(Porites Astreoides) in the Nearshore Waters of Grenada

Spatial Distribution of Benthic Habitats and Ecological Patterns of the Mustard Hill Coral (Porites astreoides) in the Nearshore Waters of Grenada

by Ryan Eagleson

A Thesis Presented to The University of Guelph

In partial fulfilment of requirements For the degree of Master of Science in Pathobiology

Guelph, Ontario, Canada © Ryan Eagleson, May, 2019

ABSTRACT

SPATIAL DISTRIBUTION OF BENTHIC HABITATS AND ECOLOGICAL PATTERNS OF THE MUSTARD HILL CORAL (PORITES ASTREOIDES) IN THE NEARSHORE WATERS OF GRENADA

Ryan Eagleson Advisor: University of Guelph, 2019 Dr. John Lumsden

The objectives of this research were: 1) create a benthic map of the Sandy Island-

Oyster Bed MPA (SIOBMPA); and 2) study populations of Porites astreoides (pa) in the

Caribbean/Grenada. A benthic map was created using 127 truthing points and eCognition software. Mapped habitats in the SIOBMPA were: dense seagrass

(25.74%); sand (23.32%); coral framework (5.88%); and others. Compiled survey data in the Caribbean revealed regional success of pa. Temporal changes in size, abundance and coverage and the relationship between benthic components, MPAs, and river outflows on pa coverage in Grenada was assessed. The abundance of pa fell

(p<0.05), and colony size/coverage increased (p<0.05). Coverage of pa was negatively correlated with MPA status, rubble, sand, macroalgae, gorgonians, and weedy corals and positively with pavement, coralline algae, and stress-tolerant/generalist corals. This study for the first time has documented the distribution of benthic habitats in the

SIOBMPA and population dynamics of pa.

ACKNOWLEDGEMENTS

First and foremost, I would like to thank my advisor Dr. John S. Lumsden for giving me the opportunity to study under him as a graduate student, an undergraduate student, and a summer employee since. Without this experience I would not be where I am in my career today. I would like to thank my advisory committee members, Dr. Lorenzo-

Alvarez Filip and Dr. Steven Schill for their expertise, flexibility and support these past years.

I would also like to thank my parents, Kathy and Peter, and my grandparents, June, Bill,

Gordon, and Francis for their love and support to follow my passions throughout my life and university career. I would like to thank Owen Harvey, Adam Maue, Kaileigh Watson,

Kiri Hardy, Tony Hoffman, Peter Macpherson, Christopher Pierce, Patrick Nantel, Diane

Blanchard, Francine Mercier, Hossein Karimi, and Abigail Bartlett for their friendship and support during my studies.

I would also like to thank the members of the Lumsden lab that I’ve worked with these past years: Zachary Millar, Dr. Ryan Horricks, Dr. Juan-Ting Liu, Dr. Ehab Misk, Paige

Vroom, Paul Huber, Dr. Elena Contador, Dr. Maureen Jarau. Mykolas Kamaitis, Sam

Renshaw, and Drayke Evans. Dr. Ryan Horricks was particularly helpful in the field, providing survey data, and advising me in statistical analyses.

Lastly, my years at the University of Guelph have been the finest in my life and the institution has and always will be my number one choice.

III

DECLARATION OF WORK PERFORMED

All of the work in this thesis was performed by myself, with the following exceptions:

1) The object-ID based benthic mapping process was conducted by Dr. Steven

Schill and his team at The Nature Conservancy using data points collected by

this study (Miami, Florida).

2) The Nature Conservancy collected all bathymetric and imagery data.

3) The dive surveys and photography of the transects in Grenada & Carriacou were

conducted by Dr. Ryan Horricks and Dr. John Lumsden (University of Guelph).

4) The species coverage data excluding the mustard hill coral used in Chapter 3

(obtained using Coral Point Count analysis) was performed by Dr. Ryan Horricks

at the University of Guelph (Guelph, Ontario).

CONTENTS

ACKNOWLEDGEMENTS ...... III LIST OF TABLES ...... VII LIST OF FIGURES ...... VIII LIST OF ABBREVIATIONS...... XIII CHAPTER 1: REVIEW OF THE LITERATURE ...... 1 1.1 INTRODUCTION ...... 1 1.2 THE MUSTARD HILL CORAL ...... 3 1.2.1 DISTRIBUTION ...... 3 1.2.2 TAXONOMY...... 4 1.2.3 BIOLOGY ...... 7 1.2.4 ECOLOGY ...... 14 1.2.5 HARDINESS & RESILIENCE ...... 18 1.3 MARINE PROTECTED AREAS ...... 22 1.4 GRENADIAN REEF COMMUNITIES ...... 26 1.5 RATIONALE ...... 33 1.6 HYPOTHESES AND OBJECTIVES ...... 34 1.6.1 HYPOTHESES: ...... 34 1.6.2 OBJECTIVES: ...... 34 LITERATURE CITED ...... 36 CHAPTER 2: HIGH RESOLUTION BENTHIC HABITAT MAPPING WITHIN THE SANDY ISLAND-OYSTER BED MPA (GRENADA) ...... 54 2.1 ABSTRACT ...... 54 2.2 INTRODUCTION ...... 55 2.3 MATERIALS AND METHODS ...... 59 2.3.1 RESEARCH SITE ...... 59 2.3.2 FIELD SAMPLING ...... 60 2.3.3 BENTHIC ANALYSIS ...... 61

2.3.4 BENTHIC HABITAT MAPPING ...... 63 2.4 RESULTS ...... 65 2.4.1 GROUND TRUTHING ...... 65 2.4.2 BENTHIC HABITAT MAP ...... 67 2.5 DISCUSSION ...... 70 2.6 ETHICS ...... 77 2.7 ACKNOWLEDGEMENTS ...... 77 2.8 LITERATURE CITED ...... 79 CHAPTER 3: THE MUSTARD HILL CORAL IN THE CARIBBEAN: LOCAL FACTORS INFLUENCING POPULATION DYNAMICS IN THE NEARSHORE WATERS OF GRENADA ...... 87 3.1 ABSTRACT ...... 87 3.2 INTRODUCTION ...... 88 3.3 METHODOLOGY ...... 91 3.3.1 DATA COLLECTION ...... 91 3.3.2 STATISTICAL ANALYSIS ...... 92 3.3.3 EXPERIMENTAL DESIGN ...... 93 3.3.4 DATA COLLECTION ...... 94 3.3.5 QUANTITATIVE MEASUREMENTS ...... 95 3.3.6 STATISTICAL ANALYSIS ...... 96 3.8 LITERATURE CITED ...... 121 GENERAL CONCLUSIONS AND DISCUSSION ...... 124 APPENDICES ...... 128 APPENDIX A ...... 128 APPENDIX B ...... 128

LIST OF TABLES

Table 1.1: Summary of differences between brooding and broadcast spawning strategies in corals (Goodbody-Gringley & de Putron 2016; Serrano et al. 2016). …………………………………………………………………………………………………...11

Table 1.2: Summary of the benefits and costs of a weedy life-history strategy in corals (Darling et al. 2012, 2013; Madin et al. 2016). …………………………………………………………………………………………………...15

Table 1.3: Predicted impacts of forecasted ocean acidification on the mustard hill coral this century at two pCO2 scenarios; 560 µatm (mid-century) and 800 µatm (late century). This is based on forecasts from IPCC 2007. ……………………………………....…………………………………………………………..22

Table 1.4 Summary of reef health in Grenada as assessed by The Nature Conservancy. Adapted from (Grenada’s Coral Reef Report Card 2016). ……………………………………....…………………………………………………………..28

Table 2.1: Benthic habitat categories used to categorize communities in Carriacou and example photographs used for habitat Identification. …………………………………………………………………………………………………..62

Table 2.2: Accuracy of modelled benthic habitat map based on benthic truthing points within the SIOBMPA. Habitats not identified in benthic truthing within the SIOBMPA are shaded black.

…………………………………………………………………………………………………..69

Table 2.3 Proportion of benthic habitats in Carriacou protected by the SIOBMPA. …………………………………………………………………………………………………..75

VII

Table 3.1: Summary of mustard hill coral survey locations and years for which survey data was available. …………………………………………………………………………………………………..92

Table 3.2: Summary of all factors measured in quadrat surveys for possible inclusion in the correlation analyses. …………………………………………………………………………………………………..98

Table 3.3: Magnitude of the relationship between mustard hill coral coverage and time at seven locations in the Caribbean (Spearman correlation). Statistical significance is indicated with an asterisk. …………………………………………………………………………………………………..99

Table 3.4: Magnitude of the relationship between mustard hill coral coverage and: 1) habitat type, 2) biotic factors, and 3) MPA status (Pearson correlations; Spearman correlation for MPA status). Significant correlations are indicated with an asterisk. ………………………………………………………………………………………………….106

Table 3.5: Magnitude of the relationship between mustard hill coral coverage and the presence of river outflows. Lower coverage was associated with rivers in 2014, and higher in 2017 (Spearman correlation; p>0.05). ………………………………………………………………………………………………….107 Table 3.6: Summarized outcome from repeated correlation analysis of mustard hill coral coverage with other benthic factors. Factors found in all four analyses are indicated with asterisks. Factors with inconsistent statistically significant correlations are shaded black. …………………………………………………………………………………………………113

VIII

LIST OF FIGURES

Figure 1.1: Global distribution of the mustard hill coral (Porites astreoides) (IUCN 2000a). ……………………………………………………………………………………………………4 Figure 1.2: A generalised taxonomic summary of the coral sublass Hexacorallia. Clade location of Poritidae is indicated with asterisks (Kitahara et al. 2010). ……………………………………………………………………………………………………6 Figure 1.3: A generalised phylogenetic tree based on mitochondrial sequence data. Porites astreoides is found within Poritidae (indicated with a black bar) (Kitahara et al. 2010). ……………………………………………………………………………………………………6 Figure 1.4: A green morph colony of Porites astreoides the mustard hill coral (left) and a magnified image of its feeding polyps, with a single polyp indicated (right). ……………………………………………………………………………………………………8 Figure 1.5: A summary of the taxonomic diversity and composition of microbiota found on samples of the mustard hill coral collected in Bocas Del Toro, Panama. The most abundant bacterial taxa found is indicated. Adapted from: (Sunagawa et al. 2010). ……………………………………………………………………………………………………9 Figure 1.6: Representative Demonstration of the 'Phoenix Effect' in Porites, arrow direction shows progression of time at the initial disturbed state (red) and after regeneration (green). Adapted from (Roff et al. 2014). …………………………………………………………………………………………………...20 Figure 1.7: Community composition within the Moliniere-Beausejour MPA (Grenada) (Molinière-Beauséjour Marine Protected Area Management Plan” 2010). …………………………………………………………………………………………………...29 Figure 1.8: Relative abundance of stony coral species groups at Jack Adam Island and Saline Island, Grenada (1976) (Goodwin et al. 1976). …………………………………………………………………………………………………...31 Figure 2.1: Shallow reef habitat (7 m in depth) near Hillsborough, Grenada. …………………………………………………………………………………………………...56

Figure 2.2: Four marine protected areas are established to date in Grenada: A) Sandy Island-Oyster Bed B) Molinier-Beasejour C) Grand Anse Bay and D) Wobourne Clarke. …………………………………………………………………………………………………...57

Figure 2.3: Location of study site within the wider Caribbean region and within the nation of Grenada (see site location in Carriacou; inset). …………………………………………………………………………………………………...59

Figure 2.4: Boundary of the Sandy Island-Oyster Bed MPA (Carriacou, Grenada). …………………………………………………………………………………………………...60

Figure 2.5: Ground truthing locations within the Sandy Island-Oyster Bed MPA, with habitat classifications. …………………………………………………………………………………………………...66

Figure 2.6: Relative proportion of benthic habitat classes for 123 ground truthing observations within the SIOBMPA. …………………………………………………………………………………………………...67

Figure 2.7: Benthic habitat map of the Sandy Island-Oyster Bed MPA (The Nature Conservancy 2017). Deep water areas are those where light does not adequately penetrate through the water column (due to depth) to allow the spectral analysis of benthic habitats. …………………………………………………………………………………………………...68

Figure 2.8: Proportion of the SIOBMPA protected sea area occupied by each of nine benthic habitat classes. …………………………………………………………………………………………………...70

Figure 3.1: A group of mustard hill coral colonies inhabiting fields of Acropora rubble (left) and a typical juvenile green morph colony (right). Both photographs were taken at a depth of 3 m.

…………………………………………………………………………………………………...89

Figure 3.2: Location of survey sites in Grenada and Carriacou including MPA boundaries. …………………………………………………………………………………………………...94

Figure 3.3: Example quadrat photo with mustard hill coral colony indicated by circle. …………………………………………………………………………………………………...96

Figure 3.4: Relationship of mustard hill coral coverage and across seven sites in the Caribbean (moving averages with dotted lines). ………………………………………………………………………………………………….100

Figure 3.5: Observed changes in mean transect coverage of the mustard hill coral from 2014 to 2017 in Carriacou and Grenada. ………………………………………………………………………………………………….101

Figure 3.6: Observed changes in mean colony size of the mustard hill coral from 2014 to 2017 in Carriacou and Grenada. ………………………………………………………………………………………………….102

Figure 3.7: Observed changes in mean colony abundance of the mustard hill coral from 2014 to 2017 in Carriacou and Grenada. ………………………………………………………………………………………………….103

Figure 3.8: Changes in size frequency distributions of mustard hill coral colonies in Carriacou (top) and Grenada (bottom) from 2014 (blue) to 2017 (red). ………………………………………………………………………………………………….104

Figure 3.9: Scleractinian species whose coverage is significantly correlated with the mustard hill coral: a) Colpophyllia natans, b) Diploria strigosa, c) Madracis auretenra, d) Porites divaricata, e) Porites porites, and f) Orbicella franksi. ………………………………………………………………………………………………….117

XII

LIST OF ABBREVIATIONS

ANOVA-Analysis of Variance

BMUB- German Federal Ministry for the Environment, Nature Conservation, Building and Nuclear Safety DOC- Dissolved Organic Carbon

DRAG-Dragon Bay Survey Site (Grenada)

FLAM- Flamingo Bay Survey Site (Grenada)

IUCN- International Union for the Conservation of Nature

JAD- Jack Adam Island Survey Site (Carriacou)

MHC- Mustard Hill Coral

MPA- Marine Protected Area

NGO- Non-governmental Organization

NORTH- Northern Abundance Survey Site (Grenada)

OGS- Ontario Graduate Scholarship

PVC- Polyvinyl Chloride

QUART- Quarter Wreck Survey Site (Grenada)

RHI- Reef Health Index

RS- River Status SAND- Sandy Island Survey Site (Carriacou)

SEAV- Seaview Survey Site (Carriacou)

SIOB- Sandy Island-Oyster Bed

SIOBMPA- Sandy Island-Oyster Bed Marine Protected Area

TNC- The Nature Conservancy

USVI- United States Virgin Islands

UVR- Ultraviolet Radiation

WHIRL- Whirlpool Survey Site (Carriacou)

XIII

CHAPTER 1: REVIEW OF THE LITERATURE

1.1 INTRODUCTION

Coral reefs are an essential part of the global marine ecosystem and provide an array of ecological services as well as support complex and diverse species assemblages. Covering only 0.2% of the total area, coral reefs are the most diverse habitat per unit area in the worlds oceans, supporting an estimated 1-3 million species

(McIntyre 2010). They are nurseries for more than a quarter of the oceans fish and nearly 1 billion people worldwide rely on these fisheries as a food source (Wilkinson

2008). Healthy reefs also contribute to tourism based economies in many developing states and are a significant source of pharmaceutical compounds (Wilkinson 2008).

Marine protected areas (MPAs) are one of the primary tools used worldwide to protect areas containing significant marine resources and diversity (Perera-Valderrama et al.

2016a). MPAs have been shown to increase the resilience of coral reef communities and decrease disease prevalence (Lamb et al. 2015).

For these reasons governments across the Caribbean have been establishing networks of MPAs (Grenada Coral Reef Report Card 2016). The tri-island state of

Grenada is in the process of establishing a network of MPAs with the goal of protecting

25% of its; land area and nearshore oceanic environment. To date, four MPAs have been established (3 in Grenada, 1 in Carriacou) with eleven more at some stage of establishment. To establish a functioning network of MPAs, sufficient benthic habitat information is necessary to assess habitat representativeness and levels of connectivity included in present and future MPAs, as well as to direct the management of those

already established (Barriteau 2011;Selig & Bruno 2010). It was determined that the most suitable way to provide this information to the Government of Grenada and the relevant conservation managers, was the creation of benthic habitat maps, particularly of the unmapped SIOBMPA in Carriacou, and other unmapped areas of Grenada and

Carriacou.

Scleractinians (stony corals) produce calcareous skeletons from calcium carbonate that form corallites around the coral polyp. This process is largely responsible for the construction of the physical structure of the reef itself (Veron 2013). Due to differences in life history strategy, habitat preference, symbiont types, competitive interactions or yet to be determined processes some species of Scleractinian coral are more resilient to local community collapse following physical disturbances and long-term environmental stressors (Perry et al. 2015). Scleractinian corals have adopted different life history strategies for survival that have had varying degrees of success in our changing oceans. These have been categorized as: weedy; stress tolerant; generalist; and competitive (Darling et al. 2013). The mustard hill coral (MHC) is a ‘weedy’

Scleractinian coral present throughout the Caribbean, the Eastern coast of Brazil, and

West Africa (IUCN 2000). Weedy corals are typically defined as species with a brooding mode of reproduction, that are comparatively small, and that rapidly colonize disturbed areas of reef (Baumann et al. 2016). In the last half century, the MHC has become one of the most abundant coral species in the region and is one of the few reef building corals that have not experienced significant declines in the Caribbean (Perry et al.

2015). This species is often more abundant in disturbed areas of reef that have been impacted by high temperature events, storm damage, and/or anthropogenic impacts

(Perry et al. 2015). This study will focus on this species of stony coral due to its life history, high abundance and ease of identification in the Caribbean region. The reefs of the Eastern Caribbean and Grenada have been understudied compared to areas such as the Bahamas, Florida, or the ABC islands (Aruba, Bonaire, and Curacao) and for this reason Grenada was selected as a field site in this study. The research objectives of this thesis are: 1) to create a benthic habitat map of the Sandy Island-

Oyster Bed MPA in Grenada; 2) to provide a temporal update on the population status of the MHC in the Caribbean; and finally 3) to assess the relative impact of a gradient of biotic and abiotic factors on population dynamics of the MHC in Grenada and the factors influencing its relative success.

1.2 THE MUSTARD HILL CORAL

1.2.1 DISTRIBUTION

Porites astreoides is found throughout the Caribbean, as well as along the coasts of Florida, Bermuda, Brazil, and West Africa (Figure 1.1). The size of its range is approximately 6.1 million km2, however it is unclear to what extent gene flow takes place between populations separated by great distances (Serrano et al. 2016; Madin

2017). This zooxanthellae-dependant species is most abundant within sheltered lagoons, shallow fore reef environments, and on the coral reefs themselves (IUCN

2000). The MHC is a shallow water coral that commonly inhabits depths of 0.5 m - 15 m but can be found as deep as 70 metres (IUCN 2000). The deepest recorded observation of this species was at a depth of 210 m and it is unclear what conditions allowed these individuals to live in this low-light environment (Madin 2017). Colonies

can typically be found on hard rocky substrate or rubble and are largely absent from areas of sandy bottom (Goreau 1959). The MHC is common throughout its range and is one of the few Scleractinian species whose populations are increasing in the face of climate change (IUCN 2000). Within its range, this species favours clear and calm water but can also tolerate a wide range of thermal and physical stressors (Perry et al. 2015;

Baumann et al. 2016) (Figure 1.1).

Figure 1.1 Global distribution of the mustard hill coral (Porites astreoides) (IUCN 2000a).

1.2.2 TAXONOMY Kingdom: Animalia Phylum: Cnidaria Class: Anthozoa Subclass: Hexacorallia Order: Scleractinia Family: Poritidae Genus: Porites

Most corals species have a high level of variation taxonomically, making the establishment of accurate taxonomic trees difficult (Nagelkerken & Bak 1998). Historical measures of taxonomic relatedness based on features like growth, skeletal form, or other physical parameters have largely given way to modern genomic techniques

(Nagelkerken & Bak 1998). This has resulted in the shifting of many cladogenic relationships between species over the last several years and decades (Powers et al.

1972). Tropical corals are members of the subclass Hexicorallia, with three orders being present: Scleractinia, Alcyonacea, and Corallimorpharia (Veron 2013). Members of the order Scleractinia have been present globally for more than 250 million years as indicated by the fossil record (Romano & Cairns 2000; Stanley 2003). Presently, more than 1400 species of Scleractinians’ have been identified across 27 different families, with far more research needed on species inhabiting greater depths (Kitahara et al.

2010). The MHC is a Scleractinian coral that was first categorized along with many

Caribbean coral species by Jean-Baptiste Lamarck in 1816 (Powers et al. 1972). It has been present in the fossil record for approximately 9.9 million years (Madin 2017). Like many groups of tropical corals today the phylogenetic position of the MHC can fluctuate based on new genetic information and its relationships with other Caribbean species are still unclear (Kitahara et al. 2010). At present no subspecies of the MHC have been identified with bar-coding analysis (Shearer & Coffroth 2008). However, two distinct colour morphs exist, green and brown, but have yet to diverge enough to be considered separate subspecies (Nagelkerken & Bak 1998). A summary of the cladogenic position of Porites within Hexicorallia can be seen in Figure 1.2 below (Kitahara et al. 2010).

Octocorallia Zoanthidae Actiniaria

Corallimorpharian Scleractinia (complex clade)

Scleractinia (complex clade) **

Scleractinia (robust clade)

Figure 1.2 A generalised taxonomic summary of the coral subclass Hexacorallia. Clade location of Poritidae is indicated with asterisks (Kitahara et al. 2010).

The phylogenetic ancestral relationships between the MHC and closely related

Scleractinian species can be seen in Figure 1.3.

Figure 1.3 A generalised phylogenetic tree based on mitochondrial sequence data. Porites astreoides is found within Poritidae (indicated with a black bar) (Kitahara et al. 2010).

1.2.3 BIOLOGY

The MHC is a shallow-water, colonial Scleractinian coral (Figure 1.4). The growth rate of the MHC is comparable to other Scleractinian corals, with a single colony reaching maturity at approximately 8-10 years of age (Madin et al. 2016). Estimates for

MHC skeleton extension are 3.67 +/- 0.65 mm/year; increases in skeletal density occur at 1.49 +/- 0.16 g/cm3/year and the rate of calcification is 0.55 +/- 0.12 g/cm/year

(Elizalde-Rendón et al. 2010). Growth in colonies is largely focused at the extension of the colony edges with limited growth at the apex, similar to what has been observed in

Porites sp. inhabiting the Pacific Ocean (Elizalde-Rendón et al. 2010). Colonies of the

MHC have differing growth rates across its large range that vary based on environmental conditions. For instance, colonies in the Gulf of Mexico have reduced growth rates with varied seasonal timing compared to the rest of the Caribbean

(Elizalde-Rendón et al. 2010).

The form of growth for the MHC is typically either plate or massive (Madin et al.

2016). It typically ranges in size from 4 cm2 – 400 cm2, with the most frequent size of adults being approximately 250 cm2 (Madin et al. 2016). MHC populations are dominated by medium (41 cm-80 cm in diameter) and large sized colonies (≥81 cm in diameter), and colony success is typically high, making it very unlikely for large colonies to reduce in size once the have reached this growth stage (Edmunds 2010). As a member of the family Poritidae, the MHC contains many small corallites filled with septa

(IUCN 2000). Typical polyp density of this species is 18 polyps/cm2, with each corallite approximately 1.2-1.6 mm in diameter (Figure 1.4) (Madin et al. 2016). As previously stated, two colour morphs are present in this species (green & brown), with the green

morph typically being found in shallow water environments and brown in the deeper portions of its range (Gleason 1993).

