative abundance between burrows and 274 surface sediments. Phyla are arranged from higher to lowest relative abundance.

Fig. D9. Pearson correlations of the effect of median grain size (µm) on bacterial main phyla’s relative abundance between 276 burrows and surface sediments. Phyla are arranged from higher to lowest relative abundance

Fig. D10. Pearson correlations of the effect of chlorophyll 278 concentration (µg/g) on bacterial main phyla’s relative abundance.

Fig. D11. Pearson correlations of the effect of organic matter 290 content (%) on selected identified and unclassified genera that described community dissimilarity.

xxv Fig. D12. Pearson correlations of the effect of mud content (%) on 291 selected identified and unclassified genera that described community dissimilarity.

Fig. D13. Pipeline comparison of rarefaction curves obtained through the DADA2 algorithm (A and C) and the original method using Usearch-Unoise (B and D). A and C show 292 rarefaction curves in surface sediments and B and C in macrofauna burrows. Red lines show samples obtained from high nitrogen conditions and blue lines show control conditions (no nitrogen added.

Fig. D14.Pipeline comparison of alfa diversity indices (richness, 293 Shannon diversity and Pielou evenness) grouped by nitrogen treatments (control and high).

Fig. D15.Pipeline comparison of alfa diversity indices (richness, Shannon diversity and Pielou evenness) grouped by 299 sediment position (Surface and macrofauna burrows). Significant differences are marked with an asterisk “*”.

xxvi

LIST OF TABLES

2. The influence of bioturbator intraspecific body size variation in sediment bioremediation: effects on survivorship and mobility of a sediment bioremediator in organically enriched sediments

Table 2.1. Summary of general linear mixed models used to investigate a) percent cockle mortality, b) lateral mobility and c) organic matter breakdown in two sediment types (ST, natural and enriched) and in different body size 50 combinations (BSC, small, medium, large, S+M, S+L, M+L and S+M+L). Results from the interaction between ST and BSC are also included. Significant differences are shown in bold.

3. Organic enrichment reduces microbial diversity and changes community structure, but these effects are not mitigated by the Sydney cockle (Anadara trapezia)

Table 3.1. Summary of general linear models (GLMs) in which sediment type (natural and enriched) and cockle treatments (No cockles, small, medium, large, all sizes) 75 were evaluated against each α-diversity index from bacterial and archaeal communities. Significant differences are shown in bold (p<0.05).

Table 3.3. Summary of a) general linear model results evaluating 81 differences in relative abundance of bacterial and archaeal phyla between natural and enriched sediment treatments xxvii

and body size combinations. b) emmeans contrasts of of bacterial and archaeal phyla for both sediment types. Significant differences (p<0.05) are shown in bold and p values in contrasts have been adjusted for multiple testing using Bonferroni correction.

4. Linking bacterial community shifts to background nutrients and organic enrichment in intertidal sediments

xxviii

Table 4.3. Summary of the GLM to determine differences in relative abundance of selected bacterial phyla (>1% relative abundance) between each phylum (11 phyla), nitrogen 117 treatments (Control and High), sediment position (Burrow and Surface) and interaction therein. Significant results are shown in bold.

xxix

A. Appendix 1

Table A.1. List of consulted literature and meta-data used for the 204 systematic review and meta-analysis

Table A.2. Hierarchical multilevel model for the effect of presence/absence of bioturbation in different meta- 217 analytic responses: total metal concentration in overlying water and ammonia fluxes to water column.

Table A.3. Hierarchical multilevel model for the effect of presence/absence of bioturbation in four different meta- 218 analytic responses: phosphorous fluxes to water column and sediment Oxygen Uptake (SOU).

Table A.4. Meta-regression fixed to a single predictor model for analysing the correlation between additional moderator variables and the effect size (SDM) obtained through a multilevel model that evaluated the presence/absence of 219 bioturbators and their interaction with 4 response variables: total metal concentration in overlying water, ammonia fluxes to water column, phosphorous fluxes to water column and sediment Oxygen Uptake (SOU).

B. Appendix 2

xxx

Table B. 3. Calculated pairwise comparisons of overall cockle lateral mobility between different body size combinations. 238 Significant differences are shown with an asterisk and in bold. Abbreviations: S = Small, M= Medium and L= Large.

Table B.4. Calculated pairwise comparisons of overall cockle lateral mobility between different body size combinations. 239 Significant differences are shown with an asterisk and in bold. Abbreviations: S = Small, M= Medium and L= Large.

Table B.5. Calculated pairwise comparisons of percent organic matter breakdown between two sediment types (i.e. natural – enriched) and different body size combinations. 239 Significant differences are shown with an asterisk and in bold. Abbreviations: NC= No cockles, S = Small, M= Medium and L= Large.

C. Appendix 3

Table C1. Bacteria and Archaea α diversity and Good’s coverage data 243 for each mesocosm in natural or enriched sediments with different body size combinations

Modulibacteria, Nitrospinae, Omnitrophicaeota, PAUC34f, Rokubacteria, Schekmanbacteria, Synergistetes, WOR-1, WS1, WS2, WS4).

Table C3. Taxonomic description of main unidentified genera that contribute <1% to bacterial community composition 249 dissimilarities between natural and enriched sediment treatments.

Table C4. Percent contribution (<1%) of specific bacterial genera to 250 dissimilarities evaluated between communities in natural and enriched sediments

Table C5. Relative abundance (%) of main archaea phyla sampled in 252 natural and enriched sediments.

Table C.6. Taxonomic description of main identified classes that contribute (%) to archaeal community composition 252 dissimilarities between natural and enriched sediment treatments.

D. Appendix 4

xxxii

Mud content (%), MGS= median grain size (µm), Chl a= Chlorophyll a concentration (µg/ g), Phaeo= Phaeo pigment concentration (µg/g), Porosity (%), MS= Macrofauna abundance (no. of species) and PWA= Porewater ammonia (µmol N L-1).

Table D4. Summary of descriptive statistics of alpha diversity indices (bacterial richness, Shannon diversity index and Pielou 269 evenness) in fixed factor levels of A)nitrogen treatments and B) sediment position.

Table D5. Descriptive summary of the relative abundance (%) of main bacterial phyla selected for the analysis between nitrogen treatments (Control and high) and sediment position (Burrows and Surface). Descriptive statistics 270 include mean, standard deviation (SD), number of cases (n) and standard error (SE). Significant differences in relative abundance between factor level are marked in bold.

Table D6. Multiple comparisons using estimated marginal means to 271 determine significant differences of relative abundance between bacterial phyla within A) Nitrogen treatments xxxiii

and B) Sediment position. Significant contrasts within factors are marked in bold.

Table D8. Summary of Planned comparative correlations were the effect of organic matter content on bacterial relative abundance was evaluated between all selected phyla 273 within burrows and surface sediments. Significant correlations and differences in slopes are marked in bold.

Table D9. Summary of Planned comparative correlations were the effect of median grain size on bacterial relative abundance was evaluated between all selected phyla 275 within burrows and surface sediments. Significant correlations and differences in slopes are marked in bold.

Table D10. Summary of Planned comparative correlations were the effect of chlorophyll a concentration in sediment on bacterial relative abundance was evaluated between 277 selected phyla. Significant correlations are marked in bold.

Table D11. Summary of planned comparative correlations were the effect of chlorophyll a concentration on bacterial relative abundance was evaluated between selected phyla. 279 Significant differences in slope are marked in bold. All p values where corrected using Tukey HSD method.

Table D12. List of selected identified and unidentified bacterial 281 genera that contributed to dissimilarities between

xxxiv

communities in nitrogen treatments and sediment position. Only genera with >1% contribution where included. Percent contribution was calculated from average proportion of contribution calculated by the simper function.

Table D14. Multiple comparisons using estimated marginal means to determine significant differences of relative abundance between selected bacterial genera within 283 burrows and surface sediments. Significant contrasts within factors are marked in bold 284

Table D16. Summary of Planned comparative correlations were the effect of organic matter content and mud content on selected identified and unclassified genera relative 285 abundance was evaluated. Significant correlations are marked in bold

xxxv

INTRODUCTION

Microbial communities and macrobiota (i.e. plants and animals) are responsible for many of the processes that maintain ecosystem functioning (Salvucci, 2016). Both communities influence biogeochemical cycles at an ecosystem scale in soils and sediments. Much research has been done to examine these communities in soil systems as they are critical to agriculture and forestry (Buscot and Varma, 2005). From these studies, I know that microbes and macroinvertebrates can change everything from nutrient cycling rates to the formation of habitats for local microbiota (e.g. bacteria, archaea or protozoa) and mesofauna (i.e. microinvertebrates) through changes in physical and chemical characteristics (Moldenke et al. 2000; Wilkinson et al. 2009). Bioturbators such as millipedes, earthworms and beetles participate actively in organic matter degradation by shredding large or complex plant matter to make it available for microbial communities for further degradation or nutrition (Moldenke et al. 2000). In addition, their overall bioturbation activities (e.g. burrowing, soil particle ingestion, faecal pellet production) can impact soil physical structure and chemical properties (e.g. redox state, water retention, pH). Through all these activities, soil macroinvertebrates can transport microbial inocula to different soil depths and regions (Kuzyakov and Blagodaskaya, 2015), determine water, oxygen and organic matter retention that promotes microbial growth (Wardle et al. 2006), that enhances soil metabolic diversity by the transport of microbial inocula and increasing microbial taxa with different enzymatic capacities (Moldenke et al. 2000). Similar to soil systems, coastal, marine, and estuarine sedimentary environments present microbial and macrofaunal communities that are known to regulate many biogeochemical cycles including nitrogen cycling. However, fewer studies have explored in detail the characteristics of each of these biotic compartments and possible interactions between them in benthic systems (Kougure and Wada, 2005; Laverock et al. 2011; Moulton et al. 2016).

Coastal marine environments and estuaries are some of the most productive natural habitats worldwide. They provide nutrients and refuge to local and migratory biota (Mclusky and Elliot, 2004), and ecosystem services such as water quality regulation and food production (Deng et al. 2015). Within these

1 INTRODUCTION systems, inputs of sediments and organic matter (i.e. plant or animal-based) are deposited regularly from multiple sources such as rivers, groundwater, land and sea (Mclusky and Elliot, 2004; Arndt et al. 2013). Organic inputs can remain as suspended particles in the water column or be deposited in the soft sediment benthos where they are available for benthic communities (Kang et al. 2017). Specifically, benthic microbial communities (including bacteria and archaea) can support primary production by oxidizing compounds in the deposited organic material to release nutrients to the overlying water column (Herbert, 1999). Sediment microbial, meiofaunal and microphytobenthic communities drive biogeochemical cycling including aerobic (e.g. ammonification, nitrification, sulphide oxidation) and anaerobic (e.g. denitrification and sulphur reduction) processes that regulate nitrogen and carbon dynamics (Petro et al., 2017). Microphytobenthic communities play a major in the control of primary production by creating nutrient sinks in the sediment surface and influencing denitrification and inorganic nitrogen rates (Sundback et al. 2000). For meiofaunal benthic communities (invertebrates smaller than 1 mm), these play an important role in nitrogen metabolism (e.g. denitrification) and nutrient processing through micro-bioturbating activities sometimes coupled with microbes or macrofauna (Bonaglia et al. 2014). Finally,benthic macrofauna plays a direct role in organic matter degradation by enhancing organic matter transport, breakdown, and nutrient release into the overlying water column and sediments where it becomes bioavailable for microbial and meiofaunal communities. (Glud et al. 2016). In this thesis, attention will be given to microbial and macrofauna communities but the essential role of meiofaunal and microphytobenthic communities needs to be acknowledged.

Sediment macrofauna bioturbation consists of a complex series of physiological and behavioral processes in which the animal changes the properties of surrounding sediment matrices and influences microbial communities (Kristensen et al. 2012). These include particle reworking (e.g. burrowing and moving on the sediment surface), bioirrigation (e.g. introduction of oxygen to sediments through ventilatory mechanisms) the release of organic

INTRODUCTION compounds through secretions or particle processing through digestion (e.g. mucus or faeces) (Kristensen et al. 2012; Dale et al. 2018; Dale et al. 2019). Through these activities, bioturbators can accumulate, transport, resuspend and regulate concentrations of organic matter and nutrients in the sediment surface as well as mobilize solutes to deeper layers (Volkenborn et al. 2007; Nicholaus et al. 2014). In addition, processes such as bioirrigation produce important oxygen oscillations in local redox conditions in the sediment and affect overall benthic metabolism (Bergström et al. 2017). Through bioturbation, macrofauna can have a direct effect on microbial communities by influencing changes in their community structure and composition (e.g change abundance of nitrifying bacteria), regulating available quantity and quality of organic matter and changing physicochemical sediment properties to favor specific microniches (Foshtomi et al. 2015; Shen et al. 2017). Examples of benthic biogeochemical pathways where macrofauna bioturbation couple with microbial communities include aerobic nitrogen and sulphur pathways (e.g. ammonification, nitrification and sulphur oxidation) and aerobic pathways (e.g. denitrification and sulphur reduction) (Foshtomi et al. 2015).

The effects of bioturbators and-microbial activities on ecosystem function and nutrient dynamics in coastal ecosystems has received some attention (Kristensen et al. 1987; Clavero et al. 1994; Holmer et al. 2008; Foshtomi et al. 2015; Glud et al. 2016; Li et al. 2019) and many environmental and biotic factors have been found to influence the outcome of this interaction (Braeckman et al. 2010; Dolbeth et al. 2019). Some of these factors include bioturbator taxonomic identity (e.g. bioturbation mechanism; see Gerino et al. 2003), density, quality of the organic matter input, sediment type and intra-specific characteristics of bioturbators (Geta et al. 2004; Karlson et al. 2007; Lohrer et al. 2010; Gilbertson et al. 2012; Sciberras et al. 2017; Douglas et al. 2018). Within these intra-specific characteristics, body size has been recognized as an important driver of bioturbator ecosystem function in soft sediment systems but not many studies have explored the interaction of this functional trait with sediment microbial communities (Clark et al. 2013; Norkko et al. 2013; Cozzoli et al. 2018). Larger

INTRODUCTION bioturbators have been found to have a greater impact on sediment properties compared to smaller organisms through higher bioirrigation rates at deeper sediment sections, increased nutrient fluxes (e.g. NO3- efflux), stronger pressure gradients in porewater and possibly providing higher surface area for microbial biofilm establishment (Norkko et al. 2013; Heisterkamp et al. 2013; Dale et al. 2018). Further research is needed to determine the impact of many of these biotic and abiotic factors on bioturbating macrofauna and sediment microbial communities, and their overall impact on ecosystem function.

Benthic ecosystems around the world provide a range of important ecosystem services that have been threatened or disturbed through a high input of contaminants (Johnston et al. 2015). Chemical compounds like herbicides, metals and hydrocarbons have historically affected many benthic systems through a constant discharge from municipal sewage, stormwater drains and farming/agriculture (Chapman and Wang, 2001; Johnston et al. 2015). However, eutrophication has also become a worldwide issue in many aquatic environments whereby a high concentration of nutrients results in problematic levels of primary production that have severe consequences for ecosystem health (Ansari et al. 2011).

Human derived organic inputs from farming, aquaculture, urban runoff and agriculture have caused eutrophication in many coastal systems affecting both water and sediment quality (Sinha et al. 2017; Li et al. 2018). In the sediments, excess nutrient and organic matter input can lead to the accumulation of organic matter in the sediment and a consequent increase in aerobic microbial activity. Sediment aerobic microbial activity will in turn increase the sediment oxygen uptake and deplete the sediment surface of oxygen (Wang et al. 2016). Hypoxic conditions in the sediment can become lethal for many benthic bioturbating macrofauna and other groups of microbial communities that rely on oxic conditions to survive (Biles et al. 2002; Altmann et al. 2004; Penha-Lopes et al. 2009; Gribben et al 2009; Gribben et al. 2017). In hypoxic conditions, anaerobic microbial activity is enhanced, and alternate routes of organic matter

INTRODUCTION degradation predominate. These include organic matter degradation by sulphate reducing bacteria (SRB) with the release of H2S to the water column (Skei et al. 2000; Mermillod-Blondin et al. 2004). The high concentration of nutrients that results from this enhanced microbial activity may further increase the primary productivity in the water column and lead to cyanobacteria or algal blooms (Sinha et al. 2017). As these populations increase, water turbidity increases, water column light availability is reduced, and high levels of dissolved organic carbon can produce blackwater events where hypoxic conditions may end in high fish and benthic macroinvertebrate mortality (King et al. 2012; Small et al. 2014; Norkko et al. 2019). The change in light regime is detrimental for many benthic metabolic processes and more importantly, increased algal populations will amplify the amount of organic matter deposition in the sediments, acting as a feedback loop that enhances hypoxic/anoxic environments in sediments (Skei et al. 2000).

The degradation of benthic systems by human-derived eutrophication can result in species loss and subsequent disruption of many of the ecosystem biogeochemical processes (Douglas et al. 2017). Organically enriched sediments have been reported to affect bioturbation activities (e.g. burrow building; Stamhuis et al. 1997) and can negatively influence bioturbator survival rate and behaviors (Osinga et al. 1997; Biles et al. 2002; Altmann et al. 2004; Penha-Lopes et al. 2009; Bartolini et al. 2009). Similar alterations in microbial communities can occur with changes in the overall sediment metabolism and production of deleterious compounds for meio and macrofauna. For example, changes in microbial communities can shift to anaerobic metabolism where sulphate reducing bacteria (SRB) increase production of sulfides (H2S) which are released to sediment porewater and the water column (Skei et al. 2000; Mermillod- Blondin et al. 2004). These deleterious effects of eutrophication to benthic communities are, however, highly variable in natural systems where biological and environmental covariates can change the system’s resilience to a stressor (Gladstone-Gallagher et al. 2019). Macrofauna species richness and the complexity of the interactions they form with other components of benthic

INTRODUCTION systems (i.e. meio and microbiota) is one of the major contributors to ecosystem resilience (Godbold and Solan, 2009). In addition, macrofauna’s variability in response to the differing magnitude of disruption can also explain a system’s capacity to recover after stress (Elmqvist et al. 2003). This characteristic of benthic systems is mainly driven by the diversity of functional traits each population has that permits them to tolerate and contribute to overall ecosystem function. For example, tolerant macrofauna such as some species of polychaetes and bivalves can withstand highly enriched sediments and promote microbial activities that sustain healthy nutrient cycling (Meksumpun et al. 1999; Carmicheal et al. 2012; Dafforn et al. 2013). The possibility of having such a range of functional traits in bioturbating macrofauna that can stimulate essential biogeochemical cycles and support resilience in a system leads to the idea of their deployment as a bioremediation strategy.

Eutrophication of coastal environments poses a risk to biodiversity and ecosystem services when systems are not able to recover to their natural state after a disruption. Under this scenario, the government, industry and the public have recognized the importance of introducing sediment remediation strategies into management plans (Khan et al. 2004). Management decisions to reduce eutrophication in many aquatic systems remain hard to judge as approaches for remediation and restoration provide variable results (Uusitalo et al. 2016). In many cases, reduction of the nutrient inputs seems to lead to a decrease in the negative impacts of ecosystem services including the once provided from benthic systems (Uusitalo et al. 2016). However, several systems with high anthropogenic disturbance remain in poor conditions and with low levels of resilience (Ansari et al. 2011). These highly disturbed systems are far from being capable of self- recovery and are in need of novel bioremediation approaches to regain ecosystem function. Many methods and technologies exist today for sediment remediation including on-site and off-site approaches such as physical (e.g. dredging, containment, thermal or electrokinetic), chemical (e.g. sediment stabilization or washing) or biological (Adriaens et al., 2006; Mesuer, 2012) remediation. Biological methods include the use of plants (e.g. phytoremediation, Ali et al.

INTRODUCTION

2013), fungi (Harms et al. 2011) and microbial inoculations (Vidali et al. 2001; Krzmarzick et al. 2018) for the treatment of pollutants such as metals and other organic compounds (e.g. Polycyclic aromatic hydrocarbons (PAHs)). In several cases, negative impacts such as loss of local biodiversity or changes in the microbial related biogeochemical process have occurred as a result of the remediation approach and limits to sediment remediation methods have been identified (Boopathy et al. 2000; Chapman et al. 2018). Among sediment remediation methods, bioremediation has been proposed as a potentially cost- effective nature-based alternative that has a less negative impact on the environment and does not introduce highly exogenous elements into an already disturbed system (Vidali, 2001; Meuser, 2013). Previously, bioremediation of eutrophic aquatic systems has concentrated on optimizing microbial inoculations to reduce hypoxia and increase organic matter breakdown (Wang et al. 2019; Shishir et al. 2019). However, many critical knowledge gaps remain regarding the long-term effects of such bioremediation approaches on local communities and the diverse range of responses reported for different contaminants (Maximov et al., 2015).

Under typical multiple-stressor circumstances, bioturbator- bioremediation has been proposed as a method that could potentially have the greatest effect on reducing contaminant bioavailability in aquatic systems (Robinson et al. 2015; Shen et al. 2016: Mandario et al. 2019). High diversity in tolerance or adaptation of bioturbating macrofauna to local eutrophic conditions is possible (Meksumpun et al. 1999: Dafforn et al. 2013) and some studies of sediment bioremediation have been attempted (Robinson et al. 2015; Ma et al. 2015; Shen et al. 2016; Mandario et al. 2019). In many cases, a shift in microbial communities in which reduction of anaerobic bacterial activity and the promotion of alternative aerobic pathways was achieved (Kristensen, 1987; Banta et al. 1999; Quintana et al. 2013; Bergström et al. 2017). However, the application of in situ sediment bioremediation using macrofauna in eutrophic systems has rarely made detailed evaluations of both the sediment microbiota and macrobiota (in situ or in mesocosm), as such a major knowledge gap still exists (Vadillo-Gonzalez et al.

INTRODUCTION

2019).

The aim of my thesis was to assess the potential use of bioturbating macrofauna for sediment remediation of eutrophic systems. In addition, this thesis also aimed to explore the coupling of bioturbators with microbial communities to assess their influence on key ecosystem functions. To do this, I used a broad analytical approach where sediment bioremediation with macrofauna was assessed through a literature review and meta-analysis, a mesocosm experiment, and a large-scale field experiment. In Chapter 1, I used a systematic review and meta-analysis to explore current knowledge of the activities of bioturbating macrofauna in contaminated sediments (i.e. metals, PAHs and organic matter/nutrients) and quantified how bioremediation affects change depending on the taxonomic group, the aquatic ecosystem and important environmental covariables. Results from this chapter will help inform environmental managers of the taxa and variables to consider in order to optimize local bioremediation methods in contaminated sediments. An associated publication was produced from the results of Chapter 1 and is acknowledged throughout the thesis as Vadillo-Gonzalez et al. 2019 (Environmental Pollution and Appendix 1). For Chapter 2, I explored how intraspecific body size variation could influence the survival and mobility of bioturbators as well as rates of sediment organic matter breakdown in experimentally enriched sediments. For this, I conducted a mesocosm experiment in which monocultures and all size combinations of three body sizes (small, medium and large) of the Sydney cockle (Anadara trapezia) were exposed to natural or organically enriched sediments. Results in this chapter help to understand the effects (i.e. lethal and sublethal) of enrichment on a possible model organism for sediment bioremediation. In Chapter 3, I explored the effect of different body sizes of the Sydney cockle (A. trapezia), from the mesocosm experiment in Chapter 2 on bacterial and archaeal richness, diversity, evenness and composition in natural and experimentally enriched sediments. This chapter tests for a link between sediment bioremediation activity and shifts in microbial communities in natural and organically enriched sediments. Finally, Chapter 4

INTRODUCTION presents a study in which experimental manipulations of nitrogen enrichment on microbial communities and bioturbators were explored in a large-scale field experiment across 10 sites on the North Island, New Zealand. Here changes in sediment bacterial communities in response to in situ porewater nitrogen enrichment were explored between sediment surfaces and macrofauna burrows. In addition, shifts in bacterial communities were compared with different environmental covariates such as organic matter (content % and algal biomass), sediment characteristics (porosity, median grain size and mud content %) and abundance of local macrofauna. The combined results of this thesis will increase our understanding of the impact of organic enrichment on benthic macrofauna and microbial communities and identify the knowledge gaps that exist if we are to optimize bioremediation strategies using sediment macrofauna.

INTRODUCTION

1. THE APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION: REVIEW AND 1 META-ANALYSIS 1.1. Abstract Human activities introduce significant contamination into aquatic systems that impact biodiversity and ecosystem function. Many contaminants accumulate, and remediation options. are now required worldwide. One method for bioremediation involves the application of macrofauna to stimulate microbial ecosystem processes including contaminant removal. However, if we are to confidently apply such a technique, we need clarity on the effect of bioturbators on different contaminants and how these vary under different environmental scenarios. Here I used a systematic review and meta-analysis to analyse current knowledge on the activities of bioturbating macrofauna in contaminated sediments and quantify how bioturbation-bioremediation changes depend on the taxonomic group, the aquatic ecosystem and important environmental variables. Three common contaminant classes were reviewed and analysed: metals, nutrients (i.e. ammonia and phosphorous) and polycyclic aromatic hydrocarbons (PAH). In addition, meta-regressions were calculated to estimate the effect of environmental and experimental design variables on effect sizes. Meta-analytic results revealed that deeper burrowing and more active sediment surface animals (e.g. polychaetes) increased metal release from sediments, nutrients and oxygen uptake by microbial fractions in comparison to bioturbators that inhabit shallower depths in sediments. In addition, there was a different effect of bioturbators on response variables in different aquatic systems. Finally, bioturbator effects on nutrient and metal release appeared modulated by context-specific variables such as temperature, pH, sediment grain size, animal density and experimental duration. Our findings highlight critical knowledge gaps such as field applications, less studied macrobenthic fauna and the incorporation of molecular approaches. Our results provide the first quantitative synthesis of the effects of bioturbators on contaminant fate and the variables that need to be considered for the optimization of this method as a viable approach for sediment remediation and contaminant management in aquatic systems.

11 CHAPTER 1 1.2. Introduction Human activities have increased the number and concentrations of contaminants entering natural aquatic ecosystems (Chapman and Wang 2001). Some of the most common contaminants include metals, hydrocarbon compounds and organic matter (Skei et al. 2000). These contaminants may have negative effects across multiple ecological levels and can interact with other environmental variables to cause synergistic impacts on aquatic ecosystem structure and function (Johnston and Roberts 2009; Johnston et al. 2015). For example, metals can be sequestered into sediments, chelate with organic matter affecting biogeochemical cycles or be absorbed by local organisms and transferred through the food chain to higher trophic levels (Tchounwou et al. 2012). In addition, high concentrations of hydrocarbons can directly influence the mortality of key species and be transferred into or magnified at other trophic levels (Grall and Chauvaud 2002). In both cases, the loss of important functional taxa can change nutrient dynamics in the system and lead to poor ecosystem health and eutrophication (Douglas et al. 2017). In these instances, environmental remediation often requires human intervention to remove or decrease the effect of contaminants and restore ecological function at degraded sites (Khan et al. 2004). However, remediation applications can themselves be damaging (Boopathy, 2000) and environmentally friendly solutions are needed to address the global problem of contamination.

Physical remediation (e.g. dredging) and chemical remediation (e.g. sediment washing) are the most common methods of removing contamination in many ecosystems (Adriaens et al. 2006). However, biological remediation (bioremediation) of contaminated sites has become an increasingly popular alternative due to its potential for cost effectiveness and lower environmental impact (Vidali 2001; Meuser, 2012). Bioremediation uses living organisms that interact directly or indirectly with the environment to neutralize or remove contaminants (Iwamoto and Nasu, 2001). Bioremediation can be applied both in situ (i.e. field conditions) and ex situ (i.e. mesocosm/controlled conditions).

THE APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION

Many small-scale examples exist of both approaches using plants, fungi and bacteria as bioremediators of organic contaminants, although results have been varied (Vidali 2001; Harms et al. 2011). However, to date there have been few applications of bioremediators at a large scale (Boopathy 2000; Adriaens et al. 2006; Ma et al. 2015).

An emerging area of bioremediation research involves the application of bioturbating macrofauna. Bioturbation comprises a series of processes driven by macrobenthic fauna that affect sediment physical and chemical properties and strongly influence bacterial communities involved in nutrient cycling (Biles et al. 2002). Bioturbation considers both particle reworking, bioirrigation and other benthic biota behaviours (i.e. feeding or grazing) that are involved in the transport and movement of porewater and particles through the water-sediment interface (Herringshaw et al. 2008; Kristensen et al. 2012). Sediment reworking and bioirrigation can affect the mobilisation of dissolved and particulate contaminants by increasing the transport and biomixing rate from overlying water and porewater to deep layers of the sediment (Timmermann et al. 2000; Teal et al. 2013). Bioturbators can also increase contaminant sequestration by the formation of complex biofilms derived from bioturbator organic waste (Lalonde et al. 2010; Qin et al. 2010). In addition, bioturbation activities can increase benthic metabolism and nutrient dynamics by the stimulation of aerobic bacterial communities that participate in contaminant degradation (Kunihiro et al. 2011; Ma et al. 2015; Boeker et al. 2016; Shen et al. 2017).

Our understanding of the effects of bioturbation on contaminant bioremediation is incomplete and the results appear to be inconsistent with a diverse range of responses reported with respect to different classes of contaminants, magnitude of contamination and geographical region (Maximov et al. 2015; Natalio et al. 2017). Further, there is a lack of information relating to the interaction of bioturbators and microbial communities, and the importance of this interaction for contaminant fate in sediments. With the increasing impact and widespread distribution of contaminants in aquatic systems, there is an urgent demand for remediation options and bioturbation may be an important

CHAPTER 1 alternative. A review and synthesis of the available literature is both timely and necessary to assess the potential for bioturbation-bioremediation as an alternative approach for contaminant removal. Information generated through this approach may provide a framework to optimize bioremediation efforts by defining which bioturbator groups have the greatest effect on different contaminants, and if their application in aquatic systems is sufficient to reduce contaminant bioavailability.

Here I used a systematic review and meta-analysis to assess the current state of knowledge of the effect of macrobenthic bioturbators on contaminant fate in aquatic sediments and how this is influenced by the taxonomic group, the aquatic system and important abiotic/biotic environmental factors. In addition, this review reveals the frequency at which different bioturbation strategies have been taken to practice for decreasing sediment contamination and identifies knowledge gaps hindering the development of bioturbation as a viable method for bioremediation.

1.3. Methods 1.3.1. Search methodology An initial literature search was done to identify the main contaminants in sedimentary systems that have had bioturbators applied for remediation. Search terms included bioturbation, bioremediation, pollutants, sediment reworking, bioirrigation and sediments. Searches were done in ProQuest, Web of Science and Scopus. From this search, three main contaminants were identified: metals, organic matter/nutrients and polycyclic aromatic hydrocarbons (Appendix 1, Fig.A.1).

Systematic searches were conducted to identify studies where the effect of bioturbation on each contaminant was evaluated. Search terms included those listed above and additionally: ecological engineer, macrobenthic, biological disturbance, remediation, microb* and bacter*. These terms were combined in

THE APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION searches with selected contaminant terms: metals (heavy metals, metals), organic matter and eutrophication (organic matter, organic enrichment, nutrient, eutrophic*) and polycyclic aromatic hydrocarbons (PAH, Polycyclic Aromatic Hydrocarbons, persistent organic pollutants). In total, 396 searches were conducted for all evaluated contaminants in each search database using all the pertinent combinations of the search terms. Searches were captured from March to August 2017. Duplicate articles were removed, and relevant articles were selected by reading titles and abstracts. Full details of the search processes and literature search are included in Appendix 1 Fig. A.2-4 and Table A.1.

1.3.2. Qualitative analysis In total, 222 research articles satisfied the criteria for inclusion in the review. From these, I found 56 articles relating to metals, 133 articles for organic matter and nutrients, and 33 articles relating to PAHs. From these, information about the main taxonomic classes, location of study, country and aquatic systems were collected as well as details of methodological approaches (ex situ or in situ). Aquatic system classification was obtained directly from each individual study in accordance with how authors described their experimental site. The number of articles where microbial analyses were used to assess the impact of bioturbation on bacterial and archaeal communities on contaminant fate was also reported. Finally, the number of cases where bioturbation was explicitly reported in a real- world bioremediation project was also documented.

1.3.3. Design criteria for meta-analysis The purpose of the meta-analysis was to determine the effect of bioturbators on contaminant fate in sediments. From the initial 222 articles found, only 49 articles met the criteria to conduct a meta-analysis for metals (13 articles) and organic matter/ nutrients (36 articles). For metals, insufficient data was available to separate particulate and dissolved fractions and therefore data from total metal concentrations in the water column were included as the specific variable for the analysis. For organic matter and nutrients, three variables representing a broad range of effects of bioturbation on nutrient release were selected: ammonia (30

CHAPTER 1 articles) and phosphorous fluxes (18 articles) from sediment surface to water column, and sediment oxygen uptake (SOU) (25 articles). The selected variables were investigated in the literature as end responses related to the presence of bioturbation and as indirect measurements of bacterial activity.

Total Ammonia (NH3 + NH4+) was evaluated as a variable for the analysis as it is an important product of bacterial aerobic metabolism during nitrogen dynamics (Martinez-Garcia et al. 2015; Zhang et al. 2014). Total phosphorous

(TP) and orthophosphate (PO43-) concentrations in the water column were also included in the analysis as these molecules represent an important source of nutrients for algal growth and thus influence eutrophication (Renz and Forster, 2014; Chen et al. 2016). SOU was evaluated as an indirect measure of bacterial aerobic metabolism in the sediment (Renz and Forster 2014; Martinez-Garcia et al. 2015). No meta-analysis was conducted for PAHs as there were not enough comparable results and methodologies.

1.3.4. Meta-analyses of the data From all studies selected for each response variable (i.e. total metals, ammonia, phosphorous and SOU,) the mean, standard deviation and number of replicates were extracted or calculated from the raw data directly from the scientific articles. Graph data were extracted using the online application WebPlotDigitizer 4.1 (Rohatgi 2018). In addition, other relevant variables like temperature, pH, salinity, experimental duration, sediment size and animal density were extracted for inclusion in analyses as covariates involved in mediating the response variables. All studies that evaluated the presence or absence of soft sediment macrobenthic and epifaunal bioturbators were considered for this analysis.

For each selected response variable, a multi-level model (package metafor R, Viechtbauer 2010) was used to account for non-independent sampling error due to correlations between effect sizes sharing the same control (Noble et al. 2017). The effect size for each study was calculated by the standardized mean difference (SMD) and sampling variances were obtained. Multiple SMD were

THE APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION obtained per study and effect size estimates were correlated through a variance- covariance matrix to account for non-independence. Absence and presence of bioturbators were selected as fixed factors for the model and animal class and system type were included as nested random factors. In all cases, familywise error rate between multiple comparisons for each random factor was calculated using a Bonferroni correction. Forest plots were constructed for each response variable using the forest function (package metafor R, Viechtbauer 2010; Appendix 1 Figures A.5-8) to visualize the individual effect size, sample variance of each study and overall effect size estimate of the model. To assess relationships between effect sizes and additional moderator variables obtained during data extraction, multiple meta-regressions were done using the function mareg from the R package MAd (Del Re and Hoyt, 2014). In addition, a funnel plot and a trim and fill method (Duval and Tweedle, 2000) were constructed to assess for publication bias.

1.4. Results 1.4.1. Qualitative assessment of contaminant types in sediments tested with bioremediation In this systematic review most studies were reported from temperate areas (e.g. USA, Australia, and European countries) with few from tropical regions (Appendix 1, Figure A.9). Across multiple aquatic systems, most studies were conducted in ex situ laboratory, indoor and outdoor conditions, although some have attempted in situ approaches (e.g. Shen et al. 2016). For metals and PAHs, the majority (~80-90%) of bioturbator applications were conducted ex situ. By contrast, organic matter/ nutrients had a greater proportion of bioremediation studies applied in situ (>40%) but still fewer than ex situ. Finally, experiments with bioturbators and PAHs were rarely reported on in situ conditions (~9%).

Estuaries, rivers and lakes are the focus of studies investigating the effect of bioturbation on contaminants. Artificial systems, such as ponds or aquaculture systems are also represented, although to a lesser extent in all selected contaminants (Fig.1.1). In addition, studies of metal bioremediation utilized

CHAPTER 1 mainly Oligochaetes (n=25, mainly Tubificids) from freshwater systems, while polychaetes (n= 18, Arenicolidae and Nereididae families), Bivalves (n=15, Tellinidae) and Malacostracans (n= 14, Corophiidae and Pontoporiidae) were used in estuarine systems (Fig.1.2). For organic matter and nutrients, polychaetes represented the most commonly used organisms to evaluate the effect of OM/nutrient and bioturbation, mainly with families such as the Arenicolidae, Capitellidae and Nereididae (n= 40). In addition, Malacostracans (n= 31) and Bivalves (n=25) were common bioturbators in OM studies (Fig.1.2). PAH fate included mostly nereidids and capitellids from the class polychaeta (n=15), followed by tubificid worms (n= 7, Oligochaeta) and several groups of Malacostracans as main groups of interest (n= 7) (Fig. 1.2).

1.4.2. Meta-analysis of bioturbator effects on selected contaminants Overall effect size determines the mean magnitude of effect of the evaluated responses in all studies included in the analysis. Results show that bioturbators were effective stimulators of organic matter degradation, contributed to nutrient release into the water column but did not affect metal release from sediments (Ammonia and Phosphorous p <0.001; metals p=0.510; Fig 1.3). Polychaetes, malacostracans, and oligochaetes increased ammonia and phosphorus fluxes from sediment (Fig 1.4). Bivalves and insects acted similarly, but only for ammonia fluxes (Fig 1.4; Appendix 1, Table A2). However, Priapulids and echinoderms (classes Ophiuroidea, Holothuroidea) did not affect nutrient release. Interestingly, the effect of bioturbators on ammonia and phosphorous release was restricted to marine, Stormwater Infiltration Systems (SWIS), lakes and estuaries (Fig. 1.5). In rivers and fjords, bioturbators increased ammonia release when present, but had no effect on phosphorous (Fig.1.5). For streams, sea inlets, ponds, mussel farms, coastal lagoons and aquaculture ponds, no release of ammonia or phosphorous was observed (Fig.1.5, Appendix 1, Table A.2- 3)

THE APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION

Fig.1.1. Frequency diagram of main aquatic systems where the effect of bioturbation on contaminant fate were studied.

Fig. 1.2. Frequency diagram of the main taxonomic groups where the effect of bioturbation on contaminant fate were studied

CHAPTER 1

Oxygen uptake into sediments increased when bioturbators were present (p<0.001, Fig. 1.3). Oligochaetes and insects increased oxygen uptake into the sediment when evaluated by class. By contrast, bivalves decreased SOU, while gastropods, priapulids, polychaetes, ophiuroideans and malacostraca didn’t affect SOU (Fig 1.4, Appendix 1, Table A.2). Bioturbation effects on SOU were variable across different ecosystems. Bioturbators increased SOU in fjords, SWIS and lakes, but decreased SOU in rivers and estuaries (Fig.1.5; Appendix 1, Table A.3). For the rest of the aquatic systems evaluated, no effect of bioturbators was documented.

Overall, the presence of bioturbators did not affect metal release from sediments (p= 0.510; Fig.1.3) although some taxa-specific effects were observed. Specifically, ray-finned fishes increased metal sequestration. In comparison, bivalves, insects and oligochaetes had no statistically significant effect on metal release to the water column, although insects tended to increase metal release from the sediments (Fig.1.4, Appendix 1, Table A.2).

1.4.3. Meta-regression analysis of effects of moderator variables Meta-regression analysis highlighted moderator variables that changed the effect of bioturbators on contaminant release. For example, higher temperatures increased the release of ammonia and decreased phosphorus release when bioturbators were present (Fig. 1.6). By contrast, metal release and SOU were not correlated to temperature when bioturbators were present, although there was a positive trend for metal release (Fig. 1.6, p= 0.110). In addition, for metals, increasing pH levels show increased metal solubility from sediments in the presence of a bioturbator (Fig. 1.6, p<0.001). Salinity was not correlated to ammonia, phosphorous or SOU (Fig. 1.6).

THE APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION

Fig. 1.3. Overall effect size of the release of evaluated responses from sediments to the overlying water and sediment oxygen uptake (SOU) in the presence of bioturbators. Point marks the overall effect size while polygon length represent overall confidence interval for all studies considered in each response. Significant effect sizes are represented by shaded polygons. Red line shows an effect size equal to 0.

CHAPTER 1

Fig.1.4. Overall effect size of the release of evaluated contaminants from sediments to the overlying water and sediment oxygen uptake in the presence of bioturbators and between different taxonomic groups. Point marks the overall effect size while polygon length and height represent range of effect the effect size for all studies considered in each response. Significant effect sizes are represented by shaded polygons. Red line shows an effect size equal to 0.

THE APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION

Fig.1.5. Overall effect size of the evaluated responses in the presence of bioturbators and between different aquatic systems. Small polygons and points mark the overall effect size while large polygon length represents the confidence interval for all studies considered in each response. Significant effect sizes are represented by shaded polygons. Red line shows an effect size equal to 0. SWIS= Stormwater Infiltration System.

CHAPTER 1

Sediment grain size was positively correlated with the release of ammonia, phosphorous and SOU, suggesting that bioturbators were more effective in coarser sediments (Fig. 1.6). In another case, increasing animal density affected phosphorous release and SOU but did not affect sediment ammonia release (Fig. 1.6). Not enough sediment grain size and animal density metadata was found for metals in order to conduct this test. Lastly, metal, ammonia and phosphorous release and SOU were greater overall for longer studies (Fig. 1.6). All meta- regression results for all the response variables are shown in Appendix 1, Fig. A.4.

1.4.4. Publication bias Funnel plots and a trim and fill method indicate no publication bias for metal and phosphorous release or SOU (Fig. 1.7). However, a possible publication bias was determined for ammonia release (k=6, p=0.0078). However, Funnel plot results could be associated to other sources of asymmetry, like differences in methodological quality between studies, heterogeneity of results or data irregularities (Egger, 1997).

1.5. Discussion The legacy and continuing contamination of sediments in aquatic systems is a pressing global issue that requires novel and effective approaches for remediation. Here I synthesized current understanding of bioturbator effects on contaminated sediments and identified knowledge gaps to progress the research field. Overall, I found that bioturbators can influence nutrient dynamics by increasing organic matter breakdown and the release of ammonia and phosphorous to the water column. Sediment oxygen uptake, analysed as a measure of microbial aerobic activity, was also enhanced in the presence of bioturbators. By contrast, I found no overall effect of bioturbators on sediment metal release, but for one group of fishes which increased metal sequestration. There were fewer studies of metals

THE APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION

Fig. 1.6. Scatter plot matrix showing the meta-regression between the obtained effect size (SMD) of each response variable and moderator variables obtained from the studies. Significant p values of the correlation are shown in bold.

` 25

CHAPTER 1

A B

C D

Fig. 1.7. Funnel plots assessing publication bias on evaluated metanalysis: A) metal release, B) Ammonia fluxes, C) Phosphorous fluxes, D) Sediment Oxygen Uptake (SOU).

THE APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION and PAHs than organic matter, which did not allow for a robust meta-analysis of bioturbator effects on PAHs and may have influenced the findings for metals.

1.5.1. Bioturbators promote nutrient release and oxygen uptake Eutrophication of the water column results from excess nutrients stimulating overgrowth of algae and the subsequent depletion of oxygen as algae die. Results from this review show that the activity of bioturbators, including polychaetes, malacostracans and oligochaetes can increase the release of both ammonia and phosphorus from sediments to the water column, while bivalves and insects stimulate the release of ammonia only. Many authors have attributed differences in nutrient release between taxa to differing bioturbation mechanisms (Michaud et al. 2006; Shen et al. 2016). For ammonia, increased fluxes are end products from aerobic microbial activity stimulated by the presence of a bioturbator. This process is especially important in burrows which have been described as OM ‘hotspots” where high levels of OM accumulation and degradation occur (Shang et al. 2013; Poulsen et al. 2014; Holmer et al. 2015; Geraldi et al. 2017; Natalio et al. 2017). Phosphorous can also be stored in burrows, however the presence of bioturbators can increase its release through bioirrigation activities (Norkko et al. 2012; Ekeroth et al. 2012; Chen et al. 2016). Released phosphorous to the overlying water column by bioturbation can potentially become an important factor that could enhance eutrophication of the water column while simultaneously improving the ecology of the sediments (Sinha et al. 2016). It is therefore important to consider the fate of nutrients post-release before applying bioturbation to the problem of enriched sediments.

It is also possible that bioturbators that produce deeper biogenic structures can introduce higher amounts of oxygen and impact a larger surface area for nutrient release in comparison to surficial bioturbators (Jordan et al. 2009). This could in turn stimulate aerobic microbial activity that changes nutrient dynamics (Poulsen et al. 2014; Shen et al. 2017). No data could be obtained for polychaetes or malacostraca in the case of SOU, yet the results comparing oligochaetes and insects demonstrate that aerobic microbial

CHAPTER 1 metabolism is increased in the presence of the oligochaetes, an active bioturbator. Bivalves increase sediment oxygen release in comparison to other bioturbators, such as polychaetes or echinoderms (e.g. Lindqvist et al. 2009; Rao et al. 2014). This could be related to intermittent bioirrigation caused by bivalve surficial bioturbation mechanisms, as compared to more periodic oxygen fluxes of other infauna (Michaud et al. 2005; Volkenborn et al. 2012). These findings suggest a differential role of surficial bioturbators in nutrient dynamics and an important example of functional bioturbation groups in benthic organic matter degradation.

1.5.2. Bioturbator effects on metals is limited to one taxonomic group The effect of bioturbation on metal fate in sediments is poorly represented in the literature, but there is evidence that epifaunal organisms such as fish may promote a higher metal sequestration in sediments than other functional groups of macrobenthic fauna (Pedro et al. 2015). Metal sequestration or bioaccumulation is favoured by sediment reworking and high bioirrigation rates from bioturbators, but this could only be evaluated in one taxonomic group in this study. Rayed finned fish appear to enhance metal sequestration in the sediment; however, results should be interpreted with caution as a low number of studies were obtained for the present meta-analysis. Epifaunal bioturbators, such as rayed finned fish, have a higher motility and cover larger distances in less time that could account for increased sediment surface resuspension and metal binding with particulate material in comparison to other bioturbators (Wall et al. 1996; He et al. 2015). Rayed finned fishes have been reported to decrease metal release through sediment reworking and bioirrigation to enhance metal complexation in sediment surface (Wall et al. 1996; Pedro et al. 2015), however the present work lacks enough power to confirm this and most studies point to the opposite. Under this scenario, it is probable that rayed finned fishes are increasing metal release but at the same time enhancing accumulation through pellet formation that would reduce metal concentrations in the water column (Hodson et al. 1988; Pedro et al. 2015). Insecta also presented a pattern of effect on metal release, however this was not statistically significant. In this case, insects

THE APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION cause a lower accumulation of metals in sediments than other macrobenthic and epibenthic fauna mainly because of their different mechanisms of feeding and reworking the sediment (Colombo et al. 2016).

1.5.3. Further research on bioturbation effect on PAHs fate in sediments Polychaetes are by far the most studied taxa in relation to PAH contamination largely because they are tolerant to PAHs and have the capacity to biotransform these compounds and reduce their toxicity (Christensen et al. 2002). Nonetheless, in other taxonomic groups PAH toxicity can affect sediment reworking and bioirrigation rates that in turn influence PAH mobilisation and bioavailability to other trophic levels (Timmermann et al. 2000). In the present study, however, not enough articles with comparable methodologies or enough information to develop a meta-analytic approach were found. Further investigations are needed in order to quantify adequately the impact of bioturbators on PAH fate and their importance as stimulators of aerobic microbial fractions that could enhance PAH degradation.

1.5.4. Factors that modulate the effect of bioturbators The effect of modulator variables on bioturbators and the release of contaminants from sediments is an important issue to account for if bioremediation strategies are to be successfully employed. The heterogeneity of results obtained in the literature depend greatly on the context of abiotic and biotic variables, as well as different experimental designs and approaches.

Salinities vary in aquatic systems both naturally and because of human activities. This can change the rate of microbial activities and concordantly the release of nutrients and metals (Nowicki et al. 1994; Rysgaard et al. 1999). In our study, no clear patterns were found between higher salinity environments (i.e. marine) and lower salinity (i.e. freshwater systems) in relation to nutrient release or SOU (Fig. 6). As each study was conducted in a specific site and different bioturbators with definite salinity tolerance thresholds and constant salinity conditions, no direct tests could be done to determine if salinity is a release

CHAPTER 1 modulating factor in the presence of bioturbators. However, results do demonstrate that in aquatic systems characterized by higher salinity, there is higher variation in ammonia release and SOU in comparison to other nutrients such as phosphorous, possibly as a result of other modulating factors (discussed below).

Meta-analytic results indicate that the effect of bioturbators on contaminant remediation varies among aquatic systems. Bioturbators tested in Stormwater Infiltration Systems (SWIS), lakes and estuaries affected ammonia and phosphorous release as well as SOU. Bioturbators tested in marine environments presented only an effect in nutrient release (i.e. ammonia and phosphorous) while rivers and fjords only affected ammonia release and SOU. For metals, no clear effects where seen and for PAHs most studies where associated to systems where PAH contamination is increased as a result of accidental release of oil (Boehm et al. 2005). Overall, this pattern may be linked to the dynamism of each system. The systems where an effect was observed, experienced greater water movement in comparison to more protected areas such as ponds, artificial systems and lagoons (Hasanudin et al. 2004).

There is a clear bias of studies being done in temperate regions such as parts of China, Denmark, USA, Germany and France for the analysed contaminants. Only 9% of reviewed articles assessed the effect of bioturbation on OM/nutrient fate in tropical regions, 4% for metals and none for PAHs. For polar regions, metals registered 4%, OM/nutrient 2% and 6% for PAHs. In both regions temperature is a key factor and meta-regression results reveal that temperature had a significant modulating role over the response variables for OM degradation and metals. Increased temperature produces an increase of ammonia release from sediments in response to the presence of bioturbation, likely due to higher bioturbation rates resulting from higher metabolism of macrobenthic organisms and increased aerobic microbial activity in the sediment (Berkenbusch and Rowden, 1999; Bernard et al. 2016). However, our SOU results do not support this explanation as no significant correlation was found with temperature

THE APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION increase in this study. For other nutrients like phosphorus, an increase in temperature was correlated with a decreased flux from the sediment. In the literature, it has been reported that increasing bioturbation rate and higher infauna metabolism increased phosphorous release (Norkko et al. 2012; Ekeroth et al. 2012), the opposite pattern to what was found here. For metals, no significant correlation was found in this study, however it is possible that higher temperatures can be involved in an increase of metal resuspension and solubility as a consequence of higher bioturbation rates (Andres 1998; Amato et al. 2016). Similar results are expected for PAHs, however the lack of information in the literature makes it hard to determine a clear effect of bioturbation and temperature on this contaminant in this study.

In the present study animal density, sediment grain size and pH were also correlated to increased contaminant release and aerobic microbial activity. Increasing bioturbator density increased the release of phosphorous and stimulated aerobic microbial activity (i.e. SOU) but did not increase ammonia release. It is possible that aerobic microbial activity is stimulated but does not enhance end products like ammonia, but instead increases other steps in OM mineralization (Nogaro et al. 2014). Animal density is a crucial aspect for further investigation as higher densities may contribute to higher OM degradation but may also increase nutrients in the overlying water which can have counterproductive effects for remediation, depending on the ecosystem studied. For metals, abiotic variables like pH may be an important modulator variable (Fig. 1.6) that determines different effects of bioturbation in multiple aquatic systems as it can change metal speciation and reactivity (Amato et al. 2016).

Sediment grain size is another important factor that determines a high degree of accumulation of OM and metals and changes the structure of microbial communities and metabolism (Jackson and Weeks 2008; Burton and Johnson 2010). In the current analysis, sediment grain size was positively correlated to ammonia, phosphorous release and SOU. Martinez-Garcia et al. determined that sediment grain size in the presence of a bioturbator was not significant at low

CHAPTER 1 levels of organic enrichment but in higher enrichments coarser sediments retain less OM and nutrients and microbial activity is increased (2015). This could explain the pattern seen in the present study and would be an indicator that the analysed studies in the meta-regression could probably derive from aquatic systems with high organic matter levels.

1.5.5. Future research directions From the qualitative literature review, 28% of the analysed studies on nutrients directly proposed the application of bioturbators to remediate eutrophic sediments, while only 5% and 15% of studies focused on strategies for metal and PAH sediment remediation, respectively. OM/nutrient bioremediation through bioturbation has mainly been reported in artificial aquatic systems such as aquaculture ponds, mussel farms, enclosed artificial lagoons and recirculation aquaculture systems (Bartoli et al. 2001; Kunihiro et al. 2008; Nicholaus et al. 2014; Zhong et al. 2015; Kang et al. 2017; Robinson et al. 2018; Lukwambe et al. 2018; Zhao et al. 2019). In these studies, water quality improvement of the aquaculture industry was the main focus to increase annual production rates (Biswas et al. 2009; Ren et al. 2010; Kunihiro et al. 2011; Shang et al. 2013; Zhang et al. 2014; Nogaro et al. 2014; Holmer et al. 2015; Robinson et al. 2016). Others have also proposed the use of bioturbation in bioreactors for the treatment of urban aquatic systems in order to treat municipal discharges (Mermillod-Blondin et al. 2005; 2008 Penha-Lopes et al. 2009). Nobbs et al. (1997) and Mayor et al. (2013) highlighted the importance of bioturbators (bivalves and malacostraca) as a component that will determine lake and estuary metal contamination from industrial wastewater. In addition, a similar proposal in stormwater infiltration systems as bioreactors has demonstrated the capacity of tubificids to immobilize metal release and even be applied in conjunction with other current bioremediation methods like phytoremediation (Nogaro et al. 2007; Hoang et al. 2018). Finally, for PAH contamination bioturbation-bioremediation was poorly represented in the literature as no study was directly linked to a bioremediation proposition for the exception of Kang et al. who evaluated the impact of mussels

THE APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION in PAH sediment removal in constructed wetlands (2019). For all analysed contaminants, there is a clear lack of reference to basic ecological theory, however results obtained in this study highlight relevant taxonomic groups and specific aquatic systems where sediment bioremediation should be effectively explored. Such is the case of artificial or urban aquatic systems in which development of a general bioturbation-bioremediation approach is still too early to apply but needs to be considered as an important tool for plausible solutions for sediment remediation in highly modified environments (Clark et al. 2015).

The present study highlights the lack of experimental data that approaches in more detail sediment macro-microbiota interactions. From the articles reviewed in this work, 25% revealed some information regarding the effect of bioturbation on the microbial communities involved in OM degradation. Some of these include studies done in recent years to establish the importance of both biological fractions for sediment bioremediation in aquaculture wastewater management (Robinson et al. 2018; Lukwambe et al. 2018; Zhao et al. 2019). The interaction between microbial communities present in the sediment and bioturbation on metals has also been poorly studied in the literature. From this systematic review, only 25% of the articles mentioned the microbial community interacting with metals and only Mayor et al. (2013) analysed directly a microbial quantifiable result (i.e. energy budgets) that can relate to the toxic effects of metals. Lastly, for PAHs, 27% of the reviewed literature analysed the effect of microbial communities stimulated by bioturbators for PAH degradation. Authors highlighted the importance of bioturbators to stimulate hydrocarbonoclastic microbial communities as a key factor for PAH degradation (Ciutat et al. 2006). In all analysed contaminants, bioturbation changed the physicochemical conditions of sediments, providing microbial communities with a higher complexity of microniches, higher resources and adequate redox environments to alter their abundance and structure (e.g. Foshtomi et al. 2015; Boeker et al. 2016). However, the lack of next generation molecular approaches has hindered the elucidation of the microbial mechanisms that can be a focus of interest to understand underlying processes involved in ecosystem function. The onset of

CHAPTER 1 new molecular approaches allows the possibility of having larger resolution and scale to analyse the mechanisms between macrobenthic biota and microbial communities in contaminated environments (Sun et al. 2012;Dafforn et al. 2014; Lawes et al. 2017). Results in this review show that bioturbation affects microbial sediment processes evidenced by SOU, however the study of macro-microbiota interactions has not been addressed in depth and new technologies make this possible and necessary to evaluate.

