*•• ^'~. L ^SOUTH PACIFIC COMMISSION

FISHERIES NEWSLETTER NUMBER 57 APRIL —JUNE 1991

IN THIS ISSUE

SPC ACTIVITIES 2 •:"f ^

NEWS FROM IN AND AROUND THE REGION 17 VA «HM •L'^l^ri

FISHERIES SCIENCE AND TECHNOLOGY 21 t*V^n

ABSTRACTS 22

REMOTE SENSING AND IMAGE ANALYSIS 24 IN PACIFIC ISLAND FISHERIES RESEARCH 1MB . by G.L Preston *1URHELE

BEGINNING OF W TAWAT& FIELD AC'itvrnES FOR IM>I

© Copyright South Pacific Commission 1991

The South Pacific Commission authorises the reproduclion of this material, whole or in part, in any form, provided appropriate acknowledgement is given.

Original text English

B.P. D5, NOUMEA CEDI-X, New Caledonia - Telephone: 2620.00 - Cable: SOUTHPACOM NOUMI-A - Telex: 3139NM SOPACOM - Fax: (687) 26 38 18 SPC ACTIVITIES

• FISH HANDLING AND PROCESSING PROJECT

Post-harvest Fisheries Study Tour to Pacific Latin America

A team from the Pacific Island sion), Santiago, Chile. It has a Fiji; Mr Satalaka Petaia, Fishe­ region undertook a study tour similar function to the Western ries Extension and Develop­ to Pacific Latin America during Pacific Fisheries Consultative ment Officer, Fisheries Divi­ April to study and evaluate the Committee, which fosters co­ sion, Tuvalu; and Mr Nadara- expertise available and the operation between the coun­ jah Rajeswaran, Senior Resour­ technologies developed in post- tries of Pacific South-East Asia ces Development Officer, De­ harvest fisheries in Mexico, and the Pacific Islands. partment of Fisheries and Ma­ Ecuador, Peru and Chile. rine Resources, Papua New The general objectives of the Guinea. The tour was also The study tour was funded by study tour were to explore areas joined by Dr Gordon Munro, the Canadian International De­ of post-harvest fisheries tech­ Coordinator PECC Task Force velopment Agency (CIDA), nology expertise in Pacific Latin on Fisheries Development and and was the first activity under America which can be of benefit Cooperation, and Senora Lola the Trans-Pacific Fisheries Con­ to Pacific Island countries and Dulanto de Soldi from Peru, sultative Committee (TPFCC) to evaluate the potential for who acted as interpreter. since its creation in July 1990 at utilising this expertise to assist a meeting in Port Moresby. The training and applied research The study tour members iden­ TPFCC is an informal body activities. tified a number of areas where designed to facilitate co-op­ the impressive level of technical eration in the area of fisheries The study tour participants expertise in the Latin American between countries of Pacific were Mr Steve Roberts (study countries visited could assist Latin America and the Pacific tour leader), Post-harvest Fish­ the Pacific Islands. These in­ Islands. It is based at the head­ eries Adviser, South Pacific clude training in quality control quarters of the Comision Per- Commission; Mr Maciu Lagi- and inspection services, parti­ menente del Pacifico Sur (Per­ balavu, Acting Principal Fish­ cularly in Ecuador and Chile, manent South Pacific Commis­ eries Officer, Fisheries Division, solar tent fish drying technol-

Samples of smoked fish produced at the National Fisheries Institute, Guayaquil, Ecuador.

SPC Fisheries Newsletter #57 p

SPC ACTIVITIES

ogy in Ecuador, low-cost fish changes among scientists and for specific projects and for smoking and salting techniques trainers. training opportunities in Latin in Ecuador and Peru, technical America. A report outlining assistance at a regional work­ The study tour representatives these opportunities will be pub­ shops (e.g. SPC's regional from the three Pacific Island lished and countries in the re­ workshop for the chilled fish countries were impressed with gion interested in any of the sector) and a whole range of li­ the high degree of commitment areas described can apply for terature and product formula­ to the post-harvest fisheries such technical assistance. Those tions (burgers, sausages, sector in each country and felt interested should inform SPC's smoked products, etc.), which that the Pacific Islands need to Post-harvest Fisheries Adviser. were collected in each country expand their activities in this visited. It was also recognised important technical area if they *?» that the establishment of the are to benefit fully from their Regional Post-harvest Fisheries marine resources. Centre in Suva, Fiji within the next two years will provide a Funding is po tentially available focal point for co-operative ap­ for technical experts (with a plied research projects and ex­ good understanding of English)

Study tour team members Maciu Lagibalavu, Fiji (top left), Satalaka Petaia, Tuvalu (bottom left) and N. Rajeswaran, PNG (third from right) watch fish research technologist Ramon Montano of the National Fisheries Institute in Guayaquil prepare mahi-mahi for salting and smoking.

Women's Fisheries Programme starts at SPC

On 15 May Shirley Steele joined For the past three years Shirley Women in the capacity of Ex­ the South Pacific Commission has been living in Kavieng, ecutive Officer. Her more re­ as the Women's Fisheries Pro­ New Ireland, Papua New cent job was as lecturer in Busi­ gramme Officer within the Guinea. Her initial assignment ness Studies, and English at the Post-Harvest Fisheries Project in PNG was with the New National Fisheries College in Ireland Provincial Council of Kavieng.

SPC Fisheries Newsletter #57 SPC ACTIVITIES

Prior to living in PNG, Shirley As the Women's Fisheries Pro­ providing support and assis­ worked in various Canadian gramme Officer, Shirley will be tance to the ICOD-funded colleges and universities, pri­ responsible for developing, co­ Women-in-Fisheries Support marily as a programme co­ ordinating and enhancing post- Project. ^^ ordinator and counsellor. She harvest fisheries activities re­ has also worked as a youth lating to women throughout the consultant and employment South Pacific. Much of her time equity consultant with the Ca­ during the next six months will nadian Federal Government. be spent in Papua New Guinea

• DEEP SEA FISHERIES DEVELOPMENT PROJECT NOTES

Canadian Peter Watt, who has Marketing of these catches by A practical SPC manual on worked with the Project as a airfreight to Hawaii and the vertical longlining and a train­ consultant for some time (see U.S. West Coast is now under ing video are presently in Fisheries Newsletter #56) was way and it appears that the preparation. >*^ engaged as a full-time Master- fishery will expand rapidly. fisherman in March, bringing the Project's field-staff comple­ ment to three. Peter brings to the Project ex­ tensive experience in Pacific Islands fisheries, having worked for more than four years in Vanuatu under the Village Fisheries Development Project. He also holds certifica­ tion as a vessel master and has wide commercial fishing expe­ rience. Apart from his native English, Peter is comfortable in both French and Bislama.

His present duty station is Western Samoa, where, fol­ lowing the success of the tuna longlining trials already re­ ported in these pages, he has developed a manual vertical longlining system suited to the local alia catamaran fleet. Fishing trials to date have proven to be more productive than those conducted from the Fisheries Division vessel. Some ten alia crews have now adopted this longlining method and are typically landing catches of prime tuna of about 200 —300 kg a trip. Masterfisherman Peter Watt and the wooden hauling drum he has developed for vertical longlining in Western Samoa.

SPC Fisheries Newsletter #57 SPG ACTIVITIES

I FISHERIES INFORMATION PROJECT

SIG Information Bulletins

The concept behind Special In­ relevant to the interests of hoped that it will increase terest Groups (SIGs) is to estab­ group members on an occa­ awareness and understanding lish networks of individuals sional basis, mainly in the form of ciguatera in Pacific Basin working in similar subject of this type of bulletin. In return, countries; it will also provide a areas, so as to encourage the SPC asks group members to forum for exchange of ideas on sharing of information and keep it informed of their own how ciguatera can be managed. ideas in the region. activities in the subject field, and

The SIGs are being established in response to the needs of Pa­ FISHERIES INFORMATION PRUJELT \ SOUTH PACIFIC COMMISSION cific Island fishery scientists and ,• K> VOX IS - NOUMEA development workers, as ex­ Li- J NEW CALEDONIA pressed at the SPC Workshop on Pacific Inshore Fishery Re­ CIGUATERA sources in March 1988 and at the 21st Regional Technical INFORMATION BUlLETltt NoiJo L • Mif L»l L*Nk. Oimohnl Dtp. rf Prirow \*kmim., Juubni Kick* tain, W. DaiTt, torpim Meeting on Fisheries in August IM*. CU 4SM, *IMH3O.

1989. In both cases, representa­ NOTE fflOM THE CO-OHDINATOK-

Ihr Qgu#b>p Wofrtaliin Bulletin ta in Importune itnv Inside this issue tives of the Commission's 22 initdlivc rf Ow South Pacific Commtalnn. This hilftfn •nmgr^MH (fa ti^dficonoe pf ths [wnMnn of dgnKia n Pacific Island member countries ifuny r*dfic blmJ CDLntrln. II Is hoped tto bulletin vM: 4) PDTB* *wmw*i and uiulentiiKlIng ct rigitMcro In Pari- hicMcby PC Hodti fie Bain countries (ID jmrMe A (brum fc*- tnrdwFgc (V H« and territories urged SPC to wl how d£u*ifis PHI b; mHiwgxlj (Jfl) bvrilblr rmr ocrfr nie icpoctinf; of twftMrnla tt dgwtnn fai the regkm; ftv) ptovtdc * ™Wdc tor fl* doanwrnnkm ami diEMiiliuikm become more active in the col­ of IHUIH cf ptiidla an C^IL-IVM In Ihe Kffcn; and M hJgMs^Tl bnfiirunt rc*nich f\n#ngi frnhrftring Jiitm b- lection, repackaging and dis­ hjrpinrfr$ bifcca^viring dguatra thtm^hrul iht? wrtl, EUJIHLII of djxihaii lull H-aw, by flJ. totafc F*w A semination of information on The Ant Silktin Afntilm a wnatin rrf jwjttka] itfumiMkni denAy dMoibing dpucjpra, ivfiirte HI juturvh finUinjjS frpouowir DI efcjatan COM and cttrmfcim (rrhvn gn^qu in tin: fn-kl To UtneJilkrOtal Wny Avntrg. by P. Dtbri the key fishery resources of the In c^nnlrm, jifff hn.HH? a ITHllkJ ii tlm! Sfwtll iHlafll Grain. Urrmunr In inn hVU [[ratt-uUrli1 >p.y Etcmlurrt region, as a step towards re­ an fcc smt si SfC fir hHuwm in b olradp rafca&ivC ttrifcm LV lu OL>J*« biHfcjgmnhlf tfafcib* IMl llhrarj*, 5p«cU r-W«tf &DLp KQv 10 ducing the problems of isola­ RTuJIy I vmnrreenrarmgrlhrsihrfibiin irf adds nlita] to dguman to SPC bl the t-U*t -KUmrt ft* bidiaim in tion and poor communication f&rtkwHifl buHetni faced by fisheries workers in the islands. Ptysu h • >*» P h F&ms. -HB ikr if RMKK b