6 cm 6 mm

Figure 1.4 A green morph colony of Porites astreoides the mustard hill coral (left) and a magnified image of its feeding polyps, with a single polyp indicated (right).

Many tropical coral species also provide a habitat to an array of zooxanthellae and other diverse microbiota communities on their surface and within their intracellular holobiont (Sunagawa et al. 2010). The composition and diversity of these microorganisms in the Caribbean is largely understudied, but initial observations have revealed very high surface diversity on Scleractinian tissue (Sunagawa et al. 2010). It is unclear what the varying composition of different microbiota communities means for individual coral species (Sunagawa et al. 2010). Like all shallow water corals, the MHC has a symbiotic relationship with photosynthetic zooxanthellae that provide much of the corals energy needs (Hauff et al. 2016).

Zooxanthellae clade composition can vary within MHC colonies and populations due to seasonal changes in gradients of environmental stress, and different habitats

(i.e. offshore vs. coastal) (Kenkel et al. 2013a; Hauff et al. 2016). Mustard hill coral colonies can contain numerous clades of zooxanthellae (i.e. A, B, and C) with A and A4 being particularly abundant in the northern Caribbean and A abundant throughout the

Caribbean (Serrano et al. 2016). Smaller colonies of MHC have been found to contain more tissue protein and higher zooxanthellae densities after bleaching as a result of acidification compared with larger colonies (Hall et al. 2015).

The taxonomic composition of the MHC’s microbiota has been found to contain a high proportion of gammaproteobacteria compared to closely related species (Figure

1.5) (Sunagawa et al. 2010). A number of environmental and ecological factors have been found to alter the taxonomic composition in the surface microbiota of corals (Vega

Thurber et al. 2012; Meyer et al. 2014). For instance, bleaching, physical creation of lesions, and macroalgal exposure have all been found to alter the microbiota in the

MHC (Vega Thurber et al. 2012; Meyer et al. 2014).

Mustard Hill Coral

Bacterial Taxa Gammaproteobacteria Alphaproteobacteria Bacteroides Cyanobacteria Unclassified Firmicutes OP11 Other Figure 1.5: A summary of the taxonomic diversity and composition of microbiota found on samples of the mustard hill coral collected in Bocas Del Toro, Panama. The most abundant bacterial taxa found is indicated. Adapted from: (Sunagawa et al. 2010).

The surface and mucus of the MHC provides important habitat for zooxanthellae and complex communities of microbiota (Rodriguez-Lanetty et al. 2013). The bacterial species Oceanospirillaceae seems to be present in all individuals and may also play some sort of symbiotic role in the organism (Glasl et al. 2016). The MHC also produces mucus for protection from ultraviolet radiation, physical damage, and to facilitate the capture of prey (Madin et al. 2016). Mucus is routinely produced and shed by the colony, with the microbial community of older mucus (mucus present at the outer layer) shifting to include more pathogenic microbes (Glasl et al. 2016). After the sloughing and replacement of older mucus, the microbial community returns to its original state (Glasl et al. 2016). The communities living within the mucus seen to be essential to the health of the colony; one study found that alteration or removal of these microbial communities resulted in colony bleaching and death (Glasl et al. 2016). The mucus of corals has also been found to have toxic qualities that may facilitate the capture of prey organisms

(Beckmann et al. 2012). The mucus of the MHC was fatal to crickets and induced vasoconstriction, however it is considered unlikely to have any profound effect on humans (Beckmann et al. 2012).

The MHC has a brooding reproductive strategy as opposed to broadcast spawning (Table 1.1), and is capable of sexual or asexual (fragmentation) reproduction

(Goodbody-Gringley & de Putron 2016). Parthenogenesis has also been recently documented in the MHC, aiding in successful reproduction in a variety of conditions

(Vollmer 2018). However, this also acts to limits genetic diversity in populations and can isolate them (Vollmer 2018). Approximately 47% of Scleractinian coral species in the

Caribbean share this mode of reproduction, with the remaining 53% being broadcast

spawners (Darling et al. 2012). Brooding species have been increasing in the Caribbean while many coral communities continue to collapse, perhaps mirroring the coral extinction event of the early Miocene, where brooding coral species experienced far higher survival rates (Table 1.1) (Edinger & Risk 1995; Darling et al. 2013; Perry et al.

2015). Although brooding corals are increasing in prominence regionally, not all areas experience these shifts in community reproductive strategy as rapidly. Spawning corals are slightly more abundant than brooding corals in the Eastern Caribbean (12% vs 10%)

(Williams et al. 2016).

Table 1.1 Summary of differences between brooding and broadcast spawning strategies in corals (Goodbody-Gringley & de Putron 2016; Serrano et al. 2016).

Coral Reproductive Strategy Brooding Broadcast Can be continuous Annual or biannual event Larvae uptake zooxanthellae directly from the Larvae uptake zooxanthellae from the parent colony water column Internal fertilization (larvae retained by colony External fertilization (larvae formed in and released) water column) Rapid larval settlement Long-range dispersal, delayed settlement High larval success rate Low larval success rate

In one study, half of MHC populations were hermaphroditic with the other half being female only, while no exclusively male colonies have been found to date

(Chornesky and Peters 1987). Inbreeding in the MHC through self-fertilization can occur at rates of 50% in populations (Chornesky & Peters 1987). Reproductive status of colonies in a population can vary with season, lunar day, polyp location/age, and colony size/age (Chornesky and Peters 1987). Individual colonies that are hermaphroditic contain a roughly equal mix of male and female polyps and, as mentioned, are capable of self-fertilization (Chornesky & Peters 1987). Gonads are most abundant at the

oldest/central regions of the colony (Goodbody-Gringley & de Putron 2016).

Spermatozoa released by one polyp are taken up by a female polyp which fertilizes its egg within (Goodbody-Gringley & de Putron 2016). Embryos are then stored within the coral for several days or weeks, until they reach a mature planula larva stage

(Goodbody-Gringley & de Putron 2016). At this stage the parent colony transfers its microbial community and zooxanthellae parent to the larval offspring (Sharp et al. 2012;

Neave et al. 2016; Serrano et al. 2016).

Like many Scleractinian species, reproduction of the MHC is a carefully timed event synchronized with the temperature of the water and the lunar phase. Typically, the closer a population of this species is to the equator, the longer their period of reproduction will last due to the relative stability of water temperatures year round (de

Putron & Smith 2011). For instance, the reproductive period for the MHC in Bermuda is only 2-3 months and in Florida it is 4-6 months, while in Bonaire it is continuous year round (Goodbody-Gringley & de Putron 2016). Release of larvae still peaks around the new moon across its range and occurs at sea temperatures of 25.6-27.5o C (McGuire

1998). The higher the latitude, e.g. Bermuda, the more pronounced climatic seasonality becomes, narrowing the time in which conditions are available for the MHC to reproduce (de Putron & Smith 2011). As temperature increases, reproductive effort and the number of planulae has been found to decrease, as the colony shifts how its energy budget is managed (de Putron & Smith 2011). Mustard hill coral colonies are equally successful in the production of larvae at depth compared with shallow waters, with any differences in the total being a result of abundance of reproductive colonies (Holstein et al. 2016). Long distance dispersal of larvae is typical of similar Caribbean species and is

largely dependent on prevailing ocean currents. For instance, gene flow is strong between MHC colonies in Florida and USVI but more limited between Florida and

Bermuda (Serrano et al. 2016). Upon release to the environment, larvae from the MHC can metamorphose into polyps within 10 hours (Goodbody-Gringley & de Putron 2016).

Although brooding corals such as the MHC have larger larvae than broadcast spawning species, size can vary across different habitats in their range (White 2014). A positive correlation exists between the size of MHC colony and the number of planulae larvae that are released (McGuire 1998).

The proper settlement of coral colonies is essential for the continuation of viable populations. Without sufficient larval settling, coral recruits will not reach adulthood to replace adult colony mortality (Olsen et al. 2016). Like most coral species, MHC larvae settle on coral rubble or hard rocky substrate in the shallow reef environment, and are impacted by algal coverage, competition and sedimentation (Olsen et al. 2016). Mustard hill coral larvae are also able to detect the intensity of ultraviolet radiation and have settlement preferences for areas of reef with a lower level of UVR such as partially sheltered overhangs (Gleason et al. 2006). It has been observed that the larvae of the

MHC are drawn to settle in areas with higher reef noise, which is the cumulative sound of the reef ecosystem including herbivore foraging, scavenger actions etc., and are often indicative of a healthy reef environment (Lillis et al. 2018). This larval preference for ‘louder’ reef areas took place regardless of light conditions (Lillis et al. 2018). It has been found that while branching species such as Acropora palmata do not settle on macroalgae, MHC larvae can but the chances of the survival of those larvae to adulthood are very limited (Ross et al. 2013; Olsen et al. 2016). Larval survival and

settlement success in the MHC is dependant on site-specific conditions (Goodbody-

Gringley et al. 2018). The larvae of colonies inhabiting upper mesophotic habitats have increased survival and settlement success (Goodbody-Gringley et al. 2018). Unlike broadcast spawning corals (i.e., A. cervicornis) that often have their larvae settle on coralline algae, the MHC settles largely on biofilms, particularly Titanoderma prototypum and Hydrolithon boergesenii and have a higher rate of post-settlement survival (Ritson-

Williams et al. 2016). This is particularly important as many coral communities in the

Caribbean have been shifting from coral or coralline dominated states to one that is dominated by macroalgae and biofilms. In the water column itself, increased water temperatures (31o C) have been found to have no effect on the on the survival of released mustard hill larvae, their ability to properly settle on the substrate, or properly undergo metamorphosis to an adult stage (Ross et al. 2013). Long-term stress events, such as incidences of extreme rainfall and storm events, can however impact the ability of larvae in the water column to settle on the substrate (Edmunds 2016).

1.2.4 ECOLOGY

As previously mentioned, the life-history strategy of the MHC is classified as

‘weedy’ in nature (Darling et al. 2012). Weedy corals are typically defined as a species with a brooding mode of reproduction, that are small and short-lived, that rapidly colonize disturbed areas of reef, and that recovery rapidly from stressors (Table 1.2)

(Baumann et al. 2016).

Table 1.2 Summary of the benefits and costs of a weedy life-history strategy in corals (Darling et al. 2012, 2013; Madin et al. 2016).

Weedy Life History Strategy Pro Con Rapid colonization Less optimal dispersal potential Resistance to wave action Shaded out by faster growing species Tolerance to a wide range of environmental Lack of colony resources for combatting conditions larger species Higher larvae success (brooding) Recovers rapidly from disturbance events

Peak densities of the MHC occur at depths of 10 and 35 metres, similar to

Porites species in the Pacific (Edmunds & Leichter 2016; Holstein et al. 2016).

Populations of the MHC at the lower reaches of its depth range provide 10% of the recruitment provided by those corals at shallower depths. Deep MHC colonies serve as a population source and future refuge in the face of climate change (Holstein et al.

2016). Across the Caribbean, live Scleractinian coral cover has been relatively stable following massive losses in the 1980s. The Caribbean-wide average of live coral cover from 2001-2005 is 16%, down from 50% in the 1970s (Schutte et al. 2010). However

16-72% of this cover is comprised solely of the MHC (Edmunds 2010). The increase in non-framework building coral species such as the MHC has resulted in a decrease in reef rugosity, net reef calcification/growth rates, and a loss of available habitat for fishes

(Alvarez-Filip et al. 2011, 2013). Therefore, despite the individual success of several non-framework building species regionally, they will be unable to ensure continued ecosystem functionality without the preservation of key framework species such as A. palmata (that do contribute significantly to reef rugosity, calcification, and fish habitat)

(Alvarez-Filip et al. 2013)

The MHC is commonly found alongside other currently abundant species in the

Caribbean such as: Sidastrea radians, Orbicella annularis, Madracis sp., and Agaricia agaricites (Miller et al. 2016; Williams et al. 2016). To date, it is one of the few species in the Caribbean with a positive rate of population growth (i.e. USVI 1997-2007)

(Edmunds 2010). The proportion of the MHC in reef communities of the Caribbean has been increasing at a rate of 1.5%/year; rising from <20% in 1975 to more than 50% of communities in 2004 (Edmunds 2010). This may not continue since the MHC is unable to return to its pre-bleaching density in a season between annual bleaching events.

Therefore, in areas that begin to experience annual bleaching events, the continued success of the MHC may be less likely (Schoepf et al. 2015).

Mustard hill populations are dominated by medium (23%) and large (20%) colonies, with 62% of colonies being less than 50 cm2 in size (Edmunds 2010). Very low mortality occurs in large colonies once individuals reach this size class (Edmunds

2010). The populations of MHCs regionally in the Caribbean are quite fragmented, with the loss of adult colonies particularly impacting regional success of the species and population connectivity (Holstein et al. 2014). For this reason, self-recruitment rather than larvae flow from other regions is the predominant method of growth for MHC populations (Holstein et al. 2014)

As a short-lived species that quickly reproduces, the MHC rapidly colonizes areas of recently disturbed reefs (Edmunds 2010). In the Pacific Ocean this behaviour in Porites sp. is termed ‘the Phoenix Effect’, in which surviving colony fragments rapidly regenerate and overgrow dead skeleton following a disturbance/bleaching event (Roff et al. 2014). Following 15 years of regeneration after the massive el-Nino event in the

Pacific in 1997/98, Porites colony coverage increased from 34.9 to 73.9%, and dead colony coverage went from 42.8% to 0% (Roff et al. 2014). Even with this fast growth, the MHC has low reef building potential, but currently accounts for 68% of total carbonate production in the Caribbean (Perry et al. 2015b). As the proportion of non- framework building coral species increases in the Caribbean, total net carbon production and reef rugosity will continue to decline (Alvarez-Filip et al. 2013; Perry et al. 2015a). In addition, calcification rates in the MHC and other common species such as Orbicella annularis, and Montastrea sp. are expected to decline in response to increasing ocean temperatures and ocean acidification (Figure 1.6) (Okazaki et al.

2016). Greater coral bleaching in areas with significantly higher levels of offshore runoff is typical across the Caribbean (Ennis et al. 2016). The MHC is often able to persist in these environments and therefore its abundance is frequently used to differentiate high stress sites with those that have more optimal conditions (Ennis et al. 2016). Colonies already thriving in high stress environments may prove to be more resilient to the forecasted impacts of climate change (Barranco et al. 2016).

As previously stated, due to their short-lived life history, MHC populations are largely driven by recruitment. Reef herbivores are essential to juvenile survival, keeping areas clear of macroalgae that would otherwise smother colonies and decrease the growth potential of MHC recruits (Burkepile et al. 2010; Vega Thurber et al. 2012). Not all herbivores impact MHC colony success equally; ocean surgeonfish have been found to increase growth of P. astreoides and P. porites by 2-3x, with redband parrotfish having no impact. Diadema antillarum is also a significant grazer in the community and have largely filled this niche after collapses in Caribbean fisheries (occurring as early as

1945 in areas such as Jamaica), however Diadema antillarum populations then subsequently also collapsed in the 1980’s (Hardt n.d.; Nimrod et al. 2017). The abundance of MHCs has been found to be positively correlated with higher abundances of these urchins, but their grazing is more damaging to young recruits than that of herbivorous fishes (Mumby et al. 2007; Ritson-Williams et al. 2016). Brooding corals such as the MHC and Agaricia sp. are often the most abundant following disturbance events; they were found to compose 30-80% of juveniles in Tobago after a large scale bleaching event in 2010 (Buglass et al. 2016). The establishment of MPAs increases the success and abundance of MHC recruits, with this species, Agaricia sp., and

Montastrea sp. responsible for much of the reef recovery following protection (Mumby &

Harborne 2010). Due to high levels of recruitment of the MHC following disturbances, a majority of colonies following protection were found to be typically small (Mumby &

Harborne 2010).

1.2.5 HARDINESS & RESILIENCE

For coral species to be successful in our rapidly changing oceans, they must be able to regenerate to their original state following stressors (resilience) and endure a wide range of environmental conditions (hardiness) (Alvarez-Filip et al. 2013). The MHC is among the hardiest and most resilient in the Caribbean, being one of the few coral species to regionally increase in population over the last several decades so that it now comprises a significant portion of total coral cover (Perry et al. 2015).

Colony morphology of the MHC provides the ability to resist damage from physical storm and wave conditions (IUCN 2000). In weedy Scleractinian species (i.e.

Siderastrea radians) the size and circularity (non-branching physiology) reduces the

susceptibility of colony breakage from physical disturbance compared to large species with a complex branching physiology such as A. palmata (Porto-Hannes et al. 2015).

The MHC has a very similar growth form to S. radians and its low lying encrusting colonies can be found in a wide range of current and storm impacted environments

(Madin et al. 2016).

Energy budgets within individual coral colonies are believed to play a major role in tolerance to chronic environmental stressors and to determine the rates of growth and recovery in colonies (Kenkel et al. 2013). This has been shown to be hereditary among certain populations of MHC distributed across habitats (Kenkel et al. 2015). Shallow water forms have been observed to be more tolerant to thermal stress as compared to those found in deeper water on the reef face (Castillo et al. 2012). Smaller colonies are more hardy and resilient to bleaching stresses, and these colonies have been found to have more protein and a higher zooxanthellae density after bleaching (Hall et al. 2015).

This plays a role in the ‘Phoenix Effect’ where damaged Porites colonies can rapidly regenerate from small surviving fragments after large-scale bleaching events (Figure

1.6), (Roff et al. 2014). Mustard hill corals can be adapted to their respective local environments, with shallower corals (more stable zooxanthellae composition) faring better following transplantation to novel environmental conditions than offshore reef colonies (more dynamic zooxanthellae communities), or shallow-dwelling green colour morphs (adapted for UVR resistance) (Gleason 1993; Hauff et al. 2016; Kenkel & Matz

2016). Even within these habitats the differences between individual colonies plays a role in thermal tolerance, with corals less than 10 km apart exhibiting significant differences in response to bleaching events.

Figure 1.6 Representative demonstration of the 'Phoenix Effect' in Porites (French Polynesia) from 1998 to 2013. Arrow direction shows progression of time at the initial disturbed state (red) and after regeneration (green). Adapted from (Roff et al. 2014).

Mustard hill corals are not resilient to all changes and are sensitive to thermal stress, however this does not seem to have reduced its success in the region. The species strong recruitment and reproductive success may reduce the impact of any thermal sensitivity on a population level (Manzello et al. 2015). Following single bleaching events, the MHC is able to make up for the resultant reduced carbon budget

(due to zooxanthellae loss) with polyp uptake of dissolved organic carbon (DOC) and zooplankton (Levas et al. 2016). A study conducted by Levas et al. (2016) found DOC uptake following a single bleaching event accounted for 11-36% of the carbon budget for sampled colonies. However, following repeat bleaching events, this was insufficient to prevent carbon budget reductions in the MHC, showing that there are limits in the resilience of colonies to repeated stressors (Levas et al. 2016).

A number of other factors related to life-history and reproduction in the MHC contribute to its high level of resilience. The MHCs ability to rapidly colonize habitats

made available by disturbance events is a major factor in its success, with its reproductive ability playing a key role. As a brooding coral, successful fertilization occurs at much higher rates for MHC colonies, compared to broadcasting species releasing their gametes into the water column (Goodbody-Gringley & de Putron 2016).

Brooding corals also typically produce fewer, and much larger larvae than broadcasting species, and these have been found to have increased resistance to rising CO2 conditions (White 2014). The forecasted long-term responses to projected ocean acidification this century in MHC reproduction can be seen in Table 1.3 (Albright &

Langdon 2011). In southern areas of the region, reproduction takes place year-round, peaking at every new moon, allowing the population of colonies to expand continually in damaged areas (Chornesky & Peters 1987). The MHC is also able to successfully reproduce in a wide range of abiotic and biotic conditions (Chornesky & Peters 1987).

Increasing depth does not appear to impact the successful production of planula larvae by P. astreoides, and this occurs well into the mesophotic zone (Holstein et al. 2016).

Planulae larvae of the MHC are resistant to temperature stresses (such as those during bleaching events), but not when cyanobacteria levels are elevated (Ritson-Williams et al. 2016). Cyanobacteria release cyanotoxins that impact many organisms when present at high abundances in seawater. The MHC has been shown to be particularly effective at both vertical and horizontal geneflow in reef environments despite its brooding form of reproduction (Serrano et al. 2016). This is significant as populations in the deeper, cooler mesophotic zone can act as a source of population rescue for shallower populations and are able to use this environment as a refugia (Serrano et al.

2016). The MHC is able to transfer its zooxanthellae and bacterial communities to its

larvae, which may play a role in in its resistance (Sharp et al. 2012). Chronic stress levels hinder the ability of mustard hill larvae to properly select and settle on adequate substrate (Edmunds 2016). Mustard hill corals have been found to settle on macroalgae unlike species such as A. palamata, with colonies unlikely to survive (Olsen et al. 2016). However, this may be offset by the fact that MHCs settles on biofilms and has higher post settlement survival than broadcasting species, which may make up for its unsuccessful settling on macroalgae (Ritson-Williams et al. 2016).

Table 1.3 Predicted impacts of forecasted ocean acidification on the mustard hill coral this century at two pCO2 scenarios; 560 µatm (mid-century) and 800 µatm (late century). This is based on forecasts from IPCC 2007 (Albright & Langdon 2011).

1.3 MARINE PROTECTED AREAS

The establishment of Marine Protected Areas (MPAs) currently ranks highly among the most widely used conservation tools to increase the resilience of tropical reef communities (Selig et al. 2012; Bruno et al. 2018). Marine conservation initiatives have focused heavily on coral reef ecosystems due to the high levels of risk to their continued survival, high biodiversity, and the essential ecosystem services that they provide (i.e. economic opportunities, storm protection, and discovery of pharmaceutical compounds)

(Schuhmann & Mahon 2015). Marine protected areas also have many direct benefits to their establishment such as: public education, research opportunities, economic opportunities, improved diversity/environmental health, and maintenance of ecosystem

services (Angulo-Valdés & Hatcher 2010). Approximately 60% of coral reefs globally are at-risk due to local anthropogenic impacts (Bridge et al. 2013). Forecasted climate impacts make a return to previous species configurations of tropical coral reefs impossible, with conservation efforts being needed to focus on preserving ecosystem functionality (Hughes et al. 2017). Marine Protected Areas have largely focused on shallow coral reefs, and it has become more apparent that protected area networks must be expanded to include whole reef ecosystems including important future refuges and larval sources such as the mesophotic zone (Bridge et al. 2013). An estimated 77% of coral reefs in the Caribbean alone extend to the mesophotic zone (Bridge et al. 2013;

Gress et al. 2018). Marine protected areas not including these habitat areas, puts them at significant risk to anthropogenic stressors such as commercial fishing, which indirectly impacts ecosystem health within protected areas (Bridge et al. 2013; Gress et al. 2018). Regional-scale networks of MPAs in the Caribbean are currently insufficient to protect the long-term sustainability of marine ecosystems facing a rapidly changing ocean (Kaplan et al. 2015). In the Caribbean, MPAs are often small in size and have inadequate penalties/enforcement (Selig et al. 2012). Any network of MPAs must take into account the connectivity of ecological communities (particularly in corals given their long-range larval dispersal patterns) to increase resilience and improve management outcomes (Schill et al. 2015). The ecological performance of individual MPAs in the

Caribbean has also been low, although the network is rapidly expanding (Dalton et al.

2012). A lack of local technical skill is also often a challenge in many small Caribbean island states (Pomeroy et al. 2005).

