Based Fisheries Management on Coral Reef Communities of American Samoa

EVALUATION OF THE EFFECTS OF COMMUNITY- BASED FISHERIES MANAGEMENT ON CORAL REEF COMMUNITIES OF AMERICAN SAMOA

by Zoë Andrews Msc Marine Environmental Protection 2004

(Rare blue coral Heliopora, Ofu lagoon, American Samoa)

University of Wales, Bangor Supervisor Dr J. Turner School of Ocean Sciences University of Wales, Bangor ABSTRACT ______

EVALUATION OF THE EFFECTS OF COMMUNITY-BASED FISHERIES MANAGEMENT ON CORAL REEF COMMUNITIES OF AMERICAN SAMOA

ABSTRACT

This study aims to investigate the effects of protection on coral reef communities at CB MPA (Community-based Marine Protected Area), federal (Statutory) MPA and non MPA sites in American Samoa. The population in American Samoa is rapidly increasing which places an increased pressure on the marine environment. A community-based fisheries management programme (CBFMP), administered by the Department of Marine and Wildlife Resources (DMWR) was implemented in 2001. By establishing MPAs in the participating villages, the programme aims to improve sustainable development of marine resources by improving habitat quality and increasing fish abundance. Key objectives of the study were to provide a quantitative description of coral reef benthic communities in terms of substrate cover, abundance of coral species and key macro-invertebrates. The Point-Intercept transect method was performed to quantify the percentage cover by substrate and hard coral species along 5 replicate 25m transects at each survey site. The ecological software PRIMER was used to understand and document the present status of each benthic community surveyed by multivariate analysis techniques. A total of 42 species of coral were recorded; Massive Porites, Porites cylindrica, Porites rus, Pocillopora damicornis, Acropora microphthalma and Pavona frondifera comprised the majority of coral cover throughout the study area. These dominating branching and massive species reflect the extreme exposure to waves on the reef flat. Coral cover ranged from 11% at Aua control to 68% at Alofau reef flat and Fagatele Bay, and was found to be positively correlated with fish abundance. No significant difference was found between CB MPA and non MPA sites in reef substrate cover and coral species abundance. This suggests that location rather than protection is responsible for differences in coral cover at the sites. Furthermore, the success of MPA performance will vary on a case-by- case basis, depending on the social, environmental, biological and physical factors involved. The aim of this study was accomplished, however not all CB MPA sites were surveyed due to time constraints. CB fisheries management should continue to be pursued. It may not initially increase habitat quality, but it is important in sustaining and enhancing fish populations.

Zoë Andrews Bala Mill 21 Ffrydan Road Bala LL23 7RY [email protected]

i DECLARATION ______

DECLARATION

This work has not previously been accepted in substance for any degree and is not being concurrently submitted in candidature for any degree.

This dissertation is being submitted to the University of Wales, Bangor in partial fulfilment of the requirements for the degree of MSc Marine Environmental protection.

This dissertation is the result of my own independent work/investigation, except where otherwise stated. Other sources are acknowledged by footnotes giving explicit references. A bibliography is appended.

I hereby give consent for my dissertation, if accepted, to be available for photocopying and for interlibrary loan, and for the title and summary to be made available to outside organisations.

Signed by candidate ……………………...... Date …………………………………………...

ii ACKNOWLEDGEMENTS ______

ACKNOWLEDGEMENTS

I would like to thank Ray Tulafono and the staff of the Department of Marine and Wildlife Resources for providing the facilities and resources during the field work of this study. Without their help and support this study would not have been possible.

In particular, special thanks to Lesley Whaylen for organising and implementing many of the field logistics, and for all her help and enthusiasm in the field. I am also grateful to Doug Fenner for all his help and advice with coral identification and for providing many of the underwater photographs; Chris Hawkins for providing project ideas and initially setting up the project; and Troy Curry for his technical help and advice with GIS. My thanks also go to Risa Grace Oram, Francesca Riolo, Selaina Vaitautolu, Pora Toliniu, Peter Craig, and Charles Birkeland for their assistance and advice during my stay in American Samoa.

I would like to thank my supervisor John Turner for his help and guidance in the preparation of this project. My deepest thanks go to Mark for his continual encouragement, support and patience, and for keeping my spirits high through such a challenging time. I am also extremely grateful to my family and friends for all their support and enthusiasm, with special thanks to Alice for all her enthusiasm, help and advice during the past 6 months and for making our trip to American Samoa such a memorable time.

Fa'afetai tele lava.

iii LIST OF CONTENTS ______

LIST OF CONTENTS

Abstract Declaration Acknowledgements List of Figures List of Plates List of Tables List of Abbreviations

1 INTRODUCTION Page 1.1 Aim 1 1.2 American Samoa 1 1.2.1 Location 1 1.2.2 Origin and Geography 1 1.2.3 Economy 3 1.2.4 Climate 3 1.2.5 American Samoan Islands 4 1.3 Marine Environment 5 1.3.1 Hydrography 5 1.3.2 Coral Reefs 7 1.3.2.1 Status of Coral Reefs 7 1.3.3 Coral Reef Fish Communities 7 1.3.3.1 Relationship between Fish and Coral 8 1.3.4 Algae 9 1.3.5 Fishing 9 1.3.6 Key Macroinvertebrates 12 1.3.6.1 Giant Clams 12 1.3.6.2 Crown-of-thorns Starfish 13 1.3.6.3 Sea Urchins 14 1.3.7 Water Quality 14 1.3.7.1 Pago Pago Harbour Special Management Area 17 1.3.8 Population Increase 17 1.3.9 Threats to Coral Reefs 18 1.4 Coastal Management 19 1.4.1 Marine Protected Areas 19 1.4.1.1 Uses of Marine Protected Areas 20 1.4.2 Marine Protected Areas in American Samoa 20 1.4.2.1 Fagatele Bay National Marine Sanctuary 21 1.4.2.2 National Park of American Samoa 23 1.4.3 Community-based Fisheries Management Programme 24 1.4.3.1 Aua 27 1.4.3.2 Alofau 27 1.4.3.3 Auto & Amaua 27 1.4.3.4 Vatia 28 1.5 Underlying Rationale 29 1.6 Objectives 30

iv LIST OF CONTENTS ______

2 MATERIALS AND METHODS 2.1 Collaboration with Outside Body 32 2.2 Survey Sites 32 2.2.1 Location of Survey Sites 32 2.2.2 Description of Survey Sites 32 2.2.3 Site Selection 36 2.3 Habitat Description Methods 36 2.3.1 Point-Intercept Transect Method 36 2.3.2 Macrobenthic Invertebrate Survey 39 2.3.3 Environmental and Physical Parameters 40 2.4 Survey Site Descriptions 41 2.5 Statistical Analysis 52

3 RESULTS 3.1 Reef Substrate 54 3.1.1 Reef Substrate Cover 54 3.1.2 Site Similarities 54 3.1.3 Environmental Parameters 60 3.1.4 Comparison of MPAs and non MPAs 61 3.2 Coral Species Abundance 63 3.2.1 Coral Cover and Species Richness 63 3.2.2 Diversity Indices 66 3.2.3 Site Similarities 68 3.2.4 Relationship between Coral Species 74 3.3 Substrate and Coral Species Abundances 75 3.4 Key Macroinvertebrates 78 3.5 Relationship between Coral and Fish 79 3.6 Site Evaluation 84

4 DISCUSSION 4.1 Reef Substrate 86 4.2 Coral Species Abundances 90 4.3 Reef Substrate and Coral Species Abundance 95 4.4 Key Macroinvertebrates 95 4.5 Relationship between Coral and Fish 96 4.6 Effects of Protection of Coral Reef Communities 99 4.6.1 Success of CB MPAs 102 4.7 Summary 103

5 CONCLUSION AND RECOMMENDATIONS 5.1 Management Recommendations 105 5.2 Limitations of the Present Study 106 5.3 Conclusion 106

6 REFERENCES 108 7 APPENDICES v LIST OF FIGURES ______

LIST OF FIGURES

Figure 1.1 Map of South Pacific showing the location of American Samoa and the adjacent independent country of Samoa. Figure 1.2 Map of Tutuila Island, American Samoa. Figure 1.3 Map of Manu’a Islands, American Samoa. Figure 1.4 Potential Total Economic Value of coral reefs in American Samoa. Figure 1.5 Destructive fishing methods damaging the coral reefs in American Samoa. Figure 1.6 Most common food collected from the reefs in American Samoa. Figure 1.7 Population growth in American Samoa. Figure 1.8 The villages currently participating in the Community-based fisheries management programme. Figure 1.9 The extension process for the Community-Based Fisheries Management Programme in American Samoa. Figure 1.10 Boundaries of Aua marine management area. Figure 1.11 Boundaries of Alofau marine management area. Figure 1.12 Boundaries of Auto & Amaua marine management area. Figure 1.13 Boundaries of Vatia marine management area. Figure 2.1 Map of the main island of Tutuila, American Samoa, showing location of each survey site. Figure 2.2 Map of the Manu’a Islands, American Samoa, showing location of survey site. Figure 2.3 Satellite image of Masefau indicating the location of the 5 transects. Figure 2.4 Satellite image of Vatia indicating the location of the 5 transects. Figure 2.5 Satellite image of Alofau indicating the location of the 5 transects. Figure 2.6 Satellite image of Alofau indicating the location of the 5 transects. Figure 2.7 Satellite image of Faga’itua indicating the location of the 5 transects. Figure 2.8 Satellite image of Auto & Amaua indicating the location of the 5 transects. Figure 2.9 Satellite image of Aua indicating the location of the 5 transects. Figure 2.10 Satellite image of Aua control indicating the location of the 5 transects. Figure 2.11 Satellite image of Nu’uuli lagoon indicating the location of the 5 transects. Figure 2.12 Satellite image of Fagatele Bay indicating the location of the 5 transects. Figure 2.13 Satellite image of Ofu lagoon indicating the location of the 5 transects. Figure 3.1 Satellite image of Tutuila and Ofu Island, American Samoa. The pie charts illustrate the percentage cover of each substrate category at the 11 sites. Figure 3.2 Cluster analysis showing classification of the 11 sites surveyed in American Samoa based on percent similarity of reef substrate.

vi LIST OF FIGURES ______

Figure 3.3 A multidimensional scaling ordination plot showing classification of the 11 survey sites in American Samoa based on percent similarity of reef substrate. Figure 3.4 Multidimensional scaling ordination bubble plots showing the classification of 11 sites in American Samoa based on percent similarity of reef substrate. Figure 3.5 Turbidity superimposed onto a Multidimensional scaling ordination plot showing the classification of 11 sites in American Samoa. Figure 3.6 Comparisons of mean cover (± SD) of reef substrate recorded at community- based MPA sites, non MPA sites and federal MPA sites in American Samoa. Figure 3.7 Coral cover and species richness (mean and standard deviation) recorded at each site surveyed in American Samoa. Figure 3.8 Cluster analysis showing classification of 5 replicate transects from each of the 11 sites surveyed based on percent similarity of coral species abundance. Figure 3.9 Diversity indices describing the coral community composition at 11 sites surveyed in American Samoa. Figure 3.10 Cluster analysis showing classification of the 11 sites surveyed in American Samoa based on percent similarity of coral species abundance. Figure 3.11 A multidimensional scaling ordination plot showing classification of 11 sites in American Samoa based on percent similarity of coral species abundance. Figure 3.12 Graph showing the percent cover of dominant coral genera recorded at each site surveyed in American Samoa. Figure 3.13 Cluster analysis showing percent similarity of coral species surveyed at 11 sites in American Samoa. Figure 3.14 Cluster analysis showing classification of the 11 sites surveyed in American Samoa based on percent similarity of reef substrate cover and coral species abundance Figure 3.15 A multidimensional scaling ordination plot showing classification of the 11 sites surveyed in American Samoa based on percent similarity of benthic substrate cover and hard coral species abundance. Figure 3.16 Fish abundance (mean and standard deviation) recorded at each site surveyed in American Samoa. Figure 3.17 Multidimensional scaling ordination bubble plots showing classification of the 11 sites surveyed in American Samoa based on percent similarity of reef substrate cover and coral species abundance. Figure 3.18 Relationships between coral and fish at 11 coral reef sites surveyed in American Samoa.

All figures not sourced are the work of the author.

vii LIST OF PLATES ______

LIST OF PLATES

Plate 1.1 Volcanic mountains and fringing reefs on Ofu and Olosega. Plate 1.2 Damselfish with its associated coral reef habitat. Plate 1.3 Fisherman on reef flat, American Samoa. Plate 1.4 Maori wrasse have declined due to overfishing. Plate 1.5 Giant clam (Tridacna maxima), Ofu Lagoon, American Samoa. Plate 1.6 Crown-of-thorns starfish, Ofu Lagoon, American Samoa. Plate 1.7 Urban and industrial expansion around Pago Pago Harbour. Plate 1.8 Tuna canneries in Pago Pago Harbour. Plate 1.9 Fagatele Bay National Marine Sanctuary, Tutuila Island. Plate 1.10 Ofu lagoon, National Park of American Samoa, Ofu Island. Plate 1.11 Group meetings between DMWR and villagers to discuss the CBFMP. Plate 2.1 Example of a transect using a 25m tape measure at Ofu Lagoon Plate 2.2 Surveyor recording substrate during the Point-Intersect transect method at Ofu Lagoon, American Samoa Plate 2.3 Vatia village and marine management area. Plate 2.4 Alofau marine management area. Plate 2.5 Alofau lagoon underwater. Plate 2.6 Faga’itua marine area. Plate 2.7 Auto & Amaua marine management area. Plate 2.8 Aua marine management area. Plate 2.9 Aua control village and marine area. Plate 2.10 Nu’uuli lagoon underwater and marine area. Plate 2.11 Fagatele Bay National Marine Sanctuary. Plate 2.12 Ofu lagoon, the National Park of American Samoa. Plate 4.1 Rubble at Auto & Amaua, American Samoa. Plate 4.2 Massive Porites at Ofu lagoon, American Samoa. Plate 4.3 High abundance of fish and high coral cover (particularly Acropora microphthalma) at Faga’itua, American Samoa. Plate 4.4 A sign designating the fishery reserve of Vatia.

All plates not sourced were taken by the author.

viii LIST OF TABLES ______

LIST OF TABLES

Table 1.1 Island and reef type, size, and human population of each island in American Samoa. Table 1.2 Marine Protected Areas of American Samoa. Table 1.3 Common fishing methods previously used by villagers. Table 2.1 Environmental, physical and social factors of each site surveyed in American Samoa during June to August 2004. Table 2.2 Description of the reef substrate categories and codes used in the surveys carried out in American Samoa. Table 2.3 SACFOR scale used for recording sessile invertebrate abundance in the American Samoa surveys. Table 3.1 The mean percent cover (±SD) of each substrate category recorded at 11 coral reef survey sites in American Samoa Table 3.2 Coral classification of the 11 sites surveyed in American Samoa during June to August 2004. Table 3.3 SIMPER analysis showing which coral species contributed the most to the similarity among each of the 11 sites surveyed in American Samoa. Table 3.4 SIMPER analysis indicating whether hard coral species or reef substrate contributed the most to the similarity among sites. Table 3.5 Key macro invertebrate abundance at 11 sites surveyed in American Samoa.

ix LIST OF ABBREVIATIONS ______

LIST OF ABBREVIATIONS

MPA Marine Protected Area CB Community-based CBFM Community-based fisheries management EEZ Exclusive Economic Zone TEV Total Economic Value ASCRTF American Samoa Coral Reef Task Force ASEPA American Samoa Environment Protection Agency COTS Crown-of-thorns Starfish IUCN The World Conservation Union UNEP United Nations Environment Programme FBNMS Fagatele Bay National Marine Sanctuary NPAS National Park of American Samoa SPC South Pacific Commission DOC Department of Commerce ASCMP American Samoa Coastal Management Program DMWR Department of Marine and Wildlife Resources CBFMP Community-based Fisheries Management Programme FMP Fisheries Management Plan GIS Geographical Information System GPS Global Positioning System SST Sea Surface Temperature SCUBA Self-Contained Underwater Breathing Apparatus ASCRAG American Samoa Coral Reef Advisory Group NOAA National Oceanic and Atmospheric Administration PRIMER Plymouth Routines In Multivariate Ecological Research ANOVA Analysis of Variance ANOSIM Analysis of Similarities MDS Multi Dimensional Scaling

Units of Measurement km Kilometers % Percentage oC Degrees oF Fahrenheit nm Nanometers m Meters + Plus US$ American dollars < Less than > More than p Probability

x 1 INTRODUCTION ______

1.1 AIM This study was carried out to investigate the effects of protection on coral reef communities in American Samoa. To achieve this, the study recorded and provided a quantitative description of coral reef benthic communities in terms of substrate cover, abundance of coral species and key macro-invertebrates at Community-based Marine Protected Areas (MPAs), federal MPAs (Statutory) and non MPAs. The benthic community data collected provided quantitative baseline information, on the status of coral reef communities of community-based fisheries management (CBFM) village MPAs, to determine the efficiency of management. The study will therefore provide the basis for establishing a long term monitoring program for the CBFM village MPAs of American Samoa. The data also allowed for future analysis to be carried out on trends in the health of the reefs.

1.2 AMERICAN SAMOA 1.2.1 Location The U.S. territory of American Samoa is a group of five volcanic islands (Tutuila, Aunu’u, and the Manu’a Islands of Ofu, Olosega and Ta’u) and two remote atolls (Rose and Swains). These Islands are located in the eastern portion of the Samoan

Archipelago in the central South Pacific Ocean between 11-14o S latitude and 168-

173o W longitude (Figure 1.1). The islands are small, ranging in size from the large populated island of Tutuila (142 km2) to the uninhabited and remote Rose Atoll (4 km2) (Craig, 2001). The combined land area of the islands is about 76.7 square miles with a population of approximately 63,000 (American Samoa Census, 2000).

1.2.2 Origin and Geography All the islands except Swains Atoll are aligned along the crest of a discontinuous submarine ridge which extends over 485km and trends roughly north-west by southeast. Swains and Rose Atolls are limestone and the other islands are composed principally of volcanic rock and are typically deep-sided with lush vegetation (IUCN/UNEP, 1988). Due to the steepness of the small islands (Plate 1.1) and their relative geological youth, shallow water habitats around the islands are limited in size and consist primarily of fringing coral reefs (85% of total coral reef area), with a few offshore banks (12%) and two atolls (3%) (Craig, 2001).

Figure 1.1 Map of South Pacific showing the location of American Samoa (between 11-14o S latitude and 168-173o W longitude) and the adjacent independent country of Samoa (Source: NPAS).

2 1 INTRODUCTION ______

Plate 1.1 Volcanic mountains and fringing reefs on Ofu and Olosega.

1.2.3 Economy The economy of American Samoa is based primarily on federal grants and exports of canned tuna (a product that is not caught within the EEZ of American Samoa). Most of the 14,000 workforce is employed by the tuna canneries (33%), which have operated in the Harbour since 1956, and by the local government (31%) (Craig et al., 2000a). There is a heavy reliance on imports for food, fuel and materials.

1.2.4 Climate American Samoa has a hot, humid tropical climate with average temperatures of 70- 90 oF (21-32 oC) and average humidity of 80%. There is a hot and wet season from October to May. June to September is the slightly drier and cooler season, characterised by 2 oC cooler air temperatures, increased easterly trade winds and reduced rainfall by half. Severe storms occur almost annually and hurricanes hit periodically.

A sharp rise in air temperature in the past decade suggests future climatic uncertainty and a probable increase in the frequency of hurricanes in the region. A broad band of high Sea Surface Temperatures (SSTs) has developed in recent summers, perhaps in conjunction with La Nina conditions, just south of American Samoa. The high SSTs

3 1 INTRODUCTION ______

has caused massive coral bleaching and mortalities in nearby Fiji and Samoa, and it would seem highly probable that corals in American Samoa may be hit as well (Craig and Basch, 2001). The record high temperature in 1998 was due in part to El Nino which also caused drought conditions on land and unusually low tides that caused mortalities among exposed corals (Craig et al., 2000a).

The southern side of the islands are exposed to the prevailing southeast Trade Winds from April to September. In contrast, the north side is more protected from the Trade Winds, but tends to be harder hit by hurricanes which occur from October to March. Pago Pago Harbour on the south side of the island tends to be relatively protected from the prevailing wind conditions (Green, 1996).

1.2.5 American Samoan Islands The present study focuses on the two volcanic islands of Tutuila and, Ofu of the Manu’a Group. These islands differ in terms of their size, age and human habitation (Table 1.1). Most of the population live on Tutuila (55,400), with a much smaller population (1,378) on the Manu’a Islands (Green, 2002). The economic and government centre Tutuila, is the oldest and largest island in American Samoa (Figure 1.2). It is approximately 32 km long and averages 4km in width. The island is highly urbanised by South Pacific standards. Ofu Island is a small volcanic island (7.5 km2) with a well-developed fringing reef that is 80-180 m wide (Figure 1.3). Located 102 km east of Tutuila, it is smaller, younger and has a much lower population density.

Pago Pago Harbour

Aunu’u Island TUTUILA

Figure 1.2 Map of the main island of Tutuila, American Samoa.

4 1 INTRODUCTION ______

Asaga Strait Nu’utele Island OFU OLESAGA

Figure 1.3 Map of Ofu and Olesaga, American Samoa.

Table 1.1 Island and reef type, size, and human population of each island in American Samoa. Where: f=fringing; ns=nonstructural reef community; and reef area is for Territorial Waters (0-3nm from shore), and 0-100m deep (American Samoa Census 2000).

Island Island Reef Island Reef Human % Human Population Type Type Area Area Population Population Density (km2) (km2) (in 2000) (in 2000) (per km2) Tutuila Volcanic f,ns 142.3 243 55,400 96.7% 389.3 Ofu Volcanic f,ns 7.5 3.2 289 0.50% 38.5

1.3 MARINE ENVIRONMENT 1.3.2 Coral Reefs Coral reefs are among the most diverse and productive ecosystems in the world that flourish in the clear, tropical waters of the South Pacific. American Samoa is fortunate to have well developed coral reefs surrounding most of the islands in the archipelago. Most of the reefs are narrow fringing reefs that are close to shore (<200m). These reefs can be divided into six recognizable habitat types, which differ in their position on the reef profile, depth and degree of wave exposure (described in detail by Green, 1996).

The total area of coral reefs in American Samoa is 296km2, with 200+ coral species, representing over half of all coral species found throughout the Indo-Pacific region (Hunter et al., 1993; Mundy, 1996). The reefs also support a diverse assemblage of 890 fishes (Wass, 1984) and 80 algal species (Spalding et al., 2001).

5 1 INTRODUCTION ______

The coral reefs of American Samoa provide significant benefits to the territory and mainland US. They are an important natural resource providing food for villagers through daily subsidence use and sales at local stores. They also play an integral role in the rich cultural heritage of the islands, and provide other important ecosystem services including shoreline protection from storm wave action (Craig, 2001). In 2004 an economic valuation of the coral reefs and adjacent habitats of American Samoa was undertaken by Spurgeon et al. (2004). Corals were estimated to be currently worth a minimum of US$8 million/year and the total current value of the coral reefs is at a minimum worth US$160 million (Spurgeon et al., 2004). The Total Economic Value (TEV) of coral reefs in American Samoa is illustrated in Figure 1.4.

Figure 1.4 Potential Total Economic Value of coral reefs in American Samoa (Source: Spurgeon et al. (2004) based on Spurgeon, 1992).

6 1 INTRODUCTION ______

1.3.2.1 Status of Coral Reefs Coral reef habitats of American Samoa are currently recovering from a series of both natural and human impacts over the past two decades. In the 1970s, the reefs were devasted by a massive outbreak of the corallivorous crown-of-thorns starfish Acanthaster planci, which resulted in the loss of up to 95% live coral cover in many areas (Wass, 1979). Since then, four hurricanes (1986, 1990, 1991 and 2004) and a mass coral bleaching event in 1994 have taken their toll on coral health.

However, coral reefs are robust ecosystems and by 1995 the coral reefs of Tutuila were beginning to recover from such "natural" disturbances. Many reefs that had been reduced to rubble by the hurricanes in 1990 and 1991 had already been consolidated by pink coralline algae and coral recruitment of juveniles was high (Mundy 1996, Birkeland et al., 1997). Coral growth has continued through 2000, but a full recovery will take more time (Craig, 2001). In the absence of further disturbances, Samoan reefs have the potential to recover to a large extent in one to two decades (Zann and Sua, 1991).

Tutuila’s reefs have also seen increasing pressure from chronic human-induced impacts such as water pollution, increased sedimentation due to poor land use practices, eutrophication and overfishing, which all have caused degradation to the reefs (Craig et al., 2000a). Live coral cover decreased from 60% in 1979 (Wass, 1982) to 3-13% in 1993. There is now concern that human impacts may be contributing to the degradation of these reefs, and inhibiting their ability to recover from natural disturbances (Craig et al. 1995; Zann and Sua, 1991).

1.3.3 Coral Reef Fish Communities Coral communities provide important habitat for their associated fish fauna. There have been some major changes in the fish communities on Tutuila over the last few decades, in response to these natural and anthropogenic impacts which have resulted in physical and biological changes to the coral communities (Green, 2002; Birkeland et al., 2003).

In the mid 1970s, the reefs of Tutuila supported a rich and diverse fish fauna, because the reefs were in good condition and fishing pressure was relatively low.

7 1 INTRODUCTION ______

When the reefs were devastated by COTS in the late 1970s, there were major impacts on some components of the fish fauna. In particular, there was a decline in abundance of species that are closely associated with the coral communities (Green, 2002). Fish communities are now recovering from the effects of the large scale disturbances over the last few decades, along with their host coral communities (Green, 2002).

1.3.3.1 Relationship between Fish and Coral Community Structure Coral reefs support a high diversity of fishes that may ultimately depend on corals for their survival (Plate 1.2). Diversity, quality, and aerial extent of habitat are among the most important environmental determinants of coral reef fish distribution, abundance and diversity (Bellwood and Hughes, 2001).

Plate 1.2 Damselfish with its associated coral reef habitat (Photo: A. Lawrence)

In the past, there has been a dichotomy of opinion over how closely fish communities are related to their habitat. A number of authors have recognised the importance of habitat complexity in structuring fish assemblages (Luckhurst and Luckhurst, 1978; Carpenter et al., 1981; Roberts and Ormond, 1987; Grigg, 1994; Friedlander and Parrish, 1998a). Many studies have found positive correlations between the percentage of live coral cover and fish diversity and abundance (Carpenter et al.,

8 1 INTRODUCTION ______

1981; Reese, 1981; Bell et al., 1985; Chabanet et al., 1997), whilst others have found little or no correlation (Luckhurst and Luckhurst, 1978; Roberts and Ormond, 1987).

1.3.4 Algae A survey by Skelton (2003) showed a fairly diverse flora for American Samoa with 239 species recorded. Information is limited, but with few exceptions, the algae found around the islands are indicative of a low-nutrient environment and/or heavy grazing by herbivores (Craig, 2001). Encrusting coralline algae cover (Porolithon) is high (40-50%), which is a good indicator of a healthy ecosystem. Coralline algae are an important group in coral reefs because of their varied roles. They play an important role during most phases of reef recovery, by consolidating loose substratum such as rubble, thus providing a stabilized reef. They also provide food for many invertebrates and specialized feeders such as parrotfish (Skelton, 2003).

