UBC Faculty Research and Publications

Fisheries catch reconstructions: islands, part II Harper, Sarah; Zeller, Dirk 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-FCRR_2011_19-4.pdf [ 1.87MB ]
JSON: 52383-1.0074737.json
JSON-LD: 52383-1.0074737-ld.json
RDF/XML (Pretty): 52383-1.0074737-rdf.xml
RDF/JSON: 52383-1.0074737-rdf.json
Turtle: 52383-1.0074737-turtle.txt
N-Triples: 52383-1.0074737-rdf-ntriples.txt
Original Record: 52383-1.0074737-source.json
Full Text

Full Text

 ISSN 1198-6727  Fisheries Centre Research Reports  2011 Volume 19 Number 4    FISHERIES CATCH RECONSTRUCTIONS: ISLANDS, PART II      Fisheries Centre, University of British Columbia, Canada  FISHERIES CATCH RECONSTRUCTIONS: ISLANDS, PART II       Edited by  Sarah Harper and Dirk Zeller                 Fisheries Centre Research Reports 19(4) 143 pages © published 2011 by  The Fisheries Centre, University of British Columbia 2202 Main Mall Vancouver, B.C., Canada, V6T 1Z4       ISSN 1198-6727  Fisheries Centre Research Reports 19(4) 2011  FISHERIES CATCH RECONSTRUCTIONS: ISLANDS, PART II  Edited by  Sarah Harper and Dirk Zeller    CONTENTS  Director‘s foreword ............................................................................................................................................... 1 Preliminary estimate of total marine fisheries catches in Corsica, France (1950-2008) ................................. 3 A brief history of fishing in the Kerguelen Islands, France .............................................................................. 15 Reconstruction of total marine fisheries catches for Madagascar (1950-2008) ............................................. 21 Reconstruction of marine fisheries catches for Mauritius and its outer islands, 1950-2008 ....................... 39 Reconstruction of Nauru‘s fisheries catches: 1950-2008 ............................................................................... 63 Marine fisheries of Palau, 1950-2008: total reconstructed catch .................................................................... 73 Reconstruction of Sri Lanka‘s fisheries catches: 1950-2008 .......................................................................... 85 From local to global: a catch reconstruction of Taiwan‘s fisheries from 1950-2007 ...................................... 97 Reconstruction of Fisheries Catches for Tokelau (1950-2009) ..................................................................... 107 Reconstructing marine fisheries catches for the Kingdom of Tonga: 1950-2007 ........................................ 119 Reconstruction of marine fisheries catches for Tuvalu (1950-2009) ............................................................ 131      A Research Report from the Fisheries Centre at UBC     Fisheries Centre Research Reports 19(4) 143 pages © Fisheries Centre, University of British Columbia, 2011  FISHERIES CENTRE RESEARCH REPORTS ARE ABSTRACTED IN THE FAO AQUATIC SCIENCES AND FISHERIES ABSTRACTS (ASFA) ISSN 1198-6727  Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  1  DIRECTOR‘S FOREWORD  As marine fisheries resources around the world are increasingly threatened by pollution, climate change, and overfishing, it is more important than ever to know the amount and types of fish and invertebrates being extracted from the marine environment. Fisheries resources, particularly for island countries, provide a crucial source of food and income. However, the very fisheries which depend on these natural goods and services—notably small-scale fisheries—are being under-represented in fisheries statistics. In many of the countries highlighted in this report, the majority of seafood consumed is taken via subsistence fisheries. This non-commercial fishing sector is largely overlooked in statistical collection systems, particularly those of several decades ago, but continues to be under-represented today. In some places this is beginning to change, as the importance of small-scale fisheries to national food security is being recognized. In many developing countries, which lack the infrastructure and resources to fish their own waters for economic development through trade with external markets, foreign access fees are collected as a key source of revenue. In exchange for a modest fee, foreign fleets are allowed to fish their waters for high valued species. While this provides much needed income for the country, it also threatens the availability of these resources for domestic sustenance. While there is a range in the quality of fisheries reporting from one country to the next, in almost all countries Illegal, Unreported and Unregulated (IUU) fisheries exist. Fisheries landings statistics, as supplied to the United Nation‘s Food and Agriculture Organization (FAO), represent mainly the commercial and larger-scale fisheries. Artisanal, subsistence and recreational fisheries are mostly overlooked. Discarded bycatch and baitfish associated with certain fishing techniques are also rarely included in the official statistics. As a follow up to Fisheries catch reconstructions: Islands, Part I, this report continues to reconstruct total marine fisheries catches of island countries around the world from 1950 to present. This edition describes fisheries for island countries in the Pacific, Indian and Atlantic Oceans, highlighting the discrepancies that exist between reported landings and likely true catches. The reconstruction approach used here, as in the previous edition, aims to estimate all marine fisheries extractions as a baseline for monitoring and management purposes in the face of continued anthropogenic pressures. The future success of these countries relies, in part, on their ability to keep pace with an increasingly gl0bal economy while maintinaing a healthy supply of resources for domestic purposes. Ussif Rashid Sumaila, Director UBC Fisheries Centre August 2011   2 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  3   PRELIMINARY ESTIMATE OF TOTAL MARINE FISHERIES CATCHES IN CORSICA, FRANCE (1950-2008)1  Frédéric Le Manacha, Delphine Durab, Anthony Perecd, Jean-Jacques Riutorte, Pierre Lejeunec, Marie-Catherine Santonif, Jean-Michel Culiolif, and Daniel Paulyg  aFaculty of Science and Technology, University of Plymouth, Drake Circus, Plymouth PL4 8AA, United Kingdom; flm@livinblue.com  bInstitut Polytechnique LaSalle Beauvais, 19 rue Pierre Waguet, BP 30313, Beauvais 60026 cedex, France; delphinedura@gmail.com cStation de Recherches Sous-marines et Océanographiques, Pointe Revellata, BP33, Calvi 20260, France; a.pere@stareso.com; pierre.lejeune@stareso.com dFaculté des Sciences et Techniques, Université de Corse Pascal Paoli, 7 avenue Jean Nicoli, BP 52, Corte 20250, France eBastia Offshore Fishing Association, 8 Parc Impérial, Furiani 20600, France;  jjriutort@gmail.com fOffice de l’Environnement Corse, département "Parc Marin International des Bouches de Bonifacio", base technique de La Rondinara, BP 507, Ajaccio 20176, France; santoni@oec.fr, culioli@oec.fr gSea Around Us Project, Fisheries Centre, University of British Columbia, 2202 Main Mall, Vancouver V6T 1Z4, Canada, d.pauly@fisheries.ubc.ca  ABSTRACT  Corsica is an island in the Western Mediterranean belonging to France, located southeast of the French mainland and west of Italy. The island covers an area of about 8,700 km2, is flanked by deep water along its west coast, and by a broad shelf along its east coast. Corsica has fisheries in its coastal lagoons, but its most commercially important fishery is the red spiny lobster (Palinurus elephas) fishery, followed by bottom trawling for finfish. Other smaller and poorly documented artisanal and recreational fisheries also occur, but overall fishing pressure appears to be low, and the number of full time fishers is declining. The total reconstructed catch from 1950 to 2008 was 118,700 tonnes - 5 times more than the 23,700 tonnes reported by France to FAO – of which 30% was unreported recreational catch by locals or tourists, 37% was bottom-trawl catch, 10% was associated bycatch (unreported, landed or discarded), and 23% was red spiny lobster and pelagic catches. The estimated mean annual catch in the 21st century was 1,300 tonnes. Field investigations are needed to improve on these data, presented here as a first approximation of total extractions from the waters surrounding Corsica.   INTRODUCTION  Corsica is the fourth largest island of the Mediterranean and a part of France. It is located southeast of the French mainland and north of Sardinia (Italy), and west of the Italian Peninsula (42° N and 9° E; Figure 1). Corsica is characterized by a mountainous landscape and a highly disparate underwater morphology, featuring a steep descent to depth along the western part of the island (down to 3,000 m, 10 km offshore). In contrast, wide expanses of shallow waters are present along the east coast, where a depth of only 150 m has been recorded 11 km offshore, and several lagoons important for the Island‘s marine fisheries are also found along the east coast (Riutort, 1994).   1  Cite as: Le Manach, F., Dura, D., Pere, A., Riutort, J.J., Lejeune, P., Santoni, M.C., Culioli, J.M., and Pauly, D. (2011) Preliminary estimate of total marine fisheries catches in Corsica, France (1950-2008). pp. 3-14. In: Harper, S. and Zeller, D. (eds.) Fisheries catch reconstructions: Islands, Part II. Fisheries Centre Research Reports 19(4). Fisheries Centre, University of British Columbia [ISSN 1198-6727].  4 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Corsican waters host numerous fish and invertebrate species (de Caraffa, 1929; Miniconi, 1989, 2001) and valuable habitats (e.g., meadows of seagrass Posidonia oceanica), of which most are protected under European Commission directives or national legislations (Anon., 1975; 1979, 1981, 1987, 1992, 1994, 1999) (Figure 1). The Réserve Naturelle des Bouches de Bonifacio is the largest marine protected area (MPA) of metropolitan France, covering approximately 800 km² (Figure 1), including a 13 km² no- take zone and 130 km² with restricted fisheries activities. Regulations and monitoring seem to be effective, as increasing catches have been reported in and around this MPA (Santoni and Culioli, unpub. data)  Since the 1950s, tourists have been attracted to Corsica for its natural beauty and pristine habitats, and the tourist population currently reaches 3 million per year (Anon., 2010), with a resident Corsican population of less than 300,000. The tourism industry has a major impact on seafood consumption and hence on marine resources, as already highlighted in the 1960s (Maurin, 1965). Currently, numerous hotels and/or charter companies offer recreational fishing opportunities, separate or in combination with consumption of local seafood.  Despite its potential attractiveness for fishers, the waters around Corsica have never experienced heavy industrial fishing pressure, and the history of Corsican resource extraction was shaped more by land-based than maritime activities. Therefore, there is almost no export of seafood out of Corsica, and a substantial fraction of the seafood consumed locally by Corsicans is imported from the French mainland or other Mediterranean countries. Currently, the number of professional fishers is declining, and Corsica likely experiences the lowest commercial fishing pressure in the Mediterranean Sea (Riutort, 1994; Relini et al., 1999). As a consequence, fisheries have generally not received much attention, and quantitative analyses of fisheries are scarce, except for the high-profile fishery for red spiny lobster (Palinurus elephas) (Pere et al., 2007; 2010) and for MPA fisheries (Rigo, 2000; Santoni, 2002; Mouillot et al., 2007; Rocklin et al., 2009).  Corsica - via France - has only supplied fisheries statistics to the Food and Agriculture Organization of the United Nations FAO since 1970. This study therefore aims to reconstruct Corsican fisheries catches back to 1950, while ensuring that all extractions due to fishing are considered, following the catch reconstruction approach of Zeller and Pauly (2007). Like most countries in the Mediterranean, France has not declared a formal Exclusive Economic Zones for its Mediterranean coast (EEZ; Anon., 1976; Santoni, 2002; Cacaud, 2005). Hense, our estimates of historical catches for the period 1950 to 2008 are deemed to have come from Corsican EEZ-equivalent waters.  Figure 1: Map of Corsica and its territorial waters (solid black line). Marine Protected Areas (blue areas and grey solid dots) are designated at a state level and aim to protect both habitats and wildlife by controlling, or even excluding human activities (no-take zones). Natura2000 zones (blue stripes) are designated at a European level and aim to protect both habitats and wildlife, without excluding human activities. The level of protection in the Natura2000 zones is lower than for MPAs. Source: www.affaires-maritimes.mediterranee.gov.fr Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  5    MATERIALS AND METHODS  Baseline data were extracted from the General Fisheries Commission for the Mediterranean (GFCM) section of the FAO FishStat database (FAO, 2009). As Corsica is remote from the French mainland, we assumed that all catches reported by France within the ‗Sardinia‘ FAO fishing area (Division 37.1.3) were Corsican. A bibliographical review of all Corsican fisheries was done to identify the ‗anchor points‘ required for inferences on historical catches back to 1950 (Zeller and Pauly, 2007). Data sources included peer-reviewed scientific articles, reports by local institutions, theses and other unpublished accounts, and local expert knowledge.  Total Corsican population  Population statistics were extracted from the National Institute of Statistics and Economic Studies (INSEE, www.insee.fr/fr/themes/theme.asp?theme=2&sous_theme=1&nivgeo=6&type=3 [accessed: October 15, 2010]). Population data were used here to indirectly estimate total catches by local residents (see ‗recreational fisheries: residents and tourists‘ sub-section; Figure 2a).  Fishers and fishing vessels in Corsica  The time-series of the population of fishers was obtained from Riutort (1994), and linear interpolations were applied between anchor points for years without data (Figure 2b). The number of fishers after the last anchor point (1993) was calculated by applying the trend in the number of fishers per vessel during the period 1950-1993 to the number of vessels for the period 1994-2008. It is worth noting that these vessels are usually smaller than 15 m, and operate close to shore (Miniconi, 1994; Riutort, 1994; Rigo, 2000; Santoni, 2002). The fishing industry in Corsica is therefore more artisanal than industrial, with small vessels (Riutort, 1994), short periods at sea, and a small supply chain (Riutort, 1989). The two time- series in Figure 2b were used to estimate bottom-trawl catches for the 1950-1970 period (see ‗artisanal demersal fishery‘ sub-section).  Lobster fishery  Red spiny lobster (Palinurus elephas) is mainly exploited along the west coast of Corsica, where its preferred hard-bottom habitats are found. The fishery for lobster is relatively small, and vessels stay close to the coast, fishing at depths not exceeding 200 m (Marin, 1987). The fishery was profitable very early on (de Caraffa, 1929), but it is not well documented. Thus, official catch data are deemed inaccurate, and is at best a refelction of trends (Marin, 1987). Here, we attempt to re-estimate lobster catches for the entire 1950-2008 period using various sources of information.  Figure 2. Basic statistics on Corsica: a) total resident Corsican population and b) trends in the number of fishers and vessels in Corsica. Anchor points are represented by closed circles. a) b) 6 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Statistics were extracted from the Office de l’Environnement Corse (2010) for the 1950-1983 period. The early values (1950s) are in accordance with catches at the beginning of the 20th century, i.e., around 300 t·year-1 (Doumenge, 1956). For the 1983-2008 period, values were extracted from studies by Riutort (1999), Marin (1987) and Pere et al. (2010).  Changes in gears had an influence on both catch per unit of effort (CPUE) and discard rates (non- marketable lobsters). In the early 1960s, the traps in common use were replaced by trammelnets (Miniconi, 1989), which had higher CPUE, but also generated a higher discarding rate. Discards for traps were estimated to be about 5% during the 1950-1964 period (Riutort, unpub. data). During the 1965-1980 period, fishers were using trammelnets for short trips, and a discard rate of 12.5% was therefore used (Riutort, unpub. data). For the 1981-1994 period, the mean value of 15.7% between the 1965-1980 period (12.5%) and the 1995-1999 period (20%) was used (Riutort, 1999; Pere et al., 2010). For the 2000-2003 period, the same value of 12.5% was used. For the 2004-2007 period, Pere et al. (2010) estimated a discarding rate of 11.4%, which was also used for the year 2008.  It is worth noting here that two types of trammelnets are used in Corsica, to target either demersal fish (since before 1950) or red spiny lobster (since the early 1960s). Even when ‗lobster trammelnets‘ are used, a considerable amount of the bycatch is fish. Thus, 55% of total catch were fish species in 2008 (Riutort, 1989; 1994; Santoni and Culioli, unpub. data). This bycatch of fish is retained and landed, and was included in the next sub-section (artisanal demersal fisheries).  Artisanal demersal fisheries  Demersal species are caught in Corsican waters with two types of gears: trammelnets and bottom- trawlers. Trammelnets have been in use for demersal fish for a long time (prior to 1950), while we assumed bottom-trawlers were introduced in the early 1950s (Riutort, 1994). Catches by trammelnets may represent 50% of total fish catches in the province of Bonifacio (Santoni and Culioli, unpub. data), and given that no other studies were available, we used this 50% ratio for the 1965-2008 time-period and the entire island. Thus, the remaining 50% of demersal fish catches were treated as caught by bottom- trawlers as of 1965. For 1950, we set bottom-trawl fish catches as zero, and interpolated linearly to 1965. FAO FishStat contains data on demersal species for the period 1970-1992 only. In the absence of any alternative, we considered these data to be realistic. Indeed, none of the documents available on Corsican fisheries allowed us to make an independent estimate of the bottom-trawl and trammelnet fisheries catches.  Catches per fisher and catches per vessel (CPUE) for the period 1970-1992 were calculated by dividing catches of bottom-dwelling fish reported to FAO by the number of fishers or vessels (Figure 2b). CPUE for the 1950-1970 and 1993-2008 periods were then estimated by extrapolation of the trends of 1970-1992 CPUE time-series. The resulting CPUE data for the 1950-1970 and 1993-2008 periods were then multiplied by the number of fishers or vessels (Figure 2). Our estimate of total catch used the average values of these two catch time-series (one based on CPUE per fisher, one on CPUE per boat), which was then split evenly to create the bottom-trawl and trammelnets components.  Trammelnet fishery  The taxonomic breakdown of trammelnet commercial species in the Bonifacio MPA (Figure 1) was studied by Mouillot et al. (2007) from 2000 to 2006. Given that there were no other studies available that included a taxonomic breakdown, we assumed that the percentage of each species remained the same for the entire 1950-2008 time-period, and were similar for the entire island. A recent study concluded that trammelnet discards were representing approximately 10% of total catches, in the MPA of Bonifacio (Rocklin et al., 2009). These discards are composed of damaged, non-marketable fish. We used this study to estimate the taxonomic breakdown of these discards.     Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  7   Bottom-trawl fishery  We assumed that the species composition of landed bottom-trawl catches were similar to the trammelnet fishery. However, we acknowledge that the species composition can vary significantly between the continental shelf and the slope. For higher depths (on the slope), many other species such as Nephrops norvegicus, Etmopterus spinax, Galeus melastomus, Merluccius merluccius or Trigla lyra can indeed account for a large part of the catch (Riutort, pers. obs.; Le Manach, pers. obs.). However, as it was not possible to estimate the percentage of each species or the importance of slope bottom-trawling, we did not take these observations into account. To estimate the bycatch by the artisanal bottom-trawl fishery, we used a bycatch rate of 40%, given by Machias et al. (2001) and Sanchez et al. (2004) for geographically close and similar fisheries. The MEDITS database (Bertrand et al., 1998) - 2009 update - was used to estimate the bycatch taxonomic breakdown. We assumed that non-commercial species occurring in this database were bycatch species, e.g., Spicara spp., Scyliorhinus spp., Raja spp., Micromesistius poutassou, Capros aper. Furthermore, as fishers land a portion of non-targeted bycatch, notably for their personal consumption and soupe de roche (‗rockfish soup‘), we conservatively assumed that 20% of the bycatch was landed, but unreported, and that the remainder (80%) was discarded.  Recreational fisheries: residents and tourists  To estimate recreational catches by Corsicans, we used a ‗Fermi solution‘, i.e., an approach pioneered by the physicist Enrico Fermi, to estimate unknown quantities from limited data (von Baeyer, 1993; Pauly, 2010). Thus, based on local knowledge (Culioli, pers. obs.; Riutort, pers. obs.), we estimated three anchor points, for 1950, 1980 and 2008. For 1980, we assumed that 30% of the total population, i.e., 76,000 out of 255,000 inhabitants, was potentially recreational fishers. Of these potential fishers, we assumed that 15% were actually fishers, and that they were on average fishing once a month, with yields of 4 kg per trip. For 2008, we used the same assumption that 30% of the total population, i.e., 84,000 out of 280,000 inhabitants, were potentially recreational fishers, but that the proportion of actual fishers increased to 25%. As local residents report that there are less fish now than in the 1980s, we assumed that fishers currently fish on average only 10 times a year, with yields of 1 kg per trip. For 1950, we assumed a stable CPUE and fishing effort compared to 1980, and derived total catches from the total population size.  Similarly, our estimate of recreational catches by tourists was based on the annual number of tourists, and assuming that sport fishing became more attractive in the 1990s. We conservatively assumed that 5% of tourists were catching on average 1 kg·year-1 for the 1950-1990 period, and that 8% of tourists were catching on average 1.5 kg·year-1 for the 1991-2008 period. Given that each tourist currently stays on average 10.3 days in Corsica, these assumptions seem reasonable (Anon., 2010).  Pelagic fisheries  Three pelagic fisheries are taking place in Corsican waters. However, information is scarce and no studies enabled us to re-estimate their total catches. Therefore, we included data as provided to FAO in our total reconstruction (except for small pelagics – see below).  Swordfish Swordfish (Xiphias gladius) started to be targeted by artisanal longliners in the 1980s (Regional Committee of Corsican Marine Fisheries, 2009; Riutort, unpub. data). However, as tonnages are likely small (15-20 t·year-1; Riutort, unpub. data), it is possible that these catches are accounted for in official FAO statistics as ‗marine fish nei‘ (FAO, 2009).  Small pelagics  ‗Blue fish‘ (i.e., sardines, anchovies and mackerels) are also fished along the Corsican coast. Several studies report substantial catches during the 1960s and 1970s in Corsica and along the French mainland (Maurin, 1965; Bonnet, 1973; Pichot and Aldebert, 1978). It is worth noting that FAO data include sardine statistics only for 1972-1976 and 2006; data for other years being either non-existent or unrealistically 8 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   low. However, older Corsican residents remember very abundant sardine and anchovy catches during the 1950-1960s, most of which were exported to the mainland (Riutort, unpub. data). For the period 1950- 1971, we therefore used the average catches for the period 1972-1976, and kept the rest of the time-period unchanged.  Tuna  Maurin (1965) reports 100 tonnes of tuna caught in 1963 by Corsicans. However, he suggests that the tourism industry already accounted for a significant part of unreported catches, although he did not elaborate on this topic. Tuna may also be reported to FAO as ‗marine fish nei‘, and annual catches are likely very low or up to 15 tonnes (Riutort, unpub. data).  RESULTS AND DISCUSSION  Lobster fishery  As expected, our lobster catch reconstruction is very different from official statistics: our values are on average 16 times higher than data provided to FAO, and show a very different pattern over time (Figure 3). Lobster catches decreased from 300 t·year-1 in 1954 to 80 t·year-1 for 1959-1960. Then, catches increased again to 300 t·year-1 by 1962. At this time, a new (unspecified) crash occurred and catches dropped to 100 tonnes annually, staying at that level until the late 1970s. By 1984, catches increased to 250 t·year-1. Since then, catches have been decreasing, reaching 80 t·year-1 by the early 2000s. However, it is worth noting that catches currently seem to be increasing. Overall, the trend of the number of fishers and vessels (Figure 2b), and lobster catches (Figure 3) show a similar pattern, which confirms that this fishery is of great importance in Corsica and largely accounts for much of the fishing pressure. Fluctuations in lobster catches may be partly explained by new policies and gear modifications, along with biological features (e.g., larval migration; Pere et al., 2011). The first crash in the 1950s likely resulted from increasing fishing pressure, and the following increase in catches is likely the result of a gear change from traps to trammelnets, which increased the CPUE. In 1968, policy- makers decided to close the lobster fishery between the 1st of October and 28th of February each year, which probably played a significant role in the stabilization of catches during the 1970s. Finally, new vessels were introduced in the early 1980s and were responsible for increased fishing effort. This increase is likely to have contributed to the increase in catches until 1984, then to the decrease in catches observed until the mid- 2000s.  Figure 3: Reconstructed catches of lobsters and associated discards for Corsica, 1950-2008. The dotted line represents official lobster landings data supplied to FAO. Figure 4: Reconstructed catches and associated bycatch (landed and discarded) for the demersal finfish fishery by the two main gears (trammelnet and bottom-trawler) in Corsica, 1950-2008. The dotted line represents official landings data supplied to FAO. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  9   In the late 1970s, nine small marine protected areas (some of them no longer existing; Figure 1) were created, which also possibly led to the catch increases in the late 1970s early 1980s (Figure 3). The current decrease, which started around 1984, may be due to several factors, such as the decreasing number of fishers, or increasing fishing effort. It seems there is an increase in catches during the last few years. Such an increase could be due to biological parameters (e.g., larval migration), but no data were available to assess the validity of this assumption.  Artisanal demersal fishery  Demersal catches totaled an estimated 56,500 tonnes, compared to only 18,800 tonnes reported to FAO (Figure 4). Catches fluctuated, but declined overall from approximately 1,300 t·year-1 in 1950 to 500 t·year-1 in the late 2000s. Bycatch followed a different trend, totaling 10,300 tonnes and peaking around 250 t·year-1 in the 1980s due to the increasing number of trawlers. Bycatch amounts in the 1950s and the 2000s are similar, slightly above 100 t·year-1 (Figure 4).  Unlike our reconstruction for the lobster fishery (above), the reconstruction of artisanal bottom-trawl fishery catches was mainly based on official statistics. The main novelty in this result comes from the gaps in time-series originally supplied to FAO being filled in. Also, a significant part of total catches (18%) were previously unreported and are now reported as bycatch (either discarded or landed; Figure 4).  Recreational fisheries: residents and tourists  Recreational fisheries were estimated to catch 35,150 tonnes, of which 80% was taken by local resident fishers, and 20% by tourists (Figure 5). These catches were previously not included in statistics provided to FAO.   Figure 5: Reconstructed catches by recreational fishers in Corsica, 1950-2008. Anchor points are indicated by closed circles. Figure 6: Catches in Corsican waters, showing a) reconstructed total catches versus landings data as supplied to FAO; and b) taxonomic breakdown (top 10 species) of reconstructed total catches in Corsica, 1950-2008. The ‗others‘ group includes Scorpaena scofa, Phycis spp., Pagellus, spp., Labridae, Serranus spp., Seriola dumerili,  Zeus faber, as well as other fish species of lower importance in term of percentage, and species of invertebrates. 10 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Based on our assumptions, we estimated that recreational catches by residents were the highest in 1950, with 612 t·year-1, and then declined to 210 t·year-1 by 2008 (Figure 5). On the other hand, recreational catches by tourists were estimated to have increased during the last two decades, increasing from 17 t·year-1 in 1950 to 360 t·year-1 by 2008 (Figure 5).  Overall reconstruction  Reconstructed total Corsican fisheries catches total over 118,700 tonnes since 1950, compared to only 23,700 tonnes reported to FAO by local fisheries authorities. Overall, total catches appear to be steadily decreasing from approximately 2,800 t·year-1 in 1950 to 1,200 t·year-1 by 2008, interrupted by a peak catch of over 3,000 t·year-1 in 1975 (Figure 6a and Appendix Table A1). This decrease seems linked to the decline in both fishers and vessel numbers, but also to declines in fish abundance along the Corsican coast.  Official statistics likely accounted for commercial (artisanal) fisheries only, that is, red spiny lobster and bottom-trawl fisheries. Recreational fisheries by Corsicans, or by tourists, were not considered by official authorities. Finally, we highlighted the existence of discards (for red lobster and bottom-trawl fisheries), which are generally not included in reported statistics (Zeller et al., 2011).  This improved accounting of total catches (versus reported commercial landings) is also evident in the improved taxonomic accounting provided by our study (Figure 6b). Data reported by FAO on behalf of Corsica, besides being of poor quality, also had a poor taxonomic breakdown. Species present in these official data were reported as arbitrary, according to the local fisheries literature. In contrast, we have been able to assign catches to over 30 taxa, of which each had catches allocated in accordance to the literature (Figure 6b and Appendix Table A2).  This study provides an estimate of total fisheries catches in Corsican waters since 1950, and although some sectors such as the pelagic fisheries have not been dealt with in detail, two major conclusions emerge from our work: (1) historical events, changes in gear and emergence of new fisheries illustrate that, despite being assumed to be one of the areas of the Mediterranean with the lowest fishing pressure (Riutort, 1994; Relini et al., 1999), Corsican waters may be exposed to higher fishing pressure than previsously assumed; and (2) our results suggest that Corsicans seem to be much more involved in marine resource exploitation than it appears in the literature and in official statistics.  ACKNOWLEDGEMENT  We thank the Sea Around Us project, a scientific collaboration between the University of British Columbia and the Pew Environment Group. We also thank Jacques Bertrand (Ifremer) for sharing information and knowledge.  REFERENCES  Anon. (1975) Décret n° 75-1128 du 9 décembre 1975 portant création de la réserve naturelle de Scandola (Corse). Journal Officiel de la République Française: 12612. Anon. (1976) Loi n°76-655 du 16 juillet 1976 relative à la zone économique au large des côtes du territoire de la république. Journal Officiel de la République Française: 4299. Anon. (1979) Council Directive 79/409/EEC of 2 April 1979 on the conservation of wild birds. Official Journal L 103: 001-0018. Anon. (1981) Décret n° 81-205 du 3 mars 1981 portant création, délimitation, réglementation et gestion (par le préfet) de la réserve naturelle des iles Cerbicale (Corse du Sud). Journal Officiel de la République Française: 704. Anon. (1987) Décret n° 87-494 du 29 juin 1987 portant création de la réserve naturelle des iles Finocchiarola (Haute Corse). Journal Officiel de la République Française: 7364. Anon. (1992) Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora. Official Journal L 206: 7. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  11   Anon. (1994) Décret n° 94-688 du 9 août 1994 portant création de la réserve naturelle de l'étang de Biguglia (Haute Corse). Journal Officiel de la République Française: 11770. Anon. (1999) Décret du 23 septembre 1999 portant création de la réserve naturelle des Bouches de Bonifacio (Corse du Sud). Journal Officiel de la République Française: 14243. Anon. (2010) Résultats et enseignements de l‘année touristique 2009. Agence du tourisme Corse. Bertrand, J.A., Gil de Sola, L., Papaconstantinou, C., Relini, G. and Souplet, A. (1998) An internation bottom trawl survey in the Mediterranean: the MEDITS programme. Actes de colloques IFREMER 26: 76-93. Bonnet, M. (1973) Les pêches maritimes sur le côtes françaises de la Méditerranée – actualités, perspectives. Science et Pêche 222: 18. Cacaud, P. (2005) Fisheries laws and regulations in the Mediterranean: a comparative study. Studies and reviews 75. General Fisheries Commission for the Mediterranean. United Nations Food and Agriculture Organization (FAO), Rome (Italy), 40 p. de Caraffa, T. (1929) poissons de mer et la pêche sur les côtes de la Corse. Laffitte Reprints (1980), Marseille, 336p. Doumenge, F. (1956) Problèmes d‘équipement de la pêche en Corse. Bulletin de la Société Languedocienne de Géographie, 28(4). FAO (2009) FishStat Plus - Universal software for fishery statistical time series. v2.3. United Nations Food and Agriculture Organization (FAO), Rome (Italy). Machias, A., Vassilopoulou, V., Vatsos, D., Bekas, P., Kallianiotis, A., Papaconstantinou, C. and Tsimenides, N. (2001) Bottom trawl discards in the northeastern Mediterranean Sea. Fisheries Research 53(2): 181-195. Marin, J. (1987) Exploitation, biologie et dynamique du stock de langouste rouge de Corse, Palinurus elephas Fabricus. Thèse de doctorat d‘état. Université d‘Aix Marseille II. Faculté des sciences de Luminy, 327p. Maurin, C. (1965) La pêche française de 1964 en Méditerrannée. Science et Pêche 134: 1-4. Miniconi, R. (1989) Les poissons et la pêche en Corse. Thèse de doctorat d‘état. Université d‘Aix Marseille II. Faculté des sciences de Luminy, 504 p. Miniconi, R. (1994) Les poissons et la pêche en Méditerranée : la Corse. Piazzola and La Marge, Ajaccio, France. Miniconi, R. (2001) Les poissons de Corse - Biologie, pêche, appellations. 262 p. Mouillot, D., Tomasini, J.A., Culioli, J.M. and Do Chi, T. (2007) Développement durable de la pêche artisanale sur le site de la Réserve Naturelle des Bouches de Bonifacio (Corse du sud). Programme MEDD LITEAU 2 : Gestion intégrée des zones côtières. Rapport final, 66 p. Pauly, D. (2010) Five easy pieces: how fishing impacts marine ecosystems. Island Press, Washington, DC, 193 p. Pere, A., Lejeune, P. and Pelaprat, C. (2010) Suivi scientifique de la pêche langoustière corse - Rapport final - Années 2004-2009. Contrat Office de l‘Environnement de la Corse - Stareso, 106 p. Pere, A., Pelaprat, C. and Lejeune, P. (2011) Overview of the corsican fishery and presentation of series of available data. 1st LanConnect Workshop. 21-23 March 2011. Palma de Mallorca (Spain). Pere, A., Pelaprat, C., Pergent-Martini, C., Livrelli, J.N. and Lejeune, P. (2007) The spiny lobster fishery in Corsica. 8th International Conference and Workshop on Lobster Biology and Management. September 23-28 2007. Charlottetown (Canada). Pichot, P. and Aldebert, Y. (1978) La pêche de la sardine en Méditerranée française. Science et Pêche 277: 16. Regional Committee of Corsican Marine Fisheries (2009) Corse - l'interdiction de pêcher l'espadon suscite des vagues. Corse Matin. Relini, G., Bertrand, J.A. and Zamboni, A. (1999) Synthesis of the knowledge on Bottom fishery resources in central Mediterranean (Italy and Corsica). Biologia Marina Mediterranea 6 (suppl. 1). 868 p. Rigo, D. (2000) La pêche professionnelle dans la Réserve Naturelle des Bouches de Bonifacio, effort et productions. Rapport de D.E.S.S. Université de Corse, 53 p. Riutort, J.J. (1989) Premières estimation des captures et de l'effort de pêche déployé par les « petits métiers » sur le littoral nord-ouest de la Corse. Etude de la biologie des principales espèces cibles. Rapport Région de Corse, 151 p. Riutort, J.J. (1994) La pêche en Corse : le rouget de roche, espèce cible. Project Stareso/EEC XIV - 1/MED/91/006, 167 p. Riutort, J.J. (1999) Pêche de la langouste rouge (Palinurus elephas), en Corse (production, sélectivité des engins de pêche, effet réserve des cantonnements à crustacés). Rapport final. 117 p. Rocklin, D., Santoni, M.C., Culioli, J.M., Tomasini, J.A., Pelletier, D. and Mouillot, D. (2009) Changes in the catch composition of artisanal fisheries attributable to dolphin depredation in a Mediterranean marine reserve. ICES Journal of Marine Science 66(4): 699-707. 12 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Sanchez, P., Demestre, M. and Martin, P. (2004) Characterisation of the discards generated by bottom trawling in the northwestern Mediterranean. Fisheries Research 67(1): 71-80. Santoni, M.C. (2002) Evolution de l‘effort et des productions de la pêche artisanale sur le site de la Réserve Naturelle des Bouches de Bonifacio. Office de l‘Environnement de la Corse : Réserve Naturelle des Bouches de Bonifacio. Rapport de D.E.S.S. Université de Corse, 64 p. von Baeyer, H.C. (1993) The Fermi solution: essays on science. Random House, New York, 176 p. Zeller, D. and Pauly, D., eds. (2007) Reconstruction of marine fisheries catches for key countries and regions (1950- 2005). Fisheries Centre Research Reports 15(2). Fisheries Centre, University of British Columbia, Vancouver, 163 p. Zeller, D., Rossing, P., Harper, S., Persson, L., Booth, S. and Pauly, D. (2011) The Baltic Sea: Estimates of total fisheries removals 1950-2007. Fisheries Research 108(2-3): 356-363. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  13    Appendix Table A1: Annual catches by Corsican fisheries. Year Domestic fisheries (t) Data reported to FAO Reconstructed catches 1950 0 2759 1951 0 2751 1952 0 2744 1953 0 2736 1954 0 2728 1955 0 2666 1956 0 2603 1957 0 2540 1958 0 2475 1959 0 2478 1960 0 2484 1961 0 2634 1962 0 2740 1963 0 2698 1964 0 2657 1965 0 2615 1966 0 2506 1967 0 2427 1968 0 2422 1969 0 2421 1970 945 2373 1971 784 2204 1972 1280 2206 1973 966 1757 1974 1719 2772 1975 2128 3195 1976 1632 2647 1977 625 1457 1978 759 1662 1979 607 1450 1980 1012 1968 1981 899 1836 1982 1100 2109 1983 986 1956 1984 967 1930 1985 1042 1934 1986 895 1746 1987 811 1672 1988 1093 2094 1989 604 1473 1990 574 1462 1991 584 1469 1992 502 1358 1993 220 1540 1994 137 1482 1995 99 1468 1996 96 1448 1997 85 1433 1998 74 1418 1999 59 1404 2000 28 1370 2001 26 1342 2002 22 1314 2003 21 1287 2004 23 1259 2005 15 1243 2006 281 1482 2007 0 1194 2008 0 1226 14 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller    Appendix Table A2: Six most important taxa caught by domestic fisheries in Corsica‘s EEZ, 1950-2008. Year Palinurus elephas Sardina pilchardus Diplodus spp. Dentex dentex Scorpaena scrofa Phycis spp. Othersa 1950 375 387 189 171 165 106 1366 1951 373 387 187 169 163 104 1368 1952 372 387 186 167 160 102 1369 1953 370 387 185 165 157 101 1371 1954 369 387 184 164 154 99 1371 1955 325 387 183 161 150 96 1363 1956 281 387 182 159 147 94 1354 1957 237 387 181 157 143 91 1343 1958 193 387 180 154 139 89 1332 1959 153 387 179 156 143 91 1368 1960 153 387 178 155 142 91 1378 1961 259 387 177 156 145 93 1416 1962 363 387 176 154 143 92 1424 1963 324 387 176 153 141 90 1426 1964 285 387 176 153 139 89 1428 1965 246 387 176 152 137 88 1429 1966 204 387 176 148 130 83 1377 1967 163 387 176 146 126 81 1347 1968 163 387 176 146 125 80 1343 1969 163 387 178 147 125 80 1342 1970 160 387 179 145 119 76 1306 1971 150 387 182 137 100 64 1184 1972 146 420 184 136 95 61 1164 1973 133 265 186 125 70 45 933 1974 184 348 189 170 164 105 1612 1975 192 548 191 178 178 114 1794 1976 179 328 188 165 154 99 1534 1977 132 70 184 124 68 44 835 1978 150 2 181 137 102 65 1025 1979 154 3 177 125 78 50 862 1980 199 15 174 147 131 84 1217 1981 209 3 172 140 118 75 1120 1982 241 0 169 152 146 93 1307 1983 248 0 167 143 129 82 1187 1984 263 6 165 140 124 79 1154 1985 191 6 162 143 134 86 1212 1986 178 13 160 133 114 73 1076 1987 169 13 165 131 105 67 1022 1988 190 13 171 154 148 95 1325 1989 150 7 179 126 80 51 879 1990 145 6 188 129 76 49 869 1991 142 7 187 129 77 50 876 1992 133 7 186 124 66 42 799 1993 141 7 185 133 86 55 932 1994 135 6 184 130 81 52 894 1995 128 8 183 129 80 51 888 1996 129 4 182 128 79 51 875 1997 130 4 181 127 78 50 864 1998 132 4 180 125 76 49 852 1999 133 4 180 124 75 48 841 2000 125 0 179 123 73 47 824 2001 118 0 178 122 71 45 809 2002 110 0 177 120 69 44 794 2003 103 0 176 119 67 43 780 2004 95 0 175 118 65 42 765 2005 98 0 174 116 63 41 751 2006 83 192 173 115 62 39 819 2007 88 0 172 114 60 38 723 2008 135 0 171 112 58 37 712 ‘Others’ comprises Pagellus, spp., Labridae, Serranus spp., Spicara spp., Raja spp., other clupeiformes, Mullus spp., other Scorpaena, Maja squinado, Sepia spp., Homarus gammarus, Lophius spp., Capros aper, Micromesistius poutassou, Scyliorhinus spp., other miscellaneous marine fish, cephalopods and crustaceans. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  15   A BRIEF HISTORY OF FISHING IN THE KERGUELEN ISLANDS, FRANCE1  M.L.D. Palomares and D. Pauly Sea Around Us Project, Fisheries Centre, University of British Columbia, 2204 Main Mall, Vancouver, V6T1Z4, Canada m.palomares@fisheries.ubc.ca; d.pauly@fisheries.ubc.ca  ABSTRACT  Catch statistics from around the (uninhabited) Kerguelen Islands, which are part of the French Antarctic and sub-Antarctic Territories, and where distant-water fisheries began in 1970, were obtained from the CCAMLR (Commission for the Conservation of Antarctic Marine Living Resources) Statistical Bulletin (Area 58.5.1) and complemented by statistics reported through the French KERPECHE program. Catches originally expressed by fishing seasons were re-expressed as calendar years, which results in a slight between-season smoothing. These catches show a general decline over a 30 year-period and an expansion of the longline fishery to deeper waters in the last 10 years.   INTRODUCTION  The Kerguelen Islands  The Kerguelen Islands (49°30‘S, 69°30‘E) are part of the French Antarctic and sub-Antarctic Territories, which also include the islands of Crozet, Amsterdam and St. Paul, and the Antarctic district of Terre Adélie (www.taaf.fr). They consist of a main island called ‗La Grande Terre‘ (6,700 km2) and a number of smaller surrounding islets. Kerguelen Island sits in the middle of the combined shelf of the Kerguelen and Heart Islands (Australia), known as the Kerguelen Plateau, which covers an area of 100,500 km2 above 500 m depth (Pruvost et al., 2005: see Figure 1). The islands are uninhabited both because of their isolated locations and the extreme climate prevailing in the area.  Fisheries and their resource species  Fishery prospecting cruises (mostly by the USSR, i.e., the Ukraine; Zeller and Rizzo 2007) in the 1960s led to the development of a modern fishery in the Kerguelen Islands starting in 1970 with about 10 Ukrainian bottom trawlers operating during 6-month fishing seasons without management or control. They targeted marbled rockcod (Notothenia rossii), mackerel icefish (Champsocephalus gunnari) and gray rockcod (Lepidonotothen squamifrons), and also caught unspecified by-catch species of the plateau, at 200-500 m depths in what is now known as CCAMLR area 58.5.1 (Pruvost et al., 2005). This unmanaged fishery continued until France declared an Exclusive Economic Zone (EEZ) around the islands as well as the  1  Cite as: Palomares, M. L. D. and Pauly, D. (2011) A brief history of fishing in the Kerguelen Islands, France. pp. 15-20. In: Harper, S. and Zeller, D. (eds.) Fisheries catch reconstructions: Islands, Part II. Fisheries Centre Research Reports 19(4). Fisheries Centre, University of British Columbia [ISSN1198-6727] Figure 1. Map of Kerguelen Islands, CCAMLR areas 58.5.1, showing French (Kerguelen Islands) and Australian (Heart Island) Exclusive Economic Zones, as well as 500 m and 1000 m depth contour.  16 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   other ‗Terres Australes et Antarctiques Françaises‘ in 1978 (TAAF; see Duhamel, 1995). Since the implementation of the French management scheme in 1980, foreign fleets could access the Kerguelen Islands‘ EEZ only through access agreements for which the French government granted quotas, limiting to 7 the number of trawlers operating at any one time (Pruvost et al., 2005; Duhamel, 1995).  The discovery of a large stock of Patagonian toothfish, Dissostichus eleginoides, by USSR bottom trawlers in the 1984-1985 fishing season on the slopes of the Kerguelen shelf led to the development of this high- value trawl fishery. In 1996, the former USSR stopped trawling in Kerguelen waters and only 2 Ukrainian longliners and 2 French bottom trawlers remained (Pruvost et al., 2005). In the same year, a joint French and Japanese prospecting cruise aboard the M/V Anyo Maru established that longlining was an effective method for catching Patagonian toothfish (Duhamel and Hautecoeur, 2009), which led to the development of this fishery, completely replacing the bottom trawl fishery in the 2000-2001 fishing season (Lord et al., 2006). The high initial abundance of this stock encouraged a rapid expansion of the longline fisheries and the subsequent proliferation of non-licensed longliners from non CCAMLR member states (Kock, 2001). The illegal fishery catch peaked between 1996 and 2004, with catches reaching four times that of the regulated catch in 1997 (Agnew, 2000). In 2005, illegal fishing was curtailed and the fishery was limited to 7 French longliners (Pruvost et al., 2005).  Overall, both trawl and longline fisheries in the Kerguelen Islands increased their effort throughout the period considered here, i.e., 1970-2005. Their catch per unit of effort has consequently strongly diminished, in spite of the expansion of the longline fishery from an average fishing depth of 500 m to 1,000 m (Lord et al., 2006). This suggests massive declines in the target fish biomass; the mean individual size of Patagonian toothfish has also declined (Duhamel and Hautecoeur, 2009).   MATERIAL AND METHODS  The Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) catch statistics for CCAMLR area 58.5.1 were used as a basis of this catch reconstruction for the period 1970-2010. These statistics were missing data for the period 1979-1987. Lord (2005) provided statistics for: (i) catches of Ukrainian and French trawlers for Patagonian toothfish, marbled rockcod, mackerel icefish and grey rockcod fisheries for the period 1979-2001; (ii) longline catches for the Patagonian toothfish fishery and its by-catch (mainly rays and grenadiers) for the period 1990-2003 using the French KERPECHE database (see Lord et al., 2006 and Pruvost et al., 2005); and (iii) catch estimates from illegal fishing operations based on recorded arrests for the period 1996-2003. In those cases where the CCAMLR and KERPECHE statistics overlapped, the KERPECHE catches ranged from 73% (marbled rockcod trawl fishery) to 204% (rays as by-catch of the longline fishery) of the CCAMLR statistics. We used the KERPECHE statistics in lieu of the CCAMLR statistics for: (a) trawlers for the period 1979-1989 for mackerel icefish, marbled and gray rockcod, and 1979-1991 for Patagonian toothfish; and (b) longliners, notably during the period of expansion for Patagonian toothfish, in order to be able to include some of the unreported catches.  Catches reported by the USSR to the CCAMLR were all assigned to the Ukraine because this part of the world‘s ocean was exploited by USSR vessels from the Ukrainian SSR during the Soviet era (see Romanov, 2003; Zeller and Rizzo 2007). Also, it is only Ukrainian vessels which exploited the Kerguelen following the breakup of the USSR (see Pruvost et al., 2005).  RESULTS AND DISCUSSION  Appendix Table A1 and A2 present a summary of the catch statistics available from the Kerguelen Islands extracted from the CCAMLR (2010) and completed with data from Lord (2005, Annex 3). The catches originally presented by CCAMLR ‗season‘, from the 1st of July of a particular year to the 30th of June of the next year, were converted to calendar years by assuming that the catch in the first half of the season (in a given year) is equal to that of the second half of the season (in the next year). This does not affect cumulative catches and, in fact, corresponds to a slight between-season smoothing.  Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  17   This brief account of the fisheries in the Kerguelen Islands is meant to present the Kerguelen Island fisheries statistics in such a way that they can be included in the Sea Around Us project‘s (www.seaaroundus.org) catch mapping procedure (see Watson et al., 2004). This is the reason why the catch is reported by calendar years and not as done in the original literature, which account for fishing ‗seasons‘. Moreover, we include estimates of illegal catches, which although highly tentative, are likely to be more correct than the statistically very precise but inaccurate estimate of zero commonly used as a replacement for difficult to estimate quantities such as illegal catches (see Zeller et al., 2011).  The resulting catch statistics for the Ukraine (see Appendix Table A2 and Figure 2a), i.e., the fishery which heavily exploited mackerel icefish over three decades, show peaks and troughs similar to patterns reported for South Georgia, South Orkney Islands, Elephant Island and South Shetland Islands (Kock, 1991). Heavy fishing pressure on the strong 1973-1974 year classes may have reduced the stock size to a level that prevented adequate recruitment and thus recovery (Anon., 2001). The trend of the peaks shows a steady decline in the catch, and Kock and Everson (2003) concluded that this decline is the result of a combination of factors, including heavy fishing pressure, changes in the abundance of icefish predators (Antarctic fur seals and penguins) and prey (krill), and warming in the northern parts of the distributional range of icefish. Commercial fishing for mackerel icefish was banned at the end of the 1980s (Kock, 1991) resulting in the tapering off of statistics reported by the Ukraine.  The increasing French catch trend (see Appendix Table A1 and Figure 2b), on the other hand, reflects an exploratory fishery tending towards expansion to deeper waters. Duhamel et al. (1997) speculated that the level of longline bycatch (mainly of rays and rattails or grenadiers, Family Macrouridae) have the potential to replace the Patagonian toothfish fishery. Although smaller and subjected to management and monitoring, this expanding fishery has effectively ‗counterbalanced‘ its decreasing catch per unit of effort. 0.0 5.0 10.0 15.0 20.0 25.0 30.0 1970 1980 1990 2000 2010 C a tc h  ( t x  1 0 3 ) Year Champsocephalus gunnari Dissostichus eleginoides Notothenia squamifrons Notothenia rossii 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 1970 1980 1990 2000 2010 C a tc h  ( t x  1 0 3 ) Year Dissostichus eleginoides Notothenia rossii Notothenia squamifrons Champsocephalus gunnari Rajiformes other Macrourus spp Macrourus carinatus 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 1970 1980 1990 2000 2010 C a tc h  ( t x  1 0 3 ) Year Ukraine FranceIllegal a) b) c) Figure 2. Reconstructed fisheries catches (in metric tonnes) for the Kerguelen Islands (CCAMLR Area 58.5.1) with statistics adapted from CCAMLR (2010) and complemented with data from Lord (2005) for ;(a) Ukrainian (i.e., former USSR and Russian Federation statistics); (b) French trawlers and long-liners; and (c) total catch by country including some estimates of illegal fishing during the period 1996-2003 from Lord (2005). 18 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Overall, it is masking the fact that no new fishing grounds have been found since all accessible shallow shelf stocks have been over exploited (Lord et al., 2006).  The illegal catch estimates shown in Figure (2c) that were reported both by the CCAMLR (2010) and by Lord (2005) may well be underestimates, i.e., catches may be twice (or more) of those reported to the CCAMLR.   ACKNOWLEDGEMENTS  The Sea Around Us Project is a collaboration between the University of British Columbia and the Pew Environment Group.  REFERENCES  Agnew, D.J. (2000) The illegal and unregulated fishery for toothfish in the Southern Ocean, and the CCAMLR catch documentation scheme. Marine Policy 24: 361-374. Anon. (2001). Report of the Workshop on Approaches to Management of Icefish. In CCAMLR of the Twentieth Meeting of the Scientific Committee, Annex 5, Appendix D, pp. 467-515. CCAMLR, Hobart, Australia. CCAMLR, (2010). CCAMLR Statistical Bulletin 2010. Volume 22 (Database Version; www.ccamlr.org). CCAMLR, Hobart, Australia. Duhamel, G. (1995). Gestion des pêches aux îles Kerguelen. Recherche Marine 13: 16-17. Duhamel, G., Hautecoeur, M. (2009). Biomass, abundance and distribution of fish in the Kerguelen Islands EZ (CCAMLR Statistical Division 58.5.1). CCAMLR Science 16: 1-32. Duhamel, G., Pruvost, P., Capdeville, D. (1997) By-catch of fish in longline catches off the Kerguelen Islands (Division 58.5.1) during the 1995/1996 season. CCAMLR Science 4: 175-193. Pruvost, P., Duhamel, G., Palomares, M.L.D. (2005) An ecosystem model of the Kerguelen Islands‘ EEZ. p. 40-64 In: M.L.D. Palomares, P. Pruvost, T.J. Pitcher and D. Pauly, D. (eds.), Modeling Antarctic Marine Ecosystems. Fisheries Centre Research Reports 13(7). Fisheries Centre, University of British Columbia, Vancouver, Canada. Kock, K.-H. (1991) The state of exploited fish stocks in the Southern Ocean – a review. Archiv für Fischereiwissenschaften 41: 1-66. Kock, K.-H. (2001) The direct influence of fishing and fishery-related activities on non-target species in the Southern Ocean with particular emphasis on longline fishing and its impact on albatrosses and petrels – a review. Reviews in Fish Biology and Fisheries 11: 31-56. Kock, K.-H., Everson, I. (2003) Shedding new light on the life cycle of mackerel icefish in the Southern Ocean. Journal of Fish Biology 63: 1-21. Lord, C. (2005) Étude d‘une population exploitée de légines (Dissosticus eleginoides) aux Îles Kerguelen. Mémoire de Master. Université Pierre et Marie Curie, Paris VI. 32 p. Lord, C., Duhamel, G., Pruvost, P. (2006) The Patagonian toothfish (Dissostichus eleginoides) fishery in the Kerguelen Islands (Indian Ocean Sector of the Southern Ocean). CCAMLR Science 13: 1-25. Romanov, E.V. (Editor) (2003). Summary and Review of Soviet and Ukrainian Scientific and Commercial Fishing Operations on the Deepwater Ridges of the Southern Indian Ocean. FAO Fisheries Circular No. 991. FAO/YugNIRO, Rome, Italy. 84 p. Watson, R., Kitchingman, A., Gelchu, A., Pauly, D. (2004) Mapping global fisheries: sharpening our focus. Fish and Fisheries 5: 168-177. Zeller, D. and Rizzo, Y. (2007) Country disaggregation of catches of the former Soviet Union (USSR). p. 157-163 In: D. Zeller, and D. Pauly (eds.) Reconstruction of marine fisheries catches by countries and regions (1950-2005). Fisheries Centre Research Reports 15(2). Fisheries Centre, University of British Columbia. Zeller, D., Rossing, P., Harper, S., Persson, L., Booth, S. and Pauly, D. (2011) The Baltic Sea: estimates of total fisheries removals 1950-2007. Fisheries Research 108: 356-363.   Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  19   Appendix Table A1: Kerguelen Islands, CCAMLR area 58.5.1, fisheries catch statistics (t) by French trawlers and longliners from 1970-2010 adjusted from fishing season to calendar year (see text and Figure 2). Year France Blue antimora Mackerel icefish Unicorn icefish Patagonian toothfish Ridge scaled rattail Rattailsa Marbled rockcod Gray rockcod Raysb 1970-1978 – – – – – – – – – 1979 – 75 – 9 – – 65 247 – 1980 – 136 – 11 – – 506 597 – 1981 – 952 – 9 – – 984 574 – 1982 – 2328 – 14 – – 645 326 – 1983 – 1832 – 15 – – 143 312 – 1984 – 855 – 379 – – 136 622 – 1985 – 1407 – 396 – – 139 548 – 1986 – 1093 – 201 – – 71 228 – 1987 – 158 – 207 – – 28 93 – 1988 – 1292 – 118 – – 16 104 – 1989 – 1295 – 152 – – 24 163 – 1990 – 188 – 208 – – 170 67 – 1991 – 8 – 1199 – – 148 5 0 1992 – 7 – 1611 – – 0 0 0 1993 – 6 – 1582 – – 0 0 0 1994 – 6 – 2960 – – 0 0 1 1995 – 42 1 4178 – – 0 1 1 1996 – 45 1 3742 – – 0 1 0 1997 – 3 2 3744 – 0 0 0 6 1998 – 0 3 3919 – 6 0 0 11 1999 – 0 1 3984 – 7 1 0 87 2000 – 0 0 5139 – 87 1 0 189 2001 – – – 5443 – 132 0 0 298 2002 – – – 4450 – 201 – – 623 2003 – – – 4722 – 472 – – 815 2004 – – – 5231 – 805 – – 383 2005 7 – – 5123 – 489 – – 477 2006 21 – – 5115 – 476 – – 428 2007 15 – – 5179 – 537 – – 351 2008 35 – – 5026 409 276 – – 186 2009 58 – – 5045 896 – – – 313 2010 23 – – 2620 488 – – – 136 Totals 157 11,826 6 81,727 1,792 3,487 3,077 3,443 4,302 aMacrourus spp.;b Raja spp. and unidentified Rajiformes, most probably Bathyraja eatonii and Bathyraja irrasa (Lord, 2005; Lord et al., 2006).   20 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller    Appendix Table A2: Kerguelen Islands, CCAMLR area 58.5.1, fisheries catch statistics (t) for the Ukrainian (i.e., former USSR and Russian Federation statistics) from 1970- 2010 adjusted from fishing season to calendar year and estimates of illegal fishing during the period 1996-2003 (see text and Figure 2). Year Ukraine Illegal Fisheriesb Mackerel icefish Patagonian toothfish Marbled rockcod Gray rockcod Other finfisha 1970 3 – – – – – 1971 8982 – – – – – 1972 17995 – – – – – 1973 9018 – – – – – 1974 410 – – – – – 1975 900 – – – – – 1976 6275 – – – – – 1977 25110 – – – – – 1978 24847 – – – – – 1979 6118 71 522 1978 – – 1980 1485 90 4045 4772 – – 1981 8019 74 7875 4595 – – 1982 18622 113 5162 2607 – – 1983 14657 123 1141 2493 – – 1984 6841 3031 1088 4979 – – 1985 11253 3170 1113 4388 – – 1986 8747 1609 571 1824 – – 1987 1261 1655 225 748 – – 1988 10338 946 126 829 – – 1989 11585 1212 172 777 13 – 1990 4494 1354 143 1397 13 – 1991 6699 3902 150 674 – – 1992 6647 3772 144 51 5 – 1993 16 4096 1 0 12 – 1994 614 1530 1 0 7 – 1995 1926 1311 – 9 31 – 1996 1312 1279 – 8 31 1000 1997 – 1008 – – – 6913 1998 – 966 – – – 8275 1999 – 739 – – – 5163 2000 – 297 – – – 5410 2001 – – – – – 5760 2002 – – – – – 5213 2003 – – – – – 2063 2004 – – – – – – 2005 – – – – – – 2006 – – – – – – 2007 – – – – – – 2008 – – – – – – 2009 – – – – – – 2010 – – – – – – Totals 214,171 32,345 22,478 28,567 111 39,795 aOsteichthyes reported in the CCAMLR statistics, most probably including incidental bycatch of southern lantern shark (Etmopterus granulosus), porbeagle (Lamna nasus), Greenland shark (Somniosus microcephalus), moray cod (Muraenolepis marmoratus) and gray rockcod (Notothenia squamifrons) (see Lord et al., 2006); bConsisting only of longlines targeting Patagonian toothfish; their bycatch consisting mainly of rays and rattails. Note that the blue antimora (Antimora rostrata), unicorn icefish (Channichthys rhinoceratus), ridge scaled rattail (Macrourus carinatus) and other rattail species and rays are also bycatch of the longline fishery which are sometimes legally reported.  Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  21   RECONSTRUCTION OF TOTAL MARINE FISHERIES CATCHES FOR MADAGASCAR (1950-2008)1  Frédéric Le Manacha, Charlotte Goughb, Frances Humberb, Sarah Harperc, and Dirk Zellerc aFaculty of Science and Technology, University of Plymouth, Drake Circus, Plymouth PL4 8AA, United Kingdom; flm@livinblue.com bBlue Ventures Conservation, Aberdeen Centre, London, N5 2EA, UK; charlie@blueventures.org; fran@blueventures.org cSea Around Us Project, Fisheries Centre, University of British Columbia 2202 Main Mall, Vancouver, V6T 1Z4, Canada ; s.harper@fisheries.ubc.ca; d.zeller@fisheries.ubc.ca  ABSTRACT  Fisheries statistics supplied by countries to the Food and Agriculture Organization (FAO) of the United Nations have been shown in almost all cases to under-report actual fisheries catches. This is due to national reporting systems failing to account for Illegal, Unreported and Unregulated (IUU) catches, including the non-commercial component of small-scale fisheries, which are often substantial in developing countries. Fisheries legislation, management plans and foreign fishing access agreements are often influenced by these incomplete data, resulting in poorly assessed catches and leading to serious over-estimations of resource availability. In this study, Madagascar‘s total catches by all fisheries sectors were estimated back to 1950 using a catch reconstruction approach. Our results show that while the Malagasy rely heavily on the ocean for their protein needs, much of this extraction of animal protein is missing in the official statistics. Over the 1950-2008 period, the reconstruction adds more than 200% to reported data, dropping from 590% in the 1950s to 40% in the 2000s. This discrepancy has profound management implications as well as consequences for current understanding of Madagascar‘s fisheries economy and communities‘ reliance on wild fish for food security.   INTRODUCTION  Madagascar is located in the western Indian Ocean, and separated from Africa by the Mozambique Channel (Figure 1). With a land area of approximately 587,000 km², it is the fourth largest island in the World and an African biodiversity hotspot, with around 80% of its terrestrial species being indigenous, and its endemic biodiversity threatened by habitat loss (Brooks et al., 2006; Anon., 2008a). Given its great size, spanning 14 degrees of latitude, Madagascar exhibits a range of geological, oceanic and climatic environments, for example, the east of the country is mountainous with a narrow continental shelf facing the prevailing trade winds and oncoming east equatorial current, while the west side is characterised by large plains in a rain shadow, with the coast fringed by a wide continental shelf (Cooke et al., 2003). The southern region is subject to more arid conditions (Jury, 2003), restricting its agricultural potential. These environmental differences have also shaped marine ecosystems: mangroves are almost exclusively present on the west coast (Giri and Muhlhausen, 2008), whereas coral reefs span the southwest, west and northeast coasts, and include one of the largest coral reef systems in the Indian Ocean, totalling approximately 2,230 km² (Spalding et al., 2001). These geographical differences have also resulted in spatial divergence in the distribution of the island‘s human population with the eastern part of the island having the highest density, while the west coast is home to the majority of fishers and therefore experiences the highest fishing pressure (Guidicelli, 1984; Bellemans, 1989; Rafalimanana, 1991; Laroche et al., 1997).   1  Cite as: Le Manach, F., Gough, C., Humber, F., Harper, S. and Zeller, D. (2011) Reconstruction of total marine fisheries catches for Madagascar (1950-2008). pp. 21-37. In: Harper, S. and Zeller, D. (eds.) Fisheries catch reconstructions: Islands, Part II. Fisheries Centre Research Reports 19(4). Fisheries Centre, University of British Columbia [ISSN1198-6727]. 22 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Historically, Madagascar has had several political regimes (Schraeder, 1995). After the Berlin Convention in 1885, which decided the fate of most of the African continent during colonisation, Madagascar was invaded by France in 1896, turning Madagascar into a French colony and finally Overseas Territory in 1946. Although the colonial power invested in national infrastructure such as trains and schools, this period was characterised by protracted political violence, with around 700,000 out of 3 million inhabitants being killed within a few decades (Rousse, 2010). Giving increasing power to national institutions, the French government withdrew step-by-step and in 1960 the First Republic was proclaimed. However, the first Malagasy President was unpopular with the country‘s people, mainly due to the continuing strong economic and political ties with France. In 1975, the Second Republic aligned itself with the USSR; key sectors of the economy were nationalized and the country experienced a radical socialist and authoritarian political regime. Ten years later, heavy opposition to this regime developed, and in 1992 the Third Republic was proclaimed. Political instability continues to the present day, following a military-backed coup in 2009. Madagascar‘s current unelected regime faces ongoing economic sanctions and is not recognised by the international community, including the European Union (EU) or the Southern African Development Community (SADC).  