A challenge facing many MPAs globally is their inability to ensure long-term community stability with the impacts of climate change. Even effectively managed MPAs sheltering reef communities from local anthropogenic impacts may not be enough to prevent further losses as a result of changing regional climate conditions, especially if there is low connectivity between these areas and other significant ecological communities in the region (Lagabrielle et al. 2014; Selig et al. 2012). More aggressive action to reduce the impacts of global climate change must be coupled with expanding protected area networks (Selig et al. 2012; Graham & McClanahan 2013). Regional scale stressors such as increasing storm frequency/intensity and water temperatures have proven to cause continual biodiversity loss and reef decline even with the presence of effectively managed MPAs (Huntington et al. 2011). The positive impact of

MPA presence on fish abundance has long been known but the impact on coral coverage and populations is more uncertain (Selig & Bruno 2010; Huntington et al.

2011; Morgan et al. 2016). A global review of MPAs has shown that they alone will be inadequate at protecting marine communities from the impacts of climate change, under a ‘business as usual’ emissions scenario. Under this scenario there will still be species losses in the tropics and low latitudes even in well-enforced no-take reserves. Sea surface temperatures would be expected to increase 2.8 C by 2100 and environmental conditions will exceed natural variation by 2050 in 42% of 309 studied no-take reserves

(Bruno et al. 2018). Marine protected areas that would otherwise be effective, will have their positive impacts dramatically reduced as a result (Bruno et al. 2018). Additional local management actions will be needed (i.e. coral farming and transplanting) when protection alone is not enough to overcome the long-term stress conditions of climate

change (Noble et al. 2013). When undertaking monitoring initiatives to assess MPA success, sufficient consideration for adequate spatial and temporal extent of these programs rarely takes place. Monitoring programs should cover at least two generations of target species (i.e. reef fishes) and non-MPA sites should be a minimum of two larvae dispersal units away for analysis. Assessing MPA success on a short timescale is inadequate (Moffitt et al. 2013). Structural changes in reef habitats in spatially mapped and monitored reefs can provide insights into benefits and drawbacks of MPAs and inform adaptive management strategies (Noble et al. 2013). Older MPAs have been shown to be more effective and can stabilize coral community losses, while those in unprotected areas or unenforced MPAs have continued to decline (Selig & Bruno 2010).

Not all MPAs are created equal. Factors such as: size, distance from adjacent protected areas, age, location, regulations, species generation time, larval dispersal, enforcement, and the presence or absence of management plans influence the relative effectiveness of these areas (Agardy et al. 2011; McClanahan & Karnauskas 2011).

Without considering these factors, confidence in a MPA network to halt biodiversity declines is misplaced, with poorly planned or managed MPAs often little more than a

‘paper park’ (Pomeroy et al. 2005; Agardy et al. 2011). MPA planning must also take into account changing ecological communities, the communities chances of regeneration, and their representation of local habitat types (McClanahan & Karnauskas

2011). More incentives for local citizens/stakeholders to participate, and adequate penalties/enforcement are key for a successful ecological outcome in any MPA (Dalton et al. 2012; Kaplan et al. 2015). Effective MPA management also requires accurate and continuous information and feedback to inform adaptive planning (Pomeroy et al. 2005).

Disturbance is an important part of building community resilience and increasing biodiversity in an area (Wilson et al. 2012). Reefs previously exposed to disturbance events such as large-scale bleaching or storms, often have higher resilience to these impacts in the future, while a minority of MPAs protect these damaged areas (Selig et al. 2012). No-take MPAs generally perform better, and increases in herbivorous fishes can also improve coral recovery as live coral coverage has a strong negative correlation with macroalgae (Wilson et al. 2012). Some MPAs have been successful at increasing benthic coral coverage but it is essential that fishing pressures on herbivorous fishes is relieved (Mumby & Harborne 2010). Population bottleneck effects of severely depleted coral populations (i.e. Acropora palmata) can severely hamper reef recovery following protection (Mumby & Harborne 2010). Only well enforced, no-take, old MPAs are effective at allowing coral recovery to take place, particularly large colony building species (Strain et al. 2019).

1.4 GRENADIAN REEF COMMUNITIES

Between the years 1975 and 2004, coral reef cover in the Caribbean declined

80%, from an average live coverage of 50% in the 1970s, to the 10% cover that we see in many areas today (Gardner et al. 2003). The reef conditions of Grenada are no exception and are typical of those found across the Eastern Caribbean, with low hard coral coverage dominated by a small number of species, low or absent Diadema antillarum populations, and a high proportion of macroalgal coverage (37% in 2014 and

17% in 2017 in Grenada; Horricks et al. 2018). Total cover in Grenada declined from

50% to 20% between 1975 and 2010 with a 14% loss taking place from 2005-2009

(Jackson et al. 2014). The most recent assessment by Horricks et al. (2018) found

mean live coral cover in the tri-island state of Grenada to be 21% (+/- 6.2%) in 2014 and

17.45% (+/-5.67%) in 2017. Historically, reefs in Grenada have been understudied compared to other regions in the Caribbean, with study numbers rising significantly in the early 2000’s. Grenada has a total of 78 km2 of reef habitat (both patch and fringing) including 33 identified species of stony coral from 10 families and 29 square kilometres of seagrasses (Burke & Maidens 2004; Molinière-Beauséjour Marine Protected Area

Management Plan 2010; Grenada’s Coral Reef Report Card 2016). These reefs have experienced the impacts of at least 12 hurricanes since the 1970s and numerous bleaching events (Grenada’s Coral Reef Report Card 2016). As a result of their degrading conditions the Government of Grenada has pledged to protect 25% of its marine ecosystems (Byrne 2005). To date, four MPAs have been created in the nearshore waters of Grenada: Sandy Island-Oyster Bed (7.87 km2), Moliniere-

Beasejour (0.8 km2), Grand Anse Bay (13.4 km2), and Woburn-Clark’s Court Bay (4.2 km2) (Figure 2.2). Eleven more proposed areas forming a total of 197 km2 (~10% of

Grenada’s nearshore marine area) have also been proposed (Byrne 2005; Turner et al.

2009). It is particularly essential that these areas are established and properly enforced given that Grenada has a high dependence on these systems that have a very low adaptive capacity to climate change (Burke et al. 2010). Impacts such as boat anchoring, overfishing, development, storms and runoff are significantly damaging

Grenada’s reefs, compounded with the forecasted impacts of climate change, including a decline in calcification of 10% and annual bleaching events by 2040 (Burke & Maidens

2004).

The Nature Conservancy, with additional funding from the German Federal

Ministry for the Environment, Nature Conservation, Building and Nuclear Safety

(BMUB), has produced six reef report cards in the Eastern Caribbean compiling data from 277 reef surveys and maps in detail of 383 km2 of reef. These reports provide an initial ecosystem health baseline and identify knowledge gaps in each nation using metrics established by CaribNode and the Healthy Reefs Initiative. The reef health index (RHI) is an integration of four parameters measured by the Nature Conservancy: coral cover, fleshy macroalgae, herbivorous fish, and commercial fish (Table 1.4; http://www.caribnode.org/; http://www.healthyreefs.org/cms/). Reef health in Grenada has been assessed by the Government of Grenada and The Nature Conservancy as being in a ‘poor’ condition with a RHI ranking of 2.5/5 (Table 1.4) (Grenada's Coral

Reef Report Card 2016).

Table 1.4 Reef health in Grenada as assessed by The Nature Conservancy. Adapted from (Grenada’s Coral Reef Report Card 2016).

The current stony coral coverage on the main island of Grenada is 14%, down from an estimated 31% in 2007, and relatively stable with the estimated coverage of

13% in 2008 (Burke & Maidens 2004; Wilkinson 2008b; Anderson et al. 2014). Levels of macroalgae have been rising across Grenada, with a measured increase from 41% benthic coverage in 2009 to 74.2% in 2011 (Anderson et al. 2014). Sea urchin numbers have plummeted on the island from an estimated 13/m2 in 1975 to 0.001-0.046/m2 in

2010 (Anderson et al. 2014; Jackson et al. 2014). The island of Carriacou’s live coral coverage has remained relatively stable at 20% in 2015 compared against an estimate of 19% in 2005, with 7 coral recruits/m2 (Coral Reefs of Carriacou Island 2005; Perry et al. 2015b). Monitoring of MPAs in Grenada from 2005-2009 found an increase in stony coral coverage, with a substantial loss in gorgonians, reef complex, and living tissue coverage (Figure 8) (Molinière-Beauséjour Marine Protected Area Management Plan

2010).

Figure 1.7: Community composition within the Moliniere-Beausejour MPA (Grenada) (Molinière-Beauséjour Marine Protected Area Management Plan 2010).

Mirroring other areas of the Caribbean, Grenada has been experiencing shifts in species assemblages in its reefs over the last several decades. A trend towards non- framework building corals has been occurring in Grenada and 73% of observed sites across the Caribbean (Perry et al. 2015b). For instance, the MHC in Grenada has increased from an absolute coverage of 3.1% in 1976 to 11.1% in 2007, with its proportion of total coral coverage increasing from 10.6% to 26.5%, respectively (Green et al. 2008). A recent analysis of reef communities in Grenada has found that 34.1%-

52.3% of corals surveyed are branching species (Anderson et al. 2014). Species once prevalent in Grenada such as A. palmata and A. cavernosa in the late 1970s have largely vanished from the reefs of Grenada (some scattered pockets remain), with weedy species such as the MHC rising in abundance (Figure 1.8). On the island of

Grenada today, the most common coral species by proportion of live coral coverage are

P. porites, P. astreoides, O. annularis, and M. auratenra (Wilkinson 2008a; Horricks et al. 2017).

Shifts in the benthic composition of non-coral species has also been occurring on the reefs of Grenada. Native seagrass species have been displaced in large areas by an invasive seagrass species Halophila stipulacea creating more habitat for inshore invertebrates like sponges and urchins, with a corresponding increase in food availability for fishes (Scheibling et al. 2017). Sea urchins such as Diadema antillarum are now largely absent from reef communities in Grenada at abundances as low as

0.001 individuals/m2 (Anderson et al. 2014). Particularly high abundances of Diadema antillarum have been noted in isolated pockets in Grenada and Carriacou such as the

‘Seaview’ site (Horricks et al. 2017; The Nature Conservancy 2017). Large numbers of

Tripneustes ventricosus have also been observed in seagrass beds (Schill, pers. comm.

2018). Due to the loss of urchin and herbivorous fish populations, abundance of

Dictyota and Halimeda macroalgae species have been on the rise and are the now the dominant form of benthic cover in Grenada (Wilkinson 2008a).

Figure 1.8 Relative abundance of stony coral species on the Southern shore of Jack Adam Island, Grenada in 1976 (left) and 2015 (right) (Goodwin et al. 1976; Horricks et al. 2017).

The spatial distribution and extent of benthic habitats have been mapped on the main island of Grenada by TNC. Fringing and patch reefs cover much of its Eastern coast and the Southwest with relatively few large areas of rubble/destroyed reef.

Benthic habitat mapping is a widely used and tremendously effective tool at identifying

the spatial distribution and extent of various marine environments (Xu & Zhao 2014).

They are of particular use in shallow tropical environments given the extremely clear water column, and that many of the coral species of interest are in the upper photic zone (excepting mesophotic reefs) (Xu & Zhao 2014). Benthic habitat mapping allows users such as conservation managers to rapidly map marine habitats using minimal resources, compared to decades past where extensive dive surveys or camera tows would have been required to map these areas with high confidence (Nature

Conservancy 2017; Porskamp et al. 2018). Having the ability to rapidly and reliably map benthic habitats allows for repeated mapping and long-term monitoring of reef environments at a regional scale (Awak et al. 2016). In addition, these maps can be used to inform management decisions (e.g. MPA zoning), support MPA establishment

(e.g. ensuring network connectivity), guide reef restoration (e.g. outplanting farmed corals to mapped rubble areas), and guide field research (e.g. site selection) (Stevens &

Connolly 2005; Nature Conservancy 2017).

Dramatic improvements in the benthic habitat mapping of tropical coral reefs have been made in the last several decades (Xu & Zhao 2014; Lecours et al. 2015).

Trends in the field have been largely moving towards increased quality in satellite imagery, and therefore increased ability to conduct spectral analyses (Hedley et al.

2018). Presently, many studies use the widely available Worldview-2 satellite, which provides imagery at a resolution of ~1.85 m per pixel (Wahidin et al. 2015). As higher resolution imagery has become available, the identification of individual coral colonies

(e.g. Orbicella sp.) is becoming possible. Sunlight glint corrections on the water surface in satellite imagery have also dramatically improved, further improving map resolution

(Anggoro et al. 2016). In addition, improvements to acoustic methods for bathymetric data collection, improved spectral analyses, and the automatic classification of benthic habitat points are key areas of improvement in the field (Lecours et al. 2015; Porskamp et al. 2018). As the technology available for benthic habitat mapping continues to improve, the confidence in mapping products increases, requiring fewer and fewer benthic truthing points.

1.5 RATIONALE

The conservation of tropical coral reef ecosystems is essential due to their immense levels of biodiversity and ecosystem services that they provide (National

Oceanic and Atmospheric Administration 2009). It is imperative to understand the ability of coral reef ecosystems to adapt to global climatic changes to focus management strategies, and to establish networks of effective protected areas. This study has selected the MHC as it is a species of increasing abundance in the wider Caribbean region (Perry et al. 2015a). As many Porites species including the MHC are considered

‘weedy’ species it is among the first to colonize an area recently disturbed by large scale bleaching or storm events (Roff et al. 2014).

Accurate benthic habitat maps of coral reef coverage and species composition is an important tool in the analysis of community assemblage changes over time in response to anthropogenic impacts or global climate change. These maps are used to monitor species composition in response to nearby land uses or impacts and to enable the rapid identification of stressors in the local coral reef environment. Linking ecological science with tools for conservation will become more imperative as the world’s oceans continue to face deleterious anthropogenic impacts including climate change.

1.6 HYPOTHESES AND OBJECTIVES

1.6.1 HYPOTHESES:

1.6.1.1 CHAPTER 2

• Benthic ground truthing surveys, together with satellite imagery and bathymetric

data, can be used to map the extent and distribution of benthic habitats in the

Sandy Island-Oyster Bed MPA with high accuracy.

1.6.1.2 CHAPTER 3

• Mustard hill coral coverage in the Caribbean has continued to increase since its

last regional assessment in 2004, given its tolerance to a wide range of

environmental conditions.

• Based on changes in other stony coral populations in Grenada observed by

Horricks et al. 2017, significant changes in mustard hill coral coverage,

abundance, and size frequency will occur between survey years in Grenada and

Carriacou.

• Given that the localized distribution of many corals depends on interspecific

interactions and larval settlement, the variation in mustard hill coral coverage

may be explained by a suite of local biotic and abiotic factors.

1.6.2 OBJECTIVES:

1.6.2.2 CHAPTER 2

• Obtain benthic truthing points covering a wide spatial extent of the MPA

boundary, and the benthic habitats that lie within.

• Categorize benthic truthing points by their respective habitat class in survey

photos.

• Use benthic truthing points, combined with pre-obtained satellite

imagery/bathymetric data for spectral analysis (conducted by TNC), to create a

benthic habitat map of relative coverage and spatial distribution of benthic habitat

classes in the Sandy Island-Oyster Bed MPA.

1.6.2.1 CHAPTER 3

• Obtain transect information (post-2000) from a minimum of 5 locations in the

wider Caribbean region with comparable survey methods, less than 15m depth,

and in a nearshore reef environment.

• Use already established 30 m transects (4 per site) at each of eight sites in

Grenada and Carriacou surveyed in 2014 and 2017.

• Measure the abundance and size of all mustard hill corals present in pre-

established transects in Grenada using ImageJ software.

• Assess changes in mustard hill coral coverage, abundance and size frequency

after two survey years in Grenada/Carriacou.

1.7 LITERATURE CITED *References in format of ‘Ecology Letters’

Agardy, T., di Sciara, G.N. & Christie, P. (2011). Mind the gap: Addressing the

shortcomings of marine protected areas through large scale marine spatial

planning. Mar. Policy, 35, 226–232.

Albright, R. & Langdon, C. (2011). Ocean acidification impacts multiple early life history

processes of the Caribbean coral Porites astreoides. Glob. Chang. Biol., 17, 2478–

2487.

Alvarez-Filip, L., Carricart-Ganivet, J.P., Horta-Puga, G. & Iglesias-Prieto, R. (2013).

Shifts in coral-assemblage composition do not ensure persistence of reef

functionality. Sci. Rep., 3, 3486.

Alvarez-Filip, L., Dulvy, N.K., Côté, I.M., Watkinson, A.R. & Gill, J.A. (2011). Coral

identity underpins architectural complexity on Caribbean reefs. Ecol. Appl., 21,

2223–2231.

Anderson, R., Morrall, C., Jossart, J., Nimrod, S., Bolda, E., Musser, K., et al. (2014).

Marine protected area monitoring in the nearshore waters of Grenada, Eastern

Caribbean: benthic cover and fish populations. Rev. Biol. Trop. (Int. J. Trop. Biol,

62, 273–286.

Anggoro, A., Siregar, V.P. & Agus, S.B. (2016). The Effect of Sunglint on Benthic

Habitats Mapping in Pari Island Using Worldview-2 Imagery. Procedia Environ.

Sci., 33, 487–495.

Angulo-Valdés, J.A. & Hatcher, B.G. (2010). A new typology of benefits derived from

marine protected areas. Mar. Policy, 34, 635–644.

Awak, D.K., Gaol, J.L., Subhan, B., Madduppa, H.H. & Arafat, D. (2016). Coral Reef

Ecosystem Monitoring Using Remote Sensing Data: Case Study in Owi Island,

Biak, Papua. Procedia Environ. Sci., 33, 600–606.

Barranco, L.M., Carriquiry, J.D., Rodríguez-Zaragoza, F.A., Cupul-Magaña, A.L.,

Villaescusa, J.A. & Calderón-Aguilera, L.E. (2016). Spatiotemporal variations of live

coral cover in the northern Mesoamerican Reef System, Yucatan Peninsula,

Mexico. Mex. Sci. Mar., 80, 143–150.

Barriteau, M. (2011). The Grenadines Network of Marine Protected Areas.

Transboundary cooperation worksop.

Baumann, J.H., Townsend, J.E., Courtney, T.A., Aichelman, H.E., Davies, S.W., Lima,

F.P., et al. (2016). Temperature regimes impact coral assemblages along

environmental gradients on lagoonal reefs in Belize. PlosOne.

Beckmann, A., Özbek, S., Ibarra-Alvarado, C., Lazcano-Pérez, F., Arreguín-Espinosa,

R., Sánchez-Rodríguez, J., et al. (2012). The Nematocyst: a molecular map of the

Cnidarian stinging organelle. Int. J. Dev. Biol., 56, 577–582.

Bonin, M.C. (2012). Specializing on vulnerable habitat: Acropora selectivity among

damselfish recruits and the risk of bleaching-induced habitat loss. Coral Reefs, 31,

287–297.

Bridge, T.C.L., Hughes, T.P., Guinotte, J.M. & Bongaerts, P. (2013). Call to protect all

coral reefs. Nat. Clim. Chang., 3, 528–530.

Bruno, J.F., Bates, A.E., Cacciapaglia, C., Pike, E.P., Amstrup, S.C., van Hooidonk, R.,

et al. (2018). Climate change threatens the world’s marine protected areas. Nat.

Clim. Chang., 8, 499–503.

Buglass, S., Donner, S.D. & Alemu I, J.B. (2016). A study on the recovery of Tobago’s

coral reefs following the 2010 mass bleaching event. Mar. Pollut. Bull., 104, 198–

206.

Burke, L. & Maidens, J. (2004). Reefs at risk in the Caribbean.

Burke, L., Reytar, K., Salding, M. & Perry, A. (2010). Reefs at risk: Revisited.

Burkepile, D.E., Hay, M.E., Carpenter, R., Carpenter, R., Hatcher, B., Larkum, A., et al.

(2010). Impact of herbivore identity on algal succession and coral growth on a

Caribbean reef. PLoS One, 5, e8963.

Byrne, J. (2005). Grenada National Protected Area System Gap Assessment. The

Nature Conservancy.

Castillo, K.D., Ries, J.B., Weiss, J.M. & Lima, F.P. (2012). Decline of forereef corals in

response to recent warming linked to history of thermal exposure. Nat. Clim.

Chang., 2.

Chornesky, E. & Peters, E. (1987). Sexual reproduction and colony growth in the

scleractinian coral Porites astreoides. Ref. Biol. BulL, 172, 161–177.

Coral Reefs of Carriacou Island, Grenadines, Grenada: Assessment of condition and

key hhreats. (2005). The Nature Conservancy;Caribbean Marine Program.

Dalton, T., Forrester, G. & Pollnac, R. (2012). Participation, process quality, and

performance of marine protected areas in the wider Caribbean. Environ. Manage.,

49, 1224–1237.

Darling, E.S., Alvarez-Filip, L., Oliver, T.A., McClanahan, T.R. & Côté, I.M. (2012).

Evaluating life-history strategies of reef corals from species traits. Ecol. Lett., 15,

1378–1386.

Darling, E.S., McClanahan, T.R. & Côté, I.M. (2013). Life histories predict coral

community disassembly under multiple stressors. Glob. Chang. Biol., 19, 1930–

1940.

Done, T.J. (1999). Coral community adaptability to environmental change at the scales

of regions, reefs and reef zones. Am. Zool., 39, 66–67.

Edinger, E.N. & Risk, M.J. (1995). Preferential survivorship of brooding corals in a

regional extinction. Paleobiology, 21, 200–219.

Edmunds, P. (2010). Population biology of Porites astreoides and Diploria strigosa on a

shallow Caribbean reef. Mar. Ecol. Prog. Ser., 418, 87–104.

Elizalde-Rendón, E.M., Horta-Puga, G., González-Diaz, P. & Carricart-Ganivet, J.P.

(2010). Growth characteristics of the reef-building coral Porites astreoides under

different environmental conditions in the Western Atlantic. Coral Reefs, 29, 607–

614.

Ennis, R.S., Brandt, M.E., Grimes, K.R.W. & Smith, T.B. (2016). Coral reef health

response to chronic and acute changes in water quality in St. Thomas, United

States Virgin Islands. MPB.

Gardner, T.A., Côté, I.M., Gill, J.A., Grant, A. & Watkinson, A.R. (2003). Long-term

region-wide declines in Caribbean corals. Science, 301, 958–60.

Glasl, B., Herndl, G.J. & Frade, P.R. (2016). The microbiome of coral surface mucus

has a key role in mediating holobiont health and survival upon disturbance. ISME

J., 10, 2280–2292.

Gleason, D.F. (1993). Differential effects of ultraviolet radiation on green and brown

morphs of the Caribbean coral Porites astreoides. Limnol. Oceanogr., 38, 1452–

1463.

Gleason, D.F., Edmunds, P.J. & Gates, R.D. (2006). Ultraviolet radiation effects on the

behavior and recruitment of larvae from the reef coral Porites astreoides. Mar. Biol.,

148, 503–512.

Goodbody-Gringley, G. & de Putron, S.J. (2016). Brooding corals: planulation patterns,

larval behavior, and recruitment dynamics in the face of environmental change. In:

The Cnidaria, Past, Present and Future. Springer International Publishing, Cham,

pp. 279–289.

Goodwin, M., Cole, M., Stewart, W. & Zimmerman, B. (1976). Species density and

associations in Caribbean reef corals. J. Exp. Mar. Bio. Ecol., 24, 19–31.

Goreau, T. (1959). The Ecology of Jamaican Coral Reefs I. Species Composition and

Zonation. Ecology, 40, 67–90.

Graham, N.A.J. & McClanahan, T.R. (2013). The last call for marine wilderness?

Bioscience, 63, 397–402.

