UNEP Report of the Meeting of Experts on the Environmental Vulnerability Index

UNEP

REPORT OF THE MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

Valletta, Malta 29 November - 3 December 1999

Organised by the

Islands and Small States Institute, Foundation for International Studies, University of Malta

in collaboration with

South Pacific Applied Geoscience Commission, Fiji

Edited by: LINO BRIGUGLIO Islands and Small States Institute, Foundation for International Studies, University of Malta URSULA KALY Environmental Adviser, Tuvalu Government, Tuvalu CRAIG PRATT South Pacific Applied Geoscience Commission, Fiji South Pacific Applied Geoscience Commission (SOPAC) 08/02/99

PROPOSAL

The Refinement of the Environmental Vulnerability Index (EVI)

1. Project Aims:

· To develop a fully functional and tested environmental vulnerability index that has been critically reviewed and refined by international experts · To build capacity within the South Pacific region to identify, analyse and collect data for the environmental vulnerability index · To develop a simple robust computer user interface for the environmental vulnerability index model to facilitate data entry and streamline EVI computation

2. Introduction

The first phase of the environmental vulnerability index project funded by the New Zealand Government and recently completed by SOPAC, achieved its primary goal of developing a conceptual framework for a workable environmental vulnerability index. The EVI calculator was also tested with limited data on 3 countries which clearly showed its potential as an operational and useful gauge for environmental vulnerability.

The environmental vulnerability index now needs to be comprehensively tested both mathematically as well as with comprehensive real data to assess its capacity to differentiate between the environmental vulnerability of several countries throughout the region with a view to extending the EVI model to the world. The environmental vulnerability index also needs to undergo critical review by international experts together with consultation with regional representatives and agencies to ensure that it is developed into an internationally acceptable and workable EVI.

Seven major tasks have been identified for the successful full development of the EVI as a tool for measuring and managing the vulnerability of countries. These are presented in a module format so that either the entire package or several different modules or parts thereof may be taken up by funding agencies. There are 4 basic modules which include: 1. Continuation, Review and Profile Raising of the EVI 2. Capacity-building and EVI database 3. EVI Model/Indicator Refinement 4. Development of EVI User Interface

1 South Pacific Applied Geoscience Commission (SOPAC) 08/02/99

The modules are currently presented only in concept with estimated budget and timeframe. Detailed proposals for each module will be provided as required. Some of the tasks within modules require others as prerequisites for their successful completion and these are highlighted below.

3. Modules required for developing the index and their justification

Module 1: Continuation, Review and Profile Raising of the EVI

Module 1.1 Continuity of EVI Work – Bridging Finance

Aim: To ensure continuity of the EVI work developed in the first phase of the project and provide expert briefing to the regional, internationalfora and New Zealand government on the outcomes of phase I.

To ensure continuity of the work developed in the first phase of the project requires the immediate provision of resources to secure a minimum SOPAC EVI study team with sufficient funding to carry through the outcomes of phase I and continue preparation for phase II of the project. This team would consist of one staff and one assistant.

Bridging finance will be needed to maintain the momentum generated by phase I and to allow presentation of Phase I results to SIDS and the international fora to maintain its high profile. This will help to raise awareness of the important research being carried out by the region and SOPAC on vulnerability indices and help in forging links with other research agencies. It will also help with the identification of support for funding for the next phase and allow the continuation of communications and collection of detailed information on the three case study countries used in Phase I.

To facilitate the special technical briefing of SIDs on the environmental vulnerability index, the SOPAC EVI team (two staff) will need funding assistance to attend the preparatory meetings for CSD7 in New York from the 1 – 5 March, 1999.

It would also be highly beneficial for a special briefing to be held between the SOPAC EVI study team (two staff) with the New Zealand Environment Minister – the Honourable Simon Upton and appropriate agencies to discuss the environmental vulnerability index and outcomes of Phase I.

All comments on the Final Report of Phase I will be collated and examined for possible usefulness in the refining of EVI indicator questions.

2 South Pacific Applied Geoscience Commission (SOPAC) 08/02/99

Outputs/Benefits:

· Maintain momentum generated in the international arena surrounding the environmental vulnerability index through the presentation of outcomes of Phase I to SIDS and the international community at CSD · Raise profile and further consideration of the environmental vulnerability index as a technically feasible measure of vulnerability through expert briefings by the SOPAC EVI study team · Raise profile of EVI at conferences and other regional and internationalfora · Develop funding proposals and source possible funding agencies · Start refinements of questions based on responses and comments from the Final Report of Phase I · Compile information for an EVI database of country information in particular the 3 case study countries used in Phase I · Forge links with other study groups · Brief the Honourable Simon Upton and NZ government agencies on the EVI and outcomes of Phase I

Budget/Timeframe:

· FJ$84,275 - 6 months

Module 1.2 Put the EVI to an International Environmental Expert “Think Tank”

Aim: To obtain international expert input to critique, refine and consolidate the environmental vulnerability index model.

The SOPAC EVI Team will host a small conference of up to 12 highly-specialised international experts in the fields of: 1. Statistical methods, particularly multivariate systems 2. Biodiversity / biogeography 3. Other types of indices that summarise very complex data 4. Weather / climate 5. Disaster research 6. Ecosystem management 7. Fisheries 8. Forestry and agriculture 9. Productivity and the flow of energy in and out of ecosystems

3 South Pacific Applied Geoscience Commission (SOPAC) 08/02/99

The purpose of this "think tank" is to expose the EVI model and its components to intense highly-technical discussion for up to 5 days. The “think tank” is considered the most cost effective way to obtain expert consideration of the EVI. We specifically envisage inputs to:

· The methods used to calculate the index, whether there may be better mathematical approaches; · The questions which form the heart of the index and "measure" vulnerability; · The response levels set for each question, and whether they should be linear or non- linear and the maxima and minima used; · Whether there are other types of data available, or which will be available soon that should be incorporated; · Identify any biases, strengths and weaknesses in the model that were not identified during Phase I.

Working together with the SOPAC EVI Team, these experts can be harnessed to ensure that the final EVI is technically excellent, will be able to properly carry out the functions required and will have the support of some of the most important scientists in their respective disciplines.

Outputs/Benefits:

· Launches the EVI into the international technical arena and provides immediate critical feedback on the strengths and weaknesses of the EVI from several of the most important scientists in environmental research · Critical refinement/development input and problem solving of the environmental vulnerability index · Publicity and profile raising of the EVI · A critiqued and workable EVI that is ratified by experts in the scientific community · Conference report

Budget/Timeframe:

· FJ$58,853 - 5 days

4 South Pacific Applied Geoscience Commission (SOPAC) 08/02/99

Module 1.3 Publication of the EVI in International Journal

Aim: To obtain international scientific peer review of the EVI work and ensure intellectual copyright

To ensure that intellectual copyright of the EVI developed by the EVI Team and its funding agent (NZODA) is established, we propose to publish the results of Phase I in an international journal. It will be necessary to fully describe the approach taken, questions used and their scoring levels, and mathematical techniques so that the EVI developed becomes a technically recognised entity. This ensures that refinements are tracked and that changes to the index by others who might take up the task are fully identified. It also ensures that the process of peer- review required for excellence of the model can be an on-going process within the international community.

SOPAC will also continue to publicise the environmental vulnerability index development work being undertaken through coverage in its own publications and exposure at variousfora.

Outputs/Benefits:

· Establish the EVI as an intellectual entity · Published intellectual ownership of the environmental vulnerability index model · Comments and constructive feedback on the EVI · Raises the EVI profile

Budget/Timeframe:

· FJ$18,645 - 30 days

Module 1 Outputs/Benefits:

· Raising of the profile of EVI · Better understanding of the EVI by New Zealand, SIDS and other countries · Publication of EVI in a recognised academic journal · Critiqued and workable EVI endorsed by scientific experts · Refinement of indicators and collation of comments/feedback on EVI · Detailed collection of EVI data on the 3 case study countries used in Phase I · Improved links with other vulnerability indices and environmental indicator researchers and institutions · Important component in the overall outputs and benefits of the project

5 South Pacific Applied Geoscience Commission (SOPAC) 08/02/99

Module 2:Capacity-building and database

Module 2.1 Environmental vulnerability index capacity-building workshop

Aim: To provide training to SIDS country representatives in the use, data collection and computation of the environmental vulnerability index.

For some countries, data may be obtainable by contacting the relevant departments and asking for assistance with data acquisition. For most SIDS, this approach is unlikely to yield adequate results. It will therefore be necessary for the EVI Team to carry out a capacity building workshop with at least 10 member country representatives to introduce them to the environmental vulnerability index concept, its workings and information required for its computation.

The workshop will be held over 5 days at the SOPAC Secretariat which is located in Suva, Fiji and is centrally situated in the South Pacific. The SOPAC Secretariat has all the necessary computer, training and workshop facilities required to facilitate such a workshop thus reducing the need to secure outside facilities at additional cost.

The workshop participants would be provided with an introduction to the environmental vulnerability index, the underlying concepts and benefits of the model and an introduction into the mechanics and computation of the EVI. The aim of the introduction would be to provide participants with the basic skills needed to determine information needs for the EVI together with the identification of possible sources for information in their respective countries and how to collect appropriate data. The knowledge gained through the workshop would allow participants to provide their countries with detailed explanations as to the need for information, the conceptual mechanics and overall benefits of the environmental vulnerability index.

Outputs/Benefits:

Budget/Timeframe:

· FJ$146,997 - 5 days

6 South Pacific Applied Geoscience Commission (SOPAC) 08/02/99

Module 2.2 Follow-up consultation and database

Aim: To compile data collected by country participants and provide in country assistance in the identification and compilation of all necessary environmental data for environmental vulnerability index calculation

The participants would be expected on their return home to collect the necessary data to compute the environmental vulnerability index. It is envisaged that the SOPAC EVI team would need to travel to all the participating countries to assist participants in the identification and collection of all data necessary for the testing of the EVI model so that real cases are included.

In so doing, the countries themselves will benefit because they will be shown the types of data that should be collected for environmental management purposes and methods for doing so will be taught to the relevant officers in each. Because it is envisaged that the EVI should be recalculated every 5 years, this capacity-building module should ensure that participating countries will be able to submit their own updates on the vulnerability status after the initial visit by the EVI Team.

Outputs/Benefits:

· Collection and compilation of each countries environmental information · Strengthening of capacity of countries to identify, collect and compile appropriate environmental information needed for the environmental vulnerability index · Improved data collection and handling · Trained personnel that understand and are able to use the environmental vulnerability index · Pacific ownership of the environmental vulnerability index · Generation of EVI values for all participating countries

Budget/Timeframe:

· FJ$89,089 - 4 months

Module 2 Outputs/Benefits: · Trained country representatives that understand the usefulness and mechanics of the environmental vulnerability index · To strengthen in country data collection and assimilation · To identify problem areas in country in the collection of data, its quality and analysis

7 South Pacific Applied Geoscience Commission (SOPAC) 08/02/99

· Collection of the most recent and best available raw data to enter into the EVI database from participating island countries for testing in the model · Generation of EVI statistics for several countries · Important component in the overall outputs and benefits of the project · Pacific ownership

Module 2 Budget/Timeframe: · FJ$236,086 - 4 months

Module 3:EVI Model/Indicator Refinement

Module 3.1 Indicator Assessment and Refinement

Aim: To consolidate comments and international expert input into a refined EVI model and expose the environmental vulnerability index to mathematical and statistical tests

Prerequisites: This task requires substantial inputs from the international technical community using the "think tank" as a major vehicle

Although consultation was made with regional experts during Phase I, not all fields of expertise could be covered. and the full range of environmental criteria necessary for proper development of the EVI could not be accessed. It will be necessary in Phase II for the questions at the heart of the EVI to be re-examined and additional ones incorporated into the model where necessary. This process will require additional work by the EVI Team through continued regional and international expert consultation.

It will also be necessary to attend and participate in regional and internationalfora to access and further the discussion process on the environmental vulnerability index issue. Special focus will be on:

· Reviewing the questions incorporated during Phase I · Reviewing the response levels set for the 1-7 scale in Phase I · Identifying any biases inherent in the questions · Identifying redundant questions · Identifying questions for which data should be collected but is currently unavailable and obtaining a time frame for inclusion · Investigating potential measures of intrinsic resilience that should be included.

8 South Pacific Applied Geoscience Commission (SOPAC) 08/02/99

Special attention will also be given to the raising of awareness of the research being carried out on the environmental vulnerability index. This will done through the presentation of work progress in seminars, an internet web page, email distribution lists, newsletters and through the provision of information collected on environmental vulnerability index development to a recognised information database such as the University of the South Pacific’s PIMRIS. These activities will assist not only in the dissemination of information regarding the environmental vulnerability index but also to generate discussion and constructive feedback on the work being carried out.

Outputs/Benefits:

· Finalisation of indicators and their levels · Refined and consolidated environmental vulnerability index which has had extensive international and regional input and consultation · Republication of the refined EVI · Creation of an internet site on the environmental vulnerability index

Module 3.2 Mathematical testing of the model developed during Phase I

Aim: To thoroughly test the model through the use of real data as well as mathematical and statistical analysis

Prerequisites: Data will have needed to have been collected from 10 countries

To test the model's performance, it is necessary to collect data from a range of countries and then subject them to rigorous mathematical modelling. Dummy data cannot be used for this purpose because it does not embody the real environmental and anthropogenic heterogeneity likely to be encountered when the model is finally applied. It is necessary to use real data so that the model, when completed, is capable of distinguishing and ranking real countries and the complexities inherent in their different vulnerabilities.

Testing of the model will include: · Checking for correlation between questions (redundancy) · The methods of accumulation of scores · Effects of differences in data availability (and confidence) · Empirical testing of the model and indices and their ability to differentiate islands · Case studies with real countries

9 South Pacific Applied Geoscience Commission (SOPAC) 08/02/99

Statistical and empirical analysis of the model is an essential component in developing a useful and robust index which is often overlooked.

Outputs/Benefits:

· A robust EVI model which has been exposed to mathematical and statistical analysis · Proper identification of the mathematical properties of the EVI · A tested EVI model on real data from case study countries · EVI values for participating countries

Module 3.3 Combine with other indices? (e.g Economic Vulnerability Index)

Aim: To further work towards the development of a fully composite vulnerability index

In Phase II of the project, we will examine possibilities for merging the environmental data with economic data to produce a composite index, which expresses overall vulnerability of states to both sets of pressures. This work will rely heavily on inputs from Professor Lino Briguglio (part of the EVI Team) and others working in the field.

Output/Benefits:

· First operational composite environmental and economic vulnerability indices · Publication of work on composite vulnerability indices in an international journal · Raise profile on composite environmental and economic vulnerability indices

Module 3 Outputs/Benefits:

· Finalisation of indicators and their levels · Comprehensive mathematical testing of the EVI model to ensure its robustness · Identification of the mathematical characteristics of the EVI · A tested EVI model on real data from case study countries · The calculation of EVI values for several countries · Expansion of the EVI model to the world · Republication of the refined EVI · Awareness raising and publicity · Regular newsletters on the EVI for circulation amongst PICs · Creation of an internet web page on EVI · Creation in PIMRIS of a comprehensive dedicated section on environmental indicators · Creation of an EVI information database for all participating countries

10 South Pacific Applied Geoscience Commission (SOPAC) 08/02/99

· Pacific ownership of the EVI · Final report on the overall EVI · Composite environmental and economic vulnerability index · Important component in the overall outputs and benefits of the project

Budget/Timeframe:

· FJ$485,528 - 18 months

Module 4:Develop user interface (Access Model)

Aim: To develop a user interface in Microsoft Access

The model developed for calculating the EVI during Phase I was built into EXCEL spreadsheets. This makes calculating the EVI cumbersome for the endpoint users and was not intended for use by them as files may be easily damaged or altered, making calculations inaccurate.

During Phase II the model will be built as a user-friendly ACCESS application with on-line help to allow for easy and stream-lined input of data. The computer interface will also make the mathematical calculations robust to damage by users and secure the generation of final reports into quick and simple formats.

Module 4 Output/Benefits:

· The development of a robust and simple user-friendly Microsoft Access interface for the environmental vulnerability index model · Simplify and streamline the entry of environmental vulnerability index data and enhance the generation of final environmental vulnerability index reports · Important component in the overall outputs and benefits of the project

Budget/Timeframe:

· FJ$14,788 - 1 month

11 South Pacific Applied Geoscience Commission (SOPAC) 08/02/99

4. Overall Expected benefits from Project

5. Overall Project Outputs

The major outputs from Phase II will be: · An internationally recognised and reviewed EVI which is transparent in its strengths, biases and functions as a way of summarising the environmental vulnerability of states · EVI data for at least 10 countries to show their relative vulnerability’s of a range of variables which affect or characterise their environments · A user-friendly computer programme which can be sent to all countries wishing to calculate their EVIs (some with assistance) · An EVI that can be easily extended and applied to other countries throughout the world · An EVI that can be combined with other vulnerability indices into a composite vulnerability index · Pacific ownership of the EVI and the work that has been carried out in the region · Capacity-building within the region in the use of the EVI and also in the identification, collection and analysis of environmental data · Final report on the EVI

12 South Pacific Applied Geoscience Commission (SOPAC) 08/02/99

6. Overall Project Budget

· Total budget for Phase II - FJ$898,175 - 24 months

Budget Summary Module 1 Continuation, Review and Profile Raising of the EVI 1.1 Continuity of the EVI – Bridging Finance $84,275 1.2 International Expert “Think Tank” $58,853 1.3 Publication of EVI $18,645 Subtotal $161,773

Module 2 Capacity-building and EVI database 2.1 Capacity-building Workshop $146,997 2.2 Follow-up Consultation and Database $89,089 Subtotal $236,086

Module 3 EVI Model/Indicator Refinement Subtotal $485,528

Module 4 Development of EVI User Interface Subtotal $14,788

PROJECT Environmental vulnerability index project TOTAL $898,175

13 20/03/00

Environmental Vulnerability Index (EVI) Project Phase III SOPAC Funding Proposal

NAME OF PROJECT Environmental Vulnerability Index (EVI) Project Phase III

TYPE OF PROJECT Development of a robust global environmental vulnerability index

AUTHORITY SUBMITTING THE PROJECT South Pacific Applied Geoscience Commission (SOPAC) 1

IMPLEMENTING ORGANISATION South Pacific Applied Geoscience Commission (SOPAC)

RECIPIENT STATES Pacific Countries

COMMITMENT PROPOSED AS A GRANT USD $604,307 DURATION OF PROJECT 2 YEARS

The project proposal has been endorsed by the SOPAC Governing Council during its 28th Annual Session for submittal for funding.

1SOPAC Member Countries: Australia, Cook Islands, Federated States of Micronesia, Fiji, French Polynesia (Associate), Guam, Kiribati, Marshall Islands, Nauru, New Caledonia (Associate), New Zealand,Niue, Papua New Guinea, Samoa, Solomon Islands, Tonga, Tuvalu and Vanuatu. South Pacific Applied Geoscience Commission 08/02/99

Environmental Vulnerability Index Study

Terms of Reference for study to be carried out by SOPAC and funded by New Zealand

Objective

To pursue the development of an ecological/environmental vulnerability index consistent with the Barbados Plan of Action and the needs enunciated by the Alliance of Small Island States (AOSIS).

Background

Two particular paragraphs of the Barbados Plan of Action, paragraphs 113 and 114, refer to the need for a vulnerability index for Small Island Developing States (SIDS).

113. Small Island Developing States, in cooperation with national, regional and international organisations and research centres, should continue to work on the development of vulnerability indices and other indicators that reflect the status of SIDS and integrate ecological fragility and economic vulnerability. Consideration should be given to how such an index, as well as relevant studies undertaken on SIDS by other international institutions, might be used in addition to other statistical measures as quantitative indicators of fragility.

114. Appropriate expertise should continue to be utilised in the development, compilation and updating of the vulnerability index. Such expertise could include scholars and representatives of international organisations that have at their disposal the data required to compile the vulnerability index. Relevant international organisations are invited to contribute to the development of the index. In addition, it is recommended that the work currently under way in the United Nations system on the elaboration of sustainable development indicators should take into account proposals on the vulnerability index.

Of particular interest is the expression of the need for a single index that is composite, and incorporates an expression of both the economic and ecological (environmental?) vulnerability of SIDS compared with other states.

Much has been written about vulnerability of small states (e.g. Commonwealth Secretariat 1997) and in particular SIDS (e.g. Briguglio 1995). However, the recent Expert Group Meeting convened by the SIDS Unit of UNDESA failed to arrive at a solution that adequately addressed the Barbados Plan of Action. The position of SIDS was presented as an address to that meeting by the Chair of AOSIS. The following is an extract from that address.

1 EVI Phase II Report

Environmental Vulnerability Index:

Development and provisional indices

and profiles for Fiji, Samoa, Tuvalu and

Vanuatu

Ursula Kaly1 and Craig Pratt2

28 February 2000

SOPAC Technical Report 306

This project was funded by the New Zealand Overseas Development Assistance (NZODA) Programme

1 Environmental Adviser, Government of Tuvalu 2 South Pacific Applied Geoscience Commission (SOPAC), Fiji Environmental Vulnerability Index (EVI)

to summarise national environmental

vulnerability profiles

By:

Ursula Kaly1, Lino Briguglio2, Helena McLeod3, Susana Schmall3, Craig Pratt3 and Reginald Pal3

4th February 1999

SOPAC Technical Report 275

This project was funded by the New Zealand Official Development Assistance (NZODA) Programme

1 Environmental Adviser, Tuvalu Government, Tuvalu 2 Islands and Small States Institute, Foundation for International Studies, University of Malta, Malta 3 South Pacific Applied Geoscience Commission (SOPAC), Fiji South Pacific Applied Geoscience Commission (SOPAC) 8/02/99

Pacific Regional Environmental Vulnerability Index

Study Project Update 20 November, 1998

Background

Concern regarding the measuring of vulnerability of small island developing states (SIDS) was first brought to international attention during the Global Summit on Small Island States held in Barbados in 1994. At this conference SIDS with the support of the United Nations expressed the desire in paragraphs 113 and 114 of the Barbados Plan of Action for the development of a vulnerability index for Small Island Developing States (SIDS).

113. Small Island Developing States, in co-operation with national, regional and international organisations and research centres, should continue to work on the development of vulnerability indices and other indicators that reflect the status of SIDS and integrate ecological fragility and economic vulnerability. Consideration should be given to how such an index, as well as relevant studies undertaken on SIDS by other international institutions, might be used in addition to other statistical measures as quantitative indicators of fragility.

“…Small Island states proceed, of course, from the basis of their relative disadvantage: geographic often oceanic isolation, small size, economic fragility, and their acknowledged vulnerability to environment, ecological and natural disasters. It has always been a matter of particular concern for our countries that the current criteria for determining socio-economic status is not comprehensive enough and not a true measure of the economic and social strength of the small island developing states. The anomaly borne out by Professor Briguglio’s work, for instance, is that many small island developing countries register relatively high gross national product per capita; yet in reality, their economies are very

1 South Pacific Applied Geoscience Commission (SOPAC) 8/02/99 susceptible to any external economic fluctuations and environmental shocks, no matter how minimal.

For us, the Barbados Plan of Action for the Sustainable Development of Small Island Developing States provides the basis and raison d’etre for the Vulnerability Index. We need full and proper understanding of all the components and technical nature of vulnerability, and so that we can plan and seek from the international community vital support for our efforts at sustainable development for our part AOSIS has endeavoured for some time to spark some momentum into the work on the Index, though we were pleased that the fourth and fifth Sessions for the Commission on Sustainable Development specifically recognised the need to accord sufficient priority to it.

Thus far, the issue of the vulnerability of the small island developing countries has tended to be projected more on the political level. In the daily lives of our Governments and citizens, there are indeed constraints, whether one calls them unique or special constraints, they are real enough obstacles which hinder the search for sustainable development. In particular contexts, as for example in the ongoing debate on global climate change, small islands countries are acknowledged to be highest among the most vulnerable and the least able to adapt to the impact of climate change. Available data, including critically important elements of the relative resilience and adaptability of small island countries, would no doubt need to be fully investigated and assessed as to their technical nature and implications. This Expert Group Meeting is therefore most timely and necessary in that we need assessment at expert and technical level on the specific variables and criteria applicable for Vulnerability Indices, and so that we can come to a more complete and proper understanding of these matters…”

Given the manner to which vulnerability indices are being used in the international arena it appears critical that vulnerability indices which address both environmental and economic dimensions be developed.

In order to address this matter it appears apparent that there is a lack of adequate data for all SIDS for both economic and environmental (e.g. natural disasters) parameters which impact on an island economy. There is also a lack of acceptable methodology to measure vulnerability. As a result there is a need to develop appropriate methodologies that highlight economic and environmental vulnerability thus strengthening the potential combination of the two sets of data into a composite vulnerability index.

2 South Pacific Applied Geoscience Commission (SOPAC) 8/02/99

The Forum Economic Ministers Meeting (8th July 1998) agreed to take a common Forum position with the objective of encouraging the UN to adopt a vulnerability index. It is hoped such an index would be included among the criteria for determining LDC status, and for deciding eligibility for concessional aid and trade treatment.

As a result of a Commonwealth Ministerial Mission the World Bank and the IMF established a Task Force for Small States on 13th July 1998 to study the legal, environmental and economic vulnerability of small states not currently eligible for IDA-funding.

The Twenty-Ninth South Pacific Forum was held in Pohnpei, Federated States of Micronesia in August 1998 where the Forum reasserted in its Communiqué (paragraph 33 & 34) the importance of the vulnerability index and commended the positive developments at the United Nations and within the Commonwealth as well as in the region towards the completion of a comprehensive vulnerability index encompassing such factors as environmental and capacity considerations.

The Commonwealth Secretariat recently completed a report, “A Study on the Vulnerability of Developing and Island States: A Composite Index” which was presented to the Commonwealth Finance Ministers meeting in Ottawa, Canada in late September. The Ministers were pleased with the progress shown in the development of a composite vulnerability index and requested that it be further refined in consultation with the World Bank.

The Commonwealth Secretariat’s proposed composite vulnerability index was later presented to the first meeting of the joint Commonwealth Secretariat/World Bank Task Force held in Washington on 8 October. Discussions were positive with the World Bank endorsing the value and need to refine the composite index in order to develop a widely acceptable index and operational tool.

The outcome of the Commonwealth Secretariat composite index study clearly showed that in general small states are more vulnerable than large developing countries. The composite index presented also had an element of environmental vulnerability but the Commonwealth Secretariat endorses the need to develop a separate composite index in this area to help reinforce the vulnerability argument. In view of this, the current environmental vulnerability index study being undertaken by SOPAC is timely.

At the recent workshop convened to discuss Pacific regional input to the seventh session of the Commission on Sustainable Development (CSD7), held in Apia, Samoa from the 13 – 15 November 1998, Pacific country representatives, regional agencies, the private sector and non-government organisation delegates prioritised key challenges to sustainable development facing the Pacific. In their final submission the workshop reiterated the urgent need to develop a vulnerability index and called for international support to achieve this:

3 South Pacific Applied Geoscience Commission (SOPAC) 8/02/99

“The need for composite vulnerability index of economic as well as ecological / environmental parameters was highlighted in the Barbados Programme of Action. The 1998 South Pacific Forum agreed the index should be included among criteria for determining Least Developed Country status and for deciding eligibility for concessional aid and trade treatment. Pacific region is now engaged in the development of the ecological aspects of a vulnerability index and will contribute to efforts of UNDP, the World Bank and Commonwealth Secretariat to development a composite vulnerability index. The CSD is requested to call for:

· UN System agencies to assign the necessary resources to support work in the region to complete the environmental vulnerability index;

· support to ensure work on a composite index to be completed at the international level by the year 2000.”

Project Aim

To pursue the development of an environmental vulnerability index (EVI) consistent with the Barbados Plan of Action and the needs enunciated by the Alliance of Small Island States (AOSIS).

Project Team

SOPAC Study Team Consultants · Russell Howorth – Geologist · Ursula Kaly – Marine Biologist · Jackson Lum – Geologist · Lino Briguglio – Economist · Craig Pratt – Environmental Scientist · Susana Schmall – Environmental Scientist · Helena McLeod – Resource Economist · Reginald Pal – Environmental Scientist

4 South Pacific Applied Geoscience Commission (SOPAC) 8/02/99

Project Progress

A wide process of consultation will be required to develop an acceptable and practical environmental vulnerability index that will be useful not only regionally but also internationally. The Study Team has attempted to inform and involve as wide a forum of people as possible within the project timeframe from August – December 1998. This includes international, regional and national experts and organisations (See Appendix 1).

Several discussions on the proposed concepts and approach to developing an environmental vulnerability index have already been undertaken by the Study Team with the last discussion being held during the Pacific Input to the Commission on Sustainable Development Workshop held in Apia, Samoa from 13 – 15 November, 1998 (See Appendix 2 for details). Feedback to date from all these meetings has been encouraging with both experts and organisations expressing their support for such an initiative.

Following an extensive literature and Internet search for relevant information the EVI study team devised a methodology for developing the environmental vulnerability index. The proposed methodology involves the formulation of a list of indicator questions through scientific consensus which reflect environmental vulnerability to risks such as cyclones, floods, pollution and other types of possible risks. Draft lists of questions have already been circulated to experts to obtain comments and input into the construction of the environmental vulnerability index and question refinement process as well as to ascertain the availability of data. The final list of agreed questions will then form the basis upon which quantification of environmental vulnerability will be established through the input of specific country data by appropriate expert representatives.

To make the input of data easier and assist in the understanding of the environmental vulnerability index a computer model has been developed. The computer model will simplify and streamline data entry as well as make calculation of the EVI easier. A test version of the computer model is now available and we would greatly value any comments to help improve it into a simple and useful tool. The computer model capabilities is also expected to be tested on several case study countries.

The project will achieve an innovative framework for an environmental vulnerability index which will need further advancement and refinement to develop it into an internationally acceptable decision-making tool. Further research priorities will also be highlighted.

Timeframe

The draft report is to be submitted to the New Zealand Government by 31 December 1998 in time for revision and submission to CSD 7.

5 South Pacific Applied Geoscience Commission (SOPAC) 8/02/99

Contact Information

Russell Howorth

Programme Manager SOPAC Private Mail Bag Suva Fiji Islands

Ph: 679 381 377 Fax: 679 370 040 Email: [email protected]

Craig Pratt Ursula Kaly

Co-ordinator Consultant Pacific Regional Environmental Vulnerability Pacific Regional Environmental Vulnerability Index Study Index Study SOPAC SOPAC Private Mail Bag Private Mail Bag Suva Suva Fiji Islands Fiji Islands

Ph: 679 381 377 x243 Ph: 679 381 377 Fax: 679 370 040 Fax: 679 370 040 Email: [email protected] Email: [email protected]

6 South Pacific Applied Geoscience Commission (SOPAC) 8/02/99

Appendix 1: Consultation List

Name Organisation

Dr Anand Tyagi Biology, USP Dr Avi Shapira Seismology Division, Israel

Dr Barrie Pittock CSIRO

Dr Bob Dunn SPC Dr Dick Watling Environmental Consultant

Dr Ferdinando Villa University of Maryland, USA

Dr Graham Sem SPREP Consultant Dr Jackson Karunasekera Commonwealth Secretariat

Dr James Terry Geography, USP

Dr Jenny Bryant-Tokalau UNDP Dr Jonathan Atkins Commonwealth Secretariat Consultant

Dr K. Koshy START-Oceania

Dr Mahendra Kumar Chemistry, USP Dr Margaret Chung UNDP

Dr Marie Ferland Marine Studies, USP

Dr Patrick Nunn Geography, USP Dr Randy Thaman Geography, USP

Dr Richard Robarts UNEP & WHO GEMS/WATER Collaborating Centre

Dr Robert Clark Canadian Wildlife Service Dr Roger Jones CSIRO

Dr Sitiveni Halapua PIDP

Dr Sonia Mazzi Commonwealth Secretariat Consultant Dr Terry Done Australian Institute of Marine Science

Dr Tim Adams SPC

Mr Arthur Dahl UNEP Earthwatch Mr Bhaska Rao Minerals Resources Department, Fiji Government

Mr Bill Hare Greenpeace

Mr Cariston B. Boucher AOSIS Mr Chuck Hesley Sea Grant Programme, University of Hawaii

Mr Datuk R. Chander UN Consultant

Mr Ed Lovell Environmental Consultant Mr Epeli Nasome Department of Environment, Fiji Government

Mr Esekia Solofa Vice Chancellor, USP

Mr Fei Tevi PCRC Mr Gerald Miles SPREP

Mr Iosefa Maiava Forum Secretariat

Mr Jack Bhusan Lands Department, Fiji Government Mr James Robertson Asia Pacific Network

Mr Johnson Seeto Marine Studies, USP

7 South Pacific Applied Geoscience Commission (SOPAC) 8/02/99

Name Organisation

Mr Jone Fereti Drainage & Irrigation Division, Fiji Government

Mr Kenneth Piddington SPREP Consultant

Mr Leone Limalevu Department of Environment, Fiji Government Mr Lionel Gibson Geography, USP

Mr Lopeti Senituli PCRC

Mr Marika Tuiwawa Biology, USP Mr Max Finlayson Vulnerability Assessment of Major Wetlands in the Asia-Pacific Region Project

Mr Michael Petterson British Geological Survey

Mr Michael Ward World Bank Mr Neale Farmer SPREP

Mr Peni Sikivou Forum Secretariat

Mr Peter Hunnam WWF Mr Pierre Encontre UNCTAD

Mr Robert Holzmann World Bank

Mr Sefa Nawadra Environmental Consultant Mr Seremaia Tuqiri IOI

Mr Stan Vandersyp Forum Secretariat

Mr Steen Jorgensen World Bank Mr Tom Crowards Caribbean Bank

Mr Tuiloma Neroni Slade AOSIS

Mr Victorio Uherbelau FFA Ms Angie Heffernan Greenpeace

Ms Milika Naqasima-Sobey Marine Studies, USP

Ms René Rebatel Agency for Water & Environment in the Pacific, New Caledonia Ms Sofia Bettencourt World Bank

Ms Vina Ram-Bidesi Marine Studies, USP

Professor Dennis Pantin UN Consultant Professor John Morrison University of Wollongong

Professor Keith Crook University of Hawaii

Professor Peter Newell Biology, USP Professor Robin South Marine Studies, USP

SOPAC 1998 Annual Session Country Delegates

SOPAC STAR Expert Consultants SPREP GEO/CSD7 Regional Workshops

SPREP Country Delegates

8 South Pacific Applied Geoscience Commission (SOPAC) 8/02/99

Appendix 2: Project Meetings Timetable

MEETINGS LOCATION DATES SPREP Annual Session SPREP 15 – 18 September 1998 · SPREP Country delegates · SPOCC agencies SOPAC Brainstorm Session SOPAC 16 September 1998 · SOPAC Staff Meeting of Experts in Fiji SOPAC 24 September 1998 · USP · UNDP · Forum Secretariat · Fiji government representatives · Non-government organisations Panel Discussion – SOPAC Annual Session SOPAC 26 September 1998 · STAR International Experts · SOPAC Country delegates · SPOCC agencies Commonwealth Finance Ministers Meeting Ottawa, Canada 29 September – 1 October 1998 (Presentation of Comm. Sec. Composite Vulnerability Index Report) World Bank / Commonwealth Secretariat Task Force on Washington 8 October 1998 Vulnerability Indices Meeting (Presentation of Composite Vulnerability Index Report for discussion) Inaugural START-OCEANIA Workshop USP 5 – 9 October 1998 · Regional university organisation representatives SPREP Regional Consultations SPREP 9 – 10 November 1998 1. Pacific Input to the GEO 2 11 – 13 November 1998 2. Pacific Input to CSD 7 · SPREP Country Representatives · Non-government organisations · Private sector · SPOCC agencies Roundtable Meeting of Pacific Island Ministers on Sustainable Auckland, NZ 17 November 1998 Development Ministerial Conference on Small Island Developing States in the Malta 24 – 27 November 1998 Indian Ocean, Mediterranean and Atlantic Regions · Country representatives from other regions World Bank / Commonwealth Secretariat Task Force on ? Washington Early December 1998 Vulnerability Indices Meeting (Refining and finalising of Composite Vulnerability Index Report) Expert Discussion Meeting in Fiji SOPAC 17 December 1998 (tentative) · EVI Consultants · UNDP · Fiji government representatives · Non-government organisations · SPOCC agencies Environmental Vulnerability Index Study Presentation to NZ SOPAC 31 December 1998 Government

9 Cataloguing-in-publication data:

Kaly U., Briguglio L., McLeod H., Schmall S., Pratt C. and Pal R. 1999. Environmental Vulnerability Index (EVI) to summarise national environmental vulnerability profiles. SOPAC Technical Report 275. 66p.; 3 annexes, 2 figures, 1 table. ISBN 982-207-009-8

1. Vulnerability index – environment

This report describes the development of a vulnerability index for the environment which could be calculated on the scale of entire states for the purpose of ranking them and providing a single-figure expression of their relative environmental vulnerabilities. This work was done in response to a call made in the Barbados Plan of Action, the Alliance of Small Island States (AOSIS) and an increasing awareness that small island developing states face disadvantages to their development associated with their remoteness, small size, dispersion, economic conditions and limited natural resources. In the past vulnerability indices have been developed which describe the risks associated with economic and social conditions, climate change, sealevel rise, natural disasters and anthropogenic impacts. Most of these indices describe the vulnerability of human systems; there have been very limited attempts to describe effects on the environment. Human systems and the environment are dependent on one another so that risks to the environment of a state will eventually translate into risks to humans. This is the first attempt to construct an index that focuses on the vulnerability of the environment.

An Environmental Vulnerability Index (the EVI) was constructed, based on a theoretical framework that identified three aspects of vulnerability: risks to the environment (natural and anthropogenic), the innate ability of the environment to cope with the risks (resilience) and ecosystem integrity (the health or condition of the environment as a result of past impacts). These three aspects correspond to three sub-indices, the REI, IRI and EDI, which are the Risk Exposure sub-Index, Intrinsic Resilience sub-Index and Environmental Degradation sub- Index respectively. The EVI was calculated as a weighted average of scores allocated in the range of 0-7 derived from a total of 57 indicators.

A preliminary EVI was calculated for three countries, Australia, Fiji and Tuvalu. The preliminary EVI value for Tuvalu was the highest of the three countries indicating that its environment is the most vulnerable. The score obtained for Fiji was intermediate in value, and that for Australia was the lowest, though the difference between Fiji and Australia was relatively smaller than that between Fiji and Tuvalu. There were similar levels of risk in each of the countries, the most degradation in Australia and the least intrinsic resilience in Tuvalu. These results, though promising are only preliminary because the EVI requires refinement and there was insufficient time to collect all of the required data from all of the countries and because there were some inconsistencies in the quality of the data. We expect that each of these problems can be overcome.

The EVI developed here will require further refinement before it becomes fully operational. The results show that it is possible to obtain a single figure measure of vulnerability and that data which were previously thought to be difficult to obtain can be obtained. The methodology selected in the computation of the index can produce results which could have operational usefulness for ranking countries according to their environmental vulnerabilities.

It is envisaged that the EVI would be recalculated every 5 years to provide updates on the vulnerability status of countries. This index highlights the need for governments to upgrade their collection and collation of environmental statistics. In addition, the breakdown of results into meteorological, geological, anthropogenic, and other categories of risk highlights areas of concern for environmental action. Acknowledgements

This project was undertaken with the aid of a grant provided by New Zealand Overseas Development Assistance (NZODA) Programme. The views expressed in this document are those of the authors and SOPAC, but are not necessarily shared by the New Zealand Government or any of the people consulted during the project.

We wish to express our thanks to Alfred Simpson (Director), Russell Howorth (Programme Manager) and Jackson Lum (Project Leader) for their support of the project. Jackson Lum also assisted with editing of this report. We also wish to thank Franck Martin for programming the draft ACCESS model for calculating the EVI.

We wish to thank the following regional and international experts for their inputs to this project: Anand Tyagi, Biology, USP (Plant Geneticist); Andrew Munro, SPREP; Arthur Dahl, UNEP Earthwatch (Ecologist); Atish Prasad, Agricultural Department, Land Use Unit (Agriculturist); Avi Shapira, Seismology Division, Israel; Barrie Pittock, CSIRO Atmospheric Research (Climate Scientist); Chuck Hesley, Sea Grant Programme, University of Hawaii; Datuk R. Chander, UN Vulnerability Index Consultant (Economist); David Scott, SOPAC, Water and Sanitation Programme (Hydrogeologist); Dick Watling, Environmental Consultant (Biologist); Ed Lovell, Environmental Consultant (Marine Biologist); Epeli Nasome, Department of Environment, Fiji Government (Town Planner); Evelyn Reigber, Pacific German Forestry Project (Forester); Ferdinando Villa, University of Maryland, USA (Ecologist); Graham Shorten, SOPAC, Disaster Reduction Unit (Geologist, Engineer); Jackson Karunasekera, Commonwealth Secretariat; James Terry, Geography, USP (Hydrologist); Joe Chung, SOPAC, Disaster Management Unit; Johnson Seeto, Marine Studies, USP (Marine Biologist); Kanayathu Koshy, START-Oceania; USP (Chemist); Keith Crook, University of Hawaii (Geologist); Lionel Gibson, Geography, USP (Geographer); Lopeti Senituli, PCRC; Mahendra Kumar, SPREP International Negotiations (Physicist); Margaret Chung, UNDP Sustainable Development Adviser (Geographer); Marie Ferland, Marine Studies, USP (Geologist); Marika Tuiwawa, Biology, USP (Botanist); Michael Petterson, British Geological Survey (Geologist); Patrick Nunn, Geography, USP (Geologist); Peter Hunnam, WWF (Marine Biologist); Philip Burgess, Environment Australia, Australia; Philipp Muller, New Zealand Government (Marine resource manager); Pierre Encontre, UNCTAD; Randy Thaman, Geography, USP (Biogeographer); Roger Jones, CSIRO Atmospheric Research (Climate Scientist); Stan Vandersyp, Forum Secretariat (Economist); Tim Adams, SPC (Fisheries Manager); Tom Crowards, Caribbean Bank (Environmental Economist). Contents

ABSTRACT ...... III

ACKNOWLEDGEMENTS ...... IV

EXECUTIVE SUMMARY...... 1

1 INTRODUCTION ...... 11

1.1 BACKGROUND OF THE PROJECT ...... 11 1.2 WHAT IS VULNERABILITY?...... 12 1.3 THE NEED FOR AN ENVIRONMENTAL VULNERABILITY INDEX...... 13 1.3 AIMS OF THIS STUDY...... 14 2 MINI REVIEW OF PREVIOUS WORK ON VULNERABILITY INDICES ...... 15

2.1 VULNERABILITY INDICES IN GENERAL...... 15 2.2 ENVIRONMENTAL VULNERABILITY INDICES...... 15 2.3 SUMMARY OF APPROACHES AND METHODS USED IN PAST VULNERABILITY INDICES AND LESSONS LEARNED FROM THEM... 16 2.3.1 Terminology...... 16 2.3.2 Framework...... 16 2.3.3 Methodology...... 16 3 A THEORETICAL FRAMEWORK FOR ENVIRONMENTAL VULNERABILITY ...... 18

3.1 THE RISKS AND ECOSYSTEM INTEGRITY...... 18 3.3.1 The risks...... 18 3.1.2 The entities at risk: 'Responders' ...... 19 3.2 THE THREE ASPECTS OF ENVIRONMENTAL VULNERABILITY...... 20 3.2.1 Risk exposure as part of vulnerability and the REI sub-index...... 20 3.2.2 Intrinsic vulnerability / resilience and the IRI sub-index...... 21 3.2.3 Extrinsic vulnerability / resilience and the EDI sub-index...... 22 3.3 FEATURES OF AN ENVIRONMENTAL VULNERABILITY INDEX (EVI)...... 22 3.4 WHY USE INDICATORS?...... 23 4 METHODOLOGY USED...... 24

4.1 INDICATORS OF VULNERABILITY...... 24 4.1.1 Criteria for the selection of indicators ...... 24 4.1.2 Indicators for the three sub-indices ...... 25 4.1.3 Collecting the data on indicators...... 26 4.2 QUANTIFYING VULNERABILITY ...... 27 4.2.1 Mapping of data on indicators on a 1 - 7 scale ...... 27 4.2.2 Setting the response levels for indicators...... 29 4.2.2 Weighting ...... 29 4.2.3 Calculating the sub-indices and the EVI...... 29 4.2.4 Confidence in and source of the data...... 31 5 PRELIMINARY RESULTS FOR AUSTRALIA, FIJI AND TUVALU...... 31

5.1 VALUES OBTAINED FOR THE EVI AND SUB-INDICES...... 32 5.1.1 Index values ...... 32 5.1.2 Nett and gross vulnerabilities ...... 33 5.1.3 Confidence in the results and data ...... 34 5.2 CONCLUSIONS...... 34 6 STRENGTHS AND WEAKNESSES OF THE EVI...... 35

6.1 STRENGTHS ...... 35 6.2 WEAKNESSES...... 35 7 CONCLUSIONS AND FUTURE DIRECTIONS ...... 36

7.1 PROMISING RESULTS...... 36 7.2 REFINING THE EVI AND EXTENDING IT GLOBALLY...... 37 7.3 INSTITUTIONAL CONSIDERATIONS ...... 38 8 CONCLUSIONS...... 39

9 DEFINITION OF ACRONYMS AND TERMS ...... 40

10 REFERENCES...... 41

ANNEX 1 TERMS OF REFERENCE FOR THIS STUDY...... 43

OBJECTIVE...... 43 BACKGROUND ...... 43 STUDY DESCRIPTION – STRATEGY ...... 45 REPORTING...... 45 ANNEX 2 THE VULNERABILITY INDICATORS USED...... 46

ANNEX 3 SUMMARY TABLE OF OTHER VULNERABILITY INDICES DEVELOPED ...... 67

Figures and Tables

FIGURE 1: GRAPH OF THE OVERALL EVI RESULTS OBTAINED FOR AUSTRALIA, FIJI AND TUVALU (NETT AND GROSS SCORES)...... 33 FIGURE 2: BREAKDOWN OF THE EVI SCORES FOR AUSTRALIA, FIJI AND TUVALU INTO THE THREE SUB-INDICES FOR NETT SCORES ONLY...... 33

TABLE 1: SUMMARY OF RESULTS OF CALCULATING VALUES FOR THE EVI AND SUB-INDICES FOR AUSTRALIA, FIJI AND TUVALU. VALUES ARE GIVEN FOR NETT AND GROSS INDEX VALUES AND A CATEGORICAL BREAKDOWN OF THE REI AND EDI. THE TOTAL NUMBER OF INDICATORS AVAILABLE FOR EACH INDEX IS GIVEN. NA=NUMBER OF INDICATORS WHICH WERE NOT APPLICABLE FOR THE COUNTRY (ASSIGNED NO VALUE IN NETT AND A 0 VALUE IN GROSS INDICES); ND=NO DATA CURRENTLY AVAILABLE (ASSIGNED NO VALUE)...... 32 Executive Summary

Background

Small Island Developing States (SIDS) face serious disadvantages to their development associated with an interplay of factors such as remoteness, geographical dispersion, vulnerability to natural disasters, a high degree of economic openness, small internal markets, limited natural resources and fragile ecosystems. These issues have been recognised and increasingly highlighted in international fora during the last decade.

Initial attempts at constructing a vulnerability index focused on economic aspects and followed a proposal by the Maltese Ambassador during a 1990 UNCTAD expert meeting on the problems of small island developing states.

Vulnerability indices have also been developed for the risks associated with climate change and sealevel rise, the El Nino-Southern Oscillation phenomenon, anthropogenic impacts and natural disasters. Most of these indices describe the vulnerability of human systems (economies, development, social systems) to these risks. Only a few studies have attempted to measure the vulnerability of the environment itself to anthropogenic and natural hazards.

Human systems and the environment are dependent on one another. Risks to the environment of a state will eventually translate into risks to humans because of their dependence on the natural environment for resources. In turn, the environment is susceptible to both natural events and appropriate management by humans. This means that overall vulnerability of a state should include measures of both human and natural systems and the risks which affect them. The development of an Environmental Vulnerability Index (EVI) will be an important step towards development of a Composite Vulnerability Index (CVI).

Definitions

Vulnerability is the potential for attributes of a system to be damaged by exogenous impacts and resilience is the potential to minimise or absorb the effects of these damaging impacts. The aim of vulnerability indices is to describe the relative vulnerability of states. Different forms of vulnerability of states are described. For example, economic vulnerability is concerned with external forces which act on the economy, while social vulnerability occurs when natural or other disasters force massive upheavals of residence, traditions and society.

The focus of this study is on vulnerability of the environment itself to both human and natural hazards. This includes effects on the physical and biological aspects of ecosystems, diversity, populations or organisms, communities and species. Unlike previously-developed vulnerability indices, human impact is considered an exogenous factor, and human systems not the recipients of the impact.

Environmental vulnerability differs from vulnerability of human systems because the environment is complex with different levels of organisation from species to interdependent ecosystems and the complex linkages between them. Because data are often not available and indicators for health and vulnerability of the environment have to be physically measured, indicators may be heterogeneous in nature and not expressible in common units. This means that developing an index for the environment will need a new approach.

The need for an EVI

The need for an environmental vulnerability index was recognised at the Global Summit on Small Island States held in Barbados in 1994 where the United Nations formally expressed the desire in paragraphs 113 and 114 of the Barbados Plan of Action for the development of a such an index. The benefits of producing an EVI are that it can attract attention to certain states which are considered 'more vulnerable' and it summarises vulnerability based on meaningful criteria which can be considered by donors when allocating financial aid and projects.

The overall aim of this study was to begin the development of an environmental vulnerability index consistent with the Barbados Plan of Action and needs of the Alliance of Small Island States (AOSIS). The EVI developed here could then be combined with an economic vulnerability index to give a composite index which in a single figure format incorporates the environmental and economic concerns of a state. It is envisaged that the EVI and CVI would be recalculated every 5 years to examine changes through time as well as relative rankings of countries.

The specific aims were to: · Review current work already completed or underway addressing environmental vulnerability of SIDS; · Build on past work on environmental vulnerability, if appropriate, or approach the problem from a new perspective where other attempts have had limited success; · Identify variables which may be used in the construction of an environmentally descriptive vulnerability index for Pacific SIDS; · Develop a logical framework and methods of calculating and index for environmental vulnerability; · Identify and collect data which would be used to calculate the environmental index; · Identify gaps in the available data; · Identify future directions for the further development of an internationally acceptable environmental vulnerability index; · Compile a report for widespread circulation and consideration prior to the Donor Round Table leading up to CSD-7 and the UN General Assembly special session on SIDS in September 1999; and · Further efforts towards the development of a composite vulnerability index as described in the Barbados Program of Action and meet the needs enunciated by AOSIS.

Mini review of previous work on vulnerability indices

Fifteen studies were reviewed which examine the relative vulnerabilities of states in terms of risks to human and natural systems. Most of the studies were concerned with risks to human economic and social systems (13), while only 5 attempted to describe effects on the natural environment. The risks of concern also varied among studies. Anthropogenic risks were considered in 11 studies; 6 studies examined climate change and sealevel rise; and 6 studies considered natural disasters as part of their risk. Only 1 study specifically examined the effects of both humans and natural hazards on the environment. It is the object of this study to fill this gap by developing an environmental index based on a wide array of environmental indicators which includes both natural and anthropogenic risks.

Features of past vulnerability indices

1. There has been some confusion with terminology. In one study, an Ecological Vulnerability Index was developed which actually looked at the vulnerability of human systems to natural disasters and inherent geographic characteristics, rather than vulnerability of the environment. We propose that vulnerability indices should be named by their responders, not risks. Human vulnerability indices include Economic, Quality of Life and Human Development Indices. This is the first study to attempt a true Environmental Index. 2. The logical framework for past indices has tended to be lacking. The successful development of an EVI requires a logical framework to ensure that the index is not just driven by data availability, terms are fully defined, appropriate indicators are found and the model can be tested. 3. The number of indicators varied among studies. Studies with only a few indicators (3-6) tended to focus on human systems. When the number of indicators used was moderate (12-15) more emphasis was placed on natural disasters and ecological variables. Only one study used a large number of indicators (60) and it was the only one which produced a list of indicators for anthropogenic pressures on the environment. The lesson here is that more indicators are required when complex ecological systems become the focus of the index being constructed. 4. Five different methods of evaluating or scoring the indicators were identified from the studies reviewed. Some of these were considered for the present study. 5. Five different methods were identified for aggregating the value of the indicators into an index or sub-indices. None of these was considered appropriate for the EVI. A modification of two methods identified from past indices was used. Theoretical framework for the EVI

The maintenance of ecosystem or ecological integrity is at the heart of the development of a vulnerability index for the environment, because it is ecosystem integrity that is threatened by natural and anthropogenic hazards. The notion of ecosystem integrity is so complex that it cannot be expressed through a single indicator, but rather requires a set of indicators at different spatial, temporal and hierarchical levels of the ecosystem. Ecosystem integrity depends on biodiversity, ecosystem functioning and resilience, all of which are such interrelated variables, that factors which affect just one of these can have far-reaching ecosystem-wide consequences.

The risks to the environment are any events or processes that can cause damage to ecosystem integrity. These include natural and human events and processes such as 'the weather' and 'pollution'. Some researchers have identified natural hazards as those in which natural environmental conditions depart from 'normal' to such an extent that systems of interest (human, environmental) may be adversely affected. The problem with this definition is that unless we identify certain natural events as being anthropogenically altered (e.g. anthropogenically-accelerated sea-level rise), all events are 'normal'. The implication from this line of reasoning is that the changes we see to the natural world as a result of natural hazards are deemed 'unacceptable' from a human perspective. This means that except in the case of anthropogenic risks, in an assessment of environmental vulnerability, what we really are examining is unacceptable departures from our (human) view of how the environment should change. For the purposes of this study, we will accept that risk events should include those which cause sudden and seemingly-negative impacts on natural systems as a way to evaluate vulnerability.

Although most identifiable risk events are capable of causing damage, it is only the larger and more intense events that are likely to cause wholesale changes in the environment, at least in the short to mid term. Some of the more important risks which can impact on the environment include meteorological events (e.g. cyclones, droughts, heatwaves), geological events (earthquakes, tsunamis, volcanoes), anthropogenic impacts (mining, habitat destruction, pollution), climate change and sealevel rise.

The entities at risk, termed the 'responders' include ecosystems, habitats, populations and communities of organisms, physical and biological processes (e.g. beach building, reproduction), energy flows, diversity, ecological resilience and ecological redundancy.

Three aspects of environmental vulnerability were identified which would need to be incorporated into an EVI. These are: 1. The level of risks (or pressures) which act on the environment within the state, forming the Risk Exposure sub-Index (REI) which examines the frequency and where possible, the intensity of risk events which may affect the environment. These are based on levels observed over the past 5-10 years for most risks, but may include data for much longer periods for geological events. The REI is a measure of potential risk only: There is no logical expectation that patterns of risk expression during the immediate history of a state will necessarily result in similar risk levels today or in the future; 2. Intrinsic vulnerability or resilience of the environment to risks, forming the Intrinsic Resilience sub-Index (IRI) which refers to characteristics of a country which would tend to make it less/more able to cope with natural and anthropogenic hazards; and 3. Extrinisic vulnerability or resilience as a result of external forces acting on the environment, forming the Environmental Degradation sub-Index (EDI) which describes the ecological integrity or level of degradation of ecosystems. The more degraded the ecosystems of a country (as a result of past natural and anthropogenic hazards), the more vulnerable it is likely to be to future risks.

Features of the EVI

In developing the EVI, we set criteria on certain features of the index to ensure that it would be able to perform the tasks for which it was developed. The criteria were that the EVI should be intuitively understandable (set within a range from which highly vulnerable states have immediate recognition as such), impartial, suitable for international comparisons and able to differentiate among countries, applicable at different spatial scales (regional, country, island), refinable, presented in breakdown and single figure formats and easy to calculate using a user-friendly computer interface.

Methodology

Because the risks are many and ecosystem resilience and integrity are complex in character, it was necessary to use indicators to characterise them. This means that not all aspects are covered, but that a subset of variables is selected which describes frequency and intensity of risks, intrinsic vulnerability, effects on ecosystems, groups of organisms, physical features of the environment and mitigators of effects.

For the purposes of the EVI the following definitions relating to indicators and indices were used: · An indicator was defined as any variable which characterises the level of risk, resilience or environmental degradation in a state; · A sub-index (the REI, IRI and EDI) was defined as an aggregated average of the scores for indicators which related separately to risk, resilience or degradation; and · An index (the EVI) was defined as an aggregated average of each of the three sub- indices (REI, IRI, EDI) to give an overall measure of the environmental vulnerability of a state. The EVI is then, a composite of each of the three sub-indices. The criteria for the selection of indicators was that they should be applicable over the entire scale of interest (countries, regions), spread over different geographic, habitat and climatic types, relatively easy to understand, well defined, have data available now or with assistance in the future and be as uncorrelated as possible (to limit redundancy).

A total of 57 indicators of environmental vulnerability were finally selected for inclusion in the index. This included 39 indicators of risk (REI), 5 indicators of resilience (IRI) and 13 indicators of environmental integrity or degradation (EDI) (Annex 2). Many of the indicators were expressed as a fraction of area of land or coast rather than simply absolute numbers because it is risk density or proportion of area degraded that is of interest from an environmental perspective.

Although a larger number of indicators would have been preferred to obtain a wider picture of risks, resilience and ecological integrity, many of the indicators initially selected were discarded because they either did not have data available and data were unlikely to be procured in the near future, they were ambiguous or bimodal in their responses; or were redundant and the information they intended to capture was present in another indicator.

Data for calculating the EVI (and initially setting the response levels) were collected for three countries: Fiji, Tuvalu and Australia to provide some initial testing of the model. These data were obtained from country reports, UN, WHO, SOPAC, SPREP, FAO and other publications from international agencies, centres for risk assessment and management (e.g. Tsunami Centre, NOAA), local experts and government officers. Without being able to go to the countries to train and focus attempts of local authorities to the task of collecting or collating the required data, some indicators were unavailable for this initial testing.

Quantifying vulnerability

Our overriding principle in constructing the EVI was not to introduce complexities into the model unless there was a justifiable reason to do so.

Environmental indicators are of a heterogeneous nature, that is they include variables for which the responses are numerical, qualitative and on different scales (linear, non-linear, or with different ranges). To deal with the heterogeneity, it was necessary to map the possible responses to the variables onto a 0-7 scale where 1-7 was used for the spread of values and 0 or N was used for 'non-applicable' and 'no-data' responses.

Response levels (maximum, minimum and intermediate divisions) for each of the indicators were set wherever possible using the technical literature or by consultation with generalists and specialists in each field. Some levels were set using the data from Tuvalu, Fiji and Australia where available, or as estimates when these sources were unavailable. Six of the 57 indicators were assigned an intrinsic weighting factor of 5, while the remaining indicators were given the default weighting of 1. This is the equivalent of giving the six weighted indicators the equivalent value of 5 indicators. This weighting was applied to indicators considered to be of central importance to the question of vulnerability in the model. To ensure that the final EVI, REI, IRI and EDI scores remained between the values of 0 and 7, it was necessary to adjust the weighting factors by dividing the intrinsic weighting value (1 or 5) by the average of all weighting values within each sub-index. The 0-7 scores were then multiplied by the adjusted weighting factor prior to accumulation in the sub-index to which they belonged.

The EVI and sub-indices were calculated using an EXCEL workbook. The workbook (Version 7-EVI-calculator.xls) is comprised of seven linked worksheets, each dealing with a different aspect of calculation and reporting. Report Level 1 was the highest level and gives the value of the EVI and sub-indices for each country and measures of confidence in the data. Report Level 2 gives a breakdown of the REI and EDI sub-indices showing relative contribution of meteorological, geological, anthropogenic risks and mitigating factors and ecosystem and biodiversity indicators. Report Level 3 gives the adjusted and raw scores for each individual indicator. A separate copy of the calculator is required to evaluate the vulnerability indices for each country.

After adjustment for intrinsic weighting, the scores for each indicator within a sub-index were averaged to produce a sub-index value of between 0 and 7. Where data were unavailable for an indicator, that indicator was omitted from the average, not given a 0 score, so that it made no contribution to the mean. Because there were also indicators for which the response was 'not applicable' (such as volcanic eruptions in Tuvalu), we calculated two types of index for the EVI and sub-indices. These we termed the Nett and Gross vulnerabilities. Nett vulnerability omits those indicators considered not applicable in a country and describes vulnerability to risks that actually apply in a country. Gross vulnerability assigns a zero value to non-applicable indicators and describes vulnerability in relation to all risk indicators used in the model and therefore vulnerability in a total sense.

In parallel with scoring each indicator against the 1-7 scale, we built into the EVI a way of assessing the reliability of data. These reliability values are reported alongside each index and should be read with them. The data reliability scores give the number indicators which were not applicable, the number with no data; the number of responses which were based on real data; and the number of responses based on 'best guess' or estimated by the operator and/or authorities. Preliminary results for Tuvalu, Fiji and Australia

The preliminary EVI value for Tuvalu was the highest of the three countries indicating that its environment is the most vulnerable. The score obtained for Fiji was intermediate in value, and that for Australia was the lowest, though the difference between Fiji and Australia was relatively smaller than that between Fiji and Tuvalu.

Australia Fiji Tuvalu EVI (Nett) 3.04 3.79 5.04

When the EVI was decomposed into the REI, IRI and EDI sub-indices and categories of risk, a more complex pattern emerged. The risk exposure index (REI) was similar for the three countries, varying only between 3.13 nett for Australia and 3.49 nett for Tuvalu. The intrinsic resilience index (IRI) varied the most among the countries with Australia having the best resilience score (1.00 nett) and Tuvalu having the worst (7.00 nett). An almost reverse pattern was seen for the EDI. In this sub-index, the highest score (worst conditions) was obtained by Australia and the lowest by Fiji. The values of the sub-indices suggests that the different aspects which form vulnerability can operate independently of one another. Tuvalu was characterised by poor intrinsic resilience, and moderate risks and degradation. In contrast, Australia was characterised by very good intrinsic resilience, but high degradation. Fiji had the greatest vulnerability to meteorological and geological events, and Australia the greatest risk to pollution.

The values obtained for nett and gross indices differed little. Although this suggests that there may be no need to separately calculate the two index types, the present study did not provide a good test because there were very few indicators which were 'not applicable' across all three countries. It is expected that the nett and gross index values will be required when the EVI is extended globally and a greater range of climatic and geographic areas are built into the model.

These results are only very preliminary. The EVI and sub-indices will only provide a reasonable measure of vulnerability if most of the indicators can be filled-in for a country. We suggest here that at least 80% of the indicators (46 of the 57 questions) should be filled out by any one country for a reasonable estimate of vulnerability. In this report, we were unable to reach this threshold for any country. The data are available for the remaining indicators, but are buried, or need to be requested from authorities and will take some time and effort to procure. There were also problems with quality of the data, as we often found conflicting estimates in the literature. These temporary deficiencies in the data means that all the EVI values obtained here are indicative: they are by no means completed estimates and should be read with caution. In conclusion, the EVI model gives single-figure measures of environmental vulnerability that appear to be able to distinguish countries and identify sources of vulnerability within a country. For the moment, it appears that of three countries tested, Australia is the least vulnerable, and Tuvalu the most. There were similar levels of risk in each of the countries, the most degradation in Australia and the least intrinsic resilience in Tuvalu. These results are only preliminary because there was insufficient time to collect all of the data required for these three countries and there were some problems with reliability of the data. We expect that each of these problems can be overcome and the results suggest that the EVI will be a useful tool for characterising vulnerabilities of states.

Strengths and weaknesses of the EVI

As for all methods of summarising and modelling data, the EVI developed here is associated with a number of strengths and weaknesses which must be understood for its proper application and use.

Strengths: The EVI is based on a theoretical framework that prompted us to find indicators for all identified aspects of vulnerability. It is able to incorporate quantitative and qualitative data on different response scales and identifies two types of vulnerability (Nett and Gross) simultaneously allowing for a world-wide comparison of states and as assessment of the real risks likely to affect a state. It also identifies areas in which local environmental agencies could improve data collection. Although at present it focuses on Oceania, it is extendable world-wide by the incorporation of new indicators.

Weaknesses: The index does not exclusively rely on published data resulting in omissions and a high cost of data collection. There is some subjectivity in assigning weights to indicators and non-linearities to the scores. The mapping of data on a 7 point scale may result in a loss of detail compared with directly using numerical data. In common with all indices, the EVI is affected by the indicators chosen and the results obtained may differ if different variables were chosen. Local variations, short and long term effects and other details could not be incorporated into the model without making it too complex. The index is also subject to problems with differences in interpretation of in-country users, though this could be minimised with training.

