Understanding the Drivers of Biodiversity on the Rocky Coast

Home , Austrocochlea constricta, Austrocochlea porcata, Austrolittorina, Lipophrys, Nerita atramentosa

Understanding the drivers of biodiversity on the rocky coast

Nina Schaefer

A thesis in fulfilment of the requirements for the degree of Doctor of Philosophy

Evolution and Ecology Research Centre School of Biological, Earth and Environmental Science Faculty of Science University of New South Wales

October 2018

Thesis/Dissertation Sheet

Surname/Family Name : Schaefer Given Name/s : Nina Abbreviation for degree as give in the : PhD University calendar Faculty : Science School : Biological, Earth and Environmental Sciences Thesis Title : Understanding the drivers of biodiversity on the rocky coast

Abstract 350 words maximum: (PLEASE TYPE) Urbanisation and climate change are pervasive stressors to natural ecosystems and have been linked to decreased biodiversity and ecological function. At the coastal fringe, intertidal rocky shores are threatened by urbanisation along shorelines and sea level rise, with the result that complex intertidal habitats are becoming rare and often replaced with simpler concrete structures. Communities that remain also experience altered light climates due to shading by engineered structures. To effectively manage and conserve urban intertidal ecosystems, we require a greater understanding of the drivers of diversity at micro- to macro scales. I begin by investigating the drivers of biodiversity on the small scale, where I was able to identify relationships between rock pool physical characteristics and associated biota around Sydney Harbour. Maximum width and depth, volume and height on shore were important drivers of biodiversity, but effects varied among organisms (i.e. sessile vs. mobile taxa) and between inner and outer locations of the lower estuary. I also found that the structure within rock pools can influence species abundances. Microhabitats such as overhangs were important to support grazers and rare species. I then investigated the effect of another potential driver of biodiversity common in urban intertidal areas: shading by engineered structures. Using a manipulative experiment, I found that the effect of varying levels of shading can influence taxa in different ways. Whereas high light intensities promoted greater algal cover, low light intensities supported higher abundances of mobile taxa. Finally, I worked at a larger scale and used predictive modelling to assess potential habitat loss of intertidal rocky shores due to sea level rise. I found that rocky shore communities in the Sydney area are likely to lose substantial habitat and may be considered ‘near threatened’ (under one category of the IUCN communities listing criteria) under the predicted 11.2mm/year sea level rise (i.e. scenario RCP8.5 (IPCC)).My research highlights the need to manage intertidal rocky shores at multiple scales. While conserving natural habitat should be set as priority, management strategies that incorporate features which support diversity, as identified in this thesis, can help mitigate biodiversity loss and help protect this important ecosystem.

Declaration relating to disposition of project thesis/dissertation I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International

(this is applicable to doctoral theses only).

…………………………………………………… ……………………………………..………… ……….……………… ……… Signature……… …… Witness 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 must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research.

FOR OFFICE USE Date of completion of requirements

ONLY for Award:

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 ……………………………………………......

COPYRIGHT STATEMENT ‘I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation. I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only). I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.' Signed ……………………………………………...... Date ……………………………………………......

AUTHENTICITY STATEMENT ‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.’ Signed ……………………………………………...... Date ……………………………………………......

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 student contributed greater than 50% of the content in the publication and is the “primary author”, ie. the student was responsible primarily for the planning, execution and preparation of the work for publication • The student 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 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. Name Signature Date (dd/mm/yy) Nina Schaefer

Acknowledgements

I firstly would like to thank my supervisors Emma Johnston, Katherine Dafforn and Mariana Mayer-Pinto for their support throughout my candidature. Thank you for your expertise, guidance, valuable comments and patience throughout my research and the process of writing this thesis. I feel very lucky for having such an inspiring team of women behind me.

I would also like to thank Graeme Clark for his statistical support and advice. Thank you for always having a spare minute and for making stats enjoyable. Thank you also to William Glamore and Kingsley Griffin for introducing me to the world of spatial modelling and giving me the chance to explore a whole new direction in my research. It’s been great fun.

I would like to thank Simone Birrer, James Lavender, Damon Bolton, Brendan Lanham, Aria Lee, Sally Bracewell, Shin Ushiama, Vivian Sim, James Lavender, Sebastian Vadillo Gonzalez, Rosie Steinberg, John Turnbull, Mark Browne, Ana Bugnot, Hayden Schilling, Chris Drummond, Valentin Heimhuber, Jamie Ruprecht, Alice Harrison for helping me with fieldwork, troubleshooting stats, and for teaching me how to handle new equipment and new programs. A big thank you also to all the other past and present members of the AMEE/Poore and the Marine Ecology Lab. You’ve been absolutely fantastic and made the time spent in the office so much fun. Thank you to Alan Millar for helping me with the identification of algae. I would also like to thank the many volunteers that helped me conducting my fieldwork.

Thank you to the Centre for Research on Ecological Impacts of Coastal Cities for providing access to Cape Banks and the pools and to the Evolution and Ecology Research Centre for supporting my research financially. Thank you to Stats Central for the statistical support.

Special thanks to my friends Talia Stelling-Wood, Janine Ledet, Jess Merrett and Ruby Garthwin for being a major source of support over the past years and your friendship. Thank you for going with me through the toughest fieldwork, for helping me with stats, for proof- reading drafts, but also for your support in every other aspect of my life. You are the most wonderful friends. Thank you also to Rachel Roberts for listening, for your encouraging words and for always finding a way to make me laugh. You’re an incredible friend. A very special thanks to my friends Julia Steffen and Kirsten Guse. Even though we’re scattered around the world, you were always there for me no matter what. You’re the best friends anyone could wish for.

My love and thanks to my family back home in Germany for your support from start to finish. Thank you for the ongoing supply with German food, for the Skype conversations and photos, for encouraging me so well despite such a great distance and for always believing in me. I could have not done this without you.

Table of contents

Acknowledgements ...... i Table of contents ...... ii List of Figures ...... v List of Tables ...... x Chapter 1: General introduction ...... 1 1.1 Intertidal rocky shores ...... 2 1.2 Anthropogenic disturbances ...... 4 1.2.1 Urbanisation ...... 4 1.2.2 Climate change ...... 5 1.3 Management ...... 6 1.4 Thesis outline ...... 6 Chapter 2: Size, depth and position influence the diversity and structure of rock pool communities in an urban estuary ...... 8 2.1 Abstract ...... 8 2.2 Introduction ...... 8 2.3 Materials and methods ...... 10 2.3.1 Study area and sampling design ...... 10 2.3.2 Rock pool characteristics ...... 12 2.3.3 Rock pool biota census ...... 13 2.3.4 Data analyses ...... 13 2.3.4.1 Multivariate analyses ...... 13 2.3.4.2 Univariate analyses ...... 13 2.4 Results ...... 16 2.4.1 Rock pool physical characteristics ...... 16 2.4.2 Local species pools ...... 16 2.4.3 Rock pool assemblages ...... 17 2.4.4 Relationships between rock pool characteristics and diversity ...... 19 2.4.5 Taxa richness ...... 19 2.4.6 Abundance of sessile and mobile taxa ...... 20 2.5 Discussion ...... 23 2.6 Conclusion and implications ...... 26 Chapter 3: The role of fine-scale habitat complexity and heterogeneity in species distributions on rocky shores ...... 28 3.1 Abstract ...... 28 3.2 Introduction ...... 28 3.3 Materials and methods ...... 31

ii Table of Contents iii

3.3.1 Study area ...... 31 3.3.2 Rock pool characteristics ...... 33 3.3.3 Rock pool biota census and habitat mapping ...... 34 3.3.4 Statistical analyses ...... 34 3.3.4.1 Differences between methodologies ...... 34 3.3.4.2 Comparison of the structural complexity of pools (heterogeneity and complexity) across an urban estuary ...... 34 3.3.4.3 Comparison of (i) the overall number of species and/or (ii) the abundance of mobile organisms between rock pools with and without microhabitats ...... 34 3.3.4.4 Comparison of the density of organisms between microhabitats and non- microhabitats in rock pools with microhabitats ...... 34 3.4 Results ...... 37 3.4.1 Comparing rock pools among locations ...... 37 3.4.2 Comparing mobile species abundance and richness among rock pools with and without microhabitats ...... 42 3.4.3 Mobile species distributions in rock pools with microhabitats ...... 44 3.4.4 Occurrence (presence/absence) of individual taxa in rock pools with and without microhabitats ...... 47 3.4.5 Mobile species distributions in rock pools with biogenic microhabitats ...... 49 3.5 Discussion ...... 49 3.6 Conclusion and implications ...... 52 Chapter 4: Effects of light intensity on mobile and sessile communities on intertidal rocky shores ...... 54 4.1 Abstract ...... 54 4.2 Introduction ...... 54 4.3 Materials and methods ...... 57 4.3.1 Study area ...... 57 4.3.2 Experimental set up ...... 58 4.3.3 Assessment of mobile and sessile colonisation ...... 60 4.3.4 Light measurements ...... 60 4.3.5 Statistical analyses ...... 61 4.3.5.1 Light measurements ...... 61 4.3.5.2 Biota ...... 61 4.4 Results ...... 62 4.4.1 Light transmission ...... 62 4.4.2 Number of species and abundance of mobile taxa ...... 63 4.4.2.1 Pools ...... 63 4.4.2.2 Emergent rock ...... 63 4.4.3 Cover and diversity of sessile taxa ...... 69

Table of Contents iv

4.4.3.1 Pools ...... 69 4.4.3.2 Emergent rock ...... 72 4.4.4 Sampling effort ...... 74 4.4.5 Debris ...... 75 4.5 Discussion ...... 76 4.6 Conclusion and implications ...... 78 Chapter 5: Conservation priorities for intertidal rocky shores assessed with remote sensing ...... 80 5.1 Abstract ...... 80 5.2 Introduction ...... 81 5.2.1 Ecosystem description ...... 82 5.2.1.1 Abiotic factors ...... 82 5.2.1.2 Biota ...... 82 5.3 Materials and methods ...... 83 5.3.1 Study location ...... 83 5.3.2 Habitat mapping ...... 84 5.3.3 Elevation data ...... 84 5.3.4 Data cleaning and spatial analyses ...... 85 5.3.5 Predicted sea level rise scenarios ...... 85 5.4 Results ...... 86 5.4.1 Classification with IUCN Red List of Ecosystems ...... 90 5.5 Discussion ...... 90 5.6 Conclusion ...... 92 Chapter 6: General discussion ...... 94 6.1 Management implications ...... 97 6.1.1 Ecological engineering ...... 97 6.1.2 Conservation ...... 98 6.2 Future directions and final remarks ...... 99 References ...... 100 Appendix S1 – Supplementary material for Chapter 2 ...... 116 Appendix S2 – Supplementary material for Chapter 3 ...... 131 Appendix S3 – Supplementary material for Chapter 4 ...... 135 Appendix S4 – Supplementary material for Chapter 5 ...... 143

List of Figures

Figure 2.1 Map of Sydney Harbour showing the location of the estuary on the coast of NSW, Australia. Rock pools (n = 9-13) were randomly selected for sampling at two sites located in the inner zone of Sydney Harbour (Balmain (B), Berry Island Reserve (BIR)) and at two sites in the outer zone of Sydney Harbour (Bradleys Head (BH), Delwood (D)). Inner and outer zones are indicated and re-drawn from Dafforn et al. (2012b)...... 11

Figure 2.2 Boxplots of ranges in maximum width, maximum depth, volume and height on shore between rock pools at the different sites, showing the first quartile (Q1) and third quartile (Q3) range of the data, the median and data outliers...... 16

Figure 2.3 Principal coordinates ordination of sessile (a) and mobile (b) assemblage composition in rock pools for all sampling events in the inner zone (filled triangles) and outer zone (empty triangles) of the estuary...... 18

Figure 2.4 Taxa richness distributions across rock pool size and height ranges for sessile (a, c, e, g) and mobile (b, d, f, h) taxa and predicted taxa richness (when part of best model) in both estuary zones (green), the inner zone (red) and the outer zone (blue). Predicted taxa richness for the parameters shown were averaged across other non-interactive factors that were part of the best model. Each point represents one rock pool at one sampling event. The line represents predicted species richness across the size range. Shaded areas represent standard error from this prediction. N=22 (outer zone),24 (inner zone)...... 21

Figure 2.5 Total cover of sessile taxa (a, b) and total abundance of mobile taxa (c-f) in both estuary zones (green), the inner zone (red) and the outer zone (blue). Predicted taxa richness for the parameters shown were averaged across other non-interactive factors that were part of the best model. Each point represents one rock pool at one sampling event. The line represents predicted species richness across the size range. Shaded areas represent standard error from this prediction. N=18 (16 for cover of sessile organisms) (outer zone),22 (inner zone)...... 22

Figure 2.6 Canopy cover (%) with increasing height on shore in the outer zone of the estuary. Each point represents one rock pool at one sampling event The line represents predicted species richness across the size range. Shaded areas represent standard error from this prediction. N=18...... 23

Figure 3.1 Map of Sydney showing the location of the estuary on the coast of NSW, Australia. Rock pools (n = 9-16) were sampled at two sites located in the inner zone of Sydney Harbour (Balmain, Berry Island Reserve), at two sites in the outer zone of Sydney Harbour (Bradleys Head, Delwood), as well as three coastal sites along the open coast (Bondi, Freshwater, Curl Curl)...... 32

v List of Figures vi

Figure 3.2 Heterogeneity of rock pools represented by the proportion of rock pools with no microhabitats, or one of two types of microhabitats at each location. N = 42 (Coastal), n = 21 (Outer Harbour), n = 27 (Inner Harbour)...... 39

Figure 3.3 Boxplot of the total estimated area (%) of microhabitats at each location (standardised by overall rock pool area), showing the first quartile (Q1) and third quartile (Q3) of the data, the median and data outliers)...... 40

Figure 3.4 Density of all mobile individuals (individuals/L) within rock pools with an overhang found under the overhang and the remaining area of the pool of the pools averaged across season. Bars are model predicted means and standard errors...... 45

Figure 3.5 Density of all mobile individuals (individuals/L) within rock pools with a pit found in the pit and other areas of the pools in each season. Bars are model predicted means and standard errors...... 45

Figure 3.6 Number of individuals per species in each habitat. Deep pits were pits with a depth greater than 10cm. Medium pits were pits deeper than 5cm. Samplings are the 4 samplings times (1+2: winter; 3+4: summer)...... 46

Figure 3.7 Frequency of occurrence across all samplings for chitons found in rock pools without and with microhabitats (expressed as a proportion). N is the total number of sampling times (number of rock pools without or with microhabitats × total number of samplings (4)...... 49

Figure 4.1 Map of Botany Bay showing the location of the embayment on the coast of NSW, Australia. Rock pools were sampled at one exposed site on mid-shore levels (site 1) and one sheltered site lower on shore (site 2)...... 57

Figure 4.2 Light treatments used in the experiment. (a) Full light – no plate, (b) procedural control – clear plate no UV transmission, (c) 75% transmission no UV, (d) 35% transmission no UV, (e) 15% transmission no UV, (f) full shade. Photos were taken during different stages of the experiment...... 59

Figure 4.3 Light levels (Lux) for each treatment within pools and on emergent rock at site 1 (a, b) and site 2 (c, d). Light measurements were averaged across time for rock pools and emergent rock at site 1. Abbreviations for treatments as shown in Table 4.1. Error bars are model predicted means and standard errors. *P < 0.05, **P < 0.01, ***P < 0.001...... 63

Figure 4.4 Abundances of mobile taxa among treatments and over time within pools at site 1 (a) and among treatments at site 2 (b). Abbreviations for treatments as shown in Table 4.1. Each point represents one replicate at one sampling. The line represents predicted abundance of mobile taxa over time. Shaded areas represent standard error from this prediction. Error bars are model predicted means and standard errors. *P < 0.05, **P < 0.01, ***P < 0.001. .... 64

List of Figures vii

Figure 4.5 Mean abundances of the four most abundant and other taxa per sampling time at site 1 (a) and site 2 (b) averaged across replicate pools. Abbreviations for treatments as shown in Table 4.1...... 65

Figure 4.6 Number of mobile taxa per treatment (averaged across time) (a, b) and over time (averaged across treatment) (c, d) on emergent rock at site 1 (a, c) and at site 2 (b, d). Abbreviations for treatments as shown in Table 4.1. Each point represents one replicate at one sampling. The line represents predicted numbers of mobile taxa over time. Shaded areas represent standard error from this prediction Error bars are model predicted means and standard errors. *P < 0.05, **P < 0.01, ***P < 0.001...... 66

Figure 4.7 Abundances of mobile taxa among treatments and over time on emergent rock at site 1 (a) and at site 2 (b). Abbreviations for treatments as shown in Table 4.1. Each point represents one replicate at one sampling. The line represents predicted abundances of mobile taxa over time. Shaded areas represent standard error from this prediction...... 67

Figure 4.8 Mean abundances of the four most abundant and other taxa per sampling time at site 1 (a) and site 2 (b) averaged across replicate plots on emergent rock. Abbreviations for treatments as shown in Table 4.1...... 68

Figure 4.9 Algal cover (%) among treatments and over time within pools at site 1 (a) and site 2 (b). Abbreviations for treatments as shown in Table 4.1. Each point represents one replicate at one sampling. The line represents predicted algal cover over time. Shaded areas represent standard error from this prediction...... 70

Figure 4.10 Mean algal cover (%) of the three most abundant and other taxa per sampling time at site 1 (a) and site 2 (b) averaged across replicate pools. Abbreviations for treatments as shown in Table 4.1...... 71

Figure 4.11 Algal cover (%) among treatments and over time on emergent rock at site 1 (a) and site 2 (b). Abbreviations for treatments as shown in Table 4.1. Each point represents one replicate at one sampling. The line represents predicted algal cover over time. Shaded areas represent standard error from this prediction...... 72

Figure 4.12 Mean algal cover (%) of the three most abundant and other taxa per sampling time at site 1 (a) and site 2 (b) averaged across replicate plots on emergent rock. Abbreviations for treatments as shown in Table 4.1...... 73

Figure 4.13 Cover of live sessile taxa (%) at site 1 for each treatment for both sampling methods (assessing the entire rock pool (base + wall) versus base only (base)). Error bars are predicted means and standard errors...... 74

List of Figures viii

Figure 4.14 Cover of live sessile taxa (%) at site 2 for both sampling methods (assessing the entire rock pool (base + wall) versus base only (base)). Error bars are predicted means and standard errors...... 74

Figure 4.15 Debris cover (%) among treatments and over time within pools at site 1 (a) and site 2 (b). Abbreviations for treatments as shown in Table 4.1. Each point represents one replicate at one sampling. The line represents predicted algal cover over time. Shaded areas represent standard error from this prediction...... 746

Figure 5.1 Map pf the coastal area (~210 km) that has been assessed for rocky shoreline (highlighted in blue) on the coast of NSW, Australia. Approximate extent of the Hawkesbury Shelf Marine Bioregion is indicated. Rocky shores that were additionally assessed are highlighted in green (BH: Bradleys Head, CB: Cape Banks, D: Delwood, F: Freshwater)...... 84

Figure 5.2 Remaining area (%) of all rocky shores combined (a) in each year for the next 50 years and (b) in 50 years...... 87

Figure 5.3 Remaining area (%) in 50 years at (a) Delwood, (b) Bradleys Head, (c) Freshwater and (d) Cape Banks...... 88

Figure 5.4 Height (m) above LAT of each LIDAR point at each site of (a-d) the whole area, (e-h) an exemplary area of rocky shoreline and (i-l) gain/loss of rocky shoreline under the predicted climate change (using median values) scenarios in 50 years. Delwood: a, e, i; Bradleys Head: b, f, j; Freshwater: c, g, k, Cape Banks: d, h, l. Untransformed datasets with greater point densities were used for plotting and points higher than 60m above LAT were excluded. Exemplary areas of rocky shoreline were restricted to points within the shapefile...... 89

List of Tables

Table 2.1 Summary of the results for all response variables and parameters examined. Results refer to changes in the response variable (i.e. richness, abundance) relative to increases in width, depth, volume or height on shore. X = factor not part of best model.  = increased.  = decreased. -- = no change...... 19

Table 3.1 Physical features measured in each rock pool and the definitions used for each feature...... 33

Table 3.2 Fixed factors included in each model. Abbreviations are as follows: P = Present, A = Absent, W = Winter, S = Summer, C = Coastal, OH = Outer Harbour, IH = Inner Harbour...... 35

Table 3.3 Rock pool volumes (Litre) at each location (Coastal, Outer and Inner Harbour). Mean, minimum and maximum values are presented for each site...... 38

Table 3.4 Results of the overall area (complexity) of microhabitats, pits and overhangs on the total number and abundance of all mobile species as well as on the abundance the three most common species. N.s.= non-significant. Results were deemed variable when contrasting patterns were found. Results were found in all of the 5 subsamplings unless stated otherwise...... 41

Table 3.5 Results of the presence of a microhabitat, pits and overhangs on the total number and abundance of all mobile species as well as on the abundance of the three most common species. N.s. = non-significant. W = Winter. S = Summer. * = magnitude of difference was greater in summer than in winter. Results were deemed variable when contrasting patterns were found. Results were found in all of the 5 subsamplings unless stated otherwise...... 43

Table 3.6 Taxa present (shaded bar) or absent (white bar) in rock pool microhabitats. Categories included none (N), one or two types of microhabitats (pits (P) and overhangs (OH)). Taxa that only occurred in a particular microhabitat within rock pools are indicated with *. Taxa that only occurred in rock pools with a microhabitat, but that, within these pools, were not found in the microhabitat area at the time of sampling are indicated with **. Taxa that only occurred in rock pools without any microhabitats are indicated with ***...... 48

Table 4.1 Estimated light transmission and experimental set up for each treatment (estimates are based on data from manufacturers). Abbreviations for treatments are shown in brackets...... 59

Table 5.1 Median values and likely ranges for projections of rate of global mean sea level rise (GMSLR) (mm/year) (Church et al., 2013) as well as projections for sea level rise (SLR) in 50 years (mm) for the four RCP scenarios used for analyses...... 86

Chapter 1

General introduction

Biodiversity underpins ecosystem functioning, which is essential for the ongoing provision of the ecosystem services on which humans rely (Hooper et al., 2005, Loreau et al., 2001). Biologically diverse systems are also considered to be more resilient to environmental disturbances and invasions (McCann, 2000). Anthropogenic activity, however, has resulted in the degradation of habitat and overexploitation of environmental resources (Vitousek et al., 1997) and has led to unprecedented levels of biodiversity loss (Johnson et al., 2017). Additionally, anthropogenically induced climate change is expected to become a growing contributor to future species loss (Johnson et al., 2017). To secure ecosystem functioning and services, it is therefore critical to incorporate strategies for supporting ecosystem resilience into management and conservation (Elmqvist et al., 2003, Oliver et al., 2015). Understanding the drivers of biodiversity, in natural and managed systems, and the mechanisms by which disturbances affect biodiversity is essential to the development of appropriate management and conservation strategies.

There is much evidence and theory to suggest that habitats with greater complexity will support greater biological diversity than homogeneous habitats (MacArthur and MacArthur, 1961, McGuinness and Underwood, 1986, Tews et al., 2004). This effect is often attributed to more complex habitats providing a greater range of resources by creating a variety of microhabitats, which can reduce interspecific competition (Klein et al., 2011), predator-prey interactions (Grabowski, 2004) and provide refugia from environmental stressors (Scheffers et al., 2014). Structural complexity can be provided by the substratum itself, but also by biogenic structures such as plants and oysters (Sebens, 1991). Maintaining habitat complexity within ecosystems is thus an important mechanism to sustain biological diversity. However, the relationship between diversity and complexity can vary depending on spatial scale (Matias et al., 2010, Tews et al., 2004) and according to the type of species and communities (Lassau et al., 2005).

Humans are a major force driving ecological processes (Corlett, 2015, Vitousek et al., 1997) and have already severely impacted more than 40% of the marine ecosystems (Halpern et al., 2008). Impacts concentrated in coastal ecosystems include: anthropogenically induced climate change, pollution, habitat loss, and fishing (Halpern et al., 2008). Climate change, in particular, has altered many processes in the marine environment, such as food web dynamics and productivity, and is driving global changes in species distributions as well as disease outbreaks (Hoegh-Guldberg and Bruno, 2010). Anthropogenic disturbances are however not evenly distributed among marine

1 Chapter 1 2

ecosystems and are strongest along coastal areas (Halpern et al., 2008), where the progressive expansion of populations has resulted in habitat degradation, species loss and the facilitation of species invasions (Lotze et al., 2006).

The overarching aim of this thesis is to determine the macro- and meso- scale drivers of diversity in a valuable natural ecosystem; intertidal rocky shores, and threats to this habitat form current climate change projections. Below, I introduce the ecosystem as well as its major threats, followed by a brief description of my thesis chapters.

1.1 Intertidal rocky shores

Intertidal rocky shores are located between the land and sea, in an area defined by the low and high spring water marks of the tidal cycle (Menge and Branch, 2001). They are the most common coastal habitat worldwide (Thompson et al., 2002) and, due to their easy accessibility, are one of the most intensely studied systems in the world (Paine et al., 1994). Intertidal rocky shores are composed of several types of microhabitats, which are related to the gradients in environmental conditions that occur in intertidal areas and the structural complexity within a rocky shore.

The environment of intertidal rocky shores is dictated by two main gradients: a horizontal and a vertical gradient (Raffaelli and Hawkins 1996). The vertical gradient is determined by the tidal cycle, which results in alternating emersion and submersion, with the number and amplitude of tides varying among times and locations (Garbary, 2007, Raffaelli and Hawkins, 2012). On an intertidal rocky shore, emersion times during low tide increase with increasing height on the shore, and thus the high shore experiences extended exposure to aerial conditions and stressors such as temperature, desiccation and irradiance (Garbary, 2007, Raffaelli and Hawkins, 2012). On the other hand, the horizontal gradient is related to exposure to wave action, with lowest exposure in sheltered bays and greatest exposure at headlands (Denny and Wethey, 2001, Raffaelli and Hawkins, 2012). Greater wave action can also increase the vertical gradient and result in extended submersion times (Druehl and Green, 1982), but also increases the probability of dislodgement or removal of species from substratum due to greater hydrodynamic forces (Blanchette, 1997, Denny, 1985).

The variation of the physical environment due to height on shore is further amplified by the habitat structure of the substratum. Habitat structure comprises of complexity, defined as the number of structural components in a habitat, and heterogeneity, which refers to the types of structural components (i.e. microhabitats) within a habitat (Bell et al., 2012). Rocky shores support a variety of microhabitats including rock pools and crevices, which provide additional variability in the physical environmental conditions present on the emergent rock bed (Garrity, 1984). The range of suitable habitat for organisms across the shore is limited by their tolerance to these

Chapter 1 3

environmental stress gradients (the environmental ‘niche’), but also influenced by biological interactions such as competition and predation that can limit the range which an organism can effectively inhabit – the ecological ‘niche’ (Connell, 1961, Paine, 1971, Underwood, 1978, Underwood, 1980). This patchiness in environmental conditions together with biological interactions results in multifaceted patterns of species distributions even on small spatial scales (Underwood and Chapman, 1996, Araújo et al., 2005).

One of the most ubiquitous microhabitats on intertidal rocky shores are rock pools. Rock pools provide a constantly submerged habitat, and thus a less physically variable environment compared to emergent rock (Metaxas and Scheibling, 1993, Garrity, 1984). Additionally, rock pools can further provide a variety of microhabitats. Rock pools have variations in temperature and salinity even on small spatial-scales (Morris and Taylor, 1983), which increases the availability of niches. Therefore, they often support greater diversity compared to emergent rock (Firth et al., 2014, Firth et al., 2013), although this effect has been found to vary spatially (Bugnot et al., 2018, Firth et al., 2014, Firth et al., 2013). Additionally, they often support unique species not found on emergent rock (Chapman and Bulleri, 2003). A recent metanalysis has shown that a variety of rock pool characteristics, such as width, depth or volume can be important drivers of diversity, although their effect on diversity was found to vary with taxa-identity and location (Bugnot et al., 2018).

The importance of rock pools for intertidal organisms has become increasingly evident for intertidal communities on artificial substrates. A recent meta-analysis found that artificial rock pools significantly increased the number of benthic taxa compared to emergent surfaces on artificial structures like seawalls (Bugnot et al., 2018). In response to this, a range of approaches are currently being taken to create artificial rock pools of different sizes in seawalls along urbanised settings (Browne and Chapman, 2014, Chapman and Blockley, 2009). However, there is a lack of knowledge about the drivers of diversity of natural rock pool communities in these urbanised settings on the large- and small-scale, which could be used as a guide for these designs and interventions.

In the Sydney area, intertidal rocky shores mostly consist of sandstone and are generally gently sloping, although some vertical rocky shores exist (Bulleri et al., 2005, Chapman, 2003, Johnston et al., 2015). Within Sydney Harbour, at least 162 taxa of benthic macrofauna and -flora can be found on natural rocky shores (Mayer-Pinto et al., 2018). Rocky shores particularly host high abundances of mobile species, including gastropods and polychaetes when compared to artificial intertidal habitats (Mayer-Pinto et al., 2018). Although there is great spatial variability in the structure of these communities (Underwood and Chapman, 1996), vertical patterns of zonation can be found. The lower intertidal is charatcerized by foliose algae, which is followed by a zone

Chapter 1 4

of polychaetes, barnacles, encrusting algae and grazing molluscs. The upper zone of the intertidal is characterized by high abundances of littorinid gastropods (Underwood, 1981).

1.2 Anthropogenic disturbances

Rocky shores are located in a region subject to the ‘coastal squeeze’ (Pontee, 2013), being vulnerable to impacts originated from both the land and the sea. In this thesis, I will focus on urbanisation and climate change, two of the most pervasive stressors threatening the diversity of intertidal rocky shores and their provision of services.

1.2.1 Urbanisation

Coastal populations around the world are rapidly increasing and are expected to exceed 1.4 billion people by 2060 (Neumann et al., 2015). The development of coastal infrastructure for residential, economic and recreational purposes and the associated demand for protective infrastructure due to increased exposure to storms and sea level rise has resulted in the transformations of coastal habitats (Bulleri and Chapman, 2010), with already up to 70% of some coastlines being modified (Dafforn et al., 2015a). A common modification in coastal environments is the replacement of natural habitat by homogeneous built ‘hard’ or ‘armoured’ infrastructure, causing the loss and fragmentation of natural habitats. This results in shores being further apart from one another in an already patchy environment (Goodsell et al., 2007). Artificial shorelines usually comprise of a smaller area than natural rocky shores, as horizontal rocky shores are condensed to a small fraction of the original extent (Perkins et al., 2015). Additionally, artificial structures provide few of the microhabitats usually found on rocky shores, such as rock pools or crevices (Chapman, 2003). Given these differences, it is becoming increasingly evident that these artificial intertidal structures are not surrogates for the habitat they have replaced. Artificial habitats such as seawalls are often characterized by lower diversity (Chapman, 2003, Firth et al., 2013, Moschella et al., 2005) and altered community structure (Bulleri and Chapman, 2004, Chapman, 2003, Lai et al., 2018) compared to natural habitats, which is attributed to their reduced complexity.

Besides the modification of the habitat itself, urbanised areas are often associated with further changes in environmental conditions. Urban waterways are often associated with more contamination such as metals and polycyclic aromatic hydrocarbons (PAHs) (Birch and Taylor, 1999, Dafforn et al., 2012b), which are known to reduce diversity across a range of marine communities (Johnston and Roberts, 2009). The expansion of coastal populations is likely to continue and thus will cause further disturbance and poses a key challenge to intertidal communities. The effects of urbanisation can vary along spatial scales and thus affect

Chapter 1 5

communities to a different extent. Yet, the effect of urbanisation is often only tested in a ‘urbanised’ versus ‘unurbanised’ context. The assessment of communities across different levels of urbanisation can help determine whether these approaches are sufficient to assess the potential impact of urbanisation on communities.

Intertidal communities with close proximity to other artificial structures along the foreshore, such as high-rise buildings, wharves and jetties may be further affected by altered light regimes and/or increased shading. Intertidal communities in shaded areas of both natural and artificial structures were found to differ from those exposed to natural light regimes and were characterized by reduced macroalgae cover and higher abundances of sessile invertebrates (Blockley, 2007, Pardal‐Souza et al., 2017). However, little is known about how different levels of shading affect both the sessile and mobile intertidal communities. Understanding how intertidal communities as a whole are affected by different levels of shading can help improve potential mitigation strategies for both types of organisms.

1.2.2 Climate change

Climate change is an inevitable consequence of anthropogenic carbon dioxide emissions into the atmosphere and is fundamentally altering the planet’s climate system (IPCC, 2013). Marine systems are the planet’s primary heat sink (Levitus et al., 2012), are most affected and are already experiencing increased temperatures, lowered PH and extreme events such as heatwaves (Doney et al., 2012, Hoegh-Guldberg and Bruno, 2010, Wernberg et al., 2013). Continuing use of fossil fuels will result in further increases in air and water temperatures, ocean acidification, sea level rise and an increase the frequency and intensity of extreme events (Doney et al., 2009, IPCC, 2013, Meehl and Tebaldi, 2004).