Green, D., Edmunds, P. & Carpenter, R. (2008). Increasing relative abundance of

Porites astreoides on Caribbean reefs mediated by an overall decline in coral

cover. Mar. Ecol. Prog. Ser., 359, 1–10.

Grenada’s Coral Reef Report Card. (2016). The Nature Conservancy. Available at:

http://caribnode.org/documents/85. Last accessed 16 November 2017.

Gress, E., Arroyo-Gerez, M.J., Wright, G. & Andradi-Brown, D.A. (2018). Assessing

mesophotic coral ecosystems inside and outside a Caribbean marine protected

area. R. Soc. Open Sci., 5, 180835.

Hall, E.R., DeGroot, B.C. & Fine, M. (2015). Lesion recovery of two scleractinian corals

under low pH conditions: Implications for restoration efforts. Mar. Pollut. Bull.

Hardt, M.J. (2009). Lessons from the past: the collapse of Jamaican coral reefs. Fish.

Fisher. 10, 143-158.

Hargreaves-Allen, V.A., Mourato, S. & Milner-Gulland, E.J. (2017). Drivers of coral reef

marine protected area performance. PLoS One, 12, e0179394.

Hauff, B., Haslun, J.A., Strychar, K.B., Ostrom, P.H. & Cervino, J.M. (2016). Symbiont

diversity of zooxanthellae (Symbiodinium spp.) in Porites astreoides and

Montastraea cavernosa from a reciprocal transplant in the lower Florida Keys. Int.

J. Biol., 8, 9.

Holstein, D.M., Paris, C.B. & Mumby, P.J. (2014). Consistency and inconsistency in

multispecies population network dynamics of coral reef ecosystems. Mar. Ecol.

Prog. Ser., 499, 1–18.

Holstein, D.M., Paris, C.B., Vaz, A.C. & Smith, T.B. (2016a). Modeling vertical coral

connectivity and mesophotic refugia. Coral Reefs.

Holstein, D.M., Smith, T.B. & Paris, C.B. (2016b). Depth-Independent reproduction in

the reef coral Porites astreoides from shallow to mesophotic zones. PLoS One.

Holstein, D.M., Smith, T.B., Paris, C.B., Riegl, B., Piller, W., Bongaerts, P., et al.

(2016c). Depth-independent reproduction in the reef coral Porites astreoides from

shallow to mesophotic zones. PLoS One, 11, e0146068.

Horricks, R.A. (2017). Tissue Regeneration of Artificially Induced Lesions in the

Caribbean Great Star Coral (Montastraea cavernosa) in the Nearshore Waters of

Grenada. University of Guelph.

Hughes, T.P., Barnes, M.L., Bellwood, D.R., Cinner, J.E., Cumming, G.S., Jackson,

J.B.C., et al. (2017). Coral reefs in the Anthropocene. Nature, 546, 82–90.

Huntington, B.E., Karnauskas, M. & Lirman, D. (2011). Corals fail to recover at a

Caribbean marine reserve despite ten years of reserve designation. Coral Reefs,

30, 1077–1085.

International Union for Conservation of Nature and Natural Resources. (2000). The

IUCN red list of threatened species. IUCN Global Species Programme Red List

Unit.

Briefing Paper: The effects of terrestrial runoff of sediments, nutrients, and other

pollutants on coral reefs. (2005). International Society for Reef Studies.

Jackson, J., Donovan, M., Cramer, K. & Lam, V. (2014). Status and Trends

of Caribbean Coral Reefs: 1970-2012.

Kaplan, K.A., Ahmadia, G.N., Fox, H., Glew, L., Pomeranz, E.F. & Sullivan, P. (2015).

Linking ecological condition to enforcement of marine protected area regulations in

the greater Caribbean region. Mar. Policy, 62, 186–195.

Kenkel, C.D., Goodbody-Gringley, G., Caillaud, D., Davies, S.W., Bartels, E. & Matz, M.

V. (2013a). Evidence for a host role in thermotolerance divergence between

populations of the mustard hill coral ( Porites astreoides ) from different reef

environments. Mol. Ecol., 22, 4335–4348.

Kenkel, C.D. & Matz, M. V. (2016). Gene expression plasticity as a mechanism of coral

adaptation to a variable environment. Nat. Publ. Gr., 1.

Kenkel, C.D., Meyer, E. & Matz, M. V. (2013b). Gene expression under chronic heat

stress in populations of the mustard hill coral ( Porites astreoides ) from different

thermal environments. Mol. Ecol., 22, 4322–4334.

Kenkel, C.D., Setta, S.P. & Matz, M. V. (2015). Heritable differences in fitness-related

traits among populations of the mustard hill coral, Porites astreoides. Heredity

(Edinb)., 115, 509–516.

Kitahara, M. V, Cairns, S.D., Stolarski, J., Blair, D. & Miller, D.J. (2010). A

comprehensive phylogenetic analysis of the Scleractinia (Cnidaria, Anthozoa)

based on mitochondrial CO1 sequence data. PLoS One, 5, e11490.

Lagabrielle, E., Crochelet, E., Andrello, M., Schill, S.R., Arnaud-Haond, S., Alloncle, N.,

et al. (2014). Connecting MPAs - eight challenges for science and management.

Aquat. Conserv. Mar. Freshw. Ecosyst., 24, 94–110.

Lamb, J.B., Williamson, D.H., Russ, G.R. & Willis, B.L. (2015). Protected areas mitigate

diseases of reef-building corals by reducing damage from fishing. Ecology.

Lecours, V., Devillers, R., Schneider, D., Lucieer, V., Brown, C. & Edinger, E. (2015).

Spatial scale and geographic context in benthic habitat mapping: review and future

directions. Mar. Ecol. Prog. Ser., 535, 259–284.

Levas, S., Grottoli, A.G., Schoepf, V., Aschaffenburg, M., Baumann, J., Bauer, J.E., et

al. (2016). Can heterotrophic uptake of dissolved organic carbon and zooplankton

mitigate carbon budget deficits in annually bleached corals? Coral Reefs, 35, 495–

506.

Lillis, A., Apprill, A., Suca, J.J., Becker, C., Llopiz, J.K. & Mooney, T.A. (2018).

Soundscapes influence the settlement of the common Caribbean coral Porites

astreoides irrespective of light conditions. R. Soc. Open Sci., 5, 181358.

Madin, J. (2017). Coral Trait Database | Growth rate: Porites Astreoides. Coral Trait

Database. Available at: https://coraltraits.org/species/1269/traits/60. Last accessed

16 November 2017.

Madin, J.S., Anderson, K.D., Andreasen, M.H., Bridge, T.C.L., Cairns, S.D., Connolly,

S.R., et al. (2016). The Coral Trait Database, a curated database of trait

information for coral species from the global oceans. Sci. Data, 3, 160017.

Manzello, D.P., Enochs, I.C., Kolodziej, G. & Carlton, R. (2015). Coral growth patterns

of Montastraea cavernosa and Porites astreoides in the Florida Keys: The

importance of thermal stress and inimical waters. J. Exp. Mar. Bio. Ecol., 471, 198–

207.

McClanahan, T. & Karnauskas, M. (2011). Relationships between benthic cover, current

strength, herbivory, and a fisheries closure in Glovers Reef Atoll, Belize. Coral

Reefs, 30, 9–19.

McGuire, M.P. (1998). Timing of larval release by Porites astreoides in the northern

Florida Keys. Coral Reefs, 17, 369–375.

McIntyre, A.D. (2010). Life in the world’s oceans : diversity, distribution, and abundance.

Meyer, J.L., Paul, V.J., Teplitski, M., Riegl, B., Bruckner, A., Coles, S., et al. (2014).

Community shifts in the surface microbiomes of the coral Porites astreoides with

unusual Lesions. PLoS One, 9, e100316.

Miller, M., Qian, P.-Y., Williams, D.E., Huntington, B.E., Piniak, G.A. & Vermeij, M.J.A.

(2016). Decadal comparison of a diminishing coral community: a study using

demographics to advance inferences of community status. PeerJ.

Grenada Ministry of Agriculture, Forestry & Fisheries. (2010). Molinière-Beauséjour

Marine Protected Area Management Plan.

Moffitt, E., White, J. & Botsford, L. (2013). Accurate assessment of marine protected

area success depends on metric and spatiotemporal scale of monitoring. Mar. Ecol.

Prog. Ser., 489, 17–28.

Morgan, K.M., Perry, C.T., Smithers, S.G., Johnson, J.A. & Daniell, J.J. (2016).

Evidence of extensive reef development and high coral cover in nearshore

environments: implications for understanding coral adaptation in turbid settings.

Nat. Publ. Gr.

Mumby, P.J. & Harborne, A.R. (2010). Marine reserves enhance the recovery of corals

on Caribbean reefs. PLoS One, 5, e8657.

Mumby, P.J. & Hastings, A. (2007). The impact of ecosystem connectivity on coral reef

resilience. J. Appl. Ecol., 45, 854–862.

Mumby, P.J., Hastings, A. & Edwards, H.J. (2007). Thresholds and the resilience of

Caribbean coral reefs. Nature, 450, 98–101.

Murdoch, T.J.T. (1991). A functional group approach for predicting the composition of

hard coral assemblages in Florida and Bermuda.

Nagelkerken, I. & Bak, R. (1998). Differential regeneration of artificial lesions among

sympatric morphs of the Caribbean corals Porites astreoides and Stephanocoenia

michelinii. Mar. Ecol. Prog. Ser., 163, 279–283.

National Oceanic and Atmospheric Administration. (2009). The value of coral and coral

ecosystems.

The Nature Conservancy (2017). Satellite Mapping of Benthic Habitats and Bathymetry

for Grenada, Caribbean.

The Nature Conservancy & Fisheries Division, G. (2007). Sandy Island/Oyster Bed

Marine Protected Area Management Plan.

Neave, M.J., Rachmawati, R., Xun, L., Michell, C.T., Bourne, D.G., Apprill, A., et al.

(2016). Differential specificity between closely related corals and abundant

Endozoicomonas endosymbionts across global scales. ISME J. Adv. online Publ.,

Nimrod, S., Easter-Pilcher, A., Aiken, K., Buddo, D.S. & Franco, C. (2017). Status of

Diadema antillarum populations in Grand Anse Bay, Grenada, 30 years after mass

mortality. Bull. Mar. Sci., 93, 917–927.

Noble, M.M., van Laake, G., Berumen, M.L. & Fulton, C.J. (2013). Community change

within a Caribbean coral reef marine protected area following two decades of local

management. PLoS One, 8, e54069.

Okazaki, R.R., Towle, E.K., van Hooidonk, R., Mor, C., Winter, R.N., Piggot, A.M., et al.

(2016). Species-specific responses to climate change and community composition

determine future calcification rates of Florida Keys reefs. Glob. Chang. Biol.

Olsen, K., Paul, V.J. & Ross, C. (2015). Direct effects of elevated temperature, reduced

pH, and the presence of macroalgae (Dictyota spp.) on larvae of the Caribbean

coral Porites astreoides. Bull. Mar. Sci., 91, 255–270.

Olsen, K., Sneed, J.M. & Paul, V.J. (2016). Differential larval settlement responses of

Porites astreoides and Acropora palmata in the presence of the green alga

Halimeda opuntia. Coral Reefs, 35, 521–525.

Perera-Valderrama, S., Hernández-Arana, H., Ruiz-Zárate, M.-Á., Alcolado, P.M.,

Caballero-Aragón, H., González-Cano, J., et al. (2016). Condition assessment of

coral reefs of two marine protected areas under different regimes of use in the

north-western Caribbean. Ocean Coast. Manag., 127, 16–25.

Perry, C.T., Steneck, R.S., Murphy, G.N., Kench, P.S., Edinger, E.N., Smithers, S.G., et

al. (2015). Regional-scale dominance of non-framework building corals on

Caribbean reefs affects carbonate production and future reef growth. Glob. Chang.

Biol., 21, 1153–64.

Edmunds, H.R.L. (2016). Cryptic regime shift in benthic community structure on shallow

reefs in St. John, US Virgin Islands. Mar. Ecol. Prog. Ser., 559, 1–12.

Pomeroy, R.S., Watson, L.M., Parks, J.E. & Cid, G.A. (2005). How is your MPA doing?

A methodology for evaluating the management effectiveness of marine protected

areas. Ocean Coast. Manag., 48, 485–502.

Porskamp, P., Rattray, A., Young, M., Ierodiaconou, D., Porskamp, P., Rattray, A., et al.

(2018). Multiscale and Hierarchical Classification for Benthic Habitat Mapping.

Geosciences, 8, 119.

Porto-Hannes, I., Zubillaga, A.L., Shearer, T.L., Bastidas, C., Salazar, C., Coffroth,

M.A., et al. (2015). Population structure of the corals Orbicella faveolata and

Acropora palmata in the Mesoamerican Barrier Reef System with comparisons over

Caribbean basin-wide spatial scale. Mar. Biol., 162.

Powers, D.A., Powers, F. & James Rohlf, F. (1972). A numerical taxonomic study of

Caribbean and Hawaiin reef corals. Syst. Zool., 21, 53–64. de Putron, S. & Smith, S. (2011). Planula release and reproductive seasonality of the

scleractinian coral Porites astreoides in Bermuda, a high-latitude reef. Bull. Mar.

Sci., 87, 75–90.

Rachello-Dolmen, P.G. & Cleary, D.F.R. (2007). Relating coral species traits to

environmental conditions in the Jakarta Bay/Pulau Seribu reef system, Indonesia.

Estuar. Coast. Shelf Sci., 73, 816–826.

Ritson-Williams, R., Arnold, S. & Paul, V. (2016). Patterns of larval settlement

preferences and post‑settlement survival for seven Caribbean corals. Mar. Ecol.

Prog. Ser., 548, 127–138.

Ritson-Williams, R., Ross, C., Paul, V.J., Hoegh-Guldberg, O., Pandolfi, J., Bradbury,

R., et al. (2016). Elevated temperature and allelopathy impact coral recruitment.

PLoS One, 11, e0166581.

Rocha, L.A., Pinheiro, H.T., Shepherd, B., Papastamatiou, Y.P., Luiz, O.J., Pyle, R.L.,

et al. (2018). Mesophotic coral ecosystems are threatened and ecologically distinct

from shallow water reefs. Science (80-. )., 361, 281–284.

Rodriguez-Lanetty, M., Granados-Cifuentes, C., Barberan, A., Bellantuono, A.J. &

Bastidas, C. (2013). Ecological Inferences from a deep screening of the Complex

Bacterial Consortia associated with the coral, Porites astreoides. Mol. Ecol., 22,

4349–4362.

Roff, G., Bejarano, S., Bozec, Y.M., Nugues, M., Steneck, R.S. & Mumby, P.J. (2014).

Porites and the Phoenix effect: Unprecedented recovery after a mass coral

bleaching event at Rangiroa Atoll, French Polynesia. Mar. Biol.

Romano, S.L. & Cairns, S.D. (2000). Molecular phylogenic hypotheses for the evolution

of scleractinian corals. Bull. Mar. Sci., 67, 1043–1068.

Ross, C., Ritson-Williams, R., Olsen, K. & Paul, V.J. (2013). Short-term and latent post-

settlement effects associated with elevated temperature and oxidative stress on

larvae from the coral Porites astreoides. Coral Reefs, 32, 71–79.

Scheibling, R.E., Patriquin, D.G. & Filbee-Dexter, K. (2017). Distribution and abundance

of the invasive seagrass Halophila stipulacea and associated benthic macrofauna

in Carriacou, Grenadines, Eastern Caribbean.

Schill, S.R., Raber, G.T., Roberts, J.J., Treml, E.A., Brenner, J. & Halpin, P.N. (2015).

No Reef Is an Island: Integrating Coral Reef Connectivity Data into the Design of

Regional-Scale Marine Protected Area Networks. PLoS One, 10, e0144199.

Schoepf, V., Grottoli, A.G., Levas, S.J., Aschaffenburg, M.D., Baumann, J.H., Matsui,

Y., et al. (2015). Annual coral bleaching and the long-term recovery capacity of coral.

Proc. R. Soc. B Biol. Sci., 282, 20151887.

Schuhmann, P.W. & Mahon, R. (2015). The valuation of marine ecosystem goods and

services in the Caribbean: A literature review and framework for future valuation

efforts. Ecosyst. Serv., 11, 56–66.

Schutte, V.G.W., Selig, E.R. & Bruno, J.F. (2010). Regional spatio-temporal trends in

Caribbean coral reef benthic communities. Mar. Ecol. Prog. Ser.

Selig, E.R. & Bruno, J.F. (2010). A global analysis of the effectiveness of marine

protected areas in preventing coral loss. PLoS One, 5, e9278.

Selig, E.R., Casey, K.S. & Bruno, J.F. (2012). Temperature-driven coral decline: the role

of marine protected areas. Glob. Chang. Biol., 18, 1561–1570.

Serrano, X.M., Baums, I.B., Smith, T.B., Jones, R.J., Shearer, T.L. & Baker, A.C.

(2016). Long distance dispersal and vertical gene flow in the Caribbean brooding

coral Porites astreoides. Nat. Publ. Gr.

Sharp, K.H., Distel, D. & Paul, V.J. (2012). Diversity and dynamics of bacterial

communities in early life stages of the Caribbean coral Porites astreoides. ISME J.,

6, 790–801.

Shearer, T.L. & Coffroth, M.A. (2008). Barcoding corals: limited by interspecific

divergence, not intraspecific variation. Mol. Ecol. Resour., 8, 247–255.

Smith, A.F., Rogers-, C.S. & Bouchon, C. (1997). Status of western atlantic coral reefs

in the Lesser Antilles. Proc 8th Int Coral Reef Syrn, 1, 351–356.

Stanley, G.D. (2003). The evolution of modern corals and their early history. Earth-

Science Rev., 60, 195–225.

Stevens, T. & Connolly, R.M. (2005). Local-scale mapping of benthic habitats to assess

representation in a marine protected area. Mar. Freshw. Res., 56, 111–123.

Strain, E.M.A., Edgar, G.J., Ceccarelli, D., Stuart-Smith, R.D., Hosack, G.R. &

Thomson, R.J. (2019). A global assessment of the direct and indirect benefits of

marine protected areas for coral reef conservation. Divers. Distrib., 25, 9–20.

Sunagawa, S., Woodley, C.M., Medina, M., Whitman, W., Coleman, D., Wiebe, W., et

al. (2010). Threatened corals provide underexplored microbial habitats. PLoS One,

5, e9554.

Turner, M., Vancouver, N. & Mitchell, K. (2009). Grenada protected area system plan

part 1 Identification and designation of protected areas.

Vega Thurber, R., Burkepile, D.E., Correa, A.M.S., Thurber, A.R., Shantz, A.A., Welsh,

R., et al. (2012). Macroalgae decrease growth and alter microbial community

structure of the reef-building coral, Porites astreoides. PLoS One, 7, e44246.

Veron, J. (2013). Corals of the World - about corals and reefs. AIMS. Available at:

http://coral.aims.gov.au/info/about.jsp. Last accessed 16 November 2017.

Wahidin, N., Siregar, V.P., Nababan, B., Jaya, I. & Wouthuyzen, S. (2015). Object-

based Image Analysis for Coral Reef Benthic Habitat Mapping with Several

Classification Algorithms. Procedia Environ. Sci., 24, 222–227.

White, K.Q. (2014). Signature page: A correlation between larval size and spat growth

and development in coral species.

Wilkinson, C. (2008). Status of coral reefs of the world: 2008.

Williams, S.M., Mumby, P.J., Chollett, I. & Cortés, J. (2015). Importance of

differentiating Orbicella reefs from gorgonian plains for ecological assessments of

Caribbean reefs. Mar. Ecol. Prog. Ser.

Williams, S.M., Sánchez-Godínez, C., Newman, S.P. & Cortés, J. (2016). Ecological

assessments of the coral reef communities in the Eastern Caribbean and the

effects of herbivory in influencing coral juvenile density and algal cover. Mar. Ecol.

Wilson, S.K., Graham, N.A.J., Fisher, R., Robinson, J., Nash, K., Chong-Seng, K., et al.

(2012). Effect of macroalgal expansion and marine protected areas on coral

recovery following a climatic disturbance. Conserv. Biol., 26, 995–1004.

Xu, J. & Zhao, D. (2014). Review of coral reef ecosystem remote sensing. Acta Ecol.

Sin., 34, 19–25.

CHAPTER 2: HIGH RESOLUTION BENTHIC HABITAT MAPPING WITHIN THE SANDY ISLAND-OYSTER BED MPA (GRENADA)

2.1 ABSTRACT

Coral reefs in the Caribbean are in decline, with live coral cover decreasing from ~50% in the 1970s to <10% today. Marine protected areas (MPAs) are one of the primary tools used to protect and restore marine ecosystems. To properly manage nearshore MPAs on a local scale, accurate spatial information regarding benthic habitats is essential. The distribution of reef habitats within the recently established

Sandy Island-Oyster Bed (SIOB) MPA in Grenada was relatively unknown. A team from

The Nature Conservancy conducted benthic surveys in March and June 2017.

Bathymetric field measurements were collected using a Lowrance Elite7Ti ® system with a xSonic P319 (50/200kHz) transducer and 10Hz GPS receiver that collected continuous depth readings at 3 pts/sec along each transect. A total of 127 survey points were collected within the MPA (21.6/km2). Benthic composition was visually assessed at each sample point and categorized into one of ten habitat classes. The benthic habitat map was created using an object-based classification process within eCognition software (v. 8.9, Trimble Inc.). Spectral properties of habitat types were identified at each of the benthic ground reference points and were used to parcel satellite imagery into segmented “objects”. All objects were then assigned a benthic habitat class based on their spectral and textural features. The final distribution of benthic habitat classes was then exported to Global Mapper. The respective habitat prominence of the resultant benthic habitat map within the SIOBMPA was: dense seagrass (25.74%); deep-water

(24.98%); sand (23.32%); hardground with turf (8.09%); coral framework (5.88%); sparse seagrass with algae (4.85%); hardground with gorgonians (3.69%); boulders

(2.38%); and reef rubble (1.04%). This GIS product will support ongoing spatial management practices within the SIOBMPA and form part of a wider regional mapping effort by The Nature Conservancy in the Eastern Caribbean.

2.2 INTRODUCTION

Coral reefs are an essential part of the global marine ecosystem and provide an array of ecological services as well as support complex and diverse species assemblages (Wilkinson 2008a). From 1975 to 2004, coral reef cover in the Caribbean declined 80%, from an average live coral coverage of 50% in the 1970s to ~10% commonly seen in many areas today (Gardner et al. 2003). Marine protected areas are one of the primary tools used worldwide to protect marine resources and diversity

(Perera-Valderrama et al. 2016). Marine protected areas have been shown to increase the resilience of coral reef communities and decrease disease prevalence among colonies (Lamb et al. 2015; Selig & Bruno 2010; Selig et al. 2012). It is therefore imperative that research objectives focus on providing conservationists with the tools necessary to properly manage existing MPAs, and to identify priority conservation areas. For these reasons governments across the Caribbean have been establishing networks of MPAs (Grenada’s Coral Reef Report Card 2016). As a signatory of the

Convention of Biological Diversity, the Government of Grenada has recognized the need to protect its natural environment and biodiversity through the creation of protected areas, both terrestrial and marine (Thomas 2016). In recognition of the acute risk facing Grenada, both as a small island and a developing country with declining

environmental health, the nation published the ‘25-25 declaration’ in 2006. This declaration pledged to protect 25% of both its terrestrial and nearshore marine environment (Figure 2.1) by 2020 (Byrne 2005).

Figure 2.1 Shallow reef habitat (7 m in depth) near Hillsborough, Grenada.

To date, four MPAs have been created in the nearshore waters of Grenada:

Sandy Island-Oyster Bed (7.87 km2), Moliniere-Beasejour (0.8 km2), Grand Anse Bay

(13.4 km2), and Woburn-Clark’s Court Bay (4.2 km2) (Figure 2.2).