1.3.5 Fishing Coral reef fisheries are an important resource (Plate 1.3), both economically and culturally, in American Samoa (Tuilagi and Green, 1995). Major components of the coral reef fishery are reef fish, giant clams and the palolo worm which are harvested in two local fisheries; subsistence catches, and small-scale artisinal fisheries where catches are sold locally (Ponwith, 1991; Craig et al., 1993).

Plate 1.3 Fisherman on reef flat, American Samoa.

9 1 INTRODUCTION ______

A limited amount of harvest information is available on the coral reef fishery in the Territory. Coral reef resources are harvested on a daily basis on Tutuila, and comprise 40-80% of the fisheries landings each year (Craig et al., 1993; Saucerman, 1996). However, as the population in American Samoa has expanded, the demand for fisheries resources has also increased. This has lead to declining subsistence catches on Tutuila (Wass, 1980; Saucerman, 1996) from around 300 tonnes yr-1 in 1980 (Ponwith, 1991) to around 42 tonnes yr-1 in 2003 (DMWR, 2002), with evidence that marine fish and invertebrate populations have diminished due to fishing pressure (ASCRTF, 1999). Green (2002) also reported that the fish populations on Tutuila are overfished. Fisheries species are much less abundant on Tutuila than in the Manu’a Islands. Large species that are particularly vulnerable to overfishing (sharks, some parrotfishes and maori wrasse - Plate 1.4) are now rare or absent on Tutuila but still occur in the Manu’a Islands (Green, 2002).

Plate 1.4 Maori wrasse have declined due to overfishing (Photo: R. Myers in Green, 2002).

No long-term catch data are available for the Manu’a islands, though the subsistence fishery clearly remains more important to the local way of life than on Tutuila. Fishing rates are presumed to be lower in the lightly populated Manu’a Islands (Green, 2002). For example, Itano and Buckley (1988) reported that the Manu’a Islands appeared to be lightly fished, based on the presence of large, unwary fish and high densities of giant clams. Other studies have shown similar decline in reef stocks there (Green, 1996). A decline in fish abundance may have contributed to the decline in catches, particularly as a result of historical over fishing (ASEPA, 2002).

10 1 INTRODUCTION ______

Despite the ongoing recovery of corals on local reefs, many fish and invertebrates are not recovering as well. At present, there is less reliance on subsistence fishing as many Samoans obtain full-time employment. In addition, fishing practises have shifted from the use of traditional methods (fish traps, nets and lures), to modern methods including the use of power boats, SCUBA (Self-Contained Underwater Breathing Apparatus) equipment and spearguns (Wass, 1980). Since 1995, the night time artisanal spear fishermen began using SCUBA gear, greatly increasing their catches. This highly efficient fishery was banned by Executive Order, by the Governor of American Samoa in April 2001 due to concerns that the reef fish populations were being overfished (Craig, 2001).

Destructive fishing practices, including the use of bleaching agents, dynamite and avaniukini (a local plant-derived poison) are also illegal in American Samoa. These methods can cause severe damage to coral reef habitats and result in an ultimate decline in the fishery (Figure 1.5). However, there is some evidence that illegal fishing practices continue to be used on Tutuila (Itano 1980; Tutuila and Green 1995; Birkeland et al., 2004). For example, in an interview survey of fishermen on Tutuila, 25% of people reported that dynamite fishing had occurred in the last year (Tuilagi and Green, 1995). There have also been various reports of fishing nets left on the reef which can damage and kill many marine species as well as the use of anchors and traps (Sauafea, 2000).

6% 3% 24% 13%

16%

19%

19% Chlorox, Avaniukini, Dynamite Outside Fishermen Avaniukini Dynamite Chlorox Fishing net Trash

Figure 1.5 Destructive fishing methods damaging the coral reefs in American Samoa (Sauafea, 2000).

11 1 INTRODUCTION ______

Coral reef biologists and geologists agree that fishing is one of the biggest human- induced factors affecting the ecology and diversity of coral reefs (Birkeland, 1997). Fishing may have little effect on the fish community other than to reduce overall numbers and biomass but it can lead to serious consequences for coral reef ecosystems (Jackson et al., 2001). For example, many key fisheries species play an important role in structuring coral reef ecosystems, and their removal can cause serious problems for reefs. In particular, removal of herbivorous fish (surgeonfish, parrotfish and rabbit fish) can lead to serious ecosystem effects, such as an increase in algae and a decrease in coral recruitment (Jackson et al., 2001). Fortunately, this does not appear to have occurred on the reefs of Samoa as yet, which is demonstrated by the fact that the reefs at most sites are still in good condition and resilient to large scale disturbances (Green, 2002).

1.3.6 Key Macroinvertebrates 1.3.6.1 Giant Clams Giant clams, locally known as faisua, are an important food item in Samoa, but their accessibility and life history characteristics make them particularly vulnerable to over-harvesting (Plate 1.5). Green and Craig (1999) examined fishing pressure on the status of giant clam populations on eight islands in the Samoan Archipelago, and concluded that they were overfished throughout most of the archipelago. This information was consistent with local fisheries statistics for Tutuila, which showed a decline in the harvest of giant clams over the last two decades, presumably as a result of overfishing (Tuilagi and Green, 1995; Green and Craig, 1996, 1999).

Furthermore, Green and Craig (1999) demonstrated a correlation between the density of clams and the size of the human population on the islands in the Samoan Archipelago. The study demonstrated that the highest clam densities were present on the uninhabited Rose Atoll, and the lowest clam densities were recorded on the most heavily populated islands of Tutuila and ‘Upolu. The Manu’a Islands, with its lower population, was intermediate in both respects. Results by Green (2002) have also confirmed that trend.

Green and Craig (1999) further demonstrated that Rose Atoll was an important refuge for giant clams that have been heavily exploited elsewhere in the archipelago.

12 1 INTRODUCTION ______

One concern is that the remaining individuals are now present in such low densities that their reproductive success, and subsequent recruitment, may be diminished (Green, 2002).

1.3.6.2 Crown-of-thorns Starfish The crown-of-thorns starfish (COTS), locally know as alamea, is a natural inhabitant of the reefs of Samoa (Plate 1.6). This species feeds on corals and is usually uncommon, where it causes minimal damage to coral communities.

COTS have been rare or uncommon on Tutuila since the major outbreak in the late 1970s, which devastated the coral communities. In contrast, several studies (Itano and Buckley 1988; Zann 1992; Maragos et al., 1994) have reported a low to moderate population of COTS on Ofu and Olosega over the last few decades. COTS were known to be quite abundant on Ofu about four years ago, when the National Park of American Samoa (NPAS) removed about 40 individuals from the lagoon (P. Craig pers comm). More recently, a study by Green (2002) recorded only a few individuals in Ofu Lagoon. The starfish may have played an important role in structuring the coral reef communities on Ofu and Olosega in the Manu’a Group (Green, 2002).

Plate 1.5 Giant clam (Tridacna maxima), Plate 1.6 Crown-of-thorns starfish, Ofu Lagoon, Ofu1.3.6.3 Lagoon, Sea AmericanUrchins Samoa.How was data collected?? American When?? Samoa (Photo: A. Lawrence).

13 1 INTRODUCTION ______

1.3.6.3 Sea Urchins To determine the problems in fishing in American Samoa, the DMWR conducted a questionnaire in 11 selected villages (Sauafea, 2000). From the 293 respondents, 22% reported that the most common food collected from the reefs in American Samoa were sea urchins. Turbo snails, octopus, clams, sea cucumbers, lobsters, snappers, groupers and surgeon fish were also common food collected (Figure 1.6)

4% 2% 4% 22% 4% 4%

10%

18% 14%

18%

Sea urchins Turbo snails Octopus Clams Sea Cucumbers lobsters snappers groupers surgeon fish Others

Figure 1.6 Most common food collected from the reefs in American Samoa. Results of a questionnaire carried out by DMWR in 2000 (Sauafea, 2000).

1.3.7 Water Quality Due to the islands steep terrain and their limited development of shallow water habitats, the reefs are continually flushed by clear oceanic waters which provide generally good water quality around the islands of American Samoa (Craig, 2001). Exceptions to this are (1) sedimentation into coastal waters after heavy rains due to poor land-use practises, (2) nutrient enrichment from human and animal wastes in populated areas, and (3) contamination in Pago Pago Harbour, which is considered a Special Management Area (Craig, 2001).

1.3.7.1 Pago Pago Harbour Special Management Area Early last century, Pago Pago Harbour was characterized by lush coral cover and minimal human impacts (Mayor, 1924). Since then, the harbour has experienced

14 1 INTRODUCTION ______some major changes and become a heavily populated urban and industrial area. As a result, the reefs have experienced major dredging and filling operations as well as chronic pollution over many decades (Green et al., 1997). Several industrial activities, all located within Pago Pago Harbour, have had a major impact on water quality in the harbour (Plate 1.7); There are two tuna canneries (Plate 1.8), a sewage treatment plant, ship repair yard, fuel tank farm and power plant whose collective discharge and runoff have resulted in declining water quality and caused several kinds of pollution (Craig et al., 2000a). Firstly fish and substrates in the harbour are contaminated with heavy metals and other pollutants from industry and agriculture (AECOS, 1991). Of particular concern has been the discharge of untreated and uncontrolled volumes of cannery wastes into the inner harbour for several decades. This waste overloaded the inner harbour with nutrients and severely depressed dissolved oxygen levels, resulting in algal blooms and occasional fish kills. Since the early 1990’s there has been an improvement in the management of waste from the tuna canneries and a subsequent dramatic reduction in the nutrient levels in the harbour (Green et al., 1997; Craig, 2001).

Sewage treatment plant

Tuna canneries

Ship repair yard

Plate 1.7 Urban and industrial expansion around Pago Pago Harbour on the main island of Tutuila, American Samoa.

The coral communities in Pago Pago Harbour are in the worst condition of all the reefs in the Territory (Maragos et al., 1994; Mundy, 1996). A study by Green et al. (1997) demonstrated that the coral reefs in the harbour have declined probably due to

15 1 INTRODUCTION ______poor water quality. In particular, approximately 97% of the reefs in the inner harbour have been completely destroyed by dredging and filling operations (IUCN/UNEP, 1988). Mundy (1996) also concluded that the poor condition of the coral communities in the harbour was probably due to the long term effects of poor water quality. Moreover, previous studies (Birkeland et al., 1987, 1997; Maragos et al., 1994) have reported that despite the stressed conditions in the harbour, these reefs are important since they support habitats and coral species otherwise unique to Samoa, in part attributed to the sheltered conditions within the harbour.

Plate 1.8 Tuna canneries in Pago Pago Harbour, American Samoa (Photo: A. Lawence).

Monitoring of coral reef communities in the harbour has demonstrated that impacted coral reefs have begun to show signs of recovery during the past 10 years in response to improved water quality. For example, the reef flat communities at Aua, located in Pago Pago Harbour (Figure 1.2), were found to be in good condition for the first time in decades (healthy coral and crustose coralline algae), which was attributed to improved water quality (Birkeland and Green, 1999). Evidence of coral recruits in the harbour included species of Acropora, (particularly Acropora hyacinthus) which are considered a good indicator of improved water quality, since they are highly sensitive to sedimentation (Birkeland et al., 2004).

16 1 INTRODUCTION ______

However, it is likely that the reefs in the harbour will not recover from the hurricanes to the same extent as the other reefs around the island, because of poor water quality (Green, 2002). In contrast, where water quality is good, the reefs of American Samoa have demonstrated that they are healthy, resilient, and able to recover from large scale disturbances. The substratum is quickly consolidated by pink coralline algae, and coral recruitment is high leading to the rapid recovery of the coral communities (Green, 1996; Green et al., 1999; Green, 2002).

1.3.8 Population Increase A serious environmental and social problem facing American Samoa is its rapid population growth rate (Figure 1.7), which is the third highest growth rate in the South Pacific region (SPC, 1994). American Samoa’s population was 57,300 in 2000. During the past 10 years, the population increased by 10,500 people, an increase of 22% (Craig, 2001). The population is increasing at a rate of 2.1% per year which equates to an increase of about 1,200 people per year (Craig et al., 2000b). It is expected to continue rising due to high birth and immigration rates.

The population has increased far beyond what the local environment can currently support (Craig et al., 2000b), which has lead to an increase in human impacts on coral reefs and their associated fisheries (Craig et al., 1995; Saucerman, 1995; Wass, 1982; Zann, 1991). Adverse human-related impacts to the marine environment will likely increase with rapid and unsustainable population growth, unless present trends are moderated by responsible policies or the course of future events (Craig et al., 2000b).

Figure 1.7 Population growth in American Samoa (Source: Craig et al., 2000b).

17 1 INTRODUCTION ______

1.3.9 Threats to Coral Reefs Craig et al. (1999) and Craig and Basch (2001) identified the following list of human-related and natural threats to coral reefs in American Samoa, ranked as being of high, medium or low concern:

High Overfishing and overharvesting of reef resources. Coastal development and habitat destruction. Oil and hazardous waste spills in Pago Pago Harbour. Hurricanes. Global warming in American Samoa (i.e. bleaching). Harvest of Samoan sea turtles in foreign waters.

Medium Increased sedimentation due to poor land use practices. Dumping/improper waste disposal. Nutrient loading/eutrophication in Pago Pago Harbour as a result of runoff from domestic sewage, piggery waste and the effluent from the tuna canneries. Increased UV radiation due to ozone depletion.

Low Nutrient loading/eutrophication other than Pago Pago Harbour. Oil and hazardous waste spills other than Pago Pago Harbour. Ship groundings. Anchor damage. Destructive fishing practices . Marine debris from marine sources. Introduction of Alien species (from ballast waters). Collections for aquarium market. Bio-prospecting/natural products. Crown-of-thorns starfish predation. Coral diseases.

18 1 INTRODUCTION ______

The heavily populated island of Tutuila appears to be the most heavily affected by human impacts (especially pollution and sedimentation: Bell, 1989; Birkeland et al., 1997; Maragos et al., 1994; Zann, 1991), and moderately high levels of fishing (Green and Hunter, 1998). The reefs of Ofu, while used for subsistence fishing by the small local population, generally receive less pressure from human activities and appear to be in better condition than the reefs on Tutuila (Green, 1996).

1.4 COASTAL MANAGEMENT American Samoa is fortunate to have a variety of means and methods of managing its coastal resources. The Department of Commerce (DOC) established the American Samoa Coastal Management Program (ASCMP) in 1980 to protect and preserve natural resources while attempting to balance and satisfy development needs of the people and Fa’asamoa (the Samoan way of life).

It is apparent that fish and shellfish catches through out the tropics have been in decline for many years (King and Faasili, 1999). Given the nutritional, cultural and economic importance of healthy fish stocks to traditional societies in the Pacific, nations and territories have been examining methods to increase catch and protect or restore habitat. One management tool becoming increasingly recognised for conserving marine resources and ecosystems is the use of MPAs.

1.4.1 Marine Protected Areas MPAs (also referred to as marine parks, marine reserves, marine sanctuaries, harvest refugia and no-take zones) are being promoted increasingly as alternative management tools for the conservation of biodiversity and fisheries management in tropical regions (Roberts and Polunin, 1993; Bohnsack, 1998; Allison et al., 1998). There are many definitions, but the IUCN defines MPAs as;

“any area of intertidal or subtidal terrain, together with its overlying water and associated flora, fauna, historical and cultural features, which has been reserved by law or other effective means to protect part or all of the enclosed environment.”

19 1 INTRODUCTION ______

1.4.1.1 Uses of MPAs MPAs have a huge range of potential functions including conserving biodiversity and ecosystem structure, tourism, protecting sensitive habitats, providing refuge for intensively fished species, enhancing the production of target species, providing a management framework for sustainable multiple use, serving as a demonstration of the extent of human impacts in coastal environments, or a combination of these goals (Allison et al., 1998). Within MPAs, some areas may exclude all fishing, collecting and mining; these are ‘highly protected’ or ‘no-take zones’.

1.4.2 MPAs in American Samoa MPAs play an important role in preserving biodiversity in Samoa by maintaining healthy coral reefs and populations of species that may be heavily impacted elsewhere in the archipelago (Green, 1996). There are three federal MPAs and one territorial MPA in American Samoa, described in Table 1.2, which account for only 6% of the Territory’s coral reefs (Craig, 2001). The three federal MPAs, Fagatele Bay National Marine Sanctuary (FBNMS), the National Park of American Samoa (NPAS) and Rose Atoll National Wildlife Refuge, all meet the definition of an MPA, and are generally accepted as useful for increasing or maintaining endangered populations, as well as important habitats and communities when properly implemented.

Table 1.2 Marine Protected Areas of American Samoa (Source: Craig, 2001). Site Year Location Size Acres % of coral reefs Km2 in territory Rose Atoll National 1973 Rose Atoll 158.8 39,251 2.4% Wildlife Sanctuary Fagatele Bay National 1986 Tutuila 0.7 161 3.1% Marine Sanctuary National Park of American 1993 Tutuila, Ofu, 42.6 10,520 0.2% Samoa Ta’u Vaoto Territorial Marine 1994 Ofu 0.5 120 0.1% Park

FBNMS was designated in 1986 in response to a proposal from the American Somoa Government to the National Marine Sanctuary Programme. This programme was

20 1 INTRODUCTION ______established by Congress in 1972 to protect nationally significant marine areas for their ecological and historical value.

The NPAS is in the Federal National Park System, but is actually a cooperative programme with the people and government of American Samoa. The Park encompasses a variety of unique natural resources, including approximately 8000 acres of paleotropical rainforest and 2,550 acres of Indo-Pacific coral reefs, located on the islands of Tutuila, Ofu and Ta'u. These reefs comprise some of the most biologically diverse and pristine habitats remaining in the US Pacific Islands (Craig and Basch, 2001).

American Samoa currently has one Territorial MPA, Ofu Vaoto Marine Park. Reserved by territorial legislation in 1994, the Park was established to protect unique coral habitats while allowing public access and enjoyment.

Rose Atoll National Wildlife Refuge is the only ‘no-take’ MPA while the NPAS and FBNMS allow traditional subsistence fishing by villagers but not commercial fishing. However, resource protection in these MPAs through regular surveillance and enforcement is generally lacking and poaching of marine organisms is an ongoing problem (Craig, 2001).

This study included sites in two federal MPAs in American Samoa: Fagatele Bay National Marine Sanctuary (FBNMS), and the Ofu Unit of the National Park of American Samoa (NPAS).

1.4.2.1 Fagatele Bay National Marine Sanctuary (FBNMS) Fagatele Bay was declared a National Marine Sanctuary in 1986 because of its isolation, spectacular beauty, and the pristine nature of its marine resources (Green et al., 1999). The sanctuary comprises a fringing coral reef ecosystem nestled within an eroded volcanic crater (Plate 1.9). The smallest and most remote of all the National Marine Sanctuaries, it is the only true tropical reef in the programme containing a moderately diverse coral reef community (150 coral species, 259 fish species) that are relatively protected from anthropogenic disturbances (Green et al., 1999).

21 1 INTRODUCTION ______

Corals, other invertebrates, fish and algae have been monitored by Birkeland et al. (1987, 1997 and 2004) in FBNMS since 1982. The Sanctuary encompasses a healthy coral reef community, which has undergone some substantial changes over the last two decades as a result of the large scale disturbances described previously. In particular, the reef flat communities had declined since 1995, probably due to a severe low tide event which resulted in a mass die-off of corals on the reef flat in 1998. Results from the long term monitoring program have demonstrated that coral abundance and cover may have increased on the reef flat from 1998 to 2001 (Birkeland et al. 2004). Green et al. (1999) also reported that the reefs in the Sanctuary are healthy, resilient, and recovering from these disturbances. In fact the Sanctuary supports some of the healthiest coral communities on Tutuila (Green, 2002).

Plate 1.9 Fagatele Bay National Marine Sanctuary, Tutuila Island, American Samoa.

Despite its protected status, dynamite fishing has been reported in the Bay on several occasions over the last few years. The case for asserting that the Sanctuary is being fished is also supported by the presence of physical damage to the reef (Birkeland et al., 2004). One contributing factor is the relative isolation of the Sanctuary, and because there is no village to protect the Sanctuary it is difficult to maintain an enforcement presence in the Bay.

22 1 INTRODUCTION ______

1.4.2.2 National Park of American Samoa (NPAS) The NPAS was authorized in 1988 and established in 1993. The Park includes the small, shallow lagoons at Ofu (Plate 1.10). They are the best developed, naturally occurring lagoons on the main volcanic islands, that support a diverse assemblage of coral and fish species (Hunter et al., 1993). The lagoons are an important natural resource as they are used for subsistence fishing and recreation.

Plate 1.10 Ofu lagoon, National Park of American Samoa, Ofu Island.

Despite chronic COTS predation, the lagoon supports spectacular coral reef communities, which are otherwise unique in American Samoa (Green, 2002). The lagoons may also play an important role in the ecology of the reefs on Ofu and Olosega, since they may act as a nursery for some important fisheries species (particularly parrotfishes). They may also play an important role in maintaining the chronic COTS population on those islands (Green, 2002). Previous surveys have also reported that lagoons at Ofu are of particular importance because the rare blue coral Heliopora coerulea is relatively abundant (Green, 1996).

Fishing pressure was identified as the potentially most significant threat to coral reefs in the NPAS. In the Ofu unit, resources like giant clams may be overharvested in the

23 1 INTRODUCTION ______

lagoon. There is also concern that an increasing number of fish may be collected island-wide to supply the larger markets on Tutuila Island. Other threats include potential recreational impacts, coral mortalities due to COTS, and pollution from a village garbage dump adjacent to the lagoon (Craig and Basch, 2001).

1.4.3 Community-based Fisheries Management Programme There is increasing awareness of the role local communities can and should play in management of marine resources (Pomeroy 1994; Ruddle and Pomeroy 1994; White et al., 1994). In the majority of developing countries, a community-based approach to MPA management involving key stakeholders is considered by many to be the only way management and enforcement will be successful in the long term (Roberts and Polunin 1993; King and Faasili 1998; Tsing et al., 1999).

The Department of Marine and Wildlife Resources’ (DMWR) sponsored Community-based Fisheries Management Programme (CBFMP) was implemented in 2001 and is currently installed in seven villages, Poloa, Alofau, Fagamalo, Vatia, Masausi, Auto & Amaua and Aua (Figure 1.8). The basis of the programme is to assist the villages in managing and conserving their own coral reef resources by a voluntary scheme of co-management with the government. The programme aims to improve fishing and sustainable development of marine resources in the participating villages as well as the Territory by establishing MPAs (Sauafea, 2002).

Masausi Vatia Auto Amaua Aua Alofau

Fagamalo Paloa

Figure 1.8 The villages currently participating in the community-based fisheries management programme with the boundaries of their marine management areas shown in purple.

24 1 INTRODUCTION ______

In the CBFM programme, the village works in establishing rules and regulations to be written in their Fisheries Management Plan (FMP), ban the use of destructive fishing methods, monitor and protect the reef area, and implement other actions to protect the environment. DMWR provides technical assistance and advice, workshops and training, and other appropriate fishery support such as the restocking of 100 giant clams in the MPA which enhances the development of good fisheries practice and management approaches (Sauafea, 2002).

Committees from each village have designed and implemented management plans which vary from village to village, but generally include some type of restriction on fishing activity on part or the entire reef fronting their villages. In addition, each village with a MPA has a monitoring and enforcement committee which have been helpful in reporting illegal fishing and other destructive activities observed in the reef area. The process from initial contact with the village to the final production and overseeing of a village FMP is summarised in Figure 1.9.

1) Initial contact and village council meeting

2) Village Group Meetings (GMs) (to identify problems and propose solutions) - includes participatory survey of marine and environment and resources

3) Fisheries Management Advisory Committee (FMAC) (to prepare a management plan with undertakings necessary to solve problems)

6) Community undertakings 4) VILLAGE 7) DMWR undertakings may include: FISHERIES may include: village rules MANAGEMENT technical assistance and banning destructive PLAN advice fishing methods (agreed to at the workshops/training size limits on fish fono meeting) appropriate fishery fish reserves support environmental protection

5) Monitoring and Enforcement Committee (MEC) (to oversee, monitor and enforce the undertakings agreed to in the management plan)

Figure 1.9 The extension process for the Community-Based Fisheries Management Programme in American Samoa (Source: Sauafea, 2002).

25 1 INTRODUCTION ______

The end result of the CBFMP will be a village with a Fisheries Management Plan with guidelines and regulations to monitor and protect its reef area, a productive and healthier reef area, improved fisheries, and an increase in awareness, motivation, consultation and participation from different stakeholders. The programme will increase conservation awareness in government, community, and the private sector. It will aid in recovering the reefs and maintain or increase fish abundance. In addition, the programme allows the community to keep a close watch on the marine resources and their condition so that the resources will continue to be healthy and productive for its people and future generations to come (Sauafea, 2002).

As the programme progresses, some villages have requested the need to have the programme started in their village because of the condition the reef area is in and just the need to improve their fisheries. The significance of the marine environment to the village, the extent of any problems with the marine environment (i.e. fish catches), and the level of concern and willingness to do something about the existing problems, are also important criteria for selecting a village for the programme (Sauafea, 2002).

Plate 1.11 Group meetings between DMWR and villagers to discuss the CBFMP (Source: DMWR).

During the present study four community-based village MPAs were surveyed; Alofau, Auto & Amaua. Vatia, and Aua.

26 1 INTRODUCTION ______

1.4.3.1 Aua The Marine Protected Area for the village of Aua was identified and marked on May 22nd, 2002. The main purpose of the CB village FMP is ‘To manage, protect and preserve the fish, shellfish and the coastal area of the village of Aua’. Currently, the status of Aua village’s participation in the program is currently unknown. The DMWR office have not been able to reach village representatives to make further progress on the Aua program.

1.4.3.2 Alofau The management area for Alofau village extends out beyond the reef flat over the drop off. The reef flat is extremely broad, with several deep areas that were dredged to mine sand and rock for construction uses. The main topic of Alofau’s Fisheries Management Plan is ‘To conserve the marine resources in the ocean and on the village reef’. The first Alofau management plan called for closing the entire area to all fishing activity beginning in May of 2001. After two years of complete closure, the village decided to open the area to fishing on Saturdays only.

Figure 1.10 Boundaries of Aua marine Figure 1.11 Boundaries of Alofau management area. marine management area.

1.4.3.3 Auto & Amaua Located on the south shore of Tutuila, Auto and Amaua villages have a shared management area. The main purpose of the CB village FMP is ‘To manage, protect and preserve the fish, shellfish and the coastal area of the village of Amaua and

27 1 INTRODUCTION ______

Auto’. No fishing of any kind has been allowed within the boundaries of the Amaua/Auto management area since February 2003.

1.4.3.4 Vatia The northernmost village of Tutuila, Vatia is geographically large (175.5 acres). The management area of the village consists of the entire northeast coast and has been closed to all fishing activities since April 2002. The main purpose of Vatia’s FMP is ‘To manage, protect and preserve the fish, shellfish and the coastal area of the village of Vatia’. Fish and shellfish in the reef area are not as common due to overfishing and the destruction of their habitats by illegal fishing activities. Following two years of closure, the villagers decided to reopen the reserve for one day earlier this year. They reported a great day of fishing, but the area was fished out and the reserve closed the next day. The council are in the process of deciding when and how to open the reserve.