The majority of studies in this review were conducted under ex situ experimental conditions and in short time periods. OM/nutrient bioturbation presented 60% ex situ experiments. For metals, 18% of the studies were approached in situ, where studies focused on the mechanisms of metal mobilisation and biomixing aided by changes in the redox state of sediments (Klerks et al. 2007; Teal et al. 2013;). Only 10% of the extracted articles applied in situ experimental designs in order to assess PAH bioturbation-bioremediation strategies (Qin et al. 2010; Timmermann et al. 2011). In most in situ studies, application of bioturbators was done in small experimental areas (Wang et al. 2010; Geraldi et al. 2017; Gammal et al: 2017) and only one was attempted at a larger scale (Ma et al. 2015). The response variables analysed generally increased over time, having the highest levels of release at 50 days from addition of bioturbators for phosphorous, metals and SOU. However, ammonia release from the sediment decreased as experiment duration increased. It is possible that available OM in experimental mesocosms might start depleting and so ammonia release also decreases. This is a case which probably would not happen in a real- world field situation although it is likely in any experimental arrangement in the laboratory or field (Hill et al 2011). Short duration studies and ex situ studies may not represent adequately the effect of bioturbation on contaminant fate in sediments for a bioremediation scenario to be assessed. In most evaluated studies, ex situ conditions cannot reproduce natural conditions entirely and most modulating factors are controlled (Burton and Johnston 2010). The present work highlights the importance of these modulating variables and together with larger spatiotemporal field based experimental designs, the applicability of bioturbation

THE APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION as a bioremediation method can be adequately analysed in future research.

1.6. Conclusion This synthesis of current studies on bioturbation highlights the possibilities of this approach for contaminant remediation. In the cases analysed in this review, bioturbation had a significant effect on contaminants through degrading, accumulating or decreasing bioavailability for other trophic levels. Few studies considered or differentiated the microbial mechanisms that are driven by bioturbating macrofauna and thus necessary for this remediation method to be effective. Context specific variables make it necessary to optimize this approach for regional ecosystems (Johnston and Keough 2005). Overall, future research directions should aim to increase the number of field or semi-field experiments that concurrently measure macro scale responses. For example, changes in bioturbator performance and toxicity that impact sediment properties (i.e. sediment redox oscillations, contaminant dynamics) and cross-scale interactions (Burton and Johnston 2010). In addition, the multiple environmental variables that have been analysed need to be considered as important modulator variables in order to optimize contaminant bioremediation. This will enable the creation of adequate strategies at longer spatiotemporal scales and assess the possibility of synergistic or antagonistic effects of multiple contaminants. The bioturbation-led remediation of contaminated aquatic habitats is possible, however, with our current understanding there is a need to develop site specific approaches using an adaptive management approach

CHAPTER 1

2. THE INFLUENCE OF BIOTURBATOR INTRASPECIFIC BODY SIZE VARIATION IN SEDIMENT BIOREMEDIATION: EFFECTS ON 2 SURVIVORSHIP AND MOBILITY OF A SEDIMENT BIOREMEDIATOR IN ORGANICALLY ENRICHED SEDIMENTS

2.1. Abstract Eutrophication is an increasing problem in aquatic systems worldwide and its influence may disrupt local processes that determine ecosystem function. In aquatic sediments, macrobenthic bioturbators are important drivers of local processes that can affect nutrient dynamics and other relevant ecosystem functions. Here I explored how intraspecific body size of macrobenthic bioturbators can affect their survival, mobility and impact on sediment organic matter loss in enriched sediments. A mesocosm experiment was conducted with two enrichment treatments (natural enrichment or spiked with organic matter) and 3 body size phenotypes of the Sydney cockle (Anadara trapezia): small (20-30mm), medium (35-50mm) and large (55-70mm). These phenotypes were used to create seven treatments: monocultures, all pairwise combinations, all three body size phenotypes together and an additional treatment with no cockles was included as a control. Results demonstrate that larger body sizes have a higher tolerance to enriched conditions than smaller body sizes. In addition, the presence of larger body sizes was shown to reduce the survival of small body sizes when combined. No clear effects of the presence of the A. trapezia were found on sediment organic matter loss. We found that intraspecific variation in body size is important for determining the survival and performance of bioturbators in eutrophic scenarios and should be considered as an important factor that could influence potential bioremediation strategies in aquatic systems.

37 CHAPTER 2

2.2. Introduction Benthic macrofauna are important drivers of essential processes in aquatic systems (e.g. organic matter accumulation, processing, or changes in sediment redox state) that can alter sediment properties and microbial communities involved in ecosystem functioning (Biles et al. 2002; Stief, 2013; Vadillo- Gonzalez et al. 2019). Macrofauna affect sediment processes through a series of behavioral activities (e.g. sediment reworking, grazing, mucus, and fecal production or bioirrigation), however these can change under high organic pollution or eutrophication scenarios (Murray et al. 2017). Eutrophication, specially the addition of nutrients such as phosphorous and nitrogen, can cause increased growth of primary producers and trigger “blooms” (e.g. algae or cyanobacteria), leading to lower dissolved oxygen levels in sediments (Wang et al. 2016) that can impact ecosystem functioning and reduce the abundance and diversity of sediment macrofauna (Vaquer-Sunyer and Duarte, 2008). However, many species of bioturbators can tolerate these conditions and are capable to stimulate aerobic microbial activities that permit high levels of nitrogen degradation (Rao et al. 2014; Kang et al. 2017). Thus, the relationship between bioturbation and organic matter degradation is complex and requires further investigation to propose novel bioremediation alternatives for organic matter pollution and eutrophication.

The body size of bioturbating organisms has important consequences for the ecosystem functions they provide. Recent studies have provided evidence that removal of specific body-sized taxa within benthic assemblages can lead to alterations in trophic functioning in aquatic systems (Seguin et al. 2014) and this effect cannot be overlooked at an intraspecific level. Intraspecific functional trait variation can determine many ecosystem processes and how a community will respond to disturbances (Cassidy et al. 2020). Body size contributes greatly as a highly variable intraspecific functional trait in benthic communities that influences ecosystem function such as processes in nutrient cycling (Cassidy et al. 2020). For example, Norkko et al. (2013) found that large organisms in sandy

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MORTALITY AND MOBILITY sediments were the most important predictors of ecosystem function due to differences in benthic metabolism associated with bioturbation. Similarly, the effects of bioturbating macrofauna on nutrient dynamics (Taylor and Brand,

1975; Carmichael et al. 2012 Clark et al. 2013), including NH4+ and PO43- fluxes, as well as the stimulation of sediment microbial fractions involved in organic matter degradation, can vary with ontogenetic stage or body size within a single species (Norkko et al. 2013). More recent studies demonstrate that intraspecific diversity (e.g. diversity of phenotypes and/or genotypes) can have important consequences for communities and ecosystem processes (Post et al. 2008; Bolnick et al. 2011; Des Roches et al. 2018; Raffard et al. 2019;). For example, subtle variation in the phenotypic traits of many aquatic organisms has been shown to influence predation (El-Sabaawi et al. 2018) and habitat selection (Duffy et al. 2010). Sessile marine invertebrates have been a particular focus of this research (for examples see Gamfeldt et al. 2005; Aguirre and Marshall 2012; Hanley et al. 2016), however, we know little about how intra-specific variation of mobile species may control benthic ecosystem functions.

Intraspecific variation in benthic macrofauna may influence bioturbation processes in several ways and these may be dependent on sediment conditions (Taylor and Brand, 1975; Cozzoli et al. 2018). Firstly, individual body sizes may respond independently to changes in sediment conditions, with all body sizes surviving in natural healthy sediments. Whereas, in eutrophic scenarios, because of their higher tolerance to poor environmental conditions, only larger body sizes may survive and provide ecosystem services (Clark et al. 2013). Alternatively, larger organisms through their bioturbation activities (e.g. higher bioirrigation or formation of aerobic microbial micro-niches) may create an associational refuge (Pfister and Hay, 1988; Barbosa et al. 2009) where more benign conditions in eutrophic sediments could occur and promote survivorship and/or movement of smaller body sizes (Vaughn and Hakenkamp 2001; Bertics and Ziebis 2010). However, eutrophic conditions coupled with resource competition may produce associational susceptibility (Barbosa et al. 2009) that can lead to increased mortality of smaller body sizes in the presence of larger individuals (Weinberg et

CHAPTER 2 al. 1985). Understanding intra-specific variation in body size and its correlation to nutrient dynamics may be critical for determining the effects of human-driven eutrophication on benthic communities and overall ecosystem function.

Infaunal bivalves (i.e. those that live within sediments) are particularly important bioturbators in coastal environments as they can dominate the benthic biomass (Vaughn and Hakenkamp 2001). These organisms often have a high tolerance to eutrophic conditions in comparison to other taxonomic groups and have the potential to be used in sediment bioremediation (Vadillo-Gonzalez et al., 2019). This tolerance arises from their unique mechanisms to cope with low concentrations of oxygen in sediments (i.e. siphon stretching or burial depth reduction; Vaquer-Sunyer and Duarte 2008; Wright et al. 2010). However, eutrophication has also been linked to bivalve mortality and sublethal effects such as growth reduction in many coastal environments (Carmichael et al. 2012). In such cases however, effects of eutrophication were only analysed between different species and intra-specific comparisons are poorly represented in the literature (some examples include Clark et al. 2013; Norkko et al. 2013).

In this study, I conducted a mesocosm experiment with the Sydney cockle, Anadara trapezia, to evaluate how sediment organic enrichment and intraspecific diversity of body size (i.e. when grown in monocultures of single body sizes vs combinations of body sizes) affected cockle survivorship and movement and how these relate to changes in sediment organic content. Sublethal effects on growth, reproduction and behaviours due to low oxygen levels in boundary water in sediments invaded by the green alga Caulerpa taxifolia, have previously been described (Gribben and Wright 2006; Wright and Gribben 2008). Here I build on these studies, to assess how bioturbating activities of these bivalves are influenced by organic matter enrichment and conversely, whether eutrophic sediments could be bioremediated with cockles. We hypothesized that 1) all body sizes would have reduced survivorship and movement and remove less organic material in enriched vs natural sediments; 2) these negative effects of enrichment would be stronger on smaller compared to

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MORTALITY AND MOBILITY large size classes, and 3) monocultures of body size classes would perform more poorly than more diverse cultures. For example, I predicted that smaller bivalves would have higher survivorship and mobility in mixtures with larger phenotypes present as the latter may alter physicochemical properties of the sediments that benefit smaller organisms. The capacity of each cockle size treatment to degrade organic matter was also evaluated to provide additional information on their role in nutrient dynamics in enriched systems and bioremediation potential.

2.3. Methods 2.3.1. Study species and collection A. trapezia is an abundant epibenthic ark shell bivalve that inhabits mud, sand and seagrass beds in intertidal and shallow subtidal areas of tropical and temperate Australian estuaries (Edgar 2001; Lloyd et al. 2012). A. trapezia can obtain shell lengths up to 75 mm and are typically shallow burrowers with at least some part of the posterior shell margin (5-10 mm) protruding from the sediment (Gribben et al. 2009; Byers et al. 2010; Wright et al. 2010; Lloyd et al. 2012). Bioremediation in sediments with high levels of organic matter enrichment has been assessed with the use of other bioturbators such as polychaetes (Ito et al. 2011) which hold as a species with higher tolerance, breeding capacity and increased richness in highly modified systems (Pearson and Rosenberg, 1975; Dafforn et al. 2013). In comparison, bivalves have higher vulnerability to these sediment conditions (Carmichael et al. 2012) and would seem inadequate for the bioremediation of highly polluted sediments. However, for this study, A. trapezia was selected due to its emerging commercial interest for the fisheries industry to evaluate its capacity to tolerate high levels of organic matter pollution that arises in many aquacultures scenarios (Chary et al. 2020 and personal communication, DPI Port Stephens Fisheries Institute). More importantly, new knowledge of different components of A. trapezia population’s (i.e. different body sizes) capacity to tolerate these conditions may provide insights into overall bivalve restoration possibilities in highly modified areas and their effect on sediment processes.

CHAPTER 2

A. trapezia were collected in silt/sandy sediments in Eraring (Pipers point, 33° 04’ 37.66’’ S, 151° 32’ 17.56’’ E) in southern Lake Macquarie NSW, Australia. Three size classes were collected based on their shell length (maximum anterior to posterior axis): a) Small: 20-30 mm, b) Medium: 35-50 mm and c) Large: 55- 70 mm. Up to 62 cockles were collected for all size classes (exact sample size for each size class given in Appendix 2, Fig.B.1). Collections were made in July (winter) in an area of ~2500 m2 in 1 to 3 m water depth and cockle density in the collection site of 0.02 individual/ m2 (52 individuals/ 25000 m2). All A. trapezia body size classes co-occurred in the same area and were collected haphazardly from the sediment. Animals were transported to the Port Stephen Fisheries Institute (PSFI) in Taylors Beach, NSW and acclimated to a water temperature of 22°C for 15 days.

2.3.2. Sediment collection and mesocosm enrichment setup To ensure sediment grain size and organic matter content were consistent across all mesocosms, sediment was collected in a single site in Tilligerry Creek, Taylors Beach NSW (32° 44’ 28.81”S, 152° 03’ 20.82” E) with known composition (Appendix 2, Table B1; average percent composed of clays (14%), silt (68%) and sands (9%); analysed with Mastersizer 2000, Malvern®, partial method; Ramaswamy and Rao, 2006) and large macrofauna were removed. Half of the sediment was enriched on-site with a pulverized commercial animal-based manure (Yates Dynamic lifter®, Total Nitrogen 3.5% w/w and Total Organic phosphorous 1.7% w/w, modified from Birrer et al. 2019). The addition of urea- based fertilizer to collected sediments did not change sediment composition and remained mostly muddy till the onset of the experimental period. (Appendix 2, Table B1). Sediments without organic enrichment had an average of 8 ± 0.25 SE% of organic matter content and enriched sediments had an average of 14 ± 0.69 SE%. Through this organic enrichment, sediments tried to mimic conditions that represent similar values of sediment organic matter reported in urban aquatic systems with high organic pollution and close to sewage or farming runoff (Filippini et al. 2019; and calculated from organic matter lost on ignition through

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MORTALITY AND MOBILITY the method described below; Wang et al. 2011). All sediment was then frozen for 24h to standardize the starting assemblages and remove the possible influence of meiofauna and any residual macrofauna on the results (No sieving was done to collected sediments; method modified from Braeckman et al. 2010 and based on similar experimental procedures in Hale et al. 2015). After freezing, ~8 L of sediment was added to 11.5 L mesocosms (n= 70; 12.5 cm (radius) x 22.5 cm (height)) in a continuous flow-through system to acclimate for 12d.

2.3.3. Mesocosm setup and experimental design Following incubation of sediments, dissolved oxygen in the upper 2 cm of the sediment was measured with a field microsensor multimeter (Unisense®) in all mesocosms. We confirmed that enriched sediments had lower levels of dissolved oxygen (66 ± 0.3 mg/L) than natural sediments (90 ± 1.5 mg/L) at the beginning of the experiment. For logistical reasons, oxygen monitoring in the sediment surface or the overlaying water was not able to be assessed from this point forward. The water flow-through system comprised of a main water storage (20,000 L) with a constant supply of filtered seawater (supplied directly from Tilligerry Creek and filtered at 1µm) and connected to each individual mesocosm through a main distributor. Water flow to each mesocosm was controlled through a flow regulator and maintained for the duration of the experiment at ~37 mL/min. Water temperature within the mesocosms was measured daily and was constant at 22 ± 2°C during the experiment. Seawater salinity remained at 33 ± 1 ppt for the duration of the trial and the systems were held under a 12h:12h light/dark regime.

A total of 14 cockle body size groups were created and divided between the natural and enriched sediment treatments: 3 groups with monocultures (i.e. Small, Medium and Large), 3 groups with pairwise comparisons (i.e. Small+Medium, Small+Large, Medium+Large) and all sizes combinations (i.e. Small+Medium+Large). A control group with no cockles was also added to determine if changes in sediment organic matter content can be attributed to the presence of the different groups of A. trapezia. Cockles were added to the

CHAPTER 2 sediment mesocosms within the different size treatments. However, to control changes in biomass with cockle size , a different number of cockles were placed in each group (refer to Appendix 2, Fig. B1) to standardize biomass across mesocosms. Biomass was standardized at 0.06 g/cm for all mesocosms and 4-5 replicas were considered for each body size treatment (to a total of 70 mesocosms for the whole experiment). Biomass variability was also accounted for in further statistical analyses within the general mixed models as an offset (see below section 2.3.5 for details). Through this experimental design it was possible to determine intraspecific variability of lethal, sublethal (i.e. mobility) and functional responses (i.e. organic matter content decrease) of different body sizes in monoculture or in combination of A. trapezia in highly organically enriched sediments compared to natural sediments.

The experiment lasted a total of 16 days after acclimation of sediments and cockles. To avoid additional stress that could affect mortality rates during the experimental period, cockles were fed every third day (Day 3, 6, 9, 12 and 15) with ~50 mL/mesocosm of an algal mix (50% Chaetoceros muelleri, 25% Tisochrysis lutea (T. Iso) and 25% Diacronema lutheri). These regimes were necessary as cockles had no substantial food source from the filtered water (1 µm), however, it needs to be acknowledge that accumulation of this organic input in the sediment or cockle biodeposits would increase the levels of organic matter content. The inclusion of a control (i.e. no cockles added) will provide some information of additional organic matter formed by cockle’s biodeposits in response to food addition when compared with treatments with cockles present. In addition, continuous flow-through from the main water deposit to all mesocosms is expected to reduce the amount of excess dissolved organic matter derived from the food. The short period of exposure (16 days) was also expected to help reduce large amounts of organic matter content from food to accumulate and impact the results for this study.

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MORTALITY AND MOBILITY

2.3.4. Quantifying mortality, mobility and changes in organic matter content Cockle mortality during the experimental period was assessed through direct touch-responsiveness of the siphon and mantle (Clark et al. 2013). Deceased organisms were recorded and immediately removed from the mesocosms. The shallow burial of A. trapezia allowed us to document their lateral movement across the sediment surface using photographs taken daily. Photographs from each mesocosm were taken at ~1 m height from all sediment surface at the same time each day (~19:00h). A sequence of 16 images per mesocosm (1,120 photographs in total) were generated and processed with Image J 1.x (Schneider et al. 2012). Image processing consisted of calculating the lateral distance travelled by each cockle in one daily cycle on the surface of the mesocosm. Total mobility was calculated by adding the distance travelled per day for each individual cockle in each mesocosm. This measure of activity was then used to infer potential sediment reworking and its effect on organic matter loss. A detailed description of mobility photogrammetric methods is given in Appendix 2, Fig. B2.

On Day 1 and 16, one sediment sample per mesocosm was collected for organic matter analysis. Collection was done by coring the top 5 cm of the sediment with a 50 mL syringe from multiple locations in each mesocosm and then pooling it into a single tube. This corresponds to the average burial depth of the cockles but includes enough range of depth to evaluate effects of cockles with large body sizes ( up to 7cm, Byers et al. 2010). Sediment samples were stored in 50 mL tubes and immediately frozen at -20 °C to avoid any further organic matter decay. Organic matter content was calculated through Loss on Ignition analysis (LOI; modified from Wang et al. 2010) from the average of three replicates per sample. This was achieved by obtaining an initial weight (g w/w) and then drying the samples at 105°C for 12 h in a drying oven. A second weight was obtained (g d/w) and samples were then heated to 500°C for 12 h in a muffle furnace to combust all available organic matter. A final weight (g ash/w) corresponding to the inorganic fraction of the samples was obtained and the % of organic matter

CHAPTER 2 content lost during ignition was calculated. From this data, total organic matter percent decrease was calculated (((Initial organic matter % - final organic matter %) / initial organic matter) *100). Selection of this method to analyze the effect of the Sydney cockle for sediment bioremediation was decided as being a practical, inexpensive, and logistically easier method to assesses changes in organic matter for a large number of mesocosms (n=70) (Shuman, 2003).

2.3.5. Statistical analysis Cockle mortality only occurred in enriched sediments therefore, mortality percentages were determined for this treatment only. Percentage cockle mortality at day 16 was calculated for all cockle size treatments and arcsine transformed to meet assumptions of normality and constant variance. To determine differences in percent mortality between each body size combination, a general linear model (GLM) was applied using the function glmmTMB from the glmTMB package in R (Brooks et al. 2017) that provided additional model distributions to account for zero-inflated datasets found in mortality and cockle mobility data (e.g. Tweedie distribution). . For the models, percent mortality was set as the response variable and body size treatments and enrichment were considered as fixed factors. Where significant effects of body size combinations were detected, least square means pairwise comparisons were calculated, and p-values adjusted (i.e. Tukey) using the function emmeans from the package emmeans (Length, 2019). For all tests, fixed factors inferences were done through a likelihood ratio test with the Anova function from the package car in R (Fox and Weisberg, 2019). All model assumptions of normality and homogeneity of variances between groups were met (Appendix 2, Fig. B4).

A general linear mixed model (GLMM; similar to the described above) was used to test for differences in cockle lateral mobility between natural and enriched sediments, across all size treatments and for individual body sizes within each body size combination. All lateral mobility data was extracted from individual cockles in each mesocosm and then pooled into each size combination category. All data was log transformed and, as many cockles registered no

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MORTALITY AND MOBILITY mobility, a Tweedie distribution for zero-inflated data was used to fit the model (Foster and Bravington, 2013). Within the models, enrichment (natural vs enriched) and size class treatments were included as fixed factors, and interactions between them were also analysed. Loss in biomass due to higher mortality in enriched treatments could have affected cockle movement, so data for biomass lost during the experiment was also included as an offset in the models. This offset term corresponds to known values of cockle biomass by day 16 and its included in the model as a covariate that corrects for population size difference due to high mortality in enriched sediments. As data for lateral movement is derived from values obtained from multiple organisms in a single mesocosm, individual mesocosm was also included as a random factor in the GLMM to account for non-independence. Fixed and random factor inferences were conducted via likelihood ratio tests as described above (Anova function) and a parametric bootstrap respectively. All model assumptions were met and diagnostics can be found in Appendix 2, Fig. B5.

To determine the effect of different body size combinations on sediment organic matter loads in enriched and natural sediments, I again employed a GLMM as above. Here organic matter percent decrease was used as the response variable and enrichment and body size treatments were assigned as fixed factors and interactions between both categorical factors were determined. All organic matter percent decrease was arcsine transformed to fulfill GLMMs assumptions (Appendix 2, Fig. B6). In addition, and similar to lateral movement, organic matter percent decrease could have been affected by the loss of cockles in the enriched sediments, so biomass was also included as an offset. To determine relationships between cockle evaluated effects of enrichment, I conducted a correlational approach that was aimed to provide additional information about the effect of mortalityand mobility of A. trapezia had on the registered decrease of organic matter.

CHAPTER 2 2.4.Results 2.4.1.Lethal effects of enrichment vary between body size combinations All cockles in the natural sediments survived to the end of the experiment. By contrast, total mortality was high in the enriched sediments (from 37.5% to 100% in some mesocosms) but did not differ among body size treatments (Fig. 2.1A, p= 0.46). This was also reflected in the mortality of large and medium individual body sizes in different combinations (Fig. 2.1B-D and Table 2.1). The exception was individual small A. trapezia which suffered greater mortality when in combination with larger body sizes compared to when they occurred in monoculture (Fig. 2.1B and Appendix 2, Table B2, i.e. SML, p= 0.05). For medium A. trapezia, there was a trend of increasing mortality when placed in size class mixtures however this was not significant (Fig. 2.1C). Overall statistical results are summarized in Table 2.1.

2.4.2. Cockle mobility decreased by enrichment in large cockles Overall, movement of cockles was reduced in enriched sediments compared to natural treatments (Fig.2.2A, p<0.001), and was reduced for large cockles in monocultures compared to all other treatment combinations (Fig.2.2B and Appendix 2, Table B.3, p<0.001). There was no interaction between enrichment and body size treatments (p=0.10).

When individual size classes are considered, there were few effects of sediment condition or size treatment combination on mobility. For small cockles, there was an interaction between sediment condition and body size combination where mobility was reduced in enriched compared to natural conditions for monocultures only (Fig.2.2C, p<0.001). For medium body sizes, sediment enrichment and body size did not influence mobility (Fig.2.2D-E). For large cockles, mobility was not affected by the enrichment (Fig. 2.2F, p=0.6) but large cockles in monocultures moved less in comparison to large cockles when in combination with smaller body sizes (Fig. 2.2G and Appendix 2, Table B4). No interaction between sediment enrichment and cockle body size combination was

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MORTALITY AND MOBILITY found in both medium and large cockles (p= 0.76 and p= 0.62). Variability in movement by multiple cockles in a single mesocosm did not influence the mobility patterns obtained (p=0.9891). All mobility statistical results are summarized in Table 2.1

CHAPTER 2

Table 2.1. Summary of general linear mixed models used to investigate a) percent cockle mortality, b) lateral mobility and c) organic matter breakdown in two sediment types (ST, natural and enriched) and in different body size combinations (BSC, small, medium, large, S+M, S+L, M+L and S+M+L). Results from the interaction between ST and BSC are also included. Significant differences are shown in bold.

Sediment type (ST) Body size Interaction Response Variable Natural-Enriched combinations (BSC) ST*BSC a) Percent cockle mortality Test Chisq df p Overall 5.73 6 0.45 Small only Not analysed 9.34 3 0.03 Not analysed Medium only 3.50 3 0.32 Large only 0.95 3 0.81 b) Lateral mobility (cm/ accumulated to day 16) Test Chisq df p Chisq df p Chisq df p Overall 12.02 1 <0.001 21.38 6 <0.001 10.59 6 0.10 Small only 12.12 1 <0.001 4.11 3 0.25 9.30 3 0.03 Medium only 3.01 1 0.08 2.42 3 0.48 1.15 3 0.76 Large only 0.31 1 0.58 19.73 3 <0.001 1.83 3 0.61 c) Percent OM breakdown Chisq df p Chisq df p Chisq df p Overall 1.38 1 0.24 21.16 7 0.003 14.12 7 0.05

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MORTALITY AND MOBILITY

Fig.2.1. Overall cockle mortality (%) between different cockle size treatments exposed to 16 days in enriched sediments only (14% OM) (A). In addition, cockle mortality (%) of only small (B), medium (C) and large cockles (D). Data are presented as mean % ± SE. In comparisons where significant differences occurred, treatments with a common letter do not differ significantly. Significant differences are shown with different letters. Abbreviations: S = Small, M= Medium and L= Large.

CHAPTER 2

Fig.2.2. Cockle mean lateral mobility (cm) evaluated as the total distance travelled during a 16 day exposure to natural (8% OM) and enriched (14% OM) sediments between all sizes (A,B) and analysed separately with only small (C, sediment type data separated between body size combinations due to the presence of an interaction), medium (D,E) and large cockles (F,G) included. All data are expressed as mean ± SE. Models to calculate these differences included biomass loss as an offset covariate to correct by high mortality in enriched treatments. Different letters represent significant differences between body size combinations within each sediment type. Asterisks indicate differences in cockle lateral movement between natural and enriched sediments. Abbreviations: S= Small, M= Medium and L= Large.

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MORTALITY AND MOBILITY

2.4.3. Effect of different cockle size treatments on sediment organic matter content decrease Sediment organic matter breakdown occurred in all mesocosms however it was similar between both sediment enrichments (p= 0.24, Natural: organic matter breakdown of 0.25 ± 0.03 % SE and Enriched: 0.28 ± 0.02 % SE). Only mesocosms with small cockles in monocultures had a higher percent decrease of organic matter breakdown in enriched sediments compared to natural (Fig. 2.3 and Appendix 2, Table B5, p= 0.005). Cockle lateral mobility (Appendix 2, Fig. B.3A) was not correlated to loss of sediment organic matter content for both natural and enriched sediments (Natural: p=0.79 and Enriched: p=0.47). Organic matter decrease was also unrelated to cockle mortality at day 16 of the experiment (Appendix 2, Fig. B.3B; Natural: p=0.92 and Enriched: p=0.29). Percent organic matter breakdown results are summarized in Table 2.1.

2.5. Discussion Intraspecific variation has important implications to benthic ecosystem functions and in the context of bioremediation strategies, body size is an important functional trait of benthic communities that can determine much of the survival rates and resilience of sediment communities against human derived disruptions such as eutrophication. Here I explored how intraspecific variation in body sizes of an abundant epibenthic ark shell bivalve, A. trapezia, affects survival, mobility and its influence on nutrient processes in enriched sediments. Overall, I found that enriched sediments had lethal and sublethal effects on the Sydney cockle, and survival was found to be body size specific. However, I found little evidence that increasing phenotypic variation increased survivorship, movement or organic material breakdown, despite recent studies indicating the benefits of intraspecific diversity to community and ecosystem functioning (Raffard et al. 2019).

CHAPTER 2

Fig.2.3. Sediment organic matter percent decrease (% ± SE) during a 16-day exposure to A) natural (8% OM) and B) enriched (14% OM) sediments. Asterisks indicate differences of organic matter breakdown in cockle body size treatments between natural and enriched sediments. Abbreviations: S = Small, M= Medium and L= Large.

2.5.1. Cockle mortality in enriched sediments is lower in large organisms Intraspecific body size variation is an important functional trait that influences multiple ecosystem processes (Seguin et al. 2014) but a loss of specific sizes within these communities has proven to be detrimental to overall ecosystem functioning (Norkko et al. 2013; Cassidy et al. 2020). Whilst I found equal mortality of cockles across all body size combinations, larger cockles appeared to have a higher tolerance to enriched sediments than smaller organisms. Larger body masses in bivalves often have an increased tolerance to low oxygen conditions in eutrophic sediments, while smaller bivalves have faster metabolisms and are more vulnerable to eutrophication (Taylor and Brand, 1975; Pechenik et al. 1999; Marsden et al. 2012: Riedel et al. 2012). An additional factor that could have caused increased mortality in small body sizes was the appearance of a bacterial community exclusive to enriched sediments. From day 10, a white coloured matt growth was observed in the surface sediment of

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MORTALITY AND MOBILITY mesocosms with enriched sediments. The appearance of bacterial mats (i.e. mainly genera Beggiatoa and Thioplaca) has been reported in similar mesocosm experiments and has been linked to a decrease in oxygen concentrations at the sediment surface (Moller et al. 1985; Weissberger et al. 2009). Unfortunately, these bacteria were not identified through the amplicon sequencing (Chapter 3) but they could be related to a dramatic difference in water quality between natural and enriched sediments that produced high overall cockle mortality.

The difference in mortality between different body sizes in bivalves can have an important repercussion in the processes in which they actively participate as bioturbators. For example, filtration and excretion rates in bivalves are known to be affected by body size through the selection of specific particle sizes that can influence the quality of nutrients locally, the particles they can retain in their biodeposits and the effect on sediment redox properties (Vaughn and Hakenkamp, 2001). Results from a loss of specific bivalve body sizes in an ecosystem can lead to severe impacts on local nutrient processing and an alteration of water quality in an ecosystem (Strayer et al. 1999). However, if some bivalve sizes can survive in disturbed conditions and participate actively in local functional processes, it may be possible to propose them as potential models for bioremediation. In this study, large cockles seem to be the most suitable candidates as they can survive high levels of eutrophication and may represent a crucial component for benthic assembles involved in sediment processes (Norkko et al. 2013). In addition, cockles placed in natural sediments with still a high level of enrichment (~8% OM) had no mortality. These results highlight the capacity of this bivalve to tolerate some levels of enrichment and the possibility to be able to reintroduce certain subgroups of their populations in areas highly modified near urban and aquaculture areas.

2.5.2. Associational susceptibility between small and large phenotypes in enriched sediments Associational interaction has been evidenced in terrestrial environments (Barbosa et al. 2009) and within benthic grazer macroinvertebrates in marine

CHAPTER 2 systems (Pfister and Hay 1988). Through these associational interspecific interactions, it was shown how a focal species can increase (associational resistance; Barbosa et al. 2009) or decrease (associational susceptibility; Barbosa et al.2009) this interaction. However, little has been attempted to determine if similar associational interactions can occur at an interspecific level in benthic communities (Creese et al. 1982) and how these can affect the responses of bioturbator communities to disturbances such as organic pollution. In this study, no differences in mortality were found for medium and large phenotypes among monocultures or size combinations. However, this was not the case for small body size phenotypes where small cockles had higher mortality when mixed with other size classes in comparison to monocultures (Fig.2.1B). Smaller organisms have a lower tolerance to enriched sediments than large organisms, but this lower tolerance could be further affected if they need to compete with larger organisms for resources such as food or available dissolved oxygen. Some reports show that intra-specific competition in benthic bivalves occurs and can lead to growth reduction, resource competition and decrease survival (Hamner, 1978; Weinberg, 1985; Vincent et al. 1994; Brichette et al. 2001; Rossi et al. 2004; Beal et al. 2006; Novais et al. 2016). In the present study, an associational susceptibility could have occurred between larger (i.e. medium/large) and small body sizes (Barbosa et al. 2009). It is possible that resource competition (i.e. available dissolved oxygen) might increase mortality of smaller organisms if larger forms are present in eutrophic conditions (Weinberg, 1985). Sublethal effects of sediment enrichment and derived hypoxia have been well documented at different behavioural, physiological or genetic levels for bivalves (Carmichael et al. 2012, Clark et al. 2013). Among the main behavioural effects of hypoxia on benthic bivalves are decreased burial depth, surfacing and siphon extension (Tallqvist, 2001). In the present study, cockle burrowing was not directly quantified however cockles exposed to enriched sediments were found to remain in the sediment surface without any burrowing behaviours detected (pers. obs.). These observations suggest that cockles may have been in a stressed state exclusively in enriched sediments and could have contributed to the reduced mobility pattern

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MORTALITY AND MOBILITY detected (Weissberger et al. 2009). Competition for oxygen in these conditions is possible and may also support the associational susceptibility that was found between small and large phenotypes. The impact of such findings could be crucial for bioremediation efforts as deployment of less diverse bioturbator assemblages could prove a more efficient approach than assemblages with high intraspecific body size variation.

2.5.3. Intraspecific variation in lateral mobility in enriched sediments Cockle overall lateral mobility was reduced in enriched sediments, but no differences in mobility were seen between body size combinations apart from when large cockles were in monocultures. Small cockles in monocultures were less mobile in enriched sediments when the data was analysed by individual size phenotype (Fig. 2.2C, D). In natural conditions, horizontal movement has been reported as a reproductive measure in epibenthic bivalves that is manifested during reproductive aggregation or as a stress response to predation (Tallqvist, 2001; Tettelbach et al. 2017). Intraspecific body size variability is a key factor that determines bivalve movement, with larger organisms being generally less mobile than smaller phenotypes (Norkko et al. 2013). This pattern was supported here, with large cockles in monoculture moving less than smaller phenotypes. Some authors propose this limited mobility in large organisms as a disadvantage for survival in eutrophic sediments (Clark et al. 2013), but this cannot be concluded from the present work as large phenotypes had the highest survival to the end of the experiment. However, when small phenotypes in monocultures were placed in enriched conditions, reduced mobility was observed suggesting that smaller cockle phenotypes are more vulnerable to enriched conditions and lateral movement may reflect a sublethal response to these conditions. As size increased with medium and large phenotypes in monoculture, this sublethal response to enriched sediments was no longer observed.

2.5.4. Large cockle’s lateral movement increases in higher intraspecific diversity When individual phenotypes in monocultures were compared with mixed size

CHAPTER 2 combinations (Fig. 2.2C-H), no differences in mobility were observed for small or medium phenotypes. However, for large phenotypes in both enrichment types, there is a clear increase in mobility when mixed with other size combinations (Fig. 2G-H). These lateral movement results cannot be explained by a density effect as biomass loss caused by mortality in enriched sediments was included in the analysis (i.e. offset in GLMMs) and remained constant in natural conditions. Tettelbach (2017) reports an increase in horizontal movement for the bivalve Mercenaria mercenaria as a response to changes in animal density but also in the presence of other individuals of the same species. This suggests that a possible chemical or environmental stimulus created by the other size phenotypes (as observed in other bivalves; Nelson and Allison 1940; Balfour and Smock 1995; Huang et al. 2007) may have increased the mobility of large phenotypes.

2.5.5. Organic matter loss was not dependant on the presence of the Sydney cockle Movement is known to be one of the components in bioturbation activities with the largest effects on sediment functional characteristics through particle transport and biodiffusion (Braeckman et al. 2010). Higher bioturbator movement on the sediment surface would mean a higher rate of transport and changes in the redox state of the sediments. In addition, high mortality rates of key body sizes may also mean an alteration to normal benthic processes regulating organic matter degradation (Cassidy et al. 2020). In this experiment, organic matter decreased in all treatments including controls with no cockles. This gives evidence of microbial activity in organic matter degradation (Kunihiro et al, 2011) in all the mesocosms, however whether bioturbation by cockles enhances this loss was not clear in natural or enriched sediments. Small cockles in monocultures showed the only evidence of a difference in organic matter degradation between natural and enriched sediments, with higher organic matter breakdown in the latter (p= 0.005). However, the decrease of organic matter was similar to values observed in control mesocosms with no cockles, so a clear link between the presence of A. trapezia and organic breakdown cannot be made. In addition, results above demonstrate that higher mobility of small sizes would

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MORTALITY AND MOBILITY have a higher impact on nutrient cycling on the sediment and pose as an active model species for bioremediation. However, their vulnerability in highly polluted sediments compared to large sizes seen through this study may prove crucial to define the most efficient body size for bioremediation. Large cockles may have a lower rate of movement, but their impact on sediment function may relate more to other bioturbation behaviours such as bioirrigation and higher excretion rates (Norkko et al. 2013). Such mechanisms in large cockles may provide the framework to define an adequate bioturbator model for eutrophic systems. However, correlation of both cockle mobility and mortality were unable to explain the pattern seen in organic matter loss when cockles are present (Appendix 2, Fig.B.2) so a direct causative link to bioturbation cannot be made.

Quantification of organic matter through LOI% is a method that has proven to have high spatial variation and sensitivity to specific equipment used, temperature for incineration and sample size (Heiri et al. 2001). Inherent methodological artifacts of this method could also be a possible explanation of the high variation seen in the percentage of organic matter decrease between replicates in the present experiment. In addition, food regimes may have also played an important role in explaining the results seen in organic matter decrease. It was expected that low and infrequent food addition and high flow- through of water through mesocosms would not affect organic matter decrease detection in the sediments. However, results show that food addition could have also played a major role in the sensitivity of detecting sediment organic matter content by increasing variability between mesocosm replicates (Fig. 2.3.). It is possible that food addition may have produced a high accumulation of organic matter in some mesocosm and this in itself altered the microbial processes involved in nutrient cycling. This in turn with LOI% organic matter detection may have obscured the resolution of the potential effect that A. trapezia may have had on sediment nutrient cycling occurring in the sediment surface.

CHAPTER 2 2.6. Conclusion This study demonstrates the capacity of large cockle phenotypes to tolerate organic enriched conditions better than smaller phenotypes and to increase mobility at higher intraspecific diversity. An associational susceptibility was also found where the presence of larger phenotypes reduced small phenotype’s survival in enriched sediments. A. trapezia did not have a significant effect on sediment organic matter loss and there was no obvious effect of intraspecific diversity on this process. Nevertheless, the higher survival and low effect of organic enrichment on the mobility of large cockles might support the use of this phenotype in sediment bioremediation. However, if potential bioremediation was attempted with large phenotypes of A. trapezia, then exposure to the enriched sediment over a longer time might be necessary to stimulate organic matter degradation (Vadillo Gonzalez et al. 2019). Measurement of nitrification, denitrification rates and sediment oxygen uptake (as suggested by results in Chapter 1) are also needed to increase the resolution in the underlying benthic processes that are involved in organic matter degradation influenced by the presence of A. trapezia. In addition, experiments focusing on a range of sediment organic enrichment (similar to the ones proposed by Gobbard and Solan, 2019) should be attempted to elucidate the role of body size in survivorship and it’s potential role in sediment bioremediation. Nonetheless, results from this study highlight that intraspecific variation of body size can affect the survival and activity of bioturbators in sediments with high organic enrichment. Higher survival of large cockles and possible effects on sediment properties are evidence that body size is a crucial factor to consider if local bioremediation wants to be attempted. As other authors have proposed (e.g. Norkko et al. 2013) large bioturbators are key functional components in natural benthic assemblages. Further research should focus on these large bioturbator sizes that could benefit current bioremediation attempts in eutrophic systems and potentially optimize their effectiveness in removing pollutants.

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MORTALITY AND MOBILITY

CHAPTER 2

3. ORGANIC ENRICHMENT REDUCES MICROBIAL DIVERSITY AND CHANGES 3 COMMUNITY STRUCTURE, BUT THESE EFFECTS ARE NOT MITIGATED BY THE SYDNEY COCKLE (ANADARA TRAPEZIA) 3.1. Abstract Eutrophication is a global issue in many aquatic systems where high input of human- derived. organic matter and nutrients have disrupted essential ecosystem processes. Much of this organic enrichment is transported to benthic soft sedimentary systems where important ecosystem processes are often driven by the coupling of macrofaunal activities and microbial communities. In these systems, it has long been thought that macrofaunal bioturbators strongly influence ecosystem processes by engineering environments that constrain microbial communities. Only recently, have next generation sequencing technologies allowed ecologists to test for the impact of bioturbators on entire microbial communities. Here I explored the effect of a common bioturbator, the Sydney cockle (Anadara trapezia), on bacterial and archaeal richness, diversity, evenness, and community composition. I manipulated both organic enrichment and intra-specific variation of bioturbator body size in a cross-factorial replicated design. A mesocosm experiment was conducted with two treatments of enrichment (natural or spiked with organic matter) fully crossed with 5 bioturbator treatments: 3 cockle body sizes in monoculture (small= 20-30 mm, medium= 35-50 mm and large= 55-70 mm), a treatment with all three sizes represented (i.e. Small+Medium+Large) and a treatment with no cockles that served as a control. Organic enrichment resulted in a substantial decrease in microbial community richness, diversity and evenness, as well as a structural shift towards microbial taxa related to anaerobic metabolism. No effect of bioturbators was detected on microbial communities and there was no difference with variation in intraspecific body size of the Sydney cockle. This study highlights the substantial impact that sediment enrichment can have on microbial community structure and potential function as well as the potentially limited influence some infaunal bivalves, such as the Sydney cockle, may have on microbial communities.

63 CHAPTER 3 3.2. Introduction Benthic macrofauna play an important role in sediments as they can accumulate high quantities of organic matter and act as ‘hot spots’ for nutrient cycling and organic matter accumulation (Shang et al. 2013; Poulsen et al. 2014; Kang et al. 2017). Changes in sediment properties produced by macrofauna include physical alterations (i.e. sediment reworking) and changes in sediment oxic state to enhance coupling of redox reactions such as nitrification and denitrification (i.e. through bioirrigation, Lavrentyev et al. 2000; Papaspyrou et al. 2006; Bertics et al. 2010;). During burrow construction, habitation and through remnant burrows casts, bioturbating macrofauna can affect the surrounding sediment by the production of mucus, nitrogen excretion and faecal pellets that help to establish small-scaled complex microhabitats for microbial communities (i.e. biofilms; Albertelli et al. 1999; Lelevield et al. 2003; Biswas et al. 2009; Wada et al. 2016; Dale et al. 2019). In addition, macrofaunal grazing activities have also been associated with changes in sediment microbial assemblages where bioturbators ingest and process microbial-rich sediments to enable specific responses upon egestion. In some cases, the ingestion and gut passage can quantitatively change microbial communities (i.e. abundance/biomass, Moriarty et al. 1985) or affect them qualitatively (i.e. metabolic diversity, Plante & Mayer, 1996). There is evidence that macrofauna can influence sediment microbial communities by increasing microbial abundances and metabolic activity with important implications for local nutrient dynamics through this digestive process or by sediment reworking or bioirrigation (Papaspyrou et al. 2006; Bertics and Ziebis, 2009; Laverock et al. 2010; Deng et al 2015). The main microbial groups affected by such activities include ammonia oxidizing bacteria (AOB) and ammonia oxidizing archaea (AOA) as well as other groups of nitrifying and denitrifying microbial communities (Altmann et al. 2004; Gilbertson et al. 2012; Quintana et al. 2013; Foshtomi et al. 2015). However, the overall effect of bioturbators on microbial communities is equivocal as some authors have found no evidence to support an increase in microbial abundance in the presence of macrofauna (Gilbertson et al. 2012; Shen et al. 2017; Ma et al. 2015). Some studies suggest that alternative environmental covariates (e.g. temperature, sediment type or salinity) may explain microbial community variation at larger scales (Albertelli et

64 INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES al. 1999; Wilde and Plante, 2002; Bertic and Ziebis 2009; Foshtomi et al. 2015; Wada et al. 2016) as well as inherent properties within a bioturbator assemblage (i.e. intraspecific trait variation; Norkko et al. 2013).

In highly organically enriched sediments, depleted oxygen will create micro-environments only suitable for anaerobic microbial activity and alternative routes of organic matter degradation (Skei et al. 2000; Mermillod-Blondin et al. 2004). Examples of this include an increase in sulphate reducing bacteria (SRB) anaerobic decomposition of organic matter, the production of sediment hydrogen sulphide as a by-product and an overall decrease in aerobic microbial and metabolic diversity (Bartolini et al. 2009; Kunihiro et al. 2011; Howarth et al. 2011; Shen et al. 2016). Recent research indicates that bioturbation enhances bacterial sulphate reduction in organic matter enriched sediments (Bertics and Ziebis, 2010; Bonaglia et al. 2013), with community shifts towards anaerobic bacterial groups that displace aerobic communities (Mchenga et al. 2007; Bonaglia et al. 2013; Boeker et al. 2016). The impact of bioturbation on shifting bacterial communities in eutrophic environments is also influenced by the specific bioturbation mechanism (Nogaro et al. 2014) and the proximity to deeper anaerobic layers where initial anaerobic bacterial communities reside (Van Duyl et al. 1992). In contrast, evidence also indicates that bioturbators can reduce the activity of SRB bacteria and promote sulphide oxidation through bioirrigation in some eutrophic conditions (Kristensen and Blackburn, 1987; Banta et al. 1999; Quintana et al. 2013; Bergström et al. 2017). If eutrophic-tolerant bioturbators reduce SRB activity in organically enriched sediments and increase oxygenation, a shift toward aerobic microbial communities may occur and support the recovery of nutrient dynamics in eutrophic sediments (Kunihiro et al. 2011). Nonetheless, these interactions between the macrofauna and the microbial community in highly enriched systems may differ not only between bioturbator species but also within a single population with high trait variation.

Bioturbators exert significant control over microbial activity however the interactions between both sediment communities and how these interactions

CHAPTER 3 influence nutrient processes and overall ecosystem functioning is not well understood (Vadillo Gonzalez et al. 2019). For example, the impact of these communities on ecosystem function may vary within a bioturbator population that displays intraspecific trait differences. Interspecific trait variation has well documented direct effects on ecosystem functions (Des Roches et al. 2018; Raffard et al. 2019; Bolnick et al. 2011) and some work has been done to investigate the impact of interspecific variability in marine communities, especially for epibenthic sessile invertebrates (Gamfeldt et al. 2005 and Aguirre and Marshall 2012; Hanley et al. 2016). In soft sedimentary systems, attention has mostly been given to interspecific body size variation of benthic macrofauna as an important functional trait that influences overall ecosystem processes (Donadi et al. 2015; Harris et al. 2016; Wrede et al. 2018). However similar research at an intraspecific level is lacking (Norkko et al. 2013). More recent studies, however, show that intraspecific variation has equally important consequences for ecosystem function including sediment ecosystem function and the services that soft sedimentary systems provide (Clark et al. 2013). For example, there have been several cases where subtle changes in phenotype within a population can lead to important changes in community processes such as predation (El-Sabaawi et al. 2018) or habitat selection (Duffy et al. 2010). In particular, several studies have suggested that large body sized organisms are important predictors of ecosystem function in comparison to smaller body sizes as they more strongly influence benthic metabolism (Taylor and Brand, 1975; Carmichael et al. 2012; Clark et al. 2013) and microbial activity (Norkko et al. 2013). The presence of a single specific body size may create more suitable sediment microhabitats for groups of bacteria or archaea that are related to specific nutrient metabolism.Through a mesocosm experiment, this study aimed to evaluate the effect of different body sizes of the Sydney cockle (Anadara trapezia) in monocultures and in combination, on the abundance, diversity, evenness and composition of bacterial and archaeal communities in natural and highly enriched sediments. Infaunal bivalves are important bioturbators in benthic environments and constitute a significant part of the local biomass

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES

(Vaugh and Hakenkamp 2001). Bivalves often possess a high tolerance to low oxygen conditions due to unique mechanisms that allow them to survive and cope with limited concentrations of oxygen (e.g. siphon stretching or burial depth reduction, Vaquer-Sunyer and Duarte, 2008; Wright et al. 2010). However, high bivalve mortality and a decrease in performance do occur in many species in eutrophic coastal areas (Carmichael et al. 2012). A. trapezia is a long-lived epibenthic ark shell bivalve that is abundant in coastal inter- and subtidal areas around south-eastern Australian estuaries from Victoria, New South Wales and South Australia (McKinnon et al 2009; Gribben and Wright, 2006; Lloyd et al. 2012). This bivalve has been known to inhabit sediments with high organic content (~8% organic matter content, see Chapter 2) and have a high mobility variation between different body sizes (see Chapter 2). Due to its high abundance and natural tolerance to organically enriched sediments, this bivalve is a suitable bioturbator to investigate for potential restoration and remediation application. This is especially the case in the aquaculture industry where A. trapezia has gained interest as a candidate for sediment bioremediation (personal communication, DPI Port Stephens Fisheries Institute),

In this study, I hypothesized that microbial diversity, abundance, and evenness to be higher in natural than enriched sediments and that microbial community composition would differ among both treatments as seen in other studies (Deng et al. 2015; Filippini et al. 2019). In addition, I expected that the presence of cockles would increase the abundance and diversity of aerobic microbial communities involved in nitrogen metabolism and organic matter degradation (Foshtomi et al. 2015). Studies in other bivalves have reported that particle aggregation through the production of mucus and fecal pellets, as well as redox changes produced from varying bioirrigation rates, can influence microbial abundance and diversity in sediments (Solan & Wigham, 2005;Volkenborn et al. 2012; Deng et al. 2015). This effect was however also hypothesized to occur between different body size treatments. For example, smaller epibenthic bivalves have been reported to have higher mobility than larger phenotypes (Norkko et al. 2013; Chapter 2). This characteristic may translate to higher sediment reworking,

CHAPTER 3 an increase in sediment oxygenation rates and enhanced organic matter redistribution that may influence microbial microhabitat formation (Deng et al. 2015). In contrast, larger bivalves may provide a unique effect due to their bioturbation activities (Norkko et al 2013) or increase microbial growth through higher surface area for shell biofilms (Heisterkamp et al. 2013). Finally, when high intraspecific diversity is present within a site (i.e. body size combinations), an associational interaction (i.e. refuge or increase susceptibility; see Pfister and Hay, 1988; Barbosa et al. 2009) may appear between bioturbator sizes that could impact the underlying sediment microbial communities. This could generate a distinct microbial habitat with higher microbial abundance, diversity, and greater organic matter loss than when placed in monoculture. Through the results obtained here it will be possible to determine if the different sizes of A. trapezia selected for this study can stimulate bacterial and archaeal communities related to nutrient cycling and assess the potential of these organisms for the bioremediation of eutrophic sediments (adding crucial experimental data to the existing literature on bioremediation, Chapter 1).

3.3. Methods 3.3.1. Study species and collection A. trapezia were collected from silt/sandy sediments in Eraring (33° 04’ 37.66’’ S, 151° 32’ 17.56’’ E, with organic matter content of ~6.4± 0.7 %) in the southern section of Lake Macquarie NSW, Australia. Three size categories were collected based on their shell length (maximum anterior to posterior axis): a) Small: 20-30 mm, b) Medium: 35-50 mm and c) Large: 55-70 mm. A total of 62 cockles were collected with an exact sample size in each size combination category described in detail in Appendix 3, Fig.C1. Further justification of this species selection for bioremediation of highly enriched sediments is given in detail in Chapter section 2.3.1. Collections were made in winter during June and July in an area of ~2500 m2 and 1 to 3 m water depth. All selected sizes co-occurred in the same area and were collected randomly from the sediment. Animals were transported to the Port Stephen Fisheries Institute (PSFI) in Taylor Beach, NSW and placed in flowing seawater (filtered at 1 µm) and acclimated to a water temperature of 22°C for 15

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES d. During this period animals were fed every 2 days with ~14.3 mL/ individual of an algal mix described below.

To investigate the influence of intra-specific body size variation of A. trapezia on bacterial and archaeal communities in enriched sediments, I exposed 3 different body sizes in monocultures (S= Small, M= medium and L= large), a combination of all of these (SML= Small + Medium + Large) and a control group with no cockles (NC). This study builds upon the experimental design described in Chapter 2 and was adjusted to a smaller balanced designed. to two levels of organic enrichment (control and enriched). Control sediments had natural background levels of organic matter content, and enriched sediments were spiked with organic matter (hereafter referred to as natural and enriched sediments, respectively). In total, 10 treatments were created; natural and enriched sediments fully crossed with four bivalve treatments: small, medium and large cockles on their own and all three body sizes together. In addition, a control group with no cockles added was also considered for both natural and enriched sediments to evaluate changes in microbial communities in the absence of A. trapezia. There were 3 replicate mesocosms per treatment and the number of cockles in each mesocosm was maintained by standardizing at 0.06 g/cm2 of biomass (for a similar method of biomass standardization between experimental treatments see Gribben et al. 2020). However, this average value of biomass did not remain constant during the whole experiment due to mortality in the final days of the experiment. To account for this, one mesocosm was eliminated from further analyses (i.e. Medium body size for enriched sediment treatments) and adjustments to statistical procedures had to be made (described below and for more details see (Appendix 3, Fig.C.1).

3.3.2. Mesocosm experimental enrichment setup To ensure sediment grain size and organic matter content were consistent across all mesocosms, sediment collection was done in a single site (i.e. Tilligery creek, Taylor beach), enriched on site and frozen to remove the effects of the majority of existing macro and meiofauna. For further details on sediment collection,

CHAPTER 3 enrichment and preparation, see the methods section (Sediment collection and mesocosm enrichment setup) of Chapter 2. Control sediments (without organic enrichment) had an average of 8 ± 0.3 SE % of organic matter content and enriched sediments had an average of 14 ± 0.7 SE % (calculated from organic matter lost on ignition method modified from Wang et al. 2010; see Chapter 2 for further details). After thawing, ~8 L of sediment was added to 11.5 L experimental mesocosms and placed in a continuous flow-through system for acclimation. After 12 d, enriched sediments were confirmed to have lower levels of dissolved oxygen (66 ± 0.3 mg/L) than natural sediments (90 ± 1.5 mg/L) and cockles were added in accordance with the established experimental design and body size combinations (Appendix 3, Fig. C1). Additional details of the experimental setup and flow-through system, including justifications of organic matter introduced through food regimes are given in the same section as above in Chapter 2. The experiment lasted a total of 16 d however microbial analysis was done on samples taken on day 10 as high cockle mortality in enriched conditions was registered after this point (see Chapter 2).