Future directions and conclusions

The environmental vulnerability index developed here will require further refinement before it becomes fully operational. This will include review of the indicators selected and the levels selected in a world-wide context, adding indicators for parts of the model which are underrepresented, and identifying indicators which should be incorporated and for which data should be collected. Refinement of the EVI will require peer review and inputs from highly-specialised experts. Mechanisms for this include running a 'think tank' and by publishing in an international reviewed journal. It will also be necessary to carry out consultations with the SIDS. A second mechanism for refining the index will be to go to, say, 10 SIDS to build their capacity to work with the EVI and to collect data which may at present be buried rather than lacking. Data should also be procured from developed countries by request. When sufficient data have been collected and the indicators refined, it will be necessary to test the performance of the model to identify biases, remove redundant variables, test its ability to differentiate countries, the method of accumulating scores and assess the availability and confidence in the data. It will also be necessary to develop a user-friendly computer calculator for the EVI. The final stage in the development of the EVI will be to combine it with other indices, such as the Economic Vulnerability Index to give an indication of overall vulnerability of states.

The results show that it is possible to obtain a single figure measure of vulnerability which incorporates the risks, intrinsic resilience and health or integrity of the environment. This study also shows that data which were previously thought to be difficult to obtain can be obtained. The methodology selected in the computation of the index can produce results which could have operational usefulness for ranking countries according to their environmental vulnerabilities. It is envisaged that the EVI would be recalculated every 5 years to provide updates on the vulnerability status of countries. This index highlights the need for governments to upgrade their collection and collation of environmental statistics. In addition, the breakdown of results into meteorological, geological, anthropogenic, mitigating and other categories of risk highlights areas of concern for environmental action. 1 Introduction

1.1 Background of the project

There has been growing international recognition that Small Island Developing States (SIDS) face serious disadvantages to their development associated with an interplay of factors such as remoteness, geographical dispersion, vulnerability to natural disasters, a high degree of economic openness, small internal markets, and limited natural resources (Briguglio 1995). These issues have been recognised and increasingly highlighted in international fora during the last decade. It was not until 1990 that the construction of a Composite Vulnerability Index (CVI) to measure the degree of overall vulnerability of developing countries was first formally proposed to the United Nations by the Maltese Ambassador.4 (Briguglio, 1997).

The issue of vulnerability of states to human and natural stressors on economics, other aspects of human development, resources and the environment is still in its development phase. Attempts have been made to provide measures of vulnerability in single figure composite index form for: · Economic vulnerability (Briguglio, 1993, 1995, 1997; Wells, 1996, 1997; Atkins et al. 1998; Chander, 1996;); · Climate change and sea-level rise (IPCC, 1991, 1994; Pernetta, 1990; Downing, 1992; Formel, 1996); · ENSO phenomenon (NOAA, 1997); · Human impacts on the environment (Erlich, 1991, UNEP, 1998; Eurostat, 1998); and · Effect of natural disasters on human systems (Albala-Bertrand, 1993; Pantin, 1997).

These above indices mostly attempt to describe the vulnerability of human systems to economic, social, climatic and other environmental factors. A few of the above studies have included some environmental responses in their indices, but no studies to date have attempted to construct a vulnerability index which describes risks to and responses of the environment, rather than on human systems.

It is now clear that vulnerability of states includes risks and their results on both the human and natural systems. Humans depend on the environment and its resources for sustaining life and for development. The environment, in turn, is dependent on both natural events and appropriate management by humans. This means that overall, vulnerability of a state5 should ultimately include measures of both human and natural systems and the risks that affect them. With the development of an Environmental Vulnerability Index (EVI), an important step towards developing a composite index will have been made.

4 The proposal was made at the Meeting of Government Experts of Island Developing Countries and Donor Countries and Organisations, held under the auspices of UNCTAD in June 1990.

5 The entities "states" being investigated in this study are politically independent territories. 1.2 What is vulnerability?

Vulnerability and resilience are two closely-related ideas. The term vulnerability refers to proneness to damage due to lack of protection or precariousness or the risk of being affected by a negative impact. Where vulnerability is considered high, resilience will be considered low and visa versa, and the two terms are used as opposites interchangeably in the field of vulnerability indices. For the purposes of this study, we will define the two terms as follows:

Vulnerability is defined as the potential for attributes of a system to respond adversely to the occurrence of hazardous events; and Resilience is defined as the potential for attributes of a system to minimise or absorb the impacts of extreme events.

States are vulnerable to many factors which can affect them economically, environmentally, socially and politically. All states are vulnerable to varying degrees. We are interested in identifying those which are the most vulnerable in a relative sense. Economic vulnerability occurs when the economy of an entity, such as an independent state, is at risk from negative impacts arising from external forces. Political vulnerability may be experienced when, for example, a state’s territorial boundaries are under threat and when forces within and outside the state exert a destabilising influence on the political and administrative establishment. Finally, social vulnerability occurs when, for example, natural disasters force massive upheavals of residence, and when external influences break down the traditional fabric of the state’s society. All of these aspects of a state have been examined from the point of view of vulnerability.

The focus of this study is on vulnerability of the environment itself, including physical and biological aspects of ecosystems, diversity, populations, communities and species. Unlike previously-developed indices, the focus is not on risks to humans or their property. Instead, we are concerned with the risks to the environment as an entity in itself and as the fundamental basis of human livelihoods. A state may be vulnerable environmentally if its ecosystems, species and processes are susceptible to damaging anthropogenic and natural pressures and these pressures are high.

Environmental vulnerability differs from economic or social vulnerability because: · “The environment” includes complex systems with different levels of organisation from species and physical features of the habitat up to entire interdependent ecosystems with a flow of organisms, energy and information between them and complex, often unpredictable and synergistic or antagonistic interactions between variables; · Quantitative data are often not available for describing these ecosystems, flows and processes; · Unlike human general indicators which can be used world-wide under the assumption that the needs of people and the thresholds for risk are similar (e.g. death toll, property damage, loss of shelter), environmental indicators may vary geographically, even within a species; · Economic indices can be expressed in money units, which can be translated as a world- wide comparable unit. Aspects of the environment represent very different processes which can not be expressed with one single unit. This means that developing an index for the environment will need a new approach.

1.3 The need for an environmental vulnerability index

113: Small Island Developing States, in co-operation with national, regional and international organisations and research centres, should continue to work on the development of vulnerability indices and other indicators that reflect the status of SIDS and integrate ecological fragility and economic vulnerability. Consideration should be given to how such an index, as well as relevant studies undertaken on SIDS by other international institutions, might be used in addition to other statistical measures as quantitative indicators of fragility.

114: Appropriate expertise should continue to be utilised in the development, compilation and updating of the vulnerability index. Such expertise could include scholars and representatives of international organisations that have at their disposal the data required to compile the vulnerability index. Relevant international organisations are invited to contribute to the development of the index. In addition, it is recommended that the work currently under way in the United Nations system on the elaboration of sustainable development indicators should take into account proposals on the vulnerability index.

There is a number of benefits that can be derived from the construction of a vulnerability index, including that: · The index can attract attention towards the issue of vulnerability of certain states and territories · It allows states to undertake self-assessment; and · It presents a single-value measure of vulnerability based on meaningful criteria which can be considered by donor countries and organisations when taking decisions regarding the allocation of financial aid and technical assistance, or for assigning special status to those states which are found to be the most vulnerable. 1.3 Aims of this study

The overall aim of this study was to develop the methodology for the construction of an environmental vulnerability index and to produce tentative vulnerability scores for a sample of countries, consistent with the Barbados Plan of Action and the needs enunciated by the Alliance of Small Island States (AOSIS). That is, to develop an index which summarises the vulnerability of the environment of states to natural and man-made hazards.

The environmental index developed in this study was not intended to describe the vulnerability of human systems or their responses to natural risks (as has been attempted in the past), but humans and their activities would be taken as part of the possible risks to the integrity of 'the environment'. This index could be combined later with an index which describes economic vulnerability so that a composite index which in a single figure incorporates economic concerns and the environment could be produced. It is envisaged that the EVI and any composite index would be recalculated for each state every 5 years to examine changes in the levels of risk and resilience through time as well as ranking the countries.

The specific aims of this project were to: · Review current work already completed or underway addressing environmental vulnerability of SIDS; · Build on past work on environmental vulnerability, if appropriate, or approach the problem from a new perspective where other attempts have had limited success; · Identify variables which may be used in the construction of an environmentally descriptive vulnerability index for Pacific SIDS; · Develop a logical framework and methods of calculating an index for environmental vulnerability; · Identify and collect data which would be used to calculate the environmental index; · Identify gaps in the available data; · Identify future directions for the further development of an internationally acceptable environmental vulnerability index; · Compile a report for widespread circulation and consideration prior to the Donor Round Table leading up to CSD-7 and the UN General Assembly Special Session on SIDS in September 1999; and · Further efforts towards the development of a composite vulnerability index as described in the Barbados Program of Action and meet the needs enunciated by AOSIS. 2 Mini review of previous work on vulnerability indices

2.1 Vulnerability indices in general

Considerable work has been carried out on the development of indices which compare the relative vulnerabilities of countries in terms of risks to human economic and social systems, notably by Briguglio (1995; 1997), Chander (1996), The Commonwealth Secretariat (Wells, 1996, 1997; and Atkins et al., 1998), Pantin (1997), and Adger (1996, 1998).

Of 15 studies on vulnerability indices reviewed in this project, only 5 attempt to describe effects of natural and/or anthropogenic risks on environmental variables. In contrast, 13 out of the 15 studies describe the effects of natural and/or anthropogenic risks on humans (see also Annex 3). The risks of concern also varied among the studies. Anthropogenic risks including those which affect economies and social systems and those which affect ecologies were of concern in the majority of studies (11 out of 15). Six of the studies were concerned with risks relating to climate change and sealevel rise, usually on both human and natural systems. Six out of the 15 studies saw natural disasters as one of the risks to which the human systems were vulnerable.

Only 1 study of the 15 examined (UNEP, 1998) looked specifically at the effects of both humans and natural disasters on the environment. It is the objective of this study to fill this gap, by developing a comprehensive environmental vulnerability index based on a wide array of environmental indicators which includes both anthropogenic and natural risks.

2.2 Environmental vulnerability indices

The most important studies on environmental vulnerability indices were those produced by Ehrlich and Ehrlich (1991), Atkins et al. (1998) and those which deal with sea-level rise and climate change (e.g. IPCC, 1991, 1992; Yamada et al., 1995; Sem et al., 1996) (see also Annex 3). An additional study being undertaken by the European Union Statistics Department (Eurostat, 1998) attempts to identify common environmental indicators which may be examined in all European countries to provide a comparison of pressures on the environment. Although these workers have not attempted (as yet) to aggregate their scores into an index, the approach they have taken so far is similar to the first stages of an environmental vulnerability index.

In addition to the above studies, is one undertaken by Pantin (1997). Although Pantin terms his work an "Ecological Vulnerability Index', the term is misleading. He has instead measured effects of natural and man-made disasters or problems (cyclones, earthquakes, anthropogenic sea-level rise etc.) on human systems. Pantin's study (1997) is really another type of human vulnerability index and does not address stresses or responses of the environment. 2.3 Summary of approaches and methods used in past vulnerability indices and lessons learned from them

2.3.1 Terminology

Most vulnerability indices developed so far are concerned with human vulnerability and do not estimate the vulnerability of the environment. There has been some confusion with terminology and we propose here that vulnerability indices should be clearly named and grouped so that they identify the responders (or exposure units) with which they are concerned, not the risks. This study is the first attempt to produce a true Environmental Vulnerability Index, for which the risks to and responses of the environment are the focus, whether of direct interest to humans, or not.

2.3.2 Framework

The logical framework for previously-developed vulnerability indices has tended to be poorly defined and developed. Largely this has been the case because the indices have dealt with relatively simple human systems the attributes for which are relatively well defined and understood. For the successful development of an environmental index, however, a logical framework is of central importance. Without one, the index will tend to be driven by data availability, terms will be poorly defined, the indicators being used may not appropriately describe the risks and results of environmental hazards and testing of the performance of the model will be impossible (no hypotheses to test).

2.3.3 Methodology

Number of indicators

The number of indicators used to quantify vulnerability fell into three categories. The majority of studies used only a small number of indicators (3-6), and these were usually the studies which focused on human systems, with only some reference to natural disasters. Four studies used a moderate number of indicators (12-15), which except for the UNDP Human Development Index (UNDP, 1998) tended to introduce more emphasis on natural disasters and ecological indicator variables (Yamada et al., 1995; Pantin, 1997; UNEP, 1998). Only one study used a large number of indicators (60). This was the Eurostat (1998) list of indicators for anthropogenic pressures on the environment and is the only study that specifically focuses on effects on ecosystem integrity. The lesson here is that more indicators are required when complex ecological systems become the focus of the index being constructed.

Methods for scoring

Various methods were used in previous studies for measuring vulnerability. The value of indicators was scored using one of the following techniques: · Raw value (not later aggregated into an index), Eurostat (1998); · Values transformed or expressed as %, fractions or additive scores to deal with dissimilar units (Pernetta, 1990; UNEP, 1998); · Numerical value standardised to a range of 0-1 using the formula:

(Xi - Min Xi)

Vi = ------

(Max Xi - Min Xi)

Where: Vij = measure of vulnerability contributed by the ith indicator in country j, Xij =

numerical value of the ith indicator in country j, Min and Max Xi = minimum and maximum value of the ith indicator across all countries being compared. (Briguglio, 1993, 1995, 1997; Chander, 1996; Wells, 1996, 1997; Pantin, 1997); · Indicator values mapped onto a categorical scale, e.g. -3 to 0 to +3 (Yamada et al, 1995, Sem, 1996); and · Conversion of all indicators to $ value (Yohe, 1991).

Most of these methods for scoring the values of indicators were developed to eliminate problems with heterogeneous data sets. That is, data which were expressed in different units (e.g. kg vs km2 vs mm) or which have different ranges (0-10 vs 560-1020).

Aggregating indicator scores into an index

The method used for aggregating the values (whether transformed as discussed above, or not) into indices also varied among studies. The main methods used were:

· No aggregation: individual indices presented separately (Eurostat, 1998); · Simple additive or multiplicative formula-style index with or without weightings (Pernetta, 1990; Ehrlich and Ehrlich, 1991; Yohe, 1991); · Numerical data for key vulnerability sub-indices assumed to represent the underlying

vulnerability factors (X1,X2,…X n), standardised to common units and averaged to obtain a composite vulnerability score (Briguglio, 1992, 1995, 1998; Chander, 1996; Wells, 1996; · Quantitative and qualitative data on vulnerability and resilience of elements of a system, standardised to common units the difference between which forms a measure of sustainable capacity (SCI) (e.g. Yamada et al. 1995); · Vulnerability was a priori assumed to be represented by an observable variable Y (e.g.

output volatility) and regressed against X1,X2,…X n (which represent the underlying

vulnerability sub-indices). The estimated coefficients on X1, X2, …. X n so obtained were then used as weights to aggregate the sub-indices (Atkins et al., 1998).

3 A theoretical framework for environmental vulnerability

3.1 The risks and ecosystem integrity

The maintenance of ecosystem or ecological integrity is at the heart of the development of a vulnerability index on the environment, because it is ecosystem integrity that is threatened by natural and anthropogenic hazards. The notion of ecosystem integrity is so complex that it cannot be expressed through a single indicator, but rather requires a set of indicators at different spatial, temporal and hierarchical levels of ecosystem organisation (Jones and Kaly, 1995; De Leo and Levin, 1997). Ecosystem integrity depends on biodiversity, ecosystem functioning and resilience, all of which are such interrelated variables, that factors which affect just one of these can have far-reaching ecosystem-wide consequences.

The first step towards the development of the EVI must be to clearly identify the risks we are concerned about and define what these risks are capable of affecting. In general, however, we will not be able to couple individual risks and effects in such complex interactive systems. The approach is this study was to examine these variables with proxy measures and indicators (see Section 4).

3.3.1 The risks

In a general sense, a risk may be defined as any event or process that can cause damage to the environment. These include natural and human events and processes such as 'the weather' and 'pollution'. For example, Campbell and Ericksen (1990) defined natural hazards as those in which natural environmental conditions depart from 'normal' to such an extent that people, property and social systems may be adversely affected. The problem with this definition is that unless we identify certain natural events as being anthropogenically altered (e.g. anthropogenically-accelerated sea-level rise), all events are 'normal'. For studies which do focus on effects on 'the environment' these natural hazards are also seen as risks to nature. The implication from this line of reasoning is that the changes we see to the natural world as a result of natural hazards are deemed 'unacceptable' from a human perspective. This means that except in the case of anthropogenic risks, in an assessment of environmental vulnerability, what we really are examining is unacceptable departures from our (human) view of how the environment should change. For the purposes of this study, we will accept that risk events should include those which cause sudden and seemingly- negative impacts on natural systems as a way to evaluate vulnerability. There is, however, increasing evidence that environmental upheaval is a natural and important part of ecosystem creation and maintenance (e.g. storms which remove coral communities and deposit them on islands and expand the land area while allowing space from where they were removed for other species to exist for some time).

Clearly, though most identifiable risk events are capable of causing damage, it is only the larger and more intense events that are likely to cause wholesale changes in the environment, at least in the short to mid term. A meaningful EVI would tend to focus on the more important events, given that quantification and even identification of all events is impossible. The following is a list of some of the more important risks which can impact on the environment:

1. Meteorological events: Cyclones, storms, surges, droughts, floods, heat waves and cold snaps, the ENSO phenomenon, tornadoes; 2. Geological events: Landslides, earthquakes, tsunamis, volcanoes, subsidence, erosion and accretion, altered tidal range; 3. Anthropogenic impacts: Exploitation of resources (e.g. mining, hydrocarbon extraction and use, fisheries), habitat destruction, human population pressure, inappropriate environmental management, developments which affect coastal processes, pollution, toxic wastes, solid wastes, urbanisation, agriculture and aquaculture, tourism, wars and civil strife. 4. Climate change: Warmer atmospheric and oceanic temperatures, changing rainfall patterns, increased incidence of extreme events (e.g. changes in frequency or increased intensity of tropical cyclones), changes in wave patterns, extinction of species unable to adapt to habitat and related changes (as a risk to ecosystem function), disruption of ecosystems, ozone depletion; 5. Sealevel rise 6. Astronomical events: Solar flares, astronomical low tides

3.1.2 The entities at risk: 'Responders'

Defining the environment which is at risk, or the 'responders', is a more difficult task and is an issue which has largely been side-stepped by most workers in the field of vulnerability. In this study, we define the environment at risk to mean all of the physical, biological and process elements of the natural world excluding humans and their structures. This includes: · Ecosystems (identifiable groupings of organisms and their habitats) · Habitats (the places in which organisms live) · Populations and communities of organisms (identifiable groups of organisms that interrelate) · Physical and biological processes (beach building, reproduction, recruitment) · Energy flows (nutrient cycling and import/export) · Species (losses of particular species e.g. mangroves can redefine the ecosystem) · Diversity (includes geographic, ecosystem, community, population, species and genetic diversity) · Ecological resilience (the ability of ecosystems to 'bounce back' after being disturbed) · Ecological redundancy (species which carry out similar functions in an ecosystem)

All of these aspects of the environment may be subject to alteration as a result of action of any of the risks identified above. It is important to note also, that complex relationships exists between the risks and the environment affected. That is, the environment is not just subject to a risk, but may modify its action either during a given event or at a later stage. For example, a cyclone may change the shape of a beach by dumping new material on it and the new material acts as a better barrier to later cyclone damage (e.g. cyclone Bebe in 1972 created a large storm bank on Funafuti).

3.2 The three aspects of environmental vulnerability

Environmental vulnerability of an entity such as a state, may be viewed as having three aspects, namely: 1. The level of risks (or pressures) which act on the environment within the state; 2. Intrinsic resilience of the environment to risks; and 3. Extrinisic or resilience as a result of external forces acting on the environment. All of these three aspects of vulnerability need to be incorporated into an EVI if we are to obtain an overall picture of the proneness to environmental changes due to risks. That is, vulnerability is made up of risk pressure (REI), intrinsic resilience (IRI) and extrinsic resilience (EDI), so that EVI=REI+IRI+EDI. These are each discussed in detail below.

3.2.1 Risk exposure as part of vulnerability and the REI sub-index

Risks are largely considered external forces which act on the environment (accepting that feedback effects can and do occur). A measurement of the amount of risk expressed as the frequency and expected intensity of risk events (exposure) is necessary within the EVI. A measurement of risk exposure allows comparisons from state to state (and time to time) of the amount of hazardous events likely to impact on the environment at any one time.

We did not attempt to extrapolate the risk levels likely to affect the environment into the future, by measuring the recent past levels of risk. This approach has two advantages: 1. It does not rely on complex mathematical modelling of risk events which requires a lot of data not currently available; and 2. By using relatively recent data (for most risks, though not those which operate on geological time scales) there is an opportunity for the EVI to change with each update (every 5 years) as levels of risk exposure change. This would be particularly important for anthropogenic risks (pollution, population pressure) and for those which may be responding to longer term cycles (such as climate).

For the EVI we calculate a 'Risk Exposure sub-Index' or REI which examines the frequency and where possible, the intensity of risk events which may affect the environment. These are based on levels observed over the past 5-10 years for most risks, but may include data for much longer periods for geological events. The REI is a measure of potential risk only; there is no logical expectation that patterns of risk expression during the immediate history of a state will necessarily result in similar risk levels today or in the future.

3.2.2 Intrinsic vulnerability / resilience and the IRI sub-index

Intrinsic resilience refers to the innate ability of natural systems to maintain their integrity when subject to disturbance (Holling, 1973; Ludwig et al., 1997). Conversely, intrinsic vulnerability is the innate fragility of a system. It is an expression of relative natural immunity to hazards.

For most environmental systems, we do not know what this natural immunity to hazards is. There are insufficient data, for example, to describe the ability of a reef to withstand a cyclone of, say, Category 2 if it were to hit directly and take 6 hours to pass. Predicting which ecological variables (e.g. species, processes) might be affected and what effect this would have on ecosystem diversity, function and future resilience is at present not possible. For the purposes of calculating an EVI we have focused on broader indicators which will give us approximate measures of resilience at the scale of entire states. The indicators used which refer to characteristics of a country which would tend to make it less/more able to cope with natural and anthropogenic hazards form the IRI (Intrinsic Resilience sub-Index). For example, the absolute land size of a state is used assuming that the larger, the more resilient because refuges from hazards are more likely and recolonisation after disturbance will be possible from these refuges. Large absolute size also means that only a small portion of the state might be affected by any single hazard event.

Clearly, indicators which show natural rates of recovery or productivity should also be included, but data for these measures are generally lacking. An exception might be measures of productivity being estimated in tropical Papua New Guinea mangroves (Robertson et al., 1991) and estimates of the tonnages/km2 of reef fishes produced in the Pacific Region (Dalzell et al., 1996). The underlying assumption if these indicators were included in the IRI would be that greater productivity rates can lead to faster rates of recovery and hence the ability for ecosystems to withstand greater disturbance. 3.2.3 Extrinsic vulnerability / resilience and the EDI sub-index

Extrinsic vulnerability / resilience refers to the ability of ecosystems to continue to maintain their integrity after already suffering impacts from the same or other hazards. We have assumed that the greater the number and intensity of hazards which have impacted on a system, the greater its level of vulnerability to future stresses is likely to be. Because neither the natural resilience nor the altered resilience of any ecosystem (impacted by, say, even a single cyclone) is known, let alone the resilience which might arise as a result of summed or interactive effects, it is impossible to estimate extrinsic vulnerability. Instead, we have opted for a proxy measure of extrinsic vulnerability for inclusion in the EVI. The indicators we have chosen attempt to describe the ecological integrity or level of degradation of ecosystems. This forms the Ecosystem Degradation sub-Index (EDI). The more degraded the ecosystems of a country (as a result of past natural and anthropogenic hazards), the more vulnerable it is likely to be to future risks.

3.3 Features of an environmental vulnerability index (EVI)

In developing the EVI, we set criteria on certain features of the index to ensure that it would be able to perform the tasks for which it was developed. All of the following criteria were built into the EVI model. The EVI index should be:

1. Intuitively comprehensible: That is, the final value of the EVI and its sub-indices should be expressed on a scale which has immediate recognition for users. We have set the range of values between 0 and 7 with higher scores indicating higher vulnerability; 2. Impartial: That is, it should provide an unbiased measure of vulnerability of states to real natural and anthropogenic risks; 3. Suitable for international comparisons: The index should encompass the range of variables found in different countries and their extremes of occurrence; 4. Able to differentiate among countries: The index had to provide a spread of values so that differences among countries would be highlighted; 5. Applicable at different spatial scales: This would allow for regional and country comparisons within the international community, as well as comparisons within countries for identifying areas of weakness at the individual government level; 6. Refinable through the indicators used but not directly by individual countries: This would make the index adaptable so that it can be applied globally and can be upgraded as additional data come to light. Limiting changes to the EVI administrator and technical review, ensures a level basis for comparisons among countries is maintained; 7. Presented in breakdown and single figure formats: The EVI is a composite index formed by combining the REI, IRI and EDI sub-indices which themselves can be separated into anthropogenic and natural hazards, risk and mitigating factors and ecosystem and biodiversity effects. By providing a breakdown of each of these, a clearer understanding of the nature of national vulnerabilities is possible. This simultaneously allows overall comparisons and the identification of specific problem areas; 8. Calculated using a user-friendly computer interface: This reduces the possibility of user errors and speeds the process of accurate calculation of the EVI.

3.4 Why use indicators?

As was noted in Section 3.1 above, examining ecosystem integrity is a complex business and cannot be measured directly or expressed through a single indicator. Neither can risk to the environment be measured as an absolute amount since many potential risks may apply, not all of which are identifiable such as low-level or diffuse impacts (e.g. non-point source impacts such as run-off of pesticides). Risks can also apply at different places, times and intensities.

In trying to determine how to measure and manage ecosystem integrity, De Leo and Levin (1997) identified reductionist and wholist approaches to examining ecosystems. A reductionist approach emphasises the structural aspects of natural systems and focuses on individual species and population dynamics of species within isolated ecosystems. A wholistic approach focuses on macro-level functional aspects such as energy flows, nutrient cycling and productivity. These structural and functional aspects of ecosystems tend to lead to different definitions of ecosystem integrity: 1. Focus on structural aspects (reductionist approach) leads to a definition in which the loss of even one species or the damage of a link between some components implies a loss of integrity because the ecosystem is no longer complete; 2. Focus on functional aspects (wholistic approach) leads to the conclusion that the loss of some species may not be important because redundancies within functional groups will maintain functional integrity (De Leo and Levin, 1997). We have attempted to include both aspects of ecosystem integrity in our indicators of environmental vulnerability.

The evaluation of environmental vulnerability requires the use of a broad base of indicators targeted at each of the components of vulnerability (the REI, IRI and EDI) and at different spatial, temporal and hierarchical levels of ecosystem organisation (Jones and Kaly, 1995; De Leo and Levin, 1997; Kaly and Jones, 1998). At the scale of an entire country, this is not an easy task. Ideally, indicators are required which describe: · Frequency and intensity of the most important risks; · Intrinsic vulnerability / resilience to risks such as characteristics of a country that render it susceptible to hazards (such as anthropogenic sea-level rise); natural rates of regeneration or productivity which make it likely to recover from disturbances quickly and more completely before the arrival of the next hazard etc; · Ecosystems: Loss of habitats, keystone species, ecosystem functions, goods and services; · Groups of organisms: Loss of diversity, populations of organisms and genetic diversity, and ecological redundancy; · Elements of the physical environment (water, storm banks, coastal processes, flood plains, lands close to sea-level); · Rare and endangered species and those of economic importance which may be more than normally targeted by human activities; and · Mitigators of the effects of hazards such as legislation which modifies human risks, monitoring programmes which provide early warning of ecosystem damage.

Attempts were made in this study to include a large number of indicators from each of these categories, though availability of data tended to set a practical limit to the number finally included in the EVI model.

4 Methodology used

4.1 Indicators of vulnerability

4.1.1 Criteria for the selection of indicators

With the preceding theoretical arguments in mind, the indicators for the calculation of the EVI and its sub-indices were finally selected on the following criteria:

· Applicable over different scales. Or they should at least be calculable over the entire scale of interest (the default being an entire country, but applicable also to regions or within countries); · Spread over the different geographic, habitat and climatic types (e.g. tropical, temperate, terrestrial, coastal, marine); · Relatively easy to understand; · Unbiased; · As well-defined as possible so that data are comparable and measure the same variable from country to country and operator to operator; · Spread over different levels of organisation (ecosystems, biodiversity, processes); · Data available and relatively reliable and collected as a matter of routine by authorities in a country. Data available in existing or on-going publications were preferred if they also met the other criteria (Section 4.1.3); · Data which should be available if a consultant could go to the country to assist in its collation or for which programmes directed at its collection or collation could be proposed and implemented; Indicators for which data that were considered unlikely to be available under each of the above two criteria were omitted from the model (Section 4.1.3); · The indicators should be as unrelated as possible to each other. Redundant indicators do not add much additional information to the EVI (though for some indicators, this criterion may only be assessed by mathematical testing - expressed as correlations in the data).

4.1.2 Indicators for the three sub-indices

Although a larger number would have been preferred to obtain a wider picture of risks, resilience and ecological integrity, many of the indicators initially selected were discarded because they either:

1. Had no data available and data were unlikely to be procured in the near future (see also Section 4.1.3); 2. Were ambiguous or bimodal in their responses; or 3. Were redundant and the information they intended to capture was present in another indicator.

No data

For example, catch-per-unit-effort data on fisheries, percentage of coastal areas eroded, tonnes of plastics produced or imported yearly, numbers of hydrocarbon spills and carriage of toxic wastes through territorial lands or waters would have all provided important indicators of risks and ecosystem integrity. All of these indicators and others like them were discarded because data were considered impossible to obtain for most countries. Bimodal response or ambiguous

The percentage of land area urbanised in contrast to decentralised was considered as a potential indicator of general human impacts and the amount of wastes and toxic compounds which would have to be absorbed and rendered harmless as an ecosystem service. It is not however clear which option represents the greater risk to the environment. In highly urbanised areas, impacts are concentrated and absorption of wastes and other substances spreads from the urbanised centre to the surrounding ecosystems often resulting in severe pollution and localised impacts or stress. In areas of low human population density, the concentrations of wastes is low, but large areas of land are disrupted by clearing, farming and other activities leading to habitat destruction.

Redundant

Production of hydrocarbons and deviations in rainfall patterns during El Nino events are examples of two indicators discarded because they are represented in other indicators which were included in the model. Although the indicators differed in their content, it was considered that the risks they represent (such as oil spills, production of greenhouse gases, floods or droughts) were already present in or too correlated with an indicator concerned with hydrocarbon usage and months with greater and less than average rainfall respectively.

4.1.3 Collecting the data on indicators

There are four categories of data available for calculating the EVI: 1. Data which are easily available and published by reputable international organisations as a matter of routine; 2. Data which are not published by reputable international organisations, but are collected by the respective governments or could easily be collected or reasonably approximated, in a matter of weeks, without the assistance of a suitably-trained consultant; 3. Data which are difficult to obtain, but could be produced or reasonably approximated in a matter of months with the assistance of suitably trained consultants; 4. Data which are very difficult or impossible to obtain or reasonably approximate, even with the assistance of a consultant. We focused on data in categories 1 and 2, and with sufficient funding available could also include data category 3. Data in category 4 were not included.

4.2 Quantifying vulnerability

Our overriding principle in constructing a vulnerability index for the environment was one of parsimony. That is, despite the inherent complexities in trying to describe the risks to and effects on integrity of the environment, we did not introduce complexities in calculating the EVI unless there was a justifiable reason to do so.

4.2.1 Mapping of data on indicators on a 0 - 7 scale

Mapping data

The indicators incorporated into the EVI model were of a heterogeneous nature. For some, responses were qualitative and took the form of 'yes' or 'no' answers or graded from 'none' to 'some' to 'a large amount'. For others, numerical data were available which could have been used in their raw state. But even for the numerical data, scales were heterogeneous occurring on a sliding linear or non-linear scale or having different maximum and minimum values. To deal with this heterogeneity, we chose to map the possible responses to each indicator on a simple scale from 1-76.

The 1-7 scale was chosen because it allows for a reasonable amount of spread among the possible values of the data. This is important because one of the central aims of the EVI is to provide spread among states in terms of their vulnerabilities: a scale which is too compressed would make the creation of spread difficult. The approach permits the processing of binary data, where only a 'yes' or 'no' answer is possible. In this case a 'yes' answer could be assigned the maximum value of 7 and a 'no' answer the minimum value of 1, or some values in between. The scale of 1-7 also has a central score which means that the well understood concepts of average, maximum and minimum can be used to anchor the responses for non-numerical data as in the following example:

SCORE

1 2 3 4 5 6 7 The lowest Significantly Slightly less Average Slightly more Significantly The highest incidence less than than than more than incidence possible average average average average possible

6 There was some discussion as to the appropriateness of the mapping scales of 0 – 7 and 1 – 7 which will be resolved in further stages of the EVI development. The 1-7 scale also allows for non-linear and discontinuous scoring. Individual indicators can be modelled to approximate exponential, quadratic or other non-linear functions. A few examples of the versatility of this scoring system are shown in the table below.

· Indicators for which effects are assumed to increase or decrease in direct response to their frequency or intensity were scored using the simple Linear Effect model. In this case, seven equal divisions are made and these are mapped directly to the 1-7 scoring scale; · Indicators for which effects at large intensities or frequencies reach a certain threshold and tend not to change much past that threshold were scored using the Diminishing Marginal Effects model. That is, indicators for which there is a decreasing rate of effect with increasing intensity or frequency; · The Increasing Marginal Effects model could be used when there was an increasing rate of effect with increasing intensity or frequency of the indicator; · The S-shaped Effect was also considered in our model, but not finally used. In this model, both ends of the indicator scale are associated with large effects, while intermediate values show little changes in response. · The Discontinuous and Part scales were used as special cases of each of the above, to truncate an indicator at a value thought to represent the point at which it has a major effect on the vulnerability scale.

Examples of the possibilities, assuming that the observed values range between 1 and 70:

SCORE Shape of underlying curve 1 2 3 4 5 6 7 Linear effect 1-10 11-20 21-30 31-40 41-50 51-60 61-70 Diminishing marginal effect 1-18 19-29 30-39 38-48 49-56 57-63 64-70 Increasing marginal effect 1-5 6-12 13-21 22-32 33-44 45-57 58-70 S-shaped effect 1-18 19-29 30-39 40-42 43-46 47-56 56-70 Discontinuous <2 2-25 26-40 41-60- 61-70 Part scale >20% 11-20% 5-10% <5%

The scoring methods used involved a degree of subjectivity since the score will depend on the assumed shape of the underlying relationship. On the positive side, however, this method allows for mapping responses of indicators that when increased by X times increase risk by less or more than X times.

A special case for non-applicable or no data responses: The values of 0 and N

Although active scoring of indicators occurs between the values of 1 to 7, we reserved the score of 0 for cases in which a question was considered not applicable for a state The value N was used for cases in which it was considered necessary to remove an indicator's signal from the index being calculated. This was done when there were no data available or, for one index type, when an indicator was considered not applicable. Please see Section 4.2.3 for full explanation of the two indices this leads to.

4.2.2 Setting the response levels for indicators

Response levels for each of the indicators selected were set wherever possible using the technical literature or by consultation with generalists and specialists in each field. Some levels were set using the data from Tuvalu, Fiji and Australia where available, or as estimates when these sources were unavailable.

4.2.2 Weighting

Six of the 57 indicators were assigned an intrinsic weighting factor of 5, while the remaining indicators were given the default weighting of 1. This is the equivalent of giving the six weighted indicators the equivalent value of 5 indicators. This weighting was applied to indicators considered to be of central importance to the question of vulnerability in the model. For example, total land area, percentage of forests remaining and human population measures were all considered key indicators which impacted significantly on the vulnerability of a state.

To ensure that the final EVI, REI, IRI and EDI scores remained between the values of 0 and 7, it was necessary to adjust the weighting factors by dividing the intrinsic weighting value (1 or 5) by the average of all weighting values within each sub-index. The 1-7 scores were then multiplied by the adjusted weighting factor prior to accumulation in the sub-index to which they belonged.

4.2.3 Calculating the sub-indices and the EVI

The EVI and sub-indices were calculated using an EXCEL workbook. The workbook (Version 7-EVI-calculator.xls) is comprised of seven linked worksheets, each dealing with a different aspect of calculation and reporting as follows: 1. Questions table: The indicators and their categorisation and levels; 2. Input screen: Responses to the indicator questions and the confidence in the data are entered here; 3. Response matrix: Results of scoring and adjustments due to weighting are calculated automatically; 4. Lookup help: Definitions of terms used; 5. Report Level 1: Highest level of reporting. Overall EVI and sub-indices calculated for two types of index (see below). Confidence in the data is also reported here; 6. Report Level 2: Breakdown of the REI and EDI sub-indices showing relative contribution of meteorological, geological, anthropogenic risks and mitigating factors and ecosystem and biodiversity indicators; 7. Report Level 3: Adjusted and raw scores are reported for each individual indicator.

A separate copy of the calculator is required to evaluate the vulnerability indices for each country.

Nett vulnerability - NA is blank. In this type of index, any indicators which were considered not applicable (scored as NA in the EXCEL calculator) were completely removed from the calculation of the indices. In EXCEL this was simply accomplished by assigning these indicators the value 'N' instead of a score between 0 and 7. This type of index measures the vulnerability experienced within a state. That is, the vulnerability of the state relative to those indicators which actually occur there. For Tuvalu, the question on volcanoes was scored NA because there are no active volcanoes in the country. The indicator volcanoes is irrelevant to Tuvalu and is not part of the suit of risks which can affect the country (except through indirect effects though gases and ash from an eruption elsewhere in the region).

Gross vulnerability - NA is 0. In this type of index, indicators which were considered not applicable were scored a 0 value. This means that the indicator contributed to the denominator of the averaging fraction, but not the numerator. This measures vulnerability in a total sense: Although Tuvalu has no volcanoes, volcanoes are considered part of the risks available in the world, and a zero score here shows that Tuvalu's risk to volcanoes is zero. Using this measure tends to increase the index value for countries which span large geographic and climatic ranges because the numerator is more likely to accumulate scores for a common denominator in all countries (not withstanding indicators for which data are not available). However, this will tend to be balanced by the fact that scores in such countries are averaged over their entire areas, including areas which might not be affected by the risk that the indicator describes, tending in turn to reduce the vulnerability value.

7 There was discussion as to which index is most appropriate for international inter-country comparison. This will be discussed further in the next phase of EVI development. 4.2.4 Confidence in and source of the data

In parallel with scoring each indicator against the 1-7 scale, we built into the EVI a way of assessing the reliability of data. These reliability values are reported alongside each index and should be read with them (like a mean with its standard deviation). In the example below, the Nett and Gross EVI and sub-indices are all reported along with the number of indicators available in each category (fixed); the number of indicators which were not applicable, the number with no data; the number of responses which were based on real data; and the number of responses based on 'best guess' or estimated by the operator and/or authorities. Nett Gross available # No Data # best guess # Not applicable # based on data Total # indicators

Environmental Vulnerability Index (EVI): 4.45 4.10 57 8 2 30 17

Risk Exposure Sub-index (REI): 3.98 3.22 39 7 2 18 12 Intrinsic Resilience Sub-index (IRI) 5.56 5.56 5 0 0 3 2 Environmental Degradation Sub-index (EDI): 3.82 3.52 13 1 0 9 3

5 Preliminary results for Australia, Fiji and Tuvalu

To provide an initial test of the EVI, we chose three countries to encompass a range of climatic, geographic, environmental and human systems. Australia, Fiji and Tuvalu were selected because data for them was relatively easy to obtain. It is expected that coverage will be extended to others in the Pacific Region, and later to all countries.

The preliminary results presented below should not be read as final EVI values for the three countries. We were unable to obtain all of the required data for all of the countries within the timeframe of this report (partly because it was the Christmas/New Year period and local authorities were unavailable for providing data). We are, however, confident that the data can be obtained. We had to proceed with the data we were able to obtain for the purposes of this report. 5.1 Values obtained for the EVI and sub-indices

5.1.1 Index values

The EVI value for Tuvalu was the highest of the three countries (5.04nett), indicating that its

environment is the most vulnerable. The score obtained for Fiji (3.79 nett) was intermediate in

value, and that for Australia was the lowest (3.04 nett), though the difference between Fiji and Australia was relatively smaller than that between Fiji and Tuvalu (Table 1 and Figure 1).

When the EVI was decomposed into the REI, IRI and EDI sub-indices, a more complex pattern emerged. The risk exposure index (REI) was similar for the three countries, varying

only between 3.13 nett for Australia and 3.49 nett for Tuvalu (Table 1). The intrinsic resilience index (IRI) varied the most among the countries with Australia having the best resilience

score (1.00 nett) and Tuvalu having the worst (7.00 nett). An almost reverse pattern was seen for the EDI. In this sub-index, the highest score (worst conditions) was obtained by Australia and the lowest by Fiji (Figure 2). The values of the sub-indices suggests that the different aspects which form vulnerability can operate independently of one another. Tuvalu was characterised by poor intrinsic resilience, and moderate risks and degradation. In contrast, Australia was characterised by very good intrinsic resilience, but high degradation.

When the sub-indices were further decomposed into different categories of risks, it was found that Fiji had the greatest vulnerability to meteorological and geological events, and Australia the greatest risk to pollution (Table 1).

Table 1: Summary of results of calculating values for the EVI and sub-indices for Australia, Fiji and Tuvalu. Values are given for Nett and Gross index values and a categorical breakdown of the REI and EDI. The total number of indicators available for each index is given. NA=Number of indicators which were not applicable for the country (assigned no value in Nett and a 0 value in Gross indices); ND=No data currently available (assigned no value).

AUSTRALIA FIJI TUVALU # Indicators Nett Gross NA ND Nett Gross NA ND Nett Gross NA ND EVI 57 3.04 3.04 0 38 3.79 3.68 1 17 5.04 4.91 2 20 REI 39 3.13 3.13 0 25 3.48 3.48 0 7 3.49 3.39 1 15 Metereological 6 1.30 1.30 0 5 2.55 2.55 0 0 1.31 1.33 0 5 Geological 4 0.65 0.65 0 2 3.27 3.27 0 2 - 0.00 1 3 Anthropogenic 29 3.70 3.70 0 18 3.73 3.73 0 5 3.58 3.64 0 7 Habitat 1 - - 0 1 - - 0 1 0.66 0.67 0 0 Agriculture 5 3.59 3.59 0 3 4.61 4.61 0 2 1.53 1.56 0 2 Fisheries 5 1.30 1.30 0 4 2.18 2.18 0 1 1.64 1.67 0 3 Government 4 0.87 0.87 0 1 4.12 4.12 0 1 4.11 4.17 0 0 Mining 3 - - 0 3 2.91 2.91 0 0 0.66 0.67 0 0 Pollution 8 3.91 3.91 0 6 2.36 2.36 0 0 2.41 2.44 0 2 Risk Factors 31 4.03 4.03 0 20 3.41 3.41 0 5 3.37 3.22 1 14 Risk mitigating Factors 8 0.65 0.65 0 5 3.76 3.76 0 2 3.76 3.81 0 1 IRI 5 1.00 1.00 0 3 4.67 4.67 0 3 7.00 7.00 0 1 EDI 13 5.00 5.00 0 10 3.22 2.90 1 7 4.63 4.35 1 4 Ecosystems 10 7.00 7.00 0 9 3.52 2.85 1 6 5.50 4.99 1 3 Biodiversity 3 4.00 4.00 0 1 2.78 3.00 0 1 2.00 2.12 0 1 Figure 1: Graph of the overall EVI results obtained for Australia, Fiji and Tuvalu (Nett and Gross scores).

EVI Nett Gross

3 Index value 2

0 Australia Fiji Tuvalu

Figure 2: Breakdown of the EVI scores for Australia, Fiji and Tuvalu into the three sub-indices for Nett scores only.

REI Sub-indices (Nett) IRI EDI 7

3 Index vale 2

0 Australia Fiji Tuvalu

5.1.2 Nett and gross vulnerabilities

The values obtained for nett and gross indices differed little. The greatest difference observed between the two methods of scoring was 0.13 on the 0-7 scale. Although this suggests that there may be no need to separately calculate the two index types, the present study did not provide a good test. There were very few indicators which were not applicable across all three countries, so the difference between the scoring types did not come into effect. It is expected that the nett and gross index values will be required when the EVI is extended globally and a greater range of climatic and geographic areas are built into the model. 5.1.3 Confidence in the results and data

As was stated above, these results are only very preliminary. The EVI and sub-indices will only provide a reasonable measure of vulnerability if most of the indicators can be filled-in for a country. We suggest here that at least 80% of the indicators (46 of the 57 questions) should be filled out by any one country for a reasonable estimate of vulnerability. In this report, we were unable to reach this threshold for any country. For Australia, Fiji and Tuvalu we were able to find data for 19, 40 and 37 indicators, respectively. We are aware that data are available for the remaining indicators, but are buried, or need to be requested from authorities and will take some time and effort to procure. The shortage of data means that all the EVI values obtained here are indicative: they are by no means completed estimates and should be read with caution.

We also had some problems with the quality of data. For example, whilst looking up values for area of land and sea for each of the countries, we found widely varying estimates in published sources. It will take some time to determine which are the correct estimates. Other problems with the data were that they were expressed in different ways in the different countries, or accumulated under different conditions (e.g. the definition of a 'drought' in Australia is not the same as in Fiji). Our partial solution to this problem was to design questions for the indicators which were independent of local definitions. For some indicators, it will be necessary to ask for the data to be supplied in different forms to those now provided by the relevant authorities.

5.2 Conclusions

The EVI model gives single-figure measures of environmental vulnerability that appear to be able to distinguish countries and identify sources of vulnerability within a country. For the moment, it appears that of three countries tested, Australia is the least vulnerable, and Tuvalu the most. There were similar levels of risk in each of the countries, the most degradation in Australia and the least intrinsic resilience in Tuvalu. These results are only preliminary because there was insufficient time to collect all of the data required for these three countries and there were some problems with reliability of the data. We expect that each of these problems can be overcome and the results suggest that the EVI will be a useful tool for characterising vulnerabilities of states. 6 Strengths and weaknesses of the EVI

As for all methods of summarising and modelling data, the EVI developed here is associated with a number of strengths and weaknesses which must be understood for its proper application and use. It should be noted here that the weaknesses in this model are inherent and relate to the methods used. Any improvements on the methods and changes to the model will lead to a new suite of strengths and weaknesses which will also have to be understood for appropriate use of the model. This work was not a search for the 'perfect' vulnerability calculator since all approaches will have associated with them certain weaknesses. At best we can hope to minimise them.

6.1 Strengths

The EVI calculator developed here is based on a theoretical framework that prompted us to find indicators for all of the identified aspects of vulnerability. In addition, it is:

· Comprehensive in its scope including indicators from a wide range of the most important risks and measures of environmental resilience and integrity; · Able to incorporate qualitative indicators for which no numerical data are available; · Innovative, since it distinguishes between risk, health (integrity) and resilience; · Inclusive of two types of vulnerability. The nett and gross vulnerabilities simultaneously provide a basis for world-wide comparisons of states and an assessment of the real risks likely to affect a state; · Able to incorporate non-linear relationships between causes and predicted effects; · Able to prompt local environmental agencies to increase and improve data collection; · Allows states to undertake self-assessment and policy refinement regarding their own environmental vulnerability; and · So far it is limited more-or-less to the area of Oceania. This is reflected in the present choice of indicators. If the project is extended world-wide, other indicators can be easily incorporated into the model to encompass the additional risks and other indicators which had no relevance for the region (e.g. risks to frosts, ice and snow avalanches).

6.2 Weaknesses

The major identified weaknesses of the EVI model developed here are:

· The index does not rely exclusively on published official data, resulting in relatively high cost of obtaining data and omissions. This means that it will initially require the presence of a consultant in developing countries to obtain some of the data which would otherwise not be easily accessible. The index is subject to problems with different in-country users responding to the indicators with their own interpretations. This could be minimised if in- country training is carried out to assist users the first time the index is calculated for that country; · There is some subjectivity in assigning weights and non-linearities to the scores. Also, long and short term effects have not been differently weighted; · The mapping of data on a seven point scale may result in loss of detail compared with using numerical data directly; · As for all indices of this kind, the index is driven by the indicators chosen. That is, a different set of indicators might result in very different vulnerability profiles and rating of countries. The index is also to some extent driven by availability of data. Some indicators had to be discarded due to a lack of data availability. In addition, while assessing the exposure and integrity of a country, local variations have not been taken into account; · Intrinsic vulnerability is not well represented due to a lack of data on carrying capacity, productivity and energy flows. Years of research are required on a range of ecosystems before these data become available; · Gross vulnerability includes signals for indicators which do not occur in countries. This tends to give lower vulnerability scores for large countries which span large geographic and climatic ranges because they are more likely to accumulate values in their numerators for a common denominator in all countries (except for indicators for which data are at present unavailable). This might be balanced by the fact that scores in such countries are averaged over their entire area, which might include areas not affected by the risk that the indicator describes, giving an apparently less vulnerable profile.

7 Conclusions and future directions

7.1 Promising results

The results described in the previous section show that it is possible to obtain a single figure measure of environmental vulnerability. This measure is a composite of the risks, intrinsic resilience and health or integrity of the environment. This study also shows that data which were previously thought to be difficult to obtain could be procured, albeit with the assistance of environmental consultants. The methodology selected in the computation of the index can produce results which could have operational usefulness for ranking countries according to their environmental vulnerabilities. This index highlights the need for governments to upgrade their collection and collation of environmental statistics. In addition, the breakdown of results into meteorological, geological, anthropogenic, mitigating and other categories of risk highlights areas of concern for environmental action. 7.2 Refining the EVI and extending it globally

The environmental vulnerability index developed here will require further refinement before it becomes fully operational. In a sense, determining the vulnerabilities of countries is always a question of examining them in the context of all possible conditions. Limiting the ranges of indicators and their response levels to the Pacific Region does not tell us much about how specific countries fare in a world-wide context. In this study we have tried to set the response ranges of the indicators selected to reflect this understanding. It has also been noted above, that the value of the EVI is dependent on the types of indicators selected and that the relative vulnerabilities of countries could change if different indicators were used. One of the most important tasks for developing the index will be to review each of the indicators and the response levels set to ensure that: · The EVI model is applicable to all geographic and climatic regions - mostly this would be achieved by the addition of new indicators; · The model may include adequate measures of risk and integrity of the environment, but more attention is required on measures of intrinsic resilience (e.g. productivity); · Response levels of the indicators include the entire range of possibilities for each risk or measure of resilience and integrity; and · Indicators for which data should be collected but are currently unavailable should also be identified.

One of the most important mechanisms by which the EVI may be refined is to subject it to an international 'think tank' and other forms of peer review. For the think tank, the aim is to assemble a small group of highly-specialised experts in the fields of statistics, biodiversity, biogeography, indices which summarise complex data, weather and climate, disaster research, ecosystem management, fisheries, forestry and productivity. It will also be necessary to carry out consultations with the countries themselves.

A second mechanism for refining the index will be to visit, say 10, countries to build their capacity to work with the EVI and collect data which at present may be available in forms not immediately useable for the EVI. For example, Tuvalu has only just begun recording its meteorological data on an electronic database (since October 1998), with all data prior to that being available only as hard copy data sheets. Data from additional developed countries should be available by request.

When sufficient data have been collected and indicators have been refined it will be necessary to test the performance of the model. This will tell us how the EVI performs for different types of countries and will highlight any biases that might be inherent in the EVI. Specifically, it will be necessary to: · Test for correlations among the indicators to remove those which are redundant and do not significantly add to our assessment of vulnerability; · Empirically test the model and indices and their ability to differentiate countries (create 'spread'); · Test the method for accumulation of scores and examine other possible options which might give better results; and · Assess the availability of data and confidence in the data procured.

To facilitate the use of the EVI by countries and regional managers, it will also be necessary to develop a user-friendly computer calculator. We have used an EXCEL workbook in this study to calculate the indices for Tuvalu, Fiji and Australia, but this is cumbersome and can easily be damaged by users. A calculator written in Microsoft ACCESS would provide a Windows-based calculator which will be easy to use and which cannot be damaged inadvertently by users.

The final stage in the development of the EVI will be to combine it with other indices, such as the Economic Vulnerability Index (Briguglio, 1993, 1995, 1997) to give an indication of overall vulnerability of states.

7.3 Institutional considerations

It is important that the Environmental Vulnerability Index be given some sort of 'ownership', so that its development can continue in a consistent and organised manner. The economic vulnerability index of Briguglio (1993, 1995, 1997) suffered in its development because it did not have a proper institutional framework at its inception making its subsequent refinement sporadic. There was a lack of consensus regarding the underlying methodological procedures that should be used for the economic vulnerability index.

The environmental vulnerability index has started off in an institutional framework within SOPAC at the technical level, and this augurs well for its consistent development. It should also be given a proper framework at the political level, and it is suggested here that it could be adopted by the Pacific Forum with the aim of strengthening its political profile.

As it continues to develop for other regions, other regional technical and political organisations could be invited to assist by computing the EVI in their respective regions, with SOPAC and the Pacific Forum remaining the co-ordinators of the process. 8 Conclusions

The construction of and environmental vulnerability index is possible. Despite certain inherent limitations (which are common to all indices) the results obtained in this study suggest that we should be able to distinguish among countries in terms of their relative vulnerabilities to natural and anthropogenic risks to the environment. To do this it is important that the index be expanded so that it is applicable on a global scale. It is only in the context of the global scale that the relative vulnerabilities of the SIDS can be assessed.

In addition to the spatial application of the EVI discussed in the previous paragraph, the EVI could also be applied temporally. That is, it could be recalculated through time to show changes in relative vulnerabilities in response to changes in the human population, legislative changes, climate change and other factors which might vary the levels of risk, resilience and integrity of the environment. We suggest that the EVI, once fully operational, should be recalculated for all participating countries every 5 years.

Much work remains to be done to refine this index to make it fully operational. Despite this, our original aim to show that the development of an EVI is possible has been achieved. This work, in addition to contributing to research on the development of country characterising indices in general, will serve to prompt governments to upgrade their capacity for the collection of environmental data. 9 Definition of acronyms and terms

Composite index An aggregated score of several indices or sub-indices Ecosystem goods Include tangible items obtainable within an ecosystem, such as fishes or sand

Ecosystem services Functions provided by an ecosystem such as biodegradation, absorption of toxins and CO2 etc. Ecosystem Organisms, their physical environment, and relationships between them and outside areas. Usually defined spatially. EDI Environmental degradation sub-index Exposure units Refers to human exposure to impacts. May be identified spatially at a number of scales - country level or island or community group. (Sem et al. 1996). In this report, this definition is extended to include exposure of all aspects of the environment. GIS Geographic information systems Index An aggregated score of sub-indices to give an overall measure of vulnerability of a state Indicator Ant variable which characterises the level of risk, resilience or environmental degradation in a state. IRI Intrinsic resilience sub-index Natural disaster A natural event of sufficient intensity which strikes natural systems or a human population that is vulnerable to it (e.g. Johnson et al., 1995) Natural hazards (i) Natural or man-made events that have a harmful effect on human beings and / or the environment or which present a risk to life or property (e.g. Johnson et al., 1995; American Geological Society, 1984, In: Alexander, 1990); (ii) A hazard may be regarded as a predisaster situation, in which some risk of disaster exists, principally because human society has placed itself in a situation of vulnerability’ (Alexander, 1990). REI Risk exposure sub-index Resilience The potential for attributes of a system to minimise or absorb the impacts of extreme events. In Sem et al. 1996, this is expressed as an integer value between 0 and +3. Risk Expected degree of loss or damage to natural or human systems (e.g. Johnson et al., 1995, p.2); Hazard x vulnerability (Johnson et al., 1995; Alexander, 1990) SIDS Small island developing states SOPAC South Pacific Applied Geosciences Commission, Suva, Fiji SPREP South Pacific Regional Environment Programme, Apia, Samoa Sub-index An aggregated separate score for indicators of risk, resilience or degradation as REI. IRI and EDI respectively VA Vulnerability assessment VI Vulnerability index Vulnerability The potential for attributes of a system to respond adversely to the occurrence of hazardous events. In Sem et al. 1996, this is expressed as an integer value between 0 and -3. 10 References

Adger, W.N. 1996. Approaches to vulnerability to climate change. CSERGE (Centre for Social and Economic Research of the Global Environment, UK) Working Paper GEC 96-05, 63pp. Adger, W.N. 1998. Indicators of social and economic vulnerability to climate change in Vietnam. CSERGE (Centre for Social and Economic Research of the Global Environment, UK) Working Paper GEC 98-02, 39pp. Atkins, J., Mazzi, S. and Ramlogan, C. 1998. A Composite Index of Vulnerability. Commonwealth Secretariat, London. Briguglio L. 1997. Alternative Economic Vulnerability Indicators for Developing Countries with Special Reference to SIDS. Report Prepared for the Expert Group on Vulnerability Indices UN-DESA, 17-19 December 1997. Briguglio, L. 1992. Preliminary Study on the Construction of an Index for Ranking Countries According to their Economic Vulnerability, UNCTAD/LDC/Misc.4. Briguglio, L. 1993. The Economic Vulnerabilities of Small Island Developing States. Study commissioned by CARICOM for the Regional Technical Meeting of the Global Conference on the Sustainable Development of Small Island Developing States, Port of Spain, Trinidad and Tobago, July 1993. Briguglio, L. 1995. Small Island States and their Economic Vulnerabilities. World Development. 23:1615-1632. Campbell, J.R. and Ericksen, N.J. 1990. Change, extreme events and natural hazards. In: Climate change impacts on New Zealand: Implications for the environment, economy and society. Ministry for the Environment, pp19-28. Chander 1996 - 43pp Chander, R. 1996. Measurement of the Vulnerability of Small States. Washington, 43pp. Dalzell, P, Adams, T.J.H., Polunin, N.V.C. 1996. Coastal fisheries in the Pacific Islands. Oceanography and Marine Biology Annual Review, 34:395. De Leo, G.A. and Levin, S. 1997. The multifaceted aspects of ecosystem integrity. Conservation Ecology 1(1):3. http://www.consecol.org/vol1/iss1/art3 Downing, T.E. 1992. Climate change and vulnerable places: Global food security and country studies in Zimbabwe, Kenya, Senegal and Chile. Research Report 1, Environmental Change Unit, University of Oxford. Ehrlich, P.R. and Ehrlich, A.H. 1991. Healing the planet. Addiison-Wesley Publication Co. Inc., Menlo Park, CA. Holling, C.S. 1973. Resilience and stability of ecological systems. Annual Review of Ecology and Systematics, 4:1-23. Houghton, J.T., Jenkins, G.J. and Ephraums, J.J. (eds). 1990. Climate change: the IPCC Scientific Assessment. Cambridge University Press, England. IPCC. 1991. The seven steps to the vulnerability assessment of coastal areas to sea-level rise - Guidelines for case studies. IPCC Report, 24pp. IPCC. 1992. Global climate change and the rising challenge of the sea. IPCC RSWG Report, 34pp. Johnson, R.W., Blong, R.J. and Ryan, C.J. 1995. Natural hazards: Their potential in the Pacific Southwest. Australian Government Publishing Service, Canberra, 60pp. Jones, G.P. and Kaly, U.L. 1995. Criteria for selecting marine organisms in biomonitoring studies. Pp. 39-56, In: "Detection of Ecological Impacts: Conceptual issues and application in coastal marine habitats", (Schmitt, R.J. and Osenberg, C.W., eds), University of California Press. Kaly, U.L. and Jones, G.P. 1998. Minimum designs for measuring ecological impacts on coral reefs. Proc. 8th Int. Coral Reef Symposium, Panama June 1996. Ludwig, D., Walker, B. and Holling, C.S. 1997. Sustainability, stability and resilience. Conservation Ecology 1(1):7. http://www.consecol.org/vol1/iss1/art7 Pantin, D. 1997. Alternative Ecological Vulnerability Indicators for Developing Countries with Special Reference to SIDS. Report Prepared for the Expert Group on Vulnerability Indices, UN-DESA, 17-19 December 1997. Pantin, D.A. 1997. Alternative ecological vulnerability indices for developing countries with special reference to small island developing states (SIDS). Report to UN Department of Economic and Social Affairs, 22pp. Pernetta, J.C. 1990. Projected climate change and sea-level rise: A relative impact rating for the countries of the Pacific Basin. In: Pernetta, J.C. and Hughes, P.J. (eds). Implications of expected climate changes in the South Pacific Region: an overview. UNEP Regional Seas Report 1990, p14-23. Robertson, A.I., Daniel, P.A. and Dixon, P. 1991. Mangrove forest structure and productivity in the Fly River Estuary, Papua New Guinea. Marine Biology, 111(1):147. Sem, G., Campbell, J.R., Hay, J.E., Mimura, N., Ohno, E., Yamada, K., Serizawa, M., and Nishioka, S. 1996. Coastal vulnerability and resilience in Tuvalu: Assessment of climate change impacts and adaptation. Integrated Coastal Zone Management Programme for Fiji and Tuvalu Phase IV, SPREP, EAJ, OECC Report, 130pp. UNCTAD. 1998. Specific Problems of Island Developing Countries. LDC/Misc/17. UNDP. 1998. Human Development Report. UNDP, New York, 228pp. UNEP Island Directory. 1998. Explanation of island indicators. http://www.unep.ch/islands/indicat.htm UNEP. 1998. Human development report 1998. UNDP Report. Oxford University Press, 228pp. United Nations - DPCSD. 1997. Vulnerability Index (Revised Background Paper). SD-SIDS Unit. United Nations. 1994. Report on the Global Conference on the Sustainable Development of Small Island Developing States, A/Conf. 167/9. Wells, J. 1996. Composite vulnerability index: A preliminary report. Commonwealth Secretariat, London. Wells, J. 1997. Composite vulnerability index: A revised report. Commonwealth Secretariat, London. Yamada, K, Nunn, P.D., Mimura, N., Machida, S. and Yamamoto, M. 1995. Methodology for the assessment of vulnerability of south Pacific Island countries to sea-level rise and climate change. Journal of global environmental engineering, 1:101-125. Yohe, G.W. 1991. The cost of not holding back the sea - Economic vulnerability. Ocean and shoreline management, 15:233-255. Annex 1 Terms of reference for this study

Objective

To pursue the development of an ecological/environmental vulnerability index consistent with the Barbados Plan of Action and the needs enunciated by the Alliance of Small Island States (AOSIS).

Background

Two particular paragraphs of the Barbados Plan of Action, paragraphs 113 and 114, refer to the need for a vulnerability index for Small Island Developing States (SIDS).

For us, the Barbados Plan of Action for the Sustainable Development of Small Island Developing States provides the basis and raison d’être for the Vulnerability Index. We need full and proper understanding of all the components and technical nature of vulnerability, and so that we can plan and seek from the international community vital support for our efforts at sustainable development for our part AOSIS has endeavoured for some time to spark some momentum into the work on the Index, though we were pleased that the fourth and fifth Sessions for the Commission on Sustainable Development specifically recognised the need to accord sufficient priority to it.

Given the manner to which vulnerability indices are being used in the international arena it appears critical that a single vulnerability index be developed expressing economic as well as environmental/ecological parameters as a matter of urgency.

In order to address this matter it appears apparent that there is a lack of adequate data for all SIDS for both economic and ecological/environmental (natural disasters) parameters which impact on an island economy. Also there is a lack of an acceptable methodology to combine the two sets of data into one composite vulnerability index. As a result there is a need to increase and strengthen all efforts to develop such an index.

The UN Expert Group in December 97 could not construct a composite index. It suggested retaining a vulnerability index based on economic parameters whilst agreeing that efforts to develop an environmentally sensitive vulnerability index continue.

New Zealand as current Chair of CSD has indicated its support to Pacific SIDS to pursue the development of an environmental vulnerability index by way of this current project being implemented by SOPAC. Support for this initiative was expressed by UNDP at the PIC Partners Meeting, 9th July 1998. The Forum Economic Ministers Meeting (8th July 1998) agreed to adopt a common Forum position with the objective of the UN adopting a vulnerability index, and with the aim of having such an index included among the criteria for determining LDC status, and for deciding eligibility for concessional aid and trade treatment.

Study Description – Strategy

1. Review and build on current work already completed or underway addressing environmental vulnerability of SIDS 2. Work with SPOCC regional agencies and others (internationally, regionally and nationally) to develop a report for New Zealand to table at CSD-7. 3. Carry out this study in conjunction with the GEO-2 exercise currently underway within the SPOCC agencies and being co-ordinated by SPREP. 4. Identify parameters which may be used in the construction of an environmentally sensitive vulnerability index for Pacific SIDS. 5. Source data. 6. Identify gaps in these data and develop concept papers for projects which focus on a common objective to ensure that these gaps are filled. 7. Convene a session on environmental vulnerability indices at the upcoming SPREP- UNEP Oceans meeting to enable widespread consultation with experts and country representatives. 8. Define future actions to: (i) further the development of an acceptable environmental vulnerability index for Pacific SIDS and (ii) further efforts towards the development of a composite vulnerability index as described in the Barbados Program of Action and meet the needs enunciated by AOSIS. 9. Compile a report for widespread circulation and consideration prior to the Donor Round Table leading up to CSD-7 and the UN General Assembly special session on SIDS in September 1999.