Figure 1.12 Boundaries of Auto & Amaua Figure 1.13 Boundaries of Vatia marine marine management area. management area.

Each village has carried out many fishing activities on its reef using different fishing methods. The most common fishing methods and activities used by the villagers are detailed in Table 1.3.

28 1 INTRODUCTION ______

Table 1.3 Common fishing methods previously used by villagers.

Village Fishing methods Auto & Amaua Fishing nets, hand-lining, free-dive, night-fishing using torches or Vatia flash lights, spear fishing, fishing with bamboo or rod, and knives Aua for gleaning

Alofau Spear fishing, gill nets, throw nets, night fishing, traditional fishing for the akules using coconut leaves, and bottom fishing using hand-lining.

1.5 UNDERLYING RATIONALE Village communities in Samoa have for centuries managed their reefs in accordance with traditional fishing and stewardship practises. However, the population in American Samoa is rapidly increasing which places an increased pressure on the marine environment. Almost all important food species on Samoan reefs have declined due to technological developments, overexploitation, the use of destructive fishing methods, poor land management and environmental disturbances (Sauafea, 2002; Hawkins et al., 2003).

A CBFMP, administered by the American Samoa Department of Marine and Wildlife Resources, was implemented in 2001. By establishing MPAs in the participating villages, the programme aims to improve sustainable development of marine resources by improving habitat quality and increasing fish abundance.

Previous studies have shown higher coral cover in sanctuary compared to non- sanctuary areas. Epstein et al. (1999) documented higher coral cover in a ‘no-use zone’ compared to open areas in the Northern Red Sea; McClanahan et al. (1999) found coral cover in northern Tanzania was 20% lower in unprotected sites than in protected sites, however, such differences were not found to be significant. They suggested that the difference in findings may reflect the different initial conditions of the reefs, and the effects of coral physical disturbance on coral abundance.

Musburger (2004) developed recommendations for methodologies that could be used to monitor the coral reef resources of the villages participating in the CBFMP in

29 1 INTRODUCTION ______

American Samoa. Many of the participating villages have reported anecdotally that they have seen improvement in their resources since the program’s commencement, but as of yet, there has been no quantitative analysis of the effectiveness of the management efforts by scientifically monitoring these areas (Musburger, 2004). Furthermore, two villagers who joined the programme have reopened their closed areas after 2 years of co-managing their reef area. It was noted that the fish abundance and catches was greater. However, no empirical data exists with which to statistically illustrate the observed difference (Hawkins et al., 2003). It is hypothesised that the effects of protection will influence the coral reef communities surveyed.

1.6 OBJECTIVES The objectives of this survey were:

1. To lay out transects 25m in length in a direction parallel to the shore at survey sites under different levels of protection, in American Samoa, during June to August 2004, in order to: i. Provide a quantitative description of the coral reef communities in terms of reef substrate cover. ii. Describe the coral reef communities in terms of the presence and relative abundance of hard coral species. iii. Estimate abundances of key macroinvertebrates that are considered to be of local and/or national importance, to include giant clams, crown- of-thorns starfish and sea urchins.

2. Using the data collected and GIS software; examine the biological, environmental, physical and social factors of each of the sites surveyed in American Samoa during June to August 2004.

3. During the analysis of the collected data, investigate the relationship between: i. Coral cover and fish abundance; ii. Coral species diversity and fish species diversity; and iii. Coral species richness and fish species richness.

30 1 INTRODUCTION ______

4. Using data collected from the above objectives, critically evaluate the effects of protection on coral reef community structure of CB MPA sites, non MPA sites and federal MPA sites in American Samoa.

5. Using the data collected, provide recommendations to DMWR of potential sites which should be taken into consideration for future participation in the CBFM programme.

31 2 MATERIALS AND METHODS ______

2.1 COLLABORATION WITH OUTSIDE BODY This study was undertaken with the cooperation of the American Samoa DMWR. It is the responsibility of DMWR to manage, protect and preserve the marine and wildlife resources in the territory. Assistance by the department included transport, staff assistance in the field and office, field equipment, office space and library resources. The American Samoa DOC supplied GPS (Global Positioning System) equipment and satellite imagery.

No permits were required during field work. However, a member of the community of each village or the Matai (village chief) was asked for permission to swim in the water.

2.2 SURVEY SITES 2.2.1 Location of Survey Sites This study will focus on the two volcanic islands of Tutuila, and Ofu of the Manu’a Islands. A total of 11 sites were surveyed; ten sites were located on the main island of Tutuila (Figure 2.1) and one site was located on Ofu Island (Figure 2.2).

Due to scuba restriction and time constraints, the focus of the sites in this project were those most readily accessible by snorkelling which meant that the main coral habitat surveyed were the reef flat and lagoons.

2.2.2 Description of Survey Sites The name, type of habitat and level of protection of each site, as well as a number of environmental, physical and social factors of each site, are detailed in Table 2.1. Area of habitat was calculated from the benthic habitat map of American Somoa using the GIS (Geographical Information System) software Arcview. Habitat width, watershed area, population density and number of streams (Appendix 1) were extracted from the American Samoan GIS databank also using GIS software. Fishing pressure was sourced from Spurgeon et al. (2004) (Appendix 2). Descriptions of each survey site are detailed in section 2.4.

32 Nu’uuli

Vatia

Masefau

Faga’itua

Fagatele Bay Auto & Amaua Aua Alofau

Aua Control

Figure 2.1 Map of the main island of Tutuila, American Samoa, showing location of each survey site. Multispectral IKONOS imagery provided under licence to the American Samoan Government by: 33 Location of the pools in Ofu Lagoon (map produced by NPAS).

OFU

Figure 2.2 Map of the Manu’a Islands, American Samoa, showing location of survey site. The site is located in lagoon pool 400 on Ofu Island. The National Park boundary is also illustrated. Multispectral IKONOS imagery provided under licence to the American Samoan Government by:

34 Table 2.1 Environmental, physical and social factors of each site surveyed in American Samoa during June to August 2004. (CB MPA = Community-based Marine Protected Area). Area of habitat was calculated from the benthic habitat map of American Somoa using the GIS (Geographical Information System) software Arcview. Habitat width, watershed area, population density and number of streams were extracted from the American Samoan GIS databank using GIS. Fishing pressure was sourced from Spurgeon et al. (2004).

Site Protection Habitat Exposure Habitat Habitat Fishing Pollution No. Watershed area Width (m) Pressure Streams Area Pop (Acres) (km2) (per km2) Alofau CB MPA Reef flat SE 36.69 Min 340 High Med 4 1.33 495 Max 320 Alofau CB MPA Lagoon SE 11.56 Min 50 High Med 4 1.33 495 Max 150 Aua CB MPA Reef flat SE 38.294 Min 200 Low High 4 2.83 2193 Max 320 Aua Control Non MPA Reef flat SE 24.492 210 Low High 4 2.83 2193

Auto & CB MPA Reef flat SE 38.29 Au 220 High Med 3 Au 0.83 Au 258 Amaua Am 250 Am 0.87 Am 102 Faga’itua Non MPA Reef flat SE 43.63 Min 410 High Med 4 1.4 483 Max 510 Vatia CB MPA Reef flat NE 34.67 Min 110 Med Low 5 5.69 648 Max 140 Masefau Non MPA Reef flat NE 33.12 Min 145 Med Low 4 4.19 435 Max 230 Nu’uuli Non MPA Lagoon SE 9.953 Min 260 High High 3 7.56 5154 Max 340 Ofu Federal MPA Lagoon SE - Min 90 Med Low 0 0 0 Max 150 Fagatele Bay Federal MPA Reef flat SW 11.28 Min 90 Med Low 0 0 0 Max 150

35 2 MATERIALS AND METHODS ______

2.2.3 Site Selection Five CB MPA sites, four non MPA sites and two federal MPA sites were surveyed. CB MPA sites (villages currently participating in CBFMP) were selected based on accessibility and tide constraints. The benthic habitat map of American Samoa (Appendix 3) and GIS were used to select the non MPA sites in order that they could be similar to CB MPA sites to enable comparisons; the main criteria used in selection were exposure, area of habitat and proximity to CB MPA sites.

2.3 HABITAT DECRIPTION METHODS Surveys were carried out during June to August 2004. At each site a reconnaissance swim was conducted by both surveyors. This was to determine suitable coral reef areas, as each transect was to be laid over an area of continuous, well-developed or dense coral reef where possible. The duration of the reconnaissance swim was between 30 minutes and an hour depending on the size of the site to be surveyed. As a safety precaution both surveyors swam together at all times.

2.3.1 Point-Intercept Transect Method At each site, the benthic communities of the coral reef habitat were described using a Point-Intercept Transect method, which is useful in quantitatively surveying reef substrate. This method was selected from a protocol previously used by Musburger (2004) (see introduction). It will provide an estimate of the percent cover of each reef substrate type on each of the transects.

In order to avoid bias when sampling, all transects were selected randomly within coral reef habitats. A total of 5 replicate transects was conducted at each survey site (Plate 2.1). It was aimed to complete 5 transects at each site but due to weather conditions and time constraints it was not always logistically possible to complete all 5 transects on the same day. Geographic co-ordinates (on WGS84 datum) were taken of each transect at every survey site using a hand held GPS (Global Positioning System). These co-ordinates (presented in Appendix 4) were used to plot the location of each transect at every survey site on rectified satellite images of the islands using Arcview GIS software (Figures 2.3 to 2.13).

36 2 MATERIALS AND METHODS ______

At large embayment sites such as Masefau and Vatia, transects were conducted on both sides of each bay to gain a broader representation of the area. Transects conducted at Auto & Amaua (two separate villages which have a shared management area) were also distanced far apart for the same reason.

At the CB MPA sites, permanent stakes delineating the beginning and end of each transect, were driven into the substrate for the first 3 transects. These permanent transect stakes were marked using luminous tape and fishing floats as they will be monitored by villager members with the assistance of DMWR in the future. Transect stakes could be located in the future by GPS also. The remaining two transects were secured using dive weights at the beginning and end of each transect. All 5 transects at non MPA sites and federal MPA sites were also secured using dive weights.

Plate 2.1 Example of a transect using a 25m tape measure at Ofu Lagoon, American Samoa.

The end of the transect tape was secured to the permanent stake marking the beginning of the transect. The fish surveyor swam along the length of the transect reeling out the transect tape 25 metres to secure it to the stake marking the end of the transect. Each transect was laid down in a direction parallel to the shore. It was aimed to keep them as straight as possible but strong currents did not always allow

37 2 MATERIALS AND METHODS ______them to be perfectly straight. The transect was left for about ten minutes as many of the fish were disturbed and swam away during the initial roll out of the transect tape. During this time the second transect was laid out by both surveyors.

On returning to the first transect, the fish surveyor would swim along the length of the transect recording fish species abundances, and lengths, that were observed within the boundaries of a 2m wide belt. Benthic communities and key macroinvertebrates were surveyed along the same transects after the fish counts were completed. The coral surveyor would swim along the transect, and identify the type of substrate found immediately below the survey tape at 0.5m intervals, yielding 50 sample points per 25m transect. At each point, the reef substrate was classified as belonging to one of 15 substrate categories which are detailed in Table 2.2. The cover of each substrate category could then be calculated as the percentage of the 50 points that it occupied on each transect. Habitat characteristics were then compared among sites based on the cover of each major substrate category.

Table 2.2 Description of the reef substrate categories and codes used in the surveys carried out in American Samoa.

Category Code Description Sand S Rubble RB Rock RK Dead coral DC Recently dead, white to dirty white Dead coral with algae DCA Dead coral, still recognisable from its original form with some algal growth Live coral LC Coralline algae CA Macroalgae MA >1 cm tall Halimeda HA Turf algae TA Lush filamentous algae, <1 cm tall Soft coral SC Soft bodied corals Sponges SP Other OTH

The 13 substrate categories in this survey were a slight modification from Musburger’s (2004) method which incorporated only 6 categories (rubble, algae, live

38 2 MATERIALS AND METHODS ______coral, sand, rock, coralline algae). The substrate category live coral was further identified to species or genus level where possible, to estimate coral abundance and diversity. All information was recorded on pre-printed survey record forms (Appendix 5), which were attached to a clipboard (Plate 2.2). On completion of the Point-Intercept transect, the coral surveyor would reel up the transect tape.

Due to the rugose nature of coral reef environments, occasionally the survey tape was not laying directly on the substrate. If it was not clear what type of substrate was laying directly below the tape at a 0.5 meter interval, to avoid bias, a small weight attached to a piece of string was dropped from a point directly above the survey mark. The type of substrate that it landed on was then recorded for that data point.

Plate 2.2 Surveyor recording substrate during the Point-Intersect transect method at Ofu Lagoon, American Samoa (Photo: A. Lawrence).

2.3.2 Macrobenthic Invertebrate Survey Information was collected on key sessile invertebrates including COTS, giant clams and sea urchins. After reaching the end of the point-intercept transect method for substrate cover, the coral surveyor would return along the transect line and record key invertebrates occurring within 1m on both sides of the transect line. Due to time

39 2 MATERIALS AND METHODS ______constraints, the abundance of the invertebrates was recorded using the SACFOR scale (Table 2.3). The intensity of the search was low and only the inverts visible on the substrate surface were recorded (i.e. no searches were made within coral structures). During the invertebrate survey, the number of trash items within a 2m wide belt was also recorded. All information was recorded on survey record forms.

Table 2.3 SACFOR scale used for recording sessile invertebrate abundance in the American Samoa surveys.

Scale Individuals Description 0 0 Absent 1 1 Rare 2 2-10 Occasional 3 11-20 Frequent 4 21-50 Common 5 51-100 Abundant 6 >100 Superabundant

2.3.3 Environmental and Physical Parameters A number of environmental and physical parameters at each site were also recorded on the survey record form. The parameters were the name of the site, date, weather, time and height of high tide, minimum and maximum depth, water temperature, salinity and secchi reading (turbidity). The information collected was recorded on the survey form. Notes and descriptions about the coral reef habitat (i.e. colony size) at each transect were also recorded.

40 2 MATERIALS AND METHODS ______

2.4 SURVEY SITE DESCRIPTIONS

MASEFAU (non MPA) Habitat type: Reef flat Location: A semi-protected embayment on the northeast side of Tutuila.

Description: Transects were conducted on the outer reef flat just before the slope. The substratum was sand and mainly rubble, with relatively high coral cover situated on irregular coralline blocks. On the west side of the bay coral colonies were larger and dominated by Porites cylindrica, P. rus and encrusting Montipora sp. On the east side, coral colonies were smaller but much more diverse. Common coral species were Millepora, Galaxea fascicularis and Pocillopora sp. Current strength in the bay was moderate with some wave action .

Figure 2.3 Satellite image of Masefau indicating the location of the 5 transects.

41 2 MATERIALS AND METHODS ______

VATIA (CB MPA) Habitat type: Reef flat Location: A semi-protected embayment on the northeast side of Tutuila.

Plate 2.3 Vatia village and marine management area.

Description: Transects were conducted by the edge of the reef flat just before the slope and the wave breaking zone. Three transect were on the west side of the bay were currents were stronger compared to the east side. On both sides of the bay reef structure consisted of a coralline blocks with high hard coral cover. Large colonies of branching Porites cylindrica and Porites rus dominated with moderate Pocillopora sp. present also (mainly small colonies). A small proportion of hard coral was dead and covered with algae. Visibility was good with turbidity ranging between 10 and 14m.

Figure 2.4 Satellite image of Vatia indicating the location of the 5 transects.

42 2 MATERIALS AND METHODS ______

ALOFAU (CB MPA) Habitat type: Reef flat dispersed with lagoon pools Location: Southeast side of Tutuila

Plate 2.4 Alofau marine management area (Photo: Y. Umezawa)

Description: Coralline blocks with very high live coral cover and much recently dead coral (branching Acropora sp.) covered with algae. Similar to the lagoon habitat, dominant species of coral were branching Porites cylindrica and Acropora microphthalma with many small colonies of Pocillopora damicornis and Pavona frondifera. Turbidity was poor to moderate ranging between 5 and 15m.

Figure 2.5 Satellite image of Alofau indicating the location of the 5 transects.

43 2 MATERIALS AND METHODS ______

ALOFAU Habitat type: Lagoon Location: Southeast side of Tutuila

Plate 2.5 Alofau lagoon underwater (Photos: D. Fenner).

Description: Sand and rubble substratum with scattered massive Porites bommies. Relatively high live coral cover and much recently dead coral (branching Acropora sp.) covered with algae. Branching Porites cylindrica and Acropora microphthalma were dominant with many small colonies of Pocillopora damicornis. Rubbish was frequently observed on the lagoon floor. Turbidity was poor to moderate ranging between 5 and 15m, and the current was weak.

Figure 2.6 Satellite image of Alofau indicating the location of the 5 transects.

44 2 MATERIALS AND METHODS ______

FAGA’ITUA (non MPA) Habitat type: Reef flat with some deeper pools Location: Southeast side of Tutuila

Plate 2.6 Faga’itua marine area (Underwater Photo: D. Fenner).

Description: Very high live coral cover situated on a sand and rubble substratum with some patches of dead coral covered in algae. Large colonies of branching Acropora sp. (mainly A. microphthalma) were dominant with frequent smaller colonies of Pocillopora and Pavona sp. Visibility was excellent due to low turbidity (20m) and currents were very strong.

Figure 2.7 Satellite image of Faga’itua indicating the location of the 5 transects.

45 2 MATERIALS AND METHODS ______

AUTO & AMAUA (CB MPA) Habitat type: Reef flat Location: Southeast side of Tutuila

Plate 2.7 Auto & Amaua marine management area (Underwater Photo: D. Fenner).

Description: Mainly dead coral rubble substratum with some coralline blocks. Hard coral cover was low, dominated by massive Porites microatoll corals and a few small colonies of branching Acropora sp. A small proportion of hard coral was dead and covered with algae. Macroalgae was common, comprised mainly of Halimeda. Sea urchins and were abundant throughout. Currents were strong with little wave action and turbidity ranged between 10 (Amaua) and 15m (Auto).

Figure 2.8 Satellite image of Auto & Amaua indicating the location of the 5 transects.

46 2 MATERIALS AND METHODS ______

AUA (CB MPA) Habitat type: Reef flat Location: East side of Pago Pago Harbour on the southeast side of Tutuila

Plate 2.8 Aua marine management area.

Description: Rubble substrate with very low coral cover. Small colonies of table Acropora and branching Pocillopora were the dominant coral species. On the outer reef flat coral colonies were very few but large. Coralline algae, turf algae and macroalgae (mainly Halimeda) were dominant and sponge was frequent throughout. Currents were (very) strong and turbidity was high ranging between 4 and 5m. Rubbish which was mainly glass bottles and aluminium cans was also frequent at this site.

Figure 2.9 Satellite image of Aua indicating the location of the 5 transects.

47 2 MATERIALS AND METHODS ______

AUA CONTROL (non MPA) Habitat type: Reef flat Location: East side of Pago Pago Harbour on the southeast side of Tutuila.

Plate 2.9 Aua control village and marine area.

Description: Sand and coral rubble substratum with a few scattered rock boulders. Live coral cover was very low with macroalgae (mainly Halimeda) being the dominant species. Coral dominated by small colonies of Pocillopora damicornis and massive Porites sp. Rubbish was frequent throughput. Turbidity was very high (<1.5m) with high suspended sediment loads and weak currents.

Figure 2.10 Satellite image of Aua control indicating the location of the 5 transects.

48 2 MATERIALS AND METHODS ______

NU’UULI (non MPA) Habitat type: Lagoon Location: Southeast side of Tutuila

Plate 2.10 Nu’uuli lagoon underwater (Photo: D. Fenner) and marine area (Photo: A. Lawrence).

Description: Mostly sand substratum with patches of rubble. Coral cover was relatively high with some dead coral (branching Acropora sp.) covered in algae. Low coral diversity dominated by Pavona frondifera and Porites cylindrica. Small colonies of Alveopora sp. not observed on any other sites were also present, as well as frequent patches of soft corals, zooanthids and occasional rubbish. The current was weak and turbidity was moderate (<9m).

Figure 2.11 Satellite image of Nu’uuli lagoon indicating the location of the 5 transects.

49 2 MATERIALS AND METHODS ______

FAGATELE BAY (federal MPA) Habitat type: Reef flat Location: Southwest side of Tutuila.

Plate 2.11 Fagatele Bay National Marine Sanctuary.

Description: Coralline rock substratum with some turf algae and very high hard coral cover. Large branching Porites cylindrica and Pavona frondifera (divaricata) were dominant with many diverse small coral colonies also. Visibility was excellent due to low turbidity (21m) and currents were weak.

Figure 2.12 Satellite image of Fagatele Bay indicating the location of the 5 transects.

50 2 MATERIALS AND METHODS ______

OFU (federal MPA) Location: Southeast coast of Ofu Island Habitat type: Reef flat

Plate 2.12 Ofu lagoon, the National Park of American Samoa.

Description: Sand and rubble substratum with irregular coralline blocks. Moderately high hard coral cover and high diversity dominated by massive Porites sp. Common coral species were Goniastrea retiformis, encrusting Montipora sp. and Porites sp. 2 which was only observed at this site as well as the rare blue coral Heliopora. Low turbidity and very sheltered in terms of waves, wind and weak currents.

Figure 2.13 Satellite image of Ofu lagoon indicating the location of the 5 transects.

51 2 MATERIALS AND METHODS ______

2.5 STATISTICAL ANALYSIS The benthic community data set was analysed using the statistical software package PRIMER (Plymouth Routines In Multivariate Ecological Research). Multivariate analysis techniques were used to understand and document the present status of each benthic community surveyed. For analysis purposes, the substrate category Halimeda was joined with the macroalgae category.

To determine which sites were similar, a cluster analysis was performed using the Bray-Curtis similarity matrix. The matrix was calculated on the square root transformed benthic substrate abundance data. The survey sites were linked into hierarchical groups that are more similar to each other using the group-average linkage technique, and was graphically represented as a dendrogram (Clark and Warwick, 2001).The similarity matrix describes the pattern of occurrences of each substrate category for each of the sample sites (Clark and Warwick, 2001).

The similarity matrix was then graphically represented using a multi-dimensional scaling (MDS) technique (Clarke and Warwick, 2001). MDS plots were created to represent the survey sites as points on a 2-dimensional configuration map to establish community differences associated with the different survey sites based on substrate abundance/cover. On these plots coral species, substrate categories, environmental parameters (i.e. turbidity) or diversity indices were superimposed as circles whose varying diameters reflect their proportions at each survey site.

Significant differences between sites based on substrate abundance were tested using PRIMER’s mulivariant equivalent of an ANOVA called ANOSIM for ‘analysis of similarities’. ANOSIM was also performed to compare differences between CB MPA sites, non MPA sites and federal MPA sites. p < 0.05 was considered significant.

The SIMPER procedure was performed to identify which substrate categories were principally responsible either for an observed clustering pattern or for differences between survey sites and confirmed to differ in community structure by the ANOSIM tests (Clark and Warwick, 2001).

52 2 MATERIALS AND METHODS ______

The same analysis was repeated using the hard coral species abundance data, and then the full benthic community data set (classifying corals by species and substrate by category) was analysed. All data was square-root transformed. Similarities of environmental variables were calculated from a normalised Euclidean distance matrix.

To examine relationships between, environmental parameters (turbidity) and substrate cover, and between coral and fish community characteristics, the statistical analysis RELATE was performed. Correlations were considered significant for relationships at p < 0.05.

A number of diversity indices were calculated for hard coral species abundance and fish species abundance at the 11 survey sites: i. Species richness (d) was calculated using Margalef’s index, and is a measure of the number of species present for a given number of individuals. ii. Diversity (H’) was calculated using the Shannon-Weiner Diversity Index. iii. Evenness (J’) was calculated using Pielou’s Index of Eveness, and describes how evenly the percent coverage was distributed throughout the community for any particular site iv. Dominance (λ) was calculated using Simpson’s Dominance Index, which emphasises changes in abundant species.

53 3 RESULTS ______

3.1 REEF SUBSTRATE 3.1.1 Reef substrate cover Figure 3.1 is a satellite image of American Samoa. The pie charts illustrate the percentage cover of each substrate category at the 11 sites surveyed, and provide a graphical representation of reef substrate cover across American Samoa. Raw data for each site are shown in Appendix 6. The dominant reef substrates found at all sites were live coral, rubble, sand, coralline algae, turf algae and macroalgae (Table 3.1).

Live coral was the dominant reef substrate category at 8 of the sites surveyed. It was recorded at all 11 sites with an overall mean of 45.6%. Coral cover was highest at Fagatele Bay, Faga’itua and Alofau reef flat (>65%), and lowest at Auto & Amaua, Aua and Aua control (11-26%). High rubble and macroalgae cover at these latter 3 sites distinguished them from all other sites. Rubble was infact the dominant substrate recorded at Auto & Amaua (31.2%) whilst macroalgae dominated at Aua control (30.4%). At all other sites macroalgae cover was <10%. Aua and Aua control were the most turbid sites with visibility <4m due to high levels of suspended sediment.

Aua, Vatia and Masefau were characterised by high coralline algae cover compared to all other sites were cover was low. Coralline algae dominated at Aua (20.4%) and was abundant at Vatia (26%) and Masefau (21.6%). Sponge cover was highest at Aua also (6.8%).

Dead coral with algae cover was highest at Faga’itua and Alofau reef flat, 18.8% and 14.4% respectively. Both Alofau and Nu’uuli lagoons were found to have a similar cover of dead coral with algae (10.4% and 10.8%) compared to Ofu lagoon were cover was low (2%). These three lagoonal sites, and Aua control were characterised by high sand cover (19-25%) compared to all other sites were cover was low (<4%).

3.1.2 Site Similarities The similarity of sites surveyed was evaluated based upon the mean abundance of the 12 substrate categories. PRIMER was used to group sites that are most similar by hierarchical clustering (Figure 3.2), and by MDS ordination (Figure 3.3). The resultant groupings are illustrated by the dashed and continuous line.

54 Figure 3.1

Aua Masefau Vatia

Alofau reef flat

Aua control

Ofu

Faga’itua

Auto & Amaua Alofau lagoon

S Rb Rk CA MA TA DC DCA SC SP OTH LC Fagatele Bay Nu’uuli

Figure 3.1 Satellite image of Tutuila and Ofu Island, American Samoa. The pie charts illustrate the percentage cover of each substrate category at the 11 sites surveyed during June to August 2004. Substrate categories were: S = sand, R = rubble, Rk = rock, CA = coralline algae, MA = Macroalgae, TA = Turf algae, DC = Dead coral, DCA = Dead coral with algae, SC = Soft coral, SP = Sponge, OTH = Other and LC = Live coral. Multispectral IKONOS imagery provided under licence to the American Samoan Government by:

Table 3.1 The mean percent cover (±SD) of each substrate category recorded at 11 coral reef survey sites in American Samoa during June-August 2004. S=Sand, RB=Rubble, RK=Rock, CA=Coralline algae, MA=Macroalgae, TA=Turf algae, LC=Live coral, DC=Dead coral, DCA=Dead coral with algae, SC=Soft coral, SP=Sponge and OTH=Other. The protection of each site is also shown (CB MPA = Community-based Marine Protected Area).

