Integrated Site Suitability and Carrying Capacity Assessment for Small-Scale Finfish Net Cage Mariculture in Marine Protected Areas, Indonesia

Hatim Albasri

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

SCHOOL OF BIOLOGICAL, EARTH AND ENVIRONMENTAL SCIENCES

FACULTY OF SCIENCE THE UNIVERSITY OF NEW SOUTH WALES December 2018

PLEASE TYPE

THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet Surname or Family Name: Albasri First name: Hatim Other names/s: - University calendar: PhD School: BEES Faculty: SCIENCE

Title: Integrated Site Suitability and Carrying Capacity Assessment for Small-Scale Finfish Net Cage Mariculture in Marine Protected Areas, Indonesia

Abstract 350 words maximum. (PLEASE TYPE)

Small-scale mariculture has been formally recognised as one of the main strategic economic developments for small- island communities. However, mariculture, especially for carnivorous fish species, is threatened by sustainability issues, that not only affect the livelihood of marginalised small-island communities, but also ecosystems in marine protected area (MPA). These issues include the lack of adequate regulation, unclear site selection criteria and unknown carrying capacity of MPA mariculture zones. This study evaluated the complex regulations that govern development of mariculture in Indonesia and assessed the sustainability of small-scale mariculture in multi-use MPAs using a sustainable livelihood analysis framework. Furthermore, it developed an integrated site selection and carrying capacity framework for better decision-making for MPA mariculture zones using environmental data and modelling. The designation of mariculture zones, local government and community approval-based permits, and third-party environmental impact assessment could substantially improve the sustainability of small-scale mariculture in MPAs. Small-scale mariculture, as a livelihood, has similar sustainability and vulnerability profiles compared to small-scale fisher and ecotourism households, which tend to be at the intermediate and medium level, respectively. Small-scale fish farmers also shared similar views with other household groups, that is, that MPA establishment is important to maintain ecosystem function, which supports their livelihoods. Site selection using a parameter-specific suitability function, with geometric mean, a stakeholder preferences sub model and MPA constraints, maintained the integrity of MPA core zones and allowed better location access for small-scale fish farmers. The Modelling – Ongrowing fish farm – Monitoring (MOM) system revealed that oxygen concentration was the main factor influencing the carrying capacity in different seasonal scenarios, feed composition and geographic location. The carrying capacity analysis indicated that the general MPA carrying capacity is contra productive to the development of small-scale mariculture in small-island MPAs and needs to be reconsidered. The site selection framework developed in this study and the MOM system could provide an efficient and easy assessment to establish mariculture zones in MPAs, and ensure fair resource and spatial allocation in Indonesia’s MPAs.

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

I also authorise University Microfilms to use the 350 words abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only).

………………………… …………………………… 11 December 2018 Signature Witness Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or condition on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of the Graduate Research

FOR OFFICE USE ONLY Date of completion of requirement for Award:

THIS SHEET IS TO BE GLUED TO THE INSIDE FRONT COVER OF THE THESIS ii

ORIGINALITY STATEMENT

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

Signed

Date

Date: 11 December, 2018.

iii

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

Signed

Date

AUTHENTICITY STATEMENT

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

Signed

Date

ACKNOWLEDGEMENTS

I remember the repeated advice from my primary supervisor, Jes Sammut, that my thesis is not meant to be a Nobel Prize. Instead, it is a solid piece of original and significant work that will lay the foundation for my future career as a researcher. This is one amongst the many pieces of advice I have received from you during my PhD candidature that have kept my spirit high to complete my thesis and my study. For that and other countless supports, I offer my sincere gratitude. I also thank Richard Lucas, Damon Bolton and Jennifer Beer as my co-supervisors for their full support during my PhD candidature. I am deeply indebted to both Damon Bolton and Jes Sammut who had spent their valuable time as editors to improve the readability of this thesis. It has been an honour to know and work with you at BEES-UNSW.

I am also grateful to AusAID who had given me a chance to study at UNSW – Australia through its Australia Awards Scholarships. I thought that only financial support was provided from the former AusAID and I was wrong. AusAID also supported other aspects of my study including academic preparation, personal wellbeing, research needs, and outreach programs which I know some other scholarships do not offer. My deep appreciation goes to people behind AusAID who work diligently and tirelessly to support our study, Mas Alkadri and Mas Ponco from CID – AAS –AusAID; Tatjana Kroll, Matthew Byron and Pia Larsson from UNSW, you guys are simply the best!

I specifically would like to thank mas Bas and mas Pran for their help with the preparation and activities during the four field visit in Anambas Archipelago. Of course, this thesis would have materialised if the Anambas community and the local government did not give their permission. For your kindness and willingness to welcome me, please accept my thankfulness.

And it is your guys’ turn, my colleagues at Aquaculture Research Group or 601 room in alphabetical order: Angela, Bayu, Brah! Havini, Irja, Justin, Lala, Shanice. I am lucky and honoured to have known you and be your friend in Sydney, Australia. Jenny Saunders, you are obviously not gone unnoticed, thank you for your help. Special thanks to Shawn Laffan who has helped me many times with GIS software and Jonathan Russel who provide almost 24/7 support for any academic inquiries.

To my better-half ‘Bunda’ and our children, Aurel ‘my Borealis light’, Zaky “the king’ and Arshad ‘the gift/ed’, I dedicate this thesis for all of you, though it cannot replace even only a small fraction of your sacrifices and unending support. Thanks to my ‘Ma’, ‘Pa’ and brothers and sisters for their supports and encouragements.

To the most gracious and mighty, Allah SWT, science has led me closer to You.

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ...... v TABLE OF CONTENTS ...... vi LIST OF TABLES ...... x LIST OF FIGURES ...... xii LIST OF APPENDIXES ...... xiv GLOSSARY OF TERMS AND ABBREVIATIONS ...... xv ABSTRACT ...... xvii CHAPTER 1. Introductory Literature Review ...... 1 1.1 Introduction ...... 1 1.2 MPAs and sustainable development ...... 3 1.3 Mariculture and other sustainable livelihoods in MPA ...... 4 1.3.1 Sustainable ecotourism...... 5 1.3.2 Small-scale fisheries...... 6 1.3.3 Mariculture in MPA ...... 7 1.4 Determining sustainability of mariculture in MPA ...... 10 1.4.1 Impacts of mariculture ...... 10 1.4.2 Site selection and carrying capacity: a safety net...... 11 1.4.3 Application of hard and soft system of thinking in site selection and carrying capacity ...... 14 1.5 Objectives and hypotheses ...... 16 1.5.1 Overall objective and research questions ...... 16 1.5.2 Hypotheses ...... 17 1.6 Research framework ...... 17 CHAPTER 2. Description of Study Area ...... 19 2.1 Geographic location, history, and administrative features ...... 19 2.2 Climate, local and regional oceanographic forces ...... 20 2.3 Water chemical properties of Anambas Archipelago ...... 24 2.4 Socio-economic condition ...... 24 2.5 Anambas MPA development and management ...... 27 CHAPTER 3. Regulations for Planning and Establishment of Mariculture Zoning in Marine Protected Area in Indonesia: A Policy Analysis ...... 29 3.1 Introduction ...... 29 3.2 Methodology ...... 30 3.2.1 Research approach ...... 30 3.2.2 Data sources and collection ...... 31 3.2.3 Data analysis ...... 31 vi

3.3 Historical overview of MPA zoning regulations in Indonesia ...... 32 3.4 Policy analysis of MPA mariculture zone planning in Indonesia ...... 37 3.4.1 The issues with MPA mariculture zoning in Indonesia ...... 39 3.4.2 Objectives, value premises, theoretical positions, and effects of the regulations ...... 47 3.4.3 Implications of the regulations for mariculture zoning and establishment in Indonesia’s MPA ...... 56 3.5 Discussion: improvement and alternative policies of the regulations ...... 59 3.5.1 Sustainable mariculture development in MPA ...... 59 3.5.2 Improved mariculture permit in MPAs ...... 62 3.5.3 Collaborative implementing agency/stakeholder ...... 66 3.6 Conclusions ...... 68 CHAPTER 4. Comparative Study on Sustainability Profiles of Different Small-scale Livelihoods within Small Island MPA-Designated Communities in the Anambas Archipelago-Indonesia ...... 70 4.1 Introduction ...... 70 4.2 Methods ...... 71 4.2.1 Study area ...... 71 4.2.2 Data collection ...... 72 4.2.3 Questionnaire ...... 74 4.2.4 Data analysis ...... 75 4.3 Results ...... 76 4.3.1 Human capital ...... 76 4.3.2 Financial capital ...... 78 4.3.3 Natural capital ...... 79 4.3.4 Physical capital...... 80 4.3.5 Social capital ...... 81 4.3.6 Overall composite indexes of capital assets ...... 82 4.4 Discussion ...... 83 4.4.1 Sustainability profile of Anambas community groups based on human capital ...... 83 4.4.2 Sustainability profile of Anambas community groups based on financial capital ...... 83 4.4.3 Sustainability profile of Anambas community groups based on natural capital ...... 85 4.4.4 Sustainability profile of Anambas community groups based on physical capital ...... 86 4.4.5 Sustainability profile of Anambas community groups based on social capital ...... 88

vii

4.5 Conclusions ...... 89 CHAPTER 5. Comparative Study on Livelihood Vulnerability Risks and Conservation Support Based on Perceptions of Different Local Communities in the Anambas Archipelago MPA, Indonesia ...... 91 5.1 Introduction ...... 91 5.2 Methodology ...... 93 5.2.1 Study area ...... 93 5.2.2 Questionnaire design ...... 94 5.2.3 Data collection ...... 95 5.2.4 Data analysis ...... 96 5.3 Results ...... 98 5.3.1 Livelihood vulnerability of Anambas local community groups ...... 98 5.3.2 Small-scale community groups’ perspectives on the Anambas MPA .. 102 5.4 Discussion ...... 105 5.4.1 Livelihood vulnerability of small scale community groups ...... 105 5.4.2 Communities acceptance of an Anambas MPA establishment ...... 108 5.5 Conclusions ...... 110 CHAPTER 6. Development of Mariculture Zone Site Selection Based on GIS Approach for Small-Scale Finfish Mariculture in MPA ...... 112 6.1 Introduction ...... 112 6.2 Materials and methods ...... 113 6.2.1 Analytical framework of GIS and remote sensing ...... 113 6.2.2 MPA finfish mariculture zone site suitability ...... 118 6.3 Results ...... 130 6.3.1 Constraint sub-model ...... 130 6.3.2 Site suitability sub-model ...... 133 6.3.3 Stakeholder preference sub-model ...... 138 6.3.4 Mariculture zone for finfish net cage mariculture...... 140 6.4 Discussion ...... 142 6.4.1 GIS application in mariculture site suitability for MPA ...... 142 6.4.2 Mariculture zone for small and medium-scale fish farmers in MPA .... 145 6.5 Conclusions ...... 147 CHAPTER 7. Predictive Carrying Capacity Modelling to Establish Small-Scale Finfish Mariculture Zone in Indonesia’s MPAs ...... 149 7.1 Introduction ...... 149 7.2 Materials and methods ...... 151 7.2.1 Study sites ...... 151 7.2.2 MOM system description and modification ...... 152 viii

7.2.3 Data collection and measurement ...... 154 7.2.4 Data analysis ...... 157 7.3 Results ...... 158 7.3.1 Current characteristics ...... 158 7.3.2 The concentration of oxygen and ammonia in the farm ...... 159 7.3.3 Fish sub-model in MOM ...... 160 7.3.4 Dispersion sub-model ...... 161 7.3.5 The benthic sub-model ...... 161 7.3.6 Water quality in fish cage sub-model ...... 162 7.4 Discussion ...... 164 7.4.1 Modification of MOM system to suit with tropical mariculture condition164 7.4.2 Factors influencing CC in MPA ...... 164 7.4.3 Comparison of the predicted and existing CC/holding density ...... 167 7.4.4 Conclusion ...... 169 CHAPTER 8. Thesis Synthesis and Conclusion ...... 171 8.1 Major findings and their implications ...... 173 8.1.1 The trade-off of legal framework, resource use, and community rights in MPA management ...... 173 8.1.2 The application multidimensional indicators in sustainability and site selection for mariculture in MPAs ...... 176 8.1.3 Future development of mariculture in Indonesia’s MPAs ...... 177 8.2 Research limitations ...... 179 8.3 Conclusion ...... 180 LIST OF REFERENCES ...... 183 APPENDIXES ...... 203

LIST OF TABLES

Table 2.1. Oil and Gas Mining Companies Operating in Anambas Area ...... 26 Table 2.2. Network of Indonesia MPAs in sSCS ...... 28 Table 3.1. List of various regulations in different levels governing Indonesia’s MPAs ...... 33 Table 3.2. Multi-use MPAs in Indonesia ...... 37 Table 3.3. Mariculture zoning in some of Indonesia’s small island MUMPAS established through MMAF decrees ...... 45 Table 3.4. Land and seascape usage for aquaculture in Indonesia during 2009 - 2013 ...... 54 Table 4.1. Distribution of respondents across villages and household livelihoods in Anambas sub-districts...... 74 Table 4.2. Sustainable indicators (SIs) of capital assets used in the study...... 74 Table 4.3. Preference of weighting in different types of boat ownership indicator .... 75 Table 4.4. Comparison of human capital SI indexes of three different household groups ...... 77 Table 4.5. Comparison of financial capital SI indexes of three different household groups ...... 78 Table 4.6. Comparison of natural capital SI indexes of three different household groups ...... 79 Table 4.7. Comparison of physical capital SI indexes of three different household groups ...... 80 Table 4.8. Comparison of social capital SI indexes of three different household groups ...... 81 Table 5.1 LVIs used in the study ...... 95 Table 5.2. Distribution of respondents based on Anambas sub-districts ...... 96 Table 5.3. Comparison of seasonal LVI index values for three different household groups ...... 98 Table 5.4. Comparison of Trend LVI index values for three different household groups ...... 99 Table 5.5. Comparison of shock LVI index values for three different household groups ...... 101 Table 5.6. NVIVO Matrix coding comparison of ‘knowingness’ and types of involvement of respondents ...... 103 Table 5.7. Combination of responses of agreement about the designation of Anambas MPA and reasons of agreement based on matrix coding comparison ...... 104 Table 6.1. Parameter-specific suitability function (PSSFs) values used in the site suitability sub-model for finfish mariculture in Anambas MPA...... 127

Table 6.2. Parameter-specific Suitability Function (PSSFs) values used in stakeholder preference sub-models for finfish mariculture in Anambas MPA ...... 129 Table 6.3. Distribution of total coverage of PSSF suitability classes in the site suitability sub-model ...... 134 Table 6.4. Mean and ± standard deviation for criteria measured in 125 sampling stations during the wet and dry season ...... 138 Table 7.1. Current characteristics and concentrations of oxygen and ammonium .... 159 Table 7.2. Farm size and number of units used in the CC/holding density based on MOM system in each location (Figure 7.1) ...... 162

Table 7.3. Results of CC/holding density based on TPFO2, TPFNH4 and TPFbentam ... 163

LIST OF FIGURES

Figure 1.1. Establishment of mariculture zones near the vicinity of core and protection zones (less than 2 km) in KJMNP, Indonesia (Maynard et al., 2010)...... 9 Figure 2.1. Monthly average wind speed in Anambas Archipelago during 6 years (BMKG-Tarempa, 2014) ...... 21 Figure 2.2. Monthly average rainfall in Anambas Archipelago during 6 years period (BMKG-Tarempa, 2014) ...... 22 Figure 3.1. Regulation hierarchy regarding MPA establishment and zonation in Indonesia ...... 48 Figure 3.2. Supply chain of hatchery-bred seed of finfish in one Indonesian MPA (Map source: Effendy, 2012) ...... 58 Figure 3.3. Existing and proposed measures for mariculture permits in MPAs ...... 64 Figure 3.4. Comparison of prioritisation of mariculture site selection/carrying capacity for (a) different regions (Ferreira et al. (2012), Ross et al. (2013b) and (b) the study’s proposed site selection/carrying capacity in MPA ...... 66 Figure 4.1. Study area and surveyed villages in Anambas Archipelago MPA ...... 72 Figure 4.2. Capital asset polygon of sustainability livelihood approach from three different household groups ...... 82 Figure 5.1. Location of study sites (villages) in Anambas MPA ...... 94 Figure 5.2. Vulnerability diagram of the LVI index values for different household groups ...... 102 Figure 5.3. Percentages of respondents who know the designation of Anambas MPA ...... 103 Figure 5.4. Level of Agreement among the household regarding the potential benefits of the Anambas MPA to the local people livelihoods ...... 105 Figure 6.1. The organisational structure of site suitability sub-models used to determine finfish mariculture zone within small-island MPA in ...... 114 Figure 6.2 The effect of attribute standardisation using geometric mean where (a) standardisation is carried out on all criteria; (b) standardisation is carried out on feasible alternative (simulated data from Malcezewski (2000) .... 116 Figure 6.3. Feasible areas from the constraint sub-model: (A) turtle nesting sites; (B) district economic zone; (C) bathymetry; (D) shipping lane ...... 131 Figure 6.4. Oil discharge into the waters from a medium size transportation ship at the regional seaport of Jemaja Island ...... 133 Figure 6.5. The composite maps of PSSF classes for wet (A) and dry (B) seasons in Anambas MPA based on site suitability sub model of 17 parameters ..... 136 Figure 6.6. Site suitability of Anambas MPA based on stakeholder preference sub- model: (A) distance to MPA core zones; (B) distance to harbours; (C)

xii

distance to village/settlements; & (D) composite map of stakeholder preference sub-model ...... 139 Figure 6.7. Site suitability maps for mariculture zone in Anambas MPA: (A) wet season, (B) dry season ...... 141 Figure 7.1. Location of carrying capacity measurement ...... 151 Figure 7.2. Current characteristics (speed and dominant direction) in the wet (A) and dry (B) seasons ...... 158

xiii

LIST OF APPENDIXES

Appendix 1. Semi-structured questionnaire for sustainability, vulnerability of small-scale households in Anambas Archipelago MPA and their perspective to the MPA establishment ...... 203 Appendix 2. Semi-Structured Questionnaire for Stakeholder Preference Sub- Model in Site Selection for Small-Scale Mariculture in Anambas Archipelago MPA ...... 215 Appendix 3. Estimated and Measured MOM Input Parameters ...... 218

xiv

GLOSSARY OF TERMS AND ABBREVIATIONS

BMKG – Badan Meteorologi, Klimatologi dan Geofisika or The Indonesian Agency of Meteorology, Climatology and Geophysics Capital asset – the basic building blocks of livelihood (natural, human, physical, financial and social capitals) that are owned, can be accessed and can be used directly or indirectly by the household to engage in productive activity to support itself. Carrying capacity – the potential maximum production of fish that can be maintained without any unacceptable impact to both the fish species and the environment Chlorophyll-a – (Chl-a) a specific form of chlorophyll used in oxygenic photosynthesis that can be used as proxy value to measure primary productivity or potential harmful algal bloom in remote sensing Critical oxygen concentration – the minimum oxygen concentration required in the net cage to sustain a fish population Critical ammonium concentration – the maximum ammonium concentration allowed in the net cage that CSR – Corporate social responsibility Directorate General of Aquaculture (DGA) – one of the six directorate general offices within MMAF whose tasks are to develop aquaculture industry in Indonesia DKP – Dinas Kelautan dan Perikanan, an office under the provincial or district structure that is tasked with aquaculture development within the administrative and geographic boundary of its corresponding administrative level DO – Dissolved oxygen

DO2 – The model computed mean oxygen consumption per kg fish production in MOM system EEZ – Exclusive economic zone FAD – Fish aggregating device built by fishermen to concentrate pelagic fish for easy capture Formzahl number – The ratio of the amplitudes of the major diurnal (AK1 and AK2) to semi-diurnal partial tides used to determine tide types at locality Geometric mean – a type of average which indicates the central tendency of typical value of a set of numbers by using the product of their values calculated as the nth root of a product of n number GT – Gross tonne HAB – Harmful algal bloom KEWEDANAAN – a government level between sub-district and district established during the Dutch colonialism in Indonesia KKPD – or Kawasan Konservasi Perairan Daerah, a type of local MPA managed by provincial government if the MPA area encompasses two or more district administrative areas and managed by district government of the MPA area is within the administrative area of the district and has strategic functions

KKPN – or Kawasan Konservasi Perairan Nasional, a type of MPA managed by central government (MMAF) with the MPA area encompasses two or more administrative provincial areas Local MPA authority – A branch office of MMAF established in the MPA locality managed by MMAF to oversee the management of the MPA MCE – Multi-Criteria Evaluation Minapolitan – a concept of integrated regional development to achieve an efficient and sustainable development of fish capture and fish culture industry MPA core zone – a no-take zone within an MPA where access and resource extraction are restricted and only research and conservation education-related activities are allowed to take place MPA Sustainable fisheries zone – a zone within an MPA dedicated for sustainable fishery and fish culture MPA Other zone – a zone within an MPA with limited economic use determined at the initial designation of the MPA NTU – Nephelometric Turbidity Unit, the surrogate value of total suspended solid (TSS) Pemekaran/ – the establishment of two or more autonomous regions (province, district, sub-district and village) from one region to improve public service quality PSSF – Parameter-specific suitability function, the classification of parameter values into a suitability index using arbitrary scale between 0 and 1, where 0 denotes a non- suitable habitat, while 1 a habitat most suitable Rekomendasi Pembudiayaan Ikan Penanaman Modal or RPIPM) – a type of formal recommendation letter issued by the DGA for large scale aquaculture activity funded by a foreign investment Surat Ijin Usaha Perikanan (SIUP) – a type of licensing permit for both fisheries a and fish culture entity issued by corresponding directorate general under MMAF to conduct the fishery businesses Toke – the Chinese alias name given to middlemen, who are mostly Chinese descents, within patronage network of small-scale fishing and fish culture. UNCLOS – United Nations Convention on the Law of the Sea UTM Zone – Universal transverse Mercator Zone, a type of conformal projection using a 2-dimensional Cartesian coordinate system which divided the Earth surface into 60 zones

xvi

ABSTRACT

Small-scale mariculture has been formally recognised as one of the main strategic economic developments for small-island communities. It serves as a trade-off in the designation of multi-use Marine Protected Areas (MPAs) in developing countries including Indonesia. With the rapid expansion of MPAs in Indonesia’s small islands, more fishermen could turn to mariculture due to its lucrative economic potential. However, mariculture, especially for carnivorous fish species, has sustainability issues that threaten MPAs as a protected ecosystem and mariculture as an important livelihood of marginalised small-island communities. Mariculture also occurs where resources are used for tourism and capture fisheries, which can lead to conflict or competition and complicate decision-making processes for MPA management. Decision-making at all levels is affected by a lack of adequate regulation, unclear or poorly developed site selection criteria and unknown carrying capacity (CC) of MPA mariculture zones. The overarching aims of this study were to assess the sustainability of small-scale mariculture in multi-use MPAs as an economic livelihood from its legal framework and livelihood sustainability and vulnerability, and to develop an integrated site selection and CC framework for sustainable MPA mariculture zones.

In order to achieve the aims, this study employed a two-stage research approach. The first stage focussed on describing the legal framework, and investigating the sustainability and vulnerability of mariculture within the context of Indonesian MPAs. Local stakeholders’ perceptions regarding the establishment of an MPA in relation to their livelihood were also studied as part of the vulnerability studies. This part was designed to provide scientific evidence whether mariculture can be legally conducted, support livelihood sustainability and reduce vulnerability risk of local coastal communities in small-island MPAs in Indonesia. A legal framework, using David Gil’s policy analysis, was used to uncover the vague yet redundant and incomplete legal regulatory framework governing mariculture development in Indonesia’s MPAs. The sustainability and vulnerability profiles of the existing mariculture activities were also studied using Livelihood Sustainability Analysis (SLA). The sustainability study developed a set of sustainable livelihood indexes based on five capital assets to compare the sustainability profile of small-scale fish farmers, small-scale fishermen and ecotourism livelihood groups. The vulnerability risks of these local livelihood groups were also compared by developing a specific livelihood vulnerability index. The perception of the livelihood groups regarding the effect of MPA establishment to their livelihoods was determined using descriptive statistics (frequency and thematic analysis).

The second stage involved establishing a site selection and carrying capacity framework to ensure that mariculture in Indonesia’s MPAs can be carried out economically and are environmentally sustainable. A GIS and remote sensing site selection framework, based on parameter-specific suitability functions (PSSFs), was employed to design mariculture zones within MPAs. The use of PSSFs within the

xvii geometric overlay process was necessary to allow the combination of the hard and soft datasets in the site suitability framework. The carrying capacity of resulting different classes of mariculture zones was then determined using the Modelling–Ongrowing fish farm–Monitoring (MOM) system. This mathematical model of carrying capacity (CC) was modified to suit the tropical condition and mariculture practices (seasonal weather, feed, and net cage specification) where the study was conducted. The results of these CC estimates were then compared to the existing holding density of mariculture activity in the area and the general CC stipulated by the Indonesian MPA regulations.

This study uncovered the complex yet imprecise regulations governing the development of mariculture in Indonesia’s MPAs. This created an opportunity for the expansion of medium-scale mariculture in MPAs that outcompetes small-scale mariculture. The designation of mariculture zones for small-scale mariculture, local government and community approval-based permits, and third-party Environmental Impact Assessment (EIA) supervision, could substantially improve the sustainability of small-scale mariculture in MPAs and curb the expansion of medium-scale mariculture. Furthermore, small-scale mariculture as a livelihood has similar sustainability and vulnerability profiles to small-scale and ecotourism households, which tend to be at the intermediate and medium level, respectively. Small-scale fish farmers have a better chance to improve their sustainability profile through capital assets and endowment support from the government as received by other household groups. Such support could directly reduce the vulnerability risk of small-scale fish farmers. The livelihood flexibility developed by small-scale fish farmers also plays an important role in reducing their vulnerability. In addition, small-scale fish farmers also shared similar views with other household groups, for example, that MPA establishment is important to maintain the function of the ecosystem that their livelihood depends upon.

Site selection using PSSFs with a geometric mean indicated that only 10% of the total area was suitable for mariculture zones, of which 30% was classified as the best class and 40% as good, for small-scale mariculture. The introduction of a stakeholder preferences sub-model and MPA constraints successfully maintained the integrity of MPA core zones and allowed better location access for small-scale fish farmers to preferred mariculture zones. In addition, the seasonal difference had minimal effect on the site suitability classification of mariculture zones, though the effect of seasonal weather could be different in other MPAs. The use of the Modelling–Ongrowing fish farm–Monitoring (MOM) system revealed that oxygen concentration was the main factor influencing CC, which corresponded positively to site suitability classes. Location 2, classified as the best suitability class, had the best CC of up to 158 kg/m3, while the lowest was location 1 (medium class) of 7.14 kg/m3, depending on seasonality, feed composition and geographic location scenarios. The MOM analysis showed a much greater value of CC for the study sites compared to both the CC limitation set by the MPA regulations and the existing holding density currently practised in the area. This indicated that the general CC limitation of 50% in MPAs has been set too high and needs to be reconsidered for a fair spatial resource use. This study

xviii suggests that the CC limit in each MPA should be determined case-by-case and agreed upon by the involved stakeholders. The site selection framework developed in this study and the MOM system could provide an efficient and easy assessment to establish mariculture zones in MPAs and utilise resources fairly and spatially allocate them for small-scale fish farmers. Future work to improve or complement the study could be in the form of comparing the site selection and CC for two or more MPAs that have different seasonal weather characteristics as well as mariculture practices.

xix

CHAPTER 1. Introductory Literature Review

1.1 Introduction

The world has seen a rapid increase in conservation areas during the past three decades. There are now more than 157,897 nationally and internationally protected areas, increasing almost fivefold since 1980 and covering approximately 24.2 million km2 of terrestrial and marine areas (IUCN and UNEP, 2013). Marine protected areas (MPAs) account for 8.1 million km2 or 33.4% (IUCN and UNEP, 2013). In the Coral Triangle (CT) region of six developing countries, there are currently 1,972 MPAs covering 2.5% (White et al., 2014) of the total MPA worldwide. The numbers will likely increase as the region holds a third of the global coral reef cover and three quarters of coral reef species (Asian Development Bank, 2014, White and Green, 2014). The region is also home to 130 million people benefiting from the coastal and marine resources (Walton et al., 2014, White et al., 2014). It is estimated that by the year of 2020, less than 3% of the total world seascape will be protected under MPA regime, from the 10% agreed upon Target 11 of the Convention on Biological Diversity (CBD) (Abdulla et al., 2013, Spalding et al., 2013). There are numerous definitions for an MPA, although the most cited is from the IUCN: “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 the entire enclosed environment” (Kelleher and Kenchington, 1991, Alder, 1996). Most countries or conservation institutions have developed their own definition for MPAs to suit their conservation efforts or types of habitat they are trying to protect. Some have even designated patches of seascape as an MPA to protect one or more specific endemic species. Although, this kind of approach is considered inadequate to ensure the protection of the whole bio-ecological processes that support the species (Baker, 2000). MPAs also vary in size from less than 0.01 km2, as found in the Philippines (Halpern, 2003), to as wide as 1 million km2 for the South Georgia and South Sandwich Islands MPA in the UK (Spalding et al., 2013). The debates over the different levels of protection and sizes of MPA, as well as the outcomes of each, have been discussed extensively (Sanderson et al., 2002, Halpern, 2003, Donner and Potere, 2007, Redford et al., 2008, Spalding et al., 2013, Green et al., 2014). Regarding MPA size, developed nations and international conservation organisations tend to advocate more and 1 expansion of MPAs, whilst developing countries are likely to consider multi-purpose MPAs. This difference is clearly seen by looking at the ownership of MPAs where 14 out of the 20 largest MPAs, that constitute 60% of total world’s MPAs, belong to developed countries (See Spalding et. al., 2013). On the contrary, developing countries, where the majority of coral reef coverage can be found (Donner and Potere, 2007, Ban et al., 2011), have numerous smaller MPAs, mostly in the form of multi-use MPAs. Indonesia, the country in which this PhD research is based, also follows the same pattern as other developing countries. The classification of Indonesian MPAs is based on the regulation of the Ministry of Marine Affairs and Fisheries (MMAF), PER.02/MEN/2009, regarding the Procedure for Determination of Marine Conservation Areas. This regulation provides a technical description as mandated by the Government Regulation No. 60, 2007 on Conservation of Fisheries Resources. Both regulations recognize only 4 types of MPAs: marine national park, marine natural preservation/marine wildlife reserve, marine nature recreational park, and marine sanctuary/fishing refuge (Wiadnya et al., 2011, WGMFDA-MMAF, 2011). Marine national and marine nature recreational parks are categorized as multi-use MPA while the other two are specifically for conservation. There are similarities with IUCN MPA classification, which are National Park (Category II), strict nature reserve (category 1), protected seascape (Category V), and wildlife reserve (Category IV), for the Indonesian MPA classification order (Wiadnya et al., 2011). From 2009 to 2013, there has been an increased coverage of MPAs, especially MPAs that allow multiple use zones within their boundaries. As much as 110,089 km2 of territorial seascape have been newly designated or approved as MPA by the MMAF alone, of which around 96% are designated as multiple use MPAs (DKKJI, 2013a). This number has almost tripled the total coverage of MPAs in Indonesia before 2009, when the Ministry of Forestry was the only authority with power to manage MPAs. Only 4% of the total MPAs in Indonesia are classified as exclusive no-take MPAs, which is a mere increase of 0.2% from the 2009 figure. In total, the latest and consistent number of MPAs in Indonesia is 108 covering 157,841 km2 which will increase in coverage if the marine areas of the 70 terrestrial protected areas are included (White et al., 2014). This study focuses on the sustainable use of multiple use zones in Indonesia’s MPAs for mariculture, to support communities settled in or near MPAs that use small- scale mariculture as part of their livelihood strategy. While ecotourism and small-scale

2 fisheries have been advocated, introduced and managed to support communities locally- affected by MPA establishment, mariculture as another potentially sustainable livelihood is often overlooked (Pomeroy et al., 2006). Where mariculture is allowed, the designation of mariculture zones within MPA multiple use zones is not based on site suitability for mariculture. Instead, it is predominantly based on management cost and connectivity with other MPA zones, as well as formal agreements between MPA stakeholders (Watts et al., 2009). Consequently, any mariculture activities in these zones might fail or pollute nearby protected MPA areas. It is to be noted that the term mariculture used throughout this thesis is defined as “cultivation, management and harvesting of marine organisms in the sea, in specially constructed rearing facilities e.g. cages, pens and long-lines” (FAO, 2008).

1.2 MPAs and sustainable development

Integrating MPAs with sustainable economic activities has been considered as important as protecting the biodiversity itself (Adams et al., 2004). According to Fountabert et al. (1999), “MPAs are coastal or oceanic management areas designed to conserve ecosystems together with their functions and resources…as such, they should not be exclusionary ecological zones, but should recognize the importance of traditional sustainable uses of local communities therein” (p.15). In fact, the importance of integrating economic development, in the framework of sustainable development, in protecting biodiversity has been included as one of the Millennium Development Goals by the United Nations (Adams et al., 2004, Redford et al., 2008). Despite its importance, disadvantages of incorporating economic development into conservation efforts have also been voiced by a number of scholars, such as Redford et al. (2008). They argue that, globally, only a small fraction of empoverished populations live near areas that need to be protected. Thus, the involvement of conservation organizations to help these populations might undermine their overall objectives to protect the biodiversity (Redford et al., 2008). Adding a level of validity is the fact that only a few cases of conservation management, combined with poverty alleviation, have had successful or positive results (Adams et al., 2004). However, looking at the local level, many protected areas with high biodiversity also have a high concentration of people, particularly in small islands areas of developing countries. For example, Wakatobi National Park, a marine conservation area in Indonesia, is home to more than 92,995 people (BPS, 2010), or Calamianes Islands 3

MPA in the Philippines hosts more than 71,000 people (Fabinyi, 2008), both with a high dependence on the marine resources around them. In MPAs located around small islands where human populations are absent or scarcely distributed, exclusive conservation effort might be the best scenarios, as it was shown by the conservation effort in the Great Barrier Reef on the east coast of Australia (Ban et al., 2011). Hence, MPAs enclosing inhabited small islands, which are common in developing countries, should be a special case involving some form of sustainable economic development along with conservation efforts. This is important for several reasons. Firstly, it potentially reduces further conflict between MPA managers and local communities due to the limitation of access and rights to use space and resources in the area (Agardy, 1994, Fabinyi, 2008, Dung, 2009, Maynard et al., 2010, Ban et al., 2011). Secondly, it provides an alternative livelihood for local communities in small island MPAs (Bottema and Bush, 2012), which in turn will reduce pressure on the biodiversity (Fabinyi, 2008). Thirdly, it decreases the economic loss and social cost borne by the local communities from the designation of an MPA. Lastly, it demonstrates social accountability of conservation entities (Adams et al., 2004).

1.3 Mariculture and other sustainable livelihoods in MPA

Indonesia’s MMAF has set a target, which was also one of Indonesia’s pledges at CBD 2006 in Brazil, to have 20 million ha of seascape managed under an MPA regime by 2020 (DKKJI, 2013a). However, the effectiveness of the current MPA establishment and management to protect natural resources remain the central issues for the Indonesian Government (Polunin et al., 1983, Fauzi and Buchary, 2002). Nevertheless, it means that more small islands will fall under the MPA management regime. The establishment of MPAs in coastal and marine areas where less or no economic or land use activities take place, usually does not encounter complex problems, and conflicts over land use have mostly been resolved (Ban et al., 2011). However, this is not the case with MPAs established around inhabited small archipelagic islands where MPA zoning reduces access for island communities to coastal resources (Adams et al., 2004, Szuster and Albasri, 2010).

Different approaches have been introduced to accommodate the needs of many coastal and marine users. Terminologies such as sustainability, integrated management, and adaptive management are being used with the main aim of reducing potential

4 conflicts over the spatial use of those areas (Agardy, 1994, Fabinyi, 2008, Dung, 2009, Maynard et al., 2010, Ban et al., 2011). Such approaches have resulted in multi- economic activities being allowed in existing or newly designated MPAs; for example, ecotourism (Walpole et al., 2001), small-scale fisheries (Fountabert et al., 1999, Fabinyi, 2008), and aquaculture (Dung, 2009, Szuster and Albasri, 2010). From these three economic activities, tourism has received more attention regarding its benefits and impacts on protected ecosystems and surrounding human populations. Ecotourism is perceived to have less negative impacts on the environment while at the same time providing a sustainable revenue for local communities (Bottema and Bush, 2012). Small-scale fisheries are also considered a good alternative livelihood for affected communities because they do not have to switch and adapt to new economic activities, while at the same time have the chance to protect their own environment (Fabinyi, 2008). For mariculture, its development in an MPA to help locally affected communities remains doubted and slow (Pomeroy et al., 2006)

1.3.1 Sustainable ecotourism

Tourism-based activities have been almost a standardised approach in the management strategy of MPAs in Indonesia. Tourism is intended to provide an alternative livelihood for locally-affected communities (Bottema and Bush, 2012), and to consider the possibility of a development-conservation strategy in MPAs (Cochrane, 2006). The benefits of tourism are clearly described in many studies, such as generating income to support the management of the conservation areas (Walpole et al., 2001), and support local economic growth through the interaction between tourists and local communities (Walpole et al., 2001, Cochrane, 2006). Ecotourism also has the potential to attract more support from local communities as well as from private entities as described by Bottema and Bush (2012). However, tourism-based activities alone are not sufficient to support local economies since the benefits of ecotourism are not proportionally shared to local communities. In one study, Walpole et al. (2001) states that only 40% of the total expenditure of tourists contributed to the local economy, and only in the form of supporting tourism transportation. However, this figure is likely inflated, as the use of motorized boats, often owned by middle and upper-class, to transport tourists around was not considered by the authors. In another study, Lee (1999), as cited in Christie (2005), found less than a quarter of jobs generated from MPA-based tourism went to 5 local people. As a result, most local people, who are considered impoverished, receives less benefit from ecotourism. In terms of supporting conservation efforts, ecotourism also has drawbacks. Tourists can potentially degrade MPA environments through littering and physically damaging fragile ecosystems (Cochrane, 2006). Tourists also contribute to biodiversity disturbances through high visitation frequencies and increased conflicts of interest among MPA stakeholders (Fabinyi, 2008). Therefore, it is important to diversify economic activities in MPAs to avoid one economic activity dominating, and reduce the pressure on marine resources (Allison and Ellis, 2001).

1.3.2 Small-scale fisheries

Fisheries have been one of the main economic activities for small island communities in Indonesia. It supports more than 2.27 million fishermen or as much as 6.07 million people whose work relates, directly and indirectly, to fisheries (MMAF, 2011). In Indonesian MPAs, fishing activities are still allowed, providing the activities are conducted by local communities in a ‘friendly manner’, within MPA sustainable fisheries zones. This ‘friendly manner’ has been interpreted as small-scale or artisanal fishing in numerous studies regarding fishing operations in MPAs. Based on the MMAF regulation No. 30/2010, regarding MPA zonation, ‘friendly manner’ is defined as fishing using static or passive fishing gears, and considers the carrying capacity (CC) of the resources. Fishing is not permitted in other MPA zones, i.e. core, common use, and other zones. However, the fishing zoning system in Indonesian MPAs, in which local communities are permitted to fish, has been controversial. Firstly, most of the resource- rich areas that were previous fishing grounds for local communities have been claimed as core or no-take zones (Christie, 2004). Despite the promising results of the no-take zones for protecting biodiversity, such as higher fish abundances, spillover effects and seed production, local communities have been treated unfairly from losing the right to access these rich areas (Christie, 2004). Secondly, the long-term management of MPAs in Indonesia has been difficult to maintain due to limited human and funding resources (Christie, 2005). These have left many MPAs in Indonesia unprotected years after their establishment. As a result, non- local fishermen freely operated in sustainable fishery zones reserved for local MPA communities and in most cases outcompete them (McLeod et al., 2009). The increased use and ownership of inboard motorised boats from 2007 to 2011 (MMAF, 2011) has 6 also created more challenges for managing fisheries activities within the fisheries sustainable zone. Motorised boats significantly increase the efficiency and productivity of fishing activities (Clifton, 2013), thus fishing has become more extractive in nature. Lastly, population growth in small island communities in Indonesia has been steadily increasing and, in some cases, faster than the national growth rate. Despite some scholars arguing that population growth should not be assumed to have caused fish stock depletion (Bottema and Bush, 2012, Clifton, 2013), fish catch needed to increase since fish are the main staple of coastal and small island communities (Fauzi and Buchary, 2002).

1.3.3 Mariculture in MPA

Mariculture within MPAs is somewhat different to ecotourism and fisheries. It is a relatively recent activity, compared to more traditional fishing-based livelihoods, that is growing as means of meeting local community needs. Mariculture has been incorporated into most of the MPA management strategies in Indonesia due to the concern that fish exploitation rates, through fishing, is faster than the ability of fish to recover (Pomeroy et al., 2006). Mariculture also supports local communities affected by the establishment of an MPA, reduces the pressure on coastal and marine resources, and diversifies economic activities in the area (Agardy, 1994). The recognition of mariculture as an important alternative livelihood for affected coastal communities has been gaining support by international conservation agencies such as The Nature Conservancy (TNC) and Conservation International (CI). Both TNC and CI have incorporated mariculture into their conservation mechanisms as an alternative economic activity for local MPA communities (TNC and CI, 2012). In other developing countries, mariculture zones have also been designated and applied in a number of MPAs, such as in Vietnam as reported by Dung (2009), and in Indonesia as reported by TNC (2004). In Vietnam, the introduction of mariculture in Hon Mun MPA resulted in several serious setbacks to conservation efforts. For example, the increased use of wild lobster seed and trash fish, as feed, to supply lobster farmers within the MPA has put more pressure on the protected areas. Management also failed to control the development of large lobster net-cages, clustered in several locations, that degraded the water conditions of the surrounding MPA (Tung, 2002). Although CC and site selection processes were planned, no information was available on whether they were implemented. Consequently, the potential for impacts was not 7 addressed. The zoning regulation was also not enforced, resulting in the uncontrolled placement of mariculture platforms in areas not designated for mariculture (Dung, 2009), thus they were not considered sustainable (Dung, 2009). In Indonesia, most of the multi-use MPAs have accommodated mariculture in the zoning system. In the MMAF regulation No.30/MEN/2010, regarding MPA zonation, mariculture occupies the same sustainable fisheries zone as fishing activities. The levels of operation of both activities are required to be reduced to a small scale. However, the government regulation provides a vague description regarding the term small-scale mariculture. The regulation only states that fish farmers should use appropriate cultured fish species, types of feed, culture technology, a certain number of farm units and consider the CC of the MPA area. The absence of clear management regulations for mariculture in MPAs, from the central government to local governments, means that mariculture development is left unmonitored. It also opens the possibility of larger scale mariculture (i.e. medium scale) to operate in the area. This, in turn, will make mariculture more susceptible to further criticisms that question its sustainability (Ross et al., 2008, De Silva, 2012). Only a handful of local regulations are in place to address mariculture in MPAs, mainly in issuing special permits for mariculture, such as in Komodo National Park (KNP) (TNC et al., 2003). In most cases, local regulations that are expected to carry technical arrangements resemble those of the higher regulation. Typically, mariculture activities conducted in small island communities in Indonesia are closed-culture systems, although the use of half-cycle culture is increasing. Half-cycle culture is mariculture that uses imported fish seeds or juveniles from non-local hatcheries. Closed-culture is using fish seeds and juveniles caught from the wild, particularly from the area, to be grown to marketable size (Pomeroy et al., 2006). A full-cycle culture system, which uses fish seeds or juveniles produced from hatcheries in the same area, is considered too difficult and expensive for small-scale mariculture farmers (Pomeroy et al., 2006). Regarded as one of the most important steps, the selection of area within Indonesian MPAs as mariculture zones has not been regulated, or even standardized. However, the use of MPA areas for mariculture is common, an arrangement between the prominent stakeholders in the area to provide alternative space and livelihoods for local communities. For example, Karimun Jawa Marine National Park (KJMNP), KNP,

8 and Raja Ampat Marine National Park (RAMNP), have several areas in sustainable fisheries zones serving as mariculture zones (Figure 1.1). Unfortunately, there is no clear evidence that MPA managers have included a process to determine the suitability of the areas based on its capability and CC for mariculture. For example, in KJMNP, the mariculture zone is less than 2 km from a core zone, likely too close considering the possible biological and chemical pollution from future mariculture activities there. Another example of uninformed zoning for mariculture in MPAs is shown in WNP. Here, zonation was created using MARXAN software, which is based on algorithms that measure the cost of management and connectivity of parcel areas under consideration (Watts et al., 2009). Although this type of zoning procedure is beneficial for creating a relatively general designation of areas for specific uses in MPAs, it is unable to determine the actual capability of a particular area of the seascape for specific objectives, such as mariculture.

Figure 1.1. Establishment of mariculture zones near the vicinity of core and protection zones (less than 2 km) in KJMNP, Indonesia (Maynard et al., 2010)

It is important to consider that introducing mariculture into an MPA requires specific requirements, that differ from those of other mariculture practices in other commonly used coastal and marine areas. Without adequate site selection and 9 determination of CC of mariculture in an MPA, promoting sustainable mariculture as a potentially sustainable livelihood for affected local communities will likely fail. Additionally, the introduction of mariculture will increase impact-related pressures to the surrounding sensitive ecosystem of MPAs.

1.4 Determining sustainability of mariculture in MPA

1.4.1 Impacts of mariculture

A number of published works on the impacts of mariculture focus on the negative environmental impacts caused by effluent, largely in the form of fish excreta and uneaten fish feed (Silva et al., Iwama, 1991, Alongi et al., 2009), and dispersion of chemical products to maintain cultured fish health (Primavera, 2006, De Silva, 2012). There are also concerns over genetic exchange between cultured species (Volpe et al., 2000, De Silva, 2012), reduced visual amenity (Falconer et al., 2013), influences on wild stock, and fish disease pathogenesis (Primavera, 2006, Bondad-Reantaso et al., 2005). Such impacts might hinder the overall benefit of mariculture in MPAs if they are not properly addressed. To be considered sustainable in MPAs, waste dispersion from mariculture platforms, both suspended and settled waste materials, should deemed insignificant (cause negligible change to biodiversity). There are numerous studies on the prediction and effects of waste dispersion from mariculture platforms (Corner et al., 2006, Alongi et al., 2009). However, most of these studies, and prediction modelling, are for commonly used coastal and marine areas and intensive mariculture systems in temperate regions (Alongi et al., 2009). In fact, there is virtually no information regarding the impacts of mariculture waste in MPAs, despite mariculture occurring in multiple use zone contexts, in countries such as Vietnam and Indonesia. Furthermore, the appearance of wild fish near mariculture platforms in tropical regions has not received sufficient attention in regards to CC and the dispersion of uneaten feed. Hevia et al. (1996) have expressed their concern on this subject because data on the effects of wild fish, attracted to the mariculture activities, is needed to determine both the local and wider geographical extent of CC of mariculture. This type of assessment is crucial to justify establishing mariculture and to determine CC of the surrounding environment (Inglis et al., 2000, Geček and Legović, 2010) in MPAs. One particular study on fish aggregation near net cages in a semi-enclosed tropical bay

10 provided information on fish behaviour near net cages, and their role in waste feed dispersion (Sudirman et al., 2009). Sudirman et al. (2009) found that wild fish concentration under and near net cages were able to consume 27% of the total uneaten feed. However, the authors conducted this research in a single location and water system, and there were no comparison of results from other similar or different water system or areas. Although the research could be accurate for that particulate water system (bay area), the results might be different for open water systems such as islets and open coastal areas, or more diverse ecosystems such as MPAs. Other critics of mariculture also raise concerns about the possibility of disease transmission from cultured fish to wild populations. Similarly, the risk of genetic diversity loss through interbreeding between less genetically diverse cultured fish with more genetically diverse wild stock is often raised. Disease risks may increase through importing infected fingerlings or seed, increased viral and bacterial loads in overstocked cages, or via fishmeal used in fish feeds (Rückert et al., 2009, Gomez et al., 2010). The current MPA management strategy in Indonesia includes the prevention of fish diseases, distributed from seeds imported from outside the MPA areas, and may be included in the newly drafted MMAF regulation, regarding procedures and permits for fishing and fish culture in MPAs (Sudarsono 2014, personal, communication, 3rd April). Aside from the fact that it is important to acknowledge the risk of these types of impacts, the benefits of mariculture in MPAs, for the affected communities, as well as to the biodiversity, should also be considered. Through best mariculture practices, the benefits should support the inclusion of mariculture in the Indonesian MPA strategy, as stated by De Silva (2012), that “…the positive impacts of grouper mariculture, whilst meeting the demand for a much sought after food commodity, far outweighs the negative impacts on biodiversity conservation and indeed, overall, indirectly aid in conservation of fragile ecosystems and biodiversity thereof” (p.3209).

1.4.2 Site selection and carrying capacity: a safety net

Site selection and CC have been unanimously considered as the main pillars in mariculture management. The Ecosystem Approach to Aquaculture (EAA), in the FAO Code of Conduct for Responsible Fisheries (CCRF), requires any mariculture development to consider the pros and cons, in terms of socio-economic, environmental and multi-sectoral goals and values, at the beginning (Ross et al., 2013b). Furthermore, since the adoption of CCRF is voluntary (FAO, 1995), countries have interpreted the 11 guidelines differently, according to their own interest in developing mariculture (Ross et al., 2013a). As a result, site selection and CC have also been interpreted, developed and applied differently amongst countries. Developed countries lean toward site selection and CC for better ecological and social management of mariculture, whilst less developed countries focus more on production (Ross et al., 2013b). Within the mariculture site selection concept, suitable site and potential area are two related aspects. Suitable site is referred to the exact location where mariculture can be carried out based on a set of a specific site selection criteria while potential area means an area where mariculture can be generally carried out (Kapetsky and Aguilar- Manjarrez, 2013). Suitable sites are usually located within a potential site where their distribution and management are controlled in different mariculture zoning areas for efficient and effective monitoring and evaluation (Kapetsky and Aguilar-Manjarrez, 2013). Within these areas, different CCs have to be determined and monitored regularly in order to ensure the sustainability of any mariculture activity. Inglis et al. (2000) developed, McKindsey et al. (2006) and Ross et al. (2013a) had identified different CCs and grouped them into four types: physical, production, ecological, and social carrying capacity. The definition of each type has been clearly described by the authors and others. Despite many research studies conducted on site selection and CC for aquaculture, there are at least three considerations that might be inadvertently neglected, which are going to be further studied under this Ph.D. research. First, site selection and CC are increasingly directed towards business-oriented aquaculture activities in inland and coastal areas. In coastal areas, site selection and CC are mostly conducted in order to issue business-based licenses (Ross et al., 2013b), or comply with the prevailing regulation of mariculture activities (Jiang and Gibbs, 2005). On the contrary, site selection and CC for small-scale mariculture has received little attention as some studies have found that small-scale mariculture does not have significant effects on the environment, such as Alongi et al. (2009), Mente et al. (2006), De Silva (2012). However, the typical development of small-scale mariculture, that tends to cluster in one location (Alongi et al., 2009), might create a cumulative effect of negative impacts in the surrounding waters. In addition, there is a tendency that it will undergo rapid expansion once other local community members see its success. This might result in increasing the total organic and inorganic effluent, spreading of fish diseases, scenic

12 disturbances and conflicting use of coastal space. In addition, regulation and licensing practices for mariculture are virtually absent at the field level, thus the coastal environment will be more likely to suffer from the negative impacts of mariculture. For example, small-scale inland aquaculture in Cirata Reservoir in West Java is almost unregulated, where the number of net cages skyrocketed, exceeding the lake’s carrying capacity by more than double (Mungkung et al., 2013). Catastrophic events, such as

H2S upwelling and fish kills in the reservoir, have increased over the years (Abery et al., 2005, Mungkung et al., 2013). This prompted a suggestion from Macbub (2018), that at least 85% of the existing cages should be removed from the reservoir. The second consideration is the generalisation of results of site selection and CC analysis for mariculture. Most of CC studies suggest a single value of CC (See Carver, 1990, Duarte et al., 2003, Jiang and Gibbs, 2005 and others), or a single final thematic suitability map (See Radiarta et al., 2008, Silva et al. 2012, Falconer et al., 2013). These site selection and CC results provide unanimous choice and easy guidance for interested parties to make an informed decision. However, at the same time, it contains unnecessary, yet easily identifiable, risks as the condition of the water body might change over a certain period. Coastal areas are globally interconnected, thus over the course of one year (the unit time commonly used to describe CC), there is a possibility that the CC value will change according to environmental conditions (McKindsey et al., 2006). For example, the CC of a particular area during the wet season might be different than that of the dry season. By disclosing this information, interested parties can make better decisions for the management of mariculture. This issue also relates to the third consideration, which is the role of decision makers/stakeholders in integrated site suitability and CC of mariculture. McKindsey et al. (2006) and Ross et al. (2013a) argued that site selection for mariculture has to apply the full set of the available CC models previously mentioned. This is needed to ensure that mariculture does not only satisfy site-specific environmental, economic and societal requirements, but also for wider areas such as regional, national and transnational. This kind of approach will involve a considerable number of variables that constitute the level of suitability and constraints limiting the suitability of an area, as well as costly expert human resources (McKindsey et al., 2006). This is probably the reason that most of the recent studies on site selection and CC for mariculture have only used one, or at the most three, combinations of CC

13 models (see Rachmansyah, 2004, Jiang and Gibbs, 2005, Radiarta and Saitoh, 2009, Geček and Legović, 2010, Byron et al., 2011). Such a widely used approach in CC research surely has valid grounds and is scientifically acceptable (Ross et al., 2013b). However, in a complex situation, such as for multiple use zones in MPAs, the full set of CC models might have to be used. The application of social carrying capacity is also limited (but see Byron et al., 2011, Whitmarsh and Palmieri, 2009). Whitmarsh and Palmieri (2009) used social CC, equipped with an Analytic Hierarchy Process (AHP), to measure local community preference regarding the development of mariculture in their areas. However, this study was carried out in a pre-existing mariculture operation, which has little influence in changing the current siting condition or for strategic MPA planning. The importance of incorporating social assessment into a CC model does not only constitute the acceptance of other users regarding a particular development plan (Byron et al. (2011); it also ensures that the development plan generates potential benefits for the community (Ross et al., 2013a)

1.4.3 Application of hard and soft system of thinking in site selection and carrying capacity The development of site selection and CC models for mariculture has mainly dealt with fish production and its resulting pollution (wastewater/effluent) (Ferreira et al., 2013). The use of Geographic Information Systems (GIS) and Remote Sensing (RS) to depict and draw conclusions from temporal/seasonal changes of suitability variables has been demonstrated by Sudarshana et al. (2013), Karthik et al. (2005), Corner et al. (2006), Longdill et al. (2008), Silva et al. (2012) and many others. Low cost, wider area, and multi-temporal coverage, as well as stunning visual representation of GIS and RS data, are the main reasons for their wide acceptance. However, the applications of GIS and RS to address several complex key parameters, such as hydrodynamic variation and nutrient dynamic, in mariculture site selection and CC have been limited (Ferreira et al., 2013). Hydrodynamic and nutrient modelling is needed to provide a complete picture of water quality and possible nutrient enrichment profiles in site selection CC estimation (Geček and Legović, 2010). The combination of hydrodynamic and nutrient modelling with GIS and RS would have been visually powerful and easy in aiding site selection and CC process. Instead, most of the work that involves hydrodynamics and nutrient dynamics focuses on the

14 production and impact of mariculture operations (Byron and Costa-Pierce, 2013, p.95; also see examples from Alongi et al., 2009, and McKinnon et al., 2010). A successful mariculture operation depends on a comprehensive siting and CC analysis, considering not only what is best for mariculture, but also its effect on other activities operating in the same space. Ross et al., (2013b) argues that mariculture can co-exist with other activities if stakeholders of a coastal area are allowed to take part in the planning process. This is where multi-criteria decision analysis (MCDA) is needed, to incorporate soft data (preferences) into the hard data (quantifiable measurement) that CC and site selection rely on. Radiarta et al., (2008) used MCDA for site selection using RS of ocean colour and GIS. This study is a good example of a successful combination of hard and soft data (literature and stakeholder preferences) analysis. However, a false participation of stakeholders is clearly seen in this study, where there was no active involvement of stakeholders in the analysis process. This has been addressed by Byron et al. (2011), who showed that hard data calculations often fail to actively engage, or communicate fairly, soft data from stakeholders into the equation, when planning mariculture. Based on the facts and available site selection and CC analyses mentioned above, the future direction of mariculture development in multi-use MPA zones in Indonesia faces several problems: a. There has never been any site selection or CC model that specifically addresses the introduction of mariculture into an MPA. b. Most of the site selection and CC studies for mariculture have been targeted to business and industry and not small-scale farming systems, that are usually found in small island MPAs. c. Integration of site selection and CC to determine mariculture zoning or site suitability has been partially addressed. In the case of mariculture in an MPA, the site selection criteria and CC assessment should be advocated as a standardized and combined approach. d. Active participation of stakeholders has been left out in most of the social CC analyses for site selection for mariculture zoning or site suitability. In the case of mariculture planning in an MPA, social CC coupled with full stakeholder involvement are the most important stage of the process.