Economically, Madagascar is one of the poorest countries in the world. Per capita GDP has declined steadily since Independence, having never exceeded $410, and currently is at less than $300 (year 2000 USD). Approximately 70% of the population currently lives below the poverty threshold, and over half of the country‘s population is dependent on the exploitation of natural resources for their livelihoods (World Bank, 2010; Horning, 2008). Subsistence fisheries are of prime importance for coastal communities, especially in the south and west of the country where agriculture is virtually impossible due to aridity.  In developed countries, scientific fisheries assessments such as stock assessments can provide robust data on which to base fisheries management decisions. However, these approaches are expensive, technically complex and often challenged (Murawski, 2010). Developing countries such as Madagascar rarely have adequate scientific capacity or resources to undertake stock assessments. Consequently, poor or incomplete catch data often serve as the only basis for policy and decision making in such countries. In the absence of effective regulations, catch statistics are thought to approximate fluctuations in fish stocks and are therefore viewed as an acceptable proxy for stock assessments. However, catch statistics generally do not account for Illegal, Unreported and Unregulated (IUU) catches, which are widely recognised as a major barrier towards sustainable fisheries management (Sumaila et al., 2006; Hosch et al., 2011). Such IUU catches often result in serious under-estimates of extracted resources and over-estimate of their sustainability (Jacquet et al., 2010). FAO FishStat (FAO, 2009) provides time-series data on marine  Figure 1: Madagascar and its Exclusive Economic Zone (solid line).   Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  23   fisheries landings starting in 1950. These data are based on statistics provided to FAO by member countries. However, it has been shown for many countries that the statistics submitted to the FAO are often incomplete, particularly with regards to artisanal and subsistence fisheries (Zeller et al., 2007; Jacquet et al., 2010; Wielgus et al., 2010). This is likely the case with Madagascar as well, where the importance of seafood for domestic food security has rarely been recognised by the various governments. Here, we re-estimated total marine catches by Madagascar within its EEZ (or EEZ-equivalent waters) for the period 1950 to 2008, with the aim of providing a more accurate baseline for use in policy decisions. This included a review of all fisheries sectors in the country, which allowed us to highlight missing or under-reported components. The present work is also published in Le Manach et al. (2012). Note that Le Manach et al. (2012) contains a mislabelled Figure (3b), which was corrected through a subsequent errata, and is correctly presented here as Figure (6).   MATERIALS AND METHODS  Human population data  Human population data were obtained from PopulStat (www.populstat.info) and various other sources (Central Intelligence Agency, 2010; Globalis, 2010) and used to derive the number of fishers for the whole time-series (1950-2008). Linear interpolations were made between years of known data (Figure 2a). A fishers‘ census conducted by the FAO documented the percentage of artisanal fishers among the population in 1988 for each district (Bellemans, 1989). In the absence of more recent estimates, we assumed that these ratios remained stable between 1988 and 2008. It is likely conservative, as the declining per capita GDP during this period would suggest a growing reliance on small-scale artisanal fishing for livelihood and food security. For the 1950-1988 period, the proportion of artisanal fishers among the total population was assumed to have doubled, increasing from approximately 2% to 4%. Indeed, Bellemans (1989) reported that the number of fishers approximately doubled during the two decades preceding the census (Figure 2b). Billé and Mermet (2002) have also indicated a two-fold increase in the number of fishers between the early 1980s and the early 2000s. Based on this, we estimated a fisher population of 100,000 individuals in 2005 (G. Hosch, pers. comm., Fisheries Planning and Management; Gough and Humber, unpub. data). For this study, the coast was divided in two areas: (1) the southwest, comprising the district of Toliara (Figure 1), where the fishing pressure is known to be the highest (Laroche et al., 1997), and (2) the remaining coastal districts, where fishing pressure is consequently thought to be somewhat lower.  Fisheries sub-sectors  The officially reported landings data which served as the baseline for the study were extracted from the FAO FishStat database (FAO, 2009), and a thorough bibliographic review of fisheries activities in Figure 2: a) Human population of Madagascar with anchor points indicated by solid circles, and b) number of artisanal fishers with the 1987-88 census data foundation (Bellemans, 1989) indicated by dots. 24 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Madagascar allowed us to determine which sectors were either missing or being under-reported. Data sources included peer-reviewed publications, reports by non-governmental organizations (NGOs), and other grey literature. Expert and local knowledge was collected for each sector in order to formulate the best assumptions possible. For each of these fisheries sectors, catches were then derived for the entire 1950-2008 period, based on anchor points found in the literature and informed knowledge-based assumptions.  Shrimp fishery  The shrimp fishery represents around 4% of Malagasy reported landings for the last decade. However, its market value is significant, reaching almost 70% of the officially recorded marine resource contribution to the GDP, mainly due to important export to Europe, Asia and North America (Soumy, 2006; Anon., 2008b; Kasprzyck, 2008).  The industrial shrimp fishery officially started in 1967, after several years of exploratory trawling (Fourmanoir, 1952a, b; Crosnier, 1965). The number of vessels steadily increased from 11 in 1967 to a maximum of 71 vessels between 1993 and 2003, when the number of licenses started to be controlled. Since then, the number of vessels has decreased, with 47 licensed vessels operating in 2008. Although linked to the economic recession of the 2000s, the increasing price of fuel and the international market being flooded by cheaper Chinese shrimp products, this drop was also due to declines in landed catch and increasing competition with artisanal fishers (Kasprzyck, 2008). Conflict between artisanal and industrial fishers continues to be a serious concern for west coast fishing communities (Cripps, 2009). Despite the intensive exploitation by industrial operators in the past, the decline in the economic viability of the fishery is causing owners to reduce their fleets (Razafindrainibe, 2010; McNeish, 2011). The Société de Pêche de Morondava (SoPeMo), a shrimp fishing company based on the west coast, stopped commercial trawling in the region in 2008 (C. Gough, pers. comm.). Unima, the biggest shrimp company in Madagascar is accusing artisanal fishers to have caused the decline in catches (McNeish, 2011).  Artisanal fishers have targeted shrimp since the 1970s. Before then, shrimp was considered incidental bycatch and was used as a complementary food, consumed as an overcooked paste. Since then, artisanal fishers have been attracted by this high-value resource and sell a large portion of their catches to local markets and processors. Due to its rapid expansion, we assumed a yearly growth rate for the artisanal shrimp fishery of 4.5% from 1970 to 2005 (Kasprzyck, 2008; C. Chaboud, pers. comm., Institut de Recherche pour le Développement). The 2005 landings of 3,500 tonnes were used as a basis to carry backward this 4.5% yearly growth rate to 1970.  Bycatch from shrimp fisheries is largely associated with industrial trawl vessels, and is known to be particularly high for tropical shrimp fisheries. Typically, discarded bycatch is not accounted for in reported landings, and we assumed this also applied to Madagascar. One of the earliest estimates for bycatch in the Malagasy shrimp fishery was an amount between 18,000-20,000 tonnes per year in the 1980s (Roullot, 1989). Kelleher (2005) proposed a 1:4.1 bycatch ratio and a discard rate of 72%, which gives tonnage values for the late 1980s similar to Roullot (1989), suggesting that the bycatch was almost entirely discarded. However, it is worth noting that a part of this bycatch is traded to local fishers who fill up their pirogues in exchange for some tobacco or a small amount of money and the local fishers land it and sell it for consumption in the local markets (A. Harris, pers. comm., Blue Ventures Conservation). Here, we applied a discard rate of 72% from 1990 to 2000, and a 90% discard rate for the 1967-1990 period (Table 1), based on Roullot (1989). Lower bycatch ratios and discard rates have been suggested for the last few years (late 2000s). Randriarilala et al. (2008) reported a bycatch ratio of 1:2.5 for 2003-2005, which suggests an annual amount of between 8,500 tonnes and 12,700 tonnes per year. This decrease, even stronger since 2005, seems to have been influenced by two developments in Malagasy regulations (Razafindrainibe, 2010). The first development was the introduction of legislation in the 1990s (decree 1999/2000) requiring industrial vessels to retain at least 50% of bycatch to supply fish to local markets. However, the effectiveness of this regulation has been questioned by Randriarilala et al. (2008), who assessed a discard rate of 62% for the period 2003-2005. The second development was the introduction of Turtle Excluder Devices (TEDs) and Bycatch Reduction Devices (BRDs) in the Malagasy shrimp fishery in 2003 (Anon, 2003; Razafindrainibe, 2010). Since 2005, all vessels are mandated to be equipped with Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  25   such BRDs (decree 2003-1101), which considerably reduce the amount of bycatch (Kasprzyck, 2008; Table 1). This reduction is estimated to be approximately 32%, depending on the year, fishing conditions and area (Fennessy and Isaksen, 2007). Therefore, we considered a bycatch ratio of 1:1.7 from 2005 onward, while discard rates were set at 62% since 2001 (Table 1). However, only approximately 30% of licensed boats carry enforcement personnel from the national fisheries surveillance authority, and it is not clear whether crews use TEDs and BRDs when not under surveillance. Indeed, without surveillance, crews have little incentive to follow legislation, since the economic conditions under which crews work create a major incentive to maximise bycatch for private sale at sea. Shrimpers work seven days a week and are paid per tonne of shrimp landed, with base salaries as little as $25 per month.  The bycatch is dominated by fin fish (e.g., Otolithes argenteus, Johnius dussumieri, Trichiurus lepturus and Pomadasys maculatus), and to a lesser extent, invertebrates such as sea urchins and jellyfish, which can represent an important component of total bycatch. Chen and Chow (2001) estimated a discard survival rate of 8% for fin fish, 35% for cephalopods and 60% for crustaceans in a tropical shrimp fishery in Asia. We assumed here that all the discarded bycatch had a similar mortality, and thus applied mortality rates of 8%, 35% and 60% to the respective bycatch amounts of fin fish, cephalopods and other crustaceans. The commercial shrimp fleet also comprises two other sectors of much lower capacity: - A small fleet of mini-trawlers with engine power less than 50 horse power. This fleet was intended to introduce more efficient gears to the artisanal fisheries. However, they were taken up by industrial fisheries societies to allow them to fish in certain areas otherwise not accessible by their large boats (Kasprzyck, 2008; Direction des Pêches in Morondava, pers. comm.). Landings reported to FAO increased from 45 tonnes to 700 tonnes per year from 1975 to 2008 (peak at 750 t in 2003), and were considered reliable and were included without modification in the ‗industrial grouping‘ described above. - A deep-sea shrimp fishery was initiated in 1992 and ended in 2005 due to technical issues, mainly due to the nature of the sea floor. No information was available concerning associated bycatch. Catches fluctuated between 100 and 150 t∙year-1 and were also included without modification in the ‗industrial grouping‘.   Shark fishery  Although consumption of sharks is common, Madagascar only reports landings of less than 10 tonnes per year for the 2001-2008 period, according to FAO. However, the FAO trade database documents shark exports of up to 85 tonnes per year since 1992, and an independent report suggests that shark meat and fins have been exported since the early 20th century (Petit, 1930). In fact, artisanal fishers target sharks for the export market of fins, but carcasses are rarely discarded, and the meat is either consumed locally or to a lesser extent, sold to Comoros. A number of endangered benthic species, such as the critically Table 1: Summary of data, assumptions and sources used to reconstruct total catches by shrimp fishing fleets in Madagascar. Time period Shrimp catches (t·year-1) Associated bycatch and discards Sources Comment Bycatch  ratio Discards  (%) 1967-2008 300 - 13,300a - - Domalain et al. (2000); Goedefroit et al. (2002); Razafindrakoto (2008), Rokotodratsimba et al. (2008) ; FAO (2009) - 1967-1989  1:4.1 90 Roullot (1989) 1990-2000  1:4.1 72 Kelleher (2005) 2001-2004    1:2.5b 62 Randriarilala et al. (2008) Decree 1999/2000 2005-2008    1:1.7b 62  BRD introduction  a Values reported to FAO were kept for the years 1966, 1994, 2000-2003, as they were deemed more representative than those reconstructed.  b Based on a 32% reduction of bycatch due to BRDs (Fennessy and Isaksen, 2007). 26 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   endangered sawfish (Pristis pectinata), were once commonly caught by artisanal fishers along the mangrove-fringed coast of western Madagascar, but are now considered extremely rare throughout the region (A. Harris, pers. comm., Blue Ventures Conservation).  Sharks are also caught as bycatch in other Malagasy commercial fisheries, such as the shrimp fishery, in which sharks have been reported as representing 1% of the bycatch (C. Chaboud, pers. comm., Institut de Recherche pour le Développement; Table 2). Due to the high price of shark fins, we assumed that all sharks are finned, and that all of the carcasses are retained for local consumption or exported to Comoros (C. Chaboud, pers. comm., Institut de Recherche pour le Développement; A. Harris, pers. comm., Blue Ventures Conservation).  Finally, shark liver oil has also been traditionally used for cooking and to waterproof wooden boats (Cooke, 1997). The quantity of oil used for this purpose is substantial in the Maldives, as Anon. (2001) reports a use of between 54 and 58 kg of oil per year, for each boat. However, as it was impossible to assess the number of sharks required to treat the whole Malagasy artisanal fleet, this component was not considered in this study. Shark liver oil is also a valuable commodity on international markets, with Madagascar‘s sharks targeted by illegal, unreported and unregulated (IUU) boats for this purpose as well as for the fins. A number of known IUU vessels, which previously targeted Patagonian toothfish in the southern Ocean (see Palomares and Pauly, this volume), are reported to have recently converted to shark fishing in southern and western Madagascar by substituting bottom trawl nets with bottom-set gillnet gear to target nurse sharks for liver oil (mainly Nebrius ferrugineus and the vulnerable Pseudoginglymostoma brevicaudatum) (SADC, 2008; Anon. 2010a and b).  Cooke (1997) and Cooke et al. (unpub. data) review the shark fishery in Madagascar, focusing on exports of fins to the Hong Kong and Singapore markets. Given the high market price of fins, we assumed that all captured sharks were finned, and therefore, trade in fins was the best proxy available to assess the minimum quantity of sharks caught each year. Three approaches were used to reconstruct total shark catches by Malagasy fishers (Table 2). Data on the trade of shark fins were used to conservatively estimate the likely minimum catches of sharks that occurred in Madagascar‘s waters during the period 1970-1994: dried fins imported between 1986 and 1995 by Honk-Kong and Singapore from Madagascar were converted to whole body, wet weight using a conversion factor of 98.5% (Cortes and Neer, 2006; Jacquet et al., 2008; Y. Sadovy, pers comm., University of Hong Kong). We assumed that the market started to greatly expand in 1980, and therefore linearly extended the 1986-1988 trend backwards to 1980. For the 1950-1979 period, we assumed that the 1980 per fisher catch rate remained constant back to 1950, and expanded it to total catches using fisher population data. For the 1996-2008 period, we conservatively assumed that the 1994 per fisher catch rate decreased by 3%∙year-1, based on literature and local knowledge (McVean et al., 2006; Y. Sadovy, pers. comm., University of Hong Kong). Currently, catches are reported to decrease, and fishers catch fewer and smaller sharks, most of the time farther from shore than before (Cooke, 1997 and 2003; McVean, 2006; Cooke et al., unpub. data). The high market demand for shark fin as a lucrative yet diminishing fisheries resource is a key factor driving Madagascar‘s nomadic Table 2: Summary of data, parameters, assumptions and sources used for the reconstruction of shark fisheries catches in Madagascar. Sector Timeperiod dry fina (t) Sources Catch (t) Comment Targeted 1950-1979 - Petit (1930) 160 - 570b Exports at least since 1930s. Backward extension of 1980 per fisher catch rate  1980-1985 -  600 - 3050 Backward extension of 1986- 1988 trend in derived catch  1986-1995 34.5 - 64.7 Cooke et al. (2001) 3,430 - 6,440  1996-2008 -  5,400 - 3,760c Decrease of 3%∙year-1 Shrimp bycatch 1967-2008 -  Up to 385 t (1998)d a Hong Kong and Singapore imports; b 1950 value of 60 t derived through keeping the 1980 catch per fisher fixed. c Values for this period were based on interpolations from the 1994 fin trade data, and an assumed 3% per year decrease in catches (McVean, 2006; Cooke et al., 2001; Y. Sadovy, pers. comm. University of Hong Kong). d Values were based on reconstructed industrial shrimp catches and an assumed 1% of total shrimp bycatch composed of sharks (C. Chaboud, pers. comm., Institut de Recherche pour le Développement). Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  27   Vezo fishers further afield during their annual migration, with shark fishers exploiting remoter regions of the west coast of Madagascar, further offshore and in larger numbers, than ever before (Cripps, 2010). As an example of this escalation in fishing effort, the recent introduction of new intensive fishing techniques in the offshore Barren Isles archipelago and around Morondava, involves teams of artisanal fishers deploying weighted ‗barrage‘ nets several kilometres in length, targeting sharks and guitarfish (Cripps, 2010). Based on this information, we applied a 3% per year decrease in catches since 1994 (Table 2).  Non-shrimp invertebrate fisheries  The remaining landings data reported to FAO have been aggregated into a miscellaneous invertebrate grouping, which includes cephalopods and other molluscs, crabs and lobsters, shells and sea cucumbers. These species are heavily targeted by men, women and children for both subsistence and commercial purposes (Rasolofonirina and Conand, 1998; Frontier Madagascar, 2003; Anderson et al., 2008; Barnes and Rawlinson, 2009; Cripps, 2009; Gough et al., 2009; Tucker et al., 2010). The under-reporting of invertebrate fisheries is visible in the statistics reported to the FAO, since reported landings are very similar to exported fisheries products (r²=0.75; not shown). Also, there are no reported invertebrate catches before 1962. However, coastal populations rely heavily on reef gleaning for invertebrates for their daily protein needs, although a significant amount is sold (Cripps, 2009; Gough et al., 2009). Indeed, invertebrate landings account for a major component of fisheries-derived income for artisanal fishers in many parts of Madagascar. Beside holothurian fisheries, octopus is the dominant commercial fishery in much of the southwest and northeast of the country, and lobster plays a crucial role in coastal livelihoods from the rocky shores of the southeast. A thriving trade in marine curios, predominantly molluscs, is also present in most coastal towns. In all these cases, catches are sold to collectors by local fishers for international export. Therefore, we assumed that invertebrate extraction by the local population was happening prior to 1962 and that this sector is missing from the official data.  In order to re-estimate the total extraction of invertebrates, product weight as it appears in the trade data was converted to (whole body) wet weight, using FAO conversion factors (Anon., 2000). A highly conservative export rate of 80% (for sea cucumber, cephalopods, crabs and lobsters) or 20% (for the other products) was then applied for the entire time period for which exports were thought to have occurred (1970s-2008) in order to calculate the domestic subsistence component. Finally, the average subsistence catch rates for the first three years of exports were applied to the number of inhabitants prior to the first year of export, in order to estimate the domestic subsistence component of invertebrate catches back to 1950.  Table 3: Summary of parameters used for the estimation of small-scale catches. Coastal Areaa Time period Number of fishers CPUE Fishing daysb Catches (t) Comment Source South-West 1950 a2,900 6.7 - a5,155 - 1988 16,000 5.8 - 24,110 - 1991 18,100  5.0*  - 23,574  - Laroche (1997) 1950-1988 - - 260 - Stable CPUE 2008 40,100 2.1 - 21,853 - 1988-2008 - - 260 - -5% CPUE∙year-1 The remainder  1950 a5,300 6.1 - a8,300 - 1988 28,500 5.7 - 42,347 - 1950-1988 -  260 - Stable CPUE 2002 55,300  4.3* - 62,178  - Doukakis (2007) 2008 71,600 3.8 - 71,344 - 1988-2008 - - 260 - -2% CPUE∙year-1  *Anchor points; a See Figure 2 for area definitions; bAssumed average number of fishing days in Madagascar (McVean, 2006; Gough and Humber, unpub. data; G. Hosch, pers. comm. Fisheries Planning and Management).  Small-scale fisheries: subsistence and artisanal catches  Between 1950 and 2008, landings of tuna-like species, narrow-barred Spanish mackerel (Scomberomorus commersoni) and miscellaneous marine fish represented the bulk of the data reported to FAO, accounting 28 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   on average for 82% of total seafood landings in Madagascar. They are exclusively caught by artisanal fishers, as there are no industrial fisheries targeting these species.Unfortunately, estuarine catches were reported by Madagascar as inland catches until 1989, after which they were accounted for in the marine landings (Stamatopoulos and Rafalimanana, 1991). The same authors also report these estuarine catches to be 30,000 tonnes per year in the late 1980s. Finally, the official records of fish caught in the 1950s would signify a highly unrealistic local consumption of approximately 0.7 kg∙person-1∙year-1 (based on a coastal population representing 90% of the total human population). Considering the change in reporting protocol and this unrealistic consumption rate, we replaced the data supplied to FAO (tuna-like species, mackerel and miscellaneous marine fish) and re-estimated small-scale catches based on CPUE and fishing effort data from independent studies (Laroche and Ramananarivo, 1995; Laroche et al., 1997; Doukakis et al., 2007), in combination with the number of fishers for the two regions as defined in Figure 2. Based on local knowledge and reports, which suggest decreases in CPUE over time (Bellemans, 1989; Laroche et al., 1997; Frontier Madagascar, 2003; Langley, 2006; Doukakis et al., 2007; Gough et al., 2009), we applied different CPUE estimates for areas and time periods as shown in Table 3.  Other fisheries  A commercial joint venture fisheries operation between Japan and Madagascar was established in the 1970s under the name of Compagnie Malgache Nippone de Pêcherie (COMANIP) for the exploitation of Madagascar‘s skipjack tunas, Katsuwonus pelamis (Marsac and Stequert, 1984; Gilbert and Rabenomanana, 1996). Independent catch data were not available; therefore we considered the data reported to FAO as reliable and included them in the final result without modification. Thus, we assumed no underreporting of skipjack tuna catches during this period. This is likely conservative, given the known occurrence of substantial and widespread underreporting of tuna catches.  An exploratory deep-sea fishery in the Malagasy EEZ started in 2001. There is only one value provided by the FAO of 4157 tonnes for the year 2002. This sector is described by Soumi (2004) to have increased without any further indication; therefore, we assumed a growth rate of 5% per year between 2002 and 2008.  In the last two decades, fishing tourism has rapidly expanded in Madagascar (Jain, 1995). Most of the catch, dominated by large pelagic species such as marlin and tuna, is not catch-and-release, and is therefore killed and retained. However, none of the 60 people contacted in this study (employees or managers of sport-fishing charter companies) were willing to share information with us. As a result we were unable to quantify extractions made by this sector. In terms of overall tonnage, this sector is likely to be small, but may have effects on the population structure of these species, especially the billfishes.  Foreign fishing in Madagascar’s waters  Since 1986, a fishing agreement has been in place with the European Commission, allowing EU purse- seine vessels to catch tuna in Malagasy waters (Gilbert and Rabenomanana, 1996). Catches of 10,000 tonnes have been declared each year since 1986. However, given that licence fees are based on this tonnage, it is highly likely that catches are largely under-reported, and may actually be around 18,000- 20,000 tonnes per year (Anon., 2002). It is also interesting to note that a substantial Asian long-line fleet has been fishing in Malagasy waters most of the time illegally without access agreements, with entirely un-reported catches of tuna. Anon. (1995) reported legal catches of 6,000-8,000 tonnes. However, some estimates are up to 50,000 t∙year-1 (Fowler, 2005). Indeed, Malagasy authorities do not possess the resources to patrol their own EEZ and therefore cannot address the problem of illegal fishing for such high-value species (Jain, 1995; Cooke, 1997; A. Harris, pers. comm., Blue Ventures Conservation; G. Hosch, pers. comm., Fisheries Planning and Management). Given that no formal access agreements exist between Madagascar and these countries, these catches are illegal under international law. For this long- line fleet, 7.5% of the bycatch is composed of sharks (Cooke, 1997; Fowler, 2005), of which only fins are retained. Finally, a longline fleet, whose catches are uncertain (Fowler, 2005), is operating from La Réunion (René et al., 1998) and is targeting tuna and other large pelagic species (e.g., swordfish, marlin). For our purposes, we assumed this longline fleet started fishing in Malagasy waters around 1990 and catches increased linearly to 5,000 t∙year-1 in 2008. Cases of this sort of unreported fishing are sometimes Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  29   covered by media journals (Anon., 2010b) but have remained largely unaddressed by Malagasy authorities or Regional Fisheries Management Organizations (Cullis-Suzuki and Pauly, 2010). These catches, important in an ecosystem sense, are estimated here and listed in Appendix Table A1, but are not included in our reconstruction of Malagasy fisheries catches, since they are made by foreign countries.   RESULTS  Reconstructed catches by sector, taxa and year are presented in Appendixes Tables (A1, A2).  Shrimp fishery  Total shrimp catches, which were very low prior to 1967, increased until the late 1990s- early 2000s, with peak catches of 12,850 t∙year-1 in 2003 (Figure 3a). Industrial landings stabilised in the 1990s at around 8,500 t∙year-1, with a peak of 9,850 t∙year-1 in 1998. They have declined substantially since 2002, with only 5,200 t caught in 2008. Similarly, small-scale catches increased from approximately 300 t∙year-1 during the 1950s, to 750 t∙year-1 in 1970 and to 3,500 t∙year-1 in 2005. Thereafter, they underwent a gradual decline to around 2,700 t∙year-1 in 2008 (Figure 3a).  Bycatch also followed a similar trend, and reached a maximum of 38,800 t∙year-1 in 1998, of which 25,500 t were discarded (72%). Since then, bycatch has decreased to around 8,000 t∙year-1 in 2008, of which 4,400 tonnes were discarded (62%; Figure 3a).  Over the 1950-2008 time-period, the artisanal component represents 27% and the industrial sector 73% of total shrimp catches.  Shark fishery  The reconstructed data for sharks conservatively suggests low catches until 1980, averaging 350 t∙year-1, followed by a rapid and substantial increase, from approximately 500 tonnes in 1980 to a peak of almost 7,000 t∙year-1 in 1992 (Figure 3b). Since then shark catches have decreased to 3,800 t∙year-1 in 2008 (Figure 3b).  Figure 3: a) Shrimp catches by the industrial and artisanal sectors, and the associated bycatch (landed and discarded); b) total shark catches by small-scale fishers and c) catches for non-shrimp invertebrates, separated by exports or commercial (light grey), and subsistence (dark grey) catches (dotted line represents the data supplied to FAO). 30 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Non-shrimp invertebrate fishery  Total invertebrate catches increased from 1,500 t∙year-1 in 1950 to 4,000 t∙year-1 in 1975 and were deemed to be exclusively for subsistence purposes during that time (Figure 3c). Since 1975, total catches have comprised both a subsistence and a commercial (export) component and, although fluctuating over time, have steadily increased to approximately 16,000 t∙year-1 by 2008 (Figure 3c). Over the 1950-2008 time-period, invertebrate catches have totalled 193,800 tonnes, of which the subsistence component represents 45% and commercial exports 55%.  Small-scale finfish fisheries: subsistence and artisanal catches  Reconstructed subsistence and artisanal catches of finfish by small-scale fishers have increased steadily over the 1950-2008 time-period (Figure 4). Total subsistence and artisanal catches in 1950 were around 13,500 t∙year-1 and have increased to around 93,000 t∙year-1 by 2008 (Figure 4).   Overall reconstruction  Over the whole 1950-2008 time-period, total catches taken by Malagasy fishers in Madagascar‘s EEZ are estimated at 4.7 million tonnes. Thisreconstructed total is twice as high as the data supplied to the FAO by the government of Madagascar (Figure 5). Significantly, the re-estimation added over 550% for the earlier time-period (1950s), but adjusted the reported data by only 39% for the 2000-2008 period (Figure 5a). The taxonomic composition of reconstructed total catches shows a constant attern over time (Figure 5b).   Figure 4: Total reconstructed catches of the small-scale finfish fishery, showing the taxonomic breakdown (based on Laroche et al., 1997). Approximately half of these catches are for subsistence, and the other half for sale on the local market (C. Gough and F. Humber, pers. obs.). Figure 5: a) Total reconstructed catches versus reported landings as supplied to FAO by Madagascar; and b) Taxonomic composition of the overall reconstructed catches. Fenneropenaeus indicus is a shrimp species; all the other groups (except the mixed-group ‗others‘) are from the small-scale, finfish fishery. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  31    Foreign fishing in Madagascar’s waters  Tuna catches taken by foreign vessels are thought to have increased substantially since the mid-1980s. Current catches are somewhat uncertain due to obvious unreported and illegal catches. A minimum estimate is over 70,000 t∙year-1 (Figure 6). Catches are likely dominated by Asian longline fleets operating illegally, with catches having increased to 50,000 t∙year-1 by 2008. The EU purse- seine fleet, although operating legally through access agreement, is known to substantially under-report by at least 100%, with official catches reported as around 10,000 t∙year-1, while estimated actual catches are around 18,000 t∙year-1 (Figure 6). Note that the publication by Le Manach et al. (2012) contains a mislabelled version of Figure (6). The present version is labelled correctly.   