Due to their location along the margins of the sea, intertidal rocky shores are particularly vulnerable climate change via its effect on sea level rise. Future sea level rise will permanently inundate lower parts of rocky shores. This will result in a vertical, upward migration of intertidal species, which has already been documented for a variety of marine organisms. For instance, Wanless (1982) has shown that the upper vertical limit of oysters and barnacles rose by 15 cm in line with sea level rise at this location. Similarly, intertidal surveys in Victoria, Australia, have shown that the tubeworm Galeloaria caespitosa showed higher upper limits on newly built infrastructure, indicating that it could be a potential indicator of sea level rise (Bird, 1988).However, sea level rise can also reduce habitat availability for intertidal organisms in places where they cannot retreat further upshore. This habitat loss can result in local extinctions and altered community structures on rocky shores (Jackson and McIlvenny, 2011). Yet, little is known about the potential extent of this risk on intertidal rocky shores. Quantifying the vulnerability of

Chapter 1 6

this ecosystem to loss of habitat via sea level rise can help to predict potential ecological outcomes so that management and conservation practices of this biologically diverse habitat can be adapted.

1.3 Management

Accelerating coastal urbanisation and increasing risk of climate change threats such as sea level rise and storms has resulted in increased foreshore developments (Firth et al., 2016). In light of the growing awareness of the negative impacts of these structures, ecological engineering designs to mitigate potential impacts of artificial structures on biodiversity are being explored. Ecological engineering incorporates ecological principles in the design of human infrastructure (Bergen et al., 2001) and is becoming an emerging research field in the marine environment. One of the frequent aims of such designs is to increase habitat complexity of otherwise homogenous built infrastructure. This has been done through the addition of microhabitats such as rock pools, crevices, and pits, and the addition of biogenic habitat formers (e.g. barnacles, bivalves), all of which have successfully enhanced diversity compared to a control, with rock pools being the most successful intervention to date (Strain et al., 2018). Advancements in technology, such as 3D printers, now allow for manipulations of designs at much finer scales. The design of these eco- engineered structures however requires basic experimental research such as surveys and experiments of natural systems, to parametrise and improve designs and maximise their positive effects (Bergen et al., 2001). Since artificial structures are mainly built in areas of heavy urbanisation, surveys of natural communities in such locations, and on a variety of-scales, have the potential to significantly improve the ecological value of a region.

1.4 Thesis outline

The main objective of this thesis is to advance the scientific knowledge of the ecology of urban intertidal rocky shores in the Sydney region. This thesis is structured in 4 data chapters which are briefly summarized below.

In the first chapter, I describe the communities of intertidal rock pools in an urban estuary. I conducted a large-scale survey and examined the influence of large-scale differences (size parameters) among rock pools (Chapter 2). Observations of different distributions of mobile organisms within rock pools have led me to further investigate what the influence of fine-scale habitat characteristics on the diversity of mobile species is (Chapter 3). Because intertidal communities in urbanised environments are additionally often affected by shading from surrounding buildings or adjoining artificial structures such as jetties and wharves, I used a manipulative experiment to investigate the effect of different light intensities on the colonisation

Chapter 1 7

of intertidal rock pools and emergent rock habitats (Chapter 4). Global climate change will additionally lead to sea level rise, which will potentially affect the size of entire intertidal rocky shores available. I used a modelling approach using novel remote sensing technology, ‘LIght Detection And Ranging (LIDAR)’, to predict the effect of different sea level rise scenarios on the availability of intertidal rocky shores in Sydney (Chapter 5). Finally, I consider the implications of my findings, including suggestions for ecological engineering designs and strategies for the mitigation of climate change impacts on intertidal rocky shores (Chapter 6).

Chapter 2-5 of this thesis have been prepared in form of stand-alone manuscripts for publication in peer-reviewed journals and may therefore be repetitive in sections. In these chapters, “we” is used instead of “I” to acknowledge the contributions of the co-authors to the manuscripts. All references can be found in a combined list at the end of the thesis. Chapter 2, “Size, depth and position influence the diversity and structure of urban rock pool communities” has been submitted for publication to Marine and Freshwater Research and is recommended for publication following minor revisions. Revisions have been submitted. The content of the manuscript has been edited and formatted from the submission to fit the structure of this thesis.

Chapter 2

Size, depth and position influence the diversity and structure of rock pool communities in an urban estuary

2.1 Abstract

Rock pools provide a range of ecological niches that can support diverse assemblages on rocky shores. As intertidal shores are increasingly lost to developments, understanding the drivers of diversity in rock pools is important for the conservation and construction of these key habitats. We investigated relationships between physical characteristics of rock pools and their biota in an urban estuary. We sampled the biota every six weeks for one year at sites in the inner and outer zones of Sydney Harbour. In the well-flushed and exposed outer zone, sessile and mobile taxa richness was positively related to rock pool width, whereas only mobile taxa richness was related to depth and volume. In the more urbanised and less exposed inner zone, mobile taxa richness was positively related to rock pool width and volume. In both zones, sessile taxa richness decreased with increasing height on shore. The results suggest that the biodiversity of intertidal rock pools varies depending on their position in Sydney Harbour and the available species pool. Therefore, restoration efforts should consider rock pool size parameters and local environmental conditions, including location, so designs can be optimised to maximise species diversity in these pools.

2.2 Introduction

Rocky shores are biologically diverse habitats, but have been increasingly degraded by coastal development (Goodsell et al., 2007, Thompson et al., 2002). Rocky shores are characterised by a tidal regime that results in alternating emersion and submersion of the intertidal zone, which can subject organisms to a high level of environmental stress, e.g. heat and desiccation (Dahlhoff et al., 2001, Harley, 2008, Helmuth and Hofmann, 2001, Raffaelli and Hawkins, 2012, Thompson et al., 2002). However, the spatial heterogeneity of rocky shores provides an array of microhabitats that can reduce some of these natural stressors. Rock pools, for example, provide areas of constant submersion, where environmental conditions are comparatively stable (Garrity, 1984, Metaxas and Scheibling, 1993); serving as an important refuge for intertidal organisms and

8 Chapter 2 9

consequently supporting higher diversity compared to surrounding emergent rock (Firth et al., 2014, Firth et al., 2013, but see Bugnot et al. 2018). However, population growth, land reclamation and increasing construction on the waterline has fragmented foreshores and reduced natural habitat, including intertidal rocky shores and rock pools (Airoldi and Beck, 2007, Mayer- Pinto et al., 2015). Understanding the drivers of diversity in intertidal rock pools is therefore important to guide conservation and restoration efforts in these urbanised settings.

Rock pools are known to support a variety of species and taxa that are not found on emergent rock (e.g. fish, ascidians, bryozoans, hydroids and sponges (Evans et al., 2016)), which has been linked to the physical characteristics of rock pools (Firth et al., 2014, Firth et al., 2013). The degree to which pools can support high diversity depends, however, on their capacity to ameliorate natural stressors, which is influenced by factors including their size and position on the shore . For example, rock pools higher on shore are exposed to aerial temperatures for a longer period of time, leading to greater fluctuations of water characteristics, including temperature, salinity and pH (Huggett and Griffiths, 1986, Morris and Taylor, 1983). Similarly, it is hypothesized that canopy cover within rock pools mitigates environmental stressors such as temperature (Bertocci et al., 2014). Differences in the physical and biological characteristics of rock pools are therefore likely to have direct consequences on the abundance and composition of resident organisms and overall local diversity. For example, the effect of rock pool depth on the biodiversity of ecological assemblages varies with location, height on shore (Firth et al., 2014), and species identity (Astles, 1993, Firth et al., 2014). Furthermore, increasing rock pool diameter has been related to increased species richness, but not to taxa density (Underwood and Skilleter, 1996). Links between physical parameters of rock pools and diversity therefore appear to be location-dependent. Despite this extensive research on rock pools at various locations, to our knowledge, no study has examined the relationship of physical parameters and diversity of natural rock pools within an urban estuary.

In urban estuaries, intertidal rocky shorelines are exposed to a variety of stressors associated with human activities, such as chemical stressors from sources such as industry (Birch and Taylor, 1999) and urban run-off (Hatje et al., 2001, Stark, 1998). The influence of anthropogenic stressors has been shown to vary within estuaries (Dafforn et al., 2012b, Feng et al., 1998). In New South Wales (Australia) for example, inner zones of estuaries are often more developed and exposed to high levels of contaminants (Dafforn et al., 2012b). This variation in anthropogenic disturbances can also impact the diversity of local biota (Bishop et al., 2002, Johnston and Roberts, 2009). In addition, natural environmental conditions also vary within estuaries, with greater flushing and exposure to wave action often occurring closer to the mouth of the estuary than further upstream (Dafforn et al., 2012b, Das et al., 2000). The multitude of stressors and variations in environmental parameters can therefore result in unique environmental conditions and variations

Chapter 2 10

in species pools within different parts of an estuary. Understanding the effects of varying environmental conditions on diversity along estuarine gradients will assist in the identification of suitable sites for rocky shore restoration. Given the ecological importance of rock pools in intertidal zones, as well as their potential to buffer natural environmental stressors, artificial mimics of pools are becoming a popular choice to ecologically improve marine infrastructure (Browne and Chapman, 2014, Evans et al., 2016). Nevertheless, such designs do not take into account how manipulations of the physical parameters of rock pools can promote species diversity. A deeper understanding of how intertidal pools affect diversity through time, taking into consideration location-specific conditions, will further enhance designs of artificial pools so these can promote greater local biodiversity (Bergen et al., 2001).

Here, we surveyed rock pools over space and time within Sydney Harbour, Australia. We assessed the abundance and diversity of sessile and mobile taxa from a total of 46 rock pools in the inner and outer zones of the estuary. We hypothesised that the location of pool within the estuary would affect the overall diversity found in the pools as well as potential relationships between diversity and the physical characteristics of rock pools. We investigated how diversity in natural rock pools differed with a variety of factors, including height on shore, size parameters of the pools and algal canopy cover within pools. We then tested whether patterns found varied within the Harbour. Specifically, we hypothesised the following: 1) Wider and deeper rock pools with a greater water volume will support greater diversity (number and abundance/cover of taxa) compared to rock pools that are narrow, shallow and with smaller water volumes. 2) Pools lower on the shore will support greater diversity than pools higher on shore. 3) Canopy cover will positively affect diversity within the pools. 4) Rock pools in the inner harbour will have lower overall diversity than rock pools in the outer harbour potentially because of a combination of natural and anthropogenic environmental differences. 5) The relationships between diversity and these physical characteristics of rock pools will vary depending on the location of pools within the estuary.

2.3 Materials and methods

2.3.1 Study area and sampling design

Sydney Harbour (Fig. 2.1) is an urbanised estuary, centred in the city of Sydney with a population of ~ 5 million (Johnston et al., 2015) and is located on the central coast of New South Wales, Australia (33° 51’ S, 126 151° 14’ E).

Chapter 2 11

The harbour has a history of industrialisation and urbanisation (Birch, 2007), with artificial structures accounting for more than 50% of the intertidal habitat (Chapman and Bulleri, 2003). Sampling was done at two sites (Bradleys Head and Delwood) in the outer zone of the harbour (hereafter referred to as outer zone) and at two sites (Balmain and Berry Island Reserve) in the inner zone of the harbour (here after referred as inner zone). The divisions in zones are based on qualitative observations and physical conditions (Dafforn et al., 2012b) (Fig. 1). The outer zone of Sydney Harbour is generally less urbanised, less polluted and has greater flushing, greater exposure to wave action and greater tidal velocities compared to the inner zone of the Harbour (Birch and Taylor, 1999, Dafforn et al., 2012b, Das et al., 2000). The entire sampling area had similar salinity and temperature conditions during sampling periods (Table S1.1), although seasonal differences in these factors between zones can occur (Johnston et al., 2015).

Sites chosen for this study were natural rocky shores with rock pools representing a range of sizes (see Figure S1.1 for detailed information of rock pool size and height ranges). At each site, a total

Chapter 2 12

number of 18 rock pools was haphazardly selected and marked with a tag. We then excluded rock pools of less than 100 mL volume, greater than 700 mm width and those located on the high shore (≥1.3m height on shore) to ensure a representative sample of the size range of most rock pools found at the sites as well as adequate replication within each size class. Additionally, some of the selected pools dried out during the period of the surveys or were located in the high intertidal and were consequently excluded from analyses. Final numbers of rock pools sampled at each site were therefore as follows: Balmain: 11; Bradleys Head: 9; Berry Island Reserve: 13; Delwood: 13.

2.3.2 Rock pool characteristics

Parameters related to size of rock pools, including maximum pool width, maximum pool depth and volume were recorded for each sampled pool, at the start of the study. Maximum width was measured as the maximum distance across the surface of the rock pool whereas maximum depth was measured as the deepest point within a rock pool. The majority of rock pools sampled has a round or oval shape (personal observations) and as such, width was assumed to be an appropriate measure of the maximum pool size. Volume was used as an additional predictor variable to max. width and max. depth, as it was assumed that this reflects the size/shape of the pool best. Volume was measured by emptying the rock pool of all water using measuring cylinders and syringes and recording the amount of water in each pool. Height on the shore of each rock pool was calculated using the time of exposure of the pool as well as the height and time of low and high tide (tidal heights referred to zero on the Fort Denison (ZFD) Tide Gauge (= -0.925 Australian Height Datum). This was done over four consecutive days (04 - 07.10.2016) with similar low tides, with one site sampled per day. A mean decrease in water level per minute was calculated to know the height of the water level for any time between high and low tide on each day: 훥퐻 퐻 퐻 퐿 = 푥̅ 훥푇퐻푇퐿

ΔHHHL: Difference in water levels between high and low tide; HL: Height of water level at low tide on the day of survey; ΔTHTL: Time difference between low and high tide

We then calculated the water level at the time a rock pool was exposed using the mean decrease in water level, which equals the height of the rock pool on the shore:

훥푇퐻푇푅푒푥푝 × 푥̅ = 퐻푅

ΔTHTL: Time difference between high tide and time of rock pool exposure; HR: height of rock pool

Algal canopy cover was estimated as percentage cover of the water surface when viewed directly from above.

Chapter 2 13

2.3.3 Rock pool biota census

Rock pool biota was surveyed during low tide over four consecutive days every 6 - 8 weeks (twice per season) between June 2015 and May 2016. For rock pools smaller than 200 mm width (3 at Bradleys Head, one at each other site), all mobile and sessile taxa present in the pool were recorded. For rock pools greater than 200 mm width, the composition and abundance of mobile and sessile species within each pool was quantified using 100 × 100 mm quadrats (Table S1.2). Each quadrat had regularly-spaced 25 point-intercepts and were haphazardly positioned within the pool. The number of quadrats assessed per pool was relative to pool size, with one quadrat assessed per 100 mm of the total width of the pool. Differences in sampling techniques used for small pools precluded the inclusion of these pools in multivariate analyses and analyses of total cover of sessile organisms and total abundance of mobile organisms. All mobile organisms within each quadrat were identified and counted. Faster moving species were counted even if they left the quadrat while being assessed. Percentage cover of sessile organisms was assessed using the point-intercepts within the quadrat. Taxa present in the pool, but not under a point (or within a quadrat), were noted down to be included in diversity analyses. When emergent algae were present (e.g. Hormosira banksii, Sargassum spp. and Codium spp.), they were accounted for in canopy cover estimates and held aside in order to quantify the primary cover in pools. All organisms were identified to the lowest possible taxonomic level (i.e. species or morphospecies, except for some sessile species which were grouped into informal categories (e.g. polychaetes of the genus ‘Hydroides’). Due to the difficulty of separating the various red algae in situ, these taxa were combined for statistical analysis and were informally referred to red fuzz. Turfing algae represent several types of micro-and macroalgae which share an extensive low-lying morphology (Connell et al., 2014).

2.3.4 Data analyses

2.3.4.1 Multivariate analyses

Permutational analysis of variance (PERMANOVA) was used to test for differences in relative abundance and composition of species/taxa between inner and outer zones in the Harbour. Only organisms that occurred in the quadrats were analysed. Point intersections of the quadrat where the rock pool assemblage composition could not be accurately sampled because of cover by emergent algae (Hormosira banksii, Sargassum spp. and Codium spp.) were classified as “unknown”. Analyses included dead organisms as well as bare rock, debris and sand. Zone (i.e. inner versus outer) was a fixed factor and sites (nested within zone), rock pools (nested within zone and site) and sampling events were random factors. As multiple rock pools were found to have no mobile taxa when sampled, a value of 1 was added to the abundance of each taxon to

Chapter 2 14

allow data standardisation by total sample abundance (Anderson, 2001). An analysis of covariance (ANCOVA) was done to test whether differences in the maximum width, maximum depth and volume of pools influenced the results. Height on shore was excluded from the ANCOVA as it significantly varied between the zones. The highest order of interaction was excluded from the analyses. Analyses used Bray-Curtis similarities from square root transformed data, with significance determined from 9999 permutations of the data. A test for homogeneity of dispersions among zones (PERMDISP) was performed using centroids of 9999 permutations.

Principal coordinates ordination was used to visualise similarities between assemblages at different zones. A Pearson correlation of 0.4 was used to identify which species were most correlated to each zone. SIMPER analysis was used to determine which species were most influential in causing differences among zones and among seasons. Due to the different sampling methods used in pools < 200 mm, those were excluded from the analyses. Two rock pools with dense canopy cover that prevented full census of primary cover were excluded from the analyses for sessile assemblages. Therefore, the total number of pools analysed from each site were: Balmain: 10; Bradleys Head: 6 (4 for cover of sessile organisms); Berry Island Reserve: 12; Delwood: 12. All multivariate analyses were done using PRIMER 6 and Permanova + add-on (PRIMER-E, Plymouth Marine Laboratory, UK) (Anderson, 2001).

2.3.4.2 Univariate analyses

To investigate how the total number of sessile and mobile species in pools (taxa richness), abundance of mobile organisms (total number of individuals) and sessile percent cover (live organisms only) varied in relation to rock pool characteristics we used generalized linear mixed effect models (GLMM) and backward model selection (as per Zuur et al., 2009). The full model (below) included interactions between the categorical factors zone (inner, outer) and continuous factors width, depth, volume and height on shore. Zone and height on shore significantly correlated, but both terms were kept in the model to test whether height on shore further influenced the variables tested even after accounting for differences between zones. Since, the number of quadrats sampled in each rock pool was highly correlated to pool size (r = 0.98) due to the sampling design, no additional offset was needed to account for varying sampling effort. Sites, rock pools and sampling events were included as categorical random factors to account for repeated sampling of the same rock pools over time and potential spatial autocorrelation of rock pools within sites (Zuur et al., 2009). A Poisson distribution was used in all models for taxa richness and for total abundance of mobile taxa. For cover of sessile organisms (alive organisms only) we assumed a binomial distribution using the cbind function, with wins (percent cover of

Chapter 2 15

organisms) and losses (100%-percent cover of organisms). Distributions were chosen based on examination of residuals vs. fitted plots, i.e. the best fit to the data. The full model was:

Response variable ~ Zone*Width + Zone*Depth + Zone*Volume + Zone*Height on shore + (1|Site) + (1|Rockpool) + (1|Sampling Event)

The full model was then simplified by sequentially removing non-significant variables until all remaining variables were significant, starting with non-significant interactions. Where interactions were removed, factors were reincluded in the model as main factors. Akaike’s Information Criterion (AIC), which measures goodness of fit and model complexity (Zuur et al., 2009), was used to determine the best model. Delta AIC was calculated for each model. Models within 2 AIC values of the best model (lowest AIC) perform similar to the best model (Burnham and Anderson, 2004). Therefore, the model that best predicted ecological patterns was selected based on two criteria; 1) it either had the lowest AIC or (2) was within 2 AIC of the model with the lowest AIC and included the lowest number of predictor variables. Only organisms found within quadrats were included in the analyses of total cover of sessile organisms and total abundance of mobile taxa. Similarly to the multivariate analyses, pools with dense canopy cover that prevented full census of primary cover were excluded from the analyses of cover of sessile organisms. Therefore, total number of pools analysed in each site were the same as those used in the multivariate analyses. Algal canopy cover only occurred on rock pools in the outer zone, and therefore model selection investigating the influence of rock pool characteristics on algal canopy was performed for sites in the outer zone only. The model was constructed as described above without interactions with sites (see full model below). The number of analysed rock pools was 6 for Bradleys Head and 12 for Delwood.

Canopy cover ~ Width + Depth + Volume + Height on shore + (1|Site) + (1|Rockpool) + (1|Sampling Event)

To investigate the predictive capacity of algal canopy cover itself on taxa richness, totalcover of sessile organisms and total abundance of mobile organisms in the outer zone, model selection was performed as above with canopy cover as an additional predictor.

Predictions were calculated with matrix multiplication of coefficients and the model matrix. Assumptions of homogeneity and normality of all tests were checked by plots of the residuals versus fitted values. For count data, we additionally tested for overdispersion (Zuur et al., 2009). A dispersion value of < 2 was considered not overdispersed. Tukey post-hoc test was used to explore significant pairwise comparisons. GLMMs were performed using the lme4 package (Bates et al., 2015). Univariate statistical analyses were conducted in R (Version 3.4.3)

Chapter 2 16

2.4 Results

2.4.1 Rock pool physical characteristics

Sampled rock pools varied in size and height on shore (Fig. 2.2). Maximum rock pool width varied among sites and ranged from a minimum of 100 mm at Bradleys Head to a maximum of 556 mm at Berry Island Reserve. Maximum depth of pools ranged from a minimum of 30 mm at Berry Island to a maximum of 149 mm at Balmain, whereas Bradleys Head had the biggest variation of volume of pools (from 0.089 L to 14.6 L). Height on shore of each pool also varied among sites, ranging from a minimum vertical height of 0.82 m above ZFD to 1.3 m above ZFD. Rock pools in both zones had a similar range of sizes, but pools within the inner zone were located higher on shore (Figure S1.1).

2.4.2 Local species pools

The total number of species/taxa recorded across sampling times and sites within each zone (i.e. local species pool) was greater in the outer zone of Sydney Harbour than in the inner zone (Table S1.3). A total of 33 sessile and 27 mobile taxa were found in the outer zone, whereas at the inner zone only 17 sessile and 17 mobile taxa were found (Table S1.4, Table S1.5). Total number of sessile taxa in each pool ranged from 0 -16 taxa in the outer zone, and 2 - 9 in the inner zone of the Harbour. The number of mobile taxa in each pool ranged from 0 - 10 in the outer and 0 - 6 in

Chapter 2 17

the inner zone. The inner zone had 3 unique sessile species/taxa (Amphibalanus variegatus, yellow anemone, ‘Hydroides’), while 19 sessile taxa were only found in the outer zone, including the species/taxa Hormosira banksii and Codium spp. (Table S1.4). Three mobile species/taxa (Bembicium auratum, Elysia species a, a ctenophore) were unique to the inner and 12 taxa unique to the outer zone, including the species Nerita atramentosa and Austrolittorina unifasciata (Table S1.5).

2.4.3 Rock pool assemblages

Assemblages of sessile taxa varied between sites and zones (Fig. 2.3a, Table S1.6). Live and dead oysters as well as Corallina spp. had greater abundances in the inner zone when compared to the outer harbour (negative scores on PCO1). In contrast, the algae Ralfsia spp., the bryozoan Watersipora spp., the polychaete Galeolaria caespitosa, the barnacle Austrobalanus imperator, the anemone Actinia tenebrosa, bare rock and the unknown category, which refers to areas covered by dense canopy, were more abundant in the outer zone (positive scores on PCO1) (Fig. 2.3a). SIMPER analysis revealed that the brown encrusting algae Ralfsia spp. (19.06 %) Corallina spp. (12.90 %) and the red algae Corallina spp. (12.90 %) contributed most to the differences in the assemblage structure of rock pools between the inner and outer zones of Sydney Harbour (Table S1.7).

The distinct patterns detected in the PCO suggest that mobile assemblages differed between zones (Fig. 2.3b). This difference was, however, not statistically significant (Table S1.6). Significant differences were only observed among sites (Table S1.6). A test for heterogeneity in dispersions between zones also revealed significant differences (df1:1, df2:318, F:79.297, P (perm) < 0.0001,). The inner harbour was characterized by greater abundances of the gastropods B. auratum, Patelloida spp. and Siphonaria spp. (positive scores on PCO1 and PCO2), whereas the outer zone had greater abundances of B. nanum, Austrocochlea porcata, Nerita atramentosa, Tenguella marginalba and Montfortula rugosa (negative scores on PCO1 and PCO2) (Fig. 2.3b). SIMPER analysis revealed that gastropods of the species B. nanum (13.35 %), the genus Patelloida spp. (12.68%) and B. auratum (11.99%) contributed most to the dissimilarities between the two zones, explaining 37.98 % of the differences (Table S1.8).

Chapter 2 18

Chapter 2 19

2.4.4 Relationships between rock pool characteristics and diversity

Detected relationships between size parameters, height on shore and diversity parameters varied with the diversity parameter examined (taxa richness, total cover of sessile organisms, total abundance of mobile organisms), the assemblage assessed (sessile, mobile) and the location (inner zone and outer zone) (Table 2.1). Results found were despite any possible differences among samplings, since this factor was included in the model.

Width Depth Volume Height Inner Outer Inner Outer Inner Outer Inner Outer Richness Mobile taxa   --    X X Sessile taxa --  X X X X   Cover/ abundance Mobile taxa   X X  --   Sessile taxa X X   X X  

2.4.5 Taxa richness

Sessile taxa richness was best predicted height on shore, and the interaction between zone and rock pool width (Table S1.9). In contrast, mobile taxa richness was best predicted by rock pool width, volume and the interaction between zone and depth (Table S1.9). Patterns of sessile and mobile taxa richness in relation to parameters of pool size varied with zone (Table S1.9). Sessile taxa richness in the outer zone increased with increasing rock pool width and approximately doubled from 200 mm width to 500 mm width, however, there was no relationship between width and number of sessile taxa in the inner zone (Fig 2.4a). Conversely, mobile taxa richness increased with increasing rock pool width (Fig. 2.4b), regardless of zone. No relationship was found between number of sessile taxa and maximum depth of rock pools in both zones (Fig. 2.4c). In contrast, the number of mobile taxa increased with increasing depth in the outer zone, with an approximate increase of 3 taxa from the shallowest to the deepest pool. However, no relationship between mobile taxa richness and pool depth was found in the inner zone (Fig. 2.4d). Although volume had no effect on sessile taxa richness (Fig. 2.4e), it was positively related with the mobile

Chapter 2 20

taxa richness in both zones of the harbour (Fig. 2.4f). Sessile taxa richness decreased with increasing height on shore (Fig. 2.4g), however no relationship was found between height on the shore and mobile taxa richness (Fig. 2.4h). Algal canopy cover had no significant predictive capacity for sessile or mobile taxa richness (Table S1.10).

2.4.6 Abundance of sessile and mobile taxa

Total cover of sessile taxa was best predicted by maximum depth of pools and the interaction between zone and height on shore (Table S1.9). Total abundance of mobile taxa was best predicted by width, height on shore and the interaction between zone and volume. Canopy cover was an additional predictor for the abundances of mobile taxa in the outer zone (Table S1.10).

Total cover of sessile organisms increased with increasing depth in both zones (Fig. 2.5a). Highest correlations of increases in cover with increasing depth were found for the red algae Corallina spp. in the inner zone, and the brown encrusting algae Ralfsia spp. in the outer zone. Depth had no significant predictive capacity for the abundance of mobileorganisms. Although width did not influence total cover of sessile organisms, increasing width was positively related to the total number of mobile organisms in both zones (Fig. 2.5c). In both zones, cover of sessile taxa decreased with increasing height on shore, with this relationship being more pronounced in the inner zone (Fig. 2.5b). In contrast, total abundance of mobile organisms increased with increasing height on shore (Fig. 2.5d). Highest correlations of increases in abundance with height on shore were found for the gastropods Patelloida spp. and Siphonaria spp. in the inner zone, and the gastropods Bembicium nanum, Austrochoclea porcata, and Nerita atramentosa in the outer zone. The effect of volume on total abundance of mobile taxa varied between zones. Whereas total abundance was negatively correlated to volume in the inner zone, no effect was found in the outer zone. (Fig. 2.5e). Algal canopy cover had no significant predictive capacity for total cover of sessile organisms, but increased the abundance of mobile organisms(Fig. 2.5f).

Algal canopy cover was best predicted by height on shore (Table S1.11). Canopy cover decreased with increasing height on shore (Fig. 2.6).

Chapter 2 21

Chapter 2 22

species richness across the size range. Shaded areas represent standard error from this prediction. N=22 (outer zone),24 (inner zone).

Chapter 2 23

2.5 Discussion

This study investigated relationships between rock pool biota and the physical characteristics of rock pools within an urban estuary. Local species pool was greater in the outer zone of Sydney Harbour, characterised by lower urbanisation and greater exposure and flushing, than the inner zone. As hypothesised, the importance of many size parameters to the diversity of rock pools varied, not only with the type of organisms (sessile vs. mobile taxa), but also with position in the estuary (i.e. inner vs outer zone). Findings from this study can be used to inform the designs of eco-engineering features such as artificial rock pools to improve the diversity of artificial structures. The lower overall diversity (local species pool) found in the inner zone of Sydney Harbour here is probably related to differences in environmental conditions between the zones. The inner zone of the Harbour is characterized by fewer rocky reefs (Johnston et al., 2015), reduced flushing and generally high contamination compared to the outer zone (Das et al., 2000, Dafforn et al., 2012b). This might impede the dispersal of species and their larvae within or in between zones. The removal or alteration of natural rocky shores results in habitat loss and/or habitat fragmentation, spatially separating the remaining patches of the natural habitat (Saunders et al., 1991) and thus reducing their connectivity. Mobile intertidal organisms, generally, cannot move across large distances due to their small size and slow pace (Hamilton, 1978). This may prevent adults from rapidly colonising distant rocky shores. Artificial structures, can further reduce mobile diversity, as they do not provide appropriate habitat for many taxa, such as sea stars, urchins or some gastropods (Chapman, 2003) and therefore cannot be used as a medium ‘stepping stone’ to

Chapter 2 24

overcome fragmentation (Goodsell et al., 2007). The distribution of mobile species in highly fragmented areas depends therefore on dispersal of their planktonic larvae (Shanks et al., 2003) or rafting (Thiel and Gutow, 2005), which is dependent on the surrounding hydrological dynamics (Gaylord and Gaines, 2000). Similarly, many sessile species rely on transport of pelagic propagules by currents, but many algae propagules only successfully settle and grow within close proximity to the parent plant (Norton, 1992). Therefore, distance between suitable habitats plays a key role for successful dispersal. Although artificial structures have been shown to facilitate the introduction and spread of non-native species (Airoldi et al., 2015), we did not find invasive species in the sampled inner zone rock pools. Therefore, it is possible that in this case, artificial structures are acting as ‘barriers’, decreasing connectivity, rather than serving as ‘stepping stones’ (Bishop et al., 2017). Furthermore, the outer zones of estuaries may experience greater influx of propagules from the ocean due to the proximity to the mouth of the Harbour and thus greater flushing (Das et al., 2000), which could positively affect the local species pool in the outer zone, increasing the differences in the overall diversity between zones observed here.

In addition to habitat modification, urbanisation is also associated with anthropogenic stressors such as pollution from urban run-off and industry (Birch and Taylor, 1999, Stark, 1998). Toxic contaminants negatively affect species richness and alter assemblage composition (Johnston and Roberts, 2009). Declines of macroalgae in urbanised areas have been reported across the globe (Benedetti-Cecchi et al., 2001, Connell et al., 2008). Growth and germination of Hormosira banksii, a common intertidal macroalgae in Sydney, is affected by antifouling biocides (Myers et al., 2006). In Sydney Harbour, previous studies have shown distinct differences in the levels of pollution between the inner and outer zones of the harbour, even though differences in concentrations of some contaminants do not differ between the studied sites here (Birch, 2017). Therefore, the overall higher concentrations of contaminants such as heavy metals and polycyclic aromatic hydrocarbon found in the inner harbour (Birch and Taylor, 1999, Dafforn et al., 2012b) may explain some of the rock pools’ diversity patterns in the present study. In addition to overall differences in the local species pool between zones, the importance of rock pool characteristics also varied between zones. Increasing depth was positively correlated with mobile taxa richness in the outer zone. Such an association can be linked to the creation of ecological niches via thermal and saline stratification in deeper rock pools, even on small spatial scales (Martins et al., 2007, Morris and Taylor, 1983). This relationship was, however, not found in the inner zone. This may be due to the overall lower number of taxa in the inner harbour, which may offset the possible benefits of ecological stratification. Percentage cover of sessile taxa also increased with increasing depth in both zones. Deeper pools tend to have smaller fluctuations in environmental conditions compared to more shallow rock pools (Metaxas and Scheibling, 1993) and may therefore provide more suitable habitat for the settlement and growth of sessile

Chapter 2 25

organisms (Davison and Pearson, 1996, Eggert, 2012). This is contrary to a study in artificially created rock pools in Botany Bay, where no difference in cover was found among rock pools of different depths (5, 15, 30 cm) and between different depth zones within pools (0-5 cm, 5-15 cm, 15-30 cm) for all species except Hildenbrandia prototypus, Austrocochlea constricta and Cellana tramoserica (Astles, 1993). The composition of the assemblages examined in the artificial rock pools, however, varied from the assemblages studied in this survey, and included a variety of species not found in Sydney Harbour. Thus, increases in total cover of sessile taxa could have derived from taxa not found in Botany Bay, but abundant in Sydney Harbour. found that wider pools hosted higher diversity of both mobile and sessile taxa, but the positive effect on sessile taxa was limited to the outer zone. These results are similar to a previous study done in artificially built rock pools in natural rock shores, where the authors found a greater number of species (25- 50%) in wider pools (Underwood and Skilleter, 1996). This pattern is also consistent with the widely accepted species-area relationship, which implicates a greater number of species with increasing habitat size (Rosenzweig, 1995). In the literature, the two major factors discussed are the overall greater area sampled, and thus greater likelihood of finding more species, and/or a greater diversity of niches in greater areas (McGuinness, 1984). Here, the area sampled was correlated to the size of the rock pool, and therefore could have led to a greater number of species found. In addition to that, the rock pools sampled were natural rock pools and therefore not entirely uniform in shape and surface structure. Wider rock pools can contain a higher number of niches, which in turn can lead to greater number of taxa. Hence, both components could have potentially influenced the findings of this study. The width of pools did not affect sessile taxa in the inner zone. This result may also be related to the difference in the overall diversity found between zones, which might be modulating the relative importance of the physical parameters of rock pools on biodiversity.