3.3.3. Sample collection, DNA extraction and sequencing Samples were taken at the sediment surface (between 0-3 cm in depth) with a 10 mL syringe constructed as a miniature sediment corer that could sample surface sediments. Three samples from each mesocosm were taken randomly from different areas of the sediment surface and pooled into a single sample per each mesocosm to give a total of 30 samples (15 natural and 15 in enriched condition, for five different cockle sizes: S= Small, M= Medium, L=Large, SML= all sizes and a control group NC= no cockles) that were immediately frozen at -80 °C for later analysis. Genomic DNA extraction from the sediment was done using the DNeasy Powersoil kit® and DNeasy PowerBiofilm® (Qiagen) following the specifications of the manufacturer’s protocol. DNA concentration and purity were measured using a Full spectrum microvolume NanoDrop 2000® (Thermo Fisher scientific). Sequencing was done on a MiSeq v3 Sequencing Platform (Illumina) following the manufacturer’s guidelines. Amplicons targeting the bacterial 16S

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES rRNA gene region V1-3 (27F-519R) and archaeal 16S (A2F-519R) were amplified and sequenced by the Ramaciotti Centre for Genomics (UNSW, Australia). All raw sequences for this study were submitted in the NCBI Sequence Read Archive (SRA) database (Bioproject accession ID for both bacteria and archaea: PRJNA596620).

3.3.4. Bioinformatic sequence analysis Sequenced data for both microbial fractions (i.e. bacteria and archaea) were attained as raw fastq sequences and quality trimmed using the software Trimmomatic v. 0.36 (Bolger et al. 2014) to remove distal low-quality bases (Min length= 100; Sliding Window= 4:10 for both microbial fractions). Trimmed sequences for both fractions were similarly merged (min length 400 and max length 600 base pairs) with a mean percent merge of 69 ± 0.6 % SE for bacteria and 78 ± 1.4 % SE for archaea, and quality filtered with maximum expected error threshold of 1. After this, all singletons were removed for both microbial fractions. All reads were denoised into zero-radius operational taxonomic units (zOTUs) to acquire the maximum possible biological resolution (Edgar et al. 2016). Chimeras from denoised unique sequences were removed initially with the UNOISE3 and to improve chimeric detection, a secondary removal was done using the UCHIME2 algorithm (Edgar et al. 2011) and with the SILVA Small Subunit rRNA reference database (SILVA 132 SSU Ref NR99). Taxonomic assignment was achieved using the same reference database as above through a BLAST algorithm to compare the processed sequences with the reference database. Processed sequences were subsequently mapped into zOTU sequences and total abundance (i.e. number of sequences) was calculated for each sample in both microbial fractions separately. Filtering of chloroplast and mitochondria in both bacterial and archaeal zOTU tables was done before proceeding with statistical analysis. Final total number of zOTU for bacterial communities was 19,874 and 12,361 for archaeal communities. All the bioinformatic pipeline was done through R using the software USEARCH v.11.0.667.

CHAPTER 3

3.3.5. Statistical analysis Microbial abundance zOTU tables were rarefied using the function rarefy function (package vegan, Oksanen et al. 2019) to eliminate any zOTU present in only one sample. To evaluate if total zOTU abundance was assessed effectively during sampling, rarefaction curves were constructed (R, vegan; Oksanen et al. 2019, Appendix 3, Fig. C2) and Good’s coverage was calculated (R, QsRutils; Quensen 2019; Appendix 3, Table C.1) for both microbial fractions.

For each sample, α-diversity indices were calculated including zOTU richness, diversity (Shannon diversity Index) and evenness (Pielou Index) using the package vegan (Oksanen et al. 2019). A general linear model (GLM) using the function lm from the basic statistical package in R (R Core team, 2019), was constructed for each α-diversity index to determine differences between enrichment treatments (i.e. natural vs enriched), body size combinations (i.e. NC, S, M, L and SML) and any interactions between both. When significant effects of the fixed factors on the α-diversity indices were found, least square means pairwise comparisons were calculated, and p values adjusted (i.e. Bonferroni) using the function emmeans from the package emmeans (Russell, 2019). For all tests in both microbial fractions, fixed factor inferences were done through a likelihood ratio test with the anova function from the package car in R (Fox and Weisberg, 2018). Bacterial Shannon diversity and evenness data were cubed transformed to fulfill GLM assumptions of normality and homoscedasticity (see Appendix 3, Fig. C3 and C4 for all statistical model assumptions diagnostics for both bacteria and archaea).

Microbial community composition was evaluated with a comparison of the relative abundance of the main bacterial and archaeal phyla present in the samples. Bacterial phyla with a percent relative abundance of less than 1% were grouped in a single category named ‘Others’ in which the following phyla were included: Acetothermia, Aegiribacteria, AncK6, Armatimonadetes, Atribacteria, BHI80-139, BRC1, Chlamydiae, CK-2C2-2, Cloacimonetes, Cyanobacteria, Dadabacteria, Deinococcus-Thermus, Dependentiae, Elusimicrobia,

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES

Entotheonellaeota, FCPU426, GN01, Halanaerobiaeota, Hydrogenedentes, Hydrothermae, Kiritimatiellaeota, LCP-89, Lentisphaerae, Margulisbacteria, Marinimicrobia, SAR406 clade, Modulibacteria, Nitrospinae, Omnitrophicaeota, PAUC34f, Rokubacteria, Schekmanbacteria, Synergistetes, WOR-1, WS1, WS2, WS4. For archaea, all phyla relative abundance data was included. Similar GLMs as done for α-diversity indices were fitted to evaluate differences in relative abundance (%) between bacterial and archaeal phyla (i.e. each domain analyzed independently), enrichment type and body size combinations. For both microbial fractions, relative abundance data was arcsine transformed to satisfy model assumptions (Appendix 3, Fig. C5 and C7). No GLMs were analysed for the bacterial category ‘Others’.

Differences in microbial community composition were assessed through a permutational multivariate analysis of variance (PERMANOVA, function adonis, R package vegan; Oksanen 2019) by using calculated Bray-Curtis dissimilarity matrices and assessing compositional differences between enrichment types and body size combinations. PERMANOVA has proven to be a robust statistical method that is not sensitive to changes in data distribution however it is highly sensitive to high grouping variance (Anderson et al. 2017). A multivariate homogeneity of group dispersions was done to evaluate variance homogeneity for the adonis multivariate model using the function betadisper and differences in the PERMANOVA were contrasted with ordination plots(R package vegan; Oksanen 2019). A multidimensional scaling method Principal Coordinate Analysis (PCoA) was done to graphically demonstrate PERMANOVA results in both microbial communities.

A similarity percentage analysis (SIMPER) was done using the calculated Bray-Curtis dissimilarity matrix to identify the specific bacterial and archaea zOTU that contributed the most to the differences between natural and enriched sediments (function simper, permutations= 999, R package vegan; Oksanen 2019). Only specific zOTUs with a contribution higher than 1% of total dissimilarity were included for bacteria. When microbial genera were not

CHAPTER 3 identified, GLMs were analysed at the lowest taxonomic level that could be resolved. Similar GLMs as the phyla analysis above were done with the relative abundance (%) of the selected genera as the response variable and genera, enrichment, and body size treatments as fixed factors. Interactions between fixed factors were also considered in the analysis and model assumptions of normality and homoscedasticity were diagnosed for both bacteria and archaea (Appendix 3, Fig. C6 and C8).For bacteria SIMPER analysis, data were logit transformed to fulfill model assumptions for analysis.

3.4. Results 3.4.1. Lower microbial α-diversity indices in enriched sediments All cockles in the natural sediments survived to the end of the experiment. By contrast, total mortality was high in the enriched sediments (from 37.5 % to 100 % in some mesocosms, details in Chapter 2) which affected the number of mesocosms included in the present analysis (Appendix 3, Fig. C1). Rarefaction curves and calculated Good’s coverage show that sampling of natural and enriched sediments effectively assessed the richness of both bacterial (Appendix 3, Table C1 and Fig. C2; coverage: natural sediments= 98 ± 0.1 % SE and enriched sediments= 99 ± 0.1 % SE) and archaeal communities (Appendix 3, Table C1 and Fig. C2; coverage: natural sediments= 99 ± 0.1 % SE and enriched sediments= 99 ± 0.1 % SE). Final total numbers of zOTUs for bacterial communities after the bioinformatic pipeline was 19,874 and 12,361 for archaeal communities.

For both bacterial and archaeal communities, zOTU richness, diversity and evenness decreased in enriched sediments compared to natural conditions, however the decrease was larger for bacteria (Fig.3.1, p<0.001; richness decrease in bacteria= 54 % and archaea= 17 %; diversity decrease in bacteria= 33 % and archaea= 4 %; evenness decrease in bacteria= 28 % and archaea= 3 %). None of the cockle treatments that included different body sizes of the Sydney cockles in monocultures (i.e. S or M or L) or in combination (e.g. S and M and L) had any effect on bacterial or archaeal α-diversity indices (Table 3.1, p>0.05). A summary

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES of all α diversity indices for both microbial fractions is shown in Appendix 3, Table C1 and corresponding statistical tests are described in Table 3.1.

Table 3.1. Summary of general linear models (GLMs) in which sediment type (natural and enriched) and cockle treatments (No cockles, small, medium, large, all sizes) were evaluated against each α-diversity index from bacterial and archaeal communities. Significant differences are shown in bold (p<0.05). Sediment type Interaction α Diversity Index Cockle treatments Sediment type*cockle Natural - Enriched treatments Bacteria df F p df F p df F p zOTU Richness 1 1182.5 <0.001 4 1.95 0.14 4 2.64 0.07 zOTU Shannon Diversity 1 1541.7 <0.001 4 0.35 0.84 4 1.70 0.19 zOTU Pielou Evenness 1 1026.1 <0.001 4 1.53 0.23 4 1.29 0.31 Archaea zOTU Richness 1 28.07 <0.001 4 0.49 0.74 4 0.58 0.68 zOTU Shannon Diversity 1 86.58 <0.001 4 0.81 0.53 4 0.94 0.46 zOTU Pielou Evenness 1 87.48 <0.001 4 1.22 0.33 4 2.58 0.07

CHAPTER 3

Fig. 3.1. α-diversity indices for A) Bacterial zOTU richness, B) archaeal zOTU richness, C) bacterial zOTU diversity, D) archaeal zOTU diversity, E) bacterial zOTU evenness and F) archaeal zOTU evenness between natural (8% OM) and enriched (14%OM) treatments. Data is expressed as mean ± SD. Significant differences between sediment types within each index is shown with an asterisk.

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES

3.4.2. Bacterial composition differed between natural and enriched conditions Bacterial community composition differed between natural and enriched sediments (PCA, Fig.3.2; PERMANOVA Table 3.2, p= 0.001). No effects of the different cockle body size combinations (PERMANOVA Table 3.2, p= 0.619) or interactions between sediment type and body size combinations on bacterial community composition were found (PERMANOVA Table 3.2, p= 0.613). A multivariate homogeneity of group dispersion analysis was then applied to the model to determine differences in community variance between natural and enriched sediments. Results show that enriched sediments had a higher variance than natural sediments (Table 3.2, p< 0.001) however there is a clear difference between both treatments as shown in the location of the centroid of both groups in the ordination (Fig. 3.2).

Table 3.2. Summary of PERMANOVA and Multivariate homogeneity of group dispersion analysis in which sediment type (natural and enriched), cockle treatments (No cockles, small, medium, large, all sizes) and interactions were evaluated against community dissimilarity (Bray-Curtis distances) from bacterial and archaeal communities. Significant differences are shown in bold (p<0.05). Sediment type Interaction Multivariate test Cockle treatments Sediment type*cockle Natural - Enriched treatments Bacteria df F R2 p df F R2 p df F R2 p PERMANOVA 1 68.56 0.73 0.001 4 0.92 0.04 0.48 4 0.81 0.03 0.57 Group dispersion 1 37.30 -- <0.001 Not evaluated Not evaluated (betadisper) Archaea PERMANOVA 1 16.78 0.39 0.001 4 0.99 0.09 0.48 4 0.98 0.09 0.45 Group dispersion 1 8.54 -- 0.01 Not evaluated Not evaluated (betadisper)

When relative abundance of bacterial phyla is analysed, results show high differences between phyla (Fig. 3.3 and Table 3.3, p<0.001) and enrichment type (Fig. 3.3 and Table 3.3, p<0.001), as well as an interactive effect between both (Table 3.3 p<0.001). Natural sediments present dominance of the phyla Proteobacteria (44.5 %), Chloroflexi (19.3 %) and Bacteroidetes (11.9 %) while in

CHAPTER 3 enriched conditions, a higher relative abundance of Bacteroidetes (54 %), Firmicutes (13.2 %) and Fusobacteria (5.1 %), and lower relative abundances of Proteobacteria (12.3 %) and Chloroflexi (2.3 %) were found (Fig.3.3 and Table 3.3). However, no differences in relative abundance were found between the body size treatments (Table 3.3, p= 0.78) or an interactive effect of this factor with sediment type (Table 3.3, p=0.91) or the different phyla (Table 3.3, p=0.63). A complete list of bacterial phyla relative abundance can be found in Appendix 3, Table C2.

Similarity analysis determined that 19827 genera contributed to 91 % of dissimilarities between bacterial communities in natural and enriched sediments. From these, 32 % were unidentified genera mainly from the phylum Proteobacteria (12 %, mainly the class Gammaproteobacteria, 5% and the family Desulfobulbaceae, 3%), Chloroflexi (6%, mainly class Anaerolinae) and Bacteroidetes (6%, mainly the class Bacteroidia and family Dysgonomonadaceae, 2%). A complete description of % contribution of unidentified bacteria to overall dissimilarities between bacterial community composition in natural and enriched sediments is shown in Appendix 3, Table C3 Twenty-one bacterial genera were identified to contribute more than 1 % to a total of 59% of community composition dissimilarity (Appendix 3, Table C4.). From these identified genera, 13 were found to increase their relative abundance in enriched sediments and 8 to decrease (Table 3.4, p<0.001). Predominantly genera of the phylum Bacteroidetes (Bacteroides, Labilibacter, Marinifilum, Draconibacterium, Sunxiuqinia, Saccharicrinis and Carboxylicivirga), Spirochaetes (Sphaerochaetes) and Proteobacteria (Desulfobacter, Halodesulfovibrio and Vibrio) increased in relative abundances in enriched conditions compared to natural sediments (Fig. 3.4 and Table 3.4, p<0.05). One genus of the phylum Fusobacteria (Fusobacterium) and one from Firmicutes (Ruminococcus 1) also increased in relative abundances in enriched sediments (Fig. 3.4 and Table 3.4, p<0.05). In contrast, six genera from the phylum Proteobacteria (Thiohalophilus, Sva0081 sediment group, Woeseia, Thiogranum, SEEP-SRB1 and Sulfurovum), one genus from Firmicutes

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES

(Fusibacteria) and one genus from Fusobacteria (Psychrilyobacter) were relatively less abundant in enriched than natural sediments (Fig. 3.4 and Table 3.4, p<0.05). No similarity analysis of cockle body size treatments or interactions with sediment type was attempted as PERMANOVA found no differences in community dissimilarity between these groups (Table 3.2).

Fig. 3.2. Metric multidimensional scaling ordination (PCA, dimensions= 2) where, coloured outlines denote discrete enrichment groups and labels indicate cockle treatment categories for each point (NC= No cockles, S=Small, M= Medium, L=Large and SML= Small + Medium + Large).

CHAPTER 3

Fig. 3.3. Relative abundances (%) of bacterial phyla in natural and enriched sediments (all phyla with smaller than 5 % of abundance where grouped in ‘Others’, see Appendix 3, Table C.2 for all bacterial phyla registered).

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES

Table 3.3. Summary of a) general linear model results evaluating differences in relative abundance of bacterial and archaeal phyla between natural and enriched sediment treatments and body size combinations. b) emmeans contrasts of of bacterial and archaeal phyla for both sediment types. Significant differences (p<0.05) are shown in bold and p values in contrasts have been adjusted for multiple testing using Bonferroni correction. a) General linear model Bacteria Archaea Factors df F p value Factors df F p value Phylum identity 20 947.41 <0.001 Phylum identity 6 4132.17 <0.001 Body size 4 0.44 0.781 Body size 4 0.15 0.963 Enrichment 1 177.93 <0.001 Enrichment 1 14.48 0.000 Phylum identity*Body size 80 0.94 0.633 Phylum identity*Body size 24 0.63 0.904 Phylum identity*Enrichment 20 310.73 <0.001 Phylum identity*Enrichment 6 194.74 <0.001 Body size*Enrichment 4 0.23 0.919 Body size*Enrichment 4 0.11 0.978 Phylum identity*Enrichment Phylum identity*Enrichment 80 0.93 0.640 24 1.20 0.259 *Body size *Body size b) Multiple comparisons for enrichment and phyla identity Enrichment Enrichment Bacterial Phylum Natural - Enriched Archaeal Phylum Natural - Enriched df t.ratio p.value df t.ratio p.value Acidobacteria 378 -13.11 <.0001 Crenarchaeota 126 23.196 <.0001 Actinobacteria 378 -8.119 <.0001 Euryarchaeota 126 -23.929 <.0001 Bacteroidetes 378 49.797 <.0001 Asgardaeota 126 3.172 0.0019 Calditrichaeota 378 10.403 <.0001 Nanoarchaeaeota 126 -8.093 <.0001 Chloroflexi 378 29.675 <.0001 Thaumarchaeota 126 -3.827 0.0002 Epsilonbacteraeota 378 -7.123 <.0001 Hadesarchaeaeota 126 -0.093 0.9257 Fibrobacteres 378 3.608 0.0004 Korarchaeota 126 -0.527 0.5988 Firmicutes 378 18.273 <.0001 Fusobacteria 378 8.108 <.0001 Gemmatimonadetes 378 -5.493 <.0001 Latescibacteria 378 -7.157 <.0001 Nitrospirae 378 -6.048 <.0001 Others 378 -4.065 0.0001 Patescibacteria 378 -5.054 <.0001 Planctomycetes 378 -11.679 <.0001 Proteobacteria 378 39.188 <.0001 Spirochaetes 378 17.445 <.0001 TA06 378 -4.353 <.0001 Tenericutes 378 2.584 0.0101 Verrucomicrobia 378 -4.678 <.0001 Zixibacteria 378 -4.616 <.0001

CHAPTER 3

Table 3.4. Summary of a) general linear model results for identified bacterial genera found to contribute to dissimilarities between natural and enriched sediment treatments and b) emmeans contrasts of each identified genera in both sediment types. Significant differences (p<0.05) are shown in bold and p values in contrasts have been adjusted for multiple testing using Bonferroni correction. All significant differences are marked in bold. Sediment type Genera Interaction General linear model Natural - Enriched (n= 21) Sediment type*genera a) General Linear model results Response variable df F p df F p df F p Relative abundance (%) 1 1160.53 <0.001 20 42.42 <0.001 20 189.98 <0.001 b) Contrasts of main identified genera (contribution >1%) in natural vs enriched sediments Genera df t ratio p Bacteroides 546 26.884 <.0001 Carboxylicivirga 546 7.922 <.0001 Desulfobacter 546 12.948 <.0001 Draconibacterium 546 9.818 <.0001 Fusibacter 546 -10.152 <.0001 Fusobacterium 546 23.628 <.0001 Halodesulfovibrio 546 18.316 <.0001 Labilibacter 546 18.863 <.0001 Marinifilum 546 11.187 <.0001 Psychrilyobacter 546 -1.299 0.1943 Ruminococcus 546 19.086 <.0001 Saccharicrinis 546 20.586 <.0001 SEEP-SRB1 546 -10.105 <.0001 Sphaerochaeta 546 22.38 <.0001 Sulfurovum 546 -8.707 <.0001 Sunxiuqinia 546 22.537 <.0001 Sva0081 546 -9.729 <.0001 Thiogranum 546 -10.991 <.0001 Thiohalophilus 546 -8.337 <.0001 Vibrio 546 9.878 <.0001 Woeseia 546 -8.6 <.0001

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES

Fig. 3.4. Relative abundance of identified genera with >1% contribution to overall community dissimilarity between natural and enriched sediments. All genera with higher relative abundance in enriched or natural sediments where found to be significant (p<0.001). All relative abundance data (%) is expressed as means ± SE.

CHAPTER 3

3.4.3. Archaeal composition differed between natural and enriched conditions Archaeal community composition differed between natural and enriched sediments (PCoA, Fig. 3.5, PERMANOVA, Table 3.2: p= 0.001). However, no evidence was found that cockle body size treatments were involved in archaeal community dissimilarities (PERMANOVA, Table 3.2: p= 0.48) or in interaction with sediment enrichment (PERMANOVA, Table 3.2: p= 0.45). Results from the multivariate homogeneity of group dispersions demonstrate that higher intragroup dispersion is present in natural sediment compared to enriched (Table 3.2: p= 0.007). This difference however is not enough to invalidate results from the PERMANOVA as inter-group differences are still noticeable with centroids of each sediment type separated in location in the PCoA ordination (Fig. 3.5).

Crenarchaeota was the dominant phyla in both sediment types but had higher relative abundances in enriched sediments (Fig. 3.6 and Table 3.3, natural= 53 % and enriched= 76 %). This was also observed in the third most abundant phyla, Asgardaeota, where a small increase in relative abundance occurred in enriched conditions with 7 % compared to 6 % in natural conditions (Fig. 3.6 and Table 3.3, p<0.001). In contrast, the second most abundant phyla Euryarchaeota, was relatively less abundant in enriched sediments (15 %) compared to natural conditions (37 %, Fig. 3.6 and Table 3.3,). Relative abundances were also lower in enriched sediments compared to natural in the phyla Nanoarchaeaeota (72 % decrease, Fig. 3.6 and Table 3.3,), Thaumarchaeota (60 % decrease, Fig. 3.6 and Table 3.3,) and Korarcaheota (73% decrease, Fig. 3.6 and Table 3.3,). No differences in relative abundances were found for the phylum Hadesarchaeaeota between both sediment types (Fig. 3.6 and Table 3.3, p= 0.93). In addition, no differences in archaeal phyla’s relative abundance were found between cockle body size treatments (Table 3.3, p= 0.96) and this factor interaction with sediment type (Table 3.3, p= 0.98) or between phyla identity (Table 3.3, p= 0.90). A summary of changes in archaeal phyla relative abundances can be seen in Appendix 3, Table. C6.

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES

Fig. 3.5. Archaea composition analysis between natural and enriched sediment treatments .Metric multidimensional scaling ordination (PCA, dimensions= 2, stress= 0.05) where, coloured outlines denote discrete enrichment groups and labels indicate cockle treatment categories for each point (NC= No cockles, S=Small, M= Medium, L=Large and SML= Small + Medium + Large).

CHAPTER 3

Fig. 3.6. Archaea composition analysis between natural and enriched sediment treatments. Relative abundances (%) of archaea phyla in natural and enriched conditions.

Similarity analysis for archaea determined 12360 zOTUs that contributed to the community dissimilarities between natural and enriched sediments. The archaeal community overall dissimilarity between sediment types was 39%. However, it was not possible to resolve the taxonomy to genus level, so classification was done at a class level with 15 classes of archaea identified to contribute 38.8% of the dissimilarity (Appendix 3, Table C6). These include the classes Bachyarchaeia (21 % contribution) and Thermoplasmata (14 % contribution) which were the most abundant, followed by Lokiarchaeia (2 % contribution), Nanohaloarchaeia (2 % contribution) and Nitrososphaeria (1 % contribution). However, 0.2% of contribution to dissimilarities between sediment types was also found to be explained by unidentified classes. Model results comparing relative abundances from identified archaeal classes demonstrate an

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES increase in relative abundances of Bachyarchaeia (Fig. 3.7 and Table 3.5, 31 % increase, Fig. 3.7 and Table 3.5, p<0.001) and Lokiarchaeia (Fig. 3.7 and Table 3.5, 22 % increase, p<0.001) in enriched sediments, while the opposite was observed for Thermoplasmata (Fig. 3.7 and Table 3.5, 58 % decrease, p<0.001) and Nanohaloarchaeia (Fig. 3.7 and Table 3.5, 72 % decrease, p<0.001) for the same sediment conditions. However, no changes in relative abundance were detected for the remaining identified archaea classes (Table 3.3). No effect of cockle body size combinations or interactions between the main treatments were evaluated in the similarity analysis as no differences in community composition were found in the PERMANOVA (Table 3.2).

Table 3.5. Summary of a) general linear model results for identified archaeal classes found to contribute to dissimilarities between natural and enriched sediment treatments and b) emmeans contrasts of each identified classes in both sediment types. Significant differences (p<0.05) are shown in bold and p values in contrasts have been adjusted for multiple testing using Bonferroni correction. Sediment type Classes Interaction General linear model Natural - Enriched (n= 15) Sediment type*classes a) General Linear model results Response variable df F p df F p df F p Relative abundance (%) 1 0.012 0.91 14 4234.01 <0.001 14 248.38 <0.001 b) Contrasts of main identified classes in natural vs enriched sediments Classes df t ratio p Bathyarchaeia 390 43.071 <.0001 Thermoplasmata 390 -39.93 <.000 Lokiarchaeia 390 2.717 0.0069 Nanohaloarchaeia 390 -4.249 <.0001 Nitrososphaeria 390 -1.511 0.1316 Group 1.1c 390 -0.054 0.957 Thermococci 390 -0.153 0.8783 Odinarchaeia 390 -0.044 0.965 Halobacteria 390 0.049 0.9607 Marine Benthic Group A 390 -0.3 0.7643 Crenarchaeota Incertae Sedis 390 -0.004 0.9971 Heimdallarchaeia 390 -0.015 0.9878 SCGC AB-179-E04 390 -0.005 0.9963 Korarchaeia 390 -0.014 0.9891 Methanococci 390 0 1

CHAPTER 3

Fig. 3.7. Relative abundance of identified archaea classes that contributed to overall community dissimilarity between natural and enriched sediments. Significant increase in class relative abundance between enrichment treatments is marked with an Asterix (8). All relative abundance data (%) is expressed as means ± SE.

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES

3.5. Discussion Macrofauna are important regulators of ecosystem processes through high impact of bioturbating activities on sediment microbial communities (Shen et al. 2017). Bioturbation can change the physicochemical properties of the sediments, act as a sink for organic matter resources and provide heterogenous microniches with different microbial driven metabolic capacities (Foshtomi et al. 2015). Little is known about the effect that different intraspecific functional traits such as different body sizes of bioturbators can have on sediment microbial communities and how can these relationships change towards disturbances such as organic enrichment. Here, a mesocosm experiment was conducted to test the effect of different body sizes in monoculture and in combination, of a highly abundant infaunal ark shell bivalve, the Sydney cockle (A. trapezia), on bacterial and archaeal communities in natural and enriched sediments. Mobility of A. trapezia has been shown to change between different body sizes with less mobility in large cockles compared to smaller sizes (Chapter 2). Higher mobility of smaller sizes could mean an increase in sediment reworking rates and possibly an impact on biogeochemical processes in the sediment. Differences in cockle body size were thus expected to influence sediment microbial communities. However, contrary to my original hypotheses, A. trapezia had no influence on alpha diversity or community composition. Organic enrichment resulted in expected substantial decreases of microbial community richness, diversity and evenness, as well as a structural shift towards microbial taxa related to anaerobic metabolism. This study highlights the substantial impact that sediment enrichment can have on microbial community diversity and potential function. In addition, this study also demonstrates the capacity of specific infaunal bivalves to mitigate the effects of enrichment and their limits under highly enriched conditions.

CHAPTER 3

3.5.1. The Sydney cockle had no influence on sediment microbial communities Bioturbation has been shown to increase microbial abundance (Papaspyrou et al. 2006; Wada et al. 2016) and metabolic activity in sediments (Bertics and Ziebis 2009; 2010; Shen et al. 2017; Vadillo Gonzalez et al. 2019). For example, an increase in sediment archaeal diversity in the presence of the arkshell, Tegillarca granosa, a similar bivalve to the Sydney cockle has been reported in intertidal mudflats sediments (Deng et al. 2015). Here, increased oxygen bioirrigation through filtration mechanisms and production of feces and organic compounds derived from T. granosa were thought to be involved in archaeal community change. Similar mechanisms of bioturbation were expected to occur in the present study, and I expected the effect to differ between body sizes of A. trapezia. However, results provided no evidence that the body size diversity of A. trapezia influences microbial communities similar to Tegillarca granosa (Deng et al. 2015). The Sydney cockle, A. trapezia, is an abundant infaunal bivalve in many subtidal habitats around Australia that can inhabit sediments with low oxygen content (Gribben et al. 2009). In addition, high intra-specific body size variation in lateral movement over the sediment surface (i.e. smaller cockles move more than large ones) has been found in natural sediments, although this mobility can be reduced in enriched conditions (Chapter 2). Differences in tolerance to hypoxic conditions and higher lateral mobility between different body sizes might predict a differing impact on benthic metabolism and a substantial influence on the sediment microbial communities by a specific body size assemblage. In this study, however, no such effect of A. trapezia was found. Results here point to the possibility that the effect of bioturbators on sediment microbial communities may be more taxa dependent than thought before as similar studies with other bioturbator taxa have had no effect on sediment microbial communities (Laverock et al. 2010; Gilbertson et al. 2012; Shen et al. 2017).

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES

For A. trapezia in this study, its presence and intra-specific body size variation did not affect microbial communities or even stimulated organic matter breakdown (see Chapter 2). It is possible that the high sediment enrichment to which A. trapezia was exposed may have been so high as to obscure a difference between body size combinations in this enrichment condition. Additional evidence of this can be seen through the PCA plots where enriched sediments produced higher variation between body size replicates in both bacterial and archaeal communities. However, another possibility lies in the experimental design chosen for this study where a single bioturbator (A. trapezia) was exposed and evaluated. In the field and other mesocosm experiments, sediments are constantly influenced by several species of macrofauna and meiofauna which can interact to influence benthic processes (Bongalia et al. 2014). It is possible that the effects expected to happen in presence of the different sizes of A. trapezia needed other interspecific interactions that could not happen with only one species of macrofauna represented. In addition, covariates like cockle density could also play a major role in shaping sediment microbial communities as seen in Chapter 1 (Vadillo Gonzalez et al. 2019), but not evaluated here. Species richness and animal density are important components that determine many of the benthic processes that microbial communities regulate. Lower diversity (only one species present) and density may impact the capacity of sediment communities to upregulate these benthic processes (Godbold and Solan, 2009). This may suggest that A. trapezia may have a small effect on sediment ecosystem processes and their intraspecific body size variation is not an important functional trait that can influence local microbial communities.

3.5.2. Sediment enrichment decreased bacterial α-diversity and changed community composition A rapid decrease in bacterial abundance in similar mesocosm experiments simulating organic enriched systems has previously been demonstrated (Riemann et al. 2000; Bunch and Bernot, 2012). It is well known that sediment microbial communities are sensitive to changes in organic matter these shifts can limit their available resources and alter the of physicochemical properties locally

CHAPTER 3

(Findlay et al. 2003). The decrease in abundance, diversity and community composition are common with increased organic matter to levels close to eutrophication (Filippini et al. 2019). Dominance of some groups over others can also occur under these conditions and determine much of the metabolic output that regulates ecosystem functions (Bai et al. 2012). In the present study, similar results to these were found with lower richness, diversity, and evenness found in sediments with high organic matter enrichment for bacterial communities. This study however extends our understanding of these effects on bacterial communities by analyzing higher levels of organic enrichment that simulate real cases of anthropogenic-derived organic pollution in urban areas (Filippini et al. 2019). Decreases in bacterial community biomass, diversity and evenness are likely a direct effect of changes in sediment properties, where high organic matter loads can produce increased sediment aerobic microbial metabolism that quickly depletes the available oxygen. Under these conditions, sediments become anoxic and alternate bacterial metabolic pathways are taken (i.e. sulphate reduction) that lead to sulphurisation of organic matter for burial or degradation (Wasmund et al. 2017). Anoxic sediments and sulphur metabolic pathways can also lead to high production of sulphides that may be toxic to certain sediment microbial species and eventually to local macrobenthic fauna (Kunihiro et al. 2011; Shen et al. 2016). The consequences of such a chemical shift should also be reflected in changes to the bacterial community composition and the overall functionality of the sediment.

The severe reduction of bacterial α-diversity indices provides strong evidence of a rapid effect of enrichment on sediment properties that shaped microbial communities. However, an even stronger microbial shift was observed for the bacterial community composition between sediment types. For natural systems, the phyla Proteobacteria and Chloroflexi remained the dominant groups, however there was a drastic reduction of both of these groups in enriched sediments and a drastic increase in the abundance of Bacteroidetes and Firmicutes. The increase in the relative abundance of Bacteroidetes and its new position as dominant phyla in enriched sediments may reflect a substantial shift

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES towards anaerobic sediment conditions that are indicative of eutrophic sediments (De Figueiredo et al. 2007; Tang et al. 2009; Wasmund et al. 2017). Bacteroidetes are the third most abundant group of bacteria in most marine systems (after Proteobacteria and Cyanobacteria) and represent a large proportion of the local bacterioplankton in coastal areas (Fernandez-Gomez et al. 2013). However, after events of high organic matter release (i.e. algal blooms or human-derived discharges), their biomass can increase dramatically as they can benefit from the availability of high molecular weight organic polymers which other groups cannot (Fernandez-Gomez et al. 2013). Under highly enriched conditions a shift to favour the dominance of this phylum is expected, and the results found here support this scenario.

The highest proportion of bacteria that were different between enrichment sediment types include unidentified genera that could largely be assigned to the families Desulfobulbaceae (class Deltaproteobacteria), Anaerolineaceae (class Anaerolineae) and Dysgonomonadaceae (class Bacteroidia). These families are often associated with hypoxic sediments and anaerobic conditions where their main metabolic pathway is sulphate reduction (Kuever et al. 2014; Gribben et al. 2017; Yan et al. 2018). In addition to these families, the dominant identified genus that contributed the most to the dissimilarities between natural and enriched bacterial communities was Bacteroides. This genus has been associated with anaerobic conditions however it is mostly known as a biomarker of fecal pollution from warm-blooded animals (Ahmed et al. 2016). As the enrichment was done using animal-based manure (i.e. chicken, Yates Dynamic lifter®) it is possible that the high abundances of the genus Bacteroides could be directly related to this source of enrichment more than the experimental conditions. However, from the total 21 genera identified to be involved with dissimilarities between bacterial community composition in natural and enriched sediments, 52 % have been reported to proliferate in anaerobic conditions or participate in sulphide reducing metabolic pathways (Sphaerochaeta, Labilibacter, Marinifilum, Fusobacterium, Thiohalophilus, Sva0081 sediment group, Carboxylicivirga, SEEP-SRB1, Sulfurovum, Halodesulfovibrio and Vibrio) (for

CHAPTER 3 specific functional roles of the genera from which interpretation was made see Sorokin et al. 2007; Fries et al. 2008; Caro-Quintero et al. 2012;Aoki et al. 2014; Chan et al. 2014; Yang et al. 2014; Lu et al. 2017; Maintinguer et al. 2017; Probandt et al. 2017; Shivani et al. 2017 and Fan et al. 2018). Thus, it appears that bacterial community composition shifted completely in enriched sediments to favour anaerobic metabolism which may represent eutrophic conditions.

3.5.3. Sediment enrichment reduces archaeal α-diversity and changes community composition In scenarios of eutrophication, archaeal communities have been reported to reduce many α-diversity metrics (Deng et al. 2015) however few studies have investigated the community compositional shifts in these conditions. In most cases, a change in dominance of archaeal groups has been seen to occur in organically enriched sediments where the availability of organic products affects the growth kinetics of some groups instead of others (Mosier and Francis, 2008; You et al. 2009; Park et al. 2010).Sediment archaeal communities evaluated for natural and enriched conditions in this study followed a similar pattern as the one observed for bacterial communities. Here, a reduction in archaeal α-diversity indices in enriched sediments was also present however, the proportion of change was less compared to bacteria α-diversity indices. Archaeal communities are ubiquitous in most marine environments and have an important role in nutrient dynamics (Park et al. 2010). The loss of archaeal α-diversity seen in this study may reflect the vulnerability of some archaeal groups to anaerobic conditions that rely on aerobic pathways (Prosser and Nicol, 2008; Mosier and Francis, 2008).

Crenarchaeota is the most common archaea phyla in most marine and freshwater systems (Prosser and Nicol, 2008) and it was the most dominant phyla in natural and enriched sediments in the present study. Within this phylum, Bathyarchaeia was the most abundant identified class which has been mostly associated with aerobic processes such as ammonification (Prosser and Nicol, 2008) and Fe-ammox (ammonia oxidation with Fe as an acceptor; Zhou et al. 2018). In this study, Bathyarchaeia increased in enriched sediments that had

INTRA-SPECIFIC BODY SIZE DIFFERENCES: MICROBIAL COMMUNITIES low oxygen concentration. This class has also been characterized as methane anaerobic heterotroph (Zhou et al. 2018) so it’s possible this group can be metabolically diverse and could explain its persistent dominance even in perturbed sediments. A similar case of higher abundances in enriched sediments was observed with the third most abundant class Lokiarchaeia (phylum Asgardeota) that has been associated with anaerobic carbon cycling in reducing habitats (Ma et al. 2015). In contrast, the classes Thermoplasmata (phylum Euryarchaeota), Nanohaloarchaeia (phylum Nanoarchaeota) decreased their abundances in enriched sediments. A decrease in the abundance of this group can be associated with a change in sediment redox conditions as Nanohaloarchaeia are known to have aerobic metabolisms that participate in nutrient cycling through ammonia oxidation (Kerou and Schleper 2015). Thermoplasmata on the other hand, prevailed as the second most abundant class in both natural and enriched sediments, however it’s decrease in abundance in enriched sediments may be related to the loss of specific obligate aerobic subgroups and the survival of subgroups that can withstand high enrichment and anoxia (Compte-Port et al. 2017). Higher taxonomic resolution was however impossible in the present study to define which genera could be contributing to archaeal community dissimilarities in enriched sediments. Still, this work represents one of the first studies in which the effect of enrichment on sediment archaeal communities has been analysed in a mesocosm experiment.

3.6. Conclusion Eutrophication is recognized as a widespread threat to many coastal environments and ecosystem services they provide. Results here highlight the drastic changes in bacterial and archaeal communities in response to organic enrichment. Loss of richness, diversity, evenness and a shift in microbial community composition (i.e. primarily to enhance anaerobic microbial activity) in the present mesocosm study provide evidence of possible microbial changes in sediment communities that could have direct negative effects on meio and macrobenthic communities and overall biogeochemical processes that are

CHAPTER 3 coupled between them. This mesocosm approach has been successful in testing the effects of enrichment, however further research is needed to elucidate if other bioturbator groups have higher impact on microbial communities in natural sediment and in polluted sites where bioremediation may be a feasible option.

CHAPTER 4

4. LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT AND SEDIMENT CHARACTERISTICS IN MACROFAUNA 4 BURROWS 4.1. Abstract Coastal intertidal sand- and mudflats play a major role in organic matter breakdown and may act as sinks for organic input. In these systems, benthic macrofauna influence microbial. communities that are responsible for organic matter degradation and nutrient transformation. However, increasing inputs of human-derived organic matter could negatively impact these communities and affect the system’s capacity to respond to other stressors. Understanding the impacts of human-derived organic inputs and their interaction with local sediment organic conditions on intertidal sand-and mudflats is crucial if irreversible effects on ecosystem functions are to be avoided. In this study, the effects of in situ nitrogen enrichment on bacterial communities from 10 intertidal sites were evaluated in surface sediments and macrofauna burrows. Seven environmental covariates were included to determine context-specific effects of sites on the nitrogen addition and interactive effects on bacterial communities. These covariates included organic matter characteristics (organic content, concentrations of algal pigments: chlorophyll a and phaeo pigments), sediment physical properties (median grain size, mud content and porosity) and abundance of local macrofauna (species counts). Small differences in organic matter and sediment physical characteristics (mud content and median grain size) were linked to differences in bacterial communities but temporary experimental nitrogen enrichment had no clear effect on bacterial diversity, abundance, or composition within macrofauna burrows or in the sediment surface. However, clear differences in bacterial communities were found between burrows and surface sediments, with the latter showing a unique bacterial composition and a possible microniche shaped by small differences of organic matter content, sediment physical properties and a potential low oxygen environment. No interaction was found between the temporary enrichment and the environmental covariates that could further influence bacterial communities in both surface sediments and burrows. These results suggest that bioturbating macrofauna may ameliorate the effect of organic enrichment in intertidal systems where background sediment organic matter levels can vary and may have an important role in shaping microbial communities and the ecosystem function they regulate.

99 CHAPTER 4

4.2. Introduction Human-derived inputs of nitrogen and nutrients to aquatic systems have dramatically increased in the last few decades because of global industrialization, synthetic nitrogen production and land use change. These have negatively impacted marine and estuarine systems (Huang et al. 2018). High levels of nutrients in coastal aquatic systems can lead to increased primary productivity that, in excess, may cause eutrophication and anoxia in the water column and sediments (Galloway et al. 2003; Kunihiro et al. 2011; Shen et al. 2016: Gilbert et al. 2017). High sediment enrichment can have a profound effect on the composition and function of both microbial communities and bioturbating macrofauna and their capacity to regulate nutrient cycling. For example, microbial abundances and composition of main groups associated with nitrogen cycling may be reduced in enriched sediments (Deng et al. 2015; Birrer et al. 2019; see Chapter 3) and even replaced by sulphate reducing bacteria (SRB) with the end point release of toxic compounds (e.g. SO43-, H2S and NH4+) to the water column (Skei et al. 2000; Mermillod-Blondin et al. 2004; Birrer et al. 2018; Filippini et al. 2019). For benthic macrofauna, enriched sediments can decrease survivorship (see Chapter 2; Penha-Lopes et al. 2009), change their community composition (Dafforn et al 2013) and alter their behaviours directly linked to bioturbation activities (e.g. sediment reworking and bioirrigation) and creation of microbial microniches (see Chapter 2; Biles et al. 2002; Bartolini et al. 2009; Bertics and Ziebis, 2010). Such enriched sediments could disrupt the essential ecosystem functions and services that coastal intertidal sand- or mudflat sediments provide (Qin et al. 2011; Howarth et al. 2011; Peñuelas et al. 2012).

The magnitude of the impact of ecological human-derived stressors can be spatially variable depending on both the abiotic and biotic context of an area (Dakos et al. 2019). This can result in changes in the ecological roles of local macrofauna, their interactions with other biota (e.g. microphytoplankton, meiofauna and microbial communities) and higher or lower functional resilience of an ecosystem to stressors (Wohglemuth et al. 2017). For example, variable input of organic matter between different areas can result in differing resource availability and allocation that directly affects microbial biomass, bioturbator feeding behaviours and the presence of mobile species (Levinton and Kelaher, 2004). Altogether, these changes in community dynamics may accumulate and

100 LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT interact to form heterogenous functional areas that respond differently to human- derived disruptors and produce different outcomes in terms of ecosystem functionality. This creates a spatially variable and dynamic system that is complex to analyze in terms of its response and its capacity to be extrapolated to a larger scale perspective (Gonzalez et al. 2020). In other words, sampling and inferring effects at a local level may not provide enough evidence to generate models that accurately assess the impacts of anthropogenic stressors such as eutrophication and nutrient pollution. Further studies are needed to assess the effect of human- derived pollutants within complex systems where environmental factors such as organic matter may determine high spatial variability and the response of biotic communities towards the stressor.

Detrital enrichment varies spatially in intertidal sand/mudflats and can have an important effect on macrofauna and microbial communities (Levinton and Kelaher, 2004). Sources for organic detritus are mainly from sediment phytoplankton but inputs from pelagic macroalgae and terrestrial sources are common (Mann, 1988). Such organic detritus input is however not always constant in supply, and it can vary between areas (delimited by physical characteristics or closeness to sources, see Filippini et al. 2019) and specific annual events (Levinton and McCartney, 1991). Detrital organic content in sediments represents an important resource for many bioturbators that can enable large accumulations near their burrows (Hughes et al. 2000; Geraldi et al. 2017) and is highly correlated to their abundance (Godbold and Solan, 2009), activity (Aller et al. 1994) and sediment composition (Arndt et al. 2013). Higher mud content and differing sediment grain size in sediments can also greatly increase the amount of organic matter accumulation, its porosity and change local physicochemical properties (Arndt et al. 2013). These conditions can stimulate increased bacterial biomass (Sander and Kalff, 1993) and determine local bacterial diversity (Filippini et al. 2019). In this study, the response of benthic bacterial communities to in situ nitrogen enrichment was explored at ten intertidal estuarine sites, each with a different environmental context including different organic matter input, varying phytoplankton biomass (i.e. measured by chlorophyll a and phaeo pigment concentration), sediment grain size with varying mud content and porosity, and different richness of local macrofauna. The effects of nitrogen enrichment on local bacterial communities were also assessed to determine differences in responses 101

CHAPTER 4 between two sediment positions: sediment surface (0-1 cm depth) and from biofilms within macrofaunal burrows. It is expected that bioturbation activities in burrows may enhance the creation of oxygenated specialized microniches for nutrient regulating bacteria that may provide high resilience to disruptors (e.g. denitrifying bacteria; Bertics and Ziebis, 2010).Under the proposed framework, I hypothesized that sediment nitrogen additions would decrease local bacterial richness, diversity, and evenness as seen in other studies where similar nutrient enrichments have been assessed (Chapter 3, and Filippini et al. 2019). Changes in bacterial community composition were also expected between control and exposed sediments suggestive of functional bacterial shifts toward anaerobic groups (Westrich and Berner 1984; Highton et al. 2016). However, I hypothesized that negative effects of nitrogen exposure to bacterial communities would be ameliorated in burrows compared to the sediment surface. Therefore, I expected that burrows would have higher richness, diversity and unique bacterial composition that could lessen the impact of the nitrogen added compared to surface sediments. In addition, I predicted that the effects of the above would interact with background environmental sediment characteristics such as organic matter (content and source), sediment physical properties (mud content, grain size and porosity) and the presence of higher abundances of macrofauna in shifting bacterial communities and regulating resilience under nitrogen exposure within burrows and in surface sediments.

4.3. Methods

4.3.1. Site selection, treatment preparation and sample collection Ten sites around the Northland, Auckland, Waikato and the Bay of Plenty regions in New Zealand were selected to investigate the effects of nitrogen pollution on sediment bacterial communities and additional interactions with macrofauna burrows and environmental context (Fig. 4.1). This research was conducted as part of the national project “Tipping points in ecosystem structure, function and services” (Sustainable Seas National Science Challenge) led by the research team of Simon Thrush (University of Auckland) and the participation of the University of Waikato (led by Rebecca Gladstone-Gallagher). Each site was in the upper intertidal zone of sandy, permeable sediments known to support abundant macrofaunal communities. This included bioturbators such as the deposit feeding

102 LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT bivalve Macomona liliensis and the polychaete Macroclymenella stewartensis (Schenone et al., 2019). Sites differed in sediment organic matter content (1-6%) and in phytoplankton biomass (chlorophyll a concentration between 1.2-23.1 µg/g of sediment and phaeo-pigment concentrations between 1.1.-19.9 µg/g of sediment). Sediments were predominantly sandy (gravel = 0-1%, sands = 76-99% and mud = 1-24%) with median grain sizes between 108-320 µm and porosity ranging between 39-61%. Due to the variability of these environmental characteristics between sites they were included as potential predictors of bacterial communities in the statistical analyses (see below). A summary of all the characteristics of each site included in this study are shown in Appendix 4, Table D1. Six experimental plots (9 m2, 3-5 m distance between plots to avoid nitrogen cross contamination) were established parallel to incoming tide in three blocks (10-15m apart) at unvegetated areas within each site and allocated in triplicates with two nitrogen porewater addition treatments: procedural control (0g N/m2) and high nitrogen addition (600g N/m2). The nitrogen addition treatment setup was modified accordingly to methods described by Douglas et al. (2018). Briefly, to create the enriched conditions, a constant density of 20 holes per m2 (3 cm diameter x 15 cm depth) were made using a hand-held corer and a 1 m2 quadrat frame (4 x 5 array) to maintain a constant spacing between holes. High nitrogen treatments were enriched with 75 g of slow release urea-based fertiliser (Nutricote®, 41-0-0% N:P: K dw, release over 140-200 days), which were added to each hole. Sediment cores were immediately replaced on top of the fertiliser to avoid additional disturbance. No fertiliser was added to procedural control holes and these were considered representative of local sediment conditions. Plots and nitrogen additions were established between March-April 2017 and context- specific variables were measured in each plot to establish baseline levels at each site before samples were taken for the present study. These site context variables included sediment grain size (hydrogen peroxide digestion 10% and measured in a Malvern Mastersizer 300; based on Singer et al. 1988), organic matter content (oven-dried at 60°C to remove moisture and then measured by weight lost on ignition at 550°C for 4h; Parker et al. 1983), Chlorophyll a and phaeo pigments (extracted in 90% buffered acetone and measured through fluorometry; Arar and Collins, 1997) and macrofauna abundance (sieved to 500µm, preserved in 70% isopropyl alcohol and counted). Porewater ammonia samples were collected and 103

CHAPTER 4 measured approximately 6 months after nitrogen enrichment in both procedural controls and high nitrogen treatments (extracted by centrifugation and measured in Lachat Quick-Chem 8000 automated flow injection analyser) to test if nitrogen enrichment increased and remained constant before microbial sampling. Porewater ammonia assays were collected and conducted by the research team of the University of Auckland. Obtained values indicated that nitrogen additions increased porewater ammonia concentrations in highly enriched treatments and remained lower in procedural controls (p<0.001, Appendix 4, Fig.D1). All environmental data collected in the ten sites for each plot are summarized in Appendix 4, Table D1 and a layout of experimental plots is described in Appendix 4, Fig. D2 (modified from Gladstone-Gallagher personal notes).

Microbial communities from each experimental plot were sampled in the surface sediment layer (~5 -10 mm) and from the biofilm of macrofauna burrows approximately 6 months following nitrogen enrichment (October-November 2017). Sampling of the sediment surface and macrofauna burrows was predicted to reflect a bacterial community response to the nitrogen additions as high porewater transport of the fertilizer in experimental plots was expected. This is in contrast to organic matter enrichment studies where less nutrient movement is expected as the nutrients are not immediately available and may also adhere to sediment particles. Four areas were randomly sampled in the surface sediment layer of each of the 6 experimental plots with a microspatula, taking ~0.5 mL of sediment and pooling them into a sterile 2 mL collection tube. Separately, microbial biofilms were collected directly from animal burrows in the plots using sterile cotton swabs and selecting for burrow openings with a diameter bigger than 0.5 cm. For each plot, three sterile swabs were used to wipe biofilm from the edge of three different burrows and then pooled into a sterile 2 mL collection tube. Each sterile swab was wiped around burrow walls for ~10 seconds to obtain as much biofilm as possible. After, collection samples were immediately stored on ice and transferred to freezers at -20 °C until analysis.

4.3.2. DNA extraction, amplification and sequencing Following the manufacturer’s protocols, genomic DNA extractions for surface sediments and for burrow biofilms were done using the DNeasy Powersoil kit ® (Qiagen) and DNeasy PowerBiofilm® (Qiagen) respectively. DNA concentration and purity were measured using a NanoDrop 2000® (Thermo Fisher Scientific). 104 LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT

Primers targeting the bacterial 16S rRNA gene region V1-3 were used (27F-519R, Kim et al. 2011) in PCRs and sequencing was done on a MiSeq v3 Sequencing Platform (Illumina) following the manufacturer’s guidelines at the Ramaciotti Centre for Genomics (UNSW, Australia). All raw sequences for this study were submitted in the NCBI Sequence Read Archive (SRA) database (BioProject accession ID: PRJNA596291).

4.3.3. Bioinformatic sequence analysis Raw fastq sequences were quality trimmed using function filterAndTrim from the dada2 package in R (Callahan et al. 2016) and adjusting the process to maximum truncation lengths of 290 for the forward reads and 270 for the reverse reads, with no filtering through maximum expected errors (maxee=0) as suggested by Prodan et al. 2020. Maximum truncation lengths were determined through the quality error profiles and a decision was taken to trim bases where the median Q score= 20 for both forward and reverse reads. Learning error rates and sample inference was done using the functions learnErrors and dada functions (dada2 package) respectively. Merging of forward and reverse unique sequences was done with the function mergePairs, (dada2 package) with a 69% ± 0.7 SE of successful sequences merged. Merged unique sequences were used to construct an amplicon sequence variant (ASV) table and a subsequent chimera removal was done through the identification and removal of bimeras (i.e. a two parent sequence chimera) across all samples using the function removeBimeraDenovo from the dada2 package. When accounting for the abundances of the identified bimeric sequences, a total of 3% of merged sequence reads were lost during this process. Taxonomic assignment was then done using the filtered ASV table through a taxonomic Naive Bayes classifier trained exclusively for DADA2 through the SILVA v.138 database. Filtering of chloroplast and mitochondria sequences in the ASV table was done and a final number of 53135 ASV was obtained. All steps of this pipeline were done using R v.3.6.

105

CHAPTER 4

Fig. 4.1. A) Map of New Zealand from which a close up (B) of the North Island and selected sites are marked with red triangles. C) Expanded map of 3 sites in the Northland region (Whangarei: Onerahai (WGR-O), Parua bay (WGR-P) and Takahiwai (WGR-T)) and 4 sites in the Auckland region (Whangateau (WTA) and Mahurangi: Lagoon bay (MAH-L) and Mandaley Bay (MAH-M). D) Expanded map of 3 sites in Waikato region (Raglan (RAG) and Whitianga: lower section (WHI-L) and upper section (WHI-U)) and 1 site in the Bay of Plenty region (Tauranga: Tuapiro (TAU-T)). Specific GPS location can be seen in Appendix 4, Table D1.

106 LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT

4.3.4. Statistical analysis The bacterial ASV abundance table was rarefied using the function rarefy (package vegan, Oksanen et al. 2019) to eliminate rare ASV present in only one sample. To evaluate if total ASV abundance was sampled effectively, rarefaction curves were constructed for each sample (R, vegan; Oksanen et al. 2019) and Good’s coverage was calculated to estimate the sampling effectivity to obtain all OTUs present in each sample (R, QsRutils; Quensen 2019). Rarefaction curves and Good’s coverage confirmed that bacterial species richness in surface sediments and burrows was assessed effectively with the sampling done in all selected sites (Good’s coverage: Appendix 4, Table D1 and rarefaction curves: Appendix 4, Fig. D3).