S^A FUL Q«iikUa pJll *v materials, and correspondence Pacific Island countries. It is

SPC Fisheries Newsletter #57 SPC ACTIVITIES

I INSHORE FISHERIES RESEARCH PROJECT Shallow reef fish stock assessment in the outer Islands of Yap State, FSM

Yap State in the Federated These experiments are de­ estimate total standing stock of States of Micronesia is com­ signed to reduce the standing fish and of different compo­ prised of the small high island stock of fish in a given area by nents of the catch, such as of Yap proper and a number of concentrated fishing effort over surgeonfish and parrotfish. Be­ atolls to the east, between Yap a short period of time and to sides the catch data, a large and Chuuk State. Fish is one of observe the resulting decline in amount of biometric data the major resources of these catch per unit of effort (CPUE). Cengths and weights) was col­ atolls and shallow reef fish have lected on individual species in been and continue to be a major The rate of decline is propor­ the catch, and some observa­ dietary staple there. However, tional to the initial standing tions of reproductive activity the remoteness of these atolls on stock and over a short time per­ were made. Estimates of the outer islands has meant that iod is unlikely to be influenced standing stocks of the fishable they have been largely ignored by the compensatory processes shallow reef biomass and some with respect to fisheries of growth and immigration. of the more common species resources investigations. How­ Amongst the many methods of components were computed ever, the abundance of fish fishing practiced by Woleaian from these data. Given the se­ stocks of the shallow reefs in the men were community leaf lectivity of the two methods, the lagoon of Woleai Atoll was sweeps ('roop' in Woleaian) and biomass estimates refer not to recently assessed by a project spear fishing ('hapungapung'). the total biomass of fish on the jointly undertaken by the Yap Both these methods can be reef, but rather to those fishes State Marine Resources Mana­ adapted readily to repeated vulnerable to the gears em­ gement Division (MRMD) and fishing on a given site. Further, ployed. These biomass esti­ the SPC Inshore Fisheries the use of familiar, traditionally mates were then expressed in Research Project (TFRP). practised fishing methods fa­ kilograms per hectare and ex­ cilitates the logistics of an en­ trapolated to the total shallow This stock assessment project terprise involving such a large inner lagoon reef area. was the culmination of two number of people. years of planning between SPC Other observations carried out (IFRP) and Yap State MRMD. During May and June, Smith over this period on Woleai in­ The idea for the project was first and Dalzell conducted on Wo­ cluded spear fishing and hand- proposed in 1989 after IFRP leai two fishing experiments on line fishing carried out by vil­ scientist Paul Dalzell met An­ reefs in the western and eastern lagers as part of their normal drew Smith, who at the time lagoons of Woleai, using both daily activities, and an inven­ was working on the collection 'roop' and "hapungapung1 tech­ tory of boats and fishing gears of traditional fisheries infor­ niques in each area on Woleai. at the atoll. Mr Smith also per­ mation for the Yap Institute of The eastern lagoon is bordered formed a trochus survey to see Natural History. Mr Smith had by the most heavily populated if the stock densities were high spent a considerable amount of islands and the most frequently enough to permit harvesting. time in the Yap outer islands, in fished shallow inner reefs. By Trochus were transplanted to an attempt to record formally contrast, the islands of the Woleai in 1984 from Yap, but as the traditional fishing knowl­ larger western lagoon are more yet none have been harvested. edge of these islanders. While sparsely populated and contain on Woleai, Mr Smith had pho­ about one third of the 770 A report of the Woleai stock as­ tographed the use of a leaf people who inhabit the atoll. As sessment project and ancillary sweep method for drive-in net a result fishing pressure on the work is currently in prepara­ fishing on shallow reefs. Smith reefs around these islands is less tion. It will include estimates of and Dalzell felt that the use of than in the eastern lagoon. standing stocks and suggested this fishing method would en­ potential yields from the shal­ able them to estimate standing Marked reductions in CPUE low reef areas, together with stocks through a series of stock were noted over the 3—5 days notes on the biology of the reduction experiments. of the different fishing experi­ various dominant families and ments. This made it possible to species groups. >^v

SPC Fisheries Newsletter #57 SPC ACTIVITIES

Workshop on Trochus Resource Assessment, Development and Management (May —June 1991)

This workshop was held in ment exercise at the island of over 1,000 juvenile trochus Vanuatu in May — June 1991, Aneityum, in the south of the reared in hatchery facilities run in response to a recommenda­ Vanuatu archipelago. Partici­ by the Vanuatu Fisheries De­ tion from the 22nd Regional pants worked in groups to partment. All Week 2 partici­ Technical Meeting on Fisheries. gather and interpret data on the pants assisted with the experi­ The workshop was structured resource, with the aim of pro­ ment, tagging the trochus, as three one-week modules: ducing a hypothetical manage­ placing them in four different ment plan for the island's fish­ release areas, and subsequently •— Week 1 consisted of presen­ ery. Eighteen individuals par­ searching for and recording tations and discussion on lo­ ticipated in this module, based those that could still be found. gistical and economic aspects of on board a chartered vessel Initial results of the experiment the international shell trade, as which sailed to Aneityum for were more encouraging than well as trochus biology, ecol­ the field-work component. expected: mortality rates were ogy, aquaculture and resource low and growth rates extremely assessment. Forty-five indi­ The workshop provided a large high — some 15 mm animals viduals participated in this body of information not cur­ had grown an additional 5 mm module, which took place at the rently available in a published after just 10 days. The Vanuatu Centre for International Rela­ form, especially on aspects of Fisheries Department will con­ tions in Port Vila. the trochus button and shell tinue to monitor the released trade, and on aquaculture work trochus over the coming — Week 2 consisted of training being carried out in Japan and months and a detailed technical and practical exercises in the elsewhere. This is presently report will be written up once application of basic survey being edited into a workshop some longer-term conclusions methods to reef organisms. report and will be published in can be drawn. Participants studied the statis­ due course. tical considerations underlying The workshop discussions also sampling design and stratifica­ In addition to the survey work, led to several recommendations tion, carried out survey work on a release experiment was per­ for future action by the Com­ selected reef areas, and per­ formed during Week 2, using mission in assisting Pacific Is- formed analysis of data using a variety of tech­ niques. Twenty- four individuals participated in Week 2, which was carried out at Erakor Island on the outskirts of Port Vila.

— Week 3 ex­ tended the Week 2 training by giving par­ ticipants an op­ portunity to carry out tro­ chus survey work in the field during a simu­ lated manage­ Vanuatu Fisheries Department Resource Officer Moses Amos (centre) shows workshop participants around the trochus hatchery.

SPC Fisheries Newsletter #57 SPC ACTIVITIES land countries to assess and — Establish a new Special In­ for trochus and other benthic manage resources of trochus terest Group (SIG) on shells and marine invertebrates in Pacific and of other marine inverte­ the shell trade, focusing espe­ Island countries, to enable brates. The main recommen­ cially on trochus; comparison of results by dif­ dations were that the Commis­ ferent workers and from dif­ sion should: — Encourage the detailed ferent areas. study of the Aitutaki trochus — Assist Pacific Island coun­ fishery as a case study, in order The recommendations will be tries to make use of remote to provide management-related discussed by the forthcoming sensing and image processing information that will be appli­ SPC Regional Technical Meet­ in survey work on trochus and cable to the developing fisheries ing on Fisheries with a view to other marine resources, espe­ in other atolls of the region; and defining the ways in which they cially benthic invertebrates; can be implemented. -. — Encourage the adoption of standardised survey techniques

A tagged juvenile trochus, about 20 mm diameter (estimated), newly released on the reef.Th e dark spot on the tip of the shell is a blob of pink-coloured cyanoacrylate glue, used to help find the tiny shell after release.

SPC Fisheries Newsletter #5? SPC ACTIVITIES

Twenty-third SPC Regional Technical Meeting on Fisheries (RTMF) to be held in August

This annual meeting, which and Pacific Island fisheries de­ Steering Committee on 1 and 2 reviews the Commission's velopment and management'. August. PIMRIS is an inter­ fisheries activities for the last 12 The workshop will consist of agency project involving SPC, months and sets future direc­ presentations and discussion on FFA, USP and SOPAC, and tions for the programme, will be the roles of traditional practices aims to provide a variety of in­ held from 5 to 9 August atSPC's and knowledge, culturalmores formation services to regional headquarters in Noumea, New and values, and of different marine resource workers. Caledonia. The draft agenda for sectors of the community, es­ the meeting, subject to modifi­ pecially women, in modern- Following the RTMF, the South cation, is shown below. day fisheries development and Pacific Regional Environment management. The aim of the Programme (SPREP) will be The meeting will review several workshop, which has attracted hosting a two-day workshop on important issues, including the a great deal of interest in the marine turtle research, man­ draft strategic plan for the Tuna region, is to identify ways in agement and conservation, and Billfish Assessment which fisheries development which will focus on SPREP's Programme for the next five and management can proceed current efforts to promote na­ years and progress on estab­ while remaining compatible tional-level turtle research and lishing a regional fishery post- and in harmony with the management activities in the harvest research and training lifestyles of Pacific Island region. Fisheries officers at­ facility. peoples. tending the RTMF are welcome to arrive early, or leave late, if The one-day special interest The RTMF will be preceded by they wish to participate in the workshop that is normally held a two-day meeting of the Pacific other two meetings. as part of the meeting will this Island Marine Resources Infor­ year focus on 'People, society mation System (PIMRIS)

Draft agenda — 23rd SPC Regional Technical Meeting on Fisheries

5 August Opening formalities Opening address Administrative arrangements Adoption of agenda and timetable General introduction SPC Fisheries Programme SPC Fisheries Programme review — open-day display (Fisheries Programme activities will be presented in the form of interactive dis­ plays and presentations, and staff will be available to discuss their activities in detail with meeting participants.) SPC Tuna and Billfish Assessment Programme review — plenary discussion session Overview Regional Tuna Tagging Project Tuna Research Project Fisheries Statistics Project Albacore Research Project Report of Fourth Standing Committee Meeting on Tuna and Billfish Draft Strategic Tuna Research Plan, 1992—1996

6 August SPC Coastal Fisheries Programme review — plenary discussion session Overview Deep Sea Fisheries Development Project Gear Development Sub-Project Offshore Fisheries Development Sub-Project Regional Purse Seine Test Fishing Sub-Project