Volume was an important predictor for the abundance and richness of mobile taxa. Whereas increasing volume increased taxa richness in both zones, it had no effect on the abundance of mobile taxa in the outer zone, and a negative effect on their abundance in the inner zone. More voluminous rock pools can be associated with greater diversity, as explained by the species-area relationship (Rosenzweig, 1995). Analyses done here however did not point to a possible explanation on why we found different, and sometimes opposing patterns, for theabundance of mobile organisms.

Height on shore was an important predictor for algal canopy cover, number and abundance of sessile taxa and abundance of mobile organisms. There were fewer taxa and reduced sessile and canopy cover with increased elevation on the shore. This can bedue to increased thermal stress in rock pools at higher shore levels due to extended exposure time compared to rock pools on the lower shore (Huggett and Griffiths, 1986, Morris and Taylor, 1983). Such stress can negatively

Chapter 2 26

affect the settlement and survival of sessile taxa (Kon-ya and Miki, 1994). Mobile taxa may be less limited by physiology because they can actively seek refuge when experiencing stressful conditions (Garrity, 1984, Williams and Morritt, 1995). Additionally, pools higher on the shore may offer refuge from predators restricted to lower shore levels (Dethier, 1980), which could have led to greater abundances of mobile taxa higher on shore.

In the present study, algal canopy cover was only ever recorded in the outer zone of the Harbour. We found a positive relationship with canopy cover and abundance of mobile organisms. The presence of macroalgae on the intertidal zone increases structural complexity and reduces environmental heat stress and desiccation (Bertness et al., 1999) and is therefore often associated with an increase in species richness and total abundance of mobile organisms (Migné et al., 2015, Schiel and Lilley, 2007). One possible reason why we did not observe a positive relationship between canopy and taxa richness may be because rock pool biota is constantly submerged, which might outweigh the beneficial effect of reduced desiccation stress. On the other hand, algal canopy can have negative effects on the recruitment of sessile species as a result of abrasion by the canopy (Hawkins, 1983). One possible explanation why we have not found such a relationship may be because algal fronds from canopy forming algae found (Hormosira banksii, Sargassum spp., Codium spp.) may be raised in the pool water, which could have prevented scouring by fronds. Nevertheless, canopy cover is known to directly host a variety of macro and microfauna (Lewis, 1987, Viejo, 1999). Future studies should consider including a microbial component when analysing drivers of rock pool biodiversity as microbial diversity may be more sensitive to a range of abiotic and biotic drivers including macroalgal cover (Tan et al., 2015).

2.6 Conclusion and implications

We revealed relationships between rock pool size as well as height on shore with biological diversity that differed between locations. In areas heavily urbanised and with low exposure and flushing (the inner zone), width of intertidal rock pools and height on shore were the only parameters related to taxa diversity. In areas of lower anthropogenic impact and greater wave action, flushing and tidal velocities (the outer zone), the maximum depth of rock pools was an additional factor positively related to diversity in these habitats. Restoration or rehabilitation practices that aim to maximise overall local diversity could prioritise the protection and/or construction of wide rock pools in the upper estuary, positioned low on shore. On the other hand, rock pools in the lower reaches of an estuary could support more biodiversity if they are built to be both deep and wide. The present study has, however solely focussed on the diversity of benthic assemblages. Previous analyses have revealed that these relationships vary among ecosystem parameters (e.g. productivity, taxon richness (e.g. fish)) (Bugnot et al., 2018). If resources permit,

Chapter 2 27

we therefore recommend the construction of rock pools of various sizes and tidal heights on artificial structures to increase overall ecosystem functioning (i.e. productivity) and overall diversity (Bugnot et al., 2018). Additionally, rock pool designs also need to take into account potential environmental disturbances. Artificial rock pools have been previously lost or broken in areas of heavy wave action (Browne and Chapman, 2014). Similarly, in areas of high sedimentation, deeper pools and/or pools lower on the shore are more likely to accumulate sediment (Waltham and Sheaves, 2018), potentially negatively affecting diversity. Additionally, the maximum weight of artificial rock pools is often constrained to avoid damages to the seawall they are attached (Browne and Chapman, 2014). Rock pools therefore need to be built to fit local environmental conditions without compromising the integrity of the seawall. These results suggest that the prospects for success of ecologically engineered designs will be dependent on the environment they are placed within. Better understanding urban rock pools is critical to determine the factors that support their ecological diversity, so that location-specific designs and priorities for protection can be optimised.

Chapter 3

The role of fine-scale habitat complexity and heterogeneity in species distributions on rocky shores

3.1 Abstract

Intertidal rock pools provide habitat for a variety of species and can thus support diverse assemblages. They are highly variable in size and shape, and there is limited knowledge on what fine-scale physical features support diversity in these habitats. Understanding the microhabitats that support diversity is important to conserve and potentially mimic these important intertidal habitats for conservation management. Here, we quantify the heterogeneity and complexity of rock pool microhabitats in different locations. We also investigate potential relationships between microhabitats and mobile species richness and abundance, and whether this changes through space and time. We surveyed natural rock pools, twice in winter and twice in summer, at 7 sites in Sydney: two sites each were located in the inner and outer zone of Sydney Harbour and 3 sites were located along the open coast. We found that rock pool heterogeneity increased from coastal > outer harbour > inner harbour, following similar patterns to mobile species richness. We found that mobile species were less abundant in rock pools with microhabitats. Patterns in individual species in relation to rock pool microhabitats were variable, but more than a third of the species only occurred in rock pools when microhabitats were present. This study indicates that the inclusion of specific microhabitats in the design of eco-engineered rock pools could enhance the abundance of certain mobile species and provide targeted restoration outcomes.

3.2 Introduction

The effects on biodiversity of different physical components of habitat structure, such as complexity and heterogeneity (McCoy and Bell, 1991), have long been a focus of ecological research (Carvalho and Barros, 2017, Dean and Connell, 1987, MacArthur and MacArthur, 1961, McGuinness and Underwood, 1986, Moschella et al., 2005, Pianka, 1966, Smith et al., 2014, Tews et al., 2004, Matias et al., 2007). The physical structure of natural habitats is widely believed to be an important driver of biodiversity (e.g. Dean and Connell, 1987, MacArthur and MacArthur, 1961, Tews et al., 2004). The prevalent hypothesis is that increased complexity and

28 Chapter 3 29

heterogeneity leads to a higher number of distinct resources, therefore allowing coexistence of different species (Menge and Sutherland, 1976). However, human activities, such as deforestation, urbanisation and logging, are leading to a global physical homogenisation of natural habitats (e.g. Jongman, 2002, McKinney, 2006, Thrush et al., 2006). This uniformity of land and seascapes is likely to lead to declines in the diversity of resources, resulting in significant biodiversity loss across habitats at global scales.

Homogenisation of coastal environments has been occurring through the introduction of artificial structures. Up to 70% of some coastlines around the world have already been modified by infrastructure such as seawalls (Dafforn et al., 2015a) and climate threats such as sea level rise and more damaging storm events will increase the demand for protective infrastructure in coastal areas. As a result, we predict increasing homogenisation of habitats, likely leading to biodiversity loss (Bulleri and Chapman, 2010, McKinney, 2006). This is of concern because coastal systems are highly productive, and it is suggested that biodiversity loss had the greatest effect on productivity at scales where environmental heterogeneity is small (Loreau et al., 2001).

Rocky shores are highly diverse and have varying levels of structural complexity. Habitat complexity (broadly referred to here as a combination of habitat complexity and heterogeneity) is an important factor in the structure and functioning of rocky shore communities and has been strongly linked to factors such as availability of resources, competitive interactions and the availability of refuge from predators and/or environmental stress (Garrity, 1984, Klein et al., 2011, McGuinness and Underwood, 1986). However, the extent of intertidal ecosystems, such as rocky shores have experienced major declines of more than 50% in many parts of the world (Perkins et al., 2015). Understanding how habitat complexity affects diversity of organisms in these coastal systems is therefore crucial if we are to devise efficient and feasible protection and restoration strategies for these threatened habitats.

On rocky shores, complexity occurs on at least two different scales: at a larger scale for example, rock pools provide complexity by providing submerged habitat on the otherwise dry rock bed, whereas the complexity at a smaller scale can result from biogenic ecosystem engineers or small- scale features of the rock bed such as micro-crevices and fractures. While the presence of rock pools is widely recognised as supporting intertidal biodiversity (e.g. Firth et al., 2014, Firth et al., 2013), little attention has been given to their finer-scale habitat structure. Consequently, there remains a lack of understanding on how structural complexity at fine spatial scales, or “microhabitats” within rock pools, influences the diversity of important marine intertidal systems.

It is common to see rocky pools being described as a “microhabitat” of the intertidal rock shelf. However, the distinction between habitats and micro-habitats depends on the scale being

Chapter 3 30

investigated (McGuinness and Underwood, 1986).The properties of a landscape or habitat should be observed at any scale that is relevant to the organisms or processes being studied (Matias et al., 2010). Moreover, some organisms may themselves provide micro-habitats for other species (Bros, 1987, Romero et al., 2015, Underwood and McFadyen, 1983). Rock pools consist of a variety of different fine-scale features including pits and rock ledges. Given the size and life- history of many species inhabiting rock pool, these fine-scale differences in physical features among pools can be important factors driving associated communities. For instance, rock pools with algal cover and rock ledge cover have been linked to fish abundances and potential predator- avoidance strategies (White et al., 2015). Additionally, some species were only found in pools with specific characteristics (e.g. loose shells), which suggested a level of habitats selectivity (White et al., 2015). Similarly, Tews et al. (2004) has found that different species groups are closely linked to so called “keystone structures” that play and important role in species diversity. The type of microhabitat therefore seems to be an important factor in assemblage composition.

Nevertheless, the effect of such structures on benthic assemblages has not been assessed. Additionally, there is limited knowledge not only on the size, type and number of finer-scale physical features occurring in natural rock pools, but also whether such features actually affect diversity. Fine-scale features within rock pools might particularly affect mobile species, as they can change their location on the intertidal rock shelf. Gastropods have been found to alter their vertical distribution seasonally (Frank, 1965) and in accordance with the tides (Dexter, 1943), and move to habitats of reduced heat and desiccation stress when not foraging (Garrity, 1984, Williams and Morritt, 1995). Personal observations from a survey on intertidal rocky shores in Sydney suggest that this is also may hold true for distributions within rock pools. Assessing fine- scale features within rock pools may therefore provide important link for more detailed habitat- diversity relationships.

As ways to combat the loss of species on homogeneous artificial structures on intertidal systems, several designs and interventions, referred to as ‘eco’, ‘blue’ and/or ‘green’-engineering, have been proposed with the main goal of increasing structural complexity in built infrastructure (Strain et al., 2018). Of those, the most widely applied intervention is the addition of water retaining features, which intends to mimic natural rock pools (Browne and Chapman, 2014, Bugnot et al., 2018, Chapman and Blockley, 2009, Evans et al., 2016). While adding artificial pools of simple shapes and different sizes and/or depth has been linked to increasing intertidal diversity of artificial structures (Browne and Chapman, 2014, Chapman and Blockley, 2009), few studies have attempted to incorporate finer-scale structural features (heterogeneity), or manipulate their abundance (complexity), in artificial rock pools. Understanding the fine-scale features of natural rock pools and their effect on diversity can help to further optimise future rock pool designs to

Chapter 3 31

maximise their ecological benefits when constructed on artificial structures.

Here, we surveyed natural rock pools to quantitatively and qualitatively assess: (1) the types and abundance of different fine-scale physical features that are commonly present in rock pools and (2) whether such features promote diversity. In particular, we tested the following hypotheses:

-The number (heterogeneity) and size (complexity) of microhabitats will vary across an urbanised estuary.

-Pools with one or more microhabitats (heterogeneity) will support a greater (i) number and (ii) abundance of mobile species compared to rock pools no microhabitats.

-In rock pools with microhabitats, pools with microhabitats that occupy a greater area (complexity) within the pool will support a greater (i) number and (ii) abundance of mobile species compared to rock pools where rock pools occupy a smaller area.

-Where present in rock pools, microhabitats will support greater densities of organisms compared to non-microhabitat areas of the pools.

-Relationships will vary with the type of microhabitat and among species, and across space (among locations) and time (between seasons).

3.3 Materials and methods

3.3.1 Study area

Intertidal rock pools were sampled in 7 sites around Sydney, New South Wales, Australia. Three sites (Bondi, Freshwater and Curl Curl) were located along the coastline of Sydney – hereafter referred to as coastal sites, and four sites were located within Sydney Harbour (33° 51’ S, 126 151° 14’ E). Of those, two sites (Bradleys Head and Delwood) were located in the outer region closer to the mouth of the estuary (hereafter referred to as Outer Harbour), and two sites (Balmain and Berry Island Reserve) were located in the inner part of the estuary (hereafter referred to as Inner Harbour) (Fig. 3.1). Sydney Harbour is a drowned-river valley, and the substrate of rock pools at all locations is Sydney sandstone (Johnston et al., 2015). All sampling locations chosen for this study are known to differ with respect to local environmental conditions. The three coastal sites are openly exposed to high wave action and well flushed, whereas the sites within Sydney Harbour are characterized by decreasing oceanic input/flushing, reduced wave action and increasing contamination from outer to inner (Dafforn et al., 2012b). All sites chosen for this study were natural rocky shores with rock pools in a range of sizes. On each of the 7 rocky shores, 18 rock pools were initially haphazardly selected and marked with a tag to capture a representative sample and due to the logistics of sampling during a low tide. For this study, rock pools were

Chapter 3 32

defined as water retaining depressions in the rock shelf and included all rock pool features under and above the waterline (Table 3.1). The sampling method turned out to be unsuitable for rock pools >800 mm width, rock pools with many small gastropods along the water edge, and rock pools with extensive oyster or algae cover, which were therefore excluded (Table S2.1). Some marked rock pools were also subsequently found to be connected to other rock pools, could not be sampled at each sampling time (due to swell) or found to partially or fully dry out during low tide and were excluded as well. Final numbers of rock pools sampled at each site were as follows: Balmain: 15; Berry Island Reserve: 12; Bradleys Head: 9; Delwood: 12; Bondi: 14; Freshwater: 13; Curl Curl: 15. To assess rock pool characteristics (volume and microhabitats) and biota, rock pools were emptied of all water using measuring cylinders and syringes for better assessment.

Chapter 3 33

Table 3.1 Physical features measured in each rock pool and the definitions used for each feature.

Physical microhabitats Definition Small pit A depression in the rock surface of 1-5 cm depth Medium pit A depression in the rock surface of 5-10 cm depth Deep pit A depression in the rock surface of >10 cm depth Rocky overhang emergent Ledges of >2cm and areas covered by a ledge >2cm above water Rocky overhang Ledges of >2cm and areas covered by a ledge >2cm under water submerged Biogenic microhabitats Oyster shell overhang An oyster shell(s) creating an overhang at the pool’s edge Sessile shell An empty oyster or barnacle shell Other Physical Features Non-microhabitat ‘Unstructured’ parts Non-microhabitat around a rock pool emergent ‘Unstructured’ parts Non-microhabitat within a rock pool submerged

3.3.2 Rock pool characteristics

The volume of each pool was measured in situ once prior to sampling by emptying the rock pool of all water using measuring cylinders and syringes and recording the amount of water in each pool. Different physical features of the pools (microhabitats) were identified (Table 3.1), and the size of each feature (% of total rock pool area) estimated in the field or from photos. Submerged features were any parts of the rock pool below the waterline, while emergent features included all surfaces above the waterline (e.g. vertical walls) directly surrounding the rock pool. Biogenic microhabitats included oyster and barnacle shells. Although algae also provide biogenic habitat, they were not considered biogenic “microhabitat” as a distinct area shaded or covered by floating algal fronds could not always be clearly determined

We also calculated overall habitat heterogeneity and complexity of each pool based on the following definitions. For this census, the term “microhabitat” refers to all pits and rocky overhangs only, as biogenic microhabitats can be of temporary nature. Heterogeneity was defined as the number of different types of microhabitats within a rock pool (Bell et al., 2012). Habitat complexity is usually defined as the abundance of structural components of a habitat (Bell et al., 2012). In the present study, we could not apply this definition for ‘overhangs’, because these microhabitats can cover the whole area of the border of the pool or multiple parts of it and therefore would not accurately reflect complexity. Therefore, habitat complexity was here defined as the combined area of all microhabitats of a given type, and its effect was investigated only among rock pools with the given microhabitat. To account for the different sizes of pools sampled

Chapter 3 34

in this study, we standardised complexity by estimating the % area of microhabitats within each rock pool (i.e. how much of the total area of each rock pool was ‘occupied’ by microhabitats).

3.3.3 Rock pool biota census and habitat mapping

The diversity and abundance of rock pool biota was surveyed twice in Austral Winter 2016 and twice in Austral Summer 2016/2017. Surveys were done every 6-8 weeks. During sampling, rock pools were emptied of all water and videos (sampling 1) and photos (samplings 2-4) of the rock pool were then taken with a handheld IPhone 5S camera (camera: 8 mexapixels with 1.5µ pixels, video recording 1080p HD (30fps)) pointing downwards approximately 20 cm from the rock pool until all areas of the pool were filmed/photographed. From the videos and photos, mobile organisms were identified to the lowest possible taxonomic level (species or morphospecies) or functional group, counted and assigned to a physical feature (Table 3.1). For example, and organisms not located in a microhabitat but covered by water were assigned to the ‘unstructured’ parts submerged category. Due to the limited resolution of images and videos, organisms smaller than ~ 0.5 cm were excluded from the analyses as they could not always be properly identified in the videos. Faster moving mobiles (fish and crustaceans) were also excluded from analyses as they were moving around. Organisms were often found in clusters and were assigned to the physical feature directly beneath.

3.3.4 Statistical analyses

3.3.4.1 Differences between methodologies

Since we used different sampling methods to record the mobile species in winter (video) and summer (video and photos), we tested for potential confounding effects. To do this, we compared a subsample of videos and photos of randomly selected rock pools (n = 1-2) at each site that were surveyed in the second winter sampling. The total number of species did not differ between methodologies, however the total number of individuals differed at Inner and Outer Harbour locations, with more individuals found in the photo. Therefore, no comparisons of biota between seasons as well as single samplings were made for these locations.

3.3.4.2 Comparison of the structural complexity of pools (heterogeneity and complexity) across an urbanised estuary

Due to the difference in numbers of rock pools sampled at each site, differences in the abundance and size of microhabitats among locations were analysed as proportions. To investigate whether

Chapter 3 35

rock pool volumes varied among locations, we used backwards model selection of a linear model with volume as a response variable, and the categorical factor location (Coastal, Outer Harbour, Inner Harbour) as a predictor variable. . Differences in the availability and numbers of microhabitats could not be compared statistically and therefore we took a qualitative approach to describing patterns observed. Differences between local species pools among the locations were determined based on presence and absence of species of all samplings and sites combined.

3.3.4.3 Comparison of (i) the overall number of species and/or (ii) the abundance of mobile organisms between rock pools with and without microhabitats

We used generalised linear mixed effects models (GLMM) to test for relationships between predictor variables and the response variables (i) total number of species (ii) total abundance of mobile organisms. All statistical analyses were performed using backwards model selection, by sequentially removing non-significant variables until all remaining variables were significant, starting with non-significant interactions. For that, all types of microhabitats were pooled into a single category of microhabitats with the levels present and absent. Fixed factors varied depending on the factor investigated (Table 3.2). Sampling and rock pool were included as random factors in the model to account for the repeated sampling of the same rock pools over time. The volume of each pool was used as a proxy for rock pool size and was included as an offset in the model. This was done so a potential increase in the number of species was not confounded by an increase rock pool area, as predicted by the species-area relationship. A Poisson distribution was used in all models. Where convergence errors occurred or overdispersion was present in residual plots, a negative binomial distribution was used. The influence of the total area of microhabitats (complexity) was only tested for rock pools in which microhabitats were present.

Table 3.2 Fixed factors included in each model. Abbreviations are as follows: P = Present, A = Absent, W = Winter, S = Summer, C = Coastal, OH = Outer Harbour, IH = Inner Harbour.

Analysis Fixed factors

Heterogeneity (Microhabitat) Microhabitat (P/A)*Season (W/S)+ Location (O, OH, IH) Complexity (Microhabitat) % -cover of microhabitat + Location + Season Heterogeneity (Overhang) Overhang (P/A)*Season (W/S)+ Location (O, OH, IH) Complexity (Overhang) % -cover of overhang + Location + Season Heterogeneity (Pit) Pit (P/A)*Season (W/S)+ Location (O, OH, IH) Complexity (Pit) % -cover of pit + Location + Season

Chapter 3 36

To investigate whether results were due to a particular type of microhabitat (i.e. whether effects were correlated with a certain type of microhabitat or whether different types of microhabitat had different effects on the diversity of organisms), tests were additionally done for each type of microhabitat separately (Table 3.2). Different sizes of pits were pooled into a category of ‘pit’ with levels present and absent. Similarly, overhangs above and below the waterline were pooled into a category of ‘overhang’ with the same levels. For these analyses, only rock pools with a single microhabitat were used. However, as there were at least twice the amount rock pools without microhabitats than with a single microhabitat, we had an unbalanced design. To balance the design, five random subsets of the rock pools without any microhabitats were taken and combined with the rock pools with the microhabitat of interest (subsets 1-5) (Table S2.2). For example, if 10 rock pools had pits, and 20 rock pools had no microhabitats, each subset contained the 10 rock pools with pits and 10 randomly chosen rock pool combinations of the 20 rock pools without any microhabitats. Due to uneven distribution of microhabitats among rock pools of different sizes, we excluded rock pools with a volume of less than 200 mL. To test whether the results were correlated with particular species, we additionally did the above-mentioned analyses for the total number of individuals of the three most abundant species/functional groups as well.

3.3.4.4 Comparisons of the density of organisms between microhabitats and non- microhabitats in rock pools with microhabitats

To determine whether organisms inhabited microhabitats within rock pools or other areas within the rock pool, we performed analyses on the distribution of organisms in rock pools with microhabitats only. For that, a new “habitat” category was created, which included the categories ‘microhabitat’ and ‘remaining area of pool’. The model included an interaction between the fixed factors microhabitat and season, as well as the random effects rock pool and sampling. Season was included as a fixed factor in the model because we hypothesised that distributions within rock pools would change between winter and summer. The volume of the habitat (microhabitat or ‘remaining area of pool’) was included as an offset to account for different proportions of the habitats within a rock pool. This was calculated using the water volume of the entire rock pool and the estimated percent area occupied by either habitat. When species distributions were analysed, organisms occupying biogenic microhabitats were assigned to the ‘remaining areas of the pool’ category. In text, numbers of individuals per habitat volumes are referred to as densities.

Generalized linear mixed models were performed using the lme4 package (Bates et al., 2015). Assumptions of all tests were checked by plots of the residuals versus fitted values. All statistical analyses were conducted in R (Version 3.4.3).

Chapter 3 37

Due to the overall low number of microhabitats (pits and overhangs) in the surveyed pools of the inner and outer zone of Sydney Harbour, formal statistical analyses for species distributions could only be done on the coastal pools, and location as a factor was subsequently dropped. Since possible confounding factors due to extensive oyster and algae cover was limited at coastal sites (3 rock pools), it is assumed that any results found can be generalized for all rock pools at coastal locations. Qualitative analyses of species diversity as well as the occurrence of types of microhabitats were however done for all studied locations.

The effect of biogenic microhabitats on the total number of species and individuals such as oyster and barnacle shells were excluded from analyses as they can be of temporary nature. Instead, their effect on diversity was tested in individual rock pools when they were present. Analyses were however only performed for biogenic microhabitats that occurred more than 10 times to ensure sufficient replication.

Additionally, species occurrences (presence in rock pools with and without microhabitats) and frequencies (number of times present in rock pools with and without microhabitats) are described qualitatively. For that, abundances of a species were summed over all samplings and seasons and then transformed into a presence/absence matrix.

3.4 Results

3.4.1 Comparing rock pools among locations

Overall, we generally found more mobile species at the coastal sites (a total of 16 species; Tale S2.3). We found 12 mobile species in the Outer Harbour and 10 mobile species in the Inner Harbour (Table S2.3). We also found that the volume of the sampled pools within each location varied greatly, but there were no significant differences among locations to explain the differences in mobile species richness (Table 3.3).

Chapter 3 38

Table 3.3 Rock pool volumes (Litre) at each location (Coastal, Outer and Inner Harbour). Mean, minimum and maximum values are presented for each site.

Location Site Mean Minimum Maximum

Bondi 0.977 0.17 4.51

Coastal Freshwater 5.327 0.34 27.92

Curl Curl 1.671 0.03 10

Bradleys Head 0.686 0.07 2.78 Outer Harbour Delwood 1.414 0.015 4

Balmain 2.971 0.090 16.42 Inner Harbour Berry Island 2.886 0.055 9.6

Similar to patterns in mobile species richness, heterogeneity and complexity also differed among locations. Specifically, the proportion of rock pools occupied by different microhabitats (i.e. heterogeneity) varied among locations with a general gradient of decreasing heterogeneity towards the inner parts of Sydney Harbour (i.e. the number of pools with one or more types of microhabitats was greater at the coastal sites, followed by Outer Harbour and Inner Harbour; Fig. 3.2). Fifty percent of rock pools at coastal locations had at least one microhabitat, whereas microhabitats were only found in ~ 30 % and ~15 % of the outer and inner rock pools, respectively (Fig. 3.2).

The presence of the different types of microhabitats also varied among locations (Table S2.4). In the Inner Harbour, the only microhabitats found were overhangs (Fig. 3.2). At the coastal sites, microhabitats occurred independently and in combination with other microhabitats in the same rock pool, whereas in the Outer Harbour, overhangs were never found independently (Fig. 2). The occurrence of pits of different sizes varied among locations as well. Pits were completely absent from the Inner harbour. Out of 26 pits found in 19 rock pools at coastal and Outer Harbour rock pools, small pits were the most common type found (17), followed by medium (7) and deep pits (2).

Chapter 3 39

Similar to the patterns of heterogeneity, we also found greater complexity (or area of microhabitat within a pool) in rock pools at the coastal sites (Fig. 3.3). Total microhabitat area (standardised by overall pool area) ranged from 2% to 65% at the coastal site, 12% to 21% in the Outer Harbour and 5% to 20% in the Inner Harbour (Fig. 3.3). However, contrary to the hypothesis that complexity would increase the abundance of mobile species, we actually found that microhabitat area (complexity), of all microhabitats combined or separately, had no predictive capacity for any of the response variables examined (Table 3.4). Further, mobile species richness wasn’t related to complexity.

Chapter 3 40

Chapter 3 41

Microhabitats

Abundance n.s. Richness n.s. Nerita atramentosa n.s. Bembicium nanum n.s. Austrolittorina unifasciata n.s.

Overhangs

Abundance n.s. Richness n.s. Nerita atramentosa n.s. Bembicium nanum n.s. Austrolittorina unifasciata n.s.

Pits

Abundance n.s. Richness n.s. Nerita atramentosa n.s. Bembicium nanum n.s. Austrolittorina unifasciata n.s.

Chapter 3 42

3.4.2 Comparing mobile species abundance and richness among rock pools with and without microhabitats

We found that the presence of microhabitats had varying effect on species richness and abundances depending on whether microhabitats were assessed in combination or separately. Specifically, we found that rock pools with microhabitats had fewer species and lower abundances compared to rock pools without microhabitats when microhabitats were analysed together. When analysed together, we found a negative correlation between the presence of microhabitats and mobile abundances in most instances.. Only the gastropod Nerita atramentosa had greater abundances with rock pools with overhangs in one out of 5 subsamplings(Table 3.5). Pits on the also had overall a negative effect on the abundance of mobile organisms, although a positive effect was found for Austrolittorina unifasciata in one instance, which was limited to summer (Table 3.5). The presence of pits and overhangs negatively influenced mobile species richness (Table 3.5). We did not observe many predatory gastropods and of those which occurred, they did not show specific patterns of distributions within rock pools.

Chapter 3 43

Table 3.5 Results of the presence of a microhabitat, pits and overhangs on the total number and abundance of all mobile species as well as on the abundance of the three most common species. N.s. = non-significant. W = Winter. S = Summer. * = same trend in seasons. Results were deemed variable when contrasting patterns were found. Results were found in all of the 5 subsamplings unless stated otherwise.

Microhabitats

Abundance decreased* Richness decreased Nerita atramentosa increased* Bembicium nanum decreased Austrolittorina unifasciata n.s.

Overhangs

Abundance decreased* Richness decreased Nerita atramentosa Increased* (1 out of 5 subsamplings) Bembicium nanum decreased* (3 out of 5 subsamplings) Austrolittorina unifasciata variable

Pits

Abundance decreased (variable effect of season) Richness Decreased (4 out of 5 subsamplings) Nerita atramentosa n.s. Bembicium nanum decreased (also * for 1 out of 5 subsamplings) Austrolittorina unifasciata variable

Chapter 3 44

3.4.3 Mobile species distributions in rock pools with microhabitats

We also investigated the distributions of mobile species in rock pools with microhabitats. Despite having an overall negative effect on overall abundances, we found greater densities under overhangs than in the non-structured parts of the pool (Fig. 3.4). We observed the same pattern when rock pools had pits, and the magnitude of this difference was greater in winter (Fig. 3.5).

When we separated pits into size classes, we also observed patterns in distribution for several mobile species. In one rock pools where deep pits were present, the gastropods N. atramentosa and Austrocochlea spp., as well as the sea star Parvulastra exigua and limpets occurred in greater numbers in deep pits compared to other parts of the pool (Fig. 3.6).

Chapter 3 45

Chapter

Figure 3.6 Number of individuals per species in each habitat. Deep pits were pits with a depth greater than 10cm. Medium pits were pits deeper than 5cm.

Samplings are the 4 samplings times (1+2: winter; 3+4: summer).

Chapter 3 47

3.4.4 Occurrence (presence/absence) of individual taxa in rock pools with and without microhabitats

The occurrence of individual taxa differed depending on whether microhabitats were available (Table 3.6). More than a third of taxa were only found in pools with microhabitats. Four gastropod species and the seastar Meridiastra calcar only occurred in rock pools with a microhabitat present and were always found in a pit or under an overhang. Gastropod sp. 5, a nudibranch and a flatworm only occurred in rock pools with microhabitats, but were generally found in the unstructured areas of these pools. In contrast, Onchelidium damelii only ever occurred in rock pools with no pits or overhangs. Furthermore, chitons occurred more frequently in rock pools with overhangs and pits. Within coastal rock pools for example, almost half of the rock pools with an overhang also supported chitons. Within Outer Harbour rock pools, this figure was even higher, with more than half of the pools with overhangs or pits supporting chitons (Fig. 3.7), while clear patterns were not observed for the other taxa, so the data has not been presented. However, the species Parvulastra exigua and Tenguella marginalba often occurred under overhangs (personal observation).

Chapter 3 48

Coastal Outer Harbour Inner Harbour

N P OH OH + P N P OH + P N OH

Gastropod 1*

Gastropod 2*

Gastropod 3*

Gastropod 4*

Meridiastra calcar*

Gastropod 5**

Nudibranch**

Flatworm**

Onchidium damelii***

Bembicium nanum

Bembicium auratum

Austrocochlea spp.

Nerita atramentosa

Tenguella marginalba

Austrolittorina unifasciata

Nodlittorina pyramidalis

Dicathais orbita

Gastropod 6

Gastropod 7

Elysia spp.