DISCUSSION  Overall, reported data show a steady increase in landings, due to the expansion of industrial fleets in the early 1980s. According to our analysis, Malagasy fisheries had been underreported by over 500% in the early time period, and seem to be underreported at present by at least 40%. The reporting is therefore improving, but current values, likely under-estimates, are still very substantial.  Our reconstructed catch time-series shows a levelling off of catches over the last two decades (Figure 5). It is worth noting that this levelling off of total catches is partly related to the improvement of bycatch handling by the shrimp industry causing decreased bycatch/discards since the 1990s. However, it is also certainly due to decreases in catches seen in various invertebrates (including shrimp) and shark fisheries, which suggest that overfishing is likely to be taking place.  The official data also fail to account for a large part of the subsistence fishery, which represents 75% of the total reconstructed catch over the whole period, and 83% for the period 1950-1980. Such marginalization of small-scale fisheries is common (Pauly, 1997), although inclusion in official statistics is crucial. As a consequence of this marginalization, total Malagasy catches may be approaching or even exceeding sustainable yields for coastal stocks, estimated at 180,000 t∙year-1 (Anon., 2008b), with it remaining undetected.  Another issue that has been dealt with in this report is the poor taxonomic information included in official statistics supplied to FAO. The major group in these official statistics represents over 80% of total catches, and is only described as ‗marine fish nowhere else included‘. Although we made a taxonomic breakdown of total catches, information related to species composition remains extremely poor. This fact justifies the importance of the implementation by FAO of taxonomic census every three to five years in order to create more reliable species composition times-series.  Although not included in the total reconstruction, we also reviewed foreign fisheries in Madagascar‘s EEZ. Current catches of tuna, billfishes and sharks are likely to be over 70,000 t∙year-1, most of which are made illegally. This situation raises serious legal questions, and also points to the issue of inappropriate low fishing access fees paid by developed countries (Kaczynski and Fluharty, 2002; Hanish and Tsamenyi, Figure 6: Estimated time-series of tuna catches in Madagascar‘s EEZ. 32 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   2009) and poor to non-existing monitoring and enforcement of such agreements (e.g., Jain, 1995). Indeed, the monitoring and enforcement system for the entirety of Madagascar is only composed of 3 monitoring vessels, 8 speedboats, 18 inspectors and 22 observers (R. Fanazava, pers. comm., Centre de Surveillance des Pêches). This lack of monitoring and enforcement capability has led to increasing illegal pirate fishing in the waters of Madagascar, as evidenced here, which likely contributes significantly to unsustainable fishing practices in the Western Indian Ocean.  For the large-scale commercial shrimp industry, no real discrepancies exist between landings reported to FAO and our re-estimated landings. However, the overall CPUE is decreasing, possibly because catches have been significantly higher (by up to 5,000 t∙year-1) than the estimated maximum sustainable yield of 8,700 tonnes∙year-1 (Kasprzyck, 2008). It is worth noting that no values were reported to the FAO before 1964, when the first exploratory trawls were conducted. Prior to this date, local people were nevertheless fishing and consuming shrimp, and our study has filled this gap by assigning a subsistence component to this sector, although this was negligible compared to total catches. Also, the significant bycatch produced by shrimp trawlers is missing from official data, and this issue needed to be addressed since this bycatch is often made unavailable to the local population, when not collected by artisanal fishers. Our study highlights the importance of such bycatch for Madagascar since the beginning of this fishery.  The reconstructed time-series of shark catches gives a very different picture to the official data. The former considers that an artisanal fishery has existed since at least 1950, while the latter show little indication that a shark fishery exists despite this being fairly well documented in the independent literature (see Petit, 1930; Cooke, 1997). These values are based on the fin trade using strong assumptions and are considered highly conservative given that Hong Kong and Singapore do not account for 100% of the fin market. However, they are likely to be closer to actual catches than the previous assumption in which a lack of data has been incorrectly interpreted as no catch.  Concerning the small-scale finfish fishery, our reconstruction provides very similar estimates to those provided to the FAO for the 1989-2008 period. For 1950-1989, our results however differ greatly, as they fill the gap made by estuarine catches accounted for in reported inland catches before 1989. This misreporting of estuarine catches is documented but has never been incorporated into official statistics. Importantly, there are increasing concerns about the rate of growth in small-scale catches slowing, eventually leading to declining catches by the small-scale, artisanal and subsistence fisheries. This finfish fishery supplies the bulk of local seafood consumption demand, as most of the catches are sold and consumed locally, and declines in catches could have significant impacts on the food security of coastal communities.  Throughout this report, historical events and changes in reporting protocols illustrate the importance of linking historical information and fisheries data to current management plans, especially in maritime developing countries such as Madagascar, where fisheries are of fundamental importance for the food security of the people. The consequences of diminishing fisheries are likely to be particularly severe in an island nation in which over 50% of children under five years of age suffer delayed development due to a chronically inadequate diet, and where chronic food insecurity affects over 65% of the population (Back- Michaud et al., 2009; Anon., 2010c).   ACKNOWLEDGMENTS  We thank of the Sea Around Us Project, a scientific collaboration between the University of British Columbia and the Pew Environment Group. We also thank Alasdair Harris, Gilles Hosch, Phaedra Doukakis, Rojo Rambonimanana, Christian Chaboud and the Wildlife Conservation Society for information and knowledge sharing.  REFERENCES  Anderson, S.C., Mills Flemming, J., Watson, R. and Lotze, H.K. (2010) Serial exploitation of global sea cucumber fisheries. Fish and Fisheries, 23 p. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  33   Anon. (1995) IOTC proceeding of the 6th Expert Consultation on Indian Ocean Tunas, Colombo (Sri Lanka), 25-29 September 1995. [in French]. 65p. Anon. (2000) Conversion factors - landed weight to live weight. FAO Fisheries Circular 847, Revision 1. FIDI/C847. United Nations Food and Agriculture Organization, Rome. 176 p. Anon. (2001) Report on the Maldivian fishing industry - 1960. In: Waheed, Z. (ed.) Maldives Marine Research Bulletin, 5.  101p. Anon. (2002) Madagascar environmental threats and opportunities assessment. United States Agency for International Development (USAID. 118p. Anon. (2003) Décret N°2003/1101 modifiant certaines dispositions du décret n°71-238 du 12 mai 1971, réglementant l'exercice de la pêche par chalutage. In: la mer territoriale malgache. [in French]. Anon. (2008a) Madagascar environmental threats and opportunities assessment - an update. United States Agency for International Development (USAID), 61 pp. Anon. (2008b) Madagascar - National Fishery Sector Overview. United Nations Food and Agriculture Organization, Rome. Anon. (2010a) Pêche illégale: Un bateau chinois arraisonné à Toliara. La Gazette, September 8th, 2010.  Available at : www.lagazette-dgi.com/index.php?option=com_content&view=article&id=5927:peche-illegale-un-bateau- chinois-arraisonne-a-toliara&catid=45:newsflash&Itemid=58. [Date accessed: February 25th 2011]. [in French]. Anon. (2010b) Une pêche exceptionnelle pour l‘Atsantsa.  Le journal de l‘ile, 18 septembre 2010. Anon. (2010c) Enquête démographique et de santé de Madagascar 2008-2009. INSTAT et ICF Macro, Antananarivo. [in French]. Barnes, D.K. and Rawlinson, K.A. (2009) Traditional coastal invertebrate fisheries in south-western Madagascar. Journal of the Marine Biological Association of the United Kingdom, 89(08): 1589-1596. Bellemans, M.S. (1989) Estimations premières de la production des pêcheries traditionnelles côtières malgaches. UNDP/FAO Project MAG/85/04. United Nations Food and Agriculture Organization. 95p. [in French]. Billé, R. and Mermet, L. (2002) Sectoralization of an integrated coastal management programme: A case study in Madagascar. Journal of environmental planning and management, 45(6): 913-926. Black-Michaud, A., de Meulder, F., Randrianasolo, H. and Conan, C. (2009). Rapport d‘évaluation de l‘intervention prolongée de secours et de redressement à Madagascar (Ipsr 10442.0) – Aide pour faire face aux catastrophes naturelles et à l‘insécurité alimentaire saisonnière (Juillet 2006 – Juin 2009). Rome: Programme Alimentaire Mondial. [in French]. Brooks, T.M., Mittermeier, R.A., da Fonseca, G.A.B., Gerlach, J., Hoffmann, M., Lamoreux, J.F., Mittermeier, C.G., Pilgrim, J.D. and Rodrigues, A.S.L. (2006) Global biodiversity conservation priorities. Science, 313: 58–61. Central Intelligence Agency (2010) The world factbook – Madagascar. Available at: www.cia.gov/library/publications/the-world-factbook/geos/ma.html. [accessed: October 12th 2010]. Chen, C. and Chow, Y. (2001) Examining on the bycatch and discard catch of the shrimp beam trawl fishery in the coastal waters off Chiayi, southwest Taiwan. Journal of the Fisheries Society of Taiwan, 28(3): 229-248. Cooke, A., Jonahson, M., Doukakis, P., Smale, M. and du Feu, T. (2001) Sharks, shark fisheries and shark fin trade in Madagascar – review and analysis, with suggestions for action. Unpublished. 9p. Cooke, A., Lutjeharms, J.R.E. and Vasseur, P. (2003) Marine and coastal ecosystems. p. 179-208. In: Goodman, S.M. and Benstead, J.P., (eds.) The natural history of Madagascar, University of Chicago Press, Chicago. Cortes, E. and Neer, J.A. (2006) Preliminary reassessment of the validity of the 5% fin to carcass weight ratio for sharks. Collective Volume of Scientific Papers ICCAT, 59(3): 1025-1036. Cripps, G. (2009) Understanding migration amongst small-scale fishers in Madagascar. Blue Ventures Conservation Report for ReCoMaP. 165 p. Cripps, G. (2010) Feasibility study on the protection and management of the Barren Isles ecosystem, Madagascar. Blue Ventures Conservation Report (2009), for WWF and the "Reseau interdisciplinaire pour une gestion durable de la biodiversite marine: diagnostic environnemental et social autour des tortues marines dans le sud-ouest de l‘Océan Indien ". 272 p. Crosnier, A. (1965) Les crevettes Pénaeidés du plateau continental Malgache, état de nos connaissances sur leur biologie et leur pêche en septembre 1964. Cahiers ORSTOM, série Océanographie (Suppl.), 3(3): 1-158. [in French]. Cullis-Suzuki, S. and Pauly, D. (2010) Failing the high seas: A global evaluation of regional fisheries management organizations. Marine Policy 34: 1036-1042. Doukakis, P., Jonahson, M., Ramahery, V., Randriamanantsoa, B.J.D. and Harding, S. (2007) Traditional fisheries of Antongil Bay, Madagascar. Western Indian Ocean Journal of Marine Science, 6(2): 175-181. Fennessy, S.T. and Isaksen, B. (2007) Can bycatch reduction devices be implemented successfully on prawn trawlers in the Western Indian Ocean? African Journal of Marine Science, 29(3): p.453-463. Food and Agriculture Organization (2009) Fishstat Plus: Universal software for fishery statistics time series, version 2.3. Fourmanoir, P. (1952a) Compte rendu des nouvelles expériences de pêche du « Gabriel II » effectuées du 24 juillet au 15 novembre 1951. Bulletin de Madagascar, 52: 17-21. [in French]. 34 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Fourmanoir, P. (1952b) Les crevettes d'intérêt économique à Madagascar. Le Naturaliste Malgache, 4(2): 163-168. [in French]. Fowler, S.L. (2005) Sharks, rays and chimaeras: the status of the Chondrichthyan fishes: status survey. World Conservation Union. 476p. Frontier-Madagascar (2003) Fin-fish resource use: artisanal fisheries of Beheloka. Woods-Ballard, A.J., Chiaroni, L.D., and Fanning, E., (eds.) Frontier-Madagascar Environmental Research Report 11. Society for Environmental Exploration, UK and the Institute of Marine Sciences, University of Toliara, Madagascar. 24 p. Gilbert, F. and Rabenomanana, L. (1996) Aperçu sur la pêche thonière à Madagascar. Appendix of the IOTC proceeding of the 6th Expert Consultation on Indian Ocean Tunas,  Colombo (Sri Lanka), 25-29 September 1995. [in French]. 3p. Giri and Muhlhausen (2008) Mangrove forest distributions and dynamics in Madagascar (1975–2005). Sensors, 8: 2104-2117. Globalis (2010) Madagascar. Available at : http://globalis.gvu.unu.edu/country.cfm?country=MG. [Date accessed: October 12th 2010]. Gough, C., Thomas, T., Humber, F., Harris, A., Cripps, G., and Peabody, S. (2009) Vezo Fishing: An Introduction to the Methods Used by Fishers in Andavadoaka Southwest Madagascar. Blue Ventures Conservation Report. 37 p. Guidicelli M. (1984). Les Pêcheries maritimes malgaches: leurs principaux potentiels et leurs besoins pour le développement. SWIOP/Document OISO, RAF/79/065/WP/17/84, Seychelles. 112 p. [in French] Hanich, Q. and Tsamenyi, M. (2009) Managing fisheries and corruption in the Pacific Islands region. Marine Policy, 33(2): 386-392. Horning, N.R. (2008) Strong support for weak performance: donor competition in Madagascar. African Affairs, 107: 405-431. Hosch, G., Ferraro, G. and Failler, P. (2011) The 1995 FAO code of conduct for responsible fisheries: adopting, implementing or scoring results? Marine Policy, 35: 189-200. Jacquet, J., Alava, J.J., Ganapathiraju, P., Henderson, S. and Zeller, D. (2008) In hot soup: sharks captured in Ecuador‘s waters. Environmental Sciences; 5(4): 269-283. Jacquet, J., Fox, H., Motta, H., Ngusaru, A. and Zeller, D. (2010) Few data but many fish: marine small scale fisheries catches for Mozambique and Tanzania. African Journal of Marine Science, 32(2): 197-206. Jain, M. (1995) Inventory of marine and coastal activities - a strategy proposal for Madagascar. knowledge and effective policies for environmental management (KEPEM). Report no. 20. 103 p. Jury, M.R. (2003) Climate: the climate of Madagascar. p. 75-87. In: Goodman, S.M. and Benstead, J.P., (eds.) The natural history of Madagascar. University of Chicago Press, Chicago. Kaczynski, V.M. and Fluharty, D.L. (2002) European policies in West Africa: who benefits from fisheries agreements? Marine Policy, 26: 75-93. Kasprzyck, Z. (2008) Shrimp fishing in Madagascar. p. 223-234. In: Gillett, R., (ed.) Global study of shrimp fisheries. FAO Fisheries Technical Paper 475. United Nations Food and Agriculture Organization, Rome. Kelleher, K. (2005) Discards in the world's marine fisheries: an update. FAO Fisheries Technical Report 470. United Nations Food and Agriculture Organization, Rome. 154 p. Langley, J. (2006). Vezo knowledge: traditional ecological knowledge in Andavadoaka, southwest Madagascar. 73 p. Laroche, J. and Ramananarivo, N. (1995) A preliminary survey of the artisanal fishery on coral reefs of the Tulear Region (southwest Madagascar). Coral Reefs, 14(4): 193-200. Laroche, J., Razanoelisoa, J., Fauroux, E. and Rabenevanana, M.W. (1997) The reef fisheries surrounding the south- west coastal cities of Madagascar. Fisheries Management and Ecology, 4: 285-299. Le Manach, F., Gough, C., Harris, A., Humber, F., Harper, S. and Zeller, D. (2012) Unreported fishing, hungry people and political turmoil: the recipe for a food security crisis in Madagascar? Marine Policy 36: 218-225. Marsac, F. and Stequert, B. (1984) Bilan des prospections sur les ressources thonières de surface effectuées depuis 1971 dans l'Océan Indien Occidental. La pêche maritime, 1271: 83-94 McNeish, H. (2011) Madagascar‘s UNIMA to close shrimp factory in March. Bloomberg. Available at: http://www.businessweek.com/news/2011-01-26/madagascar-s-unima-to-close-shrimp-factory-in- march.html. [Date accessed: February 25th 2011]. McVean, A.R., Walker, R.C. and Fanning, E. (2006) The traditional shark fisheries of southwest Madagascar: a study in the Toliara region. Fisheries Research, 82(1-3): 280-289. Murawski, S.A. (2010) Rebuilding depleted fish stocks: the good, the bad, and, mostly, the ugly. ICES Journal of Marine Science, 67(9): 1830-1840. Pauly, D. (1997) Small-scale fisheries in the tropics: marginality, marginalization, and some implications for fisheries management. In: Pikitich, E.K., Huppert, D.D. and Sissenwine, M., (eds.) Global Trends: Fisheries Management. American Fisheries Society, 20: 40-49. Petit, C. (1930) L‘industrie des pêches à Madagascar. Société d'Editions Géographiques, Maritimes et Coloniales, Paris. 392 p. [in French]. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  35   Rafalimanana, T. (1991) Estimation des captures de la pêche traditionnelle malgache en 1990. Ministère de la production animale, eaux et forêts. Direction de la pêche et de l‘aquaculture. UNDP Project MAG/85/014. Field report, 27. 84 p. [in French]. Randriarilala, F., Rafalimanana, T. and Caverivière, A. (2008) Les captures accessoires des crevettiers industriels et artisanaux. p. 237-258. In : Caverivière, A. Chaboud, C. and Rafalimanana, T., (eds.) Les crevettes côtières de Madagascar : biologie, exploitation, gestion.  IRD Editions, Marseille. [in French]. Rasolofonirina R. and Conand C. (1998) L‘exploitation des holothuries dans le sud-ouest de Madagascar, région de Toliara La bêche-de-mer. Bulletin de la CPS, 10:10-13. Razafindrainibe, H. (2010) Baseline study of the shrimp trawl fishery in Madagascar and strategies for bycatch management, Project TCP/MAG/3201 - REBYC2. United Nations Food and Agriculture Organization, Rome. 111 p. René, F., Poisson, F. and Tessier, E. (1998) Evolution de la pêcherie palangrière ciblant l'espadon (Xiphias gladius) à partir de La Réunion. In: Cayré P, Le Gall JY (eds.). Le thon : enjeux et stratégies pour l'océan Indien. ORSTOM, Paris. [in French]. Roullot, J. (1989) Utilization of shrimp fisheries by-catch in Madagascar. Project RAF/87/008. United Nations Food and Agriculture Organization, Rome. 37 p. Rousse, E. (2010) La Grande île, de 1642 à 1960. Le quotidien de la Réunion et de l'océan Indien. Available at: http://www.lequotidien.re/actualites/ocean-indien/120683-la-grande-ile-de-1642-1960.html [Date accessed: January 5th, 2011]. [in French]. Schraeder, P.J. (1995) Madagascar. In: Chapin Metz, H. (ed.) Indian Ocean: Five island countries, 3rd ed., Federal Research Division, Library of Congress, Washington D.C. Soumy, M. (2006) Country review - Madagascar. In:  De Young, C. (ed.) Review of the state of world marine capture fisheries management: Indian Ocean. United Nations Food and Agriculture Organization, Rome. 458 p. Southern African Development Community (2008) Study and analysis of the status of IUU fishing in the SADC region and an estimate of the economic, social and biological impacts. Volume 2 - Main report. 72 p. Spalding, M.D., Ravilious, C. and Green, E.P. (2001) World atlas of coral reefs. UNEP World Conservation Monitoring Centre, University of California Press, Berkeley, 416 p. Stamatopoulos, C. and Rafalimanana, T. (1991) Observations sur le système statistique de la pêche traditionnelle (maritime et continentale) à Madagascar. Project MAG/85/014. United Nations Food and Agriculture Organization, Rome. 12 p. [in French]. Sumaila, U.R., Alder, J. and Keith, H. (2006) Global scope and economics of illegal fishing, Marine Policy, 30(6): 696-703. Tucker, B., Tsimitamby, M., Humber, F., Benbow, S. and Iida, T. (2010) Foraging for development: a comparison of food insecurity, production, and risk among farmers, forest foragers, and marine foragers in southwestern Madagascar. Human Organization, 69(4): 375-386. Wielgus, J., Zeller, D., Caicedo-Herrera, D. and Sumaila, U.R. (2010) Estimation of fisheries removals and primary economic impact of the small-scale and industrial marine fisheries in Colombia. Marine Policy, 34: 506-513. World Bank (2010) Madagascar country profile. Available at: http://data.worldbank.org/country/madagascar. [Date accessed: October 12th, 2010]. Zeller, D., Booth, S., Davis, G. and Pauly, D. (2007) Re-estimation of small-scale fishery catches for U.S. flag- associated island areas in the western Pacific: the last 50 years. Fisheries Bulletin, 105: 266-277.   36 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller    Appendix Table A1: Annual catches by domestic and foreign fisheries in Madagascar‘s EEZ, 1950-2008. Year Domestic fisheries (t)  Foreign fisheries (t) Data reported to FAO Reconstructed catch Purse-seine (EU) Longline (Asia) Longline (France) 1950 2400 14295  0 0 0 1951 2500 15170  0 847 0 1952 2500 16070  0 1695 0 1953 2600 16995  0 2542 0 1954 2600 17946  0 3390 0 1955 2600 18923  0 4237 0 1956 2600 19925  0 5085 0 1957 3000 20942  0 5932 0 1958 3000 21447  0 6780 0 1959 3500 22811  0 7627 0 1960 4000 24205  0 8475 0 1961 4000 25243  0 9322 0 1962 3500 26304  0 10169 0 1963 4000 27387  0 11017 0 1964 5501 28491  0 11864 0 1965 7801 29617  0 12712 0 1966 17500 30883  0 13559 0 1967 16600 36787  0 14407 0 1968 16900 40937  0 15254 0 1969 13400 44617  0 16102 0 1970 15100 54319  0 16949 0 1971 15800 62029  0 17797 0 1972 16200 66801  0 18644 0 1973 22401 72408  0 19492 0 1974 28701 77014  0 20339 0 1975 19501 69410  0 21186 0 1976 18451 68934  0 22034 0 1977 19760 74354  0 22881 0 1978 18160 75462  0 23729 0 1979 17260 74619  0 24576 0 1980 17373 79804  0 25424 0 1981 16875 83133  0 26271 0 1982 20455 87439  0 27119 0 1983 21195 91293  0 27966 0 1984 35038 96501  0 28814 0 1985 35112 101705  0 29661 0 1986 44353 107033  520 30508 0 1987 52488 114869  1040 31356 0 1988 61141 111675  1550 32203 0 1989 67731 112374  8125 33051 0 1990 73515 115101  7400 33898 263 1991 71438 121887  8000 34746 526 1992 77021 118912  11250 35593 789 1993 84317 126392  14500 36441 1053 1994 86618 137783  14889 37288 1316 1995 85840 128033  15278 38136 1579 1996 84644 132157  15667 38983 1842 1997 86547 138809  16056 39831 2105 1998 84405 139285  16444 40678 2368 1999 87638 129942  16833 41525 2632 2000 90167 137120  17222 42373 2895 2001 93615 140240  17611 43220 3158 2002 99326 140689  18000 44068 3421 2003 99671 140276  18000 44915 3684 2004 103416 138493  18000 45763 3947 2005 99986 128148  18000 46610 4211 2006 100943 133449  18000 47458 4474 2007 115148 134454  18000 48305 4737 2008 87834 131771  18000 50000 5000 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  37     Appendix Table A2: Six most important taxa caught by domestic fisheries in Madagascar‘s EEZ, 1950-2008. Year Hemiramphidae  Scombridae  Fenneropenaeus indicus Haemulidae  Lethrinidae  Elasmobranchii  Others  1950 3043 1246 274 677 1049 605 7402 1951 3232 1323 291 719 1114 642 7850 1952 3426 1403 308 762 1181 681 8310 1953 3626 1484 325 806 1249 721 8784 1954 3831 1568 343 852 1320 761 9270 1955 4042 1655 361 899 1393 803 9770 1956 4259 1744 380 947 1468 846 10282 1957 4481 1835 389 996 1544 890 10806 1958 4587 1878 414 1020 1581 912 11054 1959 4882 1999 440 1086 1682 970 11752 1960 5184 2123 459 1153 1787 1030 12469 1961 5409 2215 479 1203 1864 1075 12998 1962 5639 2309 499 1254 1943 1121 13539 1963 5874 2405 519 1306 2024 1167 14091 1964 6113 2503 540 1359 2107 1215 14654 1965 6357 2603 561 1414 2191 1263 15228 1966 6606 2705 689 1469 2277 1313 15824 1967 7002 2867 1436 1964 2413 1395 19711 1968 7407 3033 1856 2284 2553 1477 22328 1969 7823 3203 2183 2554 2696 1561 24598 1970 8367 3426 3522 3388 2884 1675 31057 1971 8926 3655 4489 4022 3076 1790 36071 1972 9065 3711 5271 4460 3124 1821 39348 1973 9203 3768 5309 4491 3172 1849 44618 1974 9341 3824 5068 4369 3219 1875 49318 1975 9478 3880 5242 4450 3266 1902 41190 1976 9818 4020 5107 4475 3383 1970 40162 1977 10164 4161 5603 4857 3503 2041 44025 1978 10515 4305 5466 4885 3624 2110 44556 1979 11050 4524 4823 4647 3808 2214 43554 1980 11595 4747 5296 5074 3996 2324 46771 1981 12152 4975 5340 5249 4188 2621 48608 1982 12721 5208 5503 5502 4384 3003 51119 1983 13300 5445 5581 5708 4584 3381 53294 1984 13891 5687 5846 6043 4787 3758 56489 1985 14494 5934 6078 6380 4995 4141 59683 1986 14653 5999 6539 6823 5050 4394 63574 1987 14803 6061 7216 7366 5102 4860 69461 1988 15022 6151 6427 6905 5177 5013 66980 1989 15232 6236 6438 6952 5249 4834 67432 1990 15432 6318 6415 6283 5318 5156 70178 1991 15530 6358 6973 6713 5352 5411 75549 1992 15608 6391 6419 6363 5379 5969 72781 1993 15681 6420 7477 6991 5404 5824 78596 1994 16095 6590 7884 7042 5547 6258 88369 1995 16494 6753 7003 6764 5684 5257 80078 1996 16878 6910 7211 6890 5817 5829 82622 1997 17248 7062 8200 7421 5944 5809 87125 1998 17605 7208 8561 7744 6067 5788 86312 1999 17948 7348 7537 7047 6185 5756 78121 2000 18278 7483 8195 7405 6299 5736 83725 2001 18743 7674 7990 7011 6459 5735 86627 2002 19193 7858 8942 6172 6614 5728 86182 2003 19620 8033 9098 6203 6762 5727 84834 2004 20031 8201 8363 6144 6903 5725 83124 2005 20427 8363 7189 5415 7040 5715 73998 2006 20807 8519 7233 5558 7170 5714 78449 2007 20902 8558 6554 5463 7203 5665 80109 2008 20991 8594 5717 5323 7234 5615 78298 38 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller    Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  39    RECONSTRUCTION OF MARINE FISHERIES CATCHES FOR MAURITIUS AND ITS OUTER ISLANDS, 1950-20081  Lea Boistol, Sarah Harper, Shawn Booth and Dirk Zeller Sea Around Us Project, Fisheries Centre, University of British Columbia, 2202 Main Mall, Vancouver, V6T 1Z4, Canada l.boistol@fisheries.ubc.ca; s.harper@fisheries.ubc.ca; s.booth@fisheries.ubc.ca; d.zeller@fisheries.ubc.ca  ABSTRACT  Total marine fisheries catches by Mauritius and its outer dependencies were estimated from 1950 to 2008, and include unreported catches from the small-scale fisheries carried out around the islands of Mauritius, Rodrigues, Agalega and St. Brandon, recreational marine catches, estimates of catches taken by the Mauritian fleets along the Mascarene Ridge, and discards of the tuna purse-seine fishery. Summed for 1950-2008, total marine fisheries catches for Mauritius and its dependencies were estimated to be 682,392 t, which is 42 percent larger than currently reported landings of 478,305 t presented by FAO on behalf of Mauritius. This discrepancy was largely due to better accounting of small-scale catches carried out around Mauritius and Rodrigues islands by part-time fishers. This study illustrates the need for improved reporting of catches including all fisheries sectors in Mauritius, especially for the small-scale sector, which provides food security and a source of income for a large portion of the local population.  INTRODUCTION  Mauritius is an island of volcanic origin located between 20°10′S and 57°31′E about 850 km east of Madagascar (Figure 1). It covers a land area of approximately 1,860 km² and shelters a population of around 1.2 million people. Mauritius is an island state including several dependencies in the Southwest Indian Ocean, namely the island of Rodrigues, the St. Brandon (or Cargados Carajos) group of islands and islets, and the twin islands of Agalega. Mauritius was uninhabited when it was first colonized by the Dutch in 1638, and it has later been under French (1715-1810) and British (1810-1968) rule. The colonization period coincided with large-scale deforestation of the islands for sugar cane farming and the introduction of alien species which have severely damaged the islands‘ ecosystems and indigenous species (Paul, 1987; Sobhee, 2004; Turner and Klaus, 2005). At the time of independence in 1968, the economy was dominated by the sugar industry, and it has later undergone rapid growth and diversification with the development of the textile manufacturing industry and tourism.  Mauritius includes a large Exclusive Economic Zone (EEZ) that is approximately 1.7 million km2 (Figure 1). The fishing sector includes small-scale fisheries in the lagoon and non-lagoon areas around Mauritius, Rodrigues, Agalega and St. Brandon islands, offshore semi-industrial fishing on the oceanic banks along the Mascarene Ridge stretching from St. Brandon to Saya de Malha and around the Chagos Archipelago, and on the high seas targeting migratory tuna stocks. Of late, a semi-industrial fishery targeting pelagic swordfish resources has been active since 1999, and two local vessels recently operated a deep sea demersal trawl fishery in the Southwest Indian Ocean from 2000 to 2006 (Jehangeer, 2006; Anon., 2007a)  In terms of its contribution to GDP and employment (about 10,000 people), fisheries are of limited economic importance to the national economy. However, the lagoon and inshore fisheries are an important source of employment and food security to many coastal communities of Mauritius and on the island of Rodrigues (Hollup, 2000; Vogt, 2001; Sobhee, 2004; Anon., 2007b; Hardman et al., 2007).  1  Cite as: Boistol, L., Harper, S., Booth, S. and Zeller, D. (2011) Reconstruction of marine fisheries catches for Mauritius and its outer islands, 1950-2008. pp. 39-61. In: Harper, S. and Zeller, D. (eds.) Fisheries catch reconstructions: Islands, Part II. Fisheries Centre Research Reports 19(4). Fisheries Centre, University of British Columbia [ISSN 1198-6727]. 40 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Tourists, partly through their recreational fishing activities, also add to the fishing pressure on marine resources (Paul, 1987; Sobhee, 2006). However, catches of the recreational and small-scale fishing sectors are often underreported in the official statistics and especially for the Western Indian Region (Van der Elst, 2005). The FAO FishStat database, which currently offers time series data on marine fisheries landings from 1950 to the present, is based on national statistical data supplied by its member countries. Therefore, the quality of the FAO data depends on efficiency of statistical collection within these countries. FAO data have been the basis for many influential fisheries studies (Pauly et al., 1998), but they are, in fact incomplete (Zeller et al., 2006). On Mauritius and Rodrigues Islands, the lagoon and non- lagoon areas are exploited by many people from different sectors, who fish for commercial, subsistence and recreational purposes. Most are not professional but amateurs, people in search of a meal or to supplement what is generally a meager income. In addition, both professional and amateur fishers use illegal and destructive fishing techniques, such as fine meshed nets, illegal spearguns, dynamites and chemical agents (Ardill, 1979; Paul, 1987; Hollup, 2000; Sobhee, 2004, 2006).  Although such catches are not ignored and have been mentioned in recent and past studies (Paul, 1987; Pearson, 1988), they have never been estimated over a long time period, even though long time series of fisheries catches are necessary to evaluate the ecological effect of fisheries on the marine ecosystems. In this context, the purpose of the present study is to reconstruct the likely total catches of marine resources for the 1950-2008 time period following Zeller et al. (2007), to serve as a scientific baseline in the face of climate change and potential threats to food security.  METHODS  The existing reported catches were first examined. Such data were extracted from the FAO FishStat database, which currently offers time series data on marine fisheries landings from 1950 to 2008, and from national documents or from the Ministry on behalf of the Albion Fisheries Research Centre. For comparison of pelagic and non-pelagic species catches, we grouped Albacore (Thunnus alalunga), Bigeye tuna (Thunnus obesus), Black marlin (Makaira indica), Indo-Pacific sailfish (Istiophorus platypterus), ‗Marlins, sailfishes, etc. nei‘, Striped marlin (Tetrapturus audax), Swordfish (Xiphias gladius), ‗Tuna-like fishes nei‘, Skipjack tuna (Katsuwonus pelamis) and Yellowfin tuna (Thunnus albacares) as ‗pelagic species‘ and the remaining taxa as ‗non-pelagic species‘. We compared the FAO reported catches with those reported at the national level to identify discrepancies between the two. We then identified the missing components (i.e., sectors, time periods, species, gears) not covered by the existing reported catch time series through literature searches and consultations with local experts. After a search for other available and reliable sources to supply the missing catch data, we developed data anchor points in time for missing data. Time series data were reconstructed using interpolations and extrapolations. Data used to form these anchor points range from fisher and human population data, tourists arrivals, to catch per fisher data. Figure 1. Map of the Mauritius Exclusive Economic Zone which includes Mauritius, Rodrigues, Agalega, and St. Brandon Islands. