Inshore Benthic Communities of the Port Elizabeth Abalone Ranching Concession Area

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by South East Academic Libraries System (SEALS)

Inshore benthic communities of the Port Elizabeth abalone ranching concession area.

L A Moriarty

2019

Inshore benthic communities of the Port Elizabeth abalone ranching concession area.

Lauren Alethea Moriarty

Submitted in fulfilment of the requirements for the

degree of MSc Botany to be awarded at the Nelson Mandela University

April 2019

Supervisor: Dr. Paul-Pierre Steyn

Table of Contents Abstract ...... i

List of Figures ...... iii

List of Tables ...... v

List of Appendices ...... vi

Acknowledgements ...... vii

Chapter 1. Introduction ...... 1

Chapter 2. Literature Review ...... 4

2.1 The shallow coastal reef system of the Eastern Cape in South Africa...... 4

2.2 Environmental Factors ...... 4

2.2.1 Temperature ...... 4

2.2.2 Wind, Waves, and Tides ...... 5

2.3 Baseline habitat description ...... 6

2.3.1 Substrate ...... 6

2.3.2 Seaweed Communities ...... 6

2.4 The South African Commercial Abalone Industry ...... 9

2.5 Abalone Habitat Requirements ...... 10

2.5.1 Distribution ...... 10

2.5.2 Habitat ...... 10

2.5.3 Feeding ...... 11

2.6 Sampling marine benthic communities ...... 12

2.6.1 Transects and Quadrats ...... 12

2.6.2 Visual and Photographic Assessment ...... 12

2.6.3 Depth and Rugosity ...... 13

Chapter 3. Materials & methods ...... 14

3.1 Site background ...... 14

3.2 Dive Survey and Sampling ...... 17

3.3 Analysis ...... 19

3.3.1 Image Analysis ...... 19

3.3.2 Statistical Analysis ...... 20

Chapter 4. Results ...... 22

4.1 Changes in the benthic community along a depth gradient and along shore. .... 22

4.1.1 Physical Attributes ...... 22

4.1.2 Algal Composition and Diversity ...... 26

4.1.3 Macroinvertebrate Composition and Diversity ...... 29

4.1.4 Community Composition and Diversity ...... 32

4.2 Seeded vs unseeded sites ...... 37

4.2.1 Physical Attributes ...... 37

4.2.2 Algal Composition and Diversity ...... 39

4.2.3 Macroinvertebrate Composition and Diversity ...... 41

4.2.4 Community Composition and Diversity ...... 42

Chapter 5. Discussion ...... 45

5.1 Shallow benthic communities in the abalone ranching area ...... 45

5.2 Effect of abalone seeding on the benthic community ...... 49

Chapter 6. Concluding remarks ...... 52

References ...... 53

Appendices ...... 64

Abstract Abalone poaching in the Eastern Cape leading to resource depletion has resulted in the suspension of any commercial or recreational exploitation of abalone in the area. Abalone ranching trial projects have been initiated to help improve the natural stocks and potentially provide commercial yields. While seeding started in deeper water (5 – 10 m), improved abalone growth performance has been found in the shallower areas. This has resulted in the seeding effort being moved to these shallower sites (< 5m). An ecological survey of the baseline conditions of these shallow benthic areas is a requirement of the abalone ranching permit conditions (Permit No.: 1503759), aimed at detecting any impact that seeding may have on the benthic community. In addition to providing benchmark data for monitoring, the shallow benthic community (< 6m) in the Port Elizabeth ranching concession area has not been well described in terms of the requirements for abalone ranching. Information on the benthic communities in this area is limited to research on the substrate types and communities in deeper water (> 5m). This study aimed to address this information gap.

Dive surveys were conducted along 10 m long transects (~3 reps) at three depth zones (<1m; 1-2m; > 2m) for four sites along the span of the ranching concession area. Similar assessments were done at a seeded site at the Noordhoek Ski Boat Club and an unseeded site at the Willows area, in order to reveal whether seeding had any impact on the benthic community. Images from a GoPro mounted in the centre of a framer unit (0.5 m2) were taken every 0.5 m along the transect. Macroalgal and macrofaunal cover was determined from these images, and the benthic community characterised from these data. Seaweed samples were taken for species identification.

In the four baseline sites, sampling was done at three depth zones to note any changes with a depth gradient, both in terms of substrate type, as well as community composition. There was a notable trend with substrate type having a significant influence (P<0.05) on the community structure. Seaweed communities were dominated by Plocamium corallorhiza and coralline turf based on substrate types. There was also a significant relationship (P<0.05) between substrate type, dominant seaweed species and abalone presence. A Canonical Correspondence Analysis

(CCA) of the community data suggested that the benthic community does not change significantly along the distance of coastline sampled. A hierarchical cluster analysis of Bray-Curtis dissimilarity for transect data also suggested that the four sites do not represent ecologically dissimilar communities. A similar analysis showed that abalone seeding had to date not altered the benthic community in shallow seeded areas from the community described for shallow unseeded areas.

The study was used to describe the baseline benthic community in areas west of Cape Recife, examine the natural variability along the coast, and determine whether there are relationships between the benthic community composition and emergent abalone abundance. This information will be useful in selection of habitats for abalone seeding in the future. It is important that monitoring and surveying of the study area are to be continued in so allowing for long term data collection which will help in making informed decisions as well as documenting the impact in seeded areas when compared to unseeded areas.

Keywords: Benthic ecology, Haliotis midae, abalone ranching; depth gradient; macroalgae

List of Figures Figure 1. Map illustrating the concession area along the coastline (Witte, 2017)...... 2 Figure 2. Map of the study area between Schoenmakerskop and the Noordhoek Ski Boat Club, Port Elizabeth. The four sites and each depth zone are marked accordingly S1Z1; S1Z2; S1Z3; S2Z1; S2Z2; S2Z3; S3Z1; S3Z2; S3Z3; S4Z1; S4Z2; and S4Z3...... 15 Figure 3. Sites along the coast west of Noordhoek Ski Boat Club to Schoenmakerskop with marked depth zones. a1 and a2 show the Noordhoek Site (Site 1) and its depth zones. b1 and b2 show the Willows site (Site 2) and its depth zones. c1 and c2 show the Willow Grove site (Site 3) and its depth zones. Lastly, d1 and d2 show the Schoenmakerskop site (Site 4) and its depth zones...... 16 Figure 4. Sites identified to compare seeded vs unseeded areas. a1 illustrate Site 6 at the Noordhoek Ski Boat Club which has been seeded. a2 illustrates the three transects sampled at Site 6. b1 illustrates Site 5 at the Willows area and is unseeded. b2 illustrates the unseeded Site 5 amongst the three depth zones of Site 2...... 17 Figure 5. Image (a) representing those captured for analysis and used to quantify cover percentages of the different physical, algal, and faunal groups. Image (b) representing the chain used to measure distance along the bottom on the survey transect...... 20 Figure 6. Average depth (m) for three depth zones at each site (Depth zone 1: 0-1m, Depth zone 2: 1-2m, Depth zone 3: 2-4m; Bars = ± s.e., n = 36)...... 22 Figure 7. Average rugosity value for each depth zone at each site (Depth zone 1: 0- 1m, Depth zone 2: 1-2m, Depth zone 3: 2-4m; Bars = ± s.e., n = 36)...... 23 Figure 8. Average percentage cover (%) for different substrate types for Site 1, 2, 3 and 4 (figure a, b, c, and d respectively) illustrating differences between depth zones. (Depth zone 1: 0-1m, Depth zone 2: 1-2m, Depth zone 3: 2-4m; Bars = ± s.e., n = 36)...... 25 Figure 9. Average percentage cover (%) of dominant algal groups for Site 1, 2, 3 and 4 (figure a, b, c, and d respectively) illustrating differences between depth zones. (Depth zone 1: 0-1m, Depth zone 2: 1-2m, Depth zone 3: 2-4m; Bars = ± s.e., n = 36)...... 28

iii

Figure 10. Average number of individuals (Counts) of macroinvertebrates at Site 1 - 4 (a – d) illustrating changes in depth with each zone (Depth zone 1: 0-1m, Depth zone 2: 1-2m, Depth zone 3: 2-4m; Bars = ± s.e., n = 36)...... 31 Figure 11. Ordination of (a) samples (quadrats) coded for sites, (b) samples coded for depth zone and (c) species from Canonical Correspondence Analysis of log- transformed sample (quadrat) data. Sum of all Eigenvalues = 0.120, Eigenvalues axis 1 = 0.069, Eigen-value Axis 2 = 0.035...... 34 Figure 12. Dendrogram from a hierarchical cluster analysis using a complete linkage clustering method showing the classification of transects based on Bray-Curtis dissimilarity distance calculated from benthic cover. Codes: 1st digit = site, 2nd digit = depth zone, 3rd digit = transect, E.g. 4.2.1 = Site 4, Depth zone 2, Transect 1)...... 36 Figure 13. Average depth (m) for transects sampled at the seeded (Site 6) and unseeded (Site 5) sites. Bars = ± s.e., n = 6...... 38 Figure 14. Average rugosity value for seeded (Site 6) and unseeded (Site 5) sites. Bars = ± s.e., n = 6...... 38 Figure 15. Average percentage cover (%) of the substrate types along the seeded (Site 6) and unseeded (Site 5) sites. Bars = ± S.E., n =6...... 39 Figure 16. Average percentage cover (%) of algal species at the seeded (Site 6) and unseeded (Site 5) sites. (Site 5 n= 36, Bar = ± SE; Site 6 n= 36, Bar = ± SE)...... 40 Figure 17. Average number of individuals (Counts) of macroinvertebrate species at the seeded (Site 6) and unseeded (Site 5) sites. (Site 5 n= 36, Bar = ± SE; Site 6 n= 36, Bar = ± SE)...... 41 Figure 18. Ordination of (a) samples (quadrats) colour coded for sites, (b) samples coded for transects, and (c) species from Canonical Correspondence Analysis of log- transformed sample (quadrat) data. Sum of all Eigenvalues = 0.1033, Eigenvalues axis 1 = 0.0554, Eigen-value Axis 2 = 0.0277...... 43 Figure 19. Dendrogram from a cluster analysis using complete linkage clustering method showing the classification of transects based on Bray-Curtis dissimilarity distance calculated from benthic cover for seeded (site 6) and unseeded (site 5) sites. Codes: S = Site, T = Transect. E.g. S5T1 = Site 5 Transect 1...... 44 Figure 20. Abalone counts within each depth zone across the four sites...... 67 Figure 21. Abalone counts between the seeded (Site 6) and unseeded (Site 5) sites...... 68

List of Tables Table 1. Species presence/absence data across three depth zones at each of the four sites assessing changes in the benthic community both with depth and along shore (Depth zone 1 (0-1m) = Z1, Depth zone 2 (1-2m) = Z2, Depth zone 3 (2-4m) = Z3).35

List of Appendices Appendix 1. Table of GPS coordinates of sites used for the baseline study...... 64 Appendix 2. Table of GPS coordinates for the seeded vs unseeded sites...... 64 Appendix 3. Species list of the coastal section sampled from Noordhoek Ski Boat Club to Schoenmakerskop, Port Elizabeth...... 65 Appendix 4. Actual dive counts of Haliotis midae (Abalone) ...... 67 Appendix 5. R-script for Canonical Correspondence Analysis...... 69 Appendix 6. R-script for the Bray Curtis classification...... 71 Appendix 7. R-script for statistical Kruskal-Wallace and post-hoc Dunn test for multiple comparison (non-parametric data) ...... 72 Appendix 8. R-script for correlations ...... 74

Acknowledgements Firstly, I would like to acknowledge the immense support and guidance provided by my supervisor Dr Paul-Pierre Steyn, of which your assistance throughout the years has been greatly appreciated.

I would also like to thank my diving team and the Research Diving Unit (RDU), headed by Anton Cloete and Paul Baldwin. Thank you for the assistance with diving requirements, skippering, and other support which did not go unnoticed. I also very much appreciate all the divers which have assisted in the research project, spending a lot of time in the water, many days of which were not in the greatest conditions.

This project would not be possible without the support and funding from Lidomix Investments (Pty) Ltd., THRIP funding together with Wild Coast Abalone (Pty) Ltd. Providing funding in 2017 and 2018. The Nelson Mandela University Postgraduate Research Scholarship also funding in 2017 and 2018.

Lastly, I would like to thank my family and friends whom have greatly supported me throughout this process and assisted me whenever needed. Nicola Domoney you are by far the greatest support I have had, encouraging me every step of the way and for your assistance both in the field and the lab. I cannot thank you enough!

To my mom and dad who have encouraged and supported me throughout my varsity career, thank you for everything you have done to help me accomplish this great chapter in my life.

I will never truly be able to say how thankful I am for everyone that helped make this possible.

vii

Chapter 1. Introduction Abalone poaching in South Africa has become a severe problem in the Eastern Cape and stocks have subsequently demonstrated significant drops (Bartley & Bell, 2008; Raemaekers & Britz, 2009). This resource depletion has resulted in the suspension of all recreational exploitation of abalone in the area, while the area has never had a commercial fishery (Raemaekers & Britz, 2009). In order to address the diminishing abalone population, the Department of Agriculture, Forestry and Fisheries (DAFF) has authorized abalone ranching in specifically designated sections of the coast on a trial basis. The area under which Ulwandle Fishing (Pty) Ltd was granted a permit to engage in the proposed activity (Permit Number: 1503759) ranges over a portion of the coast from Cape Recife west towards Sardinia Bay, Port Elizabeth (DAFF, 2014). These abalone ranching projects could not only help improve the natural stocks but may also help facilitate the development of a commercial fishery (de Waal, 2005; Wood, 1993). Ulwandle Fishing, in partnership with Wild Coast Abalone, have since began the abalone ranching processes in this area. Ecological surveys of the benthic communities around Cape Recife have been undertaken to establish the baseline ecological conditions as a requirement of the authorisation for the proposed activity before abalone seeding took place (Steyn & du Preez, 2013). These surveys were conducted in deeper waters (Approximately 5 – 9 meters in depth). However, more recently, after determining that the abalone growth rates are greater on the shallow reefs, interest to seed in the shallow waters along the coast has increased (Steyn and Witte, 2016).

Therefore, the approach of the seeding project has changed to introducing juvenile abalone to sites in shallower waters (Approximately 1 – 3 meters in depth). It is not known how the size classes of seeded abalone might differ in terms of habitat preference in these shallow areas, and whether the benthic habitat, abalone densities, and abalone sizes vary with depth in the 0 – 3 meter depth region.

Furthermore, it is not known how much benthic habitat and benthic communities vary in structure and composition along the coast in the concession area (Figure 1) which has been designated for the development and operation of the abalone ranching program previously mentioned. Most of the research to date has been focused along the coast around Cape Recife (Wood 1993, Godfrey 2003, Witte 2017). However, the

area of interest for the abalone ranching project extends along 17 kilometres of coastline between Cape Recife and the Schoenmakerskop MPA.

Figure 1. Map of the concession area along the coastline (Witte, 2017).