1.5 Objectives and hypotheses

1.5.1 Overall objective and research questions

The overall aim of this research is to develop site selection criteria and decision processes for small-scale mariculture in or near marine protected areas based on integrated carrying capacity assessments.

The main deliverable expected from this research is a framework for site selection assessment for small-scale mariculture to be introduced within or near MPAs. To achieve the main aim and expected deliverable, several objectives and the underlying research questions for each objective are presented below:

1. To evaluate and describe policy guidelines and other legal instruments, as well as other elements of environmental decision-making frameworks, that are used in zone planning and establishment for mariculture zones in Indonesian MPAs. The following research questions are identified to achieve the abovementioned objective: a. Are the current decision-making processes, policies, and guidelines effective for sustainable mariculture zoning in Indonesian MPAs? b. What elements of these regulatory mechanisms can be improved? c. In what ways can those improvements be made so that decision-making processes are more efficient and effective? 2. To evaluate the sustainability profile of mariculture in MPAs in comparison to other livelihood activities. The following research questions will be answered to achieve and satisfy the abovementioned objective: a. How sustainable is small-scale mariculture compared to small-scale eco-tourism and small-scale fisheries in Indonesian MPAs? b. What is the level of vulnerability of households involved in mariculture compared to small-scale eco-tourism and small-scale fisheries in Indonesian MPAs? c. What are the different household groups perspectives of the MPA establishment and how has it affected their livelihood? 3. To develop a GIS-based site selection framework for the mariculture zone. The following research questions will be answered to achieve and satisfy the abovementioned aim:

a. How are the environmental, economic, and socio-cultural variables, and limiting factors of site selection for a mariculture zone in an MPA identified and quantified? b. How do these variables and limiting factors interact and influence decisions on site suitability for mariculture in or near an MPA? 4. To develop a framework of CC for mariculture zones in an MPA. The following research questions will be answered to achieve and satisfy the abovementioned aim: a. What MPA-related factors contribute to the CC of a mariculture zone within or near an MPA? b. What is the CC for small-and medium-scale mariculture zones within an MPA?

1.5.2 Hypotheses

The following hypotheses will be tested to satisfy the aims and objectives of this research: a. Small-scale mariculture within or near MPAs has economic and socio-cultural benefits for affected coastal or small island communities and supports conservation efforts in the protected area. b. Good site selection and CC will substantially reduce impacts of mariculture to the surrounding MPA zone and environment. c. Small-scale sustainable mariculture can be conducted within or near MPAs in Indonesia.

1.6 Research framework

This PhD research aims to investigate the applicability and sustainability of small-scale mariculture in multi-use zones in Indonesian MPAs by applying integrated site selection and CC assessment. To provide a thorough and step-by-step explanation of the research problem, process and eventual outputs, this thesis will be arranged into 8 chapters. Each chapter will be developed to build knowledge and answer the research objectives laid out in section 1.5.

Chapter One - The chapter provides a general overview of the development and management of MPAs worldwide, and then focuses on the Indonesian multi-use MPA situation. It will also describe the existing knowledge, and knowledge gaps, of site

17 selection and CC, and their relationships, to develop mariculture in multi-use zones in Indonesian MPAs. Chapter Two - This chapter describes the history, MPA development and management, climate, geographical layout, local and regional oceanographic forces, economic and socio-cultural conditions of the study area. This information will serve as the baseline information to highlight the importance of this study. Chapter Three - This chapter is centred on policy analysis of regulations regarding the development of mariculture within Indonesian MPAs. The current MPA regulatory framework, and possible changes to ensure sustainable mariculture within Indonesian MPAs, are discussed. Chapter Four - This chapter presents the analysis, findings, and discussion of mariculture sustainability, viewed from a sustainable livelihood analysis (SLA), in relation to the other economic activities in Indonesian MPAs. Chapter Five - This chapter covers the analysis, findings, and discussion of the level of vulnerability of small-scale mariculture household groups compared to other livelihood groups commonly found in Indonesian MPAs. The chapter also highlights the different perceptions of household groups regarding the establishment of an MPA and its effects on their livelihood. Chapter Six - This chapter develops a multiple-stage framework for mariculture site selection, specifically designed to suit MPA requirements, by integrating RS and GIS analyses. This chapter describes the selection of key variables and constraints (quantitative and qualitative) used in each stage. Several alternative site suitability schemes for mariculture zones in Indonesian MPAs are proposed. Chapter Seven - This chapter presents a CC assessment specifically designed for mariculture zones within an MPA. The analysis is determined using mathematical computations revolving around fish production and environmental CC. Chapter Eight - This chapter synthesises all the findings, results, and discussions from previous chapter to build a framework of site selection and CC of small-scale mariculture in MPAs. This chapter also provides a summary of the overall research, recommendations, and potential future research.

CHAPTER 2. Description of Study Area

This chapter provides information of the study area in terms of its geographic and administrative features, history, climate, local and regional oceanographic forces, economic and socio-cultural conditions, and MPA development and management. This information serves as a baseline of knowledge to highlight the importance of this study and underpin discussion and contexts later in the thesis.

2.1 Geographic location, history, and administrative features

Anambas Archipelago is situated in the middle of South China Sea, between Singapore, Indonesia, Malaysia, and Vietnam. It lies south of the Gulf of Thailand basin, which is part of the 1,850,000 km2 Sunda Shelf (Shi-guo et al., 1999). This study area and the rest of Sunda Shelf play a significant role in influencing the equatorial winds and sea currents, thus making it an important site for seasonal climate and oceanography studies (Shi-guo et al., 1999). The geographic extent of Anambas Archipelago covers between 2º10’0”- 3º40’0” N and 105º15’0”-106º45’0” E. It is one of the frontier areas of Indonesia bordering the Vietnamese EEZ on the North and the Malaysian EEZ on the West and East. It is closer to Malaysian (201 km) or Singaporean territorial land (215 km) compared to the nearest Indonesian land mass (221 miles from Sumatra). However, Indonesia has benefited from the UNCLOS 200-mile EEZ regime, by which Anambas, as well as Natuna archipelagos, are within the Indonesian EEZ. The total area of Anambas Archipelago is about 46,664.15 km2, of which only 1.3% is territorial land and the rest is territorial sea (Saidah, 2011). The small percentage of Anambas territorial land consists of five small main islands and 250 very small islands. The history of Anambas Archipelago District can be traced back to the Dutch Colonial period, when it was one of the central local governments called “Kewedanaan”. It was called Kewedanaan Pulau Tujuh with Tarempa (now the capital city of Anambas Archipelago District) as the main administrative city (BPS-Anambas, 2012). In 1950, Anambas Archipelago, which was under the Riau Archipelagic District, was annexed to Riau Province, and subsequently all “Kewedanaan” in Riau Archipelago District were erased in 1966 by the Governor Riau Decree, No. UP/247/5/1965, on 9 August 1964. This change in administrative structure had to be done to follow the

Indonesian Presidential Regulation #22/1963, regarding the Elimination of Keresidenan and Kewedanaan, and to give rise to the new national and local government structures. After 1 July 2004, Riau Archipelago was detached or “dimekarkan” from Riau Province to be one of 32 Indonesian provinces, where Anambas Archipelago was established as one of 17 sub-districts, under Natuna Archipelago District. The 200 km distance separating these two clusters of islands has proven to be one of the key factors for slower development in Anambas Archipelago. Unsurprisingly, four years after Riau Archipelago Province was established, Anambas Archipelago was successfully elevated to a new district, separating from Natuna District, through the enactment of Law Number 33/2008, Regarding the Formation of Anambas Archipelago District. Administratively, Anambas Archipelago District has 34 villages or “Desa” distributed across seven sub-districts or “Kecamatan”. A distinctive feature of the formation of sub-districts in Anambas is that their capital cities are all located on the Anambas main islands. This arrangement has led to slower development rates on the smaller islands, as most of the government funds are used in the sub-district capital cities. In addition, distances between islands and main administrative cities are relatively far, and the only mode of transportation available is boats or small ships. Transportation by small plane is available, but only for travelling from the district to the provincial capital city.

2.2 Climate, local and regional oceanographic forces

Anambas Archipelago is located almost in the centre of the Sunda Shelf, once a huge plain that connected all Greater Sunda Islands, Malay Peninsula, Borneo, and Indochina (Hanebuth et al., 2011). The whole region is considered as the most complex geographical area in the world (Wyrtki, 1961). The main climatic systems regulating the Southeast Asian monsoonal season, including Anambas Archipelago, are the northeast (winter) monsoon, from November to March, and the Southwest (summer) monsoon, from late May to September, with October as a transitional month (Loo et al., 2014). Because of its geographic location between two large land masses (Asia and Australia), and relatively constant air pressure, wind speeds over this region are relatively low, although some extreme fluctuations can reach 5 Beufort’s Scale (<10.5 m/s) (Wyrtki, 1961). A six-year compilation (2008 – 2013) of monthly wind speed in Anambas Archipelago (Figure 2.1) fits with Wrytki (1961) models, where wind speeds have been steadily below 5 Beufort’s Scale. At the local scale, wind directions vary annually, 20 blowing northerly from January to March, easterly from April to June, south-easterly from July to September, and westerly from October to December (Saidah, 2011). Such wind variations have affected anthropogenic activities, especially related to fisheries in the area. For example, small-scale fishing communities living on the western side of the archipelago islands must limit small boat fishing activities during relatively strong westerly winds.

2008 2009 8 2010 2011 7 2012 2013

3 Wind SpeedWind(m/s) 2

0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Figure 2.1. Monthly average wind speed in Anambas Archipelago during 6 years (BMKG-Tarempa, 2014)

Due to the relatively constant wind speed and characteristics of the climate system, rainfall in Anambas Archipelago is categorised as strongly seasonal, depending on the prevailing monsoonal seasons (Figure 2.2). The wet season lasts from October to January when the winter monsoon carries wet air masses from the northern hemisphere (Wyrtki, 1961). Although a real dry season is not obvious from rainfall data (Wyrtki, 1961; BMKG-Tarempa, 2014), less than 60 mm/month rainfall, which is considered as dry season, occurs predominantly from March to May (Saidah, 2011) (Figure 2.2). This is the effect of the prevailing summer monsoon carrying dry air from the southern hemisphere.

700 Jan

600 Feb

500 Mar Apr 400 May Jun 300 Jul 200 Aug

Averaged monthly rainfall Averagedmonthly rainfall (mm) Sep 100 Oct

0 Nov 2008 2009 2010 2011 2013

Figure 2.2. Monthly average rainfall in Anambas Archipelago during 6 years period (BMKG-Tarempa, 2014)

Surrounded by a large seawater body, annual temperature variation in Anambas Archipelago is influenced by heat transfer from solar radiation to the water mass, as well as the dry air mass coming from southern hemisphere during the summer monsoon. Typically, the temperature ranges from 28.5 OC to 30.5 OC during the winter monsoon, the peak wet season, and from 32 OC to 33.9 OC during the summer monsoon (BMKG- Tarempa, 2014). As part of the Sunda Shelf, the waters around Anambas Archipelago are characterised as shallow, ranging from 35 m to 100 m deep (Anambas and IPB, 2011). This relatively flat seabed terrain subsided at the centre to form the U-through (basin) of the Gulf of Thailand Basin, creating a low energy environment. As a result, most of the sediments found in this area are bioclasts (planktonic and benthic foraminifers as well as bivalves, gastropods, and bryozoans) (Shi-guo et al., 1999). The relatively small amount of terrigenous sediments (<15%) in this area, despite the proximity to the outlet of the Mekong and Chao Phraya Rivers (Shi-guo et al., 1999), results in a low concentration (0.2 – 5.51 mg/L) of phosphate (MMAF, 2014d). However, due to relatively shallow water in this area, continuous mixing provides a consistent supply of phosphate from deeper water (seabed) to support biodiversity within the surface water column (Wyrtki, 1961).

Because of the shallow water and tropical location, seawater temperature around Anambas Archipelago is warm, with little vertical variation annually, following the characteristics of the South China Sea (Wyrtki, 1961). Surface water temperature varies from 27 OC to 33 OC (Anambas and IPB, 2011, MMAF, 2014d). Bottom temperature is slightly cooler, ranging from 26OC to 31OC reflecting the surface temperature due to water active mixing and turbulence by wind and waves (Anambas and IPB, 2011, Ng et al., 2004). Surface water currents in Anambas region are highly influenced by the South China Sea monsoon system. The monsoon winds change direction twice a year, crossing back and forth across the equatorial zone, and at its peak forces the water mass in the region to move in the direction of the prevailing wind (Wyrtki, 1961, Chu et al., 1999). The general water current speeds for this region are between 0.12 – 0.31 m/s to the southwest during the winter monsoon, and between 0.12 – 0.25 m/s to the northeast during summer monsoon (Wyrtki, 1961; Anderson, 2014). Although measurements in the local region of Anambas Archipelago are rare, current speeds between 0.15 – 0.4 m/s have been reported (MMAF, 2014c, Purba et al., 2014). Anambas Archipelago has a diurnal tide, meaning one high tide and one low tide in a 24 hour period (DISHIDROS, 2013), confirmed with a Formzahl number of 4.98 (Purba et al., 2014). Based on the measurement by Purba et al. (2014), tidal variation within Anambas waters ranges from 0.68 m above mean sea level, during high tide, and 0.67 m below mean sea level, during low tide. The relatively low amplitude of the diurnal tide, and shallow water between islands, results in low-intensity tidal currents between islands in Anambas Archipelago. Like sea surface current, propagation of waves in Anambas Archipelago is strongly influenced by the monsoons. Relatively stronger wind during the winter monsoon generates higher waves compared to those of the summer monsoon. Typical wave heights in Anambas can range between 1 – 2 m with a period of 5 seconds during the peak of the winter monsoon (December – January), and is generally under 1m with a period of 4 seconds in the peak of the summer monsoon (August) (Anambas and IPB, 2011). However, because of the open space around Anambas Archipelago and its shallow water, significant waves can be generated during extreme wind conditions. To avoid these conditions, local communities generally live within embayments or areas sheltered by sandbars or fringing reefs. Similarly, most of the activities requiring a

23 permanent position in the sea, such as net cage mariculture and fish traps, are located in bays or between islets.

2.3 Water chemical properties of Anambas Archipelago

The chemical properties of waters in Anambas Archipelago are characterised as tropical warm water and mostly influenced by water mass from the South China Sea. The concentration of dissolved oxygen (DO) is considered medium to high from 4.34 to 6.16 mg/L (Anambas and IPB, 2011). One probable explanations for such high DO is the large supply of cold, dense and highly saline water from the Pacific Ocean, passing through the South China Sea, and arriving in Anambas during the peak winter monsoon (Wyrtki, 1961). pH levels of sea waters around Anambas Archipelago vary from 6.75 to 9.60 with turbidity ranging from 1.46 to 607 NTU (very clear water) (Anambas and IPB, 2011). Purba et al. (2014) reported similar pH values measured in various locations in waters of Anambas ranging from 6.75 to 9 with water clarity (Secchi depth) of 6 – 17.7 m. Ammonium (NH3-N), Nitrate ((N03-N) and Nitrite (N02-N) are also constituents, with ranges of 0.001 – 0.373 mg/L, 0.001 – 0.188 mg/L and 0.001 – 0.024 mg/L respectively (Anambas and IPB, 2011, MMAF, 2014d). General salinity variation in Anambas Archipelago and nearby islands (Natuna) varies between 31 – 34.5 ppt and is also mostly influenced by the monsoon system. The winter monsoon drives surface current, bringing highly saline water (>34.5 ppt) from the Pacific Ocean, reaching Anambas waters with slightly reduced salinity (between 33 – 34.5 ppt) due to mixing and precipitation along the way (Wyrtki, 1961, Ng et al., 2004). In contrast, the summer monsoon forces low salinity water mass (31 – 33 ppt, <30 ppt under extreme conditions) from the Java Sea, being diluted by freshwater discharge from big rivers on Java and Kalimantan (Borneo) Islands (Wyrtki, 1961, Ng et al., 2004). In terms of vertical variation, the difference between surface water and deep water is only around 1.5 ppt due to continuous mixing and relatively shallow water (Anambas and IPB, 2011).

2.4 Socio-economic condition

The small cluster of islands of Anambas Archipelago is a home for 44,288 people who live on 26 habitable islands out of 238 islands covering an area of 70 x 120 km2 (Pemda-Anambas and Bakosurtanal, 2011, Anambas and IPB, 2011, Mustika et al., 24

2013). The actual number of islands within Anambas archipelago varies across publications, from less than 70 (Ng et al., 2004) to 255 (BPS-Anambas, 2012), although the figure published by the government is 238 (Anambas and IPB, 2011, Mustika et al., 2013). It is estimated that the total land mass of Anambas Archipelago is less than 2% compared to more than 98% of Anambas territorial seas (Saidah, 2011). From this 2% of land mass (roughly 607.72 km2), only a fraction of the land mass of the 26 islands are habitable, due to rough terrain and exposure to sea conditions. With the current population growing at a rate of over 6% per year (Saidah, 2011), it is difficult for local communities to acquire stable land for houses and for public services near coastal areas. Despite the low population density (around 67 ind./km2) (BAPPENAS, 2012), social development is characterized by a high number of individuals per household (4.14 ind./HH) (BPS-Anambas, 2012) and relatively expensive land prices and rent. Considering the abovementioned facts, it is not surprising that the majority of Anambas people have a high dependency on coastal and marine resources. For example, the main occupations of more than 30% of households are fishing and fish farming (BPS-Anambas, 2012), not including those that indirectly benefit from fisheries (middlemen, fish traders, and other local fishery businesses). Other households rely on subsistence agriculture, forestry, small grocery, public services, transportation, financial services, construction, and domestic industries. Combining fisheries-related activities with agriculture and forestry, which seems to be the strategy of livelihood diversification of a typical household in Anambas, it would make up more than 75% of the total households (Anambas and IPB, 2011). However, gas and oil mining is the single major contributing economic sector, accounting for more than 75% of Anambas District total GDP (BAPPEDA-Anambas, 2009, Anambas and IPB, 2011), despite most households working in fisheries, agriculture and forestry. There are currently 10 companies exploring and exploiting gas and oil deposits in the northern part of Anambas sea with only three companies currently producing gas and oil (Table 2.1).

Table 2.1. Oil and Gas Mining Companies Operating in Anambas Area

No. Company Concession area Contract date Region Producing oil & gas 1 Star Energy Ltd Kakap 22 March 1975 Anambas Y 2 Premier Oil Natuna Block sea A 16 October 1979 Anambas Y 3 Conoco Philips Inc. South Natuna Sea 3 August 1990 Anambas Y Block "B" & Natuna 4 Senyen Oil and Gas Anambas 27 June 2004 Anambas N PTE LTD 5 Genting Oil Pte. Ltd. Northwest Natuna 12 December 2004 Anambas N 6 Pertamina Ep Udang Block 17 September 2006 Anambas N (Pt. Pan) 7 Indoreach PART 16 January 2007 Anambas N Exploration Ltd 8 West Natuna Duyung 16 January 2007 Anambas N Exploration Ltd 9 Lundin Oil & Gas Baronang Block and 13 November 2008 Anambas N Bv Cakarang Block 10 Pearl Oil Techlyte Kerapu Block 13 November 2008 Anambas N Ltd. Source: modified from BAPPEDA-Anambas (2009)

The revenues generated from fee-based mining concessions have significantly improved Anambas macroeconomic development indicators such as GDP, infrastructure, and government spending. However, it helps very little in improving micro-economic indicators such as reducing local unemployment or accelerating local business growth. The exclusive locations of the mining operation and highly technical skills required in oil and gas platforms have largely prevented direct economic interactions between local communities and the companies. Direct support from oil and gas companies for local livelihoods, such as supplying fishing equipment or scholarships, were intermittent and relatively insignificant. In addition, the risk of extensive mining pollution is high and could potentially destroy local community livelihoods that rely heavily on coastal and marine resources. Song (2008) has raised concern that oil drilling and mining have been polluting the waters of the South China Sea through four phases of oil and gas development (preliminary survey, setting up rig and drilling, oil and gas extraction, and transportation). Pollution resulting from these activities will not only affect local fishing communities in the area, but also newly designated national marine conservation park. The demographic structure of the Anambas population is predominately (>60%) working class age (15-55 years old) (BPS-Anambas, 2012). However, this number is not depicting a healthy indicator of community welfare in the area. For example, the unemployment level in the area is relatively high, i.e. around 38.09% of the total working age in 2012 (BPS-Anambas, 2012). This contradicts BAPPENAS (2012) (the 26

National Coordinating Agency for Development), which estimated the unemployment level in the same year at only 6%. Nonetheless, looking at the educational indicator, the majority of the Anambas population aged above 10 years old (>75%) do not attend school or only attended up to primary school. Such low levels of education are likely correlated with the high unemployment rates and the relatively high level of poverty (as high as 20% of the total population) in these remote islands (BAPPEDA-Anambas, 2009). It is clearly seen that despite the data inconsistency on social and economic conditions, small island communities in Anambas Archipelagic District are facing enormous internal and external challenges. The signs of increased vulnerability of small-scale household livelihood in the area can be seen, such as diversification of livelihood to paid labour or more seasonal non-fishing activities; increased cost for basic needs and livelihood operational costs; and, increasing competition from well- equipped and well-funded outsiders for natural resources. These aspects will be elaborated further in the later chapters of this thesis.

2.5 Anambas MPA development and management

Anambas MPA is one of the latest additions to Indonesian MPAs on the Southern South China Seas (sSCS). Together with five nearby MPAs, they form a connected MPA network in the Indonesian territorial sea within the sSCS. Essentially, Natuna MPA and other MPAs located to the east will be disconnected from nearby MPAs located in the western part of the region. Even though the distance between these MPAs is over 20 km, to allow self-replenishment (Mora et al., 2006, McLeod et al., 2008), it will reduce environmental degradation, poaching and illegal long range fishing activities in the region. Anambas MPA is managed by the national government and supported by local and other non-government institutions. MPAs managed by the national government are called National Marine Conservation Area or Kawasan Konservasi Perairan National (KKPN). A KKPN is determined based on its national or international importance and interest, which is detailed in Section II of the MMAF Ministerial regulation No.2/MEN/2009, regarding Procedure of MPA Designation. The other MPAs nearby (Table 2.2) are designated and managed by local government and called Kawasan Konservasi Perairan Daerah (KKPD). KKPD can be declared by provincial and district governments as part of their autonomous rights, stipulated in National Law No. 27

32/2004, which allow them to manage areas within 12 miles and four miles from the coastline, respectfully. However, this requires local government to submit a formal permission request document to the responsible ministries (KKP or Kemenhut), so the KKPD can be recognized by central governments, as stated in MMAF regulation No.2/MEN/2009.

Table 2.2. Network of Indonesia MPAs in sSCS

MPA Batam City Senayang Bengkayang Natuna MPA Bintan MPA Anambas MPA characteristics MPA Lingga MPA MPA Established 2007 2008 2007 2002 2004 2014 legal basis SK Bupati Natuna SK Bupati No. SK WaliKota SK Bupati SK Bupati No. KepMenKP No. No. 299/2007 261/VIII/2007 No. Kpts No.71/III/2002 220 Tahun 2004 37/KEPMEN/20 14/HK/VI/2007 14 PERDA Peraturan Bupati - - - - Kabupaten Bintan Bintan No. No. 12/2008 13/II/2009 SK Bupati Natuna SK Bupati Bintan No. 378/2008 No. 261/VIII

Size 142,997 Ha 472,905 Ha 66,876 Ha 481,748 Ha 15,300 Ha 1,262,686 Ha MPA type Managed Resource Managed Managed Managed Managed Protected (IUCN MPA Protected Area Resource Resource Resource Resource Seascape/Marine type) Protected Area Protected Area Protected Area Protected Area Nature Recreational Park Management Locally Managed Locally Managed Locally Managed Locally Locally Managed Nationally and status Marine Area Marine Area Marine Area Managed Marine Marine Area locally managed (LMMA-District) (LMMA-District) (LMMA- Area (LMMA- (LMMA- District) District) District) Zoning system Core + Buffer Core + Buffer Core + Buffer Core + Buffer Core + Buffer Multiple use Zones Zones Zones Zones Zones zones

Anambas was fully designated as an MPA through the MMAF Ministerial Decree, No. 53/KEPMEN-KP/2014, regarding Management Planning and Zonation of Anambas Marine Nature Recreational Park and Its Surrounding in Riau Province for 2014-2034. It has taken five years, four formal requests from Anambas Archipelago District and Riau Archipelago Province, as well as at least two environmental and socio-economic studies to complete the process of the MPA designation (LKKPN- Pekanbaru, 2013). Additionally, as being managed by the national government, an official branch of MMAF is stationed in the Anambas Archipelagic Districts to coordinate the management of the MPA.

CHAPTER 3. Regulations for Planning and Establishment of Mariculture Zoning in Marine Protected Area in Indonesia: A Policy Analysis

3.1 Introduction

The reasons for the low level of effectiveness and efficiency of MPA management in Indonesia vary and are site-specific (Siry, 2009). In general, low public participation, top-down management practices, regulation incompliances, and human and financial limitation in MPAs are regarded as the main problems and have been well studied (Alder et al., 1994, Fauzi and Buchary, 2002, Syarif, 2009, Kusumawati and Huang, 2015). However, specific studies addressing problems and complexity of regulations and policies related to MPA management are scarce (Siry, 2009), such as mariculture governance in MPAs, which is the central focus of this chapter. Various laws, regulations and technical guidelines regarding MPA management state that mariculture and other activities (i.e. small-scale capture fisheries and tourism) are important elements that support conservation efforts and local community livelihood (DJKKI, 2012, Setyawati, 2014). For example, MMAF regulation No. 30/2010 explicitly states that mariculture is an integral part of a sustainable fishery zone (SFZ) in an MPA area, complementing the other three zones (i.e. core, use, and other zones). However, this regulation provides more details on regulating core zones (no-take area) and fish capture activities in SFZ. Primarily due to the accelerated loss of biodiversity and increased fishing activity of communities living within or adjacent to MPAs in Indonesia (Anonymous, 2012a, Grantham et al., 2013). Such emphasis on biodiversity protection and capture fisheries has resulted in other activities, such as mariculture, to be overlooked. As a result, mariculture activities in MPAs are generally less regulated and monitored. For example, there was no exact number for live reef fish cultured in Spermonde (Glaser et al., 2015) and Anambas MPAs (Soemodinoto et al., unpublished). Compared to other zones, the description of mariculture specific zones within these regulations is vague. With such limited coastal space in small island MPAs, mariculture development could occupy areas that have high or sensitive biodiversity. For these reasons, this chapter will examine regulations and policies related to mariculture zones within MPAs. Firstly, mariculture has been practised by local communities in Indonesian small islands long before MPAs and the zoning scheme were established. Despite being recognized as a potential livelihood, and thus should be

29 regulated, mariculture has been largely overlooked and could potentially grow uncontrolled. Secondly, mariculture zones in Indonesian MPAs are classified as subzones within SFZ. As a result, no existing regulations provide enough details on regulating the development of mariculture within MPAs. Thirdly, efforts by the MMAF, since 2014, to produce specific regulations for the development of mariculture within MPAs are in their infancy, lacking scientifically-based comprehensive assessment. Thus, the objective of this chapter is to study and describe policy guidelines, legal instruments, and other elements of environmental decision-making frameworks used for zone planning and establishing mariculture zones in Indonesian MPAs. This information will underpin a better understanding, and process of revising, these regulatory mechanisms to ensure that mariculture can be undertaken sustainably within Indonesian MPAs. Several important research questions are addressed in this chapter: 1. Are the current decision-making processes, policies, and guidelines effective for sustainable mariculture zoning in Indonesian MPAs? 2. What elements of these regulatory mechanisms can be improved? 3. In what ways can those improvements be made so that decision-making processes are more efficient and effective? Following the objective and the research questions raised above, this chapter describes the historical development of MPA zoning policies and the current legal regulations and technical mechanisms related to mariculture, as part of the Indonesian MPA strategy. The latter part of this chapter addresses the limitations of current policy and other regulatory mechanisms. Drawing from experiences of well-established jurisdictional frameworks from around the world, several recommendations to address these limitations are discussed, framed specifically to suit the Indonesian context.

3.2 Methodology

3.2.1 Research approach

In order to critically and carefully answer the research questions, this chapter employed a simplified version of David Gil’s (1973) framework of policy analysis. The selection of this framework was based on the simple, logical and systematic approach to analyse specific programs or policies of interest (Macarov, 1974). These framework characteristics were advantageous in uncovering the complex enactment process, nature, scope, and intention of public policies (Gibb and Walker, 2011). Specifically, Troyna

(1994) stated that a critical policy analysis relies on disentangling the components of a policy into a more simple comprehensible structure. This was achieved in a recent study of policy analysis by Roberts et al. (2012) using David Gil’s framework. The summarised key points of the David Gil’s framework used in this chapter (Schwartz, 1974) are as follows: a. Issues or problems dealt with by the policy. b. Objectives, value premises, theoretical positions, target segments, and substantive effects of the policy. c. Implications of the policy for the key processes and the common domain of social policies. d. Development of alternative policies; comparison and evaluation.

3.2.2 Data sources and collection

Data sources used for this policy analysis were the main regulations used, or that refer directly to the governance, for MPA mariculture zoning and establishment in Indonesia. In addition, supporting laws, regulations and technical documents generally concerning the establishment and management of Indonesian MPAs were also used as references. This policy analysis also used literature information, such as the historical account of MPA zoning in Indonesia, as well as MPA management and zoning information for MPAs and coastal zones in other countries.

The data for this study were collected from various governmental archives. Newer regulations concerning MPAs and zoning were mostly acquired from the Organisational and Law Bureau and the Directorate General Marine Management Area of MMAF. Other data were collected from government websites, scholarly journals and several web pages of individuals who have compiled various databases regarding the development of MPAs in Indonesia.

3.2.3 Data analysis

Document analysis was the primary approach used, performed to compile, select and review various laws, regulations, and scholarly literature to gain an understanding of how mariculture zones are established in Indonesia. Document analysis was performed using the NVIVO 10 software, specifically to describe the historical development and overlap of laws and regulations concerning the development of MPAs

31 in Indonesia. Document analysis was also conducted to categorise various laws and regulations into themes that consider David Gil’s framework. The themes consisted of, but were not limited to, types, coverage, objectives, technical requirements for zoning, target groups, rights, effects, implications and values of each regulation being examined related to mariculture zones in MPAs. For example, four themes were used to categorise 10 MMAF decrees based on the terms used that referred to mariculture zones, types of mariculture zone boundaries, sizes of mariculture zones, and mariculture development strategy in the MPAs. Another example was the use of a ‘MPA-related regulation’ theme to classify four regulations governing mariculture development in MPAs from 42 MPA-related regulations.

Theme development in NVIVO 10 follows the Qualitative Research Software (QSR) International guides used to commercialised the NVIVO software. Various laws, regulations, and literature were converted into PDF and Word format and stored as NVIVO internal data sources. The contents of these were then coded or categorised using nodes (nodes refer to themes in NVIVO) based on the five key points of David Gil’s framework. A node can be expanded into hierarchical sub-nodes, which were used frequently in this study to highlight overlapping, redundant or non-existent articles or contents in the laws and regulations. The use of NVIVO in this study was not to quantify the qualitative information within the dataset; instead, it was used to classify and highlight the contents of the laws, regulations, and literature for further interpretation required in the document analysis.

3.3 Historical overview of MPA zoning regulations in Indonesia

Several authors have divided the history of MPAs in Indonesia differently. Wiadnya (2012) divided it into four different periods: pre-colonial, colonial, independence and reformation periods. Siry (2009), in general, divided it into 6 periods in which the independence period was divided into three distinct periods: post- independence, the old order, and new order. However, this chapter focuses more on the existence of Indonesian MPA regulations and the periods where there was significant development in MPA laws and regulation. Therefore, post-independence, the new order (Soeharto’s Regime) and reformation periods are best. The laws and regulations discussed in this chapter are listed in Table 3.1.

Table 3.1. List of various regulations in different levels governing Indonesia’s MPAs

No. Type Regulation Title 1 Law/Act Law Number 9/1985 Fisheries The ratification of the United Nations 2 Law/Act Law Number 17/1985 Convention on the Law of the Sea 3 Law/Act Law No. 5/1990 Conservation of resources and its ecosystem 4 Law/Act Law No. 31 /2004 Fisheries management 5 Law/Act Law No. 32 /2004 Regional government 6 Law/Act Law No 26/2007 MPA zone planning 7 Law/Act Law No. 27/2007 Management of coastal areas &small islands 8 Law/Act Law No 45/2009 Amendment of Law No. 31/2004 9 Law/Act Law No. 1/2014 Amendment of Law No. 27/2007 10 Law/Act Law No. 23/2014 Amendment of Law No 32/2004 11 Law/Act Law No 37/2014 Soil and water conservation 12 Law/Act Law No. 7/2016 Protection and empowerment of fisheries, fish farmers, and salt farmers 13 Govt. Regulation No 60/2007 Conservation of fish biodiversity 14 Presidential Regulation No. 78/2005 Management of outermost small islands 15 Ministerial Decree No. KEP.38/MEN/2004 Guidelines for coral reef management 16 Ministerial Regulation No. PER.22/MEN/2008 Organisation and structure of technical unit responsible for MPA management 17 Ministerial Regulation No. PER.03/MEN/2008 Procedure on fish species conservation 18 Ministerial Regulation No. PER.17/MEN/2008 MPA in coastal and small islands 19 Ministerial Regulation No. PER.23/MEN/2008 Organisation and structure of technical on the national conservation areas 20 Ministerial Regulation No. PER.02/MEN/2009 Procedure of MPA designation 21 Ministerial Regulation No. PER.03/MEN/2009 Procedure of establishment of fish species conservation 22 Ministerial Regulation No. PER.04/MEN/2010 Fish genetic and species utilisation 23 Ministerial Regulation No. PER.30/MEN/2010 Management and zone planning of MPA 24 Ministerial Regulation No. 9/PERMEN-KP/2013 Competence standards on special duty of management plan of MPA 25 Ministerial Regulation No. PER.35/PERMEN- Amendment of ministerial regulation No. KP/2013 PER.03/MEN/2009 26 Ministerial Regulation No. PER.13/PERMEN- Marine conservation area network KP/2014 27 Ministerial Regulation No. PER.21/PERMEN- Partnership of MPA management KP/2015 28 Ministerial Regulation No. 47/PERMEN-KP/2016 Utilisation of marine conservation areas 29 Ministerial Decree Various ministerial decree Conservation of Napoleon, Manta Ray, related to fish conservation Whale shark, 30 Local Regulation Various local regulations - Two MPA provincial and district (Peraturan Daerah) (provincial and district) regulations of the establishment of MPA - 17 regulations concerning coral reef management Note: Shaded rows/regulations are the specific regulations governing mariculture development in Indonesian MPAs

The history of regulations concerning conservation in Indonesia can be traced back as far as the Dutch colonial period in the early 19th century. The Dutch were under pressure from other European countries to curb transportation of several species of birds, mammals and plants from Indonesia to Europe (Cribb, 2008). After that, several colonial regulations (see Cribb, 2008) were passed concerning conservation of Indonesian biodiversity which mostly dealt with terrestrial animal and plants. It took 75

33 years before the first regulation concerning coastal and marine conservation was established through the enactment of Law No. 9/1985 regarding Fisheries. This law marked the first stepping stone on the governance of conservation of coastal and marine areas in Indonesia. This law provided the first direct statement that fishery reserves must be established (Article 8, clause 1) and a zoning system for the area should be created to restrict fishing and mariculture activities (Article 8, clause 2). Unfortunately, the law was lacking detailed description for marine conservation and zoning systems in in Indonesia. In some regards, this law can be considered a reactionary law, with the lack of details attributed to the pressure for Indonesia to ratify the UN Geneva Convention on the Law of the Sea. Within six months, the Government of Indonesia (GoI) ratified the Convention by passing Law No. 17/1985, regarding the Ratification of the United Nations Convention on the Law of the Sea. Article 10 of this law was about conservation and preservation of Indonesian territorial waters and was clearly a statement referring to Law No. 9/1985. However, both laws were inadequate for conservation guidance and thus required a set of lower level technical regulations to become more meaningful (Patlis, 2007). After five years, the GoI passed Law No. 5/1990, regarding Conservation of Biodiversity and its Ecosystem, which was the first law specifically governing conservation management in Indonesia. While this law addressed coastal and marine conservation, it was minimal, and fell under the overarching aims of land and forestry conservation. Interestingly, Presidential Decree No. 32/1990 regarding the Management of Reserve Areas was issued 3 months earlier. This presidential decree, while hierarchically lower, appeared to complement Law No. 5/1990, as the contents described the details of reserve types and conservation measures. The prior enactment of this presidential decree was arguably an effort to address the anticipated lack of details in Law No. 5/1990. Even though both legal products used different definitions and categories for conservation areas, which were confusing for stakeholders, they conveyed the message of conservation (Wiryono, 2003). In addition, both laws were created during the New Order Era (Era Orde Baru), characterised by the centralistic decision making in the management of conservation areas in Indonesia (Siry, 2009). For example, Article 38 of Law No 5/1990 appointed the central government, represented by the Forestry Department, as the sole authority of managing conservation areas. The roles of regional/local governments were limited, and only delegated by the central

34 government when deemed necessary. Article 39 of Presidential Decree No. 32/1990 limited the role of regional government (province and district) to monitoring, supervision, and enforcement. Rights for developing economic activities, such as tourism, or limited mining of earth resources, were excluded from the local government. During the New Order Era, until the fall of Soeharto, there was virtually no significant law or government regulation to manage marine conservation areas in Indonesia (Wiadnya, 2012). One exception was the ratification of two international conventions: the United Nation Convention on Biological Biodiversity ratified by Law No. 5/1994 and the RAMSAR Convention 1971, ratified by Presidential Decree No.48/1991 (Siry, 2009). These two government regulations primarily addressed the obligation of GoI to international communities and had few substantial effects on the governance of Indonesian MPAs. Despite establishing 24 MPAs by 1997, covering an area of 2.6 million hectares (Anonymous, 2012b), conflicts and management failures were common, particularly in the 6 national MPAs managed by the central government (Siry, 2009). For example, local stakeholders in Bunaken MPA rejected a new improved zoning scheme due to suspicion that the MPA authority favoured diving tourism enterprises (Lowe, 2003). As a result, MPA users, such as fishermen, continued to access protected zones in the MPA, which they regarded as their traditional fishing areas (Christie, 2004). The Reformation Era (Era Reformasi), which commenced in 1998, probably marked the beginning of a more decentralised management of MPAs in Indonesia (Dahuri and Dutton, 2000). Two of the most significant regulations issued during the early Reformation Era were Law No. 31/2004 regarding Fisheries and Law No. 32/2004 regarding Regional Government. Law No. 31/2004 tasked the Department of Marine Affairs and Fisheries (now MMAF) with managing MPAs and fisheries in Indonesia. However, Law No. 32/2004 gave partial rights to regional government to establish and manage its own MPAs (Patlis, 2007, Wiadnya, 2012). Since then, there has been a significant development in MPA regulations and management. For example, the Ministry of Forestry formally transferred authority of 8 national MPAs to the MMAF, through Records of Transfer No. BA.01/Menhut-IV/2009–BA.108/MEN.KP/III/2009 (Anonymous, 2012b). Local governments at provincial and district levels were able to exercise their right to establish MPAs (Kawasan Konservasi Laut Daerah-KKLD). As a result, there were at least 19 local regulations enacted in various locations in Indonesia

35 to establish new or legalise existing marine conservation or coastal management areas (Anonymous, 2011). Despite the myriad of laws and regulations governing MPA management in Indonesia (Table 3.1), several authors have expressed concerns over their effectiveness. Dahuri and Dutton (2000), Siry (2009), and Syarif (2009) argue that contradiction between laws/regulations are all too common, resulting in misinterpretation and confusion within the government. For example, the recently enacted Law No. 23/2014, regarding Regional Government, has stripped the right for the district government to establish or manage MPAs in their area, which was initially allowed in the superseded Law No. 32/2004. As a result, at least 108 locally-managed MPAs, established through local regulations (PERDA), were handed over to the national or provincial government. This is an enormous challenge, especially for the provincial government, considering the increasing number of MPAs in their possession, financial and human resource limitations, and the decreasing political influence of district governments. From the district government’s point of view, Law No. 23/2014 has triggered another problem where most of the currently planned and locally-managed MPAs initiated by the district governments are in status quo. The district governments have no longer rights to finalise the establishment of the MPAs and, at the same time, the provincial governments are reluctant to take over due to unavailability of resources, prioritisation of established local MPAs or simply out of disinterest. One clear example was the status quo of the finalised plan for the Buton Strait local MPA in the Southeast Sulawesi, Indonesia. The provincial government refused to take over the MPA establishment as the MPA plan was legalised by the district government and needed to be revoked first. However, the district government argued that Law No. 23/2014 had automatically nullified the district regulation and thus, it was not necessary to revoke the regulation (Sairuddin, 2018, personal, communication, 3rd December). As of today, the participation of district governments in MPA planning and establishment in their area is limited to a public hearing held by the provincial government. In addition, some provisions of existing laws and regulations regarding the participation of local communities in MPA management are vague and offer little explanation on how they should be executed. For example, MMAF regulation No. 21/2015, concerning Partnership in MPA Management, contains very general partnership programs between MPA authorities and the public (Article 10). A follow-up

36 provincial governor decree should then be enacted for every MPA, concerning the technical arrangement of the partnership, which to my knowledge does not exist. This condition, coupled with the unpreparedness of local communities, has caused infectiveness and failures in MPA management in Indonesia (Siry, 2009, Glaser et al., 2012).

3.4 Policy analysis of MPA mariculture zone planning in Indonesia

In developing countries, MPA zoning is directed more toward resolving conflicts between coastal users and reducing pressures on coastal resources and biodiversity from extractive economic activities (Agardy, 1993). Indonesia, a developing country, addresses conflicting uses in its multi-use MPAs (MUMPAs), which constitute at least 77% of total MPAs (Table 3.2), through various zoning laws and regulations. From more than 19 regulations governing MPA zoning and management in Indonesia, mariculture zones in MPAs are governed by five different regulations (Shaded regulations in Table 3.1). Within these regulations, mariculture zones are frequently mentioned as part of the MPA zones despite being regarded as part of the larger sustainable fisheries zone.

Table 3.2. Multi-use MPAs in Indonesia

Source Total Small island (SI) Partially SI Populated Not pop. MUMPA (%) MMAF-MF* 108 55 22 71 7 66% WDPA** 148 124 24 N/A N/A 80% KLN*** 106 82 24 82 24 85% Compiled from : *) DKKJI (2013a); **) IUCN and UNEP (2013); ***) Terangi et al. (2015)

Law No. 5/1990 is arguably the first piece of conservation legislation recognising multiple use zones within protected areas, including ecotourism in use zones and farming in other zones being permissible. Despite the unclear definition of mariculture zones or activities, the law formally recognises the sustainable use of space and resources within specific MPA zones (Article 9), such as fish farming (Article 17). Due to its intended use as terrestrial conservation legislation, Law No. 5/1990 only describes the possibility of establishing or allowing mariculture activities in MPAs. The second legislation is Law No. 27/2007 concerning Management of Coastal Zone and Small Islands. This law stipulates that specific zones for mariculture, in waters of small islands, must be formalised to protect local community rights (Article

23). This law recognises that mariculture is not only economically important but also socially significant for the local community of small islands in Indonesia. By protecting local community rights, this law is expected to motivate the local community to support conservation efforts. The recently issued Law No. 7/2016, regarding the protection and empowerment of fisheries, fish farmers, and salt farmers, has strengthened the position of small-scale fish farmers. The law stipulates that the central and regional governments, both provincial and district, have an obligation to provide tangible support for small-scale fish farmers from initial production to market in a written agreement. However, the law covers the mariculture activity of small-scale fish farmers in regular coastal areas which are outside the focus of this study. In fact, the law has a potential conflict with Law No. 23/2014 if applied in MPAs considering the lack of rights by district government to manage mariculture activity in MPAs. Government Regulation No. 60/2007, concerning Conservation of Fish Resources, places the importance of mariculture zones within the context of MPAs. This regulation clearly stipulates, in multiple articles, that mariculture can be undertaken within MPA sustainable fishery zones for two reasons. Firstly, mariculture can serve as a conservation tool to maintain the population and genetic diversity of unprotected and threatened wild fish within conservation areas. For example, the pressure of capture fisheries on certain fish species can be reduced by culturing the fish in the MPA or offset by restocking threatened fish species through mariculture. Secondly, it provides guidance for local communities to develop mariculture sustainably, following culture limitations on fish species used, types of feed, culture technology, culture scale and CC. The limitations of fish culture in this regulation are perceived to prevent overdevelopment of mariculture in MPA areas in Indonesia as seen in regular coastal areas. MMAF Regulation No. PER.30/MEN/2010, which carries the mandates of LAW No. 27/2007 and Government Regulation No. 60/2007 describing the technical requirements of planning, management and zonation of MPAs, contains similar general articles of mariculture management in MPAs to Law No. 27/2007. This regulation confirms that mariculture can only be undertaken within an MPA SFZ as one of the strategies for local community empowerment. Considering the lack of details of MPA usages and benefits for the local community in previous regulations, MMAF has recently passed regulation No.

47/PERMEN-KP/2016 concerning the Usage of MPAs. Within this regulation, the extent of mariculture is addressed through various articles to meet conservation goals, such as minimising the negative effects of mariculture, establishing mariculture permits (individual and corporate) or registration (small-scale fish farmer), and specifying sanctions for violations. The regulation only allows traditional and semi-intensive mariculture systems and the use of fish species deemed to have no potential risk to wild fish populations and the MPA ecosystem. The extent of mariculture activity is also strictly regulated, where only 50% of both the total CC and total SFZ can be used for fish culture. If compared to mariculture development in non-MPA areas, which reaches 20–30 % of annual production growth (Rimmer et al., 2013, Anonymous, 2014), it is clear that this limitation is intended to restrict expansion or production of mariculture in MPAs. Nevertheless, this regulation gives conflicting messages, as not only small-scale fish farmers but also individuals and private business entities can develop mariculture in MPAs. Individuals and business entities can file for a mariculture permit with no obligation to undertake environmental impact assessment (AMDAL) if the fish farming activity is < 5.0 ha or < 1.000 cage units (MenLH, 2012, Rimmer et al., 2013). MMAF bears the responsibility of determining the mariculture environmental CC for MPAs, thus it would be difficult to hold individuals and corporations accountable for any negative impact on the MPA ecosystem. Given the historical account of MPA development in Indonesia and the objectives of the mariculture-related MPA regulations, the following sections in this chapter provide analyses revolving around: 1. The issues and problems that the regulations aimed to deal with; 2. The objectives, value premises, theoretical positions, and effects of the regulations; 3. The implications of the regulations for the current MPA regime in Indonesia; and 4. Alternative policies or improvement of the regulations.

3.4.1 The issues with MPA mariculture zoning in Indonesia

3.4.1.1 Mariculture development in Indonesia’s MPAs

The concept of MPA zoning mainly deals with preventing or limiting human activities (extractive activities) carried out in protected, ecologically sensitive, valuable or damaged areas to maintain or improve ecological function (Day, 2002). This concept has been the main objective of MPA zoning plans and practices around the world,

39 including Indonesia. The enactment of Law No. 5/1990 was generally serving this purpose through mandating sustainable resource uses (particularly capture fisheries) within MPAs (Sloan, 2002). This law aimed to deal with the projected increasing pressure of capture fisheries on fish resources. Pet-Soede et al. (1999) reported sigificant fish capture in Indonesia, up 42.3% in 1996 compared to 1990, which confirmed the law’s projection. The focus on regulating capture fisheries was probably the reason this law overlooked the potential for mariculture development (specifically brackishwater and marine aquaculture) in MPAs. A decade after this law, the more regulated capture fisheries had given rise to aquaculture to fill the gap in fish supply. From 2000 to 2004, aquaculture production had increased significantly at a rate of 10.36% per year (Sukadi, 2006), compared to 5.53% per year between 1988 and 1998 (Ahmed and Lorica, 2002). This brought new concern, that aquaculture might negatively impact ecosystem sustainability (Sukadi, 2006), especially if the activity took place in an MPA. Limited fishing rights in MPA zones and over-exploitation of fish will force local communities in or adjacent to MPAs to adapt to the situation (Wiadnya et al., 2011) through intensifying fishing activities, or other means of producing fish such as mariculture. The Reformation era (post-Soeharto’s Orde Baru) marked the end of centralised management of coastal and marine areas and gave way to several new decentralised laws and regulations, such as Law No. 32/2004 regarding Regional Government, and Law No. 31/2004 regarding Fisheries. For example, Law No. 31/2004, which partially addressed aquaculture development, did not regard MPAs as unique areas to be regulated differently for mariculture. As a result, the development of mariculture in MPAs follows the same general aquaculture regulation as non-MPAs areas. The subsequent laws, Law No. 27/2007 and government regulation (Peraturan Pemerintah) No. 60/2007, were issued to deal with problems related to the right to use seascape and fish resources linked to mariculture development in small island MPAs. Law No. 27/2007 was specifically designed to address the uncontrolled development in coastal zones and small islands due to the increased control and power of local authorities (Wever et al., 2012). Most of these uncontrolled developments were the result of more than 7000 local government regulations designed to increase local government earnings with little consideration to conservation and sustainability (Dirhamsyah, 2006, Wever et al., 2012). Within Law No. 27/2007, mariculture and

40 other economic activities must be zoned, approved, and should not overlap with other conservation areas. The government regulation No. 60/2007 was arguably the first regulation that clearly stipulated that mariculture activities can take place within an MPA. Despite the regulation generally dealing with fish resource use and conservation, several articles were dedicated to mariculture in MPAs. For example, Article 30 stipulates that fish farming or mariculture are allowed in an MPA, but limited to the SFZ. This regulation provided more information by stating that the type of fish, fish feed, culture technology, the number of units and CC (Article 32:4) need to be considered. Furthermore, this regulation mandated a subsequent ministerial regulation to be issued, with the express purpose of explaining these requirements in more detail. The subsequent MMAF regulation No. PER.30/MEN/2010 failed to describe how the expansion of mariculture in Indonesian MPAs should be regulated, just restating the statements found in the original regulation (No. 60/2007). For example, Article 21 of MMAF regulation No. PER.30/MEN/2010 was worded almost the same as Article 32:4 of regulation No. 60/2007. In 2016, MMAF issued regulation No. 47/PERMEN-KP/2016 to address some of the problems of mariculture expansion in MPAs. This regulation is like previous regulations but includes some additional precautionary steps. In this regulation, mariculture activities can only occupy 50% of the total area and use 50% of the CC of the MPA SFZ. The scale of mariculture operations allowed can only be traditional (i.e. natural feed or low stocking density) and semi-intensive fish farming (i.e. formulated feed and medium stocking density). These restrictions signal the concerns of MPA stakeholders over the possibility that nutrient release from mariculture can exceed MPAs CC. However, studies such as Alongi et al. (2009) in non-MPA areas, and McKinnon et al. (2010) in an MPA, found that traditional and semi-intensive mariculture are minor contributors of nutrients in surrounding nutrient enriched areas. In addition, species of fish allowed to be cultured under this regulation must not cause damage or imbalance in the protected ecosystem. Fish diseases, genetic transfer, and invasive species are associated potential risks of certain fish species and are also addressed by this regulation. There is limited to non-existent data available regarding the likelihood that wild stocks of fish in Indonesia are impacted by mariculture. An exception might be the case of shrimp diseases in brackishwater aquaculture, as

41 reported by Primavera (2006), despite no shrimp farming allowed to take place in Indonesian MPAs. Though it is important to recognise these threats, there is no conclusive data that they have eventuated into realised impacts (Rimmer, 2010, De Silva, 2012).