Reporting

Draft report to be submitted by SOPAC to New Zealand by 31st December 1998. Annex 2 The vulnerability indicators used

In this annexe, we provide a description, categorisation and the response levels set for the questions used in this study to measure aspects of vulnerability. For many of the risk indicators, the observed value is expressed as a ratio in relation to the area of land or sea available. The reasoning behind this is that it is thedensity of risks, not the absolute number that affects the environment. The ability of ecosystems to tolerate impacts depends on how much of the ecosystem is affected at any one time (allowing for refuges for recolonisation) and how much ecosystem is available to absorb the effects of a risk. For example 1 tonne of a pollutant spread over 100 km2 is expected to have a smaller detrimental effect than the same amount spread over 1 km2 (all else being equal) because the concentration of the pollutant will be lower and any toxicity thresholds are less likely to be exceeded.

Question number: 1 Sub-index: REI Categorisation: Meteorological Factor type: Risk Factor Intrinsic weighting: 1

Deviation in average sea temperatures during moderate or greater El Nino (NOAA) (oC)

Scoring levels: 1 2 3 4 5 6 7 0 0.1-1 1.1-2 2.1-3 3.1-4 4.1-5 >5

Question number: 2 Sub-index: REI Categorisation: Meteorological Factor type: Risk Factor Intrinsic weighting: 1

Number of months over last 5 years during which rainfall is more than 20% above 30yr average for that month (flood risk)

Scoring levels: 1 2 3 4 5 6 7 0 1-10 11-20 21-30 31-40 41-50 51-60

This question examines the risk to flooding and other effects associated with high rainfall from the perspective of ecological systems. Greater than average rainfall can affect reef areas by freshwater and silt inputs, cause damage to rivers and deltas and flood inland areas, all of which can result in ecological disturbance. Number of months should be averaged for all major weather stations in the country.

Question number: 3 Sub-index: REI Categorisation: Meteorological Factor type: Risk Factor Intrinsic weighting: 1

Number of months over last 5 years during which rainfall is more than 20% below 30yr average for that month (drought risk)

Scoring levels: 1 2 3 4 5 6 7 0 1-10 11-20 21-30 31-40 41-50 51-60

This question examines drought conditions, not from a human perspective, but from the perspective of stress to ecological communities. By expressing the question in this form, easily-accessible meteorological data can be used to measure the risk to shortages of rainfall based on normal rainfall patterns. For most countries this will have to be assessed on a region by region or island by island basis. Number of months should be averaged for all major weather stations in the country.

Question number: 4 Sub-index: REI Categorisation: Meteorological Factor type: Risk Factor Intrinsic weighting: 1

Number of category 1-5 cyclones (<994 hPa central pressure) / decade / 10,000 sq. km (last decade only)

Scoring levels: 1 2 3 4 5 6 7 0 0.01-0.1 0.11-1 1.1-10 11-100 101-1,000 >1,000

The categories of cyclones referred to in this question are described in Johnson et al. (1995, p 14). We have included all categories of cyclone because even Category 1 cyclones can cause severe disruption of natural ecosystems. The calculation has been standardised to 10,000 km2 of land area to create whole number units. For this and following questions:

Number of cyclones Value = ------* 10,000 Land area in sq. km

Question number: 5 Sub-index: REI Categorisation: Meteorological Factor type: Risk Factor Intrinsic weighting: 1

Mean number of days per year (last five years) in which the maximum temperature was >5oC above the mean monthly maximum (calculated over last 30 years)

Scoring levels: 1 2 3 4 5 6 7 0 1-2 3-4 5-6 7-10 11-15 >16

Question number: 6 Sub-index: REI Categorisation: Meteorological Factor type: Risk Factor Intrinsic weighting: 1

Mean number of days per year (over last five years) in which the minimum temperature was >5oC below the mean monthly minimum (calculated over last 30 years)

Scoring levels: 1 2 3 4 5 6 7 0 1-2 3-4 5-6 7-10 11-15 >16

Question number: 7 Sub-index: REI Categorisation: Meteorological Factor type: Risk Factor Intrinsic weighting: 1

Number of severe storms and tornadoes / 10,000 sq. km / decade (last 10 years)

Scoring levels: 1 2 3 4 5 6 7 0 0.01-0.1 0.11-1 1.1-10 11-100 101-1,000 >1,000 This question refers to severe storms commonly defined as producing wind gusts of 90km/h (48 knots) or greater, hail storms that produce stones 2cm in diameter or greater, and storms that can produce flash-flooding, lightning and tornadoes (Johnson et al. 1995). The definition of severe storms varies from country to country, so for the purposes of the EVI we will define here that the Australian system for defining storms should be used.

Question number: 8 Sub-index: REI Categorisation: Geological Factor type: Risk Factor Intrinsic weighting: 1

Number earthquakes over the last 50 years / 10,000 sq. km land area with intensity of >6.0 Richter

Scoring levels: 1 2 3 4 5 6 7 0 0.01-0.1 0.11-1 1.1-10 11-100 101-1,000 >1,000

We have chosen to use the old Richter Scale rather than the newer Modified-Mercalli intensity scale which is based on non-instrument assessment (Johnson et al. 1995). By focusing on earthquakes of 6 or greater on the Richter Scale, we have omitted the majority of smaller, more frequent earthquakes that are of little significance from an environmental disturbance point-of-view. The long time scale reflects the expected frequency of these geological events - although the 50 year measurement window will move with each update of the EVI (every 5 years), it is not expected that this indicator will change much for a state, forming one of its relatively unchanging risk features.

Question number: 9 Sub-index: REI Categorisation: Geological Factor type: Risk Factor Intrinsic weighting: 1

Number tsunamis with run up 2m+ over last 50 years / 10,000 sq. km coastal area

Scoring levels: 1 2 3 4 5 6 7 0 0.001-0.01 0.011-0.1 0.11-1 1.1-4 >5

This question deals with tsunamis of Magnitude 1 or greater (Johnson et al. 1995) which are considered moderate and high tsunami potentials. Question number: 10 Sub-index: REI Categorisation: Geological Factor type: Risk Factor Intrinsic weighting: 1

Number of volcanoes with potential for explosive eruptions / 10,000 sq. km land area

Scoring levels: 1 2 3 4 5 6 7 0 0.001-0.01 0.011-0.1 0.11-1 1.1-4 >5

Volcanoes present several different types of hazards which may impact negatively on ecosystems. These include lava flows, ash clouds, sulphur dioxide gas with associated sulphuric acid aerosols, and possible landslides and tsunamis (Johnson et al. 1995). The definition of 'potential for large explosive eruptions' is as per Johnson et al. (1995).

Question number: 11 Sub-index: REI Categorisation: Anthropogenic, Habitat destruction Intrinsic weighting: 1

Percent land area burned by forest fires per year (worst year of last 5 years)

Scoring levels: 1 2 3 4 5 6 7 0 <1% 1-2.9% 2-3.9% 4-5.9% 6-10% >10%

Although most forest fires are likely to be anthropogenic in origin, it is acknowledged here that some are natural and no distinction is made in this question on origin of fires. This question does not include grasslands or shrublands which tend to be limited in their destructive potentials (Johnson et al. 1995). Forest fires can cause ecological disruption to large areas of terrestrial ecosystems which may take many years to recover.

Question number: 12 Sub-index: REI Categorisation: Anthropogenic, Agriculture Factor type: Risk Factor Intrinsic weighting: 1 Percentage of agriculture land under subsistence / organic agriculture

Scoring levels: 1 2 3 4 5 6 7 80-100% 60-79% 40-60% 21-40% 11-20% 1-10% 0%

Agricultural lands under organic and most forms of subsistence agriculture are less likely to be associated with problems of erosion, increased run-off, soil depletion, pesticides and wholesale habitat destruction than mechanised agriculture.

Question number: 13 Sub-index: REI Categorisation: Anthropogenic, Agriculture Factor type: Risk Factor Intrinsic weighting: 1

Tonnes of pesticides produced or imported / 10,000 sq. km land area / year (average last 5 years)

Scoring levels: 1 2 3 4 5 6 7 0 1-100 101-500 501-1,000 1,001- 5,001- >10,000 5,000 10,000

This question examines the loading of agricultural and urban land areas with pesticides which can then combine into further toxic compounds and/or find their way into streams, groundwater, coastal areas and therefore other ecosystems.

Question number: 14 Sub-index: REI Categorisation: Anthropogenic, Agriculture Factor type: Risk Factor Intrinsic weighting: 1

Tonnes of N,P,K fertilisers produced or imported / 10,000 sq. km land area / year (average last 5 yrs)

Scoring levels: 1 2 3 4 5 6 7 0 1-100 101-500 501-1,000 1,001- 5,001- >10,000 5,000 10,000 When these fertilisers find their way into other ecosystems (usually aquatic) they can lead to problems of algal blooms (including toxic algae such as those which lead to red tides and paralytic shellfish poisoning - PSP) and eutrophication.

Question number: 15 Sub-index: REI Categorisation: Anthropogenic, Agriculture Factor type: Risk Factor Intrinsic weighting: 5

Rate of deforestation of primary forest (% of remaining forest lost per year) (average of last 5 years)

Scoring levels: 1 2 3 4 5 6 7 0 0.1-1% 1.1-2% 2.1-3% 3.1-4% 4.1-5% >5%

Question number: 16 Sub-index: REI Categorisation: Anthropogenic, Agriculture Factor type: Risk Factor Intrinsic weighting: 1

Percentage of agriculture land which is mechanised, monoculture and/or commercial

Scoring levels: 1 2 3 4 5 6 7 0 1-10% 11-20% 21-40% 40-60% 60-79% 80-100%

Chemical farming methods includes the use of insecticides, herbicides, fungal agents, vermicides etc, for aquaculture this includes antibiotics. Also included in this question is the use of chemical fertilisers including hydroponics.

Question number: 17 Sub-index: REI Categorisation: Anthropogenic, Fisheries Factor type: Risk Factor Intrinsic weighting: 1 Number of commercial inshore fishing vessels / 10,000 sq. km coastal area / year (average of last 5 years)

Scoring levels: 1 2 3 4 5 6 7 0-100 101-250 251-500 501-1,000 1,001- 1501- >2,000 1,500 2,000

This question approximates the amount of fishing pressure in the nearshore ecosystems. Catch per unit of effort data would have been a better measure of fishing pressure, but data are unlikely to be available.

Question number: 18 Sub-index: REI Categorisation: Anthropogenic, Fisheries Factor type: Risk Factor Intrinsic weighting: 1

Number of commercial offshore fishing vessels / area of EEZ / year (average of last 5 years)

Scoring levels: 1 2 3 4 5 6 7 0-10 11-25 26-50 51-100 101-150 151-200 >200

This is an approximate measure of the amount of off-shore and pelagic fisheries pressure in the state.

Question number: 19 Sub-index: REI Categorisation: Anthropogenic, Fisheries Factor type: Risk Factor Intrinsic weighting: 1

Destructive fishing methods used? (dynamite, cyanide, muro ami, rotenone)

Scoring levels: 1 2 3 4 5 6 7 No Some Common

Destructive fishing methods usually are a result of and further exacerbate problems of not only overfishing, but also habitat destruction. When the fish habitats are destroyed the renewability of the fishery resource decreases. Question number: 20 Sub-index: REI Categorisation: Anthropogenic, Fisheries Factor type: Mitigating Factor Intrinsic weighting: 1

Number patrols run (boat or plane) / 10,000 sq. km of EEZ / year (average of last 5 years)

Scoring levels: 1 2 3 4 5 6 7 >100 11-100 1.1-10 0.11-1 0.011-0.1 >0-0.01 0

Question number: 21 Sub-index: REI Categorisation: Anthropogenic, Fisheries Factor type: Mitigating Factor Intrinsic weighting: 1

Fisheries observer programmes?

Scoring levels: 1 2 3 4 5 6 7 Yes No

Question number: 22 Sub-index: REI Categorisation: Anthropogenic, Government Factor type: Mitigating Factor Intrinsic weighting: 1

Percent of marine zone set aside as reserves (mean high tide to continental shelf)

Scoring levels: 1 2 3 4 5 6 7 >20% 11-20% 6-10% 1-5% 0%

Reserves means fully protected areas in which no fishing or collecting can occur. Other categories of zonation such as those with an open season or which allow limited fishing are not considered reserves. Question number: 23 Sub-index: REI Categorisation: Anthropogenic, Government Factor type: Mitigating Factor Intrinsic weighting: 1

Environmental Legislation

Scoring levels: 1 2 3 4 5 6 7 Law Draft None

Question number: 24 Sub-index: REI Categorisation: Anthropogenic, Government Factor type: Mitigating Factor Intrinsic weighting: 1

Percent of development projects accompanied by EIA (Environmental Impact Assessment)

Scoring levels: 1 2 3 4 5 6 7 95-100% 70-94% 50-69% 21-49% 6-20% 1-5% 0%

Question number: 25 Sub-index: REI Categorisation: Anthropogenic, Government Factor type: Mitigating Factor Intrinsic weighting: 1

Percent of terrestrial zone set aside as reserves

Scoring levels: 1 2 3 4 5 6 7 >20% 11-20% 6-10% 1-5% 0%

This question refers only to national parks and sanctuaries of natural habitat areas within which no hunting or collecting is permitted. Question number: 26 Sub-index: REI Categorisation: Anthropogenic, Mining Factor type: Risk Factor Intrinsic weighting: 1

Tonnes of coral extracted / year / 10,000 sq. km of coastal zone (average of last 5 years)

Scoring levels: 1 2 3 4 5 6 7 0-100 101-250 251-500 501-1,000 1,001- 1501- >2,000 1,500 2,000

Question number: 27 Sub-index: REI Categorisation: Anthropogenic, Mining Factor type: Risk Factor Intrinsic weighting: 1

Kilotonnes of sand / gravel extracted / year / 10,000 sq. km coastal area (average of last 5 years)

Scoring levels: 1 2 3 4 5 6 7 0-100 101-250 251-500 501-1,000 1,001- 1501- >2,000 1,500 2,000

Question number: 28 Sub-index: REI Categorisation: Anthropogenic, Mining Factor type: Risk Factor Intrinsic weighting: 1

Kilotonnes of all mining material (ore + tailings) extracted / 10,000 sq. km land area / year (average last 5 years)

Scoring levels: 1 2 3 4 5 6 7 0-100 101-250 251-500 501-1,000 1,001- 1501- >2,000 1,500 2,000 Question number: 29 Sub-index: REI Categorisation: Anthropogenic, Pollution Factor type: Risk Factor Intrinsic weighting: 1

Total tonnage of imported toxic or hazardous wastes / 10,000 sq. km land area / year (average last 10 years)

Scoring levels: 1 2 3 4 5 6 7 0 1-50 51-200 201-300 301-500 501-1,000 >1,000

Question number: 30 Sub-index: REI Categorisation: Anthropogenic, Pollution Factor type: Risk Factor Intrinsic weighting: 1

Millions of litres of hydrocarbons used / 10,000 sq. km land area / year (average over last 5 years)

Scoring levels: 1 2 3 4 5 6 7 0-100 101-200 201-300 301-400 401-500 501-600 >600

Question number: 31 Sub-index: REI Categorisation: Anthropogenic, Pollution Factor type: Risk Factor Intrinsic weighting: 1

Number of nuclear facilities (power, medical/research facilities, waste, weapons) / 10,000 sq. km land area

Scoring levels: 1 2 3 4 5 6 7 0 <0.1 0.1-1 1.1-10 11-100 >100 Question number: 32 Sub-index: REI Categorisation: Anthropogenic, Pollution Factor type: Risk Factor Intrinsic weighting: 1

Number of shipping ports which maintain and/or produce ships / 10,000 sq. km area of coastal zone

Scoring levels: 1 2 3 4 5 6 7 0 1-2 3-4 5-6 7-8 9-10 >10

Question number: 33 Sub-index: REI Categorisation: Anthropogenic, Pollution Factor type: Risk Factor Intrinsic weighting: 1

Electricity consumption kilowatt hours / capita / year

Scoring levels: 1 2 3 4 5 6 7 0-1,000 1,001- 2,001- 3,001- 5,001- 7,001- >10,000 2,000 3,000 5,000 7,000 10,000

Question number: 34 Sub-index: REI Categorisation: Anthropogenic, Pollution Factor type: Risk Factor Intrinsic weighting: 1

Number of cars / 1,000 persons

Scoring levels: 1 2 3 4 5 6 7 0-10 11-20 21-100 101-200 201-350 351-500 >500

Question number: 35 Sub-index: REI Categorisation: Anthropogenic, Pollution Factor type: Risk Factor Intrinsic weighting: 1 Percent of toxic wastes disposed of by high temperature incineration (average last 5 years)

Scoring levels: 1 2 3 4 5 6 7 100% 71-99% 41-70% 21-40% 11-20% 1-10% 0

High temperature incineration is considered a better way of disposing of toxic compounds because it bypasses the creation of further toxic compounds by low temperature burning and interactions among compounds. General solid wastes, including drugs can be incinerated at between 800-900oC. 1200-1300oC are required for hazardous toxic wastes and plastics (to bypass dioxin production), although there might still be toxins in the chimney stack and the ash residue (Andrew Munro (SPREP), pers comm).

Question number: 36 Sub-index: REI Categorisation: Anthropogenic, Pollution Factor type: Risk Factor Intrinsic weighting: 1

Percent of population with at least secondary sewage treatment

Scoring levels: 1 2 3 4 5 6 7 100 80-99 60-79 40-59 20-39 1-19 0

Question number: 37 Sub-index: REI Categorisation: Anthropogenic, Population Factor type: Risk Factor Intrinsic weighting: 5

Annual population growth rate (average over last 5 years)

Scoring levels: 1 2 3 4 5 6 7 Negative 0% 0.1-1% 1.1-2% 2.1-3% >3% Question number: 38 Sub-index: REI Categorisation: Anthropogenic, Population Factor type: Risk Factor Intrinsic weighting: 5

Total human population density (per sq. km land area)

Scoring levels: 1 2 3 4 5 6 7 0 1-100 101-200 201-300 301-400 401-500 >500

Question number: 39 Sub-index: REI Categorisation: Anthropogenic, Population Factor type: Risk Factor Intrinsic weighting: 1

Standing stock of tourists / 100 sq. km land area (Standing stock = # tourists x average # days stay / 365) (average for last 5 years)

Scoring levels: 1 2 3 4 5 6 7 0 1-50 51-100 101-150 151-200 201-250 >250

Question number: 40 Sub-index: IRI Categorisation: Country characteristics Factor type: Risk Factor Intrinsic weighting: 5

Total land area (sq. km)

Scoring levels: 1 2 3 4 5 6 7 >1,000,00 100,001- 10.001- 5,001- 1,001- 100-1,000 <100 0 1,000,000 100,000 10,000 5,000

This question has been assigned a high intrinsic weighting because it is considered one of the pivotal questions addressing resilience for a state. States which are large will tend to have larger numbers of habitats and species, refuges for recolonisation of species after an impact event and a tendency for impacts to affect only a small part of the state (rather than all of it). Question number: 41 Sub-index: IRI Categorisation: Country characteristics Intrinsic weighting: 1

Ratio of length of shoreline : total land area (fragmentation)

Scoring levels: 1 2 3 4 5 6 7 <0.05 0.06-0.1 0.2-0.5 0.5-0.9 1-1.4 1.5-2 >2

States with a large coastline to land area ratio tend to be fragmented or elongated and narrow meaning that more of the land area can be accessed by sea-based threats. On the other hand, however, fragmentation also offers some resilience because islands isolated from those which have been impacted may contain the same habitats and species and act as a refuge for recolonisation of impacted areas.

Question number: 42 Sub-index: IRI Categorisation: Country characteristics Intrinsic weighting: 1

Number of endemic species per 10,000 sq. km land area

Scoring levels: 1 2 3 4 5 6 7 0 <0.26 0.26-1 1-25 26-50 51-100 >100

Countries with large numbers of endemic species will tend to be more vulnerable to risks because localised extinctions cannot be resupplied from elsewhere by natural or augmented recolonisation. The loss of endemic species can lead to far-reaching secondary impacts on the functioning of ecosystems. This indicator includes mammals, birds, reptiles, amphibians, fishes and plants.

Question number: 43 Sub-index: IRI Categorisation: Country characteristics Intrinsic weighting: 1 Percent of land area <20m above sealevel

Scoring levels: 1 2 3 4 5 6 7 0 1-19% 20-39% 40-59% 60-79% 80-89% 90-100%

States with a large percentage of their land areas below 20m above sealevel will tend to be vulnerable to sea-level rise, tsunamis and storm surges.

Question number: 44 Sub-index: IRI Categorisation: Country characteristics Intrinsic weighting: 1

Percent of coastal land area composed of unconsolidated sediments (i.e. not native rock base)

Scoring levels: 1 2 3 4 5 6 7 0 1-19% 20-39% 40-59% 60-79% 80-89% 90-100%

This question identifies states which are largely composed of unconsolidated sediments (atolls, archipelagos, banks e.g. Bahamas, Tuvalu, St Brandon (Mauritius)). Their vulnerabilities to tsunamis and cyclones may be greatly compounded because it is possible for these states to lose land area.

Question number: 45 Sub-index: EDI Categorisation: Ecosystems Intrinsic weighting: 1

Has nuclear testing occurred?

Scoring levels: 1 2 3 4 5 6 7 No Yes

Question number: 46 Sub-index: EDI Categorisation: Ecosystems Intrinsic weighting: 1 Percent age area of land desertified since 1950

Scoring levels: 1 2 3 4 5 6 7 0 1-2% 3-4% 5-6% 7-8% 9-10% >10%

Question number: 47 Sub-index: EDI Categorisation: Ecosystems Intrinsic weighting: 1

Percentage of degraded coral reef area (ICRI Reef Check)

Scoring levels: 1 2 3 4 5 6 7 0 1-9% 10-49% 50-79% 80-100%

Question number: 48 Sub-index: EDI Categorisation: Ecosystems Intrinsic weighting: 5

Percentage of primary / old growth forests or vegetation remaining (e.g. prairies, savannah, desert, tundra)

Scoring levels: 1 2 3 4 5 6 7 90-100% 61-89% 31-60% 21-30% 11-20% 1-10% 0%

Question number: 49 Sub-index: EDI Categorisation: Ecosystems Intrinsic weighting: 1

Percent of fisheries stocks overfished

Scoring levels: 1 2 3 4 5 6 7 0 1-10% 11-20% 21-30% 31-40% 41-50% >50% Question number: 50 Sub-index: EDI Categorisation: Ecosystems Intrinsic weighting: 1

Percentage of land under agriculture including plantation / forestry (now)

Scoring levels: 1 2 3 4 5 6 7 0-10% 11-20% 21-30% 31-50% 51-60% 61-79% 80-100%

Question number: 51 Sub-index: EDI Categorisation: Ecosystems Intrinsic weighting: 1

Number of mariculture farms / 10,000 sq. km coastal area

Scoring levels: 1 2 3 4 5 6 7 0 <1 1.1-2.5 2.6-5 5.1-7.5 7.6-10 >10

Question number: 52 Sub-index: EDI Categorisation: Ecosystems Intrinsic weighting: 5

Percentage of original mangrove / saltmarsh area remaining

Scoring levels: 1 2 3 4 5 6 7 90-100% 71-89% 61-70% 51-60% 11-50% 1-10% 0%

Question number: 53 Sub-index: EDI Categorisation: Ecosystems Intrinsic weighting: 1 Number of harmful algal blooms including ciguatera, red tides etc over the last 5 years / 10,000 sq. km coastal area

Scoring levels: 1 2 3 4 5 6 7 0 1-2 3-5 5-10 >10

Question number: 54 Sub-index: EDI Categorisation: Ecosystems Intrinsic weighting: 1

Percent total land area affected by mining / quarrying

Scoring levels: 1 2 3 4 5 6 7 0 <0.1% 0.1-1% 1-3% 4-6% 7-10% >10%

Question number: 55 Sub-index: EDI Categorisation: Biodiversity Intrinsic weighting: 1

Number of species which have become extinct this century / 10,000 sq. km land and (coastal area * 0.5)

Scoring levels: 1 2 3 4 5 6 7 0 <0.1 0.1-0.5 0.6-1.0 1.1-5 6-10 >10

These figures should be available for mammals, birds, reptiles, amphibians, fishes and plants (e.g. IUCN Red List). Because the coastal area is defined as a 1km strip on either side of high tide mark, it was necessary to divide the coastal area by half to avoid overlap with the measurement of land area.

Question number: 56 Sub-index: EDI Categorisation: Biodiversity Intrinsic weighting: 1 Number of endangered and threatened species / 10,000 sq. km of land and (coastal area * 0.5)

Scoring levels: 1 2 3 4 5 6 7 0 <0.1 0.1-0.5 0.6-1.0 1.1-5 6-10 >10

Question number: 57 Sub-index: EDI Categorisation: Biodiversity Intrinsic weighting: 1

Number of introduced terrestrial species / 10,000 sq. km land area (over last 100 years)

Scoring levels: 1 2 3 4 5 6 7 0 <0.1 0.1-0.5 0.6-1.0 1.1-5 6-10 >10 Annex 3 Summary table of other vulnerability indices developed Cataloguing-in-publication data:

Kaly U. and Pratt C. 2000. Environmental Vulnerability Index: Development and provisional indices and profiles for Fiji, Samoa, Tuvalu and Vanuatu. Phase II Report for NZODA. SOPAC Technical Report 306. 89p.; 5 annexes, 8 figures, 4 tables. ISBN 982-207-010-1

1. Vulnerability index – environment

ii Abstract

The SOPAC Environmental Vulnerability Index has now been developed to the stage that the model and indicators have become stable enough that further technical review yields little change. This means that data can be collected and provisional EVI results calculated for countries seeking to examine their environmental vulnerabilities. This report describes the progress made during Phase II of the EVI project which focused on (1) obtaining rigorous peer review, (2) refining the model and indicators, (3) developing a mechanism for the collection of data, (4) calculating provisional results for Fiji, Samoa, Tuvalu and Vanuatu, (5) introducing and discussing the EVI in a range of international fora, and (6) developing criteria and a workplan for testing and refining the index to a stage where it could be internationally acceptable.

The results obtained for the four countries examined in this report demonstrates the potential of the EVI for identifying countries which are environmentally vulnerable in a general sense, while also providing a list of focal issues that could be used by the countries themselves and donors to improve their status. We recommend that a third and final phase of the EVI project is needed to test the model and indicators using a larger set of countries. It will also be necessary to continue the process of peer review, set the scoring and weighting factors, improve the capacity of countries to work with the EVI, collect EVI data and operationalise the index. It is strongly recommended that the EVI be fully globalised in Phase III, rather than remain focused on the Pacific Region.

iii iv Summary

The Environmental Vulnerability Index (EVI) was developed in early 1999 by SOPAC to provide a measure of vulnerability for the environment which could be calculated on the scale of entire states. The EVI was initially developed for the purpose of ranking countries and providing a single-figure expression of their relative environmental vulnerabilities. The work was done in response to a call made in the Barbados Programme of Action, the Alliance of Small Island States (AOSIS) and an increasing awareness that small island developing states face disadvantages to their development associated with their remoteness, small size, dispersion, economic conditions and limited natural resources. Although other types of indices which describe vulnerability of humans in relation to a range of stressors have been developed in the past, there have been few attempts to describe effects on the environment. Because human systems and the natural environment are dependent on one another, risks to the environment of a state will eventually translate into risks to humans. The EVI is a shortcut method of measuring and characterising the potential of harm to the natural environment arising from a range of risks.

Phase II of the EVI Project focused on developing and refining the draft index presented in Phase I and on calculating provisional indices for several Pacific Island countries. The specific aims were to: 1. Obtain international peer review on the model and indicators of the EVI to determine whether the approach was sound, indicators appropriate and the index acceptable to the international community; 2. Improve the draft EVI developed in Phase I to a stage where it could be used to collect data, provide provisional results for a small number of countries and allow for testing of the model when sufficient data are available; 3. Develop a procedure for data collection; 4. Calculate provisional EVI values for Fiji, Samoa, Tuvalu and Vanuatu to demonstrate the viability and uses of the EVI; 5. Introduce the EVI to international agencies, donors, governments, scientists and environmental managers to facilitate collaboration, the exchange of data and increase the profile and understanding of the EVI; and 6. Develop criteria for when the EVI could be considered operational and a work plan for testing and refining it to a stage where it could be internationally acceptable.

Technical review of the EVI was obtained through the running of a Think Tank in Fiji, a meeting in Malta, attendance at several key meetings, discussions with experts and publication of papers in the proceedings of meetings. The most significant progress was made during the Think Tank. The purpose of the think tank was to assemble a group of international experts from a range of disciplines central to the index, to subject it to critical peer review. A list of discussion topics was developed to provide a focus for debates on all aspects of the EVI, including the structure of the model, the indicators used, mathematical testing, a finishing line for deciding on when the index would be considered operational, its strengths and weaknesses and future directions for

v development. The assembled group of experts generally accepted the approach taken to the development of the EVI and made improvements to the framework and changes to most of the indicators. The recommended changes to the framework were incorporated immediately. The indicators, in contrast, were developed more slowly, in the context of additional inputs from other experts, potential users, recommendations from the Malta Meeting and the data collected from 4 Pacific countries.

The EVI model and indicators were modified during Phase II, particularly during the Think Tank, and have now stabilised. Participants at meetings after the Think Tank were less likely to be able to suggest improvements to the model that had not been previously suggested or make suggestions for upgrading or adding indicators. This means that the EVI model is now at a stage where although further improvements are likely, the basic form is sufficiently accepted to allow us to begin the process of larger scale data collection and testing. The list of indicators now numbers 47 (10 less that the original EVI in Phase I) and provisional scaling has been set for each indicator. Testing of the model and correlation amongst indicators will require data from at least 15 different types of countries from all geographic regions of the world.

It was necessary during Phase II to develop a protocol for the collection of data from participating countries. This involved a visit to each country, briefings of the nature, importance and need for the EVI, methods for obtaining and recording the data required and follow-up contact. Because the EVI is still in provisional form, and scoring levels for the indicators have not been finalised, it was necessary to circulate data sheets to the ministries and departments most likely to have access to the data required. This approach was successful in three of the four countries trialled, resulting in sufficient data to calculate a valid EVI (Fiji 81%, Samoa 83%, Tuvalu 85%). For Vanuatu this approach was less successful with only 70% of data obtained, rendering its EVI values strictly invalid. The results for Vanuatu are nevertheless presented in this report for information. Further work will be required to collect the missing data for this country.

The provisional results obtained for Fiji, Samoa, Tuvalu and Vanuatu demonstrated the great potential of the EVI as an international and national management tool. The overall unweighted EVI values for the four countries identified Tuvalu as the most vulnerable with a value of 4.5. Fiji, Samoa and Vanuatu returned similar scores in the range of 3.1-3.4. The application of weighting to the indicators made little difference to the EVIs for all countries except Vanuatu. It was difficult to separate biases caused by insufficient data from the real effects of weighting in this trial. The sub-indices calculated for risk exposure (REI), intrinsic resilience (IRI) and extrinsic resilience / environmental degradation (EDI) and the categorised indicators (meteorological, geological, biological and anthropogenic) revealed that although Tuvalu was generally the most vulnerable, the different countries were vulnerable in different ways. The most notable patterns were that the countries differed greatly in terms of their IRIs (intrinsic characteristics), and geological and biological hazards, but did not differ as much in their REIs, EDIs, meteorological and anthropogenic hazards. Again, because the EVI has not been sufficiently tested, these are provisional results. Vulnerability profiles were used to identify those indicators for which high EVI scores were

vi obtained (the value 7 or 6, indicating high vulnerability). This was a successful procedure that resulted in a report card for each country that identified the most important vulnerability issues.

The EVI was presented to international agencies, donors, governments, scientists and environmental managers to facilitate collaboration, the exchange of data and increase the profile and an understanding of the EVI. We represented the EVI at a total of 24 meetings during Phase II, two of which we hosted. We were also able to gain the support for future data collection from participants from a range of countries.

The EVI has now been developed to the stage that it clearly demonstrates its potential power for identifying countries which are environmentally vulnerable in a general sense, while also providing a list of focal issues that could be used by the countries themselves and donors to improve their status. The problem now is that the EVI presented here is still in provisional form. Although there is general agreement on the indicators and the way that the model has been constructed, there remain issues related to testing and scaling which have not been attempted here. These will have to be addressed in a third and final phase of the project.

At the end of Phase II there is now a choice to either create a global EVI or one crafted specifically for the Pacific Region (original brief). For the EVI to have its widest range of uses, it has to be globalised. This requirement was raised at every meeting we attended to introduce and discuss the EVI, and was a major recommendation of the Think Tank. A global focus is highly recommended because it seeks to develop the EVI on a world scale, providing an appropriate context for Pacific and other SIS. That is, the indicators selected and scaling set for each will encompass vulnerability issues and the range of conditions found world-wide. This is important if the SIDS wish to be identified as a group with special issues to consider. By contrasting them with all other countries, minor differences between them will be de-emphasised and the differences between the Pacific Region and others will become the focus of international procedures.

vii What is the EVI ?

A shortcut. A relatively quick and inexpensive way of characterising the vulnerability of natural systems (at the level of a region, state, province or island). By doing this as an index, the characterisation can be comparative because there is a common basis for the measurements What is the alternative ?

Ad hoc assessments done on a case-by-case basis for each country - very costly in terms of resources & time What are its uses ?

· Determination of LDC status (if globalised) · Getting a comprehensive general sense of the environmental vulnerability of a country · Predictive value for identifying vulnerability issues, types of hazards and approaches to stewardship of the environment of a state · Identify problem areas for external assistance to a country · Provides performance indicator for donor funding · Can be used as a measure of change in environmental vulnerability if repeated assessment are made (5 years) · Tool for raising awareness of environmental vulnerability and the actions that increase or decrease it · Tool for monitoring sustainable development · Useful for SOE reporting by identifying mechanisms that would tend to degrade the state of the environment

viii Acknowledgements

This project was undertaken with the aid of a grant provided by the New Zealand Overseas Development Assistance (NZODA) Programme. Additional support was provided by United Nations Environment Programme (UNEP), World Meteorological Organisation (WMO), South Pacific Regional Environment Programme (SPREP), Secretariat for the Pacific Community (SPC), Pacific Island Forum Secretariat and the governments of Fiji, Samoa, Tuvalu, Vanuatu and Australia who funded additional participants at the Expert Think Tank and assisted with data collection. We wish to express our appreciation to all participants of the Think Tank who fully engaged themselves in constructive discussions. We look forward to on-going discussions on issues still to be addressed.

The views expressed in this document are those of the authors and SOPAC, but are not necessarily shared by the New Zealand Government or any of the people consulted during the project. The work and results described in this document are often the consensus views of participants of meetings and discussions. Not all participants agreed with all of the decisions made throughout each meeting. We have tried to record alternative views along with the mainstream ones, and any omissions of these are accidental and are our responsibility.

ix x Contents

ABSTRACT...... III

SUMMARY...... V

ACKNOWLEDGEMENTS ...... IX

1 INTRODUCTION ...... 1

1.1 BACKGROUND...... 1 1.2 AIMS FOR PHASE II...... 2 2 UPDATE ON THE EVI MODEL DURING PHASE II ...... 3

2.1 CHANGES RESULTING FROM THE FIJI SEPTEMBER 99 THINK TANK...... 3 2.1.1 Purpose and approach ...... 3 2.1.2 Changes to the framework...... 4 2.1.3 Changes to the indicators...... 5 2.2 CHANGES RESULTING FROM THE MALTA NOVEMBER 99 MEETING ...... 6 2.2.1 Suggested Changes to the Indicators ...... 6 2.2.2 Suggested possible New Indicators ...... 8 2.2.3 Possible Indicators for a Subsidiary List...... 9 3 DESCRIPTION OF THE EVI BY THE END OF PHASE II ...... 9

3.1 THE APPROACH, FRAMEWORK AND MODEL...... 9 3.1.1 Purposes of the EVI ...... 9 3.1.2 Definitions...... 9 3.1.3 Theoretical framework for the EVI...... 10 3.1.4 Methodology: sub-indices and the indicators ...... 11 3.1.5 How vulnerability is quantified in the EVI ...... 13 3.1.6 Strengths and weaknesses of the EVI...... 14 3.2 STATUS OF THE EVI INDICATORS ...... 16 4 ENVIRONMENTAL VULNERABILITIES OF FIJI, SAMOA, TUVALU AND VANUATU...... 20

4.1 APPROACH AND METHODS FOR THE COLLECTION OF COUNTRY DATA...... 20 4.1.1 Approach...... 20 4.1.2 Problems encountered during data collecting ...... 21 4.2 PROVISIONAL COMPARATIVE RESULTS FOR FIJI, SAMOA, TUVALU AND VANUATU ...... 22 4.2.1 EVI scores...... 22 4.2.2 Sub-indices (REI, IRI and EDI) ...... 23 4.2.3 Categories (Meteorological, Geological, Country Characters, Biological and Anthropogenic) ...... 25 4.3 DETAILED RESULTS FOR FIJI ...... 26 4.3.1 General features of Fiji ...... 26 4.3.2 Persons met and data returned ...... 27 4.3.3 Vulnerability issues for Fiji ...... 27 4.4 DETAILED RESULTS FOR SAMOA...... 28 4.4.1 General features of Samoa...... 28 4.4.2 Persons met and data returned ...... 28 4.4.3 Vulnerability issues for Samoa...... 29

xi 4.5 DETAILED RESULTS FOR TUVALU...... 30 4.5.1 General features of Tuvalu...... 30 4.5.2 Persons met and data returned ...... 30 4.5.3 Vulnerability issues for Tuvalu...... 31 4.6 DETAILED RESULTS FOR VANUATU ...... 32 4.6.1 General features of Vanuatu ...... 32 4.6.2 Persons met and data returned ...... 32 4.6.3 Vulnerability issues for Vanuatu ...... 33 4.7 DISCUSSION AND CONCLUSIONS...... 34 5 MEETINGS ATTENDED...... 35

5.1 TECHNICAL MEETINGS FOCUSED ON THE EVI...... 35 5.2 MEETINGS AT WHICH SOPAC REPRESENTED THE EVI...... 37 6 FUTURE DIRECTIONS AND STEPS TO THE FUNCTIONAL COMPLETION OF THE EVI: PHASE III ...... 38

6.1 TESTING AND PROCEDURES REQUIRED FOR MAKING THEEVI OPERATIONAL ...... 38 6.1.1 Improving the model and indicators ...... 39 6.1.2 The importance of peer and user reviews ...... 40 6.1.3 Criteria for deciding on when the EVI is operati onal ...... 40 6.2 PROPOSED WORK PLAN FOR MAKING THE EVI FUNCTIONAL: PHASE III...... 42 7 RECOMMENDATIONS AND CONCLUSIONS ...... 44

7.1 RECOMMENDATIONS ...... 44 7.2 CONCLUSIONS...... 44 8 LITERATURE CITED...... 46

9 APPENDICES ...... 47

9.1 SUMMARY OF EVI DATA FOR FIJI, SAMOA, TUVALU AND VANUATU ...... 47 9.2 SOURCES OF EVI DATA FOR FIJI, SAMOA, TUVALU AND VANUATU...... 48 9.3 PUBLICATION (PRATT 1999): DATA COLLECTION FOR THE ENVIRONMENTAL VULNERABILITY INDEX (EVI) – DISCUSSION PAPER ……………………………………………………………………………………………………………………………….49 9.4 PUBLICATION (KALY 2000): LESSONS FROM THE EVI AND SMART INDICATORS FOR CORAL REEFS...... 53 9.5 SOPAC FUNDING PROPOSAL FOR PHASE III – OPTION 1 GLOBALISING AND TESTING THE EVI ...... 61 10. ACRONYMS & DEFINITIONS FOR INDICATORS ...... 71

xii Figures and Tables

TABLE 1: SUMMARY TABLE OF THE INDICATORS SELECTED THROUGHOUT PHASE II, INCLUDING THE THINK TANK AND MALTA MEETINGS...... 16 TABLE 2: THE DRAFT SCORING LEVELS SET FOR EACH OF THE EVI INDICATORS DURING PHASE II. THESE LEVELS WERE USED FOR CALCULATING THE DRAFT EVI SCORES IN SECTION 4 BELOW...... 18 TABLE 3: SUMMARY OF PHASES AND TASKS UNDERTAKEN SINCE SEPTEMBER 1998 AND PLANS FOR FUTURE WORK ...... 65 TABLE 4: IUCN CRITERIA FOR THE CATEGORIES ENDANGERED AND VULNERABLE, 1994 ...... 73

FIGURE 1: EVI RESULTS FOR FIJI, SAMOA, TUVALU AND VANUATU...... 23 FIGURE 2: SUB-INDEX SCORES (REI, IRI AND EDI) FOR FIJI, SAMOA, TUVALU AND VANUATU...... 24 FIGURE 3: CATEGORIES INDEX SCORES FOR FIJI, SAMOA, TUVALU AND VANUATU...... 26 FIGURE 4: VULNERABILITY PROFILE FOR FIJI. INDIVIDUAL VULNERABILITY SCORES ARE SHOWN FOR EACH INDICATOR OF THE EVI. MISSING BARS MEAN NO DATA WERE AVAILABLE ...... 27 FIGURE 5: VULNERABILITY PROFILE FOR SAMOA. INDIVIDUAL VULNERABILITY SCORES ARE SHOWN FOR EACH INDICATOR OF THE EVI. MISSING BARS MEAN NO DATA WERE AVAILABLE ...... 29 FIGURE 6: VULNERABILITY PROFILE FOR TUVALU. INDIVIDUAL VULNERABILITY SCORES ARE SHOWN FOR EACH INDICATOR OF THE EVI. MISSING BARS MEAN NO DATA WERE AVAILABLE ...... 31 FIGURE 7: VULNERABILITY PROFILE FOR VANUATU. INDIVIDUAL VULNERABILITY SCORES ARE SHOWN FOR EACH INDICATOR OF THE EVI. MISSING BARS MEAN NO DATA WERE AVAILABLE ...... 33 FIGURE 8: LOGICAL FRAMEWORK FOR THE DEVELOPMENT AND COMPLETION OF THE EVI...... 43

xiii xiv 1 Introduction

1.1 Background

Concern regarding the issue of vulnerability of Small Island Developing States (SIDS) was first brought to international attention during the Global Summit on Small Island States held in Barbados in 1994. At this conference, SIDS, with the support of the United Nations, expressed the desire for the development of a vulnerability index that reflects the status of SIDS and integrates ecological fragility and economic vulnerability. This desire was included in the Barbados Programme of Action with additional support from the Alliance of Small Island States (AOSIS). As a result of these events, the South Pacific Applied Geoscience Commission (SOPAC) was approached to develop an Environmental Vulnerability Index (EVI) for the natural environment with a focus on the Pacific Region. Its purpose was primarily to highlight an increasing awareness that SIDS face disadvantages to their sustainable development as a result of their remoteness, small size, dispersion, economic conditions, natural disasters and limited natural resources.

The impacts of natural hazards and human activities on the environment can potentially affect all countries and their ecosystems. Their impacts can influence countries both large and small, developed and developing, or land-locked and fragmented. Further, hazards may be operating under a number of guises, each with different operational definitions. An example of this are storm force winds which occur throughout the world and are known by several different names including cyclones, hurricanes or typhoons depending on which part of the world in which they occur. This means that any methodology developed for determining environmental vulnerability should be applicable to all countries on a common basis if any meaningful comparisons are to be made and if we are to determine which countries are more vulnerable than others in the world context.

Vulnerability indices have been developed in the past which describe the risks associated with economic and social conditions, climate change, sea-level rise, natural disasters, anthropogenic impacts and more recently, sustainability. Most of these indices describe the vulnerability of human systems with only limited attempts having been made to describe effects on the environment. Human systems and the environment are dependent on one another so that risks to the environment of a state will eventually translate into risks to humans and their welfare. The index described here has been the first attempt to construct an index that focuses on the vulnerability of the environment.

The environmental vulnerability index (EVI) is being developed as a robust, flexible tool aimed at providing a simple, short cut measure of the vulnerability of the environment of countries. The index will be intuitively and easily comprehensible to allow for wide usage in international processes (such as determination of LDC status) in addition to being a powerful tool for identifying vulnerability issues. That is, the main strength of the EVI will be that it can provide not only simplified summary information, but also the detailed data required to highlight specific areas of concern for environmental managers and scientists. Vigorous testing of the EVI will ensure that it is as much as possible an impartial measure which will

1 differentiate among countries and allow comparisons and determination of which countries are more vulnerable than others on the world scale.

1.2 Aims for Phase II

The SOPAC EVI project has passed through two phases of development to date. The first phase of this study begun in August 1998 and culminated in a report outlining a conceptual framework for determining environmental vulnerability and initial EVI calculations for Fiji, Australia and Tuvalu. This approach allowed us to test our draft model using three countries selected for their range in size, climatic and geological characteristics and level of economic development. The outcomes of that study were described in “ Environmental Vulnerability Index (EVI) to summarise national environmental vulnerability profiles” (Kaly et al, 1999a SOPAC Technical Report 275).

This report describes the development of the EVI during Phase II (but see also Kaly et al, 1999b). Phase II of the project was carried out between February 1999 and February 2000 with the following objectives:

1. To obtain international scientific peer review of the model and indicators to determine whether the approach was sound and to refine the way that the model was defined and the indicators which drive it. This work includes the running of the Fiji Think tank, a UNEP meeting in Malta, attendance at several other meetings to which we were invited, and the publication of two papers in the proceedings of meetings. The results of this work are described in Sections 2, 5, 9.3 and 9.4 of this report. 2. To develop the EVI model and indicators to a stage that it could be used to collect data and produce provisional results. The results of this work are described in Section 3. 3. To develop a streamlined procedure for the collection of environmental vulnerability data from participating countries (Sections 4.1 and 9.3). 4. To compile provisional EVI values and vulnerability profiles for Fiji, Samoa, Tuvalu and Vanuatu3. The results of this work are described in Section 4, with the data and sources being included in Sections 9.1 and 9.2. 5. To introduce the EVI in a range of fora to facilitate collaboration, data exchange and increase understanding of the need for and nature of the index (Section 5). 6. To develop criteria and a plan of the work required to test and refine the EVI to a stage where it could be internationally accepted (Section 6).

3 Although it was intended during Phase II to include Australia in the trial, we were unable to obtain their data.

2 2 Update on the EVI model during Phase II

The EVI model and indicators were refined during Phase II via the following mechanisms and activities. 1. The EVI Think Tank (September 1999) 2. The Malta Meeting (Nov/Dec 1999) 3. Data collection from Fiji, Samoa, Tuvalu and Vanuatu 4. On-going development carried out by the SOPAC EVI Team, including discussions with experts and inputs from other meetings.

Of these mechanisms, the most important were the Think Tank, country data collection and development by the EVI Team. Changes resulting from the Think Tank are described in detail in the report Kaly et al (1999b) and are summarised in Section 2.1 below. Results of the Malta Meeting are summarised in Section 2.2 and the details will be available soon in a report by Briguglio et al (2000). Changes to the indicators made by the EVI Team resulting from our own work, expert commentary and the collection of country data are concerned largely with the setting of the response scale (i.e. the data values that correspond to the 1-7 vulnerability mapping scale used in calculating the index) and are summarised in Section 3.2. Changes in the way that indicators are worded and the data they are intended to capture were made in a similar manner and are summarised in the same section.

2.1 Changes resulting from the Fiji Sept ember 99 Think Tank

2.1.1 Purpose and approach

The Think Tank was run between the dates of 7-10 September 1999 at Pacific Harbour Fiji. The purpose of that meeting was to assemble a group of internationally recognised scientists who were experts in fields relevant to the EVI to obtain their opinions and assistance with developing the index. The group was asked to examine all aspects of the original model designed by SOPAC (Kaly et al, 1999a); to provide inputs on improving its design and function, to make recommendations on how to test the model and the criteria it should pass before it could be introduced to the international arena. The overall aims of the Think Tank were therefore to: 1. Obtain peer-review and commentary from experts in a range of fields relevant to the development of the EVI; 2. Obtain sufficient inputs to make the EVI acceptable and/or operational in the international community; 3. Assemble information necessary to ensure that the EVI will be internationally applicable to all regions of the world; and 4. Identify directions for future work.

The scientists invited to the Think Tank were experts in the following fields: statistical methods, particularly multivariate techniques; biodiversity, biogeography and protected areas; other types of indices which summarise very complex data of an ecological or

3 environmental nature; weather and climate; disaster research; ecosystem management and impacts; fisheries; forestry and agriculture; productivity and energy flows; and environmental economics.

A range of topics designed to stimulate debate and solve technical problems was developed for the Think Tank (see Kaly et al 1999b). These were covered during the discussions, with resolution reached for most. For some of the topics, however, agreement could not be reached within the four days and a longer discussion has been on-going between participants, other experts and the EVI team.

The EVI model can be divided into two parts and it is of value here to discuss changes to these parts separately: 1. The framework which consists of the logical approach and mathematical model for calculating the index describing vulnerability of the environment; and 2. The indicators including their specific mapping on the 1-7 scoring scale and their weighting (importance) in relation to the overall model.

2.1.2 Changes to the framework

The framework originally designed by SOPAC was generally accepted by the group during the Think Tank. There were, however, several changes or areas in which additional work would be required that were suggested by the group. These were relatively minor changes to the model and were largely immediately incorporated in the updated EVI: 1. The definition of “environment” to which this EVI pertains needs careful consideration. Several of the experts felt that the EVI should not just describe the vulnerability of the natural environment, but that it should also capture vulnerability of the human environments of states (note that this suggestion would work against SOPAC’s initial brief for constructing the EVI and was not adopted); 2. Scoring of indicators should be on a scale of 1-7, not 0-7; 3. For yes/no answers, scoring should be less extreme than suggested by SOPAC. Instead of selecting the extremes of 1 and 7 for yes and no, the scale should be moderated to 2 and 6; 4. The idea of nett and gross vulnerability was dropped, with only gross vulnerability being calculated and the terminology being omitted altogether; 5. Indicators which were considered not applicable to a country should be scored as a 1 on the mapping scale; 6. Although it is useful to continue to refer to all indicators as being either a measure of risk (REI), intrinsic resilience (IRI) or extrinsic resilience (EDI), the group decided that only two sub-indices should be calculated. These would be the REI and RI which would be termed the Risk Exposure sub-Index and the Resilience sub-Index, respectively. The RI would be composed of the old Intrinsic Resilience Index and the Environmental Degradation Index from the original EVI study (this has not been adopted to date because it reduces the usefulness of the EVI sub-index information by unnecessarily pooling intrinsic and extrinsic risk); 7. The overall EVI should be calculated across all indicators irrespective of which sub-index they belong to;

4 8. Mathematical testing of the model was recommended by the group. This would include several considerations: (i) at least two methods for aggregating the scores obtained for each indicator into an index should be examined using the data from at least 15 countries (averaging as in the original EVI and the Villa modification of the Storie Index (Villa, 1995)); (ii) an independent test of the usefulness of the EVI should be made by deploying a group of uninitiated experts who would rank the test countries in terms of environmental vulnerability. Their rankings would then be compared with the EVI scores. For the EVI to pass this test, the difference between the EVI score and the mean of the group of experts should be no more than the difference among the experts themselves; (iii) a sensitivity analysis; (iv) The states used for testing should represent the extremes across the globe, including for example, Greenland and Switzerland to contrast against SIDS.

2.1.3 Changes to the indicators

The largest and most important changes to the EVI occurred, as expected, in relation to the indicators. These were: 1. Indicators should be global to provide a context for the EVI value obtained for any single country; 2. Many of the original indicators were replaced with new indicators by the experts relevant to the field; 3. The total number of indicators was reduced from 57 to 47 though this figure is not final and may move up or down with further refinement; 4. The list of criteria for selecting indicators was revised; 5. The mapping of the indicators on the 1-7 scale will be done by the appropriate individual experts in addition to the EVI Team; 6. When data from 15 countries are available, indicators will be examined for redundancy using correlation techniques. It may be possible at that stage to eliminate further indicators from the model; 7. It was decided that the indicators needed a better weighting system than initially suggested by SOPAC. The weighting of indicators will be examined fully when data for 15 countries are available, though for the purposes of this report, the weightings suggested at the Think Tank were trialled (see Sections 3 and 4). One of two methods suggested by two of the participants will be used to apply weightings after experts assign relative importance values to each indicator. This has to be done in the context of information on correlations between indicators obtained from the 15 test countries.

It should be noted here that the changes in the indicator questions suggested by the group remain provisional until it can be ascertained whether data will be available for each and whether there are still redundancies.

5 2.2 Changes resulting from the Malta November 99 Meeting

The participants at the Malta Meeting represented Caribbean, Mediterranean and Indian Ocean small island states (Malta, Mauritius, Jamaica, St Lucia and Trinidad & Tobago). Participants fully supported the approach taken by the SOPAC EVI Team for describing and measuring environmental vulnerability at the scale of states. There was, as expected, some discussion regarding the various definitions being used for the EVI and in particular that for “environment”. As a result of that discussion, it was agreed that the EVI should focus purely on attempting to describe the vulnerability of the natural environment. The term “environment” itself has been used to describe different parts of the world’s systems, including the “natural environment”, the “human environment” and others. It was therefore suggested, that to avoid confusion, the name of the EVI could be changed to the “Vulnerability Index for the Natural Environment” (VINE).

Further discussions focussed on the choice of indicators and ensuring that indicators captured all environmental vulnerability issues relevant to those states. Several general recommendations regarding the indicators included: · There was a need for an additional column of keywords with every indicator to help the respondent understand better what the indicator is measuring and to highlight the linkages between the different indicators (adopted, see Table 1); · Some of the environmental data that is derived from national sources is not properly audited and standardised for international comparisons, and additional care should be taken to reduce the dangers of comparing like with unlike; · The indicators should be accompanied by some coefficient indicating the level of confidence in the data (this was included in the model during Phase I); · The EVI needs to be further cross-examined through publication in an international peer- reviewed journal; · In the narrative describing the index, more emphasis should be made on the association between environmental vulnerability and costs and benefits to human systems and welfare; · That in the same narrative, the policy implications associated with anthropogenic indicators be given importance; · That the indicators be calculated over a span of time, so that rates of change can be calculated and thus be used as a dynamic tool towards natural environmental assessment and management (this was incorporated during Phase I); and · That clear definitions be developed for several indicators e.g. “continent” (partially addressed in this report).

2.2.1 Suggested Changes to the Indicators

Several changes were suggested by the Malta Meeting participants for several indicators as follows: · Indicator 15 relating to the percent of land area composed of unconsolidated sediments - It was suggested that the wording be changed to the following: “Percentage of land area less than 10m elevation within 2km of coast composed of unconsolidated sediments

6 (exclude coral reefs and ice)” (It was noted that this indicator may be somewhat difficult to measure); · Indicator 17 relating to the number of reported organism outbreaks - It was suggested that the word “catastrophic” be replaced by “reported and verified by appropriate authorities”. (It was noted that this indicator may still be open to interpretation especially regarding the terms “outbreak” and “appropriate authorities”); · Indicator 23 relating to tonnage of intensively-farmed animal products - It was suggested that the question should clearly indicate that intensive farming should include aquaculture; · Indicator 31 relating to generated toxic, hazardous and municipal wastes - There was some discussion relating to the difference of toxicity between domestic and industrial waste. It was agreed that this difference could be incorporated into the model by having two separate indicators instead of one. It was also agreed that given data constraints that at the very least the indicator should include imported toxic and hazardous materials; · Indicator 32 relating to mean percent of waste effectively managed or treated - There was discussion relating to the meaning of managed and treated waste. It was suggested that the term “effective management” would be more appropriate referring to the following strategies: composting; reusing; recycling; controlled incineration (including temperature control, retention time control and control of emissions), controlled landfill (involving treatment of leachate, containment, gas management, aftercare and rehabilitation i.e. recovery, planting, and post management); · Indicator 33 relating to number spills of oil and hazardous substances - It was noted that spills of oil and hazardous substances are not confined to the coastal areas. It was therefore suggested that the indicator be changed to refer to the number of spills of oil and hazardous substances greater than 1,000 litres during the last 5 years on land, in rivers or in territorial waters; · Indicator 34 relating to industrial facilities that could cause significant damage - Following discussion it was agreed that the denominator should include territorial waters as well as land area. With this new denominator, industrial facilities would therefore include such structures as oil rigs (our comment: we would include oil rigs but not necessarily need to divide the count by an area larger than that occupied by the land in a state); · Indicator 35 relating to number of cars / land area - It was agreed that the term “vehicles” should be used instead of cars, as per World Bank Definition. There was some discussion on whether to use density of vehicles on roads rather than total land area. It was noted, however, that this may not be appropriate, since the former is used as a measure of congestion, rather than of the ability of land to attenuate pollution emitted by vehicles; · Indicator 39 relating to the number of new fisheries stocks added - There was discussion regarding the term “new fisheries species added”. It was pointed out that (a) new fisheries does not take into account added effort or new technology, (b) a threshold of at least 20% in increase in catches should be set, and (c) the way the question was formulated could be misleading because it focuses on species rather than on stocks. It was suggested that the wording of this indicator be changed as follows: “Number of new fisheries stocks and added effort exploited by countries over the last 5 years”. It was also agreed to develop this indicator further to incorporate points (a) - (c);

7 · Indicator 40 relating to percentage of land area degraded since 1950 - It was noted that data for this indicator could be difficult to obtain. It was agreed that there needs to be a clear rationale to show that the indicator captures erosion, salination and desertification and that this indicator should exclude urban areas; · Indicator 41 relating to annual internal renewable water resources per capita - A properly specified indicator was required to capture quantity and quality of freshwater separately. It was suggested that an indicator to capture quantity could be the following: “Annual internal renewable water resources per capita.” This would be measured by average annual runoff and recharge of groundwater from endogenous precipitation. The rationale for measuring the renewable water supply per capita was that lower availability per head would create higher pressures on natural ecosystems (i.e. water for people is considered to have a higher priority than for ecosystem conservation). It was noted that the issue of water quality has been covered by indicators 30, 33, 37 and 38; · Indicator 43 relating to land, rivers and coastal zone affected by mining & quarrying - It was suggested that sea area should not be included in this indicator. The indicator should therefore read as follows: “Percentage of land, rivers and coastal zone affected by mining and quarrying”. It was noted that “coastal zone” needs to be defined more precisely; and · Indicator 47 relating to environmental legislation - There was considerable discussion regarding this indicator and its usefulness. It was suggested that this indicator be dropped (this was also suggested by a number of Think Tank participants), because the information might be considered sensitive, would be difficult to measure and because aspects of environmental management would be covered by other indicators.

Many, but not all, of the suggestions made by the participants at the Malta Meeting for refining indicators were adopted by the EVI Team and appear in the indicators lists included in this report (Table 1 and Table 2).

2.2.2 Suggested possible New Indicators

Participants at the Malta Meeting agreed to suggest the following indicators for possible inclusion in the EVI following further consideration by SOPAC’s EVI Team.

· Reported mass mortalities of organisms, including strandings - this indicator is related to Indicator 17, but focuses on losses of organisms and is a proxy for the outcome of pollution and other imbalances in the environment; · The number of ships/tonnage of hazardous substances carried/transiting within 100km of a country per year (averaged over the last 5 years) - this was considered to be significantly different from Indicator 18 which focuses on possible introductions of pests and disease. The purpose of this indicator would be to focus on the risk of spills, wrecks and ballast introductions caused by ships which might not register as having entered a port in the country; and · Pollution coming in currents of air or sea or rivers from outside the territory - this would require a yes/no response when a country is or is not downstream by air or water from a major pollution source, within 1,000 km. It was noted that this indicator would require

8 further development and refinement. It was also noted that this indicator may be captured by indicators 26, 33 and 34.

2.2.3 Possible Indicators for a Subsidiary List

Participants at the Malta Meeting also identified an additional group of environmental issues for which indicators might be developed in the future. These would be indicators for which data are probably not presently available, but which could provide important insights into the risks to and resilience of a country. These were:

· marine and forest productivity; · salination of groundwater, though it was pointed out that this may be captured by indicator 40; and · the effects of hail and glacial melt.

3 Description of the EVI by the end of Phase II

3.1 The approach, framework and model

3.1.1 Purposes of the EVI

The purposes to which the EVI could be put include: · Determination of Least Developed Country (LDC) status (if globalised); · State of the Environment (SOE) reporting (though the EVI is not an expression of the SOE, but an expression of its vulnerability to future hazards, it would be useful in the context of SOE because it would identify mechanisms which tend to degrade the state of the environment); · In sub-index and profile format, the EVI has predictive value for identifying the main vulnerability issues, types of hazards and approaches to stewardship of the environment of a state; · Would be a useful sustainable development tool; · Could provide a performance indicator for donor funding; · Provides a tool for identifying problem areas for internal intervention and external assistance; · Could be used to detect trends in vulnerability – it is intended that the EVI would be recalculated every 5 years.

3.1.2 Definitions

The EVI is a dimensionless numerical indicator that reflects the status of a country’s environmental vulnerability, where: · “Environment” includes those biophysical systems that can be sustained without human support (note that the Think Tank participants were divided on the question of whether the “environment” of interest for the EVI included only natural systems or whether the

9 definition should be broadened to include human systems. Although a slight majority of participants voted that human systems should be included, we have excluded them here because the brief was for an EVI of natural systems and the indicators incorporated so far do not attempt to measure vulnerability of human systems); · “Vulnerability” is the extent to which the environment is prone to damage and degradation; and · “Damage” is the loss of diversity, extent, quality and function of environments.

The definitions given here are pragmatic and only to be used for the purposes of the EVI.

3.1.3 Theoretical framework for the EVI

The maintenance of ecosystem or ecological integrity is at the heart of the development of a vulnerability index for the environment, because it is ecosystem integrity that is threatened by natural and anthropogenic hazards. The notion of ecosystem integrity is so complex that it cannot be expressed through a single indicator, but rather requires a set of indicators at different spatial and temporal scales and hierarchical levels relating to ecosystems. Ecosystem integrity depends on biodiversity, ecosystem function and resilience, all of which are such interrelated variables, that factors which affect just one of these can have far- reaching ecosystem-wide consequences.

Although most identifiable risk events are capable of causing damage, it is only the larger and more intense events that are likely to cause wholesale changes in the environment, at least in the short to mid-term. Some of the more important risks which can impact on the environment include meteorological events (e.g. cyclones, droughts, heatwaves, floods, tornadoes), geological events (earthquakes, tsunamis, volcanoes), anthropogenic impacts (mining, habitat destruction, pollution), biological events (plagues, blooms), climate change and sea-level rise.

10 The entities at risk, termed the 'responders' include ecosystems, habitats, populations and communities of organisms, physical and biological processes (e.g. beach building, reproduction), energy flows, diversity, ecological resilience and ecological redundancy.

Three aspects of environmental vulnerability have been identified and are incorporated into an EVI. These are: 1. The level of risk to hazards which act on the environment within a state. This relates to the frequency and where possible, the intensity of hazardous events which may affect the environment. These are based on levels observed over the past 5 years for most hazards, but may include data for much longer periods for geological events. These indicators measure potential risk only: There is no logical basis for an expectation that patterns of risk expression during the immediate history of a state will necessarily result in similar risk levels today or in the future; 2. Intrinsic resilience of the environment to risks refers to characteristics of a country which would tend to make it less/more able to cope with natural and anthropogenic hazards; and 3. Extrinsic resilience results from external forces acting on the environment and describes the ecological integrity or level of degradation of ecosystems. The more degraded the ecosystems of a country (as a result of past natural and anthropogenic hazards), the more vulnerable it is likely to be to future risks.

These three aspects of vulnerability form three sub-indices for the EVI as follows: REI = Risk Exposure sub-Index incorporates measures of the level of risk to hazards; IRI = Intrinsic Resilience sub-Index which examines that natural resilience of a state based on its innate characteristics (intrinsic resilience); and EDI = Environmental Degradation sub-Index which measures the degree of damage sustained by natural systems as a signal for how well those systems might be able to resist damage from future hazards (extrinsic resilience).

3.1.4 Methodology: sub-indices and the indicators

Because the risks are many and ecosystem resilience and integrity are complex in character, it was necessary to use a set of indicators to characterise them. This means that not all aspects of vulnerability are covered, but that a subset of variables was selected which describes frequency and intensity of risks, intrinsic resilience, and the health of ecosystems, organisms, physical features of the environment and mitigators of effects.

11 · An index (the EVI) was defined as an aggregated average of all indicators regardless of the sub-index to which they belong, to give an overall measure of the environmental vulnerability of a state.

The criteria for the selection of indicators were that they: · Should be applicable globally; · Would have data that were available or relatively easily obtainable; · Should be likely to measure change or be a proxy for change which would do significant harm to the environment; · Could not be selected on any political criteria but relate only to environmental vulnerability; · Could be weighted to reflect the probability of change to the environment and the amount of damage which might be done; · Should be relatively easy for users to understand; · Be well-defined; and · Be as uncorrelated as possible to limit redundancy.

47 indicators of environmental vulnerability have been selected for inclusion in the index. This includes 27 indicators of risk (REI), 7 indicators of intrinsic resilience (IRI) and 13 indicators of environmental integrity or degradation (EDI). The indicators were also divided into 5 subject categories, independently of which sub-index they belonged to: Meteorological events (6 indicators), Geological events (3 indicators), Country characteristics (7 indicators, and is the same as the IRI), Biological factors (8 indicators) and Anthropogenic factors (23 indicators).