Parvulastra exigua

Limpets Chitons

Chapter 3 49

3.4.5 Mobile species distributions in rock pools with biogenic microhabitats

Biogenic microhabitats were relatively uncommon in the rock pools sampled in this survey (found in <2 rock pools per site). The exception was Berry Island Reserve, where biogenic microhabitats (oyster shells) were found in 6 rock pools. At this inner harbour site, oyster shells were often (10 out of 18 times) occupied by a greater density of B. auratum compared to other parts of the rock pool (Table S2.5). Additionally, the limpet Patelloida. spp. generally occurred more on oysters than other surfaces (personal observation).

3.5 Discussion

It is widely accepted that habitat structure is an important driver of biodiversity, but the influence of habitat structure is often not measured at a scale that is relevant to the organisms being studied (Matias et al., 2010). This study indicates that the presence of microhabitats in intertidal rock pools is not linked to overall increases in the abundance of mobile species in these habitats, but was found to increase the abundance of certain grazers and rare species. Along the open coast, the importance of microhabitats on the abundance of mobile species in rock pools was dependent on the microhabitat present (e.g. “pit” and “overhang”) but also the species investigated. Generally, rock pools with microhabitats did not support greater abundances of mobile species, but when present, microhabitats supported a greater density of individuals than unstructured parts

Chapter 3 50

of the rock pool. The gastropod Nerita atramentosa and chitons in particular were common in rock pools with overhangs.The area of occupied by pits or overhangs was not related to either the total number of species and their abundances. Availability of different types of microhabitats also varied among locations. These results suggest that the relationship between diversity and habitat structure is more complex than usually assumed and that microhabitats are probably more important to increase the overall local diversity (by supporting more rare species, for example) than to increase the diversity within pools.

The present study highlights that the type of microhabitat present in a rock pool plays a key role in determining the presence and abundance of a variety of species. For example, we found higher abundances of the gastropod Nerita atramentosa under overhangs compared to other habitats within the rock pool, although this was limited to a single subsampling. Similarly, we found that several species were unique to rock pools with microhabitats, which they mostly also occupied at the time. This supports the concept of “keystone structures” (Tews et al., 2004), where the authors suggest that the presence of particular features provide resources or shelter crucial for single or multiple species and thus determine species diversity. Similar patterns have been found for other areas of intertidal rocky shores, where gastropods showed preferences towards particular microhabitats (crevices, tidepools or vertical surfaces) during low tide (Garrity, 1984, Williams and Morritt, 1995).

Microhabitats can reduce environmental disturbances by facilitating the development of a range of microclimates (Sebens, 1991), with the potential to buffer even extreme environmental conditions (Scheffers et al., 2014). Besides providing refuge from environmental stressors, microhabitats can also provide protection from visual predators such as birds (Cantin et al., 1974). When microhabitats were present, we found greater numbers of individuals in microhabitats than other areas of the pool, which could be linked to the predator refuge and shelter from direct sunlight, which can reduce heat stress. The increase in the abundance of mobile organisms was only correlated with increases of N. atramentosa in one of five subsamplings. Conversely, the presence of overhangs only positively influenced A. unifasciata in summer, which can be related to differences in heat tolerance among the species. The gastropod A. unifasciata is a member of the family Littorinidae, which is known to be relatively heat tolerant (Fraenkel, 1966, McMahon, 1990). Seasons are usually associated with changes in temperature, with higher temperatures and thus greater heat stress in summer. Gastropods have been found to move to habitats of reduced heat and desiccation stress when not foraging(Garrity, 1984, Williams and Morritt, 1995, Fairweather, 1988a). The distribution pattern of more individuals of A. unifasciata under overhangs compared to the rest of the rock pool was only occurring in summer, which suggests that overhangs serve as a refuge from heat stress. In this study, overhangs were either submerged by rock pool water or closely associated to it. However, similar patterns of distributions to certain

Chapter 3 51

microhabitats are found on emergent intertidal areas. The gastropods N. funiculata and N. sabricosta for example aggregate in crevices during low tide (Levings and Garrity, 1983), suggesting that it is the overhang structure providing shelter rather than the rock pool water.

We also found here that chitons occurred more often in rock pools with overhangs, however, they did not always occupy the overhang itself. This can however be due to the fact that an area was only considered ‘overhang’ when a ledge was present. Surfaces with a slight negative incline were accounted for as ‘unstructured part of the pool’. This was mostly the case in rock pools where chitons occurred in equal or higher numbers in the ‘unstructured part of the pool’ compared to ‘overhangs’. Nevertheless, more frequent occurrences of chitons in rock pools with overhangs suggests that overhangs play an important role in the distribution of this group.

One possible explanation for the negative effect of pits on diversity and abundance of species observed here is that most surveyed pits were relatively shallow (of up to 5cm) and wide (personal observation). Therefore, these microhabitats may not influence environmental conditions (e.g. attenuate temperature or wave action) or serve as refuge from predators. Water in rock pools can vary locally, with deeper parts and rock depressions to be of different temperature than water closer to the surface (Morris and Taylor, 1983). Deeper pits may differ in their temperature compared to other parts of the rock pool and may also reduce flushing and thereby physical disturbance from wave action. Pits, in particular narrow and deep pits, may also provide refuge as larger predators are prevented from entering them. Exclusion experiments with sessile species have found that greater proportions of prey organisms (sessile prey-species) survived in exposed areas where fast-moving consumers were excluded (Menge and Lubchenco, 1981). In this study, only a single rock pool had pits greater than 10cm, which were additionally relatively narrow. In this pool, more individuals (except for A. unifasciata and B. nanum) were found at the bottom of this pit. Due to the low number of replicates of deep pits the effect of depth however could not be tested.

Besides the physical (and more permanent) microhabitats, biogenic habitats increase structural complexity and can provide microhabitats (e.g. oyster shells). Here, oyster shells were an important microhabitat at one site in the Inner Harbour, affecting the abundance of B. auratum when present. Similar patterns were found in other studies, although distributions were, opposite to what was observed here, related to juvenile snails only (Crowe, 1996). As for other overhangs and pits, oyster shells may protect snails from direct sunlight and protection from predators or physical disturbance. A study investigating the influence of oysters on the predatory whelk Tenguella marginalba has shown that survival of this species was greater in habitats with many oysters (Jackson et al., 2008). This has important implications for ecological engineering designs, as oyster shells can be pre-seeded onto newly created areas such as seawalls or artificially created

Chapter 3 52

rock pools to kick-start recruitment. Their effect on the abundance of certain species at other sites however remains to be tested, as oyster shells were not abundant enough at other locations for analyses.

The area (complexity) of microhabitats had no effect on the number and abundances of species. However, in this study overall complexity was defined as the overall percent cover of all microhabitats of a certain type, rather than the number of structural components of a certain habitat, e.g. the total number of pits (Bell et al., 2012). For example, one 10cm wide pit in this study would have accounted for equal complexity as two 5cm wide pits. This could have masked potential effects of complexity.

The effect of habitat heterogeneity and habitat complexity could only be investigated quantitatively at the coastal location, as the rock pools at the Harbour locations did not have enough microhabitats for statistical analysis. To see whether the observed patterns hold true at the other locations, microhabitats will need to be added to these pools and their effect on the abundance of mobile organisms tested. Nevertheless, low numbers of microhabitats may play an important role in the differences in species pools among the locations. Microhabitats increase niche availability (Menge and Sutherland, 1976), which generally positively affects diversity. The Harbour locations had lower abundances of microhabitats, which may play an important role in the overall low species richness within the Harbour. Additionally, the local species pools may be affected by differences in environmental and anthropogenic stressors among the different locations. Rock pool communities at the study locations are exposed to different levels of contamination and exposure (Dafforn et al., 2012b), which can affect species diversity (Johnston et al., 2015).

3.6 Conclusion and implications

The present study has shown that species respond differently to the presence of microhabitats and the type of microhabitat, confirming (or supporting) findings from other studies done on different systems and/or habitats. It also revealed that links were only found when types of microhabitats were analysed separately, highlighting the need for more fine-scale studies. We found that overhangs increased the overall abundance of Nerita atramentosa and chitons, and increased rare species (e.g. Gastropod species 1-4). The presence of overhangs also meant that, within pools, species were more likely to be found under an overhang than on other unstructured areas of the pools.. Rocky reefs across the globe have become increasingly fragmented and replaced by homogeneous built artificial structures, particularly in urban environments (Dafforn et al., 2015a, Goodsell et al., 2007). Recent research however suggests that an increase in heterogeneity/complexity can benefit associated communities (for a review see Strain et al., 2018). Attaching artificial rock pools onto artificial structures have proven to be a successful way

Chapter 3 53

to increase the diversity of these habitats (Browne and Chapman, 2014, Chapman and Blockley, 2009). Advances in technologies, such as 3D laser scanning and 3D printing, enable engineers to design and build highly structured and more sophisticated artificial rock pools informed by ecologists. Where the objective is to increase abundances of particular and rare species, rock pools should therefore be constructed with overhangs and deep pits to increase the conservation efficacy of these mimics in areas of greater oyster abundances, seeding of rock pools with oyster shells can be an additional measure/action to increase the abundance of mobile organisms.

Chapter 4 54

Chapter 4

Effects of light intensity on mobile and sessile communities on intertidal rocky shores

4.1 Abstract

Artificial shading is becoming increasingly common in urban areas due to coastal development along the shoreline. Although shading is known to be an important driver of assemblage structure in intertidal ecosystems, little is known about how different levels of shading affect assemblage structure. Undertanding the effects of different levels of shading can assist in the development of mitigation strategies for artificial shading, for the benefit of both sessile and mobile assemblages. Here, we cleared intertidal rocky shore habitats (rock pools and emergent rock) of biota and assessed how different levels of shading affected the recruitment of mobile and sessile assemblages. Sessile and mobile assemblages responded differently to the level of shading. Recruitment of algae was most successful in treatments with high light transmission, but treatments with low light transmission had higher abundances of mobile taxa. Sessile taxa other than algae only occupied small amounts of space in all treatments. This experiment showed that manipulations of light intensity affects mobile and sessile communities in contrasting ways, highlighting the need to consider effects on groups of taxa when designing mitigation strategies for artificial shading.

4.2 Introduction

Urbanisation of coastal areas is accelerating (Neumann et al., 2015) and associated coastal developments have resulted in the introduction and proliferation of artificial structures into the marine environment, a process that is referred to as ‘ocean sprawl’ (Dafforn et al., 2015a, Firth et al., 2016, Duarte et al., 2013). In many regions across the globe, artificial structures already account for more than 50% of the shoreline (Chapman and Bulleri, 2003, Firth et al., 2016). One of the main habitats affected by artificial structures are intertidal rocky shores, and it is becoming increasingly evident that intertidal structures are not surrogates of the habitat they have replaced, causing losses of diversity, alterations in community structure and the spread of invasive species (Airoldi et al., 2015, Bulleri and Chapman, 2010, Chapman, 2003, Didham et al., 2007, Gacia et al., 2007, Iveša et al., 2010, Pister, 2009, Vaselli et al., 2008). As well as directly replacing natural

Chapter 4 55

habitat, the construction of artificial structures along the coastline, such as wharves, jetties and high-rise buildings, may alter the natural light regime by shading nearby habitats, in some cases permanently. Understanding how intertidal communities are affected by shading can help inform mitigation strategies that limit the ecological impact of these constructions.

Light availability is an important environmental factor that can control interactions between algae and invertebrates (Miller and Etter, 2008, Dafforn et al., 2012a) and can cause shifts in community structure from algae- to invertebrate-dominated communities when light becomes limited and vice-versa when light availability increases (Clark et al., 2015, Clark et al., 2013). Shifts in community structure have been observed in intertidal communities on both artificial structures and natural rocky shores, where shaded areas had reduced cover of macroalgae when compared to unshaded areas and were instead characterized by greater cover of sessile invertebrates (Blockley, 2007, Pardal‐Souza et al., 2017). Light is an essential factor for autotrophic organisms, providing the energy required to photosynthesise and produce sugars for respiration and growth (Franck and Loomis, 1949) and thus regulating the abundance and biomass of algae. Changes to the availaibility of light are therefore likely to change the taxanomic composition within any nearshore marine community.

Although light is an essential factor for autotrophic organisms such as algae, it can also be harmful depending on its intensity. Strong light, for example, can result in inhibition of photosynthesis of algae (Powles, 1984). In addition to that, the ultraviolet (UV) part of light can have negative effects, including reduced growth and photosynthetic rates and bleaching (Pessoa, 2012, Irving et al., 2005). Negative effects of UV light are, however, not limited to autotrophic organisms. UV radiation was found to cause mortality in encapsulated gastropod larvae (Rawlings, 1996) and slow down embryonic development (Davis et al., 2013). Both light, and UV radition are rapidly removed as they pass through water so the damaging effects are often limited to shallow waters (Dring et al., 1996, Franklin and Forster, 1997, Hargreaves, 2003). The provision of shade could potentially ameliorate the harmful effects of UV light in places of extreme intensity such as the shallow subtitdal and intertidal.

Shading can also reduce direct heat levels which may be causing stress and/or dessication. On intertidal rocky shores, canopy-forming macroalgae create natural shade that retains moisture and provides cover, thereby reducing temperature and water loss underneath (Bertness et al., 1999, Schiel and Taylor, 1999, Brawley and Johnson, 1993, Keough and Quinn, 1998, Lilley and Schiel, 2006). Additionally, the presence of shaded microhabitats, such as overhangs and crevices, can alleviate temperature stress. For example, mobile organisms inhabiting crevices in the intertidal were often found to be less thermally stressed and experience reduced water loss compared to those on exposed rock (Garrity, 1984, Gray and Hodgson, 2004). Provision of shelter or shade

Chapter 4 56

can therefore result in increased abundances of mobile organisms, although responses can be species-specific (Takada, 1999, Fairweather, 1988a, Moran, 1985). Shading by artificial structures can also lead to indirect effects, such as changes to predator-prey interactions, which in turn can strongly influence the abundance and distribution of intertidal species (Fairweather, 1988b, Fairweather and Underwood, 1991). An experiment comparing the feeding rates of fish under and besides piers has shown that feeding was reduced under the pier, and it is suggested that low light conditions under the pier prevented the visual predators from detecting their prey (Duffy-Anderson and Able, 2001). Similarly reductions in predation were found for predatory fish under wharves under dark (night) compared to light (day) and artificial light (at night) conditions (Bolton et al., 2017).

The growing evidence of the impacts of shading by artificial structures on associated communities is driving innovation in construction to mitigate such impacts. Ecological engineering incorporates ecological principles in the design of engineered structures (Bergen et al., 2001) and strategies to mitigate shading have already been trialled and incorporated into new developments, including light penetrating surfaces such as glass blocks and metal grating (PNNL 2002). Yet, little is known on how different levels of light intensities affect coastal communities, and an assessment of the effect of different levels of shading on intertidal communities is needed to optimise these designs. The intertidal zone is particularly important to study because it is likely to be hevily impacted by coastal constructions and is also likely to be strongly influenced by changes to light intensities as communities are completely exposed to direct sunlight for substantial parts of the day (Bulleri and Chapman, 2010, Raffaelli and Hawkins, 2012, Thompson et al., 2002). Other areas of the intertidal, such as shallow rock pools are also highly exposed to the effects of sunlight.

We therefore manipulated shade and investigated its effect on sessile and mobile intertidal organisms. Artificial shading plates were constructed over both emergent rock and rock pools to investigate whether these habitats respond differently to shading, and therefore require different mitigation strategies. Rock pools retain water and have lower fluctuations of physical factors such as temperature compared to emergent rock (Metaxas and Scheibling, 1993) and may therefore provide a more suitable habitat for the growth of hard substrate assemblages.

We hypothesise that

1) the development of algae assemblages (measured as percentage cover) would be faster under treatments of high light transmissions 2) treatments of low light transmission would have greater cover of sessile invertebrates and greater abundances of mobile taxa

Chapter 4 57

3) Rock pools support greater cover of sessile organisms and higher abundances of mobile taxa due to constant submergence in this habitat, and therefore we hypothesise a greater effect of shading treatments on emergent rock.

4.3 Materials and methods

4.3.1 Study area

The experiment was conducted at two sites in the Cape Banks Scientific Marine Research Area on the northern headland of the Botany Bay National Park, located on the coast of New South Wales, Australia (Fig. 4.1). Sites were approximately 175 m apart.

Each selected site had 45 artificially constructed rock pools of 5cm depth and 15cm diameter, which had been created in a previous study by drilling diamond-corers into the rock to create

Chapter 4 58

depressions of even size (Martinez (pers. communication), Underwood and Skilleter, 1996). We chose to use artificially-built pools for consistent depth and width, so any differences among treatments would not be confounded by other variables. Rocky shores at both sites were located at mid-shore levels and were submerged during every tidal cycle (personal observation). Site 1 was located at high mid-shore levels and thus emerged first during the receding tide, while site 2 was located on low mid-shore levels and therefore emerged later during the receding tide (personal observation). Additionally, site 1 had greater exposure to wave action than site 2 (personal observation).

4.3.2 Experimental set up

In the two weeks prior to the start of the experiment, 30 rock pools (from the available 45) (5cm depth and 15cm diameter) and 30 emergent rock surfaces (20 × 20 cm) were haphazardly chosen and carefully denuded of biota using hammer and chisel (without damaging the rock surface), and subsequent scrubbing with a wire brush. Removed biomass was put into natural rock pools (mobile organisms) or onto the adjacent rocky shore to be washed over by at least one tidal cycle. At the commencement of the experiment, rock pools and emergent rock surfaces were cleaned with a final scrub with bleach. After this, a photo of each experimental plot (rock pool or emergent rock) was taken to quantify any residual biota (hereafter referred to as residual cover) remaining in the plot.

Experimental plots were then randomly assigned to 5 shading treatments: 0% (full shade), 15%, 35%, 75% and 100% (full light) light transmission, as well as a procedural control (Fig. 4.2). Specifically, 50 experimental plots (25 rock pools and 25 emergent rock) were shaded with 20 × 20 cm, 4.5 mm thick acrylic plates and tinted film to simulate the different light intensities while inhibiting UV transmission (n = 5 replicates for each treatment by habitat type). To test for potential effects of the plates (i.e. changes in water flow, effect of cover), a procedural control that consisted of the attachment of a clear acrylic plate was used. Shading plates were attached 1.5-4.5 cm above the plots by cable-tying them to the lag eyes of 4 stainless steel lag eye screws (M6 X 60) positioned around the pools/emergent rock (Fig. 4.2, Table 4.1). The remaining 10 experimental plots for the ‘full light’ treatments (i.e. 100% light) were left uncovered. UV transmission was measured once just after the beginning of the experiment (12.12.2016) with a light meter, and all plates inhibited the transmission of UV to a great extent (Table 4.1, Fig. S3.1).

Experimental sites were visited regularly (maximum interval of visits was 18 days) for cleaning potential algal growth and salt aggregations on plates as well as for the replacement of any lost plates. When plates were found missing during sampling, mobile taxa were not assessed at that

Chapter 4 59

sampling time as they were exposed to full sunlight on the day. The experimental plots were, however, still photographed, for the subsequent assessment of cover of sessile organisms as missing plates were replaced in a timely manner, limiting the effect of cofounding factors. During the experiment some tags went missing and plots could therefore not be identified, which resulted in variable (n=2-5) numbers of replicates among sampling times (Table S3.1).

Table 4.1 Estimated light transmission and experimental set up for each treatment (estimates are based on data from manufacturers). Abbreviations for treatments are shown in brackets.

Treatment Light transmission Experimental set up

Full light (FL) 100%, + UV no plate Procedural control (PC) ~92%, no UV clear plate 75% light transmission (75%) ~75%, no UV clear plate + "Bolle Clear" film 35% light transmission (35%) ~35%, no UV clear plate + "Octane" film 15% light transmission (15%) ~15%, no UV clear plate + "ATP18GH" film Full shade (FS) ~0%, no UV black plate

Chapter 4 60

4.3.3 Assessment of mobile and sessile colonisation

Colonisation by mobile and sessile organisms in each experimental plot (i.e. rock pool or emergent rock) was assessed monthly for a total of 6 months. All mobile organisms present in each plot were identified to the lowest possible taxonomic level and counted. To be able to properly assess the colonisation of sessile organisms, most mobile organisms were removed from plots during sampling, then returned to the plot when sampling was finished. Limpets and chitons were not removed to avoid damage. To assess colonisation of sessile organisms, one photo of each experimental plot was taken. For assemblages within pools, the base of the pool was photographed. Photos were taken from the closest possible distance that captured the entire pool. For assemblages on emergent rock, the surrounding screws were used as a frame. At the laboratory, percentage cover of organisms in each experimental plot was estimated using a grid placed over each photograph. We used a round grid with 37 regularly spaced intersections to estimate cover of sessile taxa in rock pools and a square one with 25 intersections (square grid) for estimates on emergent rock and analysed habitats separately. As the cleared intertidal area covered by the shading plates had no clear borders when the shade was taken off for the photo, approximately 2.5 cm of the outer area within the screws was not assessed for sessile colonisation (Fig. S3.2). When identification was not possible (e.g. individual was mainly covered by algae from the pool walls), they were classified as ‘unknown’. Sessile colonisation within pools during the first 6 samplings was only assessed at the bottom of the pool. To ensure that cover of sessile organisms of the pool base was a representative measure of cover of the entire pool (including walls), an additional sampling of sessile assemblages of the pool base and pool walls was carried out. To do this, one photo of the base, and photos of four randomly placed quadrats in the wall of the pool (4.5 cm × 2 cm) were taken. Small quadrats were analysed using a 5 - point grid. Sessile organisms were identified to genus level or morphospecies. Turfing algae represent several types of micro-and macroalgae which share an extensive low-lying morphology (Connell, Foster et al. 2014). Due to the difficulty of separating turfing algae and Ralfsia spp. on walls in situ, these taxa were combined and were informally referred to algae mix on walls.

4.3.4 Light measurements

To measure the efficacy of treatments, light measurements in each experimental plot were taken using a HOBO data logger. Measurements were taken approximately halfway between each of the sampling times. Light was measured for 20 seconds underneath each shade. Measurements on control plots were taken by placing the HOBO logger in the middle of each plot for 20 seconds. These measurements were subsequently averaged per plot to get one measurement per experimental plot.

Chapter 4 61

4.3.5 Statistical analyses

4.3.5.1 Light measurements

To investigate potential differences among light treatments and the effectiveness of treatments over time, we performed backwards model selection (as per Zuur et al., 2009) using a linear mixed model. The full model included an interaction between the categorical fixed factors treatment (full light, procedural control; 75% light transmission, 35% light transmission, 15% light transmission, full shade) and the continuous fixed factor time (day of experiment of each sampling). Each replicate was included as a random factor to account for repeated sampling over time. The full model was then simplified by sequentially removing non-significant variables (based on lowest Akaike Information Criterion (AIC)). As overdispersion was present in all models, we log- transformed the data.

4.3.5.2 Biota

To investigate how mobile species richness, the abundance of mobile organisms and cover of sessile organisms changed over time and varied among treatments, analyses were performed as for light measurements using generalized linear mixed models. We assumed Poisson distribution for counts, but when there was significant overdispersion (dispersion value >2), we used negative binomial family. For cover of sessile organisms (live organisms only) we assumed a binomial distribution. Live cover was calculated by excluding points that fell on mobile species that could not be removed (limpets and chitons), debris or ‘unknown’. Due to low cover of sessile organisms other than algae, statistical analyses could only be performed for algae. Dunnett post-hoc testing was used to explore many-to-one comparisons between the full light and other treatments. GLMMs were performed using the lme4 package (Bates et al., 2015).

To assess whether sampling effort, i.e. sampling of only the base of the rock pool versus sampling of the entire pool (base plus walls), significantly affected the results of percentage cover of all sessile taxa we used a generalized linear model (GLMM). The model included an interaction of the factors surface (both versus base only; fixed) and light treatment, and rock pool as a random factor to account for the fact that the same rock pool was sampled. A binomial distribution (as described previously) was assumed. Tukey post-hoc tests were used to explore comparisons. GLMMs were performed using the lme4 package (Bates et al., 2015). Similar analyses were performed to assess whether debris accumulated within rock pools over time and whether this differed among treatments.

Chapter 4 62

Sites and habitats (rock pool and emergent rock) were analysed separately due to clear differences in the benthic assemblages, regardless of treatments. Patterns were qualitatively compared between sites and habitats.

4.4 Results

4.4.1 Light transmission

Experimental shading plates successfully manipulated the amount of light reaching the plots with some variation among sites and habitats (p < 0.05, Fig. 4.3). The black acrylic plates blocked 94 – 98 % of natural light reaching all experimental plots (Fig. 4.3). Light transmission to fully shaded (black plates), 15% light transmission and 35% light transmission plots was greatly reduced compared to full light plots (no plates) (with one exception; 35% pool plots at site 2). Plots assigned to the 75% light transmission treatment generally received less light than the full light plots (no plates), however this difference was not significant (p > 0.05; Fig. 4.3). Light transmission to procedural control plots (clear plates), did not differ statistically from the full light plots (no plates) or the 75% light transmission plots (p > 0.05; Fig. 4.3). Light transmissions for plates at site 1 further decreased over time for all treatments (Fig. S3.3)

Chapter 4 63

4.4.2 Number of species and abundance of mobile taxa

29 mobile taxa were found across all treatments, sampling times and sites. On average, pools supported more species compared to emergent rock (Table S3.2), and there was lower variability in the total number of taxa found in rock pools over time across treatments when compared to emergent rock. Overall, mobile taxa were also more abundant in rock pools than on emergent rock.

4.4.2.1 Pools

The abundance of mobile taxa differed among light treatments, but these patterns were variable among sites and times of sampling. In general, mobile taxa were more abundant in pools from treatments with no or low (≤ 15%) light transmission.

Chapter 4 64

Overall, mobile taxa at site 1 were most abundant in fully shaded rock pools (i.e. ‘no light’ treatment) with approximately 3 - 8 times more individuals in fully shaded pools than in pools under full light conditions (Fig. 4.4a). This was mainly driven by higher abundances of the species Nerita atramentosa, which, on average, showed higher abundances under this treatment (Fig. 4.5a). Furthermore, at site 1, abundances of mobile taxa increased over time in shaded rock pools (i.e. full shade and <15% light treatments). In contrast, the abundance of mobile organisms decreased over time in rock pools with greater than 15% light transmission (i.e. 35% and 75% light transmission, procedural control and full light treatments) (Fig. 4.4a).

At site 2, abundances of mobile taxa varied among treatments (P < 0.001) and there were more mobile organisms in pools that received 35% or less light transmission (i.e. full shade, 15 and 35%) than the other treatments (Fig. 4.4b). The gastropods Tenguella marginalba and Austrocochlea spp. were on average more abundant under these treatments (Fig. 4.5b). There was no effect of time.

Chapter

Figure 4.5 Mean abundances of the four most abundant and other taxa per sampling time at site 1 (a) and site 2 (b) averaged across replicate pools. Abbreviations

for treatments as shown in Table 4.1.

Chapter 4 66

4.4.2.2 Emergent rock

Patterns of the total number and abundance of mobile taxa found on emergent rock were more variable among treatments and over time compared to rock pools.

The total number of mobile taxa on emergent rock at both sites varied among treatments (P < 0.05 at site 1, P < 0.001 at site 2)) and over time (P < 0.001 at both sites). At both sites, there were significantly more taxa in the fully shaded plots than on plots that received full light (Fig. 4.6 a, b). At site 2, plots from the 15% light transmission treatment also had more taxa than the full light plots (Fig.4.6b). At both sites, the number of taxa increased in plots from all treatments with time (Fig. 4.6c, d).

Chapter 4 67

The effect of treatment on abundance of mobile taxa on emergent rock varied over time (P < 0.01 at site 1, P < 0.001 at site 2). At site 1, the abundance of mobile taxa decreased through time in fully shaded plots (Fig. 4.7a). This was mainly due to the lower abundances of N. atramentosa, Austrocochlea spp. and B. nanum found later in the year (Fig. 4.8a). All other treatments had the opposite pattern, with abundances of mobile taxa increasing over time. At site 2, there were higher abundances of mobile taxa in the fully shaded plots than in other treatment plots and this increased over time (Fig. 4.7b). There were approximately 10 times more individuals under the fully shaded treatment when compared to full light. This was mainly driven by the gastropods Austrocochlea spp. and Bembicium nanum (Fig.4.8b). Changes in abundance of mobile taxa however did not differ among other treatments and full light.

Chapter

Figure 4.8 Mean abundances of the four most abundant and other taxa per sampling time at site 1 (a) and site 2 (b) averaged across replicate plots on emergent 68 rock. Abbreviations for treatments as shown in Table 4.1.

Chapter 4 69

4.4.3 Cover and diversity of sessile taxa

A total of 12 live sessile taxa was found across all treatments, sampling times and sites. Overall, cover of algae increased at a much higher rate in pools when compared to emergent plots across all treatments and sites. Moreover, assemblages within rock pools were more diverse than those found in emergent rock. Sessile organisms other than algae were only found within pools. The effect of treatment on algal cover varied over time at both sites and both habitats (P < 0.001).

4.4.3.1 Pools

On average, pools at site 2 supported greater algal cover than those at site 1. At both sites, the most abundant sessile species were the brown encrusting algae Ralfsia spp., followed by turfing algae and the green algae Ulva spp. By the end of the experiment, algal cover was lower in the fully shaded pools when compared to pools from all other treatments at both sites (Fig. 4.9).

At site 1, rock pools from the 75% light transmission treatment had the greatest algal cover after 6 months (Fig. 4.9a), whereas at site 2, algal cover was similar among treatments (except the ‘full shade’ treatment) (Fig. 4.9b). Regardless, algal cover increased over time in all treatments, except in the fully shaded plots at site 1 (Fig. 4.9a). During the first half of the experiment, this was mainly due to increases in cover of the aforementioned most abundant species, whereas in the second half of the experiment increases in cover were mainly caused by other species that colonised the pools, in particular at site 2 (Fig. 4.10). Additionally, the types of organisms colonizing pools varied among treatments. Although the succession of sessile organisms was similar across the treatments, the coralline algae Corallina spp. was only found under full light and the procedural control (Table S3.3).

In addition to algae, there were ephemeral covers of sponges, anemones and barnacles (Fig. S3.4). Anemones and barnacles at site 1 mainly occurred under treatments of light transmissions ≥ 35%, whereas anemones, sponges and barnacles at site 2 mainly occurred under treatments with light transmissions ≤15% (Fig. S3.4)

Chapter 4 70

Chapter

Figure 4.10 Mean algal cover (%) of the three most abundant and other taxa per sampling time at site 1 (a) and site 2 (b) averaged across replicate pools.

Abbreviations for treatments as shown in Table 4.1.

Chapter 4 72

4.4.3.2 Emergent rock

Contrary to what we observed in pools, at the end of the experiment, there were no differences in algal cover between the fully shaded and fully lit emergent plots at site 1. At site 1, the greatest increase in cover was found in the plots from the less than 15% light transmission treatment, where algal cover reached ~25% after 6 months (Fig. 4.11a). However, there was high variability among replicates (Fig. 4.11a). Despite slower development of assemblages, the assemblages found in the plots from the 75% and 35% light transmission treatments were more diverse, as all other treatments were solely occupied by Ralfsia spp. (Table S3.3). At site 2, the highest increase in cover throughout the experiment was found in the control plots and in the plots from the 35% and 15% treatments (Fig. 4.11b). It is also important to note that plots from the 35% light transmission treatment had the greatest residual cover (~6%) at the beginning of the experiment (day 0 of experiment). Increases in algal cover was slowest in plots from the 75% treatment (Fig. 4.11b). As at site 2, assemblages under intermediate light levels (procedural control, 35%) were slightly more diverse (Table 2). At both sites, Ralfsia spp. had the highest abundances (Fig. 4.12). No sessile organisms other than algae were found on emergent rock.

Chapter

Chapter 4 74

4.4.4 Sampling effort

Due to loss of experimental shading plates and occasional turbidity that reduced photo clarity, the number of replicates for the final sampling was reduced (Table S.3.4). Analyses revealed that cover of sessile organisms at site 1 varied among treatments, and this effect varied depending on the sampling method (P < 0.05). For all treatments, except the procedural control and 15% light transmission, cover of sessile organisms was lower when rock pool walls were not included (Fig. 4.13). For the procedural control and 15% light transmission, cover of sessile organisms was similar between methodologies. At site 2, cover of sessile organisms only varied between sampling methods, with sessile cover being lower when base and walls were sampled together (Fig. 4.14. Diversity was similar among sampling methods, however, the algae mix as well as spirorbid polychaetes were often only found on walls (Table S.3.5).

Chapter 4 75

4.4.5 Debris

Debris, such as gastropod shells and pebbles, were only found in rock pools, but not in emergent rock plots. There was an interaction effect of treatment and time on the abundance of debris (P < 0.001) at site 1, which was mainly due to increases in debris under the fully shaded treatment (Fig. 4.15a). The fully shaded rock pools had residual debris at the beginning of the experiment (~10%), which further increased throughout the experiment to ~ 21% at the end of the diversity sampling (day 183 of experiment) and to ~26% when the sampling effort was compared (day 253) (Fig. 4.14a). Rock pools under all other treatments had little debris and did not differ from full light. At site 2, percent cover in debris did not vary among treatments, but increased over time (P < 0.001) (Fig. 4.15b).