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  41    The marine environment of Mauritius and its outer islands has been exploited since first settlement, but this study is limited to the period of global FAO reporting, i.e., from 1950 onwards. Our reconstruction comprises the following components: a) ‗unreported‘ catches of the small-scale, near-shore fisheries carried out around the islands of Mauritius, Rodrigues and St. Brandon; b) ‗recreational‘ marine catches; c) estimates of the Mauritian banks fishery catches; d) illegal catches taken in the lagoon and non-lagoon waters of Mauritius and Rodrigues islands; and e) ‗discards‘ of the industrial tuna purse-seine fishery. For the purpose of the present study, and according to the data that were available to us, we distinguished between three different categories of local fishers, namely the professional full-time fishers, the part-time professional fishers and the part-time subsistence fishers. Thus, we reconstructed the catches for each of these categories, separating the commercial and subsistence components for each. We reconstructed separately the catches for the different dependencies and fishing areas of the Mauritian state, although FAO data do not distinguish between them.  Human population  Population statistics for Rodrigues and Mauritius Islands were extracted from the Census Statistics Office (CSO) website (www.gov.mu/portal/site/cso) and reports. For years when data were unavailable, population numbers were derived by interpolating linearly between adjacent figures. Thus, a complete time series of the human population was derived from 1950 to 2008 (Figures 2a, b). The numbers of tourist arrivals on Mauritius Island were extracted from several sources (Paul, 1987; Gabbay, 1988; Anon., 2008; Anon., 2009b). Although commercial flights to Mauritius began in 1946, we assumed that tourist‘s arrivals were zero in 1950. Linear interpolations were used to estimate tourist arrivals for intervening years when no data were available (Figure 2c). While the St. Brandon islands shelter no permanent human population, some 300 people live on the twin islands of Agalega (see below).  Mauritius Island small-scale fisheries  Mauritius Island is almost totally surrounded by a fringing coral reef enclosing a lagoon of more than 300 km². Small-scale fishing takes place within the lagoon and non-lagoon areas beyond the reef on the 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1950 1960 1970 1980 1990 2000 P o p u la ti o n  ( x  1 0 6 ) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 1950 1960 1970 1980 1990 2000 P o p u la ti o n  ( x  1 0 4 ) 0 2 4 6 8 10 1950 1960 1970 1980 1990 2000 T o u ri st s (x  1 0 5 ) Year Figure 2. Human population data, 1950-2008 for: a) Mauritius Island; b) Rodrigues Island; and c) Tourist on Mauritius Island. a) b) c) 42 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   narrow shelf area around the island. Fishers use a wide spectrum of fishing gears which ranges from hand collecting to large nets, gill nets, canard nets, hooks and lines, basket traps and harpoons. To this must be added unreported catches taken by illegal fishing methods such as dynamite and pesticides. The main taxa caught include Serranidae (groupers), Siganidae (rabbitfish), Lethrinidae (emperors), Lutjanidae (snappers), Scaridae (parrotfish), Mullidae (goatfish), Mugilidae (mullets), Acanthuridae (surgeon fish), octopus and lobsters. With the use of hands, sticks and other simple gears, part-time fishers can take small fishes, molluscs, crustaceans and other edible marine species. Many use small cast and mosquito nets to catch large numbers of immature fishes (Paul, 1987).  Professional full-time fishers  Professional full-time commercial catches: On Mauritius Island, artisanal fishery landings have been monitored since 1946 (Ardill, 1979). However, the method employed as well as the managerial efficiency has fluctuated over the years, thus changing the reliability of the reported landings over time. Therefore, in order to provide a more accurate estimate of total artisanal catches, we retained the more reliable estimates, which we used as anchor points for our reconstruction of the artisanal fishers commercial catches. Linear interpolations were used between anchors points to derive a complete catch time series for the artisanal catches from 1950 to 2008.  1950-1958: From 1946 to 1958, officials have added to the ―controlled catch‖ an estimated amount of 560 t for the uncontrolled yield. This quantity included ―(1) an amount of fish artisanal fishermen retained for their own consumption, for their ‘curry’ or somewhere between 1 and 1.5 kg daily per fisher, (2) some 200 t thought to be landed at some 17 minor but uncontrolled landing stations and, (3) an estimated 50 t taken by sport fishers and countless amateurs‖ (Paul, 1987). Therefore, we used the official estimates of total catch presented by Paul (1987), from which we subtracted the official estimated amounts retained by the fishers for their own consumption (310 t) and the official estimated catch of the sport and countless amateur fishers (50 t). These will be considered separately (see below).  1959-1976: Two different samples of catch data were available to us. The first represented the artisanal catches estimated from 1960 to 1970 by multiplying the catches gathered by the Protection Service by the raising factor of 1.7. The second referred to a later method, in which the total catch was estimated from 1960 to 1977 by multiplying the controlled catch by 3.44, a correction factor arbitrarily chosen to bring total yield to ‗expected‘ levels (Ardill, 1979; Paul, 1987). Therefore, and given that the coverage of the landing stations became sporadic after 1967 (Ardill, 1979), we assumed that the earlier estimation method provided the more accurate and conservative estimates of catch from 1960 to 1966 and we retained those estimates as anchor points. For 1974, Moal (1975 in Paul, 1987) reported an estimate of 1,100 t, which was not consistent with our other data, so we assumed that this only represents a fraction of the total catch.  1977-2008: For this time period, we retained the artisanal catches estimated by the Fisheries Division as the more reliable estimates. Indeed, since 1977, a frame survey system for the collection of statistics on fish landings has been operational (Ardill, 1979), and the method was described in the literature (FAO, 1983; 1987; Samboo and Mauree, 1987; Anon., 2007b). Catch and effort data are collected on a monthly basis by a team of enumerators covering randomly selected landing stations divided into several strata, and raising factors are used to estimate catch and effort for each stratum separately. For 1982, Paul (1987) estimated the total catch of the professional artisanal fishers at 1,373 t. However, once the subsistence part of the catch was removed, the remaining amount, namely the commercial catch, was consistent with that reported by the Ministry for the same year. Therefore, landings collected by the Fisheries Division were taken as the best estimates for the artisanal commercial catches for this time period.  To estimate the taxonomic composition of the professional fishers‘ commercial catches, we used data from Paul (1987) which provided the estimated changes in species composition of total controlled catch from 1957 to 1982 (Table 1).    Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  43   Table 1. Estimated species composition of total professional full-time fishers‘ catches on Mauritius Island, based on estimated changes in species composition of total controlled catch (Paul 1987). Taxon Name Common Name 1950 - 1957 1958 – 1964a 1965 - 1971 1972 - 1981 a 1982 - 2008 a Lethrinidae Capitaine 3.9 3.4 2.0 2.2 - Lethrinidae Dame Berri 5.4 4.4 4.0 3.1 9.9 Lethrinidae Battardet/Caya 5.3 5.4 3.1 4.4 - Siganidae Cordonnier 17.4 17.0 19 10.9 12.8 Mullidae Rougets 5.2 3.9 4.1 3.1 4.1 Mugilidae Mulets 4.5 3.3 2.9 4.1 7.9 Serranidae  Vieilles 4.9 5.4 5.2 5.2 7.1 Acanthuridae Licornes 8.6 8.6 9.6 15.6 7.0 Octopodidae Octopus 13.2 16.4 17.5 15.8 11.7 Lobsters Langoustes 1.8 2.1 1.6 0.5 0.3 Miscellaneous Others 29.8 30.1 31.0 35.1 39.2 a Note change in time period coverage. The ‗others‘ category (Table 1) includes pelagic fishes. In the 1950s, the catch of clupeoids constituted an important seasonal activity of traditional fishers, but catch of those species followed a sharp decline while the catch of scombroid type fishes increased. Capture of sharks also occurred from the 1950s and 1970s, and 74 t of shark was caught in 1963 (Paul, 1987).  Professional full-time subsistence catches: The custom is for professional fishers to keep a small amount of their daily catch for their own consumption (Gonzalez Manero, 1971; Paul, 1987). This amount has been estimated between 1946 and 1958 as part of the uncontrolled landings. However, we assumed that this amount would not have been included in the official reported catches since 1958. The amount retained by fishers is fairly constant, independent of the size or value of the catch (Gameiro, 2003). Therefore, for the 1959-2008 period, rather than using a percentage of the total commercial catch, we used the number of artisanal fishers, an effort of 176 fishing days per artisanal fishers, and the average amount of 1.25 kg retained per fisher per day as reported by Paul (1987). For 1974 and 1982, we used 1,000 and 1,500 professional artisanal fishers, respectively, as reported in Paul (1987), while the number of active professional fishers from 1999 to 2008 were provided by the CSO (Anon., 2009a). For the period 1950- 1958, we rely on the description of the uncontrolled catch by Paul (1987), and we concluded that a total standard of 310 t was yearly retained by the artisanal fishers during this time period for subsistence purposes.  We used linear interpolations between the figures reported or calculated as above to establish the subsistence catches of the artisanal fishers from 1950 to 2008.  Professional part-time fishers  In addition to the professional full-time fishers, numerous part-time fishers exploit the lagoon and non- lagoon environment. The common species in their catch are from the families Siganidae (rabbitfish), Serranidae (groupers), Lethrinidae (emperors) and Carangidae (jacks) (Samboo, 1987). A report by Roullot et al. (1988) mentioned that part-time fishers also visit Fish Attracting Devices (FADs), catching pelagic fish using trolling and hand lining. From 1977 to 2008, the official reported catches for the so- called ‗amateur fishery‘ consisted of the constant amount of 300 t. In a recent survey, the Ministry of Fisheries on Mauritius found that 23,400 persons were involved in ―recreational‖ fishing in the lagoon of Mauritius, from which about 1,000 were owners of a boat. Indications were that their catch could be more than the current estimate of 300 tonnes annually (Jehangeer, 2006). (Samboo and Mauree, 1987) mentioned that ―the quantities caught by the part-time fishers may exceed the catch by the commercial fishers‖.  Paul (1987) and Moal (1975 in Paul, 1987) differentiated between two categories of part-time fishers. The first represents people who directly consume as well as sell a part of their catch; the second consists of local people who fish only for their own consumption. For clarity purposes, we will use the terms part-time professional fishers and part-time subsistence fishers for the first and second categories, respectively. Furthermore, in his estimate of catches for the part-time subsistence fishers in 1982, Paul 44 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   (1987) included catches from the fishing tourist population, but excluding the pelagic sport fishery catches. However, as we aimed at discriminating between the commercial, subsistence and recreational sectors, we rather included the fishing tourist catches in the recreational catches.  Part-time professional fishers fish for their own consumption and sell the surplus of their catch. However, we assumed that their commercial catches are not included in the recorded commercial landings, for the following reasons. First, they are fishers who generally do not operate through a middleman2 and many have developed their own sales outlets to commercialize the surplus of their catch (Paul, 1987). There are some indications that many hotels and restaurants which cater to the tourist trade are provisioned directly by contracted fishers, and that their catch escapes detection by the fisheries authorities and are unaccounted for in landings statistics (Paul, 1987; Sobhee, 2004). Moreover, many of them use illegal fishing gear. To respond to an increasing demand for seafood, reduced catches, and new regulations, many amateur fishers have currently resorted to illegal fishing methods, using fine meshed nets, illegal spearguns and landing of undersized fish (Paul, 1987; Hollup, 2000). Therefore, it was assumed that their commercial catches were not included in the reported artisanal catches.  Table 2. Data sources and method used to estimate the part-time professional fishers population of Mauritius Island from 1950 to 2008. These data were converted to percentage of the total population which we used, together with assumptions and the total human population time series, to derive estimates of the part-time professional fishers population from 1950 to 2008. Year Number of part-time professional fishers Human population Ratio (%) Source and method 1950 1,566  506,663 0.31 25% decrease in ratio from 1950 to 1974 1974 2,000 857,063 0.23 Moal 1975 (in Paul, 1987) 1982 2,000 960,994 0.21 Paul (1987) 2008 2,562  1,230,995 0.21 1982 ratio maintained unaltered  Professional part-time fishers’ population: Paul (1987) and Moal (1975, in Paul, 1987) reported 2,000 part-time professional fishers in 1974 and 1982. We converted these figures to ratios of the total population for the corresponding years. The calculated ratios, together with the total island‘s population time series and assumptions as described below, were used to derive complete time series estimates of part-time professional fisher population from 1950 to 2008 (Table 2). We assumed that the proportion of the Mauritian population involved in such fishing activities decreased by 25% from 1950 to 1974. A greater proportion of the population likely relied on fishing for food or income purposes in the earlier time period. The progressive diversification of the economy away from sugar cane after independence in 1968 (Paul, 1987; Houbert, 2009), together with new and cheap supply of frozen fish from the offshore banks fishery in the late 1960s (Christy and Greboval, 1985; Ardill, 1986) would have contributed to reduce the need to fish for food or income complement. We also assumed that the same proportion of the population was involved in this activity from 1982 to 2008. This is a conservative estimate when we consider the ever-growing population and worsening poverty in the coastal regions (Sobhee, 2004). A high unemployment rate was also of concern in the 1980s as well as in the more recent period (Paul, 1987; 2010).  Professional part-time fishers’ catch rate: According to Moal (1975 in Paul, 1987), 2,000 part-time fishers caught 700 t in 1974. This suggests a catch rate of 0.35 t∙fisher-1∙year-1 in 1974. Paul (1987) presented the change of the average catch rate for an artisanal fisher from 1948 to 1982. From the catches and number of artisanal fishers reported by the CSO (Anon., 2009b), we derived estimates of catch rates for the period 1999-2008. We assumed that the productivity of the part-time fishers would have changed similarly to that of the professional. Indeed, it is very likely that their productivity has declined similarly, as a consequence of increasing numbers of fishers exploiting the same areas. To the reported and calculated catch rates, we applied an exponential model. Using the growth constant of this model and the  2On Mauritius, a large number of professional fishers are dependent on middlemen for equipment. Middlemen usually buy the entire catch, and finance the fishers who are then indebted due to these cash advances. Locally named banyan, middlemen are usually Muslim traders from the urban areas. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  45   catch rate of 0.35 t∙fisher-1∙year-1 for 1974, we derived estimates for part-time professional fisher‘s catch rates for the whole time period (Figure 3), declining from 0.56 tonnes∙fisher-1∙year-1 in 1950 to 0.18 tonnes∙fisher-1∙year-1 in 2008.  Finally, using the fisher population and catch rates as estimated above, we derived the catches of the part- time professional fishers from 1950 to 2008.  Professional part-time fishers’ commercial vs. subsistence catches: For 1974, Moal (1975 in Paul, 1987) divided the part-time professional fishers‘ total catch into subsistence and commercial components, being 29 and 71 percent of the total catch, respectively. We assumed that the subsistence part of the catch was higher in the earlier time period, thus reflecting the increasing development of the tourism industry which likely provided new employment opportunities as well as new sale outlets for the part-time fishers. Also, in the late 1960s, the development of the banks fishery contributed to provide a supply of local frozen fish at relatively low prices (Christy and Greboval, 1985; Ardill, 1986) which would have reduced the need to fish for subsistence by the low income population. Therefore, we inverted those percentages for 1950 so that the subsistence and commercial catches represented 71 and 29 percent of the total part-time professional fishers‘ catch in 1950, respectively. Linear interpolations were used to estimate the percentages for 1950- 1974, while the 1974 percentages was carried unaltered to 2008.  Part-time subsistence (only) fishers  Part-time subsistence fishers fish only for their own consumption. Most of them are coastal residents. This category of fishers likely comprises a substantial proportion of owners of the many houses along the island‘s shores (Paul, 1987). They usually fish close to shore with rod and reel or as a group on a boat.  Part-time subsistence fishers’ population: In 1982, a sample survey of the entire coast indicated that about 10,000 people were fishing only for their own consumption (Paul, 1987). This amount was taken as a percentage of the total human population of Mauritius Island for the same year and came to about 1 percent. Moal (1975 in Paul, 1987) reported a larger estimate of 65,000 people or 7.6 percent of the island‘s population. However, Moal's estimate was considered unreliable. Indeed, to obtain this figure, Moal (1975 in Paul, 1987) simply added the coastal population based on the 1972 census and subtracted the estimated size of the vegetarian and artisanal fishing population. Therefore, we retained the one percent figure as the more conservative estimate. However, while we applied this percentage for 1982- 2008, we assumed that it would have decreased from 1950 to 1982, similarly to that of the part time professional fishers, where the percentage of the number of part-time professional fishers to the total population declined from 0.31 to 0.21 percent from 1950 to 1982. This proportional decline of 67 percent was then applied to the reported 1 percent figure for 1982 back to 1950, so that the percentage of local people fishing only for their subsistence declined from an assumed 1.55 percent in 1950 to a reported 1 percent in 1982. Linear interpolations were used to estimate the percentages in the intervening years. Model 1: y = 1.5e-0.0195x Model 2 : y = 0.6e-0.0195x 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1948 1961 1974 1987 2000 C a tc h  r a te  ( t/ fi sh er /y ea r) Year Professional full-time fishermen's productivity (model 1) Part-time professional fishermen's productivity (model 2) (0.35;1974) Figure 3. Data points and fitted models used to estimate part-time professional fishers‘ productivity for Mauritius, 1950-2008. Data points are represented by open circles (for professional full-time fishers) or dark circles (for part-time professional fishers). Data sources: Paul (1987); Moal (1975 in Paul 1987); Anon. (2009a). 46 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Using the human population data for Mauritius Island, we derived the number of local people fishing only for their own consumption for each year from 1950 to 2008.  Part-time subsistence fishers’ catch rates: For 1982, Paul (1987) estimated the catch rate of a local part time subsistence fisher at 10 kg∙fisher-1∙year-1. We maintained this catch rate unaltered from 1982 to 2008. However, we assumed that it would have been decreasing since 1950. From our precedent calculations, we derived that the part-time professional fishers‘ catch rate declined from 0.56 to 0.30 t per fisher per year from 1950 to 1982. This proportional declined of about 54 percent was then applied to the reported rate of 10 kg per fisher per year for 1982 back to 1950, so that the catch rate of the local people fishing for their subsistence declined from an assumed 18.7 kg∙fisher-1∙year-1 in 1950 to a reported rate in 1982 of 10 kg∙fisher-1∙year-1. Linear interpolations were used to estimate the catch rates in the intervening years. Using the fisher population as estimated above, these derived annual catch rates were expanded to determine the part time subsistence fishers‘ catches from 1950 to 2008.  Other unreported catches  Catches taken by illegal methods are common on Mauritius Island and are largely unreported (Ardill, 1979; Paul, 1987; Hollup, 2000; Sobhee, 2004).  Ardill (1979) mentioned that sampled landings, aside from ignoring part-time fishers‘ catches, also excluded estimates of catch taken in illegal nets and by dynamite fishing, which are landed all along the coast. Other illegal and destructives fishing practices include fine meshed nets, illegal spearguns, night fishing, capture and landing of undersized fish, and pesticides. These practices involve both part- and full-time fishers, and have been rising due to the dwindling resources, reduced catches and new regulations (Hollup, 2000; Sobhee, 2004). In the absence of quantitative information specific to Mauritius Island, we relied on the knowledge acquired for Rodrigues (see below). We assumed that these illegal fishing methods evolved similarly on the two islands, and therefore used the same method as we did for Rodrigues to estimate this component of the unreported catch. Although underwater fishing was forbidden in 1982 in Mauritius, this practice is still used today (90 interventions occurs in 2006). Underwater fishers are also said to be responsible for the use of dynamite and anesthetics, and fishing at night (Paul, 1987). Estimated catches ranged from 150 t in 1974 (1975 in Paul, 1987) to 397 t in 1982 (Paul, 1987), but we assumed that such catches were included in our estimates of unreported catches.  Rodrigues Island small-scale fisheries  Rodrigues Island (19°43‘S-63°25‘E, 110 km², 37,500 inhabitants) is located some 586 km northeast of Mauritius and it is a semi-autonomous island since 2001. Unemployment, poverty and illiteracy are high compared to the main island of Mauritius, and tourism is in its infancy. As a consequence, the lagoon fisheries are very important to Rodrigues, but they are currently severely overexploited (Blais et al., 2007). The island is surrounded by 90 km of fringing coral reef, enclosing a shallow lagoon of 240 km². Small-scale fishing takes place within and outside the lagoon, with the use of seine nets, hand-lines, basket traps and spears or harpoons for octopus and large fish.  Professional fishers  Professional commercial fishers: Total catches reported by the FAO on behalf of the Ministry do not distinguish between Mauritius and its outer dependencies. Nevertheless, comparison of the FAO data with other reports (Fisheries Division annual reports, 2003-2007; Harris, 1988) indicated that from 1977 to 2000, artisanal catches for Rodrigues have been estimated between 1,200 and 1,500 t annually. In Rodrigues, monitoring of artisanal catches started in 1994 (Reshad Jhangeer-Khan, Shoals Rodrigues, pers. comm.). Broad catch and effort data are collected at fish landing stations around the island on a regular basis by the Fisheries Research and Training Unit (FRTU), collated annually and transmitted to the Albion Fisheries Research Centre (Anon., 2007b)(E. Hardman, Shoals Rodrigues, pers. comm.). For the period 1994-2008, we obtained sampled catch data from the CSO annual digests of statistics on Rodrigues and from Dr Emily Hardman (for the octopus and lagoon fish catches reported by the CSO from 1994 to 2006). However, these reported catches refer to the registered fishers catches, and do not include unofficial catches landed by the amateurs and illegal catches (Anon., 2009c) (Dr Emily Hardman, Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  47   Shoals Rodrigues 2005-2008, pers. comm.; Reshad Jhangeer-Khan, Shoals Rodrigues, pers. comm.), and we retained those sampled catches as anchor points for the professional fishers‘ catches. For the earlier period, professional fishers‘ catches were compiled from various sources, using the more reliable estimates as anchor points. For 1955, 1962 and 1968, catches reported by Pearson (1988) were equivalent to the export data. Therefore, those amounts were only used as minimum. We interpolated linearly between our anchors points, in order to estimate the professional fishers‘ catches for the whole time period.  Taxonomic breakdown: In order to determine the taxonomic composition of the artisanal commercial catches taken on Rodrigues Island, we used data from the Fisheries Research and Training Unit (FRTU) (unpublished data) and Central Statistics Office reports. The FRTU data provided the species composition of the seine net fishery for the years 1994 to 1999, and 2004 to 2006. We interpolated linearly between the 1999 and 2004 available figures to estimate the species composition of the seine net fishery for the period 2000-2003, while the 2006 figures were maintained unaltered until 2008. For the period 1994- 2008, CSO data provided the total controlled catch on Rodrigues, discriminated by type of fishery, namely the octopus, lagoon and non-lagoon fisheries. This allowed us to calculate their respective contributions as percentages of the total controlled catch from 1994 to 2008. Similarly, for the period 1994-2006, FRTU data provided the seine net controlled catch, which we converted into percentage of the total controlled catch on Rodrigues. The calculated percentage for 2006 has been carried forward to 2008. Thus, knowing the respective contribution of the seine net, octopus, lagoon (other than seine net) and non-lagoon fisheries, we adjusted the FRTU data on the seine net species composition to estimate a breakdown of the total artisanal commercial catch on Rodrigues for the period 1994-2008 (Table 3). Furthermore, Sauer et al. (2011) indicate that the octopus fishery on Rodrigues is dominated by Octupus cyanea with the remaining catch being mainly O. vulgaris. The catches presented in Sauer et al. (2011) correspond to the CSO data for octopus.  Professional fishers’ subsistence catches: Registered fishers will also keep some of the fish and octopus that they catch for their own consumption (Dr Emily Hardman, Shoals Rodrigues, pers. comm.). Therefore, these retained amounts were estimated and added to our total subsistence catches. We used the estimate of 1.25 kg of fish retained by each fisher per fishing day, which we multiplied with Paul‘s estimate of 176 fishing days for the professional fishers to obtain a catch rate of 0.22 t∙fisher-1∙year-1. This number of fishing days was consistent with these reported by (Pearson, 1988). Indeed, for 1987, he mentioned that the number of fishing days for the trap and handline fisheries rose to about 290 and 238 days, respectively, while that of the seine-net fishery fell from about 200 days in the earlier time period to 163 in 1987, due to regulations being implemented for this fishery. The professional fishers‘ population was reconstructed using various sources (Gonzalez Manero, 1971; Moal, 1971 in Paul 1987; Ardill, 1979; Paul, 1987; Pearson, 1988; Anon., 2010). Finally, the 0.22 t∙fisher-1∙year-1 catch rate estimate as calculated above, together with the artisanal fisher population data, were used to derive the subsistence catches amounts for the whole time period.  Part-time fishers  Fishing by part-time fishers, mainly for subsistence, on Rodrigues Island is unmonitored, although it is likely to be considerable, and it currently includes in and off-lagoon fin fishing using lines and traps, in- lagoon octopus fishing and shell fishing (Reshad Jhangeer-Khan, Shoals Rodrigues, pers. comm.). In addition to the registered fishers, there are many part-time fishers (Gonzalez Manero, 1971; FAO, 1983; Paul, 1987; Anon., 2009c). Some of them are regular fishers who fish to feed their family or to sell, while others have a full-time job and fish for pleasure and their own consumption (Dr Emily Hardman, Shoals Rodrigues, pers. comm.). Therefore, their catches likely comprise commercial as well as subsistence components. However, catches of the part-time fishers were assumed to be mostly of subsistence purpose. In order to estimate these catches, we first assessed the annual number of part-time fishers, then we estimated the catch rates of those fishers, based on Pearson (1988).    Table 3. Estimated taxonomic composition of the total small-scale commercial catch for Rodrigues Island. Sources: Fisheries Research and Training Unit (FRTU) (unpublished data), Anon (2009a), Central Statistics Office of Mauritius.    Portion of total catch (%) Fishery Common name Taxon 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Seine net  Rabbitfish Siganus spp. 1.7 1.1 1.5 1.9 2.6 2.1 2.6 2.8 4.7 3.4 4.8 3.2 5.0 5.0 5.0 Spangled emperor Lethrinus nebulosus 2.1 1.1 1.3 1.1 1.0 0.2 0.6 1.0 2.3 2.1 3.8 4.1 3.3 3.3 3.3 Unicornfish Naso spp. 1.6 0.9 1.6 1.0 1.1 0.4 0.9 1.3 3.0 2.8 4.9 2.4 4.1 4.1 4.1 Mullet  Mugilidae 2.2 3.3 1.6 1.5 1.9 2.7 2.9 2.8 4.1 2.5 2.9 1.8 2.7 2.7 2.7 Trevally/jacks Carangidae 1.8 2.2 1.0 1.6 1.0 1.0 1.1 1.1 1.6 1.0 1.3 1.1 1.3 1.3 1.3 Strongspine silver- biddy Gerres longirostris 1.8 1.1 0.8 0.5 0.3 1.2 1.3 1.3 2.0 1.3 1.7 1.4 2.1 2.1 2.1 Goatfish Mullidae 1.0 1.7 1.0 0.7 0.4 1.3 1.4 1.4 2.2 1.4 1.7 1.7 3.2 3.2 3.2 Parrotfish Scaridae 0.9 0.2 0.5 0.3 0.2 0.1 0.2 0.2 0.4 0.3 0.5 0.6 0.9 0.9 0.9 Blackspot emperor Lethrinus harak 0.8 0.4 0.3 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 Yellowfin bream  Rhabdosargus sarba 0.9 0.2 0.5 0.0 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Rudderfish Kyphosus spp. 0.8 0.5 0.6 0.4 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Milkfish  Chanos chanos 0.8 0.7 0.4 0.3 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 Striped threadfin Polydactylus plebeius 0.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Others Misc. marine fishes 0.6 0.6 0.9 0.4 0.2 0.0 0.1 0.3 0.8 0.8 1.4 1.7 2.2 2.2 2.2 Octopus  Octopus Octupus spp.b 51.1 51.1 42.4 44.7 31.9 23.7 23.6 17.0 27.3 34.7 26.9 27.4 25.0 16.7 16.0 Other  Lagoon fishes Lagoon fishes 31.0 34.0 43.0 45.0 48.0 53.0 50.0 46.0 39.0 41.0 46.0 36.0 35.0 33.0 36.0 Non-lagoon fishes Pelagic and non- pelagic speciesa 1.0 1.0 2.0 - 11.0 . 15.0 15.0 25.0 13.0 9.0 4.0 18.0 15.0 25.0 23.0 aincluding Serranidae, Lethrinidae, Lutjanidae, Acanthuridae, Labridae, sharks and pelagic species such as small tunas or barracudas (from Rodrigues’ Offshore Cooperative Society Landings, 1982-1983, in Paul, 1987); b Octopus spp. caught were O. cyanea (80%) and O. vulgaris (20%) based on Sauer et al.(2011). Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  49   Part-time fishers’ population: 1950-1979: We used the estimate of 7,000 part-time fishers given by (Ardill, 1979). This represented about 26% of the island population in 1979. We assumed that this percentage would also apply to the years before 1979, as most of Rodriguan people likely relied on subsistence fishing and agriculture at this time, and thus we maintained this percentage unaltered from 1950 to 1979.  