Many of the indicators were expressed as a fraction of area of land rather than absolute numbers because it is risk density or proportion of area degraded that is of interest from an environmental perspective. A summary table describing the main features of each of the indicators selected and their rationale is provided in Section 3.2 (Table 1).

Several indicators initially selected were discarded because they either did not have data available or data were unlikely to be procured in the near future, they were ambiguous or bimodal in their responses; or were redundant and the information they intended to capture was present in another indicator. A list of discarded indicators is provided in Kaly et al 1999b and is not reproduced here.

Some of the data required for setting the response levels of each indicator were collected during Phase II for Fiji, Samoa, Tuvalu and Vanuatu. Some these data sets are not complete, a valid provisional EVI was calculated for 3 of these 4 countries (that is, 80% of indicator questions were answered). An EVI was calculated for Vanuatu, but this would not be considered valid because only 70% of the required data were made available to us.

Australia was invited to contribute a fifth set of data because of its potential to define the higher or lower values for indicators. Unfortunately our contacts there declined the invitation because of the incomplete status of the EVI and the daunting task of collecting the data

12 required for such a large country. Australia has agreed to contribute data at a later stage when the EVI is more complete and has been tested more fully.

The data for Fiji, Samoa, Tuvalu and Vanuatu were obtained by visiting the countries to work with local authorities, from country reports, UN, WHO, SOPAC, SPREP, FAO and other publications from international agencies (e.g. WRI), centres for risk assessment and management (e.g. Tsunami Centre, NOAA), local experts and from government officers.

For full testing of the EVI model, it will be necessary to obtain data for 15 countries from around the globe. These countries should represent the extremes of environmental conditions and are required to globalise the EVI. Even if the EVI is initially only to be applied to SIDS, it is necessary to globalise it from the start, since it is only in the context of the entire world that the vulnerability of any state can be fully assessed and in particular if it is to be used for LDC purposes.

3.1.5 How vulnerability is quantified in the EVI

The overriding principle in constructing the EVI was not to introduce complexities into the model unless there was a justifiable reason to do so. Environmental indicators are of a heterogeneous nature, that is they include variables for which the responses are numerical, qualitative and on different scales (linear, non-linear, or with different ranges). To deal with the heterogeneity, it was necessary to map the possible responses to the indicators onto a 1-7 scale. Where data were not available, no value was given for the indicator and the denominator of the average adjusted down by 1 value, or if weighted by the value of the indicator’s weighting. Where an indicator was considered 'non-applicable' in a state (such as volcanic eruptions in Tuvalu which has no volcanoes), the lowest vulnerability score of 1 was attributed to that indicator.

Mapping on the 1-7 scale for each of the indicators was provisionally set during Phase II using the data collected, the technical literature, and the inputs of experts, particularly those who attended the Think Tank.

Appropriate weighting of the EVI indicators was considered important by the participants at the Think Tank. The purpose of this weighting was to identify those indicators which were most important to any measurement of vulnerability and to ensure that the signals they contributed to the EVI were larger than less important indicators. In the original EVI, six of the 57 indicators were assigned an intrinsic weighting factor of 5, while the remaining indicators were given the default weighting of 1. At the Think Tank, the 47 indicators were rated in terms of relative importance in the opinion of participants and a simple weighting scale involving 3 weighting categories (high, medium and low) suggested. These weights were provisionally applied to the data collected for Fiji, Samoa, Tuvalu and Vanuatu in three strengths: 0 = No weighting - despite the weightings applied by the Think Tank participants, each indicator was given an equal weighting value of 1; Weak = Indicators considered of low importance were assigned the weighting value of 1, Medium a weighting of 2 and High a weighting of 3; and

13 Strong = Indicators considered of low importance were assigned the weighting value of 1, Medium a weighting of 5 and High a weighting of 10. The weighting category of each indicator is given in Table 1 and the results of applying these weightings are given in Section 4. Note, however, it will be necessary to repeat the process of assigning weightings (and therefore relative importance of indicators) once correlations between indicators have been identified during further development of the index.

The EVI and sub-indices were calculated using an EXCEL workbook. Normally, our workbook (Version 8-EVI-calculator.xls) is comprised of seven linked worksheets, each dealing with a different aspect of calculation and reporting. Report Level 1 is the highest, and gives the value of the EVI and sub-indices for each country and measures of confidence in the data. Report Level 2 gives a breakdown of the REI and IRI and EDI sub-indices showing the relative contribution of meteorological, geological, biological, and anthropogenic signals as well as country characteristics. Report Level 3 gives the scores for each individual indicator. In that workbook, a separate copy of the calculator is required to evaluate the vulnerability indices for each country. For the purposes of evaluating the data in Phase II, however, we constructed a single workbook for our own use which was not given out to the responding countries because the scoring scale could not be set before obtaining their data. Data for Phase II was therefore collected by issuing questionnaires which were filled out by the countries manually (see Section 4.1).

After adjustment for weighting (which does not vary by country), the scores for each indicator within a sub-index were averaged to produce a sub-index value of between 1 and 7. Where data are unavailable for an indicator, that indicator was omitted from the average, so that it made no contribution to the mean. At least 80% of the indicator questions had to be answered for a valid EVI to be calculated for a state.

In parallel with scoring each indicator against the 1-7 scale, the EVI model incorporates a way of assessing the reliability of data. In the future, these reliability values are to be reported with each index and should be read with them. In this report, the data reliability scores give the number indicators for which data are not currently available. In the future we will incorporate information on the number of responses which are based on real data; and the number of responses based on 'best guess' or estimated by the operator and/or authorities.

3.1.6 Strengths and weaknesses of the EVI

As for all methods of summarising and modelling data, the EVI developed here is associated with a number of strengths and weaknesses that must be understood for its proper application and use. After discussions held at the Think Tank and Malta Meeting as well as with expert representatives a set of strengths and weaknesses were identified and have been added to those compiled by the SOPAC team.

14 The strengths of the EVI have been identified as follows: · It is the first comprehensive and convenient measurement of environmental vulnerability; · Permits comparisons among countries; · Identifies a number of indicators which describe the features of risk and resilience for a country; · Can be used as a measure of change in environmental vulnerability; · Can be used to identify in-country vulnerability and therefore areas of major concern; · Stimulates debate at the science / policy interface at national and international levels and amongst disciplines; · It is able to incorporate quantitative and qualitative data on different response scales and non-linearities; · Is potentially globally applicable; · Could be used for awareness-raising; · Indicators and weightings were chosen by a panel of international experts; · Differences in interpretation of users can be minimised by training; · Has been designed with a set of validation tests to be performed and criteria to be met before it is made available for use by decision-makers; · Is based on a theoretical framework that prompted the EVI team and expert panel to find indicators for all identified aspects of vulnerability; · Identifies areas of environmental concern which could provide a focus for new or improved data collection.

The weaknesses of the EVI were identified as follows: · There is subjectivity in assigning weights to indicators and non-linearities to the scores (as in other indices); · Some complex environmental factors have been represented by proxy indicators because they could not be measured directly; · The EVI is affected by the indicators chosen and the results obtained may differ if different variables were chosen; · The method of aggregating the indicator scores does not allow for the contribution of a variable to be conditional on, or amplified by another variable (e.g. feedback, multiplicative or inhibitory effects). That is, for simplicity it assumes a non-interactive system; · The EVI is subject to problems with differences in the interpretation of users, although this could be minimised with training; · Some of the data may be difficult to obtain.

In addition to the above lists of strengths and weaknesses, users of the EVI will need to be aware of the following conditions whilst using the index: 1. The EVI emphasises short-term environmental change, rather than longer-term trends. This is appropriate as it allows responses to identified areas of vulnerability to be monitored; 2. It does not address climate change and sea-level rise because it is an ‘instantaneous’ expression of vulnerability, describing the risks to and resilience of the environment of a state now, rather than attempting to predict impacts expected in the future (it is not a state of the environment statement or an impact assessment);

15 3. Some local variations, short and long term effects and other details could not be incorporated into the model without making it too complex.

3.2 Status of the EVI indicators

By the end of Phase II the EVI indicators have stabilised to 47 questions covering exposure to risks, intrinsic characteristics of a country, level of environmental degradation and meteorological, geological, biological and anthropogenic factors (Table 1). Most of the changes from the original (Phase I) indicators occurred as a result of the Think Tank meeting. There were a few changes resulting thereafter from our own investigations, the Malta Meeting, or other interactions at meetings or with experts.

The indicators shown in Table 1 differ from those issued to the countries with respect to Indicator 47. The additional indicator shown here was a modification from the Malta Meeting which occurred after the process of data collection was underway. The results that follow in Section 4 are therefore in relation to an Indicator 47 which asks about the status of environmental legislation in each country. In the future this indicator will be dropped and the new indicator on safe sanitation incorporated.

Table 1: Summary table of the indicators selected throughout Phase II, including the Think Tank and Malta meetings. Each indicator is accompanied by a short form key name, detailed definition, list of key words which describe the main factors for which it is a proxy and the weighting category (WC) applied by the Think Tank participants (H=high; M=Medium; L=Low). Cat = categorisation where: REI = Risk exposure sub-index; IRI = Intrinsic resilience sub-index; EDI = Environmental degradation sub-index; Met = Meteorological; G = Geological; CC = Intrinsic country characteristics; B = biological and A = Anthropogenic indicators.

# Cat Key name Indicator text Main proxy factors WC 1 REI Sea Surface Greatest average annual deviation in Surface Sea Temperature Coral bleaching; fisheries; M Met Temperature in last 5 years from long term mean (30 years) (Centralised currents; eddies; ENSO; cyclones database) 2 REI High winds Number of days over the last 5 years during which the max Cyclones; tornadoes; storms; M Met recorded wind speed (3 second gusts) >20% higher than the erosion average maximum for that month (use 30yr average for each month as reference) (Data accumulated over all reference climate stations / # stations) 3 REI Dry periods Number of months over the last 5 years during which rainfall Droughts; dry spells; water M Met >20% lower than the 30yr average for that month. (Data resources accumulated over all reference climate stations / # stations) 4 REI Wet periods Number of months over the last 5 years during which rainfall Floods; wet spells; coral reefs; M Met >20% higher than the 30yr average for that month (Data pollution; erosion accumulated over all reference climate stations / # stations) 5 REI Heat waves Number of days over the last 5 years during which the max Heat waves; desertification; water M Met temperature >5 C higher than the mean monthly maximum for resources; temperature stress that month (use 30yr average for each month as reference) (Data accumulated over all reference climate stations / # stations) 6 REI Cold snaps Number of days over the last 5 years during which the max Cold snaps; temperature stress M Met temperature >5 C lower than the mean monthly minimum for that month (use 30yr average for each month as reference) (Data accumulated over all reference climate stations / # stations) 7 REI Volcanic eruptions Number of volcanoes with potential for eruption >= VEI 4 Eruptions; landslides; geysers; M G (Volcano explosivity Index) within 100km of country land gas; fires; ash; dust; marine kills boundary / area of land 8 REI Earthquakes Earthquakes within 100km of country land boundaries / land Earthquakes; landslides L G area with ML >=6.0 and <=15km depth over last 5 years

16 9 REI Tsunamis Number of tsunamis or storm surges with run-up >2m above Tidal waves; erosion; habitat L G MHWS / 100km coastline since 1900 disturbance and organism kills 10 IRI Land area Total land area (sq km) Richness of habitat types; refugia; H CC species redundancy and richness 11 IRI Fragmentation or Length of ocean shoreline or land border divided by total land Fragmentation; erosion; exposure H CC “islandness” area at borders or coasts 12 IRI Isolation Distance to nearest continent within 10 degrees latitude (km) Proximity to refugia; recolonisation; L CC (Australia is smallest continent) biodiversity 13 IRI Vertical relief Altitude range (Highest point - lowest point in country) Biodiversity of habitats and species M CC 14 IRI Lowlands Percent of land area <10m above sea-level Floods, areas of accumulation of H CC pollution, sensitive habitats 15 IRI Coastal vulnerability Percentage of land area <10m elevation within 2km of coast Storm surges; cyclones, erosion M CC composed of unconsolidated sediments (excluding coral reefs and ice) 16 IRI Endemic species Number of known endemic species / sq km land area (multiply Biodiversity; unique species H CC result by 1,000) 17 REI Pathogens and Number of reported (and verified) organism outbreaks over the Ecosystem stress; eutrophication; M B plagues last 5 years / land area (competitors, pathogens, blooms, pollution; introductions; plagues etc) (multiply result by 1,000) disturbance 18 REI Potential for Total tonnage of freight imported / year / sq km land area Potential for Introductions M B introductions 19 EDI Introductions Number of all introduced species / sq km land area since 1900 Past introductions; biodiversity M B (multiply result by 1,000) 20 EDI Endangered species Number of endangered and threatened species / sq km of land Biodiversity; Keystone species H B area (IUCN definitions) (multiply result by 1,000) 21 EDI Extinctions Number species which have become extinct since 1900 / 10,000 Biodiversity; Ecosystem structure H B sq km land area (IUCN definitions) (multiply result by 1,000) and function 22 EDI Natural vegetation Percentage of natural and regrowth vegetation remaining (e.g. Ecological redundancy; H B forests, mangroves, saltmarshes, prairies, savannah, desert, Biodiversity; Ecosystem services tundra) and goods 23 EDI Intensive farming Tonnage of intensively-farmed animal products / yr / sq km land Pollution; Eutrophication L B area (includes acquaculture, pigs, chickens, etc) 24 EDI Fisheries Percent of fisheries stocks overfished (FAO) Resource depletion H B 25 EDI Coastal settlements Density of people living in coastal settlements with city centre Stress on coastal ecosystems; H A within 20km of coast (people per sq km land area) pollution; eutrophication; resource depletion 26 REI Human population Total human population density (per sq km land area) All incidental damage caused by H A density human activities 27 REI Human population Annual human population growth rate (percent) (average over Potential for future incidental H A growth rate last 5 years) damage caused by human activities 28 REI Rate of loss of Net percentage of land area changed by the removal of natural Pollution attenuation; biodiversity; H A natural vegetation vegetation over last 5 years soil formation; natural resources; groundwater regeneration; CO2 fixing 29 REI Tourists Annual number of international tourists * average days stay / Additional load of all human M A 365 / sq km (last 5 years) impacts not reported in population statistics 30 REI Wastewaters Litres / sq km / day of untreated industrial and domestic Eutrophication; water pollution H A wastewater discharged 31 REI Production of Total tonnage of generated and net imported toxic, hazardous Pollution; habitat destruction; M A hazardous and and municipal wastes / sq km land area / year (average last 5 groundwater damage municipal wastes years) 32 REI Waste treatment Mean percent of hazardous, toxic and municipal waste Proportion of wastes rendered less L A effectively managed or treated / year harmful 33 REI Oil spills Number spills of oil and hazardous substances >1,000 litres Pollution L A during last 5 years on land, in rivers or within territorial waters / land area (multiply results by 1,000) 34 REI Toxic industries Number of nuclear, chemical and other major industrial facilities Pollution; acid rain L A that could cause significant damage / 10,000 sq km land area 35 REI Vehicles Number of vehicles (World Bank definition) / land area Habitat damage; habitat M A fragmentation; pollution; mining; hazardous wastes 36 REI SO2 concentration Max 24 hour SO2 concentration (micro g /cubic m) (average over Pollution; attenuation rates; acid M A last 5 years) rain 37 REI Fertilisers Tonnes of N,P,K fertilisers used / sq km agricultural land area / Eutrophication; pollution; soil L A year (average last 5 yrs) (multiply result by 1,000) damage; loss of arable land 38 REI Pesticides Tonnes of pesticides used / sq km of agricultural land / year Pollution; soil damage; damage to L A (average last 5 years) (multiply result by 1,000) reproductive systems of organisms 39 REI Fisheries stocks Number of new fisheries stocks or expanded fisheries efforts Rate of resource depletion L A (>20% increase in catches) added to country over last 5 years (within territory) 40 EDI Degradation % Land area degraded since 1950 (includes salinisation, Rate of habitat loss H

17 A desertification etc.) 41 EDI Water resources Mean rate of water usage per capita per day Use of surface free water and H A groundwater; groundwater, river and habitat damage 42 REI Sub-surface mining Tonnes of mining material (ore + tailings) extracted / sq km land Pollution; habitat disturbance; L A area / year (average last 5 years) heavy industry 43 EDI Surface mining % Land, rivers and coastal zone affected by mining and Habitat disturbance L A quarrying 44 EDI Terrestrial reserves Percent of terrestrial zone set aside as reserves Increases resilience, pollution M A attenuation, limits loss of biodiversity 45 EDI Marine Reserves Percent of marine zone set aside as reserves (mean high tide to Increases resilience, pollution L A continental shelf) attenuation, limits loss of biodiversity 46 EDI War / civil strife Number of war or civil strife years over the last 50 years within Habitat disturbance; pollution; M A the territory habitat degradation 47 EDI Legislation Environmental legislation with regulations Controls; management of goods M A (discarded) and services 47 EDI Sanitation Percentage of population with access to safe sanitation (WHO) Eutrophication; pollution A

Draft scoring levels for the indicators were set using the data collected from the four Pacific countries and additional information from international data sources (Table 2). For some indicators no data were available and no levels were set (e.g. 36). These indicators will be scaled at a later time when the required information can be obtained.

The scaling applied here is necessarily in draft form. It has been applied largely using the data returned by the four test countries (Fiji, Samoa, Tuvalu and Vanuatu). Where possible, the scaling levels were set using known ranges from international data sources and may therefore be (appropriately) outside the range indicated by the 4 test countries. In some cases (e.g. 44 & 45) the scales were set based on ecological criteria (actually the most appropriate for EVI, but not often available). For scaling to be completed, we will need access to data from a larger range of Pacific countries and a range of very different countries from other regions of the world (see discussion in Kaly et al. 1999b).

Table 2: The draft scoring levels set for each of the EVI indicators during Phase II. These levels were used for calculating the draft EVI scores in Section 4 below.

# Cat Key name 1 2 3 4 5 6 7 Units 1 REI / Met Sea Surface 0 >0-1 >1 >2 Degrees Temperature 2 REI / Met High winds 0 1-10 11-20 21-30 31-40 41-50 >50 Days 3 REI / Met Dry periods 0-5 6-10 11-15 16-20 21-25 26-30 >30 Months 4 REI / Met Wet periods 0-5 6-10 11-15 16-20 21-25 26-30 >30 Months 5 REI / Met Heat waves 0-10 11-20 21-30 31-50 51-70 71-80 81-100 Days 6 REI / Met Cold snaps 0-10 11-20 21-30 31-50 51-70 71-80 81-100 Days 7 REI / G Volcanic 0 1-2 3-4 5 >5 Volc / sq km eruptions 8 REI / G Earthquakes 0 1-2 3-4 5 >5 Eqk / sq km 9 REI / G Tsunamis 0 1-2 3-4 5 >5 Ts / 100 km 10 IRI / CC Land area >1,000,00 100,001- 10,001- 5,001- 1,001- 100- <100 Sq km 0 1,000,000 100,000 10,000 5,000 1,000 11 IRI / CC Fragmentation 0 >0-0.1 0.1-0.5 0.6-1 1.1-1.5 1.6-2 >2 km / sq km or “islandness” 12 IRI / CC Isolation 0 >0 - 500 501-1000 1001- 1501- 2001- >3000 km 1500 2000 3000 13 IRI / CC Vertical relief >3000 2001- 1001- 101-1000 11-100 <10 m 3000 2,000 14 IRI / CC Lowlands 0 >0-2 2.1-4 4.1-5 5.1-10 10.1-20 >20 % 15 IRI / CC Coastal 0 >0-2 2.1-4 4.1-5 5.1-10 10.1-20 >20 % vulnerability

18 16 IRI / CC Endemic 0 >0-10 11-30 31-50 51-70 71-100 >100 spp / sq km species 17 REI / B Pathogens and 0 >0-10 11-30 31-50 51-70 71-100 >100 outbr / sq km plagues 18 REI / B Potential for 0 >0-100 101-200 201-300 301-400 401-500 >500 t / sq km / yr introductions 19 EDI / B Introductions 0 >0-100 101-200 201-300 301-400 401-500 >500 spp / sq km 20 EDI / B Endangered 0 >0-100 101-200 201-300 301-400 401-500 >500 spp /’ sq km species 21 EDI / B Extinctions 0 >0-1 1.1-2 2.1-5 5.1-8 8.1-10 >10 spp / sq km 22 EDI / B Natural >80 61-80 41-60 31-40 21-30 11-20 0-10 % vegetation 23 EDI / B Intensive 0 >0-1 1.1-2 2.1-5 5.1-8 8.1-10 >10 t / sq km / yr farming 24 EDI / B Fisheries 0 >0-20 21-40 41-60 61-80 81-100 % 25 EDI / A Coastal 0-20 21-40 41-60 61-80 81-100 101-200 >200 People / sq settlements km 26 REI / A Human 0-20 21-40 41-60 61-80 81-100 101-200 >200 People / sq population km density 27 REI / A Human 0 >0-0.5 0.51-1 1.1-1.5 1.6-2 >2 % population growth rate 28 REI / A Rate of loss of 0 >0-2.5 2.6-5 >5 % natural vegetation 29 REI / A Tourists 0-10 11-20 21-30 31-40 41-60 61-80 >80 People / sq km / day 30 REI / A Wastewaters 0-1,000 1,001- 2,001- 3,001- 4,001- 6,001- >9,000 L / sq km / 2,000 3,000 4,000 6,000 9,000 day 31 REI / A Production of 0-10 11-20 21-30 31-40 41-50 51-60 >60 t / sq km / yr hazardous and municipal wastes 32 REI / A Waste 81-100 61-80 41-60 21-40 11-20 5-10 <5 % treatment 33 REI / A Oil spills 0 >0-0.5 .51-1 1.1-1.5 1.6-2 >2 Spills / sq km 34 REI / A Toxic industries 0 >0-0.5 .51-1 1.1-1.5 1.6-2 >2 Facilities / sq km 35 REI / A Vehicles 0-5 6-10 11-15 16-20 21-25 26-30 >30 Vehicles / sq km 36 REI / A SO2 Micro g / m3 concentration 37 REI / A Fertilisers 0-20 21-40 41-60 61-80 81-100 101-200 >200 t / sq km / yr 38 REI / A Pesticides 0-20 21-40 41-60 61-80 81-100 101-200 >200 t / sq km / yr 39 REI / A Fisheries 0 >0-1 1.1-2 2.1-5 5.1-8 8.1-10 >10 Stocks stocks 40 EDI / A Degradation 0 >0-1 1.1-2 2.1-5 5.1-8 8.1-10 >10 % 41 EDI / A Water 0-20 21-40 41-60 61-80 81-100 101-200 >200 L / capita / resources day 42 REI / A Sub-surface 0-20 21-40 41-60 61-80 81-100 101-200 >200 t / sq km / yr mining 43 EDI / A Surface mining 0 >0-1 1.1-2 2.1-5 5.1-8 8.1-10 >10 % 44 EDI / A Terrestrial >20% 11-20% 6-10% 1-5% 0 % reserves 45 EDI / A Marine >20% 11-20% 6-10% 1-5% 0 % Reserves 46 EDI / A War / civil strife 0 >0-1 1.1-2 2.1-5 5.1-8 8.1-10 >10 Years 47 EDI / A Legislation With No With No No units (discarded) legislation legislation legislation legislation , No , With , No , No controls social controls controls controls 47 EDI / A Sanitation No levels set

19 4 Environmental Vulnerabilities of Fiji, Samoa, Tuvalu and Vanuatu

4.1 Approach and methods for the collection of country data

4.1.1 Approach

The EVI has been designed to summarise a wide range of environmental vulnerability information for a country. Much of the required environmental data are at present only being collected and compiled at the national level. The international reporting and publication of environmental data is not as yet well established as the reporting of country economic data. As a result of this, we have had to rely on the often complex process of asking officers in each country to collate the data we required from diffuse, fragmented and uncomputerised databases located in a range of different government departments or ministries. This has meant that not all of the data required for the EVI could always be collected.

Several international initiatives have begun to address the problem of acquiring environmental data for a range of purposes, given the internationally recognised need for the publication and auditing of national environmental data. This includes the Global Environment Outlook (GEO) and the World Resources Institute (WRI). To date, however, none of these sources has been able to collect all of the data needed for the Pacific. Part of the aim of this project will be to work with these initiatives to identify additional types of environmental data required to assess environmental vulnerability.

The data needed for the EVI includes meteorological and geological data, information on resources available and their rate of usage, information on area of land and sea under different uses and its quality, natural hazards, and the way in which populations manage their ecosystems. The diverse and wide-ranging nature of these data means that their sources are widely dispersed and require some effort by a country to identify, collect and compile the information. Some of the indicators require information that could only be provided by the authorities or by experts in the respective country. It is therefore essential to have full government co-operation in the data gathering process to ensure success. This has largely been the case in the Pacific SIDS.

Several difficulties were experienced in the initial stages of data gathering in the four participating Pacific Island countries Fiji, Samoa, Tuvalu and Vanuatu. These included difficulties with understanding the data requirements for providing responses to EVI indicator questions, and in many cases, a lack of capacity to compile the necessary data. Overcoming these initial problems required the importation of assistance, in-country, to identify the major problems and to try to provide possible solutions so that country environmental vulnerability data files could be compiled.

The in-country approach to data gathering, while beneficial to the countries and rewarding in terms of data collection, is not a sustainable method of data gathering in the long term and would be impossible to extend globally. It became obvious that a more simple and directed

20 approach would need to be developed to assist governments in the gathering of country data. The development of an alternative approach to data gathering that is simple and which could be easily adopted by countries both large and small is critical to the overall development of the EVI. In this study, a questionnaire approach was developed and adopted to facilitate gathering. This was found to be relatively successful in the four countries approached for this study.

The approach used involved the compilation of a questionnaire which presented each EVI indicator question accompanied by an explanation of what it was trying to measure. Each indicator was also complemented with a clear indication of what the data might look like, how to step through the calculations required and guidance on which agency or sources might hold the required information.

To ensure that the data gathering process is sustainable in the future and can be applied internationally, it has been suggested that a supplementary handbook be developed to be distributed with the questionnaires. This help guide could be used to provide an in depth background to the EVI, its mechanics, and specific instructions on how to identify and gather the required information. This could reduce or eliminate the need for continued assistance and support in the data gathering process.

4.1.2 Problems encountered during data collecting

We encountered several problems during the data collection process which limited the amount and reliability of the data required. The most important of these were:

· A complete lack of data of the type required. This included Indicator 36 on SO2 concentrations and Indicator 45 which required information on the size of the continental shelf. Both of these indicators are important and should not be omitted from the EVI. It is recommended that mechanisms be sought for collecting these data in the Pacific SIDS; · Inaccuracies because the data available could only be provided in units unlikely to be internationally applicable and had to be converted. This refers to Indicator 23 on intensive farming. The data from the four participating countries could only be supplied as head of animals and not as tonnages and had to be converted using an estimated weight per animal. For most countries, these data are in tonnes; and · Omission of data that should have been available for unknown reasons. Vanuatu was unable to supply meteorological data that would undoubtedly have been available.

These problems account for the ND (no data) entries in the EVI data table for each country (Appendix 9.1).

21 4.2 Provisional comparative results for Fiji, Samoa, Tuvalu and Vanuatu

4.2.1 EVI scores

Valid EVI scores, with greater than 80% of indicator questions being answered were obtained for Fiji, Samoa and Tuvalu, but not for Vanuatu.

Tuvalu returned the highest and most different EVI score of the four countries examined. With an EVI of 4.5, it was more than 1.1 index units higher than Samoa at 3.4. This result means that in an overall sense, using the EVI, Tuvalu would be considered the most vulnerable country of those examined.

The overall EVI scores obtained for Fiji, Samoa and Vanuatu were similar and in the range of 3.1-3.4 when no weighting was applied (Figure 1). Weak weighting applied to the model resulted in little change to the results obtained for these three countries. When strong weighting was applied, however, the results for Vanuatu changed dramatically.

Because many data were missing for Vanuatu, it is unclear whether this change is a real reshuffling of results based on relative importance of indicators (as set by the Think Tank and which would be the purpose of weighting) or whether it results merely from the combination of data that were missing. Comprehensive testing of the EVI, as planned in Phase III is required to differentiate between these possibilities. Almost all of the missing data for Vanuatu were of medium or high weighting. This appears to have pulled its EVI score downwards (compared with the value calculated without weighting) when strong weighting was applied to the model. The EVI obtained for Vanuatu was a priori determined to be invalid because less than 80% of the data were returned. With the completion of its data table in the future, this country will be able to obtain a better estimate of its environmental vulnerability.

These results suggest that the four countries may fall into two vulnerability groups. A moderate vulnerability group comprising Fiji, Samoa and perhaps Vanuatu and a higher vulnerability group consisting only of Tuvalu. This result is very preliminary because the ranges and levels set for the indicators are limited by the availability of only a very small group of test countries (4) which do not encompass the conditions possible in the Pacific, let alone the globe. These results strongly support the need for more test countries from within the Pacific and from other regions before reliable EVIs can be calculated.

22 Figure 1: EVI results for Fiji, Samoa, Tuvalu and Vanuatu. For this and the following figures, the graphs show EVI index values in the range of 1 (least vulnerable) to 7 (most vulnerable). White bars are results without weighting (i.e. all indicators of equal value), pale grey bars are results with weak weighting, and dark bars with strong weighting. Numbers shown in the white bars are numbers of indicators answered of a possible 47.

EVI 7

4 40

3 38 39 33 2

1 FJ FJ FJ TUV TUV TUV VAN VAN VAN SAM SAM SAM

4.2.2 Sub-indices (REI, IRI and EDI)

The Risk Exposure sub-indices (REI) show a similar pattern to the results shown for the EVI in Section 4.2.1 above. Fiji, Samoa and Vanuatu cluster together with REIs of around 3, while Tuvalu returned a score of almost 4 (Figure 2). The arguments concerning the lack of data for Vanuatu’s EVI also apply for its REI score. In terms of risk to hazards Tuvalu appears to be the more vulnerable, with little difference among the remaining three countries.

The Intrinsic Resilience sub-index (IRI) separated the countries more markedly (Figure 2). In terms of intrinsic resilience, or country characteristics (the sub-index IRI and the category CC are the same measure), Vanuatu appears to be the most resilient (3.4), followed by Fiji (4.2) and Samoa (5.1). Tuvalu is the least resilient country in terms of natural characteristics (6.6). The large differences between the values suggest that the indicators and levels set might be good at differentiating the countries, though a larger context is required through collecting data from a wider range of country types.

In terms of environmental damage already sustained (EDI), the differences among countries are lower and the data returns better. Samoa and Vanuatu have EDI scores of around 3.3, Fiji a score of 3.7 and Tuvalu of 4.3 (Figure 2). These results may suggest that differences in the environmental damage sustained by the countries in the past do not differ markedly. Most of the differences in environmental vulnerability may be related more to risks and intrinsic resilience than to extrinsic resilience.

23 Figure 2: Sub-index scores (REI, IRI and EDI) for Fiji, Samoa, Tuvalu and Vanuatu.

REI 7

4 21 3 23 22 16 2

1 FJ FJ FJ TUV TUV TUV VAN VAN VAN SAM SAM SAM

IRI / Country characteristics 7

7 6

5 7

4 5

3 5

1 FJ FJ FJ TUV TUV TUV VAN VAN VAN SAM SAM SAM

24 EDI 7

4 12 10 3 10 12

1 F J F J F J TUV VAN TUV VAN TUV VAN SAM SAM SAM

4.2.3 Categories (Meteorological, Geological, Country Characters, Biological and Anthropogenic)

Meteorological index scores could not be calculated for Vanuatu because none of the 6 meteorological indicators were answered by the country. For the remaining three countries, unweighted scores were similar, ranging only between 2.6 (Fiji) and 3.25 (Samoa) (Figure 3). These results suggest that Samoa is the most vulnerable in terms of weather and Fiji the least. Without a wider context for comparison of the data, it is unclear whether the values obtained here indicate that the countries are indeed similar in terms of their meteorological variables as this graph suggests, or whether the levels for the indicators have been set with insufficient sensitivity.

There were marked differences among countries in terms of geological hazards (Figure 3) with a range of index values from 1 to 5.5 (unweighted). Fiji appeared generally not to be susceptible to geological phenomena, while Vanuatu recorded the highest vulnerability score. The relatively high score recorded for Tuvalu was largely carried by the number of tsunamis recorded. However, this result may require further investigation because the so- called “pencil effect” may systematically render any tsunamis that approach that country harmless. A high score under those conditions would be spurious.

The biological index scores resulted in marked differences among the countries (Figure 3). Vanuatu (score of 2) showed the least vulnerability in relation to biological phenomena such as species and disease introductions, organism outbreaks, resource utilisation and extinctions. Tuvalu returned a high score of 5.7, while Fiji and Samoa returned values of 2.7 and 3 respectively.

The categorical scores for anthropogenic factors were relatively similar for all of the countries, varying between 3.14 and 3.79 (Figure 3). The higher scores were obtained by Fiji and Tuvalu, and the relatively lower scores by Samoa and Vanuatu.

25 Figure 3: Categories index scores for Fiji, Samoa, Tuvalu and Vanuatu. Categorised scores have been calculated for meteorological (6 indicators) and geological (3 indicators) events, biological phenomena (8 indicators) and anthropogenic effects (23 indicators). Note that the category Country characteristics relates to the same indicators as the IRI and is not repeated here.

Meterological Geological 7 7

6 6

5 2 5 4 2 4 3 3 4 4 2 5 3 2 1 0 2 1 0 FJ FJ FJ FJ FJ FJ TUV TUV TUV TUV TUV TUV SAM VAN SAM VAN SAM VAN SAM VAN SAM VAN SAM VAN

Biological Anthropogenic 7 7

6 6 7 5 5

4 4 19 20 3 3 21 4 21 7 2 2 5 1 1 FJ FJ FJ FJ FJ FJ TUV TUV TUV TUV TUV TUV SAM VAN SAM VAN SAM VAN SAM VAN SAM VAN SAM VAN

4.3 Detailed results for Fiji

4.3.1 General features of Fiji

Fiji is an archipelago made up of over 300 islands located between latitudes 15 and 22o South and longitudes 177o West and 175o East. The Fiji islands have a combined land area of approximately 18,272 square kilometres of which 87 percent comprises the two main islands of Viti Levu and Vanua Levu. These two larger volcanic islands are characterised by steep, mountainous country, deeply incised by rivers and streams.

The islands of Fiji support a wide variety of ecosystems including significant areas of natural forest. Also present are a range of coastal and marine ecosystems including mangrove forests and coral formations. The climate is tropical with an average annual temperature of around 26oC and relatively heavy annual rainfall, especially on the windward sides of the larger islands. The average annual rainfall ranges from 1800 to 2600 mm. Fiji is potentially subject to natural hazardous events including earthquakes, landslides, cyclones, flooding and storm surges.

Fiji has a population in excess of 750,000 with increasing urban drift and concentration in coastal regions. Escalating population growth together with ad hoc development and a lack of a co-ordinated approach to environmental management has had serious detrimental implications for the environment.

26 4.3.2 Persons met and data returned Fiji Score 0 1 2 3 4 5 6 7 Data gathering within Fiji was greatly facilitated by the location of 1 SOPAC headquarters in that country. Helena McLeod and 3 Reginald Pal of the EVI team met with several government representatives to provide a briefing and discussion of the EVI. 5 Follow-up meetings were arranged to collect data from the 7 following Fiji Government representatives: 9 · Sadeesh Chang (Senior Health Inspector); 11 · Sakiusa Qereqeretabua (Fiji Visitors Bureau);

13 · Ifereimi Dau (Mineral Resources Department); · Sulana T. Nuison (Ministry of Lands and Mineral Resources); 15 · Yauka Soro (Hydrographic Cartographer - Marine 17 Department); 19 · Pumale Reddy (Senior Surveyor – Lands Department);

21 · Subodh Sharma (Acting Senior Fisheries Officer – Fisheries Department); 23 · Elizabeth Erasito (Acting Director – National Trust for Fiji); 25 Indicator · Nazmin Bi (Acting Senior Scientific Officer – Fiji 27 Meteorological Service);

29 · Epeli Nasome (Director – Environment Department); · Jone Feresi (Agricultural Officer – Land & Water Resources 31 Management). 33

35 Most of the follow-up effort for compiling the Fiji environmental 37 vulnerability data file has involved repeated requests to the contacts in the listed departments that they provide the 39 necessary data. To date this process has been relatively 41 successful with 81% of all required data being compiled. 43

45 Figure 4: Vulnerability profile for Fiji. Individual vulnerability scores are shown 47 for each indicator of the EVI. Missing bars mean no data were available.

Meterological Geological 4.3.3 Vulnerability issues for Fiji Country characters Biological Anthropogenic Figure 4 is a graphical representation of the scores obtained by Fiji for each indicator in the EVI. An examination of the results in this detail rapidly allows the identification of areas of particular vulnerability for the country – information that could lead to better management and possibly better vulnerability scores in the future.

The most important vulnerability issues for Fiji (those scoring a 7 on the EVI scale) and identified using the EVI relate to: · Dry spells; · Loss of natural vegetation cover;

27 · Lack of effective management of wastes; and · Lack of marine protected areas to build resilience.

Several other issues arose for Fiji that scored a 6 on the EVI scale and would represent its next most vulnerable aspects. These were: · Isolation from nearby sources of recolonisation should its natural environments become damaged; · A high discharge rate of untreated waste waters; and · A high rate of water usage by its industries and population.

4.4 Detailed results for Samoa

4.4.1 General features of Samoa

Samoa is situated between latitudes 13o and 15o South and 168o and 173o West close to the International Date Line. The country consists of two main islands Savai’i and Upolu and several smaller islands. The islands are mostly of volcanic origin with steep mountain ranges rising from narrow coastal plains. The greater part of the territory is covered by lush vegetation.

The climate is tropical with generally hot and humid conditions. The average annual temperature is about 26.5oC with an annual rainfall of around 3000mm. Samoa is positioned within the cyclone belt and storm patterns regularly affect the island group from November to January. Two strong cyclones struck the country in the early 1990 and 1991 causing extensive damage.

4.4.2 Persons met and data returned

Members of the EVI Team visited Samoa between 9 th and 14th August 1999. This visit was made by Craig Pratt and Reginald Pal to provide assistance and training in the identification, gathering and collation of the environmental information required for the calculation of Samoa’s EVI.

Experience with data collection in Vanuatu highlighted the need to provide a more detailed briefing and discussion prior to more intensive meetings. This ensured that the purpose and approach for the EVI had been fully clarified and that data identification and collection could proceed more efficiently. Follow-up meetings with the appropriate government staff were required at their places of work so that data returns were maximised during the visit. The representatives of the Samoa Government consulted included the following: · Mose Sua (Secretary for Foreign Affairs) · Sailimalo Pati Liu (Assistant Director DEC/DCSE) · Anne Trevor (Fisheries Division) · Silia Kilepoa-Ualesi (Treasury - Energy) · Siuli Tuailemafua (Samoa Visitors Bureau) · Teleiai Sapa Saufaleupolu (National University of Samoa (NUS))

28 · Sami Lemalu (Forestry) Samoa Score · Faatoia Malele (Climatology) 0 1 2 3 4 5 6 7 · Faefia Taliaoa (Climatology) 1 · Violet Wulf (DEC/DCSE) 3 · Sagato Tuiafiso (Scientific Officer - Meteorology) 5 · Malaki Iakopo (Assistant Director Forestry)

7 · Pau Ioane (Senior Draughtsman - Forestry)

9 Follow-up work upon return to the SOPAC Secretariat included 11 corresponding with the identified focal point in the Department of 13 Environment and Conservation as well as with specific

15 representatives met during the team’s visit. Most of the follow-up effort was put into maintaining a continual link with the Samoan co- 17 ordinating focal group in an effort to assemble the most complete 19 data file possible for Samoa. To date this process has yielded a 21 data collection rate of 83% for that country.

25 Figure 5: Vulnerability profile for Samoa. Individual vulnerability scores are shown Indicator

27 for each indicator of the EVI. Missing bars mean no data were available .

29 4.4.3 Vulnerability issues for Samoa 31

33 Samoa’s areas of greatest vulnerability (producing EVI scores of 7) 35 include indicators 3, 12, 28, 30, 32, 37 and 41 (Figure 5). This 37 means that if we use the EVI as a tool for identifying areas of

39 environmental vulnerability, Samoa’s most vulnerable attributes would be: 41 · Dry periods; 43 · Isolation from other land masses; 45 · Rate of loss of natural vegetation;

47 · Untreated wastewaters; · Poor management of wastes; · Use of agricultural fertilisers; and Meterological · High usage of water resources. Geological Country characters Biological Anthropogenic

29 4.5 Detailed results for Tuvalu

4.5.1 General features of Tuvalu

Tuvalu is a small fragmented atoll country comprised of nine islands and atolls spread over some 1.3 million square kilometres of ocean. The total land area is among the smallest in the world at only 25.9 km2. The country is isolated, located approximately 1,100 km north of Suva, Fiji and serviced by two flights per week and a monthly freight ship. Tuvalu is subject to a number of natural risks such as cyclones and storm surges. The islands are low-lying with most of their area below 3m above sea-level and no point higher than 4.5 metres. This feature potentially makes the country highly susceptible to erosion and inundation from ocean surges.

Because of its tiny landmass, limitations in the carrying capacity of the environment are immediately apparent. There is little land available to assimilate waste, and as population increases, risks to the environment from human pressure and poor management will continue to increase. On the main islet of Fogafale on the atoll of Funafuti, about 4,600 people are crowded together in less than 2.8 square kilometres. A third of the land is uninhabitable due to its use as an airfield and the presence of excavated borrow pits.

4.5.2 Persons met and data returned

Helena McLeod visited Tuvalu between 4th and 9th October 1999 where she teamed with Ursula Kaly to facilitate the process of identification, collection and collation of environmental vulnerability data for Tuvalu. A briefing session was held with government representatives to discuss the EVI and the data requirements. Follow-up meetings were made with the following government officers to ensure that as much of the data required could be collected: · Seluka Seluka (PICCAP) · Mataio Tekinene (Environment) · Claudia Ludescher (Funafuti Conservation Area) · Tesimita Ailesi (Lands and Survey) · Fanoanoaga Patoro (Lands & Survey) · Hilia Vavae (Meteorological Service) · Itaia Lausaveve (Agriculture) · Sautia Maluofenua (Fisheries) · Mataliki Christopher (Central Statistics Division) · Uatimani Maloo (Tourism, Trade & Commerce) · Julia Heiloa (Funafuti Town Council) · Christopher I Kae (Customs)

Tuvalu was an easy country from which to collect data for the EVI. This was because of its small size, concentration of government departments in a small area for easy access and the fact that one of our team is resident in the country and was able to increase the profile of the project. The small size of the country meant that information was usually readily accessible, easy to sift through, or could be collected as required. For example, the number of cars present on Fogafale was hand counted from the original registration records, an approach that would have been too laborious in larger countries. The most significant

30 Tuvalu problem encountered was that information was generally lacking Score for some sectors, particularly fisheries and for standard 0 1 2 3 4 5 6 7 environmental data. This necessitated the return of some data 1 based on best guess estimates made by government officers.

3 Follow-up work on return to the SOPAC Secretariat involved 5 ensuring acquisition of the data for the few indicators still to be 7 collected. 85% of data required for the EVI was collected for

9 Tuvalu.

13 Figure 6: Vulnerability profile for Tuvalu. Individual vulnerability scores are shown for each indicator of the EVI. Missing bars mean no data were available. 15

17 4.5.3 Vulnerability issues for Tuvalu 19

21 The results obtained from the EVI identified Tuvalu as the most 23 vulnerable country tested. 15 of the 40 indicators for which data

25 could be obtained resulted in an EVI score of 7, with a further Indicator single indicator scoring an EVI value of 6. Issues of high 27 vulnerability occurred in all categories of data except 29 meteorological phenomena. The main issues that country might 31 consider examining in relation to its environmental vulnerability

33 are: · Tsunamis (though this is provisional since the “pencil effect” 35 may systematically render them harmless); 37 · Small land area, high fragmentation and isolation; 39 · Low relief and 100% of land area susceptible to coastal

41 inundation and erosion; · A history of past and large potential for new introductions of 43 new species, pests and diseases; 45 · Pollution effects from intensive farming of pigs over a small 47 land area; · High human population density and growth rate; and · Poor management of wastes. Meterological Geological Country characters Biological Anthropogenic

31 4.6 Detailed results for Vanuatu

4.6.1 General features of Vanuatu

Vanuatu is an archipelago of over 80 islands with a combined land area of some 12,190 square kilometres. The country is relatively mountainous with about a quarter of the land forming rugged or steep ranges characteristically in the inland areas of the larger islands. About a third of the forest resources of the country are found in these rugged, erosion-prone inaccessible areas.

The population of Vanuatu is around 189,000 people distributed over 70 of the islands in the archipelago. The Exclusive Economic Zone (EEZ) covers an area of sea of about 710,000 square kilometres of ocean with a rich diversity of inter-tidal mangrove communities, seagrasses, lagoons, coral reefs and pelagic areas.

Vanuatu generally has a hot, wet climate with a marked difference in climate through the archipelago from tropical in the northern areas to sub-tropical in the south. Rainfall is variable with an average annual rainfall at around 1500mm and with an average temperature of about 24oC. Vanuatu is considered uncommonly prone to natural disasters such as volcanic activity, earthquakes, cyclones and the associated problems of landslides and tidal waves.

4.6.2 Persons met and data returned

A trip to collect EVI data was undertaken from 22 – 29 June 1999 to Vanuatu by Craig Pratt of the SOPAC EVI team. While there he trained technical personnel in the identification, gathering and collation of environmental information required for the calculation of Vanuatu’s EVI.

Meetings were held with government representatives in several departments and ministries. Meetings involved briefings on the EVI together with discussions to clarify the purpose of the work and the required data clearly identified. Follow-up work for the completion of data collection was co-ordinated by the Department of Environment.

Government representatives with whom meetings were held included: · Stanley Temakon (Director General – Ministry of Natural Resources); · Chris Ioan (Acting Director – Geology); · Jean Michelle (Seismology); · Jean-Phillippe (Seismology); · Ericson (Water); · Leo Moli (Principal Energy Officer – Energy Department); · Tatsuo Honda (JICA Expert); · Ernest Bani (Principal Environment Officer – Environment Department); · Emil Mael (Vanuatu Land Use Planning Office); · Peter Morris Jimmy (Statistician – Bureau of Statistics);

32 · Jacqui Caine (Deputy High Commissioner – NZ High Commission); · Ario Niki Roberts (Director - National Tourism Development Office); · Patricia Mawa (Meteorological Officer – Meteorology); · Benuel Tarilongi (Director - Vanuatu Quarantine & Inspection Service); · Moses Amos (Director - Fisheries Division); · Wesley Obed (Fisheries Licensing Officer); and · Norris Hamish (Director Ports & Marine Department).

Vanuatu Score Figure 7: Vulnerability profile for Vanuatu. Individual vulnerability scores are 0 1 2 3 4 5 6 7 shown for each indicator of the EVI. Missing bars mean no data were available. 1

3 Data collection in Vanuatu was less successful than in the 5 remaining 3 countries included in this study, with a total of only

7 70% of the required data being collected. This meant that the EVI values calculated Section 4.2 fell short of the 80% data 9 requirement for a valid EVI. The profiles discussed below which 11 provide individual scores for indicators that were answered, 13 however, could still provide useful information for in-country 15 identification of vulnerability issues. The questionnaire and meetings approach, though used in the same manner as in Fiji, 17 Samoa and Tuvalu resulted in relatively little data being collected 19 in-country in Vanuatu. We conclude that additional emphasis 21 would need to be placed in the future on a discussion of the 23 importance of the EVI, its uses and outcomes and the critical

25 need for data. Indicator

27 After leaving Vanuatu, the EVI Team continued its efforts in trying 29 to generate urgency and sufficient understanding of the EVI and 31 the critical need for environmental vulnerability data. Problems

33 with correspondence and several changes in staffing and co- ordinating focal people added to the low data returns. A follow-up 35 visit made by Helena McLeod on 22nd November 1999 yielded 37 little improvement. Despite efforts by both the SOPAC EVI team 39 and the Vanuatu Government, data targets could not be met by

41 the time of writing this report and will need to be addressed in future to create a valid EVI for this country. 43 45 4.6.3 Vulnerability issues for Vanuatu 47

Meterological Of the indicators for which data were obtained, 5 scored a value Geological of 7 on the EVI scale. There were no indicators with a score of 6 Country characters EVI units. The main environmental vulnerability issues identified Biological for Vanuatu were: Anthropogenic · Tsunamis; · Human population growth rate;

33 · Management of wastes; · Land degradation; and · Water usage. Other issues may come to light when all of the required data can be collected for this country.

4.7 Discussion and conclusions

The results obtained for the overall EVI scores for the four countries examined here (Figure 1) demonstrate the potential of the EVI as a tool for international processes such as the determination of LDC status. The EVI is able to summarise environmental vulnerability information collected for a range of indicators in a form that is comprehensible, visual, and which could be used to identify countries which have high environmental vulnerability. In addition, the vulnerability profiles provided by the sub-indices and scores for the individual indicators (Figure 2, Figure 3, Figure 4, Figure 5, Figure 6, and Figure 7) demonstrate the powerful potential of the EVI for identifying issues of vulnerability that could be used to prioritise efforts for environmental management within countries.

The results presented here are necessarily provisional. The scoring levels set by the EVI team are based on data derived from only four countries, all of which are Pacific SIDS which share characteristics of small size, isolation, developing economies, climate, biodiversity and other factors. For the EVI to be able to provide a reliable mechanism for international comparisons and a context for how well any individual country might be faring in terms of risks and intrinsic and extrinsic resilience, it is necessary that the scaling for the model be set within a much broader range of country types. Ultimately, this means that the EVI needs to be globalised so that the full range of vulnerability issues facing states can be incorporated into the model.

The results presented here are also provisional because the EVI has not been mathematically tested (see Kaly et al, 1999b for description of tests needed). The weighting of indicators is in draft form and the biases of types and strengths of weightings need to be modelled against independent assessments of vulnerability made by independent experts. It is also necessary that the effects of missing data on the outcome of the EVI be investigated. All of these issues can be addressed when data from a larger number of countries of a range of types can be collected.

Problems with a lack of data, resulting in an invalid EVI for Vanuatu, and less than complete returns for the remaining countries need to be addressed. Public awareness, political will, capacity building and the establishment of permanent data collection mechanisms are all required to address this issue in the future.

34 5 Meetings attended

5.1 Technical meetings focused on the EVI

A range of meetings were attended by the EVI team during Phase II to introduce the EVI methodology to its potential users, a range of SIDS, SOPAC member countries and potential donors. SOPAC also hosted a meeting in Fiji (Think Tank) and assisted the Foundation for International Studies (FIS) of the University of Malta to host a meeting in Malta. Both of these meetings were for the purpose of developing the EVI. The latter was also designed to introduce the EVI to the Caribbean, Indian, Mediterranean and Atlantic small island states (SIS).

Committee for Development Policy (UNCDP / UNDESA) Technical Group’s Meeting on the Vulnerability Index, London, March 1999

This meeting was hosted by CDP at Marlborough House, 1-3 March 1999. The purpose of the meeting was to explore the potential for using indices of vulnerability in the criteria for identifying LDCs. The indices discussed included the Economic Vulnerability Index being developed by the Commonwealth Secretariat, the EVI and a composite index made of these and other types of vulnerability measures.

This meeting was attended by Drs Russell Howarth and Ursula Kaly who introduced the EVI to participants, gave a seminar and held discussions on the potential role of the EVI in determining LDC status.

EVI Think Tank, Fiji, September 1999

This meeting was hosted by SOPAC and held at Pacific Harbour in Fiji, 7-10 September 1999. The purpose of the meeting was to assemble an international group of experts from a range of technical disciplines relevant to the development of the EVI and to use their expertise to rapidly and efficiently review all of the most important defining conditions for the model and its indicators. The participants invited were experts in statistics, biodiversity, biogeography, protected areas, other types of indices, weather & climate, disaster research, ecosystem management, environmental impacts, fisheries, forestry & agriculture, productivity & energy flows, and environmental economics.

The meeting was attended by the entire SOPAC EVI team, and was opened by New Zealand and the Director of SOPAC. It gave the EVI the critical peer review and development boost required to confirm that the team had basically been taking an appropriate approach as well as provided constructive criticism and suggested improvements to the selection of indicators. Almost all of the changes suggested were adopted as part of the new EVI.

35 SOPAC Annual Session Donors Presentation, Nadi, October 1999

A special seminar and discussion forum within the SOPAC Annual Session was conducted on 27th October 1999. The EVI was introduced to representatives of SOPAC’s member countries and to Regional Donors with a view to informing them of the status, progress made and requirements for the completion of the EVI. The seminar was given by Dr Ursula Kaly and Mr Craig Pratt.

UNEP Global Workshop on Vulnerability and Adaptation 4 – 5 October 1999, Nairobi, Kenya

The meeting was hosted by the United Nations Environment Programme (UNEP) in Nairobi, Kenya at UNEP headquarters. The goal of the meeting was to develop a framework for comparing vulnerability and adaptability to climate change impacts. The meeting was held over two days and divided into 5 sessions focussed on: Global vulnerability framework research and experiences of similar indices; Vulnerability framework; Sectoral issues; Adaptation to climate change; and Future directions.

The meeting was attended by international experts from all over the world. Representatives included government officials, international, university and technical institution experts from Bangladesh, Kenya, Denmark, Switzerland, England, Morocco, Tanzania, Mauritius, Zimbabwe, USA, Cuba, Uganda, Togo, Canada, Costa Rica, Brazil, Haiti and Mauritania.

Mr Craig Pratt presented an abstract and gave a seminar on the EVI. He also fielded discussion on the development of the index and how it could be of value in the development of a climate change vulnerability index. He also circulated copies of previous EVI reports.

The primary output from the meeting was a discussion paper: Climate Change Vulnerability: Toward a framework for comparing adaptability to climate change impacts. Other outcomes of the meeting included the preparation of a brief to be tabled at the Conference of the Parties (COP6) with more time and effort being put into preparing a more complete discussion paper on the development of a vulnerability index for climate change to be presented at the next COP.

UNEP / FIS University of Malta Meeting, Valetta, November/December 1999

This meeting was funded by UNEP and hosted by the Foundation for International Studies at the University of Malta with assistance by SOPAC, 29 November – 3 December 1999. The purpose of the meeting was to introduce the EVI to small island states in the Caribbean, Indian, Mediterranean and Atlantic Regions, to review and further develop the indicators selected so far and to gain support from countries in these regions in the process of data collection. Of particular concern was the question of whether the indicators developed so far were applicable to countries outside of the Pacific Region.

36 The meeting was hosted by Professor Lino Briguglio and attended by Dr Ursula Kaly and Mr Craig Pratt. Representatives of Malta, St Lucia, Trinidad & Tobago, Jamaica and Mauritius were present at the meeting.

The outcomes of the meeting were that the EVI and indicators were generally suitable for those countries and that few further changes were suggested by the participants. All of the representatives of the countries listed agreed to assist with the future development of the EVI by providing data for their countries.

UNEP / Australian Institute of Science (AIMS) Meeting on Information Management and decision support for marine biodiversity protection and human welfare: Coral Reefs, Townsville, December 1999

This meeting, hosted by UNEP and AIMS was run at AIMS headquarters in Townsville, Australia, 6 – 10 December 1999. Its purpose was to develop a programme relevant to coral reefs for GEF funding: 1) to describe elements of the process by which resource use practice is modified with a view to improved outcomes for humans and for aquatic ecosystems; 2) to identify key indicators that accurately reflect both current status and progress towards those outcomes; 3) to define the information needs necessary to develop the indicators; 4) to identify and, if necessary, specify improvements to infrastructure to: a) effectively manage information and indicators, and b) assist decision makers in scenario- testing for various policy options.

Dr Ursula Kaly attended this meeting to describe the EVI and the lessons it could provide in the development of indicators for the health of coral reefs. She supplied an abstract to the meeting, gave a seminar and produced a paper that is at present under peer review before it will be published in the proceedings (see Appendix 9.5).

5.2 Meetings at which SOPAC represented the EVI

1. SIDS Meeting in New York, US, February 1999 2. CSD 7, New York, March 1999 3. Commonwealth Secretariat/UN Task Force on Vulnerability of SIDS, London, April 1999 4. AOSIS Clean Development Mechanisms for Climate Change Meeting, Majuro, Marshall Islands, June 1999 5. Forum Economic Ministers Meeting, Apia, Samoa, July 1999 6. Pacific Science Congress, Sydney, Australia, July 1999 7. IPCC Meeting of Lead Authors of Chapter 17 on SIS, Malta, July 1999 8. IOC General Assembly Meeting, Paris, France, July 1999 9. IDNDR Meeting, Geneva, Switzerland, July 1999 10. CROP Meeting, Nadi, August 1999 11. Pacific Island Climate Change Assistance Programme (PICAPP) Meeting on Climate Change Integrated Model (PACCLIM), Auckland, New Zealand, August, 1999 12. International Workshop on Sustainable Development in SIDS, Loften, Norway, August 1999

37 13. UN General Assembly on Special Session on SIDS, New York, September, 1999 14. South Pacific Tourism Organisation (SPTO) Governing Council Annual Session, Apia, Samoa, October, 1999 15. Pacific Island Forum, Palau, October 1999 16. SOPAC Governing Council Annual Session, Nadi, Fiji, October 1999 17. Diplomatic Training Workshop for Forum Island Countries, Suva, Fiji, November 1999 18. UNEP Meeting of Experts on the Environmental Vulnerability Index (EVI), Valetta, Malta, November 1999

5.3 Upcoming meetings at which SOPAC will present the EVI

· Bermuda Workshop on Vulnerability and Risk, International Ocean Institute, Bermuda, February 2000 · Global Conference on Development Agenda for Small States (Commonwealth Secretariat/World Bank Joint Task Force on Small States), London, February 2000 · Leaders Summit, Tokyo, Japan, April 2000

6 Future directions and steps to the functional completion of the EVI: Phase III

In this section we describe the main EVI development issues for Phase III. Many of these criteria and tests were determined during the Think Tank meeting and have therefore resulted from international peer consensus. This means that we are confident that the procedures described below are both necessary and sufficient to render the EVI functional and internationally acceptable. In Section 6.1 we describe what these procedures are, and in Section 6.2 the tasks we would need to undertake to accomplish them.

6.1 Testing and procedures required for making the EVI operational

The Think Tank group identified a range of tasks that need to be completed in order to operationalise the EVI. These include improvements in the model and refinement of the indicators (including mathematical testing), peer and user reviews and logistic arrangements.

38 6.1.1 Improving the model and indicators

The main procedures for improving the EVI model or framework suggested by the Think Tank participants were to: 1. Calculate the EVI for a range of countries so that tests of its ability to summarise the environmental vulnerability of states can be undertaken; 2. Ensure that countries used in testing represent the global extremes so that testing covers all of the expected spread among countries; 3. Undertake independent ranking of countries by experts to demonstrate that the EVI is concordant with the aims, is useful for the purposes for which it was designed and provides the information required in an efficient manner; 4. Carry out a sensitivity analysis; 5. Ensure that the vulnerability breakdown (sub-indices and groupings into Meteorological, Geological, Biological, Country Characteristics and Anthropogenic signals) and profiles are logical and accompanied by diagrams to assist users in their interpretation; 6. Test alternative methods for accumulating indicator scores into the final EVI value.

By the end of the Think Tank, two methods of accumulating scores into the EVI had been considered. The first was status quo, where scores for individual indicators were accumulated as averages to produce the final index. A second method was suggested by F. Villa which is a modification of the Storie Index (Villa, 1995). This is a non-linear aggregation method which will tend to highlight differences among countries at the high end of the vulnerability scale (EVI scores close to 7). This could provide a useful way of more closely examining relative differences among highly vulnerable countries while tending to be less sensitive at distinguishing countries that are not very vulnerable. This makes operational sense since it is the more vulnerable countries that we are most concerned about at the single figure level. Using this technique would not affect profiles, so that internal identification of problem areas would still be effective for all countries. Both of these methods of accumulation could be tested as part of the process of improving the EVI model.

The list of indicators as it stands now requires additional work. Many of the original indicators (Phase I) were discarded, and the new ones suggested by the participants of the Think Tank and Malta meetings require refinement of definitions and scaling. The main tasks which will be required are: 1. Complete the operational definitions for those indicators that have not been fully defined; 2. Locate sources of data for each indicator; 3. Identify data which will need to be collected in the future and determine whether any of these can be taken up by global data-collection organisations; 4. Set the scaling for each indicator in the global context (this will require collecting data for countries at the global extremes); 5. Test indicators for redundancy (i.e. correlations); 6. Test the effects of data omissions on the final EVI scores; 7. Finalise the weightings for indicators in the context of their degree of correlation with other indicators in the model.

39 No data exist which can be used to weight indicators using ecological criteria. That is, we do not know which hazards and effects are most important in determining the vulnerability from an environmental perspective. As is common practice, it was therefore agreed at the Think Tank that indicators would be weighted in terms of their perceived importance by the experts assembled at the meeting. Two methods of weighting indicators were suggested.

The first of these, suggested by L. Briguglio, consisted of simultaneously ranking all 47 indicators with an importance rating of between 0 and 4. Indicators receiving a score of 4 were considered by a participant to be of top importance in measuring vulnerability. Indicators which were considered unimportant were given a score of 1, and so on. The score 0 was reserved for indicators thought by a participant to be irrelevant or redundant and which should be removed. The average of all of these scores were then calculated separately for each indicator and a simple weighting of High, Medium or Low applied to each indicator. We used this system of weighting in this study for Fiji, Samoa, Tuvalu and Vanuatu with zero, weak and strong weighting multipliers (see Section 3). We have not assumed, however, that this system of weighting is the best one to use for the EVI and will carry out mathematical modelling to find the best system in Phase III.

The second method, suggested by F. Villa, consisted of pair-wise ranking of each indicator against all other indicators. This method requires the ranker only to assess whether one indicator is more important than another at any one time. It is not necessary to remember the ranking given to any other indicator whilst carrying out this procedure. The resultant pair- wise ranking matrix is then analysed simultaneously for all rankers (experts), and clusters of indicators of more and less relative importance identified. Each of the clusters would then be assigned a weighting to be applied to the indicators it contains.

Either of these two techniques for weighting could be used for refining the indicators once the question of correlations has been addressed.

6.1.2 The importance of peer and user reviews

It was considered important that the EVI remain under constant review during its development. The main mechanisms for review include publication of the EVI and presentation at international meetings. The main outcomes of review would be the: · Development of a stable list of indicators in association with experts and within the context of data availability; · Acceptance of the indicators by nationals as useful and non-threatening mechanisms for describing the environmental vulnerability of states; and · Acceptance of the EVI by international organisations, such as the UN, and donors.

6.1.3 Criteria for deciding on when the EVI is operational

A list of three criteria was developed to provide guidance to experts, funding agencies and the international community on when the EVI would be technically ready for use. This list is important because it provides an independent ‘finishing line’ for the procedures described

40 above. It ensures that appropriate milestones can be developed and a completed EVI identified in relation to the funding that will be required to complete the work.

Criterion 1: Redundant indicators

The data for at least 15 states need to be collected so that redundancy among indicators can be identified if they exist. It was agreed during the Think Tank that at least 15 countries with widely ranging characteristics would need to be included in this test. Any indicator with a high correlation with one or more other indicators would at this stage be dropped or merged. The final list of indicators would then only consist of those that bring significantly new information into the EVI value. The final weighting of indicators can only occur after redundancies in the model have been limited.

Test: When the correlation coefficient among two indicators is non-significant in a standard statistical test.

Criterion 2: EVI scores for a range of country types

The EVI scores need to be evaluated for at least 15 countries with widely ranging characteristics to examine how well the model provides the spread required to distinguish them. The countries included in this test (same as for criterion 1) should include small islands, large continental masses, highly-fragmented countries, land-locked states, tropical and cold climate countries and deserts. The EVI should be able to cluster similar countries together and provide spread among countries which are very different. The response scales for each of the indicators (the mechanism in the EVI that provides the spread) can be finalised when data for these 15 test countries are available.

Test: When the spread in EVI values among the 15 test countries occupies much of the 1-7 range expected and countries considered a priori to be ‘similar’ cluster closer together than ‘dissimilar’ countries.

Criterion 3: Validation

The purpose of constructing an EVI is to simplify the task of categorising countries according to their relative environmental vulnerabilities. If personnel, funding and time were unlimited, this could be done by sending several independent teams of scientists to each country and commissioning them to carry out a vulnerability assessment for each. The replicate assessments for each country could then be used to classify countries in terms of their vulnerability. This exercise would of course be extremely expensive and it is one of the purposes of the EVI to simplify this process.

The only independent means of assessing the effectiveness of the EVI in carrying out this task in a simplified way, is to compare the results of the EVI with a full assessment for a small number of, say 5, countries. Several teams of experts would have to be mobilised in each of the test countries to provide a 'mean assessment' for each. The consultants involved should be unaware of the mechanics of the EVI to ensure that they do not unintentionally

41 bring bias into the results. These assessments could then be compared with the EVI scores obtained.