Protection Site S RB RK CA MA TA LC DC DCA SC SP OTH Vatia 0 4 0.4 26 1.2 6.8 52.4 0.4 7.6 1.2 0 0 (3.7) (0.8) (8.3) (1.6) (2.4) (7.3) (0.8) (6.0) (2.4) Auto and 3.2 31.2 1.6 7.6 16 5.6 26 0.4 8.4 0 0 0 Amaua (3.0) (10.9) (2.6) (5.5) (13.2) (6.2) (3.7) (0.9) (3.6) CB MPA Alofau 0 3.6 0 5.2 3.2 4 67.6 0.4 14.4 0 0 1.6 Reef Flat (5.0) (4.6) (3.3) (4.0) (10.2) (0.9) (10.7) (2.2) Alofau 18.8 8.8 0 0 4.8 2 50 0.8 10.4 0 0 4.4 Lagoon (5.8) (5.2) (3.9) (3.5) (7.9) (1.8) (7.8) (6.1) Aua 2.8 16.4 0 20.4 18.8 14 18.8 0 1.6 0 6.8 0.4 (3.0) (4.6) (4.6) (9.3) (3.2) (5.9) (1.7) (5.2) (0.9) Masafau 1.2 11.6 0 21.6 2 13.6 48 0.4 1.6 0 0 0 (1.8) (7.9) (11.4) (1.7) (7.4) (5.8) (0.9) (1.7) Aua 23.6 29.6 0 0.4 30.4 3.6 11.2 0 0.8 0 0.4 0 Non MPA Control (8.2) (9.9) (0.9) (5.0) (4.3) (3.3) (1.8) (0.9) Nu’uuli 21.2 9.2 0 0 0.4 6.4 45.2 0 10.8 2 1.6 3.2 (9.5) (3.6) (0.9) (4.8) (5.6) (11.2) (2.8) (2.6) (1.1) Faga’itua 2 1.2 0 1.2 4 4 66.4 0 18.8 0 0 2.4 (4.5) (2.7) (1.1) (3.2) (4.0) (5.7) (5.4) (3.3) Fagatele 0 0 0 12.4 4 12 67.6 0.4 3.6 0 0 0 Federal Bay (8.2) (2.8) (9.7) (5.4) (0.9) (2.6) MPA Ofu 24.4 8.8 1.6 6.4 5.2 3.6 48 0 2 0 0 0 (3.8) (2.3) (2.2) (1.7) (3.0) (5.0) (4.7) (3.5) % Total 8.8 11.3 0.3 9.2 8.2 6.9 0.3 7.3 0.3 0.8 1.1 45.6 3 RESULTS ______

Faga'itua Alofau reef flat1A Masefau 1B Vatia Fagatele Bay 1C Ofu 2A Nu'uuli 2B Alofau lagoon Aua 3A Auto & Amaua3B Aua control 3C 60 70 80 90 100 Bray-Curtis Similarity %

Figure 3.2 Cluster analysis showing classification of the 11 sites surveyed in American Samoa based on percent similarity of reef substrate (calculated using group-average clustering from the Bray-Curtis similarity matrix on the square-root transformed data). The three groups of sites separated at a 70% similarity (dashed line) and eight groups of sites separated at a 80% similarity (continuous line) are indicated. Substrate categories were rubble, sand, rock, dead coral, dead coral with algae, macroalgae, coralline algae, turf algae, sponge, soft coral, live coral and other.

Figure 3.3 A multidimensional scaling ordination plot showing classification of the 11 sites surveyed in American Samoa based on percent similarity of reef substrate (calculated using group-average clustering from the Bray-Curtis similarity matrix on the square-root transformed data). Groupings were determined from the cluster analysis in Figure 3.2, with similarities at the 70% level (dashed line) and at the 80% level (continuous line). Substrate categories were rubble, sand, rock, dead coral, dead coral with algae, macroalgae, coralline algae, turf algae, sponge, soft coral, live coral and other. The level of protection of each site is also shown (CB MPA = Community-based Marine Protected Area).

The MDS plot (Figure 3.3) shows the same site groupings as the dendogram in Figure 3.2. The sites are grouped according to the similarities as calculated using the

57 3 RESULTS ______

Bray-Curtis similarity matrix. The stress level 0.07 corresponds to a good ordination with no real prospect of misleading interpretation. The plot also shows the level of protection (CB MPA, non MPA or federal MPA) of each site.

Similarity among eight of the eleven sites was relatively high (>65%), reflecting the dominance by similar benthos such as live coral. Both classification and ordination produced three main groups of sites formed at around the 70% similarity level (dashed line); Group 1 (Vatia, Fagatele Bay, Masefau, Alofau reef flat and Faga’itua), Group 2 (Ofu, Nu’uuli and Alofau lagoon) and Group 3 (Aua, Aua Control and Auto & Amaua). At around the 80% similarity level, eight groups were formed (continuous line). Three of these groups were found to be the most similar sites; Faga’itua and Alofau reef flat (1A) both located on the south east side of Tutuila, (88.2% similarity). Also, Nu’uuli lagoon and Alofau lagoon (2B) on the south east side coast (85.2% similarity) and Masefau and Vatia (1B) on the north east coast (83% similarity). Ofu was most similar to both Nu’uuli and the Alofau lagoon at 77% similarity respectively. Aua, Aua control, both located in Pago Pago Harbour, and also Auto & Amaua located further along the south east coast were shown to have a similarity of 70.1%.

Bubble plots of 6 abundant substrate categories were superimposed on the MDS ordination to show the proportion of each substrate at the survey sites (Figure 3.4). The level of protection of the sites is also shown. The plots show that there appears to be relationships between some substrate categories and some sites. The blue circles in Figure 3.4 (a) and (b) represent Alofau and Faga’itua reef flat which were found to be the most similar sites with 88.2% similarity (Figure 3.2). These two sites had the highest abundance of live coral (with the exception of Fagatele Bay) and also the most dead coral with algae illustrated in Figure 3.4. The blue circles in (c) and (d) represent Aua, Aua control and Auto and Amaua which were found to have a similarity of 70.1% (Figure 3.2). These sites had the highest abundance of rubble and macro algae when compared to all other sites illustrated in Figure 3.4c and d, and also the lowest abundance of live coral illustrated in Figure 3.4a. The blue circles in (e) and (f) represent Fagatele Bay, Masefau and Vatia which were found to have a similarity of 79.9%, and Auto and Amaua and Aua which had a similarity of 76.4%

58 3 RESULTS ______

(Figure 3.2). These sites had the highest abundance of turf algae and coralline algae illustrated in Figure 3.4e and f.

(a) Live coral (b) Dead coral with algae

(e) Turf algae (f) Coralline algae

Figure 3.4 Multidimensional scaling ordination plots showing the classification of 11 sites in American Samoa based on percent similarity of reef substrate (calculated using group- average clustering from the Bray-Curtis similarity matrix on the square-root transformed data). Substrate categories = rubble, sand, rock, dead coral, dead coral with algae, macroalgae, coralline algae, turf algae, sponge, soft coral, live coral and other. The cover of each substrate category (a) – (f) is represented by the diameter of the red circles. Increasing size of circles illustrates increasing abundance of each reef substrate. The level of protection of each site is also shown (CB MPA = Community-based Marine Protected Area).

59 3 RESULTS ______

ANOSIM was performed to test for significant differences in reef substrate cover. This was tested on the 5 replicate transects per site, were indeed a significant difference was found (ANOSIM, R = 0.80, p < 0.001).

The pairwise tests detailed in Appendix 7 showed that the vast majority of sites, statistically, were significantly different in reef substrate cover. However the sites not significantly different were Vatia and Masefau (ANOSIM, R = 0.14, p = 0.175), Nu’uuli lagoon and Alofau lagoon (ANOSIM, R = 0.28, p = 0.056), and Faga’itua and Alofau reef flat (ANOSIM, R = -0.036, p = 0.46). Furthermore, these sites were also shown to be most similar according to the cluster analysis discussed previously (Figure 3.2).

3.1.3 Environmental Parameters A number of environmental parameters were measured at each survey site. Variables such as temperature and salinity were found to be fairly stable throughout the sites (Appendix 8) and were not analysed any further. The only parameter analysed in detail was Secchi disc reading which represents turbidity.

Figure 3.5 is a bubble plot with turbidity superimposed onto the MDS ordination in Figure 3.3 (showed the percent similarity of reef substrate at the 11 sites). At the time of survey, turbidity was highest at Aua control and Aua and lowest at Fagatele Bay and Faga’itua. Vatia and Masefau on the north coast also exhibited relatively clear water with good visibility. Sites with high turbidity were also characterised by low coral cover and high macroalgae cover, with rubble or sand substrata. In contrast sites with low turbidity were characterised by high coral cover and low macroalgae cover.

To determine whether turbidity was related to coral cover and macroalgae cover the statistical analysis RELATE was performed. A significant correlation was found between turbidity and coral cover (Rho = 0.376, p = 0.031). However, no significant correlation was found between turbidity and macroalgae cover (Rho = 0.046, p = 0.335).

60 3 RESULTS ______

Figure 3.5 A Multidimensional scaling ordination plot showing the classification of 11 sites in American Samoa based on percent similarity of reef substrate (calculated using group- average linking from the Bray-Curtis normalised Euclidean distance matrix on the square- root transformed data). Substrate categories = rubble, sand, rock, dead coral, dead coral with algae, macroalgae, coralline algae, turf algae, sponge, soft coral, live coral and other. Superimposed circles represent turbidity (secchi disc reading). Increasing size of circles illustrates decreasing turbidity or increasing visibility.

3.1.4 Comparisons of MPAs and non MPAs Sites were grouped according to their classification of protection; CB MPAs, non MPAs and federal MPAs. Figure 3.6 compares the percent mean cover of the 12 substrate categories in each group. The graph shows that a relationship between the different benthic substrates and the protection of the sites exists. Live hard coral cover was found to be the dominant substrate of all three groups. It was highest at federal MPAs (57.8%), followed by similar cover at CB MPAs (43%) and non MPAs (42.7%). Dead coral with algae also had similar cover at CB MPAs (8.5%) and non MPAs (8%), compared to a low 2.8% at federal MPAs. Likewise rubble was also highest at CB MPAs (12.8%) and non MPAs (12.9%) compared to 4.4% at federal MPAs. Coralline algae was highest at CB MPAs (11.8%) followed by 9.4% at federal MPAs and 5.8% at non MPAs. Macroalgae was highest at non MPAs (9.2%) followed by 8.8% at CB MPAs and 4.6% at federal MPAs. Turf algae was highest at federal MPAs (7.8%) followed by similar cover at non MPAs and CB MPAs, 6.9% and 6.5% respectively.

61 3 RESULTS ______

80.0

70.0

60.0

50.0 CB MPA

40.0 Non MPA Federal MPA Cover (%) Cover 30.0

20.0

10.0

0.0 S RbRkCAMATADCDCASCSPOTHLC Substrate

Figure 3.6 Comparisons of mean cover (± SD) of reef substrate recorded at community- based MPA sites (n=5), non MPA sites (n=4) and federal MPA sites (n=2) in American Samoa during June to August 2004. S = sand, Rb = rubble, Rk = rock, CA = coralline algae, MA = macroalgae, TA = turf algae, DC = dead coral, DCA = dead coral with algae, SC = soft coral, SP = sponge, OTH = other and LC = live coral.

An ANOSIM analysis was performed on the substrate data to determine whether there were was a significant difference between CB MPA sites, federal MPA sites and non MPA sites in relation to reef substrate cover. Statistically, no significant difference was found (ANOSIM, R = -0.245, p = 0.939).

A further SIMPER analysis was performed to determine which substrate categories contribute the most to the similarity of CB MPA sites, federal MPA sites and non MPA sites (Appendix 9). The greatest contribution to the similarity of each group of sites came from live coral; 32.9% for CB MPA sites, 35.5% for non MPA sites and 46.9 for federal MPA sites.

The substrate categories which contributed the most to the dissimilarity between CB MPA and non MPA sites were found to be coralline algae (15.8%) and sand (14.7%). Sand, rubble and coralline algae also contributed the most to the dissimilarity between CB MPA sites and federal MPA sites, and also non MPA sites and federal MPA sites.

62 3 RESULTS ______

3.2 CORAL SPECIES ABUNDANCE 3.2.1 Coral Cover and Species Richness The ASEAN-Australia Living Coastal Resource classification system describes the health of coral reefs. The 11 sites surveyed were classified according to this system shown in Table 3.2. Five of the sites were considered to be in good condition as coral cover was 50-70%. Three sites were in ‘fair’ condition (30-50%) and three sites were in ‘poor’ condition (0-30%).

Table 3.2 Coral classification of the 11 sites surveyed in American Samoa during June to August 2004. Ratings: Excellent (70-100%), Good (50-70%), Fair (30-70%) and Poor (0- 30%).

Site Rating Alofau Reef flat Good Alofau Lagoon Good Vatia Good Aua Poor Auto & Amaua Poor Masefau Fair Aua Control Poor Faga’itua Good Nu’uuli Fair Fagatele Bay Good Ofu Fair

Most of the coral communities at each site tended to be dominated by branching corals and massive corals, reflecting the extreme exposure to waves. Massive Porites, Porites cylindrica, Porites rus, Pocillopora damicornis, Acropora microphthalma and Pavona frondifera comprised the majority of coral cover throughout the study area and each one was a dominating species at one or more sites.

A total of 42 species of coral belonging to 21 genera and 11 families were recorded from all sites surveyed in American Samoa. The full list of coral species recorded at each site is provided in Appendix 10. Site species richness ranged from S = 4 at Aua control to S = 27 at Ofu shown in Figure 3.7b. The majority of sites surveyed had 12 or less species present with an overall mean of 12.5 for the 11 sites. The most

63 3 RESULTS ______common species which occurred at the majority of sites (>6) were massive Porites, P. cylindrica, P. rus, P. damicornis, A. microphthalma, P. frondifera, Pavona decussata, Montipora sp. (encrusting) and Leptastrea purpurea (see Appendix 11).

(a)

100

40 Coral cover (%) 20

(b)

Species richness (S) 5

0 VAT A&A ARF AL AUA FA AC MA NU FB OFU

Site CB MPA Non MPA Federal MPA Sites

Figure 3.7 Coral cover and species richness (mean and standard deviation) recorded at each site surveyed in American Samoa during June to August 2004 (n = 5 transects). Sites were VAT = Vatia, A&A = Auto & Amaua, ARF = Alofau reef flat, AL = Alofau lagoon, FA = Faga’itua, AC = Aua control, MA = Masefau, NU = Nu’uuli, FB = Fagatele Bay. The level of protection of each site is also shown (CB MPA = Community-based Marine Protected Area).

64 3 RESULTS ______

A cluster analysis of coral species abundance of the five replicate transects from each of the 11 sites was performed to investigate variability between transects within each site, illustrated in Figure 3.8

It is clear from the dendogram that within site variability is high at many of the sites surveyed. However, at Ofu and Aua control, transects had clustered together into a group. This indicates that these transects contain more similarity to each other than when compared to transects within other sites. At Vatia and Masefau, transects clustered together indicate that these sites were more similar to each other in terms of coral species abundances than to any other sites. Transects conducted at Aua and Aua control were clustered into one large group with about 30% similarity. The site with the greatest variability among transects appear to be Auto and Amaua. This is not surprising since Auto & Amaua are two separate villages which have a shared management area. Therefore, transects 1, 2 and 3 conducted at Auto were distanced far apart from transects 4 and 5 conducted at Amaua.

Bray-Curtis Similarity %

Sites

Figure 3.8 Cluster analysis showing classification of 5 replicate transects from each of the 11 sites surveyed in American Samoa based on percent similarity of coral species abundance (calculated using group-average clustering from the Bray-Curtis similarity matrix on the square-root transformed data). Vat = Vatia, A&A = Auto and Amaua, AL = Alofau lagoon, ARF = Alofau reef flat, AC = Aua control, MA = Masefau, FA = Faga’itua, NU = Nu’uuli, FB = Fagatele Bay and OF = Ofu.

65 3 RESULTS ______

3.2.2 Diversity Indices A range of diversity indices were calculated based on mean hard coral species abundance of the 11 sites (Appendix 12). Species richness calculated using Margalef’s index ranged from d = 8.18 at Ofu to d = 1.74 at Aua control with a mean value of d = 3.73 for the 11 sites (Figure 3.9a). The Shannon-Weiner Diversity ranged from H’ = 0.94 at Aua control to H’ = 2.8 at Ofu with an overall mean value of H’ = 1.89 (Figure 3.9b). Pielou’s evenness index ranged from J’ = 0.68 at Aua control to J’ = 0.86 at Auto & Amaua with a mean value of J’ = 0.78 for the 11 sites (Figure 3.9c). Simpson’s dominance index ranged from λ = 0.68 at Aua control to λ = 0.95 at Ofu (Figure 3.9d).

There were large differences in the dominance of coral species at the survey sites. For nine of the sites surveyed the Simpson index was greater than 0.75 with a mean value of λ = 0.83 for the 11 sites. This indicates that the majority of sites were dominated by 1 or more species. Aua control had the lowest dominance since this site also had the lowest diversity, species richness and live coral cover. Despite having moderate coral cover both Nu’uuli and Alofau lagoons also had low species richness (d = 2.25 and 2.49), diversity (H’ = 1.65 and 1.52) and dominance (λ = 0.8 and 0.69).

According to Margalef’s and Shannon’s indices the most diverse site was clearly Ofu, followed by Fagatele Bay and Masefau. Fagatele Bay had a higher species richness (d = 5.4) but slightly lower diversity (H’ = 2.16) than Masefau (d = 4.4, H’ = 2.31). Diversity was lower as live coral cover was greater at Fagatele Bay, 67.6% compared to 48% at Masefau. Aua also had high species richness (d = 4.46) and diversity (H’ = 2.06). All four sites had a high Simpson index as P. frondifera dominated at Fagatele Bay (λ = 0.86), P. rus dominated at Masefau (λ = 0.89), massive Porites dominated at Ofu (λ = 0.95) and table Acropora dominated at Aua (λ = 0.92). A high Simpson index was also found at Auto & Amaua, which was dominated by massive Porites (λ = 0.88), and Faga’itua which was dominated by A. microphthalma (λ = 0.84). Species diversity was above the mean value and species richness was below the mean value for both sites.

66 3 RESULTS ______

(a) 10 8 6 4 2 Species richness (d) richness Species 0

(b) 3

2.5

1.5

(H'loge) 1

0.5

Shannon-Weiner Diversity 0

0.6

0.4

0.2

Pieulous Eveness (J') 0

(d) 1

0.8 ) λ 0.6

0.4

0.2 ( Dominance

0 VAT A&A ARF AL AUA FA AC MA NU FB OFU

SITES CB MPA Non MPA Federal MPA

Figure 3.9 Diversity indices describing the coral community composition at 11 sites surveyed in American Samoa during June to August 2004. (a) Species richness, calculated using Margalef’s index (d), (b) Diversity, calculated using the Shannon-Weiner Diversity Index (H’), (c) Evenness, calculated using Pielou’s Index of Eveness (J’) and (d) Dominance, calculated using Simpson’s Dominance Index (λ). Sites were VAT = Vatia, A&A = Auto & Amaua, ARF = Alofau reef flat, AL = Alofau lagoon, FA = Faga’itua, AC = Aua control, MA = Masefau, N = Nu’uuli, FB = Fagatele Bay. The level of protection of each site is also shown (CB MPA = Community-based Marine Protected Area).

67 3 RESULTS ______

3.2.3 Site Similarities Cluster analysis was performed to identify similarities in coral species abundance (Figure 3.10). The MDS plot in Figure 3.11 shows the same groupings of the sites illustrated by the dashed and continuous line. The low stress of 0.06 indicates a good 2-dimensional representation of the multidimensional data.

Masefau Vatia 1 Fagatele Bay Alofau lagoon Alofau reef flat 2 Faga'itua Auto & Amaua Nu'uuli Aua control 3 Aua Ofu 4 20 40 60 80 100

Bray-Curtis Similarity % Figure 3.10 Cluster analysis showing classification of the 11 sites surveyed in American Samoa based on percent similarity of coral species abundance (calculated using group- average clustering from the Bray-Curtis similarity matrix on the square-root transformed data). The four groups of sites separated at a 45% similarity are indicated (dashed line).

Similarity was only 21.3% among all sites based on coral species abundance. This was much lower when compared to the similarity of 62.5% for substrate cover among all sites. Both classification and ordination produced four main groups of sites formed at around the 45% similarity level (dashed line). The resultant groupings show the following similarities:

Group 1 (located on the NE and SW coast of Tutuila) - Vatia, Masefau and Fagatele Bay (49% similarity). Group 2 (located on the SE coast of Tutuila) - Alofau reef flat, Alofau lagoon, Faga’itua, Nu’uuli and Auto and Amaua (58% similarity). Group 3 (located in Pago Pago Harbour on the SE coast of Tutuila) - Aua and Aua control (46.5% similarity).

68 3 RESULTS ______

The Ofu site located on the SE coast of Ofu was shown to be unique (Group 4). Furthermore, the most similar sites appear to be Alofau reef flat and Alofau lagoon (79% similarity), and Alofau reef flat, Alofau lagoon and Faga’itua (70% similarity).

Figure 3.11 A multidimensional scaling ordination plot showing classification of the 11 sites surveyed in American Samoa based on percent similarity of coral species abundance (calculated using group-average clustering from the Bray-Curtis similarity matrix on the square-root transformed data). Groupings were determined from the cluster analysis in Figure 3.10, with similarities at the 70% level (dashed line). The level of protection of each site is also shown (CB MPA = Community-based Marine Protected Area).

The location of the Ofu site on the MDS plot far from all others shows that it was distinct from all other sites. It was the least similar site which is not surprising since it is the only site not located on Tutuila Island. As discussed previously, it was the most diverse site, with the most hard coral species present (S=27). This contributed to its dissimilarity among all other sites as the majority of sites had less than 12 species present. No particular species was especially dominant at Ofu, massive Porites accounted for 22.5% of the coral community followed by Porites sp. 2 (11.7%) and Goniastrea retiformis (10.8%). Although fairly abundant at Ofu the latter two species occurred at no other sites.

Bubble plots of the nine most common coral species were superimposed on the MDS ordination (Figure 3.11) to show the proportion of each species at the 11 survey sites (Appendix 13). P. cylindrica was the most abundant hard coral with an overall mean

69 3 RESULTS ______abundance of 23.6% for the 11 survey sites (Appendix 12). It was the only species to occur at all 11 sites and was the dominant species at Vatia, Alofau lagoon and Alofau reef flat. At Alofau lagoon it accounted for a mean abundance of more than half of the total coral community, 54.4% respectively.

P. frondifera and A. microphthalma were the next two abundant hard corals accounting for an overall mean abundance of 11.7% and 10.8% respectively. A. microphthalma was recorded at 6 sites and was the dominant species at Faga’itua with a mean abundance of 35.5%. This species was also abundant at Alofau reef flat accounting for 29% of the coral cover. P. frondifera was recorded at nine sites and was found to be the dominant species at Nu’uuli lagoon and Fagatele Bay, 34.5% and 24.9% respectively.

The next abundant hard coral was P. rus with an overall mean abundance of 10%, followed by massive Porites (9.5%) and P. damicornis (7.7%). P. rus was recorded at 9 sites and was the dominant species found at Masefau (28.1%). It was also abundant at Vatia accounting for 25.2% of the coral cover. Thus, nearly 60% of the coral community at Vatia was represented by 2 species, P. cylindrica and P. rus. Massive Porites was recorded at 10 sites and was found to be the dominant species at Ofu lagoon (22.5%) and Auto and Amaua (33.8%). P. damicornis was recorded at 8 sites and was the dominant species at Aua control (53.6%). However, cover of this species was greater at Faga’itua and Alofau reef flat as live coral cover at these two sites was very high (>65%) compared to only 11% at Aua control.

Montipora sp. (encrusting) accounted for an overall mean abundance of 3.3%. Masefau and Vatia on the north side of the island had the highest abundance, 10.7% and 7.6% respectively. P. decussata was found to have an overall mean abundance of only 2%. Cover was highest at Auto & Amaua (10.8%) and Faga’itua (4.8%). Although cover of these species was generally low, they were recorded at six sites making them a common occurring species throughout the survey sites. L. purpurea was recorded at 7 sites but accounted for only 1.6% of the coral community surveyed. Thus cover of this species at the recorded sites was low, the highest abundance being at Ofu, 7.5% respectively.

70 3 RESULTS ______

Figure 3.12 is a graph showing the percentage cover of coral genera at each of the sites surveyed. It is clear that Porites sp. were the dominant genus throughout the sites surveyed in American Samoa. Acropora sp., Pocillopora sp. and Pavona sp. were also moderately abundant. Porites sp. was dominant at 8 sites accounting for more than 40% of the coral community. Cover ranged from 19.1% at Aua to 65.6% at Alofau lagoon. Acropora sp. dominated at 2 sites, Faga’itua (50%) and Aua (31.9%) but was also fairly abundant at Alofau reef flat, 33.7% respectively. Cover was lowest at Fagatele Bay and Aua control (3.6%). Pocillopora sp. dominated at Aua control accounting for 53.6% of the coral community, however live coral cover at this site was very low and therefore the abundance of Pocillopora sp. was infact low also. Abundance was greater at Vatia (19.8%) and Fagai’tua (15.1%). Other coral genera were abundant at Ofu accounting for 41.7%. This site was the most diverse site in terms of hard coral species.

100%

80%

Other 60% Montipora Pavona Acropora 40% Pocillopora Cover (%) Porites

20%

0% Nu FB FA AL AC MA ARF OFU VAT AUA A&A

CB MPA Site Non MPA Federal MPA

Figure 3.12 Graph showing the percent cover of dominant coral genera recorded at each site surveyed in American Samoa during June to August 2004 (n = 5 transects). Sites were VAT = Vatia, A&A = Auto & Amaua, ARF = Alofau reef flat, AL = Alofau lagoon, FA = Faga’itua, AC = Aua control, MA = Masefau, Nu = Nu’uuli, FB = Fagatele Bay. The level of protection of each site is also shown (CB MPA = Community-based Marine Protected Area).

71 3 RESULTS ______

An ANOSIM analysis was performed to determine whether there were site to site differences in coral species abundance. A highly significant difference was found (ANOSIM, R = 0.638, p<0.001). The pairwise tests detailed in Appendix 14 showed that the vast majority of sites were statistically significant from one another indicating that there were site differences in hard coral species abundance. However a number of sites were not significantly different; similar sites were Vatia and Masefau (ANOSIM, R = 0.18, p = 0.151), Aua and Auto & Amaua (ANOSIM, R = 0.234, p = 0.087), Faga’itua and Auto and Amaua (ANOSIM, R = 0.014, p = 0.365), Alofau lagoon and Alofau reef flat (ANOSIM, R = 0.124, p = 0.183), Faga’itua and Alofau reef flat (ANOSIM, R = 0.104, p = 0.206), Faga’itua and Alofau lagoon (ANOSIM, R = 0.316, p = 0.056) and Nu’uuli and Fagatele Bay (ANOSIM, R = 0.244, p = 0.079).

A further SIMPER analysis was performed to determine which coral species contribute the most to the similarity among sites summarised in Table 3.3 (see Appendix 15 for full list). Species found to contribute the most to one or more sites were P. cylindrica, P. damicornis, massive Porites, A. microphthalma, P. frondifera, P. rus, Montipora sp., Goniastrea retiformis and Porites sp. 2. The majority of these species, which are illustrated in the MDS bubble plots (Appendix 13), were common throughout American Samoa and have been discussed previously.