1980-2008: For 1987, Pearson (1988) estimated that 10 percent of the population practiced occasional line-fishing (or about 4,000 Rodriguans). However, it is very likely that part-time fishers also used other gears such as traps and harpoons to catch fish and octopus and so that their population was underestimated by Pearson (1988). A more recent estimate consisted of 2,000 part-time fishers in 2008 (Blais et al., 2007) or about 5% of the island‘s population. However, some believe that more than twice the number of full-time professional fishers are engaged in part-time fishing activities (Anon., 2009c), which would give an estimate of more than 3,500 people. To remain conservative, we retained the estimate of 2,000 part-time fishers provided by Blais et al. (2007) and we interpolated linearly between the percentages of the total population in 1979 (26%) and 2008 (5%). Using the total island‘s population, we estimated the number of part-time fishers on Rodrigues Island from 1982 to 2008 (Table 4).  Table 4. Data sources and method used to estimate the population of part-time fishers for Rodrigues Island from 1950 to 2008. Year Number of part-time fishers Human population Ratio (%) Sources and method 1950 3,353 12,971 25.8 1979 ratio maintained unaltered 1979 7,000 27,081 25.8 Moal (i  Paul, 1987) 2008 2,000 37,570 5.3 Blais et al. (2007)   Part-time fisher’s catch rates: 1987-2008: We used the catch rate of 0.1 t∙fisher-1∙year-1 estimated by Pearson (1988). We maintained this catch rate unaltered for the period 1987-2008, assuming that the productivity of the part-time fishers would not have changed during the last two decades. This is likely to be a conservative estimate, as Pearson‘s catch rate estimate for 1987 referred to line-fishing, while subsistence fishers also use traps and fish octopus inside the lagoon (Reshad Jhangeer-Khan, Shoals Rodrigues, pers. comm.).  1950-1986: Over-fishing in Rodrigues may date from the 1800s (Bunce et al., 2008). The island has been affected by intensive land erosion and its heavy impact on the lagoonal fauna has been reported as early as 1962 (Baissac, 1962 in Paul, 1987). It has also been illustrated by a sharp downfall in exports to Mauritius since the 1970s (Paul, 1987; Pearson, 1988; Bunce et al., 2008). Therefore, we assumed that the average productivity of the part-time fishers declined from 1950 to 1986 by a half, so that in 1950 it was equal to 0.2 tonnes∙fisher-1∙year-1. Therefore, we applied a catch rate of 0.2 t∙fisher-1∙year-1 for 1950-1987 and 0.1 t∙fisher-1∙year-1 for 1986-2008.  Other unreported catches  Additional catches taken in reserves, during the closed seasons (implemented for the seine net and sea cucumber fisheries) and using illegal gears such as undersized seine net mesh sizes, spearguns to catch octopus were unreported (Anon., 2009c) (Dr Emily Hardman, Shoals Rodrigues, pers. comm.; Reshad Jhangeer Khan, Shoals Rodrigues, pers. comm.). These catches are either consumed by the fishers themselves or sold - on the beach or directly to the markets (Reshad Jhangeer Khan, Shoals Rodrigues, pers. comm.). To estimate these unreported catches we assumed that illegal gears were not used prior to the 1960s (Reshad Jhangeer Khan, Shoals Rodrigues, pers. comm.). For 1987, Pearson (1988) estimated that 150 illegal nets may have been operating within the lagoon. Pearson (1988) conservatively estimated unreported catches taken by illegal methods amounted to 10% of total legal landings in the lagoon. In contrast, these catches may be equal to or greater than legal landings (Reshad Jhangeer Khan, Shoals Rodrigues, pers. comm.). To remain conservative, and to take into account the fact that illegal fishing practices are used by both part-time and full-time fishers, we assumed that most of this catch would be 50 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   included in our total reconstructed catch from the small-scale fishery carried out around the islands. Therefore, for the 1987-2008 period, we estimated additional unreported catches in the coastal waters of Rodrigues island as 10% of our total reconstructed small-scale fisheries catches, including both professional and part-time fishers‘ catches. Finally, we interpolated linearly between zero catch in 1960 and our first calculated value for 1987.  Recreational fisheries  Mauritius is well-known as a tourist destination. While it creates job opportunities for the local population, it also represents a potential threat to marine life (Paul, 1987; Sobhee, 2004). In Mauritius, the recreational fishery can be divided into two components, namely the extraction of marine resources for leisure purposes inside or close to the lagoon environment, and the pelagic sport fishery, which operates with boats in deeper waters. We estimated these catches separately.  Pelagic sport fishery  Mauritius is a popular destination for big game sport fishers. A number of leisure and sports fishers operate successfully around FADs (Roullot et al., 1988; Venkatasami and Sheik Mamode, 1995). For this activity, a number of lines are used with rods and outriggers baited for the large migratory carnivorous species. The catch comprises mainly blue marlin (Makaira mazara), black marlin (Makaira indica), Indo-Pacific sailfish (Istiophorus platypterus), striped marlin (Tetrapturus audax) and yellowfin tuna (Thunnus albacares), albacore (Thunnus alalunga) and bigeye tuna (Thunnus obesus). Other species like wahoo (Acanthocybium solandri), shortbill spearfish (Tetrapturus augustirostris), skipjack tuna (Katsuwonus pelamis) – used as bait for marlins, sharks (Sphyrna zygaena, Isurus oxyrinchus, Carcharinus albimarginatus, Carcharinus melanopterus) and dolphinfishes (Coryphaena hippurus) are also caught (Cayre and Stequert, 1988; Norungee et al., 2004; Jehangeer, 2006). Recently, a system of data collection has been set up at the Albion Fisheries Research Centre (Norungee et al., 2004). However, examination of the Ministry reports showed that the currently reported sport pelagic catches consist of the constant amounts of 400 t from 1977 to 1987 and 650 t from 1988 to 2008. Such amounts do not reflect the increasing trend of the tourist population that reaches the island each year. Thus, catches from the sport fishery were re-estimated. We first extracted catch estimates for the sport fishery from other reliable sources. For each reported value, we calculated a per tourist rate using the time series of number of tourists arrivals. Sport fishery likely already existed in 1950. According to Paul (1987), the sum of the countless amateurs and sport fishermen catches were estimated at 50 t by the officials from 1946 to 1958. However, to remain conservative (and in the absence of more detailed information), we assumed that the pelagic sport fishery catches were null in 1950. For the period 1950-1987, we interpolated linearly between the per tourist rates. Between 1974 and 1988, the calculated per tourist rates declines from 2.9 to 2 kg per tourist arrival. In order to reflect the decreasing catches of pelagic species since the 1990s, we carried this decreasing trend forward from the 1988 catch rate figure to derive the catch rates for the 1988-2008 period. Pelagic sport catches were finally deduced by multiplying the catch rates as estimated above by the number of tourist arrivals.  Recreational catches in the lagoon  In addition to the tourists involved in big game fishing, a substantial part of the tourist population is involved in recreational fishing in the lagoon of Mauritius. For 1982, Paul (1987) assumed that a conservative estimate of the total number of fishing tourists involved in the exploitation of the island‘s waters would be approximately 20,000 people, or about 17% of the tourist arrivals during the year. We assumed that the number of fishing tourists was proportional to the tourist arrivals, and we carried the 17 percent figures for the whole time period. Thus, using the number of tourist‘s arrivals time series, we established the number of fishing tourists from 1950 to 2008. For 1982, Paul (1987) estimated the catch rate of a fishing tourist at 5 kg∙tourist-1∙year-1. We assumed that this catch rate would not have changed from 1950 to 1982, but that the increasing degradation of marine resources of the island, together with an increasing number of people fishing in the lagoon area, would have caused the tourist‘s catch rates to decrease from 1982 to 2008. Therefore, we used the proportional decline of 60% between the part time professional fishers catch rates of 1982 and 2008, so that the tourist catch rates decreased from a reported Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  51   5 kg∙tourist-1∙year-1 in 1982 to an assumed 3 kg∙tourist-1∙year-1 in 2008. We then established the catches using the fishing tourist population time series.  Agalega Islands  The twin islands of Agalega are located some 982 km north of Mauritius between 10°28‘S and 56°40‘E. The islands are of coralline origin and cover a total land area of 70 square kilometers. Colonization took place in the early part of the 18th century and fishing was the mainstay of the settlement‘s diet (Paul, 1987).  Since then, the exploitation of the coconut trees for copra production began, and it is still the main economic activity. Fish production being very limited due to the steep drop-off outside the reef, is only sufficient for consumption of the local workers. The current reported catches for Agalega consist of the constant amount of 30 t from 1977 to 2008. To arrive at these figures, it has been assumed that the per capita consumption of fish was 100 kg∙person-1∙year-1, giving an annual catch of 30 tonnes for a population of 300. The species composition of the catches is believed to be similar to that of Mauritius Island but this remains to be examined (FAO, 1983). As the population did not change much over the 1950-2008 period (Paul, 1987; Anon., 2006, Anon., 2010) and in the absence of contradictory information, we assumed that the current reported amount was accurate and we carried it back unaltered to 1950.  The Mauritian Banks fishery  An important Mauritian fishery occurs on the shallow oceanic banks of the Mascarene Ridge. In addition to the banks located within Mauritius EEZ, the Mauritian fishing fleet fishes on the Saya de Malha bank (much of which is in international waters), and exercises traditional fishing rights within the EEZ of the British Indian Territory (BIOT, Chagos Archipelago). Since the 18th century, fishing on these grounds has been carried out by vessels engaged in inter-island trade, but it was not before the twentieth century that more systematic exploitation began when the Mauritius Fishing Development Company and its sister company the Raphael Fishing Company Limited gained control of the St. Brandon group. In the earlier years, the demersal stocks were exploited mainly for salting purposes, but since the Wheeler and Ommaney (1948-49) pioneer survey, these banks have gradually started to be the main suppliers of frozen fish to Mauritius (FAO, 1983; Ardill, 1986).  Fishing on the banks is practiced using handlines from 7-8 m dories carried by refrigerated mother-vessels, 20-60m in length. The main targeted species is Dame Berri (Lethrinus mahsena) which contributes about 80% of the total catch, while the remainder of the catch is made up of serranids, lutjanids, siganids, and carangids. The catch is mostly gilled, gutted and frozen on board. In St. Brandon, however, temporary settlements of fishers continue to fish in the lagoon using seine nets and basket traps while the mother ships are away. Their catches are salted and dried while awaiting shipping to Mauritius (Moal, 1971 in Paul 1987; Samboo and Mauree, 1987). Also, 17 vessels are allowed to operate a semi-industrial chilled fishery on the Soudan, Albatross, St Brandon, Hawkins, Saya de Malha and Nazareth Banks. Their catch is either frozen or chilled at sea and comprises mainly emperors, snappers, groupers and tunas. Catch data for the banks fishery were obtained from the Ministry on behalf of the Albion Fisheries Research Centre (AFRC). However, such data only covered the Table 5. Estimated taxonomic breakdown of the Mauritian banks handline fishery catches on the Nazareth, Saya de Malha, St Brandon and Chagos banks modified from Samboo 1989 (in Mees, 1996). In the Chagos, although it is avoided due to the potential of ciguatera, the red snapper Lutjanus bohar can represent up to 50 percent of the catch. Taxon Nazareth & Saya de Malha St Brandon Chagos Lethrinus mahsena 88.00 84.0 50 other Lethrinidae 2.00 2.0 2 Serranidae 4.00 1.5 26 Carangidae 2.00 2.0 4 Aprion virescens (Lutjanidae) 1.00 - 6 Siganidae - 2.5 - Scaridae - 2.0 - Mugilidae - 0.5 - Naso spp. (Acanthuridae) - 0.5 - Lutjanus bohar (Lutjanidae) 0.75 - (50) Pristipomoides spp. (Lutjaidae) 0.75 - 10 Tuna 0.75 - 2 Others 0.75 5.0 - 52 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   1977-2008 period. When compared to other sources, it also appeared that they represented landings in frozen weight, and for the greatest part of the time period, catches of salted- dried fish from St. Brandon were not included. Therefore, in order to assemble the most accurate estimates for the total banks fishery catches, we used data of the Ministry converted to wet weight, complemented or adjusted with data extracted from other reliable sources to supply the missing years and areas (Gonzalez Manero, 1971; Moal, 1971 in Paul 1987; Ardill, 1979; Paul, 1987; Samboo and Mauree, 1987; Mees, 1996; Anon., 2009a). Conversion factors of 1.2 and 2 were used to convert the frozen weight and salted-dried weight respectively, to wet weight as reported in Paul (1987), Gonzalez Manero (1971) and in the Conversion Factors FAO software. Although the Chagos Archipelago is part of the British Indian Ocean Territory (BIOT), we included catches from the Chagos‘ EEZ by the Mauritian fleet in our reconstruction of the banks fishery catches.  The estimated species composition of the Mauritian banks fishery catches was based on a report by Samboo (1989 in Mees, 1996) for the Nazareth, Saya de Malha, St Brandon, Albatross and Chagos banks (Table 5), while the taxonomic breakdown of the St Brandon salt and dried fish, octopus and lobsters catches was based on Gonzalez Manero (1971) and Anon. (2009a; Table 6).  Mauritian purse seine fishery  Large-scale commercial purse seining was introduced in the Indian Ocean in 1979 as a Mauritian and Japanese joint venture, using a traditional Japanese technique of fishing schools of tuna associated with logs, developed in the Pacific Ocean. The reported landings from Mauritian purse seiners reached a peak of over 10,000 t in 1991. Since then, there has been a gradual decline of this fishery until 2001 when production dropped to zero, as all the seiners left the fishery (Jehangeer, 2006). Catch data for the Mauritian purse seiners were provided by the Indian Ocean Tuna Commission (IOTC). Amande et al. (2008) provided estimates of tuna discards and bycatch by large groups of species of the European purse seiners operating in the Indian Ocean for the 2003-2007 period. Discards and bycatch were presented as tuna production ratios, for three different fishing methods. As the Mauritian purse seiners operated exclusively on artificial logs (Norungee and Lim Shung, 1995), we used the discards and bycatch ratios corresponding to the FAD-associated fishing mode. To estimate discard rates for each group of species, we multiplied the percentage of bycatch to tuna production by the percentage of discarded bycatch (Table 7). As no indications were provided regarding the amount of discarded bycatch for shark, and to remain conservative, we assumed that sharks species were accounted for in the landings and thus we did not Table 6. Estimated taxonomic breakdown of the salted-dried fish catches taken in the St Brandon lagoon (Anon., 1971; 2009a). Family or group Taxon name Common name Catch (%) Lethrinidae Lethrinus mahsena Dame Berri  75.0 Lethrinidae Lethrinus nebulosus Capitaine Gueule Longue 10.0 Siganidae Siganus spp. Cordonnier 2.5 Acanthuridae Naso spp. Licorne 2.5 Scaridae Callydon spp. Cateau 2.5 Serranidae Epinephelus spp. Vieilles/Babones 5.0 Octopus - - 2.0 Lobster - - 0.5 Table 7. Data and methods used for estimating discards of the Mauritian tuna purse seine fishery from 1979 to 2000. To calculate the discard rate (ratio to tuna production) for the bycatch species, we converted the bycatch rate expressed in tuna production ratio into a percentage, which we multiplied by the reported percentage of discarded bycatch specific to each species group (except sharks). Source: Amande et al. (2008). Species or group Bycatch/tuna production (t/1000 t) Discarded bycatch (%) Discard rate (%) Tuna nei 26.5 100 2.65 MMFa 19.7 80 1.58 Sharks 6.0 - 0.00 Billfishes 0.7 65 0.05 Rays 0.2 100 0.02 a MMF = miscellaneous marine fishes Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  53   include them in our discard estimates. Through this method, our total estimated discards for the Mauritian tuna purse seiner was consistent with the discard rate of 5% reported by Kelleher (2005).  Shark fishery  Shark catches were presented by the FAO for the years 1977-2008. A review of the independent literature found no information regarding a shark fishery in Mauritius. These might have been incidental catches associated with the longline fishery, but which do not appear in the IOTC data. However, we accept the FAO landings data for the ‗sharks, skates and rays‘ category as it is presented in FAO FishStat and add it to the industrial component of our total reconstructed catch.  Sea cucumber fishery  On Mauritius, the commercial exploitation of sea cucumbers started on a trial basis in late 2005 (Conand, 2008) and around 2006 on Rodrigues (Dr Emily Hardman, Shoals Rodrigues, pers. comm.). Species harvested are Actinopyga echinites (brownfish), A. mauritiana (surf redfish), Bohadschia marmorata (brown sandfish), Stichopus chloronatus (green fish), S. variegatus (curry fish), Holothuria scabra (sandfish), H. nobilis (black teatfish) and Holothuria spp. (Anon., 2009d). Although fishers regularly collect sea cucumbers in Mauritius Island mainly for domestic consumption (Laxminarayana, 2005), the harvesting is done by a limited section of the Mauritian population – mainly local Chinese people – and those catches were thus thought to be negligible (Chantal Conand, pers. comm.). Current management measures of the commercial sea cucumber fishery include collection restrictions, both spatial and temporal, and size limits (Anon., 2007a) (Dr Emily Hardman, Shoals Rodrigues, pers. comm.) although illegal fishing continues to occur in the closed season on Rodrigues (Dr Emily Hardman, Shoals Rodrigues, pers. comm.). We accepted the sea cucumber landings data presented by FAO on behalf of Mauritius, although it likely that this is an underestimate of actual catches.  0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Year Professional full-time Professional part-time Subsistence only Illegal gears Figure 4. Reconstructed small-scale fisheries catches for Mauritius Island. Professional full- and part-time catches include both artisanal and subsistence components. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Year Full-time (commercial) Part-time Full-time (subsistence) Illegal gears Figure 6. Reconstructed small-scale fisheries catches for Rodrigues Island, with commercial and subsistence catches taken by full-time fishers, catches taken by part-time fishers whose catch is for subsistence purposes only and unreported catch from illegal gears. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Year Reconstructed Reported Figure 5. Reconstructed small-scale fisheries catches for Mauritius Island compared to reported national landings data. 54 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   RESULTS  Total reported landings by the FAO from 1977 to 2008 followed the same pattern as those of the Ministry reports, implying a good data transfer mechanism between the Mauritius national level and FAO. The same conclusion was drawn from the comparison of the non-pelagic reported catches from both sources. The catch rate data combined with the fisher population data yielded the reconstructed small-scale catches for 1950-2008, which are presented here for Mauritius and Rodrigues Island. Also represented are estimated catches for the recreational sector including both pelagic and lagoon fisheries, the Banks fishery, and tuna catches and associated discards from the purse seine fishery. Finally, we compare our total reconstructed catch estimate for the Republic of Mauritius for 1950-2008 to total marine landings presented by the FAO on behalf of Mauritius.  Mauritius Island  Reconstructed small-scale fisheries catches for Mauritius Island, including catches by professional full- time and part-time fishers,  part-time subsistence (only) fishers and unreported catches taken by illegal gears and methods was estimated to be 170,825 tonnes over the 1950-2008 time period (Figure 4). Unreported catch taken by illegal gears and methods represented approximately 8,500 t over the period 1960-2008 (Figure 4). For the 1977-2008 period, our total reconstructed catches for the small-scale sector are about 1.6 times greater than nationally reported landings (Figure 5). Catches by part-time fishers‘ (professional and subsistence only) made up a substantial part of the reconstructed catch representing, for the whole time period, 27% of our total reconstructed catches for the small-scale sector on Mauritius Island.  The time series of catches taken by professional full-time fishers, the majority of which are artisanal, has two main periods. The first (1950-1977) represents a period where data were scarce, the second (1977-2008) corresponds to a period of improved sampling and collection of fisheries landings. Thus, the catch variability that appears in the latter time period may be more representative of a trend in catch levels. Catches taken by professional full-time fishers decreased substantially in 1979 from approximately 2,200 t to 1,500 t the following year (1980; Figure 4). The cyclone of 1980 may have played a role in this decrease e.g., through its impact on fishing effort. However, given 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Year Reconstructed Reported Figure 7. Total reconstructed catch for the small-scale fisheries of Rodrigues Island, 1950-2008 compared to the nationally reported landings, which start in 1977. 0 200 400 600 800 1000 1200 1400 1950 1960 1970 1980 1990 2000 C a tc h es  ( t) Reported Total reconstructed 0 200 400 600 800 1000 1200 1400 1950 1960 1970 1980 1990 2000 C a tc h es  ( t) Tourist lagoon Pelagic sports fishery Figure 8. a) Reported and total reconstructed recreational catches for Mauritius state 1950-2008; b) Components of total recreational catch. a) b) Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  55   that the sampling methods for monitoring artisanal catches were implemented in 1977, these dropping catches could also be the result of adjustments to the method by the officials between 1977 and 1980.  Rodrigues Island  Catches by small-scale fishers on Rodrigues Island were estimated over the 1950-2008 time period to be approximately 159,000 t (Figure 6). This estimate included commercial catches by professional full-time fishers which amounted to about 96,000 t, subsistence catches taken by full-time and part-time fishers amounting to almost 54,000 t and unreported catch from illegal gears estimated to be 8,600 t. Subsistence catches taken by part-time fishers (i.e., people who fish only for subsistence purposes) were over 2.5 times larger than subsistence catches taken by full-time fishers as the ‗take-home‘ portion (i.e., non-commercial) of their catch (Figure 6). However, this is mainly an artifact of the much larger number of part-time fishers than professional full-time fishers. The number of part-time fishers was up to 7 fold greater than the number of professional full-time fishers.  Summed for 1977-2008, reconstructed small-scale fisheries data suggested a 1.9 fold difference between reconstructed estimates and the statistics reported by the Ministry (Figure 7). As for Mauritius Island, this discrepancy between the reported and our reconstructed catches is mainly due to the inclusion of our estimated catches by part-time fishers, which, for the whole time period, represented 25% of our total reconstructed catches for the small-scale sector.  Recreational fisheries  Our total reconstructed recreational catch for Mauritius from 1950 to 2008 was estimated to be over 30,000 t, which is 1.7 times larger than the reported recreational catch (Figure 8a). Pelagic sports fishery catches accounted for approximately 21,800 t and tourist catches from the lagoon fishery representing the remaining 8,700 t (Figure 8b). It is worth noting the difference in the overall trend of the reported compared to the reconstructed recreational catch, especially since 1990 (Figure 8a). Due to the method employed, estimated catches of the fishing tourist population reflected the growing number of tourists visiting the island each year, thus showing an increasing trend for the whole time period (Figure 8b). Estimated pelagic sports fishery catches increased constantly from 1950 to a peak of 825 t in 2000, after which catches have been decreasing.  Banks fishery  Total estimated catches from the banks fisheries were estimated to be approximately 167,000 t over the study period. The St Brandon banks fishery operated from 1950-2008, the Nazareth bank fishery started in 1969 and the Albatross, Chagos and Saya de Malha banks fisheries started in the late 1970s. The Saya de Malha and Nazareth banks fisheries were the largest with catches of over 70, 000 t and 54,000 t, respectively over the study period (Figure 9).  Mauritian purse seine fishery  Total catches of tuna taken by the Mauritian purse seine fishery were estimated for the period 1977-2008 to be approximately 127,000 t (Figure 10). The discarded bycatch associated with the tuna purse seine fishery were estimated to be 4,472 t (Figure 9). 0 1 2 3 4 5 6 7 8 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Year Nazareth Saya de Malha Albatross St Brandon Chagos Figure 9. Total catches estimated for the Banks fisheries of Mauritius, 1950-2008. These include fisheries on St. Brandon, Chagos, Albatross, Saya de Malha and Nazareth banks. 56 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller    Shark and sea cucumber fisheries  Total catches of shark and sea cucumber, taken directly from the FAO data were estimated to be 2,043 t and 1,055 t, respectively, over the 1950-2008 time period. Shark catches were considered part of the industrial sector while the sea cucumbers originate from the artisanal sector.  Total reconstructed catches  Total reconstructed marine fisheries catches for Mauritius and its outer dependencies was estimated for the 1950-2008 time period to be 682,392 tonnes. This suggests our reconstructed estimate was 42% higher than the landing statistics reported by FAO on behalf of Mauritius. The discrepancy between our total estimated catches and those currently reported is most substantial for 1950-1970 (Figure 11). Summed for this period, our total reconstructed estimates represent about 2.4 times the reported catches. This implies that under-reporting of catches was particularly substantial for the earlier time period.  DISCUSSION  For a better understanding of the fisheries impact on marine ecosystems, there is a great need for improved reporting and verification of landings and catches. The present study represents an alternative approach to estimate a more comprehensive total catch for the island state of Mauritius, including estimates of unreported landings of the small-scale fisheries sector, recreational catches, and discards. Summed for 1950-2008, marine fisheries catches for Mauritius and its dependencies as estimated in our reconstruction was 682,392 t, which is 42% larger than currently reported total catches of 478,305 t presented by FAO on behalf of Mauritius. This was largely due to the under-reporting of small-scale catches for Mauritius and Rodrigues islands. Estimated small-scale catches for Mauritius and Rodrigues islands represented respectively 25% and 23% of the total reconstructed catch, thus implying an important contribution of those fisheries to total national catches. Thus, our results confirmed that currently reported catches for the Republic of Mauritius are incomplete, and especially underrepresented catches from the small-scale fishing sectors. This situation of underreporting of catches for the small-scale sector is not specific to Mauritius and has been demonstrated for other countries (Zeller et al., 2006; Jacquet and Zeller, 2007). Indeed, although they are highly important in terms of income and food security, small-scale tropical fisheries are often marginalized world-wide (Pauly, 1997). Consequently, their contribution in terms of catches is also often substantially underreported.  0 5 10 15 20 25 30 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Total reconstructed catch as supplied to FAO Figure 11. Total reconstructed catch for Mauritius Island and its dependencies and catches presented by the FAO on behalf of the Republic of Mauritius, 1950-2008.  0 2 4 6 8 10 12 14 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Year Tuna Discards Figure 10. Total catches of tuna and associated discards from the Mauritian purse seine fishery, 1950-2008. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  57   Nevertheless, it should be mentioned that the Mauritius government made efforts to decrease the pressure on lagoon fish stocks and to improve the fishers‘ livelihoods (such as the bad weather or closed net season allowances). Mauritius is also one of the few countries which accounts for recreational and ‗amateurs‘ fisheries catches in their total reported catches, even though they are represented by constant estimates over the years. However, as demonstrated in the present study, more efforts should be dedicated to monitoring recreational and small-scale fisheries.  High levels of uncertainty are associated with reconstructions such as ours. However, as long as estimates for unaccounted catches are not substantially overestimated, the catch reconstruction will present a more accurate picture of total extractions in the marine environment compared to current practices of essentially allocating ‗zero catch‘ to IUU components (Illegal, Unregulated and Unreported catches) for which no hard time series data are available.  The time series data presented do not include catches by foreign vessels which fish heavily – legally or illegally - the waters off Mauritius, which would likely add great amounts to the total extractions of marine resources in the Mauritius‘ EEZ. The Banks of the Mascarenes Ridge, for instance, have been, over the years, exploited by a number of foreign fishing units from diverse countries such as Korea, Japan, France (Reunion), Seychelles, Spain, Russia, Panama and Malaysia (Paul, 1987). Réunion based boats were fishing in the Soudan, Nazareth and Saya de Malha zone as early as 1962. Not all foreign vessels fishing on the banks use handlines but also longlines and trawls. The total catch of these foreign fishing vessels on the banks is not reported but estimated to be well in excess of 10,000 tonnes per year (Ardill, 1986; Paul, 1987).  Discards of the deep-sea demersal trawlers are also considered to be high in many deep-sea fisheries (Kelleher, 2005) and therefore they would likely add important amounts to our estimated total catches. Kelleher (2005) reported a discarding rate for deep water trawls of 39.6%. However, this discarding rate was from fisheries operating in the Northeast Atlantic and Chile, and no similar estimate has been made for the Western Indian Ocean. Therefore, attempts have not been made to quantify trawl discards.  