Test: When the difference between the value obtained by the EVI and the mean of the assessment provided for a country by several experts (who are unaware of the workings of the EVI) is about the same, or less than, the spread found among the assessments of the experts. This test should be performed for about 5 countries.

6.2 Proposed work plan for making the EVI functional: Phase III

Option 1 – Global Focus is highly recommended because it seeks to develop the EVI on a world scale, providing an appropriate context for Pacific and other SIDS and SIS. That is, the indicators selected and scaling set for each will encompass vulnerability issues and the range of conditions found world-wide. This is important if the SIDS wish to be identified as a group with special issues to consider. By contrasting them with all other countries, minor differences between them will be de-emphasised and the differences between the Pacific Region and others will become the focus of international procedures.

Option 2 – Pacific Focus is not recommended here. It would be the development path taken if the original scope of the study, to develop an EVI for Pacific SIDS, were followed. This work would necessarily require most of the same steps as Option 1, but would result in an EVI with narrower focus and applicability. By restricting the EVI to the Pacific Region, index values will not provide a global context for the conditions and issues found in Pacific SIDS and will instead emphasise the differences found among Pacific Island countries. A Pacific EVI will result only in one Pacific Island Country being compared with another.

A full description of the aims, tasks to be undertaken and outputs for Phase III is provided in Section 9.5. Only Option 1 is detailed because of the strong pressure from the international technical community to globalise the EVI. Details for Option 2 can be provided on request.

42 Figure 8: Logical framework for the development and completion of the EVI.

This framework describes the tasks completed in Phases I and II and those that will be required to complete the index in a third and final phase. Two options are provided here for completion: Global and Pacific Focus which correspond to Options 1 (recommended) and 2, respectively.

EVI Project Phase I Ÿ Review other vulnerability indices Ÿ Determine whether an EVI is possible Ÿ Construct preliminary EVI centred on Pacific SIDS Ÿ Initial data collection & calculation for 3 countries (Australia, Fiji, Tuvalu)

EVI Project Phase II Ÿ Peer review at Expert Meeting "Think Tank" Ÿ Refinement of model Ÿ Data from 4 countries in Pacific

OPTION 1 EVI Project Phase III OPTION 2

FOCUS - GLOBAL FOCUS - PACIFIC

Ÿ Collect data from 15 differing Ÿ Collect data from all SOPAC types of countries worldwide Pacific island member countries Ÿ Globalise model Ÿ Regionalise model Ÿ Test indicators Ÿ Test indicators Ÿ Test overall EVI for international application Ÿ Test overall EVI for Pacific application Ÿ Think Tank 2 Ÿ Think Tank 2 Ÿ Create user interface (computerised) Ÿ Create user interface (computerised) Ÿ Capacity-building in-country Ÿ Capacity-building in-country Ÿ Set up permanent data-collection mechanisms Ÿ Set up permanent data-collection mechanisms Ÿ Publication Ÿ Publication

43 7 Recommendations and conclusions

7.1 Recommendations

1. The EVI needs support for further development during a third phase before it could be considered functional and acceptable to the international community. 2. The EVI should be globalised. Restricting the index to the Pacific SIDS will result in a lack of context for the conditions found in those countries. It will result in a less powerful EVI because the scaling of indicators will be limited to a narrow range of conditions found in this region. The failure to globalise the EVI is likely to result in an index that will compare one Pacific State with another, rather than identifying them as a group in the world context. It is only in contrast to the global environments that vulnerability of any single state can be assessed. 3. The EVI should be considered at the next CDP session (UN Committee for Development Planning) when indicators for classification of Least Developed Countries (LDC’s) come under review. 4. The EVI should retain its ability to describe country profiles at the same time that it summarises them into a single figure index. The Think Tank participants felt that a written assessment (similar to that provided for each country in this report) should accompany the three levels of reporting already incorporated into the EVI (index, sub- indices and responses for individual indicators). 5. The EVI model should be published in an international peer-reviewed journal. Although some publications were produced during Phase II, this process should be on-going and a paper describing the full model and its indicators published. 6. The EVI Think Tank should be reconvened at a later date to critically review the work done and direct refinements.

7.2 Conclusions

The outcomes of the Think Tank and other inputs from peer review during Phase II revealed that the SOPAC EVI was generally appropriately approached and constructed. Suggested changes to the model and indicators resulting from these reviews were incorporated into the EVI during this phase of the work. This resulted in a provisional EVI that has met with the approval of a range of international experts and potential users.

The provisional EVI is now a viable tool for characterising the vulnerabilities of states. In conjunction with the Think Tank, other meetings and our own work, the EVI is now at a stage that it could be used to collect data from a larger range of countries from the Pacific and other regions for testing.

Although a procedure for the collection of data from Pacific SIDS was developed, it met with mixed success. For three of the four countries trialled in Phase II, sufficient data were collected using a single visit to the country, capacity building and questionnaires for the indicators. This was not true for Vanuatu, for which sufficient data for a valid EVI could not

44 be collected during this study. Our approach to the task of data collection will need to be reviewed during Phase III of the project.

EVI values, sub-indices and vulnerability profiles were calculated for four Pacific Island test countries with exciting results. The overall EVI scores demonstrate the potential of the EVI as a tool for international processes such as the determination of LDC status. The EVI is able to summarise environmental vulnerability information collected for a range of indicators in a form that is comprehensible, visual, and which could be used to identify countries that have high environmental vulnerability. In addition, the vulnerability profiles provided by the sub-indices and scores for the individual indicators demonstrate the powerful potential of the EVI for identifying issues of vulnerability within countries that could be used to prioritise efforts for environmental management.

The EVI was successfully introduced to a range of international bodies, donors, states, environmental managers and experts during Phase II through publication, discussions and attendance at meetings. Support for the project was widespread and opportunities for collaboration and the procurement of data for a larger range of countries were secured. We now expect to be able to collect data from Jamaica, St Lucia, Trinidad & Tobago, Malta and Mauritius in addition to the remaining Pacific Island States.

A work plan for the completion of the EVI was developed during Phase II largely in association the participants at the Think Tank. This plan calls for the globalisation of the EVI, data collection for a wide range of country types, formal testing and the setting of criteria for deciding when the EVI could be considered operational.

45 8 Literature Cited

Briguglio, L., Kaly, U.L. and Pratt, C. 2000. UNEP Report of the meeting of experts on the environmental vulnerability index. In prep. Brown, G. 1998. Selected Pacific Economies: A Statistical Summary. Report for the Secretariat of the Pacific Community, No. 14, Noumea, 9pp. CIA Factbook. 1999. CIA world factbook. http://www.odci.gov/cia/publications/factbook/country.htm FAOSTAT. 1999. FAOSTAT database collections. http://www.fao.org/cgi-bin/nph-db.pl Government of Western Samoa. 1991. Report of the Census of Population and Housing 1991. Department of Statistics, Apia, 16 pp. Paine, J.R. 1991. IUCN Directory of Protected Areas in Oceania. Gland, IUCN. Kadomura, H. 1997. (Ed). Data book of desertification / land degradation. Centre for global research (CGER) National Institute for Environmental Studies (NIES), Japan, 68pp. Kaly, U.L. 2000. Lessons from the EVI and smart indicators for coral reefs. Submitted. Proceedings of a Workshop on Information management and decision support for marine biodiversity protection and human welfare: Coral Reefs, 6-10 December 1999, AIMS, Townsville, Australia. Kaly, U.L., Briguglio, L., McLeod, H., Schmall, S., Pratt, C. and Pal, R. 1999a. Environmental Vulnerability Index (EVI) to summarise national environmental vulnerability profiles. SOPAC Report to NZODA, 38pp plus EXCEL sheets Kaly, U.L., Briguglio, L., McLeod, H., Schmall, S., Pratt, C. and Pal, R. 1999b. Report on the Environmental Vulnerability Index (EVI) Think Tank, 7-10 September 1999, Pacific Harbour, Fiji. SOPAC Technical Report 299 to NZODA, 77pp. Lane, J. 1993. Tuvalu: state of the environment report. Report for the South Pacific Regional Environment Programme, Apia, 64pp. Taule'alo, T. 1993. Western Samoa: state of the environment report. Report for the South Pacific Regional Environment Programme, Apia, 76pp. The Times. 1999. The Times comprehensive atlas of the world: Millenium edition. (10th ed.). Times Books, London. United Nations Economic and Social Commission for Asia and the Pacific (ESCAP) 1997. Statistics Division. Asia and the Pacific in Figures 1997. http://www.unescap.org/stat/statdata/apinfig.htm Villa, F. 1995. Linee guida per la rilevazione e la valutazione dei parametri ambientali richiesti dal progetto ”Rete Natura 2000''. (Italian). S.It.E. Notizie, 24(1), 67-76. Watling, D. and Chape, S. 1992. Environment Fiji: The national state of the environment report. World Conservation Union; Fiji, National Environment Management Project. IUCN, Gland, Switzerland, 154 pp. WRI. 1998. World resources 1998-1999: A guide to the global environment. Oxford University Press, New York, 369pp.

46 9 Appendices

9.1 Summary of EVI data for Fiji, Samoa, Tuvalu and Vanuatu

Indicator Fiji Samoa Tuvalu Vanuatu 1 ND ND ND ND 2 0 ND ND ND 3 35 35 21 ND 4 12.5 18 22 ND 5 0.5 0 0 ND 6 2.5 0 0 ND 7 0 0 0 1 8 0 1.418439716 ND ND 9 ND 0 9 13 10 18272 2820 25.9 12190 11 0.274 0.183 3.142 0.207 12 2650 3800 3300 1800 13 1324 1848 4.58 1877 14 ND 20 100 ND 15 ND 20 100 ND 16 64.9 63.5 38.6 1.07 17 0.547 ND 38.6 ND 18 70.71316769 50.28687943 521.24 ND 19 52.9 186 10039 ND 20 21.7 61.3 309 0.82 21 7.77 ND ND 0 22 44 ND 29 40 23 1.03707 6.58865 51.27413 0.54471 24 ND ND 50-60 0 25 44.49 57.2 438.2 14.93 26 44.49 57.2 438.2 14.93027071 27 1.2 0.68 1.66 2.45 28 7.18 16.6 1 ND 29 0.434 0.544 1.743 0.074 30 7265.21 31458.83 13146.72 4105.82 31 14.13 5.028 65.946 3.542 32 0 0 0 0 33 0.109 0 0 0 34 0 0 0 0 35 1.315 2.061 12.9 0.328 36 ND ND ND ND 37 78.7 375 ND 69.4 38 6.5 22.6 negligible 6.94 39 ND ND ND 0 40 ND 1.06 1-2 34 41 171.27 550 30 275 42 71.682 0 0 0 43 0.1 0.00284 0.0097 0.00656 44 0.31 1.06 0.24 0.595 45 0 0.0008 0.0037 0.0000211 46 ND 0 0 3 47 Draft Legislation (option 1) Draft Draft

47 9.2 Sources of EVI data for Fiji, Samoa, Tuvalu and Vanuatu

Indicator Fiji Samoa Tuvalu Vanuatu 1 ND ND ND ND 2 C ND ND ND 3 C C C ND 4 C C C ND 5 C C C ND 6 C C C ND 7 C C C C 8 C C ND ND 9 ND C C C 10 C SOE C SPC 11 C C C CIA 12 C C C WA 13 CIA C C CIA 14 ND C C ND 15 ND C C ND 16 C C C C 17 C ND C ND 18 C C C ND 19 C C C ND 20 C SOE C C 21 C ND ND C 22 SOE ND C C 23 FAOSTAT FAOSTAT FAOSTAT FAOSTAT 24 C ND C C 25 C C C C 26 C C C FAOSTAT 27 C C C FAOSTAT 28 C C C ND 29 C C C C 30 C C C C 31 C SOE C C 32 SOE SOE C C 33 C C C C 34 C C C C 35 C C C ESCAP 36 ND ND ND ND 37 C C ND C 38 C C C C 39 ND ND ND C 40 ND C C C 41 C C C C 42 C C C C 43 C C C C 44 C IUCN C C 45 C IUCN C C 46 ND C C C 47 C C C C LEGEND C Returned by country on EVI Data Sheets SOE State of the Environment Country Report FAO Food and Agricultural Organisation ESCAP United Nations Economic and Social Commission for Asia and the Pacific (Asia Pacific In figures 1997) IUCN International Union for Conservation of Nature and Natural Resources CIA CIA Factbook WA World Atlas 2000

48 9.3 Publication (Pratt 1999): Data Collection for the Environmental Vulnerability Index (EVI) – Discussion paper

Craig Pratt

1. Introduction

The SOPAC environmental vulnerability index (EVI) is still in development and has been designed as a multi-level model to describe the vulnerability of the natural environment of countries to a range of natural and anthropogenic hazards. The index is being developed in such a way that it can be broken down into sub-indices that describe levels of risk and resilience and the effects of these influences on the health or integrity of a country’s environment. Due to the variety of risks and complexities of ecosystem resilience and integrity, an indicator approach was taken to characterise them.

There are currently a total of 47 indicators in the EVI. Although the SOPAC EVI team has selected a set of indicators with extensive inputs from international reviewers, details are still being finalised on the basis of inputs by experts from different countries and regions.

The formulation and choice of indicators are based on the following criteria: · The indicators should be applicable globally; · They should be based on data already available or easily obtainable; · They should measure change or be a proxy for change which would do significant harm to the environment; · They must not be selected on any political criteria but relate only to environmental vulnerability; · They could be weighted to reflect the probability of change to the environment and the amount of damage which might be done; · They should be relatively easy for users to understand; · They should be well-defined; · They should be as uncorrelated with each other as possible to limit redundancy.

The most vital criterion used in the choice of indicators is that relating to the data and its availability. The availability of appropriate environmental vulnerability data is fundamental to both the development of the EVI and ultimately the final calculation of a country’s EVI value. The success of the EVI as a measure of vulnerability is therefore wholly dependent on accessing and obtaining relevant country environmental data for calculation of EVI values.

In light of the key role that data play in the development of the EVI, this discussion paper has been prepared to provide insight into some of the issues that have been faced in the identification and collection of data for the EVI in the Pacific, some of the lessons learnt and to provide some suggestions as to how we may progress this important process of data collection for the EVI internationally.

49 2. Approach to the Data Gathering Process

The EVI by its very essence attempts to summarise a wide variety of environmental vulnerability data for a country. Much of this environmental data is now only collected and compiled at the national level and reporting and publication of this data internationally is not as yet well-established as the frequent reporting of country economic data. Several international initiatives have been instituted to address this need for international publication of national environmental data and these include such initiatives as the Global Environment Outlook and various others.

The data needed for the EVI includes meteorological data, fisheries data, land area, natural hazard data and so on. The very diverse and wide-ranging nature of these data means that their sources are widely dispersed and require some effort by a country to identify, collect and compile the information. Some of the indicators require information that could only be provided by the authorities or by experts in the respective country. It is therefore essential to have full government co-operation in the data gathering process to ensure success, as has been the case in the Pacific.

The first major issue that arose in the initial stages of trying to gather data for the EVI in the Pacific was the difficulty in creating an understanding of the data required to provide responses to EVI indicators, and the lack of capacity to compile the necessary data. Overcoming these initial problems required the importation of assistance, in-country, to identify the major problems and to try to provide possible solutions so that country environmental vulnerability data files could be compiled.

The in-country approach to data gathering, while beneficial to the country and rewarding in terms of data collection, is not a sustainable method of data gathering in the long term and would be impossible to extend globally. It was therefore decided that a more simple and directed approach should be developed to assist the governments in the gathering of country data. This approach involved the use of detailed questionnaires for each of the EVI indicators.

Each indicator is presented with its detailed indicator question and is accompanied by an explanation of what the proxy indicator is trying to measure. All indicators require a response and guidance is provided towards the possible agency or agencies that may be sources for the information required. Each indicator is also complemented with a clear indication of what data is needed for a complete response to the indicator question.

3. Issues Relating to Data Gathering in the Pacific

During the process of data gathering in the Pacific, several important issues arose. These include difficulties in the following areas: · Data source identification · Accessibility · Availability · Quality

50 · Capacity

3.1 Data Source Identification

The identification of possible data sources and appropriate agencies to approach for the required data has been difficult. This is due to the major differences in bureaucratic structures of the various governments throughout the Pacific. Although agencies may have similar titles they can be given quite different responsibilities and hold different data sets compared with other countries. This has made the identification of appropriate sources and collection of information difficult.

Another issue is the identification of data that may be held by agencies but which may not be known to its officers or which has not be recognised by its officers as relevant to the EVI. This has largely been due to an inability to fully understand an indicator and its data requirements, the changing of staff or just a lack of knowledge of the databases held by the agency.

3.2 Accessibility

Collection, analysis and storage of data is without doubt an expensive exercise. As a result there is increasing recognition of the importance and true value of data which has had a significant impact on access to information. As many government agencies are asked to carry out these tasks on ever-reducing budgets, many are looking to recover their costs through charges for both primary data and time taken by personnel to access and compile required data.

Also in many cases, certain data may be considered sensitive by a country resulting in limited access. Although most data required for EVI indicators would not be considered sensitive, there have been several instances where access to information has required special authorisation. The support of government for the EVI and the data gathering process has therefore been essential in overcoming these problems in the Pacific.

3.3 Availability

Despite international recognition of the value and importance of environmental data in decision-making, collection and the maintenance of these data sets in Pacific Island countries have not always been given priority. In many countries there is either no data collection or it is inconsistent, or when data is collected regularly, there is no proper handling or storage of the data sets leading to incomplete databases and loss of, or poor access to the information.

3.4 Quality

51 on small-sample-biased data may have contributed to these differences. There is also the potential of inaccurate equipment, lack of proper training in measurement procedures, lack of quality control procedures and many other reasons which could all lead to inaccuracies in data reported.

3.5 Capacity

The issue of the lack of capacity is a common one throughout SIDS. In Pacific countries this has been one of the main difficulties in the facilitation of EVI data gathering. The lack of capacity is two-fold in that it involves both institutional as well as personnel capacity problems.

In the Pacific, data gathering exercises like the EVI have placed an added burden on existing institutions’ responsibilities to provide data and information. With limited resources and few trained personnel this task can often be an impossible expectation. The only way to assist these countries facilitate their country collection of environmental data has been to provide in-country input and assistance. This is extremely costly and it is imperative that alternative approaches are found to provide the assistance needed.

Another issue that has arisen is that it has been difficult to create an adequate understanding in personnel on the specific data and information requirements needed for a response to indicator questions. This is due in part to a lack of understanding of the purpose of the EVI, its mechanics and, more generally, the inadequate training of personnel in the identification, collection, analysis and manipulation of data.

4. The Future

The development of an approach to data gathering that is simple and can be easily adopted by countries both large and small is critical to the overall development of the EVI. Currently, SOPAC is attempting to develop an alternative questionnaire approach to facilitate country environmental data gathering. It still requires a lot of improvement and refinement to ensure that users of the EVI are able to get a better understanding of the purpose of the EVI, the data requirements and answers to frequently asked questions to the questionnaires.

A help handbook would prove useful as a possible way to provide a detailed background to the EVI, its mechanics and specific instructions and assistance in how to identify and gather the required information so as to reduce the need for continued assistance and support in the data gathering process.

52 9.4 Publication (Kaly 2000): Lessons from the EVI and smart indicators for coral reefs

U.L. Kaly

Government of Tuvalu, Funafuti, Tuvalu / SOPAC Secretariat, Private Mail Bag, GPO, Suva, Fiji

Abstract

This paper describes lessons learned from the development of the SOPAC Environmental Vulnerability Index (EVI) that could be used in the development of indicators of the status or ‘health’ of coral reefs. The EVI was developed to describe the vulnerability of natural environments at the scale of entire states to a range of natural and human hazards and uses an indicators and index approach. It embodies three methodological aspects that are relevant to providing information for the management of coral reefs. The first of these is the explicit definition of ‘pristine’ conditions as the direction that all management actions should take, acknowledging that different uses of reefs will keep their health at other than pristine levels. The indicators developed for coral reefs would act not only to identify the current status of reefs, but could be used to define thresholds for management. The second insight is that humans are part of the human-coral reef greater system. This means that human choices and behaviours are an integral part of the structure and function of reefs and should be included as indicators of reef health. Finally, I discuss the need for smart indicators for coral reefs. This means that indicators need to be developed that summarise the structure and function of a large number of transactions occurring on coral reefs, without our necessarily knowing how each works or what condition it is in. Smart indicators also have built in them some expression of how well a reef is faring in relation to optimum conditions and other reefs elsewhere.

Introduction

If humans are to be truly able to manage coral reefs they are going to need a reliable, simple, relatively inexpensive way of checking on ‘health’ of the reefs and being able to make some assessment of how they are going in their attempts to manage them in relation to some ideal. This would form part of an iterative system of management with the ability to constantly adjust management options in relation to real outcomes and unpredictable behaviour on the part of reefs. Problems with this adaptive management approach were highlighted recently in a special feature in Conservation Ecology (Gunderson 1999; Johnson 1999). Problems include difficulties in developing acceptable predictive models, conflicts between ecological values and management goals, inattention to non-scientific information, and unwillingness to implement long-term risky or costly programmes. By suggesting that human behaviour be incorporated into the information collected for a coral reef management model, Done and Diop (1999) have in this workshop attempted to take adaptive management a step further. They suggest the incorporation of human systems (socio- economic, use/conservation and policy/management) as part of the overall coral reef

53 complex to be managed. That is, measures of the health and viability of coral reefs includes in its greater context the way that people choose to use and manage them and the socio- economic context from which they make their choices.

The concept of health in relation to reefs is a slippery one. We have no way of indicating the ideal number of species, community characteristics, energy flows or ecosystem services for even a single reef, let alone arrive at some guidelines for the range of complex systems we are concerned with across the globe. Despite not really being ready for the challenge, we are forced by necessity to start taking action (Daily 1999) and learning as we go (Walters and Holling 1990). In the past we have not as a group of scientists, planners and managers looked into coral reef ecosystems to try and find simple measures that would indicate health and provide some way of easily measuring changes through time. We have not looked hard enough for those measures that would provide a proxy or summary for the millions of processes that must be going on within human and coral reef systems and on all kinds of time scales.

In this paper, I wish to focus on how we might go about obtaining the kinds of information we might need through indicators as a basis for management of corals reefs. The information would have to be collected in relation to defined ideal conditions for human and ecosystem measures. I will describe the approach taken for some related work done for assessing the vulnerability of natural environments at the scale of entire states. I will then suggest ways of constructing indicators in relation to targets for the health of coral reefs and the healthy behaviour of humans in relation to them. Finally, I will deal with what I will term ‘smart indicators’ which might be an approach to providing relatively simple measures of how well reefs and their humans are faring in terms of coexistence.

The EVI approach

The Environmental Vulnerability Index (EVI) was developed by the South Pacific Applied Geoscience Commission (SOPAC, Fiji) as a response to calls for measures of the vulnerability of states from the Alliance of Small Island States (AOSIS) and the Barbados Programme of Action (Kaly et al. 1999a). The EVI was designed to provide a relatively quick and inexpensive mechanism for characterising states in terms of the vulnerability of their natural environments to natural and anthropogenic hazards. The alternative would be ad hoc assessments for each state which would be costly in terms of time and resources and which would not provide a common basis for comparison.

The EVI uses 47 indicators of exposure of the natural environments of a state to hazards and their intrinsic and extrinsic resilience to hazards (table 1). The table of indicators is included here because it shows the range and complexity of indicator questions considered necessary to measure vulnerability of natural environments. The measures include both natural and human components, forming three types of sub-index: The REI (Risk Exposure Index) provides information on the types and intensity of risk to natural and anthropogenic hazards. The IRI (Intrinsic Resilience Index) measures signals relating to the innate characteristics of a state which tend to make it resilient to hazards, while the EDI (Environmental Degradation Index) provides an assessment of the present condition of the

54 natural environments assuming that those in the best condition will be most resilient to future shocks. Information collected for a state for each of the indicators is compared with ideal conditions and/or those conditions found world-wide so that data may be mapped onto a 1-7 scale. That is, the conditions for any state with respect to a single indicator will be represented somewhere on that 1-7 scale. The scale itself may be linear, non-linear or discontinuous and was developed to accommodate heterogeneous types of information (yes/no, percentage or numerical). The scores derived from this mapping are then averaged to produce an overall EVI and sub-indices. Because data are mapped on the 1-7 scale, for which a high score is considered more vulnerable, it is possible to use the individual scores as a way of identifying problem areas in terms of risk or environmental degradation (see Kaly et al. 1999a,b).

The EVI relies on four assumptions of importance to this discussion. At its basis, the EVI assumes that (i) the more pristine environments are, the better will be their resilience to natural and anthropogenic shocks. It also assumes that (ii) natural environments in good condition generally serve the needs of humans better than damaged ones. The EVI assumes that (iii) human behaviours, choices and socio-economic conditions are part of environmental vulnerability and seeks to measure these as part of the index. Finally, it assumes that (iv) indicators may be found which describe and summarise a host of complex processes which must be operating and which vary in terms of their final values in a way that relates to (the largely immeasurable) details of interest in the system being measured.

In many ways, the needs to be met by the EVI are similar to those for managing the world’s coral reefs. Although not directly applicable in its present form, the approach taken for the EVI could be adapted to the task of providing information for the protection of aquatic biodiversity, ecosystem management and identifying areas for action.

Table 1. Indicators used for calculating the Environmental Vulnerability Index (EVI). Indicators fall into three sub- indices (S-I) as follows: The REI = Risk exposure sub-index; IRI = Intrinsic Resilience Sub-index; and EDI = Environmental Degradation sub-index. Categories refer to risks to the natural environment as follows: Met = meteorological, G = geological, CC = country characteristics, B = biological, A = anthropogenic. Response levels for the indicators are not provided here.

# S-I Category Indicator 1 REI Met Greatest average annual deviation in Surface Sea Temperature in last 5 years from long term mean (30 years) (more work required to finalise form) (Centralised database) 2 REI Met Number of days over the last 5 years during which the max recorded wind speed (3 sec gusts) >20% higher than the average max for that month (use 30yr average for each month as reference) (Data accumulated over all reference climate stations / # stations) 3 REI Met Number of months over the last 5 years during which rainfall >20% lower than the 30yr average for that month (Data accumulated over all reference climate stations / # stations) 4 REI Met Number of months over the last 5 years during which rainfall >20% higher than the 30yr average for that month (Data accumulated over all reference climate stations / # stations) 5 REI Met Number of days over the last 5 years during which the max temperature >5 C higher than the mean monthly maximum for that month (use 30yr average for each month as reference) (Data accumulated over all reference climate stations / # stations) 6 REI Met Number of days over the last 5 years during which the max temperature >5 C lower than the mean monthly minimum for that month (use 30yr average for each month as reference) (Data accumulated over all reference climate stations / # stations) 7 REI G Number of volcanoes with potential for eruption >= VEI 4 (Volcano explosivity Index) within 100km of country land boundary / area of land

55 8 REI G Earthquake energy within 100km of country land boundaries / land area with ML >=6.0 and <=15km depth per 5 years 9 REI G Number of tsunamis or storm surges with run-up >2m above MHWS / 100km coastline since 1900 10 IRI CC Total land area (sq km) 11 IRI CC Ratio of length of ocean shoreline : total land area 12 IRI CC Distance to nearest continent (km) (*define continent) 13 IRI CC Altitude range (Highest point - lowest point in country) 14 IRI CC Percent of land area <10m above sea-level 15 IRI CC Percentage of land area <10m elevation within 2km of coast composed of unconsolidated sediments (excluding coral reefs) 16 IRI CC Number of known endemic species / 10,000 sq km land area 17 REI B Number of reported (and verified) organism outbreaks over the last 5 years / land area (pathogens, blooms, plagues etc) 18 REI B Total tonnage of freight imported / year 19 EDI B Number of all introduced species / 10,000 sq km land area since 1900 20 EDI B Number of endangered and threatened species / 10,000 sq km of land area (IUCN definitions) 21 EDI B Number species which have become extinct since 1900 / 10,000 sq km land area (IUCN definitions) 22 EDI B Percentage of natural and regrowth vegetation remaining (e.g. forests, mangroves, saltmarshes, prairies, savannah, desert, tundra) 23 EDI B Tonnage of intensively-farmed animal products / yr / land area (includes acquaculture, pigs, chickens, etc) 24 EDI B Percent of fisheries stocks overfished (FAO) 25 EDI A Density of people living in coastal settlements (define area) 26 REI A Total human population density (per sq km land area) 27 REI A Annual human population growth rate (average over last 5 years) 28 REI A Net percentage of land area changed by the removal of natural vegetation over last 5 years 29 REI A Annual number of international tourists * average days stay / 365 / 100 sq km (last 5 years) 30 REI A Megalitres of untreated industrial and domestic wastewater discharged to aquatic system / 1,000 km aquatic ecosystems (length coast+length rivers) 31 REI A Total tonnage of generated and net imported toxic, hazardous and municipal wastes/ 10,000 sq km land area / year (average last 10 years) 32 REI A Mean percent of hazardous, toxic and municipal waste effectively managed or treated / yr 33 REI A Number spills of oil and hazardous substances >1,000 litres during last 5 years on land, in rivers or within territorial waters / land area 34 REI A Number of nuclear, chemical and other major industrial facilities that could cause significant damage / 10,000 sq km land area (*Denominator should be territorial area?) 35 REI A Number of vehicles (World Bank definition) / land area (*Denominator could be length roading or length roads / land area) 36 REI A Max 24 hour SO2 concentration (micro g /cubic m) (average over last 5 years) 37 REI A Tonnes of N,P,K fertilisers used / 10,000 sq km agricultural land area / year (average last 5 yrs) 38 REI A Tonnes of pesticides used / 10,000 sq km of agricultural land / year (average last 5 years) 39 REI A Number of new fisheries stocks or expanded fisheries efforts (>20% increase in catches) added to country over last 5 years (within territory) 40 EDI A % Land area degraded since 1950 (includes salinisation, desertification etc.) 41 EDI A Annual internal renewable water resources per capita (average annual runoff plus recharge of groundwater from endogenous precipitation) 42 REI A Kilotonnes of mining material (ore + tailings) extracted / 10,000 sq km land area / year (average last 5 years) 43 EDI A % Land, rivers and coastal zone affected by mining and quarrying 44 EDI A Percent of terrestrial zone set aside as reserves 45 EDI A Percent of marine zone set aside as reserves (mean high tide to continental shelf) 46 EDI A Number of war or civil strife years over the last 50 years within the territory 47 EDI A Percentage of population with access to safe sanitation (WHO)

Setting targets for indices

There are three aspects of ecosystem condition of concern to us if we are to provide information and decision support for coral reefs. These are: (1) The ideal condition we would like our system to be in; (2) The present state of a reef in relation to this ideal and defined management thresholds; and in order for our management to work in the long run: (3) The

56 total viability and health of coral reefs made up from a composite picture at a range of spatial scales.

The first two assumptions of the EVI relate to explicitly setting targets for the condition of natural environments against which the performance of a particular state could be measured. This applies also to coral reefs in the present discussion. It needs to be explicitly- stated that ‘pristine’ or ‘natural’ conditions of a reef are the direction in which all management should aim, and that thresholds along that arrow may mark certain management choices (figure 1). The purpose of indicators and indices would then be to locate a particular reef or group of reefs in terms of their health along the axis from “degraded’ to ‘pristine’.

These above points relate specifically to not setting management arrows so that they point towards a particular usage type or management zone (e.g. open access fishing area). Doing so would make it difficult to assess damage sustained through human pressures because the new management target could become an artificial reef community which might not best serve our aims at preserving biodiversity and human welfare. There is also a question of being able to accumulate coral reef data over larger scales to examine overall viability and status of the world’s reefs. To do this, we need a common basis for comparison – I suggest this common basis should be some measure of ‘pristineness’ of reefs. The health of the world’s coral reefs depends on the overall picture made up of its many pixels, highlighting the value of GIS.

Figure 1. Management arrow which explicitly states that ‘natural’ or ‘pristine’ conditions are the direction of management strategies because these conditions ultimately support all coral reef processes and human welfare. Also shown are management options (A-D) as deformations of the line running from pristine to degraded. This reinforces the idea that certain uses of a reef can deform the way it is structured and functions. The solid arrow is an expression of the current state of the reef. A particular reef may be being managed along deviation line C. Along line C are some management thresholds showing the lower acceptable limit for health and an upper limit which could lead to the decision for more exploitation (horizontal lines).

PRISTINE

D A B C “Natural ”

DEGRADED

57 Humans as part of the reef system

Humans are explicitly incorporated as indicators in the EVI, and in the categories identified by Done and Diop (1999). A total of 38% of the EVI’s indicators are direct measures of human choices and behaviour in relation to environmental vulnerability. There are socio- economic indicators: 25, 26, 27, 46, 47; those dealing with use/conservation: 18, 23, 29, 30, 32, 33, 34, 35, 37, 38, 39; and policy/management: 44, 45 (table 1). In addition to these, indirect human measures are implicit in some of the remaining indicators (e.g. indicators 17, 19, 20, 21). By incorporating humans directly into the model and indicators, it is expected that we will be better able to observe interactions between humans and their natural environments and alert institutions when any part of the system starts to slide backwards along the management arrow. Clearly, the search for indicators for the management of the world’s coral reefs will need to incorporate humans as an explicit and fully integrated part of the information and decision-support system.

‘Smart indicators’ for checking the health of reefs

‘Smart indicators’ could be defined as those which capture a large number of elements in a complex interactive system while at the same time showing how the value obtained relates to some ideal or agreed-upon condition. The central aim of the EVI was to populate it with only smart indicators, and the search for appropriate smart indicators is on-going. The basic assumption of smart indicators is that the value of a chosen indicator is a culmination of perhaps millions of transactions that must have been operating appropriately to result in the value obtained. Thankfully, this does not require our full knowledge of every transaction because if this were a requirement, we would never be able to use indicators at all. Further, to some extent all indicators are ‘smart’ – this is essentially a search for the smartest and most efficient for our purposes. In the EVI for example, indicator 36 deals with the maximum SO2 concentration in the largest city of a state. This indicator is obviously intended to capture industrial discharges, but has within it the density of those discharges, choices people make about the form that discharge takes and the ability of the environment to attenuate them. In turn, the ability of the environment to attenuate discharges captures other features of the system such as forest cover, wind patterns etc.

In the case of coral reefs, a similar search for smart indicators will be required. Some possible avenues for development were identified during the workshop and possible smart indicators were examined for corals, other invertebrates, fishes, recruitment and physical features. Humans and other reef categories will need to be incorporated in the future.

Central to the concept of smart indicators is the idea that good and bad performance is inherent in the way the indicator is expressed. For the EVI this was achieved by mapping the value of an indicator on a scale ranging from 1 (relatively resilient) to 7 (very vulnerable). This approach serves two important functions. Firstly, for the managers and scientists it forces us to decide on actual values for the indicator that are ‘good’ or ‘bad’. This has not been an easy task because information on the vulnerability and limits to viability of ecosystems is generally lacking (Daily 1999). Further, for the EVI it was necessary to make

58 the index globally-applicable. This means that the range of values for an indicator not only has to indicate vulnerability, but also be applicable across all conditions found on the planet !

The second function of inherently expressing the value of an indicator is of greatest interest to the users of the EVI. Irrespective of whether the final EVI is calculated or not, any single indicator gives us a performance rating for that measure that contains within it our best understanding of how ecosystems respond to hazards. This means that the user need not be an expert to read the results - the work has already been done in the selection of the indicators and the setting of response levels. For users, smart indicators mean near-instant results, making them amenable to use by non-scientists, until now one of the weaknesses of adaptive methods of management (Shindler and Cheek 1999).

For coral reefs it may be sufficient to provide cut-off levels for each indicator, rather than mapping them onto a graded scale. One promising smart indicator suggested by Cliff Hearn at this workshop took the form of measuring the amount of incident wave energy translated across a reef. For a healthy reef, the amount of energy passing the reef should be no more than 10% of the incident energy. This smart indicator has built into it the idea of a limit so that users can understand whether their reef is healthy or not, and incorporates information on hundreds of measures and processes that must be operating well to pass the test. These processes might include good recruitment of corals and fishes, high rugosity of the reef, good cover by corals and algae, a good upwards growth rate and all of the minor and major transactions that would be going on within a healthy reef ecosystem to result in the translation of only 10% of incident wave energy. It is a range of indicators such as this that is urgently needed for better management of coral reefs.

Conclusions

Providing information and decision support for the urgent task of pulling the world’s coral reefs back from the brink requires, in part, the development of rapid methods for assessing their health. The approach taken during the development of the Environmental Vulnerability Index provides us with some direction for coral reefs. The first lesson derived from the EVI is that the indices require the explicit statement of ideal conditions that reefs should be in, and that this ideal should be on a continuum that points towards natural or pristine conditions. The second lesson is that humans are part of the coral reef system and that their choices and behaviours need to be monitored as part of a reef’s health just as importantly as any measure of diversity or ecosystem structure or function. This suggests that measures of coral reef health need to include human indicators. Finally, I have introduced the concept of smart indicators in the context of coral reefs. For indicators to provide the information necessary and to bridge the distance between science and users, it is necessary that they be end-point indicators. This requires that they be crafted so that they embody details of which we might not be aware and that they convey in their expression an immediate understanding of the status of coral reefs.

59 Acknowledgements

The development of the EVI has been generously supported by New Zealand ODA. Additional support was given by the United Nations Environment Programme (UNEP), the governments of Fiji, Samoa, Tuvalu, Vanuatu and Australia, the South Pacific Regional Environment Programme (SPREP), the Pacific Island Forum Secretariat, the Secretariat for the Pacific Community, the University of Malta and the World Meteorological Organisation (WMO). Thanks to Claudia Ludescher and Craig Pratt for comments on this manuscript.

References

Daily, G.C. 1999. Developing a scientific basis for managing earth’s life support systems. Conservation Ecology 3(2): 14 [online] http://www.consecol.org/vol3/iss2/art14 Done, T. and Diop, S. 1999. Information management and decision support for marine biodiversity protection and human welfare: coral reefs. Unpublished background paper, 14pp. Gunderson, L. 1999. Resilience, flexibility and adaptive management – Antidotes for spurious certitude? Conservation Ecology 3(1): 7 [online] http://www.consecol.org/vol3/iss1/art7 Johnson, B.L. 1999. Introduction to the special feature: Adaptive Management – Scientifically sound, socially challenged? Conservation Ecology 3(1): 10 [online] http://www.consecol.org/vol3/iss1/art10 Kaly, U.L., Briguglio, L., McLeod, H., Schmall, S., Pratt, C. and Pal, R. 1999a. Environmental Vulnerability Index (EVI) to summarise national environmental vulnerability profiles. SOPAC Report to NZODA, 38pp plus EXCEL sheets. Kaly, U.L., Briguglio, L., McLeod, H., Schmall, S., Pratt, C. and Pal, R. 1999b. Report on the Environmental Vulnerability Index (EVI) Think Tank, 7-10 September 1999, Pacific Harbour, Fiji. SOPAC Technical Report 299 to NZODA, 77pp. Shindler, B. and Cheek, K.A. 1999. Integrating citizens in adaptive management: A propositional analysis. Conservation Ecology 3(1): 9 [online] http://www.consecol.org/vol3/iss1/art9 Walters, C.J. and Holling, C.S. 1990. Large-scale management experiments and learning by doing. Ecology 71:2060-2068.

60 9.5 SOPAC funding proposal for Phase III – Option 1 globalising and testing the EVI

NAME OF PROJECT Environmental Vulnerability Index (EVI) Project Phase III

TYPE OF PROJECT Development of a robust global environmental vulnerability index

AUTHORITY SUBMITTING THE PROJECT South Pacific Applied Geoscience Commission (SOPAC) 4

IMPLEMENTING ORGANISATION South Pacific Applied Geoscience Commission (SOPAC)

RECIPIENT STATES Pacific Countries

COMMITMENT PROPOSED AS A GRANT USD $604,307 DURATION OF PROJECT 2 YEARS

The project proposal has been endorsed by the SOPAC Governing Council during its 28th Annual Session for submittal for funding.

Keywords: Pacific Regional initiative, Environmental Vulnerability, National level, Pacific Island Countries, Global issues, LDC status.

1. Background

In July 1998, the United Nations Economic and Social Council (ECOSOC) in its resolution 1998/39 requested the Committee for Development Planning to report on the usefulness of a vulnerability index as a criterion to designate least developed countries (LDC).

In CDP’s recent report to ECOSOC 26 – 30 April 1999, the CDP stated that the best approach to take explicit account of economic vulnerability in the designation of LDCs would be to utilise a composite Economic Vulnerability Index. CDP also stressed that this economic vulnerability index will give only a partial and appropriate measure of vulnerability for a country. CDP also emphasised that the vulnerability of developing countries is a much broader issue which also has ecological and social aspects which should be a priority for international research activities and work for the Committee. The CDP also noted the important work and related debates relating to the development and use of vulnerability indicates being undertaken by SOPAC, the Caribbean Development and the Commonwealth Secretariat.

4 SOPAC Member Countries: Australia, Cook Islands, Federated States of Micronesia, Fiji, French Polynesia (Associate), Guam, Kiribati, Marshall Islands, Nauru, New Caledonia (Associate), New Zealand, Niue, Papua New Guinea, Samoa, Solomon Islands, Tonga, Tuvalu and Vanuatu.

61 Currently, it is our understanding that ECOSOC has deferred decision on the CDP recommendations. SOPAC is currently seeking clarification on this and implications for Pacific SIDS. It is also our understanding that both the EU and the United States have expressed concerns about the issue of vulnerability and the CDP report. The EU position recognises vulnerability as an issue but clearly states that the EU has concerns over proposed mechanisms for its use and measurement.

It is possible that the deferral of the ECOSOC decision on the inclusion of vulnerability indices as a criterion for assessing least developing country status could be tabled at ECOSOC 2001 but a decision is not expected to be delayed any further. It is therefore essential that all development work on the EVI be completed before March 2001.

2. Project Aim

The EVI is a Pacific Regional initiative. Its purpose is to provide a measure that will simplify the otherwise expensive and difficult task of carrying out detailed assessments of countries to provide a statement of the likely vulnerability of the environment to natural and human hazards. It also provides a common basis for comparison among countries.

The overall aim of this project is to complete the development of a fully functional and tested EVI, so that it is capable of serving these functions. To achieve this aim, it will be necessary to complete the following tasks:

1. Fully define, set upper and lower limits to, and set scoring scales for the indicators of environmental vulnerability identified to date, and any new ones which might be required; 2. Collect data from a total of 15 countries with a range of characters and conditions (deserts, rainforests, tropical, temperate, large, small, etc) for testing the EVI model; 3. Test indicators for redundancy using the data from 15 widely different types of countries and from there adjust them and determine appropriate weightings; 4. Test the EVI model against a priori criteria for determining when it would be acceptable for use as an international tool. This includes an independent test of the index to be made by sending professionals to 5 countries to make independent assessments of vulnerability against which the usefulness of the EVI could be examined; 5. To develop a simple robust computer user interface for the environmental vulnerability index model to facilitate data entry and streamline calculation; 6. To build capacity within the South Pacific region to identify, analyse and collect data for the environmental vulnerability index; 7. To develop a sustainable data collection mechanism for the EVI using in-country and international resources.

3. Introduction

62 to undertake such studies on an ad hoc basis would be prohibitively expensive and/or lead to results which, for international processes, might be difficult to compare. The solution to this problem has been the attempt, over the past decade, to construct vulnerability indices that could be calculated for states on a common basis. The aim of vulnerability indices, then, is to describe and summarise the relative vulnerability of states. A range of these indices has been developed to describe vulnerability of human systems to harm from other human influences or natural ones.

Economic and quality of life indices are well on the road to completion. For natural environments, SOPAC’s work is the first of its kind. The first two phases of the environmental vulnerability index project were funded by the New Zealand Government in support of a Commission on Sustainable Development initiative as part of the implementation of the Barbados Programme of Action. The Barbados Programme of Action called for the development of a composite vulnerability index that reflects the risks and resilience of Small Island Developing States (SIDS) and integrates ecological fragility and economic vulnerability.

The SOPAC environmental vulnerability index is still in development and has been designed as a multi-level model to describe the vulnerability of the natural environment of countries to a range of natural and anthropogenic hazards. The work includes the development of an overall environmental vulnerability index, which is broken down into sub-indices that describe levels of risk and resilience and the effects of these influences on the health or integrity of a country’s environment. This information can be presented as an individual index or as sub-indices which can all be detailed into a full country environmental vulnerability profile to highlight specific areas of concern. It is important that all of these aspects of environmental vulnerability be developed simultaneously so that the final product can be widely applicable to different aspects of decision-making both among and within countries. It is also envisaged that the index would be recalculated periodically (every 5 years) to provide information on the effectiveness of actions taken by countries that have altered their relative vulnerabilities.

The SOPAC-developed environmental vulnerability index and methodology selected for its computation (SOPAC Technical Report 275), recently underwent critical peer review at an International Expert Group Meeting held in Fiji, 7–10 September 1999. At this meeting, the index and its methodology were exposed to peer review and debate to determine whether the approach taken was technically acceptable and that the index could be refined into a robust and internationally applicable measure.

63 country profiles on the hazards, intrinsic features and environmental health likely to affect environmental vulnerability; III. There is a need to look further into the development of a separate vulnerability index which examines the human environment (more content on direct impacts on human systems – vulnerability of human quality of life index); IV. The EVI should be tested using data from no less than 15 widely differing countries; and V. Should be published in an international peer-reviewed journal

The index itself was modified during the Expert Group Meeting (SOPAC Technical Report 299). Changes were made to the structure of the model (framework) and to the indicators which ‘capture’ or ‘measure’ vulnerability. Most of the structural changes can be incorporated immediately, while changes to the indicators will require work to redefine them, set their scaling within the global context and remove redundancies. The first part of this refinement will be completed during Phase II of the project using data from 5 countries centred in the Pacific Region (Australia, Fiji, Samoa, Tuvalu and Vanuatu). This third report is expected to be available in March 2000 (Table 1).

The EVI being developed by SOPAC requires further development and testing against criteria set by the expert group during the think tank before it can become fully operational (Table 1). This will involve testing of the environmental vulnerability index mathematically and with real international country data to refine the model and its indicators and to ensure its global applicability, workability and robustness. This work will also include in-country capacity-building where required, the development of a computer interface to simplify data collection and calculation of the EVI and the establishment of a permanent, international system for collection of the data. Through the process of capacity building, consultation and involvement by Pacific Island Countries, the formulation and design of an environmental vulnerability index, as a useful environmental management and policy decision-making tool that meets the needs of region will be developed. To be effective, the EVI needs to be global since it is in the global context that PIC’s can benefit the most from the index.

Several major tasks have been identified for the successful full development of the EVI as a tool for measuring and managing the vulnerability of countries (see also Table 1). These are presented in the format of modules, some of which must be run in parallel with others for the work to be undertaken efficiently.

64 Table 3: Summary of phases and tasks undertaken since September 1998 and plans for future work

Phase Description Accompanying report Phase I 1. Review other vulnerability indices SOPAC Technical Report 275 2. Determine whether an EVI is possible (January 99) 3. Construct preliminary EVI centred on Pacific SIDS 4. Initial data collection & calculation for 3 countries (Australia, Fiji, Tuvalu) Phase II 1. Peer review at Expert Meeting “Think SOPAC Technical Report 299 (October Tank 1” 1999) 2. Refinement of model Technical Report due March 2000 3. Data from 5 countries in Pacific Phase III 1. Collect data from 15 differing types of Work to begin in March 2000 countries worldwide 2. Globalise model 3. Test indicators 4. Test overall EVI 5. Think tank 2 6. Create user interface (computerised) 7. Capacity-building in-country 8. Set up permanent data-collection mechanisms 9. Publication

SOPAC is now seeking funding to take the project into Phase III and to complete the work. The funding is envisaged primarily as being assembled as a single package through co- funding from donors (though some of the modules may be identified separately for funding).

It is important to highlight the need to secure funding for all modules to ensure the complete development and refinement of the EVI. SOPAC is currently seeking immediate funding for Module 1 that will guarantee the continued development and refinement of the EVI. The acquisition of funding for other modules will enhance and expand the development and refinement process during Phase III.

The basic modules include: Module 1: Refinement and Comprehensive Testing of the Environmental Vulnerability Index (Collecting data from 15 countries and profiles from 5, Globalising the model, testing indicators, testing the overall index, Think Tank 2, publication) Module 2: Pacific EVI Country Capacity-Building Module 3: Sustainable Data Collection Process for the Environmental Vulnerability Index (through in-country and international agencies) Module 4: Computer Environmental Vulnerability Index Interface Development Module 5: Validation of the EVI The modules are currently presented in concept with estimated budgets and timeframes. Detailed proposals for each module can be provided as required.

All tasks will be carried out in consultation and partnership with all Committee for Regional Organisations in the Pacific (CROP) members including: South Pacific Regional Environment Programme, Forum Secretariat, Secretariat of the Pacific Community,

65 University of the South Pacific, Forum Fisheries Agency, Pacific Island Development Program, Tourism Council of the South Pacific, non-government organisations and other regional agencies. Particular inputs will be sought from all Pacific island governments and the international agencies with a mandate for international affairs and particularly SIDS.

4. Modules Required for EVI Development and their Justification

Module 1: Refinement and Comprehensive Testing of the EVI

Aim: To operationalise the EVI through extensive testing with real environmental vulnerability data from at least 15 countries that represent the vast variety of different environments found globally

Tasks: Collecting data from 15 countries, Profiles from 5 countries for validation, Globalising the model, Testing indicators, Testing the overall index, Think Tank 2, Publication.

Technical development of the environmental vulnerability index will require that the model and its indicators be extensively tested with environmental data from at least 15 countries with widely differing characteristics. These countries have to represent the variety of environmental and risk conditions found on the globe so that the EVI can be set within a world-wide context. The real data obtained from these countries will be used to define the parameters within which the conditions of any single country can be identified.

Examples of countries could include high mountainous countries like Nepal or Bolivia, small land-locked countries like Lesotho and Switzerland, countries with a large size embracing numerous ecosystems like the United States of America, highly industrialised countries constituted mainly of man-made environments like Germany, or countries with cold climates like Greenland or Russia. Data collection from these countries would be additional to any country information collected within the Pacific.

It is expected that the selection of 15 countries will include several developed countries whose voluntary participation and provision of data is essential to the process of data collection. It is also envisaged that the participation of several developing countries may require technical assistance to facilitate the collection of environmental vulnerability data files and this will need to be co-ordinated by the project.

The environmental vulnerability index will also need to undergo extensive mathematical and statistical analysis to fully refine and operationalise it. The testing will include correlation analyses to identify and discard redundant indicators, weighting of indicators and sensitivity analyses.

66 Outputs/Benefits:

· An index which is globally applicable while at the same time providing context for its application in the Pacific Region · Environmental vulnerability data for 15 international countries which show their relative vulnerabilities for a range of variables which affect or characterise their environments · A fully mathematically and statistically tested model using real data from a diverse range of countries leading to a refined and robust EVI

Module 2: Pacific Environmental Vulnerability Index Country Capacity-Building

Aim: To provide training to SIDS country representatives in the use, data collection and computation of the environmental vulnerability index.

For some countries, data may be obtainable by contacting the relevant departments and asking for assistance with data acquisition. For most SIDS, this approach is unlikely to yield adequate results. It will therefore be necessary for the EVI Team to carry out a capacity building workshop with representatives from at least 10 member countries to introduce them to the environmental vulnerability index concept, its workings and information required for its computation. Also included in this capacity-building exercise is follow-up work for the 10 countries.

The workshop participants would be provided with an introduction to the environmental vulnerability index, the underlying concepts, benefits and uses of the model in environmental management and decision-making, and an introduction into the mechanics and computation of the EVI. The aim of the introduction would be to provide participants with the basic skills needed to determine information needs for the EVI together with the identification of possible sources of information in their respective countries and how to collect appropriate data. The knowledge gained through the workshop would allow participants to provide their countries with detailed explanations on the need for environmental management information, the conceptual mechanics and overall benefits of the environmental vulnerability index and its application in environmental management.

To ensure the effectiveness of the workshop in providing country representatives with the appropriate skills for data manipulation, it will be necessary to undertake follow-up work. This will entail visits by the EVI Team to countries to facilitate and assist in the collation of environmental vulnerability data for all participating countries.

67 Outputs/Benefits:

· Trained country personnel in all aspects of EVI calculation · Strengthening of country capacity to identify, collect and compile environmental information · Improved understanding of the environmental vulnerability index, its uses, mechanics and applications in environmental management and policy making · Environmental vulnerability data for at least 10 Pacific countries · Pacific ownership of the EVI and the work that has been carried out in the region

Module 3: Development of a computerised EVI Interface

Aim: To develop a user interface in Microsoft Access for the EVI

The model developed for calculating the EVI during Phase I was built into EXCEL spreadsheets. This makes calculating the EVI cumbersome for the endpoint users and was not intended for use by them because files may be easily damaged or altered, making calculations inaccurate.

During Phase III, the EVI model will be built into a user-friendly ACCESS application with online help to allow for easy and streamlined input of data. The computer interface will also make the mathematical calculations robust to damage by users and secure the generation of final reports into quick and simple formats with graphical outputs.

Output/Benefits:

· The development of a robust and simple user-friendly Microsoft Access interface for the environmental vulnerability index model · Simplified and streamlined entry of environmental vulnerability index data with enhanced generation of final environmental vulnerability index values and reports

Module 4: Sustainable Data Collection Process for EVI

Aim: To develop a sustainable mechanism for EVI data collection and computation every 5 years

When the EVI is internationally accepted and set in place as a regular tool to assess the environmental vulnerability of countries world-wide, a sustainable strategy is needed to establish data collection and analysis.

Sustainable data collection has to be institutionalised, transparent and as effective and economic as possible. A suitable mechanism has to be identified which will ensure the co- ordination and collection of country environmental vulnerability data. This should also be a world-wide initiative and be accepted as such.

68 Potential mechanisms for data collection by country institutions and reporting mechanisms to international organisations responsible for international data collation and analysis will need to be investigated for applicability possible use. For example, one possible mechanism could be the collection of national environmental data through the Global Environment Outlook (GEO) process. This mechanism may be able to provide a suitable platform through which data is collected nationally and then collated internationally. Countries may then also utilise the information at the national level for their own purposes.

Outputs/Benefits:

· Identification of an appropriate international mechanism for the continual collection of environmental vulnerability data by all countries · Identification of an appropriate approach to the collection of country environmental vulnerability data for use both within country and internationally · The continual collection of environmental vulnerability data by countries which will allow EVI values to be calculated at repeated intervals by all countries

Module 5: Validation of the EVI

Aim: To independently peer review and validate the EVI

Peer review is essential throughout the development of the EVI to ensure international applicability and robustness. Because the task is multi-disciplinary and breaks new ground, it will be necessary to obtain rigorous technical inputs through a second think tank and through publication in an international academic journal. International experts from the fields of climate, geology, ecology, statistical methods, modelling, environmental assessment and management, economics and indexing will need to be consulted in a forum which allows their areas of expertise to be integrated.

An important part of this validation exercise of the EVI involves the independent assessment of 5 countries by consultants not initiated in the workings of the EVI. These detailed country vulnerability assessments will be compared against environmental vulnerability index values to ascertain their correlation. This was considered by think tank participants to be the ‘acid test’ for determining when the EVI could be considered operational.

Outputs:

· A globally applicable scientifically sound, robust, fully tested and validated EVI

5. Overall Expected benefits from the project

69 that the initiative remains with the region and that the needs and expectations of those countries are incorporated into the development of the EVI. It will also mean that the capacity of Pacific Island Countries for environmental management using tools like the environmental vulnerability index are secured.

6. Overall Project Outputs

The major outputs from Phase III will be: · Pacific ownership of the EVI and the work that has been carried out in the region · An internationally tested, robust, validated and operational EVI which is transparent in its strengths, biases and functions as a way of summarising the environmental vulnerability of states · Capacity-building within the region in the use, mechanics and applications of the EVI in environmental management and policy making · Strengthened capacity in Pacific Island Countries in the identification, collection and analysis of environmental management data · Environmental vulnerability data for at least 10 Pacific island countries and 15 international countries which show their relative vulnerability’s of a range of variables which affect or characterise their environments · A user-friendly computer programme which can be sent to all countries wishing to calculate their EVIs (some with assistance) · An EVI that can be combined with other vulnerability indices into a composite vulnerability index · Identification of an international mechanism for the continual collection of environmental vulnerability data by all countries

70 10. Acronyms & definitions for indicators

AIMS Australian Institute of Marine Science AOSIS Alliance of Small Island States CDP United Nations Committee for Development Policy COP Conference of the Parties (to the Convention on Climate Change) CROP Committee for Regional Organisations in the Pacific CSD Commission on Sustainable Development ECOSOC United Nations Economic and Social Council EVI Environmental Vulnerability Index GEF Global Environment facility GEO Global Environment Outlook IDNDR International Decade for Natural Disaster Reduction IOC Intergovernmental Oceanographic Commission IPCC Intergovernmental panel on Climate Change LDC Least Developed Country PACCLIM Pacific Island Climate and Sea Level Change Model PICCAP Pacific Islands Climate Change Assistance Programme SIDS Small Island Developing States SIS Small Island States SOPAC South Pacific Applied Geoscience Commission SPTO South Pacific Tourism Organisation UN United Nations UNCTAD United Nations Conference on Trade and Development UNDESA United Nations Department of Economic and Social Development UNDP United Nations Development Programme UNEP United Nations Environment Programme UNFCCC United Nations Framework Convention on Climate Change WRI World Resources Institute

Degraded (Indicator 40) Reduction or loss in arid, semi-arid and dry sub-humid areas, of the biological or economic productivity and complexity of rainfed cropland, irrigated cropland, or range, pasture, forest and woodlands resulting from land uses or from a process or combination of processes, including processes arising from human activities and habitation patterns, such as: soil erosion caused by wind and/or water; deterioration of the physical, chemical and biological or economic properties of soil; and long-term loss of natural vegetation. This definition is the standard for Agenda 21 and the Convention to Combat Desertification (Kadomura, 1997). Endangered (Indicator 20) A taxon is endangered when it is not critically endangered (see below) but is facing a very high risk of extinction in the wild in the near future, as defined by any of the criteria for “endangered” in Table 4. In our indicator, we are including the IUCN category of “Critically

71 endangered” which includes species which are facing an extremely high risk of extinction in the wild in the immediate future, as defined by any of the criteria for “critically endangered” in Table 4. IUCN Red List category 1994. Endemic (Indicator 16) Species which occur exclusively in the country and not elsewhere. Extinct in the wild (Indicator 21) A taxon is extinct in the wild when it is known only to survive in cultivation, in captivity or as a naturalised population (or populations) well outside the past range. A taxon is presumed extinct when exhaustive surveys in known and/or expected habitat, at appropriate times (diurnal, seasonal, annual), throughout its historic range have failed to record an individual. Surveys should be over a time frame appropriate to the taxon’s life cycle and life form. Extinct (Indicator 21) A species is extinct when there is no reasonable doubt that the last individual has died. IUCN Red List category1994. Intensive farming (Indicator 23) For the EVI, intensive animal farming includes all farming practices for which the waste products produced cannot be largely attenuated over the land on which they are conducted. This includes, but may not be limited to most poultry farming, pig farming and intensive aquaculture. For the countries we examined, data on poultry and pigs was often returned as head of animals produced per year, rather than tonnage. We used the following conversion rates for this study: chickens and ducks each weighing 2kg and pigs each weighing 100kg. ML-6 (Indicator 8) Earthquake energy units. The ML or Magnitude Local scale is equivalent to the Richter Scale. It measures the magnitude of an earthquake as the logarithm of the amplitude of waves recorded by seismographs, adjusted for variation in the distance between seismographs and the epicentre of earthquakes. For our indicator, the earthquakes of interest must be 6.0 or stronger and must occur within 15km of the earth’s surface. Vehicle (Indicator 35) This uses the World Bank definition as follows: ***** VEI-4 (Indicator 7) Expression of potential for volcanic eruption based on plume height, volume of eruption, classification of the volcano and frequency of eruption. VEI-4 is a cataclysmic volcano with a plume height of 10-25km, lava/ash volumes of 100,000,000s of m 3, of vulcanian / plinian classifications, and erupting on a time scale of 10’s of years. Vulnerable (Indicator 20) A taxon is vulnerable when it is not Critically endangered or Endangered but is facing a high risk of extinction in the wild in the medium-term future, as defined by any of the criteria for “vulnerable” in Table 4. IUCN Red List category 1994.

72 Table 4: IUCN Criteria for the categories Endangered and Vulnerable, 1994

IUCN Category Criteria Details CRITICALLY A. Population reduction in the 1. An observed, estimated, inferred or suspected a) direct observation ENDANGERED form of either of the following: reduction of at least 80% over the last 10 years or b) an index of abundance three generations, whichever is the longer, based appropriate for the taxon on (and specifying) and of the following: c) a decline in area of occupancy, extent of occurrence and/or quality of habitat d) actual or potential levels of exploitation e) the effects of introduced taxa, hybridisation, pathogens, pollutants, competitors or parasites. 2. A reduction of at least 80%, projected or suspected to be met within the next 10 years or three generations, whichever is the longer, based on (and specifying) any of b), c), d) or e) above. B. Extent of occurrence 1. Severely fragmented or known to exist only at a estimated to be less than single location 100km2 or area of occupancy estimated to be less than 10km2, and estimates indicating any two of the following: 2. Continuing to decline, observed, inferred or a) extent of occurrence projected, in any of the following: b) area of occupancy c) area, extent and/or quality of habitat d) number of locations or sub- populations e) number of mature individuals 3. Extreme fluctuations in any of the following: a) extent of occurrence b) area of occupancy c) number of locations or sub- populations d) number of mature individuals. C. Population estimated to 1. An estimated continuing decline of at least 25% number less than 250 mature within three years or one generation, whichever is individuals and either: longer or 2. A continuing decline, observed, projected, or a) severely fragmented (i.e. no inferred, in numbers of mature individuals and sub-population estimated to population structure in the form of either: contain more then 50 mature individuals) b) all individuals are in a single sub-population. D. Population estimated to number less than 50 mature individuals E. Quantitative analysis showing the probability of extinction in the wild is at least 50% within 10 years or 3 generations, whichever is the longer. ENDANGERED A. Population reduction in the 1. An observed, estimated, inferred or suspected a) direct observation form of either of the following: reduction of at least 50% over the last 10 years or b) an index of abundance three generations, whichever is the longer, based appropriate for the taxon on (and specifying) and of the following: c) a decline in area of occupancy, extent of occurrence and/or quality of habitat d) actual or potential levels of exploitation e) the effects of introduced taxa, hybridisation, pathogens, pollutants, competitors or parasites. 2. A reduction of at least 50%, projected or suspected to be met within the next 10 years or three generations, whichever is the longer, based on (and specifying) any of b), c), d) or e) above. B. Extent of occurrence 1. Severely fragmented or known to exist at no

73 estimated to be less than more than 5 locations 5000km2 or area of occupancy estimated to be less than 500km2, and estimates indicating any two of the following: 2. Continuing to decline, observed, inferred or a) extent of occurrence projected, in any of the following: b) area of occupancy c) area, extent and/or quality of habitat d) number of locations or sub- populations e) number of mature individuals 3. Extreme fluctuations in any of the following: a) extent of occurrence b) area of occupancy c) number of locations or sub- populations d) number of mature individuals. C. Population estimated to 1. An estimated continuing decline of at least 20% number less than 2500 within 5 years or 2 generations, whichever is mature individuals and either: longer, or 2. A continuing decline, observed, projected, or a) severely fragmented (i.e. no inferred, in numbers of mature individuals and sub-population estimated to population structure in the form of either: contain more then 250 mature individuals) b) all individuals are in a single sub-population. D. Population estimated to number less than 250 mature individuals E. Quantitative analysis showing the probability of extinction in the wild is at least 10% within 100 years. VULNERABLE A. Population reduction in the 1. An observed, estimated, inferred or suspected a) direct observation form of either of the following: reduction of at least 20% over the last 10 years or b) an index of abundance three generations, whichever is the longer, based appropriate for the taxon on (and specifying) and of the following: c) a decline in area of occupancy, extent of occurrence and/or quality of habitat d) actual or potential levels of exploitation e) the effects of introduced taxa, hybridisation, pathogens, pollutants, competitors or parasites. 2. A reduction of at least 20%, projected or suspected to be met within the next 10 years or three generations, whichever is the longer, based on (and specifying) any of b), c), d) or e) above. B. Extent of occurrence 1. Severely fragmented or known to exist at no estimated to be less than more than 10 locations 20,000km2 or area of occupancy estimated to be less than 2000km2, and estimates indicating any two of the following: 2. Continuing to decline, observed, inferred or a) extent of occurrence projected, in any of the following: b) area of occupancy c) area, extent and/or quality of habitat d) number of locations or sub- populations e) number of mature individuals 3. Extreme fluctuations in any of the following: a) extent of occurrence b) area of occupancy c) number of locations or sub- populations d) number of mature individuals. C. Population estimated to 1. An estimated continuing decline of at least 10% number less than 10,000 within 10 years or 3 generations, whichever is mature individuals and either: longer or 2. A continuing decline, observed, projected, or a) severely fragmented (i.e. no inferred, in numbers of mature individuals and sub-population estimated to

74 population structure in the form of either: contain more then 1000 mature individuals) b) all individuals are in a single sub-population. D. Population very small or 1. Population estimated to number less than 1000 restricted in the form of either mature individuals of the following: 2. Populations is characterised by an acute restriction in its area of occupancy (typically less than 100km2) or in the number of locations (typically less than 5). Such a taxon would thus be prone to the effects of human activities (or stochastic events whose impact is increased by human activities) within a very short period of time in an unforeseeable future, and is thus capable of becoming critically endangered or even extinct in a very short period. E. Quantitative analysis 1. Population estimated to number less than 1000 showing the probability of mature individuals extinction in the wild is at least 10% within 100 years. 2. Populations is characterised by an acute restriction in its area of occupancy (typically less than 100km2) or in the number of locations (typically less than 5). Such a taxon would thus be prone to the effects of human activities (or stochastic events whose impact is increased by human activities) within a very short period of time in an unforeseeable future, and is thus capable of becoming critically endangered or even extinct in a very short period.