Figure 2.2 Four marine protected areas have been established to date in Grenada: A) Sandy Island-Oyster Bed B) Molinier-Beasejour C) Grand Anse Bay and D) Wobourne Clarke.

These MPAs were established in 2009, 1999, 1999, and 2018, respectively

(Grenada Coral Reef Report Card 2016) and eleven more MPAs forming a total of 197 km2 (~10% of Grenada’s marine area) are currently proposed (Byrne 2005; Turner et al.

2009). It is particularly essential that these areas are properly managed and monitored given that Grenada has a high dependence on these ecosystems, which have a very

low adaptive capacity to the impacts of climate change (Burke et al. 2010). Accurate benthic habitat maps of tropical reef habitats are an important tool in the analysis of large-scale habitat shifts, research prioritization, MPA management, the establishment of a representative network of MPAs, and guiding site selection for future scientific research (Franklin et al. 2003; Pomeroy et al. 2005; Agardy et al. 2011). Linking ecological science with tools for conservation will become more imperative as the world’s oceans continue to face deleterious anthropogenic impacts such as overfishing, pollution, and climate change (Bruno et al. 2018). Benthic habitat maps of coral reef environments are used to aid conservation planning, monitoring, fisheries management, and to guide future scientific studies (Knudby et al. 2014). Through a combination of high-resolution satellite imagery, bathymetry, and ground proofing surveys, an extensive map of bottom composition throughout an area can be created (The Nature

Conservancy 2017; Xu & Zhao 2014). Benthic habitat mapping methods are rapidly moving towards further automation with improved satellite/bathymetric data (Roelfsema et al. 2018). This increasingly reduces the number and extent of truthing points needed to produce a benthic habitat map (Lecours et al. 2015). However, at this time truthing points are still required, particularly for benthic habitat classes with similar spectral signatures (i.e. coral framework vs hardground with gorgonians). Historically, reefs in

Grenada and Carriacou have been understudied, especially in regards to the spatial distribution of benthic habitat types compared to other regions in the Caribbean (Byrne

2005; Anderson et al. 2014; Grenada’s Coral Reef Report Card 2016). Spatially, the information that exists for the SIOBMPA is sparse and at a coarse level insufficient for management at a local scale (Byrne 2005). Therefore, the objectives of this study were

to: 1) Conduct benthic truthing surveys of the SIOBMPA with a high density of sample locations; 2) Create a benthic habitat map of the MPA, and 3) contribute to the mapping of the tri-island state of Grenada in its entirety.

2.3 MATERIALS AND METHODS

2.3.1 RESEARCH SITE

The tri-island nation of Grenada was selected as the location of this study due to its rapidly expanding network of MPAs, at-risk coral reef ecosystems, and a lack of existing scientific research in the area (Figure 2.3). In addition, The Nature Conservancy

(TNC) requested the analysis of benthic habitat distribution in Carriacou to complete a wider mapping initiative in the Eastern Caribbean.

Figure 2.3 Location of study site within the wider Caribbean region and within the nation of Grenada (see site location in Carriacou; inset).

Within Grenada, the island of Carriacou was selected since it has a recently established and understudied MPA (SIOB; Figure 2.4), declining reef health, and had yet to be thoroughly mapped by TNC, unlike the main island of Grenada. The marine component of the SIOBMPA is 5.75 km2. The remainder of this 7.87 km2 protected area is comprised of: mangroves (1.2 km2), islands (0.88 km2), and a cruise terminal (0.04 km2).

Figure 2.4 Boundary of the Sandy Island-Oyster Bed MPA (Carriacou, Grenada)

2.3.2 FIELD SAMPLING

All benthic truthing photographs were taken using two GoPro Hero+ Black cameras secured to a weighted PVC piping apparatus and lowered to within visual range of seafloor for photographing. A small motorized pleasure craft was used to move between each individual sampling location where the camera device was lowered from a spool. Specific drop locations were along transects haphazardly defined to cover wide

geographic areas of the MPA and to capture areas of each benthic habitat class (Figure

2.5). The approximate target areas of these transects were based on preliminary observations from satellite imagery (i.e. sand areas vs. framework areas) as well as local knowledge of the benthos. All efforts were made to leave a 100 m distance between individual sampling locations to ensure that no re-analysis of the same position took place. A portable Canmore GPS trip logger was used to mark each individual camera drop location to later match with individual benthic sample site photographs.

Bathymetric field measurements were collected using a Lowrance Elite7Ti ® system with a xSonic P319 (50/200kHz) transducer and 10Hz GPS receiver that collected continuous depth readings at 3 pts/sec along each transect. In parallel, a GPS- referenced SeaView Sea-Drop 6000 HD underwater video camera with 30 m of cable was used to record benthic habitat types along transects identified using the WorldView-

2 imagery

2.3.3 BENTHIC ANALYSIS

Previous mapping studies conducted by TNC in Grenada found that (for benthic mapping purposes) the shallow benthic environments of the area could be broadly classified as one of 10 benthic habitat classes, being: boulders, reef rubble, hardground with turf, hardground with gorgonians, sparse seagrass with algae, dense seagrass, sand, coral framework, Montastrea reef complex, and Porites reef complex (Table 2.1;

The Nature Conservancy 2015). The highest resolution photographs from each sample point were visually assessed and compared against standardized habitat class images

(Table 2.1) and assigned a habitat class. Each of these were combined with their respective GPS coordinates and input into the benthic mapping analysis. Outside of the

SIOBMPA, an additional 63 benthic truthing points were collected in Carriacou to improve the accuracy of the mapping software.

Table 2.1 Benthic habitat categories used to categorize communities in Carriacou and example photographs used for habitat Identification (The Nature Conservancy, 2015).

Benthic Habitat Class Class Description Seafloor coverage >60% of large non- skeletal boulders. Some coverage of a) Boulders macroalgae, sponges, and Scleractinian corals. Seafloor coverage >60% by coral skeletal b) Reef rubble rubble fragments (from several cm to >1 m in size. Layer of turf algae covering >60% of scoured hardground. <5% coverage of c) Hardground with turf Scleractinian corals. gorgonians, sponges, etc... Scoured hardground with coverage dominated by gorgonians (>60%). <10% d) Hardground with gorgonians coverage of Scleractinian corals, macroalgae, sponges, etc...

Sparse seagrass meadows on a soft substrate interspersed with macroalgae e) Sparse seagrass with algae (<40%). Cyanobacteria mats often cover the underlying substrate.

Dense seagrass meadows with f) Dense seagrass macroalgae cover on soft substrate (>60%). Fine sediment with <20% cover of corals, g) Sand sponges, macroalgae, etc…

Complex hardground with <30% live coral cover. Small and large colonies dominated by a few Scleractinian species. h) Coral framework <20% coverage of gorgonians, macroalgae, sponges, etc…

Framework area dominated by Montastrea coral species. Seafloor i) Montastrea reef complex dominated by a living coral layer. Gorgonians, seagrass, etc... dominate the gaps. Framework area dominated by Porites coral species. Seafloor dominated by a j) Porites reef complex living coral layer. Gorgonians, seagrass, etc...dominate the gaps.

2.3.4 BENTHIC HABITAT MAPPING

The Worldview-2 satellite (operated by Digital Globe Inc.) provides 0.48 m panchromatic resolution and 1.85 m multispectral resolution. Image strips were obtained from this satellite for Grenada/Carriacou in November and December of 2014. Some cloud cover was present in the obtained imagery, and sun glint remained problematic in

several areas. When supplemented with and cross-compared to Landsat imagery and

Google Earth imagery, it was determined to be of sufficient quality to produce an accurate benthic habitat map. Using ENVI software (RSI Inc., v. 5.2) suitable areas of the images were selected, mosaiced, colour balanced, and stitched together. Previous single-beam depth soundings were used to tune a ratio algorithm following the methods of Stumpf et al. (2003) to extract spectral bathymetry from the Worldview-2 satellite data. A digital elevation model was constructed for the study area from low-tide to a depth of 25 m with a spatial resolution of 4x4 m and a vertical resolution of 0.01 m. With this information a water column spectral correction to facilitate proper spectral analysis of benthic habitats could be applied to shallow areas not dramatically influenced by sun glare (Stumpf et al. 2003).

Benthic habitats were identified in the satellite imagery using an object- oriented approach (The Nature Conservancy 2017). For this stage of benthic habitat mapping eCognition (v. 8.9, Trimble Inc.) software was used. A total of 190 benthic truthing points in Carriacou with at least a single observation for each of the benthic habitat classes (with 124 of these being within the SIOBMPA) were input to aid the object-based identification analysis. Spectral properties of the identified habitat types were identified at each of the benthic truthing points and were used to begin the segmentation of the satellite imagery into “objects”. All objects were then assigned a benthic habitat class based on their spectral and textural features (The Nature

Conservancy 2017). Redundant divisions separating two objects of the same habitat class were removed using the “merge-region” tool in eCognition software. The final distribution of benthic habitat classes was then exported to Global Mapper (Blue Marble

Geo In., v. 15.1) for a final quality analysis and any necessary changes were manually implemented. The accuracy of the final benthic mapping product was assessed by comparing mapped habitat classes against those that were identified manually in the benthic truthing points. Each point was counted as either being mapped as the same class identified in the benthic truthing points, or not being mapped as the same class.

2.4 RESULTS

2.4.1 GROUND TRUTHING

127 benthic locations were successfully photographed in the study area (Figure

2.5). A single transect of an additional 10 survey points running East-West across the

MPA were removed from the study as there was an error in GPS data collection. In addition, three benthic truthing points fell outside the SIOBMPA boundary and were removed (leaving 124 points). Two of these points were identified as being hardground with turf and another as sparse seagrass & algae. The total marine area of the

SIOBMPA is ~5.75 km2 and was sampled at a high density of 21.6 sampling locations per km2 to aid in management planning as compared with the average sampling density of 0.52 per km2 for the map produced for the rest of Carriacou by TNC. Due to current and/or depth conditions the benthic truthing of the northwestern (rock pinnacle) and southeastern (mangrove) areas of the MPA were not sampled in the initial survey and four additional points were sampled in a second survey two months later to fill these gaps. The depth of each sample point was collected but was left out of the analysis as more accurate spectral bathymetric data was obtained for use in the creation of the benthic habitat map.

Figure 2.5 Ground truthing locations within the Sandy Island-Oyster Bed MPA, with habitat classifications.

The most dominant benthic habitat classes identified within the benthic truthing points ranked by proportion were: sand (40%); dense seagrass (17%); reef rubble

(11%); sparse seagrass with algae (11%); coral framework (8%); hardground with turf

(7%); hardground with gorgonians (5%); and Porites reef complex (2%) (Figure 2.6). No survey points were classified as being Montastrea complex or boulder.

Boulders 2% 8% 11% Reef rubble

Hardground with turf 7%

Hardground with gorgonians 5% Sparse seagrass with algae

40% Dense seagrass 11% Sand

Coral framework

Montastrea reef complex 17%

Porites reef complex

Figure 2.6 Relative proportion of benthic habitat classes for 123 ground truthing observations within the SIOBMPA.

2.4.2 BENTHIC HABITAT MAP

The benthic habitat map of the SIOBMPA was successfully created with the frequency and spatial distribution of each benthic habitat class in the area (Figure 2.7).

Figure 2.7 Benthic habitat map of the Sandy Island-Oyster Bed MPA (The Nature Conservancy 2017). Deep water areas are those where light does not adequately penetrate through the water column (due to depth) to allow the spectral analysis of benthic habitats.

An analysis of the accuracy for the final benthic map product compared against the initial benthic truthing points can be seen below (Table 2.2). Accuracy was highest for coral framework, sand, hardground with turf, and dense seagrass and lowest for Porites reef complex and sparse seagrass with algae. Boulder habitats were not identified in the truthing points but were identified in the final map product.

Table 2.2 Accuracy of modelled benthic habitat map based on benthic truthing points within the SIOBMPA. Habitats not identified in benthic truthing within the SIOBMPA are shaded black.

Habitat Type Accuracy (%) Boulders Coral framework 100 Deep water Dense seagrass 100 Hardground with gorgonians 86 Hardground with turf 100 Reef rubble 14 Sand 100 Sparse seagrass with algae 0 Porites Reef Complex 0 Montastrea Reef Complex

This map revealed that the benthic marine habitats in order of prevalence within the SIOBMPA were: dense seagrass (25.74%); deep-water (24.98%); sand

(23.32%); hardground with turf (8.09%); coral framework (5.88%); sparse seagrass with algae (4.85%); hardground with gorgonians (3.69%); boulders (2.38%); and finally reef rubble (1.04%) (Figure 2.8). Cloud cover blocked 0.03% of the MPA imagery during the analysis. Neither Porites nor Montastrea complex habitats were categorized by the mapping software within the MPA. Significant coral framework concentrations were found along the north facing areas of Jack Iron Point, Mabouya and Sandy Island, as well as around mangrove habitats (Figure 2.7).

Dense seagrass 2% 1% 4%

Deep water 5% 26% Sand 6%

Hardground with turf

8% Coral framework

Sparse seagrass with algae

Hardground with gorgonians

Boulders

Reef rubble 23% 25%

Clouds

Figure 2.8 Proportion of the SIOBMPA protected sea area occupied by each of nine benthic habitat classes.

2.5 DISCUSSION

Coral reef communities in the Caribbean have been experiencing dramatic declines and phase shifts in coral assemblages since the 1980s, with the tri-island state of Grenada being no exception (Gardner et al. 2003; Wilkinson 2008a; Anderson et al.

2014; Perry et al. 2015a). The reef conditions of Grenada are typical of those found across the Eastern Caribbean, with low hard coral coverage dominated by a small number of species, low or absent Diadema antillarum populations, and a high proportion of macroalgal coverage (Anderson et al. 2014; Horricks et al. 2017). MPAs remain a widely used tool both globally and within the Caribbean, as part of efforts to halt this decline and/or facilitate the recovery of impacted marine communities (Abelson et al.

2016). Due to the declining health of Grenada’s marine environment, a network of

MPAs is in development with four created to date (Turner et al. 2009). Established in

2010, the SIOBMPA is currently the only MPA on the island of Carriacou. The

SIOBMPA lacked accurate spatial data on the distribution of benthic habitats within its boundary (The Nature Conservancy & Fisheries Division 2007). The proportion of

Carriacou’s benthic marine habitats protected by the SIOBMPA were also unknown, with only the proportion of total island area being reported. The island of Carriacou as a whole also lacked accurate spatial data on the distribution of benthic marine habitats and its reefs have remained relatively understudied (Coral Reefs of Carriacou Island

2005; Grenada’s Coral Reef Report Card 2016). High-density benthic truthing of the

SIOBMA was therefore conducted to aid in subsequent benthic habitat mapping of the protected area and Carriacou as a whole. Accurate information on the spatial distribution of habitats is essential in the successful management and enforcement of any protected area designed to improve the health of tropical reef communities

(Hargreaves-Allen et al. 2017; Kabiri et al. 2018). Adequate representation of benthic marine habitats and understanding the proximity of key habitat areas to one another is also essential to the planning and success of any MPA network (Fava et al. 2009;

Abelson et al. 2016). The creation of a benthic habitat map of the SIOBMPA and

Carriacou will aid in the management of the protected area, and the informed expansion of the MPA system in Grenada.

Mapped distributions of benthic habitat classes were broadly characteristic of reefs in the Eastern Caribbean arc. These reefs are broadly characterised as being comprised of fore-reef areas with high coral growth, spur and grove habitat as depth

lessens and finally the reef top being dominated by sand, seagrass and macroalgae meadows (Figure 2.7) (The Nature Conservancy 2017). The benthos of the SIOBMPA was found to be largely dominated by sand, dense seagrass, and deep-water (Figure

2.7). Despite lacking the diversity or rugosity of coral framework areas, these habitats serve important ecological roles in tropical reef communities (Jackson 2001; Kahng et al. 2010; Slattery et al. 2011; Scheibling et al. 2017). Seagrass areas are particularly important for sequestering carbon, providing habitat, fish spawning, and providing feeding opportunities for numerous species (i.e. urchins and juvenile sea turtles)

(Jackson 2001; Stevens & Connolly 2005; Williams et al. 2016; Nimrod et al. 2017).

Deep-water areas are home to unique mesophotic reef communities essential for population health of shallower coral and fish species (Gress & Andradi-Brown 2018).

They are also home to unique and sensitive species assemblages, such as the black corals (Gress & Andradi-Brown 2018; Gress et al. 2018). These areas also have potential to serve as habitat refuges for shallower reef communities as temperature and storm conditions worsen at shallower depths as a result of climate change (Holstein et al. 2016a; Rocha et al. 2018). Sandy habitats often host large bacterial mats and provide a substrate in which numerous invertebrate and fish species reside (Sunagawa et al. 2010; Cameron et al. 2016).

Coral reef complex areas were largely north facing (i.e. Mabouya and Sandy

Island) and/or closely associated with nearby mangrove habitats (Figure 2.7). Clear separation between coral framework and hardground with gorgonian areas was observed, either a result of competitive dynamics or interspecific differences in habitat preferences. The mapped distribution of coral framework is likely explained by local

sediment dynamics (can inhibit the success of coral larvae), wave conditions (too weak results in sedimentation, too strong puts colony framework at risk of physical damage), and healthy herbivorous fish communities/water quality (provisioned by mangrove forests) (Anthony & Connolly 2004; Mumby & Hastings 2007; Mumby et al. 2007; Gilby et al. 2015; Stubler et al. 2016). Gorgonians with fan-like morphologies are often found in shallower more turbulent waters, while elongated narrow species tend to prefer calmer, deeper water, which may explain their distribution within the SIOBMPA

(Williams et al. 2015; Dias & Gondim 2016). Coral rubble areas were not widely distributed throughout the MPA and were predominantly isolated to the northern shore of Sandy Island. This indicates that much of the total area of coral framework distribution in the area has remained unchanged despite species composition shifts, unlike the windward shores of Carriacou where massive losses have been observed

(Grenada’s Coral Reef Report Card 2016; The Nature Conservancy 2017). It is of note that these mapped reef rubble areas are in shallow water on Sandy Island, and may have once been home to significant thickets of Acropora species (Riegl & Purkis 2015;

Bonin 2012; Perry et al. 2015; Xin et al. 2016). No Montastrea or Porites complex areas were mapped despite a benthic truthing point identifying the latter within the SIOBMPA.

The lack of Montastrea colonies is not unexpected as areas of these large colonies were not observed in the field.

The accuracy of the mapping process, assessed as the proportion of benthic truthing points where the final product mapped these points as their identified habitat class, proved to be variable. The lowest accuracy of successfully identified benthic truthing points in the final map product were: sparse seagrass with algae, Porites reef

complex, and reef rubble (Table 2.2). The highest accuracy of successfully identified benthic truthing points in the final map product were: coral framework, dense seagrass, hardground with gorgonians, hardground with turf, and sand (Table 2.2). The observation of a Porites complex area not included in the benthic habitat map may be a result of its small spatial scale not confirmed within the satellite imagery and/or imagery quality of satellite photos in the area, and therefore insufficient for inclusion by the mapping software. Reef rubble areas not included in the final map may have also suffered from excess glare in the satellite imagery (and therefore difficult spectral analysis), or very small localised distributions of this habitat were photographed. It is of note however that all benthic truthing points not correctly mapped in the final product

(apart from Porites complex) were in the immediate vicinity (<15 m) of the identified habitat or identified as a spectrally similar benthic habitat class. Truthing points identified as sparse seagrass with algae were mapped as dense seagrass or sand in most cases and may therefore have been a result of similar spectral signatures. Benthic habitat classes mapped with a high degree of accuracy likely had a strong and unique spectral signature (and/or limited glare in satellite imagery) facilitating their accurate analysis. Classes such as coral framework, sand, and dense seagrass beds are identifiable with the naked eye in shallow areas of satellite imagery so this high level of accuracy is not unexpected.

Despite the limited distance between the main island of Grenada and Carriacou, the reef communities on these island are distinct, with bathymetry and relative proportion of habitats varying considerably (Coral Reefs of Carriacou Island 2005;

Grenada’s Coral Reef Report Card 2016; Anderson et al. 2014; The Nature

Conservancy 2017). The distribution of living coral framework on Carriacou is more spatially limited than that those of Grenada, with much of the reefs on the windward side being destroyed by hurricanes (Grenada’s Coral Reef Report Card 2016). Adequate habitat and community representation are key factors in the design and success of MPA networks (Barriteau 2011; Dalton et al. 2012; Kaplan et al. 2015; Abelson et al. 2016).

The SIOBMPA is the only currently established MPA in Carriacou and protects significant amounts of total benthic habitats on the island with coral framework, boulders, and dense seagrass being highly represented (Table 2.3).

Table 2.3 Proportion of benthic habitats in Carriacou protected by the SIOBMPA.

Benthic Habitat Carriacou Habitat Representation (%) Boulders 18.8 Coral framework 19.5 Deep water 1.9 Dense seagrass 15.9 Hardground with gorgonians 8.0 Hardground with turf 5.1 Reef rubble 0.6 Sand 11.2 Sparse seagrass with algae 3.6

Benthic habitat mapping is a highly effective tool for the conservation and research of tropical reef environments but is not without its limitations. Despite the clarity of tropical waters, the depth of many areas made satellite imagery inadequate for mapping benthic habitats, particularly the ‘deep water’ habitat mapped in this study.

These mesophotic habitats, despite being largely understudied in many areas are significant sources of productivity and are important sources of recruitment for many species (Kahng et al. 2010, 2014; Andradi-Brown et al. 2017; Gress & Andradi-Brown

2018). A sufficient number of benthic truthing points and accurate satellite imagery is

necessary to ensure the creation of an accurate and viable habitat map (Stevens &

Connolly 2005; Shihavuddin et al. 2013). The benthic habitats of any areas obscured by

clouds present during satellite imagery are unknown (Rowlands et al. 2012;

Shihavuddin et al. 2013). Care must be taken in using these maps on a finer scale (i.e.

observed Porites complex not mapped), as they are a tool to observe the approximate

distribution of habitats (Stevens & Connolly 2005).

Repeating benthic habitat mapping can be used to monitor larger scale temporal shifts in habitat class coverage. Shifts in the benthic composition of non-coral species has already been occurring on the reefs of Grenada. Native seagrass species have been displaced in large areas by an invasive seagrass species Halophila stipulacea creating more habitat for inshore invertebrates like sponges and urchins, with a corresponding increase in food availability for fishes (Scheibling et al. 2017).

This study has provided information that can aid in MPA network design in

Carriacou. Information on the proportion of benthic habitats in Carriacou already protected can aid in the determination of boundaries necessary for several upcoming MPA projects in the area. The mapping of Carriacou as a whole (which the benthic truthing points of this study contributed to) will also aid conservation managers in targeting key habitat areas to protect, and unhealthy areas to regenerate (Dalton et al. 2012; Abelson et al. 2016;

Hargreaves-Allen et al. 2017). Mapped information on the frequency and distribution of sensitive coral framework and hardground with gorgonian areas can be used to concentrate limited enforcement resources and/or boundary demarcation resources. In

addition, this map can serve as a resource for researchers and conservationists to identify areas to concentrate survey resources and/or identify research sites.

This study formed a part of a wider mapping initiative of the Eastern Caribbean being conducted by TNC to aid island states in making information-based choices on the management and creation of new MPAs (Nature Conservancy & Fisheries Division 2007;

Grenada’s Coral Reef Report Card 2016). The benthic truthing points in this study will aid further mapping of the islands of Grenada and Carriacou by TNC in their entirety. The accurate spatial information mapped within the SIOBMPA will aid in the spatial and adaptive management of the area in the years to come, and contribute to accurate MPA network planning in the Grenadines (Pomeroy et al. 2005; Abelson et al. 2016).