3.4.1.2 Mariculture permit in Indonesia’s MPAs

Regulating mariculture development in the coastal zone through licensing or permits has proved difficult to implement in Indonesia due to the inadequacy of the regulatory framework and enforcement (Rimmer, 2010). Despite the drawbacks, mariculture in MPA areas currently use licensing or permit schemes. Typically, there are two mariculture licensing types under MMAF regulation No. 12/2007, which were later improved by MMAF regulation No. 49/2014 in 2014. The first is the aquaculture license permit, or Surat Ijin Usaha Perikanan (SIUP), for individual and private businesses who own a hatchery and fish farm of more than 0.5 and 2 ha, respectively. A specific letter (Recommendation for Foreign Investment in Aquaculture or Rekomendasi Pembudidayaan Ikan Penanaman Modal or RPIPM) should be obtained if the aquaculture operation is funded by foreign capital. The only difference with non-MPAs is that the Directorate General of Aquaculture (DGA) has the sole authority to approve a SIUP in an MPA. Small-scale fish farmers are obliged to report to the local district marine and fisheries office (DKP) through an aquaculture registration letter (TDKPI), or Tanda Pencatatan Usaha Pembudidayaan Ikan (TPUPI), acquired from the local MPA authority. The double obligations to register and regularly report their activity are obviously a burden for small-scale fish farmers, especially those who live far from the district’s capital city where these two offices are usually located. The additional obligations and travel distance might cause low registry and reporting participation, particularly for a TDKPI. Historically, MPA authorities have a limited relationship with local small-scale fish farmers due to their position as the central or provincial government representative. Currently, only 0.36% from 3.8 million fish farmers in Indonesia have a SIUP or were certified in 2015 (DGA, 2017). In terms of registered small-scale fish farmers, the numbers of TPUPIs are uncertain, despite the district DKP having a good relationship with local fish farmers. Thus, enforcement of a TDKPI by DGA and the MPA authority is a challenge considering their limited resources and relationship with local communities. 42

3.4.1.3 Mariculture zoning in MPAs’ SFZ

There are different interpretations of how far mariculture can be allowed or practised throughout Indonesian MPAs within MPA zoning documents found in this study. A legal document analysis of management and zoning plans from 10 national and local MPAs in Indonesia revealed that mariculture is stated as one of the short, medium and long-term strategies in the conservation framework (Table 3.3). Each document presented different statements on how mariculture zones should be established, such as by instantaneously selecting areas, boundary marking with geographic coordinates or simply a statement that mariculture is allowed in MPAs without reserving parts of an MPA for the activity. Peculiarly, MPAs that already have established mariculture subzones still devised plans for site suitability analysis for the sub-zones. This suggests that the existing mariculture sub-zone was established or designated without a proper and clear scientific and social-based process of site suitability analysis. The variation in the design of mariculture zones has triggered studies aimed at providing standardised site selection for mariculture within an MPA. Few published efforts to devise a mariculture zoning scheme through site suitability analysis in several well-established MPAs, such as Yusuf (2007) in Karimun Jawa MPA and Syadiah (2010) in Selayar Island, were focused only on the capability of the waters to support mariculture. Syadiah (2010) might have added one improved analysis in which she calculated the possible impact of a potential mariculture site on a nearby core zone. However, both studies failed to provide an integrated analysis that combines mariculture site selection with other aspects related to an MPA, such as the complex relationship with other users, visual amenity and possible overlap or conflicts with highly sensitive MPA features. Similarly, a study carried out by DKP-Anambas and LPPM-IPB (2011a) to determine site suitability of floating net cage mariculture within Anambas MPA, estimated suitable locations based only on chemical, physical and biological parameters. These studies also share similarities where the involvement of MPA stakeholders, particularly small-scale fish farmers, was absent in the site selection development process (i.e. local community preference). The inconsistent mariculture zoning and site selection practices in MPAs are associated with several regulations lacking depth, specifically to address the development of mariculture in MPAs (Law, No. 5/1990, Govt. Regulation No. 60/2007, MMAF regulation No 30/2010 and No 47/2016). Firstly, site selection for a mariculture

43 zone is not included within the MPA regulations as one of the prerequisites in establishing the SFZ. This essentially means that mariculture can be undertaken in any part of an MPA as long as it is located within the SFZ. Secondly, there are vague statements within the regulations regarding who determines the location of mariculture zones within SFZ. An analysis of the Technical Procedure of Best Practice Aquaculture Application, or Cara Budidaya Ikan yang Baik (CBIB), issued by DGA (2012), reveals that it is the responsibility of fish farmers to determine the best location for mariculture. If the process and selected location meet the CBID standard, a fish farmer can be issued a CBIB certificate. This passive strategy might be less controversial with non-MPA areas due to the absence of unique and protected ecosystems. However, MPAs have sensitive and formally protected ecosystems which require a strategy to control mariculture development. Thirdly, MPA regulations lack guidelines regarding local community participation, particularly small-scale fish farmers, in determining the location of mariculture zones within SFZ. This is contrastingly the opposite to the designation of a core zone, where multiple consultations with the local community are formally required. As a result, mariculture zones within the SFZ are usually difficult to differentiate from other zones and are less utilised.

Table 3.3. Mariculture zoning in some of Indonesia’s small island MUMPAS established through MMAF decrees

MMAF decree MPA MZ Boundaries Size Mariculture SS and CC

64/KEPMEN- Aru Archipelago Marine SFZ Location only - No actual N/S Not conducted – Planned in the KP/2014 Reserve boundaries 4th 5-year strategy 63/KEPMEN- Raja Ampat Archipelago MSz in SFZ Location with geo- 3,652 Ha (6.09%) Not conducted – Planned in the KP/2014 Marine Reserve coordinate boundaries 4th 5-year strategy

60/KEPMEN- Waigeo Archipelago Marine N/S – mariculture N/A N/S Not conducted – Planned in the KP/2014 Reserve allowed 4th 5-year strategy

06/KEPMEN- Sawu Sea Marine National SFZ (General and Location with geo- 456.36 Ha (0.01%) Not conducted–Not planned in KP/2014 Park traditional subzones coordinate boundaries MPA long-term strategy mixed with ecotourism 53/KEPMEN- Anambas Archipelago Marine MSz in SFZ Location with geo- 7,784.38 Ha Not conducted–Not planned in KP/2014 Recreational Park coordinate boundaries (0.64%) MPA long-term strategy mixed with ecotourism 59/KEPMEN- Kapoposang Archipelago SFZ (General and Location only - No actual 20,900 Ha (41.8%) Not conducted–Planned in KP/2014 traditional subzones boundaries MPA 5 and 10 years strategy

62/KEPMEN- Padaido Archipelago Marine SFZ No specified location nor N/S Not conducted–Not planned in KP/2014 Recreational Park mariculture zone MPA long-term strategy

58/KEPMEN- Banda Sea Marine Recreational MSz in SFZ Location with geo- 109.2 Ha (4.37%) Not conducted–Not planned in KP/2014 Park coordinate boundaries MPA long-term strategy

57/KEPMEN- Gili Ayer, Meno, and SFZ No specified location nor N/S Not conducted–Not planned in KP/2014 Trawangan Marine mariculture zone MPA long-term strategy Recreational Park 38/KEPMEN- Pieh Island Marine SFZ No specified location nor N/S Not conducted–Not planned in KP/2014 Recreational Park mariculture zone MPA long-term strategy Remarks: N/S=Not specified; N/A=Not available; SFZ= Sustainable fishery zone; MZ-=Mariculture zone; MSz=Mariculture sub-zone; SS=Site Suitability; CC=Carrying capacity

3.4.1.4 Implementing agency

The latest change in the organisational structure of the MPA authority, mandated by Law No.23/2014, regarding Regional Government, is the transfer of national MPA management (MPA encompassing two or more provincial areas or KKPN) to MMAF. The law also hands over the management of locally-managed MPAs (KKPD) from district governments to provincial governments. As a result, any spatial use of MPAs for economic purposes, including mariculture, should be approved by MMAF or provincial governments, regulated through MMAF Regulation No. 47/2016. For example, medium-scale fish farmers must apply SIUP directly to MMAF through the DGA by including various supporting documents. Small-scale fish farmers must apply for a TPUPI from the local MPA authority, which is under MMAF for KKPN and provincial government for KKPD. The change of administrative structure in licensing mariculture in an MPA has caused several overlapping and contradicting authorities between the central government (MMAF) and provincial and district government. Firstly, the right of provincial and district governments to issue a SIUP for mariculture in a national MPA or KKPN is taken by DGA. While there are benefits of centralising the licensing permit, there are substantial logistical issues in requesting a SIUP for fish farmers in an MPA far from the central DGA office in Jakarta. By losing the right to issue a SIUP, previously mandated in the MMAF Regulation No 47/2014, provincial and district government now have less control over the development of mariculture in an MPA under their administrative control. Secondly, district government control of small-scale fish farming within KKPD, using TPUPI, is no longer effective, as the provincial government holds the authority over the KKPD. Small- scale fish farmers will also face similar logistical issues when reporting their TPUPI regularly to the local unit of MPA authority due to distance and communication challenges. In addition, the local MPA authority unit will also face an enormous task administering the TPUPI due to limited human and financial resources. Overall, evaluating the above regulations indicates that regulations must be clearer to effectively regulate mariculture development, zoning and licensing, and to improve the organisational structure of mariculture management for Indonesian MPAs. The increasing popularity of mariculture, combined with the optimal environmental condition in MPAs,

46 are pressing factors to refine these regulations to deal with the possible future expansion of mariculture in MPAs. Reforming these regulations should also include strengthening the rights of local small-scale fish farmer who live in, or adjacent to, MPAs to be involved in decision-making regarding site selection of mariculture zones and farming activities.

3.4.2 Objectives, value premises, theoretical positions, and effects of the regulations

3.4.2.1 Regulations’ objectives

Hierarchically, Laws or Acts are the highest-level regulations in Indonesia after the Constitution and usually have broad objectives and general articles. A specific article usually exists within these laws which mandate the enactment of lower level regulations to provide further elaboration or technical explanation (Patlis, 2007). However, five of the regulations have similar objectives (shaded regulations in Table 3.1) regarding mariculture development in MPAs, despite their different levels within the regulation hierarchy (Figure 3.1.). Both Law No 5/1990 and Law No. 27/2007 have the objective to give a universal right to undertake economic activities within MPAs, although mariculture was not mentioned explicitly as one of the activities. Government regulation No. 60/2007 is the first regulation to have a clear objective of regulating mariculture development in MPAs. This objective is intended to ensure mariculture is legalised and undertaken in the SFZ with specific requirements different to mariculture activities in non-MPAs. Since this regulation covers multiple aspects of fish resource conservation, it mandates the issue of a ministerial level regulation to address the technical implementation of mariculture within MPAs. The MMAF regulation No. Per.30/MEN/2010 and newly issued No. 47/PERMEN- KP/2016 supposedly aim to expand the technical requirements of mariculture in MPAs. Surprisingly, regulation No. PER.30/MEN/2010 has a similar objective and general coverage to previous higher regulations. Three articles (Article 11, 18, and 21), with 10 clauses, have the same broad description of mariculture type allowed in an MPA, with short additional clauses mandating a mariculture permit and requirements. MMAF regulation No 47/PERMEN-KP/2016 covers almost the same aspects of mariculture to that of MMAF regulation No. PER.30/MEN/2010, except for few elaborations on mariculture requirements and permits. The intended objective of regulation No 47/PERMEN-KP/2016 was to ensure proper usage of an MPA seascape for mariculture and several other economic

47 activities. The generic objective and lack of detailed technical elaboration carried by these two MMAF regulations have resulted in different interpretation in establishing mariculture zones within MPAs. For example, all MMAF regulation (Table 3.1 and Figure 3.1) and local regulations, regarding the planning, establishment, and zonation of MPAs, have either applied different mariculture zoning within the SFZ or do not provide valid information on how the zone should be established.

PermenKP No. 47/PERMEN- KP/2016

Figure 3.1. Regulation hierarchy regarding MPA establishment and zonation in Indonesia In addition, two lower-level regulations that deal with MPA zoning and management were also issued by the MMAF Directorate General (DG) of Marine, Coastal and Small-Island or KP3K (now DG of Marine Spatial Management or PRL) responsible for managing MPAs (Figure 3.1.) One regulation (KP3K Regulation No. 02/PER-

DJKP3K/2013) provides guidelines on the spatial boundary and zoning of marine protected and small island areas. The other one (KP3K Decree No. KEP.44/KP3K/2012) deals with the monitoring of MPA management effectiveness in Indonesia. Due to the concentration of technical and scientific aspects of core zones and other protected areas in these two regulations, there are no technical instructions for spatial zonation of mariculture zones or science-based frameworks to monitor mariculture activities in MPAs. These aspects remain the challenges that MMAF must address in the near future.

3.4.2.2 Value premises underlying the regulation objectives

The management of natural resources in Indonesia is based on the Indonesian Constitution, Article 33(3) that ‘the land, waters and the natural resources within [Indonesia] shall be under the powers of the State and shall be used to the greatest benefit of the people’ (Patlis, 2007). Before the Reformation Era, this article was interpreted excessively by the central government and community rights were largely ignored in the decision making around natural resource management and use (Santosa, 2008). In the Reformation Era, the balance between state powers and community rights in natural resource management has been fairly demonstrated, at least within the regulation system. For example, local community participation has been reserved in most MPA regulations, which translated into practice through what Christie (2004) called “broad community participation”. This value premise has also been translated into the governance of MPA establishment and zoning through eight principles of coastal and marine conservation. They are principles of usage, fairness, partnership, equality, integrity, transparency, efficiency, and sustainability, which are found in the three main conservation laws (Law 27/2007, 31/2004, and 45/2009 and government regulation (PP No/ 60/2007). Similarly, MMAF regulation No. PER.30/MEN/2010 holds these overt values, which are described in the technical short, medium, and long-term implementation strategies of any MPA in Indonesia (Article 5, 6 and 7). These values were clearly adopted from the principle of Integrated Coastal Management (ICM) to ensure a balanced relationship with human-nature (Wever et al., 2012). This value premise shift, from authoritarian to a more democratic policy within these regulations, are the result of political change in Indonesia after the New Era ended (Wever et al., 2012). However, as Patlis (2007) and Wever et al. (2012) point out, policy 49 change to support coastal and marine resource management in Indonesia remains a rhetorical discourse at the field level. Both Patlis (2007) and Wever et al. (2012) attribute this problem to a constant power struggle between central and regional/local government over resource management. For example, Lowe (2003) found that local government officials and private business entities derive financial gain from MPA resource exploitation. These “under the table agreements” prevented local community gaining any benefits and diminished the central government efforts to protect the resources (Wever et al., 2012) In terms of mariculture development in an MPA, the regulations show a shift of value premises from rigid marine reserve objectives to a balance between conservation and local community resource use. This value change has also highlighted the change of Indonesian MPA management strategy toward securing local community support through economic development in MPAs (TNC, 2010). At the same time, the regulations maintain value premises that any economic development, including mariculture, should be carried out sustainably through limiting the use of seascapes, fish farming permits and the scale of activities. However, the generality of the regulations objectives and provisions is concerning (Patlis, 2007). Patlis (2007) argued that the implementation of such regulations has mostly benefited economic parties with large capital and resources, instead of the general community. Mariculture is one of the most rapidly developing sub-sectors of aquaculture in Indonesia (Rimmer et al., 2013). Developing mariculture in relatively stable and optimal environmental conditions, such as those present in an MPA, is a tempting business venture. The argument is that provisions in the regulations for environmentally-friendly mariculture might serve as a covert value premise to allow large-scale mariculture operation in an MPA. Generic statements in these regulations can be interpreted to allow large mariculture operations to occur within MPAs so long as they are sustainably sound. These covert value premises are similar to the general objectives and provisions concerning ecotourism activities in Indonesian MPA, which mostly benefit private entities with strong capital resources (Walpole et al., 2001, Christie, 2004, Ponting et al., 2005). A document analysis of a TNC report revealed that at least 9 leasing agreements, in the form of marine conservation agreements (MCA) in Indonesian MPAs, were for developing high investment pearl oyster farms. This indicates a potential bias in implementing regulations regarding the

50 development of mariculture in MPAs, which will further impair local community rights to use the seascape in an already limited SFZ.

3.4.2.3 Theories or hypotheses underlying the strategies and the concrete provisions of the regulations

The current strategies and provisions of Indonesian MPA regulations involving economic activities in MPAs are largely based on the contemporary principles of equity, power-sharing, participation and shared benefits. According to Spalding et al. (2013), these principles give rise to further strategy development of MPAs, with the view that economic and social development are equally important to the conservation of the ecosystem and the services it provides. However, this strategy was focused mostly on small-scale fisheries and ecotourism communities. For example, Kusumawati and Huang (2015) found that compliance with conservation regulations was improved by empowering local fisher communities with the responsibility of managing the MPA they work in. On the other hand, scholars such as Pomeroy et al. (2006) have doubted that mariculture can be as successful as small-scale fish capture in coastal conservation, due to its high capital and technical requirements, and potential impacts. Even though this argument has a basis in a non-fish farming community, it might be a different case in a coastal community where fish farming has been established as part of the community livelihood. In that case, the theoretical premises regarding mariculture development in Indonesian MPAs can be applied, where established mariculture activities must be supported as an alternative livelihood for local communities. If it is successful, mariculture might serve as a trade-off for destructive fishing activities and thus increase local community support for conservation objectives. Another theoretical premise of mariculture development in MPAs is that mariculture must be controlled to a certain level to avoid negative effect on biodiversity in the protected area. To ensure the sustainability of mariculture in an MPA, all the five MPA regulations unanimously require that certain limitations are set to ensure it is environmentally friendly. For example, an explicit 50% universal limit of CC of an SFZ mariculture zone is imposed by MMAF regulation No. 47/2016. This limitation will effectively prevent medium to large-scale mariculture and thus the environmental impacts that arise from them. On the contrary, the CC limitation will severely affect the

51 development of small-scale mariculture. For example, a well-flushed area can support net cage mariculture near CC, with no obvious impacts on the environment (Merceron et al., 2002). However, net cages can only occupy a maximum of 50% of the total area, or a set number of units/area (on top of the CC limitation), due to the regulations. This means that small-scale fish farming will be considered illegal, or to a lesser extent unsustainable, when exceeding 50% of the area or number of net cages. The limitation does not differentiate between the types of, and the capability of the area to sustain, mariculture activities. Even though small-scale mariculture has been shown to have marginal impacts on the surrounding ecosystem (Alongi et al., 2009), and feed waste escaping from small-scale mariculture may even benefit wild fish in high biodiversity areas such as MPAs (De Silva, 2012). For example, Dempster et al. (2006) and Sudirman et al. (2009) found a significant increase in biodiversity and fish assemblages benefiting from uneaten/escaped feed near mariculture net cages.

3.4.2.4 Target segments of society: those intended to be directly affected by the regulations Regulations discussed so far target all segments of Indonesian society, binding all communities living within or adjacent to MPAs that have direct access to resources, mostly being small-scale fish farmers and fishers. This is not surprising considering coastal water users are dominated by small-scale fishers and fish farmers, consisting of at least 2.19 million coastal small-scale fishermen and half a million coastal small-scale fish farmers (MMAF, 2014b), who are mostly below the national poverty line (Adhuri et al., 2015). As fishing activities are increasingly regulated within MPAs, mariculture will gain popularity. This will increase the number of fish farmers and areas used for mariculture activities within MPAs. The total number of small-scale fish farmers operating within or adjacent to an MPA is difficult to estimate. However, during 2009-2013, the number of households in the mariculture sub-sector, represented by finfish mariculture, had the highest increase (15.09%) compared to land-based aquaculture (6.57%) (MMAF, 2014b). Another target segment for these main regulations is medium-scale fish farmers who are also given rights to operate within SFZ of MPAs. The inclusion of this group was first vaguely indicated in the Government Regulation No. 60/2007 and MMAF regulation No. PER.30/MEN/2009. Both regulations stated that individual and business entities have

52 the right to carry out mariculture activities in an MPA with no elaboration on the scale and the details of the operators. However, MMAF regulation No. 47/2016 specified the right to undertake medium scale mariculture in MPAs for this group by allowing a SIUP to be issued for semi-intensive mariculture. This may lead to the expansion of mariculture activities in MPA SFZs, funded by foreign capital, to culture highly valuable species such as pearl oysters. This assumption can be substantiated from the TNC (2010) report which states that five out of nine pearl oyster farms, operated in nine MPAs, are funded by foreign investment from Japan and Australia. Despite foreign mariculture corporations bringing positive change, such as employment opportunities and improved infrastructure, they also create economic dependencies, cultural change, and resource use conflicts with local MPA stakeholders (McLeod et al., 2009). Looking at the future, this trend might continue as the rate of foreign investment in aquaculture (mariculture and brackishwater aquaculture) grows at a steady annual rate of 100.01%, with a total investment of 117 million USD (DJPB, 2016).

3.4.2.5 Short- and long-term effects of the regulations

Each of the five MPA regulations (Table 3.4) addresses only some aspects of mariculture, with redundancies and overlaps with the other regulations. However, the overall intended effect of these regulations was to legalise the development of mariculture, undertaken by local communities, in MPAs where conservation measures might displace or limit local livelihoods. A secondary intended effect, despite its obscurity, was to formalise the operation of medium-scale mariculture activities within MPAs. Thirdly, was that through these regulations, mariculture development can be regulated, monitored and evaluated to assure its sustainability. If these regulations were implemented correctly, there will be an expansion of small- and medium-scale mariculture activities in the SFZ, allocated for mariculture development. According to MMAF (2014b), the use of the seascape for mariculture has increased by more than 82% within only 2 years, compared to brackishwater and freshwater ponds. Both brackishwater and freshwater have shown relatively stagnant development (-1% and 34%, respectively), probably due to land limitations (Table 3.4). Using that statistical data as a proxy value, the regulations will likely boost fish culture production and seascape usage, due to favourable conditions for mariculture in MPAs compared to those of 53 non-MPAs. The drawback is that existing mariculture activities outside the SFZ zone will have to be relocated to designated areas. This is likely impossible to implement, as most of the small-scale mariculture in coastal areas take place near local community settlements for reasons of safety and easy access (Pomeroy et al., 2006). A standardised site selection and CC model specifically designed for MPAs might help ameliorate the problem by ensuring existing areas stay intact.

Table 3.4. Land and seascape usage for aquaculture in Indonesia during 2009 - 2013

Culture system (ha) Year Brackishwater Freshwater Aquatic waters Paddy-fish Mariculture

Potential 2,964,331 541,100 158,125 1,536,289 12,123,383 2009 669,738 153,316 1,686 127,679 43,804 2010 674,942 148,278 1,382 138,715 117,650 2011 652,475 126,382 1,855 151,630 169,292 2012 657,346 131,776 1,847 156,193 178,435 2013 650,509 176,509 1,564 124,057 325,825 Source: MMAF (2014b)

The way mariculture is developed and conducted within MPAs is further limited by some government regulations. For example, some government and non-government institutions that manage MPAs only support mariculture activities that use locally-bred fingerlings or lower trophic level organisms such as seaweed. Raja Ampat (Setiawan, unpublished) and Komodo national parks (TNC, 2003) are clear examples of these approaches. In other MPAs, notably smaller and locally-managed MPAs, small-scale mariculture activities are less, or unrestricted, due to the absence of regulation enforcement and non-compliance. For example, clearing coral reef patches to build fixed net cages or capture-based mariculture using highly protected fish species, such as Humphead wrasse (Cheilinus undulatus), are common (Soemodinoto et al., unpublished, Glaser et al., 2015). The generality and overlap within the regulations has also created unintended consequences. There is an increased risk of competing seascape use in the future between small-scale (local community) and medium-scale mariculture activities (non-local business entities). Small-scale mariculture has the advantage of being exempt from fees and eligible for direct incentives and economic packages. However, medium-scale mariculture usually has a strong vertical (to central government) and horizontal (to local leaders) relationship in 54 conducting their activity. In most cases, medium-scale mariculture usually wins the right to use the seascape, despite objections from the local community. For example, McLeod et al. (2009) reported that foreign-owned mariculture in Raja Ampat MPA was given area concession right by the local leaders, despite local community objection to the strict control of large areas and concern of possible disease spread. The control over large areas is particularly problematic, since most of the MPAs in Indonesia are small island MPAs with limited area suitable for sustaining small-scale mariculture. A report by TNC has highlighted the growing interest of companies to develop pearl oyster mariculture within Indonesian MPAs, due to the relatively protected, and optimum, environmental conditions (TNC, 2010). Based on a closer look at national and local regulations, regarding the establishment of each MPA, one-third mention that pearl oyster is proposed as one of the favoured species for mariculture. Although the pearl oyster is considered a lower trophic species, and its culture is an environmentally friendly mariculture activity, its establishment requires a large area. As a result, allocation of mariculture zone space, if there is such zoning, for local community within an MPA might be substantially reduced. Another concerning effect of mariculture legalisation for MPAs is shifting livelihoods, not only from fishing but also from other occupations, to fish farming. As more people become involved in mariculture, the CC of areas within the MPA mariculture zone will decrease. For instance, Dung (2009) reported not only a decline in mariculture production, but also biodiversity, in Nha Trang Bay MPA in Vietnam, due to an excessive number of cages surpassing the estimated CC of the MPA area. The protected ecosystem also faces higher pressure through increased use of wild seed for stocking cages, effluent from production, reliance on trash fish for feed, harmful chemical substances, and the possibility of disease and genetic transfer (Dempster et al., 2006, De Silva, 2012). Monitoring mariculture practices will not be effective because the task of monitoring mariculture permits (SIUP and TPUPI) has changed hands from the extension office (DKP) to the local MPA authority. Currently, the ratio between local MPA authority staff (national and local) and the total conservation area is 1 inspector per 300 km2 (MMAF, 2014b). Such a ratio makes enforcing compliance and providing support to farmers ineffective. It is also questionable whether these inspectors have adequate qualifications to ensure mariculture activities follow the regulations. Such disparities mean that the local MPA authority will

55 likely focus on the main priority, which is prevention and prosecution of MPA violations, such as illegal fishing and no-take zone intrusions, of which the inspectors are exclusively trained to do.

3.4.3 Implications of the regulations for mariculture zoning and establishment in Indonesia’s MPA

3.4.3.1 Changes in the development, management, and conservation of resources, goods and services. The use and ownership of seascapes and available resources in Indonesian coastal waters are considered open access (Adhuri et al., 2015). However, the aforementioned regulations (shaded regulations in Table 3.1) will limit the use and ownership of seascapes and the development of new business related to mariculture. A steady increase of small-scale fish farmers was predictable, as mariculture activity has been legalised and small-scale fish farmers are only required to report activity within the MPA SFZ. With successful mariculture activity, more area within SFZs is needed, as shown in the case of seaweed farming (Sievanen et al., 2005) and finfish culture (Soemodinoto et al., unpublished) in Indonesia. Only having to file a TPUPI essentially gives small-scale fish farmers more control to expand their activity within MPAs. Such rights will eliminate one of the constraints, access to water bodies, usually faced by poor small-scale fish farmers (Edwards, 2000). The rights of small-scale fish farmers to use SFZs for mariculture activities is controlled by some overly generalised limitations, such as the 50% CC threshold for any type and scale of mariculture. Although it looks like a promising measure to prevent excessive mariculture development, it is based on the assumption that mariculture activities are distributed evenly across a SFZ. A homogenous distribution of mariculture activities, such as floating net cages, within SFZ zone will avoid over-accumulation of farm wastes on the seabed. However, this is not the case as mariculture activities, especially small-scale, are usually clustered within an area (Alongi et al., 2009). Clustered mariculture activities, within the SFZ, might exceed a localised 50% CC without violating the overall 50% SFZ CC. Clustered mariculture might threaten seabed biodiversity directly under the platforms and, over time, the surrounding healthy coral reef, if it goes unmonitored (Pomeroy et al., 2006).

Small- and medium-scale fish farmers having reserved rights to conduct mariculture within MPA SFZs may further obscure the abstract administrative boundaries between villages, sub-districts, and districts. Medium-scale mariculture operators will likely benefit due to accessing other SFZ easier, as SIUP concession is granted by the central government rather than by the local government. A SIUP for medium-scale fish farmer will guarantee a 20-year lease for a particular area of seascape. Naylor et al. (2000) termed it privatisation of common resources in mariculture. This privatisation will cause small-scale fish farmers, in some villages, to have less area of SFZ for their mariculture. Small-scale fish farmers could use other villages SFZ to establish mariculture activities. However, this option would be less preferred due to security and access issues, as discussed earlier. In the end, conflicts over space amongst small-scale fish farmers, and between small-scale and medium-scale fish farmers, could arise. There is a chance that conflicts can escalate over village boundaries at sea, not only between fishers but also villages, which is currently an important issue in coastal zone management in Indonesia (Adhuri, 2005, Karin, 2015). Another possible implication of these regulations is the influx of products to support mariculture in MPAs, such as seed supply, feed and fish health medication. The sale of these products has grown into a profitable business in Indonesia. For example, there was a twofold increase in feed production between 1999 – 2013 (Laining and Kristanto, 2015), which could increase to 2.4 fold by 2030 (Tran et al., 2017). A similar trend is also shown for fish health products, with a 106% increase of registered products annually (DJPB, 2016). Efforts by MPA authorities to overcome these growing trends through advocating closed cycle mariculture and locally-bred seed have shown little success due to high capital investment in broodstock and seed production. TNC (2003) reported that a closed cycle mariculture project consumed 25% of the total conservation expenditure in Komodo MPA and was deemed the most expensive alternative livelihood project. In fact, this situation has been happening particularly due to increasing seed supply to small island areas, including small island MPAs, from seed centres such as Bali and East Java (Figure 3.2). While formulated feed can benefit an MPA, as less locally-sourced trash-fish will be used, imported seed and excessive use of fish health products will serve as additional challenges for MPA authorities.

Anambas MPA

Figure 3.2. Supply chain of hatchery-bred seed of finfish in one Indonesian MPA (Map source: Effendy, 2012)

3.4.3.2 Changes in the organisation of work and production

The additional task of managing SIUP and TPUPI, mandated by MPA regulations, particularly MMAF regulation No. 47/2016, will force local MPA authorities to seek cooperation from the local governments. However, establishing cooperation with local governments, specifically district governments, poses additional challenges, as local governments tend to prioritise economic development for their community. For example, there was a significant increase in the number of POKDAKAN (Kelompok Pembudidaya Ikan/Fish Farmer Groups) established by the local government and financed by MMAF. In some small island MPAs, the number of fish farmer groups increased up to 44%, such as Raja Ampat and Aru Archipelago MPAs, as an effort to develop small-scale mariculture (MMAF, 2014a). Sembiring et al. (2010) also reported that issuing SIUP and TPUPI for fishers and fish farmers remain conflicting interests between the local MPA authority of Taman Nasional Teluk Cendrawasih and the local government, to form a strong MPA collaboration. The strong evidence of re-centralisation of mariculture licensing in MPAs within these regulation provisions might lead to further division between central and local governments.

The current MPA regulations require small-scale fish farmers to personally visit the local MPA authority office to register their TPUPI. This task was previously carried by local government officers, who regularly visited farms to register TPUPI or collect production records. The re-centralisation of this task might reduce small-scale fish farmers willingness to report mariculture activities and fish production to local MPA authorities, due to associated cost and time. This potential increase of non-compliance has been identified, where reported production of farmed reef fish, for live sale, in Indonesian MPAs, is far less than what is distributed at the market (Soemodinoto et al., unpublished, Glaser et al., 2015). For example, this non-compliance led to low fish farmer community participation and interaction in MPA collaborative management, due to increased obligation without economic gain (Sangadji et al., 2015).

3.5 Discussion: improvement and alternative policies of the regulations

According to Gil (1973), in policy analysis, alternative policies can have the same objectives as original policies, but employ more effective or efficient measures to achieve the objectives; or have different objectives and measures as a substitute for the original. The regulations examined in this chapter clearly showed insufficient details covering the subject of mariculture management in Indonesia MPAs. Thus, it will be unfeasible to amend these regulations by purely developing efficient and effective alternative policy measures (the first approach), as suggested by Gil (1973). Instead, the following sections will discuss several recommendations and alternatives for new policy objectives and policy measures for mariculture in Indonesian MPAs. These recommendations and alternatives cover sustainable mariculture development in MPAs, local community-based mariculture permits, equitable mariculture MPA zoning, and collaborative implementing organisations.

3.5.1 Sustainable mariculture development in MPA

The messages contained in all MPA regulations, concerning the development of mariculture within MPA SFZs, are currently dominated by the assumption that mariculture is a potential threat, and thus overshadows its importance to local MPA communities. Requirements set to limit the development of mariculture, particularly, signify this assumption. One improvement that can be made is the allocation of mariculture zones within SFZs. Specifically, the regulations should state that local communities whose

59 traditional fishing areas have been designated no-take or other restrictive MPA zones are given rights to small-scale mariculture zones. The legalisation of small-scale mariculture zones, as one of the zoning system within these regulations, should have several objectives. Firstly, that mariculture zones will be one of the mandatory zones in an MPA, and not an optional zone which is currently embedded within the SFZ. During the designation of MPA zoning, procedures for assigning a mariculture zone will be as rigorous as other zones, which require substantial information. For example, MMAF must conduct a prerequisite rapid ecological assessment (REA) to determine the best locations for no-take zones before an MPA can be formally designated. With this new measure, not only potential locations for no-take zones will be surveyed, but also locations for mariculture zones. Secondly, that legalisation of small-scale mariculture zone will strengthen local MPA community rights, as the owners are responsible for sustainably managing the areas. Many MPAs have been successfully managed by local communities, such as in the Moheli (Comoro Island) and US Virgin Island MPAs, who limit fishing activities of non-local fishers and obtain leasing fees to finance local community needs (Mascia and Claus, 2009). This is particularly important in limiting the development of medium-scale mariculture, owned by private business from outside the local MPA community. Non-local fish farmers should have to go through a lengthy process and get local community consent to secure a lease within a mariculture zone reserved for local people. This consent will also be beneficial, as decision making regarding the lease transfer to medium-scale mariculture operations will not be dominated by community elites. Thirdly, requirements for developing mariculture within MPA SFZs need to be clearly elaborated. Instead of vaguely limiting the types of fish that can potentially change or damage the surrounding ecosystem, several replacement measures to address the possible impacts of mariculture (as described in Chapter 1, section 1.4.1) could be included as follows: 1. Fish species used in MPA mariculture activities are only those found in, or native to, the local MPA ecosystem. 2. Fish seed/fingerlings to be used in MPAs should be primarily produced using a closed cycle system and broodstocks from the local MPA.

3. For point (1); Fish seeds/fingerlings can be externally sourced if they are produced by nationally-certified hatcheries using genetically unmodified fish, including non-hybrid fish broodstocks. The rigid requirement to use natural and patented or registered formulated fish feed, produced mainly by big feed manufacturers, is nearly impossible for small-scale fishermen. Formulated feeds are expensive and supply to remote MPAs is uncertain, as well as requiring special storage maintenance once purchased. Currently, to the best of my knowledge, there is no certified, registered natural feed for the grow-out phase of mariculture. This requirement can be improved by the following new measures: 1. Farm-made (trash-fish, dry and moist pellets) and industrially-manufactured feeds (sinking and floating/slow-sinking) can be used by small-scale fish farmers in MPAs. 2. Medium-scale fish farmers are required to provide proof of purchase, and use, of industrially manufactured floating/slow-sinking feed for at least 40% of the total feed used. If the number of wild fish associated with net cages in an MPA is double to what Sudirman et al. (2009) predicted in non-MPAs, and they consume at least 40% uneaten feed, the use of 40% floating/slow-sinking feed, with 40% efficiency (Chu, 2002), can reduce feed waste to almost half that of trash fish.

All in all, the consideration and development of mariculture in MPA should be dependent on several reasons which have to be included in the regulations as new measures, as follows:

1. The inclusion of mariculture within the MPA zoning process is primarily to address the legalisation of mariculture activities by local small-scale fish farmer communities. This measure will secure local community rights over the mariculture zone and ensure continuous support of conservation efforts. It will also limit the expansion of medium- scale mariculture, by large capital investors, which usually ends up in seascape privatisation and loss of seascape access for local communities (as discussed in subsection 3.6.4). 2. The introduction of mariculture into MPAs with no existing mariculture should be based on practical culturing technology, segmentation of fish culturing businesses, and compatible with the social and cultural condition of local MPA communities. These

requirements will be vital to ensure that the activity is economically and environmentally sustainable (Pomeroy et al., 2006).

3.5.2 Improved mariculture permit in MPAs

Reappointing the DKP District as issuing organisation for MPA mariculture permits (SIUP and TDKPI) will simplify the process of registering and collecting production data for mariculture in MPAs managed by central and provincial governments. This task can be handled by the sub-division within the DKP, which is the Aquaculture Division. This division usually has staff and contract employees who visit fish farmers on regular basis to collect mariculture data from fish farmers operating within the district area. This arrangement will allow local MPA authorities to concentrate on monitoring and enforcement of MPA regulations within protected areas. Specifically, for SIUP, the role of DKP in issuing the SIUP for medium-scale mariculture must be shared with the central government, represented by DGA, and the provincial government, represented by DKP Province. In this arrangement, DGA or DKP Province will issue the final approval of SIUP, and DKP District will serve as the first approving and intermediary party. Arguably, these new measures will alleviate conflicting interests between district, provincial and central governments in the management of mariculture activities, and potentially other economic activities, within Indonesian MPAs. Improving the mariculture permit framework in MPAs should include local community approval when issuing a SIUP for medium-scale mariculture. Getting formal consent from small-scale fish farmers before SIUP approval is important to formalise their right over the MPA mariculture zone and prevent misuse of power from local elites (see subsection 3.6.4). However, getting local community consent over leasing areas within protected areas is logistically difficult, as customary rights over an area are unclear most of the time (Wily, 2011). Assigning mariculture zones within the SFZs to certain small-scale fish farmer groups (based on the village) will solve this logistical issue, as the consent process will involve only a limited number of local community groups. The small-scale fish farmer groups will only be responsible for approving the leasing proposal by medium-scale fish farmers, while the government will determine the feasibility of the planned activity, including an impact assessment. Medium- and small-scale fish farmers will benefit from

62 this arrangement because they will have a clear direction for negotiation. It is also easier for the government offices responsible for issuing SIUP to channel leasing fees to small-scale fish farmer groups as an exchange for their mariculture zone rights. Most importantly, giving local community rights over specific areas within MPAs will increase their sense of ownership of the overall MPA, and encourage them to manage their activities and protect the area from outsiders (Andrade and Rhodes, 2012). Another new measure in the establishment of medium-scale mariculture in MPAs is the obligation of medium-scale mariculture to conduct an environmental impact assessment (EIA), approved by the Ministry of Environment and Forestry (MEF) (see Section 3.5.3). This should be strictly imposed to medium-scale fish farmers who are allowed to have concession areas of >0.5 ha for hatchery and >2 ha for grow-out). This three-party involvement will eventually discourage private business entities, with strong capital, developing mariculture in MPAs using medium-scale mariculture permits. Therefore, a change in the scheme of mariculture permits for both small- and medium-scale mariculture activities is suggested (see Figure 3.3).

This section recommends improvements that could be included in the regulations or future strategy in establishing mariculture zones within MPAs. Starting with improved policy goals, new policy should clearly prescribe mariculture zones in an MPA SFZ are created based on integrated mariculture site selection. Integrated site selection should consider: the biological requirements of any potential, or existing, cultured fish species; the existing, and future, user needs of the SFZ seascape; the visual amenity of the area as an MPA; local community preferences of the mariculture zone location; and the existing locations of established/planned MPA zones. Furthermore, the regulations should specify that only suitable areas, determined from an integrated site selection process, are designated as mariculture zones. The integrated site selection process should be scientifically designed to suit the MPA conditions, which will be developed in Chapter 6 of this thesis. This measure would eliminate the assumption that mariculture can be undertaken in all SFZ areas. Individual mariculture zones will be easier to manage in terms of determining CC, rights of use by local small-scale fish farmers, and issuing mariculture licenses for medium- scale fish farmers. This new policy goal is in line with the FAO ecosystem approach to

63 aquaculture (EAA) that integrates aquaculture with the ecosystem, and serves as a baseline for individual mariculture permits and operations (Ross et al., 2013b).

Fish Farmer Applicants

Medium-scale Small-scale

Foreign SIUP TPUPI/TDK Invest. PI RKIPM DKP District Small-scale Approval/ Data MZ - 1 Formal Consent AMDAL Medium- MZ - 2 scale Small-scale (EIA) Group MEF MZ - n National LM-MPAs MPAs National Provincial DGA - MMAF DKP MPA MPA Provincial Authority Authority

Yes Yes No No Approved SIUP Existing regulation for mariculture permit in MPA : Proposed new measures for mariculture permit in MPA :

Figure 3.3. Existing and proposed measures for mariculture permits in MPAs

The new policy should also include a revised CC concept for the mariculture zone. The 50% limitation for both the number of cage units and total SFZ area has not been clearly defined or any scientific basis. Here, it is proposed that the CC should be determined for individually mariculture zones, based on the capability of the proposed zone to cope with potential effects of mariculture activities. Such an arrangement will prevent the risk of mariculture operations exceeding the CC of small areas, because the activities 64 have not exceeded the 50% CC of the SFZ. For example, clustered mariculture activity, within a small area, might potentially exceed the CC of the area but not the upper limit of the whole SFZ CC. On the other hand, the new CC policy will optimize the use of space within a more regulated mariculture zone. Several new measures should be introduced to anticipate the change in policy objectives. These include standardised methods for determining site selection, CC, as well as mariculture zone boundaries. Firstly, variation and prioritisation of site selection and CC for mariculture will depend on the species selected, technological requirements, data availability, and social conditions (Longdill et al., 2008, Silva et al., 2011, Meaden and Aguilar-Manjarrez, 2013, Ross et al., 2013a). These differences are described in the work of Ferreira et al. (2012) and Ross et al. (2013b), based on regional classification in the form of different pillars, namely physical, production, ecological, and social site selection and CC (Figure 3.4.a). For example, site selection for shellfish requires a different approach compared to finfish. Shellfish mariculture site selection can prioritise increased production while minimising significant effects on visual amenity or space use conflict. Site selection for finfish, on the other hand, might have to prioritise the protection of MPA ecological values due to the potential impacts of finfish culture, such as excess waste or visual amenity disturbances. The new measure proposed here are that site selection and CC analysis for mariculture in MPAs prioritise ecological functioning first. Subsequent prioritisation can then focus on production and the physical requirement of mariculture activities. The priorities mentioned above are usually tasked to a group of analysts and involve data gathering, modelling, and quantitative analysis in which local MPA communities are regularly absent. Applying only these prioritisations would be incomplete, as MPA access has changed from an open to a regulated regime, in which decisions over resource allocation and protection involve many different stakeholders. Hence, the final best location for mariculture in an MPA should be decided based on a common agreement between stakeholders (social site selection/CC), as the manifestation of principles in Indonesian MPA management. In this case, the social site selection/CC should be conducted as the final stage, separate from other site selection/CC stages that may have already been conducted (Figure 3.4.b).

US, Canada, Europe Latin America Africa Asia, China MPAs

Physical Physical Physical Physical Physical

Production Production Production Production Production

Ecological Ecological Ecological Ecological Ecological

Social Social Social Social Social (a) (b)

Figure 3.4. Comparison of prioritisation of mariculture site selection/carrying capacity for (a) different regions (Ferreira et al. (2012), Ross et al. (2013b) and (b) the study’s proposed site selection/carrying capacity in MPA

Secondly, the standardisation of boundary markers for mariculture zone using either physical or virtual signs is advised as another new measure. Current practice in Indonesian MPA boundary marking only focuses on delineating core zones to other areas. For example, a technical zoning boundary guideline issued by DKKJI (2013b) only discusses the no-take zone boundary in entire MPA, and only uses natural and man-made markings. The guideline does not describe zone boundary markings for the SFZ, use and rehabilitation zones, which has led to confusion among local MPA communities and possibly increased non-compliance and violations. Walmsley and White (2003) found that simple MPA management measures using different boundary markings, differentiating different zones, led to increased compliance by locals and improved biodiversity recovery rates. Therefore, it is proposed here that, boundary marking for mariculture zones, using specific man-made or natural markers, should also be included within regulations and applied in the zoning process of MPAs. It serves not only as visible markers for how far mariculture can be undertaken, but also as a sign of small-scale fish farmer ownership rights to conduct their activity and bargaining power with other MPA seascape users.

3.5.3 Collaborative implementing agency/stakeholder

An overview of proposed collaborative work among responsible institutions or stakeholders is presented previously in Figure 3.3 and briefly discussed in Section 3.8.2. The underlying concept of this collaboration framework is the empowerment of small-scale

66 fish farmers in the ownership of seascape rights. The framework also supports the re- decentralisation to shift responsibility back to District DKP to manage mariculture development in MPAs, both for small and medium-scale. The re-decentralisation of mariculture management responsibility to DKP District is in line with the argument by Papoutsoglou (2000) and Read and Fernandes (2003). They argue that transferring the responsibility of mariculture management to local government will ensure consistent and efficient control of the development. Such a measure will remove the local governments autonomous jurisdiction, due to its primary function as the leading agency, to over- prioritise mariculture development in MPAs, and involve other institutions in the decision- making process. The transfer of small-scale mariculture management from the central and provincial government to the district government will effectively cause Law No. 23/2014 obsolete and arguably needs to be revised. Interestingly, such, if to occur, revision based on transfer of authority to the district government will increase the applicability of Law No. 7/2016 to cover MPAs despite not intended. As such, the right of small-scale fish farmers in small- island MPAs will be more protected and the management of their activity is easier to monitor by the district government in collaboration with the central and provincial government. In addition, this collaboration concept also defines the MEF responsibility in the development of mariculture (medium-scale), previously not recognised in MPA regulations. The MEF plays a similar role, in MPA mariculture development, to the EIA practice developed in Scotland for marine farm development in sensitive areas or areas greater than 0.1 ha (Read and Fernandes, 2003). Sharing EIA responsibility with the MEF will greatly reduce the DGA liability for single-handedly overseeing the task. The proposed collaboration between these institutions, with the District DKP holding the central role, in MPA mariculture development, is similar to what has been proposed by Lauer et al. (2015) regarding the legislative process of mariculture in South Australia. They argue that a legally binding collaborative governance framework, involving responsible institutions, community participation, and scientifically-informed decision making, will lead to better management of mariculture and reduced conflict over resource

67 use. In particular, the governance framework will also reduce the long-strained relationship between central and local governments in the overall management of MPAs.

3.6 Conclusions

Mariculture is recognised as one of several alternative livelihoods for local communities living within or adjacent to MUMPAs in Indonesia. Despite the economic and cultural importance to local communities, existing Indonesian MPA regulations have not yet provided sufficient detail on how to sustainably manage the development of mariculture in MPAs. Currently, MPA regulations allow mariculture activity to be undertaken within SFZs, with very general prescription on how, where and the extent of the activity. This creates a blurry line for small-scale fish farmers regarding their rights to legally use MPA coastal waters, and the extent to which their activity might impact MPA biodiversity, and thus deeming mariculture unsustainable and damaging to the goals of marine protection. The aggressive development of aquaculture in Indonesia has attracted considerable attention from private business entities, such that MPAs, with their optimal environmental conditions, can be targeted for mariculture expansion. This move is clearly presented in the set of regulations, discussed previously, through opening up SFZs for medium-scale mariculture activities. To address the issues mentioned above, regulation objectives need to change to accomplish Indonesian MPA principles, as well as the FAO EAA (equity, power-sharing, participation, and sharing benefits). The designation of MPA mariculture zones reserved for local communities, or fish farmer groups, will increase local participation in MPA management, as the right to the allocated zone bears the responsibility of sustainable use. It will also be easier for local MPA authorities to monitor the development of mariculture activity within separated mariculture zones, instead of monitoring the whole SFZ area. Similarly, it enables better management in the absence of EIS requirements for small-to- medium scale operators. Consequently, the benefit of mariculture zone allocation is that medium-scale mariculture expansion can be effectively controlled through formally requiring local community/fish farmer group consent for mariculture permits/leases. The designated mariculture zones will also increase transparency and effectiveness of issuing mariculture permits, especially for medium-scale mariculture activities. The

68 strategic partnership between responsible institutions, including participation of small-scale fish farmers, ensures that requirements of mariculture licensing to make it sustainable is handled by an appropriate institution and within a clear decision-making framework. The involvement of MEF to oversee the EIA for medium-scale mariculture operations, and the leading role of district DKP of mariculture permit decision-making, are two important improvements that have been missing in the existing regulations. The implementation of adequate site selection and CC for mariculture in small island MPAs is different to other areas or regions due to the MPAs location, environmental sensitivity, and cultured species. Due to the existence of MPA protected zones, and the favoured carnivorous live reef cultured fish, which are net contributors to ecological functioning, ecological site selection and CC should be prioritised above other aspects, such as production. Such planning tools also need to be underpinned by science and an understanding of physical, biological and human requirements and values for sustainable mariculture. Due to the importance of site selection and CC in MPAs, this has to be addressed by MPA Authorities prior to the establishment of an MPA, or after if local communities require the creation of mariculture zones in existing MPAs. All in all, it is important to note that decision making over resource allocations in Indonesian MPAs has been a tug-of-war between multiple stakeholders. Therefore, the final step of site selection and CC should introduce stakeholder preferences. This will ensure that resource allocation will be optimal between users (small- and medium-scale fish farmers) and minimise potential conflicts. Such an approach is in line with the three key principles of the FAO EAA (Ross et al., 2013b). The lead role assumed by district DKP means that the management of mariculture development in an MPA can be streamlined, including bridging workflow between local stakeholders and central government. This, of course, will not eliminate the roles of local MPA authorities, which can focus more on the aspect of surveillance and infringement prosecution within the wider MPA area. In addition, district DKP have to be prepared to deal with usually prolonged negotiation and conflict resolution that accompanies the process of resource allocation, as also occurs with mariculture zoning in MPA.

CHAPTER 4. Comparative Study on Sustainability Profiles of Different Small-scale Livelihoods within Small Island MPA-Designated Communities in the Anambas Archipelago-Indonesia

4.1 Introduction

Chapter 1 provided a discussion regarding the development of mariculture within Indonesian MPAs, and the challenges for mariculture to support small-scale fish farmer groups and conservation efforts. Despite the development of mariculture being formalised within Indonesian regulations (see Chapter 3), its sustainability remains in doubt compared to other livelihoods, such as eco-tourism or small-scale fisheries. This has impeded the development of mariculture in MPAs, diminishing its function of improving welfare for small-scale fish farmer groups, and the overall local MPA community. Fauzi and Buchary (2002) argue that overcoming poverty and marginalisation of local coastal communities will determine the success of any MPA management in Indonesia. Christie (2004) noted that a lack of support from local communities undermines biodiversity conservation through MPA regulations in Indonesia, via low compliance and high disregard for the regulations. On the other hand, the GoI, through the MMAF, argue that designation of MPAs in Indonesia should not dismiss traditional rights of the local community. In fact, local community rights for artisanal fishers, fish farmers, and ecotourism operations are recognised fully in the MPA strategy (DKKJI, 2013a). Unfortunately, literature discussing the relationship between sustainable livelihoods and MPAs has focussed on sustainable capture fisheries management (Persoon et al., 1999, Fauzi and Buchary, 2002, Christie, 2004, McLeod et al., 2009, Adhuri et al., 2015) and ecotourism (Walpole et al., 2001, Fauzi and Buchary, 2002, Cochrane, 2006). Very few studies are available regarding the sustainability of fish farming or mariculture in Indonesian MPAs, particularly multi-purpose MPAs where economic activities can be carried out, albeit under controlled conditions. This is partly due to the perception that mariculture is not sustainable and may degrade MPA environments (Dempster et al., 2006). Further investigations challenging this claim by showing mariculture can be a sustainable livelihood, and environmentally friendly, for MPAs in Indonesia, have yet to be conducted. The complexity of the situation is exacerbated by the fact that more than 90% of Indonesian MPAs are Multi-Use MPAs (MUMPAs) and located around or near inhabited

70 small islands (DKKJI, 2013a). These local small island communities are highly dependent on coastal and marine resources and get less support from higher levels of government due to their remoteness (Abecasis et al., 2013). Access to markets (Abecasis et al., 2013), infrastructure, and land-based livelihoods (Szuster and Albasri, 2010, Ferse et al., 2014) are also limited because of the distant location and scarce land-based resources of small island communities. Therefore, neglecting livelihood diversification for small island communities, such as small-scale mariculture, will only reduce their resilience and increase existing vulnerability risks. This, in turn, will hamper the progress of MPA development, because local communities will continue to exploit nearby protected resources, or even challenge the existence of an MPA (Bennett and Dearden, 2014, Christie, 2004). The goal of this chapter is to investigate the relative sustainability of the three most common livelihoods in small-island MPA communities in Indonesia, with a particular emphasis on small-scale mariculture. This study aimed to enrich the literature by using a modified Sustainable Livelihood Approach (SLA) to study coastal community sustainability. The modification of the SLA proposed by Scoones (1998), DfID (1999) and Ellis (2000) was using capital assets and sustainable indicators (SIs) to compare the relative sustainability of small-scale fish farming, fishing and ecotourism livelihoods. The aim was also to demonstrate that small-scale mariculture, along with small-scale fisheries and ecotourism, contributes to the livelihood diversification of small island communities within or adjacent to Indonesian MPAs, and is compatible with overall MPA objectives.

4.2 Methods

4.2.1 Study area

The study area encompasses the Indonesian Anambas Archipelago, a small island cluster MPA located in the south South China Sea, comprised of 46,664.15 km2 (1.5% land and 98.7% territorial sea) (Saidah, 2011) (Figure 4.1). The archipelago was fully designated as a marine nature recreational park MPA in 2014, under the managing authority of MMAF. Administratively, the area consists of 34 villages, distributed across seven sub- districts, with a total population of over 44,288 people (BPS-Anambas, 2012). Due to its remoteness and limited land mass, the Anambas populations rely heavily on coastal and marine resources. More than 30% of total households in the Anambas

Archipelago work in various small-scale fishing activities (BPS-Anambas, 2012). With 20% of the total population living in poverty (BPS-Anambas, 2009) and 38% unemployed (BPS-Anambas, 2012), livelihood diversification plays a key role in the economic development of Anambas.

Figure 4.1. Study area and surveyed villages in Anambas Archipelago MPA

4.2.2 Data collection

The primary data (quantitative and qualitative) used in this study were collected via face-to-face interviews with respondents, whilst secondary data were collected from written and online literature. Sixty-five people responded, from households involved in various livelihoods (Table 4.1). Stratified random sampling with proportional allocation was used to select respondents and to address the uneven sample number from household groups. The respondents from the household groups were selected if they were categorised as

72 small-scale. The concept of small-scale could vary from artisanal, local, traditional, and small to subsistence, and each region or country has its own definition of each term (Natale et al., 2015). In order to maintain the consistency of data collection, the characteristics of respondents from each small-scale household group followed the Indonesian Law No. 50/2015 regarding the empowerment of small-scale fishers and fish farmers. The law defines: small-scale fisher as an individual who owns or operates less than 5 GT boat to catch fish as his/main livelihood to meet his/her daily needs; small-scale fish farmer as an individual who cultures marine finfish as his/her main livelihood to meet his/her daily needs. Additional characteristics of small-scale fish farmers were used based on Afero et al. (2010) where the number of net cages owned was less than 15 pens) and minimal production input (i.e.: use locally caught trash fish, low stocking density, no contract labour). The selection of respondents of small-scale ecotourism operator/s (hereafter called ecotourism or ecotourism operator/s) was based on the Law No. 10/2019 regarding tourism, Cater (1993), and Donohoe and Needham (2006). All respondents from ecotourism operator household were the local owners of micro and small-scale tourism businesses operating in Anambas tourism areas such as small restaurants, home stay, traders, and traditional tourism boat. The field visit to collect data lasted for over 4 months, from July to November 2014. The surveyed villages were selected to represent the five main/important islands (Siantan, Jemaja, Telaga Kiabu, and Matak islands) in the Anambas Archipelago, main settlement of populations, and sub-district administrative boundary. This research has acquired human ethics approval from the UNSW Human Research Ethics Committee #HC14191 prior to fieldwork and was considered “low impact” research on human subjects.

Table 4.1. Distribution of respondents across villages and household livelihoods in Anambas sub-districts.

Surveyed Number of respondents No. Sub-districts Villages Fisher Fish farmer Ecotourism 1 Jemaja 1 1 0 2 2 Jemaja Timur 1 2 2 1 3 Palmatak 3 12 7 0 4 Siantan 2 5 0 1 5 Siantan Selatan 3 12 3 1 6 Siantan Tengah 3 7 5 0 7 Siantan Timur 2 4 1 0 Total 15 43 18 5

4.2.3 Questionnaire

The questionnaire used was based on a set of pre-determined SIs, where each SI was grouped under corresponding capital assets (Table 4.2). All of the SIs followed the ensuing criteria: (1) exist in all three livelihood capital assets being compared, (2) quantitatively can be determined, and (3) can be standardised based on functional relationships. The questionnaire consisted of 75 closed- and open-ended questions, grouped into sub-sections, based on demographic information and the five capital assets of SLA (Appendix 1). All questions regarding the quantifiable five capital assets were designed in either of Likert- scale and Likert-type scale with multiple mixed point scales (3 and 5 points scales).

Table 4.2. Sustainable indicators (SIs) of capital assets used in the study

Human capital Financial capital Natural capital Physical capital Social capital  Education  Main livelihood  Access to  Boat ownership  Work (household income natural  Support for networking head)  Alt. livelihood resources physical asset  Quality  Education income  Conflict over support from (dependent)  Working hour resources formal  Number of  Livelihood institutions dependents) diversification  Working  Savings experience  Income stability  Input production  Household expenses (%)

The semi-structured questionnaire, with open-ended questions, was designed so that responses could be interpreted into meaningful qualitative data. This qualitative data was 74 used to highlight relationships and interactions between capital assets that the quantitative data could not capture.