75 South Pacific Applied Geoscience Commission 08/02/99

Briguglio’s work, for instance, is that many small island developing countries register relatively high gross national product per capita; yet in reality, their economies are very susceptible to any external economic fluctuations and environmental shocks, no matter how minimal.

For us, the Barbados Plan of Action for the Sustainable Development of Small Island Developing States provides the basis and raison d’etre for the Vulnerability Index. We need full and proper understanding of all the components and technical nature of vulnerability, and so that we can plan and seek from the international community vital support for our efforts at sustainable development.for our part AOSIS has endeavoured for some time to spark some momentum into the work on the Index, though we were pleased that the fourth and fifth Sessions for the Commission on Sustainable Development specifically recognised the need to accord sufficient priority to it.

2 South Pacific Applied Geoscience Commission 08/02/99

The Forum Economic Ministers Meeting (8th July 1998) agreed to adopt a common Forum position with the objective of the UN adopting a vulnerability index, and with the aim of having such an index included among the criteria for determining LDC status, and for deciding eligibility for concessional aid and trade treatment.

Study Description – Strategy

Review and build on current work already completed or underway addressing environmental vulnerability of SIDS

Work with SPOCC regional agencies and others (internationally, regionally and nationally) to develop a report for New Zealand to table at CSD-7.

Carry out this study in conjunction with the GEO-2 exercise currently underway within the SPOCC agencies and being co-ordinated by SPREP.

Identify parameters which may be used in the construction of an environmentally sensitive vulnerability index for Pacific SIDS.

Source data, and establish a database at SOPAC.

Identify gaps in these data and develop concept papers for projects which focus on a common objective to ensure that these gaps are filled.

Convene a session on environmental vulnerability indices at the upcoming SPREP-UNEP Oceans meeting to enable widespread consultation with experts and country representatives.

Define future actions to: (i) further the development of an acceptable environmental vulnerability index for Pacific SIDS and (ii) further efforts towards the development of a composite vulnerability index as described in the Barbados Program of Action and meet the needs enunciated by AOSIS.

Compile a report for widespread circulation and consideration prior to the Donor Round Table leading up to CSD-7 and the UN General Assembly special session on SIDS in September 1999.

Reporting

Draft report to be submitted by SOPAC to New Zealand by 31st December 1998.

3 20/03/00

Environmental Vulnerability Index (EVI) Project Phase III SOPAC Funding Proposal

Keywords: Pacific Regional initiative, Environmental Vulnerability, National level, Pacific Island Countries, Global issues, LDC status.

1. Background

Currently, it is our understanding that ECOSOC has deferred decision on the CDP recommendations. SOPAC is currently seeking clarification on this and implications for Pacific SIDS. It is also our understanding that both the EU and the United States have expressed concerns about the issue of vulnerability and the CDP report. The EU position recognises vulnerability as an issue but clearly states that the EU has concerns over proposed mechanisms for its use and measurement.

2. Project Aim

2 20/03/00

3. Introduction

It has become increasingly recognised that studies on, and successful characterisations of the vulnerability to harm of human and natural systems at the level of entire states would provide valuable information for international processes and management actions. However, to undertake such studies on an ad hoc basis would be prohibitively expensive and/or lead to results which, for international processes, might be difficult to compare. The solution to this problem has been the attempt, over the past decade, to construct vulnerability indices that could be calculated for states on a common basis. The aim of vulnerability indices, then, is to describe and summarise the relative vulnerability of states. A range of these indices has been developed to describe vulnerability of human systems to harm from other human influences or natural ones.

3 20/03/00

The Expert Group supported the conceptual approach to determining environmental vulnerability as developed by SOPAC and supported its development into a global tool. They also made several recommendations including that: I. The EVI be considered by the United Nations Committee on Development Policy (UN CDP) at its next meeting that considers vulnerability indices as criteria for classifying countries and assigning Least Developed Country (LDC) status; II. The environmental vulnerability index is considered a useful tool not just as a single number defining environmental vulnerability of a country but also in providing detailed country profiles on the hazards, intrinsic features and environmental health likely to affect environmental vulnerability; III. There is a need to look further into the development of a separate vulnerability index which examines the human environment (more content on direct impacts on human systems – vulnerability of human quality of life index); IV. The EVI should be tested using data from no less than 15 widely differing countries; and V. Should be published in an international peer-reviewed journal

4 20/03/00 centred in the Pacific Region (Australia, Fiji, Samoa, Tuvalu and Vanuatu). This third report is expected to be available in March 2000 (Table 1).

Table 1: Summary of phases of work undertaken since September 1998 and plans for future work

Phase Description Accompanying report Phase I 1. Review other vulnerability indices SOPAC Technical Report 275 2. Determine whether an EVI is possible (January 99) 3. Construct preliminary EVI centred on Pacific SIDS 4. Initial data collection & calculation for 3 countries (Australia, Fiji, Tuvalu) Phase II 1. Peer review at Expert Meeting “Think SOPAC Technical Report 299 (October Tank 1” 1999) 2. Refinement of model Technical Report due March 2000 3. Data from 5 countries in Pacific Phase III 1. Collect data from 15 differing types of Work to begin in March 2000 countries worldwide (Tasks can be completed within 12 months 2. Globalise model should funding be secured by March 2000) 3. Test indicators 4. Test overall EVI 5. Think tank 2 6. Create user interface (computersied) 7. Capacity-building in-country 8. Set up permanent data-collection mechanisms 9. Publication

5 20/03/00

The basic modules include: Module 1: Refinement and Comprehensive Testing of the Environmental Vulnerability Index (Collecting data from 15 countries and profiles from 5, Globalising the model, testing indicators, testing the overall index, Think Tank 2, publication) Module 2: Pacific EVI Country Capacity-Building Module 3: Sustainable Data Collection Process for the Environmental Vulnerability Index (through in-country and international agencies) Module 4: Computer Environmental Vulnerability Index Interface Development Module 5: EVI Validation Exercise The modules are currently presented in concept with estimated budgets and timeframes. Detailed proposals for each module can be provided as required.

All tasks will be carried out in consultation and partnership with all Committee for Regional Organisation of the Pacific (CROP) members including: South Pacific Regional Environment Programme, Forum Secretariat, Secretariat of the Pacific Community, University of the South Pacific, Forum Fisheries Agency, Pacific Island Development Program, Tourism Council of the South Pacific, non-government organisations and other regional agencies. Particular inputs will be sought from all Pacific island governments and the international agencies with a mandate for international affairs and particularly SIDS.

6 20/03/00

4. Modules Required for EVI Development and their Justification

Module 1: Refinement and Comprehensive Testing of the EVI

Aim: To operationalise the EVI through extensive testing with real environmental vulnerability data from at least 15 countries that represent the vast variety of different environments found globally

Tasks: Collecting data from 15 countries, Profiles from 5 countries for validation, Globalising the model, Testing indicators, Testing the overall index, Think Tank 2, Publication.

Outputs/Benefits:

7 20/03/00

Budget/Timeframe:

· US$242,660 12 months (Detailed budget attached)

Module 2: Pacific Environmental Vulnerability Index Country Capacity-Building

Aim: To provide training to SIDS country representatives in the use, data collection and computation of the environmental vulnerability index.

For some countries, data may be obtainable by contacting the relevant departments and asking for assistance with data acquisition. For most SIDS, this approach is unlikely to yield adequate results. It will therefore be necessary for the EVI Team to carry out a capacity building workshop with representatives from at least 10 member countries to introduce them to the environmental vulnerability index concept, its workings and information required for its computation. Also included in this capacity-building exercise is follow-up work for the 10 countries.

Outputs/Benefits:

8 20/03/00

· Pacific ownership of the EVI and the work that has been carried out in the region

Budget/Timeframe:

· US$156,987 9 months

Module 3: Development of a computerised EVI Interface

Aim: To develop a user interface in Microsoft Access for the EVI

The model developed for calculating the EVI during Phase I was built into EXCEL spreadsheets. This makes calculating the EVI cumbersome for the endpoint users and was not intended for use by them because files may be easily damaged or altered, making calculations inaccurate.

Output/Benefits:

Budget/Timeframe:

· US$16,714 2 months

Module 4: Sustainable Data Collection Process for EVI

Aim: To develop a sustainable mechanism for EVI data collection and computation every 5 years

9 20/03/00

Potential mechanisms for data collection by country institutions and reporting mechanisms to international organisations responsible for international data collation and analysis will need to be investigated for applicability possible use. For example, one possible mechanism could be the collection of national environmental data through the Global Environment Outlook (GEO) process. This mechanism may be able to provide a suitable platform through which data is collected nationally and then collated internationally. Countries may then also utilise the information at the national level for their own purposes.

Outputs/Benefits:

Budget/Timeframe:

· US$64,446 3 months

Module 5: EVI Validation Exercise

Aim: To independently peer review and validate the EVI

Outputs:

· A globally applicable scientifically sound, robust, fully tested and validated EVI

10 20/03/00

Budget/Timeframe:

· US$123,500 6 months

5. Overall Expected benefits from Project

All of the above modules are required to fully develop an internationally recognised and acceptable EVI. By ensuring that the model has been fully tested and that consultation has occurred at all levels (technical and political) the EVI can become an important tool for examining relative vulnerabilities of states, whether they be small islands or continental. Consultation and involvement of Pacific Island Countries throughout the project will ensure that the initiative remains with the region and that the needs and expectations of those countries are incorporated into the development of the EVI. It will also mean that the capacity of Pacific Island Countries for environmental management using tools like the environmental vulnerability index are secured.

6. Overall Project Outputs

11 20/03/00

7. Overall Project Budget

· Total budget for Phase III US $604,307

Project Budget Summary in US Dollars

Estimated Total Budget USD Amounts International Data – Refinement/Testing $242,660 Pacific Capacity-Building $156,987 EVI Computer Interface Development $16,714 Sustainable Data Collection Process $64,446 EVI Validation Exercise $123,500 TOTAL BUDGET $604,307

12 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

CONTENTS

1. INTRODUCTION 3

1.1 Date, Venue and Participants 3

1.2 Purpose of the Meeting 3

1.3 Structure 3

1.4 Agenda 3

2. OPENING SPEECHES 5

2.1 Opening Statement by Chief Executive of the F.I.S. 5

2.2 Address by Elizabeth Khaka, UNEP 5

2.3 Address by Prof. Lino Briguglio, Director, Islands and Small States Institute 6

3. THE EVI: BACKGROUND INFORMATION AND DISCUSSION 7

3.1 Presentation by Dr Ursula Kaly 7

3.2 Presentation by Mr Craig Pratt on Data Collection 8

3.3 Discussion on Dr Kaly’s and Mr Pratt’s Presentations 9

4. DISCUSSION OF BASIC CONCEPTS 10

4.1 Adoption of Definitions 10

4.2 The Impact on Human Systems 10

4.3 New Name for the Index 10

5. ENVIRONMENTAL CHARACTERISTICS OF SIDS IN THE CARIBBEAN AND IMA REGIONS 11

5.1 Presentation of Country Papers 11

5.2 Malta 11

5.3 Mauritius 13

5.4 Jamaica 16

5.5 Saint Lucia 17

5.6 Trinidad and Tobago 19

6. DISCUSSION ON THE INDICATORS 21

6.1 General Suggestions 21

1 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

6.2 Comments on Individual Indicators 22

6.3 Suggested Possible New Indicators 27

6.4 Possible Indicators for the Subsidiary List 28

6.5 Possible Weighting Schemes 28

7. STRENGTHS, WEAKNESSES, OPPORTUNITIES & THREATS 29

7.1 Strengths 29

7.2 Weaknesses 30

7.3 Opportunities 30

7.4 Threats 31

8. FRAMEWORK FOR PROCUREMENT OF DATA 32

8.1 The Role of Governments 32

8.2 The Role of International Organisations 32

8.3 The Role of Regional Organisations 32

8.4 The Role of Universities and Research Institutions 32

8.5 The Role of Private Organisations 33

9. POSSIBLE AVENUES FOR FUNDING 34

9.1 Funding at the International level 34

9.2 At the Regional Level 34

9.3 At the National Level 35

10. ADOPTION OF FINAL STATEMENT 36

APPENDICES Appendix 1: Agenda and Programme Appendix 2: List of Participants Appendix 3: List of Indicators Appendix 4: Background Paper prepared for the Workshop Appendix 5: Data collection for the EVI

2 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

1. INTRODUCTION

1.1 Date, Venue and Participants

The meeting on the Environmental Vulnerability Index (EVI) was held between 29th November and 3rd December 1999, at the Foundation for International of the University of Malta, Valletta, Malta. The meeting was attended by experts from small island developing states (SIDS) of the Pacific, Caribbean, Indian Ocean and Mediterranean regions. A participants’ list is presented as Appendix 2.

1.2 Purpose of the Meeting

The purpose of the meeting was to: 1. Introduce the SOPAC model of the Environmental Vulnerability Index, and gain support for the Index, in the Caribbean and the IMA (Indian Ocean, Mediterranean and Atlantic) regions; 2. Examine the special environmental features of a sample of Small Island Developing States in the Caribbean and IMA regions, with a view to developing EVI indicators relating to these states; 3. Consult with experts from the Caribbean and IMA regions to identify the benefits they expect from the EVI; 4. Put forward recommendations as to how the EVI can be developed further and how data collection mechanisms might be set up within the Caribbean and IMA countries; and 5. Identify the process and devise an approach for obtaining the funding needed to globalise the EVI.

1.3 Structure

The meeting consisted of (1) presentations prepared by representatives of SOPAC and by the respective country representatives, (2) discussion on these presentations, and (3) further discussion on the future development of the EVI, culminating in a final statement. The sessions were chaired by Professor Lino Briguglio and Mr Craig Pratt.

1.4 Agenda

The agenda of the meeting consisted of the following items: · Opening statements; · Presentation and discussion on the EVI as developed todate; · Discussion of concepts of Environment, Vulnerability/Resilience, Damage and Degradation. · Presentation and discussion on the environmental characteristics of SIDS in the Caribbean and IMA regions; · Consideration of the Individual Environmental Indicators; · SWOT analysis relating to the EVI; · Discussion on collection of data;

3 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

· Discussion on potential sources of funding for further development of the EVI · Final Statement.

A detailed programme is presented as Appendix 1.

4 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

2. OPENING SPEECHES

2.1 Opening Statement by Chief Executive of the F.I.S.

Mr Leslie Agius, Chief Executive of the Foundation for International Studies, inaugurated the workshop. He said that he was pleased that the Foundation for International Studies was hosting such an important meeting, intended to facilitate the further development of the Environment Vulnerability Index and to gain support for the Index in the Caribbean and IMA regions. He welcomed the participants, and wished a pleasant stay to the non-Maltese participants many of whom travelled from distant areas in the Pacific, Caribbean and Indian Ocean.

Mr Agius said that the workshop could be instrumental in extending the EVI to the Caribbean and IMA regions and to devise an approach for obtaining the funding needed to globalise the scope of the EVI.

He said that Islands and Small States Institute, which was convening the meeting, had established itself as an international centre for research and information dissemination on small island states, working very closely with international organisations in this regard.

He thanked UNEP for funding the workshop and SOPAC for agreeing to collaborate with the Islands and Small States Institute in the organisation of the workshop. He also thanked the various other institutions that provided or facilitated the participation of experts for the workshop.

2.2 Address by Elizabeth Khaka, UNEP

Ms Khaka, speaking on behalf of UNEP, said that the workshop was one of the tasks undertaken by UNEP to carry out its mandate to facilitate implementation of the Barbados Programme of Action for the Sustainable Development of SIDS, in matters relating to the environment. She said that in November 1998, UNEP collaborated with the Islands and Small States Institute and the Government of Malta, by convening a Ministerial Meeting for SIDS in the Indian, Mediterranean and Atlantic Ocean where a new SIDS grouping, named IMA- SIDS, was formed. The meeting provided a forum for these SIDS, enabling them to provide an input to the process of reviewing the Barbados Programme of Action as a group. An important document that was produced during the November 1998 meeting was the “Valletta Declaration on the Sustainable Development of IMA-SIDS”.

Ms Khaka said that the Barbados Programme of Action recognised the need for the development the Vulnerability Index, and this matter was also given a high profile in the Special Session of the UN General Assembly held in September 1999.

Ms Khaka thanked the Islands and Small States Institute and SOPAC for collaborating in this organisation of the workshop on the EVI.

5 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

2.3 Address by Prof. Lino Briguglio, Director, Islands and Small States Institute

Professor Briguglio, director of the Islands and Small States Institute, said that he was very pleased to be associated directly with the development of the EVI. He said that the Islands and Small States Institute was prepared to offer its services to extend the scope of the EVI to other small island states in the Indian Ocean, the Mediterranean and Atlantic (I.M.A.) and the Caribbean regions.

Prof. Briguglio explained how the idea of constructing an Economic Vulnerability Index to show that SIDS were more vulnerable than other states, was originally conceived by himself, within the University of Malta and gave a brief account of the developments in the conceptual framework and the statistical methods adopted for the construction of the vulnerability indices between 1992 and 1999.

Prof. Briguglio said that the Maltese Minister for the Environment, the Hon. Francis Zammit Dimech could not attend the opening session, as scheduled, but the Minister had sent a message to let the participants know that the Government of Malta was very much interested in the development of the Environmental Vulnerability Index.

Prof. Briguglio thanked UNEP for supporting this workshop and allocating funds for the development of the Environmental Vulnerability Index, and SOPAC for assisting in the organisation of this workshop. He also thanked the experts from Fiji, Tuvalu, Malta, Mauritius, Trinidad and Tobago, St Lucia and Jamaica for accepting to participate in this workshop

6 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

3. THE EVI: BACKGROUND INFORMATION AND DISCUSSION

3.1 Presentation by Dr Ursula Kaly

Dr Ursula Kaly’s presentation dealt with the rationale, development and structure of the EVI. The following is a summary of her presentation. The full text is presented in Appendix 4.

Dr Kaly said that the index was initially developed by SOPAC1 to provide an index for ranking countries in terms of environmental vulnerability, with a focus on the Pacific region. The work was carried out in response to a call made in the Programme of Action for the Sustainable Development of SIDS, adopted during the Barbados Global Conference in April- May 1994.2 and an increasing awareness that small island developing states face disadvantages to their sustainable development associated with their remoteness, small size, dispersion, and economic constraints including limited natural resources.

She said that SOPAC started the programme on the development of the Environment Vulnerability Index (EVI) in August 1998, focusing on the Pacific SIDS. A total of 57 indicators of environmental vulnerability were selected for inclusion in the index. This included 39 indicators of risk (REI), 5 indicators of intrinsic resilience (IRI) and 13 indicators of environmental integrity or degradation (EDI), to represent extrinsic resilience. Many of the indicators were expressed as a ratio of area of land or coast rather than simply absolute numbers because it is risk density or proportion of area degraded that is of interest from an environmental perspective.

These indicators were evaluated for each country and scored on a 7 point scale, thus transforming observations to permit summation of the different components of the index. She explained that this allows for heterogeneous data (including qualitative data) to be represented along a common scale. Non-linearities were built-in by mapping the data on the assumption of a particular distribution along the seven point scale, which could be unique for a particular indicator.

Dr Kaly said that on the basis of the first attempt at constructing the EVI, it was concluded that it was possible to produce single-figure measures of environmental vulnerability for three countries, namely Australia, Fiji and Tuvalu. Following this first attempt at quantifying the index, SOPAC organised a Think Tank in September 1999 with the aim of (1) obtaining peer- review and commentary from experts in a range of fields relevant to the development of the EVI, (2) to render the EVI acceptable and/or operational in the international community and (3) to identify directions for future work. Following the Think Tank meeting, the number of

1 See Kaly, U., Briguglio, L., McLeod, H., Schmall, S., Pratt, C. and Pal, R. (1999). Environmental Vulnerability Index (EVI) to summarise National Environmental Vulnerability Profiles. SOPAC Technical Report and Kaly, U., Briguglio, L., McLeod, H., Schmall, S., Pratt, C. and Pal, R. 1999. Proceedings of the Environmental Vulnerability Index (EVI) Think Tank 7-10 September 1999. SOPAC Technical Report. 2 Article 113 of the Programme of Action states: “Small island developing States, in cooperation with national, regional and international organisations and research centres, should continue work on the development of vulnerability indices and other indicators that reflect the status of small island developing States and integrate ecological fragility and economic vulnerability. Consideration should be given to how such an index, as well as relevant studies undertaken on small island developing States by other international institutions, might be used in addition to other statistical measures as quantitative indicators of fragility”. The Special Session of the UN General Assembly, held in September 1999, adopted a document, which contained three paragraphs (Section F, par. 39 to 41) dealing with the Vulnerability, and again stressing the need to compute a vulnerability index.

7 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX indicators was reduced to 47, and some modifications to the original set were introduced.

Dr. Kaly emphasised that the EVI, as proposed by SOPAC, deals with the vulnerability of the environment, on not of humans per se. But the underlying argument in SOPAC’s EVI is that the "natural environment" is the basis of all human activities. She argued that poor environmental conditions mean fewer or poorer quality natural resources and poorer ecosystem services (such as attenuation of wastes and pollution).

She said that the Index has three components namely: · the REI or Risk Exposure sub-Index; · the IRI Intrinsic Resilience sub-Index; and · the EDI, Environmental Degradation sub-Index.

Dr Kaly explained that the EVI requires further refinement. The progress so far relates to Phase I and a part of Phase II of the project, designed to determine whether it was possible to produce an EVI, and if so, expose it to peer review at the technical level. The next Phase is designed to start the process of globalising the index, enlist the support of SIDS from the IMA and Caribbean region, and find suitable funding for this purpose.

3.2 Presentation by Mr Craig Pratt on Data Collection

The following is a summary of Mr Pratt’s presentation on data collection for the EVI. The full text of the presentation is carried in Appendix 5.

Mr Pratt said that availability of appropriate environmental vulnerability data is fundamental to both the development of the EVI and ultimately the final calculation of a country’s EVI value. He said that the EVI, by its very essence, attempts to summarise a wide variety of environmental vulnerability data for a country. Much of these environmental data were only collected and compiled at the national level. Reporting and publication of these data internationally was not, until then, well-established as the reporting of country economic data.

Mr Pratt said that the criteria set for data collection, included that (a) data should preferably be already available or easily obtainable and that (b) data should measure change or be a proxy for change causing significant harm to the environment.

Mr Pratt said that the data that needed to be collected for the EVI includes meteorological data, fisheries data, land area, and natural hazards data. The very diverse and wide-ranging nature of these data means that their sources are widely dispersed and require considerable effort by a country to identify, collect and compile. Some of the indicators require information that could only be provided by the authorities or by experts in the respective country. Mr Pratt argued that it was therefore essential to have full government co-operation in the data gathering process to ensure success. This had been the case in the Pacific.

8 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

Mr Pratt discussed the problem of the lack of capacity - a common one throughout SIDS - referring to limited resources and few trained personnel, which often render data collection very difficult. Another issue that had arisen in the Pacific was that it had been difficult to create an adequate understanding in personnel of the specific data and information requirements needed for a response to indicator questions. This was due in part to a lack of understanding of the purpose of the EVI, and its mechanics and, more generally, the inadequate training of personnel in the identification, collection, analysis and manipulation of data.

Mr Pratt informed the participants that SOPAC was attempting to develop an alternative questionnaire approach to facilitate country environmental data gathering. He said that a help handbook would prove useful as a possible way to provide a detailed background to the EVI, its mechanics and specific instructions and assistance on how to identify and gather the required information. This is expected to reduce the need for continued assistance and support in the data gathering process.

3.3 Discussion on Dr Kaly’s and Mr Pratt’s Presentations

With regard to Dr Kaly’s presentation, participants discussed the problem of comparing heterogeneous countries using a single yardstick. It was argued that the EVI might be more meaningful for small countries, where climatic and environmental conditions did not vary much within territorial boundaries, as compared to large countries, where environmental conditions varied drastically between one region and another. It was noted that this problem occurs also in the case of economic comparisons between countries, e.g. when comparing GNP per capita. In spite of this concern, the GNP per capita index is used widely in comparisons of economic development among countries.

Another item of discussion related to the applicability of the index to areas within national boundaries. Dr Kaly pointed out that the EVI was intended for entire states, but has the flexibility to be used within countries.

Comments and queries were made regarding the general criteria for the selection of indicators for the EVI after Mr. Pratt’s presentation. It was argued that ease of collection should not be set as a criterion, since if an event harms the environment of a given country, its impact should be considered even if data for it is difficult to obtain. Otherwise, the index would only reflect part of the environmental damage. This comment provoked considerable discussion. It was noted amongst other things, that an “all or nothing” approach for data availability would defeat the whole exercise, and that whatever index is chosen, the problem of data limitations will occur. It was argued that sometimes proxy variables could fill the gaps, when data for a particular event is difficult or impossible to obtain.

9 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

4. DISCUSSION OF BASIC CONCEPTS

Participants discussed the concepts of “environment”, “vulnerability”, “resilience”, “damage” and “degradation” for the purposes of the EVI.

The discussion was introduced by Dr Ursula Kaly who stressed that the development of the index depended on clear and precise definitions of the underlying notions, in particular the concepts of “environment”, “vulnerability”, “resilience”, “damage” and “degradation”.

4.1 Adoption of Definitions

Following discussion, the group adopted the following definitions:

The natural environment was defined as those biophysical systems that are capable of being autonomously sustained without human inputs.

Vulnerability was defined as proneness of the natural environment to damage and degradation.

Resilience was defined as the potential for a system to minimise or absorb the effects of damage.

Damage was defined as the reduction of diversity, extent, quality and function of natural environments, which may be reversible.

Degradation was defined as irreversible damage.

4.2 The Impact on Human Systems

The discussion also dealt with the question of whether the definition of environment excluded human systems (e.g. agriculture).

Dr Kaly explained that the concept of vulnerability, as applied to the EVI, excluded human systems, and considered the natural environment. However the underlying assumption was that the health of the natural environment had ramifications for human welfare.

Following discussion, participants agreed with Dr Kaly’s explanation.

4.3 New Name for the Index

Participants considered the problems that might arise because of the different definitions of the term “environment”. It was suggested that to avoid confusion, the name for the Environmental Vulnerability Index (EVI) be changed to the Vulnerability Index for the Natural Environment (VINE). It was agreed that this suggestion was to be transmitted to SOPAC for further consideration.

10 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

5. ENVIRONMENTAL CHARACTERISTICS OF SIDS IN THE CARIBBEAN AND IMA REGIONS

5.1 Presentation of Country Papers

The representatives of Malta and Mauritius (IMA-SIDS) and of Trinidad and Tobago, St Lucia and Jamaica (Caribbean-SIDS) made presentations relating to their respective small island states.

5.2 Malta

Dr. Adriana Vella said that the more important environmental characteristics of the Maltese Islands, relate to (1) water and land shortages (2) coastal zone management and soil erosion (3) pollution (4) waste management (5) limitations, over-exploitation and degradation of natural resources and (6) loss of wildlife and genetic erosion. The following is a brief overview of these characteristics.

5.2.1 Water shortages

Natural water resources in Malta are totally dependent on rainwater, which percolates through the rocks and forms aquifers. Because of the low rate of precipitation, exacerbated by high rainwater runoff rates, natural freshwater is a scarce commodity in Malta. Such scarcity has traditionally had direct impact on human health and general well being, on agricultural production, and in recent years on tourism and certain manufacturing industries. At present about 60% of water demand in Malta is supplied by desalination plants (reverse osmosis) - a very costly method of water production. Water tariffs are heavily subsidised in Malta and do not therefore reflect the cost of production, and this may be conducive towards lack of incentives for saving water.

5.2.2 Coastal zone management

The coastal area of the Maltese islands has played a very important role in the social and economic development of the country, with a large proportion of economic activity and of the resident population occurring on the coast. The natural harbours of the islands are extensively used for commerce and for tourism related activities. Urban settlements, new industrial and tourist infrastructures and other buildings are mushrooming along the coast, leading, as expected to increase waste generation and sewage pollution.

The coastal area is also important from an ecological point of view, since it contains habitats, including sand dunes and saline marshlands, which in turn support a number of endemic plant and animal species.

The coastal area has, up to recently, been considered to consist only of the restricted strip of land or rock between the sea and the human habitations. The importance of seriously considering the coastal sea areas and the vulnerability of this marine strip to the increasing local exploitation and development has been brought forward through scientific research on specific natural resources, such as local endangered marine species.

11 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

5.2.3 Soil and coastal erosion

Soil erosion in Malta is on the increase mainly as a result of abandoned agricultural land and limited tree cover. The building sprawl has substantially decreased the number of catchment areas, thus resulting in greater rainwater run off. Local climatic conditions and the increased over exposure of areas with soil have definitely contributed to its loss. The lack of trees and shrubs together with the loss of rubble walls further exacerbate this problem.

5.2.4 Waste Management

In small islands, waste management tends to be more problematic than in larger territories, due to the limited land area. In Malta the very high population density and rapid economic growth accentuate this problem. The added load of the tourist population during different seasons of the year but in particular during the summer months is not to be underestimated.

5.2.5 Pollution

The most important considerations in this regard relate to air and marine pollution. In Malta certain environmental damage and health risks are directly associated with elevated concentrations of air pollutants. Examples of typical pollutants are sulphur dioxide, nitrogen oxides, particulate matter and dusts, ozone, carbon monoxide, benzene, polyaromatic hydrocarbons, and heavy metals (e.g. lead, mercury and cadmium). A recent scientific study has actually found strong associations between particulate air concentrations and an increasing local health problem. Such study is also starting to shed light on the dominant relationships between different weather conditions and increased presence and effects of the pollution.

As for marine pollution, the most important concern relates to solid and liquid waste disposal. A positive development in this regard is that Malta is a party to the Barcelona Convention on the Protection of the Mediterranean Sea against Pollution and its Protocols and collaborates actively in MEDPOL and there is legislation in place to control waste disposal into the sea.

5.2.6 Other environmental concerns

Loss of wildlife and biodiversity. In Malta the institutional and legal set-ups for nature conservation and protection of endemic and indigenous species are rather underdeveloped, and not adequately enforced. In addition, agricultural practices in Malta have led to a drastic reduction of forest-cover. There is also the negative impact caused by grazing goats, sheep, cattle and wild rabbits, which has led to further reduction of forest cover. Uncontrolled exploitation of natural resources is still a norm and though local expertise to undertake proper wildlife conservation assessment and monitoring exist locally the government’s will and assistance to undertake conservation monitoring and management is still very poor. Specific conservation research projects have been undertaken with the goal to highlight this

12 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX requirement. The awareness to consider the need for sustainable development and the conservation of natural and genetic resources in a scientific and professional manner is only recently developing. The need to preserve the variety of local genetic resources is increasingly becoming a requirement at both national and international level. Thus assessing the genetic variation of natural resources needs to become synonymous with resource assessment for long-term management and conservation. A local high human population density in itself creates a considerable stress on natural habitats and their biota. Local wildlife has also been exploited since time immemorial, and in most cases the rate at which individual organisms are removed from their population is well above the rate at which they are replaced, leading to loss of populations, species and biodiversity.

Lack of institutional and policy co-ordination. Over the past decade, Malta has experienced a number of improvements in environmental legislation and management. One major shortcoming at present is that environmental legislation is somewhat fragmented with different lines of command leading to inefficient enforcement and sometimes contradictory signals. There is therefore a dire need for a national strategy for co-ordinating legislation and policies.

5.2.7 Malta and the EVI

Dr. Adriana Vella said that Malta was likely to record very high risk exposure and degradation scores, and low resilience scores on the EVI with regard to indicators related to islandness (indicator number 11)3, land area (indicator 10), water shortages and droughts (indicators no. 3 and 41), natural resource monitoring and protection (indicators 16 to 24) (indicators 44, 45 & 47), soil degradation (indicator no. 40), coastal zone (indictor no. 25); tourism (indicator no. 29) removal of natural vegetation (indicators no. 22 & 28); high population density (indicator no. 26); waste generation (indicators nos. 30 and 32) and other indicators associated with economic development (indicators no. 35, 23, 18).

Dr. Vella argued that the EVI, as currently constructed, requires some refinements to the indicators to take into account vulnerability associated with water shortage and water salinity in states like Malta which experience this reality.

Also important is the emphasis on making sure that the EVI would be a reflection of the dynamic change of the environment apart from a descriptor of the static status. This would give the EVI a greater practical value over the State of the Environment Report.

5.3 Mauritius

Dr Deolall Daby said that the terrestrial, aquatic and marine environments of Mauritius are being degraded due to the socio-economic driving forces and unsustainable resource use. These problems are exacerbated by climate change and its associated impacts as well as natural disasters. The main areas of concern are (1) deforestation (2) loss of biodiversity and (3) climate change and extreme weather events. The following is a summary of Dr Daby’s presentation.

13 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

5.3.1 Deforestation and the land use

Although 31% of the total land area is under forest, woodland and scrub, only 1% of native forest remains due to three centuries of deforestation and exploitation. Approximately 45% of the total land area (90% of arable land) is under sugar cane. The forests, nature reserves and coastal Pas Geometriques (public land constituting 10% of total) is under increasing pressure from competing uses (hotel, recreation and conservation). Encroachment on environmentally sensitive areas (coastal ecosystems, forested areas, hills and mountain slopes, nature reserves and catchment areas) is also increasing.

5.3.2 Bio-diversity

Mauritius is ranked second in the world for having the highest percentage of its native plants threatened globally (39%) and nationally (71%). Natural vegetation has been largely destroyed due to clearing land for agriculture, settlements, road infrastructure, farming and pastures. Many endemic species have already become extinct. Modification of aquatic habitats (draining, pollution, competition for water), introduced species and commercial exploitation are the principal causes of decline of the freshwater fish species.

5.3.3 Climate change and extreme weather events

Mauritius is at high risk from the effects of global climate change and its associated impacts. These include sea level rise and enhanced frequency of extreme weather events such as temperature and precipitation extremes and natural disasters (10-12 annual cyclones, floods, droughts, storm surges and landslides). The changes due to these calamities represent additional stresses on the environmental systems that are already under intense and growing pressure.

5.3.4 Other areas of environmental concern

Urbanization. The rate of growth of the urban population has been 1.1% over the period 1975-1995 and this is projected to increase to 1.9% by 2015. Currently just over 40% of the population is urban and by 2025 this will increase to 60%. The environmental problems associated with urbanisation include habitat and biodiversity loss, unplanned and haphazard development, pollution of surface, underground and coastal waters, and social problems.

Fresh water. Mauritius is classified as a water poor nation by the UNDP, and further demand and consumption by the growing population with a rising standard of living may hamper economic development. Mauritius is coming under increasing water stress, defined as an annual water supply of between 1100-1700 m3 in 1995 to 1485 m3 by 2025. During the second half of 1998 and most of 1999, Mauritius faced the most severe water stress situation in the last 20 years, impacting heavily on the economy and environment. About 75% of water withdrawals are used for agriculture, 16% for domestic use and the remaining for industrial and commercial uses. These sectors of consumption are also the main polluters of both the terrestrial and coastal marine water bodies. Further, salt water intrusions into bore holes are evident.

14 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

lagoons causing hypernutrification and eutrophication problems. Enhanced soil erosion occurs in cleared areas not under sugar cane and causes siltation problems in lagoons after heavy rainfall.

Tourism. Tourism is a major earner of income for the country. It is mainly marine-based and tourist arrivals continue to increase (600,000 annually). Mauritius has invested heavily in coastal infrastructure development to accommodate the escalating tourist populations. However, such development has occurred without appropriate planning and preventive measures, and damage to coastal ecosystems (erosion, physical damage to habitats, pollution) is already evident. Expansion in the sector is often in the form of encroachment on non- tourist space.

Pollution. Oil pollution is a major source of concern in Mauritius. Oil spills can cause catastrophic effects on tourism, fisheries, coastal ecosystems and recreational activities. The region is the main transportation route of more than 475 million tonnes annually of oil from the Middle East to Europe and America. This constitutes a constant threat and at present no country in the region has the capability to deal with oil spill disasters.

Air pollution, though still low by international standards, is increasing as energy use rises. Gas emissions are particularly severe in urban centres, and dust and air pollution problems occur in areas surrounding sugar mills, from burning cane fields, emissions from boilers and release of flyash. Other sources include stone crushers, brick making plants, lime kilns, chemical processing factories, hot asphalt plants, odours from agricultural wastes, waste dump sites, coal burning for electricity and steam raising in textile industries.

Coastal zone. The intensive urbanisation of the coastal zone is a major cause for concern because of unplanned construction, land reclamation, and increasing demographic pressure. Coastal habitat degradation occurs as a result of over-fishing above the MSY level, swimming, recreation and tourism activities, siltation of lagoons, mining, quarrying, dredging, nitrogen loading and recurrent HABs. Microbial contamination of coastal waters occurs from raw sewage disposal and increasing hot spots of marine pollution result from industrial effluents, high BOD load from sugar processing and run-off of agrochemicals. Inshore fishing is detrimental in various ways such as use of undersized nets, coral breakage by poling and anchoring of boats, overfishing of aquarium fish from coral reef areas, and poor enforcement of fisheries legislation.

5.3.5 Mauritius and the EVI

Dr Daby said that Mauritius was likely to register relatively high scores in terms of risks, degradation and lack of resilience, with regard to the following aspects: sea surface temperature (indicator no. 1) cyclones (indicator no. 2), droughts (indicator no. 3), degradation of coral reefs (indicator nos. 24, 25 ) loss of critical coastal and terrestrial habitats and biodiversity (indicator nos. 20, 22), land area, soil degradation (indicator no. 40), decreasing fish catches (indicator no.28), coastal erosion and marine pollution (indicator no. 33) ecotoxicology (indicator 31), waste management (indicators nos. 32) and fresh water shortage and contamination (indicator no. 41).

Dr Daby argued that the EVI, as currently constructed, required some additional indicators or modification to existing ones, to take into account vulnerability associated with fresh water shortage and contamination in states, which, like Mauritius experience this reality.

15 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

5.4 Jamaica

Mr. Learie Miller said that the major environmental concerns in Jamaica are related to agriculture, mining and quarrying, tourism and fisheries.

5.4.1 Agriculture and forestry

The main problems associated with agriculture in Jamaica are excessive land clearing and soil erosion. There is an adverse effect on water quality and quantity of agricultural production. In addition, agriculture gives rise to a number of hazards associated with the use of chemicals, uncontrolled use of fire, loss of biodiversity and wildlife habitat, excessive siltation and risk of downstream flooding. In the case of forestry, the major concerns relate to land clearing for cultivation and fuelwood/charcoal production.

5.4.2 Mining and quarrying

Bauxite mining has resulted in a disposal problem specifically of red mud from the refining process. Air pollution from wind driven dust is common for both mining and quarrying activities. In addition, there is loss of aesthetic value of hillsides due to scarification associated with limestone quarrying and inadequate rehabilitation of mined out areas. Some beach erosion occurs due to illegal sand mining.

5.4.3 Tourism

The Jamaican representative said that tourism in Jamaica is spatially concentrated on the north coast of the island. Wetlands have been used for dumping and been filled, and beaches have been mined for sand, coastal structures such as groynes, piers and marinas established. The impact of tourism on coral reefs and sea grass beds due to poor sewage disposal and tourism associated recreational activities such as bathing and boating have been significant.

5.4.4 Fisheries

Fishing in Jamaica gives rise to considerable environmental harm especially because of overfishing, the use of dynamite, the destruction of breeding ground such as wetland areas and sea grass beds, pollution of harbours and near-shore water bodies, use of fine mesh nets and traps and dragline methods of fishing. Furthermore there is no practice to return juvenile fish to the wild.

5.4.5 Other areas of environmental concerns

Water resources. Water resource in Jamaica are affected by red mud pollution, saline intrusion where there is an over abstraction of water from some coastal aquifers, chemical, sewage pollution and sedimentation.

Energy. The bulk of the country’s energy supply is obtained from imported fuel. Depending on the sulphur content combustion can result in the production of sulphur dioxide. Charcoal and fuel wood usage is still common and usually causes deforestation and habitat destruction. Historically, the use of leaded gasoline has resulted in high leads levels in some locations.

16 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

Air quality. This is associated with motor vehicle exhaust, stack gases, and the burning of domestic and municipal garbage.

Solid, liquid and hazardous waste. There is no sanitary landfill in Jamaica and open dumping is widespread, with known attendant problems of leachates, open burning, flies, vermin and unpleasant smells. A small percentage of the country is sewered and where sewage plants occur they often malfunction. There is no hazardous waste facility to deal with PCBs, asbestos, waste oils, lead and pecholorcthylene (Perc) among others.

Other concerns. Other concerns relate to oil spills, chemical accidents, transboundary movement of hazardous waste, climate change and attendant sea level rises.

5.4.6 Jamaica and the EVI

The Environmental Vulnerability Index would have some applicability to Jamaica based on a review of the indicators. While information may not be available on all the indicators it is likely that the meteorological and geological data can be obtained. Information may not exist for some of the anthropogenic indicators such as no. 36 which deals with Max 24 hour SO2 concentration and nos. 37 and 38 which seek to determine fertilizer and pesticide usage over the last five years respectively.

The indicators seem to be deficient in evaluating water quality and quantity and this resource is being adversely affected by development in many countries. Indicator no. 47 recognises the importance of water but it seems as if work so far has not yet identified the specific parameter for which data must be sought.

Indicator no.43 requires careful weighting considering the impact that mining and quarrying can have on a small island state such as Jamaica which has extensive bauxite deposits and where limestone constitute more than 66% of the rock material.

5.5 Saint Lucia

Ms Valerie Isaac St Hill said that the most important environmental concerns in Saint Lucia related to (1) deforestation (2) coastal and marine degradation (3) unsustainable land use and (4) occurrence of natural hazards. The following is a summary of Ms Isaac St Hill’s presentation.

5.5.1 Deforestation and soil erosion

The historical pattern of land ownership and the concentration of large tracts of prime agricultural land in the hand of relatively few owners, has resulted in small farmers encroachment on forests and reserves and steep slopes unsuitable for agricultural purposes. The result has been extensive deforestation, soil erosion and siltation of inland and coastal waters and the contamination of inland water courses and coastal waters by runoff from agricultural land.

5.5.2 Coastal and marine degradation

The entire island mass of St Lucia must be considered a holistic system because natural and anthropogenic events no matter how far inland, impact on the shore and marine environment

17 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX within a very short time frame of occurrence. The aggressive development policy being pursued, particularly in tourism and agricultural sectors, has placed many of the coastal resources under stress. This is made evident by the coastal degradation occurring particularly along the Northwest coast of the island in the form of erosion, poor coast water quality, reef degradation and the loss of marine habitat.

5.5.3 Land use

The problems associated with land use are characterised by conflicts between competing uses such as forestry and agriculture, recreation and tourism, environment and tourism, and agricultural and urban environment. These pressures on land have placed all natural areas under severe pressure. The last decade has seen progressive loss of vast tracts of the central forests, almost total destruction of the remaining west coat mangroves, the loss of most areas of natural marshlands and the destruction of several sand beaches.

5.5.4 Natural hazards

Natural hazards in Saint Lucia are primarily associated with tropical cyclones, earthquakes, volcanic activity, land and rock slides and wave action. Tropical cyclones are the most common and damaging natural disasters with potential for severe impacts.

5.5.5 Other areas of environmental concern

Solid and liquid waste management. Solid waste management and the disposal of refuse is one of the most serious environmental issues facing Saint Lucia. Despite improvements in the garbage collection system, serious problems remain from poor solid waste management practices, and illegal disposal of solid wastes along roadsides, in rivers and in other sensitive habitats. In addition to the unpleasant sight and health risks posed by these practices, there is a threat to the ecology of the rivers and the mangroves, and contamination of the water supply and near shore marine eco-system. The human population residing in the coastal areas of Saint Lucia has grown over the years and is still growing, thus increasing the amounts of poorly treated or untreated sewage waste waters being discharged into the coast environment.

Use of chemicals. The intensification of agricultural activities, primarily the banana industry, has led to the increasing use of agro-chemicals to control pests and improve productivity. The regular use of these chemicals, particularly on land with relatively steep slopes, provide the potential for the contamination of streams and rivers which supply the country with drinking water.

5.5.6 Saint Lucia and the EVI

Ms Isaac St Hill said that Saint Lucia was likely to record relatively high scores in the EVI with regard to a number of indicators, in particular those referring to land use and deforestation (indicator nos. 22, 28) soil erosion and coastal misuse (indicator no. 11, 25), tourism (indicator no. 29), natural hazards (indicator nos.2, 7, 8) waste (indicator no. 30, 32) and chemicals (indicator nos. 37, 38).

She argued that the EVI, as presently constructed, required additional indicators or refinement of existing one, to take into account environmental vulnerability associated with ship based pollution, especially that associated with cruise tourism for countries, which like Saint Lucia, experience this reality.

18 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

5.6 Trinidad and Tobago

Dr John Agard said that the most important environmental problems in Trinidad and Tobago are: (1) pollution and (2) deforestation. The following is a summary of Dr Agard’s presentation.

5.6.1 Pollution

Rapid population increase and industrial development fuelled by a petroleum based economy in a small island setting have led to pronounced pollution effects on the natural environment. Although new environmental and planning regulations are being put in place, many developments have taken place without the benefit of adequate physical planning or environmental impact assessments. In the built environment there is widespread pollution of land due to improper disposal of solid and toxic wastes from industry and households. Emissions from motor vehicles and industry contribute to deteriorating air quality not only in industrial areas, but also along major roads due to the presence of the largest vehicle population per capita in Latin America and the Caribbean. Industrial effluents and malfunctioning sewage plants (especially in Tobago) have contributed to the degradation of the lower courses of rivers and nearshore coastal waters.

5.6.2 Deforestation

There has been extensive deforestation in critical watersheds, associated with shifting cultivation, hillside slash and burn agriculture, inappropriate and illegal logging, and sand and gravel extraction. These activities cause siltation in watercourses and lead to flooding. About 31.4% of Trinidad and Tobago consists of natural forests. Deforestation averaged 2600 ha per annum (or about 0.5%/yr) from 1990 - 1995. Of critical importance to the conservation of forest resources is the annual fire problem, which occurs during the dry season. Of the total of 44,850 ha of forests burnt in the ten-year period 1987 – 1996, only 230 ha or 0.5% have been replanted.

5.6.3 Other environmental issues

Coastal resources. Pelagic fisheries and shrimp are the main living resources on the continental shelf around Trinidad and Tobago. Expansion of fisheries exploitation has traditionally been hindered by over-fishing of near-coastal waters by trawlers as well as subsistence and other small scale fishers. Tobago has well developed coral reefs, which are coming under increasing pressure from coastal tourism infrastructure developments.

Disaster Proneness. Fortunately, Trinidad and Tobago is on the southern fringe of the hurricane belt. During this century, 1900 – 1998, only on seven occasions did a tropical storm or hurricane directly affect the country. Five of the tropical cyclones caused no more than torrential showers and strong gusty winds throughout Trinidad and Tobago. The remaining two attained hurricane status but only one, Hurricane Flora did substantial damage to Tobago on September 30, 1963. Trinidad and Tobago have no volcanoes. Trinidad and Tobago are on the edge of the Caribbean Plate and as such are subjected to daily minor earthquake tremors. No extensive earthquake damage has been recorded however. Even so extensive waterfront development on filled land in Port of Spain, Trinidad may make it vulnerable to earthquake induced ground failure.

19 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

5.6.4 Trinidad and Tobago and the EVI

Dr Agard said that Trinidad and Tobago is likely to record very high exposure to environmental risks, degradation and lack of resilience, and therefore likely to register high scores with respect to pollution (indicator nos. 30, 31, 32, 33, 34, 35, 36, 43), and the removal of natural vegetation (indicator nos. 22, 28).

He argued that that the EVI, as presently constructed, required additional indicators or refinement of existing one, to take into account environmental vulnerability associated with oil exploration and production of petroleum products, for countries, which like Trinidad, experience this reality.

20 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

6. DISCUSSION ON THE INDICATORS

The purpose of the discussion was to review all indicators to assess their suitability for SIDS in the Caribbean and IMA regions, and to suggest new indicators to take into account the special circumstances of these regions

Dr Ursula Kaly and Professor Briguglio gave some background information regarding the set of indicators that were being presented for discussion. It was pointed out that initially 57 indicators were selected, but after the Think Tank meeting of September 1999 (see Appendix 3), some of the original indicators were reformulated and their number was reduced to 47. Participants were asked to consider each indicator in turn, to assess its suitability, and to suggest changes. They were also asked to ensure that each indicator has a clear rationale to facilitate an understanding of why it was chosen and what it represents.

The participants then organised themselves into two groups, with one group focusing on Caribbean SIDS and the other on IMA-SIDS.

Group 1 for IMA Region Chair: A. Vella. Members: D. Daby, K. Mercieca, U. Kaly, J. Sammut, P. Gatt, A. Mallia, D. Duca, M. Tabone, L. Micallef, M. Cassar, M. Camilleri

Group 2 for Caribbean Region Chair: J. Agard. Members: V. Isaac St Hill, L. Miller, C. Pratt, L. Grima, C. Attard

Following extensive deliberations, the two groups reconvened in plenary and put forward the suggestions on which there was broad agreement within each group. The participants recommended that SOPAC takes cognisance of these suggestions.

The suggestions are reported below, and are grouped under two headings (1) general and (2) relating to specific indicators.

6.1 General Suggestions

Participants suggested that: § There was a need for an additional column of keywords with every indicator to help the respondent understand better what the indicator is measuring and to highlight the linkages between the different indicators. § Some of the environmental data that are derived from national sources are not properly audited and standardised for international comparisons, and additional care should be taken to reduce the dangers of comparing like with unlike. § The indicators should be accompanied by some coefficient indicating the level of confidence in the data. § The EVI needs to be further legitimised in a peer-reviewed international journal. § In the narrative describing the index, more emphasis should be made on the association between environmental vulnerability and costs and benefits to human systems and welfare § In the same narrative, the policy implications associated with anthropogenic indicators

21 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

should be given importance; § The indicators should be calculated over a span of time, so that rates of change can be calculated so as to be used as a dynamic tool towards natural environmental assessment and management

6.2 Comments on Individual Indicators

The participants considered the 47 indicators which were adopted following the Think Tank meeting held in Fiji in September 1999, and suggested changes where they thought appropriate, as follows:

Indicator 1 relating to deviation in surface sea temperature

No change was suggested to this indicator. It was noted that there was a discrepancy between data compiled by the National Oceanic Atmospheric Agency (NOAA) and data collected locally by different countries. It was pointed out that since NOAA data is more widely available, it should be used for the purpose of this indicator.

Indicator 2 relating to wind speed Indicator 3 relating to monthly rainfall less than or equal to the 100 year low rainfall event Indicator 4 relating to monthly rainfall greater than or equal to the 100 year high rainfall event Indicator 5 relating to temperature greater than or equal to the 100 year high temperature event Indicator 6 relating to temperature smaller than or equal to the 100 year low temperature event

Both groups agreed that using a 100 year event to capture the long term trend, could very likely create data problems. Following discussion it was agreed that the benchmark of a once in 100 year event would be changed to deviations from a 30 year average.

Indicator 7 relating to number of volcanoes with potential for eruption Indicator 8 relating to earthquake energy Indicator 9 relating to number of tsunamis or storm surges Indicator 10 relating to land area Indicator 11 relating to the ratio of the shoreline to land area

No significant changes were suggested for these indicators.

Indicator 12 relating to distance to nearest continent There was some discussion relating to the definition of “continent”. It was agreed that the term “continent” could be interpreted in different ways and that in the notes for respondents, the term should be defined as precisely as possible for consistency purposes.

22 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

Indicator 13 relating to altitude range (highest point - lowest point in country) Indicator 14 relating to percent of land area below 10m above sealevel

No significant changes were suggested for these indicators

Indicator 15 relating to the percent of land area composed of unconsolidated sediments

It was suggested that the wording be changed to “Percentage of land area less than 10m elevation within 2km of coast composed of unconsolidated sediments (exclude coral reefs and ice)”.

It was noted that this indicator is somewhat difficult to measure.

Indicator 16 relating to number of known endemic species per 10,000 sq. km land area

No significant changes were suggested for this indicator.

Indicator 17 relating to the number of reported organism outbreaks

It was suggested that the word “catastrophic” be replaced by “reported and verified by appropriate authorities”.

It was noted that this indicator is too open to interpretation especially regarding the terms “outbreak” and “appropriate authorities”.

Indicator 18 relating to total tonnage of freight imported annually

No significant changes were suggested for this indicator.

Indicator 19 relating to the number of all introduced species since 1900 Indicator 20 relating to the number of endangered and threatened species Indicator 21 relating to the number species which have become extinct since 1900

No significant changes were suggested for these indicators.

There was some discussion on whether the new IUCN definitions should be adopted regarding the status of species. Although it was generally understood that this would refer to the most recent definitions, it was agreed that this point should be made clearer to the respondents in the questionnaire.

23 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

Indicator 22 relating to percentage of natural and regrowth vegetation remaining

No significant changes was suggested for this indicator.

Indicator 23 relating to tonnage of intensively-farmed animal products

It was suggested that the question should clearly indicate that intensive farming should include aquaculture.

Indicator 24 relating to percent of fisheries stocks overfished Indicator 25 relating to density of people living in coastal settlements Indicator 26 relating to total human population density in the country Indicator 27 relating to annual human population growth Indicator 28 relating to the removal of natural vegetation

No significant changes were suggested for these indicators.

Indicator 29 relating to annual number of international tourists

It was agreed that the denominator should be changed from 100 sq. km to land area.

Indicator 30 relating to untreated industrial and domestic wastewater discharged

It was noted that this indicator requires measurement by length of coast and rivers. This could create data problems for countries like Jamaica where rivers go underground.

Indicator 31 relating to generated toxic, hazardous and municipal wastes

There was some discussion relating to the difference of toxicity between domestic and industrial waste. It was agreed that allowance was to be made for this difference by having two separate indicators instead of one.

It was agreed that the indicators should include imported toxic and hazardous materials.

24 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

Indicator 32 relating to mean percent of waste effectively managed or treated

There was discussion relating to the meaning of managed and treated waste. It was suggested that the term “Effective management” would be more appropriate referring to the following strategies: composting; reusing; recycling; controlled incineration (including temperature control, retention time control and control of emissions); and controlled landfill (involving treatment of leachate, containment, gas management, aftercare and rehabilitation i.e. recovery, planting, and post management).

Another suggestion was to change “mean percent of waste” to “cumulative percent of effectively managed waste” … per year. The term cumulative refers to total percentage of wastes managed . It was noted that this information may be very difficult to obtain.

Indicator 33 relating to number spills of oil and hazardous substances

It was noted that spills of oil and hazardous substances are not confined to the coastal areas. It was therefore suggested to change the indicator to refer to the number of spills of oil and hazardous substances greater than 1,000 litres during the last 5 years on land, in rivers or territorial waters.

Indicator 34 relating to industrial facilities that could cause significant damage

Following discussion it was agreed that the denominator should include territorial waters as well as land area. With this new denominator, industrial facilities would therefore include such structures as oil rigs.

Indicator 35 relating to number of cars / land area

It was agreed that the term “vehicles” should be used instead of cars, as per World Bank Definition. There was some discussion on whether to use density of vehicles on roads rather than total land area. It was noted that this may not be appropriate, since the former is used as a measure of congestion, rather than of the ability of land to attenuate pollution emitted by vehicles.

Indicator 36 relating to Max 24 hour SO2 concentration Indicator 37 relating to Tonnes of N,P,K fertilizers used on agricultural land Indicator 38 relating to Tonnes of pesticides used on agricultural land

No significant changes were suggested for these indicators.

25 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

Indicator 39 relating to the number of new fisheries stocks added

There was discussion regarding the term “new fisheries species added”. It was pointed that (a) new fisheries does not take into account added effort or new technology (b) a threshold of at least 20% in increase in catches should be set (c) the way the question was formulated could be misleading because it focuses on species rather than on stocks.

It was suggested that the wording of this indicator be changed as follows: “Number of new fisheries stocks and added effort exploited by countries over the last 5 years”

It was also agreed to develop this indicator further to incorporate points (a) and (b).

Indicator 40 relating to percentage of land area degraded since 1950

It was noted that data for this indicator could be difficult to obtain. It was agreed that there needs to be a clear rationale to show that the indicator captures erosion, salination and desertification and that this indicator should exclude urban areas.

Indicator 41 relating to annual internal renewable water resources per capita

A properly specified indicator was required to capture quantity and quality of freshwater separately.

It was suggested that an indicator to capture quantity would be the following: “Annual internal renewable water resources per capita.” This would be measured by average annual runoff and recharge of groundwater from endogenous precipitation. The rationale for measuring the renewable water supply per capita was that lower availability per head would create higher pressures on natural ecosystems (i.e. water for people is considered to have a higher priority than for ecosystem conservation)

It was noted that the issue of water quality has been covered by indicators 30, 33, 37 and 38.

Indicator 42 relating to kilotonnes of mining material (ore + tailings) extracted

No significant changes were suggested for this indicator.

Indicator 43 relating to land, rivers and coastal zone affected by mining & quarrying

It was suggested that sea area should not be included in this indicator. The indicator should therefore read as follows: “ Percentage of land, rivers and coastal zone affected by mining and quarrying” It was noted that “coastal zone” needs to be defined more precisely .

26 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

Indicator 44 relating to percent of terrestrial zone set aside as reserves Indicator 45 relating to percent of marine zone set aside as reserves Indicator 46 relating to number of war or civil strife years over the last 50 years

No significant changes were suggested for these indicators.

Indicator 47 relating to environmental legislation

There was considerable discussion regarding this indicator and its usefulness. It was suggested that this indicator be dropped, because aspects of environmental management are covered by other indicators.

6.3 Suggested Possible New Indicators

Participants agreed to suggest the following indicators for possible inclusion in the EVI for further consideration by SOPAC’s experts.

Reported mass mortalities of organisms, including strandings

Basically this indicator would relate to possible imbalances in the environment

. The number of ships/tonnage of hazardous substances carried/transiting within 100km of a country per year (averaged over the last 5 years)

This indicator was considered to be significantly different from Indicator 18

Pollution coming in currents of air or sea or rivers from outside the territory

This would require a yes/no response when a country is or is not downstream by air or water from a major pollution source, within 1,000 km. It was noted that this indicator would require further development and refinement.

It was noted that this indicator could possibly be captured by indicators 26, 33 and 34.

Total tonnage of chemicals and hazardous material imported and in transit within the country.

It was noted that this indicator could add useful information to indicator 31, since this indicator deals only with waste.

27 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

6.4 Possible Indicators for the Subsidiary List

Participants identified a list of important environmental issues for which indicators could eventually be developed, but could not be included now due to data limitations. These were:

Marine and forest productivity

Salination of groundwater – it was pointed out that this may be captured by indicator 40

Hail and glacial melt

6.5 Possible Weighting Schemes

Participants discussed the weighting procedures that could be adopted for the indicators to underscore their importance in terms of their contribution to the overall index.

Dr. Ursula Kaly said that in the original EVI, six of the 57 indicators were assigned a weighting factor of 5, while the remaining indicators were given the default weighting of 1.

She said that at the Think Tank, the 47 indicators were rated in terms of the importance attached to them by the experts, on a scale ranging from 0 to 4, with zero indicating no importance and 4 indicating the highest importance. The values thus obtained were averaged over all of the Think Tank experts and the final score for each indicator used to assign a low, medium or high weighting to each. High importance (weight) was assigned to 15 indicators and low importance to 14 indicators out of a total of 47. The remaining 18 indicators were assigned medium importance.3

Participants agreed that it is not possible to use weights derived on the basis of ecological criteria alone, and that a measure of subjectivity would have to be employed in this regard

Participants therefore agreed that the indicators would be weighted in terms of their perceived importance by the experts associated with the compilation of the index. This was the same conclusion arrived at during the Think Tank meeting.

It was also agreed that ultimately it will be the responsibility of the experts compiling the index to choose appropriate weighting, based on plausible criteria and following extensive consultation.

Participants agreed that the issue of weighting required further study and debate.4

3 See Appendix 3 for more details. 4 The participants at the Malta meeting were also asked to rank the indicators in order of importance. The results are reported in Appendix 3.

28 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

7. STRENGTHS, WEAKNESSES, OPPORTUNITIES & THREATS

The participants discussed the strengths, weaknesses, opportunities and threats facing the EVI. The following are the major conclusions reached by participants.

7.1 Strengths

Education . The EVI could be a useful instrument in the hands of governments, organisations and individuals to foster education about the environment. Even lack of data could be educational. For example if a country fails to collect data relating to SO2 levels in urban areas, the index will have a data gap which gives an important signal in itself. In other words, the index helps countries to find out what they do and do not know about their environment.

Information and visibility. The indicators could render environmental problems more visible, since all the indicators relate to major areas of environmental concern. They could also help governments to prioritise certain environmental problems.

Corrective measures. The index could allow governments to apply corrective measures to rehabilitate the environment, in line with the level of harm detected by the indicators. It could also serve as a useful guide for technology requirements, capacity building and development assistance. If the EVI is measured over time (say over five years), it could also produce information on whether the corrective measures are yielding positive results.

Linkages. The indicators could be used to highlight linkages. For example, tourism and agriculture could lead to undesirable levels of environmental harm. Water production, which requires electricity, which in turn is oil produced, may lead to undesirable levels of pollution. It was pointed out that these linkages may not appear clearly in the indicators themselves, and should therefore be given a high profile in the narrative accompanying the index.

Non-linearities and qualitative data. The indicators, as structured, allow the combination of quantitative and qualitative data. They also permit non-linear environmental data to be amalgamated in a form that could potentially be grasped by decision-makers. In other words, the EVI could be instrumental in simplifying complicated events into a form that is relatively easy to grasp. This appeals to decision-makers, although it may be a difficult task for those collecting and mapping the data.

Capacity building in data collection. The EVI has the potential of strengthening the process of data collection. The questionnaire itself could assist the statistical division of the Ministries for the Environment to identify statistical gaps and to try to fill them. It was pointed out that a manual on the data required would be useful in this regard.

International comparisons - fulfilling a mandate. The EVI allows comparisons within and between countries, as requested by the global community in the Barbados Programme of Action (1994) and the Special Session of the UN General Assembly (1999).

29 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

7.2 Weaknesses

Indirect measures. Many of the indicators are measured by proxy variables and not directly. This could lead to errors, since a proxy variable may capture factors not associated with the true variable which the indicator attempts to measure. If many such proxies are used, this could lead to cumulative errors. It was pointed out however that this weakness is not specific to the EVI, since it frequently occurs in compilation of indices and in quantitative estimations in general.

Weighting. The EVI is itself vulnerable to arbitrary weighting. Different weighting schemes could lead to different answers.

Lack of feedback mechanisms. The calculation method used in the EVI does not cater for feedback mechanisms or synergistic effects of one indicator to another.

Hides variation. Averaging and summarising data may hide important variations. This may lead to complacency when, for example, a positive score for one indicator cancels out a negative score for another indicator.

Comparing like with unlike. The index tries to compare large countries like Canada and Australia, with small countries like Tuvalu. The validity of such comparisons may be questioned. It was noted that this problem occurs also in the case of economic comparisons between countries.

7.3 Opportunities

Various advantages. The advantages listed in section 7.1 of this report with regard to education, information and capacity building, were also cited as opportunities.

Demand at the international level. The construction of the EVI provides an opportunity for responding to the call for such an index in various documents, including the Barbados Programme of Action (1994) and the final document of the Special Session of the UN General Assembly (September 1999).

Awareness and support by regional organisations. There was increasing awareness by regional organisations of the benefits of constructing the EVI. The Pacific regional organisations, the Indian Ocean Commission and the OECS were mentioned in this regard. There is therefore the opportunity of enlisting the support of these organisations for the further development of this index.

The National level. Some countries are interested in the construction of the EVI because, amongst other considerations, the index could help them support their claim for special status.

30 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

7.4 Threats

Lack of acceptance by Governments. Participants expressed concern about the possibility that the EVI might not be accepted by some governments because the single number which the index returns for a particular country might conceal the many facets of environmental vulnerability. It was pointed out that in the Pacific region, some countries have already expressed their preference for a vulnerability profile report, rather than a vulnerability index. It was stressed however the reporting of the results of the EVI will include both an index and profile formats, so that areas of concern can be easily identified.