2.6 ETHICS

Research permit information: IACUC 14001. A Porites colony and gorgonian colony

were damaged by the drop camera when it became ensnared during field work.

2.7 ACKNOWLEDGEMENTS

We would like to thank the Ontario Veterinary College, The Government of

Ontario (OGS), The Nature Conservancy Caribbean Division, and the German Federal

Ministry for the Environment, Nature Conservation, Building and Nuclear Safety (BMUB)

for providing funding for this research. We would also like to thank Deefer Diving

Carriacou, and The Nature Conservancy for making for assisting in data collection in the

field as well as conducting the benthic mapping in this chapter. Lastly, we would like to

thank Drs. Lorenzo-Alvarez-Filip and Steven Schill who were instrumental in reviewing the final outputs of this project.

2.8 LITERATURE CITED *References in format of ‘Ecology Letters’

Abelson, A., Nelson, P.A., Edgar, G.J., Shashar, N., Reed, D.C., Belmaker, J., et al.

(2016). Expanding marine protected areas to include degraded coral reefs.

Conserv. Biol., 30, 1182–1191.

Anderson, R., Morrall, C., Jossart, J., Nimrod, S., Bolda, E., Musser, K., et al. (2014).

Marine protected area monitoring in the nearshore waters of Grenada, Eastern

Caribbean: benthic cover and fish populations. Rev. Biol. Trop. (Int. J. Trop. Biol,

62, 273–286.

Andradi-Brown, D.A., Grey, R., Hendrix, A., Hitchner, D., Hunt, C.L., Gress, E., et al.

(2017). Depth-dependent effects of culling—do mesophotic lionfish populations

undermine current management? R. Soc. Open Sci., 4, 170027.

Anthony, K.R.N. & Connolly, S.R. (2004). Environmental limits to growth: Physiological

niche boundaries of corals along turbidity-light gradients. Oecologia, 141, 373–384.

Aronson, R.B., Macintyre, I.G., Wapnick, C.M. & O’Neill, M.W. (2004). Phase shifs,

alternative states, and the unprecedented converergence of two reef systems.

Ecology, 85, 1876–1891.

Barriteau, M. (2011). The Grenadines Network of Marine Protected Areas.

Transboundary cooperation worksop.

Bonin, M.C. (2012). Specializing on vulnerable habitat: Acropora selectivity among

damselfish recruits and the risk of bleaching-induced habitat loss. Coral Reefs, 31,

287–297.

Burke, L., Reytar, K., Salding, M. & Perry, A. (2010). Reefs at Risk: Revisited.

Byrne, J. (2005). Grenada National Protected Area System Gap Assessment. The

Nature Conservancy.

Cameron, C.M., Pausch, R.E. & Miller, M.W. (2016). Coral recruitment dynamics and

substrate mobility in a rubble-dominated back reef habitat. Bull. Mar. Sci.

Coral Reefs of Carriacou Island, Grenadines, Grenada: Assessment of condition and

key hhreats. (2005). The Nature Conservancy;Caribbean Marine Program.

Dalton, T., Forrester, G. & Pollnac, R. (2012). Participation, process quality, and

performance of marine protected areas in the wider Caribbean. Environ. Manage.,

49, 1224–1237.

Dias, T.L.P. & Gondim, A.I. (2016). Bleaching in scleractinians, hydrocorals, and

octocorals during thermal stress in a northeastern Brazilian reef. Mar. Biodivers.

Fava, F., Ponti, M., Scinto, A., Calcinai, B. & Cerrano, C. (2009). Possible effects of

human impacts on epibenthic communities and coral rubble features in the marine

Park of Bunaken (Indonesia). Estuar. Coast. Shelf Sci.

Forsman, Z.H., Page, C.A., Toonen, R.J. & Vaughan, D. (2015). Growing coral larger

and faster: micro-colony-fusion as a strategy for accelerating coral cover. PeerJ.

26500822

Gardner, T.A., Côté, I.M., Gill, J.A., Grant, A. & Watkinson, A.R. (2003). Long-term

region-wide declines in Caribbean corals. Science, 301, 958–60.

Gilby, B.L., Maxwell, P.S., Tibbetts, I.R. & Stevens, T. (2015). Bottom-up factors for

algal productivity outweigh no-fishing marine protected area effects in a marginal

coral reef system. Ecosystems.

Goreau, T. (1959). The Ecology of Jamaican Coral Reefs: Species composition and

zonation. Ecology, 40, 67–90.

Greenstein, B.J., Curran, H.A. & Pandolfi, J.M. (1998). Shifting ecological baselines and

the demise of Acropora cervicornis in the western North Atlantic and Caribbean

Province: a Pleistocene perspective. Coral Reefs, 17, 249–261.

Grenada’s Coral Reef Report Card. (2016). The Nature Conservancy. Available at:

http://caribnode.org/documents/85. Last accessed 16 November 2017.

Gress, E. & Andradi-Brown, D.A. (2018). Assessing population changes of historically

overexploited black corals (Order: Antipatharia) in Cozumel, Mexico. PeerJ, 6,

e5129.

Gress, E., Arroyo-Gerez, M.J., Wright, G. & Andradi-Brown, D.A. (2018). Assessing

mesophotic coral ecosystems inside and outside a Caribbean marine protected

area. R. Soc. Open Sci., 5, 180835.

Hargreaves-Allen, V.A., Mourato, S. & Milner-Gulland, E.J. (2017). Drivers of coral reef

marine protected area performance. PLoS One, 12, e0179394.

Holstein, D.M., Paris, C.B., Vaz, A.C. & Smith, T.B. (2016). Modeling vertical coral

connectivity and mesophotic refugia. Coral Reefs.

Horricks, R.A. (2017). Tissue Regeneration of Artificially Induced Lesions in the

Caribbean Great Star Coral (Montastraea cavernosa) in the Nearshore Waters of

Grenada. University of Guelph.

Jackson, J.B. (2001). What was natural in the coastal oceans? Proc. Natl. Acad. Sci. U.

S. A., 98, 5411–8.

Kabiri, K., Rezai, H. & Moradi, M. (2018). Mapping of the corals around Hendorabi

Island (Persian Gulf), using WorldView-2 standard imagery coupled with field

observations. Mar. Pollut. Bull., 129, 266–274.

Kahng, S.E., Garcia-Sais, J.R., Spalding, H.L., Brokovich, E., Wagner, D., Weil, E., et

al. (2010). Community ecology of mesophotic coral reef ecosystems. Coral Reefs,

29, 255–275.

Kaplan, K.A., Ahmadia, G.N., Fox, H., Glew, L., Pomeranz, E.F. & Sullivan, P. (2015).

Linking ecological condition to enforcement of marine protected area regulations in

the greater Caribbean region. Mar. Policy, 62, 186–195.

Karlson, R.H. & Hurd, L.E. (1993). Disturbance, coral reef communities, and changing

ecological paradigms. Coral Reefs, 12, 117–125.

Knudby, A., Pittman, S.J., Maina, J., Rowlands, G., Knudby, A., Pittman, S.J., et al.

(2014). Remote Sensing and Modeling of Coral Reef Resilience.

Lamb, J.B., Williamson, D.H., Russ, G.R. & Willis, B.L. (2015). Protected areas mitigate

diseases of reef-building corals by reducing damage from fishing. Ecology.

McClanahan, T. & Karnauskas, M. (2011). Relationships between benthic cover, current

strength, herbivory, and a fisheries closure in Glovers Reef Atoll, Belize. Coral

Reefs, 30, 9–19.

Mumby, P.J. & Hastings, A. (2007). The impact of ecosystem connectivity on coral reef

resilience. J. Appl. Ecol., 45, 854–862.

Mumby, P.J., Hastings, A. & Edwards, H.J. (2007). Thresholds and the resilience of

Caribbean coral reefs. Nature, 450, 98–101.

The Nature Conservancy (2017). Satellite Mapping of Benthic Habitats and Bathymetry

for Grenada, Caribbean.

The Nature Conservancy & Fisheries Division, G. (2007). Sandy Island/Oyster Bed

Marine Protected Area Management Plan.

Perera-Valderrama, S., Hernández-Arana, H., Ruiz-Zárate, M.-Á., Alcolado, P.M.,

Caballero-Aragón, H., González-Cano, J., et al. (2016). Condition assessment of

coral reefs of two marine protected areas under different regimes of use in the

north-western Caribbean. Ocean Coast. Manag., 127, 16–25.

Perry, C.T., Steneck, R.S., Murphy, G.N., Kench, P.S., Edinger, E.N., Smithers, S.G., et

al. (2015). Regional-scale dominance of non-framework building corals on

Caribbean reefs affects carbonate production and future reef growth. Glob. Chang.

Biol., 21, 1153–64.

Edmunds, P.J. (2016). Cryptic regime shift in benthic community structure on shallow

reefs in St. John, US Virgin Islands. Mar. Ecol. Prog. Ser., 559, 1–12.

Pomeroy, R.S., Watson, L.M., Parks, J.E. & Cid, G.A. (2005). How is your MPA doing?

A methodology for evaluating the management effectiveness of marine protected

areas. Ocean Coast. Manag., 48, 485–502.

Riegl, B. Purkis, S. (2015). Coral population dynamics across consecutive mass

mortality events. Glob. Chang. Biol., 21, 3995-4005.

Rocha, L.A., Pinheiro, H.T., Shepherd, B., Papastamatiou, Y.P., Luiz, O.J., Pyle, R.L.,

et al. (2018). Mesophotic coral ecosystems are threatened and ecologically distinct

from shallow water reefs. Science (80-. )., 361, 281–284.

Rowlands, G., Purkis, S., Riegl, B., Metsamaa, L., Bruckner, A. & Renaud, P. (2012).

Satellite imaging coral reef resilience at regional scale. A case-study from Saudi

Arabia. Mar. Pollut. Bull.

Scheibling, R.E., Patriquin, D.G. & Filbee-Dexter, K. (2017). Distribution and abundance

of the invasive seagrass Halophila stipulacea and associated benthic macrofauna

in Carriacou, Grenadines, Eastern Caribbean.

Selig, E.R. & Bruno, J.F. (2010). A Global analysis of the effectiveness of marine

protected areas in preventing coral loss. PLoS One, 5, e9278.

Selig, E.R., Casey, K.S. & Bruno, J.F. (2012). Temperature-driven coral decline: the role

of marine protected areas. Glob. Chang. Biol., 18, 1561–1570.

Shihavuddin, A.S.M., Gracias, N., Garcia, R., Gleason, A. & Gintert, B. (2013). Image-

based coral reef classification and thematic mapping. Remote Sens., 5, 1809–

1841.

Slattery, M., Lesser, M.P., Brazeau, D., Stokes, M.D. & Leichter, J.J. (2011).

Connectivity and stability of mesophotic coral reefs. J. Exp. Mar. Bio. Ecol., 408,

32–41.

Stevens, T. & Connolly, R.M. (2005). Local-scale mapping of benthic habitats to assess

representation in a marine protected area. Mar. Freshw. Res., 56, 111–123.

Stubler, A.D., Stevens, A.K. & Peterson, B.J. (2016). Using community-wide recruitment

and succession patterns to assess sediment stress on Jamaican coral reefs. J.

Exp. Mar. Bio. Ecol.

Stumpf, R.P., Holderied, K. & Sinclair, M. (2003). Determination of water depth with

high-resolution satellite imagery over variable bottom types. Limnol. Oceanogr., 48,

547–556.

Sunagawa, S., Woodley, C.M., Medina, M., Whitman, W., Coleman, D., Wiebe, W., et

al. (2010). Threatened corals provide underexplored microbial habitats. PLoS One,

5, e9554.

Thomas, S. (2016). National Biodiversity Strategy and Action Plan; Grenada.

Convention on Biological Diversity.

Turner, M., Vancouver, N. & Mitchell, K. (2009). Grenada Protected Area System Plan

Part 1 Identification and Designation Of Protected Areas.

Wilkinson, C. (2008). Status of coral reefs of the world: 2008.

Williams, S.M., Mumby, P.J., Chollett, I. & Cortés, J. (2015). Importance of

differentiating Orbicella reefs from gorgonian plains for ecological assessments of

Caribbean reefs. Mar. Ecol. Prog. Ser.

Williams, S.M., Sánchez-Godínez, C., Newman, S.P. & Cortés, J. (2016). Ecological

assessments of the coral reef communities in the Eastern Caribbean and the

effects of herbivory in influencing coral juvenile density and algal cover. Mar. Ecol.

Xin, L.H., Adzis, K.A.A., Hyde, J. & Cob, Z.C. (2016). Growth performance of Acropora

formosa in natural reefs and coral nurseries for reef restoration. AACL Bioflux, 9.

CHAPTER 3: THE MUSTARD HILL CORAL IN THE CARIBBEAN: LOCAL FACTORS INFLUENCING POPULATION DYNAMICS IN THE NEARSHORE WATERS OF GRENADA

3.1 ABSTRACT

Despite coral community collapse through much of the Caribbean in the last several decades, the mustard hill coral (MHC) is among the few species experiencing success. For this reason, we sought to update the current regional coverage status of this species using compiled monitoring data and assess what factors influence its populations at a local scale with transect data in Grenada. Coverage data was collected from 7 locations in the Caribbean covering the years 2000-2017 for use in correlation/regression analyses. The reefs of Grenada were selected to study the local influence of benthic factors, MPA status and river status on the abundance, size and coverage of mustard hill colonies. Four 30 m transects were photographed at 0.5 m intervals across four sites in Grenada and four sites in Carriacou in 2014 and 2017. Two sites on each island were within a MPA. Coverage, size, and abundance of colonies were assessed using ImageJ, and the coverage of all other biotic factors using stratified random sampling in Coral Point Count. The presence/absence of a MPA and/or river outflow were categorically coded (0,1). All locations (excepting Honduras) were found to have positive rates of change in MHC coverage, albeit with high interannual variation.

From 2014-2017 in Grenada the mean abundance of mustard hill coral colonies was significantly reduced (p<0.05), the mean colony size increased (p<0.05), and the overall coverage also increased (p>0.05). MHC coverage was significantly correlated with:

rubble (-), sand (-), pavement (+), macroalgae (-), coralline algae (+), sponges (+/-), gorgonians (-), generalist corals (+), weedy corals (-), and stress-tolerant corals (+).

Environmental conditions in the Caribbean continue to support populations of the mustard hill coral. On a local level, the species faces similar settlement and competitive challenges of other Scleractinian species.

3.2 INTRODUCTION

Mirroring the state of global tropical reef ecosystems, coral reefs in the

Caribbean have experienced dramatic declines in the last several decades, with live coral cover declining more than 80% since the 1970s (Gardner et al. 2003; Jackson et al. 2014). The combined impacts of stressors such as: rising sea temperatures, ocean acidification, overfishing, pollution, loss of herbivorous species, and disease outbreaks have been largely responsible for this decline (Hughes 2003; Anthony et al. 2011;

Castillo et al. 2012; Okazaki et al. 2016). However, these live coral cover declines have not impacted all Scleractinian species in the Caribbean equally, and have instead been concentrated along life-history and/or morphological lines (Perry et al. 2015).

Scleractinian species in the Caribbean can be broadly characterized as having one of four life history strategies, based on colony morphology, growth rate, and reproductive method (Darling et al. 2012). These life history strategies are: 1) weedy (high recruitment, fast growing, brooding reproduction); 2) stress tolerant (large colonies, slow growing, broadcast spawning); 3) competitive (framework, fast-growing, broadcast spawners); and 4) generalist (traits of all other categories) (Darling et al. 2012).

Competitive species such as Acropora palmata or Acropora cervicornis, responsible for

much of the complex framework building on Caribbean reefs experienced the most dramatic declines from 1970-present (Schutte et al. 2010; Rodríguez-Martínez et al.

2014). Non-framework building Scleractinian species with a ‘weedy’ life history strategy have proven successful in the region in an era of rapid environmental change, and have been among the few ‘winners’ on Caribbean reefs (Green et al. 2008). Species exhibiting this life history, such as the mustard hill coral (MHC), have increased in numbers dramatically in the region. One study showing that the MHCs’ proportion of live coral cover now ranges from 16%-72% in Caribbean reef communities, and its coverage has continued to increase in the last decade (Edmunds 2010). The MHC is found throughout the Caribbean, along the coasts of Florida, Bermuda, Brazil, and along the coasts of West Africa and is most abundant at depths of 0.5 m-15 m (Figure 3.1).

Figure 3.1 A group of mustard hill coral colonies inhabiting fields of Acropora rubble (left) and a typical juvenile green morph colony (right). Both photographs were taken at a depth of 3 m.

The MHC tolerate a wide gradient of thermal, physical, and water quality stressors

(Baumann et al. 2016). Mustard hill coral populations are dominated by medium and large sized colonies, and colony success is typically high, making it very unlikely for large colonies to reduce in size once the have reached this growth stage (Edmunds

2010). The MHC has a brooding reproductive strategy and is capable of sexual or asexual reproduction (fragmentation) facilitating a high rate of colony production

(Goodbody-Gringley & de Putron 2016). The most recent regional population assessment for the MHC was completed in 2008 (Green et al. 2008).

For highly abundant species such as the MHC, it is presently unclear what factors influence their population dynamics in the reef communities of the Caribbean

(particularly at a local scale). It is also unclear what level of resilience a weedy life history strategy confers on a species to continued changes, necessitating decadal updates of population status in the region. For these reasons, our objectives were to provide a temporal update on the rate of coverage change in MHC populations throughout the Caribbean and to assess what factors influence the coverage of the species at a local level.

To assess regional status, we compiled survey data from 8 locations for correlation and regression analysis in the Caribbean: Belize, Bonaire, United States

Virgin Islands (USVI), Florida, Guadeloupe, Honduras, Jamaica, and the Bahamas. To assess temporal changes in MHC populations and the extent of influence a gradient of biotic and abiotic factors may have on coverage, the tri-island state of Grenada was selected. To our knowledge, this is the first thorough attempt to assess the factors

influencing local population dynamics in the MHC and the most recent evaluation of its population status.

3.3 METHODOLOGY

All statistical analyses in this study were conducted using Minitab 18 Software with significance determined at a p-value <0.05.

3.3.1. REGIONAL DATA COLLECTION

Absolute coverage data of the MHC was obtained from existing survey information collected by research stations, NGOs, and individual researchers (Table 1 &

Table 3, Appendix B). Parameters set to filter data for use in the analysis were as follows: 1) data needed to have been collected since the year 2000; 2) surveys must have taken place at a depth of ≤ 15 metres; 3) surveys must have been conducted in nearshore reef habitats; 4) sites were within the wider Caribbean region; 5) site data covered a period of at least 3 years; and 6) survey locations were widely distributed across the region. Data was successfully collected from a total of 7 locations in the

Caribbean region (Table 3.1; Table 2, Appendix B) with varying numbers of transects

(Table 1, Appendix B).

Table 3.1 Summary of mustard hill coral survey locations and years for which survey data was available. Unites States Virgin Islands listed as ‘USVI’.

Site: Survey Years: Bahamas 2007, 2008, 2009, 2011, 2012,2013 Belize 2011,2013 Bonaire 2004, 2005, 2007, 2009, 2011, 2013, 2015, 2017 Grenada 2008-2012, 2014, 2015, 2017 Honduras 2010, 2011, 2012, 2013, 2014 USVI 2003-2017 Jamaica 2012, 2013, 2015

3.3.2 REGIONAL STATISTICAL ANALYSIS

In cases where data was available for multiple sites and/or transects within each location, this data was compiled into a single MHC coverage value for each year by taking the mean of all transects (Table 1, Appendix B). Where possible, standard error values were calculated for each year of available data for each location (Table 1;

Appendix B). All collected absolute coverage data was Arcsine transformed to facilitate analysis (Anderson et al. 2014). Spearman-rho correlations were used to test for any associations between the absolute cover (mean of collected transect data for each location) of P. astreoides and time at each of the 7 locations. Where a significant relationship was detected, a linear regression analysis was used to obtain the best-fit linear relationship. The slope of which was used as an estimate of the rate of change in absolute cover of P. astreoides at each location. An ANOVA with Bonferroni post-hoc analyses also conducted to assess differences between the locations themselves.

3.3.3 GRENADA EXPERIMENTAL DESIGN

A total of eight nearshore sites were selected within the tri-island state of

Grenada for an assessment of local benthic community structure. Four sites off the island of Grenada (Dragon Bay, Flamingo Bay, Northern Exposure, and Quarter

Wreck), and four off the island of Carriacou (Sandy Island, Whirlpool, Seaview, and

Jack Adam) were surveyed (Figure 3.1). On each island two sites were within a marine protected area (MPA) and two had no legislated protection (Figure 3.1). Transects delineated with rebar were previously established at the four sites on Grenada proper by Anderson et al. (2012). New transects on the island of Carriacou were established by this research group at sites that were visually similar to the sites in Grenada

(Horricks et al. 2018). Transects at each site were 30 m long and spaced 5 m apart.

Figure 3.2 Location of survey sites in Grenada and Carriacou including MPA boundaries.

3.3.4 GRENADA DATA COLLECTION

Dive surveys were conducted in 2014 and repeated in 2017 at all sites, with each transect being photographed in its entirety at 0.5 m intervals using a 0.25 m2 quadrat. A Nikon D7100 camera (Sigma 8 – 16 mm wide angle lens in an Ikelite housing) was used, and all photos were taken approximately 0.5 m above the benthos.

A total of 3,840 quadrat photos were obtained from these surveys for analysis.

3.3.5 GRENADA QUANTITATIVE MEASUREMENTS

Two methods of coverage assessment were used in the present study, one to obtain more detailed information on the MHC, and another on the coverage of all other benthic species or species functional groups. Each of the photographs could be standardized for analysis by setting the x-y distances of each quadrat at 0.5 m. ImageJ software (Abramhoff et al. 2004) was used to assess absolute MHC coverage individually in quadrat photos with the area of each individual MHC being measured

(Figure 3.3). Standardized quadrat photos were uploaded into Coral Point Count with

Excel extensions (CPCe; Kohler & Gill 2006) to estimate the absolute coverage for all other benthic species or species functional groups. Within each quadrat photograph, 18 points were distributed in a stratified random pattern (3x3 grid, with 2 randomized points per cell) (Horricks et al. 2017). This was determined to be sufficient following a species accumulation curve analysis (Horricks et al. 2017). Coverage at each point was then identified down to the species level for hard coral. Algae was identified as either macroalgae, turf algae, or coralline algae. The living group, coralline algae, was a compilation of all coralline algae species. This was also the case for macroalgae, with an addition of Dictyota sp. and turf algae. The coverage of cyanobacterial mats was not assessed during cover analysis of the quadrat photos. Remaining live benthos were categorized as sponges, gorgonians, zoanthids, and other-living. Bottom habitat of rubble, sand, and pavement was also categorized.

Figure 3.3 Example quadrat photo with mustard hill coral colony indicated by circle.

3.3.6 GRENADA STATISTICAL ANALYSIS

Abundance, size and coverage data were transformed using square root, log, and arcsine square root functions respectively to achieve normality for use of parametric tests in the analysis. A scatterplot analysis of residuals vs. predicted values for all factors was conducted to ensure homoscedasticity. Changes in mean coverage

(%), mean abundance (# of colonies), and mean colony size (cm2) of the MHC by year and islands were assessed using a two-way ANOVA. Site-specific changes from 2014 to 2017 in coverage, abundance, and size were assessed using Welch’s t-tests. Size

frequency distributions of mean colony size were created for Carriacou and Grenada and a Kolmogorov-Smirnov two-sample t-test was used to assess differences in the distributions between 2014 and 2017.