Chapter 4 76

4.5 Discussion

This experiment was the first in situ test of effects of shading on both sessile and mobile rocky shore communities, and found that effects varied between habitat types and taxa. In rock pools, algae cover was highest under intermediate light levels, whereas mobile organisms were more abundant under fully shaded treatments. On emergent rock, full shade had the highest abundance of mobile taxa at one site, but algae cover was similar among treatments. Intermediate light levels also had greater cover and similar diversity of sessile organisms to other treatments.

Optimal performance of aquatic algae under intermediate light conditions likely results from a trade-off in obtaining maximal light for photosynthesis (Cronin and Hay, 1996, Fitzpatrick and Kirkman, 1995, Ruiz and Romero, 2001). while minimising light and UV stress (Pessoa, 2012, Powles, 1984). However, as there was no significant difference between assemblages under full light conditions and the procedural control, it is likely that UV only had a limiting effect on recruitment. Additionally, shelter from sunlight under lower light treatments can potentially reduce heat stress. However, since temperature of the substrate under the experimental shades was not measured, the effect of temperature cannot be determined.

There were substantially higher increases in algal cover and greater diversity of sessile taxa in rock pools compared to emergent rock, which is likely due to reduced environmental stress (a more stable environment and less chance of desiccation) within rock pools (Martone et al., 2010)It is however important to note that at site 1, the comparison of sampling methods between “base only” and the entire rock pool (base and wall) revealed that the cover of sessile organisms differed depending on the sampling method. Sampling only the base accurately reflected the cover of

Chapter 4 77

sessile organisms of the entire rock pool for the procedural control and 15% light transmission, whereas it was slightly underestimated for all other treatments. This however only confirms that highest cover of sessile organisms was reached under intermediate conditions.

The level of shading appeared to have little influence on the diversity of algae when compared among treatments within pools and on emergent rock at each site. Diversity was similar among all treatments within each site and habitat, except for Corallina spp., barnacles and anemones within rock pools, with the former only occurring under treatments of high light intensities and the latter two mainly under fully shaded conditions. This can arise from interspecific competition. For example, higher algal cover can limit the recruitment of barnacles (Jernakoff, 1985). Larval recruitment preferences may also explain the results (Blockley and Chapman, 2006, Saunders and Connell, 2001, Keough and Downes, 1982)

Additionally, herbivory by mobile organisms can affect recruitment and abundances of some of the algae found under treatments. Mobile organisms were most abundant under treatments of lower light, which could have led to increased grazing and thus reduced algal cover or recruitment under these treatments (Bellgrove et al., 2014, Underwood, 1980, Jernakoff, 1983, Anderson and Underwood, 1997). Furthermore, the presence of grazers can also inhibit successional change by preventing the overgrowth of Ralfsia spp. by ephemeral species such as Ulva spp., as Ulva spp. is more susceptible to herbivory compared to algal crusts (Dethier, 1981, Lubchenco, 1978, McQuaid and Froneman, 1993). However, since final cover and diversity of algae were similar among most treatments in both habitats, it is unlikely that grazing limited the succession and growth of algae.

Differences in the rate of increase of algal cover can also be a result of higher residual cover, facilitating growth and settlement, or increases in debris, since cover was calculated only using the debris-free space. This is however unlikely to be the case as higher residual cover or debris did not result in higher final algal cover.

The effect of the experimental shading plates on mobile organisms was immediate in both habitats, which suggests that responses to shading plates were related to the refuge they provide and not the sessile assemblage they might feed upon, particularly since there was little residual cover of sessile organisms in the plots at the beginning of the experiment. Higher abundances of mobile species under shading plates or within shelters have also been reported in other studies on intertidal gastropods (Takada, 1999, Fairweather, 1988a). As for sessile organisms, this result can be related to the increased shelter from direct sunlight under these treatments, which can ultimately reduce heat and desiccation stress. However, as no indicators of heat stress (e.g. temperature of gastropods, heat shock proteins) were measured, the effect of temperature alone

Chapter 4 78

can not be determined. The lack of an effect of treatments with higher light intensities, in particular the 75% treatment, can also be due to the absence of a significant differences in light transmissions between this treatment and low light. Alternatively, higher abundances of mobile taxa under darker treatments could be a result of darker shading plates providing a better visual cue for shaded and thus cooler habitat, which, in turn, could have influenced species distributions among treatments. However, increases in algal cover may have facilitated the recruitment of grazers later in the experiment when food became abundant in experimental plots (Quinn, 1988b, Quinn, 1988a).

Rock pools had higher abundances of mobile taxa compared to emergent rock, which can be linked to the increased buffer provided by rock pools against desiccation. At site 1, we found higher abundances of N. atramentosa, particularly in rock pools under the low light treatment. The introduction of artificial shade over both emergent substrata and rock pools may therefore alter species-specific abundances and thus the structure of natural communities.

Overall, growth rate and abundances of mobile taxa were higher at site 2, which can be linked to the different shore levels at which the treatment plots were located. Sampled plots at site 1 are located higher on the shore, which increases desiccation and heat stress due to extended exposure time at upper shore levels (Huggett and Griffiths, 1986, Morris and Taylor, 1983). Additionally, some of the experimental plots at this site did not support any biota before the experiment, suggesting that potential mitigation of physiological stressors by shading may have not been sufficient for biota at this height on shore.

4.6 Conclusion and implications

This experiment is the first one to assess, in situ, the effect of different levels of shading on both sessile and mobile intertidal communities and has shown that sessile and mobile communities respond in opposite ways, highlighting the need to consider both taxonomic groups when mitigation strategies are evaluated and desired outcomes are established.

Our study found that treatments that had intermediate light levels had the greatest cover and a similar diversity of sessile organisms to other treatments. Strategies to mitigate the impact of artificial structures such as boardwalks and jetties on natural rocky shores can use this information to better design their structures. If the desired ecological outcome was to increase sessile biomass then using metal grids or glass blocks that aim to reduce the light to intermediate shade levels may be appropriate. If the aim was to minimise ecological impacts in any form – then the structure should try to replicate natural, full light conditions.

Chapter 4 79

The results of this study also suggest that mobile organisms benefit most from mitigation strategies that involve full shading conditions, regardless of the type of habitat (i.e. rock pools or emergent rock). Creating overhangs on rocky shores, e.g. by adding artificial overhangs, can therefore be a simple way to support high numbers of mobile individuals on the intertidal. We also found, however, that debris accumulated in some rock pools, possibly due to high abundances of mobile taxa (up to 400 organisms/ ~0.88 L), which may have impaired flushing. This limits the available space for recruitment of sessile organisms. Unfortunately, due to continuing loss of experimental shading plates the experiment could not be continued over a longer period of time and thus we were not able to test whether patterns found in the early succession persist in mature assemblages. The results of this experiment however suggest that the effect of shading on intertidal communities is complex. It is likely that the provision of a combination of different light levels can best replicate the current heterogeneity of conditions on natural rocky shores and benefit the recruitment of both sessile and mobile assemblages.

Chapter 5

Conservation priorities for intertidal rocky shores assessed with remote sensing

5.1 Abstract

Sea level rise is an inevitable consequence of climate change and threatens coastal ecosystems, particularly intertidal habitats that are constrained by landward development. Intertidal habitats support significant biodiversity, but also provide natural buffers from climate-threats such as increased storm events. Predicting the effects of climate scenarios on coastal ecosystems is important, for both understanding the degree of habitat loss for associated ecological communities and the risk of the loss of coastal buffer zones. We quantified the extent of horizontal intertidal rocky shores along ~200 kms of coastline using GIS and remote-sensing (LIDAR) and used this information to predict changes in extent under four different climate change driven sea level rise scenarios (IPCC RCP2.6, RCP4.5, RCP6.0, RCP8.5). We then applied the IUCN Red List of Ecosystems Criterion C2 to estimate the status of this ecosystem in the Hawkesbury Shelf Marine Bioregion. We also used four individual rocky shores to investigate the role of local topography in determining the severity of sea level rise impacts. We found that if the habitat loss within the study area is representative of the entire bioregion, the IUCN status of this ecosystem is ‘near threatened’, assuming that an assessment of the other criteria would return lower categories of risk. There was, however, high spatial variability in this effect. Rocky shores with gentle slopes had the highest projected losses of area whereas rocky shores with a greater range of elevation were less affected. Among the sites we surveyed in detail, the ecosystem status ranged from ‘least concern’ to ‘vulnerable’, but reached ‘endangered’ under upper estimates of the most severe scenario. This study demonstrates the application of remote sensing techniques for mapping the elevation of intertidal rocky shores, from which the impact of sea level rise can be estimated. The findings of this study further emphasise the need for widespread, high resolution monitoring data across coastal systems. They also have important implications for planning resolutions that aim to protect and enhance the resilience of coastal environments.

80 Chapter 5 81

5.2 Introduction

Climate change threatens marine ecosystems at a global scale through changes in temperature, ocean acidification and sea level rise (Brierley and Kingsford, 2009, Doney et al., 2012, Hoegh- Guldberg and Bruno, 2010). Sea level rise is a consequence of thermal expansion of the ocean and the melting of water stored in glaciers and ice-caps (Church et al., 2011, IPCC, 2013). Under climate change, sea level rise has been projected to exceed previously observed rates (IPCC, 2013), but predictions for the level of this rise vary depending on the amount of anthropogenic contributions (in the form of emissions and land-use change) to radiative forcing. Four scenarios developed by the International Panel for Climate Change (IPCC) are used to represent the effect of radiative forcing in 2100, relative to preindustrial levels: RCP2.6, RCP4.5, RCP6.0, RCP8.5 (IPCC, 2013). Under these scenarios, sea level rise is expected to increase at a rate of 4.4, 6.1, 7.4 and 11.2 mm/year (values represent median values), respectively (IPCC, 2013).

Sea level rise will have the greatest ecological impact along low lying coastlines through increasing inundation of the intertidal zone, which supports important ecological communities such as mangroves, seagrasses, saltmarshes and rocky shores (FitzGerald et al., 2008, Nicholls and Cazenave, 2010, Nicholls et al., 1999). Apart from providing habitat for intertidal biodiversity, these communities provide a buffer from destructive ocean forces, reducing the impact of storm events and mitigating erosion (Gedan et al., 2011, Shepard et al., 2011, Spalding et al., 2014). It is therefore important to quantify the risks to important coastal habitats from sea level rise. In this study we focus on intertidal rocky shores in the Sydney region, from which we estimate the threat of sea level rise to intertidal rocky shores in the Hawkesbury Shelf Marine Bioregion.

Intertidal rocky shores are the most common coastal habitat worldwide and are ecologically valuable (Thompson et al., 2002). They support a diverse array of species, which is attributed to the high structural complexity of rocky shores (Blanchard and Bourget, 1999, Chapman, 2003, Sebens, 1991). Intertidal rocky shores and the communities living on them provide numerous ecosystem functions and services. Filter-feeders such as oysters improve water quality and further promote biodiversity by creating additional habitat for other intertidal organisms (Coen et al., 2007, Grabowski et al., 2012). It is also suggested that intertidal rocky shores are important nursery and feeding grounds for fish during high tide and shorebirds during low tide (Burrows et al., 1999, Rangeley and Kramer, 1995, Cantin et al., 1974). Yet rocky shores are also amongst the most vulnerable marine systems, facing a variety of anthropogenically induced threats (Halpern et al., 2007, Thompson et al., 2002).

Because the tidal regime of alternating emersion and submersion is the key physical driver on intertidal rocky shores (Menge and Branch, 2001), rapid changes in sea level can have particularly

Chapter 5 82

severe consequences for the availability of habitat. A study of rocky shores in Scotland using 3D photogrammetry found that a rise in sea level between 0.3-1.9m will result in a loss of rocky 10%- 50% of rocky shore extent and a steepening of slope with at least 50% of shorelines becoming vertical (≥ 45o) under a 1.9m sea level rise scenario (Jackson and McIlvenny, 2011). The transition to a steeper relief from a flat rocky shore may compress organisms and increase pressure from competition, particularly in areas where static vertical barriers such as seawalls prevent a landward migration (Pontee, 2013). Yet little is known about the effect of sea level rise on intertidal rocky shores in other areas of the world or in the context of multiple climate change scenarios. Sea level rise is an inevitable consequence of climate change, and understanding the possible negative consequences is essential to inform conservation and mitigate impacts.

Here, we apply the IUCN Red List of Ecosystems criteria (Keith et al., 2013) to ~200 km of Sydney coastline in order to estimate the current status of intertidal rocky shores in the Hawkesbury Shelf Marine Bioregion and discuss potential effects of sea level rise on associated biota. Under the IUCN system the status of an ecosystem is assessed against 5 criteria, with the final ecosystem status determined based on the highest risk returned for any one category. Here, we focus on criterion C2, which involves an assessment of the extent and relative severity of habitat degradation in the next 50 years (Table S.4.1). We used a high-resolution LiDAR (light detection and ranging) survey of coastal elevation to estimate the net loss/gain of intertidal rocky shores under sea level rise scenarios. As this ecosystem is defined by the intertidal regime, the relative severity of the loss or gain of available area was assumed to be 100%.

5.2.1 Ecosystem description

5.2.1.1 Abiotic features

The rocky shores in this study and in the Sydney region in general, are gently sloping (Chapman, 2003, Chapman and Bulleri, 2003), with the landwards edge backed by cliffs. The tidal range is ~2 meters (MHL 2016) and wave exposure, although not measured, is likely to be higher along the open coast than within the two embayments (Sydney Harbour and Botany Bay).

5.2.1.2 Biota

In Sydney Harbour, approximately 162 taxa can be found on natural rocky shores (Mayer-Pinto et al., 2018). There is, however, high spatial variability in the distribution of species, which is common in Sydney Harbour (Chapter 2) and New South Wales (Underwood and Chapman, 1998). A long-term survey within Sydney Harbour revealed that the mobile fauna in rock pools

Chapter 5 83

in the outer zone of Sydney Harbour is dominated by Bembicium nanum, Austrocochlea spp. and Nerita atramentosa, whereas the inner zone of Sydney Harbour is characterised by greater abundances of B. auratum, Patelloida spp. and Siphonaria spp. (Chapter 2). Similarly, sessile species composition within rock pools changes within the Harbour, with the brown encrusting algae Ralfsia spp., the bryozoan Watersipora spp. and the polychaete Galeolaria caespitosa. being more common in the outer zone, and Corallina spp. and oysters being more abundant in rock pools in the inner zone (Chapter 2).

Besides this variability in composition over larger spatial scales, intertidal communities also show distribution patterns on smaller spatial scales, particularly across the vertical gradient. The composition of assemblages within each zone is determined by physical stressors such as temperature and desiccation stress, predation and competition (Connell, 1972, Raffaelli and Hawkins, 2012, Connell, 1961, Underwood, 1980). Intertidal rocky shores can be broadly categorized into three main zones: the low, mid- and high intertidal, which vary in their heights above sea level, representing areas of different exposure (Druehl and Green, 1982). General patterns of the vertical distribution of major groups are however found. Upper sections of rocky shores tend to separate into zones of littorinids (e.g. Austrolittorina unifasciata), followed by a barnacle (e.g. Chthalamus antennatus), and Galeolaria sp. (e.g. Galeolaria caespitosa) zone, whereas the lower intertidal is characterised by tunicates (e.g. Pyura praeputialis) (Stephenson and Stephenson, 1972). Common mobile species in the middle part of the rocky shores include various gastropods and chitons (e.g. Cellana tramoserica, Austrocochlea spp., Tenguella marginalba, Sypharochiton pelliserpentis) (Stephenson and Stephenson, 1972).

5.3 Materials and methods

5.3.1 Study location

The Hawkesbury Marine Shelf Bioregion includes the estuaries, coastline and marine waters from Newcastle to Wollongong, NSW, Australia (Fig. 1). The assessment of intertidal rocky shores in this study was limited to the coastal area between the northern site of Port Hacking (34o03’55’S, 151o08’05’E) in the south and the south end of Fishermans Beach (33o44’12’S, 151o18’22’E) in the north (Fig. 5.1), based on data availability. The assessment also included Sydney Harbour up until the Spit Bridge and the Harbour Bridge and Botany Bay up until the Captain Cook Bridge (Fig. 5.1). This represents approximately 210 km length of coastline and is considered representative of the greater bioregion.

Chapter 5 84

5.3.2 Habitat mapping

To define the extent of intertidal rocky shores a shapefile was digitised from publicly available aerial imagery (DFSI) using the editor tool in ArcMap (version 10.4.1). Upper (landward) limits of rocky shores were determined based on visible rock outcropping, while lower limits were extended well past the minimum low-tide area, for refinement later in the analysis based on tidal and elevation data. Vertical intertidal surfaces such as tall cliffs and intertidal boulder fields were avoided where possible. When shading or vegetation prevented a proper view of the intertidal areas, other sources of imagery were reviewed to obtain outlines (e.g. Google Earth). If areas could not be visually assessed from available imagery, the area was not incorporated in the shapefile.

5.3.3 Elevation data

Elevation data of the intertidal rocky shores was acquired using LIDAR, which measures reflected laser pulses (Bachman, 1979). LIDAR data were collected between the 10-24.04.2013 (LPI NSW 2013). The LIDAR data used in the following analysis were compiled from two separate datasets

Chapter 5 85

(Sydney North and Sydney South) with an average point density of 1.57 and 1.56 points per square meter, respectively. The horizontal spatial accuracy was 0.8m and the vertical spatial accuracy 0.3m. Elevation values refer to the zero level on the Fort Denison Tide Gauge (Zero Fort Denison = ZFD), being approximately the level of the Lowest Astronomical Tide (LAT). Due to the marginal differences between the datasets, they were combined without any correction. To quantify the current extent of rocky intertidal shores, rocky shores were further divided into three intertidal zones (low, mid and high intertidal). The low intertidal zone was defined as the area of less than 0.6m above LAT, the mid intertidal zone covered the area between 0.6m and less than 1.3m above LAT and the high intertidal zone reached up to a height of 1.9m above LAT.

5.3.4 Data cleaning and spatial analyses

Before analysis was conducted the data was cleaned based on the tidal height and unsupervised classification of LIDAR returns, and assigned to an intertidal zone (high, middle and low). To avoid including the elevation of water covering the intertidal surface, elevation data that was below the tidal height at the time of collection was excluded based on the local (Fort Denison) tide gauge (33° 51’ 16.8’S, 151° 13’ 32.8’E) (OEH NSW 2015). Tide data was interpolated linearly from an hourly measure to a per minute estimate to match the LIDAR timecodes. As an additional measure, LIDAR points underwent an unsupervised classification at the time of measurement, and any points classified as ‘water’ were excluded. Elevation data within the rocky shore polygons was then standardised to a 1m resolution for analysis of the total area. Because the LIDAR data was acquired at varying tide heights, the numbers of points available for assessment in the low, mid and high intertidal zones varied, and this was accounted-for in the analysis as best possible.

5.3.5 Predicted sea level rise scenarios

Four scenarios (RCP2.6, RCP4.5, RCP6 and RCP8.5) of projected global mean sea level rise were taken from the 5th Assessment Report of the Intergovernmental Panel on Climate Change (Church et al., 2013). Current sea level was set at 0m ZFD/LAT in 2013.

To satisfy the criteria for the IUCN Red List of Ecosystems, we calculated sea level rise for the year 2063 (50 years from when the data was collected) using the predicted linear rate of global mean sea level rise under all scenarios (Table 5.1) and compared the predicted intertidal area to its current extent. Additionally, we selected 4 intertidal rocky shores from the dataset to predict the loss/gain of intertidal area in relation to different topographies and spatial extent. Two of the rocky shores (Bradleys Head and Delwood) were gently sloping and backed up by beach, cliffs

Chapter 5 86

or bushland, while the other two shores were large areas with a greater range of elevation (Freshwater and Cape Banks) (personal observation).

RCP2.6 RCP4.5 RCP6.0 RCP8.5 Rate of GMSLR 4.4 [2.0 - 6.8] 6.1 [3.5 - 8.8] 7.4 [4.7 - 10.3] 11.2 [7.5 - 15.7] SLR in 50 years 220 [100 - 340] 305 [175 - 440] 370 [235 - 515] 560 [375 - 785]

5.4 Results

The current area of intertidal rocky shoreline is estimated at 374,689 m2 along approximately 210 kms of the NSW coastline. This represents 1,392 m2 in the low, 143,616 m2 in the mid, and 229,681 m2 in the high intertidal zones, though these areas are likely to be underestimates due to the lack of reliable elevation data available in the low and mid-intertidal zones (see methods: spatial analysis).

Model predictions based on median sea level rise rates suggest that the available habitat for intertidal organisms will be reduced over the next 50 years at an overall rate of ~0.11%/year, ~0.17%/year, ~0.24%/year and ~0.43%/year under scenarios RCP2.6, RCP4.5, RCP6.0 and RCP8.5, respectively. The rate is however not linear and accelerates with time (Fig. 5.2a, Fig. S4.1). In 50 years, the compounding effect of sea level rise is predicted to reduce the area of current rocky shoreline by at least ~5.85% under the most benign scenario (RCP 2.6) and a maximum of ~21.75% under the most extreme scenario (RCP 8.5) (Fig. 5.2b, Table S4.2). Losses can however exceed 30% over 50 years, based on the upper range of sea level rise projected under the RCP 8.5 scenario (Table S4.2).

Chapter 5 87

Figure 5.2 Remaining area (%) of all rocky shores combined (a) in each year for the next 50 years and (b) in 50 years.

At the four sites selected for detailed investigation, the area of rocky shore, topography, and relative severity of sea level rise effects varied. Cape Banks was the largest rocky platform assessed in detail, with a total current intertidal area of 18,304 m2, followed by Bradleys Head and Freshwater (5,864 m2 and 5,295 m2, respectively) and Delwood (2,352 m2). All of these shores were gently sloping, but the main rocky platforms along the headland of Bradleys Head and at Delwood were constrained by beaches, bushland and cliffs, whereas Cape Banks and Freshwater were predominantly rocky habitats.

Among these four sites the extent of loss of intertidal rocky shore area varied greatly, with the greatest losses projected to occur at Delwood (Fig. 5.3a, Fig. 5.4a, e, i) - at least ~14% (RCP2.6) and up to ~37% (RCP8.5) (Table S4.2). At Bradleys Head, the intertidal area is projected to stay almost constant under all predictions, except the RCP8.5 scenario where more than 30% of area is likely to be lost (Fig. 5.3b, Fig. 5.4b, f, j) (Table S4.2). Under the upper ranges of sea level rise for scenario RCP8.5, both Delwood and Bradleys Head rocky shores are reduced by more than half of their current extent (Table S4.2). Rocky shores at Freshwater (Fig. 5.3c, Fig. 5.4c, g, k) and Cape Banks (Fig. 5.3d, Fig. 5.4d, h, l) were less affected than Delwood or Bradleys Head. Losses at Freshwater ranged between ~3-19% (RCP 2.6, RCP 8.5), and the platform at Cape Banks diminished under scenario RCP2.6 (~9%) but was not constrained in available landwards rock, with smaller losses under the more severe scenarios (Table S4.2).

Chapter 5 88

Figure 5.3 Remaining area (%) in 50 years at (a) Delwood, (b) Bradleys Head, (c) Freshwater and (d) Cape Banks.

Chapter

Exemplary areas of rocky shoreline were restricted to points within the shapefile. 89

Chapter 5 90

1 5.4.1 Classification with IUCN Red List of Ecosystems

2 Using the estimates of available habitat under sea level rise scenarios, and assuming that sea level 3 rise is the most serious threat to this ecosystem, the overall threat of sea level rise to intertidal 4 rocky shores in the Hawkesbury Marine Shelf bioregion against criterion C2 of the IUCN 5 framework does not exceed the threshold for ‘vulnerable’ classification except under the upper 6 predictions of RCP8.5 and should therefore be classified as ‘near threatened’. At a site-specific 7 level, however, these predictions breach the threshold for ‘vulnerable’ under scenario RCP8.5 at 8 Delwood and Bradleys Head, and are elevated to ‘endangered’ under the upper projections of the 9 RCP8.5 scenario. Losses at Freshwater do not reach the status of ‘vulnerable’ under any of the 10 RCP scenarios, but should be classified as 'near threatened’ due to high losses under RCP8.5. 11 Losses at Cape Banks are limited, even under upper predictions of scenario RCP8.5 and should 12 therefore be classified as ‘least concern’. 13

14 5.5 Discussion

15 Coastal ecosystems are changing worldwide as a result of climate change impacts, including sea 16 level rise. Here we quantify the sea level rise threat to rocky intertidal communities along ~200 17 km of coastline in SE Australia. We found that sea level rise is likely to alter the extent of intertidal 18 rocky habitat to such a degree that rocky shores in this bioregion should be considered ‘near 19 threatened’ under the IUCN Criterion C2. However, the threat of sea level rise for rocky shores 20 is predicted to vary spatially and is linked to local topography and the landward availability of 21 rocky surfaces. Higher predicted losses of rocky shoreline occurred at locations where gently 22 sloping shores are constrained by beach and bushland, whereas shores with extended rocky slopes 23 allowed for migration of the intertidal zones. To protect coastal biodiversity and preserve buffers 24 between ocean forces and coastal infrastructure, similar efforts are required to identify suitable 25 sites for the long-term preservation of intertidal habitats, for example with non-construction zones 26 in areas of coastal development, or proactive surface preparation to allow for species migration. 27 The results of this study highlight the importance of forecasting sea level rise effects using 28 detailed spatial data, without which we may fail to notice the small-scale losses of coastal habitat 29 that are likely to act as precursors to larger, regional-scale losses of ecosystem function and 30 biodiversity. 31 The loss of intertidal rocky shores can have potential negative consequences for adjoining habitats 32 such as beaches and bushland. Rocky shores can dissipate wave action and thus prevent water 33 reaching areas higher on shore. The loss of the mitigating function of rocky shores may therefore 34 lead to flooding of adjoining habitat, for example bushland, but also to greater sediment influx

Chapter 5 91

35 into the water when adjoining to beaches, which can in turn alter adjoining subtidal ecosystems 36 (Airoldi, 2003).

37 Sea level rise threatens rocky shore communities by forcing organisms into a coastal squeeze in 38 two ways: 1) loss of upper tidal zones where barriers (e.g. cliffs, bushland) prevent a shift of 39 communities further up shore; and 2) a decrease in the overall area within each zone. Rocky shore 40 communities display vertical zonation patterns, which are a result of physiological limitations as 41 well as competition and predation (Connell, 1961, Connell, 1972, Raffaelli and Hawkins, 2012, 42 Underwood, 1980). In the first case, it is likely that only species found in the impacted zone will 43 be affected, due to reduced habitat and replacement by organisms migrating up the shore from 44 lower elevations. Species adapted to upper elevations have no room to move, and therefore could 45 experience greater predation and competition pressure (Connell, 1961). As lower vertical limits 46 are often determined by predation and competition pressure (Connell, 1972), it is likely that 47 increases in predator density and competitively more dominant species will reduce the abundance 48 and increase the risk of local extinction of former ‘upper shore species’. This would mostly affect 49 gastropods of the family Littorinidae and many barnacles, since they mainly occupy areas higher 50 on shore (Stephenson and Stephenson, 1972). In the second case, a decrease in extent in area can 51 increase biotic interactions within each zone, which can alter relative abundances of the species 52 (Underwood, 1978) and can result in overall altered communities.

53 Loss of horizontal rocky platforms can also result in the fragmentation of shoreline habitat. This 54 spatially separates remaining patches of habitat and results in a decrease in diversity along the 55 coastline (Goodsell et al., 2007). Greater distances among rocky shore communities can further 56 prevent migration and gene flow among communities (Saunders et al., 1991). Although vertical 57 rocky shores support similar suites of species (Chapman and Bulleri, 2003) and thus provide a 58 suitable medium for species dispersal and migration, those with little adhesive strength may not 59 be able to use them as a stepping medium to overcome long distances due to greater physical 60 forces from wave action on vertical walls and thus increased risk of dislodgement (Denny et al., 61 1985). Furthermore, in areas where planktonic dispersal of rocky shore larvae and propagules is 62 limited by reduced water flow (e.g. inner estuarine embayments (Dafforn et al., 2012b, Das et al., 63 2000), the distance between suitable rocky substrates could become insurmountable if some 64 shores are lost.

65 Localized losses of rocky shores may further result in local extinctions of species with limited 66 distributions. For example, in a survey of intertidal rock pools in Sydney Harbour, the gastropod 67 Bembicium auratum was found to only occur in the inner zone of Sydney Harbour (Chapter 2). 68 The inner zone of Sydney Harbour is already characterized by fewer rocky reef compared to the

Chapter 5 92

69 outer zone (Johnston et al., 2015), and thus complete inundation of rocky shores in these areas 70 would be detrimental to its persistence in Sydney Harbour.

71 Besides the overall loss of intertidal area, sea level rise can further result in reduced habitat 72 heterogeneity and complexity of the remaining rocky shore through the disappearance of 73 important intertidal microhabitats such as rock pools, which are mostly present on horizontal 74 rocky shores. A study assessing the impacts of sea level rise on intertidal rocky reefs on a marine 75 park on the north coast of NSW found losses of lower shallow pools despite low losses of lower 76 platform habitat (Thorner et al., 2014). Upper shallow pool habitats and deep pool habitats also 77 showed declines (Thorner et al., 2014).Rock pools are the main habitat of many species on natural 78 rocky shores, such as sea urchins, large whelks and starfish (Chapman, 2003), which could 79 become less abundant or absent with the loss of this microhabitat.

80 The increased rate of sea level rise may also prevent the creation of rock pools in places were 81 rocky substrate is available higher on the shore. Rock pools are formed by erosion and abrasion 82 of rock and slowly happens over time. Predicting the effects of sea-level rise is important to take 83 early measures for the conservation of these habitats (Dean and Connell, 1987, Loke and Todd, 84 2016, MacArthur and MacArthur, 1961). Reduced habitat heterogeneity and complexity can 85 therefore lead to a decrease in diversity within remaining rocky shorelines.

86 The present study estimated the loss of intertidal rocky shores in the Sydney region and found 87 that rocky shores are threatened by sea level rise. However, estimates of the current intertidal area 88 do not account for some elevations that were covered by water and therefore not included in the 89 dataset. Additionally, the slope of intertidal areas was not accounted for in the estimates. The area 90 between two adjoining points can be greater when this area is characterized by steep topography 91 in contrast to a horizontal plane. Although only horizontal rocky shores were included in the 92 analyses to limit the effect of slope, this may have contributed to a slight underestimation of the 93 current shoreline. Furthermore, calculations were based on one point per square meter and not a 94 digital surface model. Higher resolution would enhance the accuracy of the models and would 95 therefore be beneficial in future analyses. Nevertheless, this study shows that remote sensing 96 techniques can be a useful tool to assess potential effects of sea level rise on rocky shores, and 97 thus rocky shore communities. 98

99 5.6 Conclusion

100 This study highlighted the high degree of risk to intertidal rocky shore environments from sea 101 level rise. We found that the IUCN threshold for ‘vulnerable’ classification is not exceeded for 102 median estimates of sea level rise in the next 50 years, but it is certain that rising sea levels will 103 increase the inundation of rocky shores into the future. The status of ‘near threatened’ highlights

Chapter 5 93

104 the need for ongoing assessment of this habitat such that proactive management and conservation 105 strategies can be implemented. Extending sea level rise scenarios past the 50-year mark can help 106 prioritising management actions in areas that are at most risk from sea level rise. This analysis 107 may help to identify suitable sites for protection from coastal construction, to assure the 108 persistence of this important habitat.

Chapter 6 94

109 Chapter 6

110

111 General discussion

112 This research was performed to better understand the ecology of intertidal rocky shores and assess 113 their vulnerability to urbanisation and climate change. The knowledge gained from this work can 114 now be used to inform conservation and management practices worldwide, although care must be 115 taken to understand context specificity given the restricted scale of the study region. These results 116 may be used to improve current practices of ecological engineering as well as to increase the 117 awareness of the vulnerability of this important ecosystem.

118 Urbanisation is one of the most pressing stressors to coastal ecosystems worldwide. In marine 119 systems, the effect of urbanisation often occurs in the form of the replacement of natural habitats 120 by artificial structures of reduced habitat complexity and/or heterogeneity. Habitat structure, 121 which incorporates both complexity and heterogeneity (McCoy and Bell, 1991), is a major 122 mechanism that supports biological diversity in natural systems (Dean and Connell, 1987, 123 MacArthur and MacArthur, 1961, Tews et al., 2004). Understanding how intertidal communities 124 respond to habitat structure at different scales can help to predict the potential outcomes of the 125 homogenisation of habitat.

126 Intertidal rock pools are diverse habitats on the intertidal (Firth et al., 2014, Firth et al., 2013, 127 Astles, 1993, Underwood and Skilleter, 1996). The effect of habitat structure, including the effect 128 of rock pool characteristics (i.e. size and height on shore) on associated communities has been the 129 focus of research in both natural and artificial rock pools (Browne and Chapman, 2014, Chapman 130 and Blockley, 2009, Firth et al., 2014, Firth et al., 2013, Evans et al., 2016, Astles, 1993, 131 Underwood and Skilleter, 1996). However, these relationships have not been examined among 132 locations of varying environmental conditions in urban areas, making my research an important 133 step in furthering the knowledge and understanding of rock pool ecology on an increasingly 134 urbanised coast. Specifically, I examined whether local environmental conditions at two locations 135 in an urban estuary, Sydney Harbour, influence the relationships between parameters of habitat 136 structure (size and height on shore) and associated biological assemblages.