In Mauritius as in Rodrigues Island, depletion of the marine lagoon resources is of concern and indications of growth overfishing were drawn early. One of the characteristics of growth overfishing is a decline in the mean size of a fish population, and this problem has been noted for Mauritius as early as the 1920s (Paul, 1987). Pearson (1988) also defined the lagoon of Rodrigues as clearly overfished, while (Bunce et al., 2008) mentioned that overfishing in Rodrigues could have occurred since the 1800s. In Mauritius, the reduced catches have to some extent been compensated for by rising prices and a high demand for fresh fish, and as a result there has been no substantial reduction of the fishing effort (Hollup, 2000). An ever growing population and increasing tourist arrivals is leading to a greater demand for seafood, and subsequent increased pressure on fish stocks. Fishers have also resorted to illegal fishing practices that are highly destructive for the marine environment (i.e., dynamite and fine mesh nets). Combined, population pressure and destructive fishing practices suggest that ‗Malthusian overfishing‘ (Pauly, 1997) is occurring on Mauritius Island. Although regulations have existed in the legislation since the colonial days, they only concern limitations on the use of specific gears, fish reserves and closed seasons for nets (Hollup, 2000). Pearson (1988) and Hollup (2000) both mentioned that regulations should include the limitation of access to the fish resources of the lagoon area. However, alternatives are needed for the numerous fishers who depend on these resources for their livelihoods.  This study illustrates the need for better reporting of catches for all the different fisheries sectors of Mauritius state. Accounting for all fisheries components is fundamental to effectively managing the fisheries, which provide food and income for a large coastal population. We hope that the present study will encourage management agencies and policy makers to work in this direction.   ACKNOWLEDGMENTS  We thank Dr Emily Hardman, Reshad Jhangeer-Khan from the Shoals Rodrigues, the Fisheries Research and Training Unit of Rodrigues and the Fisheries Division of Mauritius, who shared information and 58 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   knowledge for this study. We acknowledge Pierre Chavance, Miguel Herrera, and Vincent Defaux, for their help and contribution, as well as the support of the Sea Around Us Project, a scientific collaboration between the University of British Columbia and the Pew Environment Group.  REFERENCES  Amande, J.M., Ariz J., Chassot, E., Chayance, P., Delago de Monlina, A., D., G., Murua, H., Pianet, R. and Ruiz, J. (2008) By-Catch and discards of the European purse seine tuna fishery in the Indian Ocean. Estimitation and Characteristics for the 2003-2007 period. Indian Ocean Tuna Commission, Bangkon, Thailand. Anon. (1987) Reports of the ad-hoc technical Commitee on fisheries research. Food and Agricultural Research Council (FARC), Maurice. Anon. (2006) Population and Vital Statistics Republic of Mauritius, Year 2005. Central Statistics Office, Republic of Maurititus, Port Louis. Anon. (2007a) Annual report 2007 Fisheries Division, Ministry of Agroculture Industry, and  Food Security, 106 p. Anon. (2008) Digest of international travel 2007. Central Statistics office, Republic of Maritius, Port Louis. Anon. (2009a) Digest of Environmental statistics 2008. Central Statistics Office, Republic of Maritius, Port Louis. Anon. (2009b) Digest of International Travel. Central Statistics Office, Republic of Maritius, Port Louis. Anon. (2009c) Digest of statistics on Rodrigues. Central Statistics Office, Repbulic of Maritius, Port Louis. Anon. (2009d) Transboundary Diagnostic Analysis of Land-based Sources and Activities Affecting the western Indian Ocean Coastal and Marine Environment.  378. United Nations Environmental Programme, Nairobi Convention Secretariat, Nairobi. Anon. (2010) Population and Vital Statistics, Republic of Mauritius, Year 2009. Central Statistics Office, Republic of Maritius. Ardill, J.D. (1979) Country Statement on the marine fisheries in Mauritius. pp. 54-62. In: Organization, U.N.D.P.- F.a.A. (ed.) Report of the FAO/ IOP workshop on the fishery resources of the western Indian Ocean south of the Equator. United Nations Development Programme, Mahe, Seychelles. Ardill, J.D. (1986) Current status of fisheries in Mauritius. Food and Agriculture Organization, United Nations, 31 p. Bunce, M., Rodwell, L.D., Gibb, R. and Mee, L. (2008) Shifting Baselines in Fishers' perceptions of island reef fishery degredation. Ocean and Coastal Management 51(4): 285-302. Christy, L. and Greboval, D. (1985) Fisheries Institutions in Mauritius. SWIOP Documents. Regional Office for Africa, Food and Agriculture Organization of the United Nations, Mahe. Conand, C. (2008) Population status, fisheries and trade of sea cucumbers in Africa and the Indian Ocean. pp. 143- 193. In: Toral-Granda, V., Lovatelli, A. and Vasconcellos, M. (eds.) Sea cucumbers: A global review of Fisheries and trade. FAO Fisheries and Aquaculture Technical papers. Food and Agrigulture Organization of the United Nations, Rome. Gabbay, R. (1988) Chapter 9: Tourism. p. 236. In: Appleyard, R.T. and Ghosh, R.N. (eds.) Indian Ocean Islands Development.  1. National Center for Development Studies, Australian National University, Canberra. Gonzalez Manero, C. (1971) Marketing of fishery products. Fishery Developmenment Project. United Nations Development Programme, Rome, 37 p. Hardman, E.R., Biais, F.E.I., Desire, M.S., Raffin, J.S.J., Perrine, S. and Gell, F.R. (2007) Marine Reserves for Sustainable Fisheries Managment in Rodrigues. Alternative Livelihood Options 1. Shoals Rodrigues & Wildlife and Conservation Division, Department of Agriculture, Fisheries and Forestry, Peel. Harris, A. (1988) Indian Ocean Islands Development. National Center for Development Studies, Australia National University, Canberra, 204 p. Hollup, O. (2000) Structrual and Sociocultural constraints for user-group participation in fisheries management in Mauritius. Marine Policy 24(5): 407-421. Houbert, J. (2009) Mauritius: a sea food hub?  In: Rumley, D., Chaturveydi, S. and Vijay, S. (eds.) Fisheries Exploitation in the Indian Ocean. Threats and opportunities. ISEAS Institute of Southeast Asian Studies. Jacquet, J.L. and Zeller, D. (2007) National conflict and fisheries: reconstructing marine fisheries catches for Mozambique. Fisheries Center Research Reports. Fisheries Centre, University of British Columbia, Vancouver. Jehangeer, I. (2006) Review of the state of world marine capture fisheries management: Indian Ocean. Country Review: Mauritius. FAO Technical Paper. Food and Agricultural Organization United Nations, Rome, 393- 413 p. Kelleher, K. (2005) Discards in the World's marine fisheries. An update. Food and Agricultural Organization United Nations, Rome, 131 p. Laxminarayana, A. (2005) Induced spawning and larval rearing of the sea cucumbers, Bohadschia marmorata and Holothuria atra in Mauritius. Information Bulletin 22. Secretariat of the Pacific Community. Mees, C. (1996) The Mauritian Bank Fishery. A review and spatial analysis, Technical Report. Available at: www.dfid.gov.uk/r4d/pdf/outputs/R5484c.pdf [accessed August 2011], 62 p. Moal, R.A. (1971) Storage, processing, and distribution of fish. Indian Ocean Programme. Indian Ocean Fishery Commission, Rome. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  59   Norungee, D. and Lim Shung, C. (1995) Analysis of the purse seine fishery of Mauritius, 1990-1994, and comparison of catch rate and species composition of catches of Mauritian purse seiners to those of the French fleet (English). Expert Consultation on Indian Ocean Tunas, Colombo (Sri Lanka). Paul, E.C. (1987) Fisheries development and the food needs of Mauritius. A. A. Balkema, Rotterdam. 216 p. Pauly, D., Christensen, V., Dalsgaard, J., Froese, R. and Torres, F. (1998) Fishing down marine foods webs. Science 279(5352): 860-863. Pearson, M.P. (1988) Rodrigues. Rapid Survey of the status of exploitation and environmental damage of the lagoon and coral reefs off Rodrigues.  Report prepared for the project assistance to artisanal fisherment and development of outer-reer fishery. . Food and Agriculture Organization, United Nations, Rome. Roullot, J., Venkatasami, A. and Soondron, S. (1988) The first three years experience in the use of Fish Aggregating Devices in Mauritius. Food and Agriculture Organization, United Nations, 45 p. Sauer, W.H.H., Potts, W., Raberinary, D., Anderson, J. and Sylvio Perrine, M.J. (2011) Assesssment of the current data for the octopus resource in Rodrigues, Western Indian Ocean. African Journal of Marine Sciences 33(1): 181-187. Sobhee, K.S. (2004) Economic Development, Income Inequality and Environmental Degradation of Fisheries Resources in Mauritius. Environmental Managment 34: 105-157. Sobhee, K.S. (2006) Fisheries biodiversity conservation and sustainable tourism in Mauritius. Ocean & Coastal Management 34: 150-157. Turner, J. and Klaus, R. (2005) Coral reefs of the Mascarenes, Western Indian Ocean. Royal Society 363: 229-250. Van der Elst, R. (2005) Fish, fishers and fisheries of the Western Indian Ocean: their diversity and status. A preliminary assessment. Royal Society 363: 263-284. Venkatasami, A. and Sheik Mamode, A. (1995) Fish- aggregating devices (FADs) as a tool to enhance production of artisanal fishermen: problems and perspectives. 5 p. Zeller, D., Craig, P. and Pauly, D. (2006) Reconstruction of coral reef fisheries catches in America Samoa, 1950- 2002. Coral Reefs 25: 144-152. Zeller, D., Booth, S., Davis, G. and Pauly, D. (2007) Re-estimation of small-scale fishery catches for U.S. flag- associated island areas in the western Pacific: the last 50 years. Fisheries Bulletin, 105: 266-277.     60 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller    Appendix Table A1: FAO landings vs. total reconstructed catch (in tonnes) for Mauritius, 1950-2008. Years FAO landings Total reconstructed catch 1950 2,000 5,786 1951 2,000 5,982 1952 2,000 5,913 1953 2,000 6,057 1954 2,000 5,958 1955 2,000 5,995 1956 2,200 6,013 1957 2,200 5,920 1958 2,200 6,121 1959 2,500 6,300 1960 2,501 6,486 1961 2,501 6,757 1962 2,801 6,263 1963 2,801 6,783 1964 2,801 6,238 1965 3,000 6,274 1966 3,001 6,343 1967 3,000 6,258 1968 3,300 6,421 1969 4,001 7,328 1970 5,400 7,586 1971 5,200 8,337 1972 6,600 9,347 1973 6,400 8,663 1974 7,679 9,536 1975 7,038 9,248 1976 6,660 8,649 1977 7,667 10,932 1978 7,108 10,697 1979 6,525 9,362 1980 6,348 9,109 1981 7,132 10,032 1982 9,780 12,638 1983 9,434 12,222 1984 10,346 13,613 1985 12,175 15,248 1986 12,848 15,817 1987 17,279 20,656 1988 17,116 20,604 1989 16,896 20,530 1990 14,098 17,603 1991 18,576 22,298 1992 18,861 22,687 1993 20,576 24,595 1994 18,145 21,911 1995 16,395 19,620 1996 11,870 15,104 1997 14,025 17,707 1998 12,093 15,068 1999 12,205 16,264 2000 9,615 13,684 2001 10,986 14,748 2002 10,706 14,414 2003 10,968 14,836 2004 9,971 13,126 2005 9,855 12,995 2006 8,681 13,653 2007 8,087 12,627 2008 6,152 11,430  Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  61     Appendix Table A2: Total reconstructed catch (in tonnes) by major taxa for Mauritius, 1950-2008. Others grouping includes 38 taxa. Year Lethrinus mahsena Octopodidae Katsuwonus pelamis Thunnus albacares Siganidae Thunnus alalunga Serranidae Acanthuridae Mugilidae Others 1950 332 1,544 -  534 - 152 261 187 2,465 1951 460 1,563 - 0 539 0 155 262 188 2,503 1952 367 1,582 - 0 538 1 153 263 189 2,511 1953 449 1,602 - 0 542 2 155 264 191 2,543 1954 329 1,621 - 0 539 3 153 264 191 2,547 1955 322 1,640 - 1 540 5 154 265 192 2,566 1956 318 1,649 - 1 541 9 154 265 192 2,575 1957 220 1,656 - 2 538 12 152 265 192 2,571 1958 339 1,762 - 3 531 17 170 266 157 2,566 1959 257 1,813 - 4 574 22 183 289 165 2,682 1960 180 1,864 - 5 618 28 196 312 174 2,799 1961 226 1,907 - 6 653 35 208 329 181 2,903 1962 158 1,840 - 7 572 42 182 289 166 2,698 1963 218 1,925 - 8 647 50 206 327 181 2,912 1964 181 1,844 - 10 551 58 175 278 163 2,668 1965 215 1,883 - 10 607 63 167 305 149 2,565 1966 218 1,896 - 14 609 87 168 306 150 2,584 1967 128 1,909 - 18 608 110 167 307 150 2,590 1968 177 1,922 - 22 612 136 168 308 151 2,614 1969 895 1,937 - 27 613 163 202 309 152 2,699 1970 1,027 1,969 - 31 614 189 209 310 154 2,749 1971 1,568 1,986 - 34 617 208 233 311 155 2,826 1972 1,884 2,229 - 37 362 224 243 507 208 3,268 1973 1,520 2,079 - 39 362 236 229 510 202 3,118 1974 2,344 2,090 - 28 363 172 266 511 203 3,205 1975 2,083 2,088 - 30 365 183 257 514 203 3,187 1976 1,506 2,077 - 39 366 235 233 518 203 3,149 1977 3,460 2,072 0 46 368 270 333 523 204 3,346 1978 3,333 2,043 15 71 360 296 304 500 199 3,283 1979 2,055 2,042 41 59 350 362 253 498 197 3,183 1980 1,553 1,926 1,004 67 277 336 192 389 169 2,851 1981 1,589 1,923 1,746 76 274 367 217 387 168 2,907 1982 3,505 1,806 2,430 78 299 368 351 163 249 3,079 1983 2,684 1,842 1,421 1,191 342 361 304 185 275 3,344 1984 2,680 1,823 2,537 1,430 349 373 332 186 276 3,377 1985 4,212 1,798 2,080 1,081 348 360 402 184 273 4,081 1986 5,063 1,773 1,899 1,039 352 359 426 184 273 4,003 1987 5,917 1,785 4,397 1,845 383 443 519 203 293 4,439 1988 5,550 1,747 5,049 1,492 377 440 505 194 284 4,551 1989 4,085 1,749 5,614 1,955 379 476 397 199 288 5,007 1990 3,360 1,758 4,195 1,624 390 480 408 206 295 4,522 1991 3,716 1,744 6,735 3,043 382 472 429 200 289 4,940 1992 5,152 1,766 6,126 2,419 415 484 509 216 308 4,969 1993 6,047 1,736 7,074 2,768 394 522 500 201 292 4,765 1994 5,961 1,742 5,209 1,987 394 536 543 208 297 4,755 1995 5,720 1,614 3,936 1,891 372 544 483 191 303 4,313 1996 4,787 1,504 1,589 727 390 601 491 204 277 4,297 1997 4,737 1,503 3,150 1,281 338 640 459 175 242 4,966 1998 4,091 1,209 1,692 1,576 333 648 368 174 253 4,550 1999 4,328 1,037 2,481 1,333 336 685 391 174 283 5,060 2000 4,208 1,042 425 835 349 725 455 184 302 4,960 2001 3,627 791 140 769 319 716 376 167 275 7,393 2002 3,936 964 140 752 348 693 420 184 309 6,529 2003 3,816 1,221 140 784 332 779 409 175 267 6,808 2004 3,215 853 140 786 313 876 344 168 252 6,099 2005 2,240 792 133 787 299 853 258 158 216 7,184 2006 3,088 752 133 838 300 838 342 161 235 6,862 2007 2,491 622 133 823 246 866 293 132 215 6,758 2008 2,449 643 133 841 260 788 252 134 225 5,705   62 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  63   RECONSTRUCTION OF NAURU‘S FISHERIES CATCHES: 1950-20081  Pablo Trujillo, Sarah Harper and Dirk Zeller Sea Around Us Project, Fisheries Centre, University of British Columbia, 2202 Main Mall, Vancouver, BC, V6T 1Z4, Canada p.trujillo@fisheries.ubc.ca; s.harper@fisheries.ubc.ca; d.zeller@fisheries.ubc.ca   ABSTRACT  Nauru is a small, single island country located in the western Pacific Ocean with a relatively large population. Historically, phosphate mining has destroyed much of the island‘s land surface. We reconstructed total fisheries catches for Nauru (1950-2008). The reconstructed catch consists of small- scale fisheries, including both commercial and subsistence components, as well as the offshore domestic catch. For 1950-2008, total reconstructed catches were 23,150 t, being 3.4 times higher than data supplied to FAO on behalf of Nauru. Of these catches, 9,000 t, 12,300 t and 1,850 t were small-scale commercial, small-scale subsistence and offshore catches, respectively.   INTRODUCTION  The Republic of Nauru is a small, single raised limestone island located at 0˚ 32' S latitude and 166˚ 55' E longitude, with a land area of approximately 21 km2 (Figure 1) and an Exclusive Economic Zone (EEZ) of 308,000 km2 (www.seaaroundus.org).  The island has an average height of 50 m above sea level, with an interior plateau that once held extensive deposits of phosphate bearing rock, resulting from the accumulation of seabird droppings over millenia. Phosphate mining was the island‘s largest source of revenue, but due to heavy mining the resource is now virtually depleted. This has left an estimated 80% of the land area uninhabitable, along with substantial environmental degradation from silt and phosphate runoff, which is believed to have impacted large parts of the island‘s marine life (Jacob, 1998). With an already degraded terrestrial environment and an eroding coastline, the impact of global sea level rise for a small island such as Nauru may be devastating (Stephen, 2011)  The original inhabitants of the island relied heavily on marine resources a source of animal protein. Nauru is surrounded by a coral belt that becomes exposed at low tide, ranging from 150 to 300 m in width, and the waters surrounding Nauru hold an abundance of both reef and pelagic fish species (Dalzell and Debao, 1994; Jacob, 1998).  1  Cite as: Trujillo, P., Harper, S. and Zeller, D. (2011) Reconstruction of Naurru‘s fisheries catches: 1950-2008. pp. 63-71. In: Harper, S. and Zeller, D. (eds.) Fisheries catch reconstructions: Islands, Part II. Fisheries Centre Research Reports 19(4). Fisheries Centre, University of British Columbia [ISSN 1198-6727]. Figure 1. Map of Nauru and its Exclusive Economic Zone (solid line). 64 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller    The colonial period  In 1888, Germany annexed the island as part of the Marshall Islands Protectorate. At the turn of the century, a British company discovered phosphate, and mining for the deposits began in 1906. Immigration to the island, starting with the European influx during the colonial period (between 1889 and 1913), followed by migrant workers to labour in the phosphate mining operation, changed the island‘s traditional make-up. These immigration patterns were linked to the mining developments of the Nauru Phosphate Corporation (NPC), yet were also relevant to the island‘s fishing capacity and history.  Initially, workers were brought to the island from other German-administered Micronesian islands, as well as New Guinea and China. Later, when Australia took control of the island in 1914, an increasing number of Chinese labourers were hired, and by 1939 their numbers (1,512) nearly equalled those of the native Nauruan population (1,733). This demographic make-up did not change until, during World War II, the Japanese occupiers of the island forcibly exiled Nauruans to Truk, where nearly 500 of them died (Underwood, 1989).  At the start of the 1950s, virtually every able-bodied Nauruan adult male was gainfully employed both and once again began to receive their respective phosphate mining royalty payments. This enabled the inhabitants to purchase imported foods as well as improving their living conditions (Viviani, 1970). This purchasing power brought on numerous changes in the structure of the island population, differentiating it from other Pacific islands.  Independence  At the time of its independence in 1966, Nauru had attained an economic status similar to that of rich oil- producing countries of the Middle East (Underwood, 1989; Vunisea et al., 2008). This in turn increased the immigration of labourers to the island to compensate for the growing number of retiring Nauruan nationals (Figure 2). The population is largely concentrated along a narrow coastal strip. Most of the non- native Pacific islanders (largely from Kiribati and Tuvalu) that worked for the Nauru Phosphate Corporation (NPC) were also at least part-time fishers (Underwood, 1989; Dalzell and Debao, 1994).  Nauru‘s economic prosperity translated into profound cultural changes. This was also reflected in the diet, not only in terms of dietary preferences, but also in the means by which dietary goods were acquired. Until the mid 1980s, the majority of fishing was done by the non-Nauruan Pacific islanders. Nonetheless, the economic decline that occurred in the early 1990s, following the downscaling of phosphate production, reduced the population‘s ability to purchase high-valued fish, and contributed to the subsequent emigration of non native fishers from the island.  The development and management of the marine resources within the Republic of Nauru falls under the jurisdiction of the Nauru Fisheries and Marine Resources Authority (NFMRA). The NFMRA does not enforce the reporting of catches or issue any fishing quotas. Marine resources are open access, and records of catches are sparse (Dalzell and Debao, 1994). The NFMRA attempted to pursue industrial-scale fisheries by purchasing two purse-seine vessels. However, one ship was lost at sea, while the other was used sparsely due to inadequate fishing gear, and was subsequently sold.  The aim of the present study was to gather available information on fisheries catches and fishing practices to reconstruct Nauru‘s total fisheries catches for the period 1950-2008. The catch reconstruction approach used here is based on the approach developed by Zeller et al. (2006; 2007).  METHODS  At present, small scale marine fisheries in Nauru can be separated into two categories:  1) Subsistence fisheries, dominated by coastal reef fisheries, beach seining and reef gleaning (mainly by women [Chapman, 1987]), are known to be traditional fishing methods practiced throughout the Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  65   Pacific islands (Gillett, 2003). I-Kiribati, Tuvaluan and Nauruan fishers commonly operate outboard powered boats ranging from 3–7 m in size, using trolling and shallow-water bottom handlining as their main fishing methods. Some Nauruans carry out spear fishing equipped with scuba gear. This type of fishing may also be done at night, using battery powered flashlights (Chapman, 1998). These coastal fishing activities do overlap with commercial fisheries, if part of the catch is sold; and  2) Commercial fisheries, using fishing methods such as mid-water handlining and drop stone fishing, targeting offshore tunas and other pelagics (Gillett, 2003). The fishers involved in this type of fishery are generally Nauru Phosphate Corporation (NPC) workers from Tuvalu and Kiribati who fish whenoff from work. This fishing is often concentrated around the NPC‘s mooring buoys used for phosphate vessels that act as Fish Aggregation Devices (FADs) (Chapman, 1998).  Data sources  Human population  Human population numbers (1950-2008) were derived with assistance from G. Beccalossi; Programme Assistant at the Demography and Statistics division of the Secretariat of the Pacific Community (SPC), complemented with data from UN databases2 and work by Underwood (1989). Every decade had at least one demographic composition anchor point, with linear interpolations performed between anchor points (Figure 2).  Catch data  Data estimates of fisheries catches, and used here as anchor points, were obtained from Dalzell and Debao (1994), Gillett and Lightfoot (2002), Vunisea et al. (2008), as well as Bell et al. (2009), for the period between 1991 and 2008. Linear interpolations were used for time periods between anchor points.  To reconstruct the 1950-1990 times series, we assumed that the per capita consumption rate for 1950 was twice the 1991 per capita catch rate of 45 kg∙person-1∙year-1  for Nauru, taken from Dalzell and Debao (1994). Hence, we assumed a consumption rate of 90 kg∙person-1∙year-1 for 1950. We also assumed negligible seafood imports for 1950. Nauruans, like most Pacific islanders, have traditionally relied on fish as a main protein source (Petit-Skinner, 1981; Sokimi and Chapman, 2001) and the assumed 1950 rate is in line with other data for Micronesia (Bell et al., 2009). We linearly interpolated rates between 1950 and 1991 and derive total catch estimates in conjunction with population data (Figure 2).  For the 1990s, Gillett and Lightfoot (2002) estimated offshore and domestic commercial landings (accounting for approximately 77% of domestic supply) as well as subsistence catches (23%; Table 1). In addition, they documented offshore pelagic catches of around 41,000 t, taken by foreign vessels in the late 1990s. This breakdown by fishing sector provides a proxy for local per capita catch rates as well as domestically sourced consumption rates. Domestic fish landings were estimated based on coastal catch  2  United Nations Department of Economic and Social Affairs Population Division: http://www.un.org/esa/population/ [Accessed: February 2010] Table 1. Fishing sector estimates for late 1990s, used as anchor points, from Gillett and Lightfoot (2002). Fishing sector Catch (t) Coastal subsistence 110 Coastal commercial 315 Offshore locally based 50 Offshore foreign based 41,000 Total 41,475 0 2 4 6 8 10 12 1950 1960 1970 1980 1990 2000 P o p u la ti o n  (1 0 3 ) Year Nauruans Pacific islanders Non Pacific islanders Figure 2. Population composition of  Nauru, 1950-2008. 66 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   estimates by Gillett and Lightfoot (2002) (425 t) summed with domestic offshore catch (50 t) for a total of 475 t (Table 1).  The economic downturn of 1999 greatly diminished the population‘s purchasing power, decreasing their ability to pay for commercially sold fish. This is reflected in the increasing proportion of fish sourced via the subsistence sector which increased to 84% of the total domestic landings by 2004 (2008). In addition, a reduction in imported fish (22 t) was reported by Gillett and Lightfoot (2002) for 1999, which is considerably lower than the 55 t of imported canned and salted fish reported a few years earlier by Dalzell and Debao (1994). Nonetheless, a two month survey of the island‘s fishers and families in 2005 estimated that canned fish consumption had risen to an average of 16 kg∙person-1∙year-1 (Vunisea et al., 2008). Because Bell et al. (2009) report that Nauruan fish consumption rates range between 55.8 and 62.3 kg∙person-1∙year-1 for the period between 2001 and 2006, we chose the latter rate to account for the canned fish consumption, along with the 425 t of small scale catch reported by Gillett and Lightfoot (2002), which translates into a demand of approximately 600 t of fish per year for Nauru in more recent years.  Catch composition  Dalzell and Debao (1994) list over 180 species of fish observed and reported in the waters of Nauru. They also report a catch breakdown by gear type collected over an eight month period between July 1992 and February 1993 where at the time, approximately 75% of the fishing was done by migrant workers and where approximately 70% of the total landings were commercially sold. Approximately 60% of the total landings were taken by trolling gear, highlighting a high prevalence of tunas (41% of total catch), mainly skipjack (Katsuwonus pelamis) and yellowfin (Thunnus albacares). Another 30% of the landings were caught using mid-water handlines, almost exclusively catching Carangidae such as rainbow runner (Elagatis bipinnulata) and using demersal handlines, that target various species such as squirrelfish (Holocentridae) and bluestripe snapper (Lutjanus kasmira). The Mid-water and demersal handline fisheries comprise 26% and 7% of the total landings, respectively. The remaining 7% corresponds to inshore reef fishing activities that are regarded as subsistence or semi-artisanal (Vunisea et al., 2008), where catches were composed predominantly of surgeonfish (Acanthuridae, 38.5%), squirrelfish (Holocentridae, 12.1%), groupers (Serranidae, 7.7%) and rainbow runners (Carangidae, 5%). More recently, lesser valued reef fish such as surgeonfish and triggerfish as well as many invertebrates found during beach and reef gleaning (octopus, turban shell and sea cucumbers etc.), account for an increasing 0 100 200 300 400 500 600 1950 1960 1970 1980 1990 2000 C a tc h  ( t) Year Figure 3a. Reconstructed catches breakdown for Nauru, compared with that reported to FAO from 1950 to 2008. Commercial Subsistence As submitted to FAO 0 100 200 300 400 500 600 1950 1960 1970 1980 1990 2000 C a tc h  ( t) Year Other taxa Lutjanidae Carangidae Scombridae Figure 3b. Taxa layout for Nauru’s landings for the period between 1950-2008. Scombridae species 56% of total catch by weight with substantial catches from the Carangidae (31%) and Lutjanidae (6%) families. Within the Scombridae family 90% of the catch were comprised of Skipjack (katsuwonus pelamis) and Yellowfin tuna (Thunnus albacares) a) b) Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  67   proportion of inshore subsistence catches (Vunisea et al., 2008). Those species were not necessarily favoured by the Nauruan locals a decade earlier (Dalzell and Debao, 1994; Gillett and Lightfoot, 2002). Presently, fishers indicated that most of their target species required further distances and hours at sea to catch, but regardless of where it came from, it was intended for their family consumption. Incidentally, all family members participated in one way or another in fishing activities (Vunisea et al., 2008; Gillett, 2009) and the increasingly important reef gleaning activity is generally undertaken mainly by women (Chapman, 1987). Overall, the more detailed breakdown by Dalzell and Debao (1994) was applied to the total catch throughout the 1950-2008 time period.  RESULTS  Data supplied to FAO for Nauru would suggest incorrectly that essentially no fish were caught prior to the early 1960s (Figure 3a). Reported landings increased steadily from 100 t∙year-1 in 1963 to around 190 t∙year-1 in 1991, before increasing substantially to around 500 t∙year-1 in 1993 and 1994 (Figure 3a). Thereafter, data supplied to FAO suggested a rapid decline throughout the 1990s to around 40 t∙year-1 in the early 2000s.  Reconstructed catch estimates suggested a distinctly different picture (Figure 3a). Overall, the 1950-2008 total catches were estimated at 23,150 t, being 3.4 times greater than the data supplied to FAO on behalf of Nauru. Note the slightly lower total estimates for 1993 and 1994 compared to data supplied to FAO. Reconstructed commercial catches increased from around 30 t∙year-1 in 1950 to around 360 t∙year-1 in 1999, before declining rapidly to just under 70 t∙year-1 by 2008 (Figure 3a). Conversely, reconstruction suggested a decline in subsistence catches from 290 t∙year-1 in 1950 to a low of around 110 t∙year-1 in 1991, before increasing rapidly to over 350 t∙year-1 by 2008 (Figure 3a).  The taxonomic breakdown applied to Nauru‘s reconstructed fisheries catches (Figure 3b) is based on the reported catch composition given by Dalzell and Debao (1994) for the commercial catches, and on Vunisea et al. (2008) for the subsistence sector. Commercial catches were dominated by Scombridae (56%), composed of 80% skipjack (Katsuwonus pelamis) 10% yellowfin tuna (Thunnus albacares) and 10% other scombrids, followed by Carangidae (30%, mainly rainbow runner, Elagatis bipinnulata) as well as Lutjanidae (6%, mainly Lutjanus kasmiri), followed by coastal reef species such as surgeonfish (5%) and triggerfish (3%). Other taxa have been increasingly targeted more recently, due to ever more intensive fishing pressure. Overall, there is a predominance of pelagic species in local landings (Figure 3b). Nonetheless, far less of these landings are sold commercially; rather, they are intended for family consumption or sharing with other family members and neighbours. This shift to increasingly non- commercial basis was driven by the recent economical crisis and the weaker financial power of the Nauruan people (Vunisea et al., 2008; Gillett, 2009).   DISCUSSION  Nauru‘s total domestic fisheries catches for the period 1950-2008 were estimated to be 23,150 t. This amount illustrates the historical importance fisheries have in meeting the island‘s dietary requirements, a fact which cannot be readily inferred from the data reported to FAO on behalf of Nauru. Small-scale fisheries are fundamental to many Pacific islands, nonetheless widespread lack of information on subsistence sector catches undervalues the social and economical importance of this sector (Zeller et al., 2006; Gillett, 2009) and may impact any successful form of ecosystem-based fisheries management (Pauly et al., 2002).  The landings reported to the global community on behalf of Nauru substantially underestimate total catches as estimated here for all but two years during the early 1990s. We assume that this peak in reported landings coincides with the publication of the seminal work by Dalzell and Debao (1994) and Dalzell et al. (1996), who presented estimates of per capita catch rates. These studies were likely used to estimate Nauru‘s 1993 and 1994 fisheries landings that were reported to FAO. However, the human population counts for that decade were overestimated (Underwood, 1989), likely resulting in the higher reported landings estimates as presented by FAO compared to our reconstructed estimates. We identified 68 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   that for the early 1950s, population statistics were underestimated by as much as 60% (Underwood, 1989) and in the most recent decade, the population was overestimated by 20% to 40%, due to the rapidly increasing emigration of foreign NPC workers (Vunisea et al., 2008). Hence it is noteworthy to mention that the per capita fish consumption of 56 kg∙person-1∙year-1 reported by Bell et al. (2009), which we used for recent time periods, is likely underestimated, due to it being based on inflated population estimates.  Despite the economic surge provided by the island‘s phosphate earnings during earlier decades, Nauru‘s fisheries did not develop in similar fashion, due in part to the absence of natural harbours to moor vessels. Three man-made channels and a small boat harbour have been excavated through the coral fringing reef, allowing only small outboard powered vessels to be launched. Industrial-scale fishing was attempted in the early 1980s involving two purse-seine vessels, but was not successful due to ineffective gear and through market loss which led to the sale of one vessel while the other was lost at sea (Chapman, 1998).  