A number of these common coral species were responsible for the similarities of those sites which were shown to be significantly similar to one another based on the pairwise tests of coral species abundance. Clearly P.rus contributed the most to site similarity at Vatia (33.6%) and Masefau (30.2%). P. damicornis contributed the most to site similarity at Aua (25.2%) and Auto & Amaua (34.5%); Faga’itua (38.4%) and Auto & Amaua (34.5%); and Faga’itua and Alofau lagoon (13.3%). P. cylindrica contributed the most to site similarity at Alofau lagoon (59.1%) and Alofau reef flat (31.3). A. microphthalma contributed the most to site similarity at Faga’itua (23.8%) and Alofau reef flat (40.8%). Finally, both P. frondifera and P. cylindrica contributed the most to site similarity at Nu’uuli and Fagatele Bay.

72 3 RESULTS ______

Table 3.3 SIMPER analysis showing which coral species contributed the most to the similarity among each of the 11 sites surveyed in American Samoa during June to August 2004.

Site Hard coral species Contribution (%) Porites rus 33.61 Vatia Porites cylindrica 29.65 Pocillopora damicornis 34.45 Auto & Amaua Massive Porites 30.61 Acropora microphthalma 40.84 Alofau reef flat Porites cylindrica 31.27 Porites cylindrica 59.07 Alofau lagoon Pocillopora damicornis 13.27 Acropora sp. 35.44 Aua Pocillopora damicornis 25.15 Pocillopora damicornis 38.35 Faga’itua Acropora microphthalma 23.82 Pocillopora damicornis 67.79 Aua control Massive Porites 32.21 Masefau Porites rus 30.15 Montipora sp. 21.83 Nu’uuli Pavona frondifera 43.29 Porites cylindrica 36.28 Fagatele Bay Porites cylindrica 32.47 Pavona frondifera 31.59 Ofu Massive Porites 27.40 Goniastrea retiformis 19.77 Porites sp. 2 17.72

A further ANOSIM analysis was performed to determine whether there were differences in hard coral species abundance between CB MPA sites, non MPA sites and federal MPA sites. A significant difference was found (ANOSIM, R = 0.156, p<0.001). The pairwise tests indicated a significant difference between federal MPA sites and CB MPA sites (ANOSIM, R = 0.333, p<0.001), and a significant difference between federal MPA sites and non MPA sites (ANOSIM, R = 0.28, p = 0.004). No significant difference was found between CB MPA sites and non MPA sites (ANOSIM, R = 0.019, p = 0.236).

A SIMPER analysis was performed to determine which hard coral species contribute the most to the similarity of CB MPA sites, federal MPA sites and non MPA sites (Appendix 16). The four species P. cylindrica, P. damicornis, P. frondifera and massive Porites were the greatest contributors to the similarity for each group of

73 3 RESULTS ______

sites. P. cylindrica contributed the most at CB MPA sites (30.3%), P. damicornis contributed the most at non MPA sites (34.5%) and massive Porites contributed the most at federal MPA sites (20.7%). P. cylindrica the most abundant hard coral throughout the study area was also found to contribute the most to the dissimilarity between CB MPA and non MPA sites, CB MPA sites and federal MPA sites, and non MPA sites and federal MPA sites.

3.2.4 Relationship between Coral Species Figure 3.13 displays the results of the cluster analysis on the percentage similarity among hard coral species. To allow interpretable clustering rarer species and those with a very low abundance were removed from the analysis (Appendix 17)

6 7

Bray Curtis % similarity

Figure 3.13 Cluster analysis showing percent similarity of coral species surveyed at 11 sites in American Samoa during June to August 2004 (calculated using group-average clustering from the Bray-Curtis similarity matrix on the square-root transformed data). The seven g roups of sites separated at a 40% similarity are indicated (dashed line).

Seven main groups of sites formed at around the 40% similarity level (dashed line). The most similar species with 98% similarity appear to be Porites sp. 2 and G. retiformis (Group 2). These two species had a similar abundance only occurring at

74 3 RESULTS ______

Ofu which accounts for their near identical similarity. This is also confirmed by the fact that these two species contributed the most to site similarity at Ofu (Table 3.3). Furthermore, other similar species were Fungia sp. and Lobophyllia hemprichii (85% similarity), and Favia sp. and Leptoria phrygia (78% similarity). Again, the former two species were only found in low abundance at Fagatele Bay and Ofu. P. cylindrica and P. frondifera, the two most abundant coral species throughout the survey sites were shown to have a similarity of 77%. Two other commonly occurring species, P. damicornis and massive Porites exhibited 71% similarity. These two species were often found together with the exception that massive Porites occurred at the federal MPA sites whereas P. damicornis did not.

Five of the most abundant and commonly occurring species, P. frondifera, P. cylindrica, P. rus, P. damicornis and massive Porites were shown to have a similarity of 56%. The third most abundant species A. microphthalma was not shown to be as similar as this species was not found at as many sites as the latter species. Alveopora sp. was found to be least similar to all other species as it only occurred at Nu’uuli lagoon.

3.3 SUBSTRATE AND CORAL SPECIES ABUNDANCE Primer was used to group sites using the full community data set (reef substrate cover and coral species abundance). Cluster analysis was performed to identify similarities among sites (Figure 3.14). The MDS plot in Figure 3.15 shows the same groupings of the sites illustrated by the dashed and continuous line. The low stress of 0.11 indicates a good 2-dimensional representation of the multidimensional data.

Sites are grouped in the same way as in the cluster analysis of reef substrate in Figure 3.2. One exception is Ofu which was found to be the least similar site among all other sites which was also the case when the data set was restricted to coral species only (Figure 3.10 showing cluster analysis). This infers that hard coral species abundance at Ofu rather than substrate cover is responsible for why Ofu was the least similar site among all other sites. The overall similarity among sites based on the full data set was almost 40%.

75 3 RESULTS ______

1A 1B

3B 4

Figure 3.14 Cluster analysis showing classification of the 11 sites surveyed in American Samoa based on percent similarity of reef substrate cover and coral species abundance (calculated using group-average clustering from the Bray-Curtis similarity matrix on the square-root transformed data). Substrate categories = rubble, sand, rock, dead coral, dead coral with algae, macroalgae, coralline algae, turf algae, sponge, soft coral and other.

Figure 3.15 A multidimensional scaling ordination plot showing classification of the 11 sites surveyed in American Samoa based on percent similarity of benthic substrate cover and hard coral species abundance (calculated using group-average clustering from the Bray-Curtis similarity matrix on the square-root transformed data). Groupings were determined from the cluster analysis in Figure 3.14, with similarities at the 55% level (dashed line) and at the 65% level (continuous line). Substrate categories were rubble, sand, rock, dead coral, dead coral with algae, macroalgae, coralline algae, turf algae, sponge, soft coral, live hard coral and other. The level of protection of each site is also shown (CB MPA = Community-based Marine Protected Area).

76 3 RESULTS ______

An ANOSIM analysis based on the 5 replicate transects per site was performed on the full community data set to determine whether there were site to site differences. A highly significant difference was found between sites in reef substrate cover and coral species abundance (ANOSIM, R = 0.787, p<0.001). The pairwise tests detailed in Appendix 18 revealed only two pairs of sites that were not significantly different. These sites were Vatia and Masefau (ANOSIM, R = 0.232, p = 0.071) and Alofau reef flat and Faga’itua (ANOSIM, R = 0, p = 0.405).

The results of the SIMPER analysis in Table 3.4 indicated a number of findings (see Appendix 19 for full list). Firstly, sand was found to be a major contributor to the similarity of Nu’uuli lagoon, Alofau lagoon and Ofu lagoon thus discriminating them from other sites. Both Rubble and macro algae contributed the greatest to the similarity of two sites with low hard coral cover, Aua and Auto & Amaua. Hard coral species contributed the most to the similarity of three sites with the highest coral cover, Fagatele Bay, Faga’itua and Alofau reef flat. These species were P.cylindrica and P. frondifera at Fagatele Bay, A. microphthalma and P. cylindrica at Alofau reef flat and A. microphthalma and P. damicornis at Faga’itua. Dead coral and algae also contributed to the similarity of Alofau reef flat and Faga’itua. Coralline algae, turf algae and P. rus contributed the most to site similarity at two significantly similar sites, Vatia and Masefau.

A further ANOSIM analysis revealed significant differences between federal MPA sites and CB MPA sites (ANOSIM, R = 0.353, p<0.001), and between federal MPA sites and non MPA sites (ANOSIM, R = 0.18, p = 0.015). However, no significant difference was found between CB MPA sites and non MPA sites (ANOSIM, R = 0.027, p = 0.203).

77 3 RESULTS ______

Table 3.4 SIMPER analysis indicating whether hard coral species or reef substrate contributed the most to the similarity among each of 11 sites surveyed in American Samoa during June to August 2004.

Site Contribution (%) Coralline algae 30.3 Vatia Turf algae 15.3 Porites rus 12.5 Rubble 28.6 Auto & Amaua Macro algae 14.6 Dead coral with algae 14.6 Acropora microphthalma 25.4 Alofau reef flat Porites cylindrica 19.7 Dead coral with algae 18.1 Porites cylindrica 26.2 Alofau lagoon Sand 21.4 Dead coral with algae 12.4 Coralline algae 19.5 Aua Rubble 17.2 Macro algae 16.2 Dead coral with algae 20.3 Faga’itua Pocillopora damicornis 20.1 Acropora microphthalma 12.8 Macro algae 28.7 Aua control Rubble 26.1 Sand 23.3 Masefau Coralline algae 18.7 Turf algae 14.9 Porites rus 14.6 Nu’uuli Sand 19.8 Pavona frondifera 16.8 Porites cylindrica 14.1 Fagatele Bay Porites cylindrica 20.3 Pavona frondifera 19.8 Turf algae 12.9 Ofu Sand 20.3 Massive Porites 12.8 Rubble 11.7

3.4 KEY MACROINVERTEBRATES The abundance of key macroinvertebrates at the 11 sites surveyed in American Samoa is presented in Table 3.5. Sea urchins were the predominant macroinvertebrates occurring at the majority of sites. They were abundant at Auto & Amaua, occasional at Aua, Nu’uuli and Masefau, and rare at all other sites.

78 3 RESULTS ______

Giant clams were very rare at the survey sites on Tutuila. They were recorded at the CB MPA sites only; one was recorded at Vatia and one at Auto and Amaua. At Ofu they were recorded occasionally. One crown-of-thorns starfish was recorded at Masefau. Belt transects resulted in extremely low abundances of urchins, giant clams, and COTS, and no meaningful analysis came from these few data.

Table 3.5 Key macro invertebrate abundance at the 11 sites surveyed in American Samoa during June to August 2004. SAFOR scale, 1 = 1 (rare), 2 = 2-10 (occasional), 3 = 11-20 (frequent), 4 = 21-50 (common), 5 = 51-100 (abundant). The level of protection of each site is also shown (CB MPA = Community-based Marine Protected Area).

Protection Site Sea urchins Giant clams COTS Vatia 1 1 0 Auto & Amaua 4 1 0 CB MPAs Alofau reef flat 1 0 0 Alofau lagoon 1 0 0 Aua 2 0 0 Faga'itua 0 0 0 Aua C 1 0 0 Non MPAs Masefau 2 0 1 Nu'uuli 0 0 0 Fagatele 1 0 0 Federal MPAs Ofu 2 2 0

3.5 RELATIONSHIP BETWEEN CORAL AND FISH A range of diversity indices were calculated based on mean fish abundance of the 11 sites surveyed in American Samoa during June to August 2004 (Appendix 20). Data was collected by A. Lawrence. Fish abundance ranged from a high of N = 131 at Faga’itua to a low of N = 39 at Aua control (Figure 3.16), with a mean value of N = 82. Species richness calculated using Margalef’s index ranged from d = 5.74 at Aua control to d = 11.04 at Masefau with a mean value of d = 6.69 for the 11 sites. The Shannon-Weiner Diversity ranged from H’ = 2.18 at Faga’itua to H’ = 3.3 at Auto & Amaua with an overall mean value of H’ = 2.73.

79 3 RESULTS ______

250

200

150

100 Fish Abundance (N) 50

0 VAT A&A ARF AL AUA FA AC MA NU FB OFU

CB MPA Non MPA Federal MPA SITES

Figure 3.16 Fish abundance (mean and standard deviation) recorded at each site surveyed in American Samoa during June to August 2004 (n = 5 transects). Sites were VAT = Vatia, A&A = Auto & Amaua, ARF = Alofau reef flat, AL = Alofau lagoon, FA = Faga’itua, AC = Aua control, MA = Masefau, NU = Nu’uuli, FB = Fagatele Bay. The protection of each site is also shown (CB MPA = Community-based Marine Protected Area). Data collected by A. Lawrence.

Bubble plots of fish and coral abundance, species richness and species diversity were superimposed on the MDS ordination (Figure 3.15) to show the proportion of each at the 11 survey sites (Figure 3.17).According to Margalef’s and Shannon’s indices the most diverse sites were Masefau, Auto & Amaua and Ofu. The highest fish species richness was recorded at Masefau (d = 11.04) followed by d = 9.48 at Auto & Amaua and d = 9.35 at Ofu. Coral species richness was also high at Masefau and Ofu (Figure 3.17c and d). The bubble plots further illustrate similar proportions of fish and coral abundance at these two sites. Auto & Amaua was characterised by very high fish species diversity (H’ = 3.3), with high diversity also at Masefau and Ofu (H’ = 3.02). Similarly hard coral species diversity was high at Masefau and Ofu. Despite high fish abundance and species richness at Auto & Amaua, this site was characterised by low coral cover and low species richness (Figure 3.17).

80 3 RESULTS ______

(a) Fish abundance (b) Hard coral cover

(e) Fish species diversity (H’) (f) Hard coral species diversity (H’)

Figure 3.17 Multidimensional scaling ordination plots showing classification of the 11 sites surveyed in American Samoa based on percent similarity of reef substrate cover and coral species abundance (calculated using group-average clustering from the Bray-Curtis similarity matrix on the square-root transformed data). Substrate categories were rubble, sand, rock, dead coral, dead coral with algae, macroalgae, coralline algae, turf algae, sponge, soft coral, live hard coral and other. The magnitude of (a) – (f) is represented by the diameter of the superimposed circles. Increasing size of circles illustrates increasing abundance (a and b) or increasing species richness and diversity (c, d, e and f). d = species richness, calculated using Margalef’s index and H’ = diversity, calculated using the Shannon-Weiner Diversity Index.

81 3 RESULTS ______

The lowest fish abundance was recorded at Aua control and Aua, N = 38.7 and 45.5 respectively. Aua control also had low species diversity (H’ = 2.52) and the lowest fish species richness (d = 5.74), followed by Aua (d = 5.76). Similarily, both these sites had very low coral cover and species richness (Figure 3.17). The highest fish abundance was recorded at Faga’itua and Alofau reef flat, N = 131 and 128.8 respectively. Both these sites also had high coral cover (Figure 3.17a and b). Faga’itua was the least diverse site in terms of fish species (H’ = 2.18) with similar low coral species diversity as well (Figure 3.17e and f).

Fish populations at Alofau lagoon and Fagatele Bay were also characterised by low species diversity, H’ = 2.36 and 2.63 respectively. Fish abundance was above the mean value at Fagatele Bay (N = 85.4), and below at Alofau lagoon (N =69). As illustrated by Figure 3.17a and b coral cover was proportionally much higher at these sites.

Fish abundance at Vatia and Nu’uuli lagoon was fairly low, N = 51.2 and 58.8, where as coral cover was proportionally higher (Figure 3.17a and b) However, Nu’uuli was characterised by high species richness (d = 9.33) compared to Vatia (d = 6.86). Low coral species richness was also found at these two sites.

Fish abundance, species richness and diversity were correlated with hard coral cover, species richness and diversity (Figure 3.18). To determine whether there was a significant relationship between fish abundance (%) and hard coral cover (%), the statistical analysis RELATE was performed. A significant positive correlation was found (Rho = 0.303, p = 0.025), illustrated in Figure 3.18a. No significant correlation however, was found between fish species richness and hard coral species richness (Rho = 0.052, p = 0.333), and between fish species diversity and hard coral species diversity (Rho = -0.137, p = 0.694).

82 3 RESULTS ______

(a) % Fish abundance and % hard coral cover.

16 14 12 Rho = 0.303, p = 0.025 10 8

Corals 6 4 2 0 05101520 Fish

(b) Species richness (d).

9 8 7 6 Rho = 0.052, p = 0.333 5 4 Corals 3 2 1 0 024681012 Fish

2.5 Rho = -0.225, p = 0.694

1.5

Corals 1

0.5

0 00.511.522.533.5 Fish

Figure 3.18 Relationships between coral and fish at 11 coral reef sites surveyed in American Samoa during June to August 2004. d = species richness, calculated using Margalef’s index and H’ = diversity, calculated using the Shannon-Weiner Diversity Index.

83 3 RESULTS ______

3.6 SITE EVALUATION A score, using a scale of 1 (low) to 5 (high), was assigned to a number of biological, physical and social factors that might vary from site to site (Table 3.6). This allows for a quantitative measure to be placed upon each site. The factors are the same as those detailed in Table 2.1 of the methods section, and also include substrate cover and fish abundance. Scale tables are detailed in Appendix 21.

Sites which were found to be significantly similar to one another in terms of reef substrate and coral species abundances (e.g. Masefau and Vatia, and Alofau reef flat and Faga’itua) were found to have very similar scores. In fact Alofau and Faga’itua had identical scores for all factors with the exception of habitat area.

84 Table 3.5 Environmental, biological, physical and social factors of the 11 sites surveyed in American Samoa during June to August 2004. (CB MPA = Community-based Marine Protected Area). Exp = Exposure, PO = Pollution, Pop = Population Density. FP = Fishing Pressure (1 = Low, 2 = Medium, 3 = High), sourced from Spurgeon et al. (2004). Reef substrates are Rb=Rubble, S= Sand, CA=Coralline algae, TA=Turf algae, MA=Macroalgae, and LC = Live coral. Fish abundance was recorded by A. Lawrence. Scale 1 = Low, 5 = High (see Appendix 21 for detail).

Site Protection Reef Substrate Habitat Habitat Exp FP PO No. Watershed Fish Rb S CA TA MA LC Area Width Streams Area Pop (per Ab. (Acres) (m) (km2) km2) Alofau CB MPA 1 1 1 1 1 5 5 4 SE 3 Med 4 1 1 6 reef flat Alofau CB MPA 1 1 1 1 1 4 2 2 SE 3 Med 4 1 1 3 lagoon Aua CB MPA 2 1 2 1 2 2 5 3 SE 1 High 4 2 4 2

Aua Non MPA 2 2 1 1 3 1 3 3 SE 1 High 4 2 4 2 Control Auto & CB MPA 3 1 1 1 2 2 5 3 SE 3 Med 3 Auto 1 Auto 1 5 Amaua Am 1 Am 1 Faga’itua Non MPA 1 1 1 1 1 5 6 4 SE 3 Med 4 1 1 6

Vatia CB MPA 1 0 2 1 1 4 4 2 NE 2 Low 5 5 1 3

Masefau Non MPA 1 1 2 1 1 4 4 2 NE 2 Low 4 4 1 5

Nu’uuli Non MPA 1 2 0 1 1 4 1 4 SE 3 High 3 6 6 3 lagoon Ofu Federal 1 2 1 1 1 4 - 2 SE 2 Low 0 0 0 4 MPA Fagatele Federal 0 0 1 1 1 5 2 2 SW 2 Low 0 0 0 4 Bay MPA

85 4 DISCUSSION ______

4.1 REEF SUBSTRATE An objective of the present study was to provide a quantitative description of the coral reef communities, in terms of reef substrate cover, at each site surveyed in American Samoa during June to August 2004. The dominant reef substrates among all sites were live coral, rubble, sand, coralline algae and macroalgae. Live coral dominated at the majority of sites surveyed. The ASEAN-Australia Living Coastal Resource classification system describes the health of coral reefs. According to this system Alofau reef flat, Fagatele Bay, Faga’itua, Vatia and Aofau lagoon were considered to be in ‘good’ condition (50-70%). Masefau, Nu’uuli and Ofu were considered to be ‘fair’ condition (30-50%) and Auto & Amaua, Aua and Aua control were considered to be in ‘poor’ condition (0-30%).

Coral cover was highest at Alofau reef flat, Faga’itua and Fagatele Bay. These sites were the most protected from physical disturbances and probably have a higher potential for maximum coral cover because of low wave disturbance. Faga’itua and Alofau are relatively large reef flats located next to one another on the south east side of Tutuila. Whilst the majority of sites surveyed were found to be significantly different based on reef substrate cover, Alofau reef flat and Faga’itua were not shown to be significantly different.

Both biotic and abiotic factors can influence coral reef development such as geomorphology (Randall, 1985), watershed size and population (West and Van Woesik, 2001), and competition for space and resources (Benzoni et al., 2003). It is not surprising then that similarities in substrate cover and the apparent good health of Alofau and Faga’itua coral reefs may be a result of very similar watershed areas, similar low population densities and also large reef flats that may serve to dilute nutrient runoff.

Fagatele Bay National Marine Sanctuary has been protected since 1986 because of its isolation, spectacular beauty, and the pristine nature of its marine resources (Green et al., 1999). It is not surprising then, why coral cover was high and why the reef was considered to be in ‘good’ condition. Green (2002) also reported that the coral reefs of FBNMS have recovered well from past large scale natural disturbances, such as a COTS outbreak, four hurricanes and a mass coral bleaching

86 4 DISCUSSION ______event, and are now in good condition. The fast recovery of corals is probably due to the sanctuaries remote location away from human interference.

Aua control and Aua located in Pago Pago harbour had the lowest live coral cover (<20%). The most likely explanation for the ‘poor’ condition of these reefs is poor water quality in the harbour due to pollution from increasing population and industry in the area. Likewise, Green (2002) reported that sites in Pago Pago harbour have been heavily impacted by pollution when compared with other sites in American Samoa.

Among the important factors determining coral abundance, growth and distribution is sedimentation (Rogers, 1990). Both Aua and Aua control were the most turbid sites with visibility <4m due to high suspended sediment loads. Sedimentation is likely to have contributed to the relatively slow recovery of corals that has been observed in the harbour, because coral recruitment, juvenile survival and growth rates all tend to be lower in areas that receive high sediment loads (Maragos, 1993; Rodgers, 1990; Richmond, 1993). These findings correspond with those of Green (1996), who also reported that where water quality is poor, coral communities in American Samoa have not recovered as rapidly, probably due to high sediment loads. Many of the villages have identified sedimentation as one of the potential causes of degradation to their coral reefs (Musburger, 2004). Therefore, an interesting idea for future scientific work would be to collect long-term information about the levels of sedimentation on the reefs. This monitoring program would help to determine how much sediment is falling on the reefs and what affect it is having.

Macroalgae cover was also found to be highest at Aua and Aua control. In a global assessment of human effects on coral reefs Hodgson (1999) chose macroalgae as an indicator of nutrient enrichment associated with sewage pollution. Industrial developments along the foreshore of Pago Pago Harbour include the sewerage plant, two tuna canneries and a fuel storage depot which may contribute nutrients to the environment enhancing algal growth. Surface water run-off from the steep mountain slopes is another potential nutrient enhancement source. High macroalgae cover at these sites is a result of the higher nutrient conditions in the harbour than elsewhere around Tutuila Island. Macroalgae may be responsible for the low coral cover at

87 4 DISCUSSION ______these sites as they can outcompete corals for space and thus inhibit recruitment of coral planulae, further endangering the ability of coral to survive (Birkeland, 1977; Lapointe et al., 1997; Belliveau and Paul, 2002).

Fish abundance at Aua and Aua control was also the lowest. Reef fish can have a critical role in structuring the benthic community. Herbivory is important as it creates frequent disturbances on reefs, reducing spatial competition of coral recruits with benthic algae, and promoting coralline algae growth which can favour coral settlement (Belliveau and Paul, 2002). A reduction of herbivorous fishes and subsequent reduced grazing can promote growth of macroalgae which may be another factor responsible for the high cover of algae at these sites.

Coralline algae are potentially a key factor in structuring species composition as they can facilitate settlement and metamorphosis of some coral species (Morse et al., 1996) and therefore promote coral recruitment and recovery (Belliveau and Paul, 2002). Coralline algae was the dominant substrate recorded at Aua which suggests that corals in the harbour are indeed beginning to recover from past disturbances. Such findings correspond with those of Birkeland (2004) who reported that despite their poor condition, coral communities at Pago Pago Harbour seem to be showing signs of recovery for the first time in decades.

The entrapment of sediment by macro algae and turf algae can also inhibit coral recruitment (Birkeland, 1977) and survivorship of CA (Fabricius and De’ath, 2001). High sediment accumulation at the Aua sites can be reduced by greater water flow rates which were observed at Aua, enhancing survival of coralline algae at this site. Coralline algae also tend to prefer exposed habitats or places where the water movement is fast (Belliveau and Paul, 2002). Currents at Aua control were less and visibility was also much worse (<1.5m) than at Aua, which may explain the very low coralline algae cover and also the dominance of macroalgae at this site.

Coralline algae cover was also abundant at Vatia and Masefau on the north east side of Tutuila. The reefs of both these sites are confined to the semi-protected waters of large embayments and are sheltered from the open ocean. However, habitat destruction by hurricanes has been more severe on the north side of the island. Thus,

88 4 DISCUSSION ______the abundance of coralline algae at these sites may play an important role in facilitating recovery of corals from past natural disturbances (particularly hurricanes) by consolidating loose carbonate structures and providing a stabilized reef.

Macroalgae cover was also very low at Vatia and Masefau (<2%). This could be a result of less nutrients from pollution at these sites due to the lower human population on the north coast compared to the highly populated south coast. The high abundance of fish recorded at Masefau, and therefore increased herbivory, may further reduce macro algae cover. In terms of benthic substrate cover, both Masefau and Vatia were not significantly different from one another. Similarities in benthic community structure of these sites are most likely a result of similar geomorphology noted.

Coral cover at Auto & Amaua was considered to be ‘poor’. The dominant substrate at this site was rubble which was also abundant at Aua control and moderately high at Aua. Auto & Amaua is a very exposed site as it is subject to the prevailing southeast trade winds. Rubble on the reef flats at this site (Plate 4.1) was most likely deposited by large waves during recent storms (Green, 1996). Overall, the benthic community structure of Auto & Amaua was similar to Aua and Aua control which is responsible for the high similarity of these three sites.

Plate 4.1 Rubble at Auto & Amaua, American Samoa (Photo: D. Fenner).

89 4 DISCUSSION ______

In this study, sites with high turbidity were characterised by low coral cover and high macroalgae cover (Aua and Aua control). In contrast sites with low turbidity were characterised by high coral cover and low macroalgae cover (Fagatele bay and Faga’itua). No significant correlation was found between turbidity and macroalgae, but a significant correlation was found between turbidity and live coral cover.

Sand was found to be highest at the three lagoonal sites and also at Aua control. Based on benthic substrate cover of the 11 sites surveyed, Nu’uuli lagoon, Alofau lagoon and Ofu lagoon were grouped together by hierarchical clustering suggesting their similarity. Sand, being an abundant benthic substrate of lagoonal habitats was found to be the major contributor to the similarity of these 3 sites. Live coral cover was also similar at these sites. It is not surprising then that Nu’uuli and Alofau lagoon were not significantly different in terms of benthic substrate cover.