137 I found that the physical features of pools associated with greater number of species varied with 138 the type of assemblage examined (sessile vs. mobile taxa) as well as with the local species pool, 139 which was highly linked to the position of pools within Sydney Harbour (i.e. inner vs. outer zone). 140 For example, the positive effect of increasing maximum width on the richness of sessile taxa, and 141 maximum depth on mobile taxa richness was limited to the outer zone of Sydney Harbour. Height

Chapter 6 95

142 on shore for sessile taxa and width and volume for mobile taxa were important factors influencing 143 richness in the inner and outer zone of the harbour. These findings suggest that local 144 environmental conditions and the local species pool play an important role in structuring rock 145 pool communities, supporting other studies that tested the effect of habitat structure at different 146 locations (Browne and Chapman, 2014, Firth et al., 2014). The context dependent nature of the 147 relationship between rock pool size and diversity emphasized that predicting outcomes of 148 manipulations of size parameters on diversity specifically requires knowledge of the surrounding 149 environmental conditions and the local species pool.

150 I further investigated the effect of smaller features of habitat structure. It is widely accepted that 151 the physical complexity of habitats is an important factor in the composition of communities 152 (Dean and Connell, 1987, Loke and Todd, 2016, MacArthur and MacArthur, 1961). To fully 153 understand the drivers of biodiversity, habitat structure needs to be assessed at the scale that is 154 appropriate to the organisms investigated (Matias et al., 2010). Despite extensive research on rock 155 pools, to my knowledge, no study has assessed how fine-scale features within rock pools (i.e. 156 microhabitats) can affect associated species. Here, I specifically investigated whether 157 microhabitats within natural rock pools affect the abundance of mobile taxa.

158 I showed that the presence of microhabitats can affect the abundance of mobile taxa within rock 159 pools, but that this effect was dependent on the type of microhabitat and was also taxa-specific. 160 These findings highlight the need to assess the effect of habitat structure at a scale that corresponds 161 to the studied organism, such that potentially important drivers of diversity are not overlooked. It 162 also highlights the need to examine the effect of habitat structure at various scales, e.g. by 163 differentiating among microhabitats and analysing overall – and taxon-specific responses, to 164 allow for a detailed understanding of patterns of biodiversity and its potential drivers. This work 165 revealed correlations between some physical features (overhangs and oyster shells) within rock 166 pools and mobile macroinvertebrate abundances. The information gathered from this survey 167 provides valuable information that can now be used as a starting point for experimental 168 manipulations to test for these effects whilst controlling for confounding factors such as the 169 abundance of sessile organisms and rock pool size.

170 The development of coastal infrastructure is not only characterised by the replacement of natural 171 surface habitats by artificial structures, but also by the shading of natural and artificial habitats by 172 artificial structures, for example high-rise buildings or structures associated with boating activities 173 such as jetties and wharves. While the effect of shading has been studied in intertidal areas 174 (Blockley, 2007, Pardal‐Souza et al., 2017), the effect of different levels of shading has not been 175 considered. Shading is an important factor in determining the abundance of algae and mobile and 176 sessile invertebrates (Miller and Etter, 2008, Takada, 1999), and thus determining thresholds of

Chapter 6 96

177 levels of shading can help to understand how communities may respond when natural light 178 regimes are altered.

179 In my research, I found that the level of shading had different effects on the cover of sessile 180 organisms and the abundance of mobile organisms. I found that algal cover was highest at 181 intermediate light levels, whereas highest numbers of mobile organisms were found under fully 182 shaded treatments. No sessile invertebrates or invaders were found. This result supported my 183 initial hypothesis and is likely driven by the fact that algae require light for photosynthesis (Franck 184 and Loomis, 1949) but may be damaged by extensive heat or UV stress and desiccation (full light 185 conditions). This experiment was conducted in summer, and thus algae exposed to sunlight were 186 exposed to high temperatures and strong sunlight. There was no increase in cover of sessile 187 invertebrates with decreasing light intensity. This is unusual as invertebrate cover is usually 188 enhanced in shaded conditions while algae cover decreases (Miller and Etter, 2008). However, 189 the experiment was conducted for 6 months only, and thus long-term effects of shading could not 190 be examined. Higher abundances of mobile organisms under fully shaded treatments are also 191 potentially linked to the fact that these areas provided better thermal refuge, as shaded areas are 192 usually cooler. Further experiments are needed to unravel the factors driving the observed 193 patterns.

194 Climate-driven sea level rise is an imminent threat to coastal ecosystems (FitzGerald et al., 2008, 195 Nicholls and Cazenave, 2010, Nicholls et al., 1999). The assessment and quantification of the 196 vulnerability of ecosystems can be used to identify ecosystems most at risk of losing biodiversity 197 and help to set priorities for conservation planning and management as well as support a 198 sustainable use of land and water (Keith, 2015). Intertidal rocky shores, a key coastal ecosystem, 199 are biologically diverse (Thompson et al., 2002), but to date, their vulnerability to sea-level rise 200 has not been assessed. Given the fact that intertidal areas are the first areas affected by sea-level 201 rise, there is an urgent need to predict the extent of the potential impacts.

202 I explored the impact of sea-level rise on the current extent of intertidal rocky shores in the Sydney 203 region in order to estimate the current status of intertidal rocky shores in the Hawkesbury Marine 204 Shelf Bioregion. To estimate the threat, I used the IUCN Red List of Ecosystems Criterion C2, 205 which determines the risk of an ecosystem towards an environmental threat using the 206 environmental degradation over a 50-year period (Keith et al., 2013). If the percentage habitat 207 loss found for intertidal rocky shores in Sydney is consistent for the bioregion of the intertidal 208 community, the regional status of this ecosystem is likely ‘near threatened’, but ‘vulnerable’ in 209 places, with very flat intertidal areas being most at risk. This will impact associated communities 210 by reducing available habitat, and will likely result in shifts in community structure. Predicting

Chapter 6 97

211 the effects of sea-level rise is important to take early measures for the conservation of these 212 habitats.

213 The results of my thesis suggest that ecological patterns within intertidal communities and 214 responses of these communities to anthropogenic disturbances are complex and that management 215 strategies will need to have clear goals from the outset, taking into account local biota and 216 environmental conditions if desired outcomes are to be achieved. 217

218 6.1 Management implications

219 Sustainable development and the conservation of biodiversity are important goals of the 220 management of coastal development policies across the globe (Dafforn et al., 2015b, Naylor et 221 al., 2012). Below, I specifically discuss how the research of my thesis can be practically used to 222 inform and improve current management practices of intertidal rocky shores.

223

224 6.1.1 Ecological engineering

225 ‘Ecological enhancement’, which aims to improve the ecological value of infrastructure built for 226 non-ecological purposes (Naylor et al., 2011), is one of the management strategies currently being 227 used to achieve these goals (Dafforn et al., 2015b, Naylor et al., 2012). For this to happen, the 228 practice of ecological enhancement should be evidence-based, and thus rigorous scientific 229 research as a basis for these practices is needed. Ecological engineering designs, which 230 incorporate ecological principles in the design of artificial structures (Bergen et al., 2001) are one 231 way to increase the ecological value of built infrastructure.

232 Artificial structures are often characterized by lower diversity when compared to natural habitats 233 (Chapman, 2003, Firth et al., 2013). It is thought that the homogeneous structure of artificial 234 structures is a major contributor to their low diversity, and in response a variety of ecological- 235 engineering designs to increase their habitat structure have been implemented (Strain et al., 2018). 236 The addition of water-retaining has been identified as one of the most successful interventions 237 that increased the diversity of artificial structures (Strain et al., 2018, Bugnot et al., 2018). Past 238 designs of artificial rock pools have however only been of simple shape and have largely 239 neglected to incorporate or manipulate the complexity within rock pools, normally observed in 240 natural habitats. Advancements in technology, such as 3D printers, now allow for more 241 sophisticated designs, which can be used to reach specific targets.

242 Findings from my thesis have shown that the size and position within the shore of artificial rock 243 pools can be manipulated to reach certain objectives, for example increasing sessile and/or mobile

Chapter 6 98

244 diversity. However, the success of manipulations depends on the local environmental conditions, 245 highlighting that site-specific designs are required to maximise the outcomes of ecological 246 engineering designs. I also found that the presence of overhangs within rock pools significantly 247 increased the abundance of the native gastropod Nerita atramentosa. Artificial structures like 248 seawalls are often characterized by low abundances of mobile organisms (Chapman, 2003) and 249 the inclusion of these fine-scale features in the design of artificial structures can help counteract 250 this problem. The increase may also help prevent colonisation by invasive algae. Artificial 251 structures have been shown to facilitate invasions (Airoldi et al., 2015, Bulleri and Airoldi, 2005, 252 Bulleri and Chapman, 2010), and increases in the abundance of native species may therefore 253 increase their biotic resistance towards invasion. Biotic resistance can form through competition, 254 predation, herbivory, disease and abiotic factors, and herbivory has been identified to have a 255 strong negative effect on invader establishment and individual performance (Levine et al., 2004). 256 Thus, increases of native grazers can potentially prevent/hinder the success of non-indigenous 257 algae.

258 In addition to modifications of habitat structure, my research also highlighted that different levels 259 of shading can affect associated communities, but that effects varied with the community assessed 260 (i.e. algae vs. mobile). These findings suggest that the effect of shading on intertidal communities 261 is complex, and that shading should be considered when artificial structures are built near 262 intertidal habitats. However, it also raises the potential for shading to be used to manipulate 263 cover/abundances of associated organisms. To increase both algal cover and the abundance of 264 mobile organisms, mitigation strategies will need to consider design approaches that suit both 265 types of organisms. For mobile organisms, this could include the addition of fully shaded 266 microhabitats such as crevices and overhangs, which have previously been shown to reduce heat 267 and desiccation stress by changing the microclimate (Garrity, 1984, Gray and Hodgson, 2004). If 268 algal productivity is desired – then mitigation strategies that reduce shading and increase light 269 intensities, such as light-penetrating surfaces, should not aim to recreate natural light conditions 270 but rather incorporate strategies that lift light to intermediate levels. 271

272 6.1.2 Conservation

273 In this thesis, I estimated the habitat loss of intertidal rocky shores in the Hawkesbury Shelf 274 Marine Bioregion using predicted habitat loss of intertidal rocky shores in the region between 275 Port Hacking and Long Reef, assuming sea level rise was the most serious threat and an 276 assessment of other criteria would return lower categories of risk. If the habitat loss found was 277 consistent for the whole bioregion, we would classify the regional status as ‘near threatened’ 278 according to IUCN Red List of Ecosystem Criterion C2. There was however spatial variability in

Chapter 6 99

279 the threat towards sea level rise among different shores. Given the progressive state of climate 280 change, the identification of high- and low-impact zones can help prioritize protection zones. 281 Intertidal rocky platforms least at risk could be targeted as protected areas, for example as no- 282 construction zones for coastal development, to conserve areas of natural habitat that are self- 283 sustaining under sea-level rise. Assessment of fine-scale features, such as the presence of 284 microhabitats, can aid conservation goals by providing information about the spatial scales of 285 protection needed for large (habitat) and fine-scale (microhabitat) levels, to increase conservation 286 goals (Banks and Skilleter, 2007).Additionally, it can help identify areas where artificial habitats 287 need to be created to provide areas for retreat of intertidal organisms and as migration corridors 288 to decrease fragmentation among communities. However, more fine-scale datasets are needed to 289 further enhance its predictive capacity. 290

291 6.2 Future directions and final remarks

292 The accelerating influx of human populations to coastal areas requires structures that protect 293 coastal areas from erosion and sea-level rise. In highly urbanised areas, habitat restoration and 294 soft-engineering approaches, which involve modifications of natural habitats rather than their 295 replacement, are not always a viable option and the construction of further coastal defence 296 infrastructure is inevitable (Mayer-Pinto et al., 2017). In these cases, ecologically engineered 297 artificial structures, which fulfil their function as protective infrastructure while simultaneously 298 supporting biodiversity and providing important services, can be a good option. Successful eco- 299 engineering interventions require, however, a greater knowledge of processes within natural 300 systems and need to be designed within a site-specific context (Bergen et al., 2001). This thesis 301 has increased new knowledge and understanding of the ecology of rocky shores, which may be 302 used to inform improvements in future management strategies.

References 100 References

AIROLDI, L. 2003. The effects of sedimentation on rocky coast assemblages. In: GIBSON, R. & ATKINSON, R. (eds.) Oceanography and Marine Biology: An Annual Review. CRC Press.

AIROLDI, L. & BECK, M. 2007. Loss, status and trends for coastal marine habitats of Europe. In: GIBSON, R. & ATKINSON, R. (eds.) Oceanography and Marine Biology: an annual review. CRC Press.

AIROLDI, L., TURON, X., PERKOL‐FINKEL, S. & RIUS, M. 2015. Corridors for aliens but not for natives: effects of marine urban sprawl at a regional scale. Diversity and Distributions, 21, 755-768.

ANDERSON, M. & UNDERWOOD, A. 1997. Effects of gastropod grazers on recruitment and succession of an estuarine assemblage: a multivariate and univariate approach. Oecologia, 109, 442- 453.

ANDERSON, M. J. 2001. A new method for non‐parametric multivariate analysis of variance. Austral ecology, 26, 32-46.

ARAÚJO, R., BÁRBARA, I., SOUSA-PINTO, I. & QUINTINO, V. 2005. Spatial variability of intertidal rocky shore assemblages in the northwest coast of Portugal. Estuarine, Coastal and Shelf Science, 64, 658-670.

ASTLES, K. 1993. Patterns of abundance and distribution of species in intertidal rock pools. Journal of the Marine Biological Association of the United Kingdom, 73, 555-569.

BACHMAN, C. G. 1979. Laser radar systems and techniques, Artech House (1979).

BANKS, S. A. & SKILLETER, G. A. 2007. The importance of incorporating fine-scale habitat data into the design of an intertidal marine reserve system. Biological Conservation, 138, 13-29.

BATES, D., MÄCHLER, M., BOLKER, B. & WALKER, S. 2015. Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67, 1-48.

BELL, S. S., MCCOY, E. D. & MUSHINSKY, H. R. 2012. Habitat structure: the physical arrangement of objects in space, Springer Science & Business Media.

BELLGROVE, A., MONK, J. & QUINN, G. P. 2014. Effects of micro-and mesograzers on intertidal macroalgal recruitment. Marine biology, 161, 1207-1216.

BENEDETTI-CECCHI, L., PANNACCIULLI, F., BULLERI, F., MOSCHELLA, P., AIROLDI, L., RELINI, G. & CINELLI, F. 2001. Predicting the consequences of anthropogenic disturbance: large-scale effects of loss of canopy algae on rocky shores. Marine Ecology Progress Series, 214, 137-150.

BERGEN, S. D., BOLTON, S. M. & FRIDLEY, J. L. 2001. Design principles for ecological engineering. Ecological Engineering, 18, 201-210.

References 101 BERTNESS, M. D., LEONARD, G. H., LEVINE, J. M., SCHMIDT, P. R. & INGRAHAM, A. O. 1999. Testing the relative contribution of positive and negative interactions in rocky intertidal communities. Ecology, 80, 2711-2726.

BERTOCCI, I., SEABRA, M., DOMINGUEZ, R., JACINTO, D., RAMÍREZ, R., COCA, J. & TUYA, F. 2014. Effects of loss of algal canopies along temperature and irradiation gradients in continental Portugal and the Canary Islands. Marine Ecology Progress Series, 506, 47-60.

BIRCH, G. 2007. A short geological and environmental history of the Sydney estuary, Australia. In: G, B. (ed.) Water Wind Art and Debate - How environmental concerns impact on disciplinary research. Sydney University Press.

BIRCH, G. 2017. Assessment of human-induced change and biological risk posed by contaminants in estuarine/harbour sediments: Sydney Harbour/estuary (Australia). Marine Pollution Bulletin, 116, 234-248.

BIRCH, G. & TAYLOR, S. 1999. Source of heavy metals in sediments of the Port Jackson estuary, Australia. Science of the Total Environment, 227, 123-138.

BIRD, E. 1988. The tubeworm Galeolaria caespitosa as an indicator of sea level rise. Victorian Naturalist, 105, 98-104.

BISHOP, M., UNDERWOOD, A. & ARCHAMBAULT, P. 2002. Sewage and environmental impacts on rocky shores: necessity of identifying relevant spatial scales. Marine Ecology Progress Series, 236, 121-128.

BISHOP, M. J., MAYER-PINTO, M., AIROLDI, L., FIRTH, L. B., MORRIS, R. L., LOKE, L. H., HAWKINS, S. J., NAYLOR, L. A., COLEMAN, R. A., CHEE, S. Y. & DAFFORN, K. 2017. Effects of ocean sprawl on ecological connectivity: impacts and solutions. Journal of Experimental Marine Biology and Ecology, 492, 7-30.

BLANCHARD, D. & BOURGET, E. 1999. Scales of coastal heterogeneity: influence on intertidal community structure. Marine Ecology Progress Series, 179, 163-173.

BLANCHETTE, C. A. 1997. Size and survival of intertidal plants in response to wave action: a case study with Fucus gardneri. Ecology, 78, 1563-1578.

BLOCKLEY, D. & CHAPMAN, M. 2006. Recruitment determines differences between assemblages on shaded or unshaded seawalls. Marine Ecology Progress Series, 327, 27-36.

BLOCKLEY, D. J. 2007. Effect of wharves on intertidal assemblages on seawalls in Sydney Harbour, Australia. Marine Environmental Research, 63, 409-427.

BOLTON, D., MAYER-PINTO, M., CLARK, G., DAFFORN, K., BRASSIL, W., BECKER, A. & JOHNSTON, E. 2017. Coastal urban lighting has ecological consequences for multiple trophic levels under the sea. Science of The Total Environment, 576, 1-9.

BRAWLEY, S. H. & JOHNSON, L. E. 1993. Predicting desiccation stress in microscopic organisms the use of agarose beads to deter,ome evaporation within and between intertidal microhabitats. Journal of phycology, 29, 528-535.

References 102 BRIERLEY, A. S. & KINGSFORD, M. J. 2009. Impacts of climate change on marine organisms and ecosystems. Current Biology, 19, R602-R614.

BROS, W. E. 1987. Effects of removing or adding structure (barnacle shells) on recruitment to a fouling community in Tampa Bay, Florida. Journal of Experimental Marine Biology and Ecology, 105, 275-296.

BROWNE, M. A. & CHAPMAN, M. G. 2014. Mitigating against the loss of species by adding artificial intertidal pools to existing seawalls. Marine Ecology Progress Series, 497, 119-129.

BUGNOT, A., MAYER-PINTO, M., JOHNSTON, E., SCHAEFER, N. & DAFFORN, K. 2018. Learning from nature to enhance Blue engineering of marine infrastructure. Ecological Engineering, 120, 611-621.

BULLERI, F. & AIROLDI, L. 2005. Artificial marine structures facilitate the spread of a non‐indigenous green alga, Codium fragile ssp. tomentosoides, in the north Adriatic Sea. Journal of Applied Ecology, 42, 1063-1072.

BULLERI, F. & CHAPMAN, M. 2004. Intertidal assemblages on artificial and natural habitats in marinas on the north-west coast of Italy. Marine Biology, 145, 381-391.

BULLERI, F., CHAPMAN, M. & UNDERWOOD, A. 2005. Intertidal assemblages on seawalls and vertical rocky shores in Sydney Harbour, Australia. Austral Ecology, 30, 655-667.

BULLERI, F. & CHAPMAN, M. G. 2010. The introduction of coastal infrastructure as a driver of change in marine environments. Journal of Applied Ecology, 47, 26-35.

BURNHAM, K. P. & ANDERSON, D. R. 2004. Multimodel inference: understanding AIC and BIC in model selection. Sociological Methods & Research, 33, 261-304.

BURROWS, M. T., KAWAI, K. & HUGHES, R. N. 1999. Foraging by mobile predators on a rocky shore: underwater TV observations of movements of blennies Lipophrys pholis and crabs Carcinus maenas. Marine Ecology Progress Series, 187, 237-250.

CANTIN, M., BEDARD, J. & MILNE, H. 1974. The food and feeding of common eiders in the St. Lawrence estuary in summer. Canadian Journal of Zoology, 52, 319-334.

CARVALHO, L. & BARROS, F. 2017. Physical habitat structure in marine ecosystems: the meaning of complexity and heterogeneity. Hydrobiologia, 797, 1-9.

CHAPMAN, M. 2003. Paucity of mobile species on constructed seawalls: effects of urbanization on biodiversity. Marine Ecology Progress Series, 264, 21-29.

CHAPMAN, M. & BLOCKLEY, D. 2009. Engineering novel habitats on urban infrastructure to increase intertidal biodiversity. Oecologia, 161, 625-635.

CHAPMAN, M. & BULLERI, F. 2003. Intertidal seawalls—new features of landscape in intertidal environments. Landscape and urban planning, 62, 159-172.

References 103 CHURCH, J., CLARK, P., CAZENAVE, A., GREGORY, J., JEVREJEVA, S., LEVERMANN, A., MERRIFIELD, M., MILNE, G., NEREM, R., NUNN, P., PAYNE, A., PFEFFER, W., STAMMER, D. & UNNIKRISHNAN, A. 2013. Sea Level Change. In: STOCKER, T., QIN, D., PLATTNER, G.-K., TIGNOR, M., ALLEN, S., BOSCHUNG, J., NAUELS, A., XIA, Y., BEX, V. & MIDGLET, P. (eds.) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.

CHURCH, J. A., GREGORY, J. M., WHITE, N. J., PLATTEN, S. M. & MITROVICA, J. X. 2011. Understanding and projecting sea level change. Oceanography, 24, 130-143.

CLARK, G. F., MARZINELLI, E. M., FOGWILL, C. J., TURNEY, C. S. & JOHNSTON, E. L. 2015. Effects of sea-ice cover on marine benthic communities: a natural experiment in Commonwealth Bay, East Antarctica. Polar Biology, 38, 1213-1222.

CLARK, G. F., STARK, J. S., JOHNSTON, E. L., RUNCIE, J. W., GOLDSWORTHY, P. M., RAYMOND, B. & RIDDLE, M. J. 2013. Light‐driven tipping points in polar ecosystems. Global Change Biology, 19, 3749-3761.

COEN, L. D., BRUMBAUGH, R. D., BUSHEK, D., GRIZZLE, R., LUCKENBACH, M. W., POSEY, M. H., POWERS, S. P. & TOLLEY, S. G. 2007. Ecosystem services related to oyster restoration. Marine Ecology Progress Series, 341, 303-307.

CONNELL, J. H. 1961. The influence of interspecific competition and other factors on the distribution of the barnacle Chthamalus stellatus. Ecology, 42, 710-723.

CONNELL, J. H. 1972. Community interactions on marine rocky intertidal shores. Annual Review of Ecology and Systematics, 3, 169-192.

CONNELL, S., FOSTER, M. & AIROLDI, L. 2014. What are algal turfs? Towards a better description of turfs. Marine Ecology Progress Series, 495, 299-307.

CONNELL, S. D., RUSSELL, B. D., TURNER, D. J., SHEPHERD, S. A., KILDEA, T., MILLER, D., AIROLDI, L. & CHESHIRE, A. 2008. Recovering a lost baseline: missing kelp forests from a metropolitan coast. Marine Ecology Progress Series, 360, 63-72.

CORLETT, R. T. 2015. The Anthropocene concept in ecology and conservation. Trends in Ecology & Evolution, 30, 36-41.

CRONIN, G. & HAY, M. E. 1996. Effects of light and nutrient availability on the growth, secondary chemistry, and resistance to herbivory of two brown seaweeds. Oikos, 93-106. CROWE, T. 1996. Different effects of microhabitat fragmentation on patterns of dispersal of an intertidal gastropod in two habitats. Journal of Experimental Marine Biology and Ecology, 206, 83-107.

DAFFORN, K. A., GLASBY, T. M., AIROLDI, L., RIVERO, N. K., MAYER-PINTO, M. & JOHNSTON, E. L. 2015a. Marine urbanization: an ecological framework for designing multifunctional artificial structures. Frontiers in Ecology and the Environment, 13, 82-90.

References 104 DAFFORN, K. A., GLASBY, T. M. & JOHNSTON, E. L. 2012a. Comparing the invasibility of experimental “reefs” with field observations of natural reefs and artificial structures. PLoS One, 7, e38124.

DAFFORN, K. A., MAYER-PINTO, M., MORRIS, R. L. & WALTHAM, N. J. 2015b. Application of management tools to integrate ecological principles with the design of marine infrastructure. Journal of Environmental Management, 158, 61-73.

DAFFORN, K. A., SIMPSON, S. L., KELAHER, B. P., CLARK, G. F., KOMYAKOVA, V., WONG, C. K. & JOHNSTON, E. L. 2012b. The challenge of choosing environmental indicators of anthropogenic impacts in estuaries. Environmental Pollution, 163, 207-217.

DAHLHOFF, E. P., BUCKLEY, B. A. & MENGE, B. A. 2001. Physiology of the rocky intertidal predator Nucella ostrina along an environmental stress gradient. Ecology, 82, 2816-2829.

DAS, P., MARCHESIELLO, P. & MIDDLETON, J. H. 2000. Numerical modelling of tide-induced residual circulation in Sydney Harbour. Marine and Freshwater Research, 51, 97-112.

DAVIS, A. R., COLEMAN, D., BROAD, A., BYRNE, M., DWORJANYN, S. A. & PRZESLAWSKI, R. 2013. Complex responses of intertidal molluscan embryos to a warming and acidifying ocean in the presence of UV radiation. PloS One, 8, e55939.

DAVISON, I. R. & PEARSON, G. A. 1996. Stress tolerance in intertidal seaweeds. Journal of Phycology, 32, 197-211.

DEAN, R. L. & CONNELL, J. H. 1987. Marine invertebrates in an algal succession. III. Mechanisms linking habitat complexity with diversity. Journal of Experimental Marine Biology and Ecology, 109, 249-273.

DENNY, M. & WETHEY, D. 2001. Physical processes that generate patterns in marine communities. In: BERTNESS, M., GAINES, S. & HAY, M. (eds.) Marine community ecology. Sinauer.

DENNY, M. W. 1985. Wave forces on intertidal organisms: A case study. Limnology and Oceanography, 30, 1171-1187.

DENNY, M. W., DANIEL, T. L. & KOEHL, M. 1985. Mechanical limits to size in wave‐swept organisms. Ecological Monographs, 55, 69-102.

DEPARTMENT OF FINANCE, SERVICES & INNOVATION, [22.02.2018]. .

DETHIER, M. N. 1980. Tidepools as refuges: predation and the limits of the harpacticoid copepod Tigriopus californicus (Baker). Journal of Experimental Marine Biology and Ecology, 42, 99- 111.

DETHIER, M. N. 1981. Heteromorphic algal life histories: the seasonal pattern and response to herbivory of the brown crust, Ralfsia californica. Oecologia, 49, 333-339.

DEXTER, R. W. D. 1943. Observations on local movements of Littorina lottorea (L.) and Thais lapillus (L.). Nautilus, 57, 6-8.

References 105 DIDHAM, R. K., TYLIANAKIS, J. M., GEMMELL, N. J., RAND, T. A. & EWERS, R. M. 2007. Interactive effects of habitat modification and species invasion on native species decline. Trends in ecology & evolution, 22, 489-496.

DONEY, S. C., FABRY, V. J., FEELY, R. A. & KLEYPAS, J. A. 2009. Ocean acidification: the other CO2 problem.

DONEY, S. C., RUCKELSHAUS, M., DUFFY, J. E., BARRY, J. P., CHAN, F., ENGLISH, C. A., GALINDO, H. M., GREBMEIER, J. M., HOLLOWED, A. B., KNOWLTON, N., POLOVINA, J., RABALAIS, N. N., SYDEMAN, W. J. & TALLEY, L. D. 2012. Climate change impacts on marine ecosystems. Annual Review of Marine Science, 4, 11-37.

DRING, M. J., WAGNER, A., BOESKOV, J. & LÜNING, K. 1996. Sensitivity of intertidal and subtidal red algae to UVA and UVB radiation, as monitored by chlorophyll fluorescence measurements: influence of collection depth and season, and length of irradiation. European Journal of Phycology, 31, 293-302.

DRUEHL, L. D. & GREEN, J. M. 1982. Vertical distribution of intertidal seaweeds as related to patterns of submersion and emersion. Marine Ecology Progress Series, 163-170.

DUARTE, C. M., PITT, K. A., LUCAS, C. H., PURCELL, J. E., UYE, S.-I., ROBINSON, K., BROTZ, L., DECKER, M. B., SUTHERLAND, K. R., MALEJ, A., MADIN, L., MIANZAN, H., GILI, J.-M., FUENTES, V., PAGÉS, F., BREITBURG, D., MALEK, J., GRAHAM, W. M. & CONDON, R. H. 2013. Is global ocean sprawl a cause of jellyfish blooms? Frontiers in Ecology and the Environment, 11, 91- 97.

DUFFY-ANDERSON, J. & ABLE, K. 2001. An assessment of the feeding success of young-of-the-year winter flounder (Pseudopleuronectes americanus) near a municipal pier in the Hudson River estuary, USA. Estuaries, 24, 430-440.

EGGERT, A. 2012. Seaweed responses to temperature. In: WIENCKE, C. & BISCHOF, K. (eds.) Seaweed Biology. Springer.

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, 488-494.

EVANS, A. J., FIRTH, L. B., HAWKINS, S. J., MORRIS, E. S., GOUDGE, H. & MOORE, P. J. 2016. Drill-cored rock pools: an effective method of ecological enhancement on artificial structures. Marine and Freshwater Research, 67, 123-130.

FAIRWEATHER, P. 1988a. Movements of intertidal whelks (Morula marginalba and Thais orbita) in relation to availability of prey and shelter. Marine Biology, 100, 63-68.

FAIRWEATHER, P. 1988b. Predation creates haloes of bare space among prey on rocky seashores in New South Wales. Australian Journal of Ecology, 13, 401-409.

FAIRWEATHER, P. & UNDERWOOD, A. 1991. Experimental removals of a rocky intertidal predator: variations within two habitats in the effects on prey. Journal of Experimental Marine Biology and Ecology, 154, 29-75.

References 106 FENG, H., COCHRAN, J. K., LWIZA, H., BROWNAWELL, B. J. & HIRSCHBERG, D. J. 1998. Distribution of heavy metal and PCB contaminants in the sediments of an urban estuary: the Hudson River. Marine environmental research, 45, 69-88.

FIRTH, L. B., BROWNE, K. A., KNIGHTS, A. M., HAWKINS, S. J. & NASH, R. 2016. Eco-engineered rock pools: a concrete solution to biodiversity loss and urban sprawl in the marine environment. Environmental Research Letters, 11, 094015.

FIRTH, L. B., SCHOFIELD, M., WHITE, F. J., SKOV, M. W. & HAWKINS, S. J. 2014. Biodiversity in intertidal rock pools: informing engineering criteria for artificial habitat enhancement in the built environment. Marine Environmental Research, 102, 122-130.

FIRTH, L. B., THOMPSON, R. C., WHITE, F. J., SCHOFIELD, M., SKOV, M. W., HOGGART, S. P., JACKSON, J., KNIGHTS, A. M. & HAWKINS, S. J. 2013. The importance of water‐retaining features for biodiversity on artificial intertidal coastal defence structures. Diversity and Distributions, 19, 1275-1283.

FITZGERALD, D. M., FENSTER, M. S., ARGOW, B. A. & BUYNEVICH, I. V. 2008. Coastal impacts due to sea-level rise. Annual Review of Earth and Planetary Sciences, 36, 601-647.

FITZPATRICK, J. & KIRKMAN, H. 1995. Effects of prolonged shading stress on growth and survival of seagrass Posidonia australis in Jervis Bay, New South Wales, Australia. Marine Ecology Progress Series, 127, 279-289.

FRAENKEL, G. 1966. The heat resistance of intertidal snails at Shirahama, Wakayama-ken, Japan. Publications of the Seto Marine Biological Laboratory, 14, 185-195.

FRANCK, J. & LOOMIS, W. E. 1949. Photosynthesis in plants, Iowa State College Press.

FRANK, P. W. 1965. The biodemography of an intertidal snail population. Ecology, 46, 831-844.

FRANKLIN, L. & FORSTER, R. 1997. Review the changing irradiance environment: consequences for marine macrophyte physiology, productivity and ecology. European Journal of Phycology, 32, 207-232.

GACIA, E., SATTA, M. P. & MARTIN, D. 2007. Low crested coastal defence structures on the Catalan coast of the Mediterranean Sea: how they compare with natural rocky shores. Scientia Marina (Barcelona).

GARBARY, D. 2007. The margin of the sea. Algae and cyanobacteria in extreme environments. Springer.