Problems with academic acceptance. The EVI may not be accepted on scientific grounds because if viewed from the perspective of a specific discipline, say, biology, it may lack rigour. Participants agreed that this threat would be reduced if a paper on the EVI is published in a scientific journal.

Participants suggested that Dr. Ursula Kaly, Mr. Craig Pratt and Prof. Lino Briguglio should try to publish a paper in the EVI in a suitable journal. Many possible journals were suggested, including the following journals: Environmental Management; International Environmental Management; Our Planet; Environmental Monitoring & Assessment; Ambio. It was also suggested that specialist studies could be published on the sub-indices in more specific journals, such as Conservation Ecology, Environmental Conservation, or Nature.

Funding. There is the threat that the further development of the EVI may be stalled due to lack of funding. The problems with procuring funds may be due to various factors, including the perception that the exercise will not produce meaningful results for policy and decision making, and the fact that other projects competing for funding may be given higher priority. To combat such a threat, participants agreed that the narrative accompanying the questionnaire should highlight the benefits of the index for decision-making.

31 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

8. FRAMEWORK FOR PROCUREMENT OF DATA

Dr Ursula Kaly and Mr Craig Pratt opened the discussion by giving some information about the manner in which data was to be collected. They said that SOPAC had prepared data sheets, to be filled by respondents, with a whole page assigned for each indicator. Space was provided for the response, the source of the data, and other information.5 The discussion which followed focused on the roles of governments, regional organisations and other institutions in data collection.

8.1 The Role of Governments

Participants agreed that the EVI development process will require formal endorsement/ approval by governments, and that some government institution should be involved officially in the procurement of the data. This would be spread across different ministries, therefore the respondent might need to approach the Prime Minister’s Office, or some high level environmental authority, to facilitate collection of data.

8.2 The Role of International Organisations

Some data would be easier to obtain from international organisations such as WMO, UNDP or UNEP. These organisations base their data on statistics provided by member countries. It was noted in this regard that (1) often data provided by international organisations has important gaps, notably with respect to small states and (2) some of the environmental data required for the EVI is not likely to be collected by international organisations. It was argued however that international organisations could act as catalysts for data procurement.

8.3 The Role of Regional Organisations

It was pointed out that in the Pacific, where the whole exercise started, SOPAC found useful backing from regional organisations, in particular the Pacific Islands Forum and the national governments. The Indian Ocean Commission and the IMA-SIDS Centre for Sustainable Development could be asked to facilitate data collection in the IMA region, while the OECS Caricom and the Association of Caribbean States (ACS) could do the same in the Caribbean region. However, national government endorsement will remains essential, even if regional organisations are willing to cooperate in data collection.

8.4 The Role of Universities and Research Institutions

It was pointed out that considerable research takes place within Universities and that some of the data required for the EVI has to be “researched”. It was agreed that Universities have an

5 Each response sheet will contain five sections as follows: (1) the text of the indicator; (2) information for the respondent regarding what the indicator represents and possible data sources; (c) a formatted space for the respondent to enter the data; (4) comments by the respondent regarding data quality and data source/s; and (5) details about the respondent and stamp of authority verifying the data.

32 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX important role to play, especially by collaborating with the Ministries of the environment for evaluating the quality of the data.

8.5 The Role of Private Organisations

Data could also be produced by private firms. There are advantages associated with such an arrangement. The product is likely to be more efficiently produced and will not depend on political exigencies. On the other hand there is the danger that the data will not be properly audited.

33 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

9. POSSIBLE AVENUES FOR FUNDING

9.1 Funding at the International level

Prof. Briguglio introduced the discussion by explaining that the initial funding for the EVI came mostly from New Zealand. He explained that the fact that the current workshop was being funded by UNEP indicates that the interest in the index has taken a global dimension. He said that there was the need to find further funding for globalising the scope of the EVI.

He said that UNEP had already indicated that it supported the EVI process by providing funds for the Malta workshop and assistance for the SOPAC Think Tank meeting. He asked whether UNEP will be prepared to provide further funding.

Ms Khaka said that there UNEP was interested in funding the further development of the EVI, as it was in line with the decision of UNEP’s Governing Council. However, this was subject to the availability of funds.

The discussion which followed dealt with a number of issues, the most important of which dealt with the possible roles of UNEP and SOPAC in the future development of the Index.

The question was raised as to whether or not, in the event that UNEP funds the further development of the index, SOPAC would still have the veto regarding possible changes in the EVI.

Participants agreed that there should be a line of continuity in the work that has been done in Fiji, before and during the SOPAC Think Tank meeting of September 1999 and that SOPAC should therefore keep the intellectual “ownership” of the development of the EVI for the next 2 years, meaning that SOPAC should remain the final arbiter of how the index is to be developed, even at the global level. After 2002, the index could be managed by some global organisation.

Participants considered the possibility that the work done by SOPAC be eventually passed into wider UNEP context, thereby anchoring the EVI within UNEP to establish a formal framework, applicable to all countries of the world. It was agreed that (1) such an arrangement would enhance the acceptance of the index at the global level and (2) before this arrangement be effected, the work related to the further development of the EVI should proceed within SOPAC to enable the experts who initiated the process to continue with the initial phases of the EVI development, and (3) that the arrangement could come into effect by the end of 2002.

9.2 At the Regional Level

Participants discussed the involvement of the Caribbean and IMA regions in the further development of the EVI.

It was pointed out that there were various regional organisations that could be instrumental in obtaining funding for the further development of the EVI in the IMA and Caribbean regions. Amongst these the following were mentioned.

34 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

IMA regions · The Indian Ocean Commission · The IMA-SIDS Centre for Sustainable Development · The Mediterranean Action Plan

Caribbean Region · The Organisation of East Caribbean States (OECS) · The Association of Caribbean States (ACS) · Caricom

9.3 At the National Level

It was noted that funding by individual SIDS is not likely. However funding could be sought from certain donor countries, such as New Zealand, Australia, Canada, Norway and others that have an interest in the sustainable development in SIDS. It was agreed that donor countries be called upon to assist in the process of further developing the EVI.

35 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

10. ADOPTION OF FINAL STATEMENT

Participants adopted the following statement during the final session of the meeting.

“ The participants at the Workshop on the Environment Vulnerability Index held in Malta between November 29 and December 3, 1999:

Agree:

(a) That an appropriate structure for the development of the EVI would be the following: (i) A suitable international organisation, such as UNEP, takes the role of international co-ordinating body for globalising the scope of the EVI; and (ii) Regional bodies, such as Association of Caribbean States, Caricom, the IMA SIDS Centre for Sustainable Development and SOPAC, act as co-ordinating centres for the development of the EVI at the regional level.

(b) That SOPAC should continue to develop the EVI at the technical level until such a task can be effectively transferred to a suitable international organisation, such as UNEP, preferably by the first quarter of the year 2002.

(c) That the appropriate national co-ordinating institution should be the respective governments, since the EVI requires endorsement by the respective governments and the government is normally the depository of most of the data required.

(d) That in countries where there is a functional environmental authority such as the EMA in Trinidad and the NRCA in Jamaica, the co-ordination at country level could be placed under the auspices of these authorities.

(e) That where appropriate, the private sector and Universities should be involved in the process.

(f) That funding should be sought to invite six additional countries from the IMA and Caribbean regions to participate directly in the procurement of data for constructing the EVI for these countries, and that such data should be collected by the end of year 2000.

(g) That by the end of 2001, data should be collected for 15 other countries to represent all country possibilities (small and large, arctic, temperate and tropic, developed and developing, etc.) to enable the carrying out of various tests on the indicators, including redundancy tests.

(h) That funding should be sought for follow-up expert meeting to be held in the Caribbean region sometime in 2001 to review progress in the development of the EVI.

(i) That a manual relating to the EVI should be written so as to give information about the purpose and usefulness of the index, the manner in which the data is to be collected and standardised, and other matters associated with the indicators.

36 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

Call upon:

(a) UNEP, to continue to support and facilitate funding for the development of the EVI according to its mandate by its Governing Council and in line with the Barbados Programme of Action and its review carried out at UNGASS (September, 1999).

(b) AOSIS, to support the development of the EVI and encourage its member governments to collaborate in the process, especially by providing data and facilitating data collection.

(c) Appropriate regional organisations, to promote interest and assistance for the development and utilisation of the index in the countries of their region.

(d) Governments, to provide the necessary data and to facilitate the collection of data for the construction of the EVI.

(e) Donor countries and organisations, to support the further development of the EVI, in line with the Barbados Programme of Action and its review carried out at UNGASS (September, 1999).

Express their gratitude to:

(a) UNEP, for supporting the development of the EVI, by providing funding for the current workshop, and for supporting participation in the Think Tank meeting organised by SOPAC in September 1999.

(b) SOPAC, for the leading role it has taken in the development of the EVI and for agreeing to collaborate with UNEP and the ISSI in the organisation of the present workshop.

(d) The OECS, the University of Mauritius, the University of the West Indies (St Augustine Campus), University of Malta, the Malta Environment Protection Department of the Ministry for the Environment, the Malta Planning Authority, Lino Bianco and Associates, Eco-Standards Services Limited, the Centre for Insular Coastal Dynamics (Malta), the International Environment Institute (Malta) and the Institute for Environment Studies (University of Toronto) for providing expertise for this workshop. ”

37 APPENDIX 1

AGENDA AND PROGRAMME

of the UNEP Meeting of Experts on the Environmental Vulnerability Index held in Malta between 29 November and 3 December 1999 at the Foundation for International Studies, Valletta, Malta organised by the Islands and Small States Institute (Malta) in collaboration with the South Pacific Applied Geoscience Commission (SOPAC) APPENDIX 1: AGENDA AND PROGRAMME

UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

AGENDA AND PROGRAMME

MONDAY 29 NOVEMBER

0930 -0940 Inauguration by Mr. Leslie Agius, Chief Executive of the Foundation for International Studies 0940 - 0950 Opening speech by Ms Elizabeth Khaka, on behalf UNEP 0950 - 1000 Address by Prof. Lino Briguglio on behalf of the Islands and Small States Institute and on behalf of Dr. Francis Zammit Dimech, Maltese Minister for the Environment. 1000 - 1100 Presentation by Dr. Ursula Kaly: “The Development of the EVI: Background Information” 1100 - 1115 Coffee Break 1115 - 1300 Presentation by Mr. Craig Pratt: “Recommendations from the Fiji Think Tank Experts regarding the EVI” 1300 - 1430 LUNCH 1430 - 1530 Discussion on Dr. Kaly and Mr. Pratt’s presentations. 1530 - 1545 Coffee Break 1545 - 1700 Discussion on basic concepts: “Environment, Vulnerability, Resilience, Damage and Degradation” 1900 Reception

TUESDAY 30 NOVEMBER

0930 - 1015 Special characteristics of SIDS in the Indian Ocean, Mediterranean and Atlantic (IMA) regions: Country statements by Malta, Mauritius and Cape Verde. 1015 - 1100 Special characteristics of SIDS in the Caribbean region: Country statements by St Lucia, Trinidad and Tobago and Jamaica 1100 - 1115 Coffee Break 1115 - 1300 Break-out in two working groups: Group 1: Indicators for the IMA regions Group 2: Indicators for the Caribbean region 1300 - 1430 LUNCH 1430 - 1600 Continuation of work in two groups 1600 - 1615 Coffee break 1615 - 1800 Continuation of work in two groups

Appendix 1: 1 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

WEDNESDAY 1 DECEMBER

0930 - 1100 Report of the two working groups. 1100 - 1115 Coffee break 1115 - 1300 Discussion on the Individual EVI indicators. 1300 - 1430 LUNCH 1430 - 1600 Discussion on the Individual EVI indicators (continued). 1600 - 1615 Coffee break 1615 - 1800 Discussion on the Individual EVI indicators (continued).

THURSDAY 2 DECEMBER

0930 - 1100 Discussion on “Strengths, Weakness. Opportunities and Threats relating to the EVI (SWOT Analysis)” 1100- 1115 Coffee break 1115 - 1300 SWOT analysis of the Model (continued) 1430 – 1530 Discussion “Framework for Procurement of Data” 1430 - 1630 Discussion “Possible Avenues for Funding”

FRIDAY 3 DECEMBER

0930 - 1100 Presentation and discussion on the structure of the report of the meeting 1100 - 1115 Coffee break 1115 - 1145 Presentation and discussion on Final Statement 1145 - 1200 Adoption of Final Statement 1200 - 1215 Closure

Appendix 1: 2 APPENDIX 2

LIST OF PARTICIPANTS

at the UNEP Meeting of Experts on the Environmental Vulnerability Index held in Malta between 29 November and 3 December 1999 at the Foundation for International Studies, Valletta, Malta organised by the Islands and Small States Institute (Malta) in collaboration with the South Pacific Applied Geoscience Commission (SOPAC)

APPENDIX 2: LIST OF PARTICIPANTS

LIST OF PARTICIPANTS

JOHN AGARD Department of Life Sciences, The University of the West Indies, St. Augustine, Trinidad and Tobago. Tel: (1-868) 645 3232; Fax: (1-868) 645 7132; Email: [email protected]

CHRISTOPHER ATTARD 14 Capuchins Street, Victoria, Gozo, Malta. Tel: (356) 2290 1606; 551780 Fax: (356) 224846; 565648; email: [email protected]

LINO BRIGUGLIO Director, Islands and Small States Institute, Foundation for International Studies, Malta, Tel: (356) 248218 Fax: (356) 248218; e-mail: [email protected] or [email protected]

MARGUERITE CAMILLERI Queens’s College, Cambridge CB3 9ET, UK. Tel: (44) 1223 767199; Fax: (44) 1223 333392; email: [email protected]

MICHELLE CASSAR Euro-Mediterranean Centre on Insular Coastal Dynamics (ICOD), Foundation for International Studies, St. Paul Street, Valletta VLT07, Malta. Tel: (356) 240746; (356) 245764; email: [email protected]

DEOLALL DABY Marine/Environmental Sciences, Faculty of Science, University of Mauritius, Redui, Mauritius. Tel: (230) 454 1041/ (230) 454 9958; Fax: (230) 465 6928/ (230) 454 9642; email: [email protected]

DEMITRIO DUCA Planning Authority, St. Francis Ravelin, Floriana, Malta. Tel: (356)240976; Fax: (356) 230589; email: [email protected]

PETER GATT 327 Manoel Dimech Street, Slimea, Malta. Tel: (356) 331807; Fax: (356) 247751; email: [email protected]

LINO GRIMA Institute of Environmental Studies, University of Toronto, 33 Wilcocks Street, Suite 1016, Toronto M5S 3E8, Canada. Tel: (1-416) 978 3486; Fax: (1-416) 978 3884 email: [email protected]

VALERIE ISAAC ST HILL Natural Resources Management Unit, OECS, Mourne Fortune, PO Box 1383, Castries, St Lucia. Tel: (758) 452 1847; Fax: (758) 452 2194; email: [email protected]

Appendix 2: 1 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

URSULA KALY Environmental Adviser to Government of Tuvalu, Funafuti, Tuvalu. Tel: (688) 20175/ (688) 20999; Fax: (688) 20664. Email [email protected]

ELIZABETH KHAKA Programme Office, Division of Environmental Policy Implementation, UNEP, Nairobi, Kenya. Tel: (254) 2 623 990; Fax: (254) 2 622788; email: [email protected]

ADRIAN MALLIA Environmental Management Unit, Planning Authority, St. Francis Ravelin, Floriana CMR01, Malta. Tel: (356) 230771; 2290 1546/7; Fax: (356) 224846; email: [email protected]

KEVIN MERCIECA Environment Management Unit, Planning Authority, St. Francis Ravelin, Floriana CMR 01. Tel: (356) 240976; Fax: (356) 224846; email: kevin.mercieca@pa- malta.org.

LAWRENCE MICALLEF Eco-Standards Services Ltd. P.O. Box 61, B’Kara, Malta. Tel: (356) 483791; Fax: (356) 443164; email: [email protected]

LEARIE MILLER Natural resources Conservation Authority (NRCA), The John McIntosh Building, 10, Caledonia Avenue, Kingston 5, Jamaica. Tel: (1-876) 754-7526, (1-876) 754-7540-4; Fax: (1-876) 754-7594-6; email: [email protected]

CRAIG PRATT South Pacific Applied Geoscience Commission (SOPAC), Private Mail Bag, GPO, Suva, Fiji. Tel: (697) 381377; Fax: (697) 370040; email: [email protected]

JOE SAMMUT Environment Protection Unit, Waste Management Section, Sant Antnin Recycling Plant, M’Scala, Malta. Tel: (356) 632654; Fax: (356) 636281; email: [email protected]

MATTHEW TABONE Pollution Control Co-ordinating Unit, Starkey Annex, Fort Saint Angelo, Vittoriosa, Malta; Tel: (356) 676395; (356) 660108; email: [email protected]

ADRIANA VELLA Department of Biology, University of Malta, Msida, Malta. Tel: (356) 3290 2970; Fax: (356) 3290 3049; email: [email protected]

Appendix 2: 2

APPENDIX 3

LIST OF INDICATORS FOR THE ENVIRONMENTAL VULNERABILITY INDEX

UNEP Meeting of Experts on the Environmental Vulerability Index organised by the Islands and Small States Institute, Malta in collaboration with SOPAC, Fiji, held at the Foundation for International Studies, Valletta, Malta 29 November – 3 December 1999 APPENDIX 3: LIST OF INDICATORS

LIST OF INDICATORS FOR THE ENVIRONMENTAL VULNERABILITY INDEX

This Appendix contains two sets of indicators, the first relates to the list compiled following the suggestions put forward during the SOPAC Think Tank meeting held in Fiji between 7 and 10 September 1999, and the second contains a modified list as proposed during the meeting of experts held in Malta between 29 November and 3 December 1999.

BACKGROUND

The process of developing the EVI started by a team of consultants appointed by SOPAC in 1998 and the first set of indicators were published in early 1999. The Think Tank meeting was convened in September 1999, with the objective of bringing together a group of experts from a range of disciplines central to the EVI, to subject the Index to critical peer review. The experts considered the structure of the model, the indicators used, procedures for mathematical testing and future directions for development of the Index. The assembled group of experts generally accepted the EVI approach, as originally proposed by the SOPAC team of consultants, but suggested changes to the indicators.6 The changes have been accepted by SOPAC in principle, and incorporated in the indicator questionnaire. The final list of indicators included 47 questions - 10 fewer that the original EVI, is presented below.

CLASSIFICATION OF THE INDICATORS

The indicators are classified into 5 categories: M = Meteorological; G = Geological; B = Biological; C = Country Characteristics; and A = Anthropogenic.

These indicators can be grouped into three sub-indices namely: REI = Exposure to natural or human risks / hazards EDI= Environmental Degradation Index. This measures present status of the 'health' of the environment It is based on the assumption that past impacts affect the ability of the environment to tolerate new impacts. IRI = Intrinsic resilience index

MAPPING OF RESPONSE

The response to each question is mapped on a 1 to 7 scale, with 1 referring to the lowest possible impact and 7 to the highest. After adjustment for weighting7, the scores are averaged to produce a summation taking the value of between 1 and 7. This summation is reported for the EVI and for its sub-indices.

Where data for a particular question are unavailable, the relevant indicator will be omitted from the average, so that it makes no contribution to the mean. At least 80% of the indicator questions have to be answered for a valid EVI to be calculated for a particular country.

6 A full report of the Think Tank meeting is given in Kaly, U., Briguglio, L., McLeod, H., Schmall, S., Pratt, C. and Pal, R. (1999). Proceedings of the Environmental Vulnerability Index (EVI) Think Tank 7-10 September 1999. SOPAC Technical Report. 7 Originally the SOPAC team assigned a weight of 1 to 51 of the original 57 indicators and a weight of 5 to the remaining six indicators, on the assumption that the latter had a much higher impact on the environment than the former. This approach was very tentative, and it was recognised that some guidance was necessary from the Think Tank experts. However, no broad agreement was forthcoming regarding this matter from the Think Tank meeting, although it was stressed that appropriate weighting of the EVI indicators was a very important requisite. At the Think Tank meeting, the 47 indicators ere rated in terms of relative importance by participants and a simple weighting scale involving 5 categories was suggested.

Appendix 3: 1 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

THE EVI INDICATORS LIST COMPILED FOLLOWING THE SUGGESTIONS PUT FORWARD DURING THE SOPAC THINK TANK MEETING HELD IN FIJI BETWEEN 7 AND 10 SEPTEMBER 1999

Importance Assigned8 Indicator Sub-Index Category Indicator Mean SD10 f(0)11 Number Score9

1 REI M Greatest average annual deviation in Surface Sea Temperature in last 5 2.48 1.12 1 years from long term mean (30 years) (more work required to finalise form) (Centralised database) 2 REI M % of reference climatological stations experiencing >= 1/100 yr. 3 2.48 1.25 1 second wind gusts in a 5 year period / land area 3 REI M % of reference climatological stations experiencing <= 1/100 yr. 2.35 1.18 1 minimum annual rainfall in a 5 year period / land area 4 REI M Cumulative number of 24 hr periods over all reference climatological 2.14 1.15 1 stations over last 5 years during which rainfall is >= 1/100 yr. event / stations 5 REI M % of reference climatological stations experiencing >= 1/100n yr. daily 2.10 1.18 1 max temp in a 5 year period 6 REI M % of reference climatological stations experiencing <= 1/100 daily min 2.05 1.16 1 temp in a 5 year period / land area 7 REI G Number of volcanoes with potential for eruption >= VEI 4 (Volcano 2.10 1.18 1 explosivity Index) within 100km of country land boundary / area of land 8 REI G Earthquake energy within 100km of country land boundaries / land area 1.95 1.32 2 with ML >=6.0 and <=15km depth per 5 years 9 REI G Number of tsunamis or storm surges with run-up >2m above MHWS / 1.95 1.24 2 100km coastline since 1900 10 IRI C Total land area (sq. km) 2.71 1.45 2 11 IRI C Ratio of length of ocean shoreline : total land area 2.62 1.32 1 12 IRI C Distance to nearest continent (km) 1.90 1.09 0 13 IRI C Altitude range (Highest point – lowest point in country) 2.19 1.08 1 14 IRI C Percent of land area <10m above sea-level 2.57 1.16 0 15 IRI C Land area below 10m elevation with unimpeded access to the coast 2.00 1.22 2 composed of unconsolidated sediments (excluding coral reefs and ice) / land area (%) 16 IRI C Number of known endemic species / 10,000 sq. km land area 2.52 1.17 1 17 REI B Number of catastrophic organism outbreaks over the last 5 years / land 2.29 1.10 0 area (pathogens, blooms, plagues etc) 18 REI B Total tonnage of freight imported / year 2.05 1.20 1 19 EDI B Number of all introduced species / 10,000 sq. km land area since 1900 2.14 1.42 3 20 EDI B Number of endangered & threatened species / 10,000 sq. km of land area 2.76 1.18 1 (IUCN definitions) 21 EDI B Number species which have become extinct since 1900 / 10,000 sq. km 2.71 0.85 0 land area (IUCN definitions) 22 EDI B Percentage of natural & regrowth vegetation remaining (e.g. forests, 2.67 1.43 2 mangroves, saltmarshes, prairies, savannah, desert, tundra) 23 EDI B Tonnage of intensively-farmed animal products / yr. / land area 1.81 1.08 2 24 EDI B Percent of fisheries stocks overfished (FAO) 2.6 1.03 0 25 EDI A Density of people living in coastal settlements (define area) 2.6 1.12 1 26 REI A Total human population density (per sq. km land area) 3.6 0.75 0 27 REI A Annual human population growth rate (average over last 5 years) 2.9 0.94 0

8 Perceived importance by 21 experts present at the Think Tank Meeting, who were asked to rank the 47 indicator in terms of the importance attached on a scale ranging from 0 to 4, with zero indicating no importance (to be discarded) and 4 indicating the highest importance. 9 This represents the mean of the scores assigned by the Think Tank experts. 10 Standard Deviation associated with the mean. 11 f(0) refers to the number of times that a 0 was assigned to the indicator.

Appendix 3: 2 APPENDIX 3: LIST OF INDICATORS

Importance Assigned8 Indicator Sub-Index Category Indicator Mean SD10 f(0)11 Number Score9

28 REI A Net percentage of land area changed by the removal of natural vegetation 3.0 1.14 1 over last 5 years 29 REI A Annual number of international tourists * average days stay / 365 / 100 2.1 1.16 1 sq. km (last 5 years) 30 REI A Megalitres of untreated industrial and domestic wastewater discharged to 2.6 1.28 1 aquatic system / 1,000 km aquatic ecosystems (length coast + length rivers) 31 REI A Total tonnage of generated and net imported toxic, hazardous & 2.0 1.14 0 municipal wastes/ 10,000 sq. km land area / year (average last 10 years) 32 REI A Mean percent of industrial and municipal waste managed or treated / yr. 2.0 1.16 2 33 REI A Number spills of oil and hazardous substances >1,000 litres during last 5 1.7 0.85 0 years within territorial waters 34 REI A Number of nuclear, chemical and other major industrial facilities that 1.7 0.96 0 could cause significant damage / 10,000 sq. km land area 35 REI A Number of cars / land area 2.4 0.97 0

36 REI A Max 24 hour SO2 concentration (micro g /cubic m) (average over last 5 2.3 0.86 0 years) 37 REI A Tonnes of N,P,K fertilisers used / 10,000 sq. km agricultural land area / 1.9 0.82 0 year (average last 5 yr.) 38 REI A Tonnes of pesticides used / 10,000 sq. km of agricultural land / year 1.9 0.84 0 (average last 5 years) 39 REI A Number of new fisheries species added to country over last 5 years 1.6 1.02 2 (within territory) 40 EDI A % Land area degraded since 1950 3.1 1.14 1 41 EDI A Question on importance of water…? 2.6 1.35 1 42 REI A Kilotonnes of mining material (ore + tailings) extracted / 10,000 sq. km 1.3 0.64 1 land area / year (average last 5 years) 43 EDI A Land and sea area affected by mining & quarrying / land area 1.7 1.10 2 44 EDI A Percent of terrestrial zone set aside as reserves 2.3 1.10 0 45 EDI A Percent of marine zone set aside as reserves (mean high tide to 1.9 1.22 2 continental shelf) 46 EDI A Number of war or civil strife years over the last 50 years within the 2.1 1.20 3 territory 47 REI A Environmentally related legislation with regulations na na na

Appendix 3: 3 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

THE EVI INDICATORS LIST COMPILED FOLLOWING THE SUGGESTIONS PUT FORWARD DURING THE MALTA MEETING BETWEEN 29 NOVEMBER AND 3 DECEMBER 199912

Importance Assigned13 Indicator Sub-Index Category Indicator Mean SD15 f(0)16 Number Score14

1 REI M Greatest average annual deviation in surface sea temperature during last 2.9 1.07 0 5 years, from long term average (use a 30 year average).17 2 REI M Number of days over the last 5 years during which the maximum 2.4 1.22 1 recorded wind speed (3 sec gusts) was at least 20% higher than the average maximum for that month (use a 30 year average for each month as reference, and average the results over all reference climate stations). 3 REI M Number of months over the last 5 years during which rainfall was at least 3.1 0.83 0 20% lower than the 30 year average for that month (use a 30 year average for each month as reference, and average the results over all reference climate stations). 4 REI Met Number of months over the last 5 years during which rainfall was at least 2.9 0.86 0 20% higher than the 30 year average for that month (use a 30 year average for each month as reference, and average the results over all reference climate stations). 5 REI M Number of days over the last 5 years during which the maximum 3.1 0.86 0 temperature was at least 50C higher than the mean monthly maximum for that month (use a 30 year average for each month as reference, and average the results over all reference climate stations). 6 REI M Number of days over the last 5 years during which the minimum 2.9 0.92 0 temperature was at least 50C lower than the mean monthly minimum for that month (use a 30 year average for each month as reference, and average the results over all reference climate stations). 7 REI G Number of volcanos (with potential for eruption) having a Volcano 2.9 1.03 0 Explosivity Index equal or greater than 4. The result to be expressed as a ratio of land area. (Volcanos situated 100 km of country land boundary). 8 REI G Earthquake energy with ML >=6.0 and <=15km depth per 5 years. (The 3.1 1.00 0 earthquake energy occurring 100 km of country land boundaries). The result to be expressed as a ratio of land area. 9 REI G Number of tsunamis or storm surges per 100 km of coastline, with run- 2.9 0.95 0 up greater than 2m above MHWS since 1900. 10 IRI C Total land area (km2) 3.4 0.93 0 11 IRI C Length of ocean shoreline (km) expressed as a ratio of total land area 3.1 1.17 0 12 IRI C Distance to nearest continent (km).18 2.0 0.88 1 13 IRI C Altitude range measured as the difference between the highest point and 2.3 0.99 0 the lowest point in country. 14 IRI C Percent of land area which is lower than 10 meters above sea-level. 3.2 0.89 0 15 IRI C Percentage of land area, composed of unconsolidated sediments, which is 2.8 0.80 0 lower than 10 meters elevation, situated within 2 km of the coast (exclude coral reefs and ice).19 16 IRI C Number of known endemic species per 10,000 km2 land area. 2.9 1.38 1 17 REI B Number of reported (and verified by appropriate authorities) organism 2.7 0.75 0 outbreaks over the last 5 years (pathogens, blooms, plagues) expressed as a ratio of land area.20 18 REI B Total tonnage of freight imported annually during the last five years, 2.2 0.97 1 expressed as a ratio of land area. 19 EDI B Number of all species introduced in the country since 1900, per 10,000 2.3 0.61 0 km2 land area. 20 EDI B Number of endangered and threatened species per 10,000 km2 of land 3.3 0.73 0 area (use IUCN definitions).

12 The EVI indicators were discussed, one by one, by the participants in the Malta meeting, and a number of changes were suggested, as detailed in Section 5.2 of the main report. The list which appears here was approved by participants. It should be noted that some of the indicators require further refinement. This task was left for completion by the SOPAC team of consultants. 13 Perceived importance by 15 experts present at the Malta Meeting, who were asked to rank the 47 indicator in terms of the importance, on a scale ranging from 0 to 4, with zero indicating no importance (to be discarded) and 4 indicating the highest importance. 14 This represents the mean of the scores assigned by the Malta meeting experts. In general the Maltese experts assigned higher scores to the indicators than the Think Tank experts (the mean difference being 22% higher). 15 Standard Deviation associated with the mean. 16 f(0) refers to the number of times that a 0 was assigned to the indicator. 17 This indicator requires further work. 18 Addition work is need regarding the definition of continent. 19 This indicator may be somewhat difficult to measure. 20 The terms verified needs further definition.

Appendix 3: 4 APPENDIX 3: LIST OF INDICATORS

Importance Assigned13 Indicator Sub-Index Category Indicator Mean SD15 f(0)16 Number Score14

21 EDI B Number species which have become extinct since 1900 per 10,000 km2 2.9 0.86 0 land area (use IUCN definitions). 22 EDI B Percentage of natural and regrowth vegetation remaining (such as: 3.5 0.65 0 forests, mangroves, saltmarshes, prairies, savannah, desert, tundra). 23 EDI B Tonnage of intensively-farmed animal products over the last five years 2.5 0.94 0 expressed as a ratio of land area (includes acquaculture, pigs, chickens). 24 EDI B Percent of fisheries stocks overfished (use FAO definitions).21 3.1 0.73 0 25 EDI A Density of people living in coastal settlements expressed as a ratio of 3.4 0.63 0 land area.22 26 REI A Total human population density expressed as a ratio of land area. 3.5 0.65 0 27 REI A Annual human population growth rate, averaged over the last 5 years. 3.5 0.85 0 28 REI A Net percentage of land area changed by the removal of natural vegetation 3.2 0.80 0 over the last 5 years 29 REI A Annual number of international tourists, multiplied by the number of 2.9 0.73 0 nights stayed, expressed as a ratio of land area, over the last 5 years. 30 REI A Megalitres of untreated industrial and domestic wastewater discharged to 3.3 0.91 0 aquatic system expressed per 1,000 km aquatic ecosystems (length of coast plus length of rivers) over the last 5 years.23 31 REI A Total tonnage of generated and net imported toxic, hazardous and 3.4 0.84 0 municipal wastes expressed per 10,000 km2 land area, expressed as an annual average over the last 5 years. 32 REI A Percent of hazardous, toxic and municipal waste effectively managed or 3.4 0.84 0 treated over the last five years.24 33 REI A Number of spills of oil and hazardous substances greater than 1,000 3.0 1.11 1 litres during last 5 years on land, in rivers or within territorial waters, expressed as a ratio of land area. 34 REI A Number of nuclear, chemical and other major industrial facilities that could 3.4 1.01 0 cause significant damage, per 10,000 km2 land area and territorial waters.25 35 REI A Number of vehicles (as per World Bank definition) expressed as a 3.0 0.88 0 percentage of land area during the last year .

36 REI A Maximum 24 hour SO2 concentration (micro g /cubic m) annual average 2.8 0.90 0 over the last 5 years. 37 REI A Tonnes of N,P,K fertilizers used per 10,000 km2 of agricultural land area, 2.9 0.66 0 annual average over the last 5 years. 38 REI A Tonnes of pesticides used per 10,000 km2 of agricultural land area, 2.8 0.70 0 annual average over the last 5 years. 39 REI A Number of new fisheries stocks or expanded fisheries efforts (measured 2.4 0.85 0 as at least 20% increase in catches) added within the territorial area of the country over last 5 years.26 40 EDI A Percentage of land area (excluding urban areas) degraded since 1950 3.6 0.74 0 (includes salinisation and desertification).27 41 EDI A Annual internal renewable water resources (average annual runoff plus 3.2 0.97 0 recharge of groundwater from endogenous precipitation) per capita, over the last five years. 42 REI A Kilotonnes of mining material (ore and tailings) extracted per 10,000 2.5 0.97 1 km2 land area. Annual average over the last 5 years. 43 EDI A Percentage of land, rivers and offshore coastal zone (1 kilometre outside 2.5 0.65 0 shore) affected by mining and quarrying.

21 Requires further work. 22 This indicator requires further work to refine the denominator. 23 Requires further work to define length of rivers. 24 This indicator requires further work to define the concepts “managed” and “treated”. 25 This indicator requires further work to refine the denominator, which could be “territorial area”. 26 This indicator requires further refinement. 27 This indicator may be somewhat difficult to measure.

Appendix 3: 5 UNEP MEETING OF EXPERTS ON THE ENVIRONMENTAL VULNERABILITY INDEX

Importance Assigned13 Indicator Sub-Index Category Indicator Mean SD15 f(0)16 Number Score14

44 EDI A Percent of terrestrial area set aside as reserves. 3.0 0.78 0 45 EDI A Percent of marine zone set aside as reserves (mean high tide to 3.0 0.78 0 continental shelf) 46 EDI A Number of years of war or civil strife over the last 50 years within the 2.7 1.14 0 country territory.28 47 EDI A Percentage of population with access to safe sanitation (use WHO 2.7 0.63 0 definitions).29

28 This indicator requires further work to define the concepts of “war” and “civil strife” 29 This index requires further refinement.

Appendix 3: 6 APPENDIX 4

EXTENDING SOPAC’S ENVIRONMENTAL VULNERABILITY INDEX TO THE IMA AND CARIBBEAN REGIONS

Background paper prepared by

Lino Briguglio and Ursula L. Kaly Islands and Small States Ministry of Natural Resources & Institute, Environment, Funafuti, Tuvalu University of Malta assisted by Craig Pratt, SOPAC, Fiji

for the UNEP Meeting of Experts on the Environmental Vulnerability Index organised by the Islands and Small States Institute, Malta in collaboration with SOPAC, Fiji, held at the Foundation for International Studies, Valletta, Malta 29 November – 3 December 1999

Appendix 4: 1 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

CONTENTS

1. INTRODUCTION 4

1.1 Background 4

1.2 The SOPAC study 4

1.3 The present report 5

2. CURRENT STATUS OF THE EVI 6

2.1 Background: purpose and history 6

2.2 Approach and progress on the EVI 7

2.3 Current status 8

2.4 Extending the SOPAC EVI Model for the Caribbean and IMA Regions 10

2.5 Adapting the existing indicators and developing new indicators 10

2.6 Extending the EVI to six new states 11

2.7 Globalising the EVI 11

2.8 Testing the EVI 12

3. SIDS IN THE IMA AND CARIBBEAN REGIONS 13

3.1 The Economies of the IMA and Caribbean SIDS 13

3.2 Population and human development 18

3.3 Main physical features 19

4. A FOCUS ON SIX IMA AND CARIBBEAN SIDS 24

4.1 Introduction 24

4.2 Cape Verde 24 4.2.1 The economy of Cape Verde 24 4.2.2 Major environmental concerns in Cape Verde 25 4.2.3 Other environmental issues in Cape Verde 25 4.2.4 Cape Verde and the EVI 26

4.3 Jamaica 26 4.3.1 The economy of Jamaica 26 4.3.2 Major environmental concerns in Jamaica 27 4.3.3 Other areas of environmental concerns in Jamaica 27 4.3.4 Jamaica and the EVI 27

4.4 Malta 28 4.4.1 The economy of Malta 28 4.4.2 Major environmental issues in Malta 29 4.4.3 Other environmental concerns in Malta 30 4.4.4 Malta and the EVI 31

Appendix 4: 2 4.5 Mauritius 31 4.5.1 The economy of Mauritius 32 4.5.2 Major environmental issues in Mauritius 32 4.5.3 Other areas of environmental concern .in Mauritius 33 4.5.4 Mauritius and the EVI 34

4.6 Saint Lucia 34 4.6.1 The economy of Saint Lucia 34 4.6.2 Major environmental concerns in Saint Lucia 35 4.6.3 Other areas of environmental concern in St Lucia 35 4.6.4 Saint Lucia and the EVI 35

4.7 Trinidad and Tobago 36 4.7.1 The economy of Trinidad and Tobago 36 4.7.2 Areas of environmental concern in Trinidad and Tobago 37 4.7.3 Other environmental issues 37 4.7.4 Trinidad and Tobago and the EVI 38

5. CONCLUSION 39

REFERENCES AND DOCUMENTS CONSULTED 39

Appendix 4: 3 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

1. INTRODUCTION 1.1 Background

During the nineties, there has been a growing interest in Small Island Developing States (SIDS), with special attention being given to their economic and environmental vulnerabilities. Two main events took place in this connection, namely the 1994 Global Conference on the Sustainable Development of SIDS and the 1999 Special Session of the United Nations General Assembly. In the Barbados Programme of Action there is a call for the development of an index that expresses vulnerability30, while in the UNGASS final statement, this call was further elaborated.31 The benefits of producing a Vulnerability Index include that it can attract attention to the plight of vulnerable states and that it attempts to quantify the extent to which some states are more vulnerable than others. The index could also provide a useful tool for environmental authorities and managers of vulnerable states. Several studies have been produced on the economic vulnerability and, more recently, on environmental vulnerability.32

1.2 The SOPAC study

The SOPAC programme on the development of the Environment Vulnerability Index (EVI) started in August 1998, focussing on the Pacific SIDS. The study attempted to develop a logical framework and methods of calculating and index for environmental vulnerability (EVI) and to identify and collect data to calculate the index. The result of this first attempt are reproduced on the Internet site http://www.unep.ch/islands/devi.htm (see also Kaly et al 1999a). A total of 57 indicators of environmental vulnerability were finally selected for inclusion in the index. This included 39 indicators of risk (REI), 5 indicators of resilience (IRI) and 13 indicators of environmental integrity or degradation (EDI). Many of the indicators were expressed as a ratio of area of land or coast rather than simply absolute numbers because it is risk density or proportion of area degraded that is of interest from an environmental perspective.

On the basis of the first attempt at constructing the EVI, it was concluded that it was possible to produce single-figure measures of environmental vulnerability across three countries33, although it was stressed that the results were tentative and preliminary because there was insufficient time to collect all of the data required for these three countries and there were some problems with reliability of the data.

Following this first attempt at quantifying the index, SOPAC organized a Think Tank in September 1999 with the aim of obtaining a peer-review and commentary from experts in a range of fields relevant to the development of the EVI and to render the EVI acceptable and/or operational in the international community (see Kaly et al, 1999b). The meeting also attempted to set criteria that would need to be met to ensure that the EVI will be internationally applicable to all regions of the world and to identify directions for future work. Following the Think Tank meeting, the number of indicators was reduced to 47 and the context of the index was shifted to global.

30 Article 113 of the Programme of Action states: “Small island developing States, in cooperation with national, regional and international organizations and research centres, should continue work on the development of vulnerability indices and other indicators that reflect the status of small island developing States and integrate ecological fragility and economic vulnerability. Consideration should be given to how such an index, as well as relevant studies undertaken on small island developing States by other international institutions, might be used in addition to other statistical measures as quantitative indicators of fragility”. 31 Three paragraphs (Section F, para. 39 to 41) dealt with the Vulnerability Index in the document adopted by the Special Session of the UN General Assembly (September 1999), wherein the need to compute a vulnerability index was again stressed. 32 A list of studies on the Vulneranbility Index is given in the references section. See also section 2.1 of this report. 33 The three countries were Tuvalu, Fiji and Australia.

Appendix 4: 4 1.3 The present report

This present report is in the context of a project initiated and funded by UNEP to extend the SOPAC’s EVI and assess the applicability of the model for the Caribbean, Indian Ocean, Mediterranean and Atlantic regions. It also puts forward preliminary recommendations as to how these regions can be involved in the process of further developing the EVI. One of the main activities of the project is an expert workshop in Malta, which the Islands and Small States Institute of the University of Malta was entrusted to organize. The Island and Small States institute was also assigned the task of producing this report, on the basis of inputs by Dr Ursula Kaly and Prof Lino Briguglio, assisted by Mr Craig Pratt.

This report is organised as follows. Section 2, which follows this introduction gives the background of the Environmental Vulnerability Index and the work done so far in the development of this index. Second 3, contains an overview of the SIDS in the IMA34 and Caribbean regions, since the focus of this study is on these SIDS. Section 4 zooms in on six SIDS in these regions, namely Cape Verde, Jamaica, Malta, Mauritus, St Lucia and Trinidad and Tobago. These small states were chosen on an experimental basis to extend the SOPAC experiment to other regions. Section 4 also discusses possible additional indicators which may be required to take into account the special circumstances in the six SIDS in the IMA and Pacific regions. Section 5 concludes the study.

34 IMA stands for Indian Ocean, Mediterranean and Atlantic Ocean

Appendix 4: 5 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

2. CURRENT STATUS OF THE EVI

This section describes the current status of the development of the Environment Vulnerability Index. This Index, which has been developed to simplify what would otherwise be a long and complex process of assessment, produces a single-figure to express the relative environmental vulnerability of entire country. This expression of vulnerability could be used as part of the argument for the ranking of countries and for assigning some special status to countries which are shown to be more vulnerable than others. It could also be used to provide estimates of the resilience of countries to particular events of concern, such as natural hazards, geological events, and particular kinds of hazards such as a volcanic eruption.

For a particular country, the EVI could be used to identify areas of weakness to particular risks and of high environmental degradation, giving the governments of that country useful information for environmental management. 2.1 Background: purpose and history

A range of indices have been developed over the past decade which attempt to describe, in a single expression, the vulnerabilities of different aspects of states to a range of internal and external stressors. The central purpose of constructing an EVI is to simplify the task of categorising countries according to their relative environmental vulnerabilities. If personnel, funding and time were unlimited, this could be done by sending several independent teams of scientists to each country and commissioning them to carry out a vulnerability assessment for each. The replicate assessments for each country could then be used to classify countries in terms of their vulnerability. This exercise would of course be extremely expensive and it is one of the purposes of the EVI to simplify this process.

Studies of vulnerability are necessarily concerned with two groups of variables. These are the "risks" which may cause unacceptable change to human and natural systems (e.g. cyclones, anthropogenic impacts), and "responders", those human and natural systems affected by the risks (e.g. economies, ecosystems).

Some studies on vulnerability indices base the index in relation to the risks involved and not the responders to those risks. For example, Pantin's (1997) Ecological Vulnerability Index was concerned with effects of natural disasters on human systems. This index was therefore more of a "Human Vulnerability Index" since after all, it is the human systems which were deemed vulnerable, not the natural disasters themselves. The EVI, as proposed by SOPAC deals with the impact on the environment, on not on humans per se.

Attempts have been made to provide indices for economic systems (Briguglio, 1992, 1993, 1995, 1997; Wells, 1996, 1997; Atkins et al. 1998; Chander, 1996; Crowards, 1998), climate change and sea-level rise (IPCC, 1991, 1992; Pernetta, 1990; Downing, 1992; Formel, 1996), ENSO phenomena (NOAA, 1997), human impacts on the environment (Ehrlich & Ehrlich, 1991, UNEP, 1998; Eurostat, 1998), and the effect of natural disasters on human systems (Albala-Bertrand, 1993; Pantin, 1997).

Prior to the development of the SOPAC EVI, the number of different vulnerability indices developed to examine the relative vulnerabilities of states in terms of these risks and responders totaled at least fifteen (see Kaly et al. 1999a). The risks of concern varied among studies. Anthropogenic risks were considered in 11 of the fifteen studies, and 6 studies focused on climate change and sea-level rise and 6 studies considered natural disasters. In terms of responders, thirteen of the indices were concerned with risks to human economic and social systems, while only 5 attempted to include effects on the natural environment. Only 1 study specifically examined the effects of both humans and natural hazards on the environment.

Appendix 4: 6 The underlying argument in SOPAC’s EVI is that the "[physical] environment" is the basis of all human activities. Poor environmental conditions mean fewer or poorer quality natural resources and poorer ecosystem services (such as attenuation of wastes and pollution). Concern for the quality of the [physical] environment is therefore well-founded because neither economic systems, quality of human life, nor human development will thrive without it. It is the object of this EVI to shift the focus of vulnerability indices off human systems and onto the foundations on which they are built. The SOPAC EVI is a true environmental vulnerability index, based on a wide array of environmental indicators which examines both natural and anthropogenic risks to environmental health (Kaly et al. 1999a,b).

2.2 Approach and progress on the EVI

The SOPAC EVI (Kaly et al. 1999a,b) was developed in response to a call made in the Barbados Programme of Action, and an increasing awareness that small island developing states face disadvantages to their development associated with their remoteness, small size, dispersion, economic conditions and limited natural resources. It was initially developed with a focus on the Small Island Developing States (SIDS) of the Pacific Region (Kaly et al., 1999a) and later peer reviewed and expanded to more global relevance during a Think Tank which utilised relevant experts from around the world (Kaly et al., 1999b).

The EVI was constructed on a theoretical framework that identified two main aspects of vulnerability, risks and resilience, that could be further broken down into a total of three distinct aspects: · risks or hazards to the environment (natural and anthropogenic); · intrinsic resilience; and · extrinsic resilience.

Intrinsic resilience is the innate ability of the environment to cope with hazards (expressed as good recovery rates, high productivity or natural resistance to damage) and extrinsic resilience is related to ecosystem health (the amount of degradation of the environment as a result of past impacts determines the ability of ecosystems to cope with future stresses).

These three aspects correspond to three sub-indices, · the REI or Risk Exposure sub-Index, and · the IRI Intrinsic Resilience sub-Index; and · the EDI ( Environmental Degradation sub-Index).

The IRI and the EDI together form the RI, or Resilience sub-Index.

A total of 47 indicators of environmental risk and resilience was assembled, 27 for the REI, 7 for the IRI and 13 for the EDI (Kaly et al., 1999b). These indicators are to be evaluated for each country and scored in the range of 1-7 in order to allow for heterogeneous data on different scales to be accumulated on a single linear scale (thus transforming them to permit summation of the different components of the index).

Each sub-index (or component) for a particular country is to be calculated as an average of scores obtained and the final EVI calculated as an average across all indicators, regardless of the sub-index to which they belong. Although weighting schemes were applied to the EVI, a final system for weighting the indicators awaits the results of more rigorous mathematical testing. It is intended that data would be collected every five years to provide an update on the vulnerability status of each country.

To ensure that the final EVI values calculated are able to identify differences in vulnerability among countries and provide an unbiased assessment for each, a series of criteria were used during their selection (Table 1).

Appendix 4: 7 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

Table 1 Criteria for the selection of indicators for the EVI

1 The data should be available or easily obtainable 2 The indicators should measure change or be a proxy for change which would do significant harm to the environment 3 A particular indicator could weighted to reflect the probability of change to the environment and the amount of damage which might be done 4 The indicators should be relatively easy for users to understand 5 Be as uncorrelated with other indicators as possible to limit redundancy 6 Be suitable for international comparisons across countries

2.3 Current status

The EVI developed by SOPAC has come a long way but requires further refinement. The progress described above constituted Phase I and some of Phase II of the project, and was designed to determine whether it was possible to produce an EVI, and if so, expose it to peer review at the technical level.

A preliminary test of the EVI based on the initial model developed by Kaly et al., (1999a) was promising and produced single figure indices for Australia, Fiji and Tuvalu that highlighted known differences in the levels of environmental degradation. That initial model and indicators have now been superseded (see Kaly et al., 1999b). A new list of proposed indicators, shown in Table 2, is awaiting complete definition, setting of the scoring scale and testing for redundancies, a process which will require data from a range of test countries. The work (and funding) remaining for Phase II is to collect data for at least 5 Pacific states for initial testing. Full testing will require a larger sample of countries from around the globe.

The methodology selected in the computation of the index appears to produce results which are expected to have operational usefulness for identifying the environmental vulnerabilities of countries. Data for the new SOPAC EVI have already been collected for Tuvalu with relative ease (one person- week of work in-country). There now remains the large task of fully globalising the model and its indicators before final testing can be completed.

Table 2 List of the indicators selected by the Think Tank group for inclusion in the EVI

Indicator Sub- Category Indicator Weight Number Index Proposed 1 REI Met Greatest average annual deviation in Surface Sea Temperature in M last 5 years from long term mean (30 years) (more work required to finalise form) (Centralised database) 2 REI Met % of reference climatological stations experiencing >= 1/100 yr 3 M second wind gusts in a 5 year period / land area 3 REI Met % of reference climatological stations experiencing <= 1/100 yr M minimum annual rainfall in a 5 year period / land area 4 REI Met Cum # of 24 hr periods over all Reference Climate Stations over M last 5 years during which rainfall is >= 1/100 yr (1/50, 1/20 ?? Less?) event / stations 5 REI Met % of reference climatological stations experiencing >= 1/100n yr M daily max temp in a 5 year period 6 REI Met % of reference climatological stations experiencing <= 1/100 M daily min temp in a 5 year period / land area 7 REI G Number of volcanos with potential for eruption >= VEI 4 M (Volcano explosivity Index) within 100km of country land boundary / area of land 8 REI G Earthquake energy within 100km of country land boundaries / L land area with ML >=6.0 and <=15km depth per 5 years

Appendix 4: 8 9 REI G Number of tsunamis or storm surges with run-up >2m above L MHWS / 100km coastline since 1900 10 IRI CC Total land area (sq km) H 11 IRI CC Ratio of length of ocean shoreline : total land area H 12 IRI CC Distance to nearest continent (km) L 13 IRI CC Altitude range (Highest point - lowest point in country) M 14 IRI CC Percent of land area <10m above sealevel H 15 IRI CC Land area below 10m elevation with unimpeded access to the M coast composed of unconsolidated sediments (excluding coral reefs and ice) / land area (%) 16 IRI CC Number of known endemic species / 10,000 sq km land area H 17 REI B Number of catastrophic organism outbreaks over the last 5 years / M land area (pathogens, blooms, plagues etc) 18 REI B Total tonnage of freight imported / year M 19 EDI B Number of all introduced species / 10,000 sq km land area since M 1900 20 EDI B Number of endangered & threatened species / 10,000 sq km of H land area (IUCN definitions) 21 EDI B Number species which have become extinct since 1900 / 10,000 H sq km land area (IUCN definitions) 22 EDI B Percentage of natural & regrowth vegetation remaining (e.g. H forests, mangroves, saltmarshes, prairies, savannah, desert, tundra) 23 EDI B Tonnage of intensively-farmed animal products / yr / land area L 24 EDI B Percent of fisheries stocks overfished (FAO) H 25 EDI A Density of people living in coastal settlements (define area) H 26 REI A Total human population density (per sq km land area) H 27 REI A Annual human population growth rate (average over last 5 years) H 28 REI A Net percentage of land area changed by the removal of natural H vegetation over last 5 years 29 REI A Annual number of international tourists * average days stay / 365 M / 100 sq km (last 5 years) 30 REI A Megalitres of untreated industrial and domestic wastewater H discharged to aquatic system / 1,000 km aquatic ecosystems (length coast+length rivers) 31 REI A Total tonnage of generated and net imported toxic, hazardous & M municipal wastes/ 10,000 sq km land area / year (avaerge last 10 years) 32 REI A Mean percent of industrial and municipal waste managed or L treated / yr 33 REI A Number spills of oil and hazardous substances >1,000 litres L during last 5 years within territorial waters 34 REI A Number of nuclear, chemical and other major industrial facilities L that could cause significant damage / 10,000 sq km land area 35 REI A Number of cars / land area M 36 REI A Max 24 hour SO2 concentration (micro g /cubic m) (average over M last 5 years) 37 REI A Tonnes of N,P,K fertilizers used / 10,000 sq km agricultural land L area / year (average last 5 yrs) 38 REI A Tonnes of pesticides used / 10,000 sq km of agricultural land / L year (average last 5 years) 39 REI A Number of new fisheries species added to country over last 5 L years (within territory) 40 EDI A % Land area degraded since 1950 H 41 EDI A Question on importance of water (to be developed) H 42 REI A Kilotonnes of mining material (ore + tailings) extracted / 10,000 L sq km land area / year (average last 5 years) 43 EDI A Land and sea area affected by mining & quarrying / land area L 44 EDI A Percent of terrestrial zone set aside as reserves M 45 EDI A Percent of marine zone set aside as reserves (mean high tide to L continental shelf)

Appendix 4: 9 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

46 EDI A Number of war or civil strife years over the last 50 years within M the territory 47 REI A Environmentally related legislation with regulations L The indicators are classified into 5 categories (Met = Meteorological, G = Geological, B = Biological, CC = Country Characteristics and A = Anthropogenic).. The final list of indicators were categorised by the Think Tank participants in terms of their perceived importance so that preliminary weightings could be assigned. Participants were asked to score each indicator with a value of between 0 and 4, where 0 showed the lowest level of importance (to be discarded) and 4 the highest. The results of this weighting exercise are shown in the last column of the table; as suggested weighting (L = low, M = Medium, H = High). Data were collected from 21 Think Tank participants.

2.4 Extending the SOPAC EVI Model for the Caribbean and IMA Regions

To extend the EVI to the Caribbean and IMA small island states, it will be necessary to adapt the indicators, their scoring scales and weights. The mechanics of calculating the EVI (framework or model) and other inherent features of the index will remain the same, regardless of the region of the world for which they are applied. The main considerations will be to determine what the particular environmental concerns relevant for the IMA and Caribbean SIDS are, to adapt the existing indicators and/or develop new indicators which cater to these needs, collect the data necessary for setting the new levels of all indicators and develop a plan of action for making data collection part of the normal reporting of each country.

The first step in adapting the EVI to include and be relevant to Caribbean and IMA regions is to select a small sample of countries from which to draw relevant indicators. It will be necessary to preface this investigation with a consideration of the explicit requirements of these countries for an environmental vulnerability index. The SOPAC EVI is specifically designed for natural environments and takes a snap-shot of the likelihood of damage from hazards, and the ability of that country as a whole to tolerate those hazards. It is envisaged that the same would be required by the other regions. To change this approach would be to eliminate the possibility of comparisons among states or regions.

A consideration of the special environmental concerns for Caribbean and IMA SIDS would require identification and discussion on all of the main natural and anthropogenic risks occurring in a range of countries from those regions. A similar process would need to occur for the questions of intrinsic resilience, expressed as permanent characteristics of a country that predispose it to environmental vulnerability and extrinsic resilience, expressed as environmental degradation. Many of the concerns will already have been incorporated into the SOPAC EVI and will not require new indicators. It is expected, however, that some new indicators will be required.

2.5 Adapting the existing indicators and developing new indicators

To explore the possibilities of adapting the existing indicators to regions other than the Pacific, SOPAC has agreed to collaborate with the Islands and Small States Institute in the convening of a UNEP workshop in Malta, between 29 November and 3 December 1999. The workshop will examine the possibility of obtaining data for six SIDS in the IMA and Caribbean regions, which are Cape Verde, Jamaica, Malta, Mauritius, St Lucia and Trinidad and Tobago. This will be a pilot experiment which could eventually be extended to include other states, in preparation for the globalisation of the index.

Once the special circumstances of these six SIDS have been fully identified, and if the existing indicators do not reflect these special circumstances it will be necessary to incorporate new indicators in the EVI.

The most important requisite will be to revise and extend the mapping of heterogeneous data onto the 1-7 scale in such a way that SIDS from the three main regions (Pacific, IMA and Caribbean) may be completely incorporated. This process is complex because for each indicator it will be necessary not only to determine the upper and lower limits of values, but also whether the mapping should be on a linear, non-linear, discontinuous, truncated or on a yes/no scale. If the scaling of an indicator is to be non-linear, it would then be necessary to determine what the assumed shape of the underlying

Appendix 4: 10 function might be. Although this process has been done for a sample of Pacific countries for many of the SOPAC indicators, it is likely to require reevaluation to take on board countries in the IMA and Caribbean regions. This process may take some time before it is completed.

In the case where new indicators would be required, these will need to be designed so that they meet the criteria in Table 1 above. New indicators would be incorporated on the basis of a need to take into account a signal which is likely to be important to the Caribbean and IMA SIDS and which is not included in any of the other indicators.

2.6 Extending the EVI to six new states

The objective of the Malta meeting is to attempt to apply the EVI, on an experimental basis, for six SIDS situated in the Caribbean and IMA regions. Testing of the EVI requires the collection of data from a range of countries worldwide.

In addition the Malta meeting will serve to introduce the concepts and gain support for the SOPAC EVI in the Caribbean and IMA SIDS and to consult with experts from the Caribbean and IMA regions in identifying their needs and the functions and benefits they expect from the EVI.

Another objective is to determine what the special considerations for the Caribbean and IMA SIDS are in relation to the EVI and attempt to extend the indicators of the EVI model, as developed by SOPAC for the Pacific region, to the Caribbean and IMA SIDS as part of the process of globalising the index.

One of the major inputs of the meeting is to put forward recommendations as to how the IMA and Caribbean regions can be involved, together with the Pacific region, in the process of further developing the EVI and to put forward recommendations on how data collection mechanisms might be set up within Caribbean and IMA countries, and in relation to global data-collection mechanisms.

A long term objective is to identify the process and devise an approach for obtaining the funding needed to complete the global EVI.

Collecting the data is outside of the scope of this meeting, although an attempt will be made to assess the difficulties of procuring suitable information relating countries in these regions. The meeting will however try to identify the probable sources of data for deriving indicators for countries in these regions.

The Malta meeting will explore how best to approach the task of presenting the indicators to the countries. This includes practical considerations of which departments to approach, requirements for capacity-building, translation of the documentation and identification of outside sources of data. It will also be necessary to finds means to incorporate the collection of data for the EVI into normal mechanisms in each country in anticipation of updating the EVI every five years.

2.7 Globalising the EVI

Globalisation of the EVI will require expanding the above processes described for the Caribbean and IMA SIDS to countries with widely differing characteristics across the world. This is a longer term goal of the EVI project and will require the cooperation of a wide variety of countries ranging from the large industrialised countries to the small developing ones, from those occupying large continental masses to those which are, highly-fragmented; and from those situated in tropical regions to others exposed to cold climates. This would ensure that the full range of environmental hazards and conditions across the globe are incorporated into the EVI. The most important aspect of this work will, once again, be the scoring onto the 1-7 scale and weighting of indicators.

Appendix 4: 11 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

2.8 Testing the EVI

Testing of the EVI can begin once the data have been collected for a sufficiently large number of countries which are considered to be a sufficiently good sample of all countries of the world. It is envisaged that about 15 countries will be sufficient in this regard.

Three tests will be performed for the purposes of determining when the EVI would be sufficiently statistically sound for operation. These are summarised in Table 3.

Table 3 Testing required to complete the globalised EVI

Test 1: When redundant indicators are eliminated. This will be tested using standard statistical tests. Test 2: When the spread in EVI values among the 15 test countries occupies much of the 1-7 range expected and countries considered a priori to be ‘similar’ cluster closer together than ‘dissimilar’ countries. Test 3: When the difference between the value obtained by the EVI and the mean of the assessment provided for a country by several independent experts is about the same or less than the spread for a country found among the experts. (this should be done for around 5 countries).

Redundancies. The first test is of redundancy among indicators. A correlation matrix would be produced using the data for each indicator and country and assessed using a suitable statistical procedures. Any indicator with a high correlation with one or more other indicators would at this stage be dropped or merged. The final list of indicators would then only consist of those which bring significantly new information into the EVI value. The final weighting of indicators can only occur after redundancies in the model have been limited.

Spread. The second test is an analysis of the ability of the EVI model to distinguish between and provide spread among countries. This means that the final EVI scores would need to be evaluated for the test 15 countries. The EVI should be able to cluster similar countries together and provide spread among countries which are considered a priori to be very different. The 1-7 mapping scales for each of the indicators (which is the mechanism in the EVI which provides the spread) can be finalised when these data are available.

Validation. The purpose of constructing an EVI is to simplify the task of categorising countries according to their relative environmental vulnerabilities. The only independent means of assessing the effectiveness of the EVI in carrying out this task in a simplified way, is to compare the results of the EVI with a full assessment for a small number of, say 5, countries. Several teams of experts would have to be mobilised in each of the test countries to provide a 'mean assessment' for each. The consultants involved should be independent and must not use the EVI approach to ensure that they do not unintentionally bring bias into the results. These assessments could then be compared with the EVI scores obtained.

Appendix 4: 12 3. SIDS IN THE IMA AND CARIBBEAN REGIONS

This section will give an overview of the SIDS in the IMA (Indian Ocean / Mediterranean / Atlantic)35 and Caribbean36 regions. As explained in section 2 of this report, it is envisaged that the EVI will be extended to cover these regions, and new indicators may be developed to cater for the special conditions in these SIDS. 3.1 The Economies of the IMA and Caribbean SIDS

The economies of the SIDS considered in this study varied according to various factors, including geographical location of the state and the structure of its economy. Summarised data in this regard is presented in Tables 4 to 7.

Table 4 shows that the highest GDP per capita were registered in Singapore, Bahamas, Cyprus, Malta and Bahrain, four of which are in the IMA-region. However, the countries with the lowest GDP per capita are also IMA-SIDS, namely Guinea Bissau, Comoros, Madagascar and Sao Tome/Principe. The Caribbean SIDS exhibited more homogeneity in their GNP per capita with Haiti and Guyana occupying the lowest levels. A closer look at the structure of the economies of the SIDS in question would show that those SIDS with the highest dependency on the agriculture sector tended to register the lowest GDP per capita. Also the countries which are proximate to main centres of Commerce (Singapore. Malta, Cyprus, and most Caribbean SIDS) tended to register the highest GDP per capita.

Table 5, which deals with the structure of merchandise structure, shows that most SIDS in the regions under consideration, with the exception of Bahrain and Trinidad and Tobago, tend to depend on exports of primary products or light industry . The table also shows that most SIDS depend on a very narrow range of exports.

A common characteristic of small island states is heavy dependence on tourism for income generation and foreign exchange inflows. Table 6 shows that many of the IMA and Caribbean SIDS fall into this category, with the Maldives, Antigua/Barbuda, St Lucia, Bahamas, Barbados, St Kitts/Nevis, Grenada, Jamaica, St Vincent, Malta, Cyprus and Seychelles registering the highest dependence in this regard.

A common feature of SIDS relates to their economic openness, which coupled with their high export concentration, render these states as very vulnerable to economic forces outside their control. Table 7 shows that most SIDS in the regions under consideration have very high dependence on imports and exports. Almost all of them are very much dependent on energy imports, and experience a negative resource balance (the difference between exports and imports of goods and services).