Marine protected areas status (MPA) and river status (RS) were included in the analysis as categorical factors. Each were coded as absent (0) and present (1), with river mouths counting as present when an outflow was measured to be within one kilometre of a transect. The sea temperature data and storm data for each site were excluded. The available satellite data was at such a course scale (i.e. 1-5 km2 per pixel) that no differences between sites could be assessed. Coverage values of each variable were arcsine square root transformed to ensure normality for the regression. An outlier test (Grubbs’ test) was conducted on each variable to remove any identified outliers from correlation analysis. A scatterplot analysis of residuals vs. predicted values for all factors was conducted to ensure homoscedasticity. Based on previous analysis by

Horricks et al. 2017 species assemblages varied between surveyed islands and years.

For this reason, a correlation was conducted on pooled data for each island (2014 and

2017), for a total of four correlation analyses. Only variables present in a minimum 123 quadrats were included in each correlation analysis. This value was based on established sample size calculations for estimated correlation coefficients of 0.2 (Bonett and Wright 2000). Pearson correlations were conducted to assess any statistically significant associations between the coverage of the MHC and habitat type (sand, rubble, and pavement), as well as other biotic factors (stony corals, sponges, algae, and gorgonians; Table 3.2). Spearman-rho correlations were conducted between MHC coverage and the presence of MPAs and river outflows (Table 3.2).

Table 3.2 Summary of all factors measured in quadrat surveys for possible inclusion in the correlation analyses.

Factor Acropora cervicornis Millipora complanata Acropora palmata Millipora squarrosa Acropora prolifera Montastraea annularis Agaricia agaricites Montastraea cavernosa Agaricia fragilis Montastrea faveolata Agaricia grahamae Montastrea franksi Agaricia lamarcki Mussa angulosa Agaricia tenuifolia Mycetophyllia aliciae Agaricia undata Mycetophyllia danaana Colpophyllia breviserialis Mycetophyllia ferox Colpophyllia natans Mycetophyllia lamarckiana Coralline Algae Oculina diffusa Dendrogyra cylindrus Porites branneri Dichocoenia stellaris Porites divaricata Dichocoenia stokesi Porites furcata Diploria clivosa Porites porites Diploria labyrinthiformis Scolymia cubensis Diploria strigosa Scolymia lacera Eusmilia fastigiata Siderastrea radians Favia fragum Siderastrea siderea Gorgonians Solenastrea bournoni Isophyllia sinuosa Solenastrea hyades Leptoseris cucullata Sponges Macroalgae Stephanocoenia michelinii Madracis decactis Tubastraea aurea Madracis mirabilis Zoanthids Manicina areolata Marine Protected Area Status Meandrina meandrites River Status Millipora alcicornis

3.4 RESULTS

3.4.1 REGIONAL ASSESSMENT

A high degree of interannual variation within the geographical locations was observed. Mustard hill coral coverage had a positive rate of increase across time in 6 of

7 studied locations (Positive spearman-rho correlation coefficients; Bonaire, USVI,

Jamaica, Bahamas, Belize, and Grenada). The positive relationship between MHC coverage and time was only found to be statistically significant in the USVI and

Bahamas (p<0.05; Table 3.3). The relationship between MHC coverage and time was not observed to be positive or negative in Honduras due to high interannual variation.

An increase in MHC coverage in 2012/2013 was observed in Jamaica, Bahamas,

Honduras, and Belize (Figure 3.4). Quality of available survey data prevented the assessment of statistical significance for this observed trend.

Table 3.3 Magnitude of the relationship between mustard hill coral coverage and time at seven locations in the Caribbean (Spearman correlation). Statistical significance is indicated with an asterisk.

Location Correlation (ρ) Bonaire 0.524 USVI 0.782* Honduras 0 Jamaica 0.5 Bahamas 0.886* Belize 1 Grenada 0.321

Figure 3.4 Relationship of absolute mustard hill coral coverage (%; Y axis) and time (year; X axis) across seven sites in the Caribbean (moving averages with dotted lines). Note the rapid increase detected in 2012/2013 coverage at several locations. Standard error bars are present for each coverage datapoint.

Linear regression analyses were conducted on sites where statistically significant correlations were identified (USVI and Bahamas). Elapsed time was found to explain

59% of the variation of MHC coverage data in the USVI and can be used to project coverages into the future (linear regression model; p<0.05). In the USVI the absolute coverage of the mustard hill coral is therefore forecasted to increase at a rate of 0.08%

100

per year. Projections of MHC coverage into the future could not be made with statistical confidence in the Bahamas (linear regression model; p>0.05).

3.4.2 MUSTARD HILL CORAL IN GRENADA

Mean percent coverage of the mustard hill coral was found to increase at all

sites in Carriacou and Grenada from 2014 to 2017 (Figure 3.5). Statistically

significant increases were only observed at the sites: FLAM, WHIRL, and JAD

(p<0.05; Figure 3.5). Absolute MHC coverage ranged from a minimum of 0% to a

maximum of 38.9% in surveyed quadrats.

* 10 *

* Mean Coverage MeanCoverage the of MHC(%) 5

0 DRAG FLAM QUART NORTH SAND WHIRL JAD SEAV Grenada Carriacou

2014 2017

Figure 3.5 Observed changes in mean transect coverage of the mustard hill coral from 2014 to 2017 in Carriacou and Grenada, with standard error bars. Statistical significance indicated with asterisks.

101

Mean colony size of the MHC significantly increased at all sites in Carriacou and Grenada from 2014 to 2017 (p<0.05; Figure 3.6). Measured colony sizes from all surveyed quadrats ranged from a minimum of 0.2 cm² to a maximum of 1600 cm².

160 * *

²) 140

120 * 100 * * 80 * *

60 * 40

MeanColony of Size theMHC (cm 20

0 DRAG FLAM QUART NORTH SAND WHIRL JAD SEAV Grenada Carriacou

2014 2017

Figure 3.6 Observed changes in mean colony size of the mustard hill coral from 2014 to 2017 in Carriacou and Grenada, with standard error bars. Statistical significance indicated with asterisks.

Mean colony abundance of the MHC significantly decreased at all sites in

Carriacou and Grenada from 2014 to 2017 (p<0.05; Figure 3.7). Mustard hill coral colony abundance within quadrats ranged from a minimum of 0 to a maximum of 134.

102

35 * 30

10 * * * * * * *

MeanColony Abundance of theMHC 5

0 DRAG FLAM QUART NORTH JAD SAND WHIRL SEAV Grenada Carriacou

2014 2017

Figure 3.7 Observed changes in mean colony abundance of the mustard hill coral from 2014 to 2017 in Carriacou and Grenada, with standard error bars. Statistical significance indicated with asterisks. The size-frequency distributions of MHC colonies in both Carriacou and

Grenada changed from 2014 to 2017 (Kolmogorov-Smirnov 2-sample tests; p<0.05).

These changes were observed to be a shift towards fewer, larger colonies in both islands from 2014 to 2017 (Figure 3.8).

103

Figure 3.8 Changes in size frequency distributions of mustard hill coral colonies in Carriacou (top) and Grenada (bottom) from 2014 (blue) to 2017 (red).

104

Statistically significant correlations between MHC coverage and other biotic factors were identified in Carriacou and Grenada for both 2014 and 2017 (p<0.05; Table

3.4). A total of four correlation analyses were conducted; one for each island (Grenada and Carriacou) at each time point (2014 and 2017). Only factors that were present in a minimum of 123 quadrats were included in each correlation analysis based on established sample size calculations. Some factors were therefore not present in all four analyses. Mustard hill coral coverage was significantly associated with sand (-), rubble

(-), and coralline algae (+) in all 4 analyses, (p<0.05; Table 3.4). Mustard hill coral coverage was found to be positively associated with sponges in Carriacou, and negatively associated in Grenada (this relationship was not statistically significant in

Grenada 2017) (Table 3.4). Mustard hill coral coverage was found to be significantly associated with macroalgae (-) in 3 of 4 correlation analyses and (+) in 1 of 4 analyses

(p<0.05; Table 3.4). Mustard hill coral coverage was found to be significantly associated with gorgonians (-) in all 3 analyses where they had sufficient numbers to be included (p<0.05; Table 3.4). Siderastrea siderea and Orbicella annularis were the only

Scleractinian species present in all four analyses, but no significant correlations were found (p>0.05; Table 3.4). Porites divaricata was present in 3 of 4 analyses and was negatively correlated with MHC coverage, although significance was only observed in 1 of the 3 analyses (Table 3.4). Other Scleractinian species were only found in 1-2 of the analyses with varying correlation relationships and statistical significance (Table 3.4).

Several species (i.e. Madracis auratenra) were only found on a single island.

Mustard hill coral coverage was found to be lower within MPAs in Carriacou and

Grenada (Spearman correlation; p<0.05; Table 3.4). Lower MHC coverage was

105

associated with rivers in 2014, and higher coverage in 2017, with no statistical significance being found (Table 3.5; Spearman correlation; p>0.05).

Table 3.4 Magnitude of the relationship between mustard hill coral coverage and: 1) habitat type, 2) biotic factors, and 3) MPA status (Pearson correlations; Spearman correlation for MPA status). Significant correlations are indicated with an asterisk. Carriacou Grenada 2014 2014 Factor r/ρ p-value Factor r/ρ p-value Rubble -0.168 0.0001* Rubble -0.101 0.002* Pavement 0.193 0.0001* Sand -0.122 0.0001* Sand -0.168 0.0001* Madracis auretenra -0.127 0.0001* Diploria strigosa 0.159 0.0001* Orbicella annularis 0.047 0.148 Orbicella annularis -0.031 0.345 Porites divaricata -0.006 0.844 Porites porites -0.118 0.0001* Porites porites -0.048 0.143 Siderastrea siderea 0.013 0.685 Siderastrea siderea -0.063 0.054 Gorgonians -0.095 0.003* Gorgonians -0.156 0.0001* Sponges 0.259 0.0001* Sponges -0.128 0.0001* Macroalgae -0.288 0.0001* Macroalgae 0.071 0.03* Coralline Algae 0.199 0.0001* Coralline Algae 0.099 0.002* MPA -0.188 0.0001* MPA -0.066 0.044* 2017 2017 Factor r/ρ p-value Factor r/ρ p-value Rubble -0.196 0.0001* Rubble -0.073 0.027* Sand -0.145 0.0001* Sand -0.114 0.001* Colpophyllia natans 0.126 0.0001* Madracis auretenra -0.101 0.002* Orbicella annularis -0.044 0.179 Orbicella annularis -0.03 0.364 Porites divaricata -0.078 0.017* Orbicella franksi 0.088 0.008* Siderastrea siderea -0.026 0.417 Porites divaricata -0.049 0.141 Sponges 0.199 0.0001* Siderastrea siderea -0.017 0.61 Macroalgae -0.238 0.0001* Gorgonians -0.14 0.0001* Coralline Algae 0.116 0.0001* Sponges -0.057 0.083 MPA -0.149 0.0001* Macroalgae -0.083 0.012* Coralline Algae 0.119 0.0001* MPA -0.115 0.0001*

106

Table 3.5 Magnitude of the relationship between mustard hill coral coverage and the presence of river outflows. Lower coverage was associated with rivers in 2014, and higher in 2017 (Spearman correlation; p>0.05).

3.5 DISCUSSION

The composition of coral reef communities in the Caribbean had remained relatively stable for much of the last 150,000 years, with species of all life history categories and taxa persisting today (Greenstein et al. 1998). However, in the last several decades increases in anthropogenic stressors such as ocean warming, ocean acidification, overfishing, and pollution has led to dramatic changes in the Caribbean with live coral cover declining by 80% since the 1970s (Gardner et al. 2003; Burke & Maidens

2004; Neal et al. 2017). These reefs have undergone a massive shift from a stable state of communities dominated by framework building corals to one that is dominated by macroalgae and non-framework building species (Aronson et al. 2004; Perry et al. 2015).

Corals with a weedy life history strategy largely make up the non-framework building species that are increasing in the region. Weedy species, like the MHC, can be broadly defined as short-lived, brooding corals that rapidly bounce back from disturbances and settle areas of disturbed reef (Baumann et al. 2016). Weedy coral species are among the most successful in present day reefs, being more abundant than Acropora or Orbicella species in up to 73% of sites surveyed (Green et al. 2008). The MHC in particular is common throughout its range and is one of the few Scleractinian species whose populations seem to be increasing in the face of regional climate change in the last

107

decades (IUCN 2000). It is unclear to what extent the success of the MHC will remain possible given the rapidly accelerating physical impacts of climate change, and therefore an updated regional population analysis was necessary given it has been more than a decade since the most recent assessment (Green et al. 2008). In addition, factors influencing MHC settlement patterns and relative success on a local level are relatively understudied in the Caribbean. Therefore, an island typical of reef conditions in the region was selected; Grenada.

Grenada has a total of 78 km2 of reef habitat and 29 km2 of seagrasses (both patch and fringing) including 33 identified species of stony coral from 10 families (Burke &

Maidens 2004; Nature Conservancy & Fisheries Division 2007). Mirroring other areas of the Caribbean, Grenada has been experiencing shifts in species assemblages in its reefs over the last several decades. Reef health in Grenada has been assessed by the

Government of Grenada and the Nature Conservancy as being in a ‘poor’ condition with a Reef Health Index ranking of 2.5/5 (Grenada’s Coral Reef Report Card 2016). The current stony coral coverage on the main island of Grenada is approximately 14%, a decline from an estimated 42% in 2003, but relatively stable compared with an estimated coverage of 13% from 2008 (Anderson et al. 2014; Green et al. 2008). The assessment of local factors influencing MHC in Grenada provided novel information on the ecological dynamics of this species. This may help inform predictions on its possible future success in the Caribbean region.

Of total living coral cover in the Caribbean, an estimated 16-72% is comprised solely of the MHC (Edmunds 2010). Mustard hill coral populations rely on their morphology to withstand physical disturbance and reproductive methods to withstand

108

long-term disturbance in the Caribbean (Darling et al. 2012, 2013; Manzello et al. 2015;

Baumann et al. 2016; Zinke et al. 2018). Increasing coverage trends of the MHC in the

Caribbean found in the present study were also identified in previous assessments of the species in the region (Green et al. 2008). Increasing MHC coverage was observed at six of seven locations, and the species therefore still appears to be experiencing success at a regional level (Green et al. 2008). At only two of these locations, the USVI and the Bahamas, there was a statistically significant correlation between MHC coverage over time. This was not unexpected as these locations were also those with the most extensive annual monitoring datasets available for inclusion in the analysis.

Once linear regression models for MHC coverage and time were performed, however, this relationship was only statistically significant for the USVI. Limited data availability made the temporal analysis of trends problematic in several cases (i.e., Jamaica and

Belize).

High interannual variation observed in survey locations was not unexpected given the temporal limitations of collected data, and is in line with present understanding of factors causing annual shifts in coral populations. Abnormal sea surface temperature events, nutrient loading, disease outbreaks, and repeated storm damage can act to rapidly shift the composition of reef communities on a local scale (Mcgill et al. 2005;

Murdoch 1991; Darling et al. 2013, 2017; Baumann et al. 2016). Extremely high interannual variation of MHC coverage in Honduras is of note and was the only location where the final measurement was less than when the monitoring began. Despite high interannual variation, all other locations had MHC coverage increase over time. More extensive data collection in the future will be necessary to improve the ability of

109

researchers to extract data for population trends of the MHC and limit the impact of high interannual variation (Miller et al. 2016).

Differences in coverage and rates of change between locations were also observed although statistical comparison was not possible given the temporal variability of monitoring data for each site. Colonies of the MHC have differing growth rates across its large geographic range that vary based on environmental conditions (Elizalde-

Rendón et al. 2010). For instance, colonies in the Gulf of Mexico have reduced growth rates when compared to the rest of the Caribbean (Elizalde-Rendón et al. 2010).

A notable regional recruitment event was observed in 2012/2013 in Jamaica,

Honduras, Belize, and the Bahamas with a spike in MHC coverage followed by a rapid decline. Given the substantial geographic spread of these locations and the similarity in coverage changes during this event, a regional shift in environmental conditions during that period was likely responsible (Schutte et al. 2010; Darling et al. 2012; Manzello et al. 2015; Neal et al. 2017; Zinke et al. 2018). Further study into climatic conditions from this time period (i.e. bleaching history, El-Nino, La-Nina) and the changes in coverage of other coral species could provide more insight into this unexpected occurrence (Mumby et al. 2007; Ennis et al. 2016; Heron et al. 2016).

The nature of the significant and rapidly changing conditions on present day reefs means that the possession of certain life-history strategies are more beneficial for some coral species, and are detrimental to others. The traits of weedy corals in particular (such as the MHC) has facilitated their success throughout the Caribbean

(Karlson & Hurd 1993; Rodríguez-Martínez et al. 2014). Changing environmental conditions can impact juvenile recruitment and/or success of individual colonies, and

110

has been observed to positively impact MHC populations in the region (including

Grenada) in the last several decades (Karlson & Hurd 1993; McClanahan & Maina

2003; Anthony & Connolly 2004; Hughes & Tanner 2013). Observed changes of MHC populations in Grenada and Carriacou from 2014 to 2017 appears to reflect this given that absolute coverage increased at all sites in Carriacou and Grenada (only statistically significant increases occurred in FLAM, WHIRL, and JAD; p<0.05).

Plotted size frequency distributions (2014 vs. 2017) in Grenada and Carriacou revealed that local populations were dominated by juvenile colonies of the MHC (<40 cm2) in both 2014 and 2017, with a shift towards fewer, larger colonies in 2017. In addition, observed large colonies (i.e. >400 cm2) in 2014 were also present in 2017.

This appears to support existing research that MHC populations (as in other weedy coral species) are largely influenced by recruitment. The observed persistence of large colonies appears to be supported by previous research that medium/large sized MHC colonies typically experience high individual success, making it very unlikely for large colonies to reduce in size once they have reached this growth stage (Edmunds 2010).

Statistically significant declines in MHC abundance were observed at all sites in Carriacou and Grenada from 2014 to 2017. In addition, statistically significant increases in MHC colony size were observed at all sites in Carriacou and Grenada for the same time period. There is a possibility these two phenomena may simply be the result of low recruit success in juvenile life history stages coupled with normal MHC colony growth in more mature colonies. The growth rate of the MHC is comparable to other Scleractinian corals (1-2 mm/year), with a single colony reaching maturity at approximately 8-10 years of age (Madin et al. 2016). However, given the similarity in

111

magnitude of measured changes in abundance and size it is also possible that numbers of juveniles are routinely high (i.e. life history strategy of the MHC) and these individuals fuse either with each other or adult colonies (Forsman et al. n.d.). Colony fusion is particularly likely given that juvenile MHC are highly likely to be genetically related to one another since inbreeding through self-fertilization is common in this species

(Chornesky and Peters 1987). Lastly, these patterns could also be reflective of a large- scale recruitment event following a disturbance that took place prior to monitoring.

Mustard hill corals are commonly most abundant in areas of high disturbance, such as areas with: overfishing, high storm frequency, and bleached areas following high SST events (Riegl & Purkis 2015). The most recent large-scale disturbance event in the area was tropical storm ‘Chantal’; eight months prior to the first survey in 2014 whose path moved through Grenada (National Weather Service 2013).

Much of the research on the MHC in the Caribbean has focused only on its absolute coverage and not the competitive factors that limit colony settlement and growth at a local level. A generalised summary of correlation analyses between MHC coverage and other benthic factors can be seen below in Table 3.6.

112

Table 3.6 Summarized outcome from repeated correlation analyses of mustard hill coral coverage with other benthic factors. Factors found in all four analyses are indicated with asterisks. Factors with inconsistent statistically significant correlations are shaded black.

Findings regarding habitat preference of the MHC are in line with other

Scleractinian coral species, being significantly correlated with pavement (+), sand (-), rubble (-), macroalgae (-), and coralline algae (+) (Table 3.6) (Schutte et al. 2010; Vega

Thurber et al. 2012a; Cameron et al. 2016; Ritson-Williams et al. 2016a). Mustard hill coral larvae are unable to settle on soft sandy substrate as a result of the constant physical disturbance resulting in poor attachment and smothering (Ross et al. 2013;

Olsen et al. 2016; Ritson-Williams et al. 2016a). Rubble habitats are also a habitat ‘sink’ for coral recruits as they provide a hard substrate that can facilitate initial attachment

113

and growth, but any elevated physical disturbance can easily shift rubble, smothering and destroying colonies (Fava et al. 2009; Hughes & Tanner 2013; Ross et al. 2013;

Cameron et al. 2016; Olsen et al. 2016; Ritson-Williams et al. 2016a). Mustard hill coral larvae are easily able to settle and thrive on pavement habitat since it is composed of permanent hard surfaces resistant to physical disturbance (Cameron et al. 2016;

Ritson-Williams et al. 2016a). The observed negative relationship between MHC coverage and macroalgae was expected given that Scleractinian corals globally are unable to compete with macroalgae (Vega Thurber et al. 2012b). As environmental conditions continue to support substantial growth of macroalgae, corals cannot compete, and this has resulted in shifts from coral dominated communities to those dominated by macroalgae (Vega Thurber et al. 2012a; Gilby et al. 2015; Olsen et al.

2015). For this reason, herbivory plays such an important role in tropical reefs. These species are essential to coral juvenile survival, keeping areas clear of macroalgae that would otherwise smother colonies (Burkepile et al. 2010; Vega Thurber et al. 2012b).

Not all herbivores impact mustard hill colony success equally; ocean surgeonfish have been found to increase growth of P. astreoides, with parrotfish having no impact.

However, herbivorous fish biomass in Grenada and Carriacou has been assessed by

TNC as being in a state of ‘poor’ health (Grenada’s Coral Reef Report Card 2016).

Herbivorous invertebrates such as Diadema antillarum are also now largely absent from reef communities in Grenada at abundances as low as 0.1 individuals/100 m²

(Anderson et al. 2014). Due to the loss of urchin and herbivorous fish populations, abundance of Dictyota and Halimeda macroalgae species have been on the rise and are the now the dominant form of benthic cover in Grenada, increasing from 41% cover

114

in 2009 to 74.2% in 2011 (Wilkinson 2008a; Anderson et al. 2014). Early monitoring survey results collected by Horricks et al. (2017), has shown that localized declines in macroalgae in Grenada are beginning to occur, perhaps a result of MPA establishment.

It is of note that site SEAV in Carriacou (2014 and 2017) had the highest MHC coverage, the lowest coverage of macroalgae, and the numbers of Diadema antillarum were observed to be much higher than all other sites (sometimes as high as 24/m²; personal observations in this study). Further study on this site is warranted, particularly in regards to the apparent health of herbivorous invertebrate populations. The positive relationship observed between MHC coverage and coralline algae was also in line with current understanding of settlement dynamics in Scleractinian coral species. Coralline algae does not outcompete coral recruits, and attracts their settlement with chemical signatures in the water (Done 1999; Ritson-Williams et al. 2016). In addition, coralline algae also attracts the larvae of herbivorous invertebrates such as D. antillarum which in turn creates low macroalgal conditions in which coralline algae and coral recruits persist

(Peter J. Edmunds 2016; Ritson-Williams et al. 2016a).

Given the limited number of available sites to compare, the lack of significant correlation between river status and MHC was not unexpected. A further expansion of survey sites and water quality sampling may improve the assessment of this relationship. The significant negative correlation observed between MHC coverage and

MPA status was surprising but likely unrelated to the presence of the protected areas.

The extremely high coverage of the MHC in SEAV skewed coverage values outside of

MPAs in Carriacou, and the only MPA assessed in Grenada has substantial river

115

outflows present, which may impact the coverage of the MHC (despite not being observed in the correlation analyses).