4.2.4 Data analysis

4.2.4.1 Quantitative data

A combination of Microsoft Excel® 10 and SPPS® 24.0 was used on quantitative data for transformation, standardisation, aggregation, and frequency distribution analysis. Due to different units of measurement for quantitative data, data analysis was performed across all SIs through the following stages:

4.2.4.1.1 Data transformation As suggested by Erenstein et al. (2007), Ranganathan et al. (2009), Erenstein et al. (2010) and Ko (2005) all data for individual indicators were transformed into simple indexes using linear transformation. This standardisation produced values from 0 - 1 for all indicators that maintain similar spread and equal weighting (Erenstein et al., 2007). This was applied to all SIs except a single physical capital, boat ownership, which received a specific value judgment weighting to accommodate different types of boat ownership (Table 4.3).

Table 4.3. Preference of weighting in different types of boat ownership indicator

Sub-indicators Weight Motorised Motorised 1 Non-motorised 0.5 Boat size >100kg 1 100kg-1 GT 2 1- 2 GT 3 5 GT 4 Property Ownership 1 Non-ownership 0.5

4.2.4.2 Individual SI composite index The second stage involved calculating individual composite indicator index by aggregating every SI of each respondent, based on livelihood groups. These indexes were used to compare the relative sustainability level between household groups directly from each SI.

4.2.4.3 Capital asset composite index All SI composite indexes grouped into a corresponding capital asset were aggregated to form a single index, the capital asset composite index, representing the overall achievement of each household group. These index values were then plotted into the 5-point scale sustainability polygon following Ko (2005), which categorised sustainability levels into bad (0-0.2), poor (0.21-0.40), medium (0.41-0.60), good (0.61-0.80) and sustainable (0.81-1). The sustainability polygon served as a means of visually interpreting and comparing the relative sustainability of livelihood groups.

4.2.4.4 Qualitative data

Thematic analysis was used in this study to distinctively characterise the three household groups accessing, and making use of, available capital assets under their control. Although the SLA’s capital assets and vulnerability context could be used as predetermined themes to analyse the interview data, a data-driven inductive approach was selected (as suggested by Boyatzis, 1998, Fereday and Muir-Cochrane, 2008). This approach started by determining any possible themes within the interview texts, during the interview transcribing process, and re-evaluated them by rescanning the text to find other possible themes or replace previously identified themes with the new ones. Sixty-five interview recordings, totalling 3,668.45 mins in length, were converted into the verbatim format and NVIVO 10 was used to analyse the texts. Through the transcribing process, and re-examination of the text to evaluate themes, 41 themes were identified, containing 181 nodes corresponding to responses.

4.3 Results

This section presents the indicator composite and overall capital indexes. This section also presents the qualitative analysis results, to complement the quantitative analysis, on how the respondents from each household group used or accessed available capital assets.

4.3.1 Human capital

There is a noticeable difference between human capital index values among the studied household groups (Table 4.4). Across the four SIs measured in human capital, fish

76 farmer households showed the highest index values for the level of education among the household heads. On the other hand, ecotourism households had the highest index value for the level of education among household members. These indexes showed that the high level of education among household heads has a positive relationship with the education level of their dependants. Fisher households scored the highest index values for the working experience indicator, despite being the lowest for both education indicators. The three household groups had a similar index value for the average number of dependents. The overall composite index for human capital was highest for fish farmer households, followed by small-scale fisher and ecotourism households (Table 4.4). The levels of education and work experience were the main indicators driving this pattern.

Table 4.4. Comparison of human capital SI indexes of three different household groups

Household group No. Sustainable indicators Fisher Fish farmer Ecotourism 1 Level of education (head) 0.18 0.41 0.27 2 Level of education (dependents) 0.27 0.42 0.43 3 Number of dependents 0.50 0.50 0.51 4 Work experience 0.84 0.70 0.27 Composite human capital 0.45 0.51 0.37

The analysis of qualitative data showed that more than 55% of fisher household heads had no formal education, nor finished elementary school. Fish farmers and ecotourism operator household heads achieved higher educational levels, where an average of 75% finished at least elementary school. Factors resulting in low levels of education among household groups, particularly heads, were: poor economic conditions; lack of school facilities, especially for middle and high schools; and no immediate benefit from attending school. In terms of work experience, most of the ecotourism households were those who recently shifted from other occupations. Their work experience was significantly low compared to fisher and fish farmer households. Less than 20% of ecotourism household respondents had work experience of more than 10 years. On the other hand, 84% of fisher and 72% fish farmer households, on average, had work experience of more than 10 years.

4.3.2 Financial capital

The overall composite indexes of financial capital among household groups did not differ significantly, although ecotourism households had the highest profile of sustainability followed by fish farmer and fisher households (Table 4.5).

Table 4.5. Comparison of financial capital SI indexes of three different household groups

Household group No. Sustainable indicators Fisher Fish farmer Ecotourism 1 Income from main livelihood 0.49 0.44 0.60 2 Income from alternative livelihood 0.09 0.46 0.60 3 Working hour 0.38 0.70 0.42 4 Livelihood diversification 0.40 0.46 0.35 5 Savings 0.16 0.29 0.72 6 Income stability 0.62 0.75 0.80 7 Input production (%) 0.49 0.53 0.48 8 Household expenses 0.69 0.70 0.73 Composite financial capital 0.42 0.54 0.59

Despite the three household groups having similar levels of income generated from the main livelihood, fisher households were classified as bad (unsustainable) for income generated from alternative livelihood activities (index value of 0.09). Both fish farmer and ecotourism households in this SI were classified as medium (intermediate sustainability) (index values of 0.46 and 0.60, respectively). Alternative income level differences are directly related to the amount of time each household group spends on their main livelihood. Fisher households devote more time to their main livelihood (working hour index value of 0.38), which then reduced their available time for other types of livelihood. Fisher households scored the lowest index (0.16) on the savings indicator (bad), ecotourism household scored the highest with an index value of 0.72 (good), and fish farmers were between them, scoring 0.29 (poor). On average, the input production indicator (ratio of income to income spent on input) showed that fishers, fish farmers, and ecotourism households spent 40.4%, 37.7% and 41.0%, respectively, of their income on subsequent production cycles, and were all classified as medium sustainability. However, fisher households consistently invested a higher amount of production inputs (ice, fuel, time) each day, during a year, to maintain their income level. Because of this spending, fisher households scored the lowest for the 78 household expenses indicator (0.69), while both ecotourism and fish farmer households scored slightly better (0.73 and 0.70, respectively). The results of the qualitative analysis showed that livelihood diversification is a strategy for household groups to cope with limited resources. The difference was that fisher households prefer to work more in construction projects funded by the government (17 respondents), compared to work in fish farming (12 respondents) or agriculture (12 respondents). Fish farmers prefer to work more in fishing activities (nine respondents) and local construction works (five respondents). Ecotourism households either worked in local construction or agriculture (two respondents). Ecotourism households had the highest index value in terms of the indicated amount of savings, whereas fisher households had the lowest. The results of the matrix coding comparison above showed some similarities for savings practices among respondents from all groups. There was fisher (20), fish farmer (8) and ecotourism (1) households that stated that they did not have any savings because their income was simply not enough.

4.3.3 Natural capital

Easier access to natural resources, either available or controlled, with less conflicts, meant Ecotourism households had the highest overall composite index for natural capital (categorised as good sustainability) (Table 4.6). On the other hand, both fish farmer an fisher households had similar lower values (categorised as medium and poor sustainability, respectively) as their livelihoods are heavily resource-based activities and influenced by persistent conflict with other users.

Table 4.6. Comparison of natural capital SI indexes of three different household groups

Household group No. Sustainable indicators Fisher Fish farmer Ecotourism 1 Access to natural resources 0.62 0.75 0.80 2 Conflict over resources 0.19 0.17 0.70 Composite natural capital 0.40 0.46 0.75

Although fisher and fish farmer households scored lower than ecotourism households for access to natural resource, both are still categorised as good sustainability. Lack of access was expressed by at least 41.9% of total fisher households who reported that

79 they could only catch fish up to 7 to 8 months per year in their fishing grounds because of climate and weather variations. Similarly, despite being able to culture fish all year round, fish farmer household suffer from frequent low water quality events, which cause mortality of the cultured fish. Conflicts over resources have been the primary concern for the three household groups in Anambas, although more so for fisher and fish farmer households. Fisher and fish farmer households regularly experience direct confrontation with the illegal fishers. For example, this study found that 41 fishers and 7 fish farmers experienced intimidation, fish aggregating device (FAD) and fishing ground overlapping claims, and harbour access. Based on the matrix coding results, the majority of respondents from fisher and fish farmer households (34 and 13 respondents, respectively) indicated a strong influence of climate and weather conditions on their livelihood. High waves and strong winds, that can last from three to seven months (September – February), restrict fishing activities in Anambas MPA. In calm weather seasons (March to June), fishers have to travel to fishing grounds farther away due to declining catch in nearshore fishing grounds. Fish farmers and ecotourism respondents were less restricted by seasonal weather. Fish farmer households cited the rainy season as the main limiting factor, as it triggered fish disease occurrences, while ecotourism household stated less customer visit during this period. High wind and waves did not affect fish farmers because their activities are located in relatively sheltered bays and straits.

4.3.4 Physical capital

Fisher households scored the highest composite index for physical capital (good) followed by ecotourism (good) and fish farmer households (medium) (Table 4.7).

Table 4.7. Comparison of physical capital SI indexes of three different household groups

Household group No. Sustainable indicators Fisher Fish farmer Ecotourism 1 Boat ownership 0.56 0.48 0.27 2 Support on physical asset 0.85 0.48 1.00 Composite physical capital 0.71 0.48 0.63

Within the capital asset, fisher households scored the highest index (medium sustainability) for boat ownership followed by fish farmer and ecotourism households, both categorised as poor sustainability. However, the result of the matrix coding comparison showed a contradiction regarding the high index of boat ownership, especially for the fisher households. Despite boats being accessible to fishers, more than 67% of fisher respondents stated that the boats were in poor condition, non-motorised, partially owned, and small. Fish farmers scored the second highest (medium) for this indicator, which implies the importance of a boat to support fish farming activities, such as catching seed and trash fish, as well as delivering the cultured fish to market. Ecotourism households scored the lowest (poor) for boat ownership, as their business operated on land. However, lack of boat ownership prevented ecotourism households from expanding their business into boat chartering and organised tours. The high index values for the physical asset index scored by the ecotourism and fisher households highlights the relative favouritism for these occupations in Anambas MPA. Ecotourism and fisher households receive most of the government endowments, such as tourism facilities, financial aid, fishing gear, and training. Fish farmers, on the other hand, receive less support in terms of physical asset provision, resulting in the low index value scored, compared to the other two.

4.3.5 Social capital

The overall composite index value of social capital showed that fisher households had the best score (good), followed by fish farmers and ecotourism households (medium) (Table 4.8).

Table 4.8. Comparison of social capital SI indexes of three different household groups

Household group No. Sustainable indicators Fisher Fish farmer Ecotourism 1 Work networking 0.74 0.69 0.70 2 Quality of support from formal institutions 0.47 0.41 0.30 Composite social capital 0.61 0.55 0.50

The data showed that each household group has developed a unique social network to manage their livelihood challenges. The relatively open and flexible social network

81 among fisher households tends to outperform the exclusive social network among fish farmers and the heavy dependence of ecotourism households on the government. Both fisher and fish farmer households scored higher (0.47 and 0.41, respectively) and were classified as medium sustainability. Ecotourism households scored the lowest index value for the quality of support from formal institutions and were classified as poor sustainability. The result of the matrix coding comparison revealed the reasons for the high index values under the social capital indicators among the households. The result showed that 28, 11 and 2 respondents from fisher, fish farmers, and ecotourism households, respectively, had developed different types of social safety net within their own group.

4.3.6 Overall composite indexes of capital assets

The composite index values for capital assets showed that all three household groups predominantly fell into medium sustainability. However, fisher and ecotourism households were classified into good sustainability on physical capital (0.71) and on natural capital (0.75), respectively. Ecotourism households scored the lowest capital index on human capital (0.37) and categorised as poor sustainability. Furthermore, the shape of overall composite index value in Figure 4.2 describes the main strength and weakness of each household group. Fisherman and ecotourism household groups have more skewed shape toward certain capital assets while fish farmer household group is more balanced.

0-0.2 = Bad (Unsustainable) 0.21-0.40 = Poor (potentially unsustainable) 0.41-0.60 = medium (intermediate) 0.61-0.80 = Good (potentially sustainable) 0.81-1.00 = Sustainable (excellent)

Figure 4.2. Capital asset polygon of sustainability livelihood approach from three different household groups

4.4 Discussion

The sustainability profile of the Anambas MPA household groups based on access to, or ownership of, capital assets is relatively similar in terms of the overall composite index. The three household groups are all categorised as medium sustainability across all capital assets. However, looking at the individual capital assets, livelihood portfolios of each household group show distinguishable differences (see Figure 4.2)

4.4.1 Sustainability profile of Anambas community groups based on human capital

In terms of human capital, the main factors that distinguish the sustainability profiles of the three livelihoods are the level of education (both head and family members) and work experience. Fisher households showed the strongest relationship between working experience and the level of education. Most of the fishers in Anambas started working at a very early age in order to support family livelihoods or sought income over education. Consequently, the head, and members, of fisher households opted to drop out or chose the lowest educational level. This finding was confirmed by the current level of education, for more than 60.5% of the total Anambas population, being equal to, or less than, primary school (BPS-Anambas, 2012). On the contrary, a higher level of education among fish farmer and ecotourism households might have influenced a shift to a more efficient livelihood. Both groups previously worked as fishers, and their willingness to engage in a new livelihood corresponded to their increased capability to search and learn new skills and information. This is in line with DfID (1999) and Ellis (1998), who suggested that formal education can influence people’s behaviour or creativity and lead to their involvement in more intensive and modern livelihood activities. Edmondson (1992) and Niehof (2004) also had a similar view, that in the long term, education plays an important role in livelihood diversification and guarantees better future income.

4.4.2 Sustainability profile of Anambas community groups based on financial capital

The levels of main income were generally similar between the household groups. However, a noticeable difference was observed in income generated from alternative livelihoods. The main income-generating activity for fish farmer and ecotourism

83 households were not time-consuming. They were able to develop and maintain consistent alternative livelihoods (fishing, agriculture farming, and construction work) and generate substantial additional incomes. On the contrary, fisher households were limited to fishing because this livelihood requires considerable time to carry out due to their conventional and small-scale nature. Consequently, for fishers, incomes generated from alternative livelihoods were insignificant due to a lack of time and resources. In the context of small-islands, fishers, fish farmers, and ecotourism operators develop different livelihood diversification strategies to cope with the geographically challenging conditions (Sievanen et al., 2005, Bene, 2009, Hampton and Jeyacheya, 2015). This study found that ecotourism and fish farmer households have a different strategy to maximise their livelihood economic return compared to fisher households. Fish farmers and ecotourism households developed an alternative livelihood as a trade-off livelihood strategy, as suggested by Scoones (1998). This means that both household groups trade off time, inputs and other assets to do alternative jobs (fisheries, agriculture and construction works) without leaving their main jobs. This livelihood diversification strategy allows them to undertake main and alternative livelihoods simultaneously. On the contrary, alternative livelihoods for fishers, such as fish farming, construction works, agriculture or ecotourism is a substitution livelihood strategy, as explained by Scoones (1998). In this sense, when access to the sea is limited, due to severe seasonal climate conditions, fishers shift completely to alternative livelihoods, with no income being generated from fishing until conditions improve. This livelihood adaptation strategy means fishers only carry out a single main livelihood at any given time, and have inadequate resources to diversify their livelihood activities. The inability of Anambas fishers to maximise livelihood diversification results in lower total annual incomes from alternative livelihoods. This has also affected fisher household financial capital indicators (savings, income stability, and household expenses). Fisher households in this study find it difficult to put aside part of their income during the productive fishing season, in the form of savings, to cover expenses during the low fishing season. As the result, fishers are highly dependent on informal moneylenders (their middlemen or regular moneylenders) to support their families and livelihood activities.

Informal loan schemes are very common for fisher and fish farmer households in Indonesia (Alexander et al., 2006), and in Anambas MPA, but not for ecotourism households. The strong patronage relationship between fishers or fish farmers with a big middleman, or ‘Toke’, in the Anambas MPA is evident through the development of an informal loan system. An informal loan has no interest, collateral or even time limit, being based on an unwritten agreement that the borrowers have to sell caught or cultured fish to the lenders at a lower price. Failure or refusal to do this might result in social punishment, such as being branded as untrustworthy among middlemen and difficulty selling fish to the market. This has been reported elsewhere in small island coastal communities in Indonesia by Ferse et al. (2014), Fitriana (2014) and Adhuri et al. (2015). In contrast, small-scale ecotourism businesses in Anambas develop a direct producer-to-customer relationship, with middleman functions not being required. As a result, ecotourism households tend to use a predominantly formal loan system through banks or government-backed systems. Despite their unstable income during low fishing seasons, fishers lack the confidence to apply for a formal loan so they can maximise profit and develop efficient fishing practices. The anxiety of fishers towards formal loans is due to several reasons: perceived higher interest rates compared to informal loans; collateral risks of losing their homes or lands; and risk of not being able to pay periodic instalments due to income instability. The phrase ‘afraid of [bank] loans’ was frequently stated by fisher respondents. On the contrary, most fish farmers showed some interests in using formal loans as a means of developing their livelihood, while ecotourism operators, as previously mentioned, have used this formal service to run their business. Waty (2011) stated in her study of a Southeast Sulawesi coastal community, that anxiety towards loans affects fishers more than other community groups.

4.4.3 Sustainability profile of Anambas community groups based on natural capital

Small island community groups in the Anambas archipelago have different levels of access to natural resources. Climate and weather conditions determine access to natural resources that sustain fisher household livelihoods. During the worst weather conditions, the sea is inaccessible, and fishers have to find land-based jobs that offer quick cash such as bricklaying, fruit picking, or other low-skill jobs. However, due to the large number of

85 fishers who temporarily stop fishing at the same time, only a handful manage to secure a short-term job. Access to natural resources is less affected by climate and weather patterns for fish farmer and ecotourism households. Most fixed net cages are located within bays or between islets and sheltered from high waves and strong winds. Nevertheless, fish farmers are challenged by annual variation in water quality and sporadic fish disease outbreaks. Ecotourism households are arguably the least affected party for this type of natural capital. This group has year-round access to the beach and other tourist destinations, such as dive spots and historical sites. During seasonal bad weather conditions, access to these sites is relatively safe through sheltered straits between the islands, except for outer lying islands. Conflict over resources caused by illegal fishing activities has placed constant pressure on fishers and fish farmer households in Anambas. Not only is it becoming more difficult to catch fish (longer distance and more frequent trips), there is also the risk of confrontation with illegal poachers resulting in loss of boats, fishing gears, and even life. It is common for illegal poachers, especially foreign, to frequently ram legal fishing boats who try to drive them out of Anambas waters (Saragih, 2014). Such conflicts have not been an issue for ecotourism households. There are rarely conflicts between ecotourism households with other coastal users in terms of coastal space use.

4.4.4 Sustainability profile of Anambas community groups based on physical capital

Boats, whether motorised or non-motorised, have important functions in the Anambas island communities, transportation of people and goods, for fishing activities, to access fish farms, and are vehicles for tourism. For this particular indicator, having a large motorised boat is a priority for fishers, because this can facilitate access to more fishing grounds and enable them to transport fresh catches over longer distances. The Anambas fishers have limited financial capital to repair or replace old small boats in their possession. In fact, during 2010 to 2011, there was a sharp 16% decline in the ownership of motorised boats over 2 GT, being replaced with non-motorised boats under 1 GT (DKP-Anambas, 2012). This has prevented fishers from travelling far and staying longer on fishing trips. A few fisher households operate bigger boats (larger than 2 GT) owned by other parties, but they have to share net profit revenue from the catch with the boat owner (2/3 for the boat owners and 1/3 for the fishers).

Fish farmers used boats to support their livelihood by catching fish seed and trash fish for feed, as well as to transport live fish to middlemen. Their dependency on a boat as a livelihood tool is not as high as for fishers. Fish farmers will usually ask middlemen to collect fish at their net cages. However, having a boat is an advantage for fish farmers because they can sell live fish to middlemen offering the best prices. This works well for fish farmers who do not have patronage relationship with a particular middleman, with only one drawback, which is limited and unguaranteed fish quota rights. This is similar with the patronage system in Indonesian small-scale fishery where middlemen prioritise to buy fish from their associated fishermen and limit fish supply from unassociated fishermen during high fishing season (Adhuri et al., 2015). For ecotourism households, a boat is not a necessity except for those running a chartered boat service for tourism. Not having a boat reduces the flexibility of these households to developing their ecotourism businesses. This group will not be able to participate in the community-based tourism program (CBT) launched by the Anambas District to provide services to local, national and overseas tourists, such as sea transportation to tourism sites (DKP-Anambas and LPPM-IPB, 2011b). Over 2013 – 2014, the District Agency for Culture and Tourism had a program to give a small canoe (jungkung) to small-scale ecotourism households to be rented by tourists (Basri, 2016, pers. comm., 15 November). However, the program was not successful because ecotourism households were not interested in taking non-motorized boats. As the result, the government redistributed the boats to small-scale fishers who operated nearshore. From the number of direct endowments to the three groups, it can be seen that the government has prioritised support for fisher groups in Anambas. The Anambas regular endowment program provides free equipment, such as GPS units, portable solar panels with battery kits, motorised boats, cool boxes and fishing gear, to improve fishing efficiency. Ecotourism households also received benefits from the endowment programs, at least in terms of tourism infrastructure. For example, over 2009 – 2013, there was a significant increase in the number of hotels (71%) and restaurants (114%) (BPS-Anambas, 2012). In addition, in 2011 the Anambas government finalised a Master Plan for Tourism Development in Anambas, in which at least 41 existing and new local sites will be developed into tourism destinations (DKP-Anambas and LPPM-IPB, 2011b). The Anambas

87 government strategy to fully support ecotourism is also evident in the Anambas MINAPOLITAN Land Use Planning, where ecotourism is stated as the future economy of the Anambas region as an MPA (BPPT and Anambas-Regency, 2011). On the other hand, the study found that the government support for fish farmer households was relatively minimal. There was intermittent distribution of nets for fish cages, built-up floating net cages and fish seeds. Unfortunately, most of the fish farmer participants considered the interventions as insufficient in volume. A fish farmer described this half-hearted endowment as “not even enough to make a mosquito net” (Respondent #B13). There was no other physical capital support from the government for fish farmers. Such limited support from the government might be the reason why fish farmers have developed a strong social relationship with middlemen. Through this relationship, fish farmers can improve their physical capital asset in a time of need, such as obtaining polystyrene, electric generators, trash fish, and antibiotics for fish.

4.4.5 Sustainability profile of Anambas community groups based on social capital

Allison and Ellis (2001) and Adhuri et al. (2015) have both implied that artisanal coastal communities, especially fishers, have developed unique social “safety nets” to overcome internal and external pressures on their livelihood. Strong social bonds between the members both Anambas fisher and fish farmer households can be clearly seen in terms of sharing information, resource, and risks, particularly within fisher households. For example, fisher groups will undertake search and rescue trips if one fisher is troubled at sea. Interestingly, all fisher groups in Anambas put aside some of the discounted fuel to be used for this emergency action. Fish farmer households also develop a strong bond within their fish-farming networks. However, fish farmer social networks are exclusive, where each fish farmer group has its own informal structure and rarely interacts with other fish farmer groups. Fish farmers belong to a group that only sell live fish to the middleman coordinating that group. An interviewed middleman said that fish farmer loyalty is highly praised, and in return is willing to support any request from the fish farmers in their group. Ecotourism households have a exclusive network similar to that of the fish farmers. However, interventions from the government and other institutions play an important role in establishing the social network of this relatively new livelihood group. The interventions

88 include creating tourism community groups, linking them with big tourism operators, and improving tourism infrastructure for the groups to operate. Nevertheless, only the first effort has shown a big improvement over time, and the latter two remain challenging for tourism community groups. Strong social networks are more likely to reduce the effect of internal and external livelihood stresses if the bonds developed are based on mutual relationships, strong emotional intensity and longer social interactions (Prell et al., 2009). Weaker social networks are evident in fish farmer and ecotourism operator households, compared to fisher households. Because of weak social networks, through the exclusivity present in fish farmer networks (Soemodinoto et al., unpublished), one fish farmer group can undercut another group by offering competitive fish price or equipment endowments. On the other hand, weak social network among small-scale ecotourism operators makes them unable to bargain with private tourism operators who dominate the ecotourism industry in Anambas. In addition, all household groups in Anambas indicated that business entities and NGOs operating in Anambas were less involved in supporting their livelihoods. For example, oil companies contribute up to 75% of Anambas gross domestic product (BAPPEDA-Anambas, 2009) and claim their CSR fund helps fishers (Sunarjanto, 2012). However, this study found that oil company’s involvement in community livelihoods is minimal or non-existed within the larger Anambas community groups. The respondents of this study reported that oil companies intermittently supplied a small number of FADs and fish seeds to community groups with limited success. Most of the respondents turned down the endowment due to the unclear selection of who were entitled to it, and obligations to return the CSR fund. Despite this social issue being common in the Anambas region, it has not escalated into a serious problem.

4.5 Conclusions

This research compared the relative sustainability profile of different small-scale livelihoods in Anambas MPA, Indonesia. Small-scale fisher households have difficulty diversifying their livelihood due to inefficient fishing practices, and low financial resources and conflict limiting access to natural resources. However, their strength in social networking and continuous support, mainly from the government, maintain their livelihood

89 sustainability. Small-scale fish farmers have developed an exclusive patronage social network with mariculture chain stakeholders, benefiting from more sustainable financial and human capital, such as savings and education. Unfortunately, the government, and other institutions, perceive this as a sign of self-sufficiency, and thus give more attention to other household groups, especially ecotourism households. Ecotourism has notably been favoured by the Anambas government, receiving a similar level of the endowment as fisher households in relation to Anambas status as an MPA. Being less affected by weather, having relatively preserved natural resources, and the ability to diversify their livelihood has been strong capital assets for ecotourism households for developing their livelihood. However, weak social networks among ecotourism households might diminish these advantages. Private business entities with strong financial capital might outcompete small- scale ecotourism households in the MPA, which has happened in other Indonesian areas, and the globe. This chapter confirms that the livelihood sustainability profile of small-scale mariculture households in small-island MPAs is comparable to small-scale fisher and ecotourism households. In fact, small-scale mariculture outperforms the other two livelihood groups in one of the capital assets. In addition, this research also concludes that small-scale fisher, fish farmer, and ecotourism households in small islands are within the intermediate sustainability category. This condition, between potential and poor sustainability, means that household groups can improve their livelihood with appropriate support of capital assets from themselves, the government and other institutions. Otherwise, lack of support, or ill-advised programs, might easily aggravate the access or ownership of capital assets, and push down sustainability levels to poor, or even lower. Finally, it should be noted here that small-scale livelihood sustainability of the community groups discussed in this chapter are consistently challenged by vulnerability risks inherent in their daily activities. Such vulnerability risks might change perceptions and resource use patterns (Schwarz et al., 2011), especially in small island areas designated as MPAs, such as this study area. Therefore, the following Chapter elaborates on the vulnerability risks of community groups discussed in this chapter, as well as their perception regarding the designation of Anambas MPA in their area.

CHAPTER 5. Comparative Study on Livelihood Vulnerability Risks and Conservation Support Based on Perceptions of Different Local Communities in the Anambas Archipelago MPA, Indonesia

5.1 Introduction

Chapter 4 dealt with the sustainability profile of different community groups within the study area. Besides the interaction of livelihood capital assets, livelihood sustainability is also influenced by vulnerability risks, which are also part of the SLA framework. The pressures from vulnerability risks will greatly change the sustainability of a livelihood in small island areas (Turvey, 2007). The resource use patterns will also change, forming a continuous feedback loop that will threaten the environment and the survival of local community groups if the roots of vulnerability are not properly addressed. Livelihood vulnerability of small island communities is generally overlooked by researchers despite the growing literature in coastal vulnerability studies (Dolan & Walker 2006). If the literature exists, studies are usually based on ecological and economic approaches dealing with external vulnerability factors, such climate change or market failure, influencing small island community livelihoods. Research measuring local community risk perception regarding their livelihood vulnerability are rare or non-existent (Adger 2006; Dolan & Walker 2006). Schwarz et al. (2011) also point out that many vulnerability studies lack understanding of the ‘subjective vulnerability’ within a diverse local community. Schwarz et al. (2011) describe subjective vulnerability as the different perceptions of people and their adaptive reaction or resilience to the impact of a particular event. Consequently, poverty programs and vulnerability interventions are too general to address, or provide a solution, to different scales of vulnerability risk within a diverse local community. Furthermore, vulnerability studies of small islands were commonly conceptualised as that of small-island developing states (SIDs), which has been extensively discussed in the literature (e.g. Briguglio, 1995; Sullivan & Meigh, 2005; Turvey, 2007) . The lack of differentiation between the vulnerability characteristics of small islands and SIDs leads to the conception that small islands and SIDs have similar economic and environmental conditions. Therefore, a global approach to measuring the sustainability of SIDs is deemed valid to be applied to small island contexts (Adrianto and Matsuda, 2002). However, within small island contexts, direct intervention from the central government (vertically) is

91 relatively slow to address vulnerability risk, due to remoteness and weak government structure. In contrast, government structure in SIDs is generally within the SIDs themselves, which allows faster response to vulnerability risks. This is in line with Kerr (2005), who argues that the strong economic and political autonomy of SIDs enable them to adapt to changing situations faster, compared to the weak autonomy of small islands which depend on the central government. The vulnerability of protected natural resources receives more attention compared to the livelihood vulnerability faced by local communities in Indonesian MPAs. For example, studies by Ferse et al. (2014) and Ferrol-Schulte et al. (2015) suggest that increased fisher livelihood vulnerabilities in small island MPAs in Indonesia, due to natural resource degradation, can be solved through population transmigration to other areas. This suggestion clearly favours the protection of damaged natural ecosystems over recognising local fisher community rights and their attachment to place and way of life. This view also fails to recognise the community diversity in which other non-fisher groups might have better adaptation and resilience to vulnerability risks. The generalisation of small island communities as fishers has been loosely used throughout Indonesia (Allison and Ellis, 2001). As a result, other community groups, which have different perceptions and stronger adaptability to address vulnerability risks, are dissolved into weaker, but dominant, community groups (Schwarz et al. 2011). Numerous studies regarding the relationship between small island local communities and MPAs in Indonesia (such as Bennett & Dearden 2014; Elliot et al. 2001; Ferse et al. 2012) have identified three main types of community group, i.e. fishers, fish farmers and ecotourism operators, without undermining the importance of other groups. With such clear differentiation, there have not been any that have attempted to study the specific vulnerability risks faced by each of these groups, nor have the groups been compared to one another. There are also no studies explaining the relationship between vulnerability risks and perception of these different community groups when establishing an MPA. This is important to study because MPA establishment has the potential to increase the vulnerability risks of local communities through prohibition or limiting access to protected resources.

This chapter aims to develop and compare the socially-constructed livelihood vulnerability indicators (LVIs) of three different community groups and their influence on the community groups toward supporting the MPA establishment. The socially-constructed LVIs used here is based on a definition of vulnerability risks by Kelly & Adger (2000) and Dolan & Walker (2006), where individual or community perceptions construct the condition of vulnerability instead of the pressure from physical events. For this purpose, vulnerability is defined as the capacity of individuals or social groups to respond to, cope with, recover from, and adapt to external stress on their livelihoods and well-being (Kelly & Adger 2000; Dolan & Walker 2006). Besides the importance of unravelling vulnerability risks from local community perceptions, the socially constructed LVIs are used here because hard data from assessing vulnerability risks in small islands in Indonesia is unavailable. This has been the challenge for many researchers conducting vulnerability assessments in developing countries (Erenstein et al., 2010). In addition, this chapter also attempts to measure local community perceptions of an MPA, and makes inferences as to whether it influenced the level of their livelihood vulnerability risks. This chapter is also intended to provide a useful tool to examine the root causes of vulnerability based on the perception of different community groups in small islands MPAs. Birkmann (2007) pointed out the need to study vulnerability risks at the local community level to balance the more common global study of vulnerability. In doing so, policy and intervention programs from the government, and other institutions, could be directed efficiently and effectively to tackle vulnerability risks at a finer level – the community level.

5.2 Methodology

5.2.1 Study area

Anambas Archipelago MPA is a district and a cluster of small islands situated in the southern South China Sea, between 2º10’0”-3º40’0” N and 105º15’0”-106º45’0” E. It is one of the sea frontier areas of Indonesia and shares its borders with the Vietnam EEZ line on the North and the Malaysia EEZ Lines on the West and East. It is closer to Malaysian (201 km) or Singaporean territorial land (215 km) compared to its own provincial capital city (Riau Province, 320 km). Survey and interview activities were carried out in 15 villages of the seven sub-districts to cover the possible variability of the population within 93

Anambas archipelago (Figure 5.1). The surveyed villages were selected to represent the five main/important islands (Siantan, Jemaja, Telaga Kiabu, and Matak islands) in the Anambas Archipelago, main settlement of populations, and sub-district administrative boundary. Fieldwork was carried out over four months, including preparation and survey activities, from July to November 2014.

Figure 5.1. Location of study sites (villages) in Anambas MPA

5.2.2 Questionnaire design

This research used perceptions of community groups to measure their livelihood vulnerability risks and their level of support for the establishment of the MPA in their area. For such purposes, predetermined LVIs, based on the Sustainable Livelihood Approach (SLA), following Scoones (1998), DfID (1999) and Ellis (2000), were embedded into the questionnaire. The LVIs were used to measure the tangible perceptions of fisher, fish

94 farmer, and ecotourism operators in Anambas MPA, regarding the influence of vulnerability risks to their livelihoods. The LVIs with similar characteristics are grouped into categories of seasonal, trend and shock (DfID, 1999, Allison and Horemans, 2006) (Table 5.1). A set of questions was also developed to address the perception of community groups to the establishment of Anambas MPA, including knowledge, level of support and perceived benefits (Appendix 1). All questions measuring livelihood vulnerability risks used the 5 points Likert-scale type response, from 1 (not important) to 5 (very important). All questions designed to capture perceptions of local community groups to the establishment of Anambas MPA used 3 points Likert-scale type response; 1 (disagree), 2 (neutral) and 3 (agree).

Table 5.1. LVIs used in the study

Seasonality (cycles) Trends Shocks  Seasonal climate  Food price increase  Bad weather  Daily weather  Fuel price increase  Local political change  Market season  Part price increase  MPA regulation change  Health price increase  Conflict over resources  Pollution increase  Ecosystem health  Climate change issues

5.2.3 Data collection

Primary data (quantitative and qualitative) was collected from direct face-to-face interviews with respondents, whilst secondary data was collected from written and online literature. Respondents totalled 65 individual people, consisting of 43 fishers, 18 fish farmers, and 5 ecotourism operators (Table 5.2). Stratified random sampling was used to select respondents, with proportional allocation to address the uneven household group numbers. The groups of selected respondents followed those outline in Chapter 4, section 4.2.2. Human ethics approval was acquired from the UNSW Human Research Ethics Committee #HC14191 prior fieldwork, and was considered “low impact” research on human subjects.

Table 5.2. Distribution of respondents based on Anambas sub-districts

Surveyed Number of respondents No. Sub-districts villages Fisher Fish farmer Ecotourism 1 Jemaja 1 1 0 2 2 Jemaja Timur 1 2 2 1 3 Palmatak 3 12 7 0 4 Siantan 2 5 0 1 5 Siantan Selatan 3 12 3 1 6 Siantan Tengah 3 7 5 0 7 Siantan Timur 2 4 1 0 Total respondents 15 43 18 5

5.2.4 Data analysis

5.2.4.1 LVI index values

The LVI dataset collected from the interviews used Likert-scale type data with different scale units, thus data normalisation into a common unit was required. The Likert- scale type data was transformed into index values using simple linear transformation (Eq.1 and 2), following the method suggested by Erenstein et al. (2007), Jones & Andrey (2007) and Ranganathan et al. (2009). This standardisation produced values from 0 - 1 (min. values = 0; max. values = 1) for all LVIs, maintaining similar spread and equal weighting (Erenstein et al., 2007). This normalisation is valid, as Likert-scale data can be interpreted into a subscale, such as indexes, and aggregated to acquire a composite value (Carifio and Perla, 2007) (Eq. 3). The following equations (Eq. 1 and 2) are the mathematical expression of transforming the original Likert-scale type data to an index value. Eq. 1 was used when higher values of original data correspond to higher index values (increase functional relationship). Eq. 2 was used when higher values of original data correspond to lower index values (decrease functional relationship).

Increase functional relationship, ……… (Equation 1)

Decrease functional relationship, ……… (Equation 2)

Where LVIij = LVI index value of respondent i in indicator j of group x; Xij = real value of response of respondent i in indicator j of group x; Min(Xij) = minimum real value of respondent’s response in indicator j of group x; Max (Xij) = maximum real value of respondent’s response in indicator j of group x. (Note: i = 1,2,3,..i ; 0 < LVIijx < 1) 96

The index composite value for each indicator was computed using simple Equation 3:

……… ……………….. (Equation 3)

Where LVIjx = LVI index value of indicator j of group x, njx = number of respondent indicator j of group x.

Classification of vulnerability has been debated in the literature due to the fuzziness between points of scale, and thus it varies greatly among scholars (Eakin & Bojórquez- Tapia 2008). For example, different scales are commonly used in the literature, such as a 3 point scale by Eakin & Bojórquez-Tapia (2008) or a 5 point scale by Hahn et al., (2009). This study used a 5-point scale to determine the LVI index values to maintain consistency of original data transformation, which ranged from 0 (least vulnerable) to 1 (most vulnerable).

5.2.4.2 Local community’s perception toward Anambas MPA establishment

Descriptive statistics using frequency distribution analysis was used to analyse the different perception of three local household groups (fisher, fish farmer, and ecotourism operator) toward Anambas MPA Establishment. Additional data analysis was performed for qualitative data collected from an open-ended questionnaire regarding the three household groups’ perception on the establishment of Anambas MPA. This qualitative analysis was to determine any possible links between vulnerability risks and the local group’s views regarding the designation of Anambas area as an MPA as well as their support to the Anambas MPA conservation efforts. Descriptive statistics, using frequency distribution analysis, was used to analyse the different perceptions of local groups (fisher, fish farmer, and ecotourism operator) to the establishment of Anambas MPA. Thematic analyses, following a data-driven inductive approach suggested by Boyatzis (1998) and Fereday and Muir-Cochrane (2008) were used for qualitative data from an open-ended questionnaire of the perception of local community groups to the establishment of Anambas MPA. This qualitative analysis was to determine possible links between vulnerability risks and the perception of local groups to the designation of Anambas area as an MPA, as well as their support for the Anambas MPA conservation efforts. NVIVO 10 for Windows was used to create a consistent theme categorisation system based on perceptions of local community groups to the MPA

97 establishment. After the coding finalised, matrix coding comparison was used to compare perception of the three local groups based on the created themes and nodes. This qualitative analysis served as the triangulation method to confirm the result of the quantitative analysis performed previously.

5.3 Results

5.3.1 Livelihood vulnerability of Anambas local community groups

5.3.1.1 Seasonality (Cycles)

In terms of seasonality risks, 73% of fisher respondents agreed that seasonal climate was very influential, compared to 50% of fish farmers and 60% of ecotourism respondents. The percentage was confirmed with the highest seasonal climate LVI index value scored by fisher households. This LVI index value can be categorised as moderately vulnerable. Fish farmers and ecotourism operators were categorised as moderately vulnerable, as seasonal climate conditions has less influence on their daily livelihood activities (Table 5.3).

Table 5.3. Comparison of seasonal LVI index values for three different household groups Household group No. Vulnerability indicators Fisher Fish farmer Ecotourism 1 Seasonal weather 0.80 0.64 0.60 2 Daily weather 0.51 0.39 0.40 3 Market season 0.55 0.67 0.50 Composite seasonality indicator 0.62 0.56 0.50

Daily weather also poses higher vulnerability risk to fisher compared to fish farmer and ecotourism households. However, the level of vulnerability risk caused by daily weather changes are not as severe as seasonal weather and were consistently categorised as moderately vulnerable across household groups. All respondents from the three groups agreed that the effect of heavy rain will diminish in a few days compared to the effect of the southern monsoon season that brings strong winds lasted at least 3 months. In terms of market seasonality, the three household groups have relatively similar vulnerability risks, all falling into the moderately vulnerable category. Despite that, the fish farmer households have the highest LVI index values compared to fisher and ecotourism households (Table 5.3). 98

5.3.1.2 Trends Among the group trend indicators: fishers expressed the most vulnerability on food and health increased costs as well as biodiversity degradation; fish farmers on three environmental related factors (pollution increase, ecosystem health, and climate change) and one economic related indicator (part price increase); and ecotourism operators felt that fuel price increase was the highest vulnerability risk to their livelihood (Table 5.4).

Table 5.4. Comparison of Trend LVI index values for three different household groups Household group No. Vulnerability indicators Fisher Fish farmer Ecotourism 1 Food price increase 0.66 0.63 0.60 2 Fuel price increase 0.74 0.67 0.80 3 Part price increase 0.51 0.56 0.50 4 Health price increase 0.22 0.19 0.20 5 Pollution increase 0.42 0.67 0.40 6 Ecosystem health 0.88 0.97 0.40 7 Climate change 0.09 0.19 0.00 Composite trend indicator 0.50 0.55 0.41

In total, 47% of the respondents from fishers expressed that they were very concerned with the consistent year by year price increase of main food items such as rice, cooking oil, eggs, and meat. In contrast, only 40% of ecotourism operator and 28% of fish farmer respondents were concerned about it. However, the LVI index values between fisher, fish farmers, and ecotourism households are similar, and categorised as moderately vulnerable. Ecotourism households scored the highest LVI index values for fuel price increase compared to fishers and fish farmers (Table 5.4). Up to 80% of ecotourism household expressed that fuel price hike has reduced the number of local and foreign tourists visiting the tourism sites. Fishers (63%) also argued that fuel price increase contributed strongly to the cost of production (catching fish and transportation). However, significant fuel discounts for fishers, as part of national policy for sustaining fisheries activity, forces down the potential vulnerability risk of fuel price increase (Muchlisin et al., 2012). Fish farmers have less dependency on fuel which could be correlated to the lowest LVI index value, with 44% of respondents stating that fuel price hike has a moderate influence on their livelihood.

The three household groups have a similar attitude toward the influence of price increases of machinery spare parts and materials on their livelihood activities. The LVI index values among the groups were similar and classified as moderately vulnerable. A slightly higher LVI index values for fish farmers may be due to their livelihood being highly dependant on materials. The LVI index values for fisher and ecotourism households were lower as both groups seldom experienced broken machinery or price increase. Only a small percentage of respondents from the three household groups (15%) expressed that health service was very important in influencing their livelihood. Very low LVI index values of the three groups (low vulnerability) confirm the relative low vulnerability associated with increased health service price (Table 5.4). Free healthcare provided by the Anambas Government have changed the perception of the groups despite the poor quality of the public clinic service. Of the three groups, fish farmer households exhibited the strongest response to the effect of water pollution on their livelihood, followed by fishers and ecotourism households (Table 5.4). On this indicator, 50% of the fish farmer respondents expressed that increased water pollution exerts high vulnerability risk to their livelihood activities, compared to only 23% of fisher and 20% of ecotourism respondents. Fisher and fish farmer households both express that biodiversity degradation poses high vulnerability risk to their livelihood compared to ecotourism (that depended mostly on non-extractive activities). Fish farmers were the most vulnerable on this indicator, followed by fishers and ecotourism (Table 5.4). The very high LVI index values of both fish farmer and fisher groups (most vulnerable), stemmed from concerns such as illegal operation of trawl nets, scarcity of fish seeds and fish stocks. A common vulnerability risk shared by the three groups was the increased degradation of coral reef systems around their island due to cyanide and unfriendly fishing practices. The term climate change was not familiar to all respondents and had to be explained in great detail; the local signs of climate change, such as prolonged drought or wet seasons, unnatural local temperature variations, or possible local sea level rise. All respondents unanimously indicated the threat of climate change to their livelihood was very low, which was confirmed by low LVI index values for all groups (least vulnerable) (Table 5.4). All groups perceived that the mentioned signs were absent in their area.

100

5.3.1.3 Shocks

The overall shock composite index values showed that both fisher and fish farmer households have a higher vulnerability (most vulnerable) than ecotourism (moderately vulnerable). Among the shock indicators, fishers were constantly classified as the most vulnerable group, except on the local political change indicator (Table 5.5). Sudden bad weather posed high vulnerability to fishers and ecotourism operators, with similar LVI index values (highly vulnerable), compared to fish farmers (moderately vulnerable). In total, 70% of fisher and 80% of ecotourism respondents reported that bad weather could prevent them carrying out their livelihood activities. In contrast, only 33% of fish farmer respondents reported that bad weather could prevent them in conducting their occupation. In terms of local politics change, fish farmers expressed higher vulnerability compared to fishers and ecotourism operators (Table 5.5). Some fish farmer households expressed that political change seldom benefits them because local candidates always target fisher households, the dominant population, in their political agenda. Despite being the least dominant population, ecotourism households expressed that change in the local political leader brings more changes to the development of community-based tourism in Anambas MPA.

Table 5.5. Comparison of shock LVI index values for three different household groups Household group No. Vulnerability indicators Fisher Fish farmer Ecotourism 1 Bad weather 0.80 0.56 0.80 2 Local politics change 0.66 0.78 0.50 3 MPA regulation change 0.89 0.81 0.40 4 Conflict over resources 0.81 0.83 0.30 Composite shock indicator 0.79 0.74 0.50

Both fishers and fish farmers expressed a higher vulnerability risk to their livelihood due to the establishment of Anambas MPA, due to access restrictions and imposed MPA regulations. On the other hand, ecotourism operators considered MPA regulations and objectives to poses low vulnerability, and in some aspects have benefited their livelihood. Conflict of resource use, particularly with illegal fishers operating in Anambas MPA waters, is a bigger influence on LVI index values for fish farmers and fishers (high vulnerability) compared to ecotourism (low vulnerability) (Table 5.5). Both 101 fishers and fish farmers consistently face conflict with non-local and foreign fishers (mostly illegal). Ecotourism households seldom experience conflict over resources, except land disputes in tourism sites. The ecotourism households have difficulty getting leasing permission from landowners who control the already limited land in the Anambas MPA.

The average LVI composite index values of the three groups shows that fisher and fish farmer households are more vulnerable compared to ecotourism households (Figure 5.2). Fishers and fish farmers expressed high vulnerability risk related to biodiversity on which they depend, as is typically experienced by resource-based livelihood activities. Low vulnerability risk faced by ecotourism households was mostly attributed to less dependence on natural resources, more local political support, and being less affected by seasonal climate and conflicts over resource use in the area.

Figure 5.2. Vulnerability diagram of the LVI index values for different household groups

5.3.2 Small-scale community groups’ perspectives on the Anambas MPA The three household groups have contrasting views about the designation of the Anambas MPA and how it affects their livelihood. The study found that both fish farmers and ecotourism respondents were more knowledgeable regarding the designation of their area as an MPA (Figure 5.3). At least 60% of the respondents from both households knew

102 that Anambas was designated as an MPA, compared to only 46.5% of respondents from fisher respondents.

Figure 5.3. Percentages of respondents who know the designation of Anambas MPA

Based on the NVIVO matrix coding comparison, the reason for respondents’ lack of knowledge was similar across all household groups; not being invited to, or present in the village during, the dissemination processes. Formally attending dissemination meetings, word of mouth from their peers, and field visits and training were the main contributing factors for respondents from household groups who knew about the designation of Anambas MPA (Table 5.6).

Table 5.6. NVIVO Matrix coding comparison of ‘knowingness’ and types of involvement of respondents

Type of response Fisher Fish farmer Ecotourism Do not know 23 7 2 Know 20 11 3 Types of involvement (times mentioned in the interview transcript) Dissemination/meeting 8 7 3 Field visit 2 2 0 Training 11 6 3 Word of mouth 11 5 0

Furthermore, there was a high level of agreement on the designation of Anambas MPA between the household groups (Table 5.7). One exception was in fisher households where unconsulted respondents (26%) showed significant disagreement. The responses of

103 those who agree with the designation of the Anambas MPA from the three household groups (44%) related their argument with the importance of protected zones as fish banks. However, 24% of responses from the household groups indicated support for the MPA provided the majority of the community support the MPA as well.

Table 5.7. Combination of responses of agreement about the designation of Anambas MPA and reasons of agreement based on matrix coding comparison Combination of response Agreement Fisher Fish farmer Ecotourism Do not agree 11 1 0 Neutral 1 1 1 Agree 31 16 4 Reason of agreement (times mentioned in the interview transcript) If the majority of community agree, I will also 16 3 1 support the MPA Preventing outsider illegal fishing 11 1 0 Protecting local fisher fishing ground 7 2 0 Relatively small to affect my livelihood 2 0 0 Serving as fish banks 16 6 3

Most of the three household groups strongly agree that the Anambas MPA will give some benefits for the island community (Figure 5.4), even though some of the respondents stated, or indicated, they did not know the existence of the Anambas MPA at the beginning. A significant number of fishers (25%) also disagreed with the establishment of Anambas MPA, yet the majority of them (72%) agreed to the potential benefits of the MPA. Fish farmers and ecotourism operators also expressed a change of attitude toward the benefits of MPA, although not as strong as in fisher households. The shift of responses among the household groups in favour of the MPA was clearly evident through the interview process.

104

Figure 5.4. Level of Agreement among the household regarding the potential benefits of the Anambas MPA to the local people livelihoods

5.4 Discussion

5.4.1 Livelihood vulnerability of small scale community groups

The influence of seasonal vulnerability risk to small island communities is arguably higher compared to land-based activities (Allison and Ellis, 2001, Adrianto and Matsuda, 2002). It is due to the high-risk nature of small island environment conditions and the highly unpredictable occurrence of those risks where the local communities reside and depend upon (Adrianto and Matsuda, 2002, Schwarz et al., 2011). This study found that the effects of seasonal vulnerability risks to the household groups are significant. Vulnerability risks of fishing during daily and seasonal bad weather conditions have forced most fishers, and some fish farmers, to suspend their activities. The effect is so severe that fishers have to switch completely to work on government construction projects in their villages or become unemployed during this period. This cutoff period forced fishers to depend more on social safety networks; for example, borrowing money or primary necessities from middlemen or other patrons (Ferse et al., 2014, Adhuri et al., 2015). Fish farmers also have a similarly high seasonal vulnerability risk, particularly with the seasonality of the market. Market access and information are controlled by the buyers from Hong Kong and their operatives (middlemen) in Anambas. Local middlemen and fish farmers are highly dependent on boat visits from Hong Kong that currently come in only twice a month. In order to accommodate

105 fish product from non-member fish farmers, Hong Kong buyers and their local middlemen set up a quota system fish (10 – 30 kg/visit) for the non-member fish farmers, provided the fish farmers who are members have met their quota. The two MMAF regulation (MMAF, 2013, MMAF, 2016), recently issued, have also tightened restrictions on the already limited market of fish in Anambas through restricting quota and direct export to overseas. Ecotourism households felt less affected by nature-based seasonality risk due to their more land-based activities. However, this group has a high dependency on intervention from formal institutions such as governments or NGOs. Ecotourism households reported a significant increase in revenue up to 400% during government-backed activities in the tourism sites. The annual school holiday season also contributed to the increase of visitors in tourism sites although not as significant as the government-sponsored events. Other than those, the market season of tourism in Anambas is relatively stable but low in visitation number. Calgaro et al. (2014) points out that out of 11 main factors of reducing tourism vulnerability, six factors can be linked to interventions by the governments or formal institutions. The relative isolation and long distance from economic centres have been the characteristic of small island communities (Adrianto and Matsuda, 2002, Abecasis et al., 2013). These disadvantages, in particular, have put more stress on the economic and ecological resilience of local island communities compared to other regular coastal communities. These disadvantages are clearly represented by the trend vulnerability indicator of fuel price increase which affected ecotourism operators and fisher households more than fish farmers. For example, fishers have to spend 40 to 74% of the total fishing trip cost on fuel, despite government fuel subsidies given to fishers (Muchlisin et al., 2012). With more fishers relying on motorised boats, an increase in fuel price will either reduce the profit per fishing trip and force fishers to shorten trips, or maximising fishing effort near shore (Sumaila et al., 2008). For ecotourism, fuel price rises means more expensive transportation, coupled with the remoteness it will discourage the flow of mass tourism in the area (Hampton and Jeyacheya, 2015). Fish farmers are less vulnerable to fuel price rises because fish farming, especially small scale, requires less consumption of fuel, and fish trading usually occurs near, or at, the farming site.

106

Continuous biodiversity degradation caused by coral mining, cyanide, and illegal fishing within Anambas MPA has continued despite it being declared an MPA (Mustika et al., 2013). This trend particularly hit fishers harder, as total capture near shore has significantly decreased and forced them to fish farther out to sea. Some fishers with boats above 1 GT have to travel between 100-200 miles to fishing grounds. Although most fish farmers do not have to travel that far, they also catch less fish seed to supply their fixed net cages. The ecotourism households expressed low vulnerability in terms of biodiversity degradation despite the current rate of biodiversity degradation in Anambas MPA. One possible reason to explain this unusual finding is that ecotourism households still view the resource as an object, rather than an asset that has to be protected and managed (Goodwin, 2002). Similar results were also reported by Walpole and Goodwin (2001), where a local community involved in ecotourism in KNP showed low support for the conservation effort. This implies that local ecotourism operators do not realise ecotourism vulnerability is strongly dependent on the health of the protected ecosystem (Ross and Wall, 1999). The perceived high level of vulnerability on LVIs related to natural resources use are clearly expressed by fishers and fish farmers. Conflicts with foreign and local fishers, MPA fishing limitations, sudden bad weather, and biodiversity degradation have been difficult to anticipate by both households. The inability to respond and adapt to these shock indicators might increase the sensitivity of the livelihood to other types of vulnerability risks (Calgaro et al., 2014). Fish farmers are, in some ways, affected similar to fishers because part of their daily routines involves fishing as an alternative livelihood. Furthermore, the apparent low vulnerability felt by ecotourism households on these LVIs is predominantly due to their activities being unrelated to fisheries resources. Conflict over resources also has comparable effects to bad weather for fisher households. Not only does it significantly decrease total fish capture, but also causes psychological damage, especially for small-scale fishers. Increased job loss and poverty were the results of psychological damage among small-scale fishers due to their inability to compete with illegal fishers (Pomeroy et al., 2007). Most of the fishers interviewed in this study were afraid and felt helpless when meeting with illegal poachers at sea. This was especially felt towards foreign fishers because the small ships operated by local fishers were intentionally rammed and sank by the foreign vessels. Some of the fisher respondents

107 said that attempts to drive these illegal fishing boats out were dangerous, as the foreign fishing crews carried firearms on board. Pomeroy et al. (2007) also reported similar findings, that conflicts between fishers using firearms are common in ‘fish wars’ in the South China Sea. On the contrary, ecotourism households reported almost no conflicts regarding natural resource use. Insignificant conflicts regarding the land use of tourism sites owned by a few individuals have been resolved through government intervention. For example, the government facilitated a tourism revenue sharing scheme between landowners and ecotourism operators in Padang Melang Beach, Jemaja Island. Landowners allowed small-scale tourism operators to build permanent guest houses and restaurants on their beach in return for small fees. The study findings show that the three household groups expressed minimal vulnerability risk posed by climate change which might uncover a hidden fact. It is likely that knowledge discrepancy regarding climate change impacts on the Anambas coastal communities are the underlying reason these views are held by all household groups. This might be caused by the observed low level of education among communities, and the absence of information and mitigation plans regarding climate change impacts on coastal communities. Being the most vulnerable communities to the impacts of climate change (McGranahan et al., 2007), these three groups are facing greater risk due to their unawareness and their dependency on the natural resources for their livelihood. However, fisher households might have a better resilience in facing vulnerability to climate change impacts compared to the other two households. Fishers have developed a more sophisticated social safety net (social capital) with other fishers that provide help during hard times (Hahn et al., 2009, Cinner et al., 2010).