In addition to the above, there is little information on the impact that seeding might have on the benthic community in these shallow areas (1 – 3 meters). The baseline data collection and ecological monitoring relating to the seeding project has focused on the seeding areas in deeper waters (3 – 9 meters).

The first baseline survey of the benthic community structure was undertaken before abalone seeding commenced and took place between 2013 and 2014. A survey conducted in 2015 by Witte and Steyn (2016), compared impacted (i.e. seeded) and control sites at deeper sites (>5 m). A similar baseline survey and monitoring protocol needs to be developed and implemented for the shallower sites in the more recent phase of the abalone ranching project.

The following were the key research questions for this project:

1. What is the community structure of shallow benthic communities along the Port Elizabeth rocky shore?

2. How do the physical conditions and community structure of shallow benthic communities change with depth?

3. Do emergent abalone abundances in shallow benthic communities change with depth? 2

4. How variable is shallow benthic community structure and composition along the shoreline of the Port Elizabeth south coast.

5. How does the benthic community structure and composition in areas seeded with abalone differ from areas which have not been seeded?

Aims & Objectives

To answer these questions, the project will have three broad aims:

Aim 1: Describe the benthic community of shallow reefs (0 – 3 meters) west of Cape Recife.

Objective 1: Determine the change in physical conditions along transects perpendicular to the shore along the coast at several sites.

Objective 2: Determine changes in algal and macroinvertebrate abundance along transects perpendicular to the shore along the coast at several sites.

Aim 2: Compare the benthic community structure of reefs in the shallow depth range along the coast.

Objective 1: Compare the benthic communities within depth ranges between different coastal sections (i.e. Noordhoek - The Willows vs. The Willows - Willow Grove vs. Willow Grove - Schoenmakerskop).

Aim 3: Determine whether shallow reefs seeded with abalone differ from shallow reefs not seeded.

Objective 1: Compare the benthic communities within depth ranges between seeded and unseeded sites.

Chapter 2. Literature Review 2.1 The shallow coastal reef system of the Eastern Cape in South Africa. The shallow coastal reefs of the Eastern Cape in South Africa are considered part of the Agulhas inshore reef system (Sink et al., 2011). The South African coastline is approximately 3000 km long and is home to a diverse range of fauna and flora, with over 800 marine seaweeds being identified for the South African coast (Branch et al., 2010; Griffiths et al., 2010; Stegenga et al., 1997). It is home to approximately 6% of the world’s coastal marine species, of which approximately 33% are endemic to South Africa (Branch et al., 2010; Griffiths et al., 2010; Witte, 2017). The study was focused in the shallow inshore region West of Cape Recife, Port Elizabeth. At depths less than 5 m where there is an expanse of sandstone and calcareous embedded rock with many crevasses and gullies that form unique habitats (McLachlan et al., 1981). The dominant algal species along this coastal region are Plocamium corallorhiza (Turner) J.D. Hooker & Harvey; Gelidium sp.; Codium sp.; Halimeda cuneata Hering and Lithophyllum sp. (McLachlan et al., 1981).

Oceanographic conditions along the stretch of coast are influenced by processes such as being exposed to southwest wave action of the Agulhas current (Goschen & Schumann, 1995; Schumann et al., 1991). Upwelling can occur west of Cape Recife bringing colder bottom waters to the surface due to easterly winds. This upwelling brings cool nutrient rich water in the shallow photic zone where it stimulates productivity (Goschen & Schumann, 1995; Schumann et al., 1991; Schumann et al., 2005). Strong thermoclines can be caused during summer months exhibiting lower winds along the continental shelf to the west (Goschen & Schumann, 1995; Schumann et al., 1991; Schumann et al., 2005).

2.2 Environmental Factors 2.2.1 Temperature The coastal temperatures in South Africa demonstrate an increase in temperatures from the west to east (Sink et al., 2012; Smit et al., 2013). The temperature changes along the coast have been described as the cool west coast; the temperate south coast and the warm temperate east coast (Branch et al., 2010; Branch & Branch, 1981; Lubke, 1988). The region west of Cape Recife and Algoa Bay falls into the temperate

south coast temperature zone. This region, as previously mentioned, is influenced by the Agulhas inshore system. Surface temperatures can range between approximately 11º C and 23º C respectively off the coast at Cape Recife (Knoop, 1988; Schumann et al., 2005; Smit et al., 2013). There is a great variability in these surface water temperatures throughout the year and exhibit the greatest surface temperature variations of the three temperature regions along the coast. There is an average annual temperature of 18.4º C. Further establishing that seasonal temperatures are projected in this region (Anderson & Stegenga, 1989; Goschen et al., 2012; Knoop, 1988; Schumann et al., 2005).

2.2.2 Wind, Waves, and Tides With respect to oceanographic processes wind is a key environmental factor. Wind is an essential driver of coastal and marine processes, as it influences the water temperatures through upwelling (Goschen et al., 2012; Schumann et al., 1991), further influencing the water column temperatures and nutrient availability (Goschen & Schumann, 1995; Knoop, 1988). The area west of Cape Recife is exposed predominantly to west-south-westerly winds compared to the dominant easterly and westerly winds around Algoa Bay. The easterly winds are responsible for the colder water as they drive the surface water away from the shore, which is then replaced by cooler waters with increased nutrients (Goschen & Schumann, 1995; Schumann et al., 1991). The average wind speed recorded for the area is 9 km/h during the autumn and winter months, and 33 km/h during spring and summer months (Schumann et al., 1991).

The South African coast experiences a semi-diurnal tidal cycle. These tides can fluctuate between 0.5 m and 2 m during spring tides (Schumann et al., 2005). Wind is also responsible for wave action and is one of the main processes that drives wave action along the coast. Wave action itself is an important ecosystem driver as waves cause disturbance and exposure to the benthic habitat increasing the available area to be established by new algal species and sessile marine fauna (Goschen & Schumann, 1995; Schumann et al., 2005). There is a notable difference between areas along the coast which are exposed directly to wave action and those that are more sheltered (Clark, 1997; Goschen & Schumann, 1995; Stegenga et al., 1997).

These different exposure levels result in different community structures in the same ecosystem. Waves also allow for constant water movement assisting with essential ecosystem processes such as influencing the structure of the ecosystem; the community composition; changes in sediment structure; exposure; and mixing which further influences phytoplankton abundance (Asmus & Duren, 2007; Clark, 1997; Goschen & Schumann, 1995; Smale et al., 2011; Stegenga et al., 1997). These changes in sediment structure and exposure play an important role in this study, as wave action influences suitable habitat and the community composition at different sites along the coastline.

2.3 Baseline habitat description 2.3.1 Substrate The coastal region west of Cape Recife is comprised of Calcareous Quartzite, however, the geological information available of the South African coastline is very limited and needs further study (Lubke et al., 1998). This is in order to fully understand coastal processes; marine ecosystems; and how their community structures are further influenced by this (Lubke et al., 1998). According to Garner (2013), the type of substrate is an important factor in determining community structure and influences what species could inhabit the area. This is due to the different types of substrate, namely, unconsolidated sandy substrate; mixed unconsolidated substrate; and consolidated rock (Garner, 2013; Witte 2017). There are notable trends in dominance of certain algal communities with each substrate type. Coralline algae will have a greater abundance in sandy substrates opposed to foliose algae, as the sand smothers the algae, whereas on consolidated rocky substrates foliose algae are able to show higher abundances (Ortega-Borges et al., 2009; Garner, 2013; Götz et al., 2009; Knoop, 1988).

2.3.2 Seaweed Communities Seaweed communities form an integral part of the subtidal community along the coastal zone. The occurrence of macroalgae allows for increased habitat and community complexity (Bégin et al., 2004). This complexity is also influenced by the substrate types along the coast, as well as, physical, chemical, and biological drivers

(Anderson & Stegenga, 1989; Denny, 1994; Hearn et al., 2001; Munnik, 1987; Walker, 2012; Wolanski et al., 2001). Seaweeds can alter their environment as they provide a food source to marine herbivores as well as a source of shelter for various marine life (Bégin et al., 2004). Plocamium corallorhiza as an example is a dominant species along the coast in shallower waters, and acts as a source of both food and coverage (Anderson & Stegenga, 1989). Kelp recruits in the temperate reefs of Australia together with wave exposure influence the community composition on exposed surfaces (Smale et al., 2011). Seaweed communities are driven by several factors, some of which have been discussed such that distribution is driven mainly by temperature and exposure (Anderson et al., 2005; Bolton & Stegenga, 2002). The South African coast has a high endemism of seaweed species with over 800 species identified. This has been attributed to the warmer temperate conditions associated with the Agulhas current together with the influence of the cooler Benguela current and isolation along the coast (Bolton & Stegenga, 2002; Stegenga et al., 1997).

Stephenson and Stephenson (1972), established three important biogeographic regions along the South African coast, namely, the cold temperate west coast; the warm temperate south coast, and the subtropical east coast. Other factors influencing seaweed distribution include previously mentioned wave action and tidal variation, as well as, zonation effects such as desiccation; light availability; and insolation (Anderson & Stegenga, 1989; Munnik, 1987). Physical, chemical and biological drivers also play a vital part in seaweed distribution (Denny, 1994; Hearn et al., 2001; Walker, 2012; Wolanski et al., 2001). This includes the aforementioned substrate types; desiccation dynamics; temperature; and exposure (Anderson & Stegenga, 1989; Denny, 1994; Walker, 2012). Salinity; oxygen; carbon dioxide and nutrient availability are some of the important chemical drivers (Hearn et al., 2001). Lastly, competition and grazers are biological drivers of seaweed distribution as well as influence community diversity (Anderson & Stegenga, 1989; Heasman et al., 2007; Munnik, 1987). Although, there is high seaweed endemism along the South African coast seaweed studies are still very limited.

Depth related information about the distribution of seaweed species in shallower regions is typically limited (Anderson & Stegenga, 1989; Ortega-Borges et al., 2009; Götz et al., 2009; Walker, 2012). Therefore, to gain insight into shallower regions and processes that influence these areas, they need to be sampled along a depth gradient.

This is best done by looking at the macrobenthos for community structure, distribution patterns and depth related changes (Anderson & Stegenga, 1989; Heyns et al., 2016). It is found that at deeper depths larger algae are more common as grazing effects present in shallower depths are lowered (Ortega-Borges et al., 2009; Heasman et al., 2007; James et al., 2017; Ojeda & Dearborn, 1989). However, according to Anderson & Stegenga (1989), grazing pressure can be seen to show an increase due to substrate changes with greater depths resulting in an increase of coralline species. Coralline turf algae are the main algal contributor to reef ecosystems as they are well adapted to intense pressures such as grazers; wave action; sand inundation; and exposure (Diez et al., 2003; Heyns et al., 2016).

There is also a greater algal diversity in shallower regions compared to deeper regions due to increased light availability allowing for a wider range of algal species to establish in these areas (Jalali et al., 2015; McKune, 2002). According to Anderson and Stegenga (1989), there are dominant algal species found at different sites due to changes with depth and other drivers such as substrate type and water movement, however, it is expected that species such as Laurencia sp.; Amphiroa sp.; and Anthrocardia sp. will be scattered along the coast at most shallower sites. However, other physical, chemical, and biological drivers such as wave action, nutrient availability, and grazing respectively also influence biomass at any given site (Denny, 1994; Hearn et al., 2001; Walker, 2012; Wolanski et al., 2001).

There is limited baseline data concerning not only the south coast of South Africa but the subtidal region on this coastline in general. Algoa Bay in particular shows a mix of algal species from warm waters found predominantly in the east coast and cold-water species typical of the west coast (Anderson & Stegenga, 1989; Stephenson & Stephenson, 1972). Algal community composition, species composition and biomass are influenced by the hard-rocky substrate of the intertidal to subtidal zones. The physical factors differ along the coastal section investigated in this study. The area West of Cape Recife shows a tropical group of species with moderate water movement that influences the type of intertidal algal groups identified here, namely, coralline turf and encrusting algae (Munnik, 1987). Anderson and Stegenga (1989), state that with increased depth and reduced water movement the benthic algal community becomes dominated by coralline species. Whereas, the coastal section west of Noordhoek to Schoenmakerskop show an affiliation to a mix of tropical and temperate algal groups

with harder more stable substrate types. However, water movement is drastically increased in this section of coast with major wave action resulting in more foliose algae such as Gelidium sp. and Plocamium sp. (Anderson & Stegenga, 1989; Munnik, 1987).

2.4 The South African Commercial Abalone Industry Abalone poaching, although a severe problem in the Eastern Cape is a global issue (Hauck, 1997; White, 2005). Due to resource depletion, aquaculture practices involving abalone are becoming more and more prevalent. The sale of abalone is now regulated by the government and these molluscs can only be removed from their natural habitat, or cultivated, by means of a permit ((Bartley & Bell, 2008; Raemaekers & Britz, 2009; Raemaekers et al., 2011). The commercial industry involving abalone began around 1949, where it saw catches decline greatly from the 1940’s to 1970’s respectively (Goga, 2014; Troell et al., 2006). This resulted in catch quotas being implemented due to unsustainable harvesting. However, before this industry took off abalone was fished sustainably in small communities (Raemaekers et al., 2011).

Intensive poaching in the Eastern Cape has caused emergent abalone stock to be severely depleted. However, with 14 abalone farms in South Africa there has been successful ranching practices initiated by private companies through regulations stipulated by the Department of Forestry and Fisheries (DAFF) and later streamlined by operation Phakisa (Sales & Britz, 2001). Operation Phakisa initiated in 2013 has helped to streamline the ranching program of the abalone Haliotis midae within a concession area along the coast from Cape Recife to Schoenmakerskop Marine Protected Area (MPA) through Ulwandle Fishing (Pty) Ltd. To date over 1 million abalone have been seeded within the area with other partnerships including Lidomix Investments pty ltd; Wildcoast Abalone; Tactical Task Force; Rhodes University and Nelson Mandela University (Witte, 2017).

The ecological health of the marine environment is crucial for sustainable fisheries. Therefore, information that will aid improving the understanding of the ecology; assist in restoring and conserving highly impacted or endangered areas; and monitoring commercially important species, is of great importance (Bailey-Brock, 1989; Claudet & Pelletier, 2004). This is especially true with the growing commercial industry for abalone ranching in South Africa.

2.5 Abalone Habitat Requirements 2.5.1 Distribution Abalone or more specifically the genus Haliotis have been recorded as far back as 80 MYA (Rhode et al., 2012; Wood, 1993). The genus Haliotis is distributed from the sub- Arctic to the Antarctic waters around the world, although, majority are found in the temperate; tropical and sub-tropical waters (Wood, 1993). Haliotis midae the abalone associated with the ranching project in South Africa forms part of a genetically diverse European-Australasian clade due to the isolating Agulhas current dynamics as previously mentioned and is distributed between St Helena Bay to Port St. Johns in South Africa (Rhode et al., 2012). Other South African species include Haliotis spadicea distributed between St Helena Bay to Durban; Haliotis speciose distributed between Port Alfred to Port St. Johns; Haliotis parva distributed between St Helena Bay to Port Alfred; and Haliotis queketti distributed between Port Alfred to Mozambique (Bester-van der Merwe et al., 2012; Branch et al., 2010; Wood, 1993).