Varying responses (+/-) of MHC coverage to sponges were observed between

Carriacou and Grenada (Table 3.6). Given that this group is a compilation of all observed sponges it is possible that the MHC has varying responses to individual sponge species and warrants possible further study. This is of particular significance as it has been forecasted that some Caribbean reefs may transition to a state of being dominated by sponges with the impacts of climate change (Villamizar et al. 2014; de

Bakker et al. 2017). The negative association between the MHC and gorgonians was not unexpected as their long-elongate form can act to block essential sunlight from many Scleractinian colonies (Williams et al. 2015). In addition, this appears to be in line with existing research on habitat types of reefs in the Caribbean, with coral framework/Orbicella reefs typically being classified as a distinctly different habitat from gorgonian plains (Williams et al. 2015; The Nature Conservancy 2017).

In Grenada today, the most common coral species by proportion of total stony coral species are P. porites, P. astreoides, O. annularis, and Madracis auratenra

(Wilkinson 2008a). In terms of interactions between the MHC and other species, statistically significant correlations were found with: Colpophylia natans (+); Diploria strigosa (+); M. auretenra (-); P. divaricata (-); P. porites (-); and O. franksi (+) (Table

3.6; Figure 3.9).

116

Figure 3.9 Scleractinian species whose coverage is significantly correlated with the mustard hill coral: a) Colpophyllia natans, b) Diploria strigosa, c) Madracis auretenra, d) Porites divaricata, e) Porites porites, and f) Orbicella franksi (Adapted from Coralpedia 2019).

It is of note that the only Scleractinian species found to be positively correlated with the MHC has a similar plate/massive colony growth pattern as opposed to the lobate or branching pattern of the other observed species (Figure 3.9). In addition, the life-history strategy of coral species may play a significant role in these observed

117

relationships. All species for which negative correlations with the MHC were identified possess a weedy life-history strategy (M. auratenra, P. porites, and P. divaricata). In addition, O. franksi (a species with a generalist life-history strategy) and stress tolerant species (C. natans, and D. strigosa) were the only corals found to have a positive correlation with the MHC. Weedy corals (such as the MHC) are commonly most abundant in areas of high disturbance levels (i.e. overfishing, high storm frequency, and bleaching events) and different species with this shared life-history strategy may therefore compete with each other for colony space within these environments (Riegl &

Purkis 2015). Stress tolerant corals can be either brooding or broadcast spawning and can better tolerate long-term environmental stressors, such as increasing sea temperature or changing pH conditions (Rachello-Dolmen & Cleary 2007). Large colony size is typical of stress-tolerant corals species, which may increase population resilience to changes over time (Moraes et al. 2006). For these reasons, colonies of stress tolerant coral species may be able to inhabit habitats normally occupied by weedy species (Karlson & Hurd 1993; Darling et al. 2012, 2013; Sabine et al. 2015). It is possible that some other factor allows these species to coexist within these habitats, such as varying settlement cues, or non-aggressive responses to colony contact. No competitive species were identified in this analysis as they have become the least abundant in the Caribbean, experiencing immense declines across the Caribbean, largely as a result of susceptibility to storm damage and coral disease (Greenstein et al.

1998; Aronson et al. 2004; Perry et al. 2015b). Generalist coral species are typically broadcast spawning corals that have a balance of weedy (disturbance tolerant), competitive, and stress tolerant traits and do not require specific conditions for their

118

survival (Darling et al. 2012; Baumann et al. 2016). Although only a small number of these species exist in the Caribbean they still persist in reef communities, despite comprising a small proportion of total coral cover (Gardner et al. 2003). Given that this group was positively correlated with the MHC (i.e. similar habitat preference), it is possible that other factors allow them to persist together such as varying settlement cues, or non-aggressive responses to colony contact. Further study regarding these preliminarily identified relationships is warranted to determine if this is merely a localised phenomenon or indicative of a relationship that takes place throughout the range of the

MHC. The addition of more survey sites and locations across a wide range of environmental conditions is essential as not all of the listed species were present in all correlation analyses due to localised distributions and/or limited observations.

In conclusion, the mustard hill coral continues to be one of the ‘winning’ coral species in Grenada and the Caribbean at large and may continue to experience success into the near future. The mustard hill coral faces similar settlement and competitive challenges as other coral species and its life-history strategy may be largely responsible for its success on a regional scale. Further work is needed to assess what magnitude of environmental stressors this species can tolerate.

3.6 ETHICS:

Research permit information: IACUC 14001. Grenada Ministry of Forestry, Fisheries and Agriculture & St. George’s University

3.7 AKNOWLEDGEMENTS

119

We would like to thank the Ontario Veterinary College, The Government of Ontario

(OGS), and St. George’s University for providing the necessary funding to make this project possible. In addition, Deefer Diving Carriacou, and EcoDive Grenada provided invaluable support in the field. We would like to thank Dr. Ryan Horricks and Dr. John

Lumsden for data collection in Grenada and Carriacou, and providing invaluable support in the drafting and review of this project. We would also like to extend thanks to Dr.

Lorenzo Alvarez-Filip, Dr. John Lumsden, Dr. Ryan Horricks, and Dr. Steven Schill for reviewing this project and providing recommendations for analyses, and lastly the numerous sources that provided survey data on mustard hill coral coverage from sites across the Caribbean.

120

3.8 LITERATURE CITED *References in format of ‘Ecology Letters’

Anderson, R., Morrall, C., Jossart, J., Nimrod, S., Bolda, E., Musser, K., et al. (2014).

Marine protected area monitoring in the nearshore waters of Grenada, Eastern

Caribbean: benthic cover and fish populations. Rev. Biol. Trop. (Int. J. Trop. Biol,

62, 273–286.

Anthony, K.R.N., Maynard, J.A., Diaz-Pulido, G., Mumby, P.J., Marshall, P.A., Cao, L.,

et al. (2011). Ocean acidification and warming will lower coral reef resilience. Glob.

Chang. Biol., 17, 1798–1808.

Bonett, D.G, Wright, T.A. (2000). Sample size requirements for estimating pearson,

kendall and spearman correlations. Psychometrika, 65, 23-28.

Castillo, K.D., Ries, J.B., Weiss, J.M. & Lima, F.P. (2012). Decline of forereef corals in

response to recent warming linked to history of thermal exposure. Nat. Clim.

Chang., 2.

Darling, E.S., Alvarez-Filip, L., Oliver, T.A., McClanahan, T.R. & Côté, I.M. (2012).

Evaluating life-history strategies of reef corals from species traits. Ecol. Lett., 15,

1378–1386.

Edmunds, P. (2010). Population biology of Porites astreoides and Diploria strigosa on a

shallow Caribbean reef. Mar. Ecol. Prog. Ser., 418, 87–104.

Gardner, T.A., Côté, I.M., Gill, J.A., Grant, A. & Watkinson, A.R. (2003). Long-term

region-wide declines in Caribbean corals. Science, 301, 958–60.

121

Green, D., Edmunds, P. & Carpenter, R. (2008). Increasing relative abundance of

Porites astreoides on Caribbean reefs mediated by an overall decline in coral

cover. Mar. Ecol. Prog. Ser., 359, 1–10.

Horricks, R.A. (2017). Tissue Regeneration of Artificially Induced Lesions in the

Caribbean Great Star Coral (Montastraea cavernosa) in the Nearshore Waters of

Grenada. University of Guelph.

Horricks, R.A., Herbinger, C.M., Lillie, B.N., Taylor, P. & Lumsden, J.S. (2018).

Differential protein abundance during the first month of regeneration of the

Caribbean star coral Montastraea cavernosa. Coral Reefs, 38, 45–61.

Hughes, T.P. (2003). Climate change, human impacts, and the resilience of coral reefs.

Science (80-. )., 301, 929–933.

Jackson, J., Donovan, M., Cramer, K. & Lam, V. (2014). Status and trends

of Caribbean coral reefs: 1970-2012.

Okazaki, R.R., Towle, E.K., van Hooidonk, R., Mor, C., Winter, R.N., Piggot, A.M., et al.

(2016). Species-specific responses to climate change and community composition

determine future calcification rates of Florida Keys reefs. Glob. Chang. Biol.

Perry, C.T., Steneck, R.S., Murphy, G.N., Kench, P.S., Edinger, E.N., Smithers, S.G., et

al. (2015). Regional-scale dominance of non-framework building corals on

Caribbean reefs affects carbonate production and future reef growth. Glob. Chang.

Biol., 21, 1153–64.

Rodríguez-Martínez, R.E., Banaszak, A.T., McField, M.D., Beltrán-Torres, A.U. &

122

Álvarez-Filip, L. (2014). Assessment of Acropora palmata in the Mesoamerican

Reef System. PLoS One, 9, e96140.

Schutte, V.G.W., Selig, E.R. & Bruno, J.F. (2010). Regional spatio-temporal trends in

Caribbean coral reef benthic communities. Mar. Ecol. Prog. Ser.

123

GENERAL CONCLUSIONS AND DISCUSSION

This research has provided insight into: 1) the current state of the literature for the MHC, 2) the distribution of benthic habitats within the Sandy Island- Oyster Bed

MPA, 3) preliminary trends of MHC status in the Caribbean, and 4) the population dynamics of the MHC in Grenada.

The mustard hill coral is not necessarily a species that is lacking in research effort. An increasing number of studies have focused their attentions on this species given that it is now one of the most abundant coral species in the Caribbean. However, to date no formal collection of the literature for this species has been conducted. This thesis provides a primer of the relevant research on the species from both a biological and ecological frame to guide further research and outline knowledge gaps.

This thesis for the first time created a map of benthic habitats within the Sandy

Island- Oyster Bed MPA in Grenada. Mapped spatial distribution of benthic habitats was found to be accurate, with limitations at a very fine spatial scale and between habitats with similar spectral signatures. The mapping software was found to be more effective for clearly defined habitats such as coral framework, sand, and dense seagrass. Other habitat types were not categorized in the same class as their respective benthic truthing points. In addition, imagery-based habitat mapping is largely limited to the upper photic zone due to rapid light refraction in water, leaving ecologically significant mesophotic communities out of the analysis. Following mapping accuracy analysis, the final mapping product was found to be of sufficient quality to support the hypothesis that

124

high-density truthing points could be used to create a benthic habitat map of the Sandy

Island-Oyster Bed MPA.

This thesis provided a preliminary temporal update of mustard hill coral population trends in the Caribbean. Coverage levels were observed to increase across six of the seven studied sites in the Caribbean, supporting the hypothesis that MHC coverage in the region has continued to increase since its last regional assessment in

2004. Analysis proved challenging given varying survey effort, methodology, and the temporal availability of survey data. In addition, there has been a shift towards reporting reef health as only ‘live coral cover’ to reduce analysis time and cost, making species level data collection far more difficult. Species data was therefore often lacking from data sets that were otherwise available. Further addition of data sources and locations would dramatically improve the confidence in these findings and lessen the impact of interannual variability.

This thesis assessed changes in populations of the MHC on the islands of

Grenada and Carriacou as well as interactions between the MHC and MPA and river status, as well as other biotic factors. Continued stability of MHC coverage was observed but this research also revealed further information on demographic makeup and recruitment of MHC populations in the area. Significant increases in colony size and decreases in colony abundance were also observed. These findings partially support the hypothesis that significant changes in mustard hill coral coverage, abundance, and size would occur between survey years in Grenada and Carriacou. MHC coverage was not found to change significantly between years, however. In addition, this thesis provides novel observations on the habitat preferences and competitive interactions of

125

the MHC. These observations supported the hypothesis that variation in mustard hill coral coverage can possibly be explained by a suite of local biotic and abiotic factors.

Limited spatial distribution of survey sites, limited temporal dispersion of data, lack of water quality samples, lack of adequate MPA enforcement, and highly localized distributions of certain coral species were factors that limited the outcomes and quality of these results. Further research is warranted in the area to fill in data gaps, address outlined challenges, and increase confidence in the findings of this thesis.

The following conclusions can be drawn from:

Chapter 2:

1. Benthic truthing points can be used to create an accurate benthic habitat map of

tropical coral reefs, with limitations regarding mesophotic communities, and

benthic habitats with similar spectral signatures. Mapping software was most

accurate in the SIOBMPA when identifying benthic habitat classes with the most

distinct spectral signatures, such as sand and coral framework.

2. The most prominent benthic habitat areas within the SIOBMPA were dense

seagrass (25.74%); deep-water (24.98%); and sand (23.32%).

3. Previous work on the distribution of benthic habitats in Carriacou dramatically

overestimated coral framework coverage on the eastern side of the island and

underestimated those within the SIOBMPA.

4. The benthic habitat map created of the SIOBMPA is of high enough accuracy to

inform management, guide research in the area, and conduct long-term

monitoring of benthic habitats within the protected area.

126

5. The SIOBMPA protects a significant proportion of the coral framework habitats in

Carriacou (19.5%) and will form an integral part of any MPA network in the area.

Chapter 3:

1. High interannual variation and limited standardized survey data availability

prevented a determination of regional trends with high confidence. Long-term

monitoring datasets utilizing standardized methods are essential to increase the

confidence in trends outlined by this study.

2. Using available data, it appears that MHC coverage has continued to increase in

the Caribbean, with a positive trend being observed in 6 of 7 locations.

3. A regional scale shift in environmental conditions likely took place in 2012/2013

in the Caribbean that influenced MHC populations in Jamaica, Bahamas,

Honduras, and Belize

4. Mustard hill coral coverage in Grenada remain healthy with increasing coverage

and may have recently experienced a substantial recruitment event, with

populations shifting towards fewer, larger colonies.

5. Mustard hill coral populations face similar habitat and interspecies competitive

challenges as other Scleractinian coral species, being negatively associated with

macroalgae, sand, and rubble; and positively associated with pavement and

coralline algae.

6. Life-history strategy may play a significant role in competitive interactions

between the MHC and other Scleractinian corals, being negatively associated

with weedy species, and gorgonians; and positively associated with stress-

tolerant and generalist species.

127

7. Study methods were inadequate to properly assess the impact of MPAs and river

outflows on the MHC in Grenada.

APPENDICES APPENDIX A

Figure 1 All benthic truthing points taken in Northern Grenada and Carriacou as part of the Nature Conservancy’s wider mapping initiative.

APPENDIX B

Table 1 Benthic survey data of absolute mustard hill coral coverage (%) at seven locations used for analysis of population trends in the Caribbean.

128

Location Total # of Transects Year % Cover SE Depth (m) Habitat Type Source 2001 1.015950294 0.174466547 <15 2002 1.198775284 0.196789437 <15 2003 1.894704228 0.229406976 <15 2004 1.685274916 0.217274055 <15 2005 1.654002595 0.158502447 <15 2006 1.053313308 0.11716895 <15 2007 1.838873321 0.256998112 <15 2008 1.806265124 0.166866347 <15 12 Patch Reef USVI Monitoring Program USVI 2009 1.917560832 0.23156309 <15 2010 1.92827097 0.20453973 <15 2011 2.453558505 0.268083824 <15 2012 2.558080737 0.323096483 <15 2013 1.780256617 0.153938358 <15 2014 2.68844566 0.30149759 <15 2015 2.191409515 0.231997621 <15 2016 2.351918656 0.242605265 <15 2010 1.6 0.033117635 <15 2011 1.990275862 0.445248704 <15 Honduras 67 2012 2.836285714 0.475539943 <15 Fore Reef AGRRA 2013 5.38775 0.986746145 <15 2014 1.56173913 0.191954639 <15 2012 1.4688 0.464475343 <15 Jamaica 23 2013 3.723909091 1.122800837 <15 Fore Reef AGRRA 2015 3.1 1.265014032 <15 2007 0.958333333 0.289180316 <15 2008 0.880571429 0.302205649 <15 2009 0.9625 0.294093345 <15 119 Fore Reef AGRRA Bahamas 2011 1.020141026 0.112079458 <15 2012 1.638833333 0.686877749 <15 2013 1.089068966 0.210975517 <15 2011 1.577611111 0.287255583 <15 2013 1.877733333 0.440504519 <15 Belize 20 Fore Reef AGRRA

2004 1.449166667 0.23860518 <15 2005 0.772959057 0.237773685 <15 2007 1.205712465 0.312644901 <15 2009 1.916443934 0.381244574 <15 12 Fore Reef Steneck et al. 2015 Bonaire 2011 1.302547774 0.26526377 <15 2013 1.744842774 0.294617399 <15 2015 1.488147758 0.333234422 <15 2017 1.527001397 0.304455052 <15 2009 1.818928615 0.18418442 <15 2010 2.238648037 0.260004423 <15 2011 1.672874787 0.226037383 <15 Anderson et al. 2014 and Horricks et al. 20 2012 1.89437043 0.235429527 <15 Patch Reef Grenada 2017 2013 1.60821424 0.237194282 <15 2014 2.09477486 0.235305156 <15 2017 3.923563334 0.258433707 <15

Table 2 Summary of benthic survey data for the mustard hill coral for Grenada and Carriacou in 2014 and 2017.

129

Table 3 GPS coordinates of all survey transects used for the regional analysis of pa.

130

Bahamas Belize Honduras Bonaire Latitude Longitude Latitude Longitude Latitude Longitude Latitude Longitude Latitude Longitude 20.92930 -73.38667 23.88732 -79.76373 16.41850 -88.04130 15.85823 -87.45339 12.193768 -68.30082 20.93535 -73.20407 23.92997 -80.47625 16.45676 -88.20930 15.86353 -87.49525 12.148736 -68.32824 20.94855 -73.15150 23.93244 -80.47313 16.50790 -87.97070 15.86541 -87.50019 12.219083 -68.352362 20.95075 -73.24772 23.95722 -80.44592 16.51746 -88.19352 15.86554 -87.50060 12.138625 -68.277028 20.97379 -73.68474 23.95870 -79.79111 16.64572 -88.06620 15.86654 -87.51341 12.171688 -68.289254 21.00912 -73.12897 23.96898 -80.38522 16.72000 -87.83870 15.86862 -87.51843 12.134542 -68.280663 21.01152 -73.70364 23.99049 -79.80727 16.76494 -88.07610 15.87149 -86.36769 12.125735 -68.288106 21.02854 -73.69021 24.00787 -80.30127 16.88532 -87.70304 15.87302 -87.52868 12.144418 -68.276761 21.03189 -73.66754 24.01154 -79.80262 16.88560 -87.81020 15.89740 -86.43987 12.164968 -68.287344 21.03934 -73.65972 24.01195 -79.81213 17.16920 -87.63170 15.91183 -87.61714 12.200347 -68.309258 21.08023 -73.64886 24.21819 -77.59741 17.19558 -87.93183 15.91781 -86.38764 12.125693 -68.28809 21.08957 -73.65251 24.23294 -77.60732 17.24508 -88.05223 15.91957 -86.55431 12.21959 -68.360165 21.13884 -73.57152 24.31187 -77.65392 17.31848 -88.04250 15.92113 -87.60552 Grenada 21.17747 -73.52147 24.33535 -77.66657 17.35678 -88.17087 15.95399 -86.51929 Latitude Longitude 21.19241 -73.43532 24.34761 -77.67262 17.38828 -88.04557 15.95947 -86.47297 12.56 -61.46 21.48096 -73.07515 24.48436 -77.69702 17.41428 -87.80936 15.98111 -86.47856 12.53 -61.45 21.66686 -73.81847 24.52161 -77.69156 17.43030 -87.45177 16.04326 -86.98087 12.14 -61.47 21.67358 -73.79047 24.63141 -77.69180 17.93406 -87.93876 16.05409 -86.97887 12.16 -61.46 21.69500 -73.85200 24.70200 -77.73993 18.08799 -87.86372 16.06433 -86.47906 12.22 -61.46 21.70200 -73.84700 24.71256 -77.74769 18.17546 -87.82800 16.06445 -86.49831 21.70942 -73.81771 24.73338 -77.76293 Jamaica 16.07962 -87.01417 22.33405 -73.00025 24.74547 -77.78220 Latitude Longitude 16.08498 -86.89317 22.37172 -73.17038 24.74589 -77.77978 16.83950 -78.08710 16.08995 -86.99433 22.43622 -73.00340 24.77186 -77.80107 16.86968 -78.10248 16.10096 -86.88094 22.57335 -73.60500 24.88698 -77.52393 16.88980 -77.99265 16.10348 -86.87947 22.57360 -73.46725 24.89468 -77.54015 16.93080 -77.91980 16.10483 -86.97240 22.61133 -73.56403 24.97425 -77.53535 16.93884 -77.83963 16.11180 -86.94998 22.62012 -73.53858 24.97475 -77.53522 17.00256 -77.78744 16.11266 -86.94912 22.62370 -73.63078 24.97956 -77.53529 17.03267 -77.53016 16.11880 -86.94077 22.62370 -73.62452 25.00542 -77.55110 17.03978 -77.68607 16.12157 -86.91516 22.70538 -73.81745 25.00743 -77.55133 17.04736 -77.51462 16.28918 -86.60270 22.74920 -73.86005 25.00822 -77.55397 17.04917 -77.72060 16.31880 -86.50160 22.75370 -74.05578 25.01408 -77.56664 17.76345 -77.05510 16.32145 -86.58442 23.05857 -73.75800 25.01535 -77.56976 17.79473 -77.07722 16.33441 -86.57124 23.07128 -73.78478 25.01565 -77.56968 17.79647 -77.06952 16.34072 -86.56174 23.07783 -73.64955 25.03990 -77.55331 17.79872 -77.00403 16.35783 -86.53253 23.08940 -73.82238 25.09311 -77.23346 17.79895 -77.01977 16.35835 -86.41237 23.09980 -73.67068 25.10426 -77.17758 17.79938 -77.04322 16.36675 -86.50686 23.10675 -73.77262 25.11876 -77.17767 17.80158 -77.07482 16.37378 -86.48287 23.42404 -79.57460 25.12923 -77.10692 17.80160 -77.07440 16.39414 -85.89658 23.43012 -79.65046 25.22235 -78.03273 17.82813 -76.98458 16.39824 -86.28248 23.48346 -79.85194 25.22997 -78.03657 17.83107 -76.92235 16.39906 -85.95850 23.49395 -79.88530 25.24057 -78.04220 17.83232 -76.92217 16.40711 -86.40711 23.49485 -79.89154 25.24977 -78.04247 18.13704 -78.03723 16.40841 -86.40711 23.55205 -77.33203 25.26412 -78.05293 18.17078 -78.05648 16.42614 -86.35575 23.55521 -79.57055 25.30747 -78.07867 USVI 16.42967 -86.09623 23.56202 -77.32854 25.31377 -78.08288 Latitude Longitude 16.43680 -86.26131 23.58089 -79.59282 25.31573 -78.07670 17.77388 -64.81350 16.44299 -85.95448 23.58253 -77.33286 25.35703 -78.10593 17.71097 -64.65221 16.44394 -85.95537 23.59069 -79.60298 25.39158 -77.82007 17.74337 -64.57160 16.45107 -86.13706 23.60288 -77.34055 25.39198 -77.82326 17.69116 -64.90008 16.46074 -85.82514 23.64930 -79.61062 25.39295 -78.12938 17.78530 -64.75940 16.47248 -85.82225 23.67619 -77.37133 25.91728 -77.27684 17.73400 -64.89540 16.48613 -85.91708 23.69257 -79.65529 26.59389 -78.39611 18.34450 -64.98595 16.49740 -85.90240 23.72853 -77.39325 26.59472 -78.38333 18.35738 -65.03442 17.39858 -83.93504 23.78030 -77.41680 26.59583 -78.39472 18.34403 -64.98435 17.40045 -83.92743 23.79921 -80.47932 26.59611 -78.37083 18.31257 -64.86058 17.40126 -83.93513 23.81126 -80.47852 26.59667 -78.38556 18.37425 -64.93438 17.40145 -83.92868 23.83794 -79.74358 26.59861 -78.37639 18.33797 -64.70402 17.40221 -83.96770 23.85814 -80.49263 18.31417 -64.76408 17.40240 -83.96818 17.41042 -83.91628 17.41054 -83.91261 17.41232 -83.94388 17.41315 -83.94455 17.41342 -83.90439 17.41401 -83.93833 17.41438 -83.93841

131