Sponge was recorded at 3 sites; cover was highest at Aua and <2% at Nu’uuli and Aua control. On some reefs, a high abundance of encrusting sponges may indicate high nutrient levels or other problems (Wilkinson, 1987). Furthermore, Brand and Aicher (1997), using chlorophyll a as an indicator of elevated nutrient inputs in coastal waters, demonstrated that the highest chlorophyll a concentrations were found at Pago Pago harbour and also Nu’uuli lagoon. High human population, and therefore a high level of impact from human activities, in both Aua and Nu’uuli would contribute to such elevated nutrient concentrations.

4.2 CORAL SPECIES ABUNDANCE Another objective of this study was to describe the coral reef benthic communities in terms of the presence and relative abundance of coral species. A total of 42 species of coral were recorded from all sites surveyed in American Samoa during June to August 2004, representing about 20% of the total number expected to occur in Samoa. This is probably due to the type of habitats surveyed. Reef flats have been characterised by low species richness while species richness in lagoons was low to moderate (Green, 1996).

90 4 DISCUSSION ______

Among all sites the relative abundance of coral cover was dominated by Porites spp., Acropora spp., Pavona spp., and Pocillopora spp., with many other corals having a smaller presence. Coral abundance on the reef flat is probably related to wave energy as most of the coral communities at each site tended to be dominated by branching corals and massive corals, reflecting the extreme exposure to waves. For example, more of the physically robust coral species such as massive Porites, Porites cylindrica, Porites rus, Pocillopora damicornis, and Pavona frondifera comprised the majority of coral cover throughout the study area and each one was a dominating species at one or more sites.

A more fragile coral species, Acropora microphthalma, was found to be an abundant species at Alofau, and was the dominant species at Faga’itua along the southeast coast. These sites are large bays that are protected from rough oceanic conditions. This protected nature allows for extensive coral community development that was observed at Alofau and Faga’itua. Coral cover at both these sites was among the highest recorded, representing more than 65% of the coral community. Such protected reefs can therefore support fragile coral species such as A. microphthalma.

The vast majority of sites were significantly different in terms of coral species abundance. However, sites that were not significantly different from one another comprised similar abundances of the same hard coral species. Faga’itua and Alofau reef flat were found to be significantly similar to one another based upon abundance of hard coral species. The species which contributed the most to the similarities of these sites was A. microphthalma. Cover was 29% at Alofau reef flat and 35.5% at Faga’itua. These sites were also found to be significantly similar in benthic substrate cover. Therefore, as previously discussed in section 4.1, similar coral community development at these sites may be a result of similar reef geomorphology, similar watershed areas and similar low population densities.

Similar species richness, and similar high dominance patterns of coral species observed, indicated that the majority of sites were dominated by one or two species. For example P. rus dominated at Masefau, massive Porites dominated at Ofu and Auto & Amaua, table Acropora dominated at Aua, and P. damicornis dominated at Aua control. P. cylindrica, the most abundant coral occurring at all 11 sites was the

91 4 DISCUSSION ______dominant species at Vatia, Alofau lagoon and Alofau reef flat. Alofau lagoon and Alofau reef flat were also found to be significantly similar to one another with P. cylindrica contributed the most to the similarities of these sites.

P. frondifera dominated at Fagatele Bay and Nu’uuli. Despite the obvious differences in the benthic community (i.e. no rubble and sand at Fagatele Bay compared to Nu’uuli), the Fagatele Bay and Nu’uuli sites were also significantly similar based on coral species abundance. In general, these sites were distinguished by having a high cover of P. frondifera compared to all other sites, and also a high cover of P. cylindrica. Both these species contributed the most to the similarities of the sites.

Vatia and Masefau were found to be significantly similar to one another based upon abundance of coral species. The species which contributed the most to the similarities of these sites was P. rus. Cover of this species was 28% at Masefau and 25% at Vatia. P. rus is a faster-growing opportunistic species which thrives in coral communities subject to continual disturbance (Quinn and Kojis, 1999). High cover at Vatia and Masefau is not surprising since these sites are located on the north coast where habitat destruction by hurricanes has been more severe. Similar coral community development at Masefau and Vatia may also be due to similar reef geomorphology, watershed area and population density as discussed in section 4.1.

According to Margalef’s and Shannon’s indices the most diverse site was clearly Ofu, followed by Fagatele Bay and Masefau. Ofu, the only site not located on Tutuila Island, was found to be least similar site according to hierarchical clustering. The coral communities at Ofu Lagoon were distinct and differed greatly from all other sites; Ten species of coral were recorded at this site only, including the very rare blue coral Heliopora and also Porites sp 2., a species unique to Samoa. No particular species was especially dominant; Massive Porites was the most abundant accounting for 22.5% of the coral community. COTS predation may have played an important role in determining the relative abundance of corals in Ofu Lagoon, since the coral communities are dominated by less preferred prey species such as massive Porites, while more preferred species such as Acropora are uncommon (Zann, 1992). High species richness and diversity at Ofu may also be a consequence of more substrate

92 4 DISCUSSION ______space being available for settlement following COTS impacts. Alternatively, it may be due to naturally higher recruitment rates compared to Tutuila (Fisk and Birkeland, 2002).

Plate 4.2 Massive Porites at Ofu lagoon, American Samoa.

Maragos et al. (1994) reported that coral reefs were generally healthier with higher live coral cover and species richness on Ofu compared to those off Tutuila. Green (2002) also reported that coral cover in the lagoon appeared to have increased over the last few years with a corresponding decrease in cover of algae and non-living substratum. It seems likely that the remoteness of the less populated Ofu, combined with the light fishing pressure and low human impacts may be responsible for why these reefs remain healthy.

The natural setting of Fagatele Bay, absent from anthropogenic disturbance must be a contributing factor to the high species richness and diversity of coral communities recorded at this site. High diversity and species richness was also shown at Masefau, a non MPA site. It was observed, that the coral community consisted of many diverse, small colonies suggesting coral recruitment was occurring. Recolonization of coral recruits maybe from past natural disturbances, particularly hurricanes. The

93 4 DISCUSSION ______high abundance of coralline algae, a major reef-builder at this site lends further support to these findings. Fisk and Birkeland (2002) also reported an increase in species richness and an increase in the density of colonies though recruitment at Masefau.

Aua control had the lowest dominance since this site also had the lowest diversity, species richness and live coral cover. The low coral cover and diversity of this site is most probably due to the poor condition of the waters in Pago Pago Harbour discussed previously. Coral cover was lowest at Aua and Aua control but consisted of many small coral colonies, suggesting that coral recolonization was occurring and the coral community beginning to recover. Furthermore, species richness and diversity at Aua was considerably more than at Aua control. The most dominant species recorded at Aua were small robust tables of Acropora species. Green (2002) and Birkeland et al. (2004) also observed a substantial increase in Acropora recruits (particularly Acropora hyacinthus) at the outer reef flat at Aua. The fact that Acropora species were abundant was considered a good indicator of improved water quality, since these species are highly sensitive to sedimentation.

Aua and Auto & Amaua, Faga’itua and Auto & Amaua, and Faga’itua and Alofau lagoon were all found to be significantly similar to one another based upon coral species abundance. The species which contributed the most to the similarities of these sites was P. damicornis. Species of Pocillopora are perceived as “weedy” species that are rapid recruiters and rapid growers (Birkeland et al., 2004).

Reef flats at Auto & Amaua were dominated by massive Porites microatolls. Coral cover at this site was ‘poor’, most likely as a result of sedimentation. Sites that receive high sediment loads such as Aua and Aua control tended to have lower coral cover than other sites. These sites comprised distinctive coral communities dominated by species that can tolerate high sediment loads such as large massive reef-building corals (e.g. Porites). Reef-building corals can rid themselves of sediment by such processes as mucus secretion and ciliary action. During this survey, mucus was observed on massive Porites colonies at Auto & Amaua, which may be an indicator of stress from sedimentation. However, visibility at this site was

94 4 DISCUSSION ______good (13m) due to a strong current flow which flushes the water from the bay daily, so sediment loads tend not to accumulate.

4.3 REEF SUBSTRATE AND CORAL SPECIES ABUNDANCE The only sites not found to be significantly different from each other, based upon reef substrate and coral species abundance, were Masefau and Vatia, and Alofau reef flat and Faga’itua. As discussed in previous sections, similarities in reef substrate and coral species abundance between sites, are most likely a result of similar reef geomorphology, watershed areas, and population densities of the sites.

4.4 KEY MACROINVERTEBRATES Another objective of this study was to estimate abundances of key macroinvertebrates that are considered to be of local and/or national importance such as giant clams, COTS and sea urchins.

Giant clams were very rare at the Tutuila sites compared to occasional sightings at Ofu. These findings correspond with those of Green (2002) who also reported that giant clams were more abundant at Ofu than the Tutuila survey sites. Given that giant clams are highly prized as food by Samoans, it seems likely that overfishing and a subsequent lack of recruitment has contributed to the low numbers of clams on Tutuila (Green, 2002). This is also supported by the results of local fisheries statistics, which showed a decline in the harvest of giant clams over the last two decades (Ponwith, 1991).

COTS were rare throughout American Samoa and only one starfish was recorded at Masefau. A study by Green (2002) also recorded no starfish on crests, reef flats or deeper reef slopes on Tutuila Island. Night surveys would probably locate more COTS because of their predominantly nocturnal behaviour (Green and Hunter, 1998).

Sea urchins were the predominant macroinvertebrates occurring at the majority of sites and were most abundant at Auto & Amaua. Sea urchins are a common food

95 4 DISCUSSION ______source collected in Samoa (Sauafea, 2000). No fishing at Auto & Amaua due to CB management may account for their large numbers compared to all other sites. However, this does not explain why large numbers were not found at the other CB MPA sites.

McClanahan (1997) reported that intense grazing by sea urchins and associated sediment production also results in a reduction of corals and coralline algae. The high numbers of sea urchins recorded at Auto & Amaua may be a reason for the low coral and coralline algae cover at this site.

4.5 RELATIONSHIP BETWEEN HARD CORAL AND FISH A further objective of this study was to investigate the relationship between coral and fish. A significant positive correlation was found between fish abundance and hard coral cover which suggests that percentage live coral cover had a strong positive influence on abundance of fish. The increase in the density of coral colonies though recruitment at Masefau may therefore be responsible for the high abundance of fish recorded. Green (2002) also concluded that some fish species have increased in abundance, in response to the recovery of the coral communities on Tutuila.

The highest fish abundance was recorded at Faga’itua and Alofau reef flat. Both these sites also had high coral cover. High fish abundance may be attributable to the high cover of branching corals at these sites, particularly branching acroporid corals such as A. microphthalma. In a study by Freidlander et al. (2003) higher fish abundance appeared to be influenced by higher rugosity, occurrence of embayments and high cover of branching corals. Embayments, typically had fewer numbers of species as did Faga’itua which was the least diverse site in terms of fish species. Faga’itua is also located between 2 CB MPAs, Alofau reef flat and Auto & Amaua, where fish abundance was also high. Higher fish abundances within these MPAs have the potential to enhance fish abundance in adjacent areas, such as Faga’itua. This may occur through increased larval export from MPAs owing to increased spawning stock biomass within MPAs (Bohnsack, 1996) and/or spillover of exploitable fish into neighbouring non-reserve areas (Roberts and Polunin, 1991; DeMartini, 1993; McClanahan and Mangi, 2000).

96 4 DISCUSSION ______

Plate 4.3 High abundance of fish and high coral cover (particularly Acropora microphthalma) at Faga’itua, American Samoa (Photo: D. Fenner).

The most diverse sites in terms of fish species richness and species diversity were Masefau, Auto & Amaua and Ofu. Coral species richness and diversity was also high at Masefau and Ofu, although no significant correlation was found between fish species richness and hard coral species richness, and between fish species diversity and hard coral species diversity. These findings are part supported by Roberts and Ormond (1987) who concluded that the amount of hard coral cover had little influence on the fish species richness and abundance, whereas this study did find a positive relationship between hard coral cover and fish abundance.

Despite high fish abundance and species richness at Auto & Amaua, this site was characterised by low coral cover and low species richness. Auto & Amaua has been closed for fishing for more than two years. High fish abundance at this site may be a result of migration from adjacent areas where fishing is permitted. This suggests that protection from fishing may be as, if not more, important than habitat quality in sustaining and enhancing fish assemblages (Freidlander et al., 2003).

In contrast, coral cover at Vatia was twice as high compared to Auto & Amaua, however fish abundance at this site was much lower. Vatia had been closed for fishing for 2 years, but one day earlier this year the villagers decided to reopen the

97 4 DISCUSSION ______reserve. They reported a great day of fishing which may be the reason for lower fish abundance at this site. Fortunately, the reserve was closed the next day. Another possible reason for low fish abundance, maybe, that when the surveys were conducted at Vatia, there were many people in the water. This may have caused greater disturbance to the fish community resulting in the lower numbers recorded at this site.

The lowest fish abundance and species richness was recorded at Aua control and Aua located in Pago Pago harbour. Both these sites also had low species diversity. Similarly, these sites had very low coral cover and species richness. As discussed previously, coral reefs in Pago Pago Harbour have been subject to anthropogenic stresses in the past, resulting in poor water quality and very low coral cover. The poor quality of these habitats may account for the low fish abundance and species richness. Green (2002) also reported that the fish communities in the Harbour reflected the poor condition of the coral communities. The species that are abundant tend to be those that are not closely associated with healthy coral communities. While those species that do rely on healthy coral communities tend to be rare or less abundant in the Harbour area (Green, 2002). Results by Jones et al. (2004) also suggest that reefs without corals, will no longer support diverse fish faunas but rather will be numerically dominated by a small subset of species preferring algal or rubble substrata. Many juvenile fish were observed at Aua which suggests that like the coral community, fish populations are also starting to recover in the Harbour.

Despite very high coral cover at Fagatele Bay, fish abundance was not so high. In contrast to the coral communities, the fish communities appear to be taking longer to recover from habitat destruction caused by the major disturbances in the past two decades. There is also evidence that illegal fishing is occurring in Fagatele Bay, which may be impacting the fish communities and impeding their recovery (Birkeland et al., 2004).

Fish abundance in the three lagoonal sites was also fairly low. These sites were also characterised by ‘fair’ coral cover. In particular, the fish communities on Ofu have been affected by the impacts of chronic COTS predation on the coral communities on those islands. Species that are dependant on corals that are the preferred food of the

98 4 DISCUSSION ______starfish (e.g. branching or plate coral) are uncommon. Chabanet et al. (1997) concluded that an environment with low coral diversity can limit fish abundance. Both Nu’uuli and Alofau lagoon had low coral diversity which may be responsible for the low fish abundance recorded. Also the relationship between the abundance of fishes and the coverage by living coral cover may be stronger in shallow reef flat habitats because fish there remain in closer proximity to the substratum (Chabanet et al., 1997).

4.6 EFFECTS OF PROTECTION ON CORAL REEF COMMUNITIES A main objective of this study was to investigate the effects of protection on coral reef community structure (substrate cover and coral species abundance) between CB MPA sites, non MPA sites and federal MPA sites. Overall reef substrate varied between CB MPA sites, non MPA sites and federal MPA sites but the differences were not found to be significant.

Rubble and dead coral with algae was highest at CB MPAs and non MPAs. Cover of these substrates was very similar at both groups of sites, and lowest at federal MPA sites. This may be because CB MPAs and non MPAs have suffered more damage and coral mortality than the federal MPAs, which may be due to anthropogenic disturbance at CB MPA and non MPA sites compared to federal MPA sites. However, a more likely explanation is that federal MPAs were protected initially as they were located in better-quality areas, compared to non MPAs and CB MPAs, which probably contained rubble to begin with.

Live coral cover was 15% greater at federal MPAs, followed by similar cover at CB MPAs and non MPAs. Again, this is not surprising, considering both federal MPAs surveyed meet the definition of a MPA and were protected to begin with because of the pristine nature of their coral reef resources. Human impact at these sites is low which contributes to the health of the reefs. Both federal MPA sites have also been protected from fishing for 10-20 years. However, regular surveillance and enforcement is generally lacking and poaching of marine organisms is an on-going problem (Craig, 2001). A significant difference in coral species abundance was

99 4 DISCUSSION ______found between federal MPA sites and CB MPA sites, and between federal MPA sites and non MPA sites.

The percentage difference of coral cover between CB MPAs and non MPAs was only 0.3%. This might have been expected since the CB MPAs have not been closed to fishing for long enough to observe a difference in the coral reef communities. Also, as discussed previously, federal MPA sites were protected because of the pristine nature of their coral reef resources. In contrast, CB MPA sites are protected regardless of whether the reef was in a good or poor condition which may account for such similarities in coral cover between CB MPA sites and non MPA sites. Several other studies (Polunin and Roberts, 1993; McClanahan et al., 1999) also failed to find any difference in live coral cover between MPA and non MPA areas. This suggests that a priori siting of MPAs in good areas may be relatively uncommon (Cote et al., 2001).

CB MPA sites are really a fishery reserve to improve reduced fish stocks which have been exploited by overfishing in the past, which is the main reason to why these sites are protected. However, Craig et al. (1999) identified overfishing as a major problem hindering recovery of local reefs in American Samoa as it threatens reef succession following disturbance. Therefore the establishment of CB MPAs to protect vulnerable herbivorous fish populations is critical to maintaining reef health and biodiversity (Belliveau and Paul, 2002).

For MPAs to act as fisheries management tools, it is also important that as much area as possible is designated as “no-take” and that fishing restrictions are effectively enforced (Green, 2002). A study by McClanahan and Arthur (2001) suggests that temporary closures of areas may not be sufficient to protect and restore the species richness or rare fish unless closures are longer than a decade. Therefore, opening and closing MPA areas to fishing (e.g. Vatia) will not be as effective in the long-term. Management of marine diversity requires the establishment and maintenance of permanent protected areas along with other fisheries management options (Allison et al., 1998).

100 4 DISCUSSION ______

The effects of MPA performance will vary on a case-by-case basis, depending on the environmental, biological and physical factors involved. As discussed in section 4.1 and 4.2, these factors play a fundamental role in determining reef structure at any particular site. Another major concern is that MPAs could be too small, too few, cover too little total area or be located in the wrong areas to be effective (Bohnsack, 1998). Therefore, regardless of whether they were protected or not, sites in this study with high coral cover were located within less populated areas exposed to low human impacts. For example, Faga’itua, a non MPA site had very high coral cover and also the highest abundance of fish. Alofau reef flat, a CB MPA site, located next to Faga’itua was also shown to have very similar coral cover and fish abundances. Masefau, a non MPA site, and Vatia, a CB MPA site, were also found to be significantly similar to one another. Aua, a CB MPA site, and Aua control, a non MPA site were both found to have low coral cover due to their location in Pago Pago harbour. These findings suggest that location rather than protection is responsible for differences in coral cover at the survey sites.

These findings are supported by McClanahan (1994) who found that protected areas had higher structural complexity and coral cover than non-protected areas. However, it was not clear whether the apparent increase in habitat quality resulted from protection, or whether protected areas were initially located in better-quality sites (Russ, 1985). Furthermore, a study carried out by Walmsley and White (2003) demonstrated that implementation of a marine sanctuary will not necessarily result in a coral reef with high percentage coral cover if the conditions at the site are not naturally favourable for it. Natural variability in recruitment and settlement, and natural climatic events, will all influence the effects of protection.

There is also growing recognition that MPAs cannot protect reefs from large-scale important threats, such as contamination from pollutants or global warming (Allison et al., 1998; Jones et al., 2004), because they do not have functional boundaries (Boersma and Parrish 1999). An exotic species that is introduced near a MPA, either accidentally or intentionally, will spread into a MPA regardless of MPA boundaries or regulations. Therefore, without adequate protection of species and ecosystems outside MPAs, effectiveness of MPAs will be severely compromised (Allison et al., 1998).

101 4 DISCUSSION ______

4.6.1 Success of CB MPAs Several of the villages in the CBFMP expressed their strong desire to develop stronger awareness of their efforts and to document the success of their management efforts (Musburger, 2004). Simple management measures (signs-Plate 4.4, boundary markers, mooring buoys) and enforcement of regulations can significantly improve coral cover and fish abundance in MPAs (Walmsley and White, 2003).

Community support, management measures and enforcement of regulations all contributed towards sanctuary success in the Philippines (Walmsley and White, 2003). Management and enforcement were most important, although community support for the sanctuary is likely to contribute to effective enforcement.

Plate 4.4 A sign designating the fishery reserve of Vatia. Such signs are an important part of increasing awareness about the programme and the fishing restrictions in the villages.

A study by Pollnac et al. (2001) was conducted at 45 CB MPAs in the Philippines. The most important predictors of success were:

Population size of the community; A perceived crisis in terms of reduced fish populations;

102 4 DISCUSSION ______

Successful alternative income projects; High levels of participation in community decision making; Continuing advice from the implementing organisation and; Inputs from local government.

The success of CB MPAs in American Samoa will depend on these social factors as well as the environmental, biological and physical factors involved. However, community-based fisheries management should continue to be pursued. It may not initially increase habitat quality, but it is important in sustaining and enhancing fish populations. It also helps indirectly by raising awareness of conservation issues amongst the general public and encouraging sustainable coastal resource use.

4.7 SUMMARY

Coral cover was highest at sites that were the most protected from physical disturbances and probably have a higher potential for maximum coral cover because of low wave disturbance. In addition coral communities tended to be dominated by branching corals and massive corals, reflecting the extreme exposure to waves on the reef flat.

Location rather than protection is responsible for differences in coral cover at the sites. Sites characterised by ‘fair’ or ‘good’ coral cover were located in less populated areas exposed to low human impacts. In contrast sites located in areas of high population and industry and therefore heavily impacted by pollution were characterised by ‘poor’ coral cover.

A significant positive relationship was found between coral cover and turbidity. Sites with high turbidity were characterised by low coral cover and sites with low turbidity were characterised by high coral cover.

Macroalgae was abundant at sites with low coral cover as it can outcompete corals for space and inhibit recruitment.

103 4 DISCUSSION ______

Massive Porites, Porites cylindrica, Porites rus, Pocillopora damicornis, Acropora microphthalma and Pavona frondifera comprised the majority of coral cover throughout the study area and each one dominating at one or more sites.

Significant similarities between sites in reef substrate and coral community development are likely to be the result of similar reef geomorphology, watershed areas, and population densities.

Coral cover was found to have a strong positive influence on abundance of fish. The majority of sites with good coral cover had a high abundance of fish compared to sites with poor coral cover which were characterised by low fish abundance. Auto & Amaua, a CB MPA site was characterised by low coral cover but high fish abundance. This suggests that protection from fishing may be as, if not more, important than habitat quality in sustaining and enhancing fish assemblages.

Overfishing as a major problem hindering recovery of local reefs in American Samoa as it threatens reef succession following disturbance. The establishment of CB MPAs to protect vulnerable herbivorous fish populations is critical to maintaining reef health and biodiversity.

No significant difference was found between CB MPA sites and non MPA sites in benthic substrate cover and hard coral species abundance. The success of MPA performance will vary on a case-by-case basis, depending on the social, environmental, biological and physical factors involved. However, community-based fisheries management should continue to be pursued. It may not initially increase habitat quality, but it is important in sustaining and enhancing fish populations. It also helps indirectly by raising awareness of conservation issues amongst the general public and encouraging sustainable coastal resource use.

104 5 CONCLUSION AND RECOMMENDATIONS ______

5.1 MANAGEMENT RECOMMENDATIONS 1. No-take MPAs are needed where there is overfishing, particularly on the main island of Tutuila. For this reason, Faga’itua and Masefau should be taken into consideration as future candidates to participate in the CBFM programme, since they represent reef areas that are still flourishing and healthy. Faga’itua because of its high live coral cover, and Masefau because of its high species richness and diversity.

2. Monitoring of the CB MPAs should be repeated on a regular basis to see if there are any improvements in the coral reef communities (coral and fish). Monitoring surveys will help to understand the natural variability and long term trends in the coral reefs being protected, and detect any threats to ecosystem health as they arise (i.e. large scale disturbances and/or human activities). Musburger (2004) recommended that surveys be conducted every 3 months.

3. Even if the villages of Faga’itua and Masefau decide not to join the CBFMP, they should, along with the CB MPA sites, be continually monitored. Alofau and Faga’itua, and Masefau with Vatia were found to be statistically similar to each other and will provide useful comparisons. Such comparisons of non MPA sites and CB MPA sites will monitor the success of CB MPAs in American Samoa to determine if their protected status is making a difference or not.

4. Since overfishing is one of the greatest threats to the long term health of the reefs in the Territory, it is important that coral reef fisheries are monitored effectively on the main islands, particularly on Tutuila. Also recommended by Green (2002).

5. CB village MPAs should be closed to fishing for minimum of 5 years to allow fish stocks to replenish.

105 5 CONCLUSION AND RECOMMENDATIONS ______

6. DMWR representatives need to meet with the village committee of Aua to discuss the outcome of their fisheries management program. Villagers have already expressed interest in the programme.

7. Educational programmes are recommended to promote public awareness on the importance of the coral reefs in Samoa, and the need to protect them.

5.2 LIMITATIONS OF THE PRESENT STUDY 1. A major limitation of this study was the not knowing what the sites were like before MPA status. Therefore it is hard to assume whether protection was responsible for the high coral cover found at some sites or whether these sites were located in better quality areas to begin with.

2. Not all villages MPAs participating in the CBFMP were surveyed due to time limitations and tide constraints.

3. It would have been of interest to survey the reef slopes but due to SCUBA restrictions it was only possible to survey reef flats and lagoons.

4. In the present study only qualitative observations were noted on size of coral colonies. Quantitatively measuring the presence of newly settled corals would have been beneficial as it measures coral recruitment and potential for recovery. This should be incorporated into the survey design for future scientific work.

5.3 CONCLUSION The main aim of this study was to investigate the effects of protection on coral reef communities at community-based Marine Protected Area sites, non MPA sites and federal MPA sites in American Samoa. This aim was accomplished by providing detailed descriptions of the coral reef communities in terms of reef substrate and coral species abundance at each survey site. GIS was used to examine the biological, environmental, physical and social factors of each of the sites, all of which will

106 5 CONCLUSION AND RECOMMENDATIONS ______influence the effects of protection. Using both the biological and GIS data, the effects of protection on coral reef communities in this study was critically evaluated, however, more data was required in order to assess what condition the coral reef community of the sites were in, before protection was set up. This would help to determine whether sites with high coral cover were a result of protection or whether such sites were located in better-quality areas to begin with.

This study provided the basis for establishing a long term monitoring program for the community-based fisheries management village marine protected areas of American Samoa. Community-based marine protected areas should be continually monitored to determine the effects of protection on coral reef communities in American Samoa.

107 APPENDICES ______

APPENDIX 1

Benthic habitat map of Tutuila, American Samoa (map produced by Francesca Riolo, DMWR, 2004).

APPENDIX 2

Fishing effort for Tutuila and Ofu, American Samoa (Source: Spurgeon et al., 2004).

APPENDIX 2

Geographic coordinates for transects 1-5 at each site surveyed in American Samoa. Sites were VAT = Vatia, A&A = Auto & Amaua, ARF = Alofau reef flat, AL = Alofau lagoon, FA = Faga’itua, AC = Aua control, MAS = Masefau, NU = Nu’uuli, FB = Fagatele Bay.