For each sample, α-diversity indices were calculated including bacterial richness (no. ASV), diversity (Shannon-Weiner diversity Index) and evenness (Pielou index) using the vegan package (Oksanen et al. 2019). To determine the effects of nitrogen addition (i.e. control and high) and sediment position (i.e. surface sediments and macrofauna burrows) on bacterial communities, a general linear mixed model (GLMM) using the function lmer (package lme4, Bates et al. 2015) was constructed for each α-diversity index and considering interactions between each of the above fixed factors. As sampling was done in three separated blocks (each with 1 control and 1 enriched plot, Appendix 4, Fig. D2) within each site, block was considered as a nested random factor to control for possible intra- site variability given by block location. Seven environmental covariates that described the inter-site variability were considered in the GLMM including organic matter content (%), mud content (%), median grain size (µm), chlorophyll a concentrations (µg/g), phaeopigment concentrations (µg/g), sediment porosity (%), and abundance of macrofauna (species counts). To determine the best model fit, covariates for each alpha diversity GLMM were selected by a backward stepwise model selection (function stepAIC, package MASS; Venables and Ripley, 2002) using the Aikaike Information Criteria (AIC). Variance inflation factors (VIF) to detect high multicollinearity between the selected covariates and fixed factors were calculated. Covariates with strong collinearity (VIF > 10, Quinn and Keough, 2003) were removed. Fixed factor (i.e. nitrogen treatments and sediment position) inferences were evaluated through F tests when the random factor was found non- significant and Wald chi-square tests when block was included as a random factor

107 CHAPTER 4

in GLMM. Both inferences were evaluated with the Anova function from the package car in R (Fox and Weisberg, 2018). If significant effects of the covariates on the alpha diversity indices were found, reduced GLMMs were constructed in which single significant covariates were evaluated with the fixed factors and their interactions. Within these new models, when a covariate was found significant, estimated marginal means of linear trends were calculated using the function emtrends (package emmeans, Russell, 2019) to determine differences in slope between factor levels. Individual Pearson correlation coefficients were also included for each generated slope and an associated p-value was calculated. All evaluated models fulfilled the GLMM requirements for normality and heteroscedasticity and were evaluated for influential outliers (Cook’s distance) (Appendix 4, Fig. D4).

A bacterial ASV composition analysis was done through a comparison of the relative abundance of the main bacterial phyla present in the samples. This included only phyla with high average relative abundance between sites (relative abundance > 1%) such as Proteobacteria, Bacteroidota, Desulfobacterota, Cyanobacteria, Actinobacteriota, Verrucomicrobiota, Acidobacteriota, Planctomycetota, Chloroflexi, Myxococcota and Firmicutes. Rare phyla (<1% relative abundance in all samples) were not included in the analysis. Such phyla include the following: Campilobacterota, Gemmatimonadota, Patescibacteria, Calditrichota, Latescibacterota, Bdellovibrionota, NB1-j, Spirochaetota, Nitrospirota, Sva0485, SAR324_clade(Marine_group_B), Nitrospinota, Fusobacteriota, Fibrobacterota, Marinimicrobia_(SAR406_clade), Cloacimonadota, Deferrisomatota, Dependentiae, Zixibacteria, Schekmanbacteria, Thermotogota, Deinococcota, Modulibacteria, Hydrogenedentes, WS2, Fermentibacterota, MBNT15, PAUC34f, LCP-89, Sumerlaeota, Elusimicrobiota, FCPU426, AncK6, Margulisbacteria, Dadabacteria, TA06, Armatimonadota, WPS-2, Acetothermia, WOR-1, Deferribacterota, NKB15, WS1, RCP2-54, Methylomirabilota, Caldisericota, CK- 2C2-2, Caldatribacteriota, Entotheonellaeota, 10bav-F6, GN01, DTB120, FW113 and Abditibacteriota. Similar GLMMs as described above were fitted for bacterial phyla to evaluate differences in phyla relative abundance (%) between nitrogen treatments and sediment position. Fixed factor (i.e. phylum identity, nitrogen

108 LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT

treatments and sediment position) inferences were evaluated through Wald chi- square tests as above GLMMs (Fox and Weisberg, 2018). If significant interactions among the different phyla, nitrogen treatment and sediment position were found, contrasts with estimated marginal means were calculated (package emmeans, Russell, 2019) and p values adjusted with Tukey HSD method. As with the community analysis, environmental covariates were evaluated in new models to determine the specific influence of these variables on the relative abundance of the selected phyla. The relative abundance of all bacterial phyla was logit transformed to fulfill requirements of normality and heteroscedasticity required for GLMMs (Appendix 4, Fig D5) and influential outliers were identified and removed.

To investigate differences among bacterial communities from different nitrogen addition treatments and sediment position, I used a permutational multivariate analysis of variances (PERMANOVA, function adonis, R package vegan; Oksanen et al. 2019) using a Bray-Curtis dissimilarity matrix (function vegdist, R package vegan; Oksanen et al. 2019). PERMANOVA is a robust statistical method not restricted to linearity or sensitive to changes in data distribution however it assumes equal variance between factor levels (Anderson et al. 2017). To fulfill this assumption, an analysis of multivariate homogeneity of group dispersions was done to evaluate variance homogeneity for the adonis multivariate model using the function betadisper (R package vegan; Oksanen et al. 2019). A non-parametric multidimensional scaling (NMDA) ordination was done to visualize the differences in bacterial composition (function metaMDS, R package vegan; Oksanen et al. 2019). Environmental covariates were fit to the NMDS ordination using the function envfit (R package vegan; Oksanen et al. 2019) and squared correlation coefficients with corresponding p-values of each covariate were calculated. A similarity percentage analysis (SIMPER) was conducted using the calculated Bray-Curtis dissimilarity matrix to discriminate the specific ASVs that contributed the most to the differences seen between nitrogen enrichment treatments and sediment position (function simper, R package vegan; Oksanen et al. 2019). SIMPER analysis was conducted at the genus level to maximize the resolution for describing differences between communities. Only specific ASVs with a contribution higher than 1% of total dissimilarity were included and analyzed to identify relevant bacterial genera. For ASVs not classified to genus 109

CHAPTER 4

level, class taxonomic level was assigned and included in the analysis as unclassified genera. A similar GLMM as in the phyla analysis above was done with the relative abundance (%) of the selected genera as the response variable and genera identity, nitrogen enrichment treatments and sediment position as fixed factors. As above, interactions between fixed factors were included in the models and similar reduced models were constructed to determine the sensibility of the selected genera to the different covariates. All evaluated models fulfilled the GLMM requirements for normality and heteroscedasticity and were evaluated for influential outliers (i.e. Cook’s distance) (Appendix 4, Fig.D6).

4.4. Results

4.4.1. Changes in alpha diversity are driven by sediment position Bacterial α-diversity (richness, diversity, and evenness) sampled from the surface sediments differed significantly from samples obtained from macrofauna burrows. (Fig. 4.2, Table 4.1). Specifically, bacterial ASV richness was higher in burrows (26% higher) than surface sediments, however for bacterial diversity and evenness, this effect was lower (an increase of 6% and 2% in burrows respectively). Nitrogen enrichment, however, did not affect bacterial α-diversity indices or interact with sediment position, when also controlling for environmental covariates (Fig. 4.2, Table 4.1). New models to evaluate the effect of environmental covariates on the alpha diversity indices did not show any interactions with the nitrogen treatments (Appendix Table D2). For all alpha diversity indices, organic matter content was an important factor explaining differences between surface and burrows (Table 4.1, p<0.001). New GLMMs results between this covariate and the alpha diversity indices show a weakly positive response with all indices increasing as organic matter content increased (Appendix 4, Table D2-D3, and Fig.D7). This relationship was only present in surface sediments, with no changes produced by organic matter content in macrofauna burrows (Appendix 4, Table D3 and Fig.D7). In addition to organic matter, mud content was also a significant predictor of bacterial richness and diversity in different sediment positions (Table 4.1, p>0.001). However, this effect was only observed for bacterial richness within burrows, as higher mud content in sediments was found to reduce bacterial richness (Appendix 4, Table D2-D3 and Fig.D7).

110 LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT

Table 4.1. Summary of general linear mixed model (GLMMs) inferences testing differences in α-diversity indices a) Bacterial Richness (number of ASV), B) Shannon Diversity Index and C) Pielou Evenness, between nitrogen treatments (Control and High), sediment position (Burrows and Surface) and interactions. The influence of the selected covariates in controlling for the alpha diversity indices Is also reported. Significant differences are marked in bold (p<0.05). A) Bacterial Richness (No. ASV) Contrasts Df F value p value Nitrogen treatment 1 33.44 0.67 Sediment position 1 38.74 <0.0001 Nitrogen*Position 1 0.19 0.41 Organic matter content (%) 1 59.24 <0.0001 Mud content (%) 1 0.69 <0.0001 B) Shannon Diversity Index Contrasts Df Chisq p value Nitrogen treatment 1 0.34 0.56 Sediment position 1 41.92 <0.0001 Nitrogen*Position 1 0.95 0.33 Organic matter content (%) 1 26.81 <0.0001 Mud content (%) 1 25.83 <0.0001 Median grain size (µm) 1 1.71 0.19 C) Pielou Evenness Contrasts Df Chisq p value Nitrogen treatment 1 0.05 0.82 Sediment position 1 18.58 <0.0001 Nitrogen*Position 1 0.35 0.55 Organic matter content (%) 1 5.59 0.02 Mud content (%) 1 2.71 0.10 Median grain size (µm) 1 2.18 0.14 Macrofauna abundance (counts) 1 1.55 0.21

111

CHAPTER 4

A B

C D

E F

Fig.4.2. Differences in alpha diversity indices of bacterial communities sampled from two nitrogen treatments (Control and High; richness: A, Shannon diversity: C and Pielou Evenness: E) and in two different sediment positions (Surface and burrows: richness: B, Shannon diversity: D and Pielou Evenness: F). Data is shown as mean ± SE. Asterisks (“*”) indicate significant (p < 0.05) lower alpha diversity values between factor levels.

112 LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT

4.4.2. Community composition dissimilarity Bacterial community composition differed significantly between nitrogen treatments and sediment position (PERMANOVA, Table 4.2., p<0.001) when the seven selected covariates (organic matter content, mud content, median grain size, chlorophyll a concentration, phaeo-pigment concentration, porosity and macrofauna abundance) were included in the model. NMDS ordination plots for both fixed-factors (Fig. 4.3, stress= 0.15; non-metric fit, R2=0.98) suggest a small separation in communities between control and high nitrogen treatments as well as for burrows and surface sediments, however with high overlap in both cases. For both grouping factors, within-group dissimilarity variance was constant (Table 4.2, betadisper, p>0.05) which supports the small differences inferred from the PERMANOVA (Anderson et al. 2017). Organic matter content, median grain size and phaeo-pigment concentration were important covariates explaining the differences in community structure between the grouping factors (PERMANOVA, Table 4.2, p<0.001). NMDS biplots and covariate fitting (i.e. envfit), also supported the influence of these covariates on bacterial community composition (Table 4.2, organic matter content: R2= 0.45, median grain size: R2= 0.46 and phaeo-pigments: R2=0.31).

4.4.3. Phylum composition analysis and specific phyla sensitivity to environmental covariates Two phyla with low relative abundances varied in response to the nitrogen treatments, Myxococcota with a decrease of 26% in sediments with high nitrogen concentrations and Firmicutes with an increase of 87% in the same conditions (Fig. 4.4 and Table 4.3; Appendix 4, Table D5 for a complete list of selected phyla’s relative abundances between nitrogen treatments). Changes in phyla relative abundances were more noticeable between burrows and surface sediments, with 64% of the selected phyla showing a significant change (Fig 4.3 and Table 4.3, p < 0.05; Appendix 4, Table D5 for a complete list of selected phyla’s relative abundances between sediment positions). Five phyla showed an increase in relative abundance in burrows compared with surface sediments, including Desulfobacterota (52% increase, p<0.001), Verrucomicrobia (22% increase, p=0.04), Acidobacteria (9% increase, p<0.001 ), Chloroflexi (27% increase, p=0.01) and Firmicutes (71% increase, p<0.001). In contrast, two phyla showed a

113 CHAPTER 4 decrease in relative abundances in burrows compared to surface sediments: Cyanobacteria (64% decrease, p<0.001) and Actinobacteria (68% decrease, p<0.001). The dominant groups, Proteobacteria and Bacteroidetes had similar relative abundances in all evaluated sites irrespective of the treatments (Proteobacteria, p=0.21 and Bacteroidetes, p=0.86). A complete summary of all multiple comparisons of the relative abundance of the selected phyla between sediment positions is shown in Appendix 4, Table D6.

Table 4.2. Summary of the multivariate analysis done to determine differences in bacterial community composition between nitrogen treatments (control and high) and sediment position (Burrows and Surfaces). A) PERMANOVA results showing compositional differences between grouping fixed factors and controlling covariates with corresponding interactions between fixed factors. R squared values for covariates calculated through the envfit function are included. B) Analysis of multivariate homogeneity of group dispersion to determine intergroup variances in both evaluated fixed factors and to validate results obtained from PERMANOVA. Bold letters indicate significant differences in community composition and significant predictors in the multivariate model. A) Permutational multivariate analysis of variance (permutations=1000) Contrasts df Mean Sqs F model R2 p value Nitrogen treatment 1 0.39 1.47 0.01 0.01 Sediment position 1 1.81 6.85 0.04 <0.001 Nitrogen*Position 1 0.15 0.57 0.00 0.91 Organic matter content (%, R2=0.45) 1 3.64 13.75 0.09 <0.001 Median grain size (µm, R2=0.46) 1 1.97 7.45 0.05 <0.001 Mud content (%, R2=0.31) 1 1.65 6.24 0.04 0.10 Macrofauna abundance (counts, R2=0.17) 1 1.55 5.87 0.04 0.47 Porosity (%, R2=0.19) 1 0.7 2.63 0.02 0.41 Chlorophyll a concentration (µg/ g, R2=0.23) 1 0.74 2.78 0.02 0.11 Phaeo-pigment concentration (µg/ g, R2=0.31) 1 1.01 3.8 0.02 0.01 B) Analysis of multivariate homogeneity of group dispersions (betadisper) Contrasts df Mean Sq F value Pr(>F) Nitrogen treatments (Control vs High) 1 0.01 3.58 0.06 Sediment position (Burrows vs Surface) 1 0.01 2.23 0.14

114 LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT

Fig. 4.3. Bacterial composition multivariate analysis (Bray-Curtis dissimilarity index) between A) nitrogen treatments (control and high) and B) sediment position (burrows and surface). Non-parametric multidimensional scaling ordination (stress= 0.15, k=2) where polygons and points represent grouping levels of fixed factors. Vectors represent fitted environmental covariates scaled by their correlation with the community composition from the grouping factor centroid. Here stronger predictors have longer arrows.

115

Chapter 4

Fig. 4.4. Relative abundance (%) of selected bacterial phyla (relative abundance >1%) between A) nitrogen treatments (Control and high) and B) sediment position (Burrows and surface). Relative abundance data is shown as mean ± SE. Significant differences between fixed factor levels are denoted with an asterisk “*” and detailed in Table 4.3 and and multiple pairwise comparisons are given in Appendix 4, Table D6.

116

LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT

Table 4.3. Summary of the GLM to determine differences in relative abundance of selected bacterial phyla (>1% relative abundance) between each phylum (11 phyla), nitrogen treatments (Control and High), sediment position (Burrow and Surface) and interaction therein. Significant results are shown in bold. Contrasts Df F value p value Phylum identity 10 518.00 <0.001 Nitrogen treatment 1 1.42 0.23 Sediment position 1 1.63 0.20 Phylum*Nitrogen 10 18.85 <0.001 Phylum*position 10 35.49 <0.001 Nitrogen*Position 1 1.43 0.23 Phylum*Nitrogen*position 10 0.84 0.59 Organic matter content (%) 1 10.98 <0.001 Mud content (%) 1 2.50 0.11 Median grain size (µm) 1 13.81 <0.001 Chlorophyll a concentration (µg/ g) 1 5.05 0.02 Phaeo-pigment concentration (µg/ g) 1 3.03 0.08 Porosity (%) 1 0.14 0.71 Macrofauna abundance (counts) 1 1.26 0.26

For all the analysis above, inter-site environmental covariates were included in the models. However only organic matter content, median grain size and chlorophyll a concentration were found to be important predictors (Table 4.3, p<0.001). Through new models and planned correlation comparisons, sensitivity of selected phyla was analyzed against the three covariates and compared between fixed factor levels (Appendix 4, Table D7). The selected covariates had no relationship with changes in phyla relative abundance for both nitrogen treatments. However, these three covariates did influence the relative abundance of some bacterial phyla in burrows and surface sediments. Relative abundances of Actinobacteria, and Chloroflexi increased in sediments with higher organic matter content irrespective of the sediment position, while the phylum Desulfobacterota increased in these organic conditions but only in burrows (Appendix 4, Table D8 and Fig D8). The phylum Acidobacteria had a similar pattern of increase in relative abundance among sediment positions but this was higher in burrows compared to surface sediments (Appendix 4, Table D8 and Fig D8). By contrast, the phylum Cyanobacteria had lower relative abundances at higher organic matter content and this pattern was stronger in surface sediments (Appendix 4, Table D8 and Fig D8). Bacterial phyla had a high sensitivity to changes in the sediment grain size, with

117

CHAPTER 4 shifts in relative abundances noticeable in 73% of the phyla. Proteobacteria and Verrucomicrobia increase their abundances at sites with larger grain sizes, but this effect was only present in burrows (Appendix 4, Table D9 and Fig D9). Relative abundances of Cyanobacteria also increased in the same conditions, but the reduction occurred at the same rate in both sediment positions (Appendix 4, Table D9 and Fig D9). In contrast, the phyla Chloroflexi and Myxococcota reduce their relative abundances in sites with larger sediment grain sizes in both burrows and surface sediments, while Actinobacteria only shows a reduction in surface sediments (Appendix 4, Table D9 and Fig D9). The phyla Desulfobacterota and Acidobacteria have a decrease of abundance in the same grain size conditions but higher rates of reduction are shown in burrows and surface sediments respectively (Appendix 4, Table D9 and Fig D9). Finally, the increase in concentration of chlorophyll a in sediments was linked to reductions in relative abundances of Cyanobacteria and Verrucomicrobia, and an increase in abundance of Chloroflexi (Appendix 4, Table D10 and D11; and Fig D10) but this was unrelated to the sediment position or nitrogen treatment.

4.4.4. Community structure dissimilarities explained by selected genera Similarity percentage analyses indicated that community composition differed among nitrogen treatments and sediment position and was explained by 17 bacterial genera (Appendix 4, Table D12). From these, only 9 genera were identified to contribute more than 1% of community dissimilarity in both levels of the fixed factors (control and high; burrows and surface) (Appendix 4, Table D12, Woeseia, Halioglobulus, Robiginatalea, Eudoraea, Candidatus Thobios, Roseobacter, Sva0081 sediment group, Ilumatobacter and Pleurocaspa PCC- 7319). In total, these identified genera explained 16% of community dissimilarity between nitrogen treatments and 17% between sediment position. Eight unidentified bacterial genera with >1% of contribution to community dissimilarity were also included in the analysis as unclassified genera and assigned to a class level (Appendix 4, Table D12). These included unclassified genera from the classes Gammaproteobacteria, Bacteroidia, Desulfobulbia, Alphaproteobacteria, Cyanobacteriia, Acidomicrobiia, Anaerolineae and Polyangia. In total, these unidentified genera contributed with 28% of community dissimilarities in both nitrogen treatments and sediment positions (Appendix 4, Table D12). Only 9.5 and

118 LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT

9.7% of contribution to community dissimilarity was not able to be assigned to a Class in nitrogen treatments and sediment position, respectively.

To further investigate the bacterial composition differences, identified and unidentified genera that contributed with more than 1% to community dissimilarity were included in a GLMM to determine differences in their relative abundance between nitrogen treatments and sediment positions. Model results show no differences in the relative abundances of selected genera between control and high nitrogen treatments (Table 4.4 and Fig. 4.5A, p=0.12) or interactions with sediment position (Table 4.4, p=0.31) or between genera (Table 4.4, p=0.91). Differences in bacterial relative abundance were found in many of the selected genera between burrows and surface sediments (p<0.001). Relative abundances of 9 genera were reduced in burrows compared to surface sediments including Woeseia (35% decrease), unclassified Alphaproteobacteria genera (32% decrease), unclassified Acidomiicrobia (68% decrease), Robiginitalea (47% decrease), Eudoraea (43% decrease), unclassified Cyanobacteriia genera (81% decrease), Roseobacter (31% decrease), Ilumatobacter (76% decrease) and Pleurocaspa PCC- 7319 (69% decrease). In contrast, only 3 genera had an increase in relative abundance in burrows compared to surface sediments: unclassified Desulfobulbia genera (63% increase), Halioglobulus (38% increase) and unclassified Ananerolineae genera (26% increase). Overall, unclassified genera of the classes Gammaproteobacteria and Bacteroidia had the highest relative abundances but did not differ between burrows and surface sediments. A complete summary of the multiple comparisons of relative abundance in burrows and surface sediments between genera is shown in Appendix 4 Table D14.

Within the models described above to determine differences in relative abundances between genera, environmental variables were also added as predictors. Here, only organic matter content (%) and mud content (%) were found to be important predictors of bacterial relative abundance (Table 4.4, organic matter: p<0.001 and mud content: p>0.001). As in the phylum composition analysis, new GLMMs were constructed to determine the interactions of each of these covariates with the evaluated fixed factors (i.e. genus identity, nitrogen treatments and sediment position). New models showed an influence of organic matter and mud content on the selected genera (Appendix 4, Table D15, organic

119

CHAPTER 4 matter: p<0.001 and mud content: p>0.001), however this influence was not different between nitrogen treatments (Appendix 4, Table D15, organic matter: p=0.34 and mud content: p=0.72), sediment positions (Appendix 4, Table D15, organic matter: p=0.63 and mud content: p=0.64) or in interaction between fixed factors (Appendix 4, Table D15, p>0.1). Planned comparative correlations, showed unclassified genera of Cyanobacteriia, Pleurocaspa PCC-7319 and Roseobacter had higher rates of decrease in relative abundance at higher levels of organic matter and mud content compared to other genera (Appendix 4, Table D16-D17 and Fig. D11-D12). Woeseia, Eudoraea and unclassified Desulfobulbia, Ananerolineae and Polyangia genera; showed slight increases in relative abundance at increasing levels of organic matter and mud content and only unclassified Alphaproteobacteria genera showed a slight decrease in relative abundance influenced by increasing levels of these covariates (Appendix 4, Table D16-D17 and Fig. D11-D12). Unclassified Acidomiicrobia genera exclusively showed a slight increase in relative abundance at higher organic matter content while only the unclassified Bacteroidia genera had slight increases at higher mud content.

Table 4.4. Summary of the GLM to determine differences in relative abundance of selected identified and unclassified genera (>1% contribution to community structure dissimilarity) between each of these genera (17 genera), nitrogen treatments (Control and High), sediment position (Burrow and Surface) and interaction therein. Significant results are shown in bold Contrast df F value p value Organic matter content (%) 1 24.6 <0.001 Mud content (%) 1 12.7 <0.001 Genus identity 16 84.7 <0.001 Nitrogen treatments 1 2.4 0.122 Sediment position 1 30.2 <0.001 Genus identity*Nitrogen treatments 16 0.6 0.913 Genus identity*Sediment position 16 16.4 <0.001 Nitrogen treatments*Sediment position 1 1.0 0.311 Genus identity*Nitrogen treatments*Sediment position 16 0.3 0.998

120 LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT

B A

Fig. 4.5. Relative abundance (%) of selected identified and unclassified genera (>1% contribution to community structure dissimilarity) between A) nitrogen treatments (Control and high) and B) sediment position (Burrows and surface). Relative abundance data is shown as mean ± SE. Significant differences between fixed factor levels are marked with an asterisk “*”, and detailed in Appendix 4 Table D13 and multiple pairwise comparisons are given in Appendix 4, Table D14.

121

CHAPTER 4

4.5. Discussion The present study investigated the effect of in situ sediment nitrogen enrichment on bacterial communities on surface sediments and within macrofauna burrows at sandy intertidal sites with different levels of organic matter content, algal biomass (i.e. chlorophyll a and phaeo pigments), varying sediment physical properties (median grain size, mud content and porosity) and abundance of macrofauna. I expected that nitrogen addition would lower bacterial alpha diversity indices and shift bacterial community structure to favour taxa linked to anaerobic processes as seen in other nutrient enrichment studies (Chapter 3 and Filippini et al. 2019). In addition, I hypothesized that this effect would be ameliorated in macrofauna burrows and sediment characteristics would play a key role as factors that determine functional resilience against the nitrogen addition. Results show no clear effect of the nitrogen exposure on bacterial communities at almost any level of analysis with only small changes in less dominant phyla. No difference in bacterial community response to the nitrogen addition exposure between surface sediments and macrofauna burrows was found. Moreover, there were no interactive effects of the environmental covariates combined with nitrogen addition suggesting the site conditions played a minor role in regulating the bacterial response towards the experimental enrichment. However, clear differences in bacterial community diversity and composition were found between both sediment positions independently of the nitrogen addition, with some relevant functional taxa shifting their abundance in burrows compared to surface sediments. Organic matter, mud content and sediment median grain size were shown to be relevant environmental predictors of bacterial communities in both sediment positions. Specifically, organic matter content was more influential on bacterial communities than the temporary experimental nitrogen exposure even in a short range of values (from 1-6% of organic content). This was particularly clear in the surface sediment communities compared to bacterial biofilms in macrofauna burrows.

124

LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT

4.5.1. Macrofauna may create low oxygen bacterial microniches in burrows The presence of benthic macrofauna can shift bacterial community composition and change the metabolic capacity of the sediments (Foshtomi et al. 2015). Macrofauna can increase the abundance of specific groups related to organic matter degradation such as ammonia oxidising bacteria (AOB) and other phyla associated with aerobic nitrogen metabolism (Altmann et al. 2004; Quintana et al. 2013). However, other studies have demonstrated the effect of macrofauna burrows on enhancing anaerobic metabolism with higher activity of sulphur reducing bacteria (SRB) as biomixing of deep anaerobic sediments appears within these biogenic structures (Mchenga et al. 2007; Bonaglia et al. 2013; Boeker et al. 2016). This study highlighted differences in bacterial communities between burrows and surface sediments, as burrows were shown to have 26% higher bacterial abundance compared to sediment surfaces. However, bacterial diversity and evenness only increased 6% and 2% in burrows, respectively. Bacterial microniches found in macrofauna burrows may not differ as much in diversity or evenness, however results show that there are clear compositional differences between sediment positions, with differences in the abundance of particular bacterial groups.

In this study, the dominant phyla, Proteobacteria and Bacteroidetes, did not differ in relative abundance between burrows and the surface sediments. Both these groups accounted for a large proportion of the assessed bacterial communities (Proteobacteria: 39% and 42%, and Bacteroidetes 23% and 22% in burrows and surface sediments respectively) and can be important functional groups that determine much of the organic matter degradation and metabolic capacity in aerobic and anaerobic sediments (Kerster et al. 2006 and Fernandez- Gomez et al. 2013). Nonetheless, some bacterial phyla showed large increases in abundance in burrows (i.e. Desulfobacterota, Verrucomicrobia, Chloroflexi, Firmicutes and Acidobacteria) and they accounted for as much as the dominant phyla (23% accumulated relative abundance of these groups) and possibly

125

CHAPTER 4 influence functional changes in the sediment when other less dominant phyla are also decreasing their abundance (i.e. Cyanobacteria: 64% decrease and Actinobacteria: 68% decrease). These changes in bacterial relative abundances are further supported by selected genera that contributed to community dissimilarities such as the increased abundance of unclassified genera of class Desulfobubia (Phylum Desulfobacterota) and Ananerolineae (Phylum Chloroflexi); and the decrease of the genera Pleurocaspa PCC-7319 (Phylum Cyanobacteria) and Ilumatobacter (Phylum Actinobacteria). The main functional role in burrows associated with the bacterial taxa that increased their abundance in this study was related to anaerobic metabolism, including processes like sulphur reduction (Desulfobacterota/unclassified genera of Desulfobulbia, Lens, 2000; Chloroflexi/unclassified genera of Ananerolineae, Sinkko et al. 2013) and organic matter degradation at low oxygen eutrophic conditions (Firmicutes, Huang et al. 2017). The only exception is the strictly aerobic Gammaproteobacteria, Halioglobulus who increased its abundance within burrows but can also dominate in some macrofauna gut microbiomes (Ding et al. 2019). In contrast, all other bacterial taxa that reduced in abundance are mostly related to nutrient cycling in aerobic conditions (Cyanobacteria/Pleurocaspa PCC-7319, Dussud et al. 2018; Actinobacteria/Ilumatobacter, Chen et al. 2016; Proteobacteria/ unclassified genera of Gammaproteobacteria and Woeseia, Probandt et al. 2017; Bacteroidetes/Eudoraea and Robignitalea, Probandt et al. 2017 and Hicks et al. 2018). Results here agree with burrow microniches that predominately have a reducing environment that may be a direct consequence of macrofauna sediment reworking from deep anoxic layers (Mchenga et al. 2007; Bonaglia et al. 2013; Boeker et al. 2016).

4.5.2. Organic matter as an important factor in burrow microniches Bioturbation by benthic macrofauna can increase bacterial abundances around burrows as these provide ideal properties for organic matter adherence and be used as an energy source by microbial fractions (Papaspyrou et al. 2006; Wada et

126

LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT al. 2016; Kang et al. 2017; Geraldi et al. 2017). These conditions permit a stable microbial community and the formation of functional metabolic microniches where high rates of organic matter breakdown occur (Bertics et al. 2010; Gilbertson et al. 2012; Foshtomi et al. 2015). I hypothesized that sampled macrofauna burrows in the present study would follow a similar trend and differences in bacterial α-diversity would remain constant between sites with differing organic matter content compared to surface sediments. Results partially support this hypothesis, as macrofauna burrows did not differ in bacterial richness, diversity and evenness compared to surface sediments. On the sediment surface, small increases in organic matter content (from 1-6%) produce slightly higher rates of bacterial richness, diversity, and evenness until reaching the same values at high organic matter content. Surface sediments may be reflecting changes in bacterial communities as a direct response to organic matter inputs while burrows may have a different organic environment with a stable accumulation or “hot spot” of organic matter (Volkenborn et al. 2007). These differences in organic matter make bioturbator burrows potential centers of bacterial organic matter degradation, and potentially beneficial for ameliorating nutrient pollution in sediments.

Sediment organic matter content is one of the most important factors affecting bacterial communities in coastal intertidal systems (Strauss and Lamberti, 2002; Wang et al. 2014). Studies demonstrate that bacterial biomass and organic matter have a strong positive correlation in marine and freshwater sediments (Yamamoto and Lopez, 1985; Sander and Kalff, 1993; Logue et al. 2004), however the increase in bacterial abundance and diversity recorded in previous studies included greater differences in organic matter content (i.e. ~ 4- 14% OMC). Compared to other studies, here, bacterial biomass increases occurred under a small range of organic matter content increase (1-6%) which indicates that sediment bacterial communities may be more sensitive to changes in organic matter by changing dominance and potentially recolonization rates than previously thought (Logue et al. 2004).

127

CHAPTER 4

Macrofauna burrows support fairly unique bacterial communities compared to the surface sediments (see section 4.5.2), which may support the hypothesis that burrows are acting as microbial microniches. However, this study also found that organic matter was the most influential environmental covariate that shaped bacterial composition in burrows and sediment surfaces (Fig. 4.3B, envfit, R2=0.45). When analyzing composition at a phylum level, organic matter increases were linked to a small increase in the relative abundance of some groups such as Actinobacteria, Acidobacteria and Chloroflexi irrespective of the position in the sediment. This pattern of small increases and in some cases decreases of abundance in response to organic matter, was repeated for some of the selected genera (Increase: Eudoraea, Woeseia and unclassified genera of Ananerolineae and Acidomiicrobia; decrease: Pleurocaspa PCC-7319, Roseobacter and unclassified genera from the classes Alphaproteobacteria) but here no differences were found between sediment positions. However, for the phylum Desulfobacterota, this increase in relative abundance as a response of organic matter increase happened exclusively within burrows. In contrast, Cyanobacteria were the only phyla whose abundance decreased as organic matter increased in surface sediments whilst remaining unchanged albeit low in burrows. It is possible that the sensitivity of these two phyla to small changes of organic matter may relate to the reducing environment proposed in section 4.5.2. These conditions may favour Desulfobacterota within burrows, limiting the growth of Cyanobacteria groups. In addition, as many of the analyzed taxa are linked to aerobic organic matter degradation (i.e. nitrification, sulphide oxidation) their population sensitivity to organic matter may also be affected by changes in organic matter quality (Strauss and Lamberti, 2002) and the redox environment. Together these may favour certain groups instead of others with similar functional roles. This is supported by observed changes in the phylum Cyanobacteria in this study, as abundances declined in response to a higher concentration of chlorophyll a in the sediment. It’s possible that higher accumulation of organic matter from different sources (i.e. plants or macroalgae)

128

LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT in the sediment and burrows contributed to an increase in abundance of some groups (for example Acidobacteria and Chloroflexi) but established unfavourable conditions for Cyanobacteria (e.g. increased turbidity to decrease cyanobacteria photosynthetic processes; Mishra et al. 2018). At even higher organic inputs and different type/quality, other groups will gain dominance (e.g. Bacteroidetes) as seen in Bai et al. 2012 and Chapter 3.

4.5.3. The role of sediment physical characteristics in shaping bacterial microniches in burrows Environmental covariates such as sediment physical characteristics (mud content, median grain size and porosity) and the abundance of local macrofauna can all tightly regulate resilience attributes of bacterial communities at a local and larger scales (Gladstone-Gallagher et al. 2019; Chen et al. 2020; Clark et al. 2020). In this study, only mud content (%) and median grain size had significant relationships with bacterial communities and their relationships varied between burrows and surface sediments. The ability of bacterial communities in sediments to form biofilms and obtain organic resources is determined by the sediment grain size (Zheng et al. 2014). Higher content of mud ( i.e. lower median grain size) increases organic matter surface area of adsorption, retention of water content and redox conditions (Wang et al. 2013). Through this study, it is clear that burrows had higher bacterial abundance however this pattern changes when mud content increases. Bacterial diversity and evenness are not affected by increased mud content, but it is clear that together with median grain size (envfit, R=0.46), these sediment variables can affect bacterial composition. Overall, the phyla Desulfobacterota, Chloroflexi and Myxococcota decreased their abundance as the sediment grain size increased, while phyla such as Cyanobacteria, Proteobacteria and Verrucomicrobia show an increase in the same conditions. When analyzing bacterial genera, it was apparent that an increase in small size sediment particles (i.e. mud) influences the same phyla (plus a member of the Bacteroidetes), as Pleurocaspa PCC-7319 (Cyanobacteria), Roseobacter (Bacteroidetes) and unclassified genera from the class Cyanobacteriia and

129

CHAPTER 4

Alphaproteobacteria; decreased their abundance. With the same increase in mud content, the genera Eudoraea (Bacteroidetes), Woeseia (Proteobacteria) and the unclassified genera from the classes Desulfobulbia, Polyangia and Ananerolineae (Chloroflexi), describe an increase as seen in some of their corresponding phyla. This result provides further evidence that physical properties of the sediment can change bacterial composition and potential functional roles. In burrows, in particular, sediment grain size and the percentage of mud will determine the amount of organic matter and resources available and, together with the organic matter quality and the redox conditions (Zheng et al. 2104), may shape these bacterial microniches and possibly change their metabolic capacity and resilience against pollutants.

4.5.4. Alternative explanations for the null effect of in situ nitrogen enrichment on bacterial communities The concentration of nitrogen enrichment chosen for this study (i.e. 600g N/m2) has been used in other projects where an increase in porewater ammonia was considered to be indicative of eutrophic sediments (Douglas et al. 2018). Nitrogen enrichment here effectively increased porewater ammonia levels below the sediment surface which may indicate eutrophic conditions (Appendix 4, Fig. D1) and effects on microbial abundance and functional shifts have reported enriched conditions similar to these (Chapter 3 and Filippini et al. 2019; Clark et al. 2020). In this study, some bacterial phyla showed a shift in relative abundance at higher nitrogen concentrations irrespective of the sediment position and showed shifts in abundance influenced by the selected environmental covariates. These included a significant increase in Firmicutes and a small decrease in Myxococcota. Firmicutes biomass increase has been associated with higher levels of eutrophication (Huang et al. 2017) and results here support this as higher abundance of Firmicutes was found at high nitrogen concentrations. However, the overall null effect of the nitrogen enrichment in this study may be explained by the presence of sufficient bacterial functional diversity and inter-site environmental differences that permit rapid use of nitrogen and greater

130

LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT community resilience (particularly in macrofauna burrows, Samuiloviene et al. 2019). Alternatively, the null result may be also explained as an artifact of the methods used. Bacterial community sampling in surface sediments and macrofaunal burrows were done at the sediment water interface where the impact of the porewater nitrogen addition may not have been as strong as predicted due to flushing. Sampling deeper sediment layers (i.e. lower interaction between sediment, water flows and atmosphere) a seen in similar studies (Clark et al. 2020) and a timely assessment from initial inoculation of organic enrichment may provide higher resolution and inform more adequately the short-term effects of the nitrogen enrichment on bacterial communities in future studies.

4.6. Conclusion Eutrophication of coastal environments is a global problem that can have important consequences for the functioning of these ecosystems. Estuarine intertidal areas are especially vulnerable to change as high inputs of allochthonous organic material (i.e. natural or human-derived) can produce large shifts in the local macro and microbial communities and disrupt the essential ecosystem services they provide. Results from this study highlight the importance of macrofaunal communities in potentially ameliorating the effect of human- derived organic disturbances in the sediments and identified how even small changes in local organic matter and sediment physical properties can drastically affect community composition and potential function. Here, nitrogen enrichment did not produce important shifts in community composition. Overall, bacterial communities in the selected study sites demonstrated high resistance to experimental nutrient manipulation but appeared to be more sensitive to the small changes to sediment organic matter content and mud content. Further research on short term impacts of organic enrichment on benthic assemblages and the interactive responses of macrofauna and microbial communities (including meiofauna and fungal communities) need to be considered to gain knowledge of the characteristics needed for a site to recover after human-derived disturbances and to propose optimized strategies for potential bioremediation of

131

CHAPTER 4 highly eutrophic systems. .

132

5. GENERAL DISCUSSION

The release of anthropogenically derived contaminants, particularly nutrients, into coastal environments is causing widespread problems for ecosystem structure and function (Sinha et al. 2017). Under the planetary boundaries framework that analyses the risk of human perturbations to the functionality of the earth system, biogeochemical flow (including nitrogen and phosphorous) is one of the boundaries that has changed the most in the last 70 years. They now sit at levels of high risk and near tipping points where recovery is uncertain (Steffen et al. 2015). The increased use of fertilizers in agriculture, higher discharges of human-derived urban wastewater and stormwater runoff have changed nutrient dynamics and biogeochemical processes in aquatic systems worldwide (Gilbert, 2017). In particular, changes to the stoichiometry of nitrogen and phosphorus (i.e. ratio of N:P, Sutton et al. 2013), the bioavailability of new forms of nitrogen (i.e. urea) and an overall increase of both nutrients have severely affected the ecosystem functions of aquatic systems (Peñuelas et al. 2012) and have led to situations of eutrophication (Howarth et al. 2011; Sinha et al. 2017), including blackwater events (King et al. 2012) and mortality of fish (Small et al. 2014) and microinvertebrates (Carmicheal et al. 2012). Many studies are now acknowledging the critical role of macro-micro interactions in controlling ecosystem functions (Laverock et al. 2011; Moulton et al. 2016; Dale et al. 2019) and how these types of interactions may be harnessed to improve restoration strategies in the context of nutrient pollution (Bergström et al. 2015; Brito et al. 2018; Mandario et al. 2019).

For coastal sediment systems, the widespread negative effects and potential tipping points associated with excess nutrient enrichment have received some attention (Gilbert et al. 2017; Sinha et al. 2017; Hewitt and Thrush, 2019). However, approaches in which both macrofaunal and microbial communities are studied, and links are made to both ecosystem structure and function remain rare (but see Kunihiro et al. 2011; Deng et al. 2018 and Nicholaus et al. 2019 for some examples where this has been studied). The integrated study of biotic fractions will increase our general ecological understanding and help predict the ecological impacts of contaminants on soft sedimentary systems. In this thesis, I have investigated the effect of nutrient enrichment at multiple biological levels and 133 GENERAL DISCUSSION explored the potential for sediment bioremediation through a systematic literature review and meta-analysis, laboratory and field experiments.

In Chapter 1 I conducted a systematic review and meta-analysis of the influence of bioturbating macrofauna on sediment contaminant fate and the effect of relevant environmental and experimental covariates. I found that the presence of bioturbators enhances the release of nutrients (i.e. ammonia and phosphorous) and other contaminants from the sediment (metals and PAHs) with possible effects on benthic metabolism (indirectly measured through increased sediment oxygen uptake). However, these results varied greatly between taxa and correlated with other environmental (e.g. temperature, type of aquatic system and sediment grain size) and biotic conditions (duration of exposure to bioturbator and animal density).

In Chapter 2 I used an experimental mesocosm approach to demonstrate that body size of bioturbators has a strong effect on their survivorship when they are exposed to highly enriched sediments. Further, small body sizes may experience apparent associational susceptibility when mixed with larger body sizes. Building on the results of Chapter 2, Chapter 3 demonstrated that high sediment organic loading leads to clear shifts of both bacterial and archaeal communities to favour anaerobic groups. Sediments with lower organic loads had higher abundance and diversity of microbial groups associated with aerobic processes. Somewhat surprisingly, the bivalve used in this study (Anadara trapezia) had little effect on sediment organic matter breakdown and associated bacterial and archaeal communities in experimental conditions.

In Chapter 4 I expanded the scale of my research to investigate the effect of an in situ nitrogen enrichment on bacterial communities in surface sediments and within macrofauna burrows. In addition, seven environmental covariates (organic matter content, mud content, porosity, median grain size, chlorophyll a concentration, phaeo pigment concentration and macrofauna abundance) were included in the analysis to investigate potential interactions with the in situ enrichment and to explore spatial differences caused by inter-site variability. The key findings of this chapter highlight that small changes in ambient organic

134

GENERAL DISCUSSION matter between sites may have a greater impact on bacterial communities than an in situ experimental nutrient enrichment. Further, bioturbator burrows were found to be areas of greater consistency with fewer effects of organic matter and mud content on bacterial communities relative to the sediment surface. Overall, the integration of approaches throughout the thesis has added information to several of the knowledge gaps identified in Chapter 1 and highlights important aspects to consider for the theory and application of sediment bioremediation.

5.1.Macrofauna sediment bioremediation: advantages and drawbacks Bioremediation approaches aimed at reducing the negative effects are an important alternative to other methods of pollution remediation in sediments. In comparison to methods such as capping or dredging, examples of bioremediation have proven to have less environmental impact, less costly and invasive (Boopathy et al. 2000). In particular, the introduction of bioturbating macrofauna in polluted sediments has proven to be a better option for bioremediation as these organisms are a natural component that provides aquatic systems with attributes for resilience against disturbances (Gladstone-Gallagher et al. 2020) and could benefit areas with long lasting effects (Lohrer et al. 2010). This capacity has been proven in many species of polychaetes that can survive harsh conditions and removed pollutants effectively (Banks et al. 2013). More importantly, this method has recently been effectively used for aquaculture waste management (Bergstörm et al. 2015; Brito et al. 2018; Mandario et al. 2019; Nicholaus et al. 2019 and Zhao et al. 2019). However clear drawbacks exist that can affect their application and incorporation in coastal management plans (Perelo et al. 2010).

5.1.1.Spatial variability and site-specific factors Spatial differences between sites (e.g. organic matter dynamics, input of pollutants and physical characteristics of sites) have contributed to a low predictability bioremediation efficiency and has led to results being constrained regionally and geographically (Arndt et al. 203). In this thesis, this was evident in Chapter 1 when analyzing different aquatic systems and when large scale effects were evaluated in the North Island New Zealand (Chapter 4). Here small changes in organic matter content between evaluated sites changed the bacterial

135

GENERAL DISCUSSION communities in the sediments. Although some bacterial functional redundancy was expected, this site-specific response to organic matter content may provide additional evidence that determines the differing thresholds of a site against a disturbance (Grall and Chevaud, 2002). In addition, these site-specific characteristics may also contribute to differing results when bioremediation is attempted as many bioturbators may not be able to survive pollutant exposure (e.g. A. trapezia in Chapter 2) or provide any beneficial effects that lead to the systems recovery (e.g. no effect of A. trapezia to microbial communities in Chapter 3).

Throughout my thesis I linked different site-specific environmental variables such as temperature, sediment grain size and background organic matter content to increased nutrient processing and associated bacterial diversity that leads to evaluate bioremediation in many contexts. I found that increases in temperature correlated positively with the release of ammonia and phosphorous as well as sediment oxygen uptake (see Chapter 1). This provides evidence that temperature may influence the applicability and effectiveness of bioremediation in different regions (i.e. tropical, temperate and polar). In tropical regions, higher temperatures may provide ideal conditions for eutrophic sediment bioremediation, however further research on the effects of temperature on bioturbators and microbial communities is needed (Berkenbusch and Rowden, 1999; Ouellette et al. 2004). This is especially true if new climatic conditions provoke an increase in temperature globally that can affect not only the bioremediation efforts but also impact the coupled bioturbators and microbial communities that influence ecosystem services and function (Godbold and Solan, 2013; Crespo et al. 2016; Isaev et al. 2017; Zhao et al. 2017). Moreover, if temperatures reach critical levels then bioturbator fitness may be reduced and bioremediation effectivity affected.

Sediment grain size and organic matter content are properties of sediments that are known to interact and influence local environmental conditions in the sediment such as redox state (Gilbert et al. 2016), abundance of bioturbators (Douglas et al. 2018), accumulation of organic compounds (Biswas et al. 2009; Geraldi et al. 2017) and the complexity of microbial microniches

136

GENERAL DISCUSSION

(Bertics and Ziebis, 2010). In my thesis I discovered clear effects of grain size on nutrient release and microbial activity (i.e. through SOU) that agree with findings from other scientific literature (see Chapter 1). Mainly, larger sediment sizes release more nutrients as less adhesion exists compared to fine sediments where a high accumulation of organic matter occurs and less space for nutrient flux is possible (Jackson and Weeks 2008; Douglas et al. 2018). In addition, I also found that that organic matter and mud content in sediments may be a crucial factor that shapes microbial communities even at small scale changes (1-6% organic matter content). (Chapter 4). Both factors, alone and in interaction, are relevant factors to consider in bioremediation as some aquatic systems with specific sediment properties and differing rates of natural organic inputs may have better prospects for bioremediation than others.

5.1.2. Importance of taxa selection: inter- and intraspecific species differences in bioremediation A large amount of scientific literature has described the variability of responses different taxa may have on bioremediation (Chapter 1) and what functional traits (e.g. mechanisms of bioturbation; Gerino et al. 2003) may predict better outcomes for a systems recovery. Interspecific differences in bioturbation mechanisms and how they influence sediment microbiota have received some attention (Papaspyrou et al. 2006; Foshtomi et al. 2015; Shen et al. 2016) and results from Chapter 1 summarize these studies in relation to contaminant fate in sediments (e.g. mainly studied for ammonia and phosphorous but also for metals and PAHs) and effects on microbial communities (e.g. SOU). However less is known about functional traits within bioturbating species and the variability of responses they may have against a disturbance (Lohrer et al. 2010; Mandario et al. 2019). In Chapters 2 and 3, I explored intraspecific differences in one bioturbator and revealed results that may inform bioremediation strategies. Here I used the common bivalve A. trapezia as a model bioturbator, and I discovered that larger individuals have higher survival in highly organically enriched sediments compared to smaller individuals. Initially this would mean that in enriched conditions, larger bioturbators from the same species would survive longer and have a greater influence on bioremediation through sediment reworking (as expected from results obtained in Chapter 1). This result is

137

GENERAL DISCUSSION consistent with most of the ecotoxicological literature that generally finds early life history stages, and small individuals to be the most sensitive to toxic conditions (Clements and Newman, 2003). However, I also found that large individuals were less mobile than smaller individuals, highlighting the complexity of bioturbation application in bioremediation. If we expect movement to reflect higher bioturbation rates that enhance organic matter breakdown then having large, immobile bivalves may not be a particularly effective treatment (Maire et al. 2008). Overall, my results from both Chapter 2 and 3, demonstrate that the identity of the bioturbator is crucial and applications should also select species with an appropriate balance of traits that increase stress tolerance (Chapter 2; Marsden et al. 2012: Riedel et al. 2012) but also increase rates of bioturbation (e.g. including increased mobility, higher oxygen bioirrigation or formation of biofilm with high microbial metabolic capacity).

The results shown throughout the thesis highlight the importance of taxa selection when attempting bioremediation. Not every bioturbating macroinvertebrate may pose as the candidate model to bioremediate specific pollutants and success will have high variability inherent to high intraspecific differences and their interaction with site-specific factors. In addition, low predictability of the outcomes of bioremediation attempts will also be associated with other benthic biotic components that inhabit same areas. This includes interactions of bioturbating macrofauna with bacterial and archaeal communities (as assessed in Chapter 2-4) but also to other assemblages such as fungal and meiofaunal communities that can shape the resilience of a system (Louati et al. 2017). As an example of the first benthic association (i.e. macrofauna- bacteria/archaea interaction) in my thesis, I did not find a link between the presence of A. trapezia, of any size, and the diversity or structure of bacterial or archaeal communities. While my results suggest that A. trapezia is not a great prospect for sediment bioremediation, they do emphasize the importance of considering inter- and intraspecific differences of bioturbating macrofauna- microbial interactions that may influence long lasting bioremediation effects

138

GENERAL DISCUSSION

5.1.3. Animal density and inter and intraspecific associational interactions Denser aggregations of bioturbators often increase oxygenation of sediments and lead to higher organic matter breakdown by microbial communities (Branch and Pringle, 1987; Hietanen et al. 2007; Papaspyrou et al. 2010). This was supported by the results in Chapter 1 where I found that animal density was correlated positively with nutrient fluxes and sediment oxygen uptake. Additionally, results from Chapter 4 highlight that macrofauna burrows may provide more environmentally stable niches for microbial settlement and possibly higher centres of organic matter breakdown if density is increased (not evaluated). Increasing animal density in an area for bioremediation however needs to be done carefully as inter and intraspecific effects may greatly influence the outcome. For example, some studies have shown an effect of highly diverse benthic assemblages on contaminant removal and overall survivorship (Ciarelli et al. 2000; Magnusson et al. 2003; Lukwambe et al. 2019). However this effect at the level of individual bioturbator’s populations has received little attention (Lohrer et al. 2010).In Chapter 2 I presented evidence suggesting that smaller bivalve body sizes may have experienced associational susceptibility when combined with larger body sizes in stressful (i.e. under high organic loads) conditions. Although I could not fully explore the mechanism behind this result, environmental stress may be associated with intraspecific interactions among phenotypes (e.g. competition). This idea is supported by a large body of research where it has been demonstrated that at the level of inter-specific interactions, important shifts from competitive to facilitative interactions along gradients of increasing environmental stress may occur (Malkinson and Tielborger, 2010; He et al. 2013; Dangles, 2019). Similar effects of stress on intra-specific interactions such as competition or facilitation on bioturbating macrofauna have not been addressed yet (Bulleri et al. 2011; Gessner and Hines, 2012) and the findings in my thesis may inform ecological theory (i.e. organic enrichment as stress and associational susceptibility by possible competition in bivalves) but also highlight an additional factor to be considered in bioremediation frameworks.

Overall, great care needs to be taken when deploying the bioturbator densities at contaminated sites as intra-specific differences may be a strong driver

139

GENERAL DISCUSSION of the final outcome. However, an additional drawback of density as a factor that can influence the effectivity of macrofauna sediment bioremediation, is the ability of the research team to have enough numbers available to be deployed and produce a positive effect. Chary et al. (2020) have suggested that although more density related studies are needed, in some cases stocking density of the model bioturbator might be problematic to obtain. Such is the example given by the study by Ma et al. (2015) in Sansha bay (southeast China) where different bioturbators were deployed (i.e. Perinereis aibuhitensis, P. nuntia and Tegillarca granosa). Here the extensive area to bioremediate and the high demand to cover densities proved logistically difficult and with high mortality rates of the bioturbators. In my thesis, density effects were not evaluated directly, however the low influence of A. trapezia on organic matter degradation and low effect on microbial communities may have correlated with low biomass deployed in the mesocosms. Careful selection of a bioturbator model with enough stocking availability for bioremediation will need to be considered if effective strategies at large scales are to be implemented.

5.2. Final remarks and future directions Through a multiple level and holistic approach (literature review, lab mesocosm and a large scale field experiment), I was able to investigate the effect of bioturbators on organically enriched sediments and vice versa, the effect of the enriched sediments on the bioturbators. My research was enhanced by the simultaneous examination of associated sediment microbial communities and assessment of the possible processes in which they may participate. I found some positive options for bioremediation applications and many possible aspects that need to be considered in future studies of sediment bioremediation. Overall, this thesis highlights the importance of environmental and biological factors that affect both benthic bioturbating macrofauna and microbial communities, and their potential to be used for the bioremediation of contaminants, specifically to solve worldwide issues of nutrient pollution. To end, I suggest some future directions for research into the use of macrofauna bioturbation as a bioremediation method in sediments:

140

GENERAL DISCUSSION

• My results point to the possibility of using bioturbating macrofauna for contaminants other than organic matter or nutrients. Results from chapter 1 demonstrate that metals and organic compounds such as PAHs can be accumulated, or biotransformed by macrofauna and their biogenic structures (i.e. burrows). The scarcity of studies on other contaminants (for example too few PAHs studies were available for a meta-analysis) makes it difficult to ascertain if the use of macrofauna can be applied to many compounds with global relevance such as metals, PAHs, PCBs, PFAS and microplastics. Further research on the toxic effects of such compounds on bioturbators and microbial communities will need to be conducted, as well as including the variables mentioned above (i.e. temperature, sediment grain size, organic matter content, etc.) as crucial factors that may constrain bioremediation efforts.

• Studies that can provide an adequate scientific basis for the selection of the bioturbating organism to be used in bioremediation. Here bioturbation mechanisms (e.g. grazers, biodiffusors, conveyor feeders or gallery diffusers; Gerino et al. 2013), tolerance to contaminants (i.e. proper background in toxicological response dosage assays), life-history stage (e.g. functional intra-specific trait consideration) and animal density (e.g. stock availability and inter assemblage interactions) should be considered. In addition, more comparative studies that compare single use of species versus complete benthic assemblages need to be considered as the interaction between macrofauna and microbial, fungal and meiofaunal communities may yield different results for bioremediation.

• Studies that establish which aquatic systems have good potential for bioremediation, as site-specific characteristics (i.e. background sediment organic matter content, mud content, grain size and temperature) may affect bioremediation outcomes. Moreover, if bioturbators release contaminants into the water column then bioremediation should be practiced in well flushed areas.

141

GENERAL DISCUSSION

• Small and large scale in situ studies where environmental variables are well monitored and microbial communities are simultaneously examined. These approaches will include spatial variability of responses to disturbances and can incorporate more complex metagenomic and metatranscriptomic approaches (such as in Birrer et al. 2019) to unravel functional traits of benthic communities regulated by biotic and biotic factors.

• Experimental manipulations (ex situ) should attend to specific questions that cannot be assessed directly in the field. For example, to determine the magnitude of impact contaminants may have on benthic communities or to evaluate microbial-derived benthic processes that would otherwise be logistically difficult to assess in the field. Continuing to assess bioremediation in ex situ conditions is likely to produce high variability in results that do not necessarily reflect real conditions in natural systems.

• Plan in situ studies where the bioremediation effects are monitored to ensure that positive outcomes remain on site after long periods of time. This will ensure that contaminant removal is complete, natural resilience is obtained and recovery is maintained.

• In highly polluted sites where bioturbators alone cannot produce a benefit for recovery (e.g. possibly organic enrichment in Chapter 2 and 3), bioremediation research should include facilitating factors such as aeration (Wang et al. 2019) or the addition of species that can assist the formation of benthic assemblages to increase recovery or enhance contaminant removal/degradation (Lukwambe et al. 2019)

The future research directions outlined above will help determine if bioremediation using macrofauna is possible in particular settings, however ongoing research will be needed to ensure approaches are efficient and optimized in a constantly changing world where multiple human derive stressors are on the rise.

142

REFERENCES

• Adriaens, P., Li, M. Y., and Michalak, A. M, (2006). Scaling methods of sediment bioremediation processes and applications. Eng. Life. Sci, 6(3), 217-227.