GARRITY, S. D. 1984. Some adaptations of gastropods to physical stress on a tropical rocky shore. Ecology, 65, 559-574.

GAYLORD, B. & GAINES, S. D. 2000. Temperature or transport? Range limits in marine species mediated solely by flow. The American Naturalist, 155, 769-789.

References 107 GEDAN, K. B., KIRWAN, M. L., WOLANSKI, E., BARBIER, E. B. & SILLIMAN, B. R. 2011. The present and future role of coastal wetland vegetation in protecting shorelines: answering recent challenges to the paradigm. Climatic Change, 106, 7-29.

GOODSELL, P., CHAPMAN, M. & UNDERWOOD, A. 2007. Differences between biota in anthropogenically fragmented habitats and in naturally patchy habitats. Marine Ecology Progress Series, 351, 15-23.

GRABOWSKI, J. H. 2004. Habitat complexity disrupts predator–prey interactions but not the trophic cascade on oyster reefs. Ecology, 85, 995-1004.

GRABOWSKI, J. H., BRUMBAUGH, R. D., CONRAD, R. F., KEELER, A. G., OPALUCH, J. J., PETERSON, C. H., PIEHLER, M. F., POWERS, S. P. & SMYTH, A. R. 2012. Economic valuation of ecosystem services provided by oyster reefs. BioScience, 62, 900-909.

GRAY, D. R. & HODGSON, A. N. 2004. The importance of a crevice environment to the limpet Helcion pectunculus (Patellidae). Journal of Molluscan Studies, 70, 67-72.

HALPERN, B. S., SELKOE, K. A., MICHELI, F. & KAPPEL, C. V. 2007. Evaluating and ranking the vulnerability of global marine ecosystems to anthropogenic threats. Conservation Biology, 21, 1301-1315.

HALPERN, B. S., WALBRIDGE, S., SELKOE, K. A., KAPPEL, C. V., MICHELI, F., D'AGROSA, C., BRUNO, J. F., CASEY, K. S., EBERT, C., FOX, H. E., FUJITA, R., HEINEMANN, D., LENIHAN, H. S., MADIN, E. M., PERRY, M. T., SELIG, E. R., SPALDING, M., STENECK, R. & WATSON, R. 2008. A global map of human impact on marine ecosystems. Science, 319, 948-952.

HAMILTON, P. 1978. Intertidal distribution and long-term movements of Littorina irrorata (Mollusca: Gastropoda). Marine Biology, 46, 49-58.

HARGREAVES, B. R. 2003. Water column optics and penetration of UVR. UV effects in aquatic organisms and ecosystems, 1, 59-108.

HARLEY, C. D. 2008. Tidal dynamics, topographic orientation, and temperature-mediated mass mortalities on rocky shores. Marine Ecology Progress Series, 371, 37-46.

HATJE, V., BIRCH, G. & HILL, D. 2001. Spatial and temporal variability of particulate trace metals in Port Jackson Estuary, Australia. Estuarine, Coastal and Shelf Science, 53, 63-77.

HAWKINS, S. 1983. Interactions of Patella and macroalgae with settling Semibalanus balanoides (L.). Journal of Experimental Marine Biology and Ecology, 71, 55-72.

HELMUTH, B. S. & HOFMANN, G. E. 2001. Microhabitats, thermal heterogeneity, and patterns of physiological stress in the rocky intertidal zone. The Biological Bulletin, 201, 374-384.

HOEGH-GULDBERG, O. & BRUNO, J. F. 2010. The impact of climate change on the world’s marine ecosystems. Science, 328, 1523-1528.

HOOPER, D. U., CHAPIN, F., EWEL, J., HECTOR, A., INCHAUSTI, P., LAVOREL, S., LAWTON, J., LODGE, D., LOREAU, M., NAEEM, S., SCHMID, B., SETÄLÄ, H., SYMSTAD, A., VANDERMEER, J. &

References 108 WARDLE, D. 2005. Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecological Monographs, 75, 3-35.

HUGGETT, J. & GRIFFITHS, C. 1986. Some relationships between elevation, physicochemical variables and biota of intertidal rock pools. Marine Ecology Progress Series, 29, 189-197.

IPCC 2013. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge, United Kingdom and New York, NY, USA, Cambridge University Press.

IRVING, A. D., CONNELL, S. D., JOHNSTON, E. L., PILE, A. J. & GILLANDERS, B. M. 2005. The response of encrusting coralline algae to canopy loss: an independent test of predictions on an Antarctic coast. Marine Biology, 147, 1075-1083.

IVEŠA, L., CHAPMAN, M., UNDERWOOD, A. J. & MURPHY, R. J. 2010. Differential patterns of distribution of limpets on intertidal seawalls: experimental investigation of the roles of recruitment, survival and competition. Marine Ecology Progress Series, 407, 55-69.

JACKSON, A., CHAPMAN, M. & UNDERWOOD, A. 2008. Ecological interactions in the provision of habitat by urban development: whelks and engineering by oysters on artificial seawalls. Austral Ecology, 33, 307-316.

JACKSON, A. & MCILVENNY, J. 2011. Coastal squeeze on rocky shores in northern Scotland and some possible ecological impacts. Journal of Experimental Marine Biology and Ecology, 400, 314- 321.

JERNAKOFF, P. 1983. Factors affecting the recruitment of algae in a midshore region dominated by barnacles. Journal of Experimental Marine Biology and Ecology, 67, 17-31.

JERNAKOFF, P. 1985. The effect of overgrowth by algae on the survival of the intertidal barnacle Tesseropora rosea Krauss. Journal of Experimental Marine Biology and Ecology, 94, 89-97.

JOHNSON, C. N., BALMFORD, A., BROOK, B. W., BUETTEL, J. C., GALETTI, M., GUANGCHUN, L. & WILMSHURST, J. M. 2017. Biodiversity losses and conservation responses in the Anthropocene. Science, 356, 270-275.

JOHNSTON, E. L., MAYER-PINTO, M., HUTCHINGS, P. A., MARZINELLI, E. M., AHYONG, S. T., BIRCH, G., BOOTH, D. J., CREESE, R. G., DOBLIN, M. A., FIGUEIRA, W., GRIBBEN, P. E., PRITCHARD, T., ROUGHAN, M., STEINBERG, P. D. & HEDGE, L. H. 2015. Sydney Harbour: what we do and do not know about a highly diverse estuary. Marine and Freshwater Research, 66, 1073-1087.

JOHNSTON, E. L. & ROBERTS, D. A. 2009. Contaminants reduce the richness and evenness of marine communities: a review and meta-analysis. Environmental Pollution, 157, 1745-1752.

JONGMAN, R. 2002. Homogenisation and fragmentation of the European landscape: ecological consequences and solutions. Landscape and urban planning, 58, 211-221.

KEITH, D. A. 2015. Assessing and managing risks to ecosystem biodiversity. Austral Ecology, 40, 337- 346.

References 109 KEITH, D. A., RODRÍGUEZ, J. P., RODRÍGUEZ-CLARK, K. M., NICHOLSON, E., AAPALA, K., ALONSO, A., ASMUSSEN, M., BACHMAN, S., BASSET, A., BARROW, E. G., BENSON, J. S., BISHOP, M. J., BONIFACIO, R., BROOKS, T. M., BURGMAN, M. A., COMER, P., COMÍN, F. A., FABER- LANGENDOEN, D., FAIRWEATHER, P. G., HOLDAWAY, R. J., JENNINGS, M., KINGSFORD, R. T., LESTER, R. E., NALLY, R. M., MCCARTHY, M. A., MOAT, J., OLIVEIRA-MIRANDA, M. A., PISANU, P., POULIN, B., REGAN, T. J., RIECKEN, U., SPALDING, M. D. & ZAMBRANO-MARTÍNEZ, S. 2013. Scientific foundations for an IUCN Red List of Ecosystems. PLOS one, 8, e62111.

KEOUGH, M. J. & DOWNES, B. J. 1982. Recruitment of marine invertebrates: the role of active larval choices and early mortality. Oecologia, 54, 348-352.

KEOUGH, M. J. & QUINN, G. 1998. Effects of periodic disturbances from trampling onrocky intertidal algal beds. Ecological Applications, 8, 141-161.

KLEIN, J. C., UNDERWOOD, A. & CHAPMAN, M. 2011. Urban structures provide new insights into interactions among grazers and habitat. Ecological Applications, 21, 427-438.

KON-YA, K. & MIKI, W. 1994. Effects of environmental factors on larval settlement of the barnacle Balanus amphitrite reared in the laboratory. Fisheries Science, 60, 563-565.

LAI, S., LOKE, L. H., BOUMA, T. J. & TODD, P. A. 2018. Biodiversity surveys and stable isotope analyses reveal key differences in intertidal assemblages between tropical seawalls and rocky shores. Marine Ecology Progress Series, 587, 41-53.

LASSAU, S. A., HOCHULI, D. F., CASSIS, G. & REID, C. A. 2005. Effects of habitat complexity on forest beetle diversity: do functional groups respond consistently? Diversity and distributions, 11, 73-82.

LEVINE, J. M., ADLER, P. B. & YELENIK, S. G. 2004. A meta‐analysis of biotic resistance to exotic plant invasions. Ecology Letters, 7, 975-989.

LEVINGS, S. C. & GARRITY, S. D. 1983. Diel and tidal movement of two co-occurring neritid snails; differences in grazing patterns on a tropical rocky shore. Journal of Experimental Marine Biology and Ecology, 67, 261-278.

LEVITUS, S., ANTONOV, J. I., BOYER, T. P., BARANOVA, O. K., GARCIA, H. E., LOCARNINI, R. A., MISHONOV, A. V., REAGAN, J., SEIDOV, D., YAROSH, E. S. & ZWENG, M. 2012. World ocean heat content and thermosteric sea level change (0–2000 m), 1955–2010. Geophysical Research Letters, 39.

LEWIS, F. 1987. Crustacean epifauna of seagrass and macroalgae in Apalachee Bay, Florida, USA. Marine Biology, 94, 219-229.

LILLEY, S. A. & SCHIEL, D. R. 2006. Community effects following the deletion of a habitat-forming alga from rocky marine shores. Oecologia, 148, 672-681.

LOKE, L. H. & TODD, P. A. 2016. Structural complexity and component type increase intertidal biodiversity independently of area. Ecology, 97, 383-393.

References 110 LOREAU, M., NAEEM, S., INCHAUSTI, P., BENGTSSON, J., GRIME, J., HECTOR, A., HOOPER, D., HUSTON, M., RAFFAELLI, D., SCHMID, B., TILMAN, D. & WARDLE, D. 2001. Biodiversity and ecosystem functioning: current knowledge and future challenges. Science, 294, 804-808.

LOTZE, H. K., LENIHAN, H. S., BOURQUE, B. J., BRADBURY, R. H., COOKE, R. G., KAY, M. C., KIDWELL, S. M., KIRBY, M. X., PETERSON, C. H. & JACKSON, J. B. 2006. Depletion, degradation, and recovery potential of estuaries and coastal seas. Science, 312, 1806-1809.

LUBCHENCO, J. 1978. Plant species diversity in a marine intertidal community: importance of herbivore food preference and algal competitive abilities. The American Naturalist, 112, 23- 39.

MACARTHUR, R. H. & MACARTHUR, J. W. 1961. On bird species diversity. Ecology, 42, 594-598.

MANLY HYDRAULICS LABORATORY 2016. NSW Ocean and river entrance tidal levels annual summary 2015–2016, MHL Report 2475, October 2016.

MARTINS, G., HAWKINS, S., THOMPSON, R. & JENKINS, S. 2007. Community structure and functioning in intertidal rock pools: effects of pool size and shore height at different successional stages. Marine Ecology Progress Series, 329, 43-55.

MARTONE, P. T., ALYONO, M. & STITES, S. 2010. Bleaching of an intertidal coralline alga: untangling the effects of light, temperature, and desiccation. Marine Ecology Progress Series, 416, 57- 67.

MATIAS, M. G., UNDERWOOD, A. & COLEMAN, R. A. 2007. Interactions of components of habitats alter composition and variability of assemblages. Journal of Animal Ecology, 76, 986-994.

MATIAS, M. G., UNDERWOOD, A., HOCHULI, D. & COLEMAN, R. 2010. Independent effects of patch size and structural complexity on diversity of benthic macroinvertebrates. Ecology, 91, 1908- 1915.

MAYER-PINTO, M., COLE, V., JOHNSTON, E., BUGNOT, A., HURST, H., AIROLDI, L., GLASBY, T. & DAFFORN, K. 2018. Functional and structural responses to marine urbanisation. Environmental Research Letters, 13, 014009.

MAYER-PINTO, M., JOHNSTON, E., BUGNOT, A., GLASBY, T., AIROLDI, L., MITCHELL, A. & DAFFORN, K. 2017. Building ‘blue’: An eco-engineering framework for foreshore developments. Journal of Environmental Management, 189, 109-114.

MAYER-PINTO, M., JOHNSTON, E., HUTCHINGS, P., MARZINELLI, E., AHYONG, S., BIRCH, G., BOOTH, D., CREESE, R., DOBLIN, M., FIGUEIRA, W., GRIBBEN, P., PRITCHARD, T., ROUGHAN, M., STEINBERG, P. & HEDGE, L. 2015. Sydney Harbour: a review of anthropogenic impacts on the biodiversity and ecosystem function of one of the world’s largest natural harbours. Marine and Freshwater Research, 66, 1088-1105.

MCCANN, K. S. 2000. The diversity–stability debate. Nature, 405, 228-233.

MCCOY, E. D. & BELL, S. S. 1991. Habitat structure: the evolution and diversification of a complex topic, Springer.

References 111 MCGUINNESS, K. & UNDERWOOD, A. 1986. Habitat structure and the nature of communities on intertidal boulders. Journal of Experimental Marine Biology and Ecology, 104, 97-123.

MCGUINNESS, K. A. 1984. Species–area curves. Biological Reviews, 59, 423-440.

MCKINNEY, M. L. 2006. Urbanization as a major cause of biotic homogenization. Biological conservation, 127, 247-260.

MCMAHON, R. F. 1990. Thermal tolerance, evaporative water loss, air-water oxygen consumption and zonation of intertidal prosobranchs: a new synthesis. Progress in Littorinid and Muricid Biology. Springer.

MCQUAID, C. D. & FRONEMAN, P. W. 1993. Mutualism between the territorial intertidal limpet Patella longicosta and the crustose alga Ralfsia verrucosa. Oecologia, 96, 128-133.

MEEHL, G. A. & TEBALDI, C. 2004. More intense, more frequent, and longer lasting heat waves in the 21st century. Science, 305, 994-997.

MENGE, B. A. & BRANCH, G. 2001. Rocky intertidal communities. In: BERTNESS, M. D., GAINES, S. D. & HAY, M. E. (eds.) Marine community ecology. Sunderland: Sinauer Associates.

MENGE, B. A. & LUBCHENCO, J. 1981. Community organization in temperate and tropical rocky intertidal habitats: prey refuges in relation to consumer pressure gradients. Ecological Monographs, 51, 429-450.

MENGE, B. A. & SUTHERLAND, J. P. 1976. Species diversity gradients: synthesis of the roles of predation, competition, and temporal heterogeneity. The American Naturalist, 110, 351-369.

METAXAS, A. & SCHEIBLING, R. E. 1993. Community structure and organization of tidepools. Marine Ecology Progress Series, 98, 187-198.

MIGNÉ, A., GOLLÉTY, C. & DAVOULT, D. 2015. Effect of canopy removal on a rocky shore community metabolism and structure. Marine Biology, 162, 449-457.

MILLER, R. J. & ETTER, R. J. 2008. Shading facilitates sessile invertebrate dominance in the rocky subtidal Gulf of Maine. Ecology, 89, 452-462.

MORAN, M. 1985. The timing and significance of sheltering and foraging behaviour of the predatory intertidal gastropod Morula marginalba Blainville (Muricidae). Journal of Experimental Marine Biology and Ecology, 93, 103-114.

MORRIS, S. & TAYLOR, A. C. 1983. Diurnal and seasonal variation in physico-chemical conditions within intertidal rock pools. Estuarine, Coastal and Shelf Science, 17, 339-355.

MOSCHELLA, P., ABBIATI, M., ÅBERG, P., AIROLDI, L., ANDERSON, J., BACCHIOCCHI, F., BULLERI, F., DINESEN, G. E., FROST, M., ESPERANÇA, G., GRANHAG, L., JONSSON, P., SATTA, M., SUNDELÖF, A., THOMPSON, R. & HAWKINS, S. 2005. Low-crested coastal defence structures as artificial habitats for marine life: using ecological criteria in design. Coastal Engineering, 52, 1053-1071.

References 112 MYERS, J. H., GUNTHORPE, L., ALLINSON, G. & DUDA, S. 2006. Effects of antifouling biocides to the germination and growth of the marine macroalga, Hormosira banksii (Turner) Desicaine. Marine Pollution Bulletin, 52, 1048-1055.

NAYLOR, L., VENN, O., COOMBES, M., JACKSON, J. & THOMPSON, R. 2011. Including ecological enhancements in the planning, design and construction of hard coastal structures: A process guide. University of Exeter: Report to Environment Agency (PID 110461).

NAYLOR, L. A., COOMBES, M. A., VENN, O., ROAST, S. D. & THOMPSON, R. C. 2012. Facilitating ecological enhancement of coastal infrastructure: the role of policy, people and planning. Environmental Science & Policy, 22, 36-46.

NEUMANN, B., VAFEIDIS, A. T., ZIMMERMANN, J. & NICHOLLS, R. J. 2015. Future coastal population growth and exposure to sea-level rise and coastal flooding-a global assessment. PloS One, 10, e0118571.

NICHOLLS, R. J. & CAZENAVE, A. 2010. Sea-level rise and its impact on coastal zones. Science, 328, 1517-1520.

NICHOLLS, R. J., HOOZEMANS, F. M. & MARCHAND, M. 1999. Increasing flood risk and wetland losses due to global sea-level rise: regional and global analyses. Global Environmental Change, 9, S69-S87.

NORTON, T. 1992. Dispersal by macroalgae. British Phycological Journal, 27, 293-301.

OLIVER, T. H., HEARD, M. S., ISAAC, N. J., ROY, D. B., PROCTER, D., EIGENBROD, F., FRECKLETON, R., HECTOR, A., ORME, C. D. L., PETCHEY, O. L., PROENÇA, V., RAFFAELLI, D., SUTTLE, K. B., MACE, G. M., MARTÍN-LÓPEZ, B., WOODCOCK, B. A. & BULLOCK, J. M. 2015. Biodiversity and resilience of ecosystem functions. Trends in Ecology & Evolution, 30, 673-684.

PACIFIC NORTHWEST NATIONAL LABORATORY 2002. Evaluation of methods to increase light under ferry terminals. PNNL-13714. Pacific Northwest Nation Laboratory.

PAINE, R. 1971. A short‐term experimental investigation of resource partitioning in a New Zealand rocky intertidal habitat. Ecology, 52, 1096-1106.

PAINE, R. T., FENCHEL, T. & KINNE, O. 1994. Marine rocky shores and community ecology: an experimentalist's perspective, Ecology Institute Oldendorf/Luhe, Germany.

PARDAL‐SOUZA, A. L., DIAS, G. M., JENKINS, S. R., CIOTTI, Á. M. & CHRISTOFOLETTI, R. A. 2017. Shading impacts by coastal infrastructure on biological communities from subtropical rocky shores. Journal of Applied Ecology, 54, 826-835.

PERKINS, M. J., NG, T. P., DUDGEON, D., BONEBRAKE, T. C. & LEUNG, K. M. 2015. Conserving intertidal habitats: what is the potential of ecological engineering to mitigate impacts of coastal structures? Estuarine, Coastal and Shelf Science, 167, 504-515.

PESSOA, M. F. 2012. Harmful effects of UV radiation in algae and aquatic macrophytes-a review. Emirates Journal of Food and Agriculture, 510-526.

References 113 PIANKA, E. R. 1966. Convexity, desert lizards, and spatial heterogeneity. Ecology, 47, 1055-1059.

PISTER, B. 2009. Urban marine ecology in southern California: the ability of riprap structures to serve as rocky intertidal habitat. Marine Biology, 156, 861-873.

PONTEE, N. 2013. Defining coastal squeeze: a discussion. Ocean & Coastal Management, 84, 204- 207.

POWLES, S. B. 1984. Photoinhibition of photosynthesis induced by visible light. Annual Review of Plant Physiology, 35, 15-44.

QUINN, G. 1988a. Ecology of the intertidal pulmonate limpet Siphonaria diemenensis Quoy et Gaimard. I. Population dynamics and availability of food. Journal of Experimental Marine Biology and Ecology, 117, 115-136.

QUINN, G. 1988b. Effects of conspecific adults, macroalgae and height on the shore on recruitment of an intertidal limpet. Marine Ecology Progress Series, 48, 305-308.

RAFFAELLI, D. & HAWKINS, S. J. 2012. Intertidal ecology, Springer Science & Business Media.

RANGELEY, R. W. & KRAMER, D. L. 1995. Use of rocky intertidal habitats by juvenile pollock Pollachius virens. Marine Ecology Progress Series, 126, 9-17.

RAWLINGS, T. A. 1996. Shields against ultraviolet radiation: an additional protective role for the egg capsules of benthic marine gastropods. Marine Ecology Progress Series, 136, 81-95.

ROMERO, G. Q., GONÇALVES‐SOUZA, T., VIEIRA, C. & KORICHEVA, J. 2015. Ecosystem engineering effects on species diversity across ecosystems: a meta‐analysis. Biological Reviews, 90, 877- 890.

ROSENZWEIG, M. L. 1995. Species diversity in space and time, Cambridge, Cambridge University Press.

RUIZ, J. M. & ROMERO, J. 2001. Effects of in situ experimental shading on the Mediterranean seagrass Posidonia oceanica. Marine Ecology Progress Series, 215, 107-120.

SAUNDERS, D. A., HOBBS, R. J. & MARGULES, C. R. 1991. Biological consequences of ecosystem fragmentation: a review. Conservation Biology, 5, 18-32.

SAUNDERS, R. & CONNELL, S. 2001. Interactive effects of shade and surface orientation on the recruitment of spirorbid polychaetes. Austral Ecology, 26, 109-115.

SCHEFFERS, B. R., EDWARDS, D. P., DIESMOS, A., WILLIAMS, S. E. & EVANS, T. A. 2014. Microhabitats reduce animal's exposure to climate extremes. Global Change Biology, 20, 495-503.

SCHIEL, D. R. & LILLEY, S. A. 2007. Gradients of disturbance to an algal canopy and the modification of an intertidal community. Marine Ecology Progress Series, 339, 1-11.

SCHIEL, D. R. & TAYLOR, D. I. 1999. Effects of trampling on a rocky intertidal algal assemblage in southern New Zealand. Journal of Experimental Marine Biology and Ecology, 235, 213-235.

References 114 SEBENS, K. 1991. Habitat structure and community dynamics in marine benthic systems. In: BELL, S. S., MCCOY, E. D. & MUSHINSKY, H. R. (eds.) Habitat structure. Dordrecht: Springer.

SHANKS, A. L., GRANTHAM, B. A. & CARR, M. H. 2003. Propagule dispersal distance and the size and spacing of marine reserves. Ecological Applications, 3, S159-S169.

SHEPARD, C. C., CRAIN, C. M. & BECK, M. W. 2011. The protective role of coastal marshes: a systematic review and meta-analysis. PloS One, 6, e27374.

SMITH, R. S., JOHNSTON, E. L. & CLARK, G. F. 2014. The role of habitat complexity in community development is mediated by resource availability. PloS One, 9, e102920.

SPALDING, M. D., RUFFO, S., LACAMBRA, C., MELIANE, I., HALE, L. Z., SHEPARD, C. C. & BECK, M. W. 2014. The role of ecosystems in coastal protection: adapting to climate change and coastal hazards. Ocean & Coastal Management, 90, 50-57.

STARK, J. S. 1998. Heavy metal pollution and macrobenthic assemblages in soft sediments in two Sydney estuaries, Australia. Marine and Freshwater Research, 49, 533-540.

STEPHENSON, T. A. & STEPHENSON, A. 1972. Life between tidemarks on rocky shores, W. H. Freeman.

STRAIN, E. M., OLABARRIA, C., MAYER‐PINTO, M., CUMBO, V., MORRIS, R. L., BUGNOT, A. B., DAFFORN, K. A., HEERY, E., FIRTH, L. B., BROOKS, P. R. & BISHOP, M. J. 2018. Eco‐engineering urban infrastructure for marine and coastal biodiversity: Which interventions have the greatest ecological benefit? Journal of Applied Ecology, 55, 426-441.

TAKADA, Y. 1999. Influence of shade and number of boulder layers on mobile organisms on a warm temperate boulder shore. Marine Ecology Progress Series, 189, 171-179.

TAN, E. L.-Y., MAYER-PINTO, M., JOHNSTON, E. L. & DAFFORN, K. A. 2015. Differences in intertidal microbial assemblages on urban structures and natural rocky reef. Frontiers in Microbiology, 6, 1276.

TEWS, J., BROSE, U., GRIMM, V., TIELBÖRGER, K., WICHMANN, M., SCHWAGER, M. & JELTSCH, F. 2004. Animal species diversity driven by habitat heterogeneity/diversity: the importance of keystone structures. Journal of Biogeography, 31, 79-92.

THIEL, M. & GUTOW, L. 2005. The ecology of rafting in the marine environment. II. The rafting organisms and community. In: GIBSON, R., ATKINSON, R. & GORDON, J. (eds.) Oceanography and Marine Biology. CRC Press.

THOMPSON, R., CROWE, T. & HAWKINS, S. 2002. Rocky intertidal communities: past environmental changes, present status and predictions for the next 25 years. Environmental Conservation, 29, 168-191.

THORNER, J., KUMAR, L. & SMITH, S. D. 2014. Impacts of climate-change-driven sea level rise on intertidal rocky reef habitats will be variable and site specific. PloS One, 9, e86130.

THRUSH, S. F., GRAY, J. S., HEWITT, J. E. & UGLAND, K. I. 2006. Predicting the effects of habitat homogenization on marine biodiversity. Ecological Applications, 16, 1636-1642.

References 115 UNDERWOOD, A. 1978. An experimental evaluation of competition between three species of intertidal prosobranch gastropods. Oecologia, 33, 185-202.

UNDERWOOD, A. 1980. The effects of grazing by gastropods and physical factors on the upper limits of distribution of intertidal macroalgae. Oecologia, 46, 201-213.

UNDERWOOD, A. 1981. Structure of a rocky intertidal community in New South Wales: patterns of vertical distribution and seasonal changes. Journal of Experimental Marine Biology and Ecology, 51, 57-85.

UNDERWOOD, A. & CHAPMAN, M. 1996. Scales of spatial patterns of distribution of intertidal invertebrates. Oecologia, 107, 212-224.

UNDERWOOD, A. & CHAPMAN, M. 1998. Spatial analyses of intertidal assemblages on sheltered rocky shores. Australian Journal of Ecology, 23, 138-157.

UNDERWOOD, A. & MCFADYEN, K. 1983. Ecology of the intertidal snail Littorina acutispira Smith. Journal of Experimental Marine Biology and Ecology, 66, 169-197.

UNDERWOOD, A. & SKILLETER, G. 1996. Effects of patch-size on the structure of assemblages in rock pools. Journal of Experimental Marine Biology and Ecology, 197, 63-90.

VASELLI, S., BULLERI, F. & BENEDETTI-CECCHI, L. 2008. Hard coastal-defence structures as habitats for native and exotic rocky-bottom species. Marine Environmental Research, 66, 395-403.

VIEJO, R. M. 1999. Mobile epifauna inhabiting the invasive Sargassum muticum and two local seaweeds in northern Spain. Aquatic Botany, 64, 131-149.

VITOUSEK, P. M., MOONEY, H. A., LUBCHENCO, J. & MELILLO, J. M. 1997. Human domination of Earth's ecosystems. Science, 277, 494-499.

WALTHAM, N. J. & SHEAVES, M. 2018. Eco-engineering rock pools to a seawall in a tropical estuary: microhabitat features and fine sediment accumulation. Ecological Engineering.

WANLESS, H. R. 1982. Sea level is rising--so what? Journal of Sedimentary Research, 52.

WERNBERG, T., SMALE, D. A., TUYA, F., THOMSEN, M. S., LANGLOIS, T. J., DE BETTIGNIES, T., BENNETT, S. & ROUSSEAUX, C. S. 2013. An extreme climatic event alters marine ecosystem structure in a global biodiversity hotspot. Nature Climate Change, 3, 78.

WHITE, G. E., HOSE, G. C. & BROWN, C. 2015. Influence of rock‐pool characteristics on the distribution and abundance of inter‐tidal fishes. Marine Ecology, 36, 1332-1344.

WILLIAMS, G. A. & MORRITT, D. 1995. Habitat partitioning and thermal tolerance in a tropical limpet, Cellana grata. Marine Ecology Progress Series, 124, 89-103.

ZUUR, A., IENO, E., WALKER, N., SAVELIEV, A. & SMITH, G. 2009. Mixed effects models and extensions in ecology Springer Science + Business Media.

Appendix S1 116 Appendix S1 – Supplementary material for Chapter 2

Table S1.1 Salinity and temperature measurements from Sydney Harbour at the four sites on each sampling event (B: Balmain, BI: Berry Island Reserve, BH: Bradleys Head, D: Delwood, S1-S8: Sampling event 1-8). Air temperatures were rounded ranges in daily maxima during the sampling periods acquired from weather station 066196 (Wedding Cake West) from the Bureau of Meteorology, Australia.

Air temperature Salinity (ppt) Water temperature (oC) (oC) Inner Zone Outer Zone Inner Zone Outer Zone Sydney Wedding B BIR BH D B BIR BH D Cake West S1 33 34 34 34 18 22 18 18 17-19 S2 34 34 34 35 14 17 14 16 17-19 S3 34 34 34 35 17 17 17 18 17-20 S4 35 35 35 35 23 22 22 20 19-26 S5 34 34 35 35 23 23 23 22 22-27 S6 30 30 33 34 26 26 23 23 25-30 S7 34 34 32 35 24 24 21 22 21-27 S8 36 35 35 35 21 22 21 20 22-25

Appendix S1 117

Figure S1.1 Size ranges of pools within each zone.

[Type here]

Table S1.2 Number of quadrats assessed per rock pool at each site. NA= no quadrat assessed. Appendix S1

Balmain Berry Island Bradleys Head Delwood Pool No. of quadrats Pool No. of quadrats Pool No. of quadrats Pool No. of quadrats B_8 2 BIR_3 2 BH_3 4 D_2 6 B_CBP1 3 BIR_CBP3 4 BH_4 2 D_CBP4 6 B_CBP8 2 BIR_CBP6 2 BH_5 4 D_CBP7 2 B_FC 4 BIR_CBR6 3 BH_R12 5 D_CBR8 6 B_FCP1 3 BIR_DFP8 3 BH_R13 4 D_FCP7 3 B_FCP3 4 BIR_FC 2 BH_WR8 2 D_R3 3 B_FCP4 2 BIR_R2 4 BH_11* NA D_R6 4 B_FCP8 4 BIR_R11 2 BH_47* NA D_R15 4 B_R4 4 BIR_WP 5 BH_DFP12* NA D_WP 3 B_R13 3 BIR_WR3 4 D_WR5 3 B_FCP5* NA BIR_WR9 4 D_WP3 4 BIR_FC1 3 D_WP4 3 BIR_CBR5* NA D_CBP11* NA

118

Appendix S1 119

Table S1.3 Mean and standard error (SE) of sessile and mobile taxa richness at the four sites on each sampling event (B: Balmain, BI: Berry Island, BH: Bradleys Head, D: Delwood, S1-S8: Sampling event 1-8).

Sessile taxa B BIR BH D Mean SE Mean SE Mean SE Mean SE S1 8.1 0.5 6.7 0.4 8.8 1.4 10.5 1.0 S2 8.6 0.6 6.5 0.4 9.1 1.1 10.5 0.6 S3 7.3 0.5 7.2 0.3 9.3 1.3 10.5 1.0 S4 8.4 0.7 6.5 0.3 9.1 1.0 11.2 1.0 S5 7.3 0.3 5.5 0.4 8.2 1.0 8.3 0.8 S6 6.5 0.5 5.8 0.3 8.3 1.2 8.7 0.9 S7 7.1 0.4 5.9 0.4 6.9 1.3 9.4 0.7 S8 6.9 0.3 6.1 0.4 7.3 1.3 8.8 0.8

Mobile taxa B BIR BH D Mean SE Mean SE Mean SE Mean SE S1 3.5 0.3 4.7 0.2 5.9 1.2 7.4 0.4 S2 3.7 0.3 4.6 0.3 5.0 0.7 6.8 0.5 S3 3.6 0.4 5.1 0.3 5.2 0.7 7.5 0.5 S4 4.0 0.2 4.7 0.2 6.3 1.0 8.4 0.3 S5 3.3 0.2 4.6 0.3 5.1 0.8 6.9 0.4 S6 2.7 0.3 3.8 0.3 5.4 0.6 7.3 0.4 S7 3.3 0.3 5.0 0.3 4.4 0.6 6.2 0.5 S8 3.5 0.2 4.5 0.2 4.2 0.8 6.4 0.5

Table S1.5 continued

Table S1.4 Presence (1)/absence (0) of sessile taxa collected in rock pools in the inner or outer Harbour in each season. Taxa unique to the inner zone are Appendix S1 indicated with *. Taxa unique to the outer zone are indicated with **.