5.4.2 Communities acceptance of an Anambas MPA establishment

Most of the problems characterising Anambas MPA until now are local community- related problems, such as MPA regulation compliance, conflicts among MPA resource users, and alternative livelihoods for the local community (LKKPN-Pekanbaru, 2013). This study found that the household groups view the establishment of Anambas MPA differently. These differences, in turn, influence regulation compliance, the level of resource use, and livelihood development of the small island communities in the area.

108

The higher percentage of fishers unaware of the Anambas MPA existence, compared to fish farmers and ecotourism operators, is intriguing, as fishers are the dominant community group and directly access marine resources. One possible reason for this could be the absence of some fishers at the time of the dissemination process, as they have to travel at sea for days to catch fish. The study findings concur with this, as most fishers operating boats above 1 GT reported fishing trips from 3-14 days at sea, with just one or two days on land before a subsequent trip. On the contrary, fish farmers and ecotourism operators have slightly better knowledge about the Anambas MPA establishment because they operate on or near land. They had a better chance of attending the meetings or receiving information that the fisher groups missed. Tam (2015) highlighted this by stating that community consultation has to consider the mobility of coastal communities to ensure a broad representation and thus guarantee the acceptance of an MPA regime. This might partly explain some fisher communities lacking knowledge of an MPA established by formal institutions in the studies by Cinner et al. (2010) and Kusumawati and Huang (2015). However, it is not always the lack of active community participation driving this. It is more rooted in the use of top down approaches, weak informal ties with fisher groups, strong community elite representatives, and lack of funding that prevent better information dissemination of MPA establishment. In terms of the potential benefits of the MPA, all household groups seemed to relate their reasoning to their livelihood. Most of the fishers mentioned the important functions of the MPA as a ‘fish bank’, preventing illegal fishing activity (non-local and foreign), and regulating the fishing activity of their own peers. The latter is particularly important to the fishers, indicating that distrust is not only between local communities and MPA management authorities in Indonesia (Kusumawati and Huang, 2015). This distrust also occurs between members or groups of a local community, who demand equitable and fair access, and compliance, in MPA multiple use zones. Interestingly, most fish farmers and ecotourism operators indicated the main benefit of Anambas MPA was the function of a fish bank. The three household groups referred to a fish bank as fishing grounds where fish are left to flourish and restock the surrounding areas through spillover effect. During the interviews, some points regarding the possible benefits of the MPA were mentioned, but no fish bank term or spillover effects were explicitly mentioned. This interesting finding could

109 be attributed to their experience of catching less fish near the shore, or that they were informed during the dissemination process of Anambas MPA.

5.5 Conclusions

Efforts to measure livelihood vulnerability of small island communities have been mostly based on economic and ecological approaches. Sometimes these studies are impossible to carry out due to unavailability of hard data in less developed countries. This chapter successfully measured and compared livelihood vulnerability of three small-scale community groups in Anambas MPA using a set of socially constructed LVIs. The chapter also extended the study to determine groups perceptions regarding the establishment of the MPA, which has, in some ways, affected their livelihood. This chapter confirms that the livelihood vulnerability risk faced by small-scale community groups in small-island MPAs can be categorised as moderately vulnerable. Despite their similar vulnerability, fisher household are the most vulnerable among the groups, followed by fish farmers and ecotourism households. Natural related risks categorised in seasonal and shock indicators severely affect the Anambas fisher and fish farmer groups. Particularly for fisher groups, vulnerability risks will likely increase if MPA regulations impose common fishing restrictions. The government could intervene by increasing the fishing efficiency of fisher groups by supporting their alternative livelihoods during times when fishing is not possible due to poor seasonal weather or other reasons. Fish farmer groups can cope with the pressures of natural risks better than fisher groups because they can be more flexible and obtain alternate livelihood inputs. Ecotourism households are the least vulnerable in the Anambas MPA due their non-extractive activity. Further government intervention to promote the MPA as a tourist destination and gradually increasing the dependency of ecotourism households from government-backed programs will strengthen the resilience of this community group. The three household groups have similar general views that the MPA has benefits directly related to their livelihoods, such as improving fish stocks and preventing illegal fishing. However, the government and MPA authority might have to invest more efforts and resources to increase fisher compliance to MPA regulations. In addition, issues of trust among the members of the groups should be a concern for the MPA authority. The view of

110 someone’s obedience to the law is someone else chance to break it is still pervasive within the three household groups, especially with fishers. With the limited resources and staff available in most of Indonesian MPAs, building trust among community members similarly dependent on MPA resource is a great challenge yet to be resolved. The findings from this chapter and the previous chapters (Chapter 3 and 4) have indicated that small-scale mariculture in Indonesia has been integrated within the MPA regulation framework. The findings also show that small-scale mariculture is economically and culturally important to small-island MPA communities and has a sustainability profile comparable to ecotourism and small-scale fisheries. Nevertheless, the current regulatory framework lacks a means to formalise the designation of mariculture zones within multi- use MPAs. To solve this issue, the subsequent chapter, Chapter 6, develops a site suitability framework to determine mariculture zones for small-scale mariculture.

111

CHAPTER 6. Development of Mariculture Zone Site Selection Based on GIS Approach for Small-Scale Finfish Mariculture in MPA

6.1 Introduction

Chapters 3, 4 and 5 demonstrated that finfish mariculture is important and can be considered a sustainable livelihood activity for local communities living within or adjacent to MPAs in Indonesia. Despite the importance and potential sustainability of finfish mariculture, careful planning is required when undertaken in an MPA, due to the sensitive ecosystem and limited areas allocated for economic development (Agardy, 1994). Determining the best location for mariculture activities, as part of fair space allocation in coastal areas, has been studied by many researchers (e.g. Silva et al. (2011); Radiarta et al. (2008); Pérez et al. (2002) Longdill et al. (2008); Falconer et al. (2013); Navas et al. (2011a); Vincenzi et al. (2006). The majority of these authors used a GIS analytical model combined with multi-criteria evaluation (MCE) to find suitable sites for different cultured fish species, and to address the needs of multiple users (Ross et al., 2013a). However, all of these studies were designed for the mariculture industry (Silva et al., 2011) and regular coastal areas where mariculture received positive acceptance from both the coastal community and governing authority. In contrast, there is a gap in the literature concerning the development of a site selection framework for mariculture in MPAs, specifically for small-scale fish farmers. This is especially important for Indonesia, where many local communities live on small-islands within multi-use MPAs and regard mariculture as an important livelihood (Chapter 3: section 3.4) This chapter addresses the gap by developing a site selection model within the GIS environment, specifically aimed to solve conflicting issues in developing sustainable small- scale finfish mariculture in MPAs. The GIS model incorporated MCE in which a parameter-specific suitability function (PSSF) was used to solve the complexity of criteria (Vincenzi et al., 2006, Longdill et al., 2008). The GIS model with PSSF, was used to develop the site suitability criteria for sustainable finfish mariculture in an MPA, hereinafter referred to as an MPA mariculture zone, using several sub-models (i.e. constraint, site suitability, visual amenity and stakeholder sub-models). Constraint and site suitability sub-models mainly deal with the environmental requirements of finfish to grow optimally and resolving conflicting uses of an MPA SFZ seascape. The visual amenity sub- 112 model deals with the need to maintain the scenic appearance of an MPA. The stakeholder preference sub-model is introduced to involve MPA stakeholders in the site selection decision-making process, designing mariculture zones within MPAs, as described in Chapter 3.

6.2 Materials and methods

6.2.1 Analytical framework of GIS and remote sensing

6.2.1.1 Conceptual model

The combination of GIS and remote sensing to aid decision-making for aquaculture site selection is increasingly common due to its efficiency and cost-effectiveness (Radiarta et al., 2008), the growth of computing power, and demand for GIS outputs (Meaden and Aguilar-Manjarrez, 2013). Most of the combinations of GIS and remote sensing analysis use MCE to overcome different measurement units of criteria (Longdill et al., 2008, Silva et al., 2011). However, MCE applications are often applied without a proper understanding of the function of weighting and the combination procedure of attribute maps (Malczewski, 2000) . In addition, it is also common that these practices involve scientists and experts to claim a participatory decision-making process, but overlook the participation of the main intended user, the local fish farmers. Site selection for MPA mariculture zones presented in this study involves a combination of GIS and remote sensing methods, based on modified site selection developed by Silva et al. (2011), but with different sub-models. The modification is introduced to account for: the application of the site suitability model to an MPA; the inclusion of local stakeholder (fish farmers); as well as to address qualitative and quantitative data scarcity in remote small island MPAs in Indonesia. In addition, this site selection approach also considers seasonal variations and common finfish mariculture practices affect site suitability in small island MPAs. For example, dry and wet seasons can significantly reduce or increase the area capable of sustaining mariculture activities. In another instance, Pomeroy et al. (2006) stated that the most preferred finfish mariculture practice in the Southeast Asian countries, including Indonesia, is using fixed net cages. However, this differs with future development plans proposed by local government, that floating net cages are the best option (TNC, 2003, DKP-Anambas and LPPM-IPB, 2011a). 113

Therefore, variations in season and farming practice are embedded into the model used here to produce MPA site suitability for mariculture floating net cages. Figure 6.1 illustrates the site selection framework with multiple sub-models used in this study.

Boolean Analysis Broad geographic area selection Legal Constraints

MPA Constraints Map constraints Basemap generation Generation/overlay Econ-Soc Constraints Physical Constraints Feasible Constraint sub-model (1)

Areas MCE based on Physical suitability Unweighted PSSF MCE Suitability Growth & survival Generation (Geometric overlay) Factor’s suitability ranges MPA visual amenity

Dry Wet Season Season Site suitability sub-model (2)

Floating Net Cage

MCE based on Distance to village Unweighted PSSF MCE Suitability Dist. to MPA features Generation (Geometric overlay)

Preferences’ Dist. to local pier suitability ranges Aggregate (Raster calc.) Distance Analysis Stakeholder preferences sub-model (3) Site Suitability for Finfish Mariculture Zone in MPA Figure 6.1. The organisational structure of site suitability sub-models used to determine finfish mariculture zone within small-island MPA in

114

6.2.1.2 Site suitability analytical framework

The diverse site suitability parameters or criteria, ranging from quantitative to semi- qualitative criteria, used in this study, require suitable classification of each criterion to produce comparable and consistent units (Vincenzi et al., 2006, Longdill et al., 2008). Classification classes have to be based on the specific requirements of the parameters to support finfish mariculture development in MPAs. This study used two procedures of data standardisation or classification, Boolean Logic and PSSFs, as shown in Figure 6.1. Boolean Logic used in the constraint sub-model differentiates between feasible alternatives, represented by an index value of “1”, and non-feasible alternatives, represented by an index value of “0” (Malcezewski, 2000, Gemitzi et al., 2007). The mathematical expression for constraint area selection and standardisation is as follows (Eq. 1):

Where: SIk = Constraint suitability index (feasible alternative), K = number of criteria in the constraint model, bj = suitability index value for each constraint criterion.

PSSFs were used to transform the original data of criteria into arbitrary standardised suitability scores (Vincenzi et al., 2006, Longdill et al., 2008). PSSF classifications were used on the site suitability and stakeholder preferences sub-models, as described in Figure 6.1. Each suitability score reflects the achievement or satisfaction of a particular criterion to the specific range needed by finfish mariculture (Further described in Table 6.1). The product of PSSFs are suitability indexes (SI) with values ranging between 0 – 1, where 0 represents a non-suitable area and 1 represents the most suitable area (best alternative). However, before PSSFs classification was performed, it was essential to incorporate the overall constraint site suitability (SIk) to original data of each criterion, as suggested by Malcezewski (2000). The objective of this important step is to ensure that raw data standardisation or site suitability classification of a criterion is performed on the feasible alternative to satisfy all identified constraints (Malcezewski, 2000). Furthermore, Malcezewski (2000) showed that an arithmetic aggregation (overlay process) of criteria produces different results if undertaken or not undertaken on the feasible alternative

115

(constraint site suitability). This effect was also observed in the MCE overlay using geometric mean in this study (Figure 6.2).

(a)

(b)

Figure 6.2 The effect of attribute standardisation using geometric mean where (a) standardisation is carried out on all criteria; (b) standardisation is carried out on feasible alternative (simulated data from Malcezewski (2000)

Therefore, this study adopted the approach used by (Malcezewski, 2000), where the set of feasible alternatives for each criterion was produced by multiplying the criterion by the overall constraint site suitability (SIk). The set of feasible alternatives was then standardised using the PSSF classes corresponding to the values required for the particular

116 criteria. Then, the overall site suitability was calculated as the geometric mean of all PSSF- standardised feasible alternatives of parameters. This approach is different to Longdill et al. (2008) and Vincenzi et al. (2006), who first standardised each parameter based on PSSF values and multiplied by the site suitability constraint to produce an overall site suitability index, violating the argument of Malcezewski (2000). The mathematical expression for the overall site suitability index for MPAs using PSSF is as follows (Eq. 2):

th Where: PSSF(x,y,i) is the standardised suitability value of criteria i at coordinates (cell) x,y; n is the number of criteria used.

The use of a geometric mean in this site suitability overlay means that the final site suitability values range between 0 – 1 and will always be lower than the arithmetic mean. The importance of the geometric mean, however, lies in the argument that if a site scores 0 for any parameter, the final score will be 0 regardless of what other parameters achieve (Longdill et al., 2008). This is particularly important in terms of site selection for MPA mariculture zones with very rigid conservation zones and sensitive biodiversity. For example, a core zone within an MPA might score a perfect geometric mean of 1, from multiple water quality and other parameters, to sustain a mariculture operation. However, a core zone might score 0 under a single access parameter, as only research activities can be conducted within the area resulting in the overall suitability class for a core zone being 0 or unsuitable. This study preferred to use PSSFs over the other well-known standardisations, such as Boolean Logic (Ehrgott et al., 2010) or Fuzzy Logic (Navas et al., 2011b), for several reasons: the arbitrary classification in PSSFs is more flexible compared to the rigid classification of Boolean logic to compensate for comparing a wide range of criteria (Longdill et al., 2008); at the same time, PSSFs offers more “firmness” compared to Fuzzy logic which uses continuous classification to incorporate uncertainty. The continuous gradual classification and cross-membership inherent in Fuzzy Logic (Navas et al., 2011b) is arguably incompatible with MPA zoning systems, which require clear-cut classes for site

117 suitability levels. For example, a buffer zone encompassing a core zone might have a gradual suitability according to Fuzzy Logic, while the buffer zone will only have one arbitrary suitability value using PSSFs. Similar to Longdill et al. (2008), the PSSFs used here do not use weighted geometric means to indicate the relative importance of a particular parameter over another. The application of weighting in the MCE overlay processes has been shown to introduce bias and subjectivity, especially from consulted third parties, such as experts and scholars (Levings et al., 1995, Nath et al., 2000, Longdill et al., 2008). Instead, this weighting is replaced by a third stage, the stakeholder sub-model incorporating the participation of the primary users, fish farmers, to include their preference for the final MPA mariculture zone. As mentioned in Chapter 3, this final stage of decision making for mariculture zones in MPAs is reserved for the fish farmers from local communities. This approach will result in the suitability classes (SI) of: good being > 0.75 – 1; medium being 0.51 - 0.75; poor being 0.26 - 0.50; and unsuitable being < 0.25. The overall site suitability index classification, produced from the PSSF model, was based on the site suitability classification developed by Longdill et al. (2008), and consistent with other classifications schemes in the site suitability literature (Cross and Kingzett, 1992, Pérez et al., 2005, Longdill et al., 2008, Silva et al., 2011). In addition, the classification is also consistent with the widest range of PSSF values of criteria used in the model.

6.2.2 MPA finfish mariculture zone site suitability

This first stage used multiple sources of spatial data including vector, raster, hardcopy maps, government documents and graphics, which varied in resolution (max. resolution 1 km). All vector datasets were converted into raster datasets with spatial resolution set at 25 m. Similarly, all raster datasets that had resolutions lower or higher than 25 m were interpolated using a Simple Kriging technique (Burrough and McDonnell, 1998) to derive the full coverage of each dataset for the whole study area (Vincenzi et al., 2006). The use of 25 m resolution was based on the lowest resolution needed to sufficiently display the area required for small boats/ships to maneuver around local harbours (200 m), used in the constraint sub-model. Hardcopies of maps, government documents and

118 graphics were scanned consistently at 600 dpi, using the industrial scanner HP DesignJet T930.

6.2.2.1 Constraint sub-model a. Legal constraint According to Silva et al. (2011), legal constraints limit the development of aquaculture in areas located outside of local government jurisdiction. Here, a similar definition of legal constraint was applied, where the areas located outside the autonomy of Anambas Archipelago District could not be used for mariculture activities. According to Law No. 32/2004, regarding Local Government Autonomy, the Anambas Archipelago District has jurisdiction over sea areas covering 4 miles, measured outward from island coastlines. Another legal constraint, included within this sub-model, was the shipping line for commercial vessels transiting the Anambas area. The geographic characteristic of the water ways between Anambas Archipelago islands is dominated by narrow straits and bays that are used by medium cargo and passenger ships. Such narrowness means that in most areas, there is only one shipping route available. The fieldwork survey tracked most of the shipping routes using a handheld GPS and an onboard GPS and determined that the maximum width of the shipping lanes in this area was approximately 500 m, which should be free from any obstacles including siting of mariculture. b. MPA zone constraint The MPA zone constraint is the areas allocated for the protection of core zones (no- take zone) and unique ecosystems within Anambas Archipelago MPA. These zones and unique ecosystems have been described in the MMAF decree No. 53/KEPMEN-KP/2014, regarding the establishment of Anambas Archipelago as an MPA. The MPA zone constraint consisted of two limiting factors, the core zone and turtle nesting areas. Since no buffer zone existed within the zoning arrangement of Anambas Archipelago MPA, this study uses a buffer zone of 1 km from the core zone boundary. This means that finfish mariculture siting is only constrained by core and buffer zones. This moderate approach is based on the possible extent of waste distribution and disease transmission, which is applied in Norway, Canada and Tasmania (Sim-Smith and Forsythe, 2013). 119

A different approach was considered to limit the development of mariculture zones in turtle nesting areas. There was an absence of literature regarding how far man-made objects should be placed from areas reserved for turtle nesting, and the effect of these structures on turtle navigational patterns when heading to and from nesting areas. Since the location of turtle nesting beaches in Anambas Archipelago MPA are far from populated areas, a 5 km buffer from identified nesting beaches was assumed sufficient to allow undisturbed nesting and navigation of turtles. c. Economic-social constraint The economic-social constraint relates to conflicting uses of the seascape in Anambas Archipelago MPA. Despite its status as an MPA, significant economic-social activities are taking place, or are currently planned and being prioritised by the local government, in the study area. These activities rely heavily on unimpeded access to harbours, which are the central economic-social hotspots in the area. Two types of harbours were identified during fieldwork and the long-term strategic spatial plan issued by the Anambas Archipelago District (Bappeda-Anambas, 2011), regional and local harbours. Considering the narrowness of bays and waters between islands in which regional harbours are mostly located, this study assumed a 1 km buffer zone around regional harbours is sufficient for medium-sized vessels to maneuver freely. Local harbours visited by small local boats require a smaller buffer area and 200 m was considered adequate. Due to the geographic location and extreme terrain profile, the western region of Anambas Archipelago, represented by Jemaja Island, has been prioritised for economic development. It is relatively close to Singapore and has the potential to be a transit harbour for international cargo vessels passing through the SCS (Anambas-Regency, 2012). The current plan for Jemaja Island, specifically the northern part, is focused on developing shipping yard services, marinas, a natural gas power plant, an international gas station, a large-scale oil and gas refinery, an international airport, and an Indonesian Navy seaport. Despite these economic developments seeming too visionary, they were declared in the 20- year development strategy of Anambas Archipelago District. Thus, any other developments within these areas are not allowed to overlap with current development plans.

120 d. Physical constraint The only physical constraint that limits the siting of any finfish net cages, considered in this study, was the depth of water in which finfish mariculture platforms could be installed. Different minimum water depths for floating finfish cages have been suggested by different authors, such as 5 m (Hargrave, 2002, Windupranata, 2007), 4 m (NACA, 1989), and 10 m (Navas et al., 2011a). In this study, a minimum of 4-m depth was chosen based on current practices for finfish mariculture in the study area, and following NACA (1989), which describe the common finfish mariculture practices in Asia. This constraint sub-model considered no upper depth limit, as depths beyond upper limits used previously (see Hargrave, 2002, Perez et al., 2003) can still be used, despite the increased costs. These values were compared with the measured tidal range and computed mean sea level to confirm water availability and construction feasibility during the lowest and highest tidal level.

6.2.2.2 Site suitability sub-model

The site suitability sub model used all criteria considered to significant affect achieving sustainable finfish mariculture activities within MPAs, in terms of the capability of physical, chemical, and biological conditions (McKindsey et al., 2006, Longdill et al., 2008). It also considers the minimisation of visual impacts of finfish mariculture platforms (Falconer et al., 2013) to the aesthetic view of the study area as an MPA. As the objective of this stage was to design a sustainable mariculture zone within an MPA, most criteria used in this sub-model cover the full spatial extent of the study area, are available, and are relatively inexpensive. However, some parameters must be measured in the field (in-situ and ex-situ), which, to cover the full extent of the MPA, would be time-consuming and resource-intensive. Thus, the previous constraint sub-model was used to exclude most areas deemed unsuitable, allowing in-situ and ex-situ measurements to be focused on the feasible areas. The constraint sub-model ensured that all important biota within the core zones are protected and received minimal effects from mariculture activity such as coral reef, seagrass and other benthic biota. This study did not include coral reef, sea grass and other benthic biota coverage outside the core zones to be considered as one of suitability criteria due to the following reasons. Firstly, coral reef, sea grass and other benthic biota are 121 distributed homogenously in Anambas Archipelago which essentially renders all areas unsuitable if applied. Secondly, the purpose of the study was to design a mariculture zone in which the exact location of a mariculture platform is the fish farmers’ decision and requires the approval from the MPA management. Thirdly, sufficient current speed in Anambas Archipelago (Anderson, 2014, Purba et al., 2014) and net cage bottom clearance, as well as the type of small-scale activity proposed by this study, indicated minimal impact of mariculture to benthic biota as suggested by Silva-Cruz et al. (2011) and Alongi et al. (2009). However, this specific approach has to be considered on case by case according to the condition of each MPA. a. Bathymetry

Outside the constraint limits, bathymetry levels have different suitability classifications for sustaining mariculture activities, with varying influence over waste dispersion from mariculture platforms (Navas et al., 2011a, Valle-Levinson, 2013). Bathymetry also significantly affects the investment cost and technical construction of mariculture cages. The bathymetry profile of Anambas Archipelago was created by on- screen digitising in ArcGIS 10.3, using two scanned hydrographic charts, #181 (1:75,000) and #182 (175.000), issued by the Hydro-Oceanography Agency, Indonesian Naval Army (TNI-AL). The density of bathymetry points was increased and refined by combining the on-screen digitising results with field measurement and a secondary bathymetry dataset produced by Pemda-Anambas and Bakosurtanal (2011). Final bathymetry points were then interpolated using Simple Kriging to identify a full bathymetry extent of the study area. Within the feasible areas, bathymetry was classified into three PSSF classes (poor, good, and best) for floating finfish net cages (Table 6.1). These classes were derived from bathymetry classification for finfish mariculture suggested by NACA (1989) and Windupranata (2007). For example, the best class for the bathymetry is a water depth between 5 – 15 m at mean sea level (MSL). Given the Anambas tidal range is between 1.35 m (Purba et al., 2014) and 1.52 m (Fieldwork measurement in 2016), there is a minimum of 2.24 m water clearance beneath floating sea cages during mean lower low water (MLLW) to allow good water circulation. The “unsuitable” class was not counted in this sub-model as it has been included within the constraint sub-model.

122 b. Remotely sensed SST, chl-a and suspended solid

The annual temperature cycle plays an important role in the assimilation rate of released faecal waste into the environment and fish metabolic processes for growth (White et al., 2010, Loka et al., 2012). However, Chl-a (chlorophyll-a) and suspended solids are rarely considered in site suitability analysis for finfish mariculture due to their indirect influence on the culturing of carnivorous species. Chl-a concentration can be linked to increased risk of anoxia and hypoxia, due to harmful algal bloom (HAB), in fish farms (Bricker et al., 2003, Wang et al., 2008). Suspended solids (SS), on the other hand, can cause severe gill damage and increase fish susceptibility to diseases (Chen et al., 1993); SS can also lead to HAB due to increased nutrient loading in the water (Brager et al., 2015). The suitability classification for the spatial and temporal temperature variation within the study area was determined using remotely sensed sea surface temperature (SST) (Level-2 dataset from the MODIS-Aqua sensor, with 1 km resolution, from the Distribution Active Archive Centre (DAAC); Goddard Space Flight Centre (GSFC); and National Aeronautics and Space Administration (NASA)). A total of 1095 daily images MODIS- Aqua images were collected, from a 4-year period over July 2010 – July 2013. Since the study area has high cloud coverage all year round, only images with less than 20% cloud coverage combined with best and good SST values were selected using SeaDAS 7.3.2 (GSFC/NASA, USA). The images were then orthorectified to local the UTM Zone and extracted into text files. The application of cloud cover, and best and good SST masks in SeaDAS 7.3.2 produced raster images with invalid and zero values in areas of high cloud coverage and bad SST. To prevent the misinterpretation and data classification in the subsequent overlay process, missing and invalid data within these areas were replaced with predicted data using Simple Kriging interpolation in ArcGIS 10.3. There were 178 SST images produced from this stage, consisting of 83 images for the dry season and 95 images for the wet season. Remotely-sensed data from the Sea-viewing Wide Field-of-View Sensor (SeaWiFS) from DAAC/GSFC/NASA were used to derive level-2 product Chl-a and six water leaving radiances, which have resolutions of 1 km. Similar steps were performed to produce full extent coverage images of Chl-a and SS distribution, as described for SST. However, an intermediate step was performed before the interpolation process for SS spatial and

123 temporal variation to replace invalid and missing data due to cloud cover. Using this step, the value of SS was a proxy value calculated from nLw (555), one of six water-leaving radiance, using Ahn’s (2001) equation, specifically designed for case-2 water (shallow and coastal waters).

0.95 SS = 3.18 * nLw(555)

Where nLw(555) is the normalised water-leaving radiance at 555 nm. In total, there were 270 images to be used in the next site suitability classification, consisting of 118 images for the dry season and 152 images for the wet season. The PSSF values for Chl-a followed the four trophic classifications for a marine aquatic system, based on Chl-a concentration, suggested by Bricker et al. (2003) and USEPA (2003). In this classification, high concentrations of Chl-a in coastal marine waters poses higher risks to finfish mariculture, as previously discussed. Furthermore, PSSF values for SS were categorised into suitability classes based on NACA (1989) classification for finfish mariculture. c. In-situ and ex-situ criteria

Due to the remoteness of, and limited access to, the study area, as well as the cost of collecting and analysing samples, only three ex-situ criteria for nutrients (nitrate, ammonia, orthophosphate) and three heavy metals (mercury, copper, and lead) were used. These nutrient and heavy metal samples have a holding time up to 28 days with freezing and acidification, which allows sufficient transportation time from the field to the laboratory for further analyses (EPA-Victoria, 2009). Another five criteria were measured in-situ, including current speed, pH, dissolved oxygen (DO), salinity and temperature. Sampling campaigns were conducted twice in 2016, at 125 sampling stations, to represent seasonal variation within Anambas MPA. The first field sampling was conducted from May to June 2016, representing the dry season (southwest monsoon), with the second conducted from August to October 2016, representing the wet season (northeast monsoon). Since the study area covers a large area (approx. 80 x 120 km2), the in-situ measurements and ex-situ sampling collections were carried out only in areas that were feasible, based on the constraint sub-model. In addition, measurements and sampling were focused in areas of incoming water mass for each monsoon period (south and north boundary), to better predict 124 the variation of water quality parameters within the inner seas of Anambas MPA. This was necessary to avoid extensive measurement over a larger area, which would be time consuming and expensive. The selected sampling stations were determined by creating a 1 km cell-grid covering the extent of the study area but limited by the feasible constraint area. All water samples for ex-situ parameters were collected using a vertical Van Dorn beta sampler at the surface depth of 5 m during daylight, consistently from 10 am to 4 pm. After collection, the water sample from each sampling station was immediately stored in 1- litre polyethylene bottles and cooled in two ice boxes that were pre-chilled with ice blocks. At the end of each sampling day, all samples were filtered using 0.45 μm cellulose acetate membrane filters (Pall Corp GN-6 Metricel® 47mm) fitted in a Nalgene® pressure filtering unit and placed into 500 ml (for heavy metals) and individual 250 ml (for nutrients) containers. After filtration, heavy metal samples were acidified using nitric acid (HNO3) to pH <2. All samples were labelled according to the collection date and location to ensure that future sample analysis was conducted within the holding period. All samples were kept frozen during storage in Anambas. Samples were also kept frozen during the 4 days of transportation from Anambas MPA to Bogor Agricultural University for further laboratory analysis using PELNI ship’s cold storage at -20oC. These methods of water sample preparation, transportation, and laboratory analysis followed the standard protocols set by EPA-Victoria (2009) and APHA (2012). The measurement of in-situ parameters was carried out after water sampling at each sampling station. Salinity, temperature, pH, and DO were measured at 5-m depth from the surface using a handheld YSI Professional Plus. Surface current velocity was measured at 5-m depth from the surface using a Valeport Current Meter 105. Each parameter was interpolated using Simple Kriging interpolation (Burrough and McDonnell, 1998, Vincenzi et al., 2006, Longdill et al., 2008) with the extent of interpolation restricted to the feasible area. The interpolation result for each parameter was then classified based on PSSF classes described in Table 6.1. Due to the limitation of several sampling stations, some extrapolation occurred in areas with no sampling stations, especially around smaller islands of the Anambas MPA. However, as the objective of this site suitability study is to design an MPA mariculture zone, the data densities and resolution are deemed sufficient to achieve the objective of the study. In addition, the in-situ and ex-

125 situ measurements completed in this study were considered the best available data for this data-poor area. d. Visual amenity The appearance of a mariculture platform is a contentious issue in relation its environmental effects, especially in production-oriented, emerging economy, countries (Falconer et al., 2013). However, the visual disturbance of mariculture structures in coastal areas is clear and felt in areas where the scenic and unspoiled view of nature is of great importance (Perez et al., 2003, Falconer et al., 2013), such as in MPAs. This study incorporated visual impact site suitability for finfish mariculture in Anambas MPA, based on three criteria; beach, road and important viewpoints. These three components of visual amenity were applied to categorise suitability classes of seascape for finfish mariculture using visibility analysis. Using this approach, the suitability class of any point on the seascape, for finfish net cages, is determined by the level of disturbances to the visual receptor viewing from beaches, roads and important viewpoints. The visibility analysis used here follows the method suggested by Falconer et al. (2013). This method assumed that the height of the observers was 2 m at every point/pixel of the MPA features. The height of finfish mariculture platforms from sea surface (floating net cages) was 5 m. Since beaches, roads and important viewpoint are located at different elevations from the sea level in Anambas, a digital elevation model (DEM), with 25 m resolution, was created in ArcGIS 10.3, for four scenes of imagery from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) Global Digital Elevation Model Version 2 (GDEM V2), acquisition date of 8 December 2014. The suitability classes developed in this sub-model used a modified visual amenity sub-model from Falconer et al. (2013). Instead of using a Boolean viewshed, as suggested by Falconer et al. (2013), this research developed PSSF values for visual amenity in which different visual distances represent different suitability classes (Table 6.1), modified from Perez et al. (2003). Despite using different distance classes compared to Perez et al. (2003), this study used the same principle, that the farther the mariculture platform is from this important MPA feature, the better the PSSF values in terms of site suitability. The dataset of beaches and important viewpoints locations (36 tourism beaches and 9 viewpoints) was derived from both spatial and aspatial datasets from the Anambas Tourism Agency (DKP-

126

Anambas and LPPM-IPB, 2011b) and field observations. A road network dataset was reproduced from DKP-Anambas and BIG (2011).

Table 6.1. Parameter-specific suitability function (PSSFs) values used in the site suitability sub-model for finfish mariculture in Anambas MPA

Parameter/criteria Data value Unit PSSF value Source Bathymetry for >0 - <3 0 (NACA, 1989, Windupranata, 2007) floating net cage >3 – <5 ; >20 m 0.33 >15 - <20 0.67 >5 - <15 1 Temperature <15 ; >35 oC 0 (NACA, 1989, Windupranata, 2007) >15 - <20 ; >33 - <35 0.33 >20 - <27 ; >31 - <33 0.67 >27 - <31 1 Suspended solids >10 mg/L 0 (NACA, 1989) <5 1 Chl-a >5 (Hypertrophic) µg/L 0 (Bricker et al., 2003, USEPA, 2003) >3 - <5 (Eutrophic) 0.33 >1 - <3 (Mesotrophic) 0.67 >0 - <1 (Oligotrophic) 1 Ammonia (NH3-N) >0.5 (Unacceptable) mg/L 0 (PHILMINAQ, 2008, Colt and >0.01- <0.5 (permissible) 0.67 Huguenin, 2002, NACA, 1989) <0.01 (best) 1 Nitrate (NO3-N) > 200 (Unacceptable) mg/L 0 (NACA, 1989, Windupranata, 2007, >100 - <200 (permissible) 0.67 PHILMINAQ, 2008) <100 (best) 1 Ortho-phosphate >0.05 (Unacceptable) mg/L 0 (PHILMINAQ, 2008) (PO4-P) <0.05 (Best) 1 Mercury (Hg) > 1 (unacceptable) mg/L 0 (ANZECC, 2000) < 1 (acceptable) 1 Copper (Cu) > 5 (unacceptable) mg/L 0 (ANZECC, 2000) < 5 (acceptable) 1 Lead (Pb) > 1 (unacceptable) mg/L 0 (ANZECC, 2000) < 1 (acceptable) 1 Current speed < 0.05 - > 1 (unacceptable) m/s 0 (NACA, 1989, Windupranata, 2007, >0.05- <0.2 (poor) 0.33 Halide et al., 2009) >0.5 - <1 (good) 0.67 >0.2 - <0.5 (best) 1 pH < 3 ; > 13 (unacceptable) 0 (NACA, 1989, Windupranata, 2007) >3 - <7 ; > 8.5 - <13 (poor) 0.33 >7 - <8.5 (best) 1 Dissolved oxygen <3 (unacceptable) mg/L 0 (NACA, 1989, Windupranata, 2007, >3 - <5 (good) 0.67 Halide et al., 2009) >5 (best) 1 Salinity <10 (unacceptable) ppt 0 (Windupranata, 2007) >10 - <15 (poor) 0.33 >15 - <25; >35 (good) 0.67 >25 - <35 (best) 1 Beach visual <1 km 0.33 (Falconer et al., 2013) and estimated amenity >1 - <2.5 0.67 visual receptor distance and radius of >2.5 1 tourism activity Road visual amenity <0.5 km 0.33 (Falconer et al., 2013) and estimated >0.5- <2.5 0.67 visual receptor distance while moving >2.5 1 Viewpoint visual <1 km 0.33 (Falconer et al., 2013) and estimated amenity >1 - <2.5 0.67 visual receptor distance and radius of >2.5 1 tourism activity

127

6.2.2.3 Stakeholder preference sub-model

As discussed in Chapter 3, local community involvement (small-scale fish farmer groups) in designing an MPA mariculture zone should be mandatory, to secure seascape access for the group and increase their bargaining position when leasing the zones to other parties. The stakeholder preference sub-model was the final stage of defining the site suitability for mariculture zones in the Anambas MPA. This sub-model considered small- scale fish farmer’s preferences important for the decision-making process, to ensure optimal and fair use of these zones. For this sub-model, a short face to face interview was conducted, from October to November 2016, to elicit preferred locations of mariculture zones in Anambas MPA from small-scale fish farmer groups. It involved 50 participants from 8 villages where most of the mariculture activities in Anambas are concentrated. All the participants were identified and selected from the previous sustainability interviews (see Chapter 3 & 4). The questionnaire was designed to collect the preference of the participants regarding the suitability of the MPA for mariculture, based on the distance from several selected features related to finfish culture, access, and MPAs. From the criteria proposed, the respondents could only provide consistent answers for three criteria; distances to the villages from floating and fixed net cages, MPA zones, and piers/jetty (Appendix 2). Each criterion has different distance ranges that measure how far any possible mariculture zone/location are from specific features representing criterion (Table 6.2). Then, the participants were asked which preferred distance range, with their responses ranked using simple ranking. For example, the distance most respondents preferred for a criterion was categorised as best, scoring the best PSSF value of 1, the least preferred distance value of a criterion was classified as 0 (unsuitable). For the next step in this sub-model, a Euclidean Distance Analysis was performed on the area deemed feasible for finfish mariculture from the first stage (constraint sub- model) for each criterion. This analysis measures the straight-path distance of every point/cell within feasible locations of the MPA to corresponding locations of villages, piers and core zones. The outputs of this step were then standardised according to the PSSF values from the stakeholder preference sub-model (Table 6.2).

128

Table 6.2. Parameter-specific Suitability Function (PSSFs) values used in stakeholder preference sub-models for finfish mariculture in Anambas MPA

Parameters/criteria Data value Unit PSSF value Source Distance to village for <0.5 ; >10 km 0 Fieldwork floating net cage >5 <10 0.33 survey 2016 >2 <5 0.67 >0.5 <2 1 Distance to piers/jetty <0.1 km 0 Fieldwork >5 0.33 survey 2016 >2 <5 0.67 >0.1 <2 1 Distance to core zone <1 km 0 Fieldwork >1 <2 0.33 survey 2016 >2 <5 0.67 >5 1

6.2.2.4 Identification of suitable areas for MPA mariculture zones

As previously described in subsection 6.2.1, to identify suitable areas for the MPA mariculture zone, the unweighted geometric mean overlay suggested by Vincenzi et al. (2006), Longdill et al. (2008) and Silva et al. (2011) was used; this produced the overall site suitability. An option to use a weighted geometric mean was available through the use of expert judgment to differentiate the relative importance of one criterion over the other. However, different experts with different backgrounds come out with different judgements (Levings et al., 1995, Nath et al., 2000, Longdill et al., 2008, Silva et al., 2011), and most of the time, have little attachment to, or authority over, the results of the decision making (Szuster and Albasri, 2010). Instead, a stakeholder preference sub-model was introduced, which serves as the final sub-model to replace the weighting effect of expert judgments. The introduction of this sub-model after the constraint and site suitability sub-models were completed, means this sub-model will greatly influence the site suitability classes produced by the two former sub-classes. The reclassification of the final model output to produce the overall site suitability used the following suitability class structures: unsuitable, poor, good, and best; with corresponding SI indexes of: <0.26, 0.26–0.50, 0.51–0.75, and, >0.75. These suitability classes were consistent with suitability classifications proposed, or used, by Cross and Kingzett (1992), Pérez et al. (2005), Longdill et al. (2008).

129

6.3 Results

6.3.1 Constraint sub-model

The constraint sub-model indicated that only 10.16% (1,199.70 km2) of the total extent of the study area (11,811.77 km2) can be categorised as feasible for mariculture (suitable area) (Figure 6.3). Bathymetry is the most significant constraint on the development of finfish mariculture and shows that 84.5% of the area is classified unsuitable due to a depth of less than 4 m. Four metres is considered the minimum depth required to install a floating net cage. Further compounding this is the 4-mile district economic zone (54%, 6,379 km2), turtle nesting sites (2.6%, 303.2 km2), MPA core zones (1.6%, 194.7 km2), industrial zone (0.8%, 9.16 km2) and harbours (0.1%, 10.4 km2). Turtle nesting sites and MPA core zones have minimal effect of reducing the overall extent of feasible area for finfish mariculture, indicated by the small coverage areas excluded from the constraint analysis. However, the inclusion of these two constraints excluded several important MPA features entirely from future development of finfish mariculture. For example, the entire Telaga Island waters (located between Jemaja and Siantan islands), which has several fishing community villages, is mostly unsuitable for finfish mariculture due to the locations of turtle nesting sites on this island. In addition, shipping lanes used by small and medium commercial ships passing through the narrow straits of the islands significantly reduced the suitable area, particularly in straits located between Air Asuk and Matak Islands (Figure 6.3). Small-scale fish farmers have built hundreds of fixed net cages along both sides of the straits. However, these areas are narrow and busy shipping lanes which regularly trigger conflicts between both users due to fixed net cages being directly hit by ships, waves produced by passing ships causing damage, and difficulty for ships to safely maneuver between the net cages.

130

(A) (B)

Figure 6.3. Feasible areas from the constraint sub-model: (A) turtle nesting sites; (B) district economic zone; (C) bathymetry; (D) shipping lane 131

(E) (F)

(G) (H)

Figure 6.3. Continued… (E) core zone; (F) harbour; (G) industrial zone; & (H) composite maps

132

An industrial zone proposed by the government of Anambas Archipelago District also limits the suitability of the MPA for finfish mariculture in most nearshore waters of northern Jemaja Island. The proposal itself might contradict the current MPA zoning plan, as there are 5 core zones located as close as 2.3 km from the nearest potential industrial sites. The proposed uses of these areas, as mentioned in section 6.2.2.1, might have adverse effects on core zone ecosystems in the form of oil spillage impacts (Figure 6.4), as well as harmful substances, invasive species and diseases from ships ballast water discharges.

Figure 6.4. Oil discharge into the waters from a medium size transportation ship at the regional seaport of Jemaja Island

6.3.2 Site suitability sub-model

In general, most of the bio-physio-chemical criteria used to determine site suitability of MPA mariculture zones in Anambas showed optimum values within feasible areas, for both of dry and wet seasons. However, the level of ammonia concentration was consistently higher during the wet season compared to the dry season. Increased level of ammonia was observed in populated areas and semi-enclosed bays on the eastern side of Siantan Island and west of Jemaja Island. The amount of area classified as best class (PSSF value of 1) is considerably reduced during the wet season due to the increased concentration of ammonia

133

(Table 6.3). Despite the ammonia parameter, most of the feasible areas still fell within the medium PSSF class during the wet season.

Table 6.3. Distribution of total coverage of PSSF suitability classes in the site suitability sub-model Total coverage of suitability classes (%) during dry and wet seasons Criteria Not suitable Poor Medium Best Dry Wet Dry Wet Dry Wet Dry Wet Bathymetry 90.19 7.34 0.96 1.51 Sea surface temperature 90.31 90.19 0.00 0.00 0.00 0.00 9.69 9.81 Suspended solids 89.82 89.84 0.00 0.00 0.00 0.00 10.18 10.16 Chlorophyll-a 91.78 90.01 0.00 0.00 0.00 0.00 8.22 9.99 Ammonia 89.84 89.84 0.00 0.00 2.18 10.16 7.98 0.001 Current speed 89.84 89.83 7.44 5.18 0.00 0.00 2. 72 4.98 Dissolved oxygen 89.87 89.84 0.00 0.00 3.45 4.50 6.68 5.65 Lead 89.84 89.84 0.00 0.00 0.00 0.00 10.16 10.16 Mercury 89.83 89.84 0.00 0.00 0.00 0.00 10.17 10.16 Copper 89.84 89.84 0.00 0.00 0.00 0.00 10.16 10.16 Nitrate 89.84 89.84 0.00 0.00 0.00 0.00 10.16 10.16 Orthophosphate 89.84 89.84 0.00 0.00 0.00 0.00 10.16 10.16 pH 89.84 89.84 0.00 0.03 0.00 0.00 10.16 10.13 Salinity 89.85 89.84 0.00 0.01 0.04 0.03 10.11 10.12 Viewpoint visual amenity 89.84 0.02 0.10 10.04 Beach visual amenity 89.84 0.25 0.41 9.49 Road visual amenity 89.89 0.04 0.28 9.79 Note: Single values of criteria in both dry and wet season columns indicate that the criteria suitability coverage remain the same during the seasons

In addition, the narrow depth ranges in the bathymetry classes for net cages have also considerably reduced the amount of suitable area. For the site suitability model, only 14.86% (178.32 km2) of the feasible areas were estimated to be the best class in terms of PSSF bathymetry parameters. Most of the areas (72.27% or 866.99 km2) were considered poor in terms of PSSF bathymetry class. For current velocity, the results show that the dry season has worse site suitability classification within feasible areas compared to the wet season. Based on the site suitability model, the area classified as the best covers only 26.8% (321 km2) for the dry season compared to 49.1% (588.8 km2) for the wet season. The suitability classification for dissolved oxygen (DO) indicates there is more area classed as best during the dry season (789 km2) compared to that of the wet season (667 km2). Most of areas of DO best class were on the southern Anambas part of Anambas during the dry season compared to more evenly distributed DO around Anambas MPA

134 during the wet season. The average DO level measured during the dry season was slightly lower (4.78 mg/L compared to the wet season (4.99 mg/L). Low DO levels were observed within the inner waters of island cluster of Anambas MPA where water flows slow down due to numerous fringing reefs and sand dunes, particularly during the dry season. The results of the distribution of other parameters (spatial and temporal) were similar for both seasons, in which the classification of PSSF best values cover 79.2% to 99.9% of the total feasible area. One exception was the temporal distribution of chlorophyll-a (Chl-a) during the month of January around Jemaja Island. During January, Chl-a concentration peaked between 2.77 - 6.59 mg/L. This particular incident might have triggered a hypertrophic condition potentially leading to a HAB (red tides). This event, during the peak of the wet season, gives an indication that excessive runoff material from land cleared for clove plantations may have occurred within these areas. The inclusion of visual amenity parameters (beach, roads and viewpoints) within the site suitability sub model gave an insignificant contribution to the overall PSSF suitability classes. In total, these three parameters classified 37.7 km2 of feasible areas as poor and 93.5 km2 as medium, due to the close proximity of feasible areas to visual amenity features. However, these parameters provided a guarantee that future finfish mariculture development will not overlap with, or occur within, the visual vicinity of these features as tourism destinations. Figure 6.5 shows that wet and dry seasons influence the distribution of best PSSF class (reddish colour) within the study area. In the wet season, most of the areas classified as best were located on the southern side of Siantan Island and the eastern side of Jemaja Island. In the dry season, the best areas shifted to the northern coast of small islands, south of Siantan and Matak Islands, while the best areas were in Jemaja Island which had a relatively consistent distribution as in the wet season. Despite the difference in the locations of best suitable areas, the distribution of the final index values for both seasons for the site suitability sub model shows high suitability for finfish net cage mariculture. For the wet season, all PSSF index values within the feasible areas ranged from 0.71 to 0.97, while for the dry season were between 0.73 – 1. In some areas, such as on the east of Jemaja Island and on the west of Siantan Island, the high suitability areas extend up to 7 km and 12 km, respectively, from the shorelines where the

135 nearest villages were located. The high site suitability bathymetry profile, in combination with the overall optimal biochemical properties of waters of these areas, has produced high values for the PSSF index despite being located far from the shoreline.

(A)

(B)

Figure 6.5. The composite maps of PSSF classes for wet (A) and dry (B) seasons in Anambas MPA based on site suitability sub model of 17 parameters 136

In addition, the areas that have best PSFF index values were mainly distributed near most of the locations of Anambas MPA core zones. In particular, consistent best PSSF index values, for both dry and wet seasons, were observed surrounding five core zones on the south and southeast of Siantan and Matak Islands. The 1 km buffer zone that was applied in the constraint sub-model did not effectively limit the impacts of finfish net cage activities on core zones. This is inevitable as the site suitability sub-model is primarily focused on what is the best area to support finfish net cage mariculture in the MPA. The result of the site suitability models presented in Table 6.3 and Figure 6.5 also contain fragments of areas produced from extrapolation results of the Kriging interpolation. As described in the method section, there were only 125 sampling stations used to measure 10 in-situ and ex-situ parameters, due to research limitations (fund and technical resources), as well as the large extent of the study area. This extrapolation occurred on the western side of Jemaja Island and around most of the small islands in the south of the study area where measurement could not be made. However, the extrapolation was forced to take place in the analysis to avoid no data or zero values in the extrapolated areas as a consequence of using a geometric mean in the overlay process. Furthermore, the variation of values shows the similarity between the measured and the predicted (extrapolated) values of each parameter. Table 6.4 confirms this, where the mean values between the measured and predicted dataset for each parameter were relatively similar. However, the standard deviations between the measured and predicted datasets of each parameter are significantly different. Lower standard deviations in the predicted data sets indicate the interpolation process (Simple Kriging Interpolation) tends to maintain homogeneity of the predicted values with the measured values used in the interpolation. This also means that the variation of predicted parameters values in the extrapolated areas will be within the normal variation of the measured ones.

137

Table 6.4. Mean and ± standard deviation for criteria measured in 125 sampling stations during the wet and dry season

Wet season Dry season Criteria Measured Predicted Measured Predicted Mean Stdev Mean Stdev Mean Stdev Mean Stdev Current 0.2 ± 0.17 0.2 ± 0.054 0.11 ± 0.07 0.123 ± 0.027 pH 8.11 ± 0.52 8.15 ± 0.11 8.11 ± 0.45 8.06 ± 0.044 DO 5 ± 1.42 5.068 ± 0.475 4.83 ± 1.33 5.19 ± 0.854 Salinity 30.94 ± 4.91 31.75 ± 0.99 29.46 ± 5.46 30.39 ± 0.92 Ammonia 0.029 ± 0.028 0.031 ± 0.008 0.01 ± 0.04 0.0173 ± 0.018 Nitrate 0.11 ± 0.022 0.107 ± 0.0088 0.11 ± 0.07 0.1 ± 0.02 Ortho-P 0.023 ± 0.0178 0.0065 ± 0.0016 0.0049 ± 0.0032 0.0045 ± 0.00075 Copper 0.003 ± 0 0.003 ± 2.115 0.003 ± 0 0.003 ± 2.11 Lead 0.007 ± 0.0013 0.0068 ± 0.00075 0.0099 ± 0.0032 0.0009 ± 0.00178 Mercury 0.00033 ± 0.00039 0.00032 ± 0.00018 0.00041 ± 0.00038 0.00045 0.00020

6.3.3 Stakeholder preference sub-model

The stakeholder preference sub-model aimed to incorporate local community preferences in the final decision making of designing a fair mariculture zone within an MPA. Besides easier access to the potential sites (distance to local villages) and better access to transportation points/market (distance to harbour), the sub-model also includes distance to core zone criterion. The latter criterion was included in this sub model to further address the possible impacts of finfish net cages near MPA core zones, described in the earlier sub-model. The result of the stakeholder preference sub-model indicates that only 2.77% (33.21 km2) of the feasible areas were classified as best areas (PSSF index value of 1) (Figure 6.6). These areas are mostly in sheltered locations such as bays and waters between islands, where villages and harbours are commonly found, which increase the site suitability potential. This sub-model approach indirectly includes a geographic site selection parameter in which sheltered areas have high suitability for finfish net cage mariculture. In addition, the sub-model successfully steered any future mariculture activity away from most of the MPA core zones.

138

(A (B ) )

(C (D ) )

Figure 6.6. Site suitability of Anambas MPA based on stakeholder preference sub-model: (A) distance to MPA core zones; (B) distance to harbours; (C) distance to village/settlements; & (D) composite map of stakeholder preference sub-model

Suitability levels decreased from 0.87 – 0.33 as feasible areas gradually increase in distance from villages and harbours, or decrease in distance to MPA core zones. As expected, most of the areas with a medium to low PSSF value (<0.67), under this sub- model, were located in areas where direct influence of wind and waves were suspected to occur, either during the wet or dry season. The areas considered highly suitable under this sub-model (PSSF values of >0.67 - <0.87) had a similar distribution to areas considered as best. Most of these areas were also located in bays and waters between islands which offered relatively better protection from wind and waves compared to the low to medium suitable areas. However, the distance of these areas to villages (2-5 km) will discourage

139 local small-scale fish farmers from using them as fish farming areas, due to increased effort and time needed to access mariculture platforms. In addition, as mentioned in the site suitability sub-model, there were areas still considered highly suitable in terms of bio-physio-chemical criteria despite their location far from the nearest villages. Through this sub-model, these areas are excluded from the being feasible areas (23.55 km2). Due to their distance, they are unreachable and difficult to maintain for local fish farmers who rely on traditional finfish mariculture technology.

6.3.4 Mariculture zone for finfish net cage mariculture

The final site suitability maps, for both dry and wet seasons, based on the geometric mean overlay of site suitability and stakeholder preference sub-models are presented in Figure 6.7.A and Figure 6.7.B, respectively. The structure of suitability classes used in these final maps was based on the range of index values representing unsuitable, poor, good and best areas, as laid out in the method section. These maps indicate that there is only a slight difference in the coverage of each site suitability class between the different seasons. Site suitability areas during the dry season showed slightly larger coverage for each suitability class compared to the wet season, thus better suitability for finfish net cage mariculture. As mentioned in Section 6.3.2 (site suitability sub-model), the reason for less suitable area during the wet season, compared to the dry season, was due to recurring high turbidity. These temporal events occurred in some areas located in the southern and eastern coasts of Jemaja Island. Despite this event rarely occurring, only being observed once during the study, the nature of the geometric mean overlay maintains these areas as unsuitable for future finfish net cage development.

140

(A)

(B ) Figure 6.7. Site suitability maps for mariculture zone in Anambas MPA: (A) wet season, (B) dry season

141

Furthermore, the areas classified as the best suitability covers a total of 32.3% (387 km2) for the dry season compared to 30.1% (361.6 km2) for the wet season. Areas with good suitability covers a total area of 41.25% (494.9 km2) for the dry season compared to 40.1% (481.65 km2) for the wet season. It is important to note that best suitability here is different to that of the best PSSF value described in the site suitability sub-model. The best PSSF values in the site suitability sub-model are based on the maximum overall index value within the sub-model achieved by one-pixel area, compared to this final map where the best suitability is based on a particular range (>0.76 – 1). This means that an area that falls under the best suitability might overall have upper medium to good PSSF values in the site suitability sub-model (>0.57 – 0.75). However, stakeholder (small-scale fish farmers) decisions that indicate an area has the best PSSF value (1), in terms of safety and easy access, will shift the area into the best suitability for MPA mariculture zones. For this reason, the best suitability class in the final site suitability maps for MPA mariculture zones covers significantly larger areas compared to the best PSSF class in the site suitability sub-model. In addition, the optimal condition of all parameters, in terms of spatial and temporal variations, has resulted in no poor suitable areas (overall index values of >0.26 – 0.5) identified in the final site suitability maps for both dry and wet seasons.