35 The IMA- SIDS group, as considered in this report, contains one country Guinea-Bissau, which is not an island. This country is included with this group because it forms part of AOSIS (the alliance of small island states), and because it is a coastal state, sharing many of the coastal problems faced by island states. The group also contains two islands which cannot be called small, namely Madagascar and Sri Lanka, but which were considered to share many features of small island states. 36 Again here the country grouping includes 3 states which are not islands, namely Guyana, Suriname and Belize, which are members of AOSIS and share many features of islands. The group also includes two island states which are not very small, namely Cuba and Dominican Republic which have often associated themselves with small island states in the past

Appendix 4: 13 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

Table 4 The GDP of SIDS in the IMA and Caribbean Regions

Sectoral Composition of GDP (%) 1994 GDP GDP per capita Agriculture Industrial Production Private and (US$) and Fishing Total Manufacturing Government (Million 1995 Services US$ 1995) IMA SIDS Bahrain 5060 9085 1 43 21 56 Cape Verde 324 862 13 15 5 72 Comoros 227 271 38 13 6 49 Cyprus 8788 11797 5 24 13 71 Guinea Bissau 257 240 45 19 7 37 Madagascar 3160 312 39 13 12 48 Maldives 271 1065 na na na na Malta 3205 8744 3 39 27 58 Mauritius 3290 2511 9 33 23 58 Sao Tome 46 342 na na na na Seychelles 474 6493 4 17 na 79 Singapore 83700 25156 - 36 27 64 Sri Lanka 12900 720 24 25 15 52

Caribbean SIDS Antigua and Barbuda 493 7585 4 17 2 79 Bahamas 3460 12409 na na na na Barbados 1740 6678 5 16 7 79 Belize 587 2753 22 27 16 29 Cuba 17827 1638 na na na na Dominica 227 3197 21 21 8 59 Dominican Rep. 11810 1510 15 22 15 64 Grenada 276 3000 11 19 6 70 Guyana 595 717 36 37 11 27 Haiti 2040 286 44 12 9 44 Jamaica 4410 1787 9 38 18 53 St Kitts and Nevis 225 5488 6 25 11 69 St Lucia 556 3915 11 21 7 68 St Vincent & Grenadines 256 2286 11 22 na 67 Trinidad & Tobago 5330 4141 2 44 9 5 Source: UNCTAD. Tables 6.1 and 6.3

Appendix 4: 14 Table 5 Export Structure (1994-95)

Major Second Third major 10 major Number Concen- Exported Major Item Export of tration Item Item items Exported As % of Items* Index** Total Exports IMA SIDS Bahrain Petroleum Aluminium Gas 93% 117 0.629 Cape Verde na na na na na na Comoros na na na na na na Cyprus Fruit/veg Garments Pharm. Prdcts 65% 96 0.228 Guinea Bissau na na na na na na Madagascar Coffee Spices Shellfish 76% 58 0.284 Maldives Fish Garments Animal feed 93% 19 0.356 Malta Electronic comp. Clothing Ships & boats 79% 130 0.499 Mauritius Sugar Garments Fish 85% 111 0.313 Sao Tome na na na na na na Seychelles Fish Spices Animal feed 98% 11 0.677 Singapore Data processing Electronic comp. Petroleum 61% 229 0.211 Sri Lanka Garments Tea Prec. Stones 69% 152 0.218

Caribbean SIDS Antigua and Barbuda na na na na na na Bahamas na na na na na na Barbados Sugar Switchgear Paper 65% 56 0.190 Belize Fruit Sugar Garments 98% 14 0.572 Cuba Sugar Metal Ores Pharm. Prdcts 92% na 0.461 Dominica Fruit Soap Pigments 94% 48 0.426 Dominican Rep. Garments Pig iron Med. Instrmnts 62% 108 0.209 Grenada Spices Fruit Cocoa 83% 26 0.326 Guyana Sugar Gold Metal ores 73% 53 0.350 Haiti Garments Coffee Oils/perfumes 73% 38 0.223 Jamaica Metal ores Garments Sugar 89% 112 0.502 St Kitts and Nevis na na na na na na St Lucia Fruit Garments Paper 93% 32 0.595 St Vincent & Grenadines na na na na na na Trinidad and Tobago Petroleum Oxides etc Gas 88% 127 0.361 Source UNCTAD. Table 4.3 and 4.5 Note: * Number of exported items at the 3 digit SITC and having a value greater than $100,000 in 1995 or more than 0.3% of the countries exports. ** The UNCTAD Export concentration takes a value of between 0 and 1, where 1 is maximum concentration. In general the concentration index for small states tends to be higher than that of large states.

Appendix 4: 15 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

Table 6 Tourism Inflows and Incomes

Revenue Number of Revenue Tourism from Tourists from Revenue as Tourism. (thousand) Tourism as % of (Million % of GDP Exports of US$, 1995) Goods and Services IMA SIDS Bahrain 25 1950 5% 5% Cape Verde 10 28 3% 12% Comoros 21 23 10% 46% Cyprus 1797 2100 21% 43% Guinea Bissau - - - - Madagascar 58 75 2% 8% Maldives 210 315 78% 66% Malta 659 1116 20% 22% Mauritius 430 422 11% 18% Sao Tome (1990) 2 2 5% na Seychelles 129 121 21% 40% Singapore 8337 6422 10% 6% Sri Lanka 225 403 2% 5%

Caribbean SIDS Antigua and Barbuda 247 212 50% 99% Bahamas 1346 1598 39% 98% Barbados 662 442 38% 58% Belize 77 131 13% 26% Cuba 977 742 5% na Dominica 34 60 15% 33% Dominican Rep. 186 1776 2% 3% Grenada 58 108 21% 49% Guyana 47 106 8% 10% Haiti 56 145 3% 22% Jamaica 1069 1147 24% 36% St Kitts and Nevis 65 79 29% 58% St Lucia 268 231 48% 70% St Vincent & Grenadines 53 60 21% 42% Trinidad and Tobago 73 260 2% 3%

Appendix 4: 16 Table 7 Other Economic Features

Trade Energy Resource Openness Depen- Balance (% (%) dence (%) of exports) IMA SIDS Bahrain 102 57 14 Cape Verde na 100 -215 Comoros na 100 -147 Cyprus 51 100 -11 Guinea Bissau 29 100 -330 Madagascar 23 93 -50 Maldives 59 100 -334 Malta 97 100 -12 Mauritius 64 99 -9 Sao Tome na 97 -337 Seychelles 59 100 -14 Singapore over 100 100 na Sri Lanka 37 88 -30

Caribbean SIDS Antigua & Barbuda na 100 na Bahamas 51 100 0 Barbados 49 83 3 Belize 60 100 -12 Cuba na na na Dominica 59 94 -49 Dominican Rep 27 96 -21 Grenada 51 100 -40 Guyana 91 99 -22 Haiti 14 94 -81 Jamaica 60 99 -8 St Kitts & Nevis 70 100 -39 St Lucia na 100 -16 St Vincent & Grenadines 63 95 -55 Trinidad & Tobago 38 15 28

· Trade openness = 5(exp+imp)/GDP (average for 1990-1995) · Energy dependence = energy imports /[imports+production] (average for 1990-1995) · Number of exported items = number of items with a value of over 0.3% of total exports (1994) · Concentration index = concentration index as computed by UNCTAD (1994) · Resource balance = Exports of goods and services less imports of goods and services as percent of exports (average for 1990-1995) · Debt ratio = long term debt expressed as a ratio of GDP (1994).

Appendix 4: 17 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

3.2 Population and human development

Population and human development indicators for the SIDS considered in this study are shown in Table 8. There is considerable variation in the IMA-SIDS in terms of population growth, population density, with the low income and those with the lowest Human Development Index countries tending to experience the highest population growth rates and the lowest life expectancy.

The population density of the IMA-SIDS, as expected, was highest in the small islands (with the exception of Singapore). The highest densities (after Singapore) occurred in Malta, Maldives and Bahrain. In the Caribbean SIDS, population growth tended to be slower than in the IMA-SIDS. The population densities also tended to be lower than in the IMA-SIDS.

As expected, the SIDS with the highest GDP per capita tended to be those with the highest scores on the Human Development Index.

There was considerable variation in the Human Development Index of the IMA SIDS, with the highest being those of Singapore, Cyprus and Malta.

The Human Development Index of the Caribbean SIDS exhibited less variation than those in the IMA-SIDS, again confirming that the Caribbean SIDS are more homogenous. The states that depart significantly from the others in terms of the Human Development Index in the Caribbean region are Haiti and Guyana.

Appendix 4: 18 Table 8 Demographic Features (1995)

Total Population Population Human Life Population growth Density Develop- Expect- (‘000) (1990-95) per ment ancy at Km sqr Index birth IMA-SIDS Bahrain 567 2.8 803 .870 72 Cape Verde 337 2.6 96 .547 65 Comoros 612 3.2 275 .412 56 Cyprus 745 1.8 81 .907 77 Guinea Bissau 1069 2.1 30 .291 43 Madagascar 14874 3.3 25 .350 57 Maldives 254 3.4 848 .611 63 Malta 367 0.7 1146 .887 76 Mauritius 1117 1.1 547 .831 71 Sao Tome 133 2.2 139 .534 67 Seychelles 73 0.8 162 .845 72 Singapore 3327 2.0 5367 .900 77 Sri Lanka 17928 1.0 273 .711 72

CARIBBEAN SIDS Antigua and Barbuda 65 0.6 150 .892 74 Bahamas 279 1.8 20 .894 73 Barbados 261 0.3 606 .907 76 Belize 213 2.6 9 .806 74 Cuba 10908 0.6 99 .723 76 Dominica 71 0.0 95 .873 72 Dominican Republic 7828 1.9 161 .718 70 Grenada 92 0.2 271 .843 72 Guyana 830 0.9 4 .649 63 Haiti 7124 1.9 257 .338 54 Jamaica 2468 0.9 225 .736 74 St Kitts and Nevis 41 -0.5 114 .853 69 St Lucia 142 1.3 229 .838 71 St Vincent and the Grenadines 112 0.9 287 .836 72 Suriname 427 1.3 3 .792 71 Trinidad and Tobago 1287 0.8 251 .880 73 3.3 Main physical features

Tables 9a present the main physical features of the SIDS located in the Indian Ocean, Mediterranean and Atlantic Regions, which here are being collectively called IMA-SIDS. It can be seen that the states in question are very heterogeneous, even those situated in proximity to each other. They differ in size (the largest being Madagascar and the smallest Maldives), climate and terrain. They also tend to face different types of natural disasters.

They vary significantly in terms of “islandness” (measured by the ratio of land area to coastal length), with Maldives and Seychelles exhibiting the highest ratios. The IMA-SIDS also vary in their natural resources endowments, terrain, climate, natural disasters and environmental concerns. The main feature which is shared by many of the islands relates to water shortages and sanitation.

Appendix 4: 19 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

Table 9b shows the main physical features of the Caribbean SIDS. Again here the states are quite heterogeneous, with some states having a large land area and others small, with the largest state being CUBA while the smallest is St Vincent and the Grenadines. As was the case with IMA-SIDS, the Caribbean SIDS differ also in their degree of islandness, natural resource endowments, terrain, climate, natural disasters and environmental concerns. The common factors in the case of the Caribbean region is that most states are susceptible to hurricanes, and they share a tropical climate. In addition, they face common problems associated with marine pollution and deforestation. In terms of terrain, many states in this area are of volcanic origin.

It would appear therefore that these two regions warrant special consideration in terms of environmental indicators. In the case of the IMA-SIDS region, it would appear that water related indicators should be given particular importance. In the case of the Caribbean SIDS, indicators related to climatic and volcanic hazards, and pollution, especially that associated with cruise shipping and industry, should be given special importance.

Appendix 4: 20 Table 9a Physical Features of the IMA-SIDS

Country Area length ratio Natural Natural disasters Terrain Climate Environmental issues (Km of area to resources sqr) coast coast (excludes fish, km timber and agricultural Products) Bahrain 680 161 0.30 oil, natural gas dust storms and mostly low desert arid with mild desertification, coastal droughts plain rising gently winters and very degradation, water to low central hot, humid shortages escarpment summers Cape Verde 4030 965 0.25 salt, basalt rock, droughts; steep, rugged, temperate with a soil degradation, volcanic ash volcanic rocky and warm, dry summer droughts movements volcanic. Comoros 2170 340 0.16 negligible cyclones and rugged tropical marine soil degradation tsunamis deforestation Cyprus 9250 648 0.07 copper, pyrites, earthquake central plain with temperate with water shortage asbestos, mountains mild, rainy winters sewage pollution gypsum, timber, scattered plains and hot, dry salt, marble summers

Guinea Bissau 36120 724 0.02 phosphates, no serious natural mostly low coastal tropical; generally soil erosion bauxite, disasters plain hot and humid; deforestation unexploited monsoonal-type deposits of rainy season petroleum Madagascar 587,040 4,828 0.01 graphite, periodic cyclones narrow coastal tropical along soil erosion chromite, coal, plain, high plateau coast, temperate endangered species bauxite, salt, and mountains in inland, arid in south quartz, tar sands, center semiprecious stones, mica Maldives 300 644 2.00 negligible sea level rise flat, with white tropical; hot, depletion of freshwater sandy beaches humid; dry, northeast rainy, southwest Malta 320 140 0.40 limestone no serious natural mostly low, temperate with water shortage, waste disasters rocky, flat with mild, rainy winters management many coastal and hot, dry cliffs summers Mauritius 2040 177 0.10 negligible cyclones and small coastal plain tropical, modified water pollution, loss of extreme climatic rising to by southeast trade biodiversity events discontinuous winds, warm, dry mountains winter and hot, encircling central humid summer. plateau. Sao Tome 960 209 0.22 negligible no serious volcanic and tropical is hot and deforestation natural mountainous humid with a rainy soil erosion disaster season Seychelles 450 491 1.08 copra, cinnamon no serious varies from island water supply problems trees natural to island. Mostly disaster granitic, and coral Singapore 623 193 0.31 negligible NA lowland; gently tropical; hot, Industrial pollution; undulating central humid, rainy; limited natural fresh plateau frequent water resources; thunderstorms s in April)

Sri Lanka 65,610 1,340 0.02 limestone, Occasional Mostly low, flat to Tropical monsoon Deforestation; soil graphite, mineral cyclones and rolling plain; erosion; coastal sands, gems, tornadoes mountains in degradation; phosphates, clay south-central freshwater pollution. interior

Appendix 4: 21 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

Table 9b Physical Features of the Caribbean SIDS

Country Area length ratio Natural Natural disasters Terrain Climate Environmental (Km of area to resources issues sqr) coast coast km Antigua and 440 153 negligible hurricanes and mostly low- tropical marine; limited natural Barbuda tropical storms; lying limestone not much fresh water periodic droughts and coral seasonal resources, islands with variation exacerbated by high some higher rainfall run off rates volcanic areas Bahamas 13940 3542 salt, aragonite, hurricanes tropical long, flat coral tropical marine; coral reef decay; timber storms, extensive formations with moderated by solid waste disposal flood and wind some low hills warm waters of damage Gulf Stream Barbados 430 97 petroleum, fish, periodic landslides relatively flat; tropical; with a pollution of coastal natural gas rises gently to rainy season waters resulting central highland (June to mostly from cruise region October) tourism; soil erosion; contamination of aquifers Belize 22800 386 timber, fish hurricanes flat, swampy tropical; very hot deforestation; water coastal plain; and humid; rainy pollution , low mountains season (May to industrial effluents, in south February) agricultural runoff

Cuba 110860 5746 cobalt, nickel, the east coast is mostly flat to tropical; pollution of Havana iron ore, copper, subject to hurricanes rolling plains moderated by Bay; overhunting of manganese, salt, from August to with rugged trade winds; wildlife; timber, silica, October, hills and rainy season deforestation petroleum prone to droughts in mountains in (May to some areas the southeast October) Dominica 750 140 timber prone to flooding; rugged tropical; NA hurricanes during the mountains of moderated by late summer months volcanic origin trade winds; heavy rainfall Dominican 48730 1288 nickel, bauxite, prone to hurricanes Rugged tropical water shortages; Republic gold, silver and severe storms highlands and maritime; soil eroding into the from June to October; mountains with seasonal sea damages coral occasional flooding fertile valleys variation in reefs; deforestation; and droughts interspersed rainfall Hurricane damage

Grenada 340 121 timber, tropical hurricanes between volcanic in tropical; fruit June and November origin with tempered by central northeast trade mountains winds Guyana 214790 459 bauxite, gold, flash floods during mostly rolling tropical; hot, water pollution diamonds, rainy seasons highlands; low humid, from sewage and hardwood coastal plain; moderated by agricultural and timber, shrimp, savanna in trade winds; two industrial fish south rainy seasons chemicals; deforestation Haiti 27750 1771 nickel, bauxite, hurricanes Mostly rough Tropical; deforestation; soil gold, silver and severe storms and semiarid erosion; clean water from June to October; mountainous (mountains in shortage periodic flooding and east cut off trade droughts; winds) earthquakes; Jamaica 10990 1022 bauxite, hurricanes mostly tropical; hot, deforestation; gypsum, mountains, with humid; pollution of coastal limestone narrow, temperate waters by industrial discontinuous interior waste, sewage, and coastal plain oil spills; damage to coral reefs

Appendix 4: 22 St Kitts and 269 135 NA hurricanes volcanic with subtropical NA Nevis mountainous tempered by sea interiors breezes; rainy season (May to November) St Lucia 620 158 minerals hurricanes and Volcanic and tropical, deforestation; soil (pumice), volcanic activity mountainous moderated by erosion mineral springs, with some trade winds; dry geothermal fertile valleys season and rainy potential season from May to August St Vincent and 340 84 NA hurricanes; volcanic Volcanic and Tropical with pollution from boat the Grenadines activity mountainous minor seasonal discharges and temperature other effluents variation; rainy season (May to November) Suriname 161820 386 timber, NA mostly tropical tropical; deforestation; hydropower, rain forest; moderated by pollution bauxite, gold, mostly rolling trade winds waterways mining and other hills; narrow activities minerals coastal plain with swamps Trinidad and 5130 362 petroleum, Not serious mostly plains tropical with a pollution; Tobago natural gas, with some hills rainy season deforestation; asphalt and low (June to overfishing mountains December)

Appendix 4: 23 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

4. A FOCUS ON SIX IMA AND CARIBBEAN SIDS 4.1 Introduction

This section zooms-in on six small island developing states in the Caribbean and IMA regions. These are Cape Verde, Jamaica, Malta, Mauritius, St Lucia and Trinidad and Tobago. These small states were chosen on an experimental basis to extend the SOPAC EVI, which has been tested for the Caribbean, to other regions. This section also discusses possible additional indicators which may be required to take into account the special circumstances in the six SIDS in the IMA and Pacific regions.

4.2 Cape Verde 37

The republic of Cape Verde, situated in the Atlantic Ocean on the west coast of Africa, off Senegal, is an archipelago consisting of ten islands with a total land area of 4030 sq km. The archipelago is divided into two groups (a) the Barlavento (windward), group is made up of Santo Antão, São Vicente, São Nicolau, Santa Luzia, Sal, and Boa Vista; and (b) the Sotavento (leeward) group, located to the south, includes São Tiago, Brava, Fogo, and Maio. The climate is tropical and dry, with an average temperature ranging from 20° to 25° C in January and 24° to 28° C in July.

The population of Cape Verde amounts to just under 400 thousand, growing, since 1990, at an annual rate of 2.5%. The population density is 96 persons per square kilometer. However the density differs from island to island.

Cape Verde was up to a few decades ago characterised by serious unrest, but it is now a working democracy, with political changes occurring within a constitutional framework. The new constitution was promulgated in 1992, leading to multiparty democracy, with free elections for president and parliament The republic has in recent years tended to politically lean to the West, but is it is considered as a stabilising factor in African politics.

4.2.1 The economy of Cape Verde

The GDP of Cape Verde, in 1995, amounted to $324 million, indicating that this republic has one of the smallest economy amongst the six countries surveyed in this section. The rate of growth of GDP was in the range of 4% per annum during the first half of the nineties.

The per capita income was US$862, which is relatively low by international standards.

The economy of Cape Verde is characterised by limited resources and scarce arable land due to the fact that the terrain is mountainous, and only 9 percent of the islands are suitable for cultivation.

About 13% of GDP is produced by the agricultural sector, a relatively high dependence when compared to the other five island states surveyed in this section, with the main products being maize and beans, sweet potatoes, coconuts, potatoes, cassava and dates. Industry contributes about 15%, of which 5% consists of manufactured products, mostly assembly and light manufacturing, fish processing, and artisanal production. The remaining 72% is contributed by the services sector, including commerce, transport and public administration.

Tourism in Cape Verde could potentially contribute more to the GDP and foreign exchange inflows of this country, but it is still very underdeveloped. Another form of production with good potential, but

37 This section draws on the following websites: http://www.umassd.edu/specialprograms/caboverde/cvgeog.html http://www.abex.co.za/Cape%20Verde/cape_verde.htm; http://www.odci.gov/cia/publications/factbook/cv.html

Appendix 4: 24 which is still relatively underdeveloped is fishing, since the ocean surrounding the islands is very productive in this regard.

The economy is heavily dependent on imports, leading to chronic trade deficits, partially offset by transfers from other countries and remittances from Cape Verdeans living overseas. The most important clients for Cape Verdean exports are Portugal, the Netherlands and Spain, while most Cape Verdeans imports originate from Portugal and the Netherlands

The most serious problem affecting Cape Verde relates to food security, caused by a structural food deficit from domestic production due to limited arable land and balance of payments constraints.

Recently, the government of Cape Verde has embarked on major changes in economic reforms and market liberalization and promotion of private sector enterprise, and reduction of foreign exchange restrictions. It is actively trying to attract foreign direct investment through tax incentives.

4.2.2 Major environmental concerns in Cape Verde

The most important environmental characteristics of Cape Verde Islands, which are mostly volcanic in origin, relate to (1) water shortages (2) soil degradation and (3) poor natural resource endowments

Water shortage. The archipelago is very windy and rainfall is low and very irregular, giving rise to frequent droughts. This has negative impacts on the quality of life in general, and on agricultural production in particular. As already explained, water shortages in Cape Verde have adverse repercussions on food production in the islands. This problem is exacerbated by relatively high water run-off rates and limited water catchment facilities. In recent years attempts have been stepped up to reduce water run-off by earth or stone-walled terraces and building of ditches or dams.

Soil degradation and desertification. The islands suffer from soil degradation (through anthropogenic factors, including such as removal of natural vegetation for fuel and construction and intensive livestock feeding and soil erosion (as a result of steep watersheds, torrential rains, rapid water runoff, and strong winds). Soil enrichment is being stepped up by growing wind-breaking crops. It is not easy to reduce demand for fuelwood given that imported fuels are very costly - although the problems could be reduced by development of alternative energy sources such as wind, solar or geothermal. Erosion control would yield major benefits to Cape Verde, since this would conserve water and enhance the fertility of agricultural land.

Limited Resources. The islands are characterised by very poor natural resources endowments. The most important minerals are pozzolana (a volcanic rock used for cement production) and salt. Water resources are very limited, and this affects agricultural production. Vegetation is sparse and consists mostly of drought-resistant species. Wildlife is very limited and includes lizards, monkeys, wild goats, and a variety of birds.

4.2.3 Other environmental issues in Cape Verde

Coastal resources. Tourism (particularly on Sal, Boa Vista and Maio) and fisheries, which have huge potential in Cape Verde, but these industries are relatively underdeveloped, and at present do not pose serious environmental impacts. No true reefs exist along the West African coast or in the archipelagos of the Gulf of Guinea and Cape Verde but there are a number of sites with rich coral communities. Again here, there does not seem to be a serious problem in this regard.

Disaster Proneness. Cape Verde islands are not among the most disaster prone among SIDS. However at least three natural disasters were recorded between 1980 and 1992. These include a

Appendix 4: 25 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

hurricane in September 1982, a huge storm in September 1984, and a volcanic eruption on the island of Fogo in April 1992.

4.2.4 Cape Verde and the EVI

The foregoing discussion on the environment of Cape Verde suggests that the country is likely to record very high risk exposure and degradation scores, and low resilience scores on the EVI with regard to indicators related to islandness (indicator number 11), land area (indicator 10), water shortages and droughts (indicators no. 3 and 41), wind gusts (indicator no. 2), soil degradation and desertification (indicator no. 40), removal of natural vegetation (indicators no. 22 & 28) and effects of past civil strife (46).

It would appear that the EVI, as currently constructed, would require some additional indicators or modification to existing ones, to take into account vulnerability associated with desertification in states like Cape Verde which experience this reality

4.3 Jamaica38

Jamaica (10,938 km2) is situated in the Caribbean Sea, west of Haiti and south of Cuba. The island`s terrain is mostly mountanous, with a fairly wide interior valleys and a narrow discontinuous coastal plain. The highest elevation on the island is the Blue Mountain peak (2256 m).

The climate is tropical (hot and humid, although interior is rather temperate) maritime, but open to northeast trade winds and land-sea breezes.

Jamaica's average annual temperature is 80°F, ranging between 78°F and 85°F. The winter season, from December to April is when Jamaica is windiest. The rainy season occurs during the fall. The island is prone to hurricanes, especially between July to November.

Jamaica gained independence from the UK in 1962 and is a Parliamanetary democracy, with elections are held every five years. Currently, the population of Jamaica is about 2.5 million, with a growth rate of around 0.9% per annum during the first half of the nineties. The life expectancy in Jamaica is about 75 years.

4.3.1 The economy of Jamaica

Jamaica was one of the most prosperous islands in the Caribbean, but its economy has experienced very slow growth rates in recent decades, with a short period of recovery between 1989 and 1992. Since 1992 high inflationary pressures, falling exchange rate and high interest rates caused particular problems to the Jamaican economy. An important contributor to Jamaica’s economic difficulties related to the decline of the bauxite mining and alumina refining industries.

Jamaica depends heavily on export of alumina and bauxite and on tourism. Tourism, h is the largest foreign exchange earner of Jamaica. Manufacturing contributes about 18% to GDP, and consists mostly of processing agricultural products, including rum, beer, cigarettes and foodstuffs. Another important sector in Jamaica is agriculture, with sugar, banana and coffee being the most important products. There is also considerable income from the export in marijuana, though this is illegal.

38 This environmental part of this section draws on information provided by Learie Miller (NRCA, Jamaica). The following websites were also consulted: http://www.odci.gov/cia/publications/factbook/jm.html; http://www.jamaicatravelnet.com/info/economy.html

Appendix 4: 26

4.3.2 Major environmental concerns in Jamaica

The most important environmental concerns in Jamaica relate to agriculture and forestry, mining and quarrying, tourism and fisheries.

Agriculture and forestry. The main problems associated with agriculture in Jamaica are excessive land clearing and soil erosion. There is an adverse effect on water quality and quantity by agricultural production. In addition, agriculture gives rise to a number of hazards associated with the use of chemicals, uncontrolled use of fire, loss of biodiversity and wildlife habitat, excessive siltation and risk of downstream flooding. In the case of forestry, the major concerns relate to land clearing for cultivation and fuelwood/charcoal production.

Mining and quarrying. Bauxite mining has resulted in a disposal problem specifically of red mud from the refining process. Air pollution from wind driven dust is common for both mining and quarrying activities. In addition, there is loss of aesthetic value of hillsides due to scarification associated with limestone quarrying and inadequate rehabilitation of mined out areas. Some beach erosion occur due to illegal sand mining.

Tourism. The Jamaican representative said that tourism in Jamaica is spatially concentrated on the north coast of the island. Wetlands have been dumped filled, beach sand mined, coastal structures such as groynes, piers and marinas established. The impact of tourism on coral reefs and sea grass beds due to poor sewage disposal and tourism associated recreational activities such as bathing and boating have been significant.

Fisheries. Fishing in Jamaica gives rise to considerable environmental harm especially because of over-fishing, use of dynamite, destruction of breeding ground such as wetland areas and sea grass beds, pollution of harbours and near-shore water bodies, use of fine mesh nets and traps and dragline method of fishing. Furthermore there is no practice to return juvenile fish to the wild.

4.3.3 Other areas of environmental concerns in Jamaica

Water resources. Water resource in Jamaica is affected by red mud pollution, saline intrusion where there is an over abstraction of water from some coastal aquifers, chemical, sewage pollution and sedimentation.

Air quality. This is associated with motor vehicle exhaust, stack gases, burning of garbage – domestic and municipal

Other concerns. Other concerns relate to oil spills, chemical accidents, transboundary movement of hazardous waste, climate change and attendant sea level rises.

4.3.4 Jamaica and the EVI

Appendix 4: 27 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

The Environmental Vulnerability Index would have some applicability to Jamaica based on a review of the indicators. While information may not be available on all the indicators it is likely that the Meteorological and geological data can be obtained. Information may not exist for some of the Anthropogenic indicators such as No. 36 which deals with Max 24 hours SO2 concentration and Nos. 37 and 38 which seeks to determine fertilizer and pesticide usage over the last five years respectively.

The indicators seem to be deficient in evaluating water quality and quantity and this resource is being adversely affected by development in many countries. Indicator No 47 recognizes the importance of water but it seems as if work so far has not yet identified the specific parameter for which data must be sought.

Indicator No.43 requires careful weighting considering the impact that mining and quarrying can have on a small island state such as Jamaica which has extensive bauxite deposits and where limestone constitute more than 66% of the rock material.

4.4 Malta39

The Republic of Malta is situated in the middle of the Mediterranean Sea, to the South of Sicily. It consists of two inhabited islands namely Malta and Gozo, and a number of smaller islands, one of which is Comino, which is inhabited mostly by tourists. The total land area of the Maltese islands is about 320 square kilometers.

The climate of Malta is typically Mediterranean, with mild, rainy winters and hot, dry summers. The islands have no rivers or mountains, with the highest point being 253 m. The annual rainfall is about 550 mm, which occurs mostly from late September to March (the Autumn and Winter seasons). The temperature is moderate, with an average minimum temperature of 10oc in the winter months (December to March) and the maximum temperature of 28oc in the summer months.

In 1995, the Maltese population amounted to about 370 thousand, of which about 330 lived in Malta and the remaining 30 thousand in Gozo. The population grew at about 1% per annum during the first half of the nineties. The population density of the Maltese islands is over 1100 persons per square kilometer, with mainland Malta registering about 3 times the density of that in Gozo.

The Republic of Malta is a multiparty democracy, with political changes occurring within a constitutional framework. There are free elections for Parliament, statutorily held every five years.

Malta has in recent years been preparing itself to join the European Union - and it will probably form part of the EU by the year 2004.

4.4.1 The economy of Malta

The GDP of Malta in 1995 amounted to $3205 million, the second highest in the six countries surveyed in this study. The rate of growth of GDP was in the range of 6% per annum during the first half of the nineties.

The per capita income in 1995 was US$8744, which is one of the highest amongst the Middle income countries of the world, and the highest in the group of six islands considered in this study.

The economy of Malta is characterised by a high degree of openness, poor natural resource endowments, and high dependence on tourism.

39 The environmental part of the section on Malta draws on information provided by Adriana Vella and Simone Borg . The following Malta government website was also consulted http://www.environment.gov.mt/ publications/soer98/

Appendix 4: 28 About 2% of GDP is produced by the agricultural sector, a relatively low percentage when compared to the other five island states surveyed in this study. In 1995, industry contributed about 39%, of which 27% consisted of manufactured products, mostly wearing apparel and electronic components. The remaining 58% was contributed by the services sector, including distributive trades, banking and finance, transport and public administration.

Tourism in Malta contributes about a fourth of all foreign exchange earnings and about 20% of GDP. There is tendency for the manufacturing sector to contract and the services sector to expand, with export oriented services, including tourism and banking and finance taking a larger share of this sector.

The economy is heavily dependent on imports, but exports are also relatively high as a percentage of GDP. However, the country faces chronic trade deficits, which are partially offset by investment income and transfers from other countries. The most important clients with regard to Maltese exports are Germany, the UK and Italy. Most Maltese imports originate from the European Union.

The most serious problem affecting the Maltese economy relate to government budget deficits and the difficulties faced a section of the manufacturing sector in competing in the international market, following the removal of trade barriers.

In recent years, the government of Malta embarked on major changes in economic reforms and market liberalisation, dismantling of state monopolies and removal of foreign exchange restrictions. This has ushered in a sink-or-swim situation for Maltese producers who have to compete with international firms in the absence of trade protection.

4.4.2 Major environmental issues in Malta

The more important environmental concerns in the Maltese Islands, relate to (1) water and land shortages (2) coastal zone management and soil erosion (3) pollution (4) waste management (5) limitations, over-exploitation and degradation of natural resources and (6) loss of wildlife and genetic erosion.

Water shortages. Natural water resources in Malta are totally dependent on rainwater, which percolates through the rocks and forms aquifers. Because of the low rate of precipitation, exacerbated by high rainwater runoff rates, natural freshwater is a scarce commodity in Malta. Such scarcity has traditionally had direct impact on human health and general well being, on agricultural production, and in recent years on tourism and certain manufacturing industries.

At present about 60% of water demand in Malta is supplied by desalination plants (reverse osmosis) - a very costly method of water production. Water tariffs are heavily subsidised in Malta and do not therefore reflect the cost of production, and this may be conducive towards lack of incentives for saving water.

Coastal zone management. The coastal area of the Maltese islands has played a very important role in the social and economic development of the country, with a large proportion of economic activity and of the resident population occurring on the coast. The natural harbours of the islands are extensively used for commerce and for tourism related activities. Urban settlements, new industrial and tourist infrastructures and other buildings are mushrooming along the coast, leading, as expected to increase waste generation and sewage pollution.

Appendix 4: 29 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

The coastal area has up to recently been considered as that restricted strip of land or rock between the sea and the human habitations. The importance of seriously considering the coastal sea areas and the vulnerability of this marine strip to the increasing local exploitation and development has been brought forward through scientific research on specific natural resources, such as local endangered marine species.

Soil and coastal erosion. Soil erosion in Malta is on the increase mainly as a result of abandoned agricultural land and limited tree cover. The building sprawl has substantially decreased the number of catchment areas, thus resulting in greater rainwater run off. Local climatic conditions and the increased over exposure of areas with soil have definitely contributed to its loss. The lack of trees and shrubs together with the loss of rubble walls further exacerbate this problem.

Waste Management. In small islands, waste management tends to be more problematic than in larger territories, due to the limited land area. In Malta the very high population density and rapid economic growth accentuate this problem. The added load of the tourist population during different seasons of the year but in particular during the summer months is not to be underestimated.

Pollution. The most important considerations in this regard relate to air and marine pollution. In Malta certain environmental damage and health risks are directly associated with elevated concentrations of air pollutants. Examples of typical pollutants are sulphur dioxide, nitrogen oxides, particulate matter and dusts, ozone, carbon monoxide, benzene, polyaromatic hydrocarbons, heavy metals (e.g. lead, mercury and cadmium). A recent scientific study has actually found strong associations between particulate air concentrations and an increasing local health problem. Such study is also starting to shed light on the dominant relationships between different weather conditions and increased presence and effects of the pollution.

4.4.3 Other environmental concerns in Malta

Other concerns related to (a) limitations, over-exploitation and degradation of natural resources (b) Loss of wildlife and biodiversity and (c) Lack of institutional and policy co-ordination.

Limitations, over-exploitation and degradation of natural resources. The initial limitations of space and biodiversity richness has imposed local limitations on the abundance of natural resources available for exploitation. This same limitation has imposed restrictions on the diversity and abundance of natural resources on and around the Maltese Islands. To avoid eradication of these natural resources serious assessment and planning of future sustainable developments and exploitations is necessary. In Malta the institutional and legal set-ups for nature conservation and protection of endemic and indigenous species are rather underdeveloped, and not adequately enforced. In addition, agricultural practices in Malta have led to a drastic reduction of forest-cover. There is also the negative impact caused by grazing goats, sheep, cattle and wild rabbits, which has led to further reduction of forest cover.

Loss of wildlife and biodiversity.Uncontrolled exploitation of natural resources is still a norm and though local expertise to undertake proper wildlife conservation assessment and monitoring exist locally the government’s will and assistance to undertake conservation monitoring and management is still very poor. Specific conservation research projects have been undertaken with the goal to highlight this requirement. The awareness to consider the need for sustainable development and the conservation of natural and genetic resources in a scientific and professional manner is only recently developing. The need to preserve the variety of local genetic resources is increasingly becoming a

Appendix 4: 30 requirement at both national and international level. Thus assessing the genetic variation of natural resources needs to become synonymous of resource assessment for long-term management and conservation. A local high human population density in itself creates a considerable stress on natural habitats and their biota. Local wildlife has also been exploited since time immemorial, and in most cases the rate at which individual organisms are removed from their population is well above the rate at which they are replaced, leading to loss of populations, species and biodiversity.

4.4.4 Malta and the EVI

Malta is likely to record very high risk exposure and degradation scores, and low resilience scores on the EVI with regard to indicators related to islandness (indicator number 11), land area (indicator 10), water shortages and droughts (indicators no. 3 and 41), natural resource monitoring and protection (indicators 16 to 24) (indicators 44, 45 & 47), soil degradation (indicator no. 40), coastal zone (indictor no. 25); tourism (indicator no. 29) removal of natural vegetation (indicators no. 22 & 28); high population density (indicator no. 26); waste generation (indicators nos. 30 and 32) and other indicators associated with economic development (indicators no. 35, 23, 18).

The EVI, as currently constructed, requires some refinements to the indicators to take into account vulnerability associated with water shortage and water salinity in states like Malta which experience this reality.

4.5 Mauritius40

The state of Mauritius (1865 km2) is situated some 800 km east of Madagascar in the Southwest Indian Ocean and forms part of the Mascarene Plateau. It has jurisdiction over a vast area of ocean space of 1.7 million km2. Mauritius comprises the main volcanic island of Mauritius and the dependencies of Rodrigues, Agalega, Tromelin, St Brandon, the Chagos Archipelago and a number of outlying islets.

The island`s topography consists of a coastal plain which rises to a central plateau (max height 670 m), with several mountain ranges and peaks. The climate is tropical, with an average temperature ranging from 12° in August to 33°C in February. Rainfall varies between 1000 mm (winter) and 5000 mm (summer) per year. Annual tropical depressions are accompanied by heavy rain and strong winds up to 300 km/h between December and May. The diverse types of coastal habitats include basaltic cliffs, numerous sandy beaches, dunes, saltwater wetlands, mangroves, estuaries, rocky shores, sheltered bays, lagoons, and coral reefs which completely surround the island.

Mauritius gained independence in 1968, holds free elections every five years and the Parliament functions on the British Westminster model. During the period 1960s - 1990s it was transformed from a typically poor, agrarian economy with high population growth to a modern low fertility

40 This section is based on information supplied by Dr Deolall Daby. Additional information about Mauritius can be found of the following websites: http://www.intnet.mu/iels/about_mauritius.htm; http://www.odci.gov/cia/publications/ factbook/mp.html

Appendix 4: 31 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS country with high levels of education and welfare. Currently the population stands at 1.1 millions with a growth rate of 1.2% and a density of 589 persons / km2.

4.5.1 The economy of Mauritius

During the past three decades the economy of Mauritius has been transformed from one characterized by low income, and agricultural production, to middle income diversified economy with growing industrial and tourist sectors. In 1995, the GDP of Mauritius amounted to $3290 million, or $2511 per capita. The average annual growth rate in recent years was 5%. The Export Processing Zone, established in 1970s, is considered to have contributed considerably to this economic success.

About 9% of the GDP of Mauritius is produced by the agricultural sector. Although declining in importance, sugar production is still a major source of income and employment, with sugarcane being grown on about 90% of the cultivated land area. About 33% of GDP is contributed by the Industrial sector, of which 23% is manufactured production. The remaining 58 per cent of GDP relate to services, including tourism, financial services and public administration.

Like many other middle income small island state, Mauritius is a very open economy and experiences chronic trade deficits. The lack of natural resources ,except for a very fertile soil, exacerbated by the very small internal market, imposes serious constraints on economic development.

The main exports of Mauritius are textiles and textile articles, representing about 60% of total exports. Sugar, which some decades ago was the main export of Mauritius, now accounts about 20% of total exports. The main export trading partners are the EU and the US. The main imported items are raw materials for the clothing industry, machines and food.

The most serious problem affecting the economy of Mauritius relate the globalisation process, implying a loss of preferential treatment for its exports of sugar and manufactured products.

4.5.2 Major environmental issues in Mauritius

The main areas of concern in Mauritius are (1) deforestation (2) loss of biodiversity and (3) climate change and extreme weather events.

Deforestation and the land use. Although 31% of the total land area is under forest, woodland and scrub, only 1% of native forest remains due to three centuries of deforestation and exploitation. Approximately 45% of the total land area (90% of arable land) is under sugar cane. The forests, nature reserves and coastal Pas Geometriques (public land constituting 10% of total) is under increasing pressure from competing uses (hotel, recreation and conservation). Encroachment on environmentally sensitive areas (coastal ecosystems, forested areas, hills and mountain slopes, nature reserves and catchment areas) is also increasing.

Bio-diversity Mauritius is ranked second in the world for having the highest percentage of its native plants threatened globally (39%) and nationally (71%). Natural vegetation has been largely destroyed due to clearing land for agriculture, settlements, road infrastructure, farming and pastures. Many endemic species have already become extinct. Modification of aquatic habitats (draining, pollution, competition for water), introduced species and commercial exploitation are the principal causes of decline of freshwater fish species.

Climate change and extreme weather events. Mauritius is at high risk from the effects of global climate change and its associated impacts. These include sea level rise and enhanced frequency of extreme weather events such as temperature and precipitation extremes and natural disasters (10-12 annual cyclones, floods, droughts, storm surges and landslides). The changes due to these calamities represent additional stresses on the environmental systems that are already under intense and growing pressure.

Appendix 4: 32

4.5.3 Other areas of environmental concern in Mauritius

Agriculture. Soil fertility is probably being maintained by the high rate of fertiliser application (600kg/ha/yr) and crop production maintained by intensive pesticide use (44 kg/ha/yr). Much of such applications run-off with red soil into inland water bodies and lagoons causing hypernutrification and eutrophication problems. Enhanced soil erosion occurs in cleared areas not under sugar cane and causes siltation problems in lagoons after heavy rainfall.

Coastal zone. The intensive urbanization of the coastal zone is a major cause for concern because of unplanned construction, land reclamation, and increasing demographic pressure. Coastal habitat degradation occurs as a result of over-fishing above the MSY level, swimming, recreation and tourism activities, siltation of lagoons, mining, quarrying, dredging, nitrogen loading and recurrent HABs. Microbial contamination of coastal waters occurs from raw sewage disposal and increasing hot spots of marine pollution result from industrial effluents, high BOD load from sugar processing and run-off of agrochemicals. Inshore fishing is detrimental in various ways such as use of undersized nets, coral breakage by poling and anchoring of boats, overfishing of aquarium fish from coral reef areas, and poor enforcement of fisheries legislation.

Appendix 4: 33 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

4.5.4 Mauritius and the EVI

Mauritius is likely to register relatively high scores in terms of risks, degradation and lack of resilience, with regard to the following aspects: sea surface temperature (indicator no. 1) cyclones (indicator no. 2), droughts (indicator no. 3), degradation of coral reefs (indicator ) loss of critical coastal and terrestrial habitats and biodiversity (indicator nos. 20, 22), land area, soil degradation (indicator no. 40), decreasing fish catches (indicator no.28), coastal erosion and marine pollution (indicator no. 33) ecotoxicology (indicatore 31), waste management (indicators nos. 32) and fresh water shortage and contamination (indicator no. 41).

The EVI, as currently constructed, would seem to require some additional indicators or modification to existing ones, to take into account vulnerability associated with fresh water shortage and contamination in states, which, like Mauritius experience this reality.

4.6 Saint Lucia41

Saint Lucia is located in the Caribbean Sea, between the islands of Martinique and Saint Vincent. It is the second largest of the Windward Islands group of the West Indies. The island has an area of 616 square kilometers and is of volcanic origin, with a rugged mountainous topography especially in the area. The Southwest area of the island, where twin volcanoes, Gros Piton and Petit Piton are located, has geothermal activity.

Saint Lucia is a parliamentary democracy with two legislative houses and is a member of the British Commonwealth. The island achieved self-government in 1967 and gained full independence in 1979. The island is of volcanic origin.

The climate is tropical with heavy rainfall and a mean temperature of 27oC. Temperatures are moderated by northeast trade winds. A dry season extends from January to May, and a rainy season from June to December.

In 1998 the population of St Lucia was estimated at 152 thousand, with an annual growth rate of about 1.3% during the current decade and a density of 229 persons per square kilometer.

4.6.1 The economy of Saint Lucia

In 1995 the GDP per capita of Saint Lucia was $3915, with a total GDP of $556 million. About 11% of GDP was contributed by the agriculture sector, where banana production feature prominently. The industrial sector contributed 21% of which about 7% was manufacturing production including wearing apparel, beverages, plastics, and electrical components. The services sector, covering public administration, tourism, and transport dominates the economy, contributing about 68% to GDP.

The Saint Lucian economy is very open, as is the case of most small island states. This renders the economy as very vulnerable to economic developments outside its control. A case in point is the impact of the removal of trade preferences on exports of bananas, which has had a large negative impact on the economy. Another source of economic vulnerability relates to the high dependence of tourism, since this industry tends to be volatile, and to a large extent depends on the interests of foreign tour operators.

Saint Lucia is at present experiencing intense competition from Latin American banana producers, and this has prompted the authorities to step up the diversification, by promoting investment in tourism, manufacturing and construction.

41 The section on Saint Lucia is based on information supplied by Ms Valerie Isaac, St Hill.

Appendix 4: 34 4.6.2 Major environmental concerns in Saint Lucia

The most important environmental concerns in Saint Lucia relates to (1) deforestation (2) Coastal and Marine degradation (3) unsustainable land use and (4) occurrence of natural hazards.

Deforestation and soil erosion. The historical pattern of land ownership and the concentration of large tracts of prime agricultural land in the hand so relatively few owners, has resulted in small farmers encroachment on forests and reserves and steep slopes unsuitable for agricultural purposes. The result has been extensive deforestation, soil erosion and siltation of inland and coastal waters and the contamination of inland water courses and coastal waters by runoff from agricultural land.

Coastal and Marine degradation.. The entire island mass must be considered a holistic system because natural and anthropogenic events no matter how far inland, impact on the shore and marine environment within a very short time frame of occurrence, The aggressive development policy being pursued, particularly in tourism and agricultural sectors, has placed many of the coastal resources under stress, This is made evident by the coastal degradation occurring particularly along the Northwest coast of the island in the form of erosion, poor coast water quality, reef degradation and the loss of marine habitat,

Land use. The problems associated with land use are characterised by conflicts between competing uses such as forestry and agriculture, recreation and tourism, environment and tourism, and agricultural and urban environment. These pressures on land have placed all natural areas under severe pressure. The last decade has seen progressive loss vast tracts of the central forests, almost total destruction of the remaining west coat mangroves, the loss of most areas of natural marshlands and the destruction of severe sand beaches.

Natural hazards. Natural hazards in Saint Lucia are primarily associated with Tropical cyclones, earthquakes, volcanic activity, land and rock slides and wave action. Tropical cyclones are the most common and damaging natural disasters with potential for severe impacts.

4.6.3 Other areas of environmental concern in St Lucia

Solid and liquid waste management. Solid waste management and the disposal of refuse is one of the most serious environmental issues facing Saint Lucia. Despite improvements in the garbage collection system, serious problems remain from poor solid waste management practices, illegal disposal of solid waste along raodsides, rivers and other sensitive habitats. In addition to the unpleasant sight and health risks posed by these practices, there is a threat to the ecology of the rivers and the mangroves, contamination of water supply and near shore marine eco-system. The human population residing in the coastal areas of Saint Lucia has grown over the years and is still growing, thus increasing the amounts of poorly treated or untreated sewage waste waters being discharged into the coast environment.

4.6.4 Saint Lucia and the EVI

From the above it appears that Saint Lucia will record relatively high scores in the EVI with regard to a number of indicators, in particular those referring to land use and deforestation (indicator nos. 22, 28) soil erosion (coastal misuse (indicator no. 11, 25), tourism (indicator no. 29), natural hazards (indicator nos.2, 7, 8) waste (indicator no. 30, 32) and chemicals (indicator no. 37, 38).

Appendix 4: 35 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

It would appear that the EVI, as presently constructed requires additional indicators or refinement of existing one, to take into account environmental vulnerability associated with ship based pollution, especially that associated with cruise tourism for countries, which like Saint Lucia, experience this reality.

4.7 Trinidad and Tobago 42

Trinidad and Tobago is an independent republic in the in the Caribbean Sea, located 12 km off the north-eastern coast of Venezuela. Tobago is some 32 km north-east of Trinidad. The country has a total area of about 5,128 sq. km, of which Tobago accounts for 300 sq. km. Port of Spain on Trinidad is the capital of the country. Trinidad and Tobago has a tropical climate. There is rainfall throughout the year with a wetter season from June to November. Trinidad and Tobago lies outside the hurricane zone, being too far south for violent tropical storms.

Trinidad and Tobago has a population of about 1,265,000 (1995 estimate), of whom about 50,230 live on the island of Tobago. The population density is about 250 per sq. km. More than 70 per cent of the population lives in urban areas. Port of Spain is the largest town, as well as the capital, with a population of about 58,400 (1990 census). Trinidad and Tobago gained its independence from Britain on August 31, 1962, and became a republic on August 1, 1976. Under the 1976 constitution, the Republic of Trinidad and Tobago is a unitary, multi-party democracy within the Commonwealth.

Trinidad’s geology is similar to that of the adjacent mainland, to which it was once attached; its soils include alluvial deposits from the Orinoco. Tobago is of volcanic origin and is a single mountain mass, although the south-west is flat or undulating and coralline.

Trinidad has a very diverse flora and fauna with both Caribbean and South American species represented. Habitats range from the rainforests of the Northern Range to the wetlands on the eastern and western coasts. About 35.5% of the country is still forested. Also about one quarter of the country is forest reserves and state lands, and many areas of both islands have been declared national parks, wildlife reserves, or protected areas.

4.7.1 The economy of Trinidad and Tobago

In 1995, Trinidad and Tobago had a gross domestic product of about US$5330, equivalent to a per capita income of US$4141, and placing it in the upper-middle-income group of countries. This relatively high per capita income reflects the fact that the country is an oil and natural gas producer.

Although well diversified structurally, with significant agricultural, manufacturing, tourism, and mineral extraction sectors, the economy is otherwise dominated by the petroleum industry, which provides about one quarter of gross domestic product (GDP), one third of government revenue, and more than two thirds of foreign exchange earnings. Industry, excluding mining and quarrying, currently accounts for about 10 per cent of GDP and about 14 per cent of employment. A major industrial estate at Point Lisas has several world scale industrial complexes producing petrochemicals, steel, ammonia, urea, and other nitrogenous fertilisers, and the synthetic fuel methanol.

Trinidad is the world's largest exporter of ammonia and methanol. Other industries include the production of liquefied natural gas, cement, garments, beer, rum, and cigarettes, and agricultural processing, notably of sugar.

Agriculture accounts for less than 3 per cent of GDP and about 11 per cent of employment. However, the soil is rich and sugar cane, rice, cacao, coffee, coconuts, citrus and tropical fruits, flowers, and vegetables are grown. The service industries, including the public, financial, and tourism sectors, are

42 The information contained in the section on Trinidad and Tobago was contributed by Dr John Agard and adapted for the purposes of this background paper. The following websites provide additional information about Trinidad and Tobago: http://www.ema.co.tt; http://www.ima.gov.tt

Appendix 4: 36 by far the most important in terms of employment, accounting for 55 per cent of jobs. Tourism is now the third-largest foreign exchange earner and the islands are being heavily promoted abroad, particularly Tobago, where there is most interest in hotel construction to take advantage of the sheltered beaches and pristine underwater environment.

Tourism in Trinidad, traditionally based on bird watching and nature tourism, has been boosted by the construction of marinas and boat-repair yards in the Chaguaramas area, which have attracted the yachting fraternity. Annual stay-over arrivals on the two islands now amount to nearly 300,000 and cruise-ship passengers' number about 30,000 per year.

Thanks to exports of crude oil and petroleum products, Trinidad and Tobago operates substantial surpluses on both the balance of trade and the current account of the balance of payments—in 1994 they totalled US$741 million and US$218 million respectively.

4.7.2 Areas of environmental concern in Trinidad and Tobago

The most important environmental problems in Trinidad and Tobago are: (1) pollution and (2) deforestation.

Pollution. Rapid population increase and industrial development fuelled by a petroleum based economy in a small island setting have led to pronounced pollution effects on the natural environment. Although new environmental and planning regulations are being put in place, many developments have taken place without the benefit of adequate physical planning or environmental impact assessments. In the built environment there is widespread pollution of land due to improper disposal of solid and toxic wastes from industry and households. Emissions from motor vehicles and industry contribute to deteriorating air quality not only in industrial areas, but also along major roads due to the presence of the largest vehicle population per capita in Latin America and the Caribbean. Industrial effluents and malfunctioning sewage (especially in Tobago) plants have contributed to the degradation of the lower courses of rivers and nearshore coastal waters.

Deforestation. There has been extensive deforestation in critical watersheds, associated with shifting cultivation, hillside slash and burn agriculture, inappropriate and illegal logging, and sand and gravel extraction. These activities cause siltation in watercourses and lead to flooding. About 31.4% of Trinidad and Tobago consists of natural forests. Deforestation averaged 2600 ha per annum (or about 0.5%/yr) from 1990 - 1995. Of critical importance to the conservation of forest resources is the annual fire problem, which occurs during the dry season. Of the total of 44,850 ha of forests burnt in the ten- year period 1987 – 1996, only 230 ha or 0.5% have been replanted.

4.7.3 Other environmental issues

Disaster proneness. Fortunately, Trinidad and Tobago is on the southern fringe of the hurricane belt. During this century, 1900 – 1998, only on seven occasions did a tropical storm or hurricane directly affect the country. Five of the tropical cyclones caused no more than torrential showers and strong gusty winds throughout Trinidad and Tobago. The remaining two attained hurricane status but only one, Hurricane Flora did substantial damage to Tobago on September 30, 1963. Trinidad and Tobago has no volcanoes. Trinidad and Tobago are on the edge of the Caribbean Plate and as such are subjected to daily minor earthquake tremors. No extensive earthquake damage has been recorded however. Even so extensive waterfront development on filled land in Port of Spain, Trinidad may make it vulnerable to earthquake induced ground failure.

Appendix 4: 37 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

4.7.4 Trinidad and Tobago and the EVI

The foregoing discussion on the environment of Trinidad and Tobago suggests that the country is likely to record very high exposure to environmental risks, degradation and lack of resilience, and therefore likely to register high scores with respect to pollution (indicator nos. 30, 31, 32, 33, 34, 35, 36, 43), removal of natural vegetation (indicator nos 22, 28).

It would appear that the EVI, as presently constructed requires additional indicators or refinement of existing one, to take into account environmental vulnerability associated with oil exploration and production of petroleum products, for countries, which like Trinidad, experience this reality.

Appendix 4: 38 5. CONCLUSION

This background paper has given a brief overview of the work in progress relating to the Environmental Vulnerability Index, highlighting the present stage in its development, namely its extension from the Pacific region to the IMA and Caribbean regions.

The paper has also given a brief overview of the economic, demographic and environmental characteristics of the small island developing states in the IMA and Caribbean region, with a focus on six SIDS, namely Cape Verde, Jamaica, Malta, Mauritius, St Lucia and Trinidad/Tobago, which have been selected on an experimental basis to consider whether the EVI model can be applied to these states.

It has been shown that these six states have characteristics which merit special consideration in terms of environmental indicators. In the case of the IMA-SIDS region, it would appear that water related indicators should be given particular importance. In the case of the Caribbean SIDS, indicators related to climatic and volcanic hazards, and pollution, especially that associated with cruise shipping and industry, should be given special importance.

REFERENCES AND DOCUMENTS CONSULTED

Albala-Bertrand, J.M. Political Economy of Large Natural Disasters with Special reference to Developing Countries. Clarendon Press, Oxford. 1993 Atkins, J., Mazzi, S. and Ramlogan, C. A Composite Index of Vulnerability. Commonwealth Secretariat, London. 1998 Axiak, V., Gauci, V., Mallia, A., Mallia, E., Schembri, P. and Vella, A. Malta: State of The Environment Report, 1998. Environment Protection Department, Malta. 1999 Borg Simone. “Country presentation on the Maltese Islands” in The Sustainable Development of in Small Island Developing States in the Indian Ocean, Mediterranean and Atlantic regions. UNEP/ISSI, Malta. 1999. Briguglio L. Alternative Economic Vulnerability Indicators for Developing Countries with Special Reference to SIDS. Report Prepared for the Expert Group on Vulnerability Indices UN-DESA. 17-19 December 1997. Briguglio, L. Small Island States and their Economic Vulnerabilities. World Development. 23:1615-1632. 1995 Briguglio, L. Preliminary Study on the Construction of an Index for Ranking Countries According to their Economic Vulnerability, UNCTAD/LDC/Misc.4. 1992 Briguglio, L. The Economic Vulnerabilities of Small Island Developing States. Study commissioned by CARICOM for the Regional Technical Meeting of the Global Conference on the Sustainable Development of Small Island Developing States, Port of Spain, Trinidad and Tobago. July 1993. Camilleri Simon. “Country presentation on the Maltese Islands” in Integrated Management of Freshwater, Coastal Areas and Marine Resources in Small Island Developing States. UNEP/ISSI. Malta. 1998. Chander, R. Measurement of the Vulnerability of Small States. Washington. 1996 Crowards, T. An Index of Economic Vulnerability for Developing Countries. (Draft). Caribbean Development Bank (1998). Crowards, T. Environmental Indicators for Barbados: A Pilot Study for 1996. Caribbean Development Bank (1997). de Cassa Sousa Barbosa, A. “Country Paper on Cape Verde” in Integrated Management of Freshwater, Coastal Areas and Marine Resources in Small Island Developing States. UNEP/ISSI. Malta. 1998. Downing, T.E. Climate change and vulnerable places: Global food security and country studies in Zimbabwe, Kenya, Senegal and Chile. Research Report 1, Environmental Change Unit, University of Oxford. 1992 Ehrlich, P.R. and Ehrlich, A.H. Healing the Planet. Addison-Wesley Publication Co. Inc., Menlo Park, CA. 1991

Appendix 4: 39 APPENDIX 4: EXTENDING SOPAC’S EVI TO THE CARIBBEAN AND IMA REGIONS

Environmental Management Authority. Trinidad and Tobago: State of the Environment 1996 Report. Zenith Services Ltd., Port of Spain. 1996 Environmental Management Authority. Trinidad and Tobago: State of the Environment 1997 Report. Trinidad and Tobago Printing Works Ltd., Port of Spain. 1997 Eurostat. Towards Environmental Pressure Indices: A first Set of Indicators for the European Union (http://www.telcom.es/tau/enviroindicators.htm). 1998 Fagoonee, I. Coastal zone of Mauritius - Sea Level rise considerations. 1989. Government of Mauritius, Agenda 21 National Report. Ministry of Local Government and Environment. 1997. Government of St Lucia. Saint Lucia Medium Term Economic Strategy: 1998-2000. Government of St Lucia. Saint Lucia National Environmental Action Plan. 1997 IPCC. Global climate change and the rising challenge of the sea. IPCC RSWG Report. 1992 IPCC. The Seven Steps to the Vulnerability Assessment of Coastal Areas To Sea-Level Rise - Guidelines for Case Studies. IPCC Report. 1991. Kaly, U.L., Briguglio, L., McLeod, H., Schmall, S., Pratt, C. and Pal, R. Environmental Vulnerability Index (EVI) to summarise national environmental vulnerability profiles. SOPAC Technical Report. 1999a Kaly, U.L., Briguglio, L., McLeod, H., Schmall, S., Pratt, C. and Pal, R. Proceedings of the Environmental Vulnerability Index (EVI) Think Tank 7-10 September 1999, Pacific Harbour, Fiji. SOPAC Technical Report. NOAA. 1999b. Langworthy, M. and Finan, T. J. Agriculture and Ecological Imbalance in Cape Verde, Lynne Rienner, USA. 1995 Nicholls, K. “Management of Freshwater, Coastal areas and Marine Resources in the OECS”, in Integrated Management of Freshwater, Coastal Areas and Marine Resources in Small Island Developing States. UNEP/ISSI, Malta. 1998. NRCA (Jamaica). State Of the Environment, Jamaica, 1995/ 96. 1996 NRCA (Jamaica). 1996 Status Report - Jamaica's National Environmental Action Plan. 1996 Pantin, D.A. Alternative ecological vulnerability indices for developing countries with special reference to small island developing states (SIDS). Report to UN Department of Economic and Social Affairs. 1997 Pernetta, J.C. “Projected climate change and sea-level rise: A relative impact rating for the countries of the Pacific Basin”. In: Pernetta, J.C. and Hughes, P.J. (eds). Implications of expected climate changes in the South Pacific Region: an overview. UNEP Regional Seas Report 1990, p. 14-23. 1990. Republique du Cape Vert. Rapport national sur l’état de la biodiversité, SEPA. 1999. Sewoobaduth, J. “Country presentation on Mauritius” in The Sustainable Development of Small Island Developing States in the Indian Ocean, Mediterranean and Atlantic Regions. UNEP/ISSI. Malta. 1999. UNDRO. Preliminary Study on the Identification of Disaster-prone countries based on economic impact. Geneva. 1990 UNEP. Human Development Report 1998. UNDP Report. Oxford University Press. 1998 UNEP. Global Environment Outlook, Earthscan. 1999. UNEP. Caribbean Environment Outlook. Chapman Bounford & Associates, London. 1999 UNEP/IOC. Western Indian Ocean Environment Outlook. 1999 Vella, Adriana. “Particulate Air Pollution in Malta: Some Causes and Effects” in International Conference on Air Pollution Proceedings, (April 1999). In press. Vella, Adriana. “Cetacean Surveys around the Maltese Islands and Maltese Sea-Users Questionnaire Study for Conservation” in Proceedings of the 12th Annual Conference of the European Cetacean Society and the First World Marine Mammal Conference, 1998. Wells, J. Composite Vulnerability Index: A Preliminary Report. Commonwealth Secretariat, London. 1996 Wells, J. Composite vulnerability index: A revised report. Commonwealth Secretariat, London. 1997 World Bank. Poverty in Cape Verde: A Summary Assessment and a Strategy for its Alleviation. June. 1994

Appendix 4: 40 APPENDIX 5

DATA COLLECTION FOR THE ENVIRONMENTAL VULNERABILITY INDEX

Discussion Paper Prepared by Craig Pratt for the UNEP Meeting of Experts on the Environmental Vulnerability Index held in Malta between 29 November and 3 December 1999 at the Foundation for International Studies, Valletta, Malta organised by the Islands and Small States Institute (Malta) in collaboration with the South Pacific Applied Geoscience Commission (SOPAC)

DATA COLLECTION FOR THE ENVIRONMENTAL VULNERABILITY INDEX

1. Introduction

The SOPAC Environmental Vulnerability Index (EVI) is still in development and has been designed as a multi-level model to describe the vulnerability of the natural environment of countries to a range of natural and anthropogenic hazards. The index is being developed in such a way that it can be broken down into sub-indices that describe levels of risk and resilience and the effects of these influences on the health or integrity of a country’s environment. Due to the variety of risks and complexities of ecosystem resilience and integrity, an indicator approach was taken to characterise them.

There are currently a total of 47 indicators in the EVI.43 However the choice is still subject to change and development, on the basis of inputs made by experts from different countries and regions.

It is proposed that he formulation and choice of indicators be based on the following criteria.

The indicators should: · be applicable globally; · be based on data already available or easily obtainable; · measure change or be a proxy for change which would do significant harm to the environment; · not be selected on any political criteria but relate only to environmental vulnerability; · be weighted to reflect the probability of change to the environment and the amount of damage which might be done; · be relatively easy for users to understand; · be well-defined; · be as uncorrelated with each other as possible to limit redundancy.

43 Initially the EVI had 57 indicators, but following a Think Tank Meeting on the EVI in September 1999, the number of indicators was reduced to 47.

Appendix 5: 1 2. Approach to the Data Gathering Process

The EVI by its very essence attempts to summarise a wide variety of environmental vulnerability data for a country. The data needed for the EVI includes meteorological data, fisheries data, land area, natural hazard data and so on. The very diverse and wide-ranging nature of these data means that their sources are widely dispersed and require some effort by a country to identify, collect and compile the information. Some of the indicators require information that could only be provided by the authorities or by experts in the respective country. It is therefore essential to have full government co-operation in the data gathering process to ensure success, as has been the case in the Pacific.

Each indicator is presented with its detailed indicator question and is accompanied by an explanation of what the proxy indicator is trying to measure. All indicators require a response and guidance is provided towards the possible agency or agencies that may be sources for the information required. Each indicator is also accompanied by a description of what data is needed for a complete response to the indicator question.

3. Issues Relating to Data Gathering in the Pacific

3.1 Data Source Identification

Appendix 5: 2 Another issue is the identification of data that may be held by agencies but which may not be known to its officers or which has not be recognised by its officers as relevant to the EVI. This has largely been due to an inability to fully understand an indicator and its data requirements, the changing of staff or just a lack of knowledge of the databases held by the agency.

3.2 Accessibility

Collection, analysis and storage of data is without doubt an expensive exercise. As a result there is increasing recognition of the importance and true value of data. This has had a significant impact on access to information. As many government agencies are asked to carry out these tasks on ever-reducing budgets, many are looking to recover their costs through charges for both primary data and time taken by personnel to access and compile required data.

3.3 Availability

3.4 Quality

The issue of data quality is not peculiar to the Pacific. In the course of data gathering, several inconsistencies between local data and international data sets have arisen. The use of different standards or methods of data collection or the use of general assessments based on small- sample-biased data may have contributed to these differences. There is also the potential of inaccurate equipment, lack of proper training in measurement procedures, lack of quality control procedures and many other reasons which could all lead to inaccuracies in data reported.

3.5 Capacity

Appendix 5: 3 in-country input and assistance. This is extremely costly and it is imperative that alternative approaches are found to provide the assistance needed.

4. The Future

Appendix 5: 4