SPC Fisheries Newsletter #57 SPC ACTTvrrrES

Inshore Fisheries Research Project Fisheries Information Project PIMRIS Steering Committee and progress report Fish Handling and Processing Project Women-in-Fisheries Project Report on Latin American study tour Post-harvest fisheries facility — progress report Fisheries Training Project

Human resource development in the fisheries sector in Pacific Island countries

7 August One-day Workshop: People, Society, and Pacific Islands Fisheries Development and Management Introduction, workshop aims and outline Traditional systems of resource management and control in the 20th century — case studies worldwide An overview of customary systems of marine resource management in Oceania, and ways in which these can be put to use in today's fisheries management context Applying traditional knowledge of marine resources to their management Developing a resource management system in Palau Traditional knowledge and management of marine resources in Tokelau Fisheries development in Papua New Guinea — involving the people Other country interventions The role of women in Pacific Island fisheries Improving opportunities for women to participate in the development process Fisheries as a part of integrated rural development The role of extension and communication skills in fisheries development Future needs in research on, and application of, traditional and social systems and knowledge in the Pacific Recommendations for action in this area by national and international agencies

8 August: Aquaculture and marine resource enhancement Introductions of marine species in the Pacific Quarantine protocols for the transfer of biological material from aquaculture facilities SPC/ SPRADP Reef Reseeding Project Trochus reseeding experiment Enhancement of spiny lobster populations through habitat modification Fifth International Conference on Artificial Habitats for Fisheries

Development of Pacific Island pearl oyster resources

Issues relating to safety at sea in Pacific Island fisheries

9 August: Reports by other organisations

Other business

Closing formalities

^^^^^^OTO^^^

10 SPC Fisheries Newsletter #57 SPC ACTIVITIES

•TUNA AND BILLFISH ASSESSMENT PROGRAMME

Regional Tuna Tagging Project (RTTP)

The tagging vessel began 1991 caught to enable a brief search ricane Sharon, After embarking field activities in Papua New around the islands and for the three Indonesian observers in Guinea in the first week of Feb­ return trip to New Ireland. One Biak, the vessel worked west- ruary after major repairs in the drifting log was located near wards finding large schools of engine-room were completed Kapingamarangi, but produced yellowfin tuna south of Waigeo in Australia. The vessel oper­ only seven releases (5 yellowfin, Island. The schools were feed- ated in PNG waters for most of 1 skipjack and 1 bigeye). On the ing actively on surface baitfish, February, apart from a brief trip return trip, 336 skipjack were so chumming attempts were to the Kapingamarangi Islands tagged. unsuccessful. Searching farther in Pohnpei towards the end of west, the ship had excellent the month. Bad weather plagued the re­ fishing results on anchored maining days of February. Both FADs and drifting logs between The first week was spent fish­ a return visit to the Dyaul Island Gag and Gebe Islands and ex­ ing along the south-east coast of seamounts and the transit from cellent baiting in Aljui Bay. the PNG mainland and in the New Ireland to Manus Island Good fishing on unassodated Solomon Sea amongst the is­ were severely hampered by schools was also encountered lands of the Louisiade Archi­ heavy seas, with the former during a trip to the south be­ pelago. A total of 6,940 tuna was prodticing 94 releases (57 yel­ tween Ceram and the Irian Jaya tagged in this time (1,576 yel­ lowfin, 32 skipjack and 5 big­ 'mainland. lowfin, 5,179 skipjack and 185 eye), and the latter 339 skipjack bigeye), with over 90 per cent of releases. The vessel then found excellent the releases coming from a net­ fishing off south-east Waigeo, work of Philippines payaos an­ The Te Tautaibegan operations but poor fishing and extremely chored in the Solomon Sea. in Indonesian waters on 6 strong north-west currents During the second week, 225 March 1991, when the vessel north of Waigeo. A return trip releases were made near crossed into Irian Jaya from to the FADs near Gag Island the Rabaul (120 yellowfin, 102 Papua New Guinea. Activities following week produced 33 skipjack and 3 bigeye) and 1,794 during the first week were recaptures by the tagging ves­ releases were made on sea- hampered by rough seas and sel in one day and a new RTTP mounts to the south of Dyaul high winds generated by Hur- record number of releases Island in the Bismark Sea (807 yellowfin, 966 skip­ jack and 21 bigeye). Over half of the Dyaul releases were double tagged.

Following a visit to Kavieng, the vessel un- dertooka long- range trip, fishing around Tench Is­ land, north to the Equator, and then west to the Kapingamarangi Islands. A total of 1,535 fish was tagged on the outward leg of this trip, comprising 7 yellowfin, 1,527 skipjack and 1 bigeye. Two nights were spent baiting in the Kapingamarangi lagoon. Small but sufficient quantities of bait were Indonesia—purchasing live bait from local lift net boats Wagons).

11 SPC Fisheries Newsletter #57 SPG ACTIVITIES

few tuna schools were sighted. Baiting in south­ ern Mindanao was also poor, although full moon conditions may have contributed to the poor catches. The vessel moved to the west and conducted a thorough survey of the Sulu Sea basin for the next ten days. A productive payao was located near Tubba- taha Reef, where 800 re­ leases were made. Several payaos near Puerto Princesa and Honda Bay in Palawan were checked out but only large quanti­ ties of dolphin fish were present. SPC specialist fisherman, Eroni Dolodai, tagging in Indonesia - Indonesian pole-and-line boat fishing on FAD. Searching north along the (1,747) for a single day. Moving Bureau of Fisheries and Aquatic northeast coast of Palawan was West to Maluku near Bacan and Resources in Manila joined the also unproductive, but the Mandioli Islands, the vessel re­ cruise and participated in tag­ Palawan baitgrounds supplied leased good numbers of tuna on ging and data collection activi­ enough bait for daily opera­ FADs south of Bacan, but bait­ ties. tions. Tuna were also absent ing was poor. Fishing farther from the fishing grounds and west off eastern Sulawesi was Three days were spent in the payaos south of the Chuyo Is­ poor, but quantities offish were west and central Moro Gulf, lands in the northern Sulu Sea. located in central and northern where dozens of anchored Tomini Bay just before the ship payaos were examined, but few An extremely productive cleared from Indonesia and held any tuna schools and very baitground was discovered in headed for the Philip­ pines. The project ac­ knowledges the vital role the Indonesian observers played in assisting with port clearances and sup­ plying the ship with ac­ curate fishing and baiting information.

The Te Tautai spent 19 days in Philippine waters between 29 March and 16 April, entering the zone from Indonesia and the Celebes Sea. Four thou­ sand tag releases were made, consisting of 830 yellowfin, 3159 skipjack and 11 bigeye tuna. Two fishery scientis ts from the SPC Fisheries Scientist, David-Itano, hands over first prize in an annual lottery for Philippine tag returns.

12 SPC Fisheries Newsletter #57 '!'

SPG AGTrVITIES

the lagoon of Cagayan Atoll, yellowfin, 1,388 skipjack and 4 Anna and Sonsorol, before the where large quantities of sprats bigeye tuna). Six days were vessel arrived in Koror. Two fi­ were caught. This bait was used spent operating out of the nal days were spent south-east to release almost 1,200 skipjack Helen Reef baitground, fishing and south-west of Koror, fish­ near Cagayan Island and over on surface schools of skipjack ing on a large unassociated 1,800 yellowfin and skipjack and yellowfin south of Helen skipjack school and one log- tuna on two payaos near Reef and on log-associated associated school where a high Tubbataha Reef. Ninety-seven schools to the north. Two days percentage of yellowfin were tuna tagged during this cruise were then spent fishing around released. >j"*x were recaptured by the Te the small islands of Merir, Pulo Tautai crew on the Tubbataha Reef payaos. On 15 April, the ship cleared from the The table below (totalas at 30/4/91) summarises the releases and recoveries Philippines, crossing the Moro to date, by specks. Gulf and searching the nearshore waters of southern Mindanao. No schools were RELEASES encountered and the vessel Yellowfin Skipjack Bigeye Other Total continued toward HelenReef in 23,221 40,994 1,111 82 65,408 southern Palau. RECOVERIES The remainder of April was Yellowfin Skipjack Bigeye Other Total spent in Palau waters, where 2,320 tuna were tagged (928 1,330 3,671 84 5,085

Joel Opnai and John Hampton, SPC Fisheries Scientists, bio-sampling tunas

13 SPC Fisheries Newsletter #57 SPC ACTIVITIES

18 16N 14N 12N ION OSN 06N 04N 02N OON 025 0*5 06S OBS 10S 12S 145 165 1£S

NSl NSl N*l NJl NOl NfiO N90 NfrO N30 NOO S30 SvO S90 S80 SOI S5 I S»l 591 S81

Places visited by Te Tautai during February — April 1991 14 SPC Fisheries Newsletter #57 SPC ACTIVITIES

South Pacific Albacore Tagging Project

The 1990 — 91 SPC Albacore Tagging Project ended in Nelson on 26 March. The final ten days of the project were spent in New Zealand (NZ) waters, coinciding with the end of the season.

Solander 3, dual purpose vessel, chartered for the programme

Solander 3, a Fijian dual-pur­ this was no surprise so early in climes, wondering what they pose pole-and-line-cum- troll the season. But if Cruise 1 was had let themselves in for. fishing vessel chartered for the to be considered disappointing, However, despite these dis­ programme, had begun its Cruise 2 at best had to be comforts and the consequent work, after a minor delay in­ thought of as depressing. This morale-sapping, meagre fish­ voked by Cyclone Sina in Fiji, cruise began on the West Coast, ing, the crew held up remark­ in Whangerei, New Zealand on South Island of New Zealand ably well. 13 December 1990. and moved into the mid- Tasman area traversing almost After a two-day break for re­ Tagging proceded first in the as far as the Australian Fishery pairs and reprovisioning in waters around the east, then Zone. It was beset by horren­ Nelson, Cruise 3 began with west, coasts of northern New dous weather conditions which baiting in the Marlborough Zealand and progressed south drove most o ther fishing vessels Sounds, then moved east to al­ to the Vitti Canyons off New to port, exposed some weak­ bacore grounds off the Plymouth. With few releases nesses in our own vessel and Wairarapa and Hawkes Bay here, the programme was off to left the Fijian crew, who mos fly regions of the North Island of a disappointing start, though lacked experience of southern New Zealand and on to the