2.5.2 Habitat Abalone (H. midae) have specific habitat requirements. They are mostly found in shallow coastal waters and require cryptic habitat such as consolidated rocky areas (Wood, 1993). Space is an important issue for these invertebrates as it is a limiting factor (Witte, 2017). They are also found in areas with high wave exposure. Therefore, they require a habitat that has many crevices; good food availability; encrusting corallines; good water circulation and protection (Day & Branch, 2000, 2002b; de Waal, 2005; Wood, 1993).

Juveniles require very cryptic habitats with many crevices as hiding places (Wood, 1993). There has been significant evidence showing larval forms settling on encrusting coralline algae (Day & Branch, 2000, 2002a). It is also vital that both drift and attached algae are present as this together with water conditions are an indicator for habitat selection (Wood, 1993; Wood & Buxton, 1996). Protection is usually associated with sea urchins (Parechinus angulosa, Leske, 1778) and specifically with respect to the protection of juvenile abalone from predators. However, while urchins do play a part in the protection of juvenile abalone they are not necessarily an indicator of suitable habitat (Day & Branch, 2002a, 2002b; deWaal, 2005). There has been no significant

selectivity to urchins or a boulder type habitat regarding protection. Habitat complexity is, therefore, a vital aspect of habitat, and represents the heterogeneity and physical structure of the habitat (Beck, 1998).

Regarding habitat complexity, depth and rugosity have an important part to play in the suitability of habitat. As previously mentioned, consolidated rock or boulder reef areas are a main indicator for suitable abalone habitat. Low rugosity and sandy substrates would, therefore, be a negative indicator of suitable habitable areas for abalone (Beck, 1998; Garner, 2013; Heyns et al., 2016).

2.5.3 Feeding Abalone are characteristically trap feeders (Wood & Buxton, 1996). They are herbivorous and feed off both attached and drift algae as previously mentioned. Urchin association has been noted that there is an increase in abalone growth rate when associated with urchins (Day & Branch, 2002a). These herbivorous grazers in turn can improve algal productivity by consumption (Wood & Buxton, 1996). Haliotis midae is known for its deep red shell colouring, this is due to it feeding predominantly on red algae. Some of which include Plocamium corallorhiza; Hypnea spicerifera (Suhr) Harvey; Ralfsia verrucosa (Areschoug) Areschoug; Neoralfsia expansa (J. Agardh) P.- E. Lim & H. Kawai ex Cormaci & G. Furnari; and Calliblepharis fimbriata (Greville) Kützing (Wood, 1993). Juveniles and adults both select for R. verrucose, while juveniles will also select for N. expansa opposed to adults selecting Ulva species (Wood, 1993; Wood & Buxton, 1996). However, there does not appear to be a strong link between the selection of seaweed and availability of these species at any site, as it has been found that there can be at least 37 species of seaweed in the stomach contents of H. midae (Barkai & Griffiths, 1986; Wood, 1993). Seaweeds typically avoided by abalone in the Eastern Cape are coralline algae such as Amphiroa ephedraea (Lamarck) Decaisne; Anthrocardia carinata (Kützing) Johansen; Mesophylum sp.; and Ecklonia radiata (C. Agardh) J. Agardh (Wood & Buxton, 1996).

2.6 Sampling marine benthic communities There are various ways to proceed with assessing or investigating the marine benthic environment. These include destructive sampling from quadrats, in situ visual estimates from quadrats, video surveys, time searches, and photographic sampling. In this study, line transects with photographic quadrats were used as a means of sampling the benthic community (Murray et al., 2001; Smale et al., 2011; Smale et al., 2011).

2.6.1 Transects and Quadrats Line transects are the most widely recognised sampling method for monitoring reef and benthic environments (Lovich et al., 2012). This method can underestimate species within the community, however, for purposes assessing algal species and sessile macroinvertebrates have proven to be successful with the belt transect being most useful (Anderson et al., 1979; Buckland et al., 1994; Lovich et al., 2012). The transect length in relation to benthic studies have varied between <10 m to >50 m. The length of the transect should be determined by the specific research project and environmental factors (Götz et al, 2009; Hart et al., 1997; Wood, 1993).

Quadrats are a recognised commonly used method to investigate benthic habitats (de Waal, 2005; Dumas et al., 2009; Murray et al., 2001. Furthermore, quadrat sizes are also determined by the specific research project and environmental factors. It is also important to note that community composition can change both seasonally and spatially which needs to be considered when determining the species diversity and abundance (de Waal, 2005; Hart et al., 1997). Thus, species composition at any given time can be variable. Therefore, transects and quadrats have been the main sampling method for benthic habitats (Lovich et al., 2012).

2.6.2 Visual and Photographic Assessment The transect and quadrat method can further have three approaches. The first is a specimen collection approach, where, divers and researchers sample the quadrats by removing the species. This is, however, destructive sampling; time consuming and can be impacted by depth; visibility; currents; and the overall safety of the site (Dumas et al., 2009). 12

The second is a species count approach, where species are counted along the transects in situ with divers doing actual counts of each quadrat. This approach allows for real values and species are not destructively sampled. However, this can be a time- consuming method and not every species will be accounted for (Dumas et al., 2009).

The third is a photographic approach. This use of photographic quadrats using a camera framer unit have become more favoured as it is a non-destructive means of sampling, and it does not require skilled personnel (Celliers et al., 2007; Murray et al., 2001). All that is required is the general protocol that divers can follow to the researcher’s specifications. The photographic approach is more cost effective, saving both time and resources (Brown et al., 2004; Hart et al., 1997). However, images distort the habitat complexity regarding crevices, cracks and overhangs. Therefore, visual assessment is also required to ensure all information is acquired. Images can then be analysed in the laboratory by the researcher to infer their results.

2.6.3 Depth and Rugosity Sampling the benthic environment along a depth gradient, it is important to establish various depth zones in order to compare the community structure and habitat type. A depth gauge attached to the camera framer unit will ensure you capture the depth in the images that will be later analysed, or a diver’s depth gauge on their watch would also provide a depth to be recorded (Heyns et al., 2016; McKune, 2002).

Rugosity is an essential measurement when looking for abalone habitat as it is essentially the measurement of surface roughness for the substrate type and indicates the availability of refuge space. The rugosity can be measured by means of a chain- and-tape measurement (Pais et al., 2013). This will give a value by contrasting the actual benthic substrate surface to the planar surface area with the tape. This is done by placing the chain of a known transect length along the benthic substrate over crevices, cracks and overhangs. Then taking a straight-line measurement from one end of the chain to the other using a measuring tape to produce a rugosity value in meters, which will help in community analysis (Beck, 1998; Jalali et al., 2015; Pais et al., 2013; Witte, 2017).

Chapter 3. Materials & methods 3.1 Site background This baseline ecological assessment looks at the marine benthic community in shallow waters between 1 m and 5 m, with the aim of providing data for future monitoring and to inform the selection of habitat for the seeding of abalone. According to Day & Branch (2002a, 2002b) and Wood (1993), abalone select for sites with adequate protection from predators, under boulders and in crevices. Predator avoidance is a vital requirement when selecting habitat. Therefore, areas with adequate rocks; boulders; crevices and other various substrate types are essential for abalone survival. Sea Urchins such as Parechinus angulosus Leske, also act as a means of protection for abalone, more so for juveniles (Day & Branch, 2002; de Waal, 2005). Abalone also require an adequate food source and water movement (Clark, 1997; Wood, 1993).

The shallow benthic community ecology in Algoa Bay has not been well described in terms of abalone ranching. Information of these benthic communities are limited to research on the substrate types and biological composition and interactions (Anderson and Stegenga 1989; Garner 2013; Godfrey 2003; Götz et al. 2009; Hutchings & Clarke, 2008; Knoop, 1988; Munnik, 1987; Wood, 1993). According to research by Knoop (1988), there are several dominant seaweed types on shallow reefs in this area, namely, Plocamium corallorhiza and Amphiroa ephedraea (Lamarck) Decaisne with others including, Delisea flaccida (Suhr) Papenfuss; Halimeda cuneata Hering and other coralline algae at deeper sites (> 7 m) (Anderson and Stegenga 1989). Research on abalone is limited to the Cape Recife area by Wood (1993) and Godfrey (2003). This area has been described as having gullies containing predominantly gravel type substrate with encrusting forms and geniculate coralline algae making up the majority of the algal community structure (Godfrey, 2003; Wood, 1993)

This study included several gullies approximately 5 m or less in depth, west of the Noordhoek ski-boat club to Schoenmakerskop, Port Elizabeth (Figure 2 and 3). The study sites are west of Cape Recife where abalone seeding is already taking place. There have been approximately over 1.4 million abalone seeded on the reefs between Cape Recife and Schoenmakerskop. The seeded areas are referred to as impacted sites, however, this survey focused on the baseline community ecology of the shallow benthic reefs where abalone will be seeded in future. Therefore, the study site

consisted of areas that are yet to be seeded with abalone. The sites were selected in order to monitor the natural benthic community along the coast (Figure 2). Thus, the benthic community at unseeded sites could then be compared to seeded sites which will also be investigated in this assessment (Figure 2 and 4). Three transects (minimum) were conducted perpendicular to the shore along gullies of each selected site. Transects were placed along three depth zones at each site, namely, <1 m; 1-2 m; and >3 m depth ranges. These transects were marked by Global Positioning System (GPS) coordinates at each site (Appendix 1 and 2).

Figure 2. Map of the study area between Schoenmakerskop and the Noordhoek Ski Boat Club, Port Elizabeth. The four sites and each depth zone are marked accordingly S1Z1; S1Z2; S1Z3; S2Z1; S2Z2; S2Z3; S3Z1; S3Z2; S3Z3; S4Z1; S4Z2; and S4Z3.

(a1) (a2)

(b1) (b2)

(c1) (c2)

(d1) (d2) Figure 3. Sites along the coast west of Noordhoek Ski Boat Club to Schoenmakerskop with marked depth zones. a1 and a2 show the Noordhoek Site (Site 1) and its depth zones. b1 and b2 show the Willows site (Site 2) and its depth zones. c1 and c2 show the Willow Grove site (Site 3) and its depth zones. Lastly, d1 and d2 show the Schoenmakerskop site (Site 4) and its depth zones.

(a1) (a2)

(b1) (b2)

Figure 4. Sites identified to compare seeded vs unseeded areas. a1 illustrate Site 6 at the Noordhoek Ski Boat Club which has been seeded. a2 illustrates the three transects sampled at Site 6. b1 illustrates Site 5 at the Willows area and is unseeded. b2 illustrates the unseeded Site 5 amongst the three depth zones of Site 2.

3.2 Dive Survey and Sampling The sampling methods are in accordance with the approved benthic community monitoring and sampling protocol submitted to DAFF in 2013 (Steyn & du Preez, 2013). The survey involved capturing images (samples) using a GoPro attached to a camera framer with a depth gauge along several transects for each site. The site coordinates were noted for future monitoring. Destructive sampling was limited to the collection of algal voucher specimens required to identify species. Dive surveys were conducted at a spring low tides, with a dive team consisting of no less than two scientific divers. Snorkeling was the main method of data collection in the shallow waters less than 2 m depth and SCUBA for depths greater than 2 m.

A transect line of approximately 10 m was set up along the selected gullies by means of an anchored chain. The benthic community was then surveyed, and images recorded at 0.5 m intervals. A camera framer unit with an attached quadrat of

0.5 x 0.5 m quadrat size was used to sample the benthic community by means of a Go-Pro mounted in the centre of the framer unit to capture the entire quadrat (i.e. The abundance of abalone; various seaweeds; urchins and other macroinvertebrates was determined by these images). Five images were taken per quadrat using the multi- image shooting mode and the best quality image used in the analysis. Emergent abalone counts were recorded by direct counts in a 1 m2 quadrat placed every 1 m along the length of the transect line.

The depth at each quadrat was recorded in each image by means of a depth gauge attached to the framer unit. Each site was sampled at three depth zones, namely, <1 m; 1-2 m; and >3 m depth ranges. Rugosity was measured using the chain-and-tape method (Pais et al., 2013). An anchored chain (i.e. transect line) was placed on the seafloor over crevices and overhangs, and a straight-line measurement was recorded as a planar surface over the chain. Rugosity was then calculated as follows:

퐷 푅 = 푐ℎ푎𝑖푛 퐷푠푡푟푎𝑖𝑔ℎ푡

Where:

푅 = rugosity (dimensionless ratio),

퐷푐ℎ푎𝑖푛 = length of chain (m)

퐷푠푡푟푎𝑖𝑔ℎ푡 = straight line distance between transect start and end (m).

Seaweed samples were removed in the general study area at each depth zone, at each site for species identification. These samples were identified in the laboratory in order to compile a comprehensive species list for the coast as well as indication of presence or absence for each depth zone at each site along the coast. Specimens that could not be identified within a few days were preserved in a 5% formaldehyde- seawater solution.

Ethics clearance (animal) was obtained for the potential harm that the sampling may cause to fauna in the samples (NMMU Ethics Approval No.: A15-SCI-BOT-001).

Species level identification was not always possible from the images. Therefore, a functional group approach to identification was adopted to describe the benthic community. Seaweed species were grouped into Coralline Turf; Codium sp.; Encrusting Algae; Foliose Algae; and Upright Coralline, with easily recognisable species such as Plocamium corallorhiza and Halimeda cuneata were described to species level. These groups play a role in abalone diet (Foliose algae, Plocamium) or settlement behaviour (Encrusting algae).

Major macrofaunal groups recognised from images include Haliotis midae. Oxystele spp., Parechinus angulosis, Porifera, Scutellastra spp., Turbo sarmaticus, and Ascidiaceae. These represent major competitors for food or space.

This functional group approach has also been used in other ecological studies employing video or photographic data collection, and has been found to be adequate for the detection of ecological impacts (Götz et al. 2009; Celliers et al. 2007; Ferrari et al. 2018; Hart et al. 2013).

Only macrofaunal and macroinvertebrate (i.e. greater than the average sea urchin size ~ 2 cm) data was recorded using both visual and photographic analysis. The substrate type in terms of rocks; boulders; gravel; shell grit; and sand were documented. Sampling and monitoring occurred over a period of a year. Final sampling and data collection took place in the last months of 2018.

Sites were labelled based on position West of Cape Recife as well as the sampled depth zones within each site. Where ‘S’ indicates the geographic location i.e. the site; ‘Z’ indicates the depth zone at a site (<1 m; 1-2 m; >2 m); ‘T’ indicates the transect line within a zone; and lastly, ‘Q’ indicates the quadrat along the transect line.

I.e. The code S3Z2T2Q5 represents Site 3, Zone 2, Transect 2, Quadrat 5.

3.3 Analysis 3.3.1 Image Analysis Several images were captured at each quadrat interval of each transect at each site (Figure 5). These images were analysed for abundance or estimated percentage cover of the benthic community. These estimates include abundance for various algal groups; abalone (Haliotis midae); sea urchins (Parechinus angulosus); and other

macroinvertebrates. Substrate types (e.g. boulder, rock, sand) were also assigned a cover percentage. Presence or absence data was recorded by comparing the percentage cover (algal groups; substrate and macroinvertebrates) in the images to the counts recorded during dives.