Site T1 T2 T3 T4 T5 Longitude Latitude Longitude Latitude Longitude Latitude Longitude Latitude Longitude Latitude VAT 170.672955 14.246551 170.673565 14.247617 170.673926 14.248373 170.671525 14.249724 170.672185 14.250117 A&A 170.627734 14.279034 170.627 14.279105 170.62734 14.278675 170.620133 14.27265 170.620628 14.272906 ARF 170.604602 14.275739 170.604877 14.275377 170.605557 14.275214 170.605684 14.275929 170.605094 14.276053 AL 170.60523 14.275149 170.60591 14.275443 170.606052 14.276356 170.605076 14.276642 170.60517 14.277427 AUA 170.6708 14.270699 170.666382 14.27258 170.67047 14.271077 170.670721 14.270394 170.666017 14.272304 FA 170.614193 14.268489 170.613826 14.268971 170.614424 14.269026 170.613756 14.269398 170.614433 14.269673 AC 170.664814 14.273497 170.66519 14.272578 170.66476 14.272967 170.665104 14.273689 170.665364 14.273218 MAS 170.628935 14.254414 170.629464 14.254646 170.629116 14.2561 170.628198 14.2563 170.627534 14.2561 NU 170.69801 14.312938 170.698097 14.312704 170.697728 14.313073 170.697309 14.312974 170.696992 14.312974 FB 170.760317 14.364125 170.760004 14.364078 170.760252 14.363478 170.760766 14.363558 170.761508 14.363105 OFU 169.654924 14.179171 169.654292 14.179119 169.654228 14.178381 169.653666 14.177741 169.653222 14.177863 APPENDICES ______

APPENDIX 3

Benthic survey form for recording data at each survey site in American Samoa during June to August 2004.

Location Temperature Date Reef habitat Salinity Weather Transect no. Secchi Time and Height of HT Min depth Max depth

0.5 17.5 1 18 1.5 18.5 2 19 2.5 19.5 3 20 3.5 20.5 4 21 4.5 21 5 5 5.5 22 6 22.5 6. 23 7 23.5 7.5 24 8 24.5 8.5 25 9 Notes 9.5 10 10.5 11 11.5 12 12.5 13 13.5 14 14.5 15 15.5 16 16.5 17

Key Macroinvertebrates Counts

Sea Urchins Giant Clams COTS APPENDICES ______

APPENDIX 6

Raw data collected during the point-intercept transects at 11 coral reef survey sites in American Samoa during June to August 2004. T = Transect.

1. VATIA

Date: 28/06/2004 Turbidity (m): 11 Salinity: 1026 Temperature (oC): 29 Min depth: 1.1 Max depth: 3.7 Time and Height of HT: 16:59 (2.4)

Substrate T1 T2 T3 T4 T5 Mean % Mean Sand 000000 0.0 Rubble 410142 4.0 Rock 001000.2 0.4 Coralline algae 15 10 14 19 7 13 26.1 Macroalgae 001020.4 0.8 Turf algae 425423.4 6.8 Dead coral 000100.2 0.4 Dead coral with algae 486103.8 7.6 Soft Coral 000030.6 1.2 Live Coral 23 29 23 24 32 26.2 52.6 Coral species Porites cylindrica 0 11 0 14 20 9 34.4 Porites rus 0162786.6 25.2 Pocillopora verrucosa 714002.4 9.2 Pocillopora meandrina 407002.2 8.4 Pocillopora eydouxi 102000.6 2.3 Acropora microphthalma 100000.2 0.8 Acropora gemmifera 102100.8 3.1 Acropora cytherea 000040.8 3.1 Acropora sp. 001000.2 0.8 Pavona frondifera 410201.4 5.3 Montipora sp. 50500 2 7.6

80 8

60 6

40 4

20 No. Species 2

Coral cover (%) 0 0 12345 12345 Transect Tra n s e c t APPENDICES ______

APPENDIX 6 continued

2. AUTO & AMAUA

Date: 2/07/2004 Turbidity (m): 13.2 Salinity:jj 1025 Temperature (oC): 29 Min depth: 0.9 Max depth: 1.1 Time and Height of HT: 08:00 (3.1)

Substrate T1 T2 T3 T4 T5 Mean % Mean Sand 1 0 4 1 2 1.6 3.2 Rubble 23 13 9 19 14 15.6 31.2 Rock 0 0 1 0 3 0.8 1.6 Coralline algae 5 1 3 2 8 3.8 7.6 Macroalgae 4 16 14 1 5 8 16 Turf algae 0 0 2 7 5 2.8 5.6 Dead coral 0 0 1 0 0 0.2 0.4 Dead coral with algae 3 7 3 5 3 4.2 8.4 Soft Coral 0 0 0 0 0 0 0 Other 0 0 0 0 0 0 0 Sponge 0 0 0 0 0 0 0 Live Coral 14 13 13 15 10 13 26 Coral species Massive Porites 7 1 1 13 0 4.4 33.8 Porites cylindrica 50 1001.2 9.2 Porites rus 00 1000.2 1.5 Porites sp. 2 00 000 0 0.0 Pocillopora damicornis 23 203 2 15.4 Acropora microphthalma 04 3001.4 10.8 Acropora aspera 00 0061.2 9.2 Pavona frondifera 02 201 1 7.7 Pavona decussata 03 3101.4 10.8 Leptastrea purpurea 00 0100.2 1.5

35 30 8 25 20 6 15 4 10 Coral cover (%) cover Coral No. Species 2 5 0 0 12345 12345 Tr a n s e ct Tra n s e c t APPENDICES ______

APPENDIX 6 continued

3. ALOFAU REEF FLAT

Date: 7/07/2004 Turbidity (m): 10.2 Salinity: 1028 Temperature (oC): 28 Min depth: 0.7 Max depth: 1.8 Time and Height of HT: 12:43 (2.1)

Substrate T1 T2 T3 T4 T5 Mean % Mean Sand 0 0 0 0 0 0 0 Rubble 2 0 1 6 0 1.8 3.6 Coralline algae 6 3 3 1 0 2.6 5.2 Macro algae 0 4 2 2 0 1.6 3.2 Turf algae 4 0 2 4 0 2 4 Dead coral 0 1 0 0 0 0.2 0.4 Dead coral with algae 2 15 6 3 10 7.2 14.4 Soft Coral 0 0 0 0 0 0 0 Other 0 2 0 0 2 0.8 1.6 Sponge 0 0 0 0 0 0 0 Live Coral 36 25 36 34 38 33.8 67.6 Coral species Massive Porites 10220 1 3.0 Porites cylindrica 3 2 16 8 33 12.4 36.7 Porites rus 10010 0.4 1.2 Porites sp. 2 00000 0 0.0 Pocillopora damicornis 57021 3 8.9 Acropora microphthalma 11 11 13 11 4 10 29.6 Acropora hyacinthus 02000 0.4 1.2 Acropora austera 21000 0.6 1.8 Acropora pulchra 00020 0.4 1.2 Pavona frondifera 92080 3.8 11.2 Pavona decussata 00300 0.6 1.8 Montipora sp. 30200 1 3.0 Leptastrea purpurea 10000 0.2 0.6

80 10 60 8 6 40 4 20 Coral cover (%) No. Species 2

0 0 12345 12345 Tr a n s e ct Tr a n s e ct APPENDICES ______

APPENDIX 6 continued

4. ALOFAU LAGOON

Date: 26/07/2004 Turbidity (m): 8 Salinity: 1028 Temperature (oC): 29.5 Min depth: 0.9 Max depth: 3.4 Time and Height of HT: 15:53 (2.2)

Substrate T1 T2 T3 T4 T5 Mean % Mean Sand 10 9 5 10 13 9.4 18.8 Rubble 5 7 0 5 5 4.4 8.8 Rock 00000 0 0 Coralline algae 0 0 0 0 0 0 0 Macroalgae 1 5 4 1 1 2.4 4.8 Turf algae 1 4 0 0 0 1 2 Dead coral 0 0 0 2 0 0.4 0.8 Dead coral with algae 4 1 11 7 3 5.2 10.4 Soft coral 0 0 0 0 0 0 0 Other 00560 2.2 4.4 Sponge 0 0 0 0 0 0 0 Live Coral 29 24 25 19 28 25 50 Coral species Massive Porites 26000 1.6 6.4 Porites cylindrica 10 11 21 12 14 13.6 54.4 Porites rus 13020 1.2 4.8 Pocillopora damicornis 54201 2.4 9.6 Acropora microphthalma 1 0 2 2 11 3.2 12.8 Acropora pulchra 20020 0.8 3.2 Pavona frondifera 70011 1.8 7.2 Pavona decussata 10000 0.2 0.8 Leptastrea purpurea 00001 0.2 0.8

70 10 60 8 50 40 6 30 4

20 No. Species (%) cover Coral 10 2 0 0 12345 12345 Tra n s e c t Tra n s e c t

APPENDICES ______

APPENDIX 6 continued

5. AUA

Date: 21/07/2004 Turbidity (m): 4 Salinity: 1026 Temperature (oC): 28 Min depth: 0.6 Max depth: 1.3 Time and Height of HT: 10:34 (2.3)

Substrate T1 T2 T3 T4 T5 Mean % Mean Sand 1 0 3 0 3 1.4 2.8 Rubble 8 7 6 12 8 8.2 16.4 Rock 00000 0 0 Coralline algae 11 9 7 13 11 10.2 20.4 Macroalgae 14 8 14 3 8 9.4 18.8 Turf algae 6 8 5 9 7 7 14 Dead coral with algae 0 2 1 0 1 0.8 1.6 Soft Coral 00000 0 0 Other 0 0 1 0 0 0.2 0.4 Sponge 0 2 4 7 4 3.4 6.8 Live Coral 10 14 9 6 8 9.4 18.8 Coral species Massive Porites 2 0 2 0 0 0.8 8.5 Porites cylindrica 0 0 0 2 0 0.4 4.3 Porites rus 1 0 1 0 1 0.6 6.4 Pocillopora damicornis 1 3 0 2 3 1.8 19.1 Pocillopora verrucosa 1 0 2 1 0 0.8 8.5 Pocillopora eydouxi 0 1 0 0 0 0.2 2.1 Acropora sp. (T) 44403 3 31.9 Pavona frondifera 0 1 0 1 1 0.6 6.4 Pavona decussata 1 2 0 0 0 0.6 6.4 Montipora sp. 0 2 0 0 0 0.4 4.3 Millepora dichotoma 0 1 0 0 0 0.2 2.1

30 25 8

20 6 15 4 10 No. Species

Coral cover (%) cover Coral 2 5 0 0 12345 12345 Tr a n s e ct Tra n s e c t

APPENDICES ______

APPENDIX 6 continued

6. FAGA’ITUA

Date: 19/07/2004 Turbidity (m): 19.5 Salinity: 1024 Temperature (oC): 29 Min depth: 0.5 Max depth: 1.1 Time and Height of HT: 09:02 (2.5)

Substrate 1 2 3 4 5 Mean % Mean Sand 00005 1 2 Rubble 0 0 0 3 0 0.6 1.2 Rock 00000 0 0 Coralline algae 1 1 0 0 1 0.6 1.2 Macroalgae 4 1 0 2 3 2 4 Turf algae 2 0 0 4 4 2 4 Dead coral 00000 0 0 Dead coral with algae 7 11 12 11 6 9.4 18.8 Soft Coral 00000 0 0 Other 1 4 1 0 0 1.2 2.4 Live Coral 35 33 37 30 31 33.2 66.4 Coral species Massive Porites 100582.8 8.4 Porites cylindrica 00195 3 9.0 Porites rus 000210.6 1.8 Pocillopora damicornis 34756 5 15.1 Acropora microphthalma 13 28 18 0 0 11.8 35.5 Acropora hyacinthus 001000.2 0.6 Acropora pulchra 16 0 4 0 0 4 12.0 Acropora nobilis 200100.6 1.8 Pavona frondifera 005393.4 10.2 Pavona decussata 001521.6 4.8 Galaxea fascicularis 010000.2 0.6

80 8 70 7 60 6 50 5 40 4 30 3 20 No. Species 2 Coral cover (%) cover Coral 10 1 0 0 12345 12345 Tra n s e c t Tr a n s e ct

APPENDICES ______

Goniastrea retiformis 3 3 3 2 2 2.6 10.8 Favia sp. 2 0 1 0 0 0.6 2.5 Favites sp. 0 0 1 1 0 0.4 1.7 Platygyra daedalea 1 1 0 1 0 0.6 2.5 Platygyra pini 0 1 0 0 0 0.2 0.8 Echinopora lamellosa 0 0 0 0 4 0.8 3.3 Heliopora 1 0 0 0 2 0.6 2.5

60 14 50 12 10 40 8 30 6 20 No. Species 4 Coral cover (%) cover Coral 10 2 0 0 12345 12345 Tr a n s e ct Tra n s e c t

APPENDICES ______

APPENDIX 6 continued

7. AUA CONTROL

Date: 27/07/2004 Turbidity (m): 1.5 Salinity: 1026 Temperature (oC): 28 Min depth: 1 Max depth: 1.5 Time and Height of HT: 16:51 (2.4)

Substrate T1 T2 T3 T4 T5 Mean % Mean Sand 9 6 15 15 14 11.8 23.6 Rubble 18 21 8 14 13 14.8 29.6 Rock 00000 0 0 Coralline algae 01000 0.2 0.4 Macroalgae 12 15 19 15 15 15.2 30.4 Turf algae 10503 1.8 3.6 Dead coral 00000 0 0 Dead coral with algae 20000 0.4 0.8 Soft Coral 00000 0 0 Other 00000 0 0 Sponge 10000 0.2 0.4 Live Coral 77365 5.6 11.2 Coral species Massive Porites 24032 2.2 39.3 Porites cylindrica 10000 0.2 3.6 Pocillopora damicornis 43332 3 53.6 Acropora sp. (T) 00001 0.2 3.6

4 15 3 10 2

5 No. Species 1 Coral cover (%) Coral 0 0 12345 12345 Transect Tr a n s e ct

APPENDICES ______

APPENDIX 6 continued

8. NU’UULI

Date: 28/07/2004 Turbidity (m): 8.5 Salinity: 1028 Temperature (oC): 29 Min depth: 1.2 Max depth: 3.6 Time and Height of HT: 17:44 (2.5)

Substrate T1 T2 T3 T4 T5 Mean % Mean Sand 7 11 13 17 5 10.6 21.2 Rubble 6 3 3 4 7 4.6 9.2 Rock 00000 0 0 Coralline algae 0 0 0 0 0 0 0 Macroalgae 0 1 0 0 0 0.2 0.4 Turf algae 3 2 6 5 0 3.2 6.4 Dead coral 00000 0 0 Dead coral with algae 4 5 1 2 15 5.4 10.8 Soft Coral 30002 1 2 Other 2 2 2 1 1 1.6 3.2 Sponge 003100.8 1.6 Live Coral 25 26 22 20 20 22.6 45.2 Coral species Massive Porites 305903.4 15.0 Porites cylindrica 5156526.6 29.2 Porites rus 010081.8 8.0 Pocillopora damicornis 001200.6 2.7 Acropora pulchra 00005 1 4.4 Pavona frondifera 1497457.8 34.5 Leptastrea purpurea 100000.2 0.9 Alveopora sp. 213001.2 5.3

60 6 50 5 40 4 30 3 20 2 No. Species Coral cover (%) 10 1 0 0 12345 12345 Tr a n s e ct Tr a n s e ct

APPENDICES ______

APPENDIX 6 continued

9. MASEFAU

Date: 5/08/2004 Turbidity (m): 9 Salinity: 1025 Temperature (oC): 28 Min depth: 0.6 Max depth: 2.1 Time and Height of HT: 12:01 (1.9)

Substrate T1 T2 T3 T4 T5 Mean % Mean Sand 201000.6 1.2 Rubble 5 7 11 0 6 5.8 11.6 Coralline algae 14 9 2 17 12 10.8 21.6 Macroalgae 0 2 1 1 2 1.2 2.4 Turf algae 5 2 12 8 7 6.8 13.6 Dead coral 010000.2 0.4 Dead coral with algae 1 0 1 0 2 0.8 1.6 Soft Coral 00000 0 0 Live Coral 23 29 22 24 22 24 48 Coral species Massive Porites 002200.8 3.3 Porites cylindrica 10 10 0 0 0 4 16.5 Porites rus 6163456.8 28.1 Pocillopora damicornis 103211.4 5.8 Pocillopora verrucosa 101431.8 7.4 Acropora hyacinthus 101000.4 1.7 Acropora gemmifera 100100.4 1.7 Acropora aspera 00500 1 4.1 Pavona frondifera 011131.2 5.0 Montipora sp. 224232.6 10.7 Leptastrea purpurea 102010.8 3.3 Millepora dichotoma 000421.2 5.0 Galaxea fascicularis 00032 1 4.1 Favia stelligera 000100.2 0.8 Leptoria phrygia 010020.6 2.5

70 12 60 10 50 8 40 6 30 20 4 No. Species Coral coverCoral (%) 10 2 0 0 12345 12345 Tr a n s e c t Tr a n s e ct APPENDICES ______

APPENDIX 6 continued

10. FAGATELE BAY

Date: 5/08/2004 Turbidity (m): 9 Salinity: 1025 Temperature (oC): 28 Min depth: 0.6 Max depth: 2.1 Time and Height of HT: 12:01 (1.9)

Substrate T1 T2 T3 T4 T5 Mean % Mean Coralline algae 4 9 9 9 0 6.2 12.4 Macroalgae 3 3 0 1 3 2 4 Turf algae 2 1 10 5 12 6 12 Dead coral 0 0 0 1 0 0.2 0.4 Dead coral with algae 3 3 0 2 1 1.8 3.6 Live Coral 38 34 31 32 34 33.8 67.6 Coral species Massive Porites 204101.4 4.1 Porites cylindrica 9115 5118.2 24.3 Porites rus 11 12 8 0 3 6.8 20.1 Acropora microphthalma 020000.4 1.2 Acropora hyacinthus 010000.2 0.6 Acropora pulchra 000010.2 0.6 Acropora sp. 000020.4 1.2 Pavona frondifera 10 6 6 5 15 8.4 24.9 Pavona decussata 000210.6 1.8 Montipora sp. 002100.6 1.8 Leptastrea purpurea 210000.6 1.8 Millepora dichotoma 100000.2 0.6 Millepora platyphylla 000200.4 1.2 Galaxea fascicularis 10270 2 5.9 Coscinaraea columna 000200.4 1.2 Fungia sp. 110010.6 1.8 Lobophyllia hemprichii 100701.6 4.7 Turbinaria reniformis 002000.4 1.2 Psammocora sp. 001000.2 0.6 Favia sp. 001000.2 0.6

80 10 60 8 40 6 4 20

Coral coverCoral (%) 2 No. Species 0 0 12345 12345 Tr a n s e ct Tr a n s e ct APPENDICES ______

APPENDIX 6 continued

11. OFU

Date: 2/08/2004 Turbidity (m): 15 Salinity: 1026 Temperature (oC): 29 Min depth: 1.1 Max depth: 1.7 Time and Height of HT: 09:18 (2.7)

Substrate T1 T2 T3 T4 T5 Mean % Mean Sand 13 12 14 13 9 12.2 24.4 Rb 4 3 5 4 6 4.4 8.8 Rk 2 2 0 0 0 0.8 1.6 CA 3 2 3 4 4 3.2 6.4 AA 2 2 3 5 1 2.6 5.2 TA 0 1 2 0 6 1.8 3.6 DC 00000 0 0 DCA 4 0 0 0 1 1 2 Soft Coral 00000 0 0 ZO 00000 0 0 Sponge 0 0 0 0 0 0 0 Live Coral 22 28 23 24 23 24 48 Coral species Massive Porites 4 8 5 6 4 5.4 22.5 Porites cylindrica 0 2 0 0 1 0.6 2.5 Porites sp. 2 3 2 5 1 3 2.8 11.7 Pocillopora verrucosa 1 0 0 0 2 0.6 2.5 Pocillopora meandrina 1 0 0 1 0 0.4 1.7 Pocillopora eydouxi 0 1 0 1 0 0.4 1.7 Acropora valida 0 1 1 1 0 0.6 2.5 Acropora austera 0 0 0 3 0 0.6 2.5 Acropora abrotanoides 0 0 1 0 0 0.2 0.8 Pavona varians 0 3 0 0 0 0.6 2.5 Pavona divaricata 0 0 0 0 1 0.2 0.8 Montipora sp. 0 3 0 3 2 1.6 6.7 Leptastrea purpurea 1 0 3 3 2 1.8 7.5 Millepora dichotoma 1 0 0 0 0 0.2 0.8 Millepora platyphylla 0 0 1 0 0 0.2 0.8 Galaxea fascicularis 1 0 0 1 0 0.4 1.7 Favia stelligera 0 2 0 0 0 0.4 1.7 Leptoria phrygia 2 0 0 0 0 0.4 1.7 Fungia sp. 0 0 2 0 0 0.4 1.7 Lobophyllia hemprichii 1 1 0 0 0 0.4 1.7 APPENDICES ______

APPENDIX 7

Pairwise test results based on reef substrate cover of all survey sites recorded in American Samoa during June to August 2004. * = significant at the 0.05 level (5%).

Sites R Statistic Significance Level % Vatia and Auto & Amaua 0.956 0.8 Vatia and Alofau reef flat 0.232 4. Vatia and Alofau lagoon 0.988 0.8 Vatia and Aua 0.944 0.8 Vatia and Faga’itua 0.756 0.8 Vatia and Aua control 1 0.8 Vatia and Masefau 0.14 17.5* Vatia and Nu’uuli 0.988 0.8 Vatia and Fagatele Bay 0.376 2.4 Vatia and Ofu 0.916 0.8 Auto & Amaua and Alofau reef flat 0.708 0.8 Auto & Amaua and Alofau lagoon 0.904 0.8 Auto & Amaua and Aua 0.744 0.8 Auto & Amaua and Faga’itua 0.968 0.8 Auto & Amaua and Aua control 0.92 0.8 Auto & Amaua and Masefau 0.744 0.8 Auto & Amaua and Nu’uuli 0.988 0.8 Auto & Amaua and Fagatele Bay 1. 0.8 Auto & Amaua and Ofu 0.92 0.8 Alofau reef flat and Alofau lagoon 0.636 0.8 Alofau reef flat and Aua 0.864 0.8 Alofau reef flat and Faga’itua -0.036 46.* Alofau reef flat and Aua control 1 0.8 Alofau reef flat and Masefau 0.44 0.8 Alofau reef flat and Nu’uuli 0.764 0.8 Alofau reef flat and Fagatele Bay 0.216 4. Alofau reef flat and Ofu 0.72 0.8 Alofau lagoon and Aua 0.964 0.8 Alofau lagoon and Faga’itua 0.66 2.4 Alofau lagoon and Aua control 0.9 0.8 Alofau lagoon and Masefau 0.876 0.8 Alofau lagoon and Nu’uuli 0.28 5.6* Alofau lagoon and Fagatele Bay 0.992 0.8 Alofau lagoon and Ofu 0.512 0.8 Aua and Faga’itua 0.996 0.8 Aua and Aua control 1. 0.8 Aua and Masefau 0.772 0.8 Aua and Nu’uuli 0.992 0.8 Aua and Fagatele Bay 1. 0.8 Aua and Ofu 0.996 0.8 Faga’itua and Aua control 1. 0.8 Faga’itua and Masefau 0.932 0.8 Faga’itua and Nu’uuli 0.88 0.8

APPENDICES ______

Appendix 7 continued

Sites R Statistic Significance Level % Faga’itua and Fagatele Bay 0.556 0.8 Faga’itua and Ofu 0.94 0.8 Aua control and Masefau 1 0.8 Aua control and Nu’uuli 0.992 0.8 Aua control and Fagatele Bay 1 0.8 Aua control and Ofu 0.996 0.8 Masefau and Nu’uuli 0.96 0.8 Masefau and Fagatele Bay 0.556 0.8 Masefau and Ofu 0.764 0.8 Nu’uuli and Fagatele Bay 1 0.8 Nu’uuli and Ofu 0.86 0.8 Fagatele Bay and Ofu 1 0.8

The pairwise R value gives an absolute measure of how separated the sites are. Sites were either well separated (R> 0.75), overlapping but clearly different (R> 0.5) or barely separable at all (R< 0.25).

APPENDICES ______

APPENDIX 8

Bar charts to show three environmental parameters measured at each survey site in American Samoa during June to August 2004. (a) Turbidity (m), (b) Temperature (oC) and (c) Salinity. Sites were VAT = Vatia, A&A = Auto & Amaua, ARF = Alofau reef flat, AL = Alofau lagoon, FA = Faga’itua, AC = Aua control, MA = Masefau, NU = Nu’uuli, FB = Fagatele Bay.

(a) Turbidity

Secchi disc (m) disc Secchi 5

0 VAT A&A ARF AL AUA FA AC MAS NU FB OFU

Site (b) Temperature

) 29.5

28.5

Temperature (oC Temperature 27.5

27 VAT A&A ARF AL AUA FA AC MAS NU FB OFU

Site (c) Salinity

1029

1028

1027

1026 1025 Salinity 1024

1023

1022 VAT A&A ARF AL AUA FA AC MAS NU FB OFU

Site

APPENDICES ______

APPENDIX 9

SIMPER analysis showing which substrate category contributed the most to the similarity at each group of sites (Community-based Marine Protected Areas = CB MPAs, federal MPAs, and non MPAs) surveyed in American Samoa. S = sand, Rb = rubble, Rk = rock, CA = coralline algae, MA = macroalgae, TA = turf algae, DC = dead coral, DCA = dead coral with algae, and LC = live coral.

Substrate Average Average Contribution category abundance similarity (%) LC 21.48 22.26 32.86 Rb 6.40 9.51 14.05 CB MPA DCA 4.24 9.23 13.63 TA 3.24 7.73 11.41 MA 4.40 7.29 10.76 CA 5.92 6.83 10.08 LC 21.35 21.77 35.49 Rb 6.45 9.04 14.73 Non MPA TA 3.45 8.74 14.24 S 6 7.58 12.36 DCA 4 6.07 9.9 MA 4.6 4.88 7.95 LC 28.90 31.52 46.91 Federal MPA CA 4.70 11.51 17.31 MA 2.30 9.10 13.54 TA 3.90 8.63 12.85

APPENDIX 10

List of coral species recorded at 11 survey sites in American Samoa during June to August 2004. Sites were A&A = Auto & Amaua, AR = Alofau reef flat, AL = Alofau lagoon, FA = Faga’itua, AC = Aua control, Mas = Masefau, FB = Fagatele Bay.