• Aguirre, J.D. and Marshall, D.J., (2012). Does genetic diversity reduce sibling competition? Evolution: International Journal of Organic Evolution, 66(1), pp.94-102.

• Ahmed, W., Hughes, B. and Harwood, V.J., (2016). Current status of marker genes of Bacteroides and related taxa for identifying sewage pollution in environmental waters. Water, 8(6), p.231.

• Albert, C.H., Grassein, F., Schurr, F.M., Vieilledent, G. and Violle, C., (2011). When and how should intraspecific variability be considered in trait-based plant ecology? Perspectives in Plant Ecology, Evolution and Systematics, 13(3), pp.217-225.

• Albertelli, G., Covazzi-Harriague, A., Danovaro, R., Fabiano, M., Fraschetti, S., & Pusceddu, A. (1999). Differential responses of bacteria, meiofauna and macrofauna in a shelf area (Ligurian Sea, NW Mediterranean): role of food availability. Journal of Sea Research, 42(1), 11-26.

• Ali, H., Khan, E. and Sajad, M.A., (2013). Phytoremediation of heavy metals—concepts and applications. Chemosphere, 91(7), pp.869-881.

• Aller, R. C. (1994). Bioturbation and remineralization of sedimentary organic matter: effects of redox oscillation. Chemical Geology, 114(3-4), 331-345.

• Altmann, D., Stief, P., Amann, R. and de Beer, D., (2004). Nitrification in freshwater sediments as influenced by insect larvae: quantification by microsensors and fluorescence in situ hybridization. Microbial ecology, 48(2), pp.145-153.

145

• Amato, E. D., Simpson, S. L., Remaili, T. M., Spadaro, D. A., Jarolimek, C. V., and Jolley, D. F. (2016). Assessing the effects of bioturbation on metal bioavailability in contaminated sediments by diffusive gradients in thin films (DGT). Environ. Sci. Technol., 50(6), 3055-3064.

• Anderson, M. J. (2017). Permutational multivariate analysis of variance (PERMANOVA). Wiley statsref: statistics reference online, 1-15.

• Andres, S., Ribeyre, F., and Boudou, A. (1998). Effects of temperature and exposure duration on transfer of cadmium between naturally contaminated sediments and burrowing mayfly nymphs (Hexagenia rigida). Arch. Environ. Con. Tox., 35(2), 295-301.

• Ansari, A.A., Singh, G.S., Lanza, G.R. and Rast, W. eds., (2010). Eutrophication: causes, consequences and control (Vol. 1). Springer Science and Business Media.

• Aoki, M., Ehara, M., Saito, Y., Yoshioka, H., Miyazaki, M., Saito, Y., Miyashita, A., Kawakami, S., Yamaguchi, T., Ohashi, A. and Nunoura, T., (2014). A long-term cultivation of an anaerobic methane-oxidizing microbial community from deep-sea methane-seep sediment using a continuous-flow bioreactor. PLoS One, 9(8), p.e105356.

• Arar, E. J., & Collins, G. B. (1997). Method 445.0: In vitro determination of chlorophyll a and pheophytin a in marine and freshwater algae by fluorescence. Washington, DC, USA: United States Environmental Protection Agency, Office of Research and Development, National Exposure Research Laboratory

• Arndt, S., Jørgensen, B.B., LaRowe, D.E., Middelburg, J.J., Pancost, R.D. and Regnier, P., (2013). Quantifying the degradation of organic matter in marine sediments: a review and synthesis. Earth-science reviews, 123, pp.53-86.

146

• Bai, Y., Shi, Q., Wen, D., Li, Z., Jefferson, W.A., Feng, C. and Tang, X., (2012). Bacterial communities in the sediments of Dianchi Lake, a partitioned eutrophic waterbody in China. PloS one, 7(5), p.e37796.

• Balfour, D.L. and Smock, L.A., (1995). Distribution, age structure, and movements of the freshwater mussel Elliptio complanata (Mollusca: Unionidae) in a headwater stream. Journal of freshwater ecology, 10(3), pp.255-268.

• Banks, J. L., Ross, D. J., Keough, M. J., Macleod, C. K., Keane, J., & Eyre, B. D. (2013). Influence of a burrowing, metal-tolerant polychaete on benthic metabolism, denitrification and nitrogen regeneration in contaminated estuarine sediments. Marine pollution bulletin, 68(1-2), 30-37.

• Banta, G.T., Holmer, M., Jensen, M.H. and Kristensen, E., (1999). Effects of two polychaete worms, Nereis diversicolor and Arenicola marina, on aerobic and anaerobic decomposition in a sandy marine sediment. Aquatic Microbial Ecology, 19(2), pp.189-204.

• Barbosa, P., Hines, J., Kaplan, I., Martinson, H., Szczepaniec, A. and Szendrei, Z., (2009). Associational resistance and associational susceptibility: having right or wrong neighbors. Annual Review of Ecology, Evolution, and Systematics, 40, pp.1-20.

• Bartoli, M., Nizzoli, D., Viaroli, P., Turolla, E., Castaldelli, G., Fano, E. A., and Rossi, R. (2001). Impact of Tapes philippinarum farming on nutrient dynamics and benthic respiration in the Sacca di Goro. Hydrobiologica, 455(1-3), 203-212.

• Bartolini, F., Penha-Lopes, G., Limbu, S., Paula, J. and Cannicci, S., (2009). Behavioural responses of the mangrove fiddler crab (Uca annulipes and U. inversa) to urban sewage loadings: Results of a mesocosm approach. Marine Pollution Bulletin, 58(12), pp.1860-1867.

147

• Bates, D., Mächler, M., Bolker, B., & Walker, S. (2014). Fitting linear mixed-effects models using lme4. arXiv preprint arXiv:1406.5823.

• Beal, B. F. (2006). Relative importance of predation and intraspecific competition in regulating growth and survival of juveniles of the soft- shell clam, Mya arenaria L., at several spatial scales. Journal of Experimental Marine Biology and Ecology, 336(1), 1-17.

• Bergström, P., Carlsson, M.S., Lindegarth, M., Petersen, J.K., Lindegarth, S. and Holmer, M., (2017). Testing the potential for improving quality of sediments impacted by mussel farms using bioturbating polychaete worms. Aquaculture research, 48(1), pp.161-176.

• Berkenbusch, K., and Rowden, A. A. (1999). Factors influencing sediment turnover by the burrowing ghost shrimp Callianassa filholi (Decapoda: Thalassinidea). J. Exp. Mar. Biol. Ecol, 238(2), 283-292.

• Bernard, G., Duchêne, J. C., Romero-Ramirez, A., Lecroart, P., Maire, O., Ciutat, A., and Grémare, A. (2016). Experimental Assessment of the Effects of Temperature and Food Availability on Particle Mixing by the Bivalve Abra alba Using New Image Analysis Techniques. PloS one, 11(4), e0154270.

• Bertics, V.J. and Ziebis, W., (2009). Biodiversity of benthic microbial communities in bioturbated coastal sediments is controlled by geochemical microniches. The ISME journal, 3(11), p.1269.

• Bertics, V.J. and Ziebis, W., (2010). Bioturbation and the role of microniches for sulfate reduction in coastal marine sediments. Environmental microbiology, 12(11), pp.3022-3034.

• Bertics, V.J., Sohm, J.A., Treude, T., Chow, C.E.T., Capone, D.G., Fuhrman, J.A. and Ziebis, W., (2010). Burrowing deeper into benthic nitrogen cycling: the impact of bioturbation on nitrogen fixation coupled to sulfate reduction. Marine Ecology Progress Series, 409, pp.1-15.

148

• Biles, C. L., Paterson, D. M., Ford, R. B., Solan, M., and Raffaelli, D. G. (2002). Bioturbation, ecosystem functioning and community structure. Hydrol. Earth Syst. Sci., 6(6), 999-1005.

• Birrer, S.C., Dafforn, K.A., Simpson, S.L., Kelaher, B.P., Potts, J., Scanes, P. and Johnston, E.L., (2018). Interactive effects of multiple stressors revealed by sequencing total (DNA) and active (RNA) components of experimental sediment microbial communities. Science of The Total Environment, 637, pp.1383-1394.

• Birrer, S.C., Dafforn, K.A., Sun, M.Y., Williams, R.B., Potts, J., Scanes, P., Kelaher, B.P., Simpson, S.L., Kjelleberg, S., Swarup, S. and Steinberg, P., (2019). Using meta‐omics of contaminated sediments to monitor changes in pathways relevant to climate regulation. Environmental Microbiology, 21(1), pp.389-401.

• Biswas, J.K., Rana, S., Bhakta, J.N. and Jana, B.B., (2009). Bioturbation potential of chironomid larvae for the sediment–water phosphorus exchange in simulated pond systems of varied nutrient enrichment. Ecological Engineering, 35(10), pp.1444-1453.

• Boehm, P. D. (2005). Polycyclic aromatic hydrocarbons (PAHs). In R.D. Morrison and B.L. Murphy (Eds.) Environmental Forensics (pp. 313- 337). San Diego, CA: Elsevier Academic Press.

• Boeker, C., Lueders, T., Mueller, M., Pander, J. and Geist, J., (2016). Alteration of physico-chemical and microbial properties in freshwater substrates by burrowing invertebrates. Limnologica, 59, pp.131-139.

• Bolger, A.M., Lohse, M. and Usadel, B., (2014). Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics, 30(15), pp.2114- 2120.

149

• Bolnick, D.I., Amarasekare, P., Araújo, M.S., Bürger, R., Levine, J.M., Novak, M., Rudolf, V.H., Schreiber, S.J., Urban, M.C. and Vasseur, D.A., (2011). Why intraspecific trait variation matters in community ecology. Trends in Ecology and Evolution, 26(4), pp.183-192.

• Bonaglia, S., Bartoli, M., Gunnarsson, J.S., Rahm, L., Raymond, C., Svensson, O., Yekta, S.S. and Brüchert, V., (2013). Effect of reoxygenation and Marenzelleria spp. bioturbation on Baltic Sea sediment metabolism. Marine Ecology Progress Series, 482, pp.43-55.

• Bonaglia, S., Nascimento, F. A., Bartoli, M., Klawonn, I., & Brüchert, V. (2014). Meiofauna increases bacterial denitrification in marine sediments. Nature communications, 5(1), 1-9.

• Boopathy, R. (2000). Factors limiting bioremediation technologies. Bioresource technol., 74(1), 63-67.

• Braeckman, U., Provoost, P., Gribsholt, B., Van Gansbeke, D., Middelburg, J.J., Soetaert, K., Vincx, M. and Vanaverbeke, J., (2010). Role of macrofauna functional traits and density in biogeochemical fluxes and bioturbation. Marine Ecology Progress Series, 399, pp.173-186.

• Branch, G.M. and Pringle, A., (1987). The impact of the sand prawn Callianassa kraussi Stebbing on sediment turnover and on bacteria, meiofauna, and benthic microflora. Journal of Experimental Marine Biology and Ecology, 107(3), pp.219-235.

• Brettar, I., Christen, R., Bötel, J., Lünsdorf, H. and Höfle, M.G., (2007). Anderseniella baltica gen. nov., sp. nov., a novel marine bacterium of the Alphaproteobacteria isolated from sediment in the central Baltic Sea. International journal of systematic and evolutionary microbiology, 57(10), pp.2399-2405.

• Brichette, I., Reyero, M.I. and Garcı́a, C., (2001). A genetic analysis of intraspecific competition for growth in mussel cultures. Aquaculture, 192(2-4), pp.155-169.

150

• Bright, M., Espada-Hinojosa, S., Lagkouvardos, I. and Volland, J.M., (2014). The giant ciliate Zoothamnium niveum and its thiotrophic epibiont Candidatus Thiobios zoothamnicoli: a model system to study interspecies cooperation. Frontiers in microbiology, 5, p.145.

• Brito, L.O., Cardoso Junior, L.D.O., Lavander, H.D., Abreu, J.L.D., Severi, W. and Gálvez, A.O., (2018). Bioremediation of shrimp biofloc wastewater using clam, seaweed and fish. Chemistry and Ecology, 34(10), pp.901-913.

• Brooks, M. E., Kristensen, K., Van Benthem, K.J., Magnusson, A., Berg, C.W., Nielsen, A., Skaug, H.J., Maechler, M. and Bolker, B. M. (2017). glmmTMB Balances Speed and Flexibility Among Packages for Zero- inflated Generalized Linear Mixed Modelling. The R Journal, 9(2), pp. 378-400.

• Bulleri, F., Cristaudo, C., Alestra, T. and Benedetti‐Cecchi, L., (2011). Crossing gradients of consumer pressure and physical stress on shallow rocky reefs: a test of the stress‐gradient hypothesis. Journal of Ecology, 99(1), pp.335-344.

• Bunch, N.D. and Bernot, M.J., (2012). Nitrate and ammonium uptake by natural stream sediment microbial communities in response to nutrient enrichment. Research in microbiology, 163(2), pp.137-141.

• Burton, G. A., and Johnston, E. L. (2010). Assessing contaminated sediments in the context of multiple stressors. Environ. Tox. Chem., 29(12), 2625-2643.

• Buscot, F. (2005). Microorganisms in soils: roles in genesis and functions, pp. 139-153, A. Varma (Ed.). Germany: Springer.

• Byers, J.E., Wright, J.T. and Gribben, P.E., (2010). Variable direct and indirect effects of a habitat‐modifying invasive species on mortality of native fauna. Ecology, 91(6), pp.1787-1798.

151

• Callahan, B. J., McMurdie, P. J., Rosen, M. J., Han, A. W., Johnson, A. J. A., & Holmes, S. P. (2016). DADA2: high-resolution sample inference from Illumina amplicon data. Nature methods, 13(7), 581-583.

• Carmichael, R.H., Shriver, A.C. and Valiela, I., (2012). Bivalve response to estuarine eutrophication: the balance between enhanced food supply and habitat alterations. Journal of Shellfish Research, 31(1), pp.1-12.

• Caro-Quintero, A., Ritalahti, K.M., Cusick, K.D., Löffler, F.E. and Konstantinidis, K.T., (2012). The chimeric genome of Sphaerochaeta: nonspiral spirochetes that break with the prevalent dogma in spirochete biology. MBio, 3(3), pp.e00025-12.

• Cassidy, C., Grange, L. J., Garcia, C., Bolam, S. G., & Godbold, J. A. (2020). Species interactions and environmental context affect intraspecific behavioural trait variation and ecosystem function. Proceedings of the Royal Society B, 287(1919), 20192143.

• Chan, Y., Li, A., Gopalakrishnan, S., Shin, P.K., Wu, R.S., Pointing, S.B. and Chiu, J.M., (2014). Interactive effects of hypoxia and polybrominated diphenyl ethers (PBDEs) on microbial community assembly in surface marine sediments. Marine pollution bulletin, 85(2), pp.400-409.

• Chapman, P. M., and Wang, F. (2001). Assessing sediment contamination in estuaries. Environ. Tox. Chem., 20(1), 3-22.

• Chapman, P. M. (2018). Sediment remediation can include no action. Ecotoxicology and Environmental Contamination, 13(2), 1-3.

• Chary, K., Aubin, J., Sadoul, B., Fiandrino, A., Covès, D. and Callier, M.D., (2020). Integrated multi-trophic aquaculture of red drum (Sciaenops ocellatus) and sea cucumber (Holothuria scabra): Assessing bioremediation and life-cycle impacts. Aquaculture, 516, p.734621.

152

• Chen, M., Ding, S., Liu, L., Wang, Y., Xing, X., Wang, D. and Zhang, C. (2016). Fine-scale bioturbation effects of tubificid worm (Limnodrilus hoffmeisteri) on the lability of phosphorus in sediments. Environ. Pollut., 219, 604-611.

• Chen, Q., Fan, J., Su, J., Ming, H., Sun, Z., Li, M., Zhao, X., Wang, Y., Zhang,Y., Zhang, H., Jin, Y., Ma, X., and Wang, B. (2020). Spatial distribution characteristics of bacterial community structure and gene abundance in sediments of the Bohai Sea. Acta Oceanologica Sinica, 39(2), 69-78.

• Christensen, M., Banta, G. T., and Andersen, O. (2002). Effects of the polychaetes Nereis diversicolor and Arenicola marina on the fate and distribution of pyrene in sediments. Mar. Ecol. Prog. Ser., 237, 159-172.

• Ciarelli, S., Kater, B. J., & van Straalen, N. M. (2000). Influence of bioturbation by the amphipod Corophium volutator on fluoranthene uptake in the marine polychaete Nereis virens. Environmental Toxicology and Chemistry: An International Journal, 19(6), 1575-1581.

• Ciutat, A., Widdows, J., and Readman, J. W. (2006). Influence of cockle Cerastoderma edule bioturbation and tidal-current cycles on resuspension of sediment and polycyclic aromatic hydrocarbons. Mar. Ecol. Prog. Ser., 328, 51-64.

• Clark, G. F., Kelaher, B. P., Dafforn, K. A., Coleman, M. A., Knott, N. A., Marzinelli, E. M., and Johnston, E. L. (2015). What does impacted look like? High diversity and abundance of epibiota in modified estuaries. Environ. Pollut., 196, 12-20.

• Clark, M.S., Husmann, G., Thorne, M.A., Burns, G., Truebano, M., Peck, L.S., Abele, D. and Philipp, E.E., (2013). Hypoxia impacts large adults first: consequences in a warming world. Global Change Biology, 19(7), pp.2251-2263.

153

• Clark, D. E., Pilditch, C. A., Pearman, J. K., Ellis, J. I., & Zaiko, A. (2020). Environmental DNA metabarcoding reveals estuarine benthic community response to nutrient enrichment–evidence from an in-situ experiment. Environmental Pollution, 115472.

• Clavero, V., Niell, F.X. and Fernandez, J., (1994). A laboratory study to quantify the influence of Nereis diversicolor in the exchange of phosphate between sediment and water. Journal of experimental marine biology and ecology, 176(2), pp.257-267.

• Clements, W. H., and Newman, M. C. (2003). Community ecotoxicology. John Wiley and Sons, pp. 23-40.

• Colombo, V., Pettigrove, V. J., Hoffmann, A. A., and Golding, L. A. (2016). Effects of Lumbriculus variegatus (Annelida, Oligochaete) bioturbation on zinc sediment chemistry and toxicity to the epibenthic invertebrate Chironomus tepperi (Diptera: Chironomidae). Environ. Pollut., 216, 198- 207.

• Compte-Port, S., Subirats, J., Fillol, M., Sànchez-Melsió, A., Marcé, R., Rivas-Ruiz, P., Rosell-Melé, A. and Borrego, C.M., (2017). Abundance and co-distribution of widespread marine archaeal lineages in surface sediments of freshwater water bodies across the Iberian Peninsula. Microbial ecology, 74(4), pp.776-787.

• Cozzoli, F., Bouma, T.J., Ottolander, P., Lluch, M.S., Ysebaert, T. and Herman, P.M., (2018). The combined influence of body size and density on cohesive sediment resuspension by bioturbators. Scientific Reports, 8(1), p.3831.

• Creese, R. G., & Underwood, A. J. (1982). Analysis of inter-and intraspecific competition amongst intertidal limpets with different methods of feeding. Oecologia, 53(3), 337-346.

154

• Crespo, D., Grilo, T.F., Baptista, J., Coelho, J.P., Lillebø, A.I., Cássio, F., Fernandes, I., Pascoal, C., Pardal, M.Â. and Dolbeth, M., (2017). New climatic targets against global warming: will the maximum 2° C temperature rise affect estuarine benthic communities? Scientific reports, 7(1), p.3918.

• Dafforn, K. A., Baird, D. J., Chariton, A. A., Sun, M. Y., Brown, M. V., Simpson, S. L., ... and Johnston, E. L. (2014). Faster, higher and stronger? The pros and cons of molecular faunal data for assessing ecosystem condition. In G. G, Woodward, A. Dumbrell, D. Baird and M. Hajibabaei (Eds.) Adv. Ecol. Res., 51, pp. 1-40). San Diego CA: Elsevier Academic Press.

• Dafforn, K.A., Kelaher, B.P., Simpson, S.L., Coleman, M.A., Hutchings, P.A., Clark, G.F., Knott, N.A., Doblin, M.A. and Johnston, E.L., (2013). Polychaete richness and abundance enhanced in anthropogenically modified estuaries despite high concentrations of toxic contaminants. PLoS One, 8(9), p.e77018.

• Dakos, V., Matthews, B., Hendry, A.P., Levine, J., Loeuille, N., Norberg, J., Nosil, P., Scheffer, M. and De Meester, L., (2019). Ecosystem tipping points in an evolving world. Nature ecology and evolution, 3(3), pp.355- 362.

• Dale, H., Taylor, J. D., Solan, M., Lam, P., and Cunliffe, M. (2018). Polychaete mucopolysaccharide alters sediment microbial diversity and stimulates ammonia-oxidising functional groups. FEMS microbiology ecology, 95(2).

• Dale, H., Solan, M., Lam, P., and Cunliffe, M. (2019). Sediment microbial assemblage structure is modified by marine polychaete gut passage. FEMS microbiology ecology, 95(5).

• Dangles, O., (2019). A dynamic model of facilitation on environmental stress gradients. Oikos. Pp. 1206-1214.

155

• De Figueiredo, D.R., Pereira, M.J., Moura, A., Silva, L., Bárrios, S., Fonseca, F., Henriques, I. and Correia, A., (2007). Bacterial community composition over a dry winter in meso-and eutrophic Portuguese water bodies. FEMS microbiology ecology, 59(3), pp.638-650.

• Del Re A. C., and Hoyt, W. T. (2014) MAd: Meta-Analysis with Mean Differences. R package version 0.8-2.

• Delacuvellerie, A., Cyriaque, V., Gobert, S., Benali, S. and Wattiez, R., (2019). The plastisphere in marine ecosystem hosts potential specific microbial degraders including Alcanivorax borkumensis as a key player for the low-density polyethylene degradation. Journal of hazardous materials, 380, p.120899.

• Deng, F., Ma, Y., Li, J., Wang, Y.Z., Yan, Q.P., Yan, X.Z. and Lin, M., (2015). Archaeal community structure and response to ark shell bioturbation in typical intertidal mudflats, Southeast coast of China. Continental Shelf Research, 106, pp.97-106.

• Deng, F., Wang, Y., Ouyang, J. and Ma, Y., (2018). Genetic Diversity of Active Ammonia Oxidizing Archaea and Variation during Macrobenthic Bioremediation in Mudflats of Sansha Bay, China. Journal of Coastal Research, 84(sp1), pp.51-57.

• Des Roches, S., Post, D.M., Turley, N.E., Bailey, J.K., Hendry, A.P., Kinnison, M.T., Schweitzer, J.A. and Palkovacs, E.P., (2018). The ecological importance of intraspecific variation. Nature Ecology and Evolution, 2(1), p.57.

• Ding, S. Y., Wang, L., Xu, H. C., Bao, X. Y., Wang, H., Chang, Y. Q., and Ding, J. (2019). Bacterial community structure and function in the intestinal tracts and culture environment of sea cucumber (Apostichopus japonicus). Chinese Journal of Ecology, 38(1), 210.

156

• Dissanayake, N.G., Frid, C.L., Drylie, T.P. and Caswell, B.A., (2018). Ecological functioning of mudflats: global analysis reveals both regional differences and widespread conservation of functioning. Marine Ecology Progress Series, 604, pp.1-20.

• Dolbeth, M., Crespo, D., Leston, S., and Solan, M. (2019). Realistic scenarios of environmental disturbance lead to functionally important changes in benthic species-environment interactions. Marine environmental research, 150, 104770.

• Donadi, S., Eriksson, B.K., Lettmann, K.A., Hodapp, D., Wolff, J.O. and Hillebrand, H., (2015). The body-size structure of macrobenthos changes predictably along gradients of hydrodynamic stress and organic enrichment. Marine biology, 162(3), pp.675-685.

• Douglas, E.J., Pilditch, C.A., Kraan, C., Schipper, L.A., Lohrer, A.M. and Thrush, S.F., (2017). Macrofaunal functional diversity provides resilience to nutrient enrichment in coastal sediments. Ecosystems, 20(7), pp.1324- 1336.

• Douglas, E.J., Pilditch, C.A., Lohrer, A.M., Savage, C., Schipper, L.A. and Thrush, S.F., (2018). Sedimentary environment influences ecosystem response to nutrient enrichment. Estuaries and coasts, 41(7), pp.1994- 2008.

• Duffy, M.A., (2010). Ecological consequences of intraspecific variation in lake Daphnia. Freshwater Biology, 55(5), pp.995-1004.

• Dussud, C., Meistertzheim, A.L., Conan, P., Pujo-Pay, M., George, M., Fabre, P., Coudane, J., Higgs, P., Elineau, A., Pedrotti, M.L. and Gorsky, G., (2018). Evidence of niche partitioning among bacteria living on plastics, organic particles and surrounding seawaters. Environmental pollution, 236, pp.807-816.

157

• Duval, S., and Tweedie, R. (2000). Trim and fill: a simple funnel‐plot– based method of testing and adjusting for publication bias in meta‐ analysis. Biometrics, 56(2), 455-463.

• Edgar, G.J., (2001). Australian Marine Habitats in Temperate Waters. New Holland.

• Edgar, R.C., (2016). UNOISE2: improved error-correction for Illumina 16S and ITS amplicon sequencing. BioRxiv, p.081257.

• Edgar, R.C., Haas, B.J., Clemente, J.C., Quince, C. and Knight, R., (2011). UCHIME improves sensitivity and speed of chimera detection. Bioinformatics, 27(16), pp.2194-2200.

• Egger, M., Smith, G. D., Schneider, M., and Minder, C. (1997). Bias in meta-analysis detected by a simple, graphical test. Bmj, 315(7109), 629- 634.

• Ekeroth, N., Lindström, M., Blomqvist, S., and Hall, P. O. (2012). Recolonisation by macrobenthos mobilises organic phosphorus from reoxidised Baltic Sea sediments. Aquat. Geochem., 18(6), 499-513.

• Elmqvist, T., Folke, C., Nyström, M., Peterson, G., Bengtsson, J., Walker, B., & Norberg, J. (2003). Response diversity, ecosystem change, and resilience. Frontiers in Ecology and the Environment, 1(9), 488-494.

• El‐Sabaawi, R.W., Bassar, R.D., Rakowski, C., Marshall, M.C., Bryan, B.L., Thomas, S.N., Pringle, C., Reznick, D.N. and Flecker, A.S., (2015). Intraspecific phenotypic differences in fish affect ecosystem processes as much as bottom–up factors. Oikos, 124(9), pp.1181-1191.

• Fan, X., Ding, S., Gong, M., Chen, M., Gao, S., Jin, Z. and Tsang, D.C., (2018). Different influences of bacterial communities on Fe (III) reduction and phosphorus availability in sediments of the cyanobacteria- and macrophyte-dominated zones. Frontiers in microbiology, 9, p.2636.

158

• Fernández-Gomez, B., Richter, M., Schüler, M., Pinhassi, J., Acinas, S.G., González, J.M. and Pedros-Alio, C., (2013). Ecology of marine Bacteroidetes: a comparative genomics approach. The ISME journal, 7(5), p.1026.

• Fierer, N. (2017). Embracing the unknown: disentangling the complexities of the soil microbiome. Nature Reviews Microbiology, 15(10), 579.

• Filippini, G., Bugnot, A.B., Johnston, E.L., Ruszczyk, J., Potts, J., Scanes, P., Ferguson, A., Ostrowski, M., Varkey, D. and Dafforn, K.A., (2019). Sediment bacterial communities associated with environmental factors in Intermittently Closed and Open Lakes and Lagoons (ICOLLs). Science of the Total Environment, 693, p.133462.

• Findlay, S.E., Sinsabaugh, R.L., Sobczak, W.V. and Hoostal, M., (2003). Metabolic and structural response of hyporheic microbial communities to variations in supply of dissolved organic matter. Limnology and oceanography, 48(4), pp.1608-1617.

• Foshtomi, M. Y., Braeckman, U., Derycke, S., Sapp, M., Van Gansbeke, D., Sabbe, K., ... and Vanaverbeke, J. (2015). The link between microbial diversity and nitrogen cycling in marine sediments is modulated by macrofaunal bioturbation. PloS one, 10(6), e0130116.

• Foster, S.D. and Bravington, M.V., (2013). A Poisson–Gamma model for analysis of ecological non-negative continuous data. Environmental and Ecological Statistics, 20(4), pp.533-552.

• Fox, J. and Weisberg, S. (2019). An {R} Companion to Applied Regression, Third Edition.

• Fries, J.S., Characklis, G.W. and Noble, R.T., (2008). Sediment–water exchange of Vibrio sp. and fecal indicator bacteria: implications for persistence and transport in the Neuse River Estuary, North Carolina, USA. Water research, 42(4-5), pp.941-950.

159

• Galloway, J.N., Aber, J.D., Erisman, J.W., Seitzinger, S.P., Howarth, R.W., Cowling, E.B. and Cosby, B.J., (2003). The nitrogen cascades. Bioscience, 53(4), pp.341-356.

• Gamfeldt, L., Wallén, J., Jonsson, P.R., Berntsson, K.M. and Havenhand, J.N., (2005). Increasing intraspecific diversity enhances settling success in a marine invertebrate. Ecology, 86(12), pp.3219-3224.

• Gammal, J., Norkko, J., Pilditch, C. A., and Norkko, A. (2017). Coastal hypoxia and the importance of benthic macrofauna communities for ecosystem functioning. Estuaries Coasts, 40(2), 457-468.

• Galand, P. E., Lucas, S., Fagervold, S. K., Peru, E., Pruski, A. M., Vétion, G., ... & Guizien, K. (2016). Disturbance increases microbial community diversity and production in marine sediments. Frontiers in microbiology, 7, 1950.

• Geraldi, N.R., Bertolini, C., Emmerson, M.C., Roberts, D., Sigwart, J.D. and Connor, N.E., (2017). Aggregations of brittle stars can perform similar ecological roles as mussel reefs. Marine Ecology Progress Series, 563, pp.157-167.

• Gérino, M., Stora, G., Francois, F., Gilbert, F., Poggiale, J.C., Mermillod- Blondin, F., Desrosiers, G. and Vervier, P., (2003). Macro-invertebrate functional groups in freshwater and marine sediments: a common mechanistic classification. Vie et milieu, 53(4), pp.221-232.

• Gessner, M.O. and Hines, J., (2012). Stress as a modifier of biodiversity effects on ecosystem processes? Journal of Animal Ecology, 81(6), pp.1143-1145.

• Geta, R., Postolache, C., and Vădineanu, A. (2004). Ecological significance of nitrogen cycling by tubificid communities in shallow eutrophic lakes of the Danube Delta. Hydrobiologia, 524(1), 193-202.

160

• Gilbert, F., Hulth, S., Grossi, V. and Aller, R.C., (2016). Redox oscillation and benthic nitrogen mineralization within burrowed sediments: An experimental simulation at low frequency. Journal of Experimental Marine Biology and Ecology, 482, pp.75-84.

• Gilbert, P.M., (2017). Eutrophication, harmful algae and biodiversity— Challenging paradigms in a world of complex nutrient changes. Marine pollution bulletin, 124(2), pp.591-606.

• Gilbertson, W.W., Solan, M. and Prosser, J.I., (2012). Differential effects of microorganism–invertebrate interactions on benthic nitrogen cycling. FEMS microbiology ecology, 82(1), pp.11-22.

• Gladstone-Gallagher, R. V., Pilditch, C. A., Stephenson, F., & Thrush, S. F. (2019). Linking traits across ecological scales determines functional resilience. Trends in ecology & evolution.

• Glud, R. N., Berg, P., Stahl, H., Hume, A., Larsen, M., Eyre, B. D., and Cook, P. L. (2016). Benthic carbon mineralization and nutrient turnover in a Scottish sea loch: an integrative in situ study. Aquatic Geochemistry, 22(5-6), 443-467.

• Godbold, J. A., & Solan, M. (2009). Relative importance of biodiversity and the abiotic environment in mediating an ecosystem process. Marine Ecology Progress Series, 396, 273-282.

• Godbold, J.A. and Solan, M., (2013). Long-term effects of warming and ocean acidification are modified by seasonal variation in species responses and environmental conditions. Philosophical Transactions of the Royal Society B: Biological Sciences, 368(1627), p.20130186.

• Gonzalez, A., Germain, R. M., Srivastava, D. S., Filotas, E., Dee, L. E., Gravel, D., ... & Montoya, J. (2020). Scaling‐up biodiversity‐ecosystem functioning research. Ecology Letters, 23(4), 757-776.

161

• Grall, J., and Chauvaud, L. (2002). Marine eutrophication and benthos: the need for new approaches and concepts. Glob. Change Biol., 8(9), 813- 830.

• Gribben, P.E. and Wright, J.T., (2006). Invasive seaweed enhances recruitment of a native bivalve: roles of refuge from predation and the habitat choice of recruits. Marine Ecology Progress Series, 318, pp.177- 185.

• Gribben, P.E. and Wright, J.T., (2006). Sublethal effects on reproduction in native fauna: are females more vulnerable to biological invasion? Oecologia, 149(2), p.352.

• Gribben, P.E., Byers, J.E., Clements, M., McKenzie, L.A., Steinberg, P.D. and Wright, J.T., (2009). Behavioural interactions between ecosystem engineers control community species richness. Ecology Letters, 12(11), pp.1127-1136.

• Gribben, P.E., Nielsen, S., Seymour, J.R., Bradley, D.J., West, M.N. and Thomas, T., (2017). Microbial communities in marine sediments modify success of an invasive macrophyte. Scientific reports, 7(1), p.9845.

• Hale, R., Jacques, R. O., & Tolhurst, T. J. (2015). Cryogenic defaunation of sediments in the field. Journal of Coastal Research, 31(6), 1537-1540.

• Hale, R., Godbold, J. A., Sciberras, M., Dwight, J., Wood, C. L., Hiddink, J., and Solan, M. (2017). Mediation of macronutrients and carbon by post-disturbance shelf sea sediment communities. Biogeochemistry, 135(1-2), 121–133

• Haller, L., Tonolla, M., Zopfi, J., Peduzzi, R., Wildi, W. and Pote, J., (2011). Composition of bacterial and archaeal communities in freshwater sediments with different contamination levels (Lake Geneva, Switzerland). Water research, 45(3), pp.1213-1228.

162

• Hamdan, H.Z., Salam, D.A. and Saikaly, P.E., (2019). Characterization of the microbial community diversity and composition of the coast of Lebanon: Potential for petroleum oil biodegradation. Marine pollution bulletin, 149, p.110508.

• Hamner, W.M., (1978). Intraspecific competition in Tridacna crocea, a burrowing bivalve. Oecologia, 34(3), pp.267-281.

• Hanley, T.C., Hughes, A.R., Williams, B., Garland, H. and Kimbro, D.L., (2016). Effects of intraspecific diversity on survivorship, growth, and recruitment of the eastern oyster across sites. Ecology, 97(6), pp.1518- 1529.

• Harms, H., Schlosser, D., and Wick, L. Y. (2011). Untapped potential: exploiting fungi in bioremediation of hazardous chemicals. Nat. Rev. Microbiol., 9(3), 177-192.

• Harris, R.J., Pilditch, C.A., Greenfield, B.L., Moon, V. and Kröncke, I., (2016). The influence of benthic macrofauna on the erodibility of intertidal sediments with varying mud content in three New Zealand estuaries. Estuaries and coasts, 39(3), pp.815-828.

• Hasanudin, U., Fujita, M., Kunihiro, T., Fujie, K., and Suzuki, T. (2004). The effect of clams (Tapes philippinarum) on changes in microbial community structure in tidal flat sediment mesocosms, based on quinone profiles. Ecol. Eng., 22(3), 185-196.

• He, Q., Bertness, M.D. and Altieri, A.H., (2013). Global shifts towards positive species interactions with increasing environmental stress. Ecology letters, 16(5), pp.695-706.

• He, Y., Men, B., Yang, X., and Wang, D. (2015). Bioturbation/bioirrigation effect on thallium released from reservoir sediment by different organism types. Science of the Total Environment, 532, 617-624.

163

• Heilskov, A. C., and Holmer, M. (2001). Effects of benthic fauna on organic matter mineralization in fish-farm sediments: importance of size and abundance. ICES Journal of Marine Science, 58(2), 427-434.

• Heiri, O., Lotter, A. F., & Lemcke, G. (2001). Loss on ignition as a method for estimating organic and carbonate content in sediments: reproducibility and comparability of results. Journal of paleolimnology, 25(1), 101-110.

• Heisterkamp, I.M., Schramm, A., Larsen, L.H., Svenningsen, N.B., Lavik, G., de Beer, D. and Stief, P., (2013). Shell biofilm‐associated nitrous oxide production in marine molluscs: processes, precursors and relative importance. Environmental microbiology, 15(7), pp.1943-1955.

• Herbert, R. A. (1999). Nitrogen cycling in coastal marine ecosystems. FEMS microbiology reviews, 23(5), 563-590.

• Herringshaw, L. G., and Solan, M. (2008). Benthic bioturbation in the past, present and future. Aquatic Biology, 2(3), 201-205.

• Hewitt, J.E. and Thrush, S.F., (2019). Monitoring for tipping points in the marine environment. Journal of environmental management, 234, pp.131-137.

• Hicks, N., Liu, X., Gregory, R., Kenny, J., Lucaci, A., Lenzi, L., Paterson, D.M. and Duncan, K.R., (2018). Temperature driven changes in benthic microbial diversity influences biogeochemical cycling in coastal sediments. Frontiers in microbiology, 9, p.1730.

• Hietanen, S., Laine, A.O. and Lukkari, K., (2007). The complex effects of the invasive polychaetes Marenzelleria spp. on benthic nutrient dynamics. Journal of Experimental Marine Biology and Ecology, 352(1), pp.89-102.

164

• Highton, M.P., Roosa, S., Crawshaw, J., Schallenberg, M. and Morales, S.E., (2016). Physical factors correlate to microbial community structure and nitrogen cycling gene abundance in a nitrate fed eutrophic lagoon. Frontiers in microbiology, 7, p.1691.

• Hill, N. A., Johnston, E. L., King, C. K., and Simpson, S. L. (2011). Physico-chemical changes in metal-spiked sediments deployed in the field: Implications for the interpretation of in situ studies. Chemosphere, 83(4), 400-408.

• Hoang, T.K., Probst, A., Orange, D., Gilbert, F., Elger, A., Kallerhoff, J., Laurent, F., Bassil, S., Duong, T.T. and Gerino, M., (2018). Bioturbation effects on bioaccumulation of cadmium in the wetland plant Typha latifolia: A nature-based experiment. Sci. Total Environ., 618, pp.1284- 1297.

• Hodson, P. V. (1988). The effect of metal metabolism on uptake, disposition and toxicity in fish. Aquat. Toxicol., 11(1-2), 3-18.

• Holmer, M., and Heilskov, A. C. (2008). Distribution and bioturbation effects of the tropical alpheid shrimp Alpheus macellarius in sediments impacted by milkfish farming. Estuarine, Coastal and Shelf Science, 76(3), 657-667.

• Holmer, M., Thorsen, S. W., Carlsson, M. S., and Kjerulf, P. J. (2015). Pelagic and benthic nutrient regeneration processes in mussel cultures (Mytilus edulis) in a eutrophic coastal area (Skive Fjord, Denmark). Estuaries Coasts, 38(5), 1629-1641.

• Howarth, R., Chan, F., Conley, D.J., Garnier, J., Doney, S.C., Marino, R. and Billen, G., (2011). Coupled biogeochemical cycles: eutrophication and hypoxia in temperate estuaries and coastal marine ecosystems. Frontiers in Ecology and the Environment, 9(1), pp.18-26.

165

• Huang, D., Todd, P.A. and Guest, J.R., (2007). Movement and aggregation in the fluted giant clam (Tridacna squamosa). Journal of Experimental Marine Biology and Ecology, 342(2), pp.269-281.

• Huang, W., Chen, X., Jiang, X., & Zheng, B. (2017). Characterization of sediment bacterial communities in plain lakes with different trophic statuses. MicrobiologyOpen, 6(5), e00503

• Huang, Y., Huang, J., Ervinia, A., Duan, S., and Kaushal, S. S. (2018). Land use and climate variability amplifies watershed nitrogen exports in coastal China. Ocean and Coastal Management.

• Hughes, D. J., Atkinson, R. J. A., & Ansell, A. D. (2000). A field test of the effects of megafaunal burrows on benthic chamber measurements of sediment-water solute fluxes. Marine Ecology Progress Series, 195, 189- 199.

• Isaev, A.V., Eremina, T.R., Ryabchenko, V.A. and Savchuk, O.P., (2017). Model estimates of the impact of bioirrigation activity of Marenzelleria spp. on the Gulf of Finland ecosystem in a changing climate. Journal of Marine Systems, 171, pp.81-88.

• Ito, K., Nozaki, M., Kunihiro, T., MIURA, C., & MIURA, T. (2011). Study of sediment cleanup using polychaetes. Interdisciplinary Studies on Environmental Chemistry—Marine Environmental Modeling & Analysis, Eds., K. Omori X, Guo N, Yoshie N, Fujii IC, Handoh A, Isobe and S. Tanabe, 133-139.

• Iwamoto, T., and Nasu, M. (2001). Current bioremediation practice and perspective. J. Biosci. Bioeng., 92(1), 1-8.

• Jackson, C. R., and Weeks, A. Q. (2008). Influence of particle size on bacterial community structure in aquatic sediments as revealed by 16S rRNA gene sequence analysis. Appl. Environ. Microbiol., 74(16), 5237- 5240

166

• Johnston, E. L., and Keough, M. J. (2005). Reduction of pollution impacts through the control of toxicant release rate must be site-and season-specific. J. of Exp. Mar. Biol. and Ecol., 320(1), 9-33.

• Johnston, E. L., and Roberts, D. A. (2009). Contaminants reduce the richness and evenness of marine communities: a review and meta- analysis. Environ. Pollut., 157(6), 1745-1752.

• Johnston, E. L., Mayer‐Pinto, M., and Crowe, T. P. (2015). Chemical contaminant effects on marine ecosystem functioning. J. App. Ecol., 52(1), 140-149.

• Jones, C. G., and Lawton, J. H. (2012). Linking species and ecosystems. Springer Science and Business Media.

• Jordan, M. A., Welsh, D. T., Dunn, R. J., and Teasdale, P. R. (2009). Influence of Trypaea australiensis population density on benthic metabolism and nitrogen dynamics in sandy estuarine sediment: a mesocosm simulation. J. Sea Res., 61(3), 144-152.

• Kadouche, D., Ducatez, M., Cenci, U., Tirtiaux, C., Suzuki, E., Nakamura, Y., Putaux, J.L., Terrasson, A.D., Diaz-Troya, S., Florencio, F.J. and Arias, M.C., (2016). Characterization of Function of the GlgA2 Glycogen/Starch Synthase in Cyanobacterium sp. Clg1 Highlights Convergent Evolution of Glycogen Metabolism into Starch Granule Aggregation. Plant physiology, 171(3), pp.1879-1892.

• Kang, Y., Zhang, J., Xie, H., Guo, Z., Ngo, H. H., Guo, W., and Liang, S. (2017). Enhanced nutrient removal and mechanisms study in benthic fauna added surface-flow constructed wetlands: The role of Tubifex tubifex. Biores. Technol., 224, 157-165.

• Kang, Y., Xie, H., Li, B., Zhang, J., Ngo, H.H., Guo, W., Guo, Z., Kong, Q., Liang, S., Liu, J. and Cheng, T., (2019). Performance of constructed wetlands and associated mechanisms of PAHs removal with mussels. Chem. Eng. J., 357, pp.280-287.

167

• Karlson, K., Hulth, S., and Rosenberg, R. (2007). Density of Monoporeia affinis and biogeochemistry in Baltic Sea sediments. Journal of Experimental Marine Biology and Ecology, 344(2), 123-135.

• Kerou, M. and Schleper, C., (2015). Nitrososphaeraceae. Bergey's Manual of Systematics of Archaea and Bacteria, pp.1-2.

• Kersters, K., De Vos, P., Gillis, M., Swings, J., Vandamme, P., & Stackebrandt, E. (2006). Introduction to the Proteobacteria. In The Prokaryotes: a handbook on the biology of bacteria (Vol. 5, pp. 3-37). Springer.

• Khan, F. I., Husain, T., and Hejazi, R. (2004). An overview and analysis of site remediation technologies. J. Environ. Manag., 71(2), 95-122.

• Kim, J.H., Kim, K.Y., Hahm, Y.T., Kim, B.S., Chun, J. and Cha, C.J., (2008). Actibacter sediminis gen. nov., sp. nov., a marine bacterium of the family Flavobacteriaceae isolated from tidal flat sediment. International journal of systematic and evolutionary microbiology, 58(1), pp.139-143.

• King, A. J., Tonkin, Z., and Lieshcke, J. (2012). Short-term effects of a prolonged blackwater event on aquatic fauna in the Murray River, Australia: considerations for future events. Marine and Freshwater Research, 63(7), 576-586.

• Klerks, P. L., Felder, D. L., Strasser, K., and Swarzenski, P. W. (2007). Effects of ghost shrimp on zinc and cadmium in sediments from Tampa Bay, FL. Mar. Chem., 104(1-2), 17-26.

• Koblížek, M., (2015). Ecology of aerobic anoxygenic phototrophs in aquatic environments. FEMS Microbiology Reviews, 39(6), pp.854-870.

• Kogure, K., and Wada, M. (2005). Impacts of macrobenthic bioturbation in marine sediment on bacterial metabolic activity. Microbes and Environments, 20(4), 191-199.

168

• Kristensen, E. and Blackburn, T.H., (1987). The fate of organic carbon and nitrogen in experimental marine sediment systems: influence of bioturbation and anoxia. Journal of Marine Research, 45(1), pp.231-257.

• Kristensen, E., Penha-Lopes, G., Delefosse, M., Valdemarsen, T., Quintana, C.O. and Banta, G.T., (2012). What is bioturbation? The need for a precise definition for fauna in aquatic sciences. Marine Ecology Progress Series, 446, pp.285-302.

• Krzmarzick, M. J., Taylor, D. K., Fu, X., and McCutchan, A. L. (2018). Diversity and niche of archaea in bioremediation. Archaea, 2018.

• Kuever, J., (2014). The family Desulfobacteraceae. The prokaryotes: Deltaproteobacteria and epsilonproteobacteria, pp.45-73.

• Kunihiro, T., Takasu, H., Miyazaki, T., Uramoto, Y., Kinoshita, K., Yodnarasri, S., Hama, D., Wada, M., Kogure, K., Ohwada, K. and Tsutsumi, H., (2011). Increase in Alphaproteobacteria in association with a polychaete, Capitella sp. I, in the organically enriched sediment. The ISME Journal, 5(11),

• Kuzyakov, Y., and Blagodatskaya, E. (2015). Microbial hotspots and hot moments in soil: concept and review. Soil Biology and Biochemistry, 83, 184-199.

• Lalonde, S. V., Dafoe, L. T., Pemberton, S. G., Gingras, M. K., and Konhauser, K. O. (2010). Investigating the geochemical impact of burrowing animals: Proton and cadmium adsorption onto the mucus lining of Terebellid polychaete worms. Chem. Geol., 271(1-2), 44-51.

• Laverock, B., Smith, C.J., Tait, K., Osborn, A.M., Widdicombe, S. and Gilbert, J.A., (2010). Bioturbating shrimp alter the structure and diversity of bacterial communities in coastal marine sediments. The ISME journal, 4(12), p.1531.

169

• Laverock, B., Gilbert, J. A., Tait, K., Osborn, A. M., and Widdicombe, S. (2011). Bioturbation: impact on the marine nitrogen cycle.

• Lavrentyev, P.J., Gardner, W.S. and Yang, L., (2000). Effects of the zebra mussel on nitrogen dynamics and the microbial community at the sediment-water interface. Aquatic Microbial Ecology, 21(2), pp.187-194.

• Lawes, J. C., Dafforn, K. A., Clark, G. F., Brown, M. V., and Johnston, E. L. (2017). Multiple stressors in sediments impact adjacent hard substrate habitats and across biological domains. Sci. Total Environ., 592, 295-305.

• Lelieveld, S. D., Pilditch, C. A., & Green, M. O. (2003). Variation in sediment stability and relation to indicators of microbial abundance in the Okura Estuary, New Zealand. Estuarine, Coastal and Shelf Science, 57(1-2), 123-136.

• Length, R. (2019). emmeans: Estimated Marginal Means, aka Least- Squares Means. R package version 1.4. https://CRAN.R- project.org/package=emmeans.

• Lenk, S., Moraru, C., Hahnke, S., Arnds, J., Richter, M., Kube, M., Reinhardt, R., Brinkhoff, T., Harder, J., Amann, R. and Mußmann, M., (2012). Roseobacter clade bacteria are abundant in coastal sediments and encode a novel combination of sulfur oxidation genes. The ISME journal, 6(12), p.2178.

• Lens, P. N. L., Pol, L. H., and Tichy, R. (2000). Sulfur cycle. In Encyclopedia of Microbiology 4 (pp. 495-505).

• Levin, L.A., Boesch, D.F., Covich, A., Dahm, C., ErsÚus, C., Ewel, K.C., Kneib, R.T., Moldenke, A., Palmer, M.A., Snelgrove, P. and Strayer, D., (2001). The function of marine critical transition zones and the importance of sediment biodiversity. Ecosystems, 4(5), pp.430-451.

170

• Levinton, J., & Kelaher, B. (2004). Opposing organizing forces of deposit- feeding marine communities. Journal of Experimental Marine Biology and Ecology, 300(1-2), 65-82.

• Levinton, J. S., & McCartney, M. (1991). Use of photosynthetic pigments in sediments as a tracer for sources and fates of macrophyte organic matter. Mar. Ecol. Prog. Ser, 78, 87-96.

• Li, M., Lundquist, C. J., Pilditch, C. A., Rees, T. A. V., and Ellis, J. (2019). Implications of nutrient enrichment for the conservation and management of seagrass Zostera muelleri meadows. Aquatic Conservation: Marine and Freshwater Ecosystems, 29(9), 1484-1502.

• Li, J., Hu, R., Guo, Y., Chen, S., Xie, X., Qin, J. G., and Pei, S. (2019). Bioturbation of peanut worms Sipunculus nudus on the composition of prokaryotic communities in a tidal flat as revealed by 16S rRNA gene sequences. MicrobiologyOpen, e802.

• Lindqvist, S., Norling, K., and Hulth, S. (2009). Biogeochemistry in highly reduced mussel farm sediments during macrofaunal recolonization by Amphiura filiformis and Nephtys sp. Mar. Environ. Res., 67(3), 136-145

• Lloyd, H.B., Murray, B.R. and Gribben, P.E., (2012). Trait and abundance patterns in two marine molluscs: the influence of abiotic conditions operating across multiple spatial scales. Marine Ecology Progress Series, 463, pp.205-214.

• Logue, J.B., Robinson, C.T., Meier, C. and der Meer, J.R.V., (2004). Relationship between sediment organic matter, bacteria composition, and the ecosystem metabolism of alpine streams. Limnology and Oceanography, 49(6), pp.2001-2010.

171

• Lohrer, A. M., Halliday, N. J., Thrush, S. F., Hewitt, J. E., and Rodil, I. F. (2010). Ecosystem functioning in a disturbance-recovery context: contribution of macrofauna to primary production and nutrient release on intertidal sandflats. Journal of Experimental Marine Biology and Ecology, 390(1), 6-13.

• Lu, D.C., Zhao, J.X., Wang, F.Q., Xie, Z.H., Du, Z.J. and Du, Z.J., (2017). Labilibacter aurantiacus gen. nov., sp nov., isolated from sea squirt (Styela clava) and reclassification of Saccharicrinis marinus as Labilibacter marinus comb. nov.

• Lucena, T., Ruvira, M.A., Garay, E., Macián, M.C., Arahal, D.R. and Pujalte, M.J., (2012). Actibacterium mucosum gen. nov., sp. nov., a marine alphaproteobacterium from Mediterranean seawater. International journal of systematic and evolutionary microbiology, 62(12), pp.2858-2864.

• Ludwig, M. and Bryant, D.A., (2012). Synechococcus sp. strain PCC 7002 transcriptome: acclimation to temperature, salinity, oxidative stress, and mixotrophic growth conditions. Frontiers in microbiology, 3, p.354.

• Lukwambe, B., Yang, W., Zheng, Y., Nicholaus, R., Zhu, J. and Zheng, Z., (2018). Bioturbation by the razor clam (Sinonovacula constricta) on the microbial community and enzymatic activities in the sediment of an ecological aquaculture wastewater treatment system. Sci. Total Environ., 643, pp.1098-1107.

• Lukwambe, B., Nicholaus, R., Yang, W. and Zheng, Z., (2019). Blood clams (Tegillarca granosa) bioturbation alter succession of bacterioplankton community and nutrient removal performance in an aquaculture wastewater bioremediation system. Aquaculture, p.734520.

• Ma, Y., Hu, A., Yu, C. P., Yan, Q., Yan, X., Wang, Y., and Xiong, H. (2015). Response of microbial communities to bioturbation by artificially introducing macrobenthos to mudflat sediments for in situ

172

bioremediation in a typical semi-enclosed bay, southeast China. Mar. Pollut. Bull., 94(1-2), 114-122.

• Magnusson, K., Agrenius, S., & Ekelund, R. (2003). Distribution of a tetrabrominated diphenyl ether and its metabolites in soft-bottom sediment and macrofauna species. Marine Ecology Progress Series, 255, 155-170.

• Maintinguer, S.I., Sakamoto, I.K., Adorno, M.A.T. and Varesche, M.B.A., (2016). Diversity of anaerobic bacteria in sediments from a subtropical reservoir. Lakes and Reservoirs: Research and Management, 21(4), pp.351-361.

• Maire, O., Lecroart, P., Meysman, F., Rosenberg, R., Duchêne, J.C. and Grémare, A., (2008). Quantification of sediment reworking rates in bioturbation research: a review. Aquatic Biology, 2(3), pp.219-238.

• Malkinson, D. and Tielbörger, K., (2010). What does the stress‐gradient hypothesis predict? Resolving the discrepancies. Oikos, 119(10), pp.1546- 1552.

• Mandario, M.A.E., Alava, V.R. and Añasco, N.C., (2019). Evaluation of the bioremediation potential of mud polychaete Marphysa sp. in aquaculture pond sediments. Environmental Science and Pollution Research, 26(29), pp.29810-29821.

• Mann, K. H. (1988). Production and use of detritus in various freshwater, estuarine, and coastal marine ecosystems. Limnology and Oceanography, 33(4part2), 910-930.

• Marsden, I.D., Shumway, S.E. and Padilla, D.K., (2012). Does size matter? The effects of body size and declining oxygen tension on oxygen uptake in gastropods. Journal of the Marine Biological Association of the United Kingdom, 92(7), pp.1603-1617.

173

• Martinez-Garcia, E., Carlsson, M. S., Sanchez-Jerez, P., Sánchez-Lizaso, J. L., Sanz-Lazaro, C., and Holmer, M. (2015). Effect of sediment grain size and bioturbation on decomposition of organic matter from aquaculture. Biogeochemistry, 125(1), 133-148.

• Maximov, A., Bonsdorff, E., Eremina, T., Kauppi, L., Norkko, A., and Norkko, J. (2015). Context-dependent consequences of Marenzelleria sp. (Spionidae: Polychaeta) invasion for nutrient cycling in the Northern Baltic Sea. Oceanologia, 57(4), 342-348.

• Mayor, D. J., Gray, N. B., Elver-Evans, J., Midwood, A. J., and Thornton, B. (2013). Metal-macrofauna interactions determine microbial community structure and function in copper contaminated sediments. PLoS One, 8(5), e64940.