(a) (b)

Figure 5. Image (a) representing those captured for analysis and used to quantify cover percentages of the different physical, algal, and faunal groups. Image (b) representing the chain used to measure distance along the bottom on the survey transect.

All data were tested for normality using the r-package “nortest” (Gross & Ligges, 2015), using the Anderson-Darling test (ad.test function).

Mean abundances for the various groups were compared. The data not found to have a normal distribution was tested using Kruskal-Wallis (Kruskal.test function) and Wilcoxon rank sum tests (wilcox.test function) respectively in the r-package stats to look for differences within groups. Post hoc Dunn tests (Ogle et al., 2018) with p-values adjusted with the Benjamini-Hochberg method were used to test for differences between sites and depth zones (dunnTest function). Normally distributed data for the seeded vs unseeded sites was tested using T-tests in Excel.

Correlation analyses were done between selected algal groups, macroinvertebrates, and physical attributes (i.e. substrate and depth) using the r-package “ggpubr” (Kassambara, 2018), where the scatter plots (ggscatter function) and correlation tests (cor.test function) were used.

Community data for sites were analysed using various community ecology metrics provided for in the r-package “vegan” (Oksanen et al., 2015). A comparison of the physical attributes, algal groups, and macroinvertebrates of the different sites along the coast, as well as the changes in these physical variables with depth were examined where the average percentage cover of algae, and average individual counts of macroinvertebrates for each site (n=9) were also plotted with indication of standard error. Depth and rugosity were also plotted for each depth zone (n=3) along each site.

Benthic community data were analysed using multivariate analysis techniques to determine whether discrete reef communities were present along the coast, and at different depths (0 – 5 m). Canonical Correspondence Analysis (CCA), was used to show patterns in community composition for transects and depth zone, and also made it possible to explain these differences in terms of specific environmental drivers using the r-package “vegan” (Oksanen et al., 2015), where ordination plots (cca function and ordiplot function) were used.

Lastly, ecological dissimilarity was assessed by means of Bray-Curtis similarity using the r-package “vegan” (Oksanen et al., 2015), where a dendrogram (vegdist function, method “bray”), was used and sites were classified using a hierarchical clustering with the hclust function in the r-package vegan. The result of these analyses were plotted in a classification dendrogram.

Chapter 4. Results

4.1 Changes in the benthic community along a depth gradient and along shore. 4.1.1 Physical Attributes Figure 6 illustrates the average depth (m) of each depth zone for each site. Although, site 2 depth zone 2 was shallower than the three other sites, it still fell into the depth range specified for that depth zone 2 (1-2 m).

4,5 4 3,5 3 Site 1 2,5 Site 2 2 Site 3

1,5 Site 4 Average Depth (m) Depth Average 1 0,5 0 Zone 1 Zone 2 Zone 3 Depth Zones

Figure 6. Average depth (m) for three depth zones at each site (Depth zone 1: 0-1m, Depth zone 2: 1-2m, Depth zone 3: 2-4m; Bars = ± s.e., n = 36).

Rugosity was compared between site and the three depth zones at each site (Figure 7). Rugosity shows that with increasing depths the benthic substrate demonstrates a lower surface complexity. However, this is most evident in depth zone 3. General trend shows the rugosity was higher in the shallow depth zone compared to the deeper zone.

0,45 0,4 0,35 0,3 Site 1 0,25 Site 2 0,2

0,15 Site 3 Average Rugosity Average 0,1 Site 4 0,05 0 Zone 1 Zone 2 Zone 3 Depth Zones

Figure 7. Average rugosity value for each depth zone at each site (Depth zone 1: 0- 1m, Depth zone 2: 1-2m, Depth zone 3: 2-4m; Bars = ± s.e., n = 36).

Figure 8(a – d) shows the average substrate type composition [boulder; rock; gravel (pebbles/cobbles); shell grit; and sand] of the four sites across the three depth zones. Kruskal-Wallis test was used to test for differences in the physical variables across sites and depth zones, where P < 0.05 was the threshold for significance. There was a difference in the mean cover of all the respective substrate types between the four sites along the coast (KW: df = 3; P < 0.05). Dunn post-hoc tests also showed that each substrate type had a significant difference between the four sites with each producing a value of P < 0.05 across the depth zones for all four sites. Boulder substrate cover (71 ± 3,33%; 73 ± 5,81%; 89 ± 9,65%; 33 ± 8,90%) for the four sites made up the majority of substrate cover.

The cover of rock differed between sites (P < 0.05, n = 36), and between the three depth zones. Gravel substrate showed a significant difference (P < 0.05, n = 36) between all sites, and between all three depth zones. Shell grit was only recorded in sites 2 and 4, and showed significant difference (P < 0.05, n = 18) between the two sites and between the three depth zones (P < 0.05, n = 18).

Sandy substrate cover showed a significant difference (P < 0.05) between all sites and across all the three depth zones (P < 0.05). Gravel, shell grit and sand show lower

cover percentages over the four sites. There is a trend between gravel and shell grit over the four sites, whereby, as the gravel cover increases, shell grit shows an increase across the four sites too. Shell grit and sand also seems to covary over the four sites. Overall, all four sites showed significant differences (P < 0.05) in terms of the cover by the respective substrate types, with depth zone 1 showing the most variability between the substrate types present.

100 80 Boulder 60 Rock 40 Gravel 20 Cover Cover (%) Shell Grit 0

Average Average Percentage Zone 1 Zone 2 Zone 3 Sand Depth Zones a)

100 80 Boulder 60 Rock 40 Gravel

Cover Cover (%) 20 Shell Grit 0 Average Average Percentage Zone 1 Zone 2 Zone 3 Sand Depth Zones b)

100 80 Boulder 60 Rock 40 Gravel

Cover Cover (%) 20 Shell Grit 0 Average Average Percentage Zone 1 Zone 2 Zone 3 Sand Depth Zones c)

100 80 Boulder 60 Rock 40 Gravel 20 Cover Cover (%) Shell Grit 0

Average Average Percentage Zone 1 Zone 2 Zone 3 Sand Depth Zones d) Figure 8. Average percentage cover (%) for different substrate types for Site 1, 2, 3 and 4 (figure a, b, c, and d respectively) illustrating differences between depth zones. (Depth zone 1: 0-1m, Depth zone 2: 1-2m, Depth zone 3: 2-4m; Bars = ± s.e., n = 36).

4.1.2 Algal Composition and Diversity Figure 9 (a-d) shows the average abundance (% cover) for the main algal groups identified at each of the four sites and between the three depth zones. A Kruskal-Wallis (KW) test revealed that there were significant differences (P < 0.05) in the abundance of several of the algal groups between the different depth zones along the four sites.

There are noticeable trends in figure 9 (a-d), where sites with a higher abundance of coralline turf cover show a lower abundance of foliose algae and Plocamium corallorhiza abundance. However, foliose algae (KW: df = 3; P < 0.05) and Plocamium corallorhiza (KW: df = 3; P < 0.01) showed significant difference between the four sites and along the depth zones (KW: df = 2; P < 0.05).

Foliose algae and Plocamium corallorhiza have a weak but significant correlation (R =0.12; P < 0.05). Foliose algae can be seen with its highest abundance in depth zone 2 and Plocamium corallorhiza showed increases in abundance as upright corallines increase. Foliose algae and upright coralline had a weak but significant correlation (R = 0.17; P < 0.05).

Upright coralline also showed significant differences (KW: df = 3; P < 0.05) between the sites and between the depth zones (KW: df = 2; P < 0.05). However, there was no correlation between Plocamium corallorhiza and upright coralline (R = 0.03; P = 0.43). There was also a trend where coralline turf abundance increases, encrusting algae decreases and vice versa.

Encrusting algae showed a significant difference between sites (KW: df = 3; P < 0.05), however, did not show any difference across depth zones (KW: df = 2; P = 0.49).

Coralline turf also shows a higher abundance in shallower depths, decreasing to deeper depth zones while upright coralline increases. Coralline turf shows a significant difference between the four sites (KW: df = 3; P <0.05) and across the depth zones (KW: df = 2; P < 0.05).

Codium sp. showed significant differences (KW: df = 3; P < 0.04) between the sites and across the depth zones (KW: df = 2; P <0.01). Lastly, Halimeda cuneata showed significant differences between the sites and across the depth zones (KW: P < 0.05). There is a weak but significant correlation between Halimeda cuneata and sand substrate (R = 0.23; P < 0.05).

Post-hoc Dunn tests showed that Codium sp. cover differed between all sites (P < 0.05). Codium sp. also showed significant difference between the three depth zones (P < 0.05). Coralline turf cover showed significant differences between all sites and the three depth zones (P < 0.05), depth zone 3 more so than depth zones 1 and 2. Foliose algal cover showed significant differences between all sites and between all three depth zones (P < 0.05). Encrusting algal cover differed significantly between all sites (P < 0.05). Encrusting algae however, showed no difference between the three depth zones (P > 0.05). The mean cover of Plocamium corallorhiza showed significant difference between sites 3 and 4 (P < 0.05) and depth zones 1 and 3, and depth zones 2 and 3 (P < 0.05). Plocamium corallorhiza and depth showed a significant but weak negative correlation but was significant (R = -0.1; P < 0.01). Plocamium corallorhiza also shows a weak positive correlation with boulder substrate (R = 0.19, P < 0.05). Upright coralline showed significant differences between all sites excluding site 3 (P=0.28). Upright coralline also showed significant differences between all sites and three depth zones (P < 0.05). Lastly, Halimeda cuneata cover showed significant differences between all sites (P < 0.05). H. cuneata cover also showed significant differences between depth zones 1 and 3, and depth zones 2 and 3 (P < 0.05).

Overall there was a substantial difference in the cover of the respective algal groups between all the sites. Although a trend is noticeable between the algal groups and sites (e.g. Figure 9 shows decreases in coralline turf with increases in upright coralline with depth) there was no significant correlation between the groups. The correlation tests also showed that the community composition within the three depth zones are also significantly different between each other with no significant correlation between the algal groups.

60 Codium sp. 50 Coralline Turf 40 30 Encrusting Algae 20 Halimeda cuneata

Cover Cover (%) 10 0 Plocamium sp. Average Average Percentage Zone 1 Zone 2 Zone 3 Foliose Algae Depth Zones Upright Coralline a)

60 Codium sp. 50 40 Coralline Turf 30 Encrusting Algae 20 Halimeda cuneata

Cover Cover (%) 10 0 Plocamium sp. Average Average Percentage Zone 1 Zone 2 Zone 3 Foliose Algae Depth Zones Upright Coralline b)

60 Codium sp. 50 40 Coralline Turf 30 Encrusting Algae 20 Halimeda cuneata

10 Cover Cover (%) 0 Plocamium sp.

Average Average Percentage Zone 1 Zone 2 Zone 3 Foliose Algae Depth Zones Upright Coralline c)

60 Codium sp. 50 Coralline Turf 40 30 Encrusting Algae 20 Halimeda cuneata

Cover Cover (%) 10 Plocamium sp. 0

Average Average Percentage Zone 1 Zone 2 Zone 3 Foliose Algae Depth Zones Upright Coralline d) Figure 9. Average percentage cover (%) of dominant algal groups for Site 1, 2, 3 and 4 (figure a, b, c, and d respectively) illustrating differences between depth zones. (Depth zone 1: 0-1m, Depth zone 2: 1-2m, Depth zone 3: 2-4m; Bars = ± s.e., n = 36).

4.1.3 Macroinvertebrate Composition and Diversity Figures 10 (a – d) illustrates the average macroinvertebrate cover for the four sites. Kruskall-Wallace Tests were used to test for differences in the mean invertebrate abundances between sites, as well as between the three depth zones at the four sites. The analyses also showed significant differences in the abundances of the invertebrate groups between sites, and the abundances of several invertebrate groups were found to differ between depth zones. The invertebrate groups that showed significant differences between the four sites were Oxystele sp. (KW: df = 3; P < 0.05), Parechinus angulosus (KW: df = 3; P < 0.05), and Scutellastra sp. (KW: df = 3; P < 0.01). Haliotis midae showed no significant difference between sites, however, did show a difference between depth zones (KW: df = 2; P < 0.01). The Haliotis midae (abalone) counts are shown in Appendix 4, figure 20. Parechinus angulosus and Oxystele sp. were the dominant macroinvertebrates across the sites.

Post-hoc Dunn’s tests were conducted for Scutellastra sp. showed significant difference in the abundance of these invertebrates between sites 1 and 3 (P < 0.01), and between sites 1 and 4 (P < 0.02). There were also significant differences in their abundance between depth zones 1 and 3, and between depth zones 2 and 3 (P < 0.05).

Post-hoc Dunn’s test for Haliotis midae showed no significant differences between the four sites (P > 0.05). There was a significant difference between depth zones 1 and 2 (P < 0.02) and between depth zones 1 and 3 (P < 0.01) for Haliotis midae. There was a weak but significant correlation between Haliotis midae and depth (R = 0.096; P = 0.02). Haliotis midae also showed a weak positive but significant correlation with boulder substrates (R = 0.12; P < 0.01). However, there was no correlation between Haliotis midae and Plocamium corallorhiza (R = 0.01; P = 0.76).

Parechinus angulosus post-hoc Dunn’s test show significant differences between site 4 with each of the other sites (P < 0.05). There were also significant differences between depth zones 1 and 3 (P = 0.03), and depth zones 2 and 3 (P < 0.01). There is a trend noticed when comparing figures 10 (a-d), where Haliotis midae is present at greater numbers with Parechinus angulosus. However, there was no correlation between the two (R = 0.008; P = 0.84).

Haliotis midae also showed a general trend where their numbers increase as Oxystele sp. numbers decrease, however, there was no correlation between the two (R = 0.01; P = 0.81). Turbo sarmaticus showed no significant difference between sites or depth zones (P > 0.05). Oxystele sp. post-hoc Dunn’s test show significant differences between all sites and depth zones (P < 0.05).

10,00 8,00 Haliotis midae 6,00 4,00 Oxystele sp.

2,00 Parechinus angulosus Individuals

0,00 Scutellastra sp. Average Average Number of Zone 1 Zone 2 Zone 3 Turbo sarmaticus Depth Zones a)

10,00 8,00 Haliotis midae 6,00 Oxystele sp. 4,00

2,00 Parechinus angulosus Individuals 0,00 Scutellastra sp. Average Average Number of Zone 1 Zone 2 Zone 3 Turbo sarmaticus Depth Zones b)

10,00 8,00 Haliotis midae 6,00 Oxystele sp. 4,00 Parechinus angulosus

2,00 Individuals 0,00 Scutellastra sp. Average Average Number of Zone 1 Zone 2 Zone 3 Turbo sarmaticus Depth Zones c)

10,00 8,00 Haliotis midae 6,00 Oxystele sp. 4,00 Parechinus angulosus 2,00 Individuals Scutellastra sp. 0,00

Average Average Number of Zone 1 Zone 2 Zone 3 Turbo sarmaticus Depth Zones d) Figure 10. Average number of individuals (Counts) of macroinvertebrates at Site 1 - 4 (a – d) illustrating changes in depth with each zone (Depth zone 1: 0-1m, Depth zone 2: 1-2m, Depth zone 3: 2-4m; Bars = ± s.e., n = 36).