Family Genus Species Vatia A&A AR AL Aua FA AC Mas Nu'uuli FB Ofu Acroporidae Acropora microphthalma x x x x x x Acroporidae Acropora hyacinthus x x x x Acroporidae Acropora gemmifera x x Acroporidae Acropora valida x Acroporidae Acropora cytherea x Acroporidae Acropora austera x x Acroporidae Acropora aspera x x Acroporidae Acropora pulchra x x x x x Acroporidae Acropora nobilis x Acroporidae Acropora abrotanoides x Acroporidae Acropora sp. x x Acroporidae Montipora sp. x x x x x x Agariciidae Pavona frondifera x x x x x x x x x Agariciidae Pavona decussata x x x x x x Agariciidae Pavona varians x Agariciidae Pavona divaricata x Dendrophylliidae Turbinaria reniformis x Faviidae Leptastrea purpurea x x x x x x x Faviidae Goniastrea retiformis x Faviidae Favia sp. x x Faviidae Favites sp. x APPENDIX 10 continued

Family Genus Species Vatia A&A AR AL Aua FA Aua Mas Nu'uuli FB Ofu Faviidae Platygyra daedalea x Faviidae Platygyra pini x Faviidae Echinopora lamellosa x Faviidae Favia stelligera x x Faviidae Leptoria phrygia x x Fungiidae Fungia sp. x x Milleporidae Millepora dichotoma x x x x Milleporidae Millepora platyphylla x x Milleporidae Heliopora sp. Mussidae Lobophyllia hemprichii x x Oculinidae Galaxea fascicularis x x x x Poritidae Alevopora sp. x Poritidae Porites sp. (massive) x x x x x x x x x x Poritidae Porites cylindrica x x x x x x x x x x x Poritidae Porites rus x x x x x x x x x Poritidae Porites sp. 2 x Pocilloporidae Pocillopora damicornis x x x x x x x x Pocilloporidae Pocillopora verrucosa x x x x Pocilloporidae Pocillopora meandrina x x Pocilloporidae Pocillopora eydouxi x x x Siderastreidae Coscinaraea columna x Siderastreidae Psammacora sp. x

APPENDICES ______

APPENDIX 11

The most common coral species recorded throughout the 11 survey sites in American Samoa during June to August, 2004.

Porites cylindrica Massive Porites

(Photo: Eva, NPAS) (Photo: A Lawrence)

Pocillopora damicornis Acropora microphthalma

(Photo: Eva, NPAS) (Photo: D Fenner)

Porites rus Pavona frondifera

( (Photo: Eva, NPAS) (Photo: D. Fenner) APPENDICES ______

Pavona decussate Montipora sp.

(Photo: Eva, NPAS) (Photo: D. Fenner)

Leptastrea purpurea

(Photo: Eva, NPAS) APPENDIX 12

Mean abundance of hard coral species recorded at 11 survey sites in American Samoa during June to August 2004 (n = 5 transects per site). Sites were VAT = Vatia, A&A = Auto & Amaua, ARF = Alofau reef flat, AL = Alofau lagoon, FA = Faga’itua, AC = Aua control, MAS = Masefau, NU = Nu’uuli, FB = Fagatele Bay.

VAT A&A ARF AL AUA FA AC MAS NU FB OFU Total %Total Massive Porites 04.411.60.82.82.20.8 3.41.45.423.8 9.5 Porites cylindrica 91.212.413.60.430.24 6.68.20.659.2 23.6 Porites rus 6.60.20.41.20.60.606.8 1.86.80 25 10.0 Porites sp. 2 00000000 002.82.8 1.1 Pocillopora damicornis 0 2 3 2.4 1.8 5 3 1.4 0.6 0 0 19.2 7.7 Pocillopora verrucosa 2.4 0 0 0 0.8 0 0 1.8 0 0 0.6 5.6 2.2 Pocillopora meandrina 2.2 0 0 0 0 0 0 0 0 0 0.4 2.6 1.0 Pocillopora eydouxi 0.6 0 0 0 0.2 0 0 0 0 0 0.4 1.2 0.5 Acropora microphthalma 0.2 1.4 10 3.2 0 11.8 0 0 0 0.4 0 27 10.8 Acropora hyacinthus 0 00.40 00.200.4 00.20 1.2 0.5 Acropora gemmifera 0.8 0 0 0 0 0 0 0.4 0 0 0 1.2 0.5 Acropora valida 00000000 000.60.6 0.2 Acropora cytherea 0.80000000 0000.8 0.3 Acropora austera 000.600000 000.61.2 0.5 Acropora aspera 01.2000001 0002.2 0.9 Acropora pulchra 000.40.80400 10.206.4 2.6 Acropora nobilis 000000.600 0000.6 0.2 Acropora abrotanoides 00000000 000.20.2 0.1 Acropora sp. (Table) 0000300.20 0003.2 1.3 Acropora sp. (unidentified) 0.2 0 0 0 0 0 0 0 0 0.4 0 0.6 0.2

APPENDIX 12 continued

VAT A&A ARF AL AUA FA AC MAS NU FB OFU Total %Total Pavona frondifera 1.4 1 3.8 1.8 0.6 3.4 0 1.2 7.8 8.4 0 29.4 11.7 Pavona decussata 01.40.60.20.61.60 0 00.60 5 2.0 Pavona varians 00000000 000.60.6 0.2 Pavona divaricata 00000000 000.20.2 0.1 Montipora sp. 2 0 1 0 0.4 0 0 2.6 0 0.6 1.6 8.2 3.3 Leptastrea purpurea 0 0.2 0.2 0.2 0 0 0 0.8 0.2 0.6 1.8 4 1.6 Millepora dichotoma 0 0 0 0 0.2 0 0 1.2 0 0.2 0.2 1.8 0.7 Millepora platyphylla 0 0 0 0 0 0 0 0 0 0.4 0.2 0.6 0.2 Galaxea fascicularis 000000.201 020.43.6 1.4 Favia stelligera 0 0 0 0 0 0 0 0.2 0 0 0.4 0.6 0.2 Leptoria phrygia 0 0 0 0 0 0 0 0.6 0 0 0.4 1 0.4 Alveopora sp. 00000000 1.2001.2 0.5 Coscinaraea columna 00000000 00.400.4 0.2 Fungia sp. 0 0 0 0 0 0 0 0 0 0.6 0.4 1 0.4 Lobophyllia hemprichii 0 0 0 0 0 0 0 0 0 1.6 0.4 2 0.8 Turbinaria reniformis 00000000 00.400.4 0.2 Goniastrea retiformis 00000000 002.62.6 1.0 Psammocora sp. 00000000 00.200.2 0.1 Favia sp. 0 0 0 0 0 0 0 0 0 0.2 0.6 0.8 0.3 Favites sp. 00000000 000.40.4 0.2 Platygyra daedalea 00000000 000.60.6 0.2 Platygyra pini 00000000 000.20.2 0.1 Echinopora lamellosa 00000000 000.80.8 0.3 Heliopora 00000000 000.60.6 0.2 APPENDICES ______

APPENDIX 13

Multidimensional scaling ordination plots showing the classification of 11 sites in American Samoa based on percent similarity of hard coral species abundance. The cover of each common hard coral species (a) – (i) is represented by the diameter of the red circles. Increasing size of circles illustrates increasing abundance of each species.

(a) Porites cylindrica (b) Acropora microphthalma

(e) Massive Porites (f) Pocillopora damicornis

APPENDICES ______

APPENDIX 13 continued

(g)Montipora sp. (encrusting) (h) Pavona decussata

(i) Leptastrea purpurea.

APPENDICES ______

APPENDIX 14

Pairwise test results based on coral species abundance of all survey sites recorded in American Samoa during June to August 2004. * = significant at the 0.05 level (5%).

Sites R Statistic Significance Level % Vatia and Auto & Amaua 0.636 0.8 Vatia and Alofau reef flat 0.536 0.8 Vatia and Alofau lagoon 0.406 0.8 Vatia and Aua 0.582 0.8 Vatia and Faga’itua 0.638 0.8 Vatia and Aua control 0.932 0.8 Vatia and Masefau 0.18 15.1* Vatia and Nu’uuli 0.422 0.8 Vatia and Fagatele Bay 0.284 3.2 Vatia and Ofu 0.836 0.8 Auto & Amaua and Alofau reef flat 0.282 1.6 Auto & Amaua and Alofau lagoon 0.346 0.8 Auto & Amaua and Aua 0.234 8.7* Auto & Amaua and Faga’itua 0.014 36.5* Auto & Amaua and Aua control 0.306 0.8 Auto & Amaua and Masefau 0.674 0.8 Auto & Amaua and Nu’uuli 0.404 0.8 Auto & Amaua and Fagatele Bay 0.692 0.8 Auto & Amaua and Ofu 0.822 0.8 Alofau reef flat and Alofau lagoon 0.124 18.3* Alofau reef flat and Aua 0.82 0.8 Alofau reef flat and Faga’itua 0.104 20.6* Alofau reef flat and Aua control 0.984 0.8 Alofau reef flat and Masefau 0.856 0.8 Alofau reef flat and Nu’uuli 0.624 0.8 Alofau reef flat and Fagatele Bay 0.672 0.8 Alofau reef flat and Ofu 1 0.8 Alofau lagoon and Aua 0.808 0.8 Alofau lagoon and Faga’itua 0.316 5.6 Alofau lagoon and Aua control 0.908 0.8 Alofau lagoon and Masefau 0.8 0.8 Alofau lagoon and Nu’uuli 0.468 2.4 Alofau lagoon and Fagatele Bay 0.564 0.8 Alofau lagoon and Ofu 1 0.8 Aua and Faga’itua 0.56 0.8 Aua and Aua control 0.618 0.8 Aua and Masefau 0.664 0.8 Aua and Nu’uuli 0.816 0.8 Aua and Fagatele Bay 0.928 0.8 Aua and Ofu 0.972 0.8 Faga’itua and Aua control 0.64 0.8 Faga’itua and Masefau 0.728 0.8 Faga’itua and Nu’uuli 0.38 4

APPENDICES ______

APPENDIX 14 continued

Sites R Statistic Significance Level % Faga’itua and Fagatele Bay 0.524 0.8 Faga’itua and Ofu 0.956 0.8 Aua control and Masefau 0.976 0.8 Aua control and Nu’uuli 0.88 0.8 Aua control and Fagatele Bay 1 0.8 Aua control and Ofu 1 0.8 Masefau and Nu’uuli 0.88 0.8 Masefau and Fagatele Bay 0.536 0.8 Masefau and Ofu 0.984 0.8 Nu’uuli and Fagatele Bay 0.244 7.9* Nu’uuli and Ofu 1 0.8 Fagatele Bay and Ofu 0.988 0.8

APPENDICES ______

APPENDIX 15

SIMPER analysis showing which coral species contributed the most to the similarity among each of the 11 sites surveyed in American Samoa during June to August 2004.

Site Hard coral species Average Average Contribution abundance similarity (%) Porites rus 6.60 12.87 33.61 Porites cylindrica 9.00 11.36 29.65 Vatia Pocillopora verrucosa 2.40 3.54 9.25 Pavona frondifera 1.40 3.41 8.91 Acropora gemmifera 0.80 2.76 7.19 Montipora sp. 2.0 1.86 4.86 Pocillopora damicornis 2.00 12.53 34.45 Auto & Amaua Massive Porites 4.40 11.13 30.61 Pavona decussata 1.40 4.84 13.32 Pavona frondifera 1.00 4.56 12.54 Acropora microphthalma 10.00 22.09 40.84 Alofau reef flat Porites cylindrica 12.40 16.92 31.27 Pocillopora damicornis 3.00 6.32 11.68 Pavona frondifera 3.80 3.98 7.36 Porites cylindrica 13.60 34.35 59.07 Alofau lagoon Pocillopora damicornis 2.40 7.72 13.27 Acropora microphthalma 3.20 7.54 12.97 Porites rus 1.20 3.30 5.67 Acropora sp. 3.00 16.04 35.44 Aua Pocillopora damicornis 1.80 11.38 25.15 Pocillopora verrucosa 0.80 5.02 11.09 Porites rus 0.60 4.83 10.68 Pavona frondifera 0.60 4.66 10.31 Pocillopora damicornis 5.00 16.38 38.35 Faga’itua Acropora microphthalma 11.80 10.17 23.82 Pavona frondifera 3.40 4.18 9.80 Massive Porites 2.80 3.26 7.64 Porites cylindrica 3.00 3.12 7.32 Pavona decussata 1.60 2.51 5.88 Aua control Pocillopora damicornis 3.00 48.24 67.79 Massive Porites 2.20 22.93 32.21 Porites rus 6.80 15.48 30.15 Montipora sp. 2.60 11.21 21.83 Pocillopora verrucosa 1.80 4.95 9.64 Masefau Pocillopora damicornis 1.40 4.73 9.21 Pavona frondifera 1.20 4.57 8.91 Porites cylindrica 4.00 2.80 5.45 Leptastrea purpurea 0.80 2.30 4.47 Massive Porites 0.80 0.99 1.93 Nu’uuli Pavona frondifera 7.80 24.73 43.29 Porites cylindrica 6.60 20.73 36.28 Massive Porites 3.40 5.95 10.42 Alveopora sp. 1.20 3.51 6.15

APPENDICES ______

APPENDIX 15 continued

Site Hard coral species Average Average Contribution abundance similarity (%) Porites cylindrica 8.20 16.96 32.47 Fagatele Bay Pavona frondifera 8.40 16.50 31.59 Porites rus 6.80 9.71 18.59 Galaxea fascicularis 2.00 2.18 4.18 Massive Porites 1.40 2.18 4.17 Massive Porites 5.40 13.83 27.40 Goniastrea retiformis 2.60 9.98 19.77 Porites sp. 2 2.80 8.95 17.72 Ofu Leptastrea purpurea 1.80 5.12 10.15 Montipora sp. 1.60 2.96 5.86 Acropora valida 0.60 1.97 3.91 Platygyra daedalea 0.60 1.85 3.67 Favites sp. 0.40 0.68 1.35 Favia sp. 0.60 0.67 1.34

APPENDICES ______

APPENDIX 16

SIMPER analysis showing which coral species contributed the most to the similarity at each group of sites (Community-based Marine Protected Areas = CB MPAs, federal MPAs, and non MPAs) surveyed in American Samoa during June to August 2004.

Group Hard coral species Average Average Contribution abundance similarity (%) Porites cylindrica 7.32 9.82 30.29 Pocillopora damicornis 1.84 6.23 19.21 CB MPA Pavona frondifera 1.72 4.25 13.11 Acropora microphthalma 2.96 3.44 10.63 Porites rus 1.80 3.11 9.59 Massive Porites 1.56 2.57 7.94 Pocillopora damicornis 2.50 11.31 34.51 Massive Porites 2.30 6.42 19.58 Non MPA Porites cylindrica 3.45 5.03 15.35 Pavona frondifera 3.10 4.93 15.03 Porites rus 2.30 2.40 7.33 Massive Porites 3.40 6.76 20.67 Porites cylindrica 4.40 5.71 17.45 Pavona frondifera 4.20 3.67 11.21 Leptastrea purpurea 1.20 2.66 8.14 Federal MPA Goniastrea retiformis 1.30 2.22 6.78 Porites rus 3.40 2.16 6.60 Porites sp. 2 1.40 1.99 6.08 Montipora sp. 1.10 1.85 5.67 Galaxea fascicularis 1.20 1.46 4.46 Fungia sp. 0.50 0.94 2.89 Lobophyllia hemprichii 1.00 0.82 2.51

APPENDIX 17

Mean abundance data used in the PRIMER multivariate analysis of coral species relationships, recorded at 11 survey sites in American Samoa (n = 5 transects per site). Sites were VAT = Vatia, A&A = Auto & Amaua, ARF = Alofau reef flat, AL = Alofau lagoon, FA = Faga’itua, AC = Aua control, MAS = Masefau, NU = Nu’uuli, FB = Fagatele Bay.

VAT A&A ARF AL AUA FA AC MAS NU FB OFU Massive Porites 0 4.4 1 1.6 0.8 2.8 2.2 0.8 3.4 1.4 5.4 Porites cylindrica 9 1.2 12.4 13.6 0.4 3 0.2 4 6.6 8.2 0.6 Porites rus 6.6 0.2 0.4 1.2 0.6 0.6 0 6.8 1.8 6.8 0 Porites sp. 2 0 000000 0002.8 Pocillopora damicornis 0 2 3 2.4 1.8 5 3 1.4 0.6 0 0 Pocillopora verrucosa 2.4 0000.800 1.8000.6 Pocillopora meandrina 2.2 000000 0000.4 Pocillopora eydouxi 0.6 0000.200 0000.4 Acropora microphthalma 0.2 1.4 10 3.2 0 11.8 0 0 0 0.4 0 Acropora hyacinthus 0 0 0.4 0 0 0.2 0 0.4 0 0.2 0 Acropora gemmifera 0.8 000000 0.4000 Acropora austera 0 00.60000 0000.6 Acropora aspera 0 1.200000 1000 Acropora pulchra 0 00.40.8040 010.20 Acropora sp. 0.2 000300.2 000.40 Pavona frondifera 1.4 1 3.8 1.8 0.6 3.4 0 1.2 7.8 8.4 0 Pavona decussata 0 1.4 0.6 0.2 0.6 1.6 0 0 0 0.6 0 Montipora sp. 2 0100.400 2.600.61.6

APPENDIX 17 continued

VAT A&A ARF AL AUA FA AC MAS NU FB OFU Leptastrea purpurea 0 0.2 0.2 0.2 0 0 0 0.8 0.2 0.6 1.8 Millepora sp. 0 0000.200 1.200.60.4 Galaxea fascicularis 0 00000.20 1020.4 Leptoria phrygia 0 000000 0.6000.4 Alveopora sp. 0 000000 01.200 Fungia sp. 0 000000 000.60.4 Lobophyllia hemprichii 0 000000 001.60.4 Goniastrea retiformis 0 000000 0002.6 Favia sp. 0 000000 0.600.21

APPENDICES ______

APPENDIX 18

Pairwise test results based on reef substrate cover and coral species abundance of all survey sites recorded in American Samoa during June to August 2004. * = significant at the 0.05 level (5%).

Sites R Statistic Significance Level % Vatia and Auto & Amaua 0.852 0.8 Vatia and Alofau reef flat 0.62 0.8 Vatia and Alofau lagoon 0.852 0.8 Vatia and Aua 0.748 0.8 Vatia and Faga’itua 0.788 0.8 Vatia and Aua control 1 0.8 Vatia and Masefau 0.232 7.1* Vatia and Nu’uuli 0.804 0.8 Vatia and Fagatele Bay 0.416 0.8 Vatia and Ofu 0.936 0.8 Auto & Amaua and Alofau reef flat 0.604 0.8 Auto & Amaua and Alofau lagoon 0.8 0.8 Auto & Amaua and Aua 0.668 0.8 Auto & Amaua and Faga’itua 0.52 1.6 Auto & Amaua and Aua control 0.644 0.8 Auto & Amaua and Masefau 0.836 0.8 Auto & Amaua and Nu’uuli 0.952 0.8 Auto & Amaua and Fagatele Bay 0.976 0.8 Auto & Amaua and Ofu 0.992 0.8 Alofau reef flat and Alofau lagoon 0.44 2.4 Alofau reef flat and Aua 0.88 0.8 Alofau reef flat and Faga’itua 0 40.5* Alofau reef flat and Aua control 0.964 0.8 Alofau reef flat and Masefau 0.732 0.8 Alofau reef flat and Nu’uuli 0.752 0.8 Alofau reef flat and Fagatele Bay 0.576 0.8 Alofau reef flat and Ofu 0.972 0.8 Alofau lagoon and Aua 0.968 0.8 Alofau lagoon and Faga’itua 0.424 4 Alofau lagoon and Aua control 0.856 0.8 Alofau lagoon and Masefau 0.92 0.8 Alofau lagoon and Nu’uuli 0.444 2.4 Alofau lagoon and Fagatele Bay 0.92 0.8 Alofau lagoon and Ofu 0.992 0.8 Aua and Faga’itua 0.812 0.8 Aua and Aua control 0.948 0.8 Aua and Masefau 0.748 0.8 Aua and Nu’uuli 0.984 0.8 Aua and Fagatele Bay 0.996 0.8 Aua and Ofu 1 0.8 Faga’itua and Aua control 0.864 0.8 Faga’itua and Masefau 0.84 0.8 Faga’itua and Nu’uuli 0.6 0.8

APPENDICES ______

APPENDIX 18 continued

Sites R Statistic Significance Level % Faga’itua and Fagatele Bay 0.64 0.8 Faga’itua and Ofu 0.96 0.8 Aua control and Masefau 0.98 0.8 Aua control and Nu’uuli 0.936 0.8 Aua control and Fagatele Bay 1 0.8 Aua control and Ofu 1 0.8 Masefau and Nu’uuli 0.988 0.8 Masefau and Fagatele Bay 0.628 0.8 Masefau and Ofu 0.964 0.8 Nu’uuli and Fagatele Bay 0.924 0.8 Nu’uuli and Ofu 0.988 0.8 Fagatele Bay and Ofu 1 0.8

APPENDICES ______

APPENDIX 19

SIMPER analysis showing which reef substrate or coral species contributed the most to the similarity among each of the 11 sites surveyed in American Samoa during June to August 2004. S = sand, Rb = rubble, Rk = rock, CA = coralline algae, MA = macroalgae, TA = turf algae, DC = dead coral, DCA = dead coral with algae, and LC = live hard coral.

Site Coral species of reef Average Average Contribution substrate abundance similarity (%) CA 13.00 15.91 30.28 TA 3.40 8.04 15.30 Porites rus 6.60 6.57 12.49 Vatia Porites cylindrica 9.00 5.72 10.89 DCA 3.80 4.68 8.91 Rb 2.00 3.66 6.96 Pocillopora verrucosa 2.40 1.91 3.63 Pavona frondifera 1.40 1.77 3.37 Rb 15.60 17.32 28.56 MA 8.00 8.87 14.62 DCA 4.20 8.85 14.60 Auto & Amaua CA 3.80 6.88 11.34 Pocillopora damicornis 2.00 4.26 7.02 Massive Porites 4.40 3.91 6.45 S 1.60 3.11 5.13 TA 2.80 2.41 3.97 Acropora microphthalma 10.00 13.60 25.43 Porites cylindrica 12.40 10.51 19.65 DCA 7.20 9.67 18.08 Alofau reef flat Pocillopora damicornis 3.00 3.92 7.32 CA 2.60 3.75 7.00 Pavona frondifera 3.80 2.49 4.65 TA 2.00 2.16 4.04 MA 1.60 1.98 3.71 Massive Porites 1.00 1.54 2.89 Porites cylindrica 13.60 16.57 26.17 S 9.40 13.53 21.36 Alofau lagoon DCA 5.20 7.88 12.44 Rb 4.40 6.41 10.12 MA 2.40 5.50 8.68 Pocillopora damicornis 2.40 3.92 6.18 Acropora microphthalma 3.20 3.64 5.75 CA 10.20 13.71 19.52 Rb 8.20 12.16 17.32 MA 9.40 11.40 16.24 Aua TA 7.00 11.31 16.11 Acropora sp. 3.00 5.04 7.18 SP 3.40 4.72 6.72 Pocillopora damicornis 1.80 3.53 5.02 S 1.40 1.71 2.44

APPENDICES ______

APPENDIX 19 continued

Site Hard coral species Average Average Contribution abundance similarity (%) DCA 9.40 14.18 28.31 Pocillopora damicornis 5.00 10.09 20.14 Acropora microphthalma 11.80 6.40 12.76 MA 2.00 3.75 7.47 Faga’itua Pavona frondifera 3.40 2.71 5.42 TA 2.00 2.24 4.47 Porites cylindrica 3.00 1.98 3.95 Massive Porites 2.80 1.95 3.89 OTH 1.20 1.67 3.33 Pavona decussata 1.60 1.61 3.21 MA 15.20 23.03 28.71 Aua Control Rb 14.80 20.90 26.05 S 11.80 18.72 23.33 Pocillopora damicornis 3.00 9.95 12.41 CA 10.80 11.10 18.71 TA 6.80 8.83 14.87 Porites rus 6.80 8.68 14.62 Montipora sp. 2.60 6.29 10.60 Masefau Rb 5.80 6.21 10.46 Pocillopora verrucosa 1.80 2.84 4.78 Pocillopora damicornis 1.40 2.70 4.55 Pavona frondifera 1.20 2.60 4.39 MA 1.00 2.60 4.39 Porites cylindrica 4.00 1.48 2.50 Leptastrea purpurea 0.80 1.25 2.11 S 10.60 12.99 19.82 Pavona frondifera 7.80 11.01 16.80 Nu’uuli Porites cylindrica 6.60 9.22 14.07 Rb 4.60 8.90 13.59 DCA 5.40 6.97 10.64 OTH 1.60 5.38 8.20 TA 3.20 4.70 7.17 Porites cylindrica 8.20 11.48 20.27 Pavona frondifera 8.40 11.18 19.75 TA 6.00 7.31 12.90 Fagatele Bay CA 6.20 6.70 11.83 Porites rus 6.80 6.65 11.73 MA 2.00 3.75 6.63 DCA 1.80 3.40 6.01 Massive Porites 1.40 1.50 2.65 S 12.20 12.69 20.26 Massive Porites 5.40 8.00 12.78 Rb 4.40 7.33 11.71 Ofu CA 3.20 6.24 9.97 Goniastrea retiformis 2.60 5.13 9.21 Porites sp.2 2.80 4.89 8.20 MA 2.60 2.92 7.82 Leptastrea purpurea 1.80 1.75 4.67 Montipora sp. 1.60 1.34 2.79 APPENDICES ______

APPENDIX 20

Diversity indices describing the coral reef fish community at 11 sites surveyed in American Samoa during June to August 2004. S = total number of species at each site, N = total fish abyndance, d = species richness, calculated using Margalef’s index, J’ = evenness, calculated using Pielou’s Index of Eveness and H’ = diversity, calculated using the Shannon-Weiner Diversity Index. Data collected by A. Lawrence.

S N d J’ H' Vatia 28 51.2 6.86 0.85 2.82 Auto & Amaua 45 103.6 9.48 0.87 3.30 Alofau reef flat 34 128.8 6.79 0.77 2.73 Alofau lagoon 3068.956.850.692.36 Aua 2345.455.760.852.65 Faga'itua 32 131 6.36 0.63 2.18 Aua control 22 38.7 5.74 0.82 2.52 Masefau 53 111 11.04 0.76 3.02 Nu'uulii 39 58.8 9.33 0.77 2.81 Fagatele Bay 32 85.4 6.97 0.76 2.63 Ofu 42 80.4 9.35 0.81 3.02

APPENDICES ______

APPENDIX 21

Scale values used for Table 3.4 - Environmental, biological, physical and social factors of the 11 sites surveyed in American Samoa during June to August 2004.

1. Substrate

Scale % cover 0 0 1 1 - 15 2 >15 - 30 3 >30 - 45 4 >45 - 60 5 >60 - 75 6 >75

2. Habitat Area

Scale Area (Acres) 1 0-10 2 >10-20 3 >20-30 4 >30-35 5 >35-40 6 >40-50

3. Habitat width

Scale Width (m) 0 0 1 1 - 100 2 >100 - 200 3 >200 - 300 4 >300 - 400 5 >400 - 500 6 >500

4. Fish Abundance

Scale Numbers 1 1 - 25 2 >25 -50 3 >50 -75 4 >75 - 100 5 >100 - 125 6 >125 -150

APPENDICES ______

5. Watershed Area

Scale Area (km2) 0 0 1 1 - 2 2 >2 - 3 3 >3 - 4 4 >4 - 5 5 >5 - 6 6 >6

6. Population Density

Scale Pop. Density (per km2) 0 0 1 1 - 500 2 >500 - 1000 3 >1000 - 2000 4 >2000 - 3000 5 >3000 - 5000 6 >5000