• Mchenga, I.S., Mfilinge, P.L. and Tsuchiya, M., (2007). Bioturbation activity by the grapsid crab Helice formosensis and its effects on mangrove sedimentary organic matter. Estuarine, Coastal and Shelf Science, 73(1-2), pp.316-324.

• McKinnon, J.G., Gribben, P.E., Davis, A.R., Jolley, D.F. and Wright, J.T., (2009). Differences in soft sediment macrobenthic assemblages invaded by Caulerpa taxifolia compared to uninvaded habitats. Marine Ecology Progress Series, 380, pp.59-71.

• McLusky, D.S. and Elliott, M., (2004). The estuarine ecosystem: ecology, threats and management. OUP Oxford, pp. 1-18.

• Meksumpun, C., and Meksumpun, S. (1999). Polychaete–sediment relations in Rayong, Thailand. Environmental Pollution, 105(3), 447- 456.

• Mermillod‐Blondin, F., Gaudet, J.P., Gerino, M., Desrosiers, G., Jose, J. and Chatelliers, M.C.D., (2004). Relative influence of bioturbation and predation on organic matter processing in river sediments: a microcosm experiment. Freshwater Biology, 49(7), pp.895-912.

174

• Mermillod-Blondin, F., Nogaro, G., Datry, T., Malard, F., and Gibert, J. (2005). Do tubificid worms influence the fate of organic matter and pollutants in stormwater sediments? Environ. Pollut., 134(1), 57-69.

• Mermillod-Blondin, F., Nogaro, G., Vallier, F., and Gibert, J. (2008). Laboratory study highlights the key influences of stormwater sediment thickness and bioturbation by tubificid worms on dynamics of nutrients and pollutants in stormwater retention systems. Chemosphere, 72(2), 213-223.

• Mestre, M., Borrull, E., Sala, M.M. and Gasol, J.M., (2017). Patterns of bacterial diversity in the marine planktonic particulate matter continuum. The ISME journal, 11(4), p.999.

• Meuser, H. (2012). Soil remediation and rehabilitation: treatment of contaminated and disturbed land, 23, New York: Springer Science and Business Media.

• Michaud, E., Desrosiers, G., Mermillod-Blondin, F., Sundby, B., and Stora, G. (2005). The functional group approach to bioturbation: the effects of biodiffusers and gallery-diffusers of the Macoma balthica community on sediment oxygen uptake. J. of Exp. Mar. Biol. Ecol., 326(1), 77-88.

• Michaud, E., Desrosiers, G., Mermillod-Blondin, F., Sundby, B., and Stora, G. (2006). The functional group approach to bioturbation: II. The effects of the Macoma balthica community on fluxes of nutrients and dissolved organic carbon across the sediment–water interface. J. Exp. Mar. Biol. Ecol., 337(2), 178-189.

• Mishra, A. K., Tiwari, D. N., & Rai, A. N. (Eds.). (2018). Cyanobacteria: From Basic Science to Applications. Academic Press.

175

• Mohit, V., Archambault, P., & Lovejoy, C. (2015). Resilience and adjustments of surface sediment bacterial communities in an enclosed shallow coastal lagoon, Magdalen Islands, Gulf of St. Lawrence, Canada. FEMS microbiology ecology, 91(5).

• Moldenke, A., Pajutee, M., and Ingham, E. (2000). The functional roles of forest soil arthropods: the soil is a lively place. In Proceedings of the California Forest Soils Council Conference on Forest Soils Biology and Forest Management. USDA Forest Service, Pacific Southwest Research Station, Berkeley, California, pp. 7-22.

• Moller, M.M., Nielsen, L.P. and Jørgensen, B.B., (1985). Oxygen responses and mat formation by Beggiatoa spp. Applied and Environmental Microbiology, 50(2), pp.373-382.

• Mosier, A.C. and Francis, C.A., (2008). Relative abundance and diversity of ammonia‐oxidizing archaea and bacteria in the San Francisco Bay estuary. Environmental microbiology, 10(11), pp.3002-3016.

• Moulton, O. M., Altabet, M. A., Beman, J. M., Deegan, L. A., Lloret, J., Lyons, M. K., and Pfister, C. A. (2016). Microbial associations with macrobiota in coastal ecosystems: patterns and implications for nitrogen cycling. Frontiers in Ecology and the Environment, 14(4), 200-208.

• Murray, F., Solan, M., and Douglas, A. (2017). Effects of algal enrichment and salinity on sediment particle reworking activity and associated nutrient generation mediated by the intertidal polychaete Hediste diversicolor. Journal of Experimental Marine Biology and Ecology, 495, 75-82.

• Natálio, L. F., Pardo, J. C., Machado, G. B., Fortuna, M. D., Gallo, D. G., and Costa, T. M. (2017). Potential effect of fiddler crabs on organic matter distribution: A combined laboratory and field experimental approach. Estuar. Coast. Shelf S., 184, 158-165.

176

• Nelson, T.C. and Allison, J.B., (1940). On the nature and action of diantlin; a new hormone-like substance carried by the spermatozoa of the oyster. Journal of Experimental Zoology, 85, pp.299-338.

• Nicholaus, R., and Zheng, Z. (2014). The effects of bioturbation by the Venus clam Cyclina sinensis on the fluxes of nutrients across the sediment–water interface in aquaculture ponds. Aquacult. Int, 22(2), 913-924.

• Nicholaus, R., Lukwambe, B., Zhao, L., Yang, W., Zhu, J. and Zheng, Z., (2019). Bioturbation of blood clam Tegillarca granosa on benthic nutrient fluxes and microbial community in an aquaculture wastewater treatment system. International Biodeterioration and Biodegradation, 142, pp.73-82.

• Nobbs, D. M., Barford, J. P., Carroll, B. I., and Peters, G. M. (1997). An integrated investigation of anthropogenic selenium contamination in Lake Macquarie, NSW. Pure Appl. Chem., 69(11), 2387-2402

• Noble, D. W., Lagisz, M., O'dea, R. E., and Nakagawa, S. (2017). Nonindependence and sensitivity analyses in ecological and evolutionary meta‐analyses. Mol. Ecol., 26(9), 2410-2425.

• Nogaro, G., Mermillod-Blondin, F., Montuelle, B., Boisson, J. C., Lafont, M., Volat, B., and Gibert, J. (2007). Do tubificid worms influence organic matter processing and fate of pollutants in stormwater sediments deposited at the surface of infiltration systems? Chemosphere, 70(2), 315-328.

• Nogaro, G. and Steinman, A.D., (2014). Influence of ecosystem engineers on ecosystem processes is mediated by lake sediment properties. Oikos, 123(4), pp.500-512.

• Nogaro, G., and Burgin, A. J. (2014). Influence of bioturbation on denitrification and dissimilatory nitrate reduction to ammonium (DNRA) in freshwater sediments. Biogeochemistry, 120(1-3), 279-294.

177

• Norkko, J., Reed, D. C., Timmermann, K., Norkko, A., Gustafsson, B. G., Bonsdorff, E., ... and Conley, D. J. (2012). A welcome can of worms? Hypoxia mitigation by an invasive species. Glob. Change Biol., 18(2), 422-434.

• Norkko, A., Villnäs, A., Norkko, J., Valanko, S. and Pilditch, C., (2013). Size matters: implications of the loss of large individuals for ecosystem function. Scientific Reports, 3, p.2646.

• Norkko, J., Pilditch, C. A., Gammal, J., Rosenberg, R., Enemar, A., Magnusson, M., ... and Norkko, A. (2019). Ecosystem functioning along gradients of increasing hypoxia and changing soft-sediment community types. Journal of Sea Research, 153, 101781.

• Norling, K., Rosenberg, R., Hulth, S., Grémare, A. and Bonsdorff, E., (2007). Importance of functional biodiversity and species-specific traits of benthic fauna for ecosystem functions in marine sediment. Marine Ecology Progress Series, 332, pp.11-23.

• Nowicki, B. L. (1994). The effect of temperature, oxygen, salinity, and nutrient enrichment on estuarine denitrification rates measured with a modified nitrogen gas flux technique. Estuar. Coast. Shelf S., 38(2), 137- 156.

• Oksanen, J., Blanchet, F.G., Friendly, M., Kindt, R., Legendre, P., McGlinn, D., Minchin, P. R., O’hara, R.B., Simpson, G.L., Solymos, P., Stevens, M.H.H., Szoecs, E. and Wagner, H., (2019). Vegan: Community ecology package. R package version 2.5-51.

• Osinga, R., Kop, A. J., Malschaert, J. F., and Van Duyl, F. C. (1997). Effects of the sea urchin Echinocardium cordatum on bacterial production and carbon flow in experimental benthic systems under increasing organic loading. Journal of Sea Research, 37(1-2), 109-121.

178

• Ouellette, D., Desrosiers, G., Gagne, J.P., Gilbert, F., Poggiale, J.C., Blier, P.U. and Stora, G., (2004). Effects of temperature on in vitro sediment reworking processes by a gallery biodiffusor, the polychaete Neanthes virens. Marine Ecology Progress Series, 266, pp.185-193.

• Paliy, O., & Shankar, V. (2016). Application of multivariate statistical techniques in microbial ecology. Molecular ecology, 25(5), 1032-1057.

• Papaspyrou, S., Gregersen, T., Kristensen, E., Christensen, B. and Cox, R.P., (2006). Microbial reaction rates and bacterial communities in sediment surrounding burrows of two nereidid polychaetes (Nereisdiversicolor and N. virens). Marine Biology, 148(3), pp.541-550.

• Papaspyrou, S., Thessalou-Legaki, M. and Kristensen, E., (2010). The influence of infaunal (Nereis diversicolor) abundance on degradation of organic matter in sandy sediments. Journal of Experimental Marine Biology and Ecology, 393(1-2), pp.148-157.

• Park, B.J., Park, S.J., Yoon, D.N., Schouten, S., Damsté, J.S.S. and Rhee, S.K., (2010). Cultivation of autotrophic ammonia-oxidizing archaea from marine sediments in coculture with sulfur-oxidizing bacteria. Appl. Environ. Microbiol., 76(22), pp.7575-7587.

• Parker, J. G. (1983). A comparison of methods used for the measurement of organic matter in marine sediment. Chemistry in Ecology, 1(3), 201- 209.

• Pearson, T. H., & Rosenberg, R. (1978). Macrobenthic succession in relation to organic enrichment and pollution of the marine environment. Oceanogr. Mar. Biol. Ann. Rev, 16, 229-311.

• Pechenik, J.A., (1999). On the advantages and disadvantages of larval stages in benthic marine invertebrate life cycles. Marine Ecology Progress Series, 177, pp.269-297.

179

• Pedro, S., Duarte, B., Reis, G., Pereira, E., Duarte, A. C., Costa, J. L., and de Almeida, P. R. (2015). Metal partitioning and availability in estuarine surface sediments: changes promoted by feeding activity of Scrobicularia plana and Liza ramada. Estuar. Coast. Shelf S., 167, 240-247.

• Penha-Lopes, G., Bartolini, F., Limbu, S., Cannicci, S., Kristensen, E., and Paula, J. (2009). Are fiddler crabs potentially useful ecosystem engineers in mangrove wastewater wetlands? Mar. Pollut. Bull., 58(11), 1694-1703.

• Peñuelas, J., Sardans, J., Rivas‐ubach, A., and Janssens, I. A. (2012). The human‐induced imbalance between C, N and P in Earth's life system. Global Change Biology, 18(1), 3-6.

• Perelo, L. W. (2010). In situ and bioremediation of organic pollutants in aquatic sediments. Journal of hazardous materials, 177(1-3), 81-89.

• Petro, C., Starnawski, P., Schramm, A., and Kjeldsen, K. U. (2017). Microbial community assembly in marine sediments. Aquatic Microbial Ecology, 79(3), 177-195.

• Pfister, C.A. and Hay, M.E., (1988). Associational plant refuges: convergent patterns in marine and terrestrial communities result from differing mechanisms. Oecologia, 77(1), pp.118-129.

• Pita, L., Rix, L., Slaby, B. M., Franke, A., and Hentschel, U. (2018). The sponge holobiont in a changing ocean: from microbes to ecosystems. Microbiome, 6(1), 46.

• Plante, C. J., & Mayer, L. M. (1996). Seasonal variability in the bacteriolytic capacity of the deposit feeder Arenicola marina: environmental correlates. Aquatic Microbial Ecology, 11(2), 101-109.

• Post, D.M., Palkovacs, E.P., Schielke, E.G. and Dodson, S.I., (2008). Intraspecific variation in a predator affects community structure and cascading trophic interactions. Ecology, 89(7), pp.2019-2032.

180

• Poulsen, M., Kofoed, M. V., Larsen, L. H., Schramm, A., and Stief, P. (2014). Chironomus plumosus larvae increase fluxes of denitrification products and diversity of nitrate-reducing bacteria in freshwater sediment. Syst. App. Microbiol., 37(1), 51-59.

• Probandt, D., Knittel, K., Tegetmeyer, H.E., Ahmerkamp, S., Holtappels, M. and Amann, R., (2017). Permeability shapes bacterial communities in sublittoral surface sediments. Environmental Microbiology, 19(4), pp.1584-1599.

• Prodan, A., Tremaroli, V., Brolin, H., Zwinderman, A. H., Nieuwdorp, M., & Levin, E. (2020). Comparing bioinformatic pipelines for microbial 16S rRNA amplicon sequencing. Plos one, 15(1), e0227434.

• Prosser, J.I. and Nicol, G.W., (2008). Relative contributions of archaea and bacteria to aerobic ammonia oxidation in the environment. Environmental microbiology, 10(11), pp.2931-2941.

• Qin, X., Sun, H., Wang, C., Yu, Y., and Sun, T. (2010). Impacts of crab bioturbation on the fate of polycyclic aromatic hydrocarbons in sediment from the Beitang estuary of Tianjin, China. Environ. Tox. Chem., 29(6), 1248-1255.

• Quensen, J. (2019). QsRutils: R Functions Useful for Community Ecology. R package version 0.1.3.

• Queen, J. P., Quinn, G. P., & Keough, M. J. (2002). Experimental design and data analysis for biologists. Cambridge university press.

• Quintana, C.O., Kristensen, E. and Valdemarsen, T., (2013). Impact of the invasive polychaete Marenzelleria viridis on the biogeochemistry of sandy marine sediments. Biogeochemistry, 115(1-3), pp.95-109.

• Queen, J. P., Quinn, G. P., and Keough, M. J. (2002). Experimental design and data analysis for biologists. Cambridge university press.

181

• R Core Team. (2019). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.

• Raffard, A., Santoul, F., Cucherousset, J. and Blanchet, S., (2019). The community and ecosystem consequences of intraspecific diversity: a meta‐analysis. Biological Reviews, 94(2), pp.648-661.

• Ramaswamy, V. and Rao, P.S., (2006). Grain size analysis of sediments from the northern Andaman Sea: comparison of laser diffraction and sieve-pipette techniques. Journal of Coastal Research, pp.1000-1009.

• Rao, A. M., Malkin, S. Y., Montserrat, F., and Meysman, F. J. (2014). Alkalinity production in intertidal sands intensified by lugworm bioirrigation. Estuar., Coast. Shelf S., 148, 36-47.

• Ren, Y., Dong, S., Wang, F., Gao, Q., Tian, X., and Liu, F. (2010). Sedimentation and sediment characteristics in sea cucumber Apostichopus japonicus (Selenka) culture ponds. Aquacul. Res., 42(1), 14-21.

• Renz, J. R., and Forster, S. (2014). Effects of bioirrigation by the three sibling species of Marenzelleria spp. on solute fluxes and porewater nutrient profiles. Mar. Ecol. Prog. Ser., 505, 145-159.

• Riedel, B., Zuschin, M. and Stachowitsch, M., (2012). Tolerance of benthic macrofauna to hypoxia and anoxia in shallow coastal seas: a realistic scenario. Marine Ecology Progress Series, 458, pp.39-52.

• Riemann, L., Steward, G.F. and Azam, F., (2000). Dynamics of bacterial community composition and activity during a mesocosm diatom bloom. Appl. Environ. Microbiol., 66(2), pp.578-587.

• Robinson, G., Caldwell, G. S., Jones, C. L., Slater, M. J., and Stead, S. M. (2015). Redox stratification drives enhanced growth in a deposit-feeding invertebrate: implications for aquaculture bioremediation. Aquacul. Environ. Interact., 8, 1-13.

182

• Robinson, G., MacTavish, T., Savage, C., Caldwell, G.S., Jones, C.L., Probyn, T., Eyre, B.D. and Stead, S.M., (2018). Carbon amendment stimulates benthic nitrogen cycling during the bioremediation of particulate aquaculture waste. Biogeosciences, 15(6), pp.1863-1878.

• Rodríguez‐Echeverría, S., Lozano, Y. M., and Bardgett, R. D. (2016). Influence of soil microbiota in nurse plant systems. Functional ecology, 30(1), 30-40.

• Rohatgi, R. (2018). WebPlotDigitizer: HTML5 based online tool to extract numerical data from plot images. Version 4.1. [WWW document] URL https://automeris.io/WebPlotDigitizer (accessed between September- November 2017).

• Russell, L., 2018. Emmeans: estimated marginal means, aka least- squares means. R package version, 1(2).

• Rysgaard, S., Thastum, P., Dalsgaard, T., Christensen, P. B., and Sloth, N.

P. (1999). Effects of salinity on NH4+ adsorption capacity, nitrification, and denitrification in Danish estuarine sediments. Estuaries, 22(1), 21- 30.

• Salvucci, E. (2016). Microbiome, holobiont and the net of life. Critical reviews in microbiology, 42(3), 485-494.

• Samuiloviene, A., Bartoli, M., Bonaglia, S., Cardini, U., Vybernaite- Lubiene, I., Marzocchi, U., Petkuviene, J., Politi, T., Zaiko, A. and Zilius, M., (2019). The effect of chironomid larvae on nitrogen cycling and microbial communities in soft sediments. Water, 11(9), p.1931.

• Sander, B.C. and Kalff, J., (1993). Factors controlling bacterial production in marine and freshwater sediments. Microbial Ecology, 26(2), pp.79-99.

183

• Schenone, S., Meara, T.A. and Thrush, S.F., (2019). Non-linear effects of macrofauna functional trait interactions on biogeochemical fluxes in marine sediments change with environmental stress. Marine Ecology Progress Series, 624, pp.13-21.

• Schneider, C.A., Rasband, W.S. and Eliceiri, K.W., (2012). NIH Image to ImageJ: 25 years of image analysis. Nature Methods, 9(7), p.671.

• Sciberras, M., Tait, K., Brochain, G., Hiddink, J. G., Hale, R., Godbold, J. A., and Solan, M. (2017). Mediation of nitrogen by post-disturbance shelf communities experiencing organic matter enrichment. Biogeochemistry, 135(1-2), 135-153.

• Séguin, A., Harvey, É., Archambault, P., Nozais, C., & Gravel, D. (2014). Body size as a predictor of species loss effect on ecosystem functioning. Scientific reports, 4, 4616.

• Shang, J., Zhang, L., Shi, C. and Fan, C., (2013). Influence of Chironomid Larvae on oxygen and nitrogen fluxes across the sediment-water interface (Lake Taihu, China). Journal of Environmental Sciences, 25(5), pp.978-985.

• Sharpton, T. J. (2018). Role of the gut microbiome in vertebrate evolution. MSystems, 3(2), e00174-17.

• Shen, H., Thrush, S. F., Wan, X., Li, H., Qiao, Y., Jiang, G., and He, P. (2016). Optimization of hard clams, polychaetes, physical disturbance and denitrifying bacteria of removing nutrients in marine sediment. Mar. Pollut. Bull., 110(1), 86-92.

• Shen, H., Jiang, G., Wan, X., Li, H., Qiao, Y., Thrush, S., and He, P. (2017). Response of the microbial community to bioturbation by benthic macrofauna on intertidal flats. Journal of Exp. Mar. Biol. Ecol., 488, 44- 51.

184

• Sheng, P., Yu, Y., Zhang, G., Huang, J., He, L. and Ding, J., (2016). Bacterial diversity and distribution in seven different estuarine sediments of Poyang Lake, China. Environmental Earth Sciences, 75(6), p.479.

• Shishir, T. A., and Mahbub, N. (2019). Review on Bioremediation: A Tool to Resurrect the Polluted Rivers. Pollution, 5(3), 555-568.

• Shivani, Y., Subhash, Y., Sasikala, C. and Ramana, C.V., (2017). Halodesulfovibrio spirochaetisodalis gen. nov. sp. nov. and reclassification of four Desulfovibrio spp. International journal of systematic and evolutionary microbiology, 67(1), pp.87-93.

• Shuman, C. S., & Ambrose, R. F. (2003). A comparison of remote sensing and ground‐based methods for monitoring wetland restoration success. Restoration Ecology, 11(3), 325-333.

• Singer, J. K., Anderson, J. B., Ledbetter, M. T., McCave, I. N., Jones, K. P. N., and Wright, R. (1988). An assessment of analytical techniques for the size analysis of fine-grained sediments. Journal of Sedimentary Research, 58(3), 534-543.

• Sinha, E., Michalak, A. M., and Balaji, V. (2017). Eutrophication will increase during the 21st century as a result of precipitation changes. Science, 357(6349), 405-408.

• Sinkko, H., Lukkari, K., Sihvonen, L. M., Sivonen, K., Leivuori, M., Rantanen, M., Paulin L. and Lyra, C. (2013). Bacteria contribute to sediment nutrient release and reflect progressed eutrophication-driven hypoxia in an organic-rich continental sea. PloS one, 8(6), e67061.

• Skei, J., Larsson, P., Rosenberg, R., Jonsson, P., Olsson, M., and Broman, D. (2000). Eutrophication and contaminants in aquatic ecosystems. AMBIO, 29(4), 184-194.

185

• Small, K., Kopf, R. K., Watts, R. J., and Howitt, J. (2014). Hypoxia, blackwater and fish kills: experimental lethal oxygen thresholds in juvenile predatory lowland river fishes. PLoS One, 9(4), e94524.

• Solan, M., & Wigham, B. D. (2005). Biogenic particle reworking and bacterial-invertebrate interactions in marine sediments. Interactions between macro-and microorganisms in marine sediments. American Geophysical Union, Washington, DC, 105-124.

• Sorokin, D.Y., Tourova, T.P., Bezsoudnova, E.Y., Pol, A. and Muyzer, G., (2007). Denitrification in a binary culture and thiocyanate metabolism in Thiohalophilus thiocyanoxidans gen. nov. sp. nov.–a moderately halophilic chemolithoautotrophic sulfur-oxidizing Gammaproteobacterium from hypersaline lakes. Archives of microbiology, 187(6), pp.441-450.

• Stamhuis, E. J., Schreurs, C. E., and Videler, J. J. (1997). Burrow architecture and turbative activity of the thalassinid shrimp Callianassa subterranea from the central North Sea. Marine Ecology Progress Series, 151, 155-163.

• Steffen, W., Richardson, K., Rockström, J., Cornell, S. E., Fetzer, I., Bennett, E. M., ... and Folke, C. (2015). Planetary boundaries: Guiding human development on a changing planet. Science, 347(6223), 1259855.

• Stief, P., and de Beer, D. (2002). Bioturbation effects of Chironomus riparius on the benthic N-cycle as measured using microsensors and microbiological assays. Aquatic Microbial Ecology, 27(2), 175-185.

• Stief, P., (2013). Stimulation of microbial nitrogen cycling in aquatic ecosystems by benthic macrofauna: mechanisms and environmental implications. Biogeosciences, 10(12), pp.7829-7846.

• Strauss, E.A. and Lamberti, G.A., (2002). Effect of dissolved organic carbon quality on microbial decomposition and nitrification rates in stream sediments. Freshwater Biology, 47(1), pp.65-74.

186

• Strayer, D. L., Caraco, N. F., Cole, J. J., Findlay, S., & Pace, M. L. (1999). Transformation of freshwater ecosystems by bivalves: a case study of zebra mussels in the Hudson River. BioScience, 49(1), 19-27.

• Sun, J., Wang, M. H., and Ho, Y. S. (2012). A historical review and bibliometric analysis of research on estuary pollution. Mar. Pollut. Bull., 64(1), 13-21.

• Sun, M.Y., Dafforn, K.A., Johnston, E.L. and Brown, M.V., (2013). Core sediment bacteria drive community response to anthropogenic contamination over multiple environmental gradients. Environmental Microbiology, 15(9), pp.2517-2531.

• Sun, Y., Wang, T., Peng, X., Wang, P. and Lu, Y., (2016). Bacterial community compositions in sediment polluted by perfluoroalkyl acids (PFAAs) using Illumina high-throughput sequencing. Environmental Science and Pollution Research, 23(11), pp.10556-10565.

• Sundbäck, K., Miles, A., & Göransson, E. (2000). Nitrogen fluxes, denitrification and the role of microphytobenthos in microtidal shallow- water sediments: an annual study. Marine Ecology Progress Series, 200, 59-76.

• Sutton, M. A., Bleeker, A., Howard, C. M., Erisman, J. W., Abrol, Y. P., Bekunda, M., and Zhang, F. S. (2013). Our nutrient world. The challenge to produce more food and energy with less pollution. Centre for Ecology and Hydrology.

• Tallqvist, M., (2001). Burrowing behaviour of the Baltic clam Macoma balthica: effects of sediment type, hypoxia and predator presence. Marine Ecology Progress Series, 212, pp.183-191.

• Tang, X., Gao, G., Qin, B., Zhu, L., Chao, J., Wang, J. and Yang, G., (2009). Characterization of bacterial communities associated with organic aggregates in a large, shallow, eutrophic freshwater lake (Lake Taihu, China). Microbial ecology, 58(2), pp.307-322.

187

• Taylor, A.C. and Brand, A.R., (1975). Effects of hypoxia and body size on the oxygen consumption of the bivalve Arctica islandica (L.). Journal of Experimental Marine Biology and Ecology, 19(2), pp.187-196.

• Tchounwou, P. B., Yedjou, C. G., Patlolla, A. K., and Sutton, D. J. (2012). Heavy metal toxicity and the environment. In Luch, A. (Ed.) Molecular, clinical and environmental toxicology (pp. 133-164). Basel: Springer.

• Teal, L. R., Parker, E. R., and Solan, M. (2013). Coupling bioturbation activity to metal (Fe and Mn) profiles in situ. Biogeosciences, 2365-2378.

• Tettelbach, S.T., Europe, J.R., Tettelbach, C.R., Havelin, J., Rodgers, B.S., Furman, B.T. and Velasquez, M., (2017). Hard clam walking: Active horizontal locomotion of adult Mercenaria mercenaria at the sediment surface and behavioral suppression after extensive sampling. PloS one, 12(3), p.e0173626.

• Thompson, J. R., Rivera, H. E., Closek, C. J., and Medina, M. (2015). Microbes in the coral holobiont: partners through evolution, development, and ecological interactions. Frontiers in cellular and infection microbiology, 4, 176.

• Timmermann, K., Christensen, M., Banta, G. T., and Andersen, O. (2000). Bioavailability and metabolism of pyrene in sediments with the polychaete Arenicola marina. WIT T. Ecol. Environ., 86-94.

• Timmermann, K., Banta, G. T., Klinge, L., and Andersen, O. (2011). Effects of bioturbation on the fate of oil in coastal sandy sediments–An in situ experiment. Chemosphere, 82(10), 1358-1366.

• Uusitalo, L., Korpinen, S., Andersen, J. H., Niiranen, S., Valanko, S., Heiskanen, A. S., & Dickey-Collas, M. (2016). Exploring methods for predicting multiple pressures on ecosystem recovery: a case study on marine eutrophication and fisheries. Continental Shelf Research, 121, 48- 60.

188

• Vadillo Gonzalez, S., Johnston, E., Gribben, P.E. and Dafforn, K., (2019). The application of bioturbators for aquatic bioremediation: Review and meta-analysis. Environmental pollution. 250, pp. 426-436.

• Vandenkoornhuyse, P., Quaiser, A., Duhamel, M., Le Van, A., and Dufresne, A. (2015). The importance of the microbiome of the plant holobiont. New Phytologist, 206(4), 1196-1206.

• Van der Hel, S., Hellsten, I. and Steen, G., (2018). Tipping points and climate change: Metaphor between science and the media. Environmental Communication, 12(5), pp.605-620.

• Van Duyl, F.C., Kop, A.J., Kok, A. and Sandee, A.J.J., (1992). The impact of organic matter and macrozoobenthos on bacterial and oxygen variables in marine sediment boxcosms. Netherlands journal of sea research, 29(4), pp.343-355.

• Vaquer-Sunyer, R. and Duarte, C.M., (2008). Thresholds of hypoxia for marine biodiversity. Proceedings of the National Academy of Sciences, 105(40), pp.15452-15457.

• Vaughn, C.C. and Hakenkamp, C.C., (2001). The functional role of burrowing bivalves in freshwater ecosystems. Freshwater Biology, 46(11), pp.1431-1446.

• Venables, W. N., and Ripley, B. D. (2013). Modern applied statistics with S-PLUS. Springer Science & Business Media.

• Vidali, M. (2001). Bioremediation. an overview. Pure. App. Chem., 73(7), 1163-1172.

• Viechtauber, W. (2010). Conducting meta-analyses in R with the metafor package. J. Stat. Softw, 1-48.

189

• Vincent, B., Joly, D. and Harvey, M., (1994). Spatial variation in growth of the bivalve Macoma balthica (L.) on a tidal flat: effects of environmental factors and intraspecific competition. Journal of Experimental Marine Biology and Ecology, 181(2), pp.223-238.

• Volkenborn, N., Polerecky, L., Hedtkamp, S. I. C., van Beusekom, J. E., and De Beer, D. (2007). Bioturbation and bioirrigation extend the open exchange regions in permeable sediments. Limnology and Oceanography, 52(5), 1898-1909.

• Volkenborn, N., Meile, C., Polerecky, L., Pilditch, C. A., Norkko, A., Norkko, J., and Woodin, S. A. (2012). Intermittent bioirrigation and oxygen dynamics in permeable sediments: an experimental and modeling study of three tellinid bivalves. J. Mar. Res., 70(6), 794-823

• Wada, M., Urakawa, T. and Tamaki, A., (2016). Dynamics of bacterial community structure on intertidal sandflat inhabited by the ghost shrimp Nihonotrypaea harmandi (Decapoda: Axiidea: Callianassidae) in Tomioka Bay, Amakusa, Japan. Gene, 576(2), pp.657-666.

• Wall, S. B., La Point, T. W., and Isely, J. J. (1996). Fish bioturbation of cadmium‐contaminated sediments: Factors affecting Cd availability to Daphnia magna. Environ. Tox. Chem., 15(3), 294-298.

• Wang, J. Q., Zhang, X. D., Jiang, L. F., Bertness, M. D., Fang, C. M., Chen, J. K., and Li, B. (2010). Bioturbation of burrowing crabs promotes sediment turnover and carbon and nitrogen movements in an estuarine salt marsh. Ecosystems, 13(4), 586-599.

• Wang, Q., Li, Y. and Wang, Y., (2011). Optimizing the weight loss-on- ignition methodology to quantify organic and carbonate carbon of sediments from diverse sources. Environmental Monitoring and Assessment, 174(1-4), pp.241-257.

190

• Wang, L., Liu, L., Zheng, B., Zhu, Y., and Wang, X. (2013). Analysis of the bacterial community in the two typical intertidal sediments of Bohai Bay, China by pyrosequencing. Marine pollution bulletin, 72(1), 181-187.

• Wang, L., Zheng, B., Nan, B. and Hu, P., (2014). Diversity of bacterial community and detection of nirS-and nirK-encoding denitrifying bacteria in sandy intertidal sediments along Laizhou Bay of Bohai Sea, China. Marine pollution bulletin, 88(1-2), pp.215-223.

• Wang, H., Dai, M., Liu, J., Kao, S.J., Zhang, C., Cai, W.J., Wang, G., Qian, W., Zhao, M. and Sun, Z., (2016). Eutrophication-driven hypoxia in the East China Sea off the Changjiang Estuary. Environmental Science and Technology, 50(5), pp.2255-2263.

• Wang, L., Shao, X., Xu, M., and Chen, S. (2019). Bioremediation of nitrogen-and phosphorus-polluted aquaculture sediment by utilizing combined immobilized effective microorganisms and. International Journal of Agricultural and Biological Engineering, 12(6), 192-201.

• Wardle, D. A. (2006). The influence of biotic interactions on soil biodiversity. Ecology letters, 9(7), 870-886.

• Wasmund, K., Mußmann, M. and Loy, A., (2017). The life sulfuric: microbial ecology of sulfur cycling in marine sediments. Environmental microbiology reports, 9(4), pp.323-344.

• Weinberg, J.R., (1985). Factors regulating population dynamics of the marine bivalve Gemma gemma: intraspecific competition and salinity. Marine Biology, 86(2), pp.173-182.

• Weissberger, E.J., Coiro, L.L. and Davey, E.W., (2009). Effects of hypoxia on animal burrow construction and consequent effects on sediment redox profiles. Journal of Experimental Marine Biology and Ecology, 371(1), pp.60-67.

191

• Westrich, J.T. and Berner, R.A., (1984). The role of sedimentary organic matter in bacterial sulfate reduction: The G model tested 1. Limnology and oceanography, 29(2), pp.236-249.

• Wilde, S.B. and Plante, C.J., (2002). Spatial Heterogeneity of Bacterial Assemblages in Marine Sediments: The Influence of Deposit Feeding by Balanoglossus aurantiacus. Estuarine, Coastal and Shelf Science, 55(1), pp.97-107.

• Wilkinson, M. T., Richards, P. J., and Humphreys, G. S. (2009). Breaking ground: pedological, geological, and ecological implications of soil bioturbation. Earth-Science Reviews, 97(1-4), 257-272.

• Wohlgemuth, D., Filip, J., Hillebrand, H., & Moorthi, S. D. (2017). Prey diversity effects on ecosystem functioning depend on consumer identity and prey composition. Oecologia, 184(3), 653-661.

• Wrede, A., Beermann, J., Dannheim, J., Gutow, L. and Brey, T., (2018). Organism functional traits and ecosystem supporting services–A novel approach to predict bioirrigation. Ecological Indicators, 91, pp.737-743.

• Wright, J.T. and Gribben, P.E., (2008). Predicting the impact of an invasive seaweed on the fitness of native fauna. Journal of Applied Ecology, 45(5), pp.1540-1549.

• Wright, J.T., Byers, J.E., Koukoumaftsis, L.P., Ralph, P.J. and Gribben, P.E., (2010). Native species behaviour mitigates the impact of habitat- forming invasive seaweed. Oecologia, 163(2), pp.527-534.

• Yamamoto, N. and Lopez, G., (1985). Bacterial abundance in relation to surface area and organic content of marine sediments. Journal of Experimental Marine Biology and Ecology, 90(3), pp.209-220.

192

• Yan, W., Chi, J., Wang, Z., Huang, W. and Zhang, G., (2009). Spatial and temporal distribution of polycyclic aromatic hydrocarbons (PAHs) in sediments from Daya Bay, South China. Environmental Pollution, 157(6), pp.1823-1830

• Yan, Y.W., Jiang, Q.Y., Wang, J.G., Zhu, T., Zou, B., Qiu, Q.F. and Quan, Z.X., (2018). Microbial Communities and Diversities in Mudflat Sediments Analyzed Using a Modified Metatranscriptomic Method. Frontiers in microbiology, 9, p.93.

• Yang, Sung-Hyun, Hyun-Seok Seo, Jung-Hee Woo, Hyun-Myung Oh, Hani Jang, Jung-Hyun Lee, Sang-Jin Kim, and Kae Kyoung Kwon, (2014), Carboxylicivirga gen. nov. in the family Marinilabiliaceae with two novel species, Carboxylicivirga mesophila sp. nov. and Carboxylicivirga taeanensis sp. nov., and reclassification of Cytophaga fermentans as Saccharicrinis fermentans gen. nov., comb. nov. International journal of systematic and evolutionary microbiology 64(4), pp. 4 1351-1358.

• You, J., Das, A., Dolan, E.M. and Hu, Z., (2009). Ammonia-oxidizing archaea involved in nitrogen removal. Water research, 43(7), pp.1801- 1809.

• Zhang, X., Liu, Z., Jeppesen, E., and Taylor, W. D. (2014). Effects of deposit-feeding tubificid worms and filter-feeding bivalves on benthic- pelagic coupling: implications for the restoration of eutrophic shallow lakes. Water Res., 50, 135-146.

• Zhao, D., Cao, X., Zeng, J., Huang, R. and Yu, Z., (2017). Effects of macrobenthic bioturbation on the abundance and community composition of ammonia-oxidizing prokaryotes under different temperatures. Journal of Freshwater Ecology, 32(1), pp.405-414.

193

• Zhao, L., Zheng, Y., Nicholaus, R., Lukwambe, B., Zhu, J., Yang, W. and Zheng, Z., (2019). Bioturbation by the razor clam Sinonovacula constricta affects benthic nutrient fluxes in aquaculture wastewater treatment ecosystems. Aquacult. Env. Interact., 11, pp.87-96.

• Zheng, B., Wang, L., & Liu, L. (2014). Bacterial community structure and its regulating factors in the intertidal sediment along the Liaodong Bay of Bohai Sea, China. Microbiological research, 169(7-8), 585-592.

• Zhong, D., Wang, F., Dong, S., and Li, L. (2015). Impact of Litopenaeus vannamei bioturbation on nitrogen dynamics and benthic fluxes at the sediment–water interface in pond aquaculture. Aquacult. Int., 23(4), 967-980.

• Zhou, Z., Pan, J., Wang, F., Gu, J.D. and Li, M., (2018). Bathyarchaeota: globally distributed metabolic generalists in anoxic environments. FEMS microbiology reviews, 42(5), pp.639-655.

•

194

APPENDIX 1: APPLICATION OF BIOTURBATORS FOR AQUATIC BIOREMEDIATION: REVIEW AND META- A ANALYSIS

Fig. A.1. Percentage of the main contaminants found in peer reviewed scientific articles where contaminant-bioturbation interaction is analysed.

195

APPENDIX 1

Fig.A.2. Flow diagram of the specific search, selection and article exclusion for metal systematic analysis

196

APPENDIX 1

Fig. A.3. Flow diagram of the specific search, selection and article exclusion for organic matter/nutrients systematic analysis

197

APPENDIX 1

Fig. A.4. Flow diagram of the specific search, selection and article exclusion for Polycyclic Aromatic Hydrocarbons (PAH) systematic analysis

198

APPENDIX 1

Fig. A.5. Forest plot of the effect size (Standard Mean Difference, SMD) in the selected studies where the interaction between bioturbators and metal release to overlying water was evaluated. Black squares denote the SMD of each individual study and the whiskers the sampling variance per each study; these values are also noted on the right side of the figure. On the bottom right, the overall effect size is noted which is also marked by a red polygon at the centre of the plot. The dotted line indicates an effect size equal to 0.

199

APPENDIX 1

Fig. A.6. Forest plot of the effect size (Standard Mean Difference, SMD) in the selected studies where the interaction between bioturbators and ammonia fluxes (NH4+) in sediment to the overlying water was evaluated. Black squares denote the SMD of each individual study/treatment and the whiskers the sampling variance per each study, these values are also noted on the right side of the figure. On the bottom right, the overall effect size is noted which is also marked by a red polygon at the centre of the plot. The dotted line indicates an effect size equal to 0.

200

APPENDIX 1

Fig. A.7. Forest plot of the effect size (Standard Mean Difference, SMD) in the selected studies where the interaction between bioturbators and phosphorous fluxes in sediment to the overlying water was evaluated. Black squares denote the SMD of each individual study/treatment and the whiskers the sampling variance per each study, these values are also noted on the right side of the figure. On the bottom right, the overall effect size is noted which is also marked by a red polygon at the centre of the plot. The dotted line indicates an effect size equal to 0.

201

APPENDIX 1

Fig. A.8. Forest plot of the effect size (Standard Mean Difference, SMD) in the selected studies where the interaction between bioturbators and sediment oxygen uptake (SOU) was evaluated. Black squares denote the SMD of each individual study/treatment and the whiskers the sampling variance per each study, these values are also noted on the right side of the figure. On the bottom right, the overall effect size is noted which is also marked by a red polygon at the centre of the plot. The dotted line indicates an effect size equal to 0. Analysed oxygen fluxes where only selected as an additional variable found during the literature search for organic matter and nutrients: original search did not include bioirrigation or oxygen fluxes.

202

APPENDIX 1

Fig.A.9. Main countries found in the systematic review where the effect of bioturbation and metals, organic matter/nutrients and PAHs was evaluated.

203

APPENDIX 1

Table A.1. List of consulted literature and meta-data used for the systematic review and meta-analysis Authors year Journal Country System Class Type Contaminant Used for Aller 1994 Chemical Geology USA Pond Model ex situ Nutrients Systematic review Stream, Artificial Altmann et al. 2004 Microbial Ecology Germany Insecta ex situ Nutrients Systematic review brook Amato et al. 2016 Environmental Science and Technology Australia Coastal wetland Bivalva ex situ Metals Meta-analysis Andersen and Kristian 1992 Limnology and Oceanography Denmark Coastal lagoon Benthic assemblage ex situ Nutrients Systematic review Transactions on Ecology and the Andersen et al. 2001 Denmark Fjord Polychaeta ex situ PAH Systematic review Environment Archives of Environmental Contamination and Andres et al. 1998 France Lake Insecta ex situ Metals Systematic review Toxicology Angeler et al. 2001 Hydrobiologia Spain Wetland Malacostraca in situ Nutrients Systematic review Atkinson et al. 2007 Chemosphere Australia Lake Bivalva ex situ Metals Meta-analysis Banta et al. 1999 Aquatic Microbial Ecology Germany Marine Polychaeta ex situ Nutrients Systematic review Bartoli et al. 2001 Hydrobiologia Italy Coastal lagoon Bivalvia in situ Nutrients Systematic review Journal of Experimental Marine Biology and Bartoli et al. 2009 Belgium Marine Polychaeta ex situ NH3+, PO43- Meta-analysis Ecology Bartoli et al. 2000 Hydrobiologia USA Lake Polychaeta ex situ Nutrients Systematic review Bartolini et al. 2009 Marine Pollution Bulletin Tanzania Mangrove Malacostraca ex situ Nutrients Systematic review NH3+, PO43- Benelli et al 2017 Aquatic Ecology Italy River Bivalvia ex situ Meta-analysis , SOU Knowledge and Management of Aquatic Berezina et al. 2011 China Lake Benthic assemblage in situ Nutrients Systematic review Ecosystems Bergström et al. 2017 Aquaculture Research USA Estuary Polychaeta ex situ Nutrients Systematic review Bertics et al. 2010 Marine Ecology Progress Series USA Coastal lagoon Malacostraca in situ Nutrients Systematic review Malacostraca, Biles et al. 2002 Hydrology and Earth System Sciences Sweden Fjord ex situ Nutrients Systematic review Polychaeta, Bivalva Table A. 1. continued

204

APPENDIX 1

Authors year Journal Country System Class Type Contaminant Used for Polychaeta, Biles et al. 2002 Hydrology and Earth System Sciences UK Estuary Gastropoda, Bivalva, in situ Nutrients Systematic review Malacostraca Biswas et al. 2009 Ecological Engineering India Pond Insecta ex situ Nutrients Systematic review Blankson and Klerks 2017 Ecotoxicology USA Lake Oligochaeta ex situ Metals Systematic review Blankson et al. 2016 Environmental Toxicology and Chemistry USA Lake Oligochaeta ex situ Metals Meta-analysis Blankson et al. 2017 Chemosphere USA Lake Oligochaeta ex situ Metals Systematic review Bivalvia, Insecta, Boeker et al 2016 Limnologica Germany River ex situ Nutrients Systematic review Oligochaeta Bonaglia et al. 2013 Marine Ecology Progress Series Sweden Marine Polychaeta ex situ Nutrients Systematic review Bosch et al. 2015 Marine Ecology Progress Series Belgium Marine Polychaeta ex situ NH3+, SOU Meta-analysis Botto and Iribarne 2000 Estuarine, Coastal and Shelf Science Argentina Coastal lagoon Malacostraca in situ Nutrients Systematic review Botto et al. 2006 Marine Ecology Progress Series Argentina Coastal lagoon Malacostraca in situ Nutrients Systematic review Bivalva, Malacostracam Bradshaw et al. 2006 Estuarine, Coastal and Shelf Science Sweden Estuary ex situ Metals Systematic review Priapulida Braeckman et al. 2010 Marine Ecology Progress Series China Estuary Bivalvia in situ Nutrients Systematic review Journal of experimental marine biology and Chan et al. 2013 Canada Marine Bivalvia in situ Nutrients Systematic review ecology. Chandler et al. 2014 Environmental Science and Technology USA Lake, Dam Malacostraca ex situ Metals Systematic review Polychaeta, Chen and Mayer 1998 Environmental Science and Technology USA Dam, River ex situ Metals Systematic review Holothuroidea Polychaeta, Chen and Mayer 1999 Marine Ecology Progress Series USA River ex situ Metals Systematic review Holothuroidea Chen et al. 2016 Science of Total Environment China Lake Oligochaeta, Bivalva ex situ Nutrients Systematic review Christensen et al. 2002 Oil and Hydrocarbon Spills III Denmark Fjord Polychaeta ex situ PAH Systematic review

205

APPENDIX 1

Table A.1. continued Authors year Journal Country System Class Type Contaminant Used for The Ciarelli et al. 1999 Environmental Toxicology and Chemistry Estaury Malacostraca ex situ PAH Systematic review Netherlands The Ciarelli et al. 2000 Environmental Toxicology and Chemistry Estuary Polychaeta ex situ PAH Systematic review Netherlands Ciutat and Boudou 2003 Environmental Toxicology and Chemistry France Estuary, River Bivalva, Insecta ex situ Metals Meta-analysis Ciutat et al. 2006 Marine Ecology Progress Series UK Estaury Bivalva ex situ PAH Systematic review River, Coastal Ciutat et al. 2005 Ecotoxicology and Environmental Safety France Oligochaeta ex situ Metals Meta-analysis lagoon Estuary, Sheltered Ciutat et al. 2005 Environmental Toxicology and Chemistry France Oligochaeta ex situ Metals Systematic review loch Ciutat et al. 2007 Ecotoxicology and Environmental Safety France Estuary, River Bivalva, Oligochaeta ex situ Metals Meta-analysis Journal of Experimental Marine Biology and Clavero et al. 1994 Spain River Polychaeta ex situ Nutrients Systematic review Ecology Clavero et al. 1992 Hydrobiologia Spain Estuary Polychaeta in situ Nutrients Systematic review Archives of Environmental Contamination and Clements et al. 1994 USA Lake Insecta ex situ PAH Systematic review Toxicology Colombo et al. 2016 Environmental Pollution Australia River, Pond Oligochaeta, Insecta ex situ Metals Systematic review Bulletin of Environmental Contamination and Costa et al. 2017 Brazil Tidal flat Malacostraca In situ Metals Systematic review Toxicology De carvahlo et al. 2016 Marine Pollution Bulletin Brazil Lake Malacostraca in situ Metals Systematic review De Haas et al. 2005 Freshwater Biology Netherlands Estuary Insecta ex situ Metals Meta-analysis De Vaugelas and 1990 Hydrobiologia Jordan Marine Malacostraca in situ Nutrients Systematic review Buscail Delmotte et al. 2007 Geochimica et Cosmochimica Acta France River Oligochaeta ex situ Metals Systematic review Duport et al. 2007 Estuarine, Coastal and Shelf Science France Coastal lagoon Benthic assemblage in situ Nutrients Systematic review Ekeroth et al. 2012 Aquatic Geochemistry Finland Marine Malacostraca ex situ Nutrients Systematic review

206

APPENDIX 1

Table A.1. continued Authors year Journal Country System Class Type Contaminant Used for Fanjul et al. 2015 Journal of Sea Research Argentina Marine Malacostraca in situ Nutrients Systematic review Fanjul et al. 2007 Marine Ecology Progress Series Argentina Coastal lagoon Malacostraca in situ Nutrients Systematic review Fernandes et al. 2009 Estuarine, Coastal and Shelf Science Portugal Estuary Polychaeta ex situ Metals Systematic review Fernandes et al. 2006 Marine Chemistry Portugal Estuary Polychaeta ex situ Metals Systematic review Ferro et al. 2003 Vie et milieu Netherlands Wetlands Polychaeta ex situ Metals Systematic review Fonseca et al. 2003 International Review of Hydrobiology Brazil Lake Insecta ex situ NH3+, SOU Meta-analysis Foshtomi et al. 2015 Plos One Belgium Marine Benthic assemblage in situ Nutrients Systematic review Fukuhara and 1987 Oikos Japan Lake Oligochaeta ex situ Nutrients Systematic review Sakamoto Gammal et al. 2017 Estuaries and Coasts Finland Marine Benthic assemblage in situ Nutrients Systematic review Geraldi et al. 2017 Marine Ecology Progress Series UK Marine Benthic assemblage in situ Nutrients Systematic review Geta et al. 2004 Hydrobiologia Romania Delta Oligochaeta in situ Nutrients Systematic review Journal of Experimental Marine Biology and Gilbert et al. 2016 Finland Marine Bivalvia ex situ Nutrients Systematic review Ecology Polychaeta, Gilbertson et al. 2012 FEMS Microbiology Ecology UK Estuary Malacostraca, ex situ Nutrients Systematic review Gastropoda Glud et al. 2016 Aquatic Geochemistry UK Sea inlet Ophiuroidea in situ Nutrients Systematic review Granberg and Selck 2005 Marine Ecology Progress Series Sweden Marine Equinodermata ex situ PAH Systematic review Granberg et al. 2005 Marine Ecology Progress Series Sweden Marine Ophiuroidea ex situ PAH Systematic review Hansen and Journal of Experimental Marine Biology and 1998 Denmark Fjord Polychaeta ex situ Nutrients Systematic review Kristensen Ecology Hansen et al. 1997 Hydrobiologia Denmark Lake Insecta ex situ NH3+, SOU Meta-analysis Hasanudin et al. 2004 Ecological Engineering Japan Marine Bivalvia ex situ Nutrients Systematic review

207

APPENDIX 1

Table A.1. continued Authors year Journal Country System Class Type Contaminant Used for Oligochaeta, Insecta, He et al. 2015 Science of the Total Environment China Wetlands ex situ Metals Meta-analysis Actinopteryigii Heilskov et al. 2001 ICES Journal of Marine Science Denmark Fjord Polychaeta ex situ Nutrients Systematic review Journal of Experimental Marine Biology and Hietanen et al. 2007 Finland Marine Polychaeta ex situ NH3+, PO43- Meta-analysis Ecology Hines et al. 1984 Marine Chemistry USA Lake Polychaeta, Bivalva in situ Metals Systematic review Holmer and Heilskov 2008 Estuarine, Coastal and Shelf Science Philippines Marine Malacostraca in situ NH3+, SOU Meta-analysis Holmer et al. 2015 Estuaries and Coasts Denmark Mussel farm Benthic assemblage in situ Nutrients Systematic review Holmer et al. 1997 Marine Biology Denmark Estuary Polychaeta ex situ PAH Systematic review Horng et al. Journal of Experimental Marine Biology and 1999 USA Marsh Polychaeta ex situ PAH Systematic review Ecology Huettel et al. 1990 Marine ecology progress series. Oldendorf Germany Marine Polychaeta in situ Nutrients Systematic review Malacostraca, NH3+, PO43- Hughes et al. 2000 Marine Ecology Progress Series UK Sea inlet in situ Meta-analysis Polychaeta , SOU Iribarne et al. 1997 Marine Ecology Progress Series Argentina Coastal Lagoon Malacostraca in situ Nutrients Systematic review Jelbart et al. 2011 Aquaculture. Australia Bivalve farm Benthic assemblage in situ Nutrients Systematic review Jordan et al. 2009 Journal of Sea Research Australia Coastal Lagoon Malacostraca ex situ NH3+, SOU Meta-analysis Joyni et al. 2011 Aquaculture India Pond Actinopterygii ex situ Nutrients Systematic review Kang et al. 2017 Bioresource Technology China Artificial wetland Oligochaeta ex situ Nutrients Systematic review Journal of Experimental Marine Biology and Karlson 2007 Sweden Marine Malacostraca ex situ NH3+, PO43- Meta-analysis Ecology Karlson et al. 2007 Marine Ecology Progress Series Sweden Marine Malacostraca ex situ NH3+, PO43- Meta-analysis Klerks et al. 2007 Marine Chemistry USA River, Estuary Malacostraca in situ Metals Systematic review

208

APPENDIX 1

Table A.1. continued Authors year Journal Country System Class Type Contaminant Used for Canadian Journal of Fisheries and Aquatic Krantzberg and Stokes 1985 Canada Estuary Benthic assemblage ex situ Metals Systematic review Sciences Kristensen and 1987 Journal of Marine Research Denmark Fjord Polychaeta ex situ Nutrients Systematic review Blackburn Kristensen et al. 1992 Limnology and Oceanography Denmark Coastal lagoon Polychaeta ex situ Nutrients Systematic review Kristensen et al. 1991 Journal of Marine Research USA Marine Anthozoa in situ Nutrients Systematic review Kunihiro et al. 2008 Marine Pollution Bulletin Japan Marine Polychaeta in situ Nutrients Systematic review Kunihiro et al. 2011 ISME Journal Japan Marine Polychaeta in situ Nutrients Systematic review Kure et al. 1997 Marine Ecology Progress Series Denmark Fjord Polychaeta ex situ PAH Systematic review Lalonde et al. 2010 Chemical geology Canada Estuary Polychaeta ex situ Metals Systematic review Archives of Environmental Contamination and Landrum et al. 2002 USA River Oligochaeta ex situ PAH Systematic review Toxicology Ophiuroidea, Lindqvist et al. 2009 Marine Environmental Research Sweden Mussel farm ex situ NH3+ Meta-analysis Polychaeta Journal of Experimental Marine Biology and Lohrer et al. 2010 New Zeland Estuary Benthic assemblage in situ Nutrients Systematic review Ecology Journal of Experimental Marine Biology and Lohrer et al. 2005 New Zeland Marine Echinoidea in situ Nutrients Systematic review Ecology Lotufo et al. 1996 Environmental Toxicology and Chemistry USA Drainage system Oligochaeta ex situ PAH Systematic review Zhongguo Huanjing Kexue/China Stormwater Lu and Yan 2010 China Polychaeta in situ Metals Systematic review Environmental Science infiltration basin Ma et al. 2015 Marine Pollution Bulletin France River Oligochaeta ex situ Nutrients Systematic review MacTavish et al. 2012 Plos One New Zeland Marine Holothuroidea ex situ NH3+, PO43- Meta-analysis Madsen et al. 1997 Marine Ecology Progress Series Denmark Fjord Polychaeta ex situ PAH Systematic review Marinelli and Williams 2003 Estuarine, Coastal and Shelf Science USA Marine Bivalvia ex situ NH3+ Meta-analysis