4.1.4 Community Composition and Diversity Figure 11 shows the Ordination of (a) samples (quadrats) coded for sites, (b) samples coded for depth zone and (c) species, from Canonical Correspondence Analysis of log-transformed sample (quadrat) data. Environmental parameters used as constraining variables included substrate type and depth. While the figures in the previous section examined differences in the cover of specific ecosystem components between sites and depth zones, a CCA of the community data was used to reveal broader ecological differences between the four sites.

The ordination shows complete overlap of the samples in ordination space. Sites 1 and site 4 depth zones 3 being influenced by sandy substrate and depth. The benthic community composition at site 4 zone 1, site 1 zone 2, and site 2 zone 2, are influenced more by shell grit and gravel type substrates and not depth. The other sites were mainly influenced by rock and boulder type substrates.

The clustering of samples in the ordination space, while noticeable, overlap entirely which suggests that the samples are not from significantly different communities (i.e. the benthic community at each site and depth zone does not differ significantly between each other).

The species plot showed similar grouping to that shown in the sample plot, whereby, site 1 and site 4 depth zones 3 is more affiliated with sandy substrate types, and the coralline turf and Halimeda cuneata algal groups. While sites 4 zone 1; site 1 zone 2 and site 2 zone 2 shows a discrete affiliation towards gravel; shell grit; and rock substrate types, with coralline turf and Codium sp. Foliose algae, including Plocamium corallorhiza with boulder type substrates; and Haliotis midae were more affiliated with depth zone 3 at all four sites.

Table 1 shows species level presence/absence data across three depth zones at each of the four sites assessing changes in the benthic community both with depth and along shore. The foliose algae are represented by species such as Calliblepharis fimbriata (Greville) Kützing; Delisea flaccida (Suhr) Papenfuss; Dictyota dichotoma var. intricata (C.Agardh) Schmidt; Dictyota naevosa (Suhr) Montagne; Gelidium abbottiorum R.E.Norris; Gelidium pteridifolium R.E.Norris, Hommersand & Fredericq; Gigartina insignis (Endlicher & Diesing) F.Schmitz; Hypnea viridis Papenfuss; Laurencia flexuosa Kützing; Laurencia natalensis Kylin and Sargassum elegans Suhr.

The Upright coralline group represents species such as Amphiroa anceps (Lamarck) Decaisne; Amphiroa ephedraea (Lamarck) Decaisne; Arthrocardia carinata (Kützing) Johansen; Arthrocardia flabellata (Kützing) Manza; Jania adhaerens J.V. Lamouroux; and Jania verrucosa J.V. Lamouroux.

Coralline turf was defined as a group of algal species of both small algae and the growing forms of larger algae species. These included Amphiroa ephedraea (Lamarck) Decaisne; Arthrocardia carinata (Kützing) Johansen; Dictyota naevosa (Suhr) Montagne; Gelidium abbottiorum R.E. Norris; Gelidium pteridifolium R.E. Norris, Hommersand & Fredericq; Halimeda cuneata Hering; Hypnea viridis Papenfuss; Laurencia flexuosa Kützing; Laurencia natalensis Kylin; and Ulva rigida C. Agardh.

Codium species were identified as Codium lucasii subsp. capense P.C. Silva and Codium stephensiae C.I. Dickinson.

Encrusting algae was identified as Heydrichia woelkerlingii R.A. Townsend, Y.M. Chamberlain & Keats; Mesophyllum sp.; and Ralfsia expansa (J. Agardh) J. Agardh.

Figure 11. Ordination of (a) samples (quadrats) coded for sites, (b) samples coded for depth zone and (c) species from Canonical Correspondence Analysis of log- transformed sample (quadrat) data. Sum of all Eigenvalues = 0.120, Eigenvalues axis 1 = 0.069, Eigen-value Axis 2 = 0.035.

Table 1. Species presence/absence data across three depth zones at each of the four sites assessing changes in the benthic community both with depth and along shore (Depth zone 1 (0-1m) = Z1, Depth zone 2 (1-2m) = Z2, Depth zone 3 (2-4m) = Z3).

Macroalgae Site 1 Site 2 Site 3 Site 4 Species Z1 Z2 Z3 Z1 Z2 Z3 Z1 Z2 Z3 Z1 Z2 Z3 Upright Coralline Amphiroa anceps (Lamarck) Decaisne * * Amphiroa ephedraea (Lamarck) Decaisne * * * * * * * * Arthrocardia carinata (Kützing) Johansen * * * * * * * * * * * Arthrocardia flabellata (Kützing) Manza * * * * * * * Halimeda cuneata Hering * * * * * Jania adhaerens J.V. Lamouroux * * * * * * * * Jania verrucosa J.V. Lamouroux * * * * * * * * Foliose Algae Calliblepharis fimbriata (Greville) Kützing * * Caulerpa filiformis (Suhr) Hering * * * * * Champia compressa Harvey * * * * Delisea flaccida (Suhr) Papenfuss * * Dictyota dichotoma var. intricata (C. Agardh) Schmidt * * Dictyota naevosa (Suhr) Montagne * * Gelidium pteridifolium R.E. Norris, Hommersand & * * Fredericq Gelidium pristoides (Turner) Kützing, 1843 * * Hypnea viridis Papenfuss * Laurencia flexuosa Kützing * * * * Laurencia natalensis Kylin * * * * * * Plocamium corallorhiza (Turner) J.D. Hooker & Harvey * * * * * * * * * * Plocamium suhrii Kützing * * * * * Porphyra capensis Kützing, 1843 * Portieria hornemannii (Lyngbye) P.C. Silva, 1987 * Rhodophyllis reptans (Suhr) Papenfuss, 1956 * Sargassum elegans Suhr * * * * Trematocarpus sp. (Kützing, 1843) * Codium Species Codium duthieae P.C. Silva, 1956 Codium lucasii subsp. capense P.C. Silva * * * * * * Codium stephensiae C.I. Dickinson * * * * Encrusting Algae Heydrichia woelkerlingii R.A. Townsend, Y.M. * * * * Chamberlain & Keats Mesophyllum sp. * * * Ralfsia expansa (J. Agardh) J. Agardh * * * * * * * * Ulva rigida C. Agardh, 1823 * Invertebrate Taxa Ascidiaceae * * * * (Sea squirts) Haliotis midae (Linnaeus, 1758) * * * * * * * * Oxystele sp. (Philippi, 1847) * * * * * * * Parechinus angulosus (Leske, 1778) * * * * * * * Porifera * * * Scutellastra sp. (H. Adams & A. Adams, 1854) * * * * Turbo sarmaticus (Linnaeus, 1758) *

Figure 12 shows a hierarchical cluster analysis of Bray-Curtis dissimilarity scores of each site transects along each depth zone. The data shows five defined clusters. However, there seems to be a lot of natural variability within the data. While the first cluster on the left shows an affiliation to the shallower depths, the classification is inconclusive showing no clear groupings according to site or depth zones. The CCA ordination provides more realistic representation of the benthic community.

Figure 12. Dendrogram from a hierarchical cluster analysis using a complete linkage clustering method showing the classification of transects based on Bray-Curtis dissimilarity distance calculated from benthic cover. Codes: 1st digit = site, 2nd digit = depth zone, 3rd digit = transect, E.g. 4.2.1 = Site 4, Depth zone 2, Transect 1).

4.2 Seeded vs unseeded sites The sites shown in the map of the study area (Chapter 3 - Figure 4) were selected within the ranching concession area at the Noordhoek Ski Boat Club (seeded) and the Willows area (unseeded), with respect to specific environmental attributes (i.e. accessibility, safety of the divers). These sampling biases can be noticed in the data. Site 5 was an area which had not been seeded (i.e. Willows area) with juvenile abalone, while site 6 was a seeded area (i.e. Noordhoek Ski Boat Club).

4.2.1 Physical Attributes Figure 13 shows the average depth (m) of each of the transects for the two sites. The depth ranged between 1 and 3 metres and was similar for both sites with Wilcoxon rank sum test showing no significant difference in depth (P = 0.28).

Rugosity was compared between the two sites using Wilcoxon rank sum test (Figure 14). Rugosity was also similar between the two sites showing no significant difference (P = 0.19).

Figure 15 shows the substrate types of the two sites in terms of: boulder; rock; gravel (pebbles/cobbles); shell grit; and sand. The data was tested for normality and the mean cover for seeded and unseeded sites was compared using Wilcoxon rank sum test. The two sites did not differ in terms of the cover of any of the substrate types (P > 0.05, n = 6).

While the mean boulder substrate cover was 77 ± 3,07% and 54 ± 27,49%, for the unseeded (site 5) and seeded (site 6) sites respectively, this was found to be a significant difference (W = 2906, P < 0.05, n = 6). At both sites, the substrate was dominated by boulder and rock substrate, with gravel, shell grit and sand showing lower cover percentages at both sites.

3,5

2,5

1,5 Average Average Depth(m) 1

0,5

0 Unseeded Seeded Sites

Figure 13. Average depth (m) for transects sampled at the seeded (Site 6) and unseeded (Site 5) sites. Bars = ± s.e., n = 6.

0,25

0,2

0,15

0,1

Average Average Transect LineRugosity 0,05

0 Unseeded Seeded Sites

Figure 14. Average rugosity value for seeded (Site 6) and unseeded (Site 5) sites. Bars = ± s.e., n = 6.

60 Boulder 50 Rock 40 Gravel 30 Shell Grit

20 Sand Average Average Percentage Cover (%)

0 Unseeded Seeded Sites

Figure 15. Average percentage cover (%) of the substrate types along the seeded (Site 6) and unseeded (Site 5) sites. Bars = ± S.E., n =6.

4.2.2 Algal Composition and Diversity Figure 16 shows the average abundance (% cover) for the main algal groups identified at each of the sites. Wilcoxon rank sum tests for non-parametric data and T-tests for normally distributed data revealed significant differences (P < 0.05) in the abundance of some of the algal groups between the two sites. Foliose algal cover differed significantly between seeded (5 ± 1.2%) and unseeded (14 ± 2.4%) sites (W = 3271, P < 0.05). Halimeda cuneate also showed significant difference between the sites (W = 2177, P < 0.05). A T-test revealed that encrusting algae showed no significant differences between the sites (P = 0.06).

25 Codium sp. Coralline Turf 20 Encrusting Algae 15 Halimeda cuneata 10 Plocamium sp.

5 Foliose Algae Average Average Percentage Cover (%) Upright Coralline 0

-5 Unseeded Seeded Sites

Figure 16. Average percentage cover (%) of algal species at the seeded (Site 6) and unseeded (Site 5) sites. (Site 5 n= 36, Bar = ± SE; Site 6 n= 36, Bar = ± SE).

Plocamium corallorhiza cover was significantly higher in the unseeded site (21 ± 3.5%) than the site seeded with abalone (14 ± 0.9%) (W = 1211, P < 0.05). Upright coralline cover also showed significant differences between the two sites (W = 2652, P < 0.05) with the seeded site having much greater Upright Coralline cover (17 ± 3%) than site 6 (11 ± 1.5%).

The cover data shown in figure 16 suggest a negative relationship between encrusting algae and Plocamium corallorhiza and upright coralline and coralline turf cover. Site with high cover of the former two groups appear to have low cover for the latter two. However, correlation analysis showed that only encrusting algae had weak negative but significant negative correlation with upright coralline algae (R = -0.44; P < 0.05) and coralline turf (R = -0.3; P < 0.01) respectively. Lastly, where Plocamium corallorhiza has higher cover there is lower foliose algal cover.

4.2.3 Macroinvertebrate Composition and Diversity Figure 17 shows the mean macroinvertebrate numbers for the two sites (individuals per 0.5 m2). Data was tested for normality using an Anderson-Darling test and Wilcoxon rank sum tests were used to test for differences in the mean cover for invertebrates between the two sites.

10,00 9,00 8,00 7,00

6,00 Haliotis midae 5,00 Oxystele sp. 4,00 Parechinus angulosus 3,00 Scutellastra sp. Turbo sarmaticus

Average Average NumberIndividuals of 2,00 1,00 0,00 Unseeded Seeded Sites

Figure 17. Average number of individuals (Counts) of macroinvertebrate species at the seeded (Site 6) and unseeded (Site 5) sites. (Site 5 n= 36, Bar = ± SE; Site 6 n= 36, Bar = ± SE).

Parechinus angulosus abundance showed significant differences between sites (P < 0.05). Turbo sarmaticus abundance showed no significant difference between the sites (P = 0.9). There were no significant differences found for Scutellastra sp. (P = 0.82) or Oxystele sp. (P = 0.13) between the two sites.

Haliotis midae (abalone) counts were recorded by visual assessment per m2 along the transect lines. The actual abalone counts are shown in Appendix 4, figure 21. While the mean Haliotis midae abundance appears to be higher in the seeded site compared to that of the unseeded site there is no difference between the two sites (P = 0.11).

4.2.4 Community Composition and Diversity Figure 18 shows the Ordination plot for Canonical Correspondence Analysis (CCA) of abundance data from quadrats (samples) grouped by (a) site and (b) transects, and (c) species from Canonical Correspondence Analysis of log-transformed sample (quadrat) data. Environmental parameters which were used as constraining variables included substrate type and depth.

Distinct differences between specific ecosystem components were noted when compared on an individual basis in section 4.2.2 above. The ordination of CCA fir samples of the community data showed to some differences between the two sites. The samples from site 5 plot towards one side of the ordination, and the samples from site 6 towards the other, some overlap evident. When considered on a transect basis this pattern is also evident. This suggests that the samples are not from significantly different communities and do not show clear grouping (i.e. the benthic community at each site does not differ between each other). However, due to some variability or noise in the environmental data has resulted in a certain differentiation between the unseeded (site 5) and seeded (site 6) sites and is due to the higher abundance of Plocamium corallorhiza and abalone counts in the seeded site (site 6). The species plot showed similar grouping as that which was shown in the site plot, whereby, the seeded site (site 6) shows a discrete affiliation towards rock; gravel; and shell grit substrate types, with foliose algae including Plocamium corallorhiza and the macroinvertebrate Haliotis midae.

Figure 18. Ordination of (a) samples (quadrats) colour coded for sites, (b) samples coded for transects, and (c) species from Canonical Correspondence Analysis of log-transformed sample (quadrat) data. Sum of all Eigenvalues = 0.1033, Eigenvalues axis 1 = 0.0554, Eigen-value Axis 2 = 0.0277.

Figure 19 shows a hierarchical cluster analysis of Bray-Curtis dissimilarity scores for transects for the seeded and unseeded sites.

Transects form clusters, however the clusters are not representative of the respective sites. This suggests that, based on Bray-Curtis scores for transects, the seeded and non-seeded communities do not represent ecologically distinct communities.