Inner Zone Outer Zone

Taxon Winter Spring Summer Autumn Winter Spring Summer Autumn Corallina spp. 1 1 1 1 1 1 1 1

Non-geniculate Corallina** 0 0 0 0 1 1 1 1

Red fuzz 1 1 1 1 1 1 1 1

Hormosira banksii** 0 0 0 0 1 1 1 1

Sargassum spp.** 0 0 0 0 1 1 1 1

Petalonia spp. 1 1 0 0 1 1 0 0

Colpomenia spp.** 0 0 0 0 1 1 0 0

Dictyota spp.** 0 0 0 0 1 1 0 0

Padina spp.** 0 0 0 0 0 0 1 0

Codium spp.** 0 0 0 0 0 1 1 1

Ulva spp. 1 1 1 1 1 1 1 1

Turfing 1 1 1 1 1 1 1 1

Ralfsia spp.** 0 0 0 0 1 1 1 1

Encrusting coralline** 0 0 0 0 1 1 1 1

Amorphous crust 1 1 1 1 1 1 1 1

Oulactis muscosa 1 1 0 1 1 1 1 1

Actinia tenebrosa** 0 0 0 0 1 1 1 1

White anemone** 0 0 0 0 1 1 0 0

120 Yellow anemone* 1 1 0 0 0 0 0 0

Table S1.5 continued

Balanus trigonus 0 1 0 0 1 1 0 0 Appendix S1

Tetraclitella purpurascens** 0 0 0 0 1 1 1 1

Amphibalanus variegatus* 1 1 1 0 0 0 0 0

Austrobalanus imperator** 0 0 0 0 1 1 1 1

Megabalanus coccopoma** 0 0 0 0 0 1 0 0

other barnacles 1 1 1 1 1 1 1 1

Oysters 1 1 1 1 1 1 1 1

Mytilus spp. 1 1 1 1 1 1 0 1

Galeolaria caespitosa 1 0 0 1 1 1 1 1

Spirorbid polychaetes** 0 0 0 0 1 1 1 1

Hydroides* 1 0 0 1 0 0 0 0

Polychaete species a** 0 0 0 0 1 1 1 1

Watersipora spp.** 0 0 0 0 1 1 1 1

Cryptosula pallasiana** 0 0 0 0 0 1 1 1

Ascidian a** 0 0 0 0 1 1 1 0

Ascidian b 0 0 0 1 1 0 0 0

Sponges 0 1 0 1 1 1 1 1

121

Table S1.5 Presence (1)/absence (0) of mobile taxa collected in rock pools in the inner or outer Harbour in each season. Taxa unique to the inner zone Appendix S1 are indicated with *. Taxa unique to the outer zone are indicated with **.

Inner Zone Outer Zone

Taxon Winter Spring Summer Autumn Winter Spring Summer Autumn Bembicium nanum 1 1 1 1 1 1 1 1

Bembicium auratum* 1 1 1 1 0 0 0 0

Austrocochlea porcata 1 1 1 1 1 1 1 1

Nerita atramentosa** 0 0 0 0 1 1 1 1

Austrolittorina unifasciata** 0 0 0 0 1 1 1 0

Afrolittorina acutispira** 0 0 0 0 0 1 1 1

Gastropod species a** 0 0 0 0 0 1 0 0

Onchidium damelii 0 0 0 1 1 0 0 0

Nudibranchia species a** 0 0 0 0 0 1 0 0

Elysia species a* 1 0 0 1 0 0 0 0

Elysia species b** 0 0 0 0 0 0 0 1

Cellana tramoserica 1 0 1 0 1 1 1 1

Montfortula rugosa 1 0 0 1 1 1 1 1

Patelloida spp. 1 1 1 1 1 1 1 1

Notoacmea spp. 1 0 0 0 0 1 0 0

Siphonaria spp. 1 1 1 1 1 1 1 0

Tenguella marginalba 1 0 0 1 1 1 1 1

Gastropod species b** 0 0 0 0 0 1 0 0

122 Sypharochiton pelliserpentes 0 0 0 1 1 1 1 1

Table S1.6 continued

Acantochitona spp.** 0 0 0 0 1 1 1 1 Appendix S1

Ischnochiton spp.** 0 0 0 0 1 1 0 0

Parvulastra exigua 0 0 0 1 1 1 1 1

Brittlestar** 0 0 0 0 0 1 0 1

Platyhelminthes** 0 0 0 0 1 0 1 0

Gammaridae** 0 0 0 0 0 1 1 1

Crab 1 1 1 1 1 1 1 1

Shrimps 0 1 0 0 0 1 1 0

Fish 1 1 1 1 1 1 0 0

Jellyfish 0 1 0 0 0 0 0 0

Ctenophora* 0 0 0 0 0 0 1 0

123

Appendix S1 124

Table S1.6 PERMANOVA results for differences in the structure of sessile and mobile assemblages. Analyses used Bray-Curtis similarities from square root transformed data, with significance determined from 9999 permutations of the data under a reduced model (* = P < 0.05, ** = P < 0.01, *** = P < 0.001). Abbreviations: maxW=maximum Width, maxD=maximum Depth, Vo=Volume, Zo=Zone, Si=Site, Sa=Sampling Event, Res=Residuals.

Sessile assemblages Mobile assemblages

Pseudo- Pseudo- Source df F P(perm) df F P(perm) maxW 1 29.598 0.001*** 1 17.369 0.0001***

maxD 1 4.5497 0.0019** 1 2.0653 0.0975 Vo 1 2.041 0.1071 1 1.3362 0.2528

Zo 1 6.82 0.0067** 1 3.1962 0.0628

Sa 7 1.9065 0.0201 7 2.0279 0.0081

Si(Zo) 2 30.37 0.0001*** 2 52.413 0.0001*** ZoxSa 7 1.7474 0.0362* 7 1.3625 0.1657* Si(Zo)xSa 14 0.79391 0.85 14 0.53199 0.9966 Res 269 285

Appendix S1 125 Table S1.7 Results of SIMPER analysis for dissimilarities between zones for sessile taxa and debris, sand and bare rock.

Group Inner Group Outer Taxon Av.Abund Av.Abund Av.Diss Diss/SD Contrib% Cum.% Ralfsia spp. 0.94 5.37 4.7 3.05 19.06 19.06 Corallina spp. 4.05 1.12 3.18 1.66 12.9 31.96 Bare rock 3.62 4.35 2.15 1.44 8.73 40.69 Live oyster. 2.75 1.13 1.89 1.4 7.67 48.36 Turfing 2.43 1.63 1.85 0.98 7.49 55.85 Dead oyster. 2.63 1.01 1.77 1.46 7.18 63.03 Watersipora spp. 0.94 1.83 1.05 0.68 4.25 67.28 Galeolaria caespitosa 0.94 1.76 0.95 1.01 3.85 71.13 Debris 1.64 1.02 0.88 0.81 3.57 74.7 unknown 0.94 1.63 0.8 0.93 3.26 77.96 Amorphus crust 1.33 1.03 0.6 0.7 2.45 80.41 Sand 1.13 0.9 0.34 0.49 1.37 81.78 Austrobalanus imperator 0.94 1.13 0.33 0.77 1.33 83.11 Non-geniculate Corallina 0.94 1.06 0.31 0.58 1.25 84.36 Red.fuzz 1.07 0.93 0.29 0.6 1.17 85.54 Ulva 1.01 0.99 0.26 0.77 1.04 86.57 Sponges 0.94 1.03 0.25 0.64 1.03 87.6 Other barnacles 0.98 0.97 0.24 0.66 0.98 88.58 Actinia tenebrosa 0.94 1.04 0.24 0.82 0.97 89.54 Encrusting coralline 0.94 0.94 0.18 0.66 0.73 90.28 Colpomenia spp. 0.94 0.93 0.17 0.68 0.7 90.98 Oulactis muscosa 0.94 0.93 0.17 0.91 0.69 91.67 Mytilus spp.. 0.98 0.87 0.16 0.89 0.67 92.33 Petalonia spp. 0.95 0.87 0.13 1.12 0.54 92.88 White anemone 0.94 0.88 0.13 1.11 0.53 93.41 Ascidian a 0.94 0.87 0.13 1.15 0.52 93.93 Dictyota spp. 0.94 0.87 0.13 1.22 0.51 94.45 Polychaete species a 0.94 0.87 0.13 1.24 0.51 94.96 Padina spp. 0.94 0.87 0.12 1.25 0.5 95.46 Yellow anemone 0.94 0.87 0.12 1.25 0.5 95.96 Balanus trigonus 0.94 0.87 0.12 1.25 0.5 96.47 Tetraclitella purpurascens 0.94 0.87 0.12 1.25 0.5 96.97 Amphibalanus variegatus 0.94 0.87 0.12 1.25 0.5 97.48 Megabalanus coccopoma 0.94 0.87 0.12 1.25 0.5 97.98 Spirorbid polychaetes 0.94 0.87 0.12 1.25 0.5 98.49 Hydroides 0.94 0.87 0.12 1.25 0.5 98.99 Cryptosula pallasiana 0.94 0.87 0.12 1.25 0.5 99.5 Ascidian b 0.94 0.87 0.12 1.25 0.5 100

Appendix S1 126 Table S1.8 Results of SIMPER analysis for dissimilarities between zones for mobile taxa.

Group Group Inner Outer Av.Dis Diss/S Contrib Cum. Av.Abund Av.Abund Taxon s D % % Bembicium nanum 1.65 3.26 1.55 1.13 13.35 13.35 Patelloida spp. 3.1 1.51 1.47 1.47 12.68 26.03 Bembicium auratum 2.99 1.49 1.39 1.27 11.99 38.02 Austrocochlea porcata 1.8 2.57 0.94 1.15 8.1 46.12 Parvulastra exigua 1.59 2.05 0.55 0.99 4.72 50.84 Tenguella marginalba 1.56 2.01 0.51 1 4.38 55.22 Montfortula rugosa 1.56 2 0.49 0.9 4.2 59.41 Siphonaria spp. 1.88 1.52 0.41 0.88 3.52 62.93 Nerita atramentosa 1.56 1.82 0.35 0.64 3.03 65.96 Cellana tramoserica 1.57 1.68 0.27 1.02 2.28 68.24 Sipharochiton 1.56 1.64 0.23 1.07 1.97 70.21 pelliserpentes Acantochitona spp. 1.56 1.54 0.17 1.22 1.51 71.72 Austrolittorina unifasciata 1.56 1.52 0.17 0.81 1.5 73.22 Crab 1.58 1.49 0.17 1.08 1.43 74.66 Brittlestar 1.56 1.49 0.16 1.23 1.36 76.02 Afrolittorina acutispira 1.56 1.49 0.15 1.25 1.33 77.35 Nodlittorina pyramidalis 1.56 1.49 0.15 1.25 1.33 78.68 Cerinthium corallium 1.56 1.49 0.15 1.25 1.33 80.02 Gastropod species a 1.56 1.49 0.15 1.25 1.33 81.35 Gastropod species b 1.56 1.49 0.15 1.25 1.33 82.68 Gastropod species c 1.56 1.49 0.15 1.25 1.33 84.01 Onchidium damelii 1.56 1.49 0.15 1.25 1.33 85.34 Nudibranchia species a 1.56 1.49 0.15 1.25 1.33 86.68 Elysia species a 1.56 1.49 0.15 1.25 1.33 88.01 Elysia species b 1.56 1.49 0.15 1.25 1.33 89.34 Notoacmea spp. 1.56 1.49 0.15 1.25 1.33 90.67 Ischnochiton spp. 1.56 1.49 0.15 1.25 1.33 92.01 Platyhelminthes 1.56 1.49 0.15 1.25 1.33 93.34 Gammaridae 1.56 1.49 0.15 1.25 1.33 94.67 Shrimps 1.56 1.49 0.15 1.25 1.33 96 Fish 1.56 1.49 0.15 1.25 1.33 97.34 Jellyfish 1.56 1.49 0.15 1.25 1.33 98.67 Ctenophora 1.56 1.49 0.15 1.25 1.33 100

Table S1.10 continued

Table S1.9 Model selection table using GLMMs to predict the total number of sessile and mobile taxa, total cover of sessile taxa and the total abundance Appendix S1 of mobile taxa in both zones. X’s denote variables used in each model. Lower AIC values indicate higher support for the model. The best model is highlighted in bold. Abbreviations: W=Width, D=Depth, V=Volume, H=Height on shore, Z=Zone.

Model W D V H Z Z*W Z*D Z*V Z*H AIC ΔAIC

Sessile taxa richness M1 X X X X X 1517.8 0.0 M2 X X X X X X X 1518.3 0.5 M3 X X X X X X 1519.4 1.6 M4 X X X X 1519.5 1.8 M5 X X X X X X X X 1519.9 2.1 M6 X X X X X X X X X 1521.8 4.1

Mobile taxa richness M1 X X X X X 1365.4 0.0 M2 X X X X X X 1366.5 1.0 M3 X X X X X X X 1367.2 1.8 M4 X X X X X X X X 1367.9 2.4 M5 X X X X X X X X X 1369.9 4.4

Total cover of sessile taxa M1 X X X X X 3849.6 0.0 M2 X X X X 3850.3 0.7 M3 X X X X X X 3850.8 1.2 M4 X X X X X X X 3852.8 3.2

M5 X X X X X X X X 3854.8 5.2 127

M6 X X X X X X X X X 3856.8 7.2

Table S1.10 continued

Appendix S1

Total abundance of mobile taxa M1 X X X X X 1928.7 0.0

M2 X X X X X X 1929.7 1.0 M3 X X X X X X X 1931.6 2.9 M4 X X X X X X X X 1933.1 4.4 M5 X X X X X X X X X 1935.1 6.4

128

Appendix S1 129

Table S1.10 Model selection table using GLMMs to predict the total number of sessile and mobile taxa, the total cover of sessile taxa, and the total abundance of mobile taxa in the outer zone only with canopy cover as an additional predictor. X’s denote variables used in each model. Lower AIC values indicate higher support for the model. The best model is highlighted in bold. Abbreviations: W=Width, D=Depth, V=Volume, H=Height on shore, C=Canopy cover).

Model W D V H C AIC ΔAIC

Sessile taxa richness M1 X X 793.1 0.0 M2 X X X X 793.3 0.3 M3 X X X 793.7 0.7 M4 X X X X X 795.1 2.1

Mobile taxa richness M1 X X X 723.8 0.0 M2 X X X X 725.4 1.6 M3 X X 725.5 1.7 M4 X X X X X 727.3 3.6

Total cover of sessile taxa M1 260.1 0 M2 X 260.9 0.8 M3 X X 262.5 2.4 M4 X X X 264.4 4.3 M5 X X X X 266.3 6.2 M6 X X X X X 268.3 8.2

Total abundance of mobile taxa M1 X X X X 887.6 0.0 M2 X X X 888.8 1.2 M3 X X X X X 889.5 1.9

Appendix S1 130

Table S1.11 Model selection table using GLMMs to predict canopy cover in the outer zone only. X’s denote variables used in each model. Lower AIC values indicate higher support for the model. The best model is highlighted in bold. Abbreviations: W=Width, D=Depth, V=Volume, Si= Site, Se=Season, H=Height on shore, C=Canopy cover).

Model W D V H AIC ΔAIC

Canopy cover M1 X 1628.4 0.0 M2 X X 1628.9 0.5 M3 X X X 1630.8 2.4 M4 X X X X 1632.8 4.4

Appendix S2 131

Appendix S2 – Supplementary material for Chapter 3

Table S2.1 Number of rock pools excluded for which the sampling method was unsuitable Excluded due to Coastal Outer Harbour Inner Harbour >800mm width 3 2 1 Gastropods 1 0 0 Oyster cover 0 0 3 Algae cover 3 9 0

Table S2.2 List of rock pools with pits and overhangs used in all analyses and lists of randomly selected rock pools without any microhabitat in each Appendix S2 subset.

Pits sub1 sub2 sub3 sub4 sub5

With pits Without pits Without pits Without pits Without pits Without pits curlcurl.6 curlcurl.4 curlcurl.4 curlcurl.8 curlcurl.4 curlcurl.1 curlcurl.7 curlcurl.8 curlcurl.8 curlcurl.13 curlcurl.8 curlcurl.4 freshwater.5 curlcurl.13 curlcurl.15 curlcurl.15 curlcurl.11 curlcurl.11 freshwater.7 curlcurl.15 freshwater.1 freshwater.1 curlcurl.15 curlcurl.13 freshwater.9 freshwater.6 freshwater.6 freshwater.16 freshwater.1 curlcurl.15 freshwater.10 freshwater.17 freshwater.17 bondi.1 freshwater.6 freshwater.1 freshwater.11 bondi.3 bondi.1 bondi.3 freshwater.17 freshwater.6 freshwater.18 bondi.5 bondi.3 bondi.5 bondi.3 bondi.1 Bondi.16 bondi.14 bondi.14 bondi.14 bondi.5 bondi.3

Overhangs sub1 sub2 sub3 sub4 sub5 With overhangs Without overhangs Without overhangs Without overhangs Without overhangs Without overhangs curlcurl.2 curl curl.1 curlcurl.4 curlcurl.1 curlcurl.1 curlcurl.4 curlcurl.5 curlcurl.4 curlcurl.8 curlcurl.4 curlcurl.8 curlcurl.8 curlcurl.10 curlcurl.8 curlcurl.11 curlcurl.8 curlcurl.11 curlcurl.13 curlcurl.14 curlcurl.11 curlcurl.15 curlcurl.13 curlcurl.13 curlcurl.15 freshwater.2 curlcurl.13 freshwater.17 curlcurl.15 curlcurl.15 freshwater.1 freshwater.8 curlcurl.15 freshwater.6 freshwater.1 freshwater.1 freshwater.6 freshwater.12 freshwater.1 bondi.1 freshwater.6 freshwater.17 freshwater.17 bondi.2 freshwater.17 bondi.3 freshwater.17 bondi.1 bondi.1 bondi.8 bondi.3 bondi.5 bondi.14 bondi.5 bondi.3 bondi.9 bondi.5 bondi.14 bondi.5 bondi.14 bondi.5

132

Appendix S2 133

Table S2.3 Presence (1)/absence (0) of mobile species at the different locations. Species/morphospecies unique to the coastal location are indicated with *. Species/morphospecies unique to the Outer Harbour are indicated with **. Species/morphospecies unique to the Inner Harbour are indicated with ***.

Location Coastal Outer Harbour Inner Harbour Season winter summer winter summer winter summer Bembicium nanum 1 1 1 1 0 1 Bembicium auratum*** 0 0 0 0 1 1 Austrocochlea 1 1 1 1 1 1 Nerita atramentosa 1 1 1 1 0 0 Tenguella marginalba 1 1 1 1 0 1 Austrolittorina unifasciata 1 1 1 1 0 0 Nodlittorina pyramidalis 1 1 0 1 0 0 Dicathais orbita 1 1 0 0 0 0 Cerinthium corallium 0 0 0 0 0 0 Gastropod 1* 0 1 0 0 0 0 Gastropod 2* 1 0 0 0 0 0 Gastropod 3* 1 0 0 0 0 0 Gastropod 4* 1 0 0 0 0 0 Gastropod 5** 0 0 0 1 0 0 Gastropod 6*** 0 0 0 0 1 0 Gastropod 7*** 0 0 0 0 1 0 Nudibranch 1 0 0 0 0 0 Elysia spp.*** 0 0 0 0 1 1 Onchidium damelii 0 0 1 1 0 1 Parvulastra exigua 1 1 1 1 0 0 Meridiastra calcar* 1 0 0 0 0 0 Flatworm** 0 0 1 0 0 0 Limpets 1 1 1 1 1 1 Chitons 1 1 1 1 1 1

Appendix S2 134

Table S2.4 Number of rock pools with no, one, or two microhabitats at each location. Categories included none (N), small pit (SP), medium pit (MP), deep pit (DP) and overhangs (OH)) as well as combinations of them.

Location N SP MP OH MP + DP OH + SP OH + SP + MP Coastal 20 5 3 10 1 2 1 Outer Harbour 14 3 1 0 0 3 0 Inner Harbour 24 0 0 4 0 0 0

Table S2.5 Density of B. auratum (ind/L) in oyster shells or habitats other than oyster shells in rock pools when oyster shells were present.

Rock pool Sampling Season Oyster shell Other habitat berryisland.3 1 winter 93.75 7.5 Oyster > Other berryisland.FC 1 winter 209.3 20.93 Oyster > Other berryisland.FC1 1 winter 46.3 6.94 Oyster > Other berryisland.R11 1 winter 200 24.62 Oyster > Other berryisland.R2 1 winter 51.12 4.5 Oyster > Other berryisland.3 2 winter 0 14.06 berryisland.FC 2 winter 23.26 23.84 berryisland.FC1 2 winter 0 16.67 berryisland.R2 2 winter 0 8.18 berryisland.3 3 summer 41.67 29.69 Oyster > Other berryisland.FC 3 summer 29.07 32.56 berryisland.FC1 3 summer 92.59 17.01 Oyster > Other berryisland.R2 3 summer 30.67 12.68 Oyster > Other berryisland.DFP8 3 summer 218.18 70.91 Oyster > Other berryisland.3 4 summer 0 25.31 berryisland.FC 4 summer 0 46.51 berryisland.R2 4 summer 0 12.68 berryisland.DFP8 4 summer 218.18 63.64 Oyster > Other

Appendix S3 – Supplementary material for Chapter 4 Appendix S3

Figure S3.1 UVA (a, b) and UVB (c, d) measurements under full light and under experimental shading plates at site 1 (a, c) and site 2 (b, d). Abbreviations 135 as shown in Table 4.1

Table S3.1 Number of replicates per sampling for each treatment at each site. Abbreviations as shown in Table 4.1. Appendix S3 Site 1 Mobile taxa Sessile taxa

Pool FL PC 75% 35% 15% FS FL PC 75% 35% 15% FS

Sampling 1 3 5 5 5 5 5 3 5 5 5 5 5

Sampling 2 5 5 5 5 5 5 5 5 5 5 5 5

Sampling 3 5 5 4 4 4 3 5 5 5 4 5 4

Sampling 4 5 5 4 4 3 5 5 5 5 4 3 5

Sampling 5 5 5 5 5 3 5 5 5 5 5 4 5

Sampling 6 5 5 5 4 4 4 5 5 5 4 4 4

Site 2 Mobile taxa Sessile taxa

Pool FL PC 75% 35% 15% FS FL PC 75% 35% 15% FS

Sampling 1 5 5 5 5 5 5 5 5 5 5 5 5

Sampling 2 5 5 5 5 5 5 5 5 5 5 5 5

Sampling 3 5 5 5 5 5 5 5 5 5 5 5 5

Sampling 4 5 5 5 5 5 5 5 5 5 5 5 5

Sampling 5 5 5 5 5 5 4 5 5 5 5 5 5

Sampling 6 5 5 5 5 5 4 5 5 5 5 5 4

136

Table S3.1 continued

Site 1 Mobile taxa Sessile taxa Appendix S3

Emergent rock FL PC 75% 35% 15% FS FL PC 75% 35% 15% FS

Sampling 1 5 5 5 5 5 5 5 5 5 5 5 5

Sampling 2 5 5 5 5 5 3 5 5 5 5 5 4

Sampling 3 5 3 4 5 3 4 5 5 5 5 4 5

Sampling 4 5 2 5 5 4 3 5 4 5 5 4 5

Sampling 5 5 4 5 5 4 4 5 5 5 5 5 5

Sampling 6 5 2 5 5 1 2 5 3 5 5 4 3

Site 2 Mobile taxa Sessile taxa

Emergent rock FL PC 75% 35% 15% FS FL PC 75% 35% 15% FS

Sampling 1 5 5 5 5 5 5 5 5 5 5 5 5

Sampling 2 5 5 5 5 5 5 5 5 5 5 5 5

Sampling 3 5 5 5 4 5 4 5 5 5 5 5 4

Sampling 4 5 5 5 5 5 5 5 5 5 5 5 5

Sampling 5 5 5 5 5 5 5 5 5 5 5 5 5

Sampling 6 4 5 5 5 5 4 5 5 5 5 5 4

137

Appendix S3 138

Figure S3.2 Shaded area (20 × 20 cm) versus sampled area (~ 15 × 15 cm) of experimental plots on emergent rock.

Figure S3.3 Light measurements over time above pools (a) and on emergent rock (b) at site 1.

Table S3.2 Mean number and standard error (SE) of mobile taxa per sampling per habitat at each site.

Site 1 Site 2 Pool Emergent rock Pool Emergent rock Mean SE Mean SE Mean SE Mean SE Sampling 1 4.25 0.28 1.03 0.24 5.30 0.20 1.37 0.23 Sampling 2 4.37 0.25 0.86 0.18 5.43 0.22 1.77 0.25 Sampling 3 5.08 0.24 2.38 0.35 5.50 0.23 2.36 0.28 Sampling 4 5.12 0.22 2.13 0.26 5.33 0.21 2.30 0.25 Sampling 5 4.89 0.24 1.96 0.23 5.10 0.21 2.60 0.26 Sampling 6 5.15 0.20 2.48 0.23 5.52 0.24 2.61 0.23

Table S3.3 Presence of sessile taxa under at least one replicate per treatment for each treatment at the start of the experiment and at each sampling. Appendix S3 R=Ralfsia spp., P=Petalonia spp., U=Ulva spp., T=Turfing algae, Co=Corallina spp., Cp=Colpomenia spp., Al= algae species 1, B=Barnacle, S=Sponge, An=Anemone, Ge=Green encrusting. Abbreviations for treatments as shown in Table 4.1.

Site 1 (pool) Start Sampling 1 Sampling 2 Sampling 3 Sampling 4 Sampling 5 Sampling 6

FL R R R, T R, T R, T R, U, T, S R, U, T PC R R, T R, T R, U, T, Ge R, U, P, T, Ge, S R, U, P, Ge, S R, U, P, Ge, S 75% R R, U, T, Al R, U, T, Ge R, U, T, Ge R, U, Ge, B R, U, Ge R, U, Ge, S 35% R R, T R, U, T, S R, U, T, S R, U, T, S R, U R, U 15% R R, T R, U, T R, U, T R, U R, U R, U FS R R R, U R, U R R R

Site 2 (pool) Start Sampling 1 Sampling 2 Sampling 3 Sampling 4 Sampling 5 Sampling 6 FL R R, T R, T R, Co, T R, U, Co, T, Ge, S R, Co, P, T, Ge R, Co, P, T, Ge PC R R, T R, U, Cp, T R, U, Cp, T R, U, Cp, T, Ge R, U, Co, P, Cp, T R, U, Co, P, Cp, T, Ge 75% R R, U, T R, U, T R, U, T R, U, Cp, T, Ge R, U, P, Cp, T R, U, P, T 35% R, T R, U, T R, U, Cp R, U, T R, U, Cp, T R, U, P, T R, U, P, T 15% R R, Cp, T R, Cp, T R, T, Ge R, U, Cp, T, Ge R, U, P, Cp, T, Ge, S R, U, P, Cp, T, Ge, S FS R, T R, T R, Cp, T R, U, Cp, An, B R, U, B R, U, Ge, B R, T, An, B

139

Table S3.3 continued

Start Sampling 1 Sampling 2 Sampling 3 Sampling 4 Sampling 5 Sampling 6 Appendix S3 Site 1 (emergent rock)

FL R

PC R R R

75% R R, U, T R, U R, U, T R, U, P, T

35% R R, P, T R, P, T R, T

15% R R R R R, T

FS R R R R R

Site 2 (emergent rock) Start Sampling 1 Sampling 2 Sampling 3 Sampling 4 Sampling 5 Sampling 6

FL R T R R R R

PC R, T R R, T R, U R, T

75% R R R R R R R 35% R R R, Ge R R R, U R 15% R R R R R R R FS R R R R R R R

140

Appendix S3

Figure S3.4 Mean cover of sessile taxa besides algae per sampling time at site 1 (a) and site 2 (b) averaged across replicate plots within rock pools. Abbreviations for treatments as shown in Table 4.1.

141

Appendix S3 142

Table S3.4 Number of replicates for comparison between sampling methods within rock pools. Abbreviations for treatments as shown in Table 4.1. Site 1 Site 2 FL 5 4 PC 4 4 75% 5 5 35% 3 2 15% 4 5 FS 2 3

Table S3.4 Presence of taxa under at least one replicate per treatments for each treatment for each sampling method (assessing the entire rock pool (base + wall) versus base only (base)). R=Ralfsia spp., P=Petalonia spp., U=Ulva spp., T=Turfing algae, Cp=Colpomenia spp., B=Barnacle, An=Anemone, Ge=Green encrusting, Am= Algae matrix, Sp= Spirorbid polychaete. Abbreviations for treatments as shown in Table 4.1. Site 1 (pool) Base Base + wall FL R, U, T R, U, T, Am PC R, P R, P 75% R, U, P, Ge R, U, P, Ge 35% R, U R, U 15% R, U R, U FS R,Sp R

Site 2 (pool) Base Base + wall FL R, C, P, T, Ge R, U, Co, P, T, Ge, Sp, Am PC R, U, Co, P, Cp, T R, U, Co, P, Cp, T, Ge, Sp 75% R, U, P, T R, U, P, T, Ge, Sp, Am 35% R, U, P, T, An R, U, P, T, An, Am 15% R, U, P, Cp, T, Ge, An R, U, P, Cp, T, Ge, An, Sp, Am FS R, Ge, An, B R, Ge, An, B, Sp

Appendix S4 143

Appendix S4 – Supplementary material for Chapter 5

Table S4.1 IUCN Red List of Ecosystems Criterion Critically Endang Vulner C2 (Keith et al., 2013).C2 endangered ered able Environmental degradation over the next 50 years, or ≥80% severity ≥50% ≥50% any 50-year period including the present and future, with ≥80% severity severity based on change in an abiotic variable affecting… relative with with severity ≥80% ≥50% relative relative severity severity ≥80% ≥80% severity severity with with ≥50% ≥30% relative relative severity severity ≥30% severity with ≥80% relative severity

Figure S4.1 Rate of change of intertidal rocky shore area (in %) over time under each climate change scenario.

Table S4.2 Predicted remaining area of all rocky shores and selected sites for each climate change scenario. Appendix S4 All rocky shores Intertidal area (m2) Intertidal area (m2) Intertidal area (m2) Intertidal area (%) Intertidal area (%) Intertidal area (%) Projection (median) (lower likely range) (upper likely range) (median) (lower likely range) (upper likely range)

RCP2.6 352760 366776 334742 94.15 97.89 89.34 RCP4.5 341643 358130 317970 91.18 95.58 84.86 RCP6.0 329461 350983 302041 87.93 93.67 80.61 RCP8.5 293183 328701 253863 78.25 87.73 67.75

Bradleys Head Intertidal area (m2) Intertidal area (m2) Intertidal area (m2) Intertidal area (%) Intertidal area (%) Intertidal area (%) Projection (median) (lower likely range) (upper likely range) (median) (lower likely range) (upper likely range) RCP2.6 6108 6009 6115 99.54 98.27 100.00 RCP4.5 6136 6085 5720 100.00 99.51 93.54 RCP6.0 6084 6115 4940 99.15 100.00 80.78 RCP8.5 4281 6071 2804 69.77 99.28 45.85

Delwood Intertidal area (m2) Intertidal area (m2) Intertidal area (m2) Intertidal area (%) Intertidal area (%) Intertidal area (%) Projection (median) (lower likely range) (upper likely range) (median) (lower likely range) (upper likely range) RCP2.6 2030 2248 1915 86.31 95.58 81.42 RCP4.5 1950 2106 1780 82.91 89.54 75.68 RCP6.0 1872 2017 1616 79.59 85.76 68.71 RCP8.5 1482 1867 839 63.01 79.38 35.67

Intertidal area (m2) Intertidal area (m2) Intertidal area (m2) Intertidal area (%) Intertidal area (%) Intertidal area (%) Freshwater (median) (lower likely range) (upper likely range) (median) (lower likely range) (upper likely range) Projection 5104 5296 4764 96.39 100.00 89.97 RCP2.6 4898 5210 4528 92.50 98.38 85.51 RCP4.5 4705 5053 4388 88.86 95.41 82.87 144 RCP6.0 4307 4684 3811 81.34 88.44 71.97

Table S4.1 continued

RCP8.5 Appendix S4 Intertidal area (m2) Intertidal area (m2) Intertidal area (m2) Intertidal area (%) Intertidal area (%) Intertidal area (%) Cape Banks (median) (lower likely range) (upper likely range) (median) (lower likely range) (upper likely range) Projection 16678 17309 17162 91.12 94.56 93.76

RCP2.6 16957 16811 17634 92.64 91.84 96.34 RCP4.5 17340 16701 17738 94.73 91.24 96.91 RCP6.0 17820 17355 18085 97.36 94.82 98.80 RCP8.5

145