15 SPC Fisheries Newsletter #57 SPC ACTIVITIES

Chatham Islands. The release ability of a pole-and-line fish­ length-frequency and driftnet/ rate started to pick up, with ery, had, until this stage, met troll vessel interaction data.) some particularly good fishing with unsuccessful results. Cruises 6 and 7 were short and on the South Madden Bank. Schools encountered had took place back in New Zealand Although the cruise had to be mostly seemed to be invulner­ on the east and west coasts re­ interrupted for further repairs able to that method of fishing. spectively. to an errant radar, prospects for The day arrived, however, the project were looking when a suitable school was met A total of 2,701 albacore and 22 brighter. A last-night baiting and fished most successfully skipjack was tagged. The pro­ before the ship left the coast also until the remaining bait re­ gramme was conducted jointly produced excellent results, with serves were exhausted. Unfor­ with the MAF research vessel . a fine haul of pilchards (Sardi- tunately, evolving appropriate Kaharoa, resulting in more than nops neopikkardis), some sprats methods to gather bait in the 3,100 albacore being released (Sprattus antipodum) and some STCZ was never quite achieved for the 1990 —91 season. As yet Jack mackerel (Trachurus no- and so this very productive day no recoveries have been re­ vaezelandiae). was never repeated. ported, but we eagerly await returns as the fish recruit into The last land-based refuelling, The exercise did demonstrate the Asian longline fishery. The before the major cruise of the very clearly the well-held belief project has already provided a programme, took place in the that fish released after capture valuable base on which to de­ Chatham Islands. Land would on pole-and-Iine gear are in far velop improved strategies for not be sigh ted again for another superior condition to those re­ obtaining the mark-recapture 40 days. Reprovisioning during leased after troll capture. There data necessary to assess the the extended Subtropical Con­ is a very real need for South status of the resource and in­ vergence Zone (STCZ) cruise Pacific albacore mark-recapture teractions among the fisheries. was to be made at sea. During data and if a means to release this extended portion of the more pole-and-line-captured Recovered tags should be re­ project, advantage was taken of albacore could be developed, it turned to a local fisheries office an offer from Solander 3's would help to achieve that end. or sent to: South Pacific Com­ owners to place a tagger aboard mission, B.P. D5, Noumea another of their vessels engaged An effort was also made to ob­ Cedex, New Caledonia. If pos­ in commercial operations in the serve interaction between troll sible, each tag should be ac­ same fishery. fishing and driftnet fishing companied by a record which fleets and all fish landed were includes the tag number, a de­ A crewman, Jona Ravasakula, recorded for presence (or not) of scription of where, when and who had worked on the SPC driftnet marks. Overall there how the tuna was caught, the skipjack tagging project aboard were very few encounters with fork length and the name and the Hatsutori Maru 5 in the late driftnetters, whether Solander 3 address of the tag recoverer. 1970s, was swiftly recruited and was irking alone or amongst Rewards will be sent for every hLi work aboard the Daniel the troll fishing fleet, as was the tag recovered. ,x*v Solander 0 eferred to as Cruise 5) norm. SPC staff who had made a substantial contribution worked the two previous sea­ to the overall success of the sons in the fishery noted that project. Cruises 4 and 5 ex­ there was markedly less visible tended east of the Chathams in driftnet activity this year. the STCZ to the south of French Polynesia. More than half the There were enough occurrences total releases for the project of marked fish, however, to in­ were made during these dicate that driftnetters were still cruises, the bulk of them in a 5- an influence in the fishery. Data degree square 700 miles south­ are in the process of being west of French Polynesia and combined with those from SPC more than 1,500 miles east of observers aboard commercial New Zealand. vessels for further analysis. Giuropean Community-funded A secondary objective of the observers were placed aboard programme, to assess the vi­ commercial vessels to collect

16 SPC Fisheries Newsletter #57 NEWS FROM IN AND AROUND THE REGION

[SURVEY OF SAFETY-AT-SEA ISSUES IN PACIFIC ISLAND ARTISANAL FISHERIES The FAO/UNDP Regional officials concerned with search designed for people involved in Fishery Support Programme and rescue in the countries sur­ commercial or industrial fish­ recently carried out a survey to veyed. Because of problems of eries, not artisanal activities. determine how large the prob­ communication in many coun­ Many officials in countries sur­ lem of safety at sea is for small- tries and remoteness of many veyed believe that the public scale fishermen and what could islands and villages, this prob­ does not really consider the be done to improve the safety ably represents only a fraction safety problem for fishermen to situation. The Programme sur­ of the actual incidents. Like­ be large, and therefore educa­ veyed 16 Island countries and wise, the known fatalities at­ tional programmes aimed at territories in the Pacific region tributed to these incidents, increasing public awareness and interviewed 167 fishermen about 60 per year, is probably would be worthwhile. and government officials. far less than the actual total. It is estimated that the region A report of this study is avail­ The survey estimated that there spends close to US$1 million on able from the Regional Fishery is a total of 25,000 non- efforts to locate lost fishermen. Support Programme (UNDP mo torised and 16,000 mo tor­ Mail Bag, Suva, Fiji). ised artisanal fishing vessels in In the past several programmes the region. Approximately one have attempted to address (Source: R. Gillett — FAO/ distress incident per day related problems of outboard engine UNDP) ^ to the operations of these ves­ repair. However most institu­ sels comes to the attention of tional courses in the region are

[FAO PACIFIC ISLANDS FISHERIES CONSULTANCIES During the period January 1986 Fishery Support Programme or available from the FAO/UNDP to February 1991, 111 consul­ indirectly through regional or­ office. Enquiries for country- tancies and staff missions were ganisations. The following table specific reports should be made carried out either directly by summarises this work Reports to the country concerned. the FAO/UNDP Regional for the regional missions are

Area Consultant/Staff Work

Regional Cillauren (1990) Lectures at USP course Cook Islands Savins (1990) Collect information for re-design of flying fish boats Cook Islands Savins (1990) Conduct boatbuilding course

Cook Islands Gulbrandsen (1990) Design flying fish vessel Fed. States of Micronesia Sibert(1990) Assess tuna fishery Fed. States of Micronesia McCoy (1990) Advise Yap State on industrial fisheries development Fiji Gulbrandsen (1990) Produce plans for 28-ft vessel

Fiji Watts, Maclnnes (1990) Review EKA Corporation manpower situation

Kiribati Savins (1990) Build new-design vessel and carry out emergency sail rig trials 17 SPC Fisheries Newsletter #57 NEWS FROM IN AND AROUND THE REGION

Kiribati Gulbrandsen (1990) Produce plans for emergency sail rig Marshall Islands Medlin (1990) Install new computer system and provide training; in conjunction with FFA

Niue McCoy (1990) Advise on fisheries development plans

Palau Crossland (1990) Advise on fisheries planning Papua New Guinea Hartong (1990) Produce information for re-designing vessels Papua New Guinea Gulbrandsen (1990) Design plank on frame fishing vessel, propose design changes to existing plans

Solomon Islands Gulbrandsen (1990) Re-design fishing catamaran Tokelau McCoy (1990) Advise on fisheries development plans Tonga Walton (1990) Review boatbuilding project

Tonga Dunlop (1990) Review longliner design Tonga Mead (1991) Instruct in FAD planning and deployment

Tuvalu Savins (1990) Advise on boatbuilding training requirements and on repair of Government vessel

Tuvalu Savins (1991) Provide boatbuilding training, introduce new fishing vessel designs

Vanuatu Savins (1991) Provide boatbuilding training

Western Samoa Gillett (1990) Transplant trochus Western Samoa Crossland (1990) Advise on structure and function of Fisheries Division (Source: R. Gillett — FAO/UNDP) ^£^

HAWAIIAN ORNAMENTAL REEF FISH REARED IN CAPTIVITY COULD BOOST WORLDWIDE MARKET

Hawaiian ornamental reef fish There's an obvious need for this sity. "That industry's practices for the aquarium trade are bring because there is a very large are badly depleting the wild successfully raised in captivity industry based on the collection population, so somebody for the first time at the Univer­ and export of wild ornamental should work on rearing reef fish sity of Ha waii's Hawaii Ins titute reef fish', said Christopher in captivity.' of Marine Biology (HIMB) on Brown, associate professor of Coconut Island in Gahu. marine biology at the Univer- 18 SPC Fisheries Newsletter #57 NEWS FROM IN AND ABOUND THE REGION

The worldwide market for or­ Brown and Danilowicz have are placed in the tank with the namental reef fish is estimated tried a different approach. In­ active larvae culture, so the food at US$100 million annually. Up stead of the conventional chain remains intact and its to 250,000 fish are taken from method, they are culturing the components nutrient-rich. Hawaiian reefs each year, pri­ algae, along with small live marily for export, with about 30 foods such as rotifers and The Hawaiian damselfish was species of Hawaiian reef fish copepods, in the same tank as chosen as a model for initial exported regularly. the larval fish. experiments because it is readily available in Kaneohe While rearing methods for Bright lights are used to stimu­ Bay, its eggs are relatively large, some species such as clownfish late the algae growth and sus­ it is an inexpensive mainstay of and neon gobies have been de­ tain an entire food chain, the aquarium trade and it is veloped, none of the Hawaiian thereby giving the larvae a popular as an export. reef fish have ever been raised choice of assorted food organ­ in captivity. isms from which to choose. 'What we've done here is a demonstration on a very small Now, two species of damsel- This differs from the conven­ scale to show that rearing or­ fish, Dasq/llus albisela from Ha­ tional method of daily feeding namental reef fish in captivity is waii and D. arunus from Guam, in which food organisms may not impossible. It's just a mat­ have been reared to marketable not be eaten immediately and, ter of giving them the food they size under laboratory condi­ without their own nutrition need.' tions in just ten weeks. In the source, begin to starve and lose wild, these fish take a full year their nutritional value. With the (Source: The Garden Island) to reach market size. new method, food organisms

Brown and Bret Danilowicz, a graduate student at Duke Uni­ versity, have collaborated in developing new feeding tech­ niques for the damselfishat the critical larval stage of develop­ ment. Following larval biology research funded by the UH Sea Grant College Program, Brown and Danilowicz's experiments demonstrated that reef fish can survive in-tank rearing from the egg to the hardier juvenile stage if a complete food chain is provided during early devel­ opment.

The conventional method of rearing larval fish has been to set up a small tank with the fish eggs in it. Upon hatching, the larvae are placed in a large tank. In still another tank, under controlled lighting, cultures of algae and other microorgan­ isms are grown as food and transferred into the larval tank daily.