209

APPENDIX 1

Table A.1. continued Authors year Journal Country System Class Type Contaminant Used for Journal of the Marine Biological Association Martinetto et al. 2011 China Artificial pond Holothuroidea in situ Nutrients Systematic review of the United Kingdom Mussel farm , NH3+, Martinez-Garcia et al. 2015 Biogeochemistry Denmark, Spain Polychaeta ex situ Meta-analysis Marine PO43-, SOU Mayor et al. 2013 PLoS ONE UK Estuary Malacostraca ex situ Metals Systematic review McElroy et al. 2016 Marine Pollution Bulletin Australia Estuary NA in situ Metals Systematic review Mchenga et al. 2007 Estuarine, Coastal and Shelf Science Japan Wetland Malacostraca in situ Nutrients Systematic review Meksumpun and 1999 Environmental Pollution Thailand Coastal Polychaeta in situ Nutrients Systematic review Meksumpun Mendez et al. 2001 Marine Biology Denmark Fjord Polychaeta ex situ PAH Systematic review Mermillod-Blondin et al. 2008 Chemosphere France Estuary Oligochaeta ex situ Metals Meta-analysis Mermillod-Blondin et al. 2013 Science of the Total Environment France River Oligochaeta ex situ PAH Systematic review Stormwater Mermillod-Blondin et al. 2008 Chemosphere France Oligochaeta ex situ PAH Systematic review Infiltration System Journal of Experimental Marine Biology and Bivalvia, Malacsotraca, NH3+, Mermillod-Blondin et al. 2005 France Fjord ex situ Meta-analysis Ecology Polychaeta PO43-, SOU Journal of Experimental Marine Biology and Storm Water NH3+, Mermillod-Blondin et al. 2005 France Oligochaeta ex situ Meta-analysis Ecology Infiltration System PO43-, SOU Bivalvia, Polychaeta, NH3+, Mermillod-Blondin et al. 2004 Freshwater Biology Sweden Fjord ex situ Meta-analysis Malacostraca PO43-, SOU Storm Water NH3+, Mermillod-Blondin et al. 2008 Chemosphere France Oligochaeta ex situ Meta-analysis Infiltration System PO43-, SOU Mermillod-Blondin et al. 2004 Aquatic Microbial Ecology Japan Marine Polychaeta ex situ Nutrients Systematic review Mermillod-Blondin et al. 2008 Freshwater Biology USA Estuary Benthic assemblage in situ Nutrients Systematic review Canadian Journal of Fisheries and Aquatic Mermillod-Blondin et al. 2001 Italy Pond Polychaeta ex situ Nutrients Systematic review Sciences

210

APPENDIX 1

Table A.1. continued Authors year Journal Country System Class Type Contaminant Used for Journal of Experimental Marine Biology and Michaud et al. 2006 Canada Estuary Bivalvia, Polychaeta ex situ NH3+, PO43- Meta-analysis Ecology Michio et al. 2003 Marine Pollution Bulletin Japan Marine Holothuroidea ex situ Nutrients Systematic review Montgomery et al. 2008 Bioremediation Journal USA Creek Benthic assemblage in situ PAH Systematic review Moore et al. 2007 Ecology USA River Actinopterygii in situ Nutrients Systematic review Morrisey et al. 1999 Marine Ecology Progress Series New Zeland Estuary Malacostraca in situ Metals Systematic review British virgin Murphy and Kremer 1992 Journal of Marine Research islands Marine Malacostraca in situ Nutrients Systematic review British virgin Murphy and Kremer 1992 Journal of Marine Research Marine Malacostraca in situ Nutrients Systematic review islands Naes et al. 1995 Marine and Freshwater Research Norway Fjord Malacostraca ex situ PAH Systematic review Natalio et al. 2017 Estuarine, Coastal and Shelf Science Brazil Mangrove Malacostraca ex situ Nutrients Systematic review Needham et al. 2011 Ecosystems New Zeeland Estuary Malacostraca in situ Nutrients Systematic review NH3+, PO43- Nicholaus and Zheng 2014 Aquaculture International China Aquaculture Pond Bivalvia ex situ Meta-analysis , SOU Nobbs et al. 1997 Pure and Applied Chemistry Australia Estuary Polychaeta, Bivalva ex situ Metals Systematic review Stormwater Nogaro et al. 2007 Chemosphere France Infiltration System, Oligochaeta ex situ PAH Systematic review Lake, Estuary Storm Water Nogaro et al. 2008 Aquatic Sciences France Insecta ex situ Nutrients Systematic review Infiltration System Storm Water Nogaro et al. 2007 Chemosphere France Oligochaeta ex situ Nutrients Systematic review Infiltration System Oligochaeta, Insecta, Nogaro et al. 2014 Biogeochemistry USA Lake ex situ Nutrients Systematic review Bivalva Norkko et al. 2012 Global Change Biology Finland Marine Polychaeta ex situ Nutrients Systematic review

211

APPENDIX 1

Table A.1. continued Authors year Journal Country System Class Type Contaminant Used for Priapulida, Bivalva, NH3+, PO43- Norling et al. 2007 Marine Ecology Progress Series Sweden Marine ex situ Meta-analysis Polychaeta , SOU Novais et al. 2016 Science of Total Environment Portugal Estuary Bivalvia in situ Nutrients Systematic review Journal of Experimental Marine Biology and O'Brien et al. 2009 Germany Marine Polychaeta in situ Nutrients Systematic review Ecology The Osinga et al. 1997 Journal of Sea Research Marine Equinoidea ex situ Nutrients Systematic review Neatherlands Palmqvist et al. 2006 Aquatic Toxicology Denmark Fjord Polychaeta ex situ PAH Systematic review Pang et al. 2012 Ecotoxicology and Environmental Safety China Creek Oligochaeta ex situ PAH Systematic review Papaspyrou et al. 2007 Estuarine, Coastal and Shelf Science Denmark Fjord Polychaeta in situ Nutrients Systematic review Papaspyrou et al. 2004 Marine Ecology Progress Series Greece Marine Malacostraca ex situ NH3+, SOU Meta-analysis Journal of Experimental Marine Biology and Pascal et al. 2016 France Marine Malacostraca ex situ Nutrients Systematic review Ecology Pelegri and Blackburn 1995 Aquatic Microbial Ecology Denmark Stream Oligochaeta ex situ NH3+, SOU Meta-analysis Pelegri et al. 1994 Marine Biology Denmark Fjord Malacostraca ex situ NH3+, SOU Meta-analysis Penha-Lopez et al. 2009 Marine Pollution Bulletin Tanzania Mangrove Malacostraca ex situ Nutrients Systematic review Pennifold et al. 2001 Hydrological Processes Australia Estuary Benthic assemblage ex situ Nutrients Systematic review Perdo et al. 2015 Estuarine, Coastal and Shelf Science Portugal River Actinopterygii, Bivalva ex situ Metals Systematic review Polychaeta, Petersen et al. 1998 Marine Environmental Research Denmark River, Estuary ex situ Metals Systematic review Malacostraca Peterson et al. 1996 Environmental Toxicology and Chemistry USA Estuary Oligochaeta ex situ Metals Systematic review Petrash et al. 2011 Palaios Canada Not defined Polychaeta ex situ Metals Systematic review Point et al. 2007 Estuarine, Coastal and Shelf Science France Not defined Benthic assemblage in situ Metals Systematic review Poulsen et al. 2014 Systematic and Applied Microbiology Germany Lake Insecta ex situ Nutrients Systematic review

212

APPENDIX 1

Table A.1. continued Authors year Journal Country System Class Type Contaminant Used for Knowledge and Management of Aquatic Puigagut et al. 2014 Canada Aquaculture Pond Oligochaeta ex situ Nutrients Systematic review Ecosystems Qiao et al. 2011 ISWREP 2011 China Estuary, River Oligochaeta ex situ Metals Systematic review Qin et al. 2010 Environmental Toxicology and Chemistry China Estuary Malacostraca in situ PAH Systematic review Quintana et al. 2013 Biogeochemistry Denmark Fjord Polychaeta ex situ NH3+, SOU Meta-analysis The NH3+, PO43- Rao et al. 2014 Estuarine, Coastal and Shelf Science Estuary Polychaeta ex situ Meta-analysis Neatherlands , SOU Rasmussen et al. 1998 Marine Ecology Progress Series Denmark River Polychaeta ex situ Metals Systematic review Rasmussen et al. 2000 Environmental Toxicology and Chemistry Denmark River Polychaeta ex situ Metals Systematic review Reible et al. 1996 Water Research USA Marsh Oligochaeta ex situ PAH Systematic review Remaili et al. 2016 Environmental Pollution Australia Lake Bivalva, Malacostraca ex situ Metals Meta-analysis Remaili et al. 2017 Environmental Pollution Australia Lake Malacostraca ex situ PAH Systematic review Ren et al. 2010 Aquaculture Research China Marine Holothuroidea ex situ Nutrients Systematic review Sweden, NH3+, PO43- Renz and Forster 2014 Marine Ecology Progress Series Marine Polychaeta ex situ Meta-analysis Germany , SOU Riise and Roos 1997 Aquaculture Thailand Aquaculture Pond Actinopterygii in situ Nutrients Systematic review Ritvo et al. 2004 Aquaculture Israel Aquaculture Pond Actinopterygii in situ Nutrients Systematic review Robinsn et al. 2016 Aquaculture Environment Interactions Finland Marine Malacostraca ex situ Nutrients Systematic review No Roche et al. 2013 American Geophysical Union Fall Meeting USA No access Oligochaeta Metals Systematic review access The Rossi et al. 2013 PLoS One Estuary Polychaeta in situ Nutrients Systematic review Neatherlands Schaffner et al. 1997 Environmental Science and Technology USA Marine Benthic assemblage ex situ PAH Systematic review Schaller 2013 Ecological Engineering Germany No access Malacostraca ex situ Metals Systematic review

213

APPENDIX 1

Table A.1. continued Authors year Journal Country System Class Type Contaminant Used for Schaller and Planer- 2017 Chemosphere Germany Estuary Bivalva ex situ Metals Meta-analysis Friedrich Schaller et al. 2014 Chemosphere Germany Lake, Estuary Insecta ex situ Metals Systematic review Shang et al. 2013 Journal of Environmental Sciences (China) China Lake Insecta ex situ NH3+, SOU Meta-analysis Journal of Experimental Marine Biology and Shelley et al. 2008 UK Marine Polychaeta ex situ NH3+ Meta-analysis Ecology Shen et al. 2016 Marine Pollution Bulletin China Coastal Bivalvia, Polychaeta in situ Nutrients Systematic review Journal of Experimental Marine Biology and Shen et al. 2017 China Estuary Benthic assemblage in situ Nutrients Systematic review Ecology Archives of Environmental Contamination Simao et al. 2010 Portugal Lake, Estuary Bivalva ex situ Metals Systematic review and Toxicology Simpson and Batley 2003 Environmental Toxicology and Chemistry Australia River Polychaeta, Bivalva ex situ Metals Systematic review Environmental Science and Technology, Simpson et al. 2002 Australia Estuary Polychaeta, Bivalva ex situ Metals Systematic review ES and T Soster et al. 1992 Hydrobiologia USA Coastal Oligochaeta ex situ Metals Systematic review The Stamhuis et al. 1997 Marine Ecology Progress Series Marine Malacostraca ex situ Nutrients Systematic review Neatherlands Canadian Journal of Fisheries and Aquatic Starkel et al. 1985 USA Coastal Insecta in situ Metals Systematic review Sciences Stauffert et al. 2014 FEMS Microbioogy Ecology France Marine Polychaeta ex situ PAH Systematic review Stief and Hoelker 2006 Ecology Germany Lake Actinopterygii, Insecta ex situ Nutrients Systematic review Stief et al. 2002 Aquatic Microbial Ecology Germany Lake, Stream Insecta ex situ Nutrients Systematic review Sun et al. 2017 Environmental Science Pollution Research China Estuary Malacostraca ex situ PAH Systematic review Sundelin et al. 2001 Environmental Toxicology and Chemistry Sweden Estuary Malacostraca ex situ Metals Systematic review

214

APPENDIX 1

Table A.1. continued Authors year Journal Country System Class Type Contaminant Used for Svensson 1998 Aquatic Microbial Ecology Sweden Lake Insecta ex situ SOU Meta-analysis Svensson and 1996 Freshwater Biology Sweden Lake Benthic assemblage ex situ Nutrients Systematic review Leonardson Swan et al. 2007 Hydrobiologia China Estuary Polychaeta in situ Nutrients Systematic review Tarvainen etal. 2005 Freshwater biology Finland Lake Actinopterygii ex situ Nutrients Systematic review Teal et al. 2013 Biogeosciences UK Commercial Benthic assemblage in situ Metals Systematic review Timmermann et al. 2008 Aquatic Microbial Ecology Denmark Fjord Polychaeta ex situ PAH Systematic review Timmermann et al. 2003 Marine Ecology Progress Series Denmark Fjord Polychaeta ex situ PAH Systematic review Timmermann et al. 2011 Chemosphere Denmark Fjord Polychaeta in situ PAH Systematic review Timmermann et al. 2000 Oil and Hydrocarbon Spills II Denmark Fjord Polychaeta ex situ PAH Systematic review Turek and Hoellein 2015 Freshwater Science USA River Bivalvia in situ SOU Meta-analysis Turner and Bishop 2006 Estuarine Coastal and Shelf Science UK Coastal Polychaeta ex situ Metals Systematic review The Van Duyl et al. 1992 Netherlands Journal of Sea Research Marine Benthic assemblage ex situ Nutrients Systematic review Neatherlands Volkenborn et al. 2007 Limnology and Oceanography Germany Marine Polychaeta in situ Nutrients Systematic review Wall et al. 1996 Environmental Toxicology and Chemistry USA Coastal Actinopterygii ex situ Metals Meta-analysis Wang et al. 2010 Ecosystems China Estuary Malacostraca in situ Nutrients Systematic review Wendelboe et al. 2013 Journal of Sea Research Denmark Fjord Polychaeta ex situ Nutrients Systematic review Wood and Shelley 1999 Ecological Engineering USA Stream NA Model Metals Systematic review Journal of Experimental Marine Biology and Ophiuroidea, Wrede et al. 2017 Germany Marine ex situ Nutrients Systematic review Ecology Equinoidea, Bivalvia Yahel et al. 2008 Marine Ecology Progress Series Canada Sea inlet Benthic assemblage in situ Nutrients Systematic review Yamamuro et al. 1993 Limnology and Oceanography Japan Lake Bivalvia ex situ NH3+, PO43- Meta-analysis

215

APPENDIX 1

Table A.1. continued Authors year Journal Country System Class Type Contaminant Used for Zemlys et al. 2006 Helgoland marine research Lithuania Coastal lagoon Bivalvia in situ Nutrients Systematic review Zhang et al. 2014 Water Research China Lake Oligochaeta, Bivalvia ex situ Nutrients Systematic review NH3+, PO43- Zhang et al. 2011 Water, Air and Soil Pollution China Lake Bivalvia ex situ Meta-analysis , SOU Zheng et al. 2009 Clean-Soil Air Water Belgium Marine Bivalvia ex situ Nutrients Systematic review Zheng et al. 2011 Clean-Soil Air Water China Artificial pond Gastropoda in situ Nutrients Systematic review Actinopterygii, Zhong et al. 2015 Aquaculture International China Aquaculture Pond Malacsotraca in situ Nutrients Systematic review

216

APPENDIX 1

Table A.2. Hierarchical multilevel model for the effect of presence/absence of bioturbation in different meta-analytic responses: total metal concentration in overlying water and ammonia fluxes to water column. Abbreviation: SWIS: Stormwater Infiltration system, AP= Aquaculture pond, CL= Coastal Lagoon, MF= Mussel farm. Models with random factors have adjusted p-values by Bonferroni correction. Number of (Fixed factors) x Random factor Response variable Effect size z value p value studies Presence x Absence Metals 13 -0.090 -0.659 0.510 (Presence x Absence) x Metals 2 -0.697 -2.872 0.016 Actinopterygii (Presence x Absence) x Bivalvia Metals 5 0.098 0.434 1.000 (Presence x Absence) x Insecta Metals 3 0.759 2.271 0.092 (Presence x Absence) x Oligochaeta Metals 7 -0.075 -0.268 1.000 (Presence x Absence) x Dam Metals 1 0.374 0.643 1.000 (Presence x Absence) x Estuary Metals 3 0.205 0.889 1.000 (Presence x Absence) x Lake Metals 2 0.632 0.792 1.000 (Presence x Absence) x River Metals 5 -0.395 -2.137 0.162 (Presence x Absence) x SWIS Metals 1 0.533 0.641 1.000 Presence x Absence Ammonia 30 1.369 11.377 <0.001 (Presence x Absence) x Bivalvia Ammonia 9 1.264 5.656 <0.001 (Presence x Absence) x Ammonia 1 0.8967 1.2089 1.000 Holothuroidea (Presence x Absence) x Insecta Ammonia 3 1.312 3.590 0.0086 (Presence x Absence) x Ammonia 9 1.009 4.303 <0.001 Malacostraca (Presence x Absence) x Oligochaeta Ammonia 2 2.210 5.874 <0.001 (Presence x Absence) x Ammonia 1 1.6720 2.8916 0.099 Ophiuroidea (Presence x Absence) x Polychaeta Ammonia 12 1.453 7.788 <0.001 (Presence x Absence) x Priapulida Ammonia 1 0.7867 1.3404 1.000 (Presence x Absence) x AP Ammonia 2 0.278 0.562 1.000 (Presence x Absence) x CL Ammonia 1 1.889 2.201 0.333 (Presence x Absence) x Estuary Ammonia 3 1.785 4.551 <0.001 (Presence x Absence) x Fjord Ammonia 4 1.235 2.881 0.047 (Presence x Absence) x Lake Ammonia 5 1.558 4.951 <0.001 (Presence x Absence) x Marine Ammonia 10 1.110 5.491 <0.001 (Presence x Absence) x MF Ammonia 2 0.273 0.661 1.000 (Presence x Absence) x Pond Ammonia 1 2.283 2.175 0.355 (Presence x Absence) x River Ammonia 1 2.628 4.909 <0.001 (Presence x Absence) x Sea inlet Ammonia 1 0.630 1.171 1.000 (Presence x Absence) x Stream Ammonia 1 1.660 2.470 0.162 (Presence x Absence) x SWIS Ammonia 1 2.461 2.470 <0.001

217

APPENDIX 1

Table A.3. Hierarchical multilevel model for the effect of presence/absence of bioturbation in four different meta-analytic responses: phosphorous fluxes to water column and sediment Oxygen Uptake (SOU). Abbreviation: SWIS: Stormwater Infiltration system, AP= Aquaculture pond, CL= Coastal Lagoon, MF= Mussel farm. Models with random factors have adjusted p-values by Bonferroni correction. (Fixed factors) x Random factor Response variable Number of studies Effect size z value p value Presence x Absence Phosphorous 18 1.829 4.276 <0.001 (Presence x Absence) x Bivalvia Phosphorous 10 0.599 2.654 0.206 (Presence x Absence) x Holothuroidea Phosphorous 1 1.158 1.516 1.000 (Presence x Absence) x Malacostraca Phosphorous 5 1.326 4.224 <0.001 (Presence x Absence) x Oligochaeta Phosphorous 2 1.796 4.791 <0.001 (Presence x Absence) x Polychaeta Phosphorous 9 0.933 5.130 <0.001 (Presence x Absence) x Priapulida Phosphorous 1 -0.058 -0.098 1.000 (Presence x Absence) x AP Phosphorous 7 0.261 0.659 1.000 (Presence x Absence) x Estuary Phosphorous 4 1.860 3.721 0.0019 (Presence x Absence) x Fjord Phosphorous 6 1.323 2.591 0.096 (Presence x Absence) x Lake Phosphorous 6 2.076 4.468 <0.001 (Presence x Absence) x Marine Phosphorous 13 1.322 4.783 <0.001 (Presence x Absence) x MF Phosphorous 4 0.140 0.341 1.000 (Presence x Absence) x Pond Phosphorous 1 1.040 1.196 1.000 (Presence x Absence) x River Phosphorous 1 0.209 0.532 1.000 (Presence x Absence) x Sea Inlet Phosphorous 2 0.325 0.832 1.000 Presence x Absence x SWIS Phosphorous 3 1.814 4.447 <0.001 Presence x Absence SOU 25 2.018 4.005 <0.001 Presence x Absence x Bivalvia SOU 6 -1.326 -4.356 <0.001 Presence x Absence x Gastropoda SOU 1 0.240 0.344 1.000 Presence x Absence x Insecta SOU 4 2.597 6.708 <0.001 (Presence x Absence) x Malacostraca SOU 8 0.472 1.850 1.000 Presence x Absence x Oligochaeta SOU 2 2.409 6.016 <0.001 (Presence x Absence) x Ophiuroidea SOU 1 -0.193 -0.346 1.000 (Presence x Absence) x Polychaeta SOU 10 -0.027 -0.135 1.000 (Presence x Absence) x Priapulida SOU 1 0.734 -1.234 1.000 (Presence x Absence) x AP SOU 2 0.822 2.150 0.347 (Presence x Absence) x CL SOU 1 -2.764 -2.421 0.170 (Presence x Absence) x Estuary SOU 2 -2.327 -4.567 <0.001 (Presence x Absence) x Fjord SOU 4 1.788 4.100 <0.001 (Presence x Absence) x Lake SOU 5 2.481 7.049 <0.001 (Presence x Absence) x Marine SOU 5 0.4736 1.347 1.000 (Presence x Absence) x River SOU 2 -4.132 -7.173 <0.001 (Presence x Absence) x Sea Inlet SOU 2 -0.662 -1.763 0.858 (Presence x Absence) x Stream SOU 1 1.589 2.359 0.201 Presence x Absence x SWIS SOU 1 2.857 5.737 <0.001

218

APPENDIX 1

Table A.4. Meta-regression fixed to a single predictor model for analysing the correlation between additional moderator variables and the effect size (SDM) obtained through a multilevel model that evaluated the presence/absence of bioturbators and their interaction with 4 response variables: total metal concentration in overlying water, ammonia fluxes to water column, phosphorous fluxes to water column and sediment Oxygen Uptake (SOU). Significant correlations are shown in bold.

Response Meta-regression Factors p value variable coefficient estimate

Effect size (SDM) x temperature (°C) Metals -0.135 0.110

Effect size (SDM) x temperature (°C) Ammonia 0.046 0.003

Effect size (SDM) x temperature (°C) Phosphorous -0.046 0.018

Effect size (SDM) x temperature (°C) SOU 0.033 0.093

Effect size (SDM) x pH Metals 1.621 <0.001

Effect size (SDM) x sediment grain size (mm) Ammonia 0.746 0.006

Effect size (SDM) x sediment grain size (mm) Phosphorous 1.543 <0.001

Effect size (SDM) x sediment grain size (mm) SOU 0.784 <0.001

Effect size (SDM) x Salinity (ppt) Ammonia 0.025 0.190

Effect size (SDM) x Salinity (ppt) Phosphorous -0.023 0.146

Effect size (SDM) x Salinity (ppt) SOU 0.019 0.201

Effect size(SDM) x Animal density (ind-m2) Ammonia 0.000 0.065

Effect size(SDM) x Animal density (ind-m2) Phosphorous 0.000 0.043

Effect size(SDM) x Animal density (ind-m2) SOU 0.000 <0.001

Effect size (SDM) x experiment duration (days) Metals 0.031 <0.001

Effect size (SDM) x experiment duration (days) Ammonia -0.023 <0.001

Effect size (SDM) x experiment duration (days) Phosphorous 0.015 0.010

Effect size (SDM) x experiment duration (days) SOU 0.033 <0.001

219

APPENDIX 1

Attached publication of Chapter 1 results acknowledged throughout the thesis as Vadillo-Gonzalez et al. 2019.

220

APPENDIX 1

221

APPENDIX 1

222

APPENDIX 1

223

APPENDIX 1

224

APPENDIX 1

225

APPENDIX 1

226

APPENDIX 1

227

APPENDIX 1

228

APPENDIX 1

229

APPENDIX 1

230

APPENDIX 2: THE INFLUENCE OF BIOTURBATOR INTRASPECIFIC BODY SIZE VARIATION IN SEDIMENT BIOREMEDIATION: B EFFECTS ON SURVIVORSHIP AND MOBILITY OF A SEDIMENT BIOREMEDIATOR IN ORGANICALLY ENRICHED SEDIMENTS

Fig.B.1. Summary of the experimental design developed for the main experiment for two enrichment treatments (natural or enriched sediments) crossed with 8 cockle single and mixed body size treatments. Total mesocosm number = 70).

231 APPENDIX 2

Table. B.1. Sediment grain size percent of three main sediment categories taken from a subsample from natural and enriched treatments. Sediment was collected in Tilligerry Creek, Taylors Beach NSW. Percentages represent values obtained directly after sediment collection and before enrichment was done. Clay % Silt % Sand % Mesocosm OM Size (<0.002 – 0.004 mm) (0.005-0.060 mm (0.070-1.90mm) 8 Natural No cockles 4.10 23.81 72.09 29 Natural Small 3.84 24.92 71.23 23 Natural Medium 3.52 29.60 66.88 25 Natural Large 3.74 20.96 75.30 18 Natural Small+Medium 5.47 27.41 67.11 20 Natural Small+Large 4.15 24.29 71.56 1 Natural Medium+Large 5.62 33.32 61.06 14 Natural All 5.49 28.51 65.99 44 Enriched No cockles 3.15 23.44 73.41 54 Enriched Small 5.40 31.47 63.13 36 Enriched Medium 5.89 31.51 62.60 45 Enriched Large 6.42 28.03 65.56 51 Enriched Small+Medium 4.60 32.49 62.91 53 Enriched Small+Large 5.22 29.92 64.86 64 Enriched Medium+Large 7.88 32.64 59.47 42 Enriched All 4.70 31.53 63.77

232 APPENDIX 2

A B

Fig. B2. Image processing graphic protocol to describe calculation of cockle mobility using the Image J 1.x. A) A scale in pixels is measured with the photograph using the known diameter of the mesocosm (i.e. 25 cm). As distance change slightly with every daily photo an average of this distance in pixels was calculated. B)After establishing the scale, a straight line from the center of each individual cockle was measured from Day 0 to its new position next day. This was done for each cockle for the 16 days of experiment. From this data, accumulated mobility per cockle in each mesocosm was obtained in both natural and enriched treatments.

233

APPENDIX 2

Fig. B.3. Pearson correlation of A) cockle mortality (%) and B) cockle lateral movement (cm/16 days) and the percentage of organic matter decrease in sediments in natural and sediments exposed to 16 days of enrichment. Coloured area around regression line correspond to calculated confidence intervals.

234 APPENDIX 2

Fig.B.4. General Mixed model diagnostics (QQ plots and residual vs fitted plots) to determine differences in percent cockle mortality between body size combinations in two sediment enrichment types.

235

APPENDIX 2

Fig.B5. General Mixed model diagnostics (QQ plots and residual vs fitted plots) to determine differences in cockle lateral movement between cockle body size combinations in two sediment enrichment types.

236 APPENDIX 2

Fig.B6. General Mixed model diagnostics (QQ plots and residual vs fitted plots) to determine differences in sediment organic matter decrease between cockle body size combinations in two sediment enrichment types.

237

APPENDIX 2

Table B.2. Calculated pairwise comparisons of small cockle mortality between different body size combinations in enriched sediments. Significant differences are shown with an asterisk and in bold. NA= Not analysed. Abbreviations: S = Small, M= Medium and L= Large. % mortality Small only body size combinations S SM SL SML S ------SM 0.314 ------SL 0.5869 0.9472 ------SML 0.0544* 0.6557 0.3664 ----

Table B. 3. Calculated pairwise comparisons of overall cockle lateral mobility between different body size combinations. Significant differences are shown with an asterisk and in bold. Abbreviations: S = Small, M= Medium and L= Large. Overall body size combinations S M L SM SL ML SML S ------M 1.00 ------L 0.03* 0.005* ------SM 0.17 1.00 0.0002* ------SL 0.52 1.00 0.001* 1.00 ------ML 1.00 1.00 0.004* 1.00 1.00 ------SML 1.00 1.00 0.004* 1.00 1.00 1.00 ---

238 APPENDIX 2

Table B.4. Calculated pairwise comparisons of overall cockle lateral mobility between different body size combinations. Significant differences are shown with an asterisk and in bold. Abbreviations: S = Small, M= Medium and L= Large. Large only: Body size combinations L SL ML SML L ------SL 0.0006 ------ML 0.0325 1 ------SML 0.0047 1 1 ---

Table B.5. Calculated pairwise comparisons of percent organic matter breakdown between two sediment types (i.e. natural – enriched) and different body size combinations. Significant differences are shown with an asterisk and in bold. Abbreviations: NC= No cockles, S = Small, M= Medium and L= Large. Sediment type Natural-enriched NC 0.108 S 0.005* M 0.533 L 0.659 SM 0.697 SL 0.112 ML 0.393 SML 0.678

239

APPENDIX 2

240

APPENDIX 3: ORGANIC ENRICHMENT REDUCES MICROBIAL DIVERSITY AND CHANGES COMMUNITY STRUCTURE, BUT THESE EFFECTS ARE NOT MITIGATED BY C THE SYDNEY COCKLE (ANADARA TRAPEZIA)

Fig.C.1. Summary of the experimental design developed for the experiment for two enrichment treatments (natural and enriched sediments) crossed with 3 monocultures (S, M and L), 1 mixed body size combination (SML) and 1 control group (NC, no cockles). Total mesocosm number = 28).

241 APPENDIX 3

Fig. C2. Rarefaction curves obtained from A) Bacterial and B) Archaea zOTUs table of abundance. Blue lines indicate rarefaction curves obtained from mesocosms in natural sediment conditions and green indicate from mesocosms in enriched sediments.

242

APPENDIX 3

Table C1. Bacteria and Archaea α diversity and Good’s coverage data for each mesocosm in natural or enriched sediments with different body size combinations Bacteria Archaea Mesocosm Sediment Body size Good's ID type combinations Richness Diversity Evenness Richness Diversity Evenness Good's coverage coverage 4 Natural No cockles 12186 8.04 0.85 97.5 8630 8.10 0.89 98.0 6 Natural No cockles 13289 8.15 0.86 98.4 9462 8.16 0.89 98.4 7 Natural No cockles 13030 8.25 0.87 98.3 9656 8.13 0.89 99.1 19 Natural Small 12126 8.06 0.86 97.7 9170 8.13 0.89 98.6 22 Natural Small 13175 8.11 0.86 98.3 9389 8.11 0.89 98.5 29 Natural Small 12979 8.22 0.87 97.9 9888 8.22 0.89 99.1 13 Natural Medium 12253 8.10 0.86 97.7 10005 8.17 0.89 99.0 23 Natural Medium 13527 8.32 0.87 98.5 8195 8.06 0.89 97.7 35 Natural Medium 13554 8.26 0.87 98.7 9020 8.16 0.90 98.4 11 Natural Large 12430 8.08 0.86 98.1 9294 8.12 0.89 98.6 21 Natural Large 13273 8.15 0.86 98.5 8678 8.22 0.91 98.0 33 Natural Large 12584 8.07 0.86 97.8 8632 8.20 0.90 98.0 14 Natural Small+Medium+Large 13101 8.08 0.85 98.4 8646 8.07 0.89 98.7 17 Natural Small+Medium+Large 13154 7.99 0.84 98.4 8659 8.13 0.90 98.2 9 Natural Small+Medium+Large 13188 8.32 0.88 97.9 9211 8.11 0.89 98.7 37 Enriched No cockles 4778 4.98 0.59 99.3 6736 7.64 0.87 98.6 40 Enriched No cockles 6308 5.36 0.61 98.4 7268 7.79 0.88 98.3 50 Enriched Small 5361 5.45 0.63 98.5 8295 8.00 0.89 98.5 54 Enriched Small 5459 4.84 0.56 98.9 7865 7.78 0.87 98.9 58 Enriched Small 5886 5.79 0.67 98.6 7771 7.77 0.87 98.8 38 Enriched Medium 5006 5.02 0.59 98.7 7105 7.60 0.86 98.9 62 Enriched Medium 5213 5.27 0.62 98.8 7927 7.69 0.86 99.3

243

APPENDIX 3

Table C.1. continued Bacteria Archaea Mesocosm Sediment Body size Good's ID type combinations Richness Diversity Evenness Richness Diversity Evenness Good's coverage coverage 41 Enriched Large 6780 5.96 0.68 98.4 8870 7.95 0.88 98.8 45 Enriched Large 6654 5.63 0.64 98.7 8483 7.89 0.87 98.9 68 Enriched Large 7059 5.46 0.62 98.8 7458 7.77 0.87 98.9 42 Enriched Small+Medium+Large 5196 5.30 0.62 99.0 7600 7.82 0.87 98.7 43 Enriched Small+Medium+Large 6583 5.46 0.62 98.8 7689 7.79 0.87 98.9 70 Enriched Small+Medium+Large 6174 5.87 0.67 98.8 7548 7.81 0.88 98.1

244

APPENDIX 3

Fig.C3. General linear model diagnostics (residual plots and normality QQ plots) for bacterial alpha diversity indices compared with different sediment types (natural vs. enriched), cockle body size treatments (i.e. No cockles, small, medium, large and small+medium+large) and interaction of both variables.

245

APPENDIX 3

Fig.C4. General linear model diagnostics (residual plots and normality QQ plots) for archaeal alpha diversity indices compared with different sediment types (natural vs. enriched), cockle body size treatments (i.e. No cockles, small, medium, large and small+medium+large) and interaction of both variables.

246

APPENDIX 3

Table C2. Relative abundance (%) of main bacterial phyla sampled in natural and enriched sediments. Within ‘Others’, bacterial phyla with less than 5% were grouped (Acetothermia, Aegiribacteria, AncK6, Armatimonadetes, Atribacteria, BHI80-139, BRC1, Chlamydiae, CK-2C2-2, Cloacimonetes, Cyanobacteria, Dadabacteria, Deinococcus-Thermus, Dependentiae, Elusimicrobia, Entotheonellaeota, FCPU426, GN01, Halanaerobiaeota, Hydrogenedentes, Hydrothermae, Kiritimatiellaeota, LCP-89, Lentisphaerae, Margulisbacteria, Marinimicrobia, SAR406 clade, Modulibacteria, Nitrospinae, Omnitrophicaeota, PAUC34f, Rokubacteria, Schekmanbacteria, Synergistetes, WOR-1, WS1, WS2, WS4). Phylum Natural Phylum Enriched Proteobacteria 44.05 Bacteroidetes 54.60 Chloroflexi 19.32 Firmicutes 13.23 Bacteroidetes 11.97 Proteobacteria 12.34 Acidobacteria 3.81 Spirochaetes 7.71 Firmicutes 3.46 Fusobacteria 5.11 Planctomycetes 2.77 Chloroflexi 3.11 Fusobacteria 2.16 Others 0.79 Calditrichaeota 1.80 Epsilonbacteraeota 0.48 Epsilonbacteraeota 1.77 Fibrobacteres 0.47 Actinobacteria 1.72 Acidobacteria 0.47 Others 1.56 Tenericutes 0.41 Spirochaetes 1.09 Planctomycetes 0.34 Latescibacteria 0.99 Actinobacteria 0.28 Patescibacteria 0.74 Patescibacteria 0.15 Nitrospirae 0.71 Calditrichaeota 0.13 Gemmatimonadetes 0.69 Latescibacteria 0.10 TA06 0.47 Gemmatimonadetes 0.09 Verrucomicrobia 0.44 TA06 0.07 Zixibacteria 0.34 Nitrospirae 0.07 Tenericutes 0.08 Verrucomicrobia 0.05 Fibrobacteres 0.06 Zixibacteria 0.02

247

APPENDIX 3

Fig.C5. General linear model diagnostics (residual plots and normality QQ plots) for main bacterial phyla compared with different sediment types (natural vs. enriched), cockle body size treatments (i.e. No cockles, small, medium, large and small+medium+large) and interaction of both variables.

248

APPENDIX 3

Table C3. Taxonomic description of main unidentified genera that contribute <1% to bacterial community composition dissimilarities between natural and enriched sediment treatments. % Phylum Class Order Family contribution Proteobacteria Gammaproteobacteria B2M28 Unidentified bacterial family 5 Chloroflexi Anaerolineae Anaerolineales Anaerolineaceae 3 Proteobacteria Deltaproteobacteria Desulfobacterales Desulfobulbaceae 3 Actinobacteria WCHB1-81 uncultured bacterium Unidentified bacterial family 2 Bacteroidetes Bacteroidia Bacteroidales Dysgonomonadaceae 2 Chloroflexi Anaerolineae Ardenticatenales Unidentified bacterial family 2 Firmicutes Clostridia Clostridiales Lachnospiraceae 1 Bacteroidetes Bacteroidia Bacteroidales Marinifilaceae 1 Bacteroidetes Bacteroidia Bacteroidales Bacteroidetes BD2-2 1 Proteobacteria Deltaproteobacteria Syntrophobacterales Syntrophobacteraceae 1 Bacteroidetes Bacteroidia Cytophagales Cyclobacteriaceae 1 Proteobacteria Gammaproteobacteria Thiotrichales Thiotrichaceae 1 Calditrichaeota Calditrichia Calditrichales Calditrichaceae 1 Bacteroidetes Bacteroidia Bacteroidales Marinilabiliaceae 1

249

APPENDIX 3

Table C4. Percent contribution (<1%) of specific bacterial genera to dissimilarities evaluated between communities in natural and enriched sediments

% contribution Genus (Natural vs enriched) Bacteroides 16% Sphaerochaeta 4%

Labilibacter 3%

Marinifilum 3% Draconibacterium 2% Fusobacterium 2% Thiohalophilus 2% Carboxylicivirga 1% 1% Desulfobacter Fusibacter 1%

Halodesulfovibrio 1%

Psychrilyobacter 1% Ruminococcus 1 1% Saccharicrinis 1% SEEP-SRB1 1% 1% Sulfurovum Sunxiuqinia 1%

Sva0081 sediment group 1%

Thiogranum 1% Vibrio 1% Woeseia 1%

250

APPENDIX 3

Fig.C6. General linear model diagnostics (residual plots and normality QQ plots) for identified genera (>1% contribution to overall community dissimilarity) compared with natural and enriched sediments.

251

APPENDIX 3

Table C5. Relative abundance (%) of main archaea phyla sampled in natural and enriched sediments. Phylum Natural Phylum Enriched Crenarchaeota 53.00 Crenarchaeota 75.88 Euryarchaeota 36.67 Euryarchaeota 15.29 Asgardaeota 5.54 Asgardaeota 7.30 Nanoarchaeaeota 3.14 Nanoarchaeaeota 0.88 Thaumarchaeota 1.63 Thaumarchaeota 0.66 Hadesarchaeaeota 0.00 Hadesarchaeaeota 0.00 Korarchaeota 0.01 Korarchaeota 0.00

Table C.6. Taxonomic description of main identified classes that contribute (%) to archaeal community composition dissimilarities between natural and enriched sediment treatments. Phylum Class percent Crenarchaeota Bathyarchaeia 20.8 Euryarchaeota Thermoplasmata 13.6 Asgardaeota Lokiarchaeia 2.0 Nanoarchaeaeota Nanohaloarchaeia 1.5 Thaumarchaeota Nitrososphaeria 0.6 Asgardaeota uncultured archaeon 0.2 Thaumarchaeota Marine Benthic Group A 0.1 Euryarchaeota Thermococci 0.1 Asgardaeota Odinarchaeia 0.0 Euryarchaeota Halobacteria 0.0 Thaumarchaeota Group 1.1c 0.0 Asgardaeota Heimdallarchaeia 0.0 Crenarchaeota Crenarchaeota Incertae Sedis 0.0 Korarchaeia Korarchaeia 0.0 Thaumarchaeota SCGC AB-179-E04 0.0 Euryarchaeota Methanococci 0.0

252

APPENDIX 3

Fig.C7. General linear model diagnostics (residual plots and normality QQ plots) for main archaeal phyla compared with different sediment types (natural vs. enriched), cockle body size treatments (i.e. No cockles, small, medium, large and small+medium+large) and interaction of both variables.

253

APPENDIX 3

Fig.C8. General linear model diagnostics (residual plots and normality QQ plots) for identified archaeal classes that contributed to overall community dissimilarity between natural and enriched sediments.

254

APPENDIX 4: LINKING BACTERIAL COMMUNITY SHIFTS TO NITROGEN ENRICHMENT AND SEDIMENT CHARACTERISTICS IN D MACROFAUNA BURROWS

Fig. D1. Difference in porewater ammonia concentration in sediments within control (no nitrogen addition) and high nitrogen addition (600g N/ m2). A previous statistical analysis (general linear mixed model) was done with nitrogen treatment as fixed factor and plot as a random factor. Main effects inferences were done with a Wald Chi test (χ2=215.5, df=1 and p<0.001). Data is shown as mean ± SE. Asterisks (“*”) indicate significant (p < 0.05) lower porewater ammonia concentration.

255 APPENDIX 4

Fig. D2. Arrangement of N enriched plots within a sampling site. Circles represent the area where the plots were placed and coloured squares the different levels of enrichment within an experimental area. Diagram modified from Rebecca Gallagher’s personal notes.

256 APPENDIX 4

Table D1. Description of the 10 selected intertidal zones in the North island, New Zealand where the effect of two nitrogen treatments (N trt, Control= No nitrogen and High= 600g N/m2) on bacterial communities was assessed within two sediment positions (SP, Burrows and surface sediments). Sites include Mahurangi: Lagoon bay (MAH-L) and Mandaley Bay (MAH-M); Raglan (RAG); Tauranga: Tuapiro (TAU-T); Whangarei: Onerahai (WGR-O), Parua bay (WGR-P) and Takahiwai (WGR-T) and Whangateau (WTA). Bacterial alpha diversity indices and community coverage are expressed as follows: Bacteria richness (BR, no. zOTU), diversity (BD, Shannon diversity index), evenness (BE, Pielou index) and Good’s Coverage (GC, %). Covariate abbreviations: OMC= Organic matter content(%), MC= Mud content (%), MGS= median grain size (µm), Chl a= Chlorophyll a concentration (µg/ g), Phaeo= Phaeo pigment concentration (µg/g), Porosity (%), MS= Macrofauna abundance (no. of species) and PWA= Porewater ammonia (µmol N L-1). Site N trt SP Plot Latitude Longitude Date sampled OMC MC MGS Chl a Phaeo Porosity MA PWA BR BD BE GC MAH-L Control Burrows 1 36.2939° S 174. 4415° E 20/10/2017 1.92 1.4 241 3.92 1.79 43 21 105 1338 6.13 0.85 99.98 MAH-L Control Surface 1 36.2939° S 174. 4415° E 20/10/2017 1.92 1.4 241 3.92 1.79 43 21 105 846 5.99 0.89 99.92 MAH-L Control Burrows 2 36.2939° S 174. 4415° E 20/10/2017 2.53 5.84 207 2.24 1.72 50 17 225 1397 6.85 0.95 99.95 MAH-L Control Surface 2 36.2939° S 174. 4415° E 20/10/2017 2.53 5.84 207 2.24 1.72 50 17 225 1168 6.26 0.89 99.92 MAH-L Control Burrows 3 36.2939° S 174. 4415° E 20/10/2017 2.56 4.6 241 4.46 3.57 49 20 289.3 1818 7.06 0.94 99.96 MAH-L Control Surface 3 36.2939° S 174. 4415° E 20/10/2017 2.56 4.6 241 4.46 3.57 49 20 289.3 921 6.17 0.9 99.93 MAH-L High Burrows 1 36.2939° S 174. 4415° E 20/10/2017 1.85 1.23 239 4.8 1.47 48 17 1392.9 1464 6.75 0.93 99.97 MAH-L High Surface 1 36.2939° S 174. 4415° E 20/10/2017 1.85 1.23 239 4.8 1.47 48 17 1392.9 1516 6.68 0.91 99.95 MAH-L High Burrows 2 36.2939° S 174. 4415° E 20/10/2017 2.61 7.01 265 4.2 2.23 46 22 5378.6 1417 6.54 0.9 99.95 MAH-L High Surface 2 36.2939° S 174. 4415° E 20/10/2017 2.61 7.01 265 4.2 2.23 46 22 5378.6 943 6.04 0.88 99.94 MAH-L High Burrows 3 36.2939° S 174. 4415° E 20/10/2017 1.97 1.27 246 3.69 1.29 53 8 2807.1 1191 5.94 0.84 99.98 MAH-L High Surface 3 36.2939° S 174. 4415° E 20/10/2017 1.97 1.27 246 3.69 1.29 53 8 2807.1 899 6.37 0.94 99.97 MAH-M Control Burrows 1 36.2657° S 174.4247° E 20/10/2017 2.29 2.91 274 6.72 2.48 44 16 430.7 1686 6.93 0.93 99.98 MAH-M Control Surface 1 36.2657° S 174.4247° E 20/10/2017 2.29 2.91 274 6.72 2.48 44 16 430.7 1008 6.33 0.91 99.99 MAH-M Control Burrows 2 36.2657° S 174.4247° E 20/10/2017 2.62 7.29 240 5.76 3.32 46 14 183.6 1468 6.71 0.92 99.98 MAH-M Control Surface 2 36.2657° S 174.4247° E 20/10/2017 2.62 7.29 240 5.76 3.32 46 14 183.6 1242 6.57 0.92 99.93 MAH-M Control Burrows 3 36.2657° S 174.4247° E 20/10/2017 2.66 4.68 217 9.29 3.48 50 22 185.7 1129 6.52 0.93 99.91 MAH-M Control Surface 3 36.2657° S 174.4247° E 20/10/2017 2.66 4.68 217 9.29 3.48 50 22 185.7 1058 6.1 0.88 99.96 MAH-M High Burrows 1 36.2657° S 174.4247° E 20/10/2017 2.83 7.82 320 4.94 2.3 47 16 2150 1622 6.84 0.92 99.96 MAH-M High Surface 1 36.2657° S 174.4247° E 20/10/2017 2.83 7.82 320 4.94 2.3 47 16 2150 1079 6.39 0.92 99.96 MAH-M High Burrows 2 36.2657° S 174.4247° E 20/10/2017 2.71 7.39 234 7.14 2.48 47 10 2178.6 2093 7.18 0.94 99.98 MAH-M High Surface 2 36.2657° S 174.4247° E 20/10/2017 2.71 7.39 234 7.14 2.48 47 10 2178.6 1196 6.51 0.92 99.95 MAH-M High Burrows 3 36.2657° S 174.4247° E 20/10/2017 2.81 3.65 212 9.95 3.69 51 18 728.6 1408 6.7 0.92 99.93 MAH-M High Surface 3 36.2657° S 174.4247° E 20/10/2017 2.81 3.65 212 9.95 3.69 51 18 728.6 1247 6.45 0.9 99.96

257 APPENDIX 4 Table D1. continued Site N trt SP Plot Latitude Longitude Date sampled OMC MC MGS Chl a Phaeo Porosity MA PWA BR BD BE GC RAG Control Burrows 1 37.4814° S 174.5203° E 31/10/2017 4.17 19.29 124 19.17 10.34 50 25 21.1 1236 6.48 0.91 99.94 RAG Control Surface 1 37.4814° S 174.5203° E 31/10/2017 4.17 19.29 124 19.17 10.34 50 25 21.1 1195 6.46 0.91 99.91 RAG Control Burrows 2 37.4814° S 174.5203° E 31/10/2017 3.62 14.71 133 16.9 8.03 57 17 28.7 977 6.33 0.92 99.93 RAG Control Surface 2 37.4814° S 174.5203° E 31/10/2017 3.62 14.71 133 16.9 8.03 57 17 28.7 909 6.23 0.91 99.91 RAG Control Burrows 3 37.4814° S 174.5203° E 31/10/2017 3.88 16.76 132 23.1 3.33 51 20 86.4 1217 6.46 0.91 99.92 RAG Control Surface 3 37.4814° S 174.5203° E 31/10/2017 3.88 16.76 132 23.1 3.33 51 20 86.4 940 6.25 0.91 99.92 RAG High Burrows 1 37.4814° S 174.5203° E 31/10/2017 3.96 17.22 135 19.87 6.65 58 15 21214.3 1067 6.41 0.92 99.93 RAG High Surface 1 37.4814° S 174.5203° E 31/10/2017 3.96 17.22 135 19.87 6.65 58 15 21214.3 898 6.42 0.94 99.98 RAG High Burrows 2 37.4814° S 174.5203° E 31/10/2017 3.92 14.71 140 14.85 10.61 50 15 1585.7 1803 6.95 0.93 99.97 RAG High Surface 2 37.4814° S 174.5203° E 31/10/2017 3.92 14.71 140 14.85 10.61 50 15 1585.7 1386 6.63 0.92 99.92 RAG High Burrows 3 37.4814° S 174.5203° E 31/10/2017 4.25 18.95 131 17.01 5.47 56 17 7285.7 1085 6.33 0.91 99.94 RAG High Surface 3 37.4814° S 174.5203° E 31/10/2017 4.25 18.95 131 17.01 5.47 56 17 7285.7 780 6.26 0.94 99.97 TAU-T Control Burrows 1 37.2930° S 175.5707° E 23/10/2017 1.28 4.66 197 8.01 2.64 43 23 10.7 1302 6.3 0.88 99.95 TAU-T Control Surface 1 37.2930° S 175.5707° E 23/10/2017 1.28 4.66 197 8.01 2.64 43 23 10.7 611 5.48 0.85 99.96 TAU-T Control Burrows 2 37.2930° S 175.5707° E 23/10/2017 1.34 4.24 204 12.15 1.12 44 15 12.6 1106 6.18 0.88 99.95 TAU-T Control Surface 2 37.2930° S 175.5707° E 23/10/2017 1.34 4.24 204 12.15 1.12 44 15 12.6 699 5.53 0.84 99.97 TAU-T Control Burrows 3 37.2930° S 175.5707° E 23/10/2017 1.33 3.98 201 10.86 3.61 39 18 7.0 1198 6.21 0.88 99.92 TAU-T Control Surface 3 37.2930° S 175.5707° E 23/10/2017 1.33 3.98 201 10.86 3.61 39 18 7.0 568 5.35 0.84 99.96 TAU-T High Burrows 1 37.2930° S 175.5707° E 23/10/2017 1.05 1.57 204 6.61 1.42 42 17 1120.2 1190 6.27 0.89 99.95 TAU-T High Surface 1 37.2930° S 175.5707° E 23/10/2017 1.05 1.57 204 6.61 1.42 42 17 1120.2 739 5.64 0.85 99.97 TAU-T High Burrows 2 37.2930° S 175.5707° E 23/10/2017 1.24 3.22 206 4.16 1.69 42 16 2165.6 1471 6.3 0.86 99.97 TAU-T High Surface 2 37.2930° S 175.5707° E 23/10/2017 1.24 3.22 206 4.16 1.69 42 16 2165.6 919 5.77 0.85 99.95 TAU-T High Burrows 3 37.2930° S 175.5707° E 23/10/2017 1.07 4.05 209 5.14 1.58 44 17 6740.8 978 5.99 0.87 99.97 TAU-T High Surface 3 37.2930° S 175.5707° E 23/10/2017 1.07 4.05 209 5.14 1.58 44 17 6740.8 721 5.47 0.83 99.97

258 APPENDIX 4

Table D1. continued Site N trt SP Plot Latitude Longitude Date sampled OMC MC MGS Chl a Phaeo Porosity MA PWA BR BD BE GC WGR-O Control Burrows 1 35.0765° S 174.3569° E 5/11/2017 1.94 6.34 116 11.23 2.57 49 7 149.3 1176 6.46 0.91 99.93 WGR-O Control Surface 1 35.0765° S 174.3569° E 5/11/2017 1.94 6.34 116 11.23 2.57 49 7 149.3 887 6.12 0.9 99.93 WGR-O Control Burrows 2 35.0765° S 174.3569° E 5/11/2017 1.54 5.37 118 9.68 2.89 49 8 667.9 897 6.24 0.92 99.93 WGR-O Control Surface 2 35.0765° S 174.3569° E 5/11/2017 1.54 5.37 118 9.68 2.89 49 8 667.9 861 6.11 0.9 99.93 WGR-O Control Burrows 3 35.0765° S 174.3569° E 5/11/2017 1.64 4.91 115 11.23 3.51 48 6 90.7 1315 6.58 0.92 99.94 WGR-O Control Surface 3 35.0765° S 174.3569° E 5/11/2017 1.64 4.91 115 11.23 3.51 48 6 90.7 830 6.06 0.9 99.91 WGR-O High Burrows 1 35.0765° S 174.3569° E 5/11/2017 1.95 6.87 114 12.08 3.88 48 7 1064.3 1332 6.67 0.93 99.91 WGR-O High Surface 1 35.0765° S 174.3569° E 5/11/2017 1.95 6.87 114 12.08 3.88 48 7 1064.3 738 5.95 0.9 99.92 WGR-O High Burrows 2 35.0765° S 174.3569° E 5/11/2017 1.66 7.04 113 12 3.85 51 8 9071.4 1141 6.57 0.93 99.89 WGR-O High Surface 2 35.0765° S 174.3569° E 5/11/2017 1.66 7.04 113 12 3.85 51 8 9071.4 798 6.05 0.91 99.92 WGR-O High Burrows 3 35.0765° S 174.3569° E 5/11/2017 1.54 5.85 108 11.96 3.42 50 4 950.0 1209 6.48 0.91 99.92 WGR-O High Surface 3 35.0765° S 174.3569° E 5/11/2017 1.54 5.85 108 11.96 3.42 50 4 950.0 906 6.16 0.9 99.92 WGR-P Control Burrows 1 35.7799° S 174.4344° E 5/11/2017 2.07 3.84 213 6.97 2.96 46 15 218.6 1382 6.7 0.93 99.91 WGR-P Control Surface 1 35.7799° S 174.4344° E 5/11/2017 2.07 3.84 213 6.97 2.96 46 15 218.6 1005 6.36 0.92 99.94 WGR-P Control Burrows 2 35.7799° S 174.4344° E 5/11/2017 2.33 9.37 209 5.1 2.85 47 7 489.3 1204 6.53 0.92 99.91 WGR-P Control Surface 2 35.7799° S 174.4344° E 5/11/2017 2.33 9.37 209 5.1 2.85 47 7 489.3 1100 6.53 0.93 99.90 WGR-P Control Burrows 3 35.7799° S 174.4344° E 5/11/2017 2.48 7.33 267 6.63 3.07 43 17 785.7 1176 6.5 0.92 99.92 WGR-P Control Surface 3 35.7799° S 174.4344° E 5/11/2017 2.48 7.33 267 6.63 3.07 43 17 785.7 989 6.35 0.92 99.91 WGR-P High Burrows 1 35.7799° S 174.4344° E 5/11/2017 1.92 5.32 206 7.73 3.02 43 10 3392.9 1284 6.65 0.93 99.93 WGR-P High Surface 1 35.7799° S 174.4344° E 5/11/2017 1.92 5.32 206 7.73 3.02 43 10 3392.9 927 6.36 0.93 99.91 WGR-P High Burrows 2 35.7799° S 174.4344° E 5/11/2017 2.19 6.8 204 5.44 3.35 49 14 18857.1 1271 6.64 0.93 99.88 WGR-P High Surface 2 35.7799° S 174.4344° E 5/11/2017 2.19 6.8 204 5.44 3.35 49 14 18857.1 1097 6.39 0.91 99.91 WGR-P High Burrows 3 35.7799° S 174.4344° E 5/11/2017 2.25 9.11 226 6.76 2.8 47 7 3192.9 973 6.37 0.93 99.89 WGR-P High Surface 3 35.7799° S 174.4344° E 5/11/2017 2.25 9.11 226 6.76 2.8 47 7 3192.9 1131 6.5 0.92 99.88

259 APPENDIX 4 Table D1. continued Site N trt SP Plot Latitude Longitude Date sampled OMC MC MGS Chl a Phaeo Porosity MA PWA BR BD BE GC WGR-T Control Burrows 1 35.4953° S 174.2543° E 6/11/2017 1.48 0.53 233 7.12 2.18 44 20 445.0 1164 6.42 0.91 99.95 WGR-T Control Surface 1 35.4953° S 174.2543° E 6/11/2017 1.48 0.53 233 7.12 2.18 44 20 445.0 1129 6.17 0.88 99.95 WGR-T Control Burrows 2 35.4953° S 174.2543° E 6/11/2017 1.34 0.35 224 5.8 2.22 48 20 757.1 1444 6.58 0.9 99.96 WGR-T Control Surface 2 35.4953° S 174.2543° E 6/11/2017 1.34 0.35 224 5.8 2.22 48 20 757.1 841 5.88 0.87 99.89 WGR-T Control Burrows 3 35.4953° S 174.2543° E 6/11/2017 1.4 3.47 214 8.21 2.47 48 16 600.0 1179 6.43 0.91 99.94 WGR-T Control Surface 3 35.4953° S 174.2543° E 6/11/2017 1.4 3.47 214 8.21 2.47 48 16 600.0 884 6.04 0.89 99.95 WGR-T High Burrows 1 35.4953° S 174.2543° E 6/11/2017 1.7 3.34 222 10.45 5.21 47 23 2357.1 882 6.45 0.95 99.98 WGR-T High Surface 1 35.4953° S 174.2543° E 6/11/2017 1.7 3.34 222 10.45 5.21 47 23 2357.1 818 5.9 0.88 99.94 WGR-T High Burrows 2 35.4953° S 174.2543° E 6/11/2017 1.39 2.28 220 6.69 2.78 46 22 13642.9 1590 6.8 0.92 99.98 WGR-T High Surface 2 35.4953° S 174.2543° E 6/11/2017 1.39 2.28 220 6.69 2.78 46 22 13642.9 769 5.82 0.88 99.93 WGR-T High Burrows 3 35.4953° S 174.2543° E 6/11/2017 1.37 0.46 223 6.72 2.81 46 22 1000.0 1528 6.63 0.9 99.96 WGR-T High Burrows 3 35.4953° S 174.2543° E 6/11/2017 1.37 0.46 223 6.72 2.81 46 22 1000.0 877 5.97 0.88 99.95 WHI-L Control Burrows 1 36.1901° S 174.4611° E 2/11/2017 3.54 10.33 156 3.22 17.21 43 19 16.4 1485 6.77 0.93 99.94 WHI-L Control Surface 1 36.1901° S 174.4611° E 2/11/2017 3.54 10.33 156 3.22 17.21 43 19 16.4 1246 6.57 0.92 99.95 WHI-L Control Burrows 2 36.1901° S 174.4611° E 2/11/2017 3.38 9.95 158 10.62 7.28 46 13 5.7 1151 6.48 0.92 99.94 WHI-L Control Surface 2 36.1901° S 174.4611° E 2/11/2017 3.38 9.95 158 10.62 7.28 46 13 5.7 1251 6.58 0.92 99.96 WHI-L Control Burrows 3 36.1901° S 174.4611° E 2/11/2017 4.01 13.36 151 18.45 6.91 55 21 10.9 2309 7.32 0.94 99.99 WHI-L Control Surface 3 36.1901° S 174.4611° E 2/11/2017 4.01 13.36 151 18.45 6.91 55 21 10.9 1195 6.55 0.92 99.93 WHI-L High Surface 1 36.1901° S 174.4611° E 2/11/2017 3.6 8.59 157 18.33 3.13 46 9 1057.3 1205 6.51 0.92 99.97 WHI-L High Burrows 2 36.1901° S 174.4611° E 2/11/2017 3.6 8.59 157 18.33 3.13 46 7 1057.3 1168 6.45 0.91 99.94 WHI-L High Surface 2 36.1901° S 174.4611° E 2/11/2017 3.79 9.51 156 22.59 5.13 51 7 1513.7 1425 6.63 0.91 99.96 WHI-L High Burrows 3 36.1901° S 174.4611° E 2/11/2017 3.79 9.51 156 22.59 5.13 51 8 1513.7 1289 6.54 0.91 99.95 WHI-L High Surface 3 36.1901° S 174.4611° E 2/11/2017 4.14 7.87 160 10.88 19.87 49 8 4219.6 1230 6.52 0.92 99.96

260 APPENDIX 4

Table D1. continued Site N trt SP Plot Latitude Longitude Date sampled OMC MC MGS Chl a Phaeo Porosity MA PWA BR BD BE GC WHI-U Control Burrows 1 36.5029° S 175.4242° E 2/11/2017 4.47 28.53 191 1.17 15.79 58 12 10.5 935 6.06 0.89 99.90 WHI-U Control Surface 1 36.5029° S 175.4242° E 2/11/2017 4.47 28.53 191 1.17 15.79 58 12 10.5 763 5.88 0.89 99.91 WHI-U Control Burrows 2 36.5029° S 175.4242° E 2/11/2017 5.5 19.67 215 14.91 5.82 60 16 5.0 1062 6.25 0.9 99.90 WHI-U Control Surface 2 36.5029° S 175.4242° E 2/11/2017 5.5 19.67 215 14.91 5.82 60 16 5.0 777 5.92 0.89 99.91 WHI-U Control Burrows 3 36.5029° S 175.4242° E 2/11/2017 5.06 27.91 157 5.73 9.03 61 14 17.3 990 6.02 0.87 99.90 WHI-U Control Surface 3 36.5029° S 175.4242° E 2/11/2017 5.06 27.91 157 5.73 9.03 61 14 17.3 866 6.04 0.89 99.92 WHI-U High Burrows 1 36.5029° S 175.4242° E 2/11/2017 4.88 23.2 235 3.06 15.85 54 9 1301.0 1155 6.3 0.89 99.94 WHI-U High Surface 1 36.5029° S 175.4242° E 2/11/2017 4.88 23.2 235 3.06 15.85 54 9 1301.0 755 5.73 0.86 99.92 WHI-U High Burrows 2 36.5029° S 175.4242° E 2/11/2017 5.6 22.81 225 20.06 1.98 55 13 647.9 1189 6.4 0.9 99.91 WHI-U High Surface 2 36.5029° S 175.4242° E 2/11/2017 5.6 22.81 225 20.06 1.98 55 13 647.9 843 6.08 0.9 99.90 WHI-U High Burrows 3 36.5029° S 175.4242° E 2/11/2017 4.51 27.26 148 9.8 9.6 55 5 365.3 126 4.46 0.92 100.00 WHI-U High Surface 3 36.5029° S 175.4242° E 2/11/2017 4.51 27.26 148 9.8 9.6 55 5 365.3 875 6.1 0.9 99.92 WTA Control Burrows 1 36.5234° S 175.4138° E 18/10/2017 1.21 3.3 253 7.36 2.15 47 19 520.0 1420 6.36 0.88 99.97 WTA Control Surface 1 36.5234° S 175.4138° E 18/10/2017 1.21 3.3 253 7.36 2.15 47 19 520.0 781 5.83 0.88 99.96 WTA Control Burrows 2 36.5234° S 175.4138° E 18/10/2017 1.4 2.57 260 8.21 2.42 46 24 771.4 1680 6.81 0.92 99.94 WTA Control Surface 2 36.5234° S 175.4138° E 18/10/2017 1.4 2.57 260 8.21 2.42 46 24 771.4 818 5.85 0.87 99.96 WTA Control Burrows 3 36.5234° S 175.4138° E 18/10/2017 1.56 4.62 241 9.04 3.8 45 21 460.7 1277 6.64 0.93 99.97 WTA Control Surface 3 36.5234° S 175.4138° E 18/10/2017 1.56 4.62 241 9.04 3.8 45 21 460.7 750 5.85 0.88 99.97 WTA High Burrows 1 36.5234° S 175.4138° E 18/10/2017 1.47 2.81 257 12.19 4 47 22 7500.0 1271 6.32 0.88 99.97 WTA High Surface 1 36.5234° S 175.4138° E 18/10/2017 1.47 2.81 257 12.19 4 47 22 7500.0 767 5.68 0.86 99.94 WTA High Burrows 2 36.5234° S 175.4138° E 18/10/2017 1.25 3.01 251 12.19 3.31 47 20 197.1 1134 6.33 0.9 99.95 WTA High Surface 2 36.5234° S 175.4138° E 18/10/2017 1.25 3.01 251 12.19 3.31 47 20 197.1 820 5.83 0.87 99.94 WTA High Burrows 3 36.5234° S 175.4138° E 18/10/2017 1.45 2.99 251 13.73 4.88 49 22 127.9 1430 6.51 0.9 99.96 WTA High Surface 3 36.5234° S 175.4138° E 18/10/2017 1.45 2.99 251 13.73 4.88 49 22 127.9 665 5.57 0.86 99.98

261 APPENDIX 4

Fig. D3. Rarefaction curves of bacterial community samples in A) macrofauna burrows and B) surface sediment. Blue lines indicate curves from bacterial communities with natural levels of N (i.e. control) and red lines from communities with a high N enrichment treatment.