6.4 Discussion

6.4.1 GIS application in mariculture site suitability for MPA

Integrating various criteria to assess and design MPA mariculture zones have shown that GIS and remote sensing, coupled with an MCE technique, can produce valuable information and solve decision-making problems relating to the incompatibility and scarcity of datasets. This finding is consistent with other studies, such as Vincenzi et al. (2006), Longdill et al. (2008), Pérez et al. (2005), Silva et al. (2011), and Radiarta et al. (2008). Despite this, previous studies mostly dealt with mariculture site selection in regular coastal zones, with little involvement of the main stakeholder (fish farmers) in decision making. Accordingly, the current study contributes new knowledge on the integration of methods in a mariculture decision-making context for MPAs. The incorporation of the stakeholder preference sub-model to design mariculture zones in the MPA will increase key stakeholder acceptance of the results for the following

142 reasons. Firstly, one of the primary concerns of legalising mariculture development in MPAs is the potential overlap of seascape use with protected MPA ecosystems (core zones) and ecotourism industries embedded in most Indonesian MPAs. The constraint sub-model developed here addressed this problem by limiting the development of mariculture to certain distances from core zones and turtle nesting areas, two of the most common features in Indonesian MPAs. This is a different approach compared to other studies such as Longdill et al. (2008) and Silva et al. (2011) who considered a conservation area as a single area with no distinctive sub-zones within it. Furthermore, the site suitability sub-model with visual amenity analysis, based on the visual receptor of tourists, will limit or eliminate visual disturbance to tourism activities from mariculture developments. This particular concern has been steadily gaining attention due to the expansion of net-cage mariculture, which alters the scenic beauty of coastal areas (Falconer et al., 2013) and affects the value of the ecosystem to the tourism industry (Perez et al., 2003). However, the influence of visual amenity within the site suitability sub-model was considered low in terms of reducing the suitability of areas within the sub model. This can be attributed to the distant location of most visual amenity features (beaches, roads and viewpoints) from each other, reducing the overlapping field of view. As a result, the less suitable classes of visual amenity criteria only cover relatively small areas. One exception might be the waters off the Padang Melang Beach, where the significant length of the beach (>8 km) provides an overlapping field of view, reducing the suitability of most of the coastal waters within the visual limit to poor or even unsuitable. However, as more tourism development takes place in the future, more locations will be designated tourism destinations (beaches and viewpoints), with longer road networks, which will increase the overlapping field of view. The effects of overlapping field of view, in areas with dense tourism features, that significantly reduced the site suitability of a coastal area was shown by Falconer et al. (2013). They found that a particular point at sea has a significantly higher visual impact if it can be seen by different user groups from different viewpoints. Secondly, using GIS in tandem with remote sensing facilitates developing a more comprehensive site selection process, in terms of spatial and temporal coverage, and offers a more cost-effective approach for aquaculture (Radiarta et al., 2008, Longdill et al., 2008, Meaden and Aguilar-Manjarrez, 2013), particularly net cage mariculture. GIS and remote

143 sensing could be used to narrow down the areas suitable for net cage mariculture, both spatially and temporally, involving various quantitative and qualitative suitability parameters (Ross et al., 2013b). However, freely available remotely-sensed data, such as MODIS and SeaWiFs, are mostly used in site selection for oyster and scallop culture. This is due to the data derived from the sensors being compatibility for oyster and scallop as sessile and passive organisms (e.g. Silva et al. (2011), Longdill et al. (2008), Radiarta et al. (2008)). For net cages, some remote sensing datasets need to be transformed to be usable for the analysis described in this study. Here, Chl-a and water-leaving radiance (nLw555) were used as proxy values representing the trophic state risk and suspended solid condition of the waters, respectively, to support finfish net cage mariculture. A potentially increased risks of HAB was found during the north monsoon period, shown by the high Chl-a concentration in the month of January 2011. The high occurrence of Chl-a was sporadic and localised only in one area. Nevertheless, waters around Anambas are classified as having a high risk of HAB occurrence due to seasonal flow patterns and nutrient run-off from the islands (Purba et al., 2014). Thirdly, the stakeholder preference sub-model used here is specifically aimed to involve local small-scale fish farmers in the decision-making process for creating mariculture zones, of which they will be the main user. This involvement was achieved by deliberately combining the composite map of the stakeholder preference sub-model with the final composite maps from the site suitability sub-model. Despite no specific weights applied for either composite maps in the overlay process, adding the stakeholder preference sub-model to the final site suitability classification had a strong influence on the classification results. This is due to the fact that most of the biological, chemical and physical parameters have spatially (geographic location) and temporally (dry and wet seasons) consistent optimum values for supporting finfish net cage requirements. As a result of narrow variations in the high index values, the discrete index values within the stakeholder preference are prioritised in the aggregation process. The increased priority of the stakeholder preference in the aggregation process and the generally optimum values of criteria in the site suitability sub model, have resulted in no poor site suitability class identified in the final composite maps. Using similar methods, Longdill et al. (2008) also found comparable results where poor suitability class only accounted for 0.26% (6 km2) of

144 the total area (2338 km2). Although the present model successfully incorporates and indirectly prioritises fish farmer preferences, regarding the location of mariculture zones within Anambas MPA, this prioritisation might have some drawbacks. The most obvious is probably the ineffective use of the model in areas where site suitability parameters are highly variable due to, for example, freshwater input and runoff such as in coastal areas of big land mass or estuaries. As the variation in suitability class, within the site suitability sub-model, increases, the influence of the discrete suitability class, the stakeholder preference sub-model, decreases. In this situation, a weighted linear combination, as suggested by Malcezewski (2000), should be used to prioritise any of the sub-models or parameters deemed more important or dominant. Another drawback is that the over prioritisation of the stakeholder preference sub-model might produce sub- or over-optimal site suitability classes. For example, an area close to villages might be considered highly suitable despite poor environmental conditions, or areas farther away could be deemed to have poor suitability. This has been a conflicting issue in GIS models that involve a combination of qualitative and quantitative data sets (Perez et al., 2003, Portman, 2007, Longdill et al., 2008). One way to address this drawback is through controlling culture stage and activities, such as stocking of new fish seeds during known good seasons (i.e. dry period) for locations near settlements. For locations far from settlements, regular watch shifts over one fish farm cluster can be arranged among the fish farmers. This is called reactive culture management by Tarunamulia (2014), who argues that social and environmental issues affecting site suitability can be solved through direct intervention.

6.4.2 Mariculture zone for small and medium-scale fish farmers in MPA

Based on the final composite maps, most of the areas categorised as the best class are distributed near the settlements/villages. On the other hand, the locations of good and unsuitable areas are located farther away, in a direction perpendicular to the shorelines. As one of the objectives of site selection is to design mariculture zones for the MPA, the final composite maps can be used to assign fair mariculture zones for both local fish farmers and medium-scale business entities, as discussed in Chapter 3. The first scenario is that the best site suitability class can be allocated to small-scale fish farmers. These areas are located within a distance fish farmers are willing to travel every day (from harbours and

145 settlements). The relative short distance from settlements and market features (harbours) will advantage small-scale mariculture, enabling it to be technologically and economically feasible (Pomeroy et al., 2006). In addition, the best site suitability class areas are also the farthest from the core zones. This is an important consideration, as small-scale fish farmers are usually numerous in number and difficult to manage due to educational, financial and attitudinal limitations (Rimmer et al., 2013). Even if you disregard the environmental effects from net cage mariculture, which would be present, at least the core zones and the surrounding areas are free from visual amenity disturbances caused by the net cages. On the other hand, the moderate site suitability class areas can be allocated to medium- scale fish farmers. The consideration of mariculture sites in these areas is for the following reasons. Medium-scale mariculture is usually limited in numbers (size and distribution) and regularly supervised by the local Fisheries Extension Office. Therefore, despite being relatively closer to core zones compared to small-scale fish farms, the environmental effects of medium-scale mariculture would be minimal, as reported by Alongi et al. (2009). In addition, the scale of operation and financial capability of medium- scale mariculture allows the construction of mariculture platforms and self-sustained operations to be situated some distance from the shorelines. Unlike small-scale fish farmers, medium-scale fish farming is relatively well equipped, with electricity for net cage maintenance and boats for transportation. This separate site arrangement for small- and medium-scale net cages will significantly minimise conflict over the seascape and ensure optimal use of potential sites for small-scale fish farmers within the MPA. Another site arrangement can also be devised for which both fish farmer types can use all suitable sites. In this second scenario, it is proposed that medium-scale fish farmers must purchase leasing rights for the best sites from small-scale fish farmers. As the location of best sites are correlated with the location of small-scale fish farmer villages, transferring rights of site use to medium-scale fish farmers will be easier and more transparent. Medium-scale mariculture operators can directly contact and negotiate with the right group of small-scale fish farmers regarding leasing rights, area size used, culture system and management of the sites. Only after leasing agreement has been finalised, medium-scale fish farmers can inquiries about fish farming permits through the responsible organisations (MMAF and MEF), as discussed in Chapter 3.

146

The site selection model developed here used various quantitative and qualitative datasets, of which some have high spatial and temporal resolution for coastal waters. However, with the aforementioned drawbacks, and the focus on designing mariculture zones within an MPA, the model results should be treated as a general guideline only for designing MPA zoning in which mariculture zones are included. The actual best location for building mariculture net cages should be conducted separately (Perez et al., 2003, McLeod et al., 2002), especially for medium-scale mariculture. The model also does not include the actual CC of each mariculture zone, which is important for providing more rigorous management of activities with respect to the sensitivity of MPAs. As discussed in Chapter 3, the CC should be determined for each mariculture zone to avoid an under- or over-estimation of CC, which usually occurs in CC estimation of large water bodies. Chapter 7 of this thesis will address this issue through the development of a specific CC model for mariculture zones within MPAs.

6.5 Conclusions

The use of the three sub-models in the sequenced stage of mariculture site selection analysis in this study resulted in a balanced approach for combining multiple criteria, both quantitative and qualitative, within the GIS framework. The overall site selection model developed here successfully addressed the potential overlapping use of the seascape within MPA SFZs. The model also able to exlude areas near a protected or unique MPA ecosystem to be developed for mariculture activities. However, the most important aspect of the model was the direct involvement of local fish farmers in the site selection decision- making process, which has been missing in most site selection models to date. The stakeholder preference sub-model was developed in order to avoid the dominance of environmental criteria in the mariculture site selection analysis. With the sequenced stages starting from developing a constraint sub-model, followed by a site suitability sub-model and finalised with the stakeholder preference sub-model, the prioritisation of the stakeholder preference sub-model can be maintained in an unweighted GIS linear combination. The consistent and optimal environmental conditions within an MPA also support the use of this model, as site suitability sub-model criteria variations were too small to overpower the significance of criteria within the stakeholder preference sub-model.

147

Applying a GIS environment within the model and the use of a geometric mean in the overlay process addresses conflicting space allocation within an MPA, especially with the rigid space allocation for MPA core zones. The geometric mean maintains a site as unsuitable for mariculture based on one parameter, regardless of any other parameters involved showing otherwise. This is of utmost importance within an MPA management regime, where seascape allocation has been designated for a particular purpose. As a result, it is concluded here that an arithmetic mean should not be used in site selection process for MPA zone classification, as the addition of this function can violate the allocation of seascape for other purposes. Furthermore, the overlay process using geometric mean to determine suitable sites has to be conducted within feasible cells determined previously from constraint sub-models. This is to avoid the overlay calculation performed in non- feasible cells which will influence the final overlay result. This proccess is regularly used in the GIS process using arithmetic mean overlay. However, it has not been standardised in GIS process based on geometric mean. The use of different freely-available temporal and spatial datasets (quantitative and qualitative) in this study shows that site selection for mariculture in a data-poor environment, such as an MPA in Indonesia, can still be conducted. However, this approach might introduce some bias as longer temporal and wider spatial datasets have to be compensated with shorter and narrower datasets through extrapolation in the GIS environment. Considering the application of the model to design mariculture zones within MPAs, and the limited availability of datasets in remote islands in Indonesia, the result should fairly describe the actual condition. In the end, a specific site selection analysis has to be conducted by fish farmers to determine the best locations based on their financial and technical capabilities. Finally, the site selection model for net-cage mariculture in MPAs proposed here has suggested two different scenarios for mariculture zoning within MPAs. The first scenario offers strict delineation between small- and medium-scale fish farming zones, with limited trade-off and thus little benefit to small-scale fish farmers. The second scenario offers more flexibility, as medium-scale fish farmers can still use the best sites by purchasing usage rights from small-scale fish farmers. Both scenarios are still within the regulations regarding the use of SFZs for mariculture activities.

148

CHAPTER 7. Predictive Carrying Capacity Modelling to Establish Small- Scale Finfish Mariculture Zone in Indonesia’s MPAs

7.1 Introduction

Briefly described in Chapter 1, and revisited in Chapter 6, carrying capacity (CC) has to be determined to complement the aquaculture site selection process. CC is even more important for establishing small-scale mariculture in MPAs to ensure that the activity can sustain itself and has minimal or no unacceptable impacts on the ecosystem. The contemporary concepts and application of CC develop from solely based on maximum allowable fish production to a more integrated approach with the addition of ecological and social aspects into the framework (Duarte et al., 2003, Ross et al., 2013b, Ross et al., 2013a). Various researchers suggested that CC should combine all four models (physical, production, ecological and social CC) (Inglis et al., 2000, McKindsey et al., 2006, Byron and Costa-Pierce, 2013). However, Ross et al. (2013a) argue that each CC model could be used as a free-standing decision tool to validate the feasibility of suitable sites for mariculture. This depends on the: type and level of the mariculture system, location, environmental condition, social-economic conditions, and overall objective of the mariculture industry (Ross et al., 2013a), which were discussed in Chapter 3, Section 3.4.4.2. Despite their rapid development (Ross et al., 2013a), most CC models are primarily established, and have evolved, around the prediction and measurement of CC based on oxygen and nutrient dynamics within the farming system and its surrounding environment. This biogeochemical-based CC modelling has also been developed in tandem with hydrodynamics-based models to provide a better spatial and temporal estimation of CC (Duarte et al., 2003), and this increasingly becoming the standard procedure for estimating CC (Geček and Legović, 2010). However, this fully coupled model is currently underutilised, specifically for finfish mariculture, due to its complexity, limited spatial coverage and data-intensive requirements, particularly for the hydrodynamics model (Geček and Legović, 2010). Thus, the use of only biogeochemical models remain common due to the relatively simple application, less data-intensive need, and sufficient ability to predict fish farming CC. These models can estimate CC of fish farming through oxygen balance and nutrient dynamics using mathematical simulations (Stigebrandt et al., 2004).

149

This mathematical approach relies less on the availability of data and is suitable for coastal zone planning that includes mariculture as one of the managed economic activities (Stigebrandt et al., 2004). Mathematical models of CC have been developed and used to compute various aspects of fish farming, such as oxygen flux between farmed fish and its environment (Stigebrandt, 1999), the volume and dispersion of effluent (wasted feed and faecal matter) (Hevia et al., 1996, Duarte et al., 2003, Corner et al., 2006), and the effects of fish farming effluent on benthic ecosystems (Stigebrandt et al., 2004). The objectives of these models vary across determining the best sites, managing potential farming waste, and estimating fish holding densities (Halide et al., 2009). With various applications combined with relatively simple, low cost and lower data requirements, the mathematical model-based approaches for CC have been used in temperate and tropical areas (Stigebrandt, 2011, Zhang et al., 2009). Nevertheless, mathematical models of CC for fish farming in MPAs have not been used or considered a part of MPA methodology for designing MPA zoning systems. For example, as discussed in Chapter 3, Indonesian MPA regulations have set arbitrary and universal CC limits for mariculture in all MPAs. As each MPA water body, and associated ecosystem, are different to each other, these universal CC limits might be too high for some MPAs and low for others. This might cause suppressed or overdevelopment of mariculture in some MPAs. Instead of using such general limits, Stigebrandt et al. (2004) and Ross et al. (2013a) suggested the use of specific environmental quality standards (EQS) to estimate CC of mariculture activities. This chapter develops a set of EQS and identifies the main factors influencing CC for mariculture in an Indonesian MPA using the Modelling–Ongrowing fish farm– Monitoring (MOM) system, developed by Stigebrandt (1999) and refined by Stigebrandt et al. (2004). As this method was originally developed for temperate areas and salmon farming, it is investigated and modified here to suit tropical areas and mariculture practices. The CC estimate from this study was then compared to the current holding capacity of mariculture practices in the MPA, and the general 50% CC limitation/threshold set by the MPA regulations (See Chapter 3, Section 3.4.2.3). C. It was hypothesised that the current CC regime for MPA mariculture has been set too high, and thus will potentially limit the development of sustainable small-scale mariculture in Indonesian MPAs.

150

7.2 Materials and methods

7.2.1 Study sites

The study was conducted in the Anambas Archipelago MPA. The description of the study site was outlined in detail in Chapter 2. For this CC analysis, site #1, #2 and #3 were selected (Figure 7.1) and were located in medium, best, and unsuitable site classes, previously determined in Chapter 6. Site #1 is located in the northern areas of Matak Island currently partially used for fixed grouper net cages. Site #2 is located off the coast of Nyamuk Island where net cage mariculture has not been developed yet. Site #3 is situated within the Niulwan Bay of Matak Island which has been traditionally used by small- and medium-scale fish farmers despite its shallow water and significant freshwater runoff and other inputs.

Mubur Island Matak Island 1 3

Siantan Island

Nyamuk Island 2

Locations of carrying capacity measurement

Figure 7.1. Location of carrying capacity measurement

In-situ measurements and ex-situ sample collections of CC parameters were carried out at the same time as the site selection field surveys in the wet (May 20016) and dry (September 2016) seasons.

151

7.2.2 MOM system description and modification

The MOM system comprises of two components; the first component is used to monitor and evaluate existing mariculture activities, and the second determines the CC for future mariculture operations (Ervik et al., 1997, Stigebrandt, 2011). Both MOM components deal with three basic controls of ensuring a sustainable mariculture operation, which are: controlling the accumulation of organic material under the farm, ensuring optimum water quality for cultured fish, and preventing deterioration of water quality around the fish farm (Stigebrandt et al., 2004). As previously discussed in Chapter 2 and 3, there is no existing CC estimate in coastal areas used for finfish farming in the Anambas Archipelago MPA. Also, most of the finfish farming in the study area use fixed net cages and virtually no floating net cages exist (one exception is the array of floating net cages used by middlemen to temporarily hold bought fish from fish farmers). For these reasons, the MOM system has been applied here to predict the CC or holding density in different site classes, determined previously in Chapter 6. Within the MOM system, there are four interlinked sub-models, each generates a set of data that feed into the final calculations of different CC or holding densities (Ervik et al., 1997, Stigebrandt, 1999, Stigebrandt et al., 2004, Stigebrandt, 2011). The fish farm sub- model deals with the energy transfer from feed to fish, as well as fish metabolism and growth. This sub-model also determines the amount of feed and faecal material released into the environment. The second sub-model, the dispersion sub-model, computes the dispersion of waste feed and faecal material based on the dimension of the farm, the form and settling velocities of the waste, and the dominant current speed. The benthic sub-model computes the interaction between the rate of organic matter sedimentation and oxygen concentration just above the seabed surface that can affect the diversity of benthic organisms. The benthic sub-model prescribes that at least two benthic organisms remain living under the farm (Ervik et al., 1997, Stigebrandt et al., 2004). In order to set the model to follow the standardised mariculture operation applied in Indonesia, the Indonesian EQS for finfish mariculture were used, except for the benthic EQS values which still adopted those of Stigebrandt et al. (2004). The last sub-model is water quality within the fish cages, which keeps the minimum oxygen concentration above, and maximum ammonium concentration below, critical levels,

152 while still supporting maximum fish production (Stigebrandt et al., 2004). In this sub- model, the availability of oxygen in the water inside net cages is strongly correlated with the dimensions and position of the net cage relative to the current speed and direction. On the other hand, ammonium concentration is influenced by the ammonium released within net cages and the time needed for a parcel of water to enter and exit the net cages (water residence time). The data generated from these four sub-models are used to compute the CC or holding density of the selected sites. Here, the study adopted Stigebrandt et al. (2004) definition of CC or holding density as the maximum total fish production in relation to the minimum oxygen availability, within the nets and at the seabed under the nets, and the maximum acceptable ammonia concentration during the minimum flushing period. For this purpose, three different CC or holding density equations were developed by Stigebrandt et al. (2004) as follows:

1. The maximum fish production based on the availability of oxygen in the farm or TPFO2 (kg/m3) is computed using the equation:

…………………. Equation 4

Where: UMIN = the minimum mean current 3 O2MIN = O2 concentration of waters flowing into the farm (kg/m ), 3 O2MIN = the critical concentration of O2 in the farm (kg/m ) Lf = the length of the farm (m) D = the depth of the net cage used in the farm (m) Pf =Permeability of the farm, the fraction of reduced current due to resistance of net cage DO2 = Respiratory oxygen demand of fish

2. The maximum fish production based on the limitation of NH4 concentration in the farm 3 not exceeding the critical value or TPFNH4 (kg/m ) is computed using the equation:

…………………… Equation 5

3 Where: NH4MAX = the critical concentration of MH4 in the farm (kg/m ) 3 NH4IN = MH4 concentration of waters flowing into the farm (kg/m ), Lf = the length of the farm (m) D = the depth of the net cage used in the farm (m)

Pf =permeability of the farm, the fraction of reduced current due to the resistance of net cage

153

3 DNH4 = the mean ammonium production per kg fish production (kg/m ) 3. The maximum fish production in the farm that does not cause detrimental effect to 3 bethic community or TPFbentam (kg/m ) is computed using the equation:

…………………. Equation 6

Where: β = the critical concentration of MH4 in the farm (kg/m3) 2 A = the total cage area of the farm (m ) Ubent = horizontal current speed at the bottom O2i = oxygen concentration just above turbulent benthic boundary layer O2min = lowest oxygen concentration allowing benthic infauna to survive α = the fraction of the particulate organic matter from the farm that is oxidised within the farm area FCR = the actual feed conversion ratio FCRt = the theoretical feed conversion ratio

4. The CC/holding capacity of the location (PROD, kg/m3) is determined by the minimum value of the three TPFs computed from Equation 7, 8 and 9 above:

PROD = min (TPFbentam, TPFO2, TPFNH4)…………. …….Equation 7

The selection of the minimum values as the acceptable CC/holding density of a site implies the use of worst-case scenarios in the MOM system. This means that any TPF value used as the selected CC/holding density does not violate the CC/holding densities of the other two TPFs. In relation to the possible value of TPFbentam, Stigebrandt (2011) has advised that if the standard deviation of the bottom current is greater than 3.5 cm/s, sediment will be resuspended and carried away from the farm. Thus, the CC will be essentially unlimited. It is to be noted that, each of the CC/holding density equations has components which are generated from the four sub-models. The explanations of these components are described in detail in Stigebrandt (1999), and Stigebrandt et al. (2004).

7.2.3 Data collection and measurement

As with the site selection described in Chapter 6, the in-situ measurements for CC were conducted in both dry and wet seasons to determine if different seasons influence the level of CC within Anambas MPA mariculture zones. The following sections describe the in-situ parameter measurements during the fieldwork.

154

7.2.3.1 Current measurement

Current characteristics are important to determine the flux of oxygen and nutrient dispersion within and around the farm and thus greatly influence the CC prediction. The measurement of current speed and direction was conducted using a Valeport Current Meter 105 (SN #21161). The first measurement was conducted from 17 to 18 May 2016 representing the wet season. The second measurement was conducted from 17 to 18 September 2016 representing the dry season. Each measurement lasted for 24 hours, where current speed and direction as well as temperature were recorded every 15 min, at a constant depth of 5 m. Due to limited time and resources, current measurements were conducted at a single point located at 3°15'8.73"N and 106°19'54.56"E. The current measurement was not conducted in each location or in existing farms. Thus, a 0.8 value was applied to the minimum mean current speed (UMIN) to adjust the speed for net cage resistance, as if the measurement of current characteristic was conducted within a net cage (Stigebrandt et al., 2004). The measurements of current characteristics near the bottom sediment were done at the three locations (Figure 7.1) to determine the mean speed of bottom currents (Ubent). Each measurement was conducted for 1 hour where current speed and direction were recorded every 15 min. The measurement depth was limited between 8 - 10 m because the accompanying measurements (oxygen and sample collection for ammonium) were limited to that depth. The short measurement period of current speed and direction is one of the limitations of this study. Stigebrandt (2011) suggested that a continuous current measurement of over 2 weeks should be done to record a valid minimum current profile to feed into the MOM system. However, due to research limitations, as described earlier, as well as the purpose of this study to describe the general CC of MPA mariculture zones, such current measurement was considered sufficient.

7.2.3.2 Oxygen (O2) and ammonium (NH4) measurements

It is essential to know the existing concentration levels of O2 and NH4 to determine the flux of oxygen and ammonium within and under the farm to feed into the MOM system. Oxygen concentration in the upper water column (± 1 m) and at the bottom (approximately 0.10 m above the substrate) was recorded with a measurement interval of 15 min, and the

155 values were averaged. The maximum depth of oxygen measurement was 10 m as limited by the length of the oxygen probe cable of the YSI Professional Plus used in the study. Sample collection, preparation, storage and analysis for ammonium concentration followed the method described in Chapter 6.

7.2.3.3 Fish farming management and practices

As required in the MOM system, information regarding fish species, feed composition, feeding practices, as well as the typical dimension and arrangement of the fish farms are also essential. Within the Anambas MPA, groupers and Napoleon wrasse are the most common finfish cultured in net cages of small-scale fish farmers. Due to their threatened status, wild seed capture, and tightly controlled selling quotas, Napoleon wrasse (Cheilinus undulates) culture has been in decline. Therefore, the CC or holding density discussed in this chapter was based on grouper finfish culture. Information on feed types, feeding practices, stocking densities, and cage dimension and arrangement of grouper finfish, was collected during the first fieldwork conducted over July to November 2014 (Chapter 4). The net cage dimension per unit farm used in the MOM system was the average dimension due to various farm sizes owned by small-scale fish farmers in the Anambas MPA. Furthermore, this study was focused on providing a theoretical CC for grouper finfish culture. Thus, the additional information needed by the MOM system, such as feed, fish composition and energy density, and feed conversion ratio were sourced from the available literature of grouper culture in Indonesia. This allowed more efficient resource use and research time, which were the limitations of the MPA authority in designing an MPA zone.

7.2.3.4 Environmental quality standards (EQS)

Different EQS are used in finfish net cage mariculture, which usually follow the national strategy of a country to develop mariculture (Stigebrandt, 2011, Ross et al., 2013a). Due to their importance, EQS have to be known to estimate CC of an area to develop mariculture or monitor the effects of an activity (Ross et al., 2013a). Indonesia has not specifically developed a set of mariculture EQS for MPAs. As a replacement, a literature review of the general Indonesian EQS for grouper mariculture was conducted. The review collected the EQS based on Indonesian National Standard (SNI) for grouper

156 culture, Indonesian Ministry of Environment and Forestry, and two prominent literature sources (i.e. Romimohtarto, 1985, Stigebrandt, 2011).

7.2.4 Data analysis

As described earlier, the four sub-models comprising the MOM system produced datasets used for computing the CC/holding densities of the three different site suitability class locations. To determine if there were variations in the CC or holding densities between different site suitability classes, determined previously, the mathematical computation of the four MOM sub-models were followed by calculations of TPFO2, TPFNH4 and TPFbentam for the three selected locations. The lowest value among the three TPF’s was selected as the CC/holding densities for the corresponding site classes (Stigebrandt et al.,

2004). However, the computation of TPFbentam depends on the characteristic of current speed fluctuation (Ustd). This was explored in detail in the result section. Data analysis was also performed to determine if there were differences in CC/holding densities in different seasons (dry and wet seasons). These different scenarios of CC/holding densities according to seasonality are important as this dissertation focuses on a small-island community whose livelihoods are significantly influenced by their remote location and environmental condition. The worst value of CC/holding densities based on different seasons could be also established as the worst-case scenario to control the development of mariculture within an MPA. The use of trash fish as fish feed has been considered to contribute to increased nitrogenous effluent from a fish farm, which degrades water quality due to its faster decomposition rate (Huang et al., 2011). Other researchers, such as Alongi et al. (2009) and De Silva (2012), argue that the effect of such effluent from small-scale fish farming is almost insignificant. Despite that, a theoretical use of trash fish and commercial feed is embedded within the data analysis of the MOM system for this research. This was to elicit a comparison between the CC/holding densities of mariculture areas within an MPA based on different feeds used.

157

7.3 Results

7.3.1 Current characteristics

As indicated by Equation 7, 8 and 9, knowing the current characteristics at cage depth (UMIN) and on the seabed (Ubent) are essential to determine the flux of oxygen and dispersion of effluent within and underneath a fish farm. Between the dry and wet season, current speed differs slightly. The current speed at cage depth (UMIN) during the wet season was slightly faster, with the mean speed of 0.96 m/s, with the dominant current direction from NW to SE (Table 7.1.; Figure 7.2. Current characteristics (speed and dominant direction) in the wet (A) and dry (B) seasons The current speed during the dry season was 0.94 m/s, with the dominant current direction from SE to NW (Table 7.1; Table 7.1). The values for Ubent was highest at location #2 (7 cm/s) due to its close proximity to open sea, followed by location #3 (6.2 cm/s), and the lowest at location #1 (4.5 cm/s). Location #1 had the lowest Ubent due to the presence of fringing reefs around the site serving as natural barriers. Location #1 is, in general, a shallow coastal area with water depth ranging from 1 to 10 m and has poor water exchange.

A B

Figure 7.2. Current characteristics (speed and dominant direction) in the wet (A) and dry (B) seasons

The fluctuating component of current speed (Ustd), determined from the standard deviation of the current speeds recorded, had values of 4.71 cm/s for wet season and 5.83

158 cm/s for the dry season (Table 7.1). Both Ustd values show that the study area has a higher probability of regular sediment resuspension.

Table 7.1. Current characteristics and concentrations of oxygen and ammonium

Location #1 Location #2 Location #3 Parameters Symbol Unit Sources Wet Dry Wet Dry Wet Dry

Average incoming DO O2in mg./l 4.56 4.15 5.30 6.40 4.47 4.33 Fieldwork, 2016 into the farm

Critical DO in the farm O2min mg./l 4.00 4.00 4.00 4.00 4.00 4.00 KemenLH EQS, SNI EQS Average incoming NH4in mg/l 0.03 0.01 0.02 0.01 0.02 0.01 Fieldwork, 2016 ammonia into the farm

Critical ammonia in the NH4max mg./l 0.10 0.10 0.10 0.10 0.10 0.10 KemenLH EQS, farm SNI EQS

Minimum mean surface Umin cm/s 9.57 9.38 9.57 9.38 9.57 9.38 Fieldwork, 2016 current

Fluctuating component Ustd cm/s 4.71 5.83 4.71 5.83 4.71 5.83 Fieldwork, 2016 of current (STD)

Horizontal bottom Ubent cm/s 4.50 4.50 7.00 7.00 6.20 6.20 Fieldwork, 2016 current velocity

7.3.2 The concentration of oxygen and ammonia in the farm

Concentrations of oxygen measured during the fieldwork differed according to season (wet and dry season) and locations. In the wet season, dissolved oxygen (DO) concentrations in the upper water column were relatively higher compared to the dry season (Table 7.1). One exception was location #1, which had lower DO concentration during the wet season compared to the dry season. However, DO concentrations at location #2 in both wet and dry seasons were the highest compared to the other locations. Overall, the recorded DO concentrations at all three locations were consistent with typical DO concentrations measured generally in the South China Sea, and particularly in the Anambas Archipelago, ranging between 4.4 – 6.8 mg/L (Purba et al., 2014). The DO concentrations at all locations were above the minimum EQS for DO (O2min >4 mg/L) for mariculture activities, set by the Indonesian Ministry of Environmental and Forestry (KemenLH, 2004) and Indonesian National Standard for mariculture (SNI). Ammonia concentrations were slightly higher during the wet season compared to the dry season in all three locations (Table 7.1). Location #1 had the highest ammonia concentration, corresponding to the measured low DO concentration at this site. Location

159

#2 and #3 had similar ammonia concentrations in both wet and dry season, despite both locations having different geographic characteristics (location #2 is near open sea and location #3 is within a bay). The overall ammonia concentrations at all locations were well under the maximum EQS for ammonia (NH4max<0.1 mg/L) for mariculture activities set by the Indonesian Ministry of Environmental and Forestry (KemenLH, 2004) and Indonesian National Standard for mariculture (SNI).

7.3.3 Fish sub-model in MOM

The fish sub-model in MOM computes the mean oxygen consumption needed by -1 one kg of fish (DO2, kg O2 day ), which is then used to calculate the TFPO2 for water quality in the cage sub-model. The mathematical model to compute DO is presented in Equation 17 of Stigebrandt (1999, p.18) . The parameters to determine DO are presented in Appendix 3. DO differed according to the types of feed used in the hypothetical fish farm.

Here, the mathematical calculation of DO used two theoretical feed types: trash fish and commercial feed. Both feeds contained different levels of protein, fat, and carbohydrate which affected the amount of oxygen required by fish for metabolism (respiration) to break down the ingredients (Appendix 3). Using the dataset published by Tacon et al. (1991) and Tucker Jr (2012) for grouper culture in Indonesia, a typical fraction of protein, fat (lipid), and carbohydrate in commercial feed were higher compared to that of trash fish (Appendix

3- row A:4,5,6). DO in fish farms using commercial feed were considerably higher (19.53 grO2 day-1kg fish-1) compared to farms using trash fish (19.53 grO2 day-1kg fish-1). The fish sub-model also estimates the mean ammonium production per one kg of -1 fish (DHN4, kg NH4 day ). DHN4 was determined using Equation 15 in Stigebrandt (1999, p.17) . The parameters to determine DNH4 are listed in Appendix 3. To determine DNH4, the amount of wasted/excess feed calculated from the difference between the factual feed conversion ratio (FCR) and theoretical feed conversion ratio (FCRt), the ratio between the amount of ingested feed and resulting growth, must be known. To reach a typical harvest weight of 0.7 kg with the average culture period of 210 days (Fieldwork, 2016), the fish model produced an FCRt of 1.51 (trash fish) and 0.89 (commercial feed). The calculation

FCRt value here was intended to feed the model in determining the unused food based on an theoretical assumption that all energy intake was used for weight gain. The greater

160 efficiency of commercial feed (FCRt) was supported by the high predicted assimilated protein factor adopted from Stigebrandt (1991). The value of FCRt calculated in this study was at the upper range of salmonids FCRt, predicted in Stigebrandt (1999) study (7 – 9.2). According to Tacon et al. (1991), the typical factual FCR in grouper culture in Indonesia was 3.53 using trash fish and 1.44 using commercial feed. Thus, the amount of wasted/excess feed was 1.42 for trash fish and 0.385 for commercial feed. This means that for every 1 kg of fish production, there was 1.42 kg of trash fish and 0.385 kg of commercial feed deposited underneath the fish farm and dispersed in the surrounding environment.

Using the value of FCR-FCRt and the fraction of nitrogen (N) in the protein fraction -1 -1 in the feed (Fp/6), the DHN4 was determined to be approximately 0.206 kg NH4 day kg -1 -1 fish fed with trash fish and 0.217 kg NH4 day kg fish fed with commercial feed. The fraction of protein in the feed used in grouper culture in Indonesia in this model was based on Tacon et al. (1991) and Tucker Jr (2012), which were 0.3% for trash fish and 0.43% for commercial feed. The DNH4 values were used to determine the carrying capacity/holding density based on TPFNH4 of the water quality in the cage sub-model.

7.3.4 Dispersion sub-model

Based on the type of cultured fish (grouper), type of feed (trash fish and commercial feed), and culture period (210 days), the spatial and temporal distribution of waste from the farm was estimated at 52.641 (feed) and 0.781 (faeces) g/m2/day using trash fish, and 14.324 (feed) and 0.781 (faeces) g/m2/day using commercial feed. The lower difference between factual FCR and FCRt of commercial feed contributed to the low values of excess feed and faeces produced from farms using commercial feed. The dispersion sub model estimated that as much as 1,415 kg of wasted trash fish was released from farms when producing 1,000 kg of fish, over a 210-day culture period, compared to fish farms using commercial feed releasing only 385.02 kg of wasted feed.

7.3.5 The benthic sub-model

Despite the high amount of wasted feed, particularly for trash fish, the fluctuating components of current speed (Ustd) were more than 3.5 cm/s (Table 7.1) in all study locations. As described previously in Section 7.3.1., such high value of Ustd means that 161 resuspension occurs regularly which will carry away organic materials (excess feed and faeces) from the sites. This result means that the TFPbentam does not need to be computed as these sites can have virtually unlimited CC based on the requirement that fish production does not lead to the extinction of benthic fauna.

7.3.6 Water quality in fish cage sub-model

Water quality in the fish cage sub-model computes the TPFO2 and TPFNH4 using the results of the fish sub-model calculations (DO and DNH4). The objective of this study was to predict the CC of newly created MPA mariculture zones using the commonly practised mariculture activity in the area. Thus, the size and number of net cage units used in the calculations were the same in all locations and based on the typical mariculture activity in the Anambas Archipelago MPA (Table 7.2). These similarities in sizes were needed to ensure that the results for CC were comparable between locations (Figure 7.1) and seasons.

The flow factor reduction (Pf) (of 0.8) was similar to that of the CC prediction by

Stigebrandt et al. (2004), Halide et al. (2009) and Anonymous (2007). The higher the Pf value, the more freely seawater flows in and out of the net cages. The current direction was set parallel to the longest side of the net cages to simulate commonly arranged fish cages in the Anambas Archipelago MPA.

Table 7.2. Farm size and number of units used in the CC/holding density based on MOM system in each location (Figure 7.1)

No. of Netpen Flow factor Sites No. rows Side length Spacing cages depth reduction (Pf) Location #1 8 2 4 2 3 0.8 Location #2 8 2 4 2 3 0.8

Location #3 8 2 4 2 3 0.8

The results of the CC prediction show that most of the TPFO2 values consistently had the minimum value based on season and feed types, compared to TPFNH4 and TPFbentam

(Table 7.3). One exception was the value of TPFNH4 in location #2 in the dry season (158.78 kg/m3), which was the lowest CC/holding density among the three TPFs for that season.

162

CC/holding density based on TPFbentam was considered unlimited due to the high probability of sediment resuspension. Based on feed types, the MOM system predicted commercial feed to have the lowest CC/holding density, compared to the use of trash fish, in terms of different season, location, and both TPFO2 and TPFNH4. Based on seasonal differences, the MOM system predicted CC, based on TPFO2, was the highest during the wet season compared to the dry season. This was expected as the concentration of oxygen was the highest during the wet season, as described in Chapter 3 and Chapter 6. For CC based on TPFNH4, similar predictions were also produced by the MOM system, where the lowest CC was in the fish farms using commercial feed. The high percentage of protein fraction contained in commercial feed, compared to trash fish, contributed to this low predicted CC (Appendix 3

- row A4). The high protein fraction also contributed to the low CC based on TPFO2 using commercial feed. This is because cultured fish need substantially more oxygen to metabolise the higher protein fraction in commercial feed. Based on the locations, the minimum PROD values were the highest in location #2, followed by location #3 and the lowest was location #1 (Table 7.3). The result of PROD corresponds positively to the site suitability location of location #2 which was classified as the best suitable site. However, location #3 had better minimum PROD values compared to location #1, despite being located within a bay and classified as unsuitable. The average

NH4 values in location #1 was the highest (0.034 mg/L) compared to location #2 and #3.

The high NH4 values recorded in location #1 was also translated into the lowest TPFNH4, both in terms of season and feed types. However, as the PROD values take only the minimum values, the CC value for PROD was based on TPFO2 instead.

Table 7.3. Results of CC/holding density based on TPFO2, TPFNH4 and TPFbentam Location #1 Location #2 Location #3 Carrying Unit Wet Dry Wet Dry Wet Dry capacity TF CF TF CF TF CF TF CF TF CF TF CF 3 TPFO2 kg/m 51.38 27.76 13.49 7.14 119.09 64.35 215.44 116.41 49.93 26.98 34.05 18.40

3 TPFNH4 kg/m 113.26 107.58 158.10 150.18 133.85 127.15 158.78 150.82 168.89 160.43 183.07 173.90

TPFbentam kg/m3 unlimited unlimited unlimited

Note: Bold values are the minimum values of CC and selected as the CC/holding density for the corresponding season location, and feed type (TF: trash fish; CF: commercial feed)

163

7.4 Discussion

7.4.1 Modification of MOM system to suit with tropical mariculture condition

Zhang et al. (2009) clearly indicated that the MOM system offers an efficient, easy and cost-effective method to determine the impact of aquaculture on itself and the environment. Nevertheless, the MOM system needs to be modified to suit the local conditions in Indonesia. To do so, minimal modifications were introduced in the model, including the use of lower protein and higher carbohydrate content, which corresponded to the respiration rate that typically occurs in tropical marine finfish (grouper). The CC estimate was also modified using seasonal holding density, where the original MOM system used an annual holding density (kg/year). Observation during fieldwork indicated there were significant differences in fish production during the wet and dry seasons. Through this modification, CC could be estimated for each season allowing fish farmers to decide when to start farming fish or when to expect fish production to be low. For example, stocking of grouper fry could be started during the best CC season, as grouper juveniles require more optimum environmental conditions. These CC estimates, based on season, is suitable for tropical finfish mariculture as culture cycles are usually shorter than one year, ranging between 7 - 10 months (Fieldwork, (2016), Afero et al. (2010). Other parameters within the MOM system were maintained, as described by Stigebrandt (1999) and Stigebrandt et al. (2004). These parameters, including standard feed energy content and fish metabolic energy requirement, were considered compatible for tropical marine finfish used in this study.

7.4.2 Factors influencing CC in MPA

The MOM system in this study strongly suggested that different feed types, seasons, and locations had significant influences on the CC. Feed types contributed to how high nitrogen loading was on the environment through two pathways: uneaten feed and faecal excretion. The MOM system in this study predicted that grouper fed with trash fish and commercial feed, cultured within the hypothetical MPA small-scale mariculture environment, could disperse 249.04 kg N and 67.76 kg N per 1-tonne fish production to the MPA environment, respectively. These values assume that the nitrogen content in the protein fraction of feed is 17.6% (Verdegem, 2013). The estimated values of N loading

164 produced in this MOM system were consistent with the result of Islam’s (2005) study of grouper culture, where the use of trash fish released up to 320.6 kg of N, and commercial feed from 47.3 to 130 kg of N, per 1-tonne kg fish production. These numbers suggest that the use of commercial feed for grouper culture within any MPA mariculture offers the best potential reduction of N released to the surrounding environment. The lower N loading on the MPA environment from mariculture activities, using commercial feed, will cause less nutrient enrichment stress in coral and other benthic fauna, as shown by De Silva (2012) and Loya et al. (2004). However, as discussed in Chapter 3, commercial feeds are difficult to obtain in many remote Indonesian MPAs, and there is a need to increase fish farmer knowledge of how to safely store and efficiently use this feed. Thus, a combination of trash fish and commercial feed is advised to reduce the N loading on the MPA environment, and at the same time alleviate some of the technical challenges in the use of commercial feed. Industry development would benefit from education of farmers on the technical aspects of feeding fish. Seasonal differences on the predicted CC in this study were mostly influenced by the availability of DO within MPA waters. Higher CC during the wet season was apparently caused by a cold-water mass, with high oxygen content, from the north (Chapter 2, section 2.4.), while during the dry season, a low oxygen water mass from the Java sea flows into Anambas waters. The low oxygen concentration measured during this study (wet and dry seasons) is similar to the findings of DKP-Anambas and LPPM-IPB (2011b) for the Anambas area, and Roseli and Akhir (2014) for the larger sSCS waters, ranging between 4.34 and 6.5 mg/L.

The result of MOM calculations for respiration rate (DO2), in relation to feeding contents (protein, lipid and carbohydrate), also influences the final CC result. This influence occurs at the same magnitude across all variables used in this research, e.g. season and location. The MOM system calculated the oxygen requirement of fish to metabolise commercial feed was almost twice as high as if fed with trash fish. This is due to higher protein content in commercial feed compared to trash fish. To be noted, the DO values predicted in this model agreed with the oxygen consumption range of grouper cultured in tropical areas (between 6.86 – 12.72 g/day) (Chua and Teng, 1980, Sullivan and Garine-Wichatitsky, 1994). The difference in protein contents of commercial feed and trash

165 fish was not that significant (only 13% difference, Appendix 3 - row A.4.). However, doubling the required oxygen for metabolism shows the potential drawbacks of using commercial feed. If the availability of oxygen in fish farms is insufficient for supporting fish metabolism, low protein retention from commercial feed will result. This will then lead to an increased N loading from fish excretion, as implied by Islam (2005), and Mungkung et al. (2013). To avoid this drawback, the use of lower protein content in the commercial feed, between 35-37% per unit feed weight (Marzuqi and Anjusary, 2013, Sim et al., 2005), should be encouraged in Indonesian MPAs. Another important factor influencing the CC estimate using the MOM system in this study is the limit of critical oxygen concentration within the farm, or the EQS value of 4 mg/L. This EQS value was considerably higher compared to that of Halide et al. (2009), who used 3 mg/L for their oxygen EQS value to estimate CC of a fish farm in Indonesia. However, using a higher oxygen EQS limit in this study was intentional, to prevent low oxygen conditions around shallow MPA mariculture zones adversely affect protected biota, such as coral and other sessile fauna. More importantly, the oxygen EQS value is based on the current standard oxygen concentration for mariculture activities in Indonesia, set by KemenLH (2004) and the Indonesian National Standard (SNI). Sea surface current speed also contributed to the differences, as the current continuously supplies fish farms with sufficient DO in the incoming seawater. The notable difference between sea surface current speed during wet and dry seasons found in this study is consistent with the results of Purba et al. (2014). They found that there was a decrease in current speed during the dry season due to the influence of water mass momentum from the Pacific Ocean. Despite the difference between wet and dry seasons, it was evident that the bottom current speed is high enough to cause resuspension of sediment, marked by the high value of the current fluctuating component (> 3.5 m/s) measured in this study. Such high variation is caused by complex seabed geography, the threshold of motion of particulates and sediments, the shallow water around the Anambas Archipelago, and the water mass exchange during monsoon periods (Purba et al., 2014). As a result, surface and bottom current velocities were not the limiting factors in determining CC or holding density in the study area.

166

Nevertheless, the use of EQS to maintain at least two benthic organisms remaining underneath floating net cages in the benthic sub-model might not be sensitive to important habitat or species and certain conditions of MPAs. This is due to the wider tolerance of benthic organisms to degraded water transparency and DO compared to, for example, coral reef and seagrass. Stachowitsch et al. (2007) revealed that at least 5 benthic organisms were still alive after 2 days of a complete anoxic condition with high H2S concentration and 2 remained active after 4 days. In contrast, Villanueva et al. (2006) and Loya et al. (2004) found an adverse effect of a prolonged exposure of high nutrient loading and sedimentation from mariculture on coral reef reproduction. Similarly, Rountos et al. (2012) found a decrease in seagrass density due to increased number of benthic organisms as an indirect effect of high mariculture effluent release. These findings suggested that the benthic sub- model might have to use coral reef tolerance to mariculture nutrient loading as the EQS particularly for the EQS requirement of medium-scale mariculture in MPA. Moreover, future research is needed to determine the relative resistance of coral reef to effluent from both small- and medium-scale mariculture in MPA areas. Such research is important yet, it is beyond the focus and scale of this current study.

Like current speed, the concentration of ammonium produced from the hypothesised fish farm did not limit the CC in the study area, in comparison to the oxygen concentration in the farm. One exception was the CC value in location #2 during the dry season with fish fed with trash fish. The optimal conditions in location #2 (both oxygen and current speed) meant that CC could be increased up to the point the ammonium build-up will adversely affect the cultured finfish. Other than that, CC was limited by the oxygen concentration in the farm, despite this study using the lower limit of ammonium EQS value (Table 7.1). The ammonium EQS used by Halide et al. (2009) in their MOM computation was much higher (0.2 mg/L). This essentially means that fish production is the main objective in their study and thus, it is argued to be incompatible with the objectives of this study – to maintain a balance between mariculture and conservation efforts in MPA mariculture zones.

7.4.3 Comparison of the predicted and existing CC/holding density

This predictive CC used a hypothetical floating net cage farm (4 pens with total farm area of 128 m2). This net cage farm is similar to the study by Afero et al. (2010),

167 which found that the average size of small-scale mariculture net cage farms in Indonesia was 135 m2, consisting of 4 net pens. Existing net cages owned by Anambas fish farmers were on average 4 pens/farm, with an average size of 50 m2 (Fieldwork, 2017). However, these existing cages were mostly fixed net cages built along the narrow strip of the Anambas coastline by clearing coral reef coverage. Such practice has been deemed destructive to the MPA ecosystem and site availability is limited (DKP-Anambas and LPPM-IPB, 2011a). Thus, the use of common floating net cages in Indonesia was more compatible for this study. The values for CC using this net cage size and layout indicated a positive correlation with the site suitability classification, at least for location #1 and #2 (classified medium and best suitability). Location #3 still yielded measurable CC despite being classified as unsuitable based on the parameters not related to the MOM parameters. The range of CC values for location #1 and #3 (7.41 – 51.38 kg/m3) were similar to the CC range computed by Halide et al. (2009) in three different locations in Indonesia (26.67 – 53.2 kg/m3). One exception was location #2, where CC values were up to three times higher in all scenarios (Table 7.3). This is related to the geographical advantage of location #2, which is situated near the open sea where the conditions for water exchange (current speed) and oxygen concentration are the best. If compared to the existing CC/holding density practised by fish farmers in Anambas MPA, the result for the predicted CC in this study were much higher. Based on the two field trips in 2016 and 2017, the average holding density practised by the small- scale fish farmers was 2.19 kg/m3. Afero et al. (2010), who compared the economic viability of small, medium- and large-scale fish farmers in Indonesia, found a slightly higher holding density of 3.9 kg/m3 for small-scale fish farming. Afero et al. (2010) indicate that the use of low holding density is a risk management strategy to minimise production costs and increase survival rates. However, it is argued here that this strategy does not increase fish production and an improvement in quality and frequent grading of grouper fry is suggested. The huge gap between predicted CC and the existing CC/holding densities indicate that the current mariculture activities in Anambas MPA, and most likely in other MPAs in Indonesia, are far below the CC of each area. For example, the worst performing site,

168 location #3, still has at least 4 times greater CC, even after being limited by the 50% CC of MPA regulations, compared to the average Anambas mariculture CC/holding density. As such, the imposed 50% CC for mariculture, universally across all MPAs (discussed in Chapter 3), and disregarding their environmental characteristics and potential, might be too rigid for the development of small-scale mariculture in Indonesian MPAs. In some MPAs, with poor environmental characteristics (even though highly unlikely), the 50% CC might be too low, so mariculture activities might adversely affect the protected MPA biodiversity. These findings highlight that each location, or mariculture zone, within an MPA has a different CC, which is influenced by the fish farming practices and environmental characteristics. Thus, the control of mariculture development in an MPA should be on a case-by-case, or site-by-site, basis. This management technique will provide a comprehensive strategy to control the use of MPA resources for mariculture to a point that it is economically sustainable for the community, and environmentally sustainable for the protected MPA biodiversity.

7.4.4 Conclusion

This study confirms that the MOM system could be used as a predictive model to determine the CC/holding density of mariculture zones within an MPA. The MOM system provides a systematic analysis of CC yet is efficient and effective in terms of using limited research resources. The Indonesian MPA authority could use the MOM system to verify the usage level of MPA mariculture zones for mariculture activities. This was not possible before due to the limited resources possessed by the MPA authority, and the complexity of other CC models. However, it is noted that some parameters in this CC prediction, using the MOM system, utilised hypothetical data and shorter temporal datasets due to research limitations. For example, using a single surface current measurement within a short period of time to represent three locations might not depict the actual surface current at each location, particularly if the geographic characteristic between the reference current is different to that of studied locations. Thus, the CC results should only be used to set a general benchmark regarding the future usage level of MPA mariculture zones for MPA users (fish farmers) and MPA authority consideration. Regular monitoring to evaluate the

169

CC post establishment of mariculture operation using the MOM system, or similar CC model, is advised (Stigebrandt et al., 2004, Stigebrandt, 2011). In terms of the study results, DO plays the major role in determining the CC at different site classes in waters of the Anambas Archipelago MPA. The increase of ammonium concentration and nutrient loading on the seabed can be tolerated by the fish and the benthic community, to the point the increases surpass the CC based on DO availability. This signifies the role of current speed, the second dominant factor, in dispersing nutrient loading and bringing a constant supply of water into net cages. Still, it is of the utmost importance to control feeding practices of mariculture in an MPA through the balanced and wise use of trash fish and medium protein-content commercial feed. High protein-content commercial feeds could cause excessive N loading under the cages, especially in location where water exchange is limited. As a general conclusion, I argue that the current CC limitation for mariculture in Indonesian MPAs is somewhat too excessive or unfair to local fish farmer communities. Each MPA has different environmental conditions, and thus the limitation of mariculture development in MPAs should be treated differently. Here, it is suggested that controlling mariculture development in Indonesian MPAs through CC limitation should be based on: mariculture zones being established within MPA SFZs through site selection followed up by CC estimation for each site suitability class or zone; the CC estimates should take into account the different feeds used, cultured fish species, and seasonality; and CC is a trade- off between environmental capacity of the mariculture location and economic capacity of the mariculture operation. Therefore, it is equally reasonable to say that CC limitation for each MPA mariculture zone is a trade-off between the small-scale fish farmers and the managing MPA authorities.