Either the introduction of seeded abalone has not had an impact on the benthic community in seeded areas, or due to the natural variability of these benthic communities the effect of seeding on the benthic environment are concealed by the variability.

Figure 19. Dendrogram from a cluster analysis using complete linkage clustering method showing the classification of transects based on Bray-Curtis dissimilarity distance calculated from benthic cover for seeded (site 6) and unseeded (site 5) sites. Codes: S = Site, T = Transect. E.g. S5T1 = Site 5 Transect 1.

Chapter 5. Discussion The discussion has been written in two parts. Section 5.1 discusses the general patterns of benthic communities along the coast in the abalone ranching concession area. It also explored the change in these communities in response to water depth (0 – 5m).

Section 5.2 of the discussion focusses on the effect of abalone seeding on the benthic community is discussed. The abalone seed in this area had only been introduced approximately one year prior to the ecological sampling, and it is possible that the impact of seeding may not yet have been apparent at the time of sampling.

5.1 Shallow benthic communities in the abalone ranching area This research focussed on the marine benthic environment in shallow waters (<5 m) with the aim of providing baseline information to serve as a benchmark for future ecological monitoring. The ongoing monitoring of potential ecological impacts associated with ranching is a condition of the permit issued for experimental ranching in Port Elizabeth by the Department of Agriculture Forestry and Fisheries.

The information will also serve to inform future selection of suitable habitat for abalone seeding, as limited information is available on the benthic community in this depth range (Anderson and Stegenga, 1989). Published work by Wood (1993) and Godfrey (2003), considered several indicators for suitable abalone habitat; such as substrate type, refuge space, food availability, depth, and water movement. In this study, four typical sites within the ranching concession area were characterised with respect to substrate types, rugosity, depth, seaweed cover, and macroinvertebrates; in order to assess suitability for abalone ranching.

A key aim of this study was to determine how variable reef communities are along the shore within the concession area. It is for this reason that the sampling sites were selected at roughly equally spaced distances along the section of coast that is being used for the abalone ranching project. The change in seaweed communities along the South African coast is well documented (Bolton & Anderson 1990; Bolton & Anderson, 2004; Bolton et al., 2004). However, these studies took place over a fairly large

geographic range and the associated change in inshore coastal seawater temperatures has been widely accepted as the driver for these changes. On a more local scale, other factors are responsible for changes in the seaweed community. Factors such as storm damage or herbivory (Dayton et al. 1984, Harrold 1985), local oceanographic conditions (Bermejo et al., 2015), and geomorphology (Sangil et al., 2011) could all cause local variability in the benthic community composition.

The data comparing the benthic communities at different sites along the abalone ranching concession area showed that, in terms of ecological similarity, there is very little overall change in the benthic communities along the coastal section. Ferrari et al. (2018), similarly, found no differences in the benthic community along a ca. 10 km stretch of coast in Southeastern Australia. They found that substrate had an influence on the distribution of the different communities, as well as their complexity (Ferrari et al., 2018).

However, there are significant differences in terms of discrete macroalgal and macrofaunal groups between sites along the coast. The differences in the benthic community can be attributed to the difference in substrate types at these sites. A permutation test for the CCA showed that all the substrate types except gravel, had a significant influence over the benthic community composition.

Coralline turf cover was higher at sites that had greater sand cover, while foliose algal cover and that of Plocamium corallorhiza was higher at sites with more boulder/rock substrate cover. These sandy turf dominated sites were gently deeper, with the rocky Plocamium dominated quadrats being located in deeper water, which is in accordance with findings by Knoop (1988). Götz et al. (2009), speculated that higher water turbidity and less available light decreased algal abundance, particularly foliose algae with increases in upright coralline. Haliotis midae shows the same trend showing low abundances where there is higher sand compared to sites with greater boulder/rock substrate cover and Plocamium corallorhiza cover. Plocamium corallorhiza, is a main food source for abalone in this sampling area and showed significant differences in abundance between the four sites. This relationship between abalone numbers and the abundance of Plocamium corallorhiza was also recognised by Knoop (1988) and Wood (1993).

The role of substrate type in structuring marine benthic reef communities has been shown in a number of studies (Baynes, 1999; Guidetti et al., 2004; Piazzi et al., 2002), and is most likely responsible for the variability found in the reef communities in the study area. This has also been shown in studies focussing on abalone habitat, where substrate type and depth exhibited the most influence on the community composition (Godfrey, 2003; Heasman et al., 2007; Zeeman et al., 2013).

A comparison of the benthic communities at different depths showed that in terms of overall ecological similarity the communities do not differ. Multivariate analysis (classification of Bray Curtis scores) showed no distinct clustering of transects reflecting different depth zones, and no clear grouping was not noticed in the canonical correspondence analysis (CCA) for quadrats from different depths. A permutation test for the CCA showed that depth did play a role in the benthic communities. However, since sampling was basically done along a continuous depth gradient, the parameter did not result in distinct groupings in the ordination plot. Abalone abundance was one of the few faunal groups that showed a significant correlation with depth, albeit a weak relationship. The poor correlation may be due to the cryptic nature of these molluscs and potential underestimation of abalone numbers in the counts along the transects.

At a local scale community composition, at least in terms of benthic invertebrates, has been found to be dependent on depth (James et al. 2017, Ojeda & Dearborn 1989). According to Walker (2012), the Northern Florida Reef Tract shows distinct community differences across the coastal section with increased depth, while Ortega-Borges et al. (2009), found that shallow subtidal habitats are different from each other.

In this study depth did not appear to exert a strong influence over community structure. However, it has been shown that depth along with other factors such as rugosity; substrate; and disturbance influence species distribution (Heyns et al, 2016; McKune, 2002). Rugosity is a measurement of the roughness of the benthic surface and is often used in reef studies. Rugosity is an important parameter for ecological assessment, as it essentially describes habitat complexity (Beck, 1998). Suitable habitat for abalone will have a high habitat complexity as this is an indication of high species diversity (Wood, 1993). These habitats with higher rugosity values allow for greater algal diversity along with macroinvertebrates and is also impacted by disturbance of the

area (Heasman et al., 2007; Ojeda & Dearborn, 1989). The rugosity values in this study are similar to values considered typical of good abalone habitat in the study by Witte (2017).

The suitability of the sites investigated in this study for abalone seeding can be inferred from the community data and the physical variables measured. Key identifiers for good abalone seeding sites include the presence of rocky substrate; Plocamium corallorhiza and encrusting algae, and water movement (Chapman, 2002; Wood, 1993). Species associated with Haliotis midae such as Plocamium corallorhiza; Parechinus angulosus; algal turf; and encrusting coralline are, therefore, useful in identifying suitable habitat for abalone seeding. While variable, these species / groups were all common in the study area, and the area is expected to offer good abalone habitat considering the above criteria. However, in terms of these groups, there was no overall significant difference in faunal abundances between sites or across depth zones. Day & Branch (2002b) and De Waal (2005), showed a strong relationship between juvenile abalone and the sea urchin Parechinus angulosus. However, the results from this study did not show a similar pattern.

5.2 Effect of abalone seeding on the benthic community Little information is available on the impact that abalone ranching has on benthic communities, with the current research limited to reports on the ranching at Port Nolloth, on the South African west coast (Hutching & Clark, 2008), and work by Hart et al. (2013), on the stock-enhancement of Greenlip abalone in Australia. Studies elsewhere, involving environmental monitoring of benthic communities with, and without abalone, have found no change within the benthic community when compared to control sites. However, these have noted that increased abalone abundances may result in a decline in urchin numbers, which may only become evident over long time- scales (Hart et al., 2013; Wing et al., 2015; and Zeeman et al., 2014). Over 1 million abalone have been seeded during the ranching operation in Port Elizabeth, it is therefore, important to monitor the impact that this sudden increase in the abalone density may have on the benthic ecosystem in the area.

Ecological monitoring of sites seeded with abalone have been taking place since 2015 (Witte, 2017). It is also not only important to monitor seeded and unseeded sites to determine any variations between habitat and community composition, but also to monitor seeded sites after abalone have been harvested. Harvesting involves destructive methods to obtain the abalone and the impacts thereof need to be noted and monitored to see how the habitat recovers.

The seeded and unseeded sites in this study were similar in terms of depth range, rugosity, and substrate composition, so it is unlikely that these factors could be responsible for differences in the community with the two sites. Notwithstanding the physical similarities, these sites did differ in terms of both seaweed cover and abalone abundance. Plocamium sp. and encrusting coralline algal groups were more abundant at the seeded site, while the foliose and upright coralline groups were better represented at the unseeded site.

These differences are reflected in the CCA ordination that plots the quadrats from the seeded site to the one side and the unseeded site to the other. However, there was substantial overlap which suggests a great deal of ecological similarity between the two sites. The classification of transects for the two sites do not support an ecological distinction between the communities at the seeded and unseeded site.

A difference between the two sites was not anticipated. The abalone feed largely on drift algae, therefore, do not significantly change the benthic environment by removing algae cover (Wood & Buxton, 1996). The introduction of hatchery reared juvenile abalone was therefore not expected to alter algal cover. This is further supported by the work of Zeeman et al. (2014), where abalone exclusion from the benthos did not result in any change in the benthic community structure. Hart et al. (2013), monitored benthic communities for two years after abalone seeding and found no ecological difference between seeded and unseeded sites.

While this appears to be the case for abalone, it is not always the case as physical conditions may have a more important role that structures the benthic community than the introduction of abalone. Gotz et al. (2009), in the Goukamma reserve, found altered algal cover in response to fishing pressure. The increase in algae was believed to be the result of reduced herbivory due to lower numbers of grazing fish species in the fished area. Estes & Duggins (1995), showed a similar effect when sea urchin numbers declined in the presence of increase predation by sea otters. The removal or urchins benefitted algal recruitment and increased kelp abundance. The reason that effects such as these are not seen with abalone introductions is most likely related to the fact that these molluscs feed on drift algae (Wood & Buxton, 1996; Zeeman, 2014).

A number of questions related to the impact of ranching were not answered in this study, and the sampling strategy employed was most likely inappropriate to address these.

The first is relevant in where very small abalone size classes are seeded or if larval seeding should take place. These young abalone feed on biofilms and the microphytobenthos. Won et al. (2007), showed that while adult abalone fed on macroalgae, juveniles that feed on microalgae could compete with other taxa, such as sea cucumbers or amphipods, which belong to similar feeding guilds. The effect that this increased competition will have on the benthic ecosystem should be considered in future work.

Another question relates to refugia and competition for space. Competition between abalone and other species like sea urchins have been shown to affect the relative abundances of these invertebrates (Centoni, 2018; Lowry & Pearce, 1973). On the reefs in Port Elizabeth abalone would compete with sea urchins, giant periwinkle and

possibly chitons for refuge space. The impact of abalone ranching on these interactions were not addressed in this study.

Chapter 6. Concluding remarks

Abalone farming is a growing industry and has been listed as a priority area for research and development in Operation Phakisa, which aims to grow the blue economy. Abalone ranching in South Africa is still largely untested, and the project in Port Elizabeth is one of the first of its kind in the country. The project has resulted in more than 1 million abalone seeded in the concession area to date.

When juvenile abalone are seeded into the marine environment it is important to focus on areas that are suitable for abalone survival. In order to ensure this, baseline ecological studies that characterise the benthos in terms this ‘suitability’ are required. There are specific requirements of abalone that constitute a suitable habitat. The main requirements being habitat complexity, depth, and substrate.

The results from this study show that the benthic communities along the section of coast earmarked for ranching does not change much in ecological terms but is highly variable on a local scale. This variability has been shown to be related to patchiness in the substrate types within the study area.

The data also shows that, in the shallow reef area sampled in this research (0 – 4 m water depth), abalone abundance appears to increase with depth. This can be attributed to subtle changes in the cover of certain algal groups between the different depth zones. However, other factors that were not accounted for in this study could also play a role (e.g. water movement).

In order to detect subtle changes in community structure, particularly with respect to smaller taxa, future studies should include permanent transects and more fine scale sampling. Recruitment level impacts, specifically the impact that increased grazer densities may have on local recruitment of macroalgae and sessile invertebrates, has not been investigated. This could have longer-term impacts, that have been alluded to in a few studies on abalone impacts, but still not examined in any detail.

The role of physical drivers in these shallower reaches could also be explored further, as it is clear that local small-scale oceanographic conditions, and the indirect effect that this has on sediment dynamics and substrate stability also plays a role in benthic community structure and abalone habitat suitability.

References

Amus, H. & van Duren, L. 2007. Introduction: hydrodynamic control of aquatic ecosystem processes – how does water movement affect different levels of organisation? Estuarine, Coastal and Shelf Science 75, 279-280.Anderson, R., McKune, C., Bolton, J., Declerck, O., and Tronchin, E., 2005. Patterns in subtidal seaweed communities on coral-dominated reefs at Sodwana Bay on the KwaZulu-Natal coast, South Africa. African Journal of Marine Science 27, 529-537.

Anderson, D. R., Laake, J. L., Crain, B. R., Burnham, K. P. 1979. Guidelines for line transect sampling of biological populations. The Journal of Wildlife Management 43, 70-78.

Anderson, R.J., and Stegenga, H., 1989. Subtidal algal communities at Bird Island, Eastern Cape, South Africa. Botanica Marina 32, 299-311.

Aued, A.W., Smith, F., Quimbayo, J.P., Candido, D.V., Longo, G.O., Ferreira, C.E.L., Witman, J.D., Floeter, S.R., and Segal, B., 2018. Large-scale patterns of benthic marine communities in the Brazilian Province. PLoS One 13, e0198452.

Bailey-Brock, J.H., 1989. Fouling community development on an artificial reef in Hawaiian waters. Bulletin of Marine Science 44, 580-591.

Baynes, T.W., 1999. Factors structuring a subtidal encrusting community in the southern Gulf of California. Bulletin of Marine Science 64, 419-450.

Beaudry, S., Daher, H., Nolan, P., and Petersen, M., 2012. Colonization rates of algae on artificial substrates in Douglas Lake. University of Michigan. Pp 1-12.

Barkai, R., and Griffiths, C.L., 1986. Diet of the South African abalone Haliotis midae. South African Journal of Marine Science 4, 37-44.

Bartley, D.M., and Bell, J.D., 2008. Restocking, stock enhancement, and sea ranching: arenas of progress. Reviews in Fisheries Science 16, 357-365.

Beck, M.W., 1998. Comparison of the measurement and effect of habitat structure on gastropods in rocky intertidal and mangrove habitats. Marine Ecology Progress Series 169, 165-178.

Bégin, C., Johnson, L.E. and Himmelman, J.H., 2004. Macroalgal canopies: distribution and diversity of associated invertebrates and effects on the recruitment and growth of mussels. Marine Ecology Progress Series 27, 121- 132.

Bermejo, R., Ramirez-Romero, E., Vergara, J.J. and Hernández, I., 2015. Spatial patterns of macrophyte composition and landscape along the rocky shores of the Mediterranean–Atlantic transition region (northern Alboran Sea). Estuarine, Coastal and Shelf Science 155, 17-28.