19 SPC Fisheries Newsletter #57 NEWS FROM m AND AROUND THE REGION

[SEDIMENTATION THE MOST CRITICAL REEF PROBLEM

The most critical problem af­ the marine life unless proper where improvements can fecting coral reefs on Guam and precautions are taken. work. The use of adequate throughout the Pacific Islands is ponding basins, slope stabilisa­ sedimentation. With the recent Sediments affect reef animals in tion, phased clearing, mulching passage over Guam of ty­ several ways. For corals, the and grading are all methods phoons Mike, Owen and Russ, most obvious damage comes that can be used and modified the enormous amount of sedi­ from burial and eventual death. to fit the conditions of a par­ ment being dumped in the Sediment also stops sunlight ticular project. coastal environments was vis­ penetrating the water. Because ible to all. corals contain plant cells and Timing also is important, con­ use them as their energy source centrating certain activities Red plumes were seen orig­ via photosynthesis, corals can during the dry season. Com­ inating from streams, rivers and starve if they are shaded by silt munication among developers, outfalls. In addition to the vi­ in the water. government agencies, technical sible sediments, many other experts and legislators is critical substances toxic to marine life Fish also respond to sediment to the success of preserving were present in the runoff. by avoiding plumes and mov­ Pacific Islands reef resources ing away from dead areas. Fish while enabling development to This article summarises the that normally spawn in reef occur. problems caused by sedimen­ channels were found to stop all tation and runoff, how these activity or move elsewhere on Mistakes have been made, but affect marine animals, and what the reefs of southern Japan they can be valuable if we leam can be done to prevent unnec­ when sediment was present. what went wrong and direct essary damage to precious Once a reef dies, the related efforts toward solutions. We coastal resources. edible reef resources disappear know that sedimentation and as well. runoff kill coral reefs and the All rock, coral limestone and mechanisms by which the sand grains that make up an The question then becomes: can death occurs are known. island are subject to erosion. anything be done to prevent Erosion is the process by which sedimentation and control (Source: Pacific Daily News) material is broken down and runoff? Fortunately, the answer moved by the forces of nature. is yes. Project design is one area For most islands, the eventual destination of runoff and sedi­ ment is the coastal zone, in­ cluding bays, lagoons and reefs. After heavy rain, bo th dissolved and solid materials travel down streams and enter the ocean. The goal of most engineers and planners is to move the water as quickly as possible off the roads and land.

The runoff enters the ocean di­ rectly through the aquifer (the underground water supply). On islands like Guam, this can have serious and damaging ef­ fects. If the water contains tox­ ins, such as herbicides, pesti­ cides and petrochemicals, it will poison the drinking water or

20 SPC Fisheries Newsletter #57 FISHERIES SCIENCE AND TECHNOLOGY

I OTOLITH REVEALS ALL TO FISH RESEARCHERS Canadian researchers say that The most stressful times of • Age — even when the tradi­ fisheries management world­ the fish's life. Check lines in­ tional method of counting wide can benefit from research dicate major stress. Ex­ rings cannot be used. work on the otolith—the half- amples include spawning, inch long 'earbones' of fish. and times when the fish may (Source: Fishing News Interna­ have been caught and then tional) A. Patterns on the otholith seen thrown back into the water. under a compound mi­ croscope are like the swirls of a thumbprint, say scientists with the Canadian Department of Fisheries and Oceans (DFO).

A mass of information is available when the otolith is examined closely. This includes:

• Fish's age. Up to one year of age, a ring is formed for each day of life. After one year, the daily rings continue but become intermit­ tent. However yearly rings are also formed and can be read right up until the time of death. Otoliths of bigeye tuna (Thutmus obesus) x 5.5 • Date of hatch. By iden­ tifying the very first ring at the otolith's core, research scientists can calculate the fish's birth-date.

• Fish's growth rate: the thicker the daily ring, the greater the growth on that corresponding day. • Migration patterns. By analysing the chemical composition of the dif­ ferent rings of the oto­ lith, scientists can dis­ cover when the fish swam in warm water (such as near shore) and when it swam in cooler water (such as Otoliths of red snapper {Etelis sp.) x 6.5 offshore). 21 SPC Fisheries Newsletter #57 ABSTRACTS

Report on a market survey on giant clam products in selected countries by Y.C. Shang, C. Tisdell and P. Leung The natural stocks of giant the price and taste of the pro­ dam products are not restricted clams have been so seriously duct. If Tridacna derasa were .to for trade, it is difficult to distin­ depleted, mainly because of be exported to Okinawa, the guish between naturally bred over-harvesting, that most spe­ price for a two-year-old dam and cultured clams. Therefore, cies are now listed as threat­ would have to be much lower certification that the giant dam ened. The recent advances in than T. crocea'S current price of products were cultured may be culture techniques make giant US$1.43 to $2.38 to be accept­ necessary, and an import per­ clam farming possible. With the able. mit is required to ship giant anticipated expansion in giant clam products to the member clam production, marketing In Taiwan, a market for fresh or countries of CITES. information is urgently needed. frozen giant clain adductor The lack of published trade data niusdes exists,' with a on giant clam products made it potential of about 2401 necessary to collect basic mar­ annually. The eco­ ket information by field inter­ nomic feasibility of views. producing the large- size clam musdes at The purpose of this study was competitive prices to evaluate the existing market needs to be studied for giant clam products in Ja­ further. pan, Taiwan, Hong Kong, Australia and the United States. In Australia, there is a Most of the information for this market for giant clams study was obtained through among immigrants interviews with seafood im­ from the Pacific Is­ porters/distributors, restaurant " lands and their owners, aquarium species dis­ descendants.The use tributors and shell distributors of giant clams for food who have had experience in is unknown in the Ja­ handling giant dam products. pan mainland, Hong The interviews were conducted Kong and the United during the latter part of 1989 States; the market po­ and early 1990. tential should be ex­ plored. The preliminary market survey, ascertained that markets exist A limited market po­ for five types of giant clam tential exists for giant products: food, aquarium dams as aquarium specimens, seedstock, brood- specimens in Japan, stock and shell. The largest Australia and the potential market for giant dam United States. The products appears to be as food market for giant clam shells The book is published by the appears good. A limited market Center for Tropical and Sub­ Markets for giant clams as food also exists for giant dams as tropical Aquaculture, Univer­ already exist in Okinawa, Tai­ seedstock and brood-stock in sity of Hawaii, Makapu'u Point, wan, Australia and the Pacific the Pacific Islands. Waimanalo, HI 96795, USA. Islands. The food market in Okinawa may absorb up to Giant clams are treated as 5>0001 of giant clams in-shell— threatened spedes, and trade of mainly for use in sashimi and naturally caught giant dam sushi dishes. People in products between the signato­ Okinawa prefer Tridacna crocea.rie s of the Convention on In­ Whether other giant clam spe­ ternational Trade in Endan­ cies could penetrate the Oki­ gered Spedes (CITES) is res­ nawa market would depend on tricted. Although cultured giant 22 SPC Fisheries Newsletter #57 ABSTRACTS

Artificial reefs for marine habitat enhancement in Southeast Asia by Alan T. White et al.

Can sunken junk replace coral and use of artificial reefs and is essential if they are to be more reefs? Old tyres, bamboo bar­ recommends that these be em­ efficiently used.' ges, junk cars and other scrap ployed within the broader con­ materials are gaining wide­ text of coastal resources man­ The book compares the features spread use as artificial reefs in agement. of natural and artificial reefs Southeast Asia, where they help from the fishes' point of view improve commercial and small- 'Artificial reefs can and do at­ and illustrates some basic scale fisheries by attracting fish. tract fish; they can be deployed structures and construction for commercial, subsistence and materials. The authors say that Within Asia, countries like recreational purposes. But the 'proper planning and design of Brunei Darussalam, Indonesia, present state of knowledge artificial reefs are a must for a Malaysia, the Philippines, Sin­ cannot as yet give a clear un­ successful outcome'. gapore, Taiwan and Thailand derstanding of their biological have established national and ecological functions, which The book is published by the programmes to construct and International Center for Living submerge artificial reefs in se­ Aquatic Resources Manage­ lected areas. ment aCLARM). Contact ad­ dress is: ISBS, PO Box 1632, A new primer on artificial reefs, Beaverton, Oregon 97075, USA. Artificial reefs for marine habitat enhancement in South­ east Asia, however, cautions that while man-made reefs offer immediate benefits, they may also have adverse effects on the environment and should be used judi­ ciously, with careful at­ tention to environmen­ tal, economic and so­ cial considerations.

The book offers guidelines for the effective design

An artificial reef module of 36 tyres anchored with weights. Old tyres are the most common material used because they are available at low cost, are easy to handle and are physically and chemically stable under water. 23 SPC Fisheries Newsletter #57 Although the resolution of REMOTE SENSING AND IMAGE these instruments can be very high, the filmha s to be returned ANALYSIS IN PACIFIC ISLAND to the ground for processing. The difficulty of doing this in FISHERIES RESEARCH some cases, especially with sat­ ellites, has made it necessary to use technology that permits the Introduction transmission of images in the by G.L.. Preston form of signals. Most recently, Since 1988, when I attended an Soufli Pacific Commission the development of digital Australian conference on 'Re­ . Noumea, New Caledonia technology has allowed the mote Sensing and the Coastal rapid transmission of large Zone', I have been trying to in­ personal experience of applying volumes of data with relatively form myself about the ways in remote sensing to fisheries low rates of error. which my work and the work work. of my Pacific colleagues could Digital images can be produced benefit from the use of remote Basic concepts by scanning ordinary photo­ sensing. At that point my graphs, but in the case of remote knowledge of remote sensing What is remote sensing? Es­ sensing it is more usual to em­ was nil, but there seemed to be sentially it is observing, making ploy an instrument that cap­ all kinds of potential applica­ measurements of, and deriving tures the image directly in a tions for this technology and I information about the earth digital form. Typically, the in­ wanted to know how practi­ from a distance. Remote sens­ strument scans the image a tiny cable these were. Since that time ing is not a new technology: for bit at a time, dividing it into a I have attended further techni­ many years man has been as­ grid made up of numerous cal conferences dealing with cending from the earth in order picture elements, or 'pixels'. For remote sensing, and become to observe it from above. Aerial each pixel, the instrument involved in using remote sens­ photography is the oldest and records one or more numbers ing and digital image analysis best established form of remote that describe the intensity of the technology in quantitative sensing. However, the devel­ radiation being reflected at fisheries work on two occa­ opment of new types of sensors, given frequencies. These nu­ sions, with a third project under digital instruments and data merical data can then be trans­ way. transmission techniques, and mitted to earth, where the im­ satellites, have all combined to age can be recompiled, based It seems that I am not the only offer new remote sensing pos­ on the size of the pixels and one interested in this subject sibilities. their characteristics. Fisheries departments, envi­ ronmental agencies and uni­ Any airborne platform offers Pixel size has a critical effect on versities in the region are be­ the possibility of remote sens­ the resolution of the image, and coming more aware of the ing, and aerial photographs therefore its usefulness for dif­ availability of remote sensing have been and continue to be ferent purposes. The pixel size technology, and fisheries-re­ taken from balloons, aircraft, that the instrument uses in lated projects that make use of space craft and satellites. Aerial capturing the image determines it are under way in French photographs use conventional how finely detail can be seen by Polynesia, New Caledonia, optical components (lenses and the user — whether he can Papua New Guinea and mirrors) and films that are sen- gather information on parts of Vanuatu, with others planned si tive to light or other sorts of trees, or whole trees, or groups in Palau, Tuvalu and elsewhere. radiation. A photosensitive film of trees, or forests. Some users However, I believe we are still is coated with tiny crystals that need high resolution, while at an early stage in the process undergo a permanent reaction others may be able to sacrifice of learning what this technol­ on exposure to radiation. The precise detail in favour of the ogy can do for the region, and crystals are distributed ran­ large 'footprints' of satellite in­ what it cannot. In this article I domly on the film and the struments. A typical satellite will try to summarise some of combination of the colours that image would be a square whose the more important points that they produce go to make up the side is between 60 km and I have discovered from my image that our eye perceives. 200 km, giving an area coverage