170

CHAPTER 8. Thesis Synthesis and Conclusion

This thesis aimed to develop an MPA mariculture zone site selection and carrying capacity framework to sustainably support small-scale finfish farming in Indonesia’s small- island MPAs. Undertaking such complex research involving various social and environmental factors was challenging yet necessary to open a discourse on, and fill the gap of knowledge in, this subject. The focus of this thesis on small-scale mariculture in Indonesian small island MPAs was based on the premise that: 1) the recent Indonesian conservation policy indicates the expansion of MPAs in inhabited small-island seascapes of high biodiversity; 2) programs of sustainable livelihood in MPAs are centred on ecotourism and small-scale fisheries, while mariculture is often overlooked; 3) the MPA strategy and legal framework are not accompanied by clear-cut regulations and science-based technical guidance for a sustainable small-scale mariculture operations, i.e. site selection and carrying capacity; 4) the expansion of industrialised mariculture (medium- to large-scale) by non-local fish farmers poses a significant threat to small-scale fish farmers who traditionally have the usage rights of the small-island MPA seascapes; and, 5) mariculture limitations in MPAs had presumably been based on an arbitrary CC, and thus disregards the capability of each area to sustain these activities. This study demonstrated that small-scale finfish mariculture can be carried out sustainably by local fish farmers within Indonesian MPAs by satisfying social, environmental and technical aspects of the activity. Chapter 3 highlighted the need for regulation change regarding, resource allocation for mariculture, the formulation of mariculture zones, increased transparency of mariculture leasing licenses based on community approval, and redistributing authority to control mariculture development back to the local government. These regulation improvements are needed to ensure the sustainability of mariculture activities in MPAs. Additionally, the improvements were also required to secure formal local community usage rights for MPA seascapes to support livelihoods. Chapter 4 addressed the ongoing debate regarding the sustainability of small- scale mariculture in MPA compared to much-preferred livelihood activities, namely ecotourism and small-scale fisheries. This chapter concluded that small-scale mariculture has the potential to be as sustainable as ecotourism and small-scale fisheries. However, small-scale mariculture needed equal support for capital assets, as received by the other 171 livelihood groups, and fair treatment from the government. This is essential to prevent small-scale fish farmers becoming trapped in a vicious cycle of patronage relationships commonly found in most small-island communities in Indonesia (Glaser et al., 2015, Adhuri et al., 2015). In general, each of the livelihood groups discussed in this chapter has a different sustainability level defined by its ownership or access profile to capital assets. Thus, any devised strategy to support these livelihoods in Indonesian MPAs should aim to strengthen or solve problems in the access to, or ownership of, capital assets. Chapter 5 confirmed the vulnerability of small-island MPA communities is strongly influenced by its geographical disadvantages, such as seasonal weather, conflict over limited resources, cost of production inputs and change in the status of the area (i.e. designated as MPA). This study concluded that the three household groups being compared here have a moderate vulnerability, where each has specific vulnerability issues and needs specific interventions from responsible institutions. Chapter 5 also concluded that the household groups support the MPA establishment, despite knowing that some aspects of their livelihood will be affected. Fish farmers have the highest agreement regarding the MPA establishment, which indicates the potential of this household group to be involved in supporting conservation efforts. Chapter 3, 4, and 5 have established that small-scale mariculture can be as sustainable as the other livelihoods and be developed in MPAs through the improvement of regulations and technical guidance. This is significant for remote island communities where access to other resources and livelihood opportunities are limited. Chapter 4 and 5 provided a conclusion that the livelihood sustainability and vulnerability of small-scale mariculture are equally comparable to the other livelihoods in terms of economic, social and environmental benefits and challenges. With this conclusion and the fact that small-scale mariculture has been carried out in most multi-use MPAs in Indonesia, a specific mariculture zone should accommodate this activity and control its development. Chapter 6 and 7 provided the scientific information on how mariculture zones can be established in MPAs, and the requirements to make it sustainable. Chapter 6 concluded that a geometric mean is a better overlay for mariculture site selection in MPAs as it maintains full protection of sensitive MPA ecosystems and zones. The stakeholder model, developed in this site selection method, successfully incorporates local community preferences, which is

172 absent in other site selection methods. This site selection framework could give priority of best sites to small-scale mariculture household and still accommodate the needs of medium- scale mariculture operators for suitable fish farming sites. Chapter 7 found that the estimated values for CC of site suitability classes were all above the currently practiced stocking density of small-scale mariculture operations. In addition, DO concentration limited the CC of the area. However, this condition is site specific and could not be used to define CC for other MPAs in Indonesia. Thus, Chapter 7 concluded that the current 50% limitation of CC was set too high for mariculture development in MPAs and should be reconsidered. This finding confirmed the initial study hypothesis that a universal CC limitation for MPAs has been determined arbitrarily and has not considered the specific capability of each MPA. Based on all the study findings, it is concluded that mariculture can be sustainably developed in an MPA. To sustain the activity, specific mariculture zones should be created as one of the zoning regimes in MPAs, different to the SFZs, and entitled to local communities as the main users. This specific mariculture zone could be determined based on site selection designed specifically for MPAs, as this study suggested. The calculation of CC for each mariculture zone, before any mariculture operations take place, should be conducted and followed by regular monitoring. All of this should be included in a specific regulation which guarantees the rights of local MPA communities over their seascape. The information from this thesis can inform policymakers, encoraging them to revisit or change the current prevailing regulations regarding zoning and operation of mariculture in MPAs. MPA authorities will also benefit from the proposed technical design of MPA mariculture zones and could effectively manage the level of the operation. The multiplier effects of the above will significantly strengthen local community rights over the fair use of an MPA seascape for small-scale mariculture.

8.1 Major findings and their implications

8.1.1 The trade-off of legal framework, resource use, and community rights in MPA management This research highlighted the existing complex and redundant, yet arguably unclear, regulations that underpin mariculture management in Indonesian MPAs. This echoes the similar complexity of the overall MPA regulations found by Dirhamsyah (2006) and Wever 173 et al. (2012). The findings in Chapter 3 suggested that the same principles of EAA and MPA (equity, power-sharing, participation and shared benefits) are difficult to be realised due to this complexity. Syarif (2009) has highlighted this problem as “the great disparity between written law and practised law” (p.31), resulting in what Rife et al. (2013) argue as “more paper parks created…offering a false sense of protection” (p.200) of coastal and marine resources in Indonesia. I argue that the designation of mariculture zones, licensing based on local community approval, increasing the roles of local DKP and standardised MPA mariculture zone site selection might offer a miniaturised resolution of the problems. These key recommendations, from Chapter 3, could be used to tackle the problems in achieving the EAA and MPA principles through several pathways. Firstly, the proposed specific mariculture zones within MPAs in Chapter 3, which differs from SFZs, allows local fish farmers to exercise their rights to use the seascape without fear of prosecution. This is similar to the idea proposed by Adhuri et al. (2015) regarding the development of a specific regulation that small-scale fishermen are given exclusive formal rights to sustainably catch the spillover fish stock from MPA no-take zones. This embodies the message of equity where small-scale fish farmers, despite currently being the smallest population group in most MPAs, are acknowledged for their rights. In addition, mariculture zones also discourage the long-held open-access regime which Allison and Ellis (2001) and Adhuri et al. (2015) argue still dominates coastal and marine resource use. An open-access regime essentially means that coastal and marine resources are freely available to anyone. However, to hold stakeholders accountable for negative impacts of resource use is difficult or even impossible. Exclusive mariculture zone access for local MPA communities is the opposite of the open-access regime and could be used to minimise the expansion of medium- and large-scale mariculture activities by entrepreneurs from outside an MPA. The proposed community-based mariculture leasing approval in Chapter 3 fosters the bargaining position of small-scale fish farmers. Having their rights of access to the mariculture zone at stake also increases their participation in the decision-making process. Rimmer et al. (2013) and McLeod et al. (2009) both implied the need of a balanced power-sharing by, and active participation from, local fish farmers to equalise the dominance of non-local entrepreneurs backed by large portfolios and strong ties. The example of the concession-based area or resource use by local communities has

174 been observed by Rife et al. (2013) in Mexico, and had a positive result toward achieving MPA objectives. Secondly, the complex local situation in most Indonesian MPAs requires a balance between bottom-up and top-down management approaches, as suggested by Rife et al. (2013), to resolve conflicts over fair resource use. Failure to balance these two management approaches have been shown to cause confusion and distrust among MPA stakeholders, as described by Bennett and Dearden (2014). They found that the dominant roles of central government, in some Thailand MPAs, have made the local leaders felt they were being used. This ended up in a complete deterioration of communication between the central government and the community and no further MPA development could be discussed. To a lesser extent, this research found a similar situation regarding the government endowment programs to small-scale fish farmers described in Chapter 4 and 5. Fish farmers felt that the endowment programs were not what they needed and low in quality and quantity, rendering them unusable. Such treatment, added with the strict MPA law enforcement, led to exclusive patronage relationships between small-scale fish farmers and middlemen. The re- distribution of mariculture control of MPAs, from the central government (MMAF) to the local government (local DKP), might represent the mixed approach above. As described in Chapter 3, the central government could have focused more on improving compliance. At the same time, the local government, with its extensive knowledge of the social and cultural background of the local communities, will likely be more effective and efficient at controlling small-scale mariculture development. In addition, the more direct involvement of small-scale fish farmers in the decision making to design mariculture zones (Chapter 6) could also promote a balanced top-down and bottom-up approach. As discussed in Chapter 6, adding simple fish farmer preferences into the GIS site selection model, changed the suitable locations significantly. However, it also must be noted that biodiversity protection, represented mainly by the top-down approach, is still maintained in the model using MPA constraints, and other MPA related factors. The principle of community participation within the site selection model may also be used to design other zones or to the full extent of planning an MPA. Thirdly, site selection and CC of the MPA mariculture zone, discussed extensively in Chapter 3, 6, and 7, should be formalised as a standard procedure in multi-use MPA

175 zoning. The predominant use of freely available GIS and remote sensing databases in this decision-making framework could significantly overcome the limited resources of MPA authorities for zoning design. According to Claussen and Green (2007), the Indonesian government has to spend 2-3 times more (5.5 USD funding gap) than it could afford to establish a new multi-use MPA. Using this estimate, just to establish the planned 20 million hectares by 2020, the Indonesian government will require at least 85 million USD. This efficient yet effective framework could reduce the projected expense yet maintain an ideal participatory zoning process. Furthermore, this framework also allows a smoother flow of information and objective-based decision making, resourced from the local stakeholder, which is difficult to collect through formal and large-scale events due to the dominance of some fish farmer groups or individuals (Essen et al., 2013). More importantly, these site selection and CC models provide a clear and easy step-by-step process of which an analyst could follow. However, it is important to note that the human resources of the MPA authority and local government have to be upgraded through training, education, and improved communication and a conflict resolution (Bennett and Dearden, 2014). This is needed to deal with the specific mariculture tasks related to MPAs, such as mariculture zoning, monitoring, and evaluation of CC and leasing permits.

8.1.2 The application multidimensional indicators in sustainability and site selection for mariculture in MPAs This study illustrated that multidimensional indicators, both quantitative and qualitative, can be used to measure the sustainability and vulnerability level of different community groups, and determine the site suitability for mariculture. In mapping the sustainability of household groups, both quantitative and qualitative indicators are used to understand the components of building sustainability and their interaction among these groups. The numerical proxy values of qualitative indicators have complemented the SLA frmaework in this data scarce region. A similar study by Erenstein et al. (2010) in India faced similar data scarcity issues in which proxy qualitative and aggregated numerical data were successfully combined to determine the underlying causes of poverty and the influence of vulnerability. Here, the numerical transformation of multidimensional data is based on the argument that stakeholders are more comfortable with the numerical data format as suggested by Ko (2005) and implied by Erenstein (2011).

176

Similarly, the use of standardised multidimensional data in this study was to address: 1) data scarcity persisting in many developing countries as argued by Silva et al. (2011) and confirmed by this study; 2) the difficulty to extract numerical data from local communities, which can be used within the GIS environment; 3) the complexity of data having different measurement units (McKindsey et al., 2006); and, 4) the knowledge dominance of analysts and experts in the site selection decision making processes. Local community participation in the site selection process has been limited to reviewing the outputs of the site selection process (Kapetsky and Aguilar-Manjarrez, 2013). However, their minimal knowledge of the subject, and the dominance of other parties, prevents meaningful changes to the site selection output. Thus, it is argued here that other site selection models rely too heavily on expert knowledge and analyst capabilities. This study developed a stakeholder sub-model that successfully interprets local community preferences into workable GIS data. At the same time, the sub-model also influenced the site selection outputs from within the GIS framework. As a result, site selection for designing mariculture zones applied in this study could have greater local community acceptance, as the result reflects their own preferences as the main users of the seascape. This is in line with the site selection principle described by Ross et al. (2013b), that increased stakeholder participation and better policy regulation will increase local stakeholder acceptance of fish farming zones.

8.1.3 Future development of mariculture in Indonesia’s MPAs

This study has provided first evidence that mariculture could be as sustainable as other livelihoods in small-island MPAs in Indonesia. It is important to note that the sustainability discussed here only applies to small-scale mariculture activities carried out by the local community. The rights of the local community to develop mariculture in MPAs can be regarded as a trade-off for limiting the access to some coastal resources imposed by MPA regimes. However, it is argued here that existing small-scale fish farmers in an MPA are more likely to have better success rates compared to newly establish ones. These existing small-scale fish farmers usually have the required experience, technical knowledge, determination and pre-established social and market networks. This group consisted of individuals who rely on mariculture as their only source of income and those who switch back and forth between fishing and fish farming (Martin et al., 2013), which is 177 also described in Chapter 4 and 5. Thus, they need support from the government and MPA authorities in the form of improved regulation, endowment support, and protection from larger fish farmer competitors. This argument is synonymous with the view of Pomeroy et al. (2006) who implied that the introduction of mariculture into a fishing community has a higher chance of failure. Newly established mariculture activities targeting small-scale fishers requires significantly more resources from the government and MPA authorities to reduce risks such as capital investment, fish culture and financial trainings, and most importantly establishing social and market alternatives (Pomeroy et al., 2006). If mariculture zones suggested in this study were applied, the usage rights could switch to larger fish farming entrepreneurs when inexperienced fish farmers failed or simply did not have the consistent determination to develop their activities. These small-scale fish farmers might also end up as hired labour of these entrepreneurs in this failed scenario (Pomeroy et al., 2006). The mariculture zone and its reserved rights belonging to the local community will likely lead to increased production of finfish in small-island MPAs in Indonesia. However, newly established regulations (MAAF regulations No. 15/PERMEN-KP/2016 and No. 32/PERMEN-KP/2016) limit the access of Hong Kong-based ships to these remote areas, for the collection of live reef fish, a valuable market. The remoteness of most small-island MPAs means that transportation and market guarantee of live grouper are limited. The limitation of export shipping schedules, as well as the obligation of fish farmers to bring fish products to the export transit harbours, imposed by these regulations, could essentially render small-scale mariculture economically inefficient. The market chain could get longer which causes the already diminutive margin of small-scale fish farmers to get even smaller, as argued by Pomeroy et al. (2006) and Dirhamsyah (2012). Shortening the market chain in this scenario is impossible, as the role of middlemen to consolidate enough fish product from various remotely and widely distributed locations is essential (Muldoon and Johnston, 2006). MMAF efforts to shorten the market chain by subsidising transportation cost through procurement of live fish transport vessels has been suspended due to high financial liability. The program also receives strong resistance from local middlemen due to the inefficiency in consolidating fish products (Wayan Sudja, 2018 pers. comm., April 17). Instead, strengthening the cooperation between fish farmers and improved flow of market

178 information could be more sensible and increase the bargaining power of small-scale fish farmers in price negotiations (Muldoon and Johnston, 2006). The formal usage rights of MPA mariculture zones by fish farmer groups suggested in this thesis might be of use to strengthen this horizontal integration between local fish farmers. This improved cooperation among local small-scale fish farmers could also lead to better bargaining power with medium-scale entrepreneurs in the use of mariculture zones, as discussed in Chapter 3 and 6. Another concerning aspect of more formal mariculture zones established in these remote small-island MPAs is the possibility of increasing the volume of smuggled illegal, unreported, unregulated (IUU) caught live reef fish along cultured live reef fish trade. Dirhamsyah (2012) argued that IUU fishing occurs in most of the live reef fish market chain due to transhipment of live fish at sea and the absence of onboard observers to record fish data. This might be true for live fish trade based on capture but could be too generalised for live reef fish trade from mariculture. Chapter 4 and 5 described foreign export ships (Hong Kong) having a clear and regular presence when collecting fish at local middlemen net cages. Transactions and fish transfers happened at the easily accessible sites and could take up to one week. Local middlemen also must store the fish in their net cages before the ship arrived from Hong Kong. It should be easy to perform inspections or record fish production data. Transhipment at sea increases the risks of transportation, such as increased fish stress, loss, injury and mortality (Muldoon and Johnston, 2006), which are the last thing an exporter wants. It is argued here that a classic problem of regulation implementation, weak monitoring and control over the market chain, and the insufficient ratio of field observers may contribute to IUU fish farming problems.

8.2 Research limitations

One of the study limitations lies in the fact that absolute livelihood sustainability and vulnerability of the household groups is probably not measured. Yet, livelihood sustainability and vulnerability have been extensively used throughout the document. However, one of the intentions of this thesis was to determine the sustainability profile of different groups of the household. Due to differences in possession of, or access to, capital assets, I argue that it is empirically possible to compare the relative sustainability state of a

179 system, for example, a household group, with other systems (other livelihood). Such an approach is viable only if all SIs used in the analysis are the same, which was the case with this current research. Another limitation was the use of value judgements, in the form of Likert type data, in expressing sustainability and vulnerability of the household groups. Böhringer and Jochem (2007) argued that the process of producing the index and an index composite to measure sustainability introduces some biases due to inconsistent normalisation, weighting and aggregation. The decision to use value judgements in this study has been influenced by the fact that some SIs could only be measured on a categorical scale due to the inability of respondents to express them in numerical values. For example, most of the respondents were not able to provide their own income during the interview due to the large variation of annual income. As a result, ranges of values were used to better aid responses. This study also recognises the limitation in the use of a small number of respondents (Chapter 4 and 5), proxy values of site suitability replacing expensive in-field measurements (Chapter 5), and shorter measurement periods of current speeds (Chapter 6) due to the limited research resources. However, it is argued the data collected (social and environmental parameters) is the most comprehensive available for this data-scarce area. Thus, future research with larger numbers of respondents and distribution across different household groups, denser sampling points and a longer period of measurement of environmental parameters are advised to improve the results of this study.

8.3 Conclusion

This study tackled a thorough discussion and successfully filled the knowledge gaps regarding the compatibility of small-scale finfish mariculture to be developed in MPAs from various points of view. It is imperative that the Government of Indonesia, represented by the MMAF, must make substantial and detailed changes to the regulatory framework governing activities in multi-use MPAs in Indonesia. These changes include concession of specific mariculture zones to local MPA fish farmer communities, community-based approval of mariculture permits for medium- and large-scale mariculture entrepreneurs and their obligation to seek environmental impacts assessment for their activities.

180

From the perspective of livelihood sustainability and vulnerability, small-island MPA communities with resource extraction-based activities are influenced by conflicts over resource and spatial uses due to geographical remoteness and extreme weather patterns. Resource and spatial use conflicts are also exacerbated by minimal government presence. Finfish mariculture activities have been developed in most small-island MPAs and can provide a solution to these problems. At the same time, it offers a potential reduction of future stress on the surrounding MPA ecosystem when fishers, or other extractive livelihood groups, switch to, or complement their main job with, finfish mariculture. Despite its relative infancy compared to other long-established activities, mariculture, as a livelihood, has the potential to perform better in terms of sustainability and vulnerability in MPAs if equal support from the national and local government is given. As an economic activity in MPAs, the sustainability of small-scale finfish mariculture depends on appropriate site selection of mariculture zones as well CC determination and monitoring. In this study, the developed site selection method based on GIS and remote sensing for MPA mariculture zones provides an improved and efficient framework, specifically tailored for MPAs through the development of indicators based on parameters site suitability functions (PSSFs). The framework also addresses the lack of local stakeholder participation in the decision-making process through the inclusion of a stakeholder sub-model and visual amenity indicators. This robust site selection framework is efficient in dealing with data scarcity in remote small-island MPAs using freely available remote sensing and proxy data. This framework could substantially reduce resource requirements for designing not only mariculture zones, but also other zones given a set of indicators for the zones are identified, verified and agreed on by MPA stakeholders. In addition, the attractiveness of successful finfish mariculture could lead to a “boom and bust” mariculture development in MPAs, as has happened at many of non-protected coastal areas worldwide. The efficiency of resources and effectiveness of the MOM system to determine CC is essential and can be used in this scenario of MPA mariculture zones. The MOM system also has proven that arbitrary limitation of CC in Indonesian MPA should be revised case-by-case, depending on the quality of the MPA environment and the level of existing and/or planned mariculture activities. All the above-mentioned results are

181 interlinked and need to be integrated into one solid strategy. This is to ensure mariculture development in MPAs is safe for the environment and benefits the ever-marginalised small- scale fish farmers in Indonesia.

182

LIST OF REFERENCES

ABDULLA, A., OBURA, D., BERTZKY, B. & SHI, Y. 2013. Marine Natural Heritage and the World Heritage List: Interpretation of World Heritage criteria in marine systems, analysis of biogeographic representation of sites, and a roadmap for addressing gaps, IUCN. Gland, Switzerland. xii + 52pp. ABECASIS, R. C., SCHMIDT, L., LONGNECKER, N. & CLIFTON, J. 2013. Implications of community and stakeholder perceptions of the marine environment and its conservation for MPA management in a small Azorean island. Ocean & Coastal Management, 84, 208-219. ABERY, N. W., SUKADI, F., BUDHIMAN, A. A., KARTAMIHARDJA, E. S., KOESHENDRAJANA, S. & DE SILVA, S. S. 2005. Fisheries and cage culture of three reservoirs in west Java, Indonesia; a case study of ambitious development and resulting interactions. Fisheries Management & Ecology, 12, 315-330. ADAMS, W. M., AVELING, R., BROCKINGTON, D., DICKSON, B., ELLIOT, J., HUTTON, J., ROE, D., VIRA, B. & WORMER, W. 2004. Biodiversity conservation and the eradication of poverty. Science. ADHURI, D. S. 2005. Perang-perang atas laut, menghitung tantangan pada manajemen sumber daya laut di era otonomi: pelajaran dari Kepulauan Kei. Maluku Tenggara. ADHURI, D. S., RACHMAWATI, L., SOFYANTO, H. & HAMILTON-HART, N. 2015. Green market for small people: Markets and opportunities for upgrading in small- scale fisheries in Indonesia. Marine Policy. ADRIANTO, L. & MATSUDA, Y. 2002. Developing economic vulnerability indices of environmental disasters in small island regions. Environmental Impact Assessment Review, 22, 393-414. AFERO, F., MIAO, S. & PEREZ, A. 2010. Economic analysis of tiger grouper Epinephelus fuscoguttatus and humpback grouper Cromileptes altivelis commercial cage culture in Indonesia. Aquaculture International, 18, 725-739. AGARDY, M. 1993. Accommodating ecotourism in multiple- use planning of coastal and marine protected areas. Ocean Coastal Manage., 20, 219-239. AGARDY, M. 1994. Advances in marine conservation: the role of marine protected areas. Trends in Ecology & Evolution, 9, 267-270. AHMED, M. & LORICA, M. H. 2002. Improving developing country food security through aquaculture development—lessons from Asia. Food Policy, 27, 125–141. AHN, Y.-H., MOON, J.-E. & GALLEGOS, S. 2001. Development of suspended particulate matter algorithms for ocean color remote sensing. Korean Journal of Remote Sensing, 17, 285-295. ALDER, J. 1996. Have tropical marine protected areas worked? An initial analysis of their success. Coastal Management, 24, 97-114. ALDER, J., SLOAN, H. & UKTOLSEYA, H. 1994. A comparison of management planning and implementation in three Indonesian marine protected areas. Ocean and Coastal Management, 24, 179-198. ALEXANDER, B., CHAN‐HALBRENDT, C. & SALIM, W. 2006. Sustainable livelihood considerations for disaster risk management: Implications for implementation of the

183

Government of Indonesia tsunami recovery plan. Disaster Prevention and Management: An International Journal, 15, 31-50. ALLISON, E. H. & ELLIS, F. 2001. The livelihoods approach and management of small- scale fisheries. Marine Policy, 25, 377-388. ALLISON, E. H. & HOREMANS, B. 2006. Putting the principles of the Sustainable Livelihoods Approach into fisheries development policy and practice. Mar. Pol., 30, 757-766. ALONGI, D. M., MCKINNON, A. D., BRINKMAN, R., TROTT, L. A., UNDU, M. C., MUAWANAH & RACHMANSYAH 2009. The fate of organic matter derived from small-scale fish cage aquaculture in coastal waters of Sulawesi and Sumatra, Indonesia. Aquaculture, 295, 60-75. ANAMBAS-REGENCY 2012. Facing the future of Anambas as the new national rising centre of of internationally recognised economic development area (menyongsong mada depan Anambas sebagai pusat pertumbuhan economi baru nasional berskala internasional). Presented at the meeting of the Coordinating Ministry for Economics of GoI, 5 December 2012. In: DISTRICT, A. A. (ed.). Tarempa: Anambas District. ANAMBAS, D. & IPB, L. 2011. Studi Kelayakan Pengembangan Budidaya Rumput Laut. Program Pengembangan Budidaya Perikanan Kabupaten Kepulauan Anambas. Tarempa: Dinas Kelautan dan Perikanan Kabupaten Anambas & LPPM IPB ANDERSON, S. J. 2014. HF Radar Network Design for Remote Sensing of the South China Sea. In: MARGHANY, P. M. (ed.) Advanced Geoscience Remote Sensing. InTech. ANDRADE, G. & RHODES, J. 2012. Protected areas and local communities: An inevitable partnership toward successful conservation strategies? Ecology and Society, 17. ANONYMOUS 2007. Local environmental effects of aquaculture farming sea bream/sea bass in the Mediterranean Sea; A MOM (Modelling- Ongrowing fish farm- Monitoring system) model study. ANONYMOUS 2011. Kumpulan Peraturan Perundangan-Undangan di Bidang Pengelolaan Terumbu Karang. In: KARANG, S. R. D. P. T. & DIREKTORAT JENDERAL KELAUTAN, P. D. P.-P. K. (eds.). Jakarta: MMAF. ANONYMOUS 2012a. Pedoman Teknis Evaluasi Efektivitas Pengelolaan Kawasan Konservasi Perairan, Pesisir dan Pulau-Pulau Kecil (E-KKP3K). In: MMAF, D.-. (ed.) Nomor KEP.44/KP3K/2012. Jakarta: DKKJI - MMAF. ANONYMOUS. 2012b. Sejarah Perkembangan Kawasan Konservasi Perairan Indonesia [Online]. Jakarta: DKKJI - MMAF. Available: http://kkji.kp3k.kkp.go.id/index.php/informasi-konservasi/87-sejarahperkembangan [Accessed 16 December 2015 2015]. ANONYMOUS 2014. Value Chain Analysis of Marine Fish Aquaculture in Indonesia : Bussiness Opportunities for Norwegian Companies. Jakarta: Innovation Norway and Spire. ANZECC 2000. Australian and New Zealand Guidelines for Fresh and Marine Water Quality. NATIONAL WATER QUALITY MANAGEMENT STRATEGY. Canberrra: Australian and New Zealand Environment and Conservation Council, Agriculture and Resource Management Council of Australia and New Zealand.

184

APHA 2012. Standard methods for the examination of water and wastewater, Washington D.C., American Public Health Association. 22nd. ASIAN DEVELOPMENT BANK 2014. State of the Coral Triangle: Indonesia. State of the Coral Triangle. Mandaluyong City, Philippines: Asian Development Bank, Coral Triangle Initiative, Global Environment Facility. BAKER, J. L. 2000. Guide to Marine Protected Areas. In: BRIDGLAND, J. (ed.). Adelaide: Office for Coast and Marine, Department for Environment and Heritage. BAN, N. C., ADAMS, V. M., ALMANY, G. R., BAN, S., CINNER, J. E., MCCOOK, L. J., MILLS, M., PRESSEY, R. L. & WHITE, A. 2011. Designing, implementing and managing marine protected areas: Emerging trends and opportunities for coral reef nations. Journal of Experimental Marine Biology and Ecology, 408, 21-31. BAPPEDA-ANAMBAS 2009. Profil Kabupaten Kepulauan Anambas Tahun 2009. Profil Kabupaten Kepulauan Anambas. Tarempa, Kabupaten Anambas: Badan Perencanaan Pembangunan Penelitian Pengembangan dan Penanaman Modal Daerah Kabupaten Kepulauan Anambas. BAPPEDA-ANAMBAS 2011. Rencana Tata Ruang Wilayah Kabupaten Kepulauan Anambas Tahun 2011 - 2031, Tarempa, Pemerintah Kabupaten Kepulauan Anambas. BAPPENAS. 2012. Profil Pembangunan Kepulauan Riau. Profil Daerah di Indonesia [Online]. [Accessed 10 December 2014]. BAYRON, C. J. & COSTA-PIERCE, B. A. 2013. Carrying capacity tools for use in the implementation of an ecosystem approach to aquaculture. In: ROSS, L. G., TELFER, T. C., FALCONER, L., SOTO, D. & AGUILAR-MANJARREZ, J. (eds.) Site selection and carrying capacities for inland and coastal aquaculture, pp. 129– 146. FAO/Institute of Aquaculture, University of Stirling, Expert Workshop, 6–8 December 2010. Stirling, the United Kingdom of Great Britain and Northern Ireland. FAO Fisheries and Aquaculture Proceedings. Rome: FAO. BENE, C. 2009. Are Fishers Poor or Vulnerable? Assessing Economic Vulnerability in Small- Scale Fishing Communities. J. Dev. Stud., 45, 911-933. BENNETT, N. & DEARDEN, P. 2014. Why local people do not support conservation: Community perceptions of marine protected area livelihood impacts, governance and management in Thailand. Mar. Pol., 44, 107-116. BMKG-TAREMPA 2014. Meteorological dataset at the Tarempa Station. In: BMKG- TAREMPA (ed.). Tarempa. BÖHRINGER, C. & JOCHEM, P. E. 2007. Measuring the immeasurable—A survey of sustainability indices. Ecological economics, 63, 1-8. BONDAD-REANTASO, M. G., SUBASINGHE, R. P., ARTHUR, J. R., OGAWA, K., CHINABUT, S., ADLARD, R., TAN, Z. & SHARIFF, M. 2005. Disease and health management in Asian aquaculture. Vet Parasitol, 132, 249-72. BOTTEMA, M. J. M. & BUSH, S. R. 2012. The durability of private sector-led marine conservation: A case study of two entrepreneurial marine protected areas in Indonesia. Ocean and Coastal Management, 61, 38-48. BOYATZIS, R. E. 1998. Transforming qualitative information: Thematic analysis and code development, sage. BPPT & ANAMBAS-REGENCY 2011. Masterplan Pembangunan Kawasan Minapolitan di Kabupaten Kepulauan Anambas Provinsi Kepulauan Riau 2011-2016. Jakarta: Anambas Regency. 185

BPS-ANAMBAS 2009. Kepuluan Anambas dalam Angka (Anambas Archipelago in Figures) 2009, Tarempa, Indonesia, BPS Kabupaten Kepulauan Anambas. BPS-ANAMBAS 2012. Kepuluan Anambas dalam Angka (Anambas Archipelago in Figures) 2013, Tarempa, Indonesia, BPS Kabupaten Kepulauan Anambas. BPS. 2010. Population of Indonesia by the Village in 2010 (Penduduk Indonesia menurut Desa tahun 2010). Kependudukan [Online]. [Accessed 19 August 2013]. BRAGER, L. M., CRANFORD, P. J., GRANT, J. & ROBINSON, S. M. C. 2015. Spatial distribution of suspended particulate wastes at open-water Atlantic salmon and sablefish aquaculture farms in Canada. 6, 135-149. BRICKER, S. B., FERREIRA, J. G. & SIMAS, T. 2003. An integrated methodology for assessment of estuarine trophic status. Ecological Modelling, 169, 39-60. BRIGUGLIO, L. 1995. Small island developing states and their economic vulnerabilities. World Development, 23, 1615-1632. BURROUGH, P. A. & MCDONNELL, R. A. 1998. Principles of GIS. Oxford University Press: London, UK. BYRON, C., BENGTSON, D., COSTA-PIERCE, B. & CALANNI, J. 2011. Integrating science into management: Ecological carrying capacity of bivalve shellfish aquaculture. Marine Policy, 35, 363-370. BYRON, C. J. & COSTA-PIERCE, B. A. 2013. Carrying capacity tools for use in the implementation of an ecosystems approach to aquaculture. Site selection and carrying capacity for inland and coastal aquaculture, 87-101. CALGARO, E., LLOYD, K. & DOMINEY-HOWES, D. 2014. From vulnerability to transformation: a framework for assessing the vulnerability and resilience of tourism destinations. Journal of Sustainable Tourism, 22, 341-360. CARIFIO, J. & PERLA, R. J. 2007. Ten common misunderstandings, misconceptions, persistent myths and urban legends about Likert scales and Likert response formats and their antidotes. Journal of Social Sciences, 3, 106-116. CARVER, C. E. A. & MALLET, A. L. 1990. Estimating the carrying capacity of a coastal inlet for mussel culture. Aquaculture, 88, 39–53. CATER, E. 1993. Ecotourism in the third world: problems for sustainable tourism development. Tourism Management, 14, 85-90. CHEN, S., TIMMONS, M. B., ANESHANSLEY, D. J. & BISOGNI, J. J. 1993. Suspended solids characteristics from recirculating aquacultural systems and design implications. Aquaculture, 112, 143-155. CHRISTIE, P. 2004. Marine Protected Areas as Biological Successes and Social Failures in Southeast Asia. American Fisheries Society Symposium, 42, 155–164. CHRISTIE, P. 2005. Is Integrated Coastal Management Sustainable? Ocean and Coastal Management, 48, 208-232. CHU, J. C. W. 2002. Environmental management of mariculture: the effect of feed types on feed waste. Report of the Regional Workshop on Sustainable Seafarming and Grouper Aquaculture APEC/NACA/BOBP/GOI 2002. Collaborative APEC Grouper Research and Development Network (FWG 01/99) (2002). CHU, P. C., EDMONS, N. L. & FAN, C. W. 1999. Dynamical mechanisms for the South China Sea seasonal circulation and thermohaline variabilities. J. Phys. Oceanogr., 29, 2971-2989. CHUA, T. E. & TENG, S. K. 1980. Economic production of estuary grouper, Epinephelus salmoides Maxwell, reared in floating net cages. Aquaculture, 20, 187-228. 186

CINNER, J. E., MCCLANAHAN, T. R. & WAMUKOTA, A. 2010. Differences in livelihoods, socioeconomic characteristics, and knowledge about the sea between fishers and non-fishers living near and far from marine parks on the Kenyan coast. Marine Policy, 34, 22-28. CLAUSSEN, M. J. D. & GREEN, M. S. J. 2007. A Quick Analysis of MPA financing and A Financial Management Framework for MPA Networks. CLIFTON, J. 2013. Refocusing conservation through a cultural lens: Improving governance in the Wakatobi National Park, Indonesia. Marine Policy, 41, 80-86. COCHRANE, J. 2006. Indonesian national parks: Understanding Leisure Users. Annals of Tourism Research, 33, 979-997. COLT, J. & HUGUENIN, J. E. 2002. Design and operating guide for aquaculture seawater systems, Elsevier. CORNER, R. A., BROOKER, A. J., TELFER, T. C. & ROSS, L. G. 2006. A fully integrated GIS-based model of particulate waste distribution from marine fish-cage sites. Aquaculture, 258, 299-311. CRIBB, R. 2008. CONSERVATION IN COLONIAL INDONESIA. International Journal of Postcolonial Studies, 9, 49-61. CROSS, S. & KINGZETT, B. 1992. Biophysical criteria for shellfish culture in British Columbia: A site capability evaluation system. Victoria, Canada: British Columbia Ministry of Agriculture. Food and Fisheries. DAHURI, R. & DUTTON, I. M. 2000. Integrated coastal and marine management enters a new era in Indonesia. Integrated Coastal Zone Management, 1, 1-16. DAY, J. 2002. Zoning - lessons from the Great Barrier Reef Marine Park. Ocean & Coastal Management, 45, 139-156. DE SILVA, S. S. 2012. Aquaculture: a newly emergent food production sector-and perspectives of its impacts on biodiversity and conservation. Biodivers. Conserv., 21, 3187-3220. DEMPSTER, T., SANCHEZ-JEREZ, P., TUYA, F., FERNANDEZ-JOVER, D., BAYLE- SEMPERE, J., BOYRA, A. & HAROUN, R. 2006. Coastal aquaculture and conservation can work together. Marine Ecology Progress Series, 314, 309–310. DFID, U. 1999. Sustainable livelihoods guidance sheets. London: DFID. DGA 2012. Technical Procedure of Aquaculture Best Practice Application : Floating Net Cage Mariculture. In: DGA (ed.). Jakarta: DGA-MMAF. DGA 2017. Information System of Aquaculture Management (SIMAqua). Jakarta: DGA. DIRHAMSYAH 2012. IUU fishing in Indonesia’s live reef fisheries. Australian Journal of Maritime & Ocean Affairs, 4, 44-52. DIRHAMSYAH, D. 2006. Indonesian legislative framework for coastal resources management: A critical review and recommendation. Ocean & Coastal Management, 49, 68-92. DISHIDROS 2013. Daftar Pasang Surut (Tide Tables) 2013 Kepulauan Indonesia, Jakarta, Dinas Hidrografi dan Oseanografi TNI Angkatan Laut. DJKKI 2012. Pedoman Umum Pemanfaatan Kawasan Konservasi Perairan Untuk Budidaya Perikanan. In: DKKJI (ed.). Jakarta: DKKJI - MMAF. DJPB 2016. Laporan Kinerja (LKj) Direktorat Jenderal Perikanan Budidaya Tahun 2015. In: DJPB (ed.). Jakarta: DJPB-MMAF.

187

DKKJI 2013a. Mengenal Potensi Kawasan Konservasi Perairan, Pesisir dan Pulau-pulau Kecil di Indonesia : profil 76 kawasan konservasi perairan, dan pulau-pulau kecil, Kementerian Kelautan dan Perikanan. DKKJI 2013b. Petunjuk teknis penataan batas kawasan konservasi perairan pesisir dan pulau-pulau kecil. In: IKAN, D. K. K. D. J. (ed.). Jakarta: Direktorat Konservasi Kawasan dan Jenis Ikan, Direktotat Jenderal Kelautan, Pesisir dan Pulau-pulau Kecil, Kementerian Kelautan dan Perikanan RI. DKP-ANAMBAS 2012. Fisheries Recapitulation Database for Anambas Archipelago Regency. In: ANAMBAS, D. (ed.) Fisheries Recapitulation Database 2011. Tarempa, Anambas. DKP-ANAMBAS & BIG 2011. Spatial database of small islands in Anambas Regency (Database spasial pulau pulau di Kabupaten Anambas) In: DKP-ANAMBAS & BIG (eds.). Anambas & Jakarta: DKP-Anambas. DKP-ANAMBAS & LPPM-IPB 2011a. Laporan Kemajuan Penyusunan Studi Kelayakan Keramba Jaring Apung Perikanan di Kabupaten Kepulauan Anambas. Tarempa, Kabupaten Kepulauan Anambas: Dinas Kelautan dan Perikanan, Kabupaten Kepulauan Anambas. DKP-ANAMBAS & LPPM-IPB 2011b. Studi Kelayakan Pengembangan Budidaya Rumput Laut. Program Pengembangan Budidaya Perikanan Kabupaten Kepulauan Anambas. Tarempa: Dinas Kelautan dan Perikanan Kabupaten Anambas & LPPM IPB. DONNER, S. D. & POTERE, D. 2007. The inequity of the global threat to coral reefs.(Viewpoint). BioScience, 57, 214. DONOHOE, H. M. & NEEDHAM, R. D. 2006. Ecotourism: The Evolving Contemporary Definition. Journal of Ecotourism, 5, 192-210. DUARTE, P., MENESES, R., HAWKINS, A. J. S., ZHU, M., FANG, J. & GRANT, J. 2003. Mathematical modelling to assess the carrying capacity for multi-species culture within coastal waters. Ecological Modelling, 168, 109-143. DUNG, L. D. 2009. Nha Trang Bay marine protected area, Vietnam: Initial trends in coral structure and some preliminary linkages between these trends and human activities (2002–2005). Aquatic Ecosystem Health & Management, 12, 249-257. EDMONDSON, J. C. 1992. Bali Revisited: Rural Economy Intergenerational Exchanges, and the Transition to Smaller Family Sizes, 1977-90, Pusat Kajian Budaya, Universitas Udayana. EDWARDS, P. 2000. Aquaculture, poverty impacts and livelihoods, Overseas Development Institute. EHRGOTT, M., FIGUEIRA, J. E. R. & GRECO, S. (eds.) 2010. Trends in Multiple Criteria Decision Analysis, New York: Springer. ELLIS, F. 1998. Household strategies and rural livelihood diversification. The Journal of Development Studies, 35, 1-38. ELLIS, F. 2000. Rural livelihoods and diversity in developing countries, New York, USA, Oxford University Press Inc. EPA-VICTORIA 2009. Sampling and analysis of waters, wastewatersm soils and wastes. In: VICTORIA, E. (ed.). Victoria: EPA Victoria. ERENSTEIN, O. 2011. Livelihood Assets as a Multidimensional Inverse Proxy for Poverty: A District-level Analysis of the Indian Indo-Gangetic Plains. J. Hum. Dev. Capab., 12, 283-302. 188

ERENSTEIN, O., HELLIN, J. & CHANDNA, P. 2007. Livelihoods, poverty and targeting in the Indo-Gangetic Plains: A spatial mapping approach, CIMMYT. ERENSTEIN, O., HELLIN, J. & CHANDNA, P. 2010. Poverty mapping based on livelihood assets: A meso-level application in the Indo-Gangetic Plains, India. Appl. Geogr., 30, 112-125. ERVIK, A., HANSEN, P. K., AURE, J., STIGEBRANDT, A., JOHANNESSEN, P. & JAHNSEN, T. 1997. Regulating the local environmental impact of intensive marine fish farming I. The concept of the MOM system (Modelling-Ongrowing fish farms- Monitoring). Aquaculture, 158, 85-94. ESSEN, L.-M. V., FERSE, S. C. A., GLASER, M. & KUNZMANN, A. 2013. Attitudes and perceptions of villagers toward community-based mariculture in Minahasa, North Sulawesi, Indonesia. Ocean & coastal management. FABINYI, M. 2008. Dive tourism, fishing and marine protected areas in the Calamianes Islands, Philippines. Marine Policy, 32, 898-904. FALCONER, L., HUNTER, D.-C., TELFER, T. C. & ROSS, L. G. 2013. Visual, seascape and landscape analysis to support coastal aquaculture site selection. Land Use Policy, 34, 1-10. FAO 1995. Code of Conduct for Responsible Fisheries, Rome, FAO. 41p. FAO 2008. Glossary of aquaculture. Online: http://www.fao.org/fi /glossary/aquaculture [accessed 17.08.18]. FAUZI, A. & BUCHARY, E. A. 2002. A Socioeconomic Perspective of Environmental Degradation at Kepulauan Seribu Marine National Park, Indonesia. Coastal Management, 30, 167-181. FEREDAY, J. & MUIR-COCHRANE, E. 2008. Demonstrating rigor using thematic analysis: A hybrid approach of inductive and deductive coding and theme development. International journal of qualitative methods, 5, 80-92. FERREIRA, J. G., AGUILAR-MANJARREZ, J., BACHER, C., BLACK, K., DONG, S., GRANT, J., HOFMANN, E. E., KAPETSKY, J., LEUNG, P. & PASTRES, R. Progressing aquaculture through virtual technology and decision-support tools for novel management. Global Conference on Aquaculture 2010, 2012. FERREIRA, J. G., RAMOS, L. & COSTA-PIERCE, B. A. 2013. Key drivers and issues surrounding carrying capacity and site selection, with emphasis on environmental components. In: ROSS, L. G., TELFER, T. C., FALCONER, L., SOTO, D. & AGUILAR-MANJARREZ, J. (eds.) Site selection and carrying capacities for inland and coastal aquaculture, pp. 129–146. FAO/Institute of Aquaculture, University of Stirling, Expert Workshop, 6–8 December 2010. Stirling, the United Kingdom of Great Britain and Northern Ireland. FAO Fisheries and Aquaculture Proceedings. Rome: FAO. FERSE, S., GLASER, M., NEIL, M. & MANEZ, K. S. 2014. To cope or to sustain? Eroding long- term sustainability in an Indonesian coral reef fishery. Reg. Envir. Chang., 14, 2053-2065. FITRIANA, R. 2014. Assessing the impact of a marine protected area on coastal livelihoods : a case study from Pantar Island, Indonesia. Charles Darwin University. FOUNTABERT, A. C. D., DOWNES, D. R. & TUNDI AGARDY, S. 1999. Biodiversity in the Seas : Implementing the Convention on Biological Diversity in Marine and Coastal Habitats. IUCN Environmental Policy and Law Paper No. 32. IUCN Gland and Cambridge. 189

GEČEK, S. & LEGOVIĆ, T. 2010. Towards carrying capacity assessment for aquaculture in the Bolinao Bay, Philippines: A numerical study of tidal circulation. Ecological Modelling, 221, 1394-1412. GEMITZI, A., TSIHRINTZIS, V. A., VOUDRIAS, E., PETALAS, C. & STRAVODIMOS, G. 2007. Combining geographic information system, multicriteria evaluation techniques and fuzzy logic in siting MSW landfills. Environmental Geology, 51, 797-811. GIBB, T. & WALKER, J. 2011. Educating for a high skills society? The landscape of federal employment, training and lifelong learning policy in Canada. Journal of Education Policy, 26, 381-398. GIL, D. G. 1973. Unravelling social policy: Theory, analysis, and political action towards social equality, Schenkman Books. GLASER, M., BAITONINGSIH, W., FERSE, S., NEIL, M. & DESWANDI, R. Whose sustainability? Top-down participation in MPA management in Indonesia. Proceeding of the 12th International Coral Reef Symposium: Cairns, Australia. Available at http://www. icrs2012. com/proceedings/manuscripts/ICRS2012_22A_2. pdf[Links], 2012. GLASER, M., BRECKWOLDT, A., DESWANDI, R., RADJAWALI, I., BAITONINGSIH, W. & FERSE, S. C. 2015. Of exploited reefs and fishers–A holistic view on participatory coastal and marine management in an Indonesian archipelago. Ocean & Coastal Management, 116, 193-213. GOMEZ, D. K., MORI, K.-I., OKINAKA, Y., NAKAI, T. & PARK, S. C. 2010. Trash fish can be a source of betanodaviruses for cultured marine fish. Aquaculture, 302, 158- 163. GOODWIN, H. 2002. Local Community Involvement in Tourism around National Parks: Opportunities and Constraints. Current Issues in Tourism, 5, 338-360. GRANTHAM, H., AGOSTINI, V., WILSON, J., MANGUBHAI, S., HIDAYAT, N., MULJADI, A., MUHAJIR, ROTINSULU, C., MONGDONG, M., BECK, M. & POSSINGHAM, H. 2013. A comparison of zoning analyses to inform the planning of a marine protected area network in Raja Ampat, Indonesia. Marine Policy, 38, 184-194. GREEN, A. L., FERNANDES, L., ALMANY, G., ABESAMIS, R., MCLEOD, E., ALIÑO, P. M., WHITE, A. T., SALM, R., TANZER, J. & PRESSEY, R. L. 2014. Designing Marine Reserves for Fisheries Management, Biodiversity Conservation, and Climate Change Adaptation. Coastal Management, 42, 143-159. HAHN, M. B., RIEDERER, A. & FOSTER, S. 2009. The Livelihood Vulnerability Index: A pragmatic approach to assessing risks from climate variability and change-A case study in Mozambique. Glob. Environ. Change-Human Policy Dimens., 19, 74-88. HALIDE, H., STIGEBRANDT, A., REHBEIN, M. & MCKINNONA, A. D. 2009. Developing a decision support system for sustainable cage aquaculture. Environ. Modell. Softw., 24, 694-702. HALPERN, B. S. 2003. The impact of marine reserves: do reserves work and does reserve size matter? Ecological applications : a publication of the Ecological Society of America., 13, 117 - 137. HAMPTON, M. P. & JEYACHEYA, J. 2015. Power, Ownership and Tourism in Small Islands: Evidence from Indonesia. World Development.

190

HANEBUTH, T. J. J., VORIS, H. K., YOKOYAMA, Y., SAITO, Y., OKUNO, J. & ICHI 2011. Formation and fate of sedimentary depocentres on Southeast Asia's Sunda Shelf over the past sea-level cycle and biogeographic implications. Earth-Science Reviews, 104, 92-110. HARGRAVE, B. T. 2002. A traffic light decision system for marine finfish aquaculture siting. Ocean & Coastal Management, 45, 215-235. HEVIA, M., ROSENTHAL, H. & GOWEN, R. J. 1996. Modelling benthic deposition under fish cages. Journal of Applied Ichthyology, 12, 71-74. HUANG, Y. C. A., HSIEH, H. J., HUANG, S. C., MENG, P. J., CHEN, Y. S., KESHAVMURTHY, S., NOZAWA, Y. & CHEN, C. A. 2011. Nutrient enrichment caused by marine cage culture and its influence on subtropical coral communities in turbid waters. Marine Ecology Progress Series, 423, 83-93. INGLIS, G. J., HAYDEN, B. J. & ROSS, A. H. 2000. An Overview of Factors Affecting the Carrying Capacity of Coastal Embayments for Mussel Culture. Christchurch Client Report CHC00/69. New Zealand: NIWA. ISLAM, M. S. 2005. Nitrogen and phosphorus budget in coastal and marine cage aquaculture and impacts of effluent loading on ecosystem: review and analysis towards model development. Marine Pollution Bulletin, 50, 48-61. IUCN & UNEP 2013. The World Database on Protected Areas (WDPA). Cambridge, UK: UNEP-WCMC. IWAMA, G. K. 1991. Interactions between aquaculture and the environment. Critical Reviews in Environmental Control, 21, 177-216. JIANG, W. & GIBBS, M. T. 2005. Predicting the carrying capacity of bivalve shellfish culture using a steady, linear food web model. Aquaculture, 244, 171-185. KAPETSKY, J. M. & AGUILAR-MANJARREZ, J. 2013. From estimating global potential for aquaculture to selecting farm sites: perspectives on spatial approaches and trends. In: ROSS, L. G., TELFER, T. C., FALCONER, L., SOTO, D. & AGUILAR-MANJARREZ, J. (eds.) Site selection and carrying capacities for inland and coastal aquaculture, pp. 129–146. FAO/Institute of Aquaculture, University of Stirling, Expert Workshop, 6–8 December 2010. Stirling, the United Kingdom of Great Britain and Northern Ireland. FAO Fisheries and Aquaculture Proceedings. Rome: FAO. KARIN, P. N. P. 2015. Implementasi penentapan batas desa di Desa Nanga Betung Kecamatan Boyan Tanjung Kabupaten Kapuas Hulu. GOVERNANCE, 3. KARTHIK, M., SURI, J., SAHARAN, N. & BIRADAR, R. S. 2005. Brackish water aquaculture site selection in Palghar Taluk, Thane district of Maharashtra, India, using the techniques of remote sensing and geographical information system. Aquacultural Engineering, 32, 285-302. KELLEHER, G. & KENCHINGTON, R. 1991. Guidelines for Establishing Marine Protected Areas. In: ZHU, M. (ed.) A Marine Conservation and Development Report. Gland, Switzerland: IUCN. KEMENLH 2004. Keputusan Menteri Negara Lingkungan Hidup No. 51/2004 Tentang Baku Mutu Air Laut (Ministerial Decree Number 51/2004 Regarding the Standards of Marine Water Quality). In: ENVIRONMENT, M. O. (ed.) 51/2004. Jakarta: KemnLKH. KERR, S. A. 2005. What is small island sustainable development about? Ocean & Coastal Management, 48, 503-524. 191

KO, T. G. 2005. Development of a tourism sustainability assessment procedure: a conceptual approach. Tourism Management, 26, 431-445. KUSUMAWATI, I. & HUANG, H.-W. 2015. Key factors for successful management of perception of two MPAs in ׳marine protected areas: A comparison of stakeholders Weh island, Sabang, Aceh, Indonesia. Marine policy. LAINING, A. & KRISTANTO, A. H. Aquafeed development and utilization of alternative dietary ingredients in aquaculture feed formulations in Indonesia. Development and Use of Alternative Dietary Ingredients or Fish Meal Substitutes in Aquaculture Feed Formulation: Proceedings of the ASEAN Regional Technical Consultation on Development and Use of Alternative Dietary Ingredients or Fish Meal Substitutes in Aquaculture Feed Formulation, 9-11 December 2014, Nay Pyi Taw, Myanmar, 2015. Aquaculture Department, Southeast Asian Fisheries Development Center, 1- 3. LAUER, P., LÓPEZ, L., SLOAN, E., SLOAN, S. & DOROUDI, M. 2015. Learning from the systematic approach to aquaculture zoning in South Australia: A case study of aquaculture (Zones – Lower Eyre Peninsula) Policy 2013. Marine Policy, 59, 77-84. LEE, V. 1999. An assessment of ecotourism’s abilities to provide local benefits, Bunaken National Park, Indonesia. Report for the Collaborative Environmental Project in Indonesia (CEPI). Funded by the Canadian International Development Agency. LEVINGS, C., ERVIK, A., JOHANNESSEN, P. & AURE, J. 1995. Ecological criteria used to help site fish farms in fjords. Estuaries, 18, 81-90. LKKPN-PEKANBARU 2013. Rencana Pengelolaan dan Zonasi Wisata Perairan Kepulauan Anambas dan Laut Sekitarnya. Pekanbaru, Riau Archieplago, Indonesia: LKKPN Pekanbaru, KP3K, MMAF. LOKA, J., VAIDYA, N. & PHILIPOSE, K. 2012. Site and species selection criteria for cage culture. LONGDILL, P. C., HEALY, T. R. & BLACK, K. P. 2008. An integrated GIS approach for sustainable aquaculture management area site selection. Ocean & Coastal Management, 51, 612-624. LOO, Y. Y., BILLA, L. & SINGH, A. 2014. Effect of climate change on seasonal monsoon in Asia and its impact on the variability of monsoon rainfall in Southeast Asia. Geoscience Frontiers. LOWE, C. 2003. Sustainability and the question of “enforcement” in integrated coastal management: the case of Nain Island, Bunaken National Park. Indonesian Journal of Coastal and Marine Resources Special Edition, 1, 49-63. LOYA, Y., LUBINEVSKY, H., ROSENFELD, M. & KRAMARSKY-WINTER, E. 2004. Nutrient enrichment caused by in situ fish farms at Eilat, Red Sea is detrimental to coral reproduction. Marine Pollution Bulletin, 49, 344-353. MACAROV, D. 1974. Review of Unravelling social policy: Theory, analysis, and political action toward social equality. American Journal of Orthopsychiatry, 44, 463-464. MACBUB, B. 2018. Model perhitungan daya tampung beban pencemaran air danau dan waduk. Jurnal Sumber Daya Air, 6, 129-144. MALCEZEWSKI, J. 2000. On the Use of Weighted Linear Combination Method in GIS: Common and Best Practice Approaches. Transactions in GIS, 4, 5. MARTIN, S. M., LORENZEN, K. & BUNNEFELD, N. 2013. Fishing farmers: Fishing, livelihood diversification and poverty in rural Laos. Human ecology, 41, 737-747.

192

MARZUQI, M. & ANJUSARY, D. N. 2013. Kecernaan nutrien pakan dengan kadar protein dan lemak berbeda pada juvenil ikan kerapu pasir (Epinephelus corallicola). Jurnal Ilmu dan Teknologi Kelautan Tropis, 5, 311-323. MASCIA, M. B. & CLAUS, C. A. 2009. A property rights approach to understanding human displacement from protected areas: The case of marine protected areas. Conservation Biology, 23, 16-23. MAYNARD, J. A., ANTHONY, K. R. N., AFATTA, S., DAHL-TACCONI, N. & HOEGH-GULDBERG, O. V. E. 2010. Making a Model Meaningful to Coral Reef Managers in a Developing Nation: a Case Study of Overfishing and Rock Anchoring in Indonesia. Diseño de un Modelo Significativo para Manejadores de Arrecifes de Coral en un País en Desarrollo: un Estudio de Caso de Sobrepesca y Anclaje con Rocas en Indonesia., 24, 1316-1326. MCGRANAHAN, G., BALK, D. & ANDERSON, B. 2007. The rising tide: assessing the risks of climate change and human settlements in low elevation coastal zones. Environment and Urbanization, 19, 17-37. MCKINDSEY, C. W., THETMEYER, H., LANDRY, T. & SILVERT, W. 2006. Review of recent carrying capacity models for bivalve culture and recommendations for research and management. Aquaculture, 261, 451-462. MCKINNON, A. D., TROTT, L. A., BRINKMAN, R., DUGGAN, S., CASTINE, S., O’LEARY, R. A. & ALONGI, D. M. 2010. Seacage aquaculture in a World Heritage Area: The environmental footprint of a Barramundi farm in tropical Australia. Marine Pollution Bulletin, 60, 1489-1501. MCLEOD, E., SALM, R., GREEN, A. & ALMANY, J. 2008. Designing marine protected area networks to address the impacts of climate change. Frontiers in ecology and the environment. MCLEOD, E., SZUSTER, B. & SALM, R. 2009. Sasi and Marine Conservation in Raja Ampat, Indonesia. Coastal Management, 37, 656-676. MCLEOD, I., PANTUS, F. & PRESTON, N. 2002. The use of a geographical information system for land-based aquaculture planning. Aquaculture Research, 33, 241-250. MEADEN, G. J. & AGUILAR-MANJARREZ, J. 2013. Advances in geographic information systems and remote sensing for fisheries and aquaculture. Summary version. FAO Fisheries and Aquaculture Technical Paper No. 552. [Online]. MENLH 2012. Ministerial decree concerning plans of business and or activity required to have an environmental impact assessment. In: ENVIRONMENT, S. M. F. (ed.) Number 05 Tahun 2012. Jakarta: MenLH. MENTE, E., PIERCE, G. J., SANTOS, M. B. & NEOFITOU, C. 2006. Effect of feed and feeding in the culture of salmonids on the marine aquatic environment: a synthesis for European aquaculture. Aquaculture International, 14, 499-522. MERCERON, M., KEMPF, M., BENTLEY, D., GAFFET, J. D., LE GRAND, J. & LAMORT-DATIN, L. 2002. Environmental impact of a salmonid farm on a well flushed marine site: I. Current and water quality. Journal of Applied Ichthyology, 18, 40-50. MMAF 2011. Marine and Fisheries Statistics 2011, Jakarta, Indonesia, Ministry of Marine Affairs and Fisheries. MMAF 2013. MMAF Decree regarding Conservation Status of Humpback Wrasse (Cheilinus undulatus) In: FISHERIES, M. O. M. A. A. (ed.). Jakarta: MMAF.