Bester-van der Merwe, A. E., DÁmato, M. E., Swart, B. L., Roodt-Wilding, R. 2012. Molecular phylogeny of South African abalone, its origin and evolution as revealed by two genes. Marine Biology Research 8, 727-736.

Bolton, J.J. and Anderson, R.J., 1990. Correlation between intertidal seaweed community composition and sea water temperature patterns on a geographical scale. Botanica Marina 33, 447-458.

Bolton, J.J. and Anderson, R.J., 2004. Marine vegetation 15. Vegetation of southern Africa, Pp 348.

Bolton, J.J., Leliaert, F, DeClerck, O, Anderson, R.J., Stegenga, H., Engledow, H.E. and Coppejans, E., 2004. Where is the western limit of the tropical Indian Ocean seaweed flora? An analysis of intertidal seaweed biogeography on the east coast of South Africa. Marine Biology 144, 51-59.

Bolton, J.J., and Stegenga, H., 2002. Seaweed species diversity in South Africa. South African Journal of Marine Science 24, 9-18.

Branch, G.M., Griffiths, C.L., Branch, M.L., and Beckley, L.E., 2010. Two Oceans; A guide to the marine life of southern Africa. Struik Nature, Cape town. Pp 371- 419.

Branch, M., and Branch, G.M., 1981. The Living Shores of Southern Africa, 1st ed. Struik Publishers, Cape Town.

Brown, E., Cox, E., Jokiel, P., Rodgers, K., Smith, W., Tissot, B., Coles, S.L., and Hultquist, J., 2004. Development of benthic sampling methods for the coral reef assessment and monitoring program (CRAMP) in Hawai’i. Pacific Science 58, 145-158.

Buckland, S. T., Anderson, D. R., Burnham, K. P., Laake, J. L. 1994. Distance sampling: estimating abundance of biological populations. Biometrics 50, 891- 892.

Celliers, L., Mann, B., Macdonald, A., and Schleyer, M., 2007. A benthic survey of the rocky reefs off Pondoland, South Africa. African Journal of Marine Science 29, 65-77.

Centoni, J., 2018. Negatively correlated abundance suggests competition between red abalone (Haliotis rufescens) and red sea urchins (Mesocentrotus franciscanus) inside and outside established MPAs closed to commercial sea urchin harvest in northern California. Unpublished MSc Thesis. Humbold State University. Pp 1-143.

Chapman, M.G., 2002. Early colonization of shallow subtidal boulders in two habitats. Journal of Experimental Marine Biology and Ecology 275, 95-116

Chu, D.A., Jaufman, Y.J., Zibordi, G., Chern, J.D., Mao, Jietai., Li, C., and Holben, B.N. 2003. Global monitoring of air pollution over land from the earth observing system-terra moderate resolution imaging spectoradiometer (MODIS). Journal of Geophysical Research 108, 1-18.

Clark, B.M., 1997. Variation in surf-zone fish community structure across a wave- exposure gradient. Estuarine, Coastal and Shelf Science 44, 659–674.

Claudet, J., and Pelletier, D., 2004. Marine protected areas and artificial reefs: A review of the interactions between management and scientific studies. Aquatic Living Resources 17, 129-138.

DAFF., 2014. DAFF Announces Abalone TAC for 2014/15 Fishing Season.

Day, E., and Branch, G.M., 2000. Relationships between recruits of abalone Haliotis midae, encrusting corallines and the sea urchin Parechinus angulosus. South African Journal of Marine Science 22, 137-144.

Day, E., and Branch, G.M., 2002a. Effects of sea urchins (Parechinus angulosus) On recruits and juveniles of abalone (Haliotis midae). Ecological Monograph 72, 133-149.

Day, E., and Branch, G.M., 2002b. Influences of the sea urchin Parechinus angulosus (Leske) on the feeding behaviour and activity rhythms of juveniles of the South African abalone Haliotis midae Linn. Journal of Experimental Marine Biology and Ecology 276, 1-17.

Dayton, P.K., Currie, V., Gerrodette, T., Keller, B.D., Rosenthal, R. and Tresca, D.V., 1984. Patch dynamics and stability of some California kelp communities. Ecological Monographs 54, 253-289.

Denny, M.W., 1994. Extreme drag forces and the survival of wind- and water-swept organisms. Journal of Experimental Biology 194, 97-115. de Waal, S., 2005. Boulders or urchins? Selecting Seeding Sites for Juvenile Haliotis midae along the Namaqualand Coast of South Africa. African Journal of Marine Science 27, 501-504.

Diez, I., Santolaria, A., and Gorostiaga, J.M., 2003. The relationship of environmental factors to the structure and distribution of subtidal seaweed vegetation of the western Basque coast (N Spain). Estuarine, Coastal and Shelf Science 56, 1041-1054.

Dumas, P., Bertaud, A., Peignon, C., Léopold, M., and Pelletier, D., 2009. A “quick and clean” photographic method for the description of coral reef habitats. Journal of Experimental Marine Biology and Ecology 368, 161-168.

Estes, J.A. and Duggins, D.O., 1995. Sea otters and kelp forests in Alaska: generality and variation in a community ecological paradigm. Ecological Monographs 65, 75-100.

Ferrari, R., Marzinelli, E.M., Ayroza, C.R., Jordan, A., Figueira, W.F., Byrne, m., Malcome, H.A., Williams, S.B., and Steinberg, P.D., 2018. Large-scale assessment of benthic communities across multiple marine protected areas using an autonomous underwater vehicle. PLoS One 13, e0193711.

Garner, C.J., 2013. Characterisation and biotic classification of Eastern Cape mixed substrate shores. Nelson Mandela Metropolitan University. Pp 1-395.

Gersun, L., Anderson, R.J., Hart, J.R., Maneveldt, G.W., and Bolton, J.J., 2016. Sublittoral seaweed communities on natural and artificial substrata in a high- latitude coral community in South Africa. African Journal of Marine Science 38, 303-316.

Godfrey, B., 2003. The potential of abalone stock enhancement in the Eastern Cape Province of South Africa. Rhodes University. Pp 1-170.

Goga, K., 2014. The Illegal Abalone Trade in the Western Cape. Pretoria. Pp 1-18.

Goschen, W.S., and Schumann, E.H., 1995. Upwelling and the occurrence of cold water around Cape Recife, Algoa Bay, South Africa. South African Journal of Marine Science 16, 57-67.

Goschen, W.S., Schumann, E.H., Bernard, K., Bailey, S., and Deyzel, S., 2012. Upwelling and ocean structures off Algoa Bay and the south-east coast of South Africa. African Journal of Marine Science 34, 525-536.

Götz, A., Kerwath, S.E., Attwood, C.G., and Sauer, W., 2009. Effects of fishing on a temperate reef community in South African 2: benthic invertebrates and algae. African Journal of Marine Science 31, 253-262.

Griffiths, C.L., Robinson, T.B., Lange, L., and Mead, A., 2010. Marine biodiversity in South Africa: an evaluation of current states of knowledge. PloS One 5, 1-13.

Gross, J., and Ligges, U., 2015. nortest: Tests for Normality. R package version 1.0- 4. https://CRAN.R-project.org/package=nortest

Guidetti, P., Bianchi, C.N., Chiantore, M., Schiaparelli, S., Morri, C. and Cattaneo- Vietti, R., 2004. Living on the rocks: substrate mineralogy and the structure of subtidal rocky substrate communities in the Mediterranean Sea. Marine Ecology Progress Series 274, 57-68.

Harrold, C. and Reed, D.C., 1985. Food availability, sea urchin grazing, and kelp forest community structure. Ecology 66, 1160-1169.

Hart, A.M., Fabris, F., Strain, L.W.S., Davidson, M., and Brown, J., 2013. Stock enhancement in greenlip abalone part II: population and ecological effects. Reviews in Fisheries Science 21, 310-320.

Hart, A.M., Gorfine, H.K., and Callan, M.P., 1997. Abundance estimation of blacklip abalone (Haliotis rubru) I. An analysis of diver-survey methods used for large- scale monitoring. Fisheries Research 29, 159-169.

Hauck, M., 2009. Crime, environment and power: revisiting the abalone fishery. South African Journal of Criminal Justice 22, 229.

Hearn, C.J., Atkinson, M.J., and Falter J.L., 2001. A physical derivation of nutrient uptake rates in coral reefs: effects of roughness and waves. Coral Reefs 20, 347-356.

Heasman, M.P., Liu, W., Goodsell, P.J., Hurwood, D.A., and Allan, G.L., 2007. Development and delivery of technology for production, enhancement and aquaculture of blacklip abalone (Haliotis rubra) in New South Wales. NSW Department of Primary Industries – Fisheries Final Report Series No. 95. ISSN 1449-9967. Pp 1-228.

Heyns, E.R., Bernard, A.T.F., Richoux, N.B., and Götz, A., 2016. Depth-related distribution patterns of subtidal macrobenthos in a well-established marine protected area. Marine Biology 163, 1-15.

Hutchings, K., and Clarke, B.M., 2008. Environmental assessment of Port Nolloth Marine, Seafarm’s abalone (Haliotis midae) ranching operation. Cape Town. Pp 1-24

Jalali, M.A., Ierodiaconou, D., Monk, J., Gorfine, H., and Rattray, A., 2015. Predictive mapping of abalone fishing grounds using remotely sensed LiDAR and commercial catch data. Fisheries Research 169, 26-36.

James, L.C., Marzloff, M.P., Barrett, N., Friedman, A., and Johnson, C.R., 2017. Changes in deep reef benthic community composition across a latitudinal and environmental gradient in temperate Eastern Australia. Marine ecology progress series 565, 35-52.

Kassambara, A., 2018. ggpubr: 'ggplot2' Based Publication Ready Plots. R package version 0.2. https://CRAN.R-project.org/package=ggpubr.

Kendra, M., Renaud, P.E., Andrade, H., Goszczko, I., and Ambrose, W.G., 2013. Benthic community structure, diversity, and productivity in the shallow Barents Sea bank (Svalbard Bank). Marine Biology 160, 805-819.

Knoop, W.T., 1988. Subtidal benthic macroalgal production in the south west corner of Algoa Bay. Botany Department. University of Port Elizabeth. Pp 1-319.

Leliaert, F., Anderson, R.J., Bolton, J.J., and Coppejans, E., 2000. Subtidal understorey algal community structure in kelp beds around the Cape Peninsula (Western Cape South Arica). Botanica Marina 43, 359-366.

Lovich, R. E., Hayes, W. K., Mushinsky, H., and Rodda, G. H. 2012. Transect surveys, including line distance. Pp. 227-234 in R. W. McDiarmid, M. S. Foster, C. Guyer, J. W. Gibbons, and N. Chernoff (eds.), Reptile Biodiversity: Standard Methods for Inventory and Monitoring. University of California Press, Berkeley, California.

Lowry, L.F. and Pearse, J.S., 1973. Abalones and sea urchins in an area inhabited by sea otters. Marine Biology 23, 213-219.

Lubke, R.A., 1988. The Coastal Environment, in: Lubke, R.A., de Moor, I. (Eds.), Field Guide to the Eastern and Southern Cape Coasts. University of Cape Town Press, Cape Town. Pp 1-63.

McKune, C., 2002. An analysis of subtidal seaweed communities on a depth gradient at Sodwana Bay, on the KwaZulu-Natal coast, South Africa. University of Cape Town. Pp 1-35.

McLachlan, A., Lombard, A.T., and Louwrens, S., 1981. Trophic structure and biomass distribution on two East Cape rocky shores. South African Journal of Zoology 16, 85-89.

Munnik, L., 1987. Identification of intertidal rocky-shore macroalgal communities in the vicinity of Port Elizabeth. University of Port Elizabeth. Pp 1-253.

Murray, S.N., Ambrose, R.F., and Deithier, M.N., 2001. Methods for performing monitoring, impact, and ecological studies on rocky shores. MMS OCS Study 2001-070. Santa Barbara, California. Pp 1-227.

Mustafa, S., 2003. Stock enhancement and sea ranching: objectives and potential. Reviews in Fish Biology and Fisheries 13, 141-149.

Ogle, D.H., Wheeler, P., and Dinno, A., 2018. FSA: Fisheries Stock Analysis. R package version 0.8.22, https://github.com/droglenc/FSA.

Ojeda, F.P., and Dearborn, J.H., 1989. Community structure of macroinvertebrates inhabiting the rocky subtidal zone in the Gulf of Maine: seasonal and bathymetric distribution. Marine ecology progress series 57, 147-161.

Oksanen, J., Blanchet, F.G., Friendly, M., Kindt, R., Legendre, P., McGlinn, D., Minchin, P.R., O'Hara, R.B., Simpson, G.L., Solymos, P., Stevens, M.H., Szoecs, E., and Wagner, H., 2018. vegan: Community Ecology Package. R package version 2.5-3. https://CRAN.R-project.org/package=vegan.

Ortega-Borges, L., Tuya, F., and Haroun, R.J., 2009. Does depth and sedimentation interact with sea urchins to affect algal assemblage patterns on eastern atlantic reefs?. Journal of Shellfish Research 28, 947-955.

Pais, M.P., Henriques, S., Costa, M.J., and Cabral, H.N., 2013. Improving the “chain and tape” method: A combined topography index for marine fish ecology studies. Ecological Indicators 25, 250-255.

Piazzi, L., Pardi, G., Balata, D., Cecchi, E. and Cinelli, F., 2002. Seasonal dynamics of a subtidal north-western Mediterranean macroalgal community in relation to depth and substrate inclination. Botanica Marina 45, 243-252.

R Core Team, 2018. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R- project.org/.

Raemaekers, S., Hauck, M., Bürgener, M., Mackenzie, A., Maharaj, G., Plagányi, É.E., and Britz, P.J., 2011. Review of the causes of the rise of the illegal South African abalone fishery and consequent closure of the rights-based fishery. Ocean & Coastal Management 54, 433-445.

Raemaekers, S.J., and Britz, P.J., 2009. Profile of the illegal abalone fishery (Haliotis midae) in the Eastern Cape Province, South Africa: Organised pillage and management failure. Fisheries Research 97, 183-195.

Rhode, C., 2013. Signatures of Selection in Natural and Cultured Abalone (Haliotis midae): A Population Genomics Study. Stellenbosch University. Pp 1-299.

Sales, J., and Britz, P.J., 2001. Research on abalone (Haliotis midae L.) cultivation in South Africa. Aquaculture Research 32, 863-874.

Sangil, C., Sansón, M. and Afonso-Carrillo, J., 2011. Spatial variation patterns of subtidal seaweed assemblages along a subtropical oceanic archipelago: thermal gradient vs herbivore pressure. Estuarine, Coastal and Shelf Science 94, 322-333.

Schumann, E.H., Churchill, J.R.S., and Zaayman, H.J., 2005. Oceanic variability in the western sector of Algoa Bay, South Africa. African Journal of Marine Science 27, 65-80.

Schumann, E.H., Illenberger, W.K., and Goschen, W.S., 1991. Surface winds over Algoa Bay, South Africa. South African Journal of Science 87, 202-207.