24 SPC Fisheries Newsletter #57 f

REMOTE SENSING AND IMAGE ANALYSIS

of over 40,000 km2. As aerial colours. By recombining these numerical format, it can also be photographs are more typically three, it is possible to generate assimilated by computers and 2 km by 2 km, or 4 km2, the just about every shade and thereby subjected to various economies of scale become colour the human eye is capsfele kinds of analysis routines and quickly apparent. of perceiving, as well as some enhancement. things, such as infra-red wave­ Although they often use ana­ lengths, that it cannot perceive. Image analysis has no exclusive logue rather than digital tech­ The higher the resolution of the connection with remote sens­ nology, home video cameras screen, the better the digital ing. A computer can be used to record information in a similar image will be reconstituted and study and alter the characteris­ way to a satellite instrument/ the easier it will be to recognise tics of any digital image,, pixel by pixel, and the image is or delineate the features that whether it's a satellite view of reconstituted when played make up the image. the Gobi desert or a picture of back to a TV or monitor. Look Marilyn Monroe. However, the closely a t the TV screen and you types of study and alteration can see that the image is made Satellite remote sensing tech­ that might be wanted would be up of lots of tiny elements that nology makes use of digital different in the two cases: in the contairi varying intensities of imagery because it can be former case the purpose of three colours (blue, green and transmitted rapidly and re­ analysis might be to look for red). The video camera has re­ ceived electronically in large certain spectral characteristics corded, for each pixel, the volumes by satellites and that indicate the presence of brightness of each of these ground stations. Because of its water or vegetation, while in the latter the aim could be to improve skin tones, remove an unwanted pimple, and gener­ ate a final image suitable for publication in a glossy maga­ zine. As a result, some specialised image analysis techniques and procedures have been developed that are mainly used in the remote sensing field.

Remote sensing technology

Instruments and satellites As well as optical cameras on aeroplanes and spacecraft, the heavens are full of satellites carrying various types of re­ mote sensing instrument. These transmit data to ground receiv­ ing stations, sometimes via another satellite. The station records the data on tape or other media and from there it is made available to users, some­ times free, sometimes com­ mercially, and often "on a re­ stricted basis. In some cases the amount of data being transmit­ ted is so great that it would cost a fortune for the media needed Remote sensing is observing, making measurements of, and deriv­ to store it all, so unless someone ing information about the earth from a distance... 25 SPC Fisheries Newsletter #57 REMOTE SENSING AND IMAGE ANALYSIS has specifically ordered the both civilian and military. Re­ the ocean or coastal zones, the data, or it is of known useful­ mote sensing comprises only a data they gather, and the satel­ ness, it is not kept. part of the overall use of many lites that carry them. The most satellites, and remote sensing potentially useful for inshore Modern satellites may carry instruments are often consid­ and coastal work are the The­ several instruments which are ered distinct from the satellites matic Mapper (TM), carried on used variously for telephone that carry them. each of the LANDSAT series of communications, data trans­ satellites, and the two High mission, TV transmission, as­ Table 1 shows a list of some Resolution Visible (HRV) in­ tronomy, experimental work, instruments used in remote struments carried on each and for many other purposes, sensing projects that relate to SPOT satellite.

Table 1: Remote sensing satellites and instruments

Satellite Instrument Nominal scene Resolu- Type of radiation size tion detected

LANDSAT 1—3 Multi-spectral 189 km x 189 km 79 m 2 visible Scanner (MSS) 2 near infra-red

LANDSAT 4 — 5 Multi-spectral 189 km x 189 km 83 m 2 visible Scanner (MSS) 2 near infra-red Thematic Mapper (TM) 185kmxl70km 30 m 6 visible 120 m 1 near infra-red

SPOT High Resolution 60 km x 60 km 20 m 3 visible Visible (HRV) 10 m 1 visible

NOAA Polar Advanced Very 2,600 km x 4 km 1 visible Orbiting Operational High Resolution 2,600 km 1.1 km 1 near infra-red Satellites Radiometer (AVHRR) 3 thermal infra-red

NIMBUS-7 Coastal Zone Color 5 visible/ near infra-red Scanner (CZCS)* 1 thermal infra-red

* no longer operational

LANDSAT lites (multi-spectral scanner merged features in coastal ar­ [MSS] and thematic mapper eas. Other light frequencies The American LANDSAT [TM]), the latter is of interest to penetrate water less well than programme has been operating marine workers because of the the blue band: red and near- since the 1970s, although Pacific blue band. TM data comprise infra-red light are quickly coverage in the early days of the information on seven frequen­ attenuated by water, so pen­ programme was poor. cies of radiation collected for etration is poor, and infra-red LANDSAT 4 and 5 are pres­ each pixel sampled. One of radiation is totally absorbed by ently fully operational, and these frequencies, within the less than 1 cm depth of water. LANDSAT 6 is planned for blue part of the visible spec­ launch in 1992. trum, is close to optimum for Unfortunately, historical TM water penetration, and so this - data is not available for much of Of the two instruments carried frequency gives the greatest the Pacific region, since it was on the recent LANDSAT satel- amount of information on sub- out of range of the region's

26 SPC Fisheries Newsletter #57 t

nearest ground receiving sta­ HRV instrument does not have high-resolution colour monitor tion, in Australia. However, a blue band, and therefore can­ (often in addition to the stan­ new ground receiving stations not achieve very good water dard monitor), and will prob­ and relay satellites have been penetration. Of the three fre­ ably also need a large hard disc, deployed and LANDSAT can quencies used in multispectral a graphics card and and, for now provide coverage of the mode—yellow/green, red and hard-copy output, a colour whole region. near infra-red, the first achieves printer or plotter. the best water penetra lion. If the The TM sampling unit, or pixel water is exceptionally clear, Important image analysis pro­ size, is a 30 m x 30 m square penetration may be up to 25 m; cedures for remotely sensed (except for the thermal infra-red 12—15 m is more typical, how­ data include: band, which is 120 m x 120 m). ever , and this may be reduced A full TM scene is 185 x 170 km, further if the water has a heavy — Rectification and correction or about 32,000 sq. km, and sediment load. of images to conform to existing consists of about 3,000 pixels in maps or to fit together into mo­ each direction, or about 9 mil­ saics. Because satellites and air­ lion pixels in total. These di­ Data types, availability and craft are moving when images mensions vary somewhat de­ application are taken, because the earth's pending on the latitude at surface is not flat, and because which the image was taken, and The two main types of practi­ the image may have been taken according to customer needs, cally useful remotely sensed from an oblique angle instead of since various types of sub-scene data for our purposes are: from overhead, there are vari­ can be ordered. TM data are ous corrections that may need good for looking at large scale digital data: these include all to be made to make the image features or fairly extensive computer-compatible data de­ conform to cartographic stan­ coastal areas, but because of rived from remote sensing dards or to topographic reality; their poorer resolution, are less instruments, as well as photo­ useful than SPOT for looking at graphs that have been digitised — Classification of the image small-scale features. using a scanner; according to the characteristics of its pixels. It would normally SPOT hard-copy data: these include be expected that features that prints of aerial photographs, look the same or similar on the The SPOT satellite programme and the colour outputs from ground will have die same or was established in 1986, and digital image processing, such similar spectral characteristics now consists of two operational as thematic maps and charts on the image (in the same way satellites, with SPOT 3 due for and products derived from that two pixels with the same launch in 1992, and SPOT 4 in these. values of blue, red and green on the planning stages. SPOT data a TV screen will give the same offer better resolution than These two data types are dealt colour to the eye). Using com­ LANDSAT, but each image with separately below. puters it is quite easy to classify covers a smaller area, about 60 pixels into related groups based km by 60 km. In multispectral Digital data on their similarities, and then to ("colour') mode, SPOT pixel size count them up to obtain, for is 20 m x 20 m, improving to Digital data are usually deliv­ instance, estimates of the area of 10 m x 10 m in panchromatic ered in a computer-compatible a given type of ground feature; ('black-and-white') mode. At format (tapes or discs) and this latter resolution, small loaded onto a computer hard — Image enhancement (for features such as buildings and disc for study and analysis. example by the production of gardens can be clearly seen, Image analysis packages are false-colour images to emphas­ while in multispectral mode it now commonly available for ise certain features). If different still remains possible to detect micro- and mini-computers. colours are assigned to repre­ small coral patches and general Depending on the sophistica­ sent pixels that have particular features of reefs. tion, the cost may range from a sets of characteristics, different few hundred to a few thousand types of ground features can be The biggest disadvantage of dollars. In addition, in order to made more apparent and dis­ SPOT data for those interested use them effectively, the com­ tinguishable to the eye. By in the coastal zone is that the puter must be equipped with a masking out certain groups of