262 APPENDIX 4

A) B) Shannon C) Pielou Evenness Richness diversity Fig. D4. Model diagnostics for alpha diversity index analysis. For all indices, QQ normality plots, residual plots (Fitted vs observed residuals, and Fitted vs Studentized residuals) are shown. For models where plot was not found to be a significant random factor and a GLM approach was done (A), residual vs leverage plots and Cook’s distance plots are also shown.

263 APPENDIX 4

Fig. D5. Model diagnostics for the GLM of Phylum composition analysis. QQ normality plots, residual plots (fitted vs observed residuals, and fitted vs studentized residuals) are shown. As plot was not found to be a significant random factor, a GLM approach was used and a residual vs leverage plot and Cook’s distance were included

264 APPENDIX 4

Fig. D6. Model diagnostics for the GLM of SIMPER genera community dissimilarity. QQ normality plots, residual plots (fitted vs observed residuals, and fitted vs studentized residuals) are shown. As plot was not found to be a significant random factor, a GLM approach was used and a residual vs leverage plot and Cook’s distance were included.

265 APPENDIX 4

Table D2. Summary of reduced planned GLMs were specific covariates in the original model were found as important predictors. Here the effects of organic matter content (%) and mud content (%) are evaluated on all fixed factors in A) bacterial richness (No. ASV), B)Shannon diversity index and C)Pielou evenness. Significant effects of covariates and interactions with the fixed factors are marked in bold. A) Bacterial Richness (No. ASV) Contrasts: Organic matter content (%) df F value p value Organic matter content 1 0.19 0.67 Nitrogen treatments 1 0.06 0.80 Sediment position 1 43.89 <0.001 Organic matter content:Nitrogen treatments 1 0.07 0.80 Organic matter content:Sediment position 1 4.21 0.04 Organic matter content:Nitrogen treatments:Sediment position 1 0.05 0.83 Contrasts: Mud content (%) df F value p value Mud content 1 4.40 0.04 Nitrogen treatments 1 0.08 0.78 Sediment position 1 45.12 <0.001 Mud content:Nitrogen treatments 1 0.47 0.49 Mud content:Sediment position 1 4.68 0.03 Mud content:Nitrogen treatments:Sediment position 1 0.03 0.87 B) Shannon diversity index Contrasts: Organic matter content (%) df F value p value Organic matter content 1 2.11 0.15 Nitrogen treatments 1 0.19 0.66 Sediment position 1 26.66 <0.001 Organic matter content:Nitrogen treatments 1 0.10 0.76 Organic matter content:Sediment position 1 7.93 0.01 Organic matter content:Nitrogen treatments:Sediment position 1 0.25 0.62 Contrasts: Mud content (%) df F value p value Mud content 1 1.34 0.25 Nitrogen treatments 1 0.19 0.66 Sediment position 1 25.93 <0.001 Mud content:Nitrogen treatments 1 0.47 0.49 Mud content:Sediment position 1 7.79 0.01 Mud content:Nitrogen treatments:Sediment position 1 0.38 0.54 B) Pielou Evenness Contrasts: Organic matter content (%) df F value p value Organic matter content 1 10.85 <0.001 Nitrogen treatments 1 0.01 0.92 Sediment position 1 10.08 <0.001 Organic matter content:Nitrogen treatments 1 0.34 0.56 Organic matter content:Sediment position 1 3.75 0.055 Organic matter content:Nitrogen treatments:Sediment position 1 0.02 0.90

266 APPENDIX 4

Table D3. Summary of reduced GLM to determine through planned comparative correlations, the effect of significant covariates found in original GLMM. Within each section (A-C), description of Pearson correlations and estimated marginal means of linear trends (Slope comparison) are given to compare linear trends and slopes of alpha diversity indices and covariates between sediment positions. Significant differences are marked in bold (p<0.05). A) Bacterial Richness (No. ASV) Covariate Pearson correlation Slope comparison Sediment position R p value Contrast df t.ratio p.value Organic matter content (%) Surface 0.30 0.02 Burrow- Surface 115 -2.07 0.04 Burrows -0.12 0.34 Surface 0.01 0.92 Mud content (%) Burrow- Surface 115 -2.18 0.03 Burrows -0.33 0.01 B) Shannon Diversity Index Covariate Pearson correlation Slope comparison Sediment position R p value Contrast df t.ratio p.value Organic matter content (%) Surface 0.42 0.001 Burrow- Surface 114 -2.28 0.02 Burrows 0.046 0.73 Surface 0.16 0.21 Mud content (%) Burrow- Surface 114 -1.64 0.10 Burrows -0.14 0.29 C) Pielou Evenness Covariate Pearson correlation Slope comparison Sediment position R p value Contrast df t.ratio p.value Organic matter content (%) Surface 0.44 <0.001 Burrow- Surface 115 -1.96 0.05 Burrows 0.13 0.32

267 APPENDIX 4

A) Organic Matter content (%) B) Mud content (%)

Richness (No. ASV) (No. Richness

Shannon Diversity Index Diversity Shannon

NOT EVALUATED Pielou Pielou Evenness

Fig. D7. Planned comparative Pearson correlations of selected covariates and significant fixed factors to determine the effect of A) Organic matter content (%) and B) Mud content (%) on alpha diversity indices (Richness, Shannon diversity and Pielou index).

268 APPENDIX 4

Table D4. Summary of descriptive statistics of alpha diversity indices (bacterial richness, Shannon diversity index and Pielou evenness) in fixed factor levels of A)nitrogen treatments and B) sediment position. A) Nitrogen treatments Alpha diversity index Nitrogen trt mean sd n SE Control 1121 304.75 60 39.34 Richness High 1114 323.62 59 42.13 Control 6.31 0.37 60 0.05 Diversity High 6.29 0.43 59 0.06 Control 0.90 0.02 60 0.00 Evenness High 0.90 0.03 59 0.00 B) Sediment position Alpha diversity index Position mean sd n SE Burrows 1279 313.76 60 40.51 Richness Surface 953 210.96 59 27.46 Burrows 6.47 0.39 60 0.05 Diversity Surface 6.13 0.34 59 0.04 Burrows 0.91 0.02 60 0.00 Evenness Surface 0.90 0.03 59 0.00

269 APPENDIX 4

Table D5. Descriptive summary of the relative abundance (%) of main bacterial phyla selected for the analysis between nitrogen treatments (Control and high) and sediment position (Burrows and Surface). Descriptive statistics include mean, standard deviation (SD), number of cases (n) and standard error (SE). Significant differences in relative abundance between factor level are marked in bold. A) Nitrogen treatments

Phylum Control High mean SD n SE mean SD n SE Acidobacteria 2.39 1.47 60 0.19 2.27 2.22 59 0.29 Actinobacteria 4.49 3.12 60 0.40 4.23 3.26 59 0.42 Bacteroidetes 22.91 4.48 60 0.58 21.99 3.63 59 0.47 Chloroflexi 2.12 1.06 60 0.14 2.16 1.33 59 0.17 Cyanobacteria 5.46 5.75 60 0.74 5.80 7.04 59 0.92 Desulfobacterota 9.20 5.08 60 0.66 8.83 5.37 59 0.70 Firmicutes 0.28 0.35 60 0.04 2.11 3.95 59 0.51 Myxococcota 1.57 0.73 60 0.09 1.15 0.60 59 0.08 Planctomycetes 2.31 0.83 60 0.11 2.20 1.13 59 0.15 Proteobacteria 41.11 4.90 60 0.63 40.50 5.21 59 0.68 Verrucomicrobia 3.48 1.03 60 0.13 3.55 1.00 59 0.13 B) Sediment position Burrows Surface Phylum mean SD n SE mean SD n SE Acidobacteria 2.43 1.22 59 0.16 2.22 2.34 60 0.30 Actinobacteria 2.09 1.29 59 0.17 6.59 2.89 60 0.37 Bacteroidetes 22.59 3.72 59 0.48 22.32 4.44 60 0.57 Chloroflexi 2.48 1.30 59 0.17 1.80 0.98 60 0.13 Cyanobacteria 2.94 4.09 59 0.53 8.27 7.14 60 0.92 Desulfobacterota 12.22 5.31 59 0.69 5.86 2.47 60 0.32 Firmicutes 1.85 3.97 59 0.52 0.53 0.87 60 0.11 Myxococcota 1.36 0.64 59 0.08 1.36 0.76 60 0.10 Planctomycetes 2.21 0.88 59 0.11 2.30 1.09 60 0.14 Proteobacteria 38.85 5.30 59 0.69 42.73 3.94 60 0.51 Verrucomicrobia 3.95 1.01 59 0.13 3.09 0.82 60 0.11

270 APPENDIX 4

Table D6. Multiple comparisons using estimated marginal means to determine significant differences of relative abundance between bacterial phyla within A) Nitrogen treatments and B) Sediment position. Significant contrasts within factors are marked in bold. A) Nitrogen treatments (Control vs High) Phylum df t.ratio p.value Acidobacteria 1258 1.63 0.10 Actinobacteria 1258 1.11 0.27 Bacteroidetes 1258 0.42 0.67 Chloroflexi 1258 0.41 0.68 Cyanobacteria 1258 0.26 0.80 Desulfobacterota 1258 0.61 0.54 Firmicutes 1258 -13.02 <0.001 Myxococcota 1258 3.03 <0.001 Planctomycetes 1258 0.96 0.34 Proteobacteria 1258 0.26 0.79 Verrucomicrobia 1258 -0.12 0.90 B) Sediment Position (Burrows vs Surface) Phylum df t.ratio p.value Acidobacteria 1258 4.02 <0.001 Actinobacteria 1258 -9.54 <0.001 Bacteroidetes 1258 0.18 0.86 Chloroflexi 1258 2.55 0.01 Cyanobacteria 1258 -9.71 <0.001 Desulfobacterota 1258 5.92 <0.001 Firmicutes 1258 10.34 <0.001 Myxococcota 1258 -0.19 0.85 Planctomycetes 1258 -0.15 0.88 Proteobacteria 1258 -1.26 0.21 Verrucomicrobia 1258 2.06 0.04

271 APPENDIX 4

Table D7. Summary of reduced planned GLMs were specific covariates in the original model were found as important predictors. Here the effects of A) organic matter content (%), B) Median grain size (µm) and C) Chlorophyll a concentration (µg/ g) on the relative abundance of bacterial phyla within the different fixed factors (Nitrogen treatment and sediment position) are evaluated. Significant effects of organic matter and interactions with the fixed factors are marked in bold. A) Organic matter content (%) Contrasts df F value p value Organic matter content 1 12.02 <0.001 Phylum 10 568.93 <0.001 Nitrogen treatments 1 2.59 0.11 Sediment position 1 1.79 0.18 Organic matter content *Phylum 10 14.62 <0.001 Organic matter content *Nitrogen treatments 1 0.15 0.70 Organic matter content *Sediment position 1 3.04 0.08 Organic matter content *Phylum*Nitrogen treatments 10 0.23 0.99 Organic matter content *Phylum*Sediment position 10 2.67 <0.001 Organic matter content *Nitrogen treatments*Sediment position 1 0.03 0.87 Organic matter content *Phylum*Nitrogen treatments*Sediment position 10 0.04 1.00 B) Median grain size (µm) Contrasts df F value p value Median grain size 1 13.93 <0.001 Phylum 10 593.28 <0.001 Nitrogen treatment 1 3.26 0.07 Sediment position 1 1.83 0.18 Median grain size*Phylum 10 18.45 <0.001 Median grain size*Nitrogen treatment 1 0.09 0.77 Median grain size*Sediment position 1 0.15 0.70 Median grain size*Phylum*Nitrogen treatment 10 0.29 0.98 Median grain size*Phylum*Sediment position 10 4.88 <0.001 Median grain size*Nitrogen treatment*Sediment position 1 0.03 0.86 Median grain size*Phylum*Nitrogen treatment*Sediment position 10 0.08 1.00 C) Chlorophyll a concentration (µg/ g) Contrasts df F value p value Chlorophyll a 1 1.96 0.16 Phylum 10 521.95 <0.001 Nitrogen treatment 1 1.94 0.16 Sediment position 1 1.54 0.21 Chlorophyll a:Phylum 10 4.48 <0.001 Chlorophyll a:Nitrogen treatment 1 1.19 0.28 Chlorophyll a:Sediment position 1 0.78 0.38 Chlorophyll a:Phylum:Nitrogen treatment 10 0.98 0.46 Chlorophyll a:Phylum:Sediment position 10 1.04 0.41 Chlorophyll a:Nitrogen treatment:Sediment position 1 0.08 0.78 Chlorophyll a:Phylum:Nitrogen treatment:Sediment position 10 0.54 0.87

272 APPENDIX 4

Table D8. Summary of Planned comparative correlations were the effect of organic matter content on bacterial relative abundance was evaluated between all selected phyla within burrows and surface sediments. Significant correlations and differences in slopes are marked in bold.

Phylum Pearson correlation Slope comparison Sediment position R p value Contrast df t.ratio p.value Surface -0.054 0.68 Proteobacteria Burrow- Surface 1221 -0.04 0.97 Burrows -0.067 0.62 Surface -0.12 0.69 Bacteroidetes Burrow- Surface 1221 -0.06 0.95 Burrows -0.052 0.38 Surface 0.21 0.11 Desulfobacterota Burrow- Surface 1221 0.38 0.70 Burrows 0.25 0.054 Surface -0.5 <0.001 Cyanobacteria Burrow- Surface 1221 3.05 <0.001 Burrows -0.15 0.25 Surface 0.3 0.019 Actinobacteria Burrow- Surface 1221 0.12 0.90 Burrows 0.27 0.042 Surface 0.048 0.71 Verrucomicrobia Burrow- Surface 1221 -0.76 0.45 Burrows -0.21 0.11 Surface 0.54 <0.001 Acidobacteria Burrow- Surface 1221 -3.44 <0.001 Burrows 0.64 <0.001 Surface 0.08 0.54 Planctomycetes Burrow- Surface 1221 -0.25 0.80 Burrows -0.1 0.44 Surface 0.6 <0.001 Chloroflexi Burrow- Surface 1221 -1.62 0.11 Burrows 0.28 0.031 Surface 0.24 0.068 Myxococcota Burrow- Surface 1221 -1.43 0.15 Burrows 0.08 0.55 Surface -0.045 0.73 Firmicutes Burrow- Surface 1221 -1.74 0.08 Burrows -0.11 0.43

273 APPENDIX 4

Fig. D8. Pearson correlations of the effect of organic matter content (%) on bacterial relative abundance between burrows and surface sediments. Phyla are arranged from higher to lowest relative abundance.

274 APPENDIX 4

Table D9. Summary of Planned comparative correlations were the effect of median grain size on bacterial relative abundance was evaluated between all selected phyla within burrows and surface sediments. Significant correlations and differences in slopes are marked in bold.

Phylum Pearson correlation Slope comparison Sediment position R p value Contrast df t.ratio p.value Surface 0.19 0.14 Proteobacteria Burrow- Surface 1221 0.41 0.68 Burrows 0.35 0.006 Surface 0.053 0.69 Bacteroidetes Burrow- Surface 1221 0.15 0.88 Burrows 0.18 0.17 Surface -0.26 0.05 Desulfobacterota Burrow- Surface 1221 -1.94 0.05 Burrows -0.57 <0.001 Surface 0.53 <0.001 Cyanobacteria Burrow- Surface 1221 -1.39 0.17 Burrows 0.39 0.002 Surface -0.42 <0.001 Actinobacteria Burrow- Surface 1221 2.12 0.03 Burrows 0.099 0.46 Surface 0.13 0.34 Verrucomicrobia Burrow- Surface 1221 0.69 0.49 Burrows 0.44 <0.001 Surface -0.51 <0.001 Acidobacteria Burrow- Surface 1221 3.49 <0.001 Burrows -0.47 <0.001 Surface -0.19 0.14 Planctomycetes Burrow- Surface 1221 0.77 0.44 Burrows 0.026 0.84 Surface -0.35 0.007 Chloroflexi Burrow- Surface 1221 -0.08 0.93 Burrows -0.34 0.008 Surface -0.51 <0.001 Myxococcota Burrow- Surface 1221 1.68 0.09 Burrows -0.36 0.005 Surface 0.12 0.37 Firmicutes Burrow- Surface 1221 -4.67 <0.001 Burrows -0.066 0.62

275 APPENDIX 4

Fig. D9. Pearson correlations of the effect of median grain size (µm) on bacterial main phyla’s relative abundance between burrows and surface sediments. Phyla are arranged from higher to lowest relative abundance

276 APPENDIX 4

Table D10. Summary of Planned comparative correlations were the effect of chlorophyll a concentration in sediment on bacterial relative abundance was evaluated between selected phyla. Significant correlations are marked in bold. Pearson Phylum correlation R p value Acidobacteria 0.40 <0.001 Actinobacteria 0.10 0.30 Bacteroidetes -0.09 0.31 Chloroflexi 0.28 0.002 Cyanobacteria -0.19 0.04 Desulfobacterota 0.16 0.09 Firmicutes -0.04 0.66 Myxococcota 0.17 0.06 Planctomycetes 0.12 0.21 Proteobacteria -0.16 0.09 Verrucomicrobia -0.12 0.18

277 APPENDIX 4

Fig. D10. Pearson correlations of the effect of chlorophyll concentration (µg/g) on bacterial main phyla’s relative abundance.

278 APPENDIX 4

Table D11. Summary of planned comparative correlations were the effect of chlorophyll a concentration on bacterial relative abundance was evaluated between selected phyla. Significant differences in slope are marked in bold. All p values where corrected using Tukey HSD method. Slope comparison Phylum contrast df t.ratio p.value Acidobacteria - Actinobacteria 1221 2.17 0.53 Acidobacteria - Bacteroidetes 1221 2.90 0.12 Acidobacteria - Chloroflexi 1221 1.18 0.98 Acidobacteria - Cyanobacteria 1221 5.77 <0.001 Acidobacteria - Desulfobacterota 1221 1.65 0.86 Acidobacteria - Firmicutes 1221 3.29 0.04 Acidobacteria - Myxococcota 1221 1.40 0.95 Acidobacteria - Planctomycetes 1221 2.01 0.64 Acidobacteria - Proteobacteria 1221 3.05 0.08 Acidobacteria - Verrucomicrobia 1221 2.94 0.11 Actinobacteria - Bacteroidetes 1221 0.73 1.00 Actinobacteria - Chloroflexi 1221 -0.99 1.00 Actinobacteria - Cyanobacteria 1221 3.59 0.01 Actinobacteria - Desulfobacterota 1221 -0.52 1.00 Actinobacteria - Firmicutes 1221 1.12 0.99 Actinobacteria - Myxococcota 1221 -0.77 1.00 Actinobacteria - Planctomycetes 1221 -0.16 1.00 Actinobacteria - Proteobacteria 1221 0.88 1.00 Actinobacteria - Verrucomicrobia 1221 0.76 1.00 Bacteroidetes - Chloroflexi 1221 -1.72 0.83 Bacteroidetes - Cyanobacteria 1221 2.87 0.14 Bacteroidetes - Desulfobacterota 1221 -1.25 0.98 Bacteroidetes - Firmicutes 1221 0.39 1.00 Bacteroidetes - Myxococcota 1221 -1.50 0.92 Bacteroidetes - Planctomycetes 1221 -0.89 1.00 Bacteroidetes - Proteobacteria 1221 0.15 1.00 Bacteroidetes - Verrucomicrobia 1221 0.04 1.00 Chloroflexi - Cyanobacteria 1221 4.58 <0.001 Chloroflexi - Desulfobacterota 1221 0.47 1.00 Chloroflexi - Firmicutes 1221 2.11 0.57 Chloroflexi - Myxococcota 1221 0.22 1.00 Chloroflexi - Planctomycetes 1221 0.82 1.00 Chloroflexi - Proteobacteria 1221 1.87 0.74 Chloroflexi - Verrucomicrobia 1221 1.75 0.81 Cyanobacteria - Desulfobacterota 1221 -4.11 <0.001 Cyanobacteria - Firmicutes 1221 -2.47 0.32 Cyanobacteria - Myxococcota 1221 -4.36 <0.001 Cyanobacteria - Planctomycetes 1221 -3.76 0.01 Cyanobacteria - Proteobacteria 1221 -2.71 0.20 Cyanobacteria - Verrucomicrobia 1221 -2.83 0.15

279 APPENDIX 4

Table D11. continued Slope comparison Phylum contrast df t.ratio p.value Desulfobacterota – Firmicutes 1221 1.64 0.87 Desulfobacterota - Myxococcota 1221 -0.25 1.00 Desulfobacterota - Planctomycetes 1221 0.35 1.00 Desulfobacterota - Proteobacteria 1221 1.40 0.95 Desulfobacterota - Verrucomicrobia 1221 1.28 0.97 Firmicutes - Myxococcota 1221 -1.89 0.72 Firmicutes - Planctomycetes 1221 -1.29 0.97 Firmicutes - Proteobacteria 1221 -0.24 1.00 Firmicutes - Verrucomicrobia 1221 -0.36 1.00 Myxococcota - Planctomycetes 1221 0.61 1.00 Myxococcota - Proteobacteria 1221 1.65 0.86 Myxococcota - Verrucomicrobia 1221 1.53 0.91 Planctomycetes - Proteobacteria 1221 1.04 0.99 Planctomycetes - Verrucomicrobia 1221 0.93 1.00 Proteobacteria - Verrucomicrobia 1221 -0.12 1.00

280 APPENDIX 4

Table D12. List of selected identified and unidentified bacterial genera that contributed to dissimilarities between communities in nitrogen treatments and sediment position. Only genera with >1% contribution where included. Percent contribution was calculated from average proportion of contribution calculated by the simper function. Nitrogen treatments Sediment position Genus % contribution Unclassified genera to class level 9.50 9.74 Gammaproteobacteria_unclassified 9.45 9.65 Bacteroidia_unclassified 5.53 5.63 Woeseia 5.01 5.04 Desulfobulbia_unclassified 3.32 3.66 Alphaproteobacteria_unclassified 2.96 2.97 Halioglobus 2.58 2.67 Cyanobacteriia_unclassified 1.96 2.02 Robiginitalea 1.93 1.95 Acidimicrobiia_unclassified 1.77 1.81 Eudoraea 1.70 1.70 Anaerolineae_unclassified 1.65 1.69 Candidatus_Thiobios 1.26 1.28 Roseobacter 1.23 1.23 Sva0081_sediment_group 1.19 1.23 Polyangia_unclassified 1.01 1.03 Ilumatobacter 1.01 1.08 Pleurocapsa_PCC-7319 0.96 1.00

281 APPENDIX 4

Table D13. Descriptive summary of the relative abundance (%) of selected identified and unclassified genera that described community dissimilarity between the levels of A) Nitrogen treatments (Control and high) and B) sediment position (Burrows and Surface). Descriptive statistics include mean, standard deviation (SD), number of cases (n) and standard error (SE). Significant differences in relative abundance between factor level are marked in bold. A) Nitrogen treatments Control High Genus mean sd n SE mean sd n SE Acidimicrobiia_unclassified 3.47 2.53 60 0.33 3.29 2.59 59 0.34 Alphaproteobacteria_unclassified 4.86 1.98 60 0.26 5.03 1.91 59 0.25 Anaerolineae_unclassified 1.73 0.82 60 0.11 1.80 1.09 59 0.14 Bacteroidia_unclassified 12.57 3.34 60 0.43 11.59 2.74 59 0.36 Candidatus_Thiobios 2.04 1.09 60 0.14 1.78 0.98 59 0.13 Cyanobacteriia_unclassified 2.51 3.97 60 0.51 2.26 4.53 59 0.59 Desulfobulbia_unclassified 4.80 3.51 60 0.45 4.16 3.06 59 0.40 Eudoraea 2.78 1.61 60 0.21 2.34 1.45 59 0.19 Gammaproteobacteria_unclassified 23.82 3.77 60 0.49 22.34 4.67 59 0.61 Halioglobus 3.31 1.83 60 0.24 3.15 2.28 59 0.30 Ilumatobacter 1.29 1.00 60 0.13 1.25 1.10 59 0.14 Pleurocapsa_PCC-7319 1.03 1.26 60 0.16 1.22 2.03 59 0.26 Polyangia_unclassified 1.44 0.71 60 0.09 1.04 0.57 59 0.07 Robiginitalea 3.02 2.24 60 0.29 2.53 1.93 59 0.25 Roseobacter 1.56 1.02 60 0.13 1.80 1.19 59 0.16 Sva0081_sediment_group 1.78 0.91 60 0.12 1.36 0.72 59 0.09 Woeseia 6.83 3.35 60 0.43 6.79 3.81 59 0.50 B) Sediment position Burrows Surface Genus mean sd n SE mean sd n SE Acidimicrobiia_unclassified 1.62 1.17 59 0.15 5.11 2.34 60 0.30 Alphaproteobacteria_unclassified 4.00 1.71 59 0.22 5.88 1.69 60 0.22 Anaerolineae_unclassified 2.03 1.00 59 0.13 1.50 0.84 60 0.11 Bacteroidia_unclassified 11.25 2.29 59 0.30 12.90 3.53 60 0.46 Candidatus_Thiobios 1.97 1.14 59 0.15 1.85 0.93 60 0.12 Cyanobacteriia_unclassified 0.75 1.20 59 0.16 3.99 5.41 60 0.70 Desulfobulbia_unclassified 6.57 3.43 59 0.45 2.44 1.25 60 0.16 Eudoraea 1.85 1.11 59 0.14 3.25 1.60 60 0.21 Gammaproteobacteria_unclassified 22.71 4.06 59 0.53 23.45 4.50 60 0.58 Halioglobus 4.00 1.80 59 0.23 2.48 2.03 60 0.26 Ilumatobacter 0.49 0.31 59 0.04 2.03 0.95 60 0.12 Pleurocapsa_PCC-7319 0.53 1.49 59 0.19 1.71 1.67 60 0.22 Polyangia_unclassified 1.21 0.60 59 0.08 1.27 0.74 60 0.10 Robiginitalea 1.93 1.08 59 0.14 3.61 2.49 60 0.32 Roseobacter 1.37 0.70 59 0.09 1.98 1.34 60 0.17 Sva0081_sediment_group 1.47 0.76 59 0.10 1.68 0.92 60 0.12 Woeseia 5.38 2.59 59 0.34 8.22 3.85 60 0.50

282 APPENDIX 4

Table D14. Multiple comparisons using estimated marginal means to determine significant differences of relative abundance between selected bacterial genera within burrows and surface sediments. Significant contrasts within factors are marked in bold. Genus df t ratio p value Acidimicrobiia_unclassified 1953 -8.748 <0.001 Alphaproteobacteria_unclassified 1953 -2.882 0.004 Anaerolineae_unclassified 1953 2.219 0.027 Bacteroidia_unclassified 1953 -0.952 0.341 Candidatus_Thiobios 1953 0.168 0.866 Cyanobacteriia_unclassified 1953 -11.907 <0.001 Desulfobulbia_unclassified 1953 6.412 <0.001 Eudoraea 1953 -3.667 <0.001 Gammaproteobacteria_unclassified 1953 -0.306 0.759 Halioglobus 1953 4.193 <0.001 Ilumatobacter 1953 -10.218 <0.001 Pleurocapsa_PCC-7319 1953 -15.424 <0.001 Polyangia_unclassified 1953 -0.476 0.634 Robiginitalea 1953 -3.896 <0.001 Roseobacter 1953 -1.920 0.055 Sva0081_sediment_group 1953 -0.946 0.344 Woeseia 1953 -2.946 0.003

283 APPENDIX 4

Table D15. Summary of reduced planned GLMs were specific covariates in the original model were found as important predictors. Here the effects of A) organic matter content (%) and B) Mud content (%) on the relative abundance of selected identified and unclassified genera that described community dissimilarity within the different fixed factors (genus identity, nitrogen treatment and sediment position) Significant effects of covariates and interactions with the fixed factors are marked in bold. A) Organic matter content (%) Contrasts df F value p value Organic matter 1 18.24 <0.001 Genus identity 16 254.40 <0.001 Nitrogen treatments 1 16.86 <0.001 Sediment position 1 178.45 <0.001 Organic matter*Genus identity 16 23.99 <0.001 Organic matter*Nitrogen tretaments 1 0.90 0.343 Organic matter*Sediment position 1 0.23 0.630 Organic matter*Genus identity*Nitrogen tretaments 16 0.28 0.998 Organic matter*Genus identity*Sediment position 16 0.61 0.878 Organic matter*Nitrogen tretaments*Sediment position 1 0.66 0.415 Organic matter*Genus identity*Nitrogen tretaments*Sediment position 16 0.53 0.935 B) Mud content content (%) Contrasts df F value p value Mud content 1 4.16 0.041 Genus identity 16 243.36 <0.001 Nitrogen tretaments 1 16.30 <0.001 Sediment position 1 169.28 <0.001 Mud content*Genus identity 16 18.20 <0.001 Mud content*Nitrogen tretaments 1 2.00 0.158 Mud content*Sediment position 1 0.22 0.638 Mud content*Genus identity*Nitrogen tretaments 16 0.35 0.992 Mud content*Genus identity*Sediment position 16 0.81 0.678 Mud content*Nitrogen tretaments*Sediment position 1 0.01 0.916 Mud content*Genus identity*Nitrogen tretaments*Sediment position 16 0.62 0.870

284 APPENDIX 4

Table D16. Summary of Planned comparative correlations were the effect of organic matter content and mud content on selected identified and unclassified genera relative abundance was evaluated. Significant correlations are marked in bold Organic matter (%) Mud content (%) Genus Pearson correlation Pearson correlation R p value R p value Acidimicrobiia_unclassified 0.24 0.01 0.09 0.32 Alphaproteobacteria_unclassified -0.19 0.04 -0.27 <0.001 Anaerolineae_unclassified 0.37 <0.001 0.20 0.03 Bacteroidia_unclassified 0.10 0.30 0.26 0.01 Candidatus_Thiobios -0.11 0.22 0.09 0.31 Cyanobacteriia_unclassified -0.37 <0.001 -0.29 0.002 Desulfobulbia_unclassified 0.21 0.02 0.30 <0.001 Eudoraea 0.35 <0.001 0.43 <0.001 Gammaproteobacteria_unclassified 0.00 0.98 0.13 0.16 Halioglobus -0.08 0.38 -0.05 58.00 Ilumatobacter 0.17 0.06 0.15 0.10 Pleurocapsa_PCC-7319 -0.28 0.002 -0.32 <0.001 Polyangia_unclassified 0.19 0.04 0.23 0.01 Robiginitalea -0.12 0.19 0.07 0.46 Roseobacter -0.54 <0.001 -0.35 <0.001 Sva0081_sediment_group -0.09 0.32 -0.05 0.61 Woeseia 0.21 0.03 0.25 0.01

285 APPENDIX 4

Table D17. Summary of planned comparative correlations were the effect of organic matter content (%) and mud content (%) on selected identified and unclassified genera relative abundance was evaluated. Significant differences in slopes are marked in bold. All p values where corrected using Tukey HSD method. Organic matter content (%) Mud content (%) Slope contrast df t ratio p value df t ratio p value Acidimicrobiia_unclassified - Alphaproteobacteria_unclassified 1887 2.24 0.695 1887 1.223 0.999 Acidimicrobiia_unclassified - Anaerolineae_unclassified 1887 -1.10 1.000 1887 -1.584 0.978 Acidimicrobiia_unclassified - Bacteroidia_unclassified 1887 1.06 1.000 1887 -0.695 1.000 Acidimicrobiia_unclassified - Candidatus_Thiobios 1887 2.45 0.539 1887 -0.220 1.000 Acidimicrobiia_unclassified - Cyanobacteriia_unclassified 1887 10.43 0.000 1887 6.682 0.000 Acidimicrobiia_unclassified - Desulfobulbia_unclassified 1887 -0.69 1.000 1887 -2.554 0.461 Acidimicrobiia_unclassified - Eudoraea 1887 -1.60 0.976 1887 -3.156 0.124 Acidimicrobiia_unclassified - Gammaproteobacteria_unclassified 1887 1.28 0.998 1887 -0.342 1.000 Acidimicrobiia_unclassified - Halioglobus 1887 1.40 0.994 1887 0.141 1.000 Acidimicrobiia_unclassified - Ilumatobacter 1887 0.02 1.000 1887 -1.110 1.000 Acidimicrobiia_unclassified - (Pleurocapsa_PCC-7319) 1887 9.90 0.000 1887 8.419 0.000 Acidimicrobiia_unclassified - Polyangia_unclassified 1887 0.33 1.000 1887 -0.963 1.000 Acidimicrobiia_unclassified - Robiginitalea 1887 2.37 0.605 1887 -0.436 1.000 Acidimicrobiia_unclassified - Roseobacter 1887 4.89 0.000 1887 1.891 0.899 Acidimicrobiia_unclassified - Sva0081_sediment_group 1887 1.74 0.949 1887 0.031 1.000 Acidimicrobiia_unclassified - Woeseia 1887 -0.65 1.000 1887 -1.945 0.876 Alphaproteobacteria_unclassified - Anaerolineae_unclassified 1887 -3.34 0.073 1887 -2.806 0.288 Alphaproteobacteria_unclassified - Bacteroidia_unclassified 1887 -1.18 0.999 1887 -1.917 0.888 Alphaproteobacteria_unclassified - Candidatus_Thiobios 1887 0.21 1.000 1887 -1.442 0.992 Alphaproteobacteria_unclassified - Cyanobacteriia_unclassified 1887 8.18 0.000 1887 5.460 0.000 Alphaproteobacteria_unclassified - Desulfobulbia_unclassified 1887 -2.94 0.214 1887 -3.777 0.017 Alphaproteobacteria_unclassified - Eudoraea 1887 -3.85 0.013 1887 -4.378 0.002 Alphaproteobacteria_unclassified - Gammaproteobacteria_unclassified 1887 -0.96 1.000 1887 -1.564 0.981 Alphaproteobacteria_unclassified - Halioglobus 1887 -0.84 1.000 1887 -1.081 1.000 Alphaproteobacteria_unclassified - Ilumatobacter 1887 -2.23 0.709 1887 -2.332 0.630 Alphaproteobacteria_unclassified - (Pleurocapsa_PCC- 7319) 1887 7.66 0.000 1887 7.196 0.000 Alphaproteobacteria_unclassified - Polyangia_unclassified 1887 -1.91 0.890 1887 -2.186 0.737 Alphaproteobacteria_unclassified - Robiginitalea 1887 0.12 1.000 1887 -1.659 0.966 Alphaproteobacteria_unclassified - Roseobacter 1887 2.64 0.396 1887 0.668 1.000 Alphaproteobacteria_unclassified - Sva0081_sediment_group 1887 -0.51 1.000 1887 -1.191 0.999 Alphaproteobacteria_unclassified - Woeseia 1887 -2.89 0.238 1887 -3.167 0.120

286 APPENDIX 4

Table D17. continued Organic matter content (%) Mud content (%) Slope contrast df t ratio p value df t ratio p value Anaerolineae_unclassified - Bacteroidia_unclassified 1887 2.16 0.754 1887 0.889 1.000 Anaerolineae_unclassified - Candidatus_Thiobios 1887 3.55 0.038 1887 1.364 0.995 Anaerolineae_unclassified - Cyanobacteriia_unclassified 1887 11.52 0.000 1887 8.266 0.000 Anaerolineae_unclassified - Desulfobulbia_unclassified 1887 0.40 1.000 1887 -0.971 1.000 Anaerolineae_unclassified - Eudoraea 1887 -0.51 1.000 1887 -1.572 0.980 Anaerolineae_unclassified - Gammaproteobacteria_unclassified 1887 2.38 0.596 1887 1.242 0.998 Anaerolineae_unclassified - Halioglobus 1887 2.50 0.504 1887 1.725 0.952 Anaerolineae_unclassified - Ilumatobacter 1887 1.11 1.000 1887 0.474 1.000 Anaerolineae_unclassified - (Pleurocapsa_PCC-7319) 1887 11.00 0.000 1887 10.002 0.000 Anaerolineae_unclassified - Polyangia_unclassified 1887 1.43 0.993 1887 0.620 1.000 Anaerolineae_unclassified - Robiginitalea 1887 3.46 0.050 1887 1.147 0.999 Anaerolineae_unclassified - Roseobacter 1887 5.98 0.000 1887 3.474 0.048 Anaerolineae_unclassified - Sva0081_sediment_group 1887 2.83 0.271 1887 1.615 0.974 Anaerolineae_unclassified - Woeseia 1887 0.45 1.000 1887 -0.361 1.000 Bacteroidia_unclassified - Candidatus_Thiobios 1887 1.39 0.994 1887 0.475 1.000 Bacteroidia_unclassified - Cyanobacteriia_unclassified 1887 9.36 0.000 1887 7.377 0.000 Bacteroidia_unclassified - Desulfobulbia_unclassified 1887 -1.76 0.944 1887 -1.860 0.911 Bacteroidia_unclassified - Eudoraea 1887 -2.67 0.379 1887 -2.461 0.532 Bacteroidia_unclassified - Gammaproteobacteria_unclassified 1887 0.22 1.000 1887 0.353 1.000 Bacteroidia_unclassified - Halioglobus 1887 0.34 1.000 1887 0.836 1.000 Bacteroidia_unclassified - Ilumatobacter 1887 -1.04 1.000 1887 -0.415 1.000 Bacteroidia_unclassified - (Pleurocapsa_PCC-7319) 1887 8.84 0.000 1887 9.113 0.000 Bacteroidia_unclassified - Polyangia_unclassified 1887 -0.73 1.000 1887 -0.269 1.000 Bacteroidia_unclassified - Robiginitalea 1887 1.30 0.997 1887 0.258 1.000 Bacteroidia_unclassified - Roseobacter 1887 3.82 0.014 1887 2.585 0.438 Bacteroidia_unclassified - Sva0081_sediment_group 1887 0.67 1.000 1887 0.726 1.000 Bacteroidia_unclassified - Woeseia 1887 -1.71 0.955 1887 -1.250 0.998 Candidatus_Thiobios - Cyanobacteriia_unclassified 1887 7.97 0.000 1887 6.902 0.000 Candidatus_Thiobios - Desulfobulbia_unclassified 1887 -3.15 0.127 1887 -2.334 0.629 Candidatus_Thiobios - Eudoraea 1887 -4.05 0.006 1887 -2.936 0.216 Candidatus_Thiobios - Gammaproteobacteria_unclassified 1887 -1.17 0.999 1887 -0.122 1.000 Candidatus_Thiobios - Halioglobus 1887 -1.05 1.000 1887 0.361 1.000 Candidatus_Thiobios - Ilumatobacter 1887 -2.43 0.554 1887 -0.890 1.000 Candidatus_Thiobios - (Pleurocapsa_PCC-7319) 1887 7.45 0.000 1887 8.639 0.000 Candidatus_Thiobios - Polyangia_unclassified 1887 -2.12 0.779 1887 -0.743 1.000 Candidatus_Thiobios - Robiginitalea 1887 -0.09 1.000 1887 -0.216 1.000 Candidatus_Thiobios - Roseobacter 1887 2.43 0.552 1887 2.111 0.786 Candidatus_Thiobios - Sva0081_sediment_group 1887 -0.71 1.000 1887 0.251 1.000 Candidatus_Thiobios - Woeseia 1887 -3.10 0.143 1887 -1.725 0.952

287 APPENDIX 4

Table D17. continued Organic matter content (%) Mud content (%) Slope contrast df t ratio p value df t ratio p value Cyanobacteriia_unclassified - Desulfobulbia_unclassified 1887 -11.12 0.000 1887 -9.236 0.000 Cyanobacteriia_unclassified - Eudoraea 1887 -12.03 0.000 1887 -9.838 0.000 Cyanobacteriia_unclassified - Gammaproteobacteria_unclassified 1887 -9.14 0.000 1887 -7.024 0.000 Cyanobacteriia_unclassified - Halioglobus 1887 -9.02 0.000 1887 -6.541 0.000 Cyanobacteriia_unclassified - Ilumatobacter 1887 -10.41 0.000 1887 -7.792 0.000 Cyanobacteriia_unclassified - (Pleurocapsa_PCC-7319) 1887 -0.52 1.000 1887 1.737 0.949 Cyanobacteriia_unclassified - Polyangia_unclassified 1887 -10.09 0.000 1887 -7.645 0.000 Cyanobacteriia_unclassified - Robiginitalea 1887 -8.06 0.000 1887 -7.118 0.000 Cyanobacteriia_unclassified - Roseobacter 1887 -5.54 0.000 1887 -4.791 0.000 Cyanobacteriia_unclassified - Sva0081_sediment_group 1887 -8.69 0.000 1887 -6.651 0.000 Cyanobacteriia_unclassified - Woeseia 1887 -11.07 0.000 1887 -8.627 0.000 Desulfobulbia_unclassified - Eudoraea 1887 -0.91 1.000 1887 -0.601 1.000 Desulfobulbia_unclassified - Gammaproteobacteria_unclassified 1887 1.98 0.861 1887 2.213 0.718 Desulfobulbia_unclassified - Halioglobus 1887 2.10 0.794 1887 2.696 0.359 Desulfobulbia_unclassified - Ilumatobacter 1887 0.71 1.000 1887 1.445 0.991 Desulfobulbia_unclassified - (Pleurocapsa_PCC-7319) 1887 10.60 0.000 1887 10.973 0.000 Desulfobulbia_unclassified - Polyangia_unclassified 1887 1.03 1.000 1887 1.591 0.977 Desulfobulbia_unclassified - Robiginitalea 1887 3.06 0.159 1887 2.118 0.781 Desulfobulbia_unclassified - Roseobacter 1887 5.58 0.000 1887 4.445 0.001 Desulfobulbia_unclassified - Sva0081_sediment_group 1887 2.43 0.553 1887 2.586 0.437 Desulfobulbia_unclassified - Woeseia 1887 0.05 1.000 1887 0.610 1.000 Eudoraea - Gammaproteobacteria_unclassified 1887 2.88 0.243 1887 2.814 0.283 Eudoraea - Halioglobus 1887 3.00 0.183 1887 3.297 0.083 Eudoraea - Ilumatobacter 1887 1.62 0.973 1887 2.046 0.824 Eudoraea - (Pleurocapsa_PCC-7319) 1887 11.50 0.000 1887 11.574 0.000 Eudoraea - Polyangia_unclassified 1887 1.93 0.881 1887 2.192 0.732 Eudoraea - Robiginitalea 1887 3.97 0.008 1887 2.719 0.343 Eudoraea - Roseobacter 1887 6.49 0.000 1887 5.047 0.000 Eudoraea - Sva0081_sediment_group 1887 3.34 0.073 1887 3.187 0.114 Eudoraea - Woeseia 1887 0.95 1.000 1887 1.211 0.999 Gammaproteobacteria_unclassified - Halioglobus 1887 0.12 1.000 1887 0.483 1.000 Gammaproteobacteria_unclassified - Ilumatobacter 1887 -1.26 0.998 1887 -0.768 1.000 Gammaproteobacteria_unclassified - (Pleurocapsa_PCC-7319) 1887 8.62 0.000 1887 8.760 0.000 Gammaproteobacteria_unclassified - Polyangia_unclassified 1887 -0.95 1.000 1887 -0.622 1.000 Gammaproteobacteria_unclassified - Robiginitalea 1887 1.08 1.000 1887 -0.095 1.000 Gammaproteobacteria_unclassified - Roseobacter 1887 3.60 0.031 1887 2.232 0.704 Gammaproteobacteria_unclassified - Sva0081_sediment_group 1887 0.46 1.000 1887 0.373 1.000 Gammaproteobacteria_unclassified - Woeseia 1887 -1.93 0.882 1887 -1.603 0.975

288 APPENDIX 4

Table D17. continued Organic matter content (%) Mud content (%) Slope contrast df t ratio p value df t ratio p value Halioglobus – Ilumatobacter 1887 -1.38 0.995 1887 -1.251 0.998 Halioglobus - (Pleurocapsa_PCC-7319) 1887 8.50 0.000 1887 8.278 0.000 Halioglobus - Polyangia_unclassified 1887 -1.07 1.000 1887 -1.104 1.000 Halioglobus - Robiginitalea 1887 0.96 1.000 1887 -0.577 1.000 Halioglobus - Roseobacter 1887 3.48 0.047 1887 1.750 0.946 Halioglobus - Sva0081_sediment_group 1887 0.34 1.000 1887 -0.110 1.000 Halioglobus - Woeseia 1887 -2.05 0.821 1887 -2.086 0.801 Ilumatobacter - (Pleurocapsa_PCC-7319) 1887 9.88 0.000 1887 9.529 0.000 Ilumatobacter - Polyangia_unclassified 1887 0.31 1.000 1887 0.147 1.000 Ilumatobacter - Robiginitalea 1887 2.35 0.619 1887 0.674 1.000 Ilumatobacter - Roseobacter 1887 4.87 0.000 1887 3.001 0.185 Ilumatobacter - Sva0081_sediment_group 1887 1.72 0.954 1887 1.141 0.999 Ilumatobacter - Woeseia 1887 -0.67 1.000 1887 -0.835 1.000 (Pleurocapsa_PCC-7319) - Polyangia_unclassified 1887 -9.57 0.000 1887 -9.382 0.000 (Pleurocapsa_PCC-7319) - Robiginitalea 1887 -7.54 0.000 1887 -8.855 0.000 (Pleurocapsa_PCC-7319) - Roseobacter 1887 -5.01 0.000 1887 -6.528 0.000 (Pleurocapsa_PCC-7319) - Sva0081_sediment_group 1887 -8.16 0.000 1887 -8.387 0.000 (Pleurocapsa_PCC-7319) - Woeseia 1887 -10.55 0.000 1887 -10.363 0.000 Polyangia_unclassified - Robiginitalea 1887 2.04 0.830 1887 0.527 1.000 Polyangia_unclassified - Roseobacter 1887 4.56 0.001 1887 2.854 0.260 Polyangia_unclassified - Sva0081_sediment_group 1887 1.41 0.993 1887 0.995 1.000 Polyangia_unclassified - Woeseia 1887 -0.98 1.000 1887 -0.981 1.000 Robiginitalea - Roseobacter 1887 2.52 0.486 1887 2.327 0.634 Robiginitalea - Sva0081_sediment_group 1887 -0.63 1.000 1887 0.468 1.000 Robiginitalea - Woeseia 1887 -3.02 0.178 1887 -1.508 0.987 Roseobacter - Sva0081_sediment_group 1887 -3.15 0.127 1887 -1.860 0.911 Roseobacter - Woeseia 1887 -5.54 0.000 1887 -3.836 0.014 Sva0081_sediment_group - Woeseia 1887 -2.39 0.588 1887 -1.976 0.861

289 APPENDIX 4

Fig. D11. Pearson correlations of the effect of organic matter content (%) on selected identified and unclassified genera that described community dissimilarity.

290 APPENDIX 4

Fig. D12. Pearson correlations of the effect of mud content (%) on selected identified and unclassified genera that described community dissimilarity.

291 APPENDIX 4

A B

C D

Fig. D13. Pipeline comparison of rarefaction curves obtained through the DADA2 algorithm (A and C) and the original method using Usearch-Unoise (B and D). A and C show rarefaction curves in surface sediments and B and C in macrofauna burrows. Red lines show samples obtained from high nitrogen conditions and blue lines show control conditions (no nitrogen added.

292 APPENDIX 4

Fig. D14.Pipeline comparison of alfa diversity indices (richness, Shannon diversity and Pielou evenness) grouped by nitrogen treatments (control and high).

293 APPENDIX 4

Fig. D15.Pipeline comparison of alfa diversity indices (richness, Shannon diversity and Pielou evenness) grouped by sediment position (Surface and macrofauna burrows). Significant differences are marked with an asterisk “*”.

294