Sink, K., Holness, S., Harris, L., Majiedt, P., Atkinson, L., Robinson, T., Kirkman, S., Hutchings, L., Leslie, R., Lamberth, S., Kerwath, S., von der Heyden, S., Lombard, A., Attwood, C., Branch, G., Fairweather, T., Taljaard, S., Weerts, S., Cowley, P., Awad, A., Halpern, B., Grantham, H., and Wolf, T., 2012. National Biodiversity Assessment 2011: Technical Report. Volume 4: Marine and Coastal Component. South African National Biodiversity Institute, Pretoria. Pp 325.

Smale, D.A., Langlois, T.J., Kendrick, G.A., Meeuwig, J.J., Harvey, E.S., 2011. From fronds to fish: The use of indicators for ecological monitoring in marine benthic ecosystems, with case studies from temperate Western Australia. Reviews in Fish Biology and Fisheries 21, 311-337.

Smale, D. A., Wernberg, T., & Vance, T. 2011. Community development on subtidal temperate reefs: the influences of wave energy and the stochastic recruitment of a dominant kelp. Marine Biology 158, 1757-1766.

Stephenson, T.A., and Stephenson, A., 1972. Life between tide marks on rocky shores. W.H. Freeman and Co. San Francisco. Pp 1-425.

Stegenga, H., Bolton, J.J., and Anderson, R.J., 1997. Seaweeds of the South African West Coast, 1st ed. Contributions to the Bolus Herbarium18, Cape Town. Pp 1-655.

Steyn, P., and du Preez, D.R., 2013. A baseline survey of dominant benthic biota in a proposed experimental abalone seeding area at Cape Recife, Port Elizabeth. Port Elizabeth.

Steyn, P., and Witte, A. D., 2016. A survey of the effect of abalone seeding on the marine benthic community at Cape Recife, Port Elizabeth (2016 Survey). Port Elizabeth. Pp 1-29.

Troell, M., Robertson-Anderson, D., Anderson, R.J., Bolton, J.J., Maneveldt, G., Halling, C. and Probyn, T., 2006. Abalone farming in South Africa: an overview with perspectives on kelp resources, abalone feed, potential for on-farm seaweed production and socio-economic importance. Aquaculture 257: 266- 281.

Walker, B.K., 2012. Spatial analyses of benthic habitats to define coral reef ecosystem regions and potential biogeographic boundaries along a latitudinal gradient. PLoS One 7.

Wing, S.R., Jack, L.C. and Fujita, Y., 2015. Overthrowing a regime shift: displacement of sea urchin by abalone in a kelp forest ecosystem. Ecosphere 6:1-13.

Witte, A. D., 2017. Benthic algal communities of shallow reefs in the Eastern Cape: availability of abalone habitat. Nelson Mandela Metropolitan University. Pp 1- 183.

Wolanski, E., Fabricius, K., Spagnol, S., and Brinkman, R., 2005. Fine sediment budget on an inner-shelf coral fringed island, Great Barrier Reef of Australia. Estuarine, Coastal and Shelf Science 65, 153-158.

Won, N.I., Kawamura, T., Onitsuka, T., Hayakawa, J., Watanabe, S., Horii, T., Takami, H. and Watanabe, Y., 2007. Community and trophic structures of

abalone Haliotis diversicolor habitat in Sagami Bay, Japan. Fisheries science 73, 1123-1136.

Wood, A., 1993. Aspects of the Biology and Ecology of the South African Abalone Haliotis midae Linnaeus, 1758 (Mollusca: Gastropoda) along the Eastern Cape and Ciskei coast. Rhodes. Pp 1-168.

Wood, A.D., and Buxton, C.D., 1996. Aspects of the biology of the abalone Haliotis midae (Linne, 1758) on the east coast of South Africa. 1. Feeding biology. South African Journal of Marine Science 17, 61-68.

Zeeman, Z., Branch, G.M., Farrell, D., Maneveldt, G.W., Robertson-Andersson, D., and Pillay, D., 2013. Comparing community structure on shells of the abalone Haliotis midae and adjacent rock: implications for biodiversity. Marine Biology 160, 107-117.

Zeeman, Z., Branch, G.M., Pillay, D., and Jarre, A., 2014. An experimental test of the effect of the abalone Haliotis midae on benthic community composition. Marine Biology 161, 329-337.

Appendices Appendix 1. Table of GPS coordinates of sites used for the baseline study.

Site Depth Zone Latitude Longitude Site 1 Zone 1 34° 2'40.79"S 25°38'11.19"E Zone 2 34° 2'41.73"S 25°38'11.96"E Zone 3 34° 2'42.40"S 25°38'12.53"E Site 2 Zone 1 34° 2'47.85"S 25°37'23.69"E Zone 2 34° 2'49.03"S 25°37'21.53"E Zone 3 34° 2'51.12"S 25°37'21.86"E Site 3 Zone 1 34° 2'53.87"S 25°35'35.04"E Zone 2 34° 2'55.21"S 25°35'37.11"E Zone 3 34° 2'56.17"S 25°35'38.40"E Site 4 Zone 1 34° 2'29.70"S 25°32'25.09"E Zone 2 34° 2'30.68"S 25°32'28.45"E Zone 3 34° 2'31.24"S 25°32'29.37"E

Appendix 2. Table of GPS coordinates for the seeded vs unseeded sites.

Site Transect Latitude Longitude Site 5 (Unseeded) NA 34° 2'50.05"S 25°37'21.22"E Site 6 (Seeded) 1A 34° 2'26.52"S 25°38'26.16"E 2A 34° 2'27.60"S 25°38'24.00"E 3A 34° 2'29.40"S 25°38'24.00"E

Appendix 3. Species list of the coastal section sampled from Noordhoek Ski Boat Club to Schoenmakerskop, Port Elizabeth.

Algae Amphiroa anceps (Lamarck) Decaisne Amphiroa ephedraea (Lamarck) Decaisne Arthrocardia carinata (Kützing) Johansen Arthrocardia flabellata (Kützing) Manza Calliblepharis fimbriata (Greville) Kützing Caulerpa filiformis (Suhr) Hering Champia compressa Harvey Codium lucasii subsp. capense P.C.Silva Codium stephensiae C.I.Dickinson Delisea flaccida (Suhr) Papenfuss Dictyota dichotoma var. intricate (C.Agardh) Schmidt Dictyota naevosa (Suhr) Montagne Gelidium abbottiorum R.E.Norris Gelidium pteridifolium R.E.Norris, Hommersand & Fredericq Gigartina insignis (Endlicher & Diesing) F.Schmitz Halimeda cuneata Hering Heydrichia woelkerlingii R.A.Townsend, Y.M.Chamberlain & Keats Hypnea viridis Papenfuss Jania adhaerens J.V.Lamouroux Jania verrucosa J.V.Lamouroux Laurencia flexuosa Kützing Laurencia natalensis Kylin Mesophyllum sp. Plocamium beckerii F.Schmitz ex Simons Plocamium corallorhiza (Turner) J.D.Hooker & Harvey Plocamium suhrii Kützing Ralfsia expansa (J.Agardh) J.Agardh Sargassum elegans Suhr

Invertebrate Taxa Ascidiaceae Bryozoa Haliotis midae (Linnaeus, 1758) Hydrozoa Marthasterias glacialis (Linnaeus, 1758) Oxystele sp. (Philippi, 1847) Parechinus angulosus (Leske, 1778) Callopatiria granifera (Gray, 1847) Porifera Pyura stolonifera (Heller, 1878) Scutellastra sp. (H. Adams & A. Adams, 1854) Trematocarpus sp. (Kützing, 1843) Turbo sarmaticus (Linnaeus, 1758)

Appendix 4. Actual dive counts of Haliotis midae (Abalone)

Figure 20 Illustrates the abalone counts for each site and at the three depth zones. There is a notable trend that with increasing depth there is an increase in abalone abundance. Depth zone 2 shows the most variation regarding abalone counts and notices presence from this depth zone for sites 2; 3; and 4. Site 3 abalone numbers remain low in depth zone 3. While site 4 shows no counts as site 1 and site 2 numbers increase.

30 Site 1 25

20 Site 2

15 AbaloneCounts Site 3 10

5 Site 4

0 Zone 1 Zone 2 Zone 3 Depth Zones

Figure 20. Abalone counts within each depth zone across the four sites.

Figure 21 illustrates the abalone counts for each of the two sites (seeded vs unseeded). It is noticed that the seeded site (site 6) has higher abalone abundance compared to that of the unseeded site (site 5).

10 Transect 1 8 Transect 2

AbaloneCounts 6 Transect 3

0 Unseeded Seeded Sites

Figure 21. Abalone counts between the seeded (Site 6) and unseeded (Site 5) sites.

Appendix 5. R-script for Canonical Correspondence Analysis.

##Canonical Correspondence Analysis

#Load required libraries library(vegan)

#Set the working directory setwd("C:/Users/Lauren/Desktop/Final Results/Analysis/csv/FinalFinal")

#Import the data from a CSV file CCAdata <- read.csv("LMall1.csv", row.names=1) CCAcodes <- read.csv("LMcodes1.csv", row.names=1) CCAenv <- read.csv("LMenv1.csv", row.names=1)

#Log transform the data CCAdata1 <- log1p(CCAdata)

##Run CCA with stats Cord <- cca(CCAdata1 ~ Boulder+Rock+Gravel+Shell.Grit+Sand+Depth, data=CCAenv) Cord

#Determine the significance of the constraining variables anova(Cord, by="term", permutation=499)

#Determine the significance of the axes anova(Cord, by="axis", permutations=499) windows()

#Plot the CCA plot <- plot(Cord) plot(Cord, display="sites") plot(Cord, display="species")

#Plot only the arrows ordiplot(Cord, display="sites",type="n") text(Cord, scaling = 3, display = "bp", col="red")

#Plot colour-coded sites ordiplot(Cord, display="sites",type="n")

#Set grouping Environmental Variable & colours & plot points Class <- as.numeric(CCAcodes$Zone)

#Assign as many colours as there are grouping variables colvec <- c("chartreuse", "darkolivegreen1", "darkgreen", "chocolate", "chocolate1", "darkgoldenrod1", "deeppink", "coral3", "lightpink", "cadetblue", "aquamarine1", "azure3", "cornflowerblue")

points(Cord, display="sites", pch=16, cex=0.5, col=colvec[Class]) text(Cord, scaling = 6, display = "bp", col="red")

#Add Legend #topright with(CCAcodes, legend("topright", legend=levels(Zone), bty="n",col=colvec, pch=21, pt.bg=colvec, adj = c(0.1, 0.6), cex=0.7))

#Choose the points to label on the sites graph #Run the next line to choose the sites by clicking on the dots in the plot #Press escape when all the points of interest are selected identify(plot,"sites", cex=0.4)

#Plot a CCA hull-diagram showing domains #Run up to line 46 ordihull(Cord, groups=CCAcodes$Tsct, lty="dotted") #Or with elipses ordiellipse(Cord,groups=CCAcodes$Tsct, draw="polygon") #ordiellipse(Cord,groups=CCAcodes$Tsct, col="", draw="polygon", label=T)

#Do a species plot ordiplot(Cord, display="species",type="n") points(Cord, display="species", pch=16, cex=1.2) text(Cord, scaling = 3, display = "bp", col="red")

#Add the species names (all on top of each other) text(Cord, display="species", cex=0.8, col="blue")

#Or choose which points to label #Run the next line to choose the sites by clicking on the dots in the plot #Press escape when all the points of interest are selected identify(plot,"species")

Appendix 6. R-script for the Bray Curtis classification.

# Set the working directory #setwd("C:/Users/Lauren/Documents/Rscripts/Data") setwd("C:/Users/Lauren/Desktop/Final Results/Analysis/csv/FinalFinal")

# Import the data from CSV files - choose ONE of the following rows #Data should be formatted so that each row is a quadrat spe <- read.csv("LMallav1.csv", row.names=1) spe2 <- log1p(spe) env<- read.csv("LMenv1.csv", row.names=1) env2<- log1p(env) sites<- read.csv("LMcodes1.csv", row.names=1) library(vegan) mat = spe2 d = (1 - vegdist(mat, method="bray")) * 100 h = hclust(d) windows() plot(h, main = "Clustering samples using Bray Curtis method", sub = "", xlab="", cex=0.20 , ylab="Bray Curtis Similarity", axes = FALSE, hang = -1) lines(x = c(0,0), y = c(0,100), type = "n") # force extension of y axis axis(side = 2, at = seq(0,100,10), labels = seq(100,0,-10)) rect.hclust(h, h=80, border= c("orange", "red", "deeppink", "purple", "blue","green"))

Appendix 7. R-script for statistical Kruskal-Wallace and post-hoc Dunn test for multiple comparison (non-parametric data)

# Set the working directory #setwd("C:/Users/Lauren/Documents/Rscripts/Data") setwd("C:/Users/Lauren/Desktop/Final Results/Analysis/csv/FinalFinal")

# Import the data from CSV files - choose ONE of the following rows #Data should be formatted so that each row is a quadrat my_data <- read.csv("LMenvANOVA.csv", row.names=1) my_data2 <- read.csv("LMANOVA.csv", row.names=1)

###################################################################

##Shapiro-Wilks normaility test ## If p<0.05 then the data are not normally distributed and a ##non-parametric approach should be used ##the first part inthe bracket is the parameter to be tested ##the second part is the level e.g. Dzone or Site. attach(my_data) tapply(Rock, Zone, boxplot) tapply(Rock, Zone, shapiro.test)

###################################################################

####NON-PARAMETRIC DATA #Kruskall-Wallace Anova for non-parametric data kruskal.test(Rock ~ Site, data = my_data) kruskal.test(Halimeda.cuneata ~ Site, data = my_data2)

###################################################################

##Dunn test for multiple comparison post-hoc (non-parametric data) #to look for post-hoc differences between factors library(FSA)

###ENVIRO Res.Dunn.Test = dunnTest(Sand ~ Dzone, data=my_data, method="bh")

Res.Dunn.Test ####ALGAE Res.Dunn.Test = dunnTest(Upright.Coraline ~ Site, data=my_data2, method="bh") Res.Dunn.Test

Res.Dunn.Test = dunnTest(Upright.Coraline ~ Dzone, data=my_data2, method="bh") Res.Dunn.Test

Appendix 8. R-script for correlations

#Set the working directory setwd("C:/Users/Lauren/Desktop/Final Results/Analysis/csv/FinalFinal")

# Import the data from CSV files - choose ONE of the following rows #Data should be formatted so that each row is a quadrat my_data <- read.csv("LMcor.csv", row.names=1)

install.packages("ggpubr") library("ggpubr")

##This is the correlation plot ## It also inclues the correlation coeficient (R) and the confidence limits (p) ggscatter(my_data, x = "Depth", y = "Plocamium.sp.", add = "reg.line", conf.int = TRUE, cor.coef = TRUE, cor.method = "pearson", xlab = "Depth (m)", ylab = "Abundance (%)")

##This is just the correlation cor.test(Depth, Plocamium.sp., data = my_data, conf.level = 0.95)