27^ SPC Fisheries Newsletter #57 REMOTE SENSING AM) IMAGE ANALYSIS pixels, or emphasising their workers, the photographs may working in a single country or boundaries, line maps or con­ not extend beyond the coast area, I would make it a priority tour maps can be produced. into lagoon and inshore waters. to organise my own set, or ac­ •Where they are available, how­ cess to a set, of existing aerial The cost of carrying out in- ever, I believe aerial photo­ photographs of the coastal and house image analysis is now graphs, especially colour ones, inshore area involved. within the means of govern­ are by far the most valuable ment departments and small type of remotely sensed data, The uses of standard aerial companies, and some are start­ and are useful for all kinds of photographs in fisheries are ing to use the technology. The things, especially the detailed similar to the uses of marine real cost, however, and one that planning of field work, as well charts and maps. They are nor­ is often overlooked, is not nec­ as for getting simply a feel for mally consulted when planning essarily in setting up the equip­ the nature of the area. visits by vessels and field ment, but in training the per­ projects, to estimate likely pro­ sonnel to carry out sometimes Aerial photographs are nor­ ductive fishing spots, to identify quite sophisticated analysis mally held by the lands and physical characteristics such as work, and in acquiring data. surveys department or equiva­ passes, areas of coastal access, lent agency in the country watercourses, centres of popu­ Hard copy data concerned. However, I have lation density, etc., all of which found on more than one occa­ are (or should be) used in plan­ Hard copy is needed so that sion that aerial photographs of ning and developing fisheries data can be annotated, exam­ an area in which I am working activities. ined by groups of people, car­ are effectively not available ried into the field, and incorpo­ because they have been bor­ There are some purposes for rated into documents. Addi­ rowed, misplaced or deterio­ which maps and aerial photo­ tionally, most people who are rated in storage. graphs are not commonly used used to working with maps are by fisheries departments in the more comfortable working One could always try to get region. In particular, an instru­ with a paper product derived replacement pho tographs from ment called a planimeter (Fig­ from digital data, than they are the agency that took them in the ure 1), which measures the pe­ with a screen image that may first place. It is normal for these rimeter or area of irregularly require the use of sophis ticated bodies to archive negatives so shaped objects on a flat surface, computer software, to look at that new prints can be made makes it possible to es timate the and manipulate. Producing when needed. Dr Dan Qaasen surface areas of certain marine hard copy by prmting out has recently published a list of features (coral formations of products based on the digital potential sources of aerial pho­ various types, seagrass beds, data mentioned above is an tographs and other data types mangroves, bathymetric con­ essential part of image process­ in the region1. However, I have tours, etc.). This is a laborious ing. Photographic output is never tried to obtain replace­ and time-consuming proce­ better than what can be ment photos, having been de­ dure, especially where there are achieved using a colour printer terred by the potential effort, numerous ground features in or plotter, but is also more ex­ time and cost involved. Work­ small patches, as is often the pensive. ing in a regional body, I tend to case with coral reefs. However, be involved mainly in short- the information obtained is The other major category of term work and do not focus on valuable for resource assess­ hard-copy data is aerial pho­ any one area for more than a ment and management, as it tography. Aerial photographs short time. I therefore have no permits us to quantify the exist, mostly in black-and- need to establish my own col­ habitat of certain resources. If white, in many countries of the lection of aerial photographs we have an idea of the density Pacific Islands region. How­ (nor do I usually have a lot of and distribution of the animals ever, they may be old, the cov­ time before and after a project in their habitat, then we can use erage may be incomplete, and, to devote to following up such this to estimate their likely stock importantly for fisheries things). However, if I were size.

1. Qaasen, D. (1990). A guide to remote sensing data types and sources for the developing countries of the south-west Pacific region. Produced by the Institute of Natural Resources, University of the South Pacific and the ESC AP/UNDP Regional Remote Sensing Programme (UNDP RAS/86/141)

28 SPC Fisheries Newsletter #57 REMOTE SENSING AND IMAGE ANALYSIS

satellite data and actually get­ ting it, so allow plenty of lead time. Many things can inter­ vene, one of the most important of which is cloud. Like LANDSAT, SPOT cannot see through clouds, and only 1. Roller axle passes by once every few days. 2. Plug Depending on how the weather 3. Display is, many weeks may pass 4. Tracer arm without a suitable image being 5. Roller acquired. Competing pro­ gramming requests for the sat­ 6. Operation keys ellite (unlikely over Palmerston 7. Tracer lens but a big problem over devel­ oped countries) can also mean Figure 1: Digital planimeter that if s a long time before it has a chance to look your way. Both these uses are also appli­ Palmerston Atoll survey Having missed our chance in cable to digital data, but the use the field, we at least tried to get of image processing software In 19881 assisted with a survey something useful from the im­ can improve the efficiency with of the marine resources of age. I asked the LATICAL staff which the required information Palmerston atoll in the Cook whether they could derive a can be extracted from the im­ Islands. The other 12 partici­ bathymetric map of the lagoon age. In the former case (plan­ pants in the survey included an from the SPOT data, and the ning), features of interest can be FAO consultant, fisheries staff result of their efforts is shown enhanced by the application of from the Cook Islands Ministry at Figure 2. false colours to emphasise of Marine Resources, the Forum contrast between areas, Fisheries Agency and the Fish­ For comparative purposes, a boundary demarcations, etc. In eries Divisions of Kiribati, bathymetric map based on the latter case, the surface areas Tuvalu and Fiji. As part of that echo-sounding and produced occupied by different ground exercise, we ordered a SPOT by the New Zealand Oceano- features are represented by image of Palmerston. Dr Willy graphic Institute (NZOI) is similar pixels. Once the features Bour of the New Caledonian shown in Figure 3 (this was the of interest have been Image Analysis Laboratory only map of Palmerston we characterised, the pixels that (LATTCAL), based atORSTOM had). The SPOT data indicate represent them can be quickly in Noumea, agreed to carry out the approximate positions tallied by the computer to give some preliminary classifica­ (these were not verified in the a measure of surface area that tions to identify major features field for the reasons mentioned will probably be a good deal of the reef and lagoon for sur­ above) of the 5 m and 10 m more accurate than one derived vey planning purposes. After contours in two colours of pur­ using a planimeter. our visit, when we had had a ple, as shown on the New chance to gather data in the Zealand map. Because of the field, we would look at the rather milky water in the la­ Applications for inshore possibility of doing further goon, SPOT cannot distinguish fisheries work to quantify the areas anything below this depth, but covered by the features we had it does show major coral pin­ How can all these techniques observed on the ground. nacles in the lagoon (the purple and technologies that are avail­ spots). able to us be put to work in the Unfortunately, the image tape inshore fisheries field? To il­ arrived too late for the classifi­ Like the NZOI map of lustrate the possibilities, and cations to be done in time for Palmerston, most standard also some of the problems, I will the survey. As a result, the plans maps and charts do not show describe two examples of work for the data were not fulfilled. all the pinnacles and coral I have been involved in, then Lesson number one: a long time heads that exist in a lagoon. several from other people. can elapse between ordering These are hard to cover using

29 SPC Fisheries Newsletter #57 REMOTE SENSING AND IMAGE ANALYSIS

'5 Xi •;"•*••,': i 9 isKflRj^MT? -"--"Si:. '•••?•- [•ij^UrJr^.&Tf-Sq-"" r?

Figure 2: SrOT unage of Falmerston 30 SPC Fisheries Newsletter #57 *

REMOTE SENSING AND MACE ANALYSIS

nth r—m l*

J MM

X. V

Figure 3: NZOI battiymetric map of Palmeiston

31 SPC Fisheries Newsletter #57 REMOTE SENSING AND IMAGE ANALYSIS standard echo-sounding tech­ rently available, and as such is Finally, we recently received an niques. (Where they are shown, a useful information source for enquiry from Cook Islands to they have usually been added fisheries workers. ask if SPC had any information from aerial photographs.) There on the actual surface areas of may nevertheless be several Returning to the Palmerston coral reef present in that coun­ fishery-related reasons why example, we also attempted to try. While no information of this people want to know where quantify a feature of the reef kind existed at SPC, it was a they are. that appeared to be important relatively simple matter for for the fishery in Palmerston: LATICAL scientist Didier Lille In French Polynesia, the num­ the emergent reef-crest of the to measure the reef surface at ber, type and location of coral eastern side of the atoll that Palmerston. By manually heads in a lagoon is used as one appeared to be a major feeding- masking out non-reef areas, of the factors used in deter­ ground for parrotfish at high and then having the computer mining the number of pearl tide, and where all the count the number of pixels left, oyster farming licences to be parrotfish fishing was carried he made an estimate of reef authorised in that lagoon, since out. That particular zone was down to 2 m deep in about 2 the pinnacles represent the characterised by the presence of hours. (The figure, by the way, major part of the habitat for an encrusting coralline alga, was 2,139 ha.) adult pearl oysters. In Kiribati, Lithothamnion,' which we the national pole-and-line fish­ thought would be distin­ Survey of beche-de-mer re­ ing company Te Mautari at one guishable on a SPOT image. sources in Ha'apai, Tonga point (before it closed down) enquired about obtaining SPOT In fact, there were a couple of In July 1990, a fishery scientist image data of lagoons in problems. The alga is mainly from Papua New Guinea and I Kiribati to help pinpoint good distinguished by the fact that it assisted the staff of the Tonga baiting sites, since its boats contains chlorophyll, which Fisheries Division to carry out baited very dose to pinnacles. gives a characteristic spectral a survey of beche-de-mer re­ signature. Unfortunately, this sources in the Ha'apai group of Another reason for wanting to signature is at the red end of the islands. To plan for the field know about coral pinnacles is spectrum, and red light does work we had three British Ad­ navigation. The information in not penetrate water well. As it miralty marine charts, which many marine charts is incom­ happens, the tide must have cover the whole area of interest. plete or dated and dangers to been fairly high when the sat­ boats are not shown in full. ellite took that particular image The charts were used to decide Remotely sensed data are a of Palmerston. Most of the alga the areas in which we would quick, easy way to gather was submerged, and thus in­ dive to gather field data on sea supplementary information on visible. There is a strong line of cucumber abundance, and the shallow water features for ad­ breakers, which, being pure selected sites were then checked dition to maps and charts. This white, reflect very brightly, and with an echo-sounder and if is also considerably cheaper prevent observation of what is necessary changed before we than sending a survey ship to underneath. And, where the went in the water. Standard resurvey an area, or flying alga does emerge, its spectral navigational equipment (com­ planes around to take aerial signature seems to be quite pass, sounder and rangefinder) photographs. similar to that of the vegetation was used to determine our po­ on the small islets around the sitions, which were then plotted The Polynesian Remote Sensing atoll. on the charts. Station (SPT) in Tahiti is pres­ ently using just this approach to It may be possible to solve this During the 6 days of actual field improving maps of the 120 latter difficulty by further work, we completed 45 SCUBA atolls of the Territory, many of ground-truthing of the image, dives and sampled transects in which are not very compre­ but it could be a while before a way that allowed us to esti­ hensive. A sample of one of anyone is able to go to mate the densities of commer­ SPT's maps is shown at Fig­ Palmerston to do that. In the cially important sea cucumbers. ure 4: while not strictly a fish­ meantime, these problems There was a fairly dear depth ery-related product, it repre­ combined to prevent LATICAL stratification for the most im­ sents an improvement on the from estimating the surface of portant spedes which, together marine charts and maps cur- the emerged eastern reef edge. with the limits on individual