193

MMAF 2014a. Fish Farmer Group Database of Aquaculture Rural Bussiness Development Recipients 2011-2014. In: MMAF (ed.). Jakarta: MMAF. MMAF 2014b. Marine and Fisheries Statistics 2014, Jakarta, Indonesia, Ministry of Marine Affairs and Fisheries. MMAF 2014c. Rencana Pengelolan dan Zonasi Taman Wisata Perairan Kepulauan Anambas dan Laut Sekitarnya di Provinsi Kepuluan Riau Tahun 2014-2034. In: KKP (ed.) 53/KEPMEN-KP/2014. Jakarta: KKP. MMAF 2014d. Rencana Pengelolan dan Zonasi Taman Wisata Perairan Kepulauan Anambas dan Laut Sekotarnya di Provinsi Kepuluan Riau Tahun 2014-2034. In: KKP (ed.) 53/KEPMEN-KP/2014. Jakarta: KKP. MMAF 2016. MMAF Decree regarding Transhipment of Life Reef Fish In: FISHERIES, M. O. M. A. A. (ed.). Jakarta: MMAF. MORA, C., COSTELLO, M. J., ROLLO, A., MYERS, R. A., ANDRÉFOUËT, S., KRANENBURG, C., VERON, J. & GASTON, K. J. 2006. Coral reefs and the global network of marine protected areas. MUCHLISIN, Z. A., FADLI, N., NASUTION, A. M., ASTUTI, R., MARZUKI & MUSNI, D. 2012. Analysis of fuel subsidy (diesel) for fishermen in Aceh Besar District, Aceh Province. DEPIK Journal, 1, 107-113. MULDOON, G. & JOHNSTON, B. Market chain analysis for the trade in live reef food fish. Proceeding of a workshop in Economic and Marketing of Live Reef Food Fish Trade in AsiaPacific, ACIAR Working Paper, 2006. MUNGKUNG, R., AUBIN, J., PRIHADI, T. H., SLEMBROUCK, J., WERF, H. M. G. V. D. & LEGENDRE, M. 2013. Life Cycle Assessment for environmentally sustainable aquaculture management: a case study of combined aquaculture systems for carp and tilapia. Journal of cleaner production, 57, 249-256. MUSTIKA, P. L., GUNAWAN, T. & ERDMANN, M. V. (eds.) 2013. Kajian Cepat Kelautan (Marine Rapid Assessment – MRAP) di Taman Wisata Perairan Kepulauan Anambas 3-31 Mei 2012, Denpasar: Kementerian Kelautan dan Perikanan, Lembaga Ilmu Pengetahuan Indonesia (LIPI), Pemerintah Daerah Kabupaten Anambas, The Nature Conservancy, Conservation International Indonesia. NACA. 1989. Site selection criteria for marine finfish netcage culture in Asia. UNDP/FAO Regional Seafarming Development and Demonstration Project [Online]. [Accessed 26 June 2017]. NATALE, F., CARVALHO, N. & PAULRUD, A. 2015. Defining small-scale fisheries in the EU on the basis of their operational range of activity The Swedish fleet as a case study. Fisheries Research, 164, 286-292. NATH, S. S., BOLTE, J. P., ROSS, L. G. & AGUILAR-MANJARREZ, J. 2000. Applications of geographical information systems (GIS) for spatial decision support in aquaculture. Aquacultural Engineering, 23, 233-278. NAVAS, J. M., TELFER, T. C. & ROSS, L. G. 2011a. Application of 3D hydrodynamic and particle tracking models for better environmental management of finfish culture. Continental shelf research, 31, 675–684. NAVAS, J. M., TELFER, T. C. & ROSS, L. G. 2011b. Spatial modeling of environmental vulnerability of marine finfish aquaculture using GIS-based neuro-fuzzy techniques. Marine Pollution Bulletin, 62, 1786-1799.

194

NAYLOR, R. L., GOLDBURG, R. J., PRIMAVERA, J. H., KAUTSKY, N., BEVERIDGE, M. C. M., CLAY, J., FOLKE, C., LUBCHENCO, J., MOONEY, H. & TROELL, M. 2000. Effect of aquaculture on world fish supplies. Nature, 405, 1017-1024. NG, P. K. L., ILAHUDE, A. G., SIVASOTHI, N. & YEO, D. C. J. 2004. Expedition Anambas: An overview of the scientific marine exploration of the Anambas and Natuna Archipelago, 11- 22 march 2002. Raffles Bulletin of Zoology, 1-17. NIEHOF, A. 2004. The significance of diversification for rural livelihood systems. Food Policy, 29, 321-338. PAPOUTSOGLOU, S. E. 2000. Monitoring and regulation of marine aquaculture in Greece: licensing, regulatory control and monitoring guidelines and procedures. Journal of Applied Ichthyology, 16, 167-171. PATLIS, J. 2007. Indonesia's new fisheries law: Will it encourage sustainable management or exacerbate over-exploitation? Bulletin of Indonesian Economic Studies, 43, 201- 226. PEMDA-ANAMBAS & BAKOSURTANAL 2011. Laporan Akhir Penyusunan Database Pulau-Pulau Gugusan Kepulauan Anambas 2011, Cibinong, Pemda-Anambas and Bakosurtanal. PEREZ, O. M., TELFER, T. C. & ROSS, L. G. 2003. Use of GIS-based models for integrating and developing marine fish cages within the tourism industry in Tenerife (Canary Islands). Coastal management., 3, 355–366. PÉREZ, O. M., TELFER, T. C. & ROSS, L. G. 2002. Geographical information systems‐based models for offshore floating marine fish cage aquaculture site selection in Tenerife, Canary Islands. Aquaculture Research, 36, 946-961. PÉREZ, O. M., TELFER, T. C. & ROSS, L. G. 2005. Geographical information systems- based models for offshore floating marine fish cage aquaculture site selection in Tenerife, Canary Islands. Aquaculture Research, 36, 946-961. PERSOON, G., DE IONGH, H. & WENNO, B. 1999. Exploitation, management and conservation of marine resources: the context of the Aru Tenggara Marine Reserve (Moluccas, Indonesia). Ocean and Coastal Management, 32, 97-122. PET-SOEDE, C., MACHIELS, M., STAM, M. & VAN DENSEN, W. 1999. Trends in an Indonesian coastal fishery based on catch and effort statistics and implications for the perception of the state of the stocks by fisheries officials. Fisheries Research, 42, 41-56. PHILMINAQ 2008. Water Quality Criteria and Standards for Freshwater and Marine Aquaculture. Mitigating Impact from Aquaculture in the Philippines. Manila: PHILMINAQ. POLUNIN, N. V. C., HALIM, M. K. & KVALVÅGNAES, K. 1983. Bali Barat: An Indonesian marine protected area and its resources. Biological Conservation, 25, 171-191. POMEROY, R., PARKS, J., POLLNAC, R., CAMPSON, T., GENIO, E., MARLESSY, C., HOLLE, E., PIDO, M., NISSAPA, A., BOROMTHANARAT, S. & THU HUE, N. 2007. Fish wars: Conflict and collaboration in fisheries management in Southeast Asia. Marine Policy, 31, 645-656. POMEROY, R. S., PARKS, J. E. & BALBOA, C. M. 2006. Farming the reef: is aquaculture a solution for reducing fishing pressure on coral reefs? Marine Policy, 30, 111–130. 195

PONTING, J., MCDONALD, M. & WEARING, S. 2005. De- constructing Wonderland: Surfing tourism in the Mentawai Islands, Indonesia. Loisir et Societe, 28, 141-162. PORTMAN, M. E. 2007. Zoning design for cross-border marine protected areas: The Red Sea Marine Peace Park case study. Ocean and Coastal Management, 50, 499-522. PRELL, C., HUBACEK, K. & REED, M. 2009. Stakeholder Analysis and Social Network Analysis in Natural Resource Management. Society & Natural Resources, 22, 501- 518. PRIMAVERA, J. H. 2006. Overcoming the impacts of aquaculture on the coastal zone. Ocean & Coastal Management, 6, 531–545. PURBA, N. P., KELVIN, J., WIDIARTI, R. & SYADIAH, N. 2014. Inner Water Characteristics of Anambas Island, Riau Province-Indonesia. International Journal of Science and Research (IJSR), 3, 5. RACHMANSYAH. 2004. Carrying capacity analysis of Awarange Bay for milkfish culture in floating net cage. Doctor of Philosophy, Bogor Agricultural University. RADIARTA, I. & SAITOH, S. I. 2009. Biophysical models for Japanese scallop, Mizuhopecten yessoensis, aquaculture site selection in Funka Bay, Hokkaido, Japan, using remotely sensed data and geographic information system. Aquac. Int., 17, 403-419. RADIARTA, I. N., SAITOH, S.-I. & MIYAZONO, A. 2008. GIS-based multi-criteria evaluation models for identifying suitable sites for Japanese scallop (Mizuhopecten yessoensis) aquaculture in Funka Bay, southwestern Hokkaido, Japan. Aquaculture, 284, 127-135. RANGANATHAN, C., SINGH, N., BANTILAN, M., PADMAJA, R. & RUPSHA, B. 2009. Quantitative assessment of Vulnerability to Climate Change: Computation of Vulnerability indices. Unpublished. READ, P. & FERNANDES, T. 2003. Management of environmental impacts of marine aquaculture in Europe. Aquaculture, 226, 139-163. REDFORD, K. H., LEVY, M. A., SANDERSON, E. W. & SHERBININ, A. D. 2008. What is the role for conservation organizations in poverty alleviation in the world's wild places? Oryx. RIFE, A. N., BRAD, E., ALEXANDRA, S. & OCTAVIO, A. O. 2013. When good intentions are not enough … Insights on networks of “paper park” marine protected areas. Conservation Letters, 6, 200-212. RIMMER, M. 2010. Mariculture development in indonesia: Prospects and constraints. Indonesian Aquaculture Journal, 5, 15. RIMMER, M. A., SUGAMA, K., RAKHMAWATI, D., ROFIQ, R. & HABGOOD, R. H. 2013. A review and SWOT analysis of aquaculture development in Indonesia. Rev. Aquac. ROBERTS, A. R., MILLER, D. B. & HOKENSTAD JR, M. C. T. 2012. Long Term Care Insurance Beyond the CLASS Program. J. Soc. & Soc. Welfare, 39, 85. ROMIMOHTARTO, K. 1985. Kualitas Air Dalam Budidaya Laut (Water Quality in Mariculture. SEAFARMING WORKSHOP REPORT BANDAR LAMPUNG 28 OCTOBER - 1 NOVEMBER 1985 PART II - TECHNICAL REPORT. Lampung: Food Agriculture Organisation - United Nation. ROSELI, N. H. & AKHIR, M. F. M. 2014. Variations of Southern South China Sea Characteristics near Pahang. Sains Malaysiana, 43, 1389-1396.

196

ROSS, L. G., PALACIOS, C. A. M. & MORALES, E. J. 2008. Developing native fish species for aquaculture: the interacting demands of biodiversity, sustainable aquaculture and livelihoods. Aquaculture research. ROSS, L. G., TELFER, T. C., FALCONER, L., SOTO, D. & AGUILAR-MANJARREZ, J. Site selection and carrying capacities for inland and coastal aquaculture. FAO/Institute of Aquaculture, University of Stirling, Expert Workshop, 6–8 December 2010., 2013a Stirling, the United Kingdom of Great Britain and Northern Ireland.: FAO Fisheries and Aquaculture Proceedings No. 21. Rome. ROSS, L. G., TELFER, T. C., FALCONER, L., SOTO, D., AGUILAR-MANJARREZ, J., ASMAH, R., BERMÚDEZ, J., BEVERIDGE, M. C. M., BYRON, C. J., CLÉMENT, A., CORNER, R., COSTA-PIERCE,, B.A., C., S., DE WIT, M., DONG, S., FERREIRA, J.G., KAPETSKY, J. M., KARAKASSIS, I., LESCHEN, W. L., D., LUNDEBYE, A.-K., MURRAY, F. J., PHILLIPS, M., RAMOS, L., SADEK, S., SCOTT, P. C., VALLE-LEVINSON, A., WALEY, D., WHITE, P. G. & ZHU, C. 2013b. Carrying capacities and site selection within the ecosystem approach to aquaculture. In: ROSS, L. G., TELFER, T. C., FALCONER, L., SOTO, D. & AGUILAR-MANJARREZ, J. (eds.) Site selection and carrying capacities for inland and coastal aquaculture. Rome: FAO/Institute of Aquaculture, University of Stirling, Expert Workshop, 6–8 December 2010. Stirling, the United Kingdom of Great Britain and Northern Ireland. FAO Fisheries and Aquaculture Proceedings No. 21. Rome, FAO. ROSS, S. & WALL, G. 1999. Evaluating ecotourism: The case of North Sulawesi, Indonesia. Tourism Management, 20, 673-682. ROUNTOS, K. J., PETERSON, B. J. & KARAKASSIS, I. 2012. Indirect effects of fish cage aquaculture on shallow Posidonia oceanica seagrass patches in coastal Greek waters. Aquaculture Environment Interactions, 2, 105-115. RÜCKERT, S., KLIMPEL, S., AL-QURAISHY, S., MEHLHORN, H. & PALM, H. W. 2009. Transmission of fish parasites into grouper mariculture (Serranidae: Epinephelus coioides (Hamilton, 1822)) in Lampung Bay, Indonesia. Parasitology Research, 104, 523. SAIDAH, S. E. 2011. Kajian Pengembangan Wilayah Pesisir Kabupaten Kepulauan Anambas. Jakarta: BPPT Jakarta. SANDERSON, E. W., JAITEH, M., LEVY, M. A., REDFORD, K. H., WANNEBO, A. V. & WOOLMER, G. 2002. The human footprint and the last of the wild. Bioscience, 52, 891-904. SANGADJI, M. N., SUMARDJO, S., ASNGARI, P. S. & SOEWITO, S. H. 2015. Strategi Penyuluhan di Kawasan Konservasi (Kasus Taman Nasional Kepulauan Togean). Jurnal Penyuluhan, 7. SANTOSA, A. (ed.) 2008. Indonesian Conservation, A Potrait of Management and Policy (in Bhs: Konservasi Indonesia, Sebuah Potret Pengeloaan & Kebijakan), Bogor, Indonesia: Kemenhut-PKK-USAID-ESP. SARAGIH, B. B. 2014. War on illegal fishing begins. The Jakarta Post. SCHWARTZ, E. 1974. Social Service Review, 48, 599-601. SCHWARZ, A.-M., BÉNÉ, C., BENNETT, G., BOSO, D., HILLY, Z., PAUL, C., POSALA, R., SIBITI, S. & ANDREW, N. 2011. Vulnerability and resilience of remote rural communities to shocks and global changes: Empirical analysis from Solomon Islands. Global Environmental Change, 21, 1128-1140. 197

SCOONES, I. 1998. Sustainable rural livelihoods: a framework for analysis. SEMBIRING, E., BASUNI, S. & SOEKMADI, R. 2010. Resolusi Konflik Pengelolaan Taman Nasional Teluk Cenderawasih di Kabupaten Teluk Wondama. Jurnal Manajemen Hutan Tropika, 16, 84-91. SETIAWAN, F. unpublished. Kepulauan Raja Ampat. Surabaya: Airlangga University. SETYAWATI 2014. Managing marine protected areas in Indonesia. Journal of the Marine Biological Association of India, 56, 13. SHI-GUO, W., WONG, H., YOU-LANG, L. & ZHI-RONG, L. 1999. Distribution and origin of sediments on the northern Sunda Shelf, South China Sea. Chin. J. Ocean. Limnol., 17, 28-40. SIEVANEN, L., CRAWFORD, B., POLLNAC, R. & LOWE, C. 2005. Weeding through assumptions of livelihood approaches in ICM: Seaweed farming in the Philippines and Indonesia. Ocean Coastal Manage., 48, 297-313. SILVA-CRUZ, Y., CASTAÑEDA-CHÁVEZ, M., LANGO-REYNOSO, F. & LANDEROS-SÁNCHEZ, C. 2011. Environmental impact of fish farming in floating cages in Isla Arena, Campeche. Tropical and Subtropical Agroecosystems, 3, 291-298. https://www.redalyc.org/articulo.oa?id=93920942005. SILVA, C., FERREIRA, J. G., BRICKER, S. B., DELVALLS, T. A., MARTÍN-DÍAZ, M. L. & YÁÑEZ, E. 2011. Site selection for shellfish aquaculture by means of GIS and farm-scale models, with an emphasis on data-poor environments. Aquaculture, 318, 444-457. SILVA, C., YÁÑEZ, E., MARTÍN-DÍAZ, M. L. & DELVALLS, T. A. Assessing a bioremediation strategy in a shallow coastal system affected by a fish farm culture – Application of GIS and shellfish dynamic models in the Rio San Pedro, SW Spain. Marine Pollution Bulletin, 64, 751-765. SILVA, C., YÁÑEZ, E., MARTÍN-DÍAZ, M. L. & DELVALLS, T. A. 2012. Assessing a bioremediation strategy in a shallow coastal system affected by a fish farm culture – Application of GIS and shellfish dynamic models in the Rio San Pedro, SW Spain. Marine Pollution Bulletin, 64, 751-765. SIM-SMITH, C. & FORSYTHE, A. 2013. Comparison of the international regulations and best management practices for marine finfish farming. In: LTD, N. I. O. W. A. R. (ed.). New Zealand: Ministry for Primary Industries. SIM, S.-Y., RIMMER, M., WILLIAMS, K., TOLEDO, J. D., SUGAMA, K., RUMENGAN, I. & PHILLIPS, M. J. 2005. Pedoman praktis pemberian dan pengelolaan pakan untuk ikan kerapu yang di budidaya. Asia-Pacific Marine Finfish Aquaculture Network, 2. SIRY, H. Y. 2009. Making decentralized coastal zone management work in Indonesia : case studies of Kabupaten Konawe and Kabupaten Pangkajene dan Kepulauan. Doctor of Philosophy ANU PhD Digital Theses, Australian National University. SLOAN, N. A. 2002. History and Application of the Wilderness Concept in Marine Conservation. Conservation Biology, 16, 294-305. SOEMODINOTO, A., DJUNAIDI, A. & NUR, J. M. unpublished. Community-based ranching of Napoleon wrasse in Anambas Islands, Riau Islands Province: Its Evolution, Rancher Network, and Business Feasibility. SONG, Y.-H. 2008. The potential marine pollution threat from oil and gas development activities in the disputed South China sea/ spratly area: A role that Taiwan can play. Ocean Development and International Law, 39, 150-177. 198

SPALDING, M. D., MELIANE, I., MILAM, A., FITZGERALD, C. & HALE, L. Z. 2013. Protecting Marine Spaces: Global Targets and Changing Approaches. Ocean Yearbook 27. STACHOWITSCH, M., RIEDEL, B., ZUSCHIN, M. & MACHAN, R. 2007. Oxygen depletion and benthic mortalities: the first in situ experimental approach to documenting an elusive phenomenon. Limnology and Oceanography: Methods, 5, 344-352. STIGEBRANDT, A. 1991. Computations of oxygen fluxes through the sea surface and the net production of organic matter with application to the Baltic and adjacent seas. Limnology and Oceanography, 36, 444-454. STIGEBRANDT, A. 1999. MOM(Monitoring- Ongrowing fish farms- Modelling) turnover of energy and matter by fish- A general model with application to salmon. Fisken og Havet, 26. STIGEBRANDT, A. 2011. Carrying capacity: general principles of model construction. Aquaculture Research, 42, 41-50. STIGEBRANDT, A., AURE, J., ERVIK, A. & HANSEN, P. K. 2004. Regulating the local environmental impact of intensive marine fish farming: III. A model for estimation of the holding capacity in the Modelling–Ongrowing fish farm–Monitoring system. Aquaculture, 234, 239-261. SUDARSHANA, R., BHAN, S. K., SIDDIQUI, M. S., VOHRA, T. P. S., PRASAD, P. N., BHAT, U. G. & NEELAKANTAN, B. 2013. Experiments in site selection for coastal aquaculture using Indian remote- sensing satellite - IRS-1a. Int. J. Remote Sens., 14, 2275-2284. SUDIRMAN, HALIDE, H., JOMPA, J., ZULFIKAR, ISWAHYUDIN & MCKINNON, A. D. 2009. Wild fish associated with tropical sea cage aquaculture in South Sulawesi, Indonesia. Aquaculture, 286, 233-239. SUKADI, M. F. Sustainable aquaculture in Indonesia. Technical paper presented at FFTC- RCA International Workshop on Innovative Technologies for Ecofriendly Fish Farm Management and Production of Safe Aquaculture Foods, 2006 Bali, Indonesia. RCA-MMAF, 4-8. SULLIVAN, C. & MEIGH, J. 2005. Targeting attention on local vulnerabilities using an integrated index approach: the example of the climate vulnerability index. Water Science and Technology, 51, 69. SULLIVAN, K. & GARINE-WICHATITSKY, D. 1994. Energetics of juvenile Epinephelus groupers: impact of summer temperatures and activity patterns on growth rates. SUMAILA, U., TEH, L., WATSON, R., TYEDMERS, P. & PAULY, D. 2008. Fuel price increase, subsidies, overcapacity, and resource sustainability. ICES J. Mar. Sci., 65, 832-840. SUNARJANTO, D. 2012. Peranan Industri Migas dalam Pengembangan Wilayah Kepulauan Anambas dan Wilayah Perbatasan Negara. Minyak dan Gas Bumi Journal, 10, 47-57. SYADIAH, N. 2010. Zoning of Local Marine Conservation Areas for Marine Mariculture (A Case Study at Pasi Island, District of Kepulauan Selayar, South Sulawesi Province). Master, Bogor Agricultural Insititute.

199

SYARIF, L. M. 2009. Promotion and Management of Marine Fisheries in Indonesia. In: WINTER, G. (ed.) Towards Sustainable Fisheries Law. Gland, Switzerland: IUCN. xiv + 340 pp. SZUSTER, B. W. & ALBASRI, H. 2010. Mariculture and Marine Spatial Planning: Integrating Local Ecological Knowledge at Kaledupa Island, Indonesia. Island Studies Journal,, 5, 237-250. TACON, A. G. J., RAUSIN, N., KADARI, M. & CORNELIS, P. 1991. The food and feeding of tropical marine fishes in floating net cages: Asian seabass, Lates calcarifer (Bloch), and brown-spotted grouper, Epinephelus tauvina (Forskal). Aquaculture Research, 22, 165-182. TAM, C.-L. 2015. Timing exclusion and communicating time: A spatial analysis of participation failure in an Indonesian MPA. Marine Policy, 54, 122-129. TARUNAMULIA 2014. Improving the utility of GIS for land suitability assessment for sustainable extensive brackishwater aquaculture in Indonesia. In: SAMMUT, J. & BEER, J. (eds.). University of New South Wales. Biological, Earth & Environmental Sciences. TERANGI, Y., RCI & WWF 2015. Kawasan Konservasi Laut INdonesia. http://www.konservasi- laut.net/index.php?option=com_fabrik&view=table&tableid=3&calculations=1&r esetfilters=1&Itemid=5. 2 June 2015 ed. Jakarta: Konservasi-Laut-Net. TNC 2003. Komodo Fish Culture Project : A pilot project to establish a multi-species reef fish hatchery in Loh Mbongi and village-based grow-out farms in communities surrounding Komodo National Park, West Flores, Indonesia. In: MOUS, P. J., SUDARYANTO & MEYER, T. (eds.). Bali, Indonesia: TNC Southeqast Asia Center for Marine Protected Areas. TNC 2004. Progress report on the Komodo fish culture project: A pilot project to establish a multi-species reef fish hatchery in Loh Mbongi and village-based grow-out farms in communities surrounding Komodo National Park, West Flores, Indonesia. Flores: TNC. TNC 2010. Analisis Kelayakan Kesepakatan Konservasi Laut (MCA, Marine Conservation Agreement) untuk Segitiga Karang - Indonesia ; Temuan-Temuan Sementara – Versi Publik (V.2). Bali: The Nature conservancy. TNC & CI 2012. Practitioner’s Field Guide for Marine Conservation Agreements: Best Practices for Integrating Rights-based Incentive Agreements into Ocean and Coastal Conservation Efforts. Narragansett, Rhode Island: TNC &CI. TNC, WWF & RAJA-AMPAT, P. 2003. Rapid ecological assessment of the Raja Ampat Islands, Papua, Eastern Indonesia. In: DONNELLY, R., NEVILLE, D. & MOUS, D. P. J. (eds.). Indonesia: TNC. TRAN, N., RODRIGUEZ, U. P., CHAN, C. Y., PHILLIPS, M. J., MOHAN, C. V., HENRIKSSON, P. J. G., KOESHENDRAJANA, S., SURI, S. & HALL, S. 2017. Indonesian aquaculture futures: An analysis of fish supply and demand in Indonesia to 2030 and role of aquaculture using the AsiaFish model. Marine Policy, 79, 25-32. TROYNA, B. 1994. Critical social research and education policy. British Journal of Educational Studies, 42, 70-84. TUCKER JR, J. W. 2012. Marine fish culture, Springer Science & Business Media.

200

TUNG, H. 2002. Improving local livelihood through sustainable aquaculture in Hon Mun marine protected area. Hon Mun Marine Protected Area Pilot Project. Vietnam: Thun San Tinh - Kanh Hoa - IUCN. TURVEY, R. 2007. Vulnerability Assessment of Developing Countries: The Case of Small-island Developing States. Development Policy Review, 25, 243-264. USEPA 2003. Ambient water quality criteria for dissolved oxygen, water clarity and chlorophyll a for the Chesapeake Bay and its tidal tributaries. Washington, DC. VALLE-LEVINSON, A. 2013. Some basic hydrodynamic concepts to be considered for coastal aquaculture. In: ROSS, L. G., TELFER, T. C., FALCONER, L., SOTO, D. & AGUILAR-MANJARREZ, J. (eds.) Site selection and carrying capacities for inland and coastal aquaculture, pp. 147–158. Rome: FAO/Institute of Aquaculture, University of Stirling. VERDEGEM, M. C. 2013. Nutrient discharge from aquaculture operations in function of system design and production environment. Reviews in Aquaculture, 5, 158-171. VILLANUEVA, R. D., YAP, H. T., MONTA, XF & O, M. N. E. 2006. Intensive fish farming in the Philippines is detrimental to the reef-building coral Pocillopora damicornis. Marine Ecology Progress Series, 316, 165-174. VINCENZI, S., CARAMORI, G., ROSSI, R. & DE LEO, G. A. 2006. A GIS- based habitat suitability model for commercial yield estimation of Tapes philippinarum in a Mediterranean coastal lagoon (Sacca di Goro, Italy). Ecol. Model., 193, 90-104. VOLPE, J. P., TAYLOR, E. B., RIMMER, D. W. & GLICKMAN, B. W. 2000. Evidence of Natural Reproduction of Aquaculture-Escaped Atlantic Salmon in a Coastal British Columbia River. Conservation Biology, 14, 899-903. WALMSLEY, S. & WHITE, A. 2003. Influence of social, management and enforcement factors on the long-term ecological effects of marine sanctuaries. Environmental Conservation, 30, 388-407. WALPOLE, M. J. & GOODWIN, H. J. 2001. Local attitudes towards conservation and tourism around Komodo National Park, Indonesia. Envir. Conserv., 28, 160-166. WALPOLE, M. J., GOODWIN, H. J. & WARD, K. G. R. 2001. Pricing Policy for Tourism in Protected Areas: Lessons from Komodo National Park, Indonesia. Conservation Biology, 15, 218-227. WALTON, A., WHITE, A. T., TIGHE, S., ALIÑO, P. M., LAROYA, L., DERMAWAN, A., KASASIAH, A., HAMID, S. A., VAVE-KARAMUI, A., GENIA, V., DE JESUS MARTINS, L. & GREEN, A. L. 2014. Establishing a Functional Region- Wide Coral Triangle Marine Protected Area System. Coastal Management, 42, 107- 127. WANG, S., TANG, D., HE, F., FUKUYO, Y. & AZANZA, R. V. 2008. Occurrences of harmful algal blooms (HABs) associated with ocean environments in the South China Sea. Hydrobiologia, 596, 79-93. WATTS, M. E., BALL, I. R., STEWART, R. S., KLEIN, C. J., WILSON, K., STEINBACK, C., LOURIVAL, R., KIRCHER, L. & POSSINGHAM, H. P. 2009. Marxan with Zones: Software for optimal conservation based land- and sea-use zoning. Environ. Modell. Softw., 24, 1513-1521. WATY, T. E. 2011. Evaluasi program nasional pemberdayaan masyarakat mandiri kelautan perikanan terhadap pemberdayaan masyarakat pesisir (studi kasus kecamatan Awangpone Kabupaten Bone Sulawesi selatan). Master, Hasanuddin University. 201

WEVER, L., GLASER, M., GORRIS, P. & FERROL-SCHULTE, D. 2012. Decentralization and participation in integrated coastal management: Policy lessons from Brazil and Indonesia. Ocean & Coastal Management, 66, 63-72. WGMFDA-MMAF 2011. Marine and Fisheries in Figures 2011 (Kelautan dan Perikanan dalam Angka 2011), Jakarta, Centre of Data Statistics and Information, Ministry of Marine Affairs and Fisheries. WHITE, A. T., ALIÑO, P. M., CROS, A., FATAN, N. A., GREEN, A. L., TEOH, S. J., LAROYA, L., PETERSON, N., TAN, S., TIGHE, S., VENEGAS-LI, R., WALTON, A. & WEN, W. 2014. Marine Protected Areas in the Coral Triangle: Progress, Issues, and Options. Coastal Management, 42, 87-106. WHITE, A. T. & GREEN, A. L. 2014. Introduction. Coastal Management, 42, 81-86. WHITE, P., PHILLIPS, M. & BEVERIDGE, M. C. M. 2010. Environmental impact, site selection and carrying capacity estimation for small-scale aquaculture in Asia. In: ROSS, L. G., TELFER, T. C., FALCONER, L., SOTO, D. & AGUILAR- MANJARREZ, J. (eds.) Site selection and carrying capacities for inland and coastal aquaculture. Rome: FAO Fisheries and Aquaculture WHITMARSH, D. & PALMIERI, M. G. 2009. Social acceptability of marine aquaculture: The use of survey-based methods for eliciting public and stakeholder preferences. Marine Policy, 33, 452–457. WIADNYA, D. G. R. 2012. Roadmap Kawasan Konservasi Perairan di Indonesia. [Accessed 16 December 2015]. WIADNYA, D. G. R., SYAFAAT, R., SUSILO, E., SETYOHADI, D., ARIFIN, Z. & WIRYAWAN, B. 2011. Recent Development of Marine Protected Areas (MPAs) in Indonesia: Policies and Governance. Journal of Applied Environmental and Biological Sciences, 1, 608-613. WILY, L. A. 2011. ‘The Law is to Blame’: The Vulnerable Status of Common Property Rights in Sub‐Saharan Africa. Development and change, 42, 733-757. WINDUPRANATA, W. 2007. Development of a Decision Support System for Suitability Assessment of Mariculture Site Selection. Ph.D. Dissertasion, Christian-Albrechts Universit t zu Kiel. WIRYONO. 2003. Klasifikasi Kawasan Konservasi Indonesia. Warta Kebijakan [Online], 11. [Accessed 16 December 2015]. WYRTKI, K. 1961. Physical Oceanography of the Southeast AsianWaters. NAGA REPORT. La Jolla, California: The University of California, Scripps Institution of Oceanography. YUSUF, M. 2007. Kebijakan pengelolaan sumberdaya pesisir dan laut kawasan taman nasional karimunjawa secara berkelanjutan. ZHANG, J., HANSEN, P. K., FANG, J., WANG, W. & JIANG, Z. 2009. Assessment of the local environmental impact of intensive marine shellfish and seaweed farming— Application of the MOM system in the Sungo Bay, China. Aquaculture, 287, 304- 310.

202

APPENDIXES

Appendix 1. Semi-structured questionnaire for sustainability, vulnerability of small-scale households in Anambas Archipelago MPA and their perspective to the MPA establishment

INTERVIEW QUESTIONS (Your participation in this survey will be kept strictly confidential and no stored and published information can be used to identify the individual source of the information)

A. General Information Q.1 Date : Q.2 Location : District :…………………………. Sub-district :…………………………. Village :…………………………. Q.3 What is your gender? Male Female Q.4 Please specify your age? 18 – 24 45 - 54 25 – 34 55 - older 35 - 44 Q.5 Please indicate your educational achievement? Never attending formal school Primary school (completing or no completing, please mark) Junior high school (completing or no completing, please mark) Senior high school (completing or no completing, please mark) Graduate diploma (completing or no completing, please mark) Bachelor degree (completing or no completing, please mark) Post graduate Master PhD Q.6 What is your role in your family? Head of the family Household wife Immediate family member…………………… Q.7 What is your main occupation in relation to Anambas Marine Protected Area (MPA)? Fish farmer Tourist operator/businessman

203

Fisherman NGO, government officer, or other, please specify…………………………. Q.8 Besides your main occupation, do you have other alternative source of income?...... Q.9 Do you have other immediate family members helping you to support the economy of your family?...... if yes, could you please explain? …………………………………………...... ………………………………………………………………………………………………… ………………………………………………………………………………………………… Q.10 Have you ever considered to shift from your current job to another one? ………, if yes, can you explain why? ……………………………………………………………………………... ………………………………………………………………………………………………… …………………………………………………………………………………………………

B. Stakeholder’s perspective about Anambas MPA Q.11 Do you know about the recent designation of Anambas archipelago as an MPA?...... If yes, could please explain what do you know about it?………………………………………… ………………………………………………………………………………………………… ………………………………………………………………………………………………… Q.12 Have you or others that you might know, ever been consulted or notified regarding the MPA and its zonation?...... , if yes, could you please elaborate ………………………… ………………………………………………………………………………………………… ………………………………………………………………………………………………… Q.13 If you HAVE BEEN consulted or notified, please indicate your agreement with the designation of the MPA and its zonation. Agree Neutral Do not agree Why?...... Q.14 If you HAVE NOT BEEN consulted or notified, please indicate your agreement with the designation of the MPA and its zonation. Agree Neutral Do not agree Why?......

Q.15 In relation to your current job or knowledge, do you think the MPA will benefit you and the community as a whole? Strongly Agree Agree Disagree Do not know

204

Except “do not know”, please explain why? ……………………………………………………. ………………………………………………………………………………………………………… …………………………………………………………………………………………………………

205

C. Ecotourism Sustainability Q.C.1 How long have you been working in ecotourism?

1-4 years 5-9 years 10-14 years more than 15 yrs Q.C.2 What aspect or service of ecotourism you are working for? ………………………………… Q.C.3 How many working hours do you spend daily?...... Q.C.4 How many working hours do you spend each week for the other jobs if any? ……(hours) Q.C.5 Could you please indicate the range of your monthly earning from your main job? Less than Rp. 1.000.000 Rp. 4.000.000 – Rp. 6.000.000 Rp. 1.000.000 – Rp. 2.000.000 Rp. 6.000.000 – Rp. 10.000.000 Rp. 2.000.000 – Rp. 4.000.000 Rp. 10.000.000 – more Q.C.6 Could you please indicate the range of your monthly earning from your alternate job? Less than Rp. 1.000.000 Rp. 4.000.000 – Rp. 6.000.000 Rp. 1.000.000 – Rp. 2.000.000 Rp. 6.000.000 – Rp. 10.000.000 Rp. 2.000.000 – Rp. 4.000.000 Rp. 10.000.000 – more Q.C.7 Could you please indicate the range of your current savings Less than Rp. 1.000.000 Rp. 4.000.000 – Rp. 6.000.000 Rp. 1.000.000 – Rp. 2.000.000 Rp. 6.000.000 – Rp. 10.000.000 Rp. 2.000.000 – Rp. 4.000.000 Rp. 10.000.000 – more Q.C.8 Please indicate the stability of your earning for one-year period? Very Stable (11 - 12 months stable income) Stable enough (9 – 10 months stable income) Stable (7 – 8 months stable income) Somewhat stable (5 – 6 months stable income) Not stable (1 – 4 months stable income) Q.C.9 Have you or your family members received any supports from the local/national government/NGO which, in any way, relate to the designation of Anambas region as an MPA? a. Training ……………. times b. Funding scheme ……… times c. Scholarship …………… times e. Equipment …………… times f. Other, …………………………… ……….. times Q.C.10 Have you or your immediate family members ever apply for credit from bank or other financial institution? ……….., if yes, what is your impression on the process of getting

206

it?……………………………………………………...... Q.C.11 Were there any studies or activities conducted by institutions that specifically address the relationship between your type of livelihood and Anambas MPA?...... , if yes, could you please elaborate………………………………………………………………………...... Q.C.12 Are there any specific national or local regulations/traditional laws/community agreements that intersect with your daily livelihood activities?...... , if yes.. can you please explain: What kind ………………………………………………………………………………….. Who issued …………………….…………………………………………………………… How long has it been enforced………………………………..…………………………….. Do you agree or disagree or do not know, with the regulation? Except do not know, why? ……………………………………………..…………………. ……………………………………………………………………………………………… ……………………………………………………………………………………………… Q.C.13 Have you experienced any clash/conflict/disagreement with other community members or groups regarding the use or access of coastal space and natural resources? ………… If yes, can you please elaborate? ……………………………………………………………… ……………………………………………………………………………………………… …………...…………………………………………………………………………………… Q.C.14 In doing your daily livelihood, are you working : Independently as a group as a network of groups Q.C.15 What are the levels of education of your immediate family members? Husband : ……………….. Wife : …………………….. Children : …………………. Q.C.16 In your opinion, do you think that your daily livelihood activity is affecting the MPA environment? ……., can you please describe in what ways………………………………... ……………………………………………………………………………………………… ……………………………………………………………………………………………… ………………………...………………………………………………………………………

207

Q.C.17 Regarding other livelihoods (fishing and fish farming), in your opinion, what are their effects to the MPA environment?...... ……………………………………………………………………………………………… ………………………………………………………………………………………………

Q.C.18 From the list below, please indicate and rate the factors that you consider as risks to your livelihood from 1 (most important) to 5 (least important). Vulnerability Factors 1 2 3 4 5 A. Cycles Climate season Seasonal weather condition Seasonality of market Other,…………………………………………… Other,…………………………………………… B. Trends Increasing food price Increasing cost of fuel Increasing price of equipment parts Increasing cost of health services Increasing water pollution Decreasing coverage of coral reef and other rsc. Climate change and sea level rise Other,…………………………………………… Other,…………………………………………… C. Shocks Storm and bad weather occurrence Change in the local political situation New regulation regarding access to the MPA areas Conflicts regarding access to the MPA areas Other,…………………………………………… Other,……………………………………………

D. Small-scale Fisheries Q.D.1. How long have you been working in small scale fisheries?

1-4 years 5-9 years 10-14 years more than 15 years Q.D.2. What aspect or service of small scale fisheries you are working for? ………………………………… Q.D.3. How many working hours do you spend daily?...... Q.D.4. How many working hours do you spend each week for the other jobs if any? ……(hours) Q.D.5. Could you please indicate the range of your monthly earning from your main job ? Less than Rp. 1.000.000 Rp. 4.000.000 – Rp. 6.000.000

208

Rp. 1.000.000 – Rp. 2.000.000 Rp. 6.000.000 – Rp. 10.000.000 Rp. 2.000.000 – Rp. 4.000.000 Rp. 10.000.000 – more Q.D.6. Could you please indicate the range of your monthly earning from your alternate job ? Less than Rp. 1.000.000 Rp. 4.000.000 – Rp. 6.000.000 Rp. 1.000.000 – Rp. 2.000.000 Rp. 6.000.000 – Rp. 10.000.000 Rp. 2.000.000 – Rp. 4.000.000 Rp. 10.000.000 – more Q.D.7. Could you please indicate the range of your current savings Less than Rp. 1.000.000 Rp. 4.000.000 – Rp. 6.000.000 Rp. 1.000.000 – Rp. 2.000.000 Rp. 6.000.000 – Rp. 10.000.000 Rp. 2.000.000 – Rp. 4.000.000 Rp. 10.000.000 – more

Q.D.8. Please indicate the stability of your earning for one year period? Very Stable (11 - 12 months stable income) Stable enough (9 – 10 months stable income) Stable (7 – 8 months stable income) Somewhat stable (5 – 6 months stable income) Not stable (1 – 4 months stable income) Q.D.9. Have you or your family members received any supports from the local/national government/NGO which, in any way, relate to the designation of Anambas region as an MPA?

a. Training ……………………………………………….……………………. b. Funding scheme ………………………………………………………………… c. Scholarship ……………………………………………………………………… e. Equipment ………………………………………………………………………. f. Other, …………………………………………..………………………………… Q.D.10. Have you or your immediate family members ever apply for credit from bank or other financial institution? …………….., if yes, what is your impression on the process of getting it?……………………………………………………...... Q.D.11. Were there any studies or activities conducted by institutions that specifically address the relationship between your type of livelihood and Anambas MPA?...... , if yes, could you please elaborate………………………………………………………......

209

Q.D.12. Are there any specific national or local regulations/traditional laws/community agreements that intersect with your daily livelihood activities?...... , if yes.. can you please explain: What kind ……………………………………………….…………………… Who issued ……………………………………………………………….….. How long has it been enforced ………………………………………………. Do you agree or disagree or do not know, with the regulation?..... Except do not know, why? ………………………………………………………….………. ………………………………………………………………………………………..……… ……………………………………………………………………………………………….. ……………………………………………………………………………………………….. Q.D.13. Have you experienced any clash/conflict/disagreement with other community members or groups regarding the use or access of coastal space and natural resources? ……………. If yes, can you please elaborate? …………………………… ………………………………………………………………………………………… ……..…………...……………………………………………………………………… Q.D.14. In doing your daily livelihood, are you working : Independently as a group as a network of groups Q.D.15. What are the levels of education of your immediate family members? Husband : …………………….. Wife : …………………….. Children : …………………….. Q.D.16. In your opinion, do you think that your daily livelihood activity is affecting the MPA environment? ……., can you please describe in what ways…………………………. ………………………………………………………………………………………… ………………………………………………………………………………………… ……………………...…………………………………………………………………. Q.D.17. Regarding other livelihoods, eco-tourism and fish farming, in your opinion, what are their effects to the MPA environment?...... ………………………………………………………………………………………… ………………………………………………………………………………………… ………………………………………………………………………………………… Q.D.18. From the list below, please indicate and rate the factors that you consider as risks to your livelihood from 1 (most important) to 5 (least important). Vulnerability Factors 1 2 3 4 5 D. Cycles Climate season Seasonal weather condition Seasonality of market E. Trends Increasing food price 210

Increasing cost of fuel Increasing price of equipment parts Increasing cost of health services Increasing water pollution Decreasing coverage of coral reef and other rsc. Climate change and sea level rise F. Shocks Storm and bad weather occurrence Change in the local political situation New regulation regarding access to the MPA areas Conflicts regarding access to the MPA areas

211

E. Small-scale Mariculture Q.E.1. How long have you been working in fish farming?

1-4 years 5-9 years 10-14 years more than 15 years Q.E.2. What aspect or service of fish farming you are working for? ………………………………… Q.E.3. How many working hours do you spend daily?...... Q.E.4. How many working hours do you spend each week for the other jobs if any? ………….(hours) Q.E.5. Could you please indicate the range of your monthly earning from your main job Less than Rp. 1.000.000 Rp. 4.000.000 – Rp. 6.000.000 Rp. 1.000.000 – Rp. 2.000.000 Rp. 6.000.000 – Rp. 10.000.000 Rp. 2.000.000 – Rp. 4.000.000 Rp. 10.000.000 – more Q.E.6. Could you please indicate the range of your monthly earning from your alternate job? Less than Rp. 1.000.000 Rp. 4.000.000 – Rp. 6.000.000 Rp. 1.000.000 – Rp. 2.000.000 Rp. 6.000.000 – Rp. 10.000.000 Rp. 2.000.000 – Rp. 4.000.000 Rp. 10.000.000 – more Q.E.7. Could you please indicate the range of your current savings? Less than Rp. 1.000.000 Rp. 4.000.000 – Rp. 6.000.000 Rp. 1.000.000 – Rp. 2.000.000 Rp. 6.000.000 – Rp. 10.000.000 Rp. 2.000.000 – Rp. 4.000.000 Rp. 10.000.000 – more

Q.E.8. Please indicate the stability of your earning for one-year period? Very Stable (11 - 12 months stable income) Stable enough (9 – 10 months stable income) Stable (7 – 8 months stable income) Somewhat stable (5 – 6 months stable income) Not stable (1 – 4 months stable income) Q.E.9. Have you or your family members received any supports from the local/national government/NGO which, in any way, relate to the designation of Anambas region as an MPA?

a. Training ……………. times b. Funding scheme ……… times c. Scholarship …………… times e. Equipment …………… times f. Other, ……………, ……….. times Q.E.10. Have you or your immediate family members ever apply for credit from bank or other financial institution? …………….., if yes, what is your impression on the 212

process of getting it?…………………….……………………………...... Q.E.11. Were there any studies or activities conducted by institutions that specifically address the relationship between your type of livelihood and Anambas MPA?...... , if yes, could you please elaborate……………...... Q.E.12. Are there any specific national or local regulations/traditional laws/community agreements that intersect with your daily livelihood activities?...... , if yes.. can you please explain: What kind …………………………………………………………………………… Who issued ………..……………………………………………………………….. How long has it been enforced ………….………………………………………….. Do you agree or disagree or do not know, with the regulation Except do not know, why? ………………………………………………………………………………………………. ………………………………………………………………………………………………. Q.E.13. Have you experienced any clash/conflict/disagreement with other community members or groups regarding the use or access of coastal space and natural resources? ……………. If yes, can you please elaborate? …………………………. ………………………………………………………………………………………… ………………...……………………………………………………………………… Q.E.14. In doing your daily livelihood, are you working : Independently as a group as a network of groups Q.E.15. What are the levels of education of your immediate family members? Husband : ……………….. Wife : …………………….. Children : …………………. Q.E.16. In your opinion, do you think that your daily livelihood activity is affecting the MPA environment? ……., can you please describe in what ways…………………………. ………………………………………………………………………………………… ………………………………………………………………………………………… Q.E.17. Regarding other livelihoods fishing and ecotourism, in your opinion, what are their effects to the MPA environment?...... ………………………………………………………………………………………… …………………………………………………………………………………………

213

Q.E.18. From the list below, please indicate and rate the factors that you consider as risks to your livelihood from 1 (most important) to 5 (least important). Vulnerability Factors 1 2 3 4 5 G. Cycles Climate season Seasonal weather condition Seasonality of market H. Trends Increasing food price Increasing cost of fuel Increasing price of equipment parts Increasing cost of health services Increasing water pollution Decreasing coverage of coral reef and other rsc. Climate change and sea level rise I. Shocks Storm and bad weather occurrence Change in the local political situation New regulation regarding access to the MPA areas Conflicts regarding access to the MPA areas

214

Appendix 2. Semi-Structured Questionnaire for Stakeholder Preference Sub-Model in Site Selection for Small-Scale Mariculture in Anambas Archipelago MPA

INTERVIEW QUESTIONS (Your participation and any personal information collected and published from this survey are treated confidentially and would not be able to be used to identify your involvement in this study)

A. General Information Q.1 Date : Q.2 Location : Sub-district :…………………………. Village :…………………………. Q.3 Gender? Male Female Q.4 Age? 18 – 24 45 - 54 25 – 34 Above 55 years old 35 - 44 Q.5 Marital status in the family? Husband Wife Dependent B. Fish culture practices Q.6. Describe the fish species that were or are cultured Farming cage type (fixed, Fish Species Reason to use the species floating)

Q.7. Type and net cage specification that were or are used Type of Netpen size Netpen mesh Floating Net cage structure net cage and number size (mm) device material Size: For seed: Number: For grow out:

215

Q.8. How did/do you access your net cage? Distance from Vehicle Time needed Access frequency Time spent on the net cage home (m) used (min) (times/day) (hrs)

C. Stakeholder site suitability preference

C.1. Distance preference Q.9. Choose your distance preferences of your net cage location to a specific place/feature in the MPA where #1 is the most preferred and #3 is the least preferred …. 9.1. Distance from the net cage location to your village?

0 – 500 m ç 500 – 1 km ç Above 1 km ç Please give your reasoningÇ of your preferences………………………………………………....Ç Ç √ √ √ ……………………………………………………………………………………………………

……………………………………………………………………………………………………

9.2. Distance from the net cage location to the net cages belonged to other fish farmers

0 – 500 m ç 500 – 1 km ç Above 1 km ç Please give your reasoningÇ of your preferences………………………………………………....Ç Ç √ √ √ …………………………………………………………………………………………………… …………………………………………………………………………………………………… 9.3. Distance from the net cage location to middlemen location 0 – 1 km 1 km – 5 km Above 5 km ç ç ç Please give your reasoningÇ of your preferencesÇ ………………………………………...... Ç .. √ √ √ …………………………………………………………………………………………………… …………………………………………………………………………………………………… 9.4. Distance from the net cage location to the nearest harbor/jetty in your village 0 – 1 km 1 km – 5 km Above 5 km ç ç Please give your Çreasoning of your preferencesÇ ………………………………………………....ç √ √ ……………………………………………………………………………………………………Ç √ …………………………………………………………………………………………………… 9.5. Distance from the net cage location to coral reed coverage 0 – 500 m 500 m – 1 km Above 1 km ç ç ç Ç Ç Ç √ √ √ 216

Please give your reasoning of your preferences ………………………………………………....

…………………………………………………………………………………………………… …………………………………………………………………………………………………… 9.6. Distance from the net cage location to potential seed source location 0 – 1 km 1 km – 5 km Above 5 km c ç ç Please give your reasoningç of your preferences………………………………………………....Ç Ç Ç √ √ ……………………………………………………………………………………………………√ ……………………………………………………………………………………………………

Q.10. Please give your preference of net cage type and preferred coastal geographical location for the net cage 10.1. Please indicate your preferred net cage type

c a. Fixed net cage, describe briefly why………………...………………..……………. ç b. Floating net cage, describe briefly why ……………………………..……………... c Ç √ç 10.2.Ç Please give your preference regarding the best geographical location for net cage in your √ area with #1 as the most preferred and #4 as the least preferred Bay c Strait (between islands/islets) çc Çç Coastline (open) √c çÇ Coastline (protected) c√ çÇ Ç√ √

217

Appendix 3. Estimated and Measured MOM Input Parameters

Value No. Estimated and observed parameters Eq. symbol Unit Sources Trash fish Comm. feed A. Feed Energy Content 1 Protein energy content Cp kcal/kg 5,650 Parsons et al. (1979), Stigebrandt 2 Fat energy content Cf kcal/kg 9,450 et al.(1999), Usman et al (2016), 3 Carbohydrate energy content Cc kcal/kg 4,100 Marzuki et al. (2013), Tacon et al. 4 Protein fraction in feed Fp % 0.300 0.430 (1991), John W. Tucker Jr. (2012) 5 Fat fraction in feed Ff % 0.090 0.090 6 Carbohydrate fraction in feed) Fc % 0.220 0.240 7 Total feed energy content δ kcal/kg 3448 4264 8 specific protein energy content in feed Ep kcal/kg 0.492 0.570 9 specific fat energy content in feed Ef kcal/kg 0.247 0.199 10 specific carbohydrate energy content in feed Ec kcal/kg 0.262 0.231 B. Grouper fish energy content 11 Protein fraction in fish Pp % 0.192 0.181 Tacon et al (1991) and Tucker 12 Fat fraction in fish Pf % 0.0622 0.0648 (2012), Stigebrandt (1999) 13 Specific energy of grouper Cfi kcal/kg 1,672.59 1,635.01 14 Assimilated fraction of protein Ap % 0.890 Stigebrandt et al. (1999) 15 Assimilated fraction of fat Af % 0.920 Stigebrandt et al. (1999) 16 Assimilated fraction of carbohydrate Ac % 0.500 Stigebrandt et al. (1999) C. Net cage Construction in the MPA 14 Depth of the pen D m 3 Fieldwork, 2016 15 Number of rows R 2 Fieldwork, 2016 16 Side length of the pen (square) L m 4 Fieldwork, 2016 17 Number of pens Nf unit 8 Fieldwork, 2016 18 Separation between the pens S m 2 Fieldwork, 2016 20 Total area of farm (Af=Nf * L^2) Af m2 16 Fieldwork, 2016 21 The length of the farm ((Nf/R)*(L+S) - S ) Lf m 5.333 Fieldwork, 2016 22 The width of the farm (R*(L+S)-S ) Wf m 22 Fieldwork, 2016 24 Average harvested weight W kg 0.7 Fieldwork, 2016 25 Culture period day 210 Fieldwork, 2016, Ismi et al. (2013) D. Fish Sub Model 26 Assimilated protein factor 1.650 1.480 Tacon et al. (1991) 27 Protein consumption per fish to reach 0.7 kg kg 0.317 0.268 Calculated from Eq. 28 Total feed required (P consump/Protein % Fish) kg 1.056 0.623 Calculated from Eq. 29 Theoretical Feed Convertion Ratio FCRt 1.509 0.890 Calculated from Eq. 30 Feed Convertion Ratio FCR 3.530 1.440 Calculated from Eq. 31 Excess feed (FCR - FCRt) per kg fish FCR - FCRt kg 1.415 0.385 Calculated from Eq. 32 Total excess feed of the farm (Tp*ind excess feed) kg 1415 385.023 Calculated from Eq. 33 The spatial and temporal excess feed from the farm F1feed kg/m2/day 0.053 0.014 Calculated from Eq. 34 The spatial and temporal of feaces from the farm F1faeces kg/m2/day 0.001 0.001 Calculated from Eq. 35 Constant 1 (non-dimensional) α 11 Stigebrandt et al. (1999) 36 Constant 2 (non-dimensional) g 0.8 Stigebrandt et al. (1999) 37 The average water temperature at the location T C 27 Fieldwork, 2016 38 The inverse value of T (e.g. 1/T) t 0.08 Stigebrandt et al. (1999) Estimated (for stable tropical 39 The factor of appetie and growth e 1 condition)

218