32 SPC Fisheries Newsletter #57 e REMOTE SENSING AND IMAGE ANALYSIS

Figure 4: SPT map of Marrihi, incorporating remotely sensed data

33 SPC Fisheries Newsletter #57 REMOTE SENSING AM) IMAGE ANALYSIS

3B ci •M, we Sftr^2 a ™ * 5 » KVIK 41 24 35 u ""n Sin -*>--,_..--3' a

*<=** AVA AHANOA ME -|^>

"1, /rf" v^r#

» 31

•-JH.' n M

W ^ —* mj 1BO

vta in/

™ „••-- •*• / « " ® « ^ fo *f « / -,*" " JJ ' fV-jV x ™ "«, v ./ ?sz& „ a ?*/ jF

Lrid R"fef

« /•;;-; 1 ? «,'"S =h-6--._ 5U ! JO (&•• .; •--1(>0 :» ; V3 j Id

11 aj

so as \ Onir*"? fes^^^#^

— •™ • -'- .H-'Jan i

^..- -2»V .." Q^

JZ2

Figure 5: British Admiralty marine chart of Uiha island in Tonga's Ha'apai group fishermen's ability to free-dive, In November 1990 I found out 30 m contour around Uiha, ex­ effectively meant that the fish­ that the Polynesian Remote cluding the land and emergent ery would be confined to a Sensing Station (SPT) in Tahiti reef surface, at 4,575 ha (1 ha = maximum depth of about 30 m. had a SPOT image that covered 100 x 100 m, or 10,000 m2). Using the Admiralty charts, we part of the Ha'apai area, in­ therefore estimated the extent cluding the island of Uiha as Dr Loubersac subsequently of the 30 m depth contour shown on the chart in Figure 5. processed that part of the image around each island and reef in I discussed with station director covering Uiha to produce the the area, then used a planime- Dr Lionel Loubersac the possi­ bathymetric contour maps ter to measure the area within bility of using this data to repeat shown in Figure 6, and in do­ each contour. This process was the work for the Uiha area that ing so estimated the surface laborious, but allowed us to es­ we had carried out manually, area within the contours by timate commercial beche-de- and comparing the results with counting the pixels. Judging mer abundance at about 1.01 our own. Our manual estima­ from the marine charts, the million animals. tion had put the extent of the SPOT contours appear to cor-

34 SPC Fisheries Newsletter #57 REMOTE SENSING AND IMAGE ANALYSIS

*

Figure 6: (a) SPOT image of Uiha island and surrounding area, and (b) bathymetric contour map derived from it 35 SPC Fisheries Newsletter #57 REMOTE SENSING AND IMAGE ANALYSIS respond roughly to the 10 m areas of coastal land suitable for cessing of the data from areas and 20 m isobaths, but this can- prawn culture and other forms where gravitational anomalies no t be verified without some of aquaculture in New Cale­ have been detected permits the field-testing ('ground-truthing'). donia. Dr Willy BOUT'S in­ local bathymetry to be deter­ Light penetration in the fre­ volvement in remote sensing mined. Finally, the assignation quencies used by SPOT is rarely began with his carrying out a of false colours to these areas good enough for bottom fea­ project to estimate the area of makes thebathymetric features tures much beyond this depth trochus habitat on Tetembia stand out: the deeper the red, to be distinguished. Within the reef in New Caledonia. Dr the shallower the seamount. apparent contours, the estim­ Pascale Joannot, of the Noumea ated surface areas are: 0—10 m: aquarium, has recently been A few features of Figure 8 are 2,372 ha; 10—20 m: 1,394 ha; involved with Dr Bour in a worth noting. The white con­ total (0—20 m): 3,766 ha, similar project to use SPOT data tour lines shown in the figure to estimate the standing stock of are not derived from the satel­ The whole process took Dr corals of the family Fayidae in lite data, but are superimposed Loubersac less than one after­ the same area (these corals are on the image from a world ba- noon and gave, I feel, much commercially exploited in New thymetric map, maintained as a more accurate area estimates. Caledonia). LATICAL has also computerised database by the SPT could probably estimate assisted Dr Pierre Thollot of U.S. Defence Mapping Agency. the areas of the other islands, in ORSTOM by making an esti­ Thismap essentially represents another afternoon's work — mate of the extent of different the tbtal of what was previously about a fifth of the time it took types of mangrove in New known about the bathymetry of me to do it manually, and again Caledonia. Recently, as a public the area in question. At least a with better results, as I was us­ information exercise, the re­ dozen seamounts or sea floor ing estimated contours and an gional government of New features that are no t on the map instrument, the planimeter, that Caledonia's Southern Province are shown in the satellite image is very sensitive to operator issued a poster of the SPOT The seamount shown on the inaccuracy. composite image, shown in map at 24.2°S, 201.2°E does not Figure 7, which illustrates the appear on the sa tellite image— If I had known beforehand that areas of existing marine re­ i.e. it does not exist. In fact the SPOT imagery was available for serves in the Territory. seamount shown on the map is even part of the area, I would probably a seamount located have taken some printed out­ A somewhat different applica­ two degrees further east, and puts, of the type shown in Fig­ tion is demonstrated in Figure the error in recording its posi­ ure 6, with me. By measuring 8, which shows a false colour tion dates from the beginning of the depth accurately at known image produced by Dr Nicolas the century. The group of sea­ points on the printed image, we Baudry and his colleagues us­ mounts around 24°S, 212°E could have calibrated the depth ing data from a radar instru­ show several mapping inaccu­ contours more precisely than ment carried on the SEASAT racies. Modelling the bathym­ has been possible from the satellite. The instrument, which, etry of these seamounts using Admiralty charts, and refined unlike HRV, MSS, TM and the satellite data allows their the area estimates still further. others, is not passive but active positions to be corrected to (it generates its own signal), is within 5 km — better than the Other examples able to measure the exact dis­ resolution of most of the map­ tance between the satellite itself ping surveys used to compile There are several other ex­ and the sea surface. After pro­ the global bathymetric database amples from the region of digi­ cessing to remove the effects of — and their depths to be esti­ tal image analysis being used in wave height, what remain are mated to within 10 per cent of fisheries projects, or having an irregularities in the average the true depth. obvious interest to fisheries. In height of the surface of the sea. the early 1980s, workers in These deviations from the av­ There is a clear application for Papua New Guinea were using erage are caused by variations this information in assessing, LANDSAT imagery to make in the earth's gravitational field, developing, and even finding, area estimates for different and the variations are in turn deep-bottom fishing stocks in types of reef around Kavieng. A associated with the presence of some countries of the region. pilot project for SPOT, also in underwater features like sea- With the vastlyimproved radar the early 1980s, was to identify mounts. More detailed pro­ data coverage that will become

36 SPC Fisheries Newsletter #57 REMOTE SENSING AND IMAGE ANALYSIS

Figure 7: Poster publicising marine reserves in New Caledonia, composed of a mosaic of 26 SPOT images 37 SPC Fisheries Newsletter #57 REMOTE SENSING AND IMAGE ANALYSIS

Figure 8: Seamounts in French Polynesia, mapped using radar satellite data (Photo courtesy of Seafloor Imaging, Inc, Noumea, New Caledonia) 38 SPC Fisheries Newsletter #57 REMOTE SENSING AND IMAGE ANALYSIS available from the ERS-1 and The UNESCO book on coral with a 1/1000 second strobe TOPEX/ POSEIDON satellites reef research methods2 de­ shutter) under a light aircraft, as of late 1991, satellite-based scribes a survey of coral reefs and thereby take an airborne sea floor ba thymetric modelling from above using an ordinary video sequence of a long stretch will be easier and more precise camera suspended from a of beach that he was studying5. throughout the world's oceans. small, helium-filled balloon Afterwards, the video machine anchored on the reef flat. In the was hooked up to an IBM PC Philippines, Dr John McManus equipped with a graphics card Conclusions used some of his research bud­ and a video frame grabber, and get to buy a second-hand ultra­ the tape stopped when the de­ People are often both im­ light aircraft (cost: about 3 sired part of the beach was in pressed by the beautiful images US$5,000) . He then fitted his view. The frame grabber then produced by image analysis of pocket camera to the bottom of converted the scene to a digital remo tely sensed da ta, and at the the plane and, in calm weather, image amenable to computer- same time intimidated by the overflew the areas he was in­ based image processing meth­ sophisticated technology in­ terested in, taking pictures to ods. volved. There is no doubt that make up a mosaic. In this way, the use of digital image tech­ in a short time and for the ad­ These examples illustrate that, nology requires costly equip­ ditional cost of a few litres of as well as the conventional ment, costly data and even fuel and a roll of film, he was sources, there are many more costly training, and tha t in able to get a quick measure of unconventional ways of ob­ some cases, the production of the number of boats out fishing taining remotely sensed data, pretty pictures from satellite that he would never have been often at low cost, and these are data has been more aestheti­ able to get by conventional sur­ often limited only by imagina­ cally pleasing than practically veillance on the ground. Most tion and perseverance. As the useful. Hopefully, as people people, however, acquire their availability of remotely sensed come to realise that remotely data by the more conventional information grows, and tech­ sensed data really are just data method of buying it from the nical fishery people become — with associated advantages local lands department. more familiar with its potential and limitations like any other applications, this kind of data is data set — the use of the asso­ The use of digital technology sure to be used more and more ciated technology will lose its allows more information to be in the development and man­ mystique and start to conform extracted from an aerial photo­ agement of the region's marine more to real needs than to graph than can be seen with the resources. >a^ imaginary possibilities. naked eye. Drs David Hopley and Pauline Ca tt of James Cook It should not be forgotten that University have for some time simple visual interpretation of been using digitised infra-red aerial photography is still aerial photographs taken from probably the most useful form a low-flying light aircraft to of remote sensing for the monitor coral growth and de­ 4 region's workers on inshore velopment in Queensland . In fisheries. There have been some Scotland, Robert Wright was innovative approaches to ac­ able to mount an ordinary quiring aerial photographs by home video camera (equipped fisheries workers.

2. Stoddart, D.R. and R.E. Johannes (eds) (1978). Coral reefs: Research methods. UNESCO monographs on oceanographic methodology No 5. United Nations Educational, Scientific and Cultural Organization, 7 Place de Fontenoy, 75007 Paris, France. 3. McManus, J. (1988). Pcre. comm. 4. Hopley, D. and P. Cart (1988). Use of near infra-red aerial photography for monitoring ecological changes to coral reef flats on the Great Barrier Reef. Proceedings of the 6th International Coral Reef Symposium, Townsville, Australia, August 1988, 5. Wright, R. and T.A. Harris (1988).Change-detection in physically fragile coastal areas of Scotland by remote sensing. Paper presented at a conference on 'Remote sensing of the Coastal Zone', Gold Coast, Australia, September 1988.

39 SPC Fisheries Newsletter #57 \