Nauru has been a party to the US Multilateral Fisheries Treaty since it came into effect in June 1988. In July of 1994, Nauru entered into a bilateral fishing access agreement with Japan, granting four Japanese fishing vessels access to Nauruan waters and in June 1997, the first fishing access agreement was signed with the Philippines (Chapman, 1998). It is estimated that Nauru received about US$3.4 million in access fees in 1999 (Gillett and Lightfoot, 2002) and US$5.4 million in access fees in 2008 (Gillett, 2009), with 131 foreign fishing vessels (10 countries), licensed to fish in Nauru‘s EEZ, catching approximately 66,000 t of tuna.  Attempts to farm milkfish (Chanos chanos) and tilapia (Oreochromis mossambicus) in the past had failed; these fish were introduced into the Baduan Lagoon. Milkfish, although part of the traditional Nauruan diet and culture, were used as bait for tuna fishing (Spennemann, 2002). Tilapia was introduced in the early 1960s but was not accepted as a staple food option and eventually infested all the milkfish ponds causing many farmers to abandon their traditional practice of raising milkfish (Gillett, 2009). Programs to eradicate the introduced fish from the island‘s lagoon have failed (Dalzell et al., 1996). In 2000, 10,000 milkfish fry from Kiribati were introduced into Buada Lagoon, reaping 5,000 adult fish (Gillett, 2009). At present, several milkfish grow-out ponds exist; these are backyard and mostly subsistence operations. No accurate production estimates exist (Gillett, 2009).  Inshore fishing pressure appears to have increased dramatically since the late 1990s, with almost all households involved in fishing; women and children glean the beaches and reefs, collecting all invertebrate and finfish species they come across. Vunisea et al. (2008) surveyed invertebrate catches, estimating the total catch to be 23 t·year-1, dominated by genera such as Etisus, Octopus, Turbo, Thais, Tripneustes and Cardisoma and to a lesser extent Actinopyga, Panulirus, Grapsus, and Cypraea. All species and types are targeted and consumed, with the exception of lobsters, which are destined for sale (Vunisea et al., 2008). All sizes of fish are caught and consumed, however fishers have observed a decrease in size and volume of catches (Vunisea et al., 2008), suggesting that overfishing is occurring. There has been a steady increase in the intensity and frequency of fishing since the economic crisis in 1999, driven mainly by increasing subsistence efforts. Presently, pelagic fishing is dominated by canoes operated by Tuvaluans and I-Kiribati (Vunisea et al., 2008).  Fish continue to form a large part of the Nauruan diet, increasingly so with the island‘s recent economic downturn. Food security may be jeopardized, leading to substantial dietary changes. Imported produce such as meat and poultry have been replaced with canned sardines and mackerel from abroad, seafood provides the main source of protein for more than 98% of Nauruan households (Vunisea et al., 2008). The Nauru Fisheries and Marine Resource Authority (NFMRA) have the responsibility of overseeing, managing and conserving the country‘s natural marine resources and environment. Yet, with fishing being the only major fallback option for the population, the task presents several challenges, especially when marine resources are vulnerable to overexploitation and the livelihoods of an entire nation are at risk.     Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  69   ACKNOWLEDGEMENTS  We thank Ms. Gladys Beccalossi (SPC) for her assistance with population estimates and the Sea Around Us Project, a scientific collaboration between the University of British Columbia and the Pew Environment Group.  REFERENCES  Bell, J., Kronen, M., Vunisea, A., Nash, W., Keeble, G., Demmke, A., Pontifex, S. and Andréfouët, S. (2009) Planning the use of fish for food security in the Pacific. Marine Policy 33: 64-76. Chapman, L. (1998) Feasibility study on infrastructure requirements and vessel parameters for tuna longlining in Nauru (17–22 November 1997). Noumea, New Caledonia 25 p. Chapman, M. (1987) Women's Fishing in Oceania. Human Ecology 15(3): 267-288.Dalzell, P., Adams, T. and Polunin, N. (1996) Coastal fisheries in the Pacific Islands. Oceanography and marine biology 34: 395-531. Dalzell, P. and Debao, A. (1994) Coastal fisheries production on Nauru. Noumea: Inshore Fisheries Research Project, Country Assignment Report, South Pacific Commission, Noumea, New Caledonia, 19 p. Gillett, R. (2003) Subregional review - Small island developing states of the Southwest Pacific. FAO, Rome, 121-139 p. Gillett, R. (2009) Fisheries in the economies of the Pacific isalnd countries and territories. Asian Development Bank, Mandaluyong City, Philippines, 484 p. Gillett, R. and Lightfoot, C. (2002) The contribution of fisheries to the economies of Pacific Island countries. Asian Development Bank, Manila, 242 p. Jacob, P. (1998) The status of marine resources and coral reefs of Nauru. Coral reefs in the Pacific: Status and monitoring, resources and management: 207-316. Pauly, D., Christensen, V., Guénette, S., Pitcher, T., Sumaila, U., Walters, C., Watson, R. and Zeller, D. (2002) Towards sustainability in world fisheries. Nature 418: 689-695. Petit-Skinner, S. (1981) The Nauruans. MacDuff Press, 321 p. Sokimi, W. and Chapman, L. (2001) Small-scale tuna longlining assistance and training for the Republic of Nauru. Fisheries Development Section. Fiend Report, Noumea, New Caledonia, 27 p. Spennemann, D. (2002) Traditional milkfish aquaculture in Nauru. Aquaculture International 10: 551-562. Stephen, M. (2011) On Nauru, a sinking feeling. New York Times Opinion Pages, July 18, 2011. Available at: www.nytimes.com/2011/07/19/opinion/19stephen.html?_r=1&ref=todayspaper [accessed July 21, 2011] Underwood, J. (1989) Population history of Nauru: A cautionary tale. Micronesica 22: 3-22. Viviani, N. (1970) Nauru: Phosphate and political progress. Australian National University Press, 215 p. Vunisea, A., Pinca, S., Friedman, K., Chapman, L., Magron, F., Sauni, S., Pakoa, K., Awira, R. and Lasi, F. (2008) Nauru country report: Profile and results from in-country survey work. SPC, Secretariat of the Pacific Community, Noumea, New Caledonia, 68 p. Zeller, D., Booth, S., Craig, P. and Pauly, D. (2006) Reconstruction of coral reef fisheries catches in American Samoa, 1950–2002. Coral Reefs 25: 144-152. Zeller, D., Booth, S., Davis, G. and Pauly, D. (2007) Re-estimation of small-scale fishery catches for US flag-associated island areas in the western Pacific: the last 50 years. US Fishery Bulletin 105: 266-277. 70 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller    Appendix Table A1: Total FAO landings vs. total reconstructed catch for Nauru, 1950-2008, in metric tonnes. Year  FAO landings  Total reconstructed 1950 0  296 1951 0  301 1952 0  306 1953 0  311 1954 0  315 1955 0  319 1956 0  323 1957 0  327 1958 0  331 1959 0  334 1960 0  337 1961 0  342 1962 0  347 1963 100  352 1964 100  356 1965 100  360 1966 100  364 1967 100  367 1968 100  370 1969 100  373 1970 100  375 1971 100  378 1972 100  380 1973 100  381 1974 100  382 1975 120  384 1976 120  384 1977 120  385 1978 130  394 1979 130  392 1980 140  391 1981 140  389 1982 140  387 1983 140  385 1984 150  388 1985 150  392 1986 150  394 1987 170  397 1988 170  398 1989 180  399 1990 180  452 1991 190  444 1992 377  441 1993 500  447 1994 500  455 1995 400  464 1996 300  472 1997 250  481 1998 200  490 1999 150  493 2000 109  443 2001   61  442 2002   22  440 2003   44  438 2004   19  437 2005   39  437 2006   39  447 2007   39  447 2008   39  447 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  71    Appendix Table A2: Total reconstructed catch (t) by major taxa for Nauru, 1950-2008. Others category contains 7 taxonomic groups including miscellaneous marine fishes. Year Katsuwonus pelamis Elagatis bipinnulata Thunnus albacares Scombridae Carangidae Lutjanus kasmiri Crustaceans nei Others 1950 130 72 16 16 18 14 6 24 1951 132 73 17 17 18 14 6 25 1952 134 74 17 17 19 14 6 25 1953 136 75 17 17 19 15 6 25 1954 138 76 17 17 19 15 6 26 1955 140 77 17 17 19 15 6 26 1956 142 78 18 18 20 15 7 26 1957 143 79 18 18 20 15 7 27 1958 145 80 18 18 20 16 7 27 1959 146 81 18 18 20 16 7 27 1960 147 82 18 18 20 16 7 28 1961 150 83 19 19 21 16 8 28 1962 152 84 19 19 21 16 8 28 1963 171 73 21 21 18 14 8 25 1964 173 74 22 22 18 14 8 25 1965 175 75 22 22 19 14 8 25 1966 176 75 22 22 19 15 9 26 1967 178 76 22 22 19 15 9 26 1968 179 77 22 22 19 15 9 26 1969 180 78 23 23 19 15 9 26 1970 181 78 23 23 20 15 10 27 1971 182 79 23 23 20 15 10 27 1972 183 79 23 23 20 15 10 27 1973 183 79 23 23 20 15 10 27 1974 184 80 23 23 20 15 10 27 1975 184 80 23 23 20 15 11 27 1976 185 80 23 23 20 15 11 27 1977 185 80 23 23 20 15 11 27 1978 188 82 24 24 21 16 12 28 1979 188 82 23 23 20 16 12 28 1980 187 81 23 23 20 16 12 28 1981 186 81 23 23 20 16 12 28 1982 185 80 23 23 20 16 12 27 1983 184 80 23 23 20 15 13 27 1984 185 80 23 23 20 16 13 28 1985 187 81 23 23 20 16 13 28 1986 188 82 23 23 20 16 14 28 1987 188 82 24 24 21 16 14 28 1988 189 82 24 24 21 16 15 28 1989 189 83 24 24 21 16 15 28 1990 213 96 27 27 24 18 15 33 1991 209 93 26 26 23 18 16 32 1992 207 93 26 26 23 18 16 32 1993 210 94 26 26 24 18 16 32 1994 214 96 27 27 24 19 16 33 1995 218 98 27 27 25 19 16 34 1996 221 100 28 28 25 19 16 34 1997 225 102 28 28 26 20 16 35 1998 229 105 29 29 26 20 16 36 1999 230 105 29 29 26 20 16 36 2000 190 105 24 24 26 20 16 37 2001 190 105 24 24 26 20 16 37 2002 189 105 24 21 26 20 16 39 2003 189 104 24 12 26 20 16 48 2004 188 104 24 21 26 20 16 38 2005 188 104 24 21 26 20 15 39 2006 193 107 24 21 27 21 15 39 2007 193 107 24 21 27 21 15 39 2008 193 107 24 21 27 21 15 39  72 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller    Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  73   MARINE FISHERIES OF PALAU, 1950-2008: TOTAL RECONSTRUCTED CATCH1  Stephanie Lingard, Sarah Harper, Yoshi Ota and Dirk Zeller Sea Around Us Project, Fisheries Centre, University of British Columbia, 2202 Main Mall, Vancouver, BC, V6T 1Z4, Canada s.lingard@fisheries.ubc.ca; s.harper@fisheries.ubc.ca; d.zeller@fisheries.ubc.ca   ABSTRACT  The small Pacific Island nation of Palau has a long history of human settlement. Palau maintained a predominantly traditional lifestyle until the post-war modernization after 1950s, with fishing being a pre- occupation for the majority of its male population. This study estimated Palau‘s total marine fisheries catches for the 1950-2008 period to be just over 200,000 tonnes. This total was 43% higher than the official reported data as presented by the FAO on behalf of Palau. The discrepancy was mainly due to subsistence catches which were under-reported in the official statistics. The total coastal catches including subsistence and artisanal, were estimated to be 103,480 t 45% higher than the 46,615 t of coastal catches reported over the period.  Our findings illustrate the importance of the subsistence sector, with catches representing 60% of coastal fisheries catches. Better monitoring or at least regular comprehensive estimation of the subsistence sector is key to properly account for the social and economic importance of fishing in Palauan society.   INTRODUCTION  Palau, a small country in the Western Pacific, is comprised of 340 islands which lie between 131°-135°E and 2°- 8°N, 500 kilometers east of the Philippines (Figure 1). Palau is located within FAO statistical area 71, the Western Central Pacific, has a land mass of approximately 488 km2 and an exclusive economic zone (EEZ) of around 604,289 km2 (www.seaaroundus.org). The five main inhabited islands of Palau are Kayangel, Babeldoab, Korror, Peliliu, Angaur, and it has two de-populated outer islands, being Sonsorol and Hatohobei (Figure 1). The history of human settlement in Palau is long, with the earliest archaeological findings in Palau dating back over 2400 years (Clark, 2005). In recent human history, Palau has been successively colonized and under external stewardship by different states; Palau was ruled by Spain in the 1880s, Germany from 1899-1913 and Japan from 1914-1944. Following World War II, Palau was administered by the United  1  Cite as: Lingard, S., Harper, S. Ota, Y. and Zeller, D. (2011) Marine Fisheries of Palau,1950-2008: Total reconstructed catch. pp.73- 84. In: Harper, S. and Zeller, D. (eds.) Fisheries catch reconstructions: Islands, Part II. Fisheries Centre Research Reports 19(4). Fisheries Centre, University of British Columbia [ISSN 1198-6727]. Figure 1. Map of Palau and its Exclusive Economic Zone (solid line). 74 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   States. In 1994, Palau was designated as an independent country under the Compact of Free Association (Anon., 2003a) (Ota, 2006). The Compact of Free Association entails the United States providing military defense to Palau, development aid in the form of annual grants, and scholarship programs for Palauan residence to attend pos- secondary education in the United States. Due to this external support gradual development of some local industries, such as the tourism sector, Palau has the highest per capita GDP among Micronesia, and higher than most small island states in other Pacific region2.  Traditionally, Pacific islanders have relied on marine resources as their main source of food, and fishing skills and knowledge were recognized as the status symbol of both wisdom and masculinity among many Pacific cultures (Johannes, 1981; Colbert, 2000). Likewise, in Palau, fishers were highly esteemed members of the community (Ota, 2006). Both Johannes (1981) and Ota (2006) have described fishing to be central to the organization of Palauan communities, embodying the gender dichotomy and social stratification, which still supports the basis of the socio-cultural dynamics of the society. In Palauan tradition, for instance, women have traditionally engaged in reef gleaning activities (Chapman, 1987), and farming of taro, while men capture fin fish which make up the majority of Palauan diet (Johannes, 1981; Mathews and Oiterong, 1991). Sharks and larger pelagic species are usually only caught for special occasions (Johannes, 1981).  Prior to colonization, Palauan society was organized into separate communities, each with its own respective chief.  Between communities, land and sea barriers were continuous, and each community had proprietorship of sections of the coast line and reef. From ownership of and responsibility for a defined geographical area came a natural conservation ethic (Johannes, 1981). With the succession of colonizing countries, this system was increasingly disturbed by external political control and eventually replaced by modern democracy, which maintains the traditional chieftainship merely as a façade (Ota, 2006). More recently, the increase in the importance of the cash economy has led rapid urbanization of the country as people started seeking work in the capital, Koror (Johannes, 1981; Ota, 2006). In 1990, an estimated 70% of Palauans were living in Korror (Nichols, 1991). Even in the midst of this modernization, fishing has remained both economically and culturally important and is practiced regularly for subsistence purposes, though not commercially, by many Palauans both in their urban and rural settings. However, the centralization of marine resource management to the contemporary governing body has had negative effects on Palauan fishing culture and fish populations from the early days of the Palauan modernization as it slowly replaced the power of traditional community based management (TCBM) system which was based on the indigenous chieftainship (Johannes, 1981). For instance, Johannes (1981) reports that species of the Serranidae (e.g., groupers) were quickly overexploited due to fishers targeting spawning aggregations, which were previously controlled through the traditional community-based management (TCBM) system. Dynamite, poison, and other unsustainable fishing practices have also been employed since the breakdown of the TCBM, resulting in negative impacts on the reef fisheries of Palau (Johannes, 1981).  Since the Japanese occupation of Palau in 1914, tuna have been exploited in Palauan waters, and in recent decades, this has provided an important source of income for Palau, as fishing access fees for foreign vessels were introduced. Gillett(2009) estimated that in 2007 approximately 1.2 million USD were paid to Palau for access to fish in their waters. However, during WWII, all off-shore tuna fishing was halted until 1964 when an American company, Van Camp Seafoods, opened a processing facility in Koror (Lawson, 1991). Van Camp Seafoods operated until 1982. Since then, there has only been one locally based off-shore pole-and-line vessel operating under the Palauan Flag (Nichols, 1991).  The subsistence fishery has been largely unaccounted for in the FAO data for many island countries, despite its significant contribution to food security and local economies (Zeller et al., 2006). It has been noted that Palau‘s inshore coral reef fishery continues to provide the main source of protein, and financial income for the majority of Palauan people (Johannes, 1981; Ota, 2006). In the recent period it has been estimated that 87% of Palau‘s population is engaged in coastal artisanal and subsistence fishing activities (Palau International Coral Reef Center, unpublished data in Golbuu et al, 2005). The artisanal and subsistence reef fisheries are carried out with a variety of gears. Fish pots, drop lines, trolling, hand  2  http://www.adb.org/Documents/Fact_Sheets/PAL.pdf [accessed June, 2011] Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  75   spears, spear guns, gill nets, set nets (kesokes), and cast nets are the major gear types employed. The major species fished for in the small-scale fishery are snappers (Lutjanidae), emperors (Lethrinidae), groupers (Serranidae), parrot fishes (Scaridae), wrasses (Labridae), rabbitfishes (Siganidae), surgeon fishes (Acanthuridae), trevallies (Carangidae) and herring (Clupeidae) (Nichols, 1991).  Invertebrate fisheries are important components of the local diet and economy, but are poorly represented in official fisheries statistics. In earlier time periods most of the invertebrate fisheries were reported as being handled by women and children, predominantly for subsistence use (Johannes, 1981; Mathews and Oiterong, 1991), but they are now caught by commercial fishers and were largely sold at the local market mainly to the tourism sector (Anon., 2003b; Pakoa et al., 2009). Bêche-de-mer is an important part of the Palauan diet, and is often collected by women during reef gleaning activities (Johannes, 1981; Mathews and Oiterong, 1991; Pakoa et al., 2009). Bêche-de-mer is considered a boom and bust fishery, as the catches can fluctuate substantially year-to-year due to its ease of harvest and the open access nature of this fishery (Dalzell et al., 1996). Other important invertebrate resources in Palau are the mud crab (Scylla serrata), land crab (Cardiosoma hirtipes and Cardisoma carnifex), and coconut crab (Birgus latro) (Johannes, 1981; Nichols, 1991; Dalzell et al., 1996).  Since the 1990s, there have been some improvements to the documentation of the offshore and artisanal sectors of Palauan fisheries also reflected in better taxonomic breakdown in FAO data; however, to date no studies exist that include all fisheries components in a single estimate with a complete estimated time series (although Gillet [2009] does provide a very comprehensive estimate for the recent time period). The aim of this study is to make a comprehensive estimate of total marine fisheries catches for Palau that includes the invertebrate, subsistence, artisanal, offshore, and baitfish fisheries sectors over the 1950- 2008 time period.   METHODS  For this report, officially reported landings were acquired from FAO (FishStat), and data were obtained from government reports, and independent reports published by the Secretariat of the Pacific Community (SPC) and Asian Development Bank (ADB). For the time period 1990-2007, several estimates of subsistence, artisanal and offshore commercial fisheries catches were available for some years (Nichols, 1991; Adams and Dalzell, 1994; Kitalong and Dalzell, 1994; Dalzell et al., 1996; 2001; Gillett and Lightfoot, 2002; Gillett, 2009). Prior to 1990, few studies on commercial reef and off-shore fisheries were available (Johannes, 1981; Perron et al., 1983; 1984). Data for certain years and for certain sectors were taken from Johannes (2009), Kitalong and Dalzell (1994), Dalzell et al.(1996), Gillett and Lightfoot (2002), and Gillett (2009). To derive a complete time series of data (1950-2008), we interpolated linearly between years of known data. Human population data and per capita fish consumption rates were used to calculate total seafood demand and secondarily derive subsistence sector catches for the 1950-2008 time period.  Human population  Human population census data were obtained from the Palau Office of Planning and Statistics (http://www.palaugov.net/stats) and used in combination with per capita consumption rates to calculate Palau‘s seafood demand. Linear interpolations were used to create a continuous time series of human population data from 1950 to 2008, as data were not available for all years (Figure 2). To account for the temporary increase in population due to tourists, we converted number of tourists 0 5 10 15 20 25 1950 1960 1970 1980 1990 2000 P o p u la ti o n  ( x  1 0 3 ) Visitors (full-time equivalents) Resident population Figure 2. Population of Palau, 1950-2008 with the resident population and the number of visitors converted to full-time equivalents. 76 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   into full-time equivalents (Figure 2). To do this, we multiplied the number of visitors by the average number of nights stayed. The length of stay was derived from the estimated number of full-time equivalents presented in Gillette (2009) for the 1990s and the average number of annual visitors for the 1990s taken from the Palau Visitors Authority, published in Yamashita (2000). We estimated the average length of stay to be 3.6 days∙visitor-1∙year-1 and applied this to the time series of visitors, which was for the period 1980-2008. The number of full-time equivalents was added to the resident human population and the total population was then used in conjunction with per capita seafood consumption rates to calculate the total demand of fishery products over the study period.  Commercial Fisheries  Inshore fisheries  Several independent studies on commercial inshore fisheries estimate annual catches (Table 1). The PCS (2000) estimate of 865 t has been the most widely agreed upon estimate for coastal fish production in Palau (Gillett, 2009). This estimate was compared to total catches presented by FAO for all species except large pelagics (tuna and billfishes). In 2007, FAO (non-pelagic) landings were found to be 11% higher than the PCS estimate of 865 t∙year-1 of coastal commercial catch. We assumed that this additional catch (11%) presented in the FAO data was an estimate of non- commercial (subsistence) catches. Therefore, we assumed that 89% of FAO reported landings (excluding large pelagics) represent the coastal commercial (artisanal) catch, while the remaining 11% were considered the reported or estimated component of subsistence catches.  Offshore fisheries  Palau‘s offshore fishery began in the 1920s when Japan occupied the islands. After WWII, tuna fisheries in Palau ceased and it was not until the early 1960s, after a fisheries development program was launched to jump start Palauan offshore fisheries, that tuna fishing resumed. In 1964, a joint-venture company (Van Camp Seafood of the United States) began operating a pole-and-line tuna fishery in Palau, which lasted until 1982 (Anon., 1984). These catches were considered domestic as they were caught and landed by Palau, even though the financing for this fishery was from the United States. Aside from Van Camp Seafood, offshore tuna fisheries in Palau have been mainly foreign fleets, which land their catches outside Palau (Anon., 1944; Gillett and Lightfoot, 2002). After the closing of Van Camp Seafood, only a single domestic pole-and-line vessel remained in operation. Based on our comparison of FAO tuna and billfish landings with other reports (e.g., Gillett, 2010), we concluded that the FAO data were the best available representation of domestic, large pelagic fishery catches.  Due to the minimal bycatch associated with pole-and-line fisheries (Bailey and Williams, 1996; Kelleher, 2005) we did not estimate bycatch, either landed or discarded. However, pole-and-line fisheries do require considerable amounts of live bait which are often caught in reef areas adjacent to the tuna fishing grounds. These catches are rarely accounted for in catch statistics, and we assumed that these were not estimated or included in the FAO data. Therefore, we estimated the amount of fish that was likely caught in order to provide bait for the domestic tuna fishery. Gillette(2011) gives an average tuna-to-baitfish ratio of 26:1 for Palau‘s pole-and-line fishery between 1964 and 1972. We assumed a similar ratio for the entire 1964-1982 period, when Van Camp Seafood was in operation. This ratio was applied to the total tuna landings from 1964-1982. In 1982, the bait fishery in Palau ceased operations, likely as a result of Van Camp Seafood discontinuing operations in Palau (Anon., 1984). Although one domestic tuna vessel continued to fish after 1982, we did not make any further estimates of baitfish catches. Catch of baitfish was likely dominated by short head anchovy (Encrasicholina heteroloba), delicate round herring (Spratelloides delicatulus), and Samoan silverside (Hypoatherina temmincki). From the estimated catch by species given in the SPC (1984) report on the bait fishery, we derived a species breakdown of 56% Encrasicholina heteroloba, 21% Spratelloides delicatulus and 8% Hypoatherina temminckim, with the remaining 15% being Clupeiformes. Table 1. Comparison of commercial inshore fisheries catch estimates for the recent period (1990s & 2000s) from various independent sources. Catch (t/year) Source 300-400 Shimada (1987 in Nichols, 1991) 250 Kitalong and Dalzell (1994) 865 PCS (2000 in Gillett and Lightfoot, 2002) 736 Anon. (1993 in Dalzell et al., 1996) Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  77    Subsistence fishery  To independently estimate the subsistence catch in Palau, we used per capita fish consumption estimates for three separate years, and interpolated linearly between these to derive a complete time series of per capita fish consumption rates. The per capita consumption rates used were 122 kg∙person-1∙year-1 for 1974 (Johannes, 1981), 135 kg∙person-1∙year-1 for 1999 (Gillett and Lightfoot, 2002), and 115 kg∙person-1∙year-1 for 2007 (Gillett, 2009). Johannes‘ (1981) estimate of per capita fish consumption was calculated using the weight of protein consumed daily, and the landed weight of fish (with weight of bones and scales removed). The estimate from Gillett and Lightfoot (2002) was calculated using the PCS (2000) in-shore catch estimate of 2,115 t adjusted for imports, exports and full time visitor equivalents. The consumption rate calculated from Gillett (2009), which used the same methodology as the Gillett and Lightfoot(2002) and population and visitation data from 2007, was adjusted for imported and exported fishery products and excluded pelagic species which do not form a significant portion of the Palauan diet. Other estimates of per capita fish consumption were available (Perron et al., 1983; Preston, 1990) but were general estimates for the South Pacific region or were partially derived using FAO or commercial landings data. We disregarded these estimates based on information regarding the Palauan diet which suggested that although imported food, pigs, and bats contribute in small parts to the Palauan diet, fish still remains the main source of protein (Johannes, 1981; Anon., 2003b; Ota, 2006). Our earliest estimate of per capita fish consumption (Johannes, 1981) was carried back fixed from 1974 to 1950 and linear interpolations were used between 1974 and the two later estimates to derive a complete time series of per capita fish consumption rates. These rates were then multiplied by the annual human population to derive the total domestic seafood demand. This was then used to determine whether the demand was met through the supply of reported landings. The discrepancy found between these two numbers was then considered unreported subsistence catch. To this we added the 11% of FAO non-pelagic catches that were considered reported subsistence catches. The total subsistence catch (reported + unreported) was then compared to our estimate of reconstructed coastal commercial catches. On average, subsistence catches represented approximately 60% of the total coastal catches (subsistence and artisanal combined). For the South Pacific in general, the ratio of subsistence to total catch can be as high as 80% (Adams and Dalzell, 1994) with the average being around 70%(Gillett, 2009). We thus considered our approach to be conservative.  Invertebrate fisheries  Estimates of invertebrate fisheries catches were not readily available; however, sea cucumber (bêche-de- mer) has been a component of the Palauan diet for centuries (Pakoa et al., 2009). Sea cucumber fisheries often fluctuate considerably from year-to-year (Dalzell et al., 1996); however, catches reported by FAO for this fishery are minimal (<0.5 t), despite it being an important component of the subsistence fishery. Sea cucumber and invertebrates are collected in reef gleaning activities by women (Mathews and Oiterong, 1991). Sea cucumber landings were estimated to be 11.3 t∙year-1 between 1989 and 1998 (Pakoa et al., 2009). Over 50% of this was for subsistence consumption, 48% was sold in local markets and less than 1% Table 2. Taxonomic composition of the non-pelagic catches (as proportion of total catch) by family or grouping based on Kitalong and Dalzell (1994) for 1950- 1990 and Friedman et al. (2007) for 2007-2008. A linear interpolation was used to derive a complete time series between 1990 and 2007. Family 1950-1983 1984-1990 1991-2006 2007-2008 Acanthuridae 0.141 0.129 linear interpolation 0.104 Carangidae 0.026 0.040 linear interpolation 0.022 Gerridae 0.007 0.002 linear interpolation 0.011 Haemulidae 0.002 0.001 linear interpolation 0.015 Holocentridae 0.004 0.001 linear interpolation 0.011 Labridae 0.010 0.008 linear interpolation 0.008 Lethrinidae 0.119 0.139 linear interpolation 0.268 Lutjanidae 0.085 0.135 linear interpolation 0.140 Mugilidae 0.019 0.005 linear interpolation 0.024 Mullidae 0.010 0.012 linear interpolation 0.023 Scaridae 0.163 0.187 linear interpolation 0.154 Serranidae 0.091 0.093 linear interpolation 0.105 Siganidae 0.116 0.105 linear interpolation 0.105 Others 0.185 0.123 linear interpolation 0.005 Crustaceans 0.020 0.020 linear interpolation 0.005 78 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   was exported. Pakoa et al. (2009) estimate that in 2007, sea cucumber catches were approximately 65.5 t∙year-1. Given the substantial fluctuations characteristic of sea cucumber fisheries and the limited catch data, we used the average of Pakoa et al.‘s (2009) estimate for the 1990s (11.3 t) and for 2007 (65.5 t). The average of these two was approximately 38.4 t (19.2 t each from artisanal and subsistence sectors), which we applied as the catch for the year 2000. This was converted to a per capita rate, and using the population time series, sea cucumber catches were estimated for the entire study period. Sea cucumber catch estimates were then subtracted from the ‗other species‘ category of the artisanal and subsistence catch. (Anon., 2003a)  Crustacean fisheries are also an important part of subsistence and, to a certain extent, commercial fisheries in the South Pacific (Dalzell et al., 1996). Palau‘s catches of crustacean in the early to mid-1990s were estimated by Dalzell et al. (1996) to be 14.4 t∙year-1, which represents approximately 2% of artisanal catches. Prior to 1990, FAO catches for Palau present all catches as either miscellaneous marine fishes or pelagic fish species. As of 1990, the FAO provides greater taxonomic detail, which includes several invertebrate categories. Between 1990 and 2008, FAO crustacean categories represent on average 0.8% of the total non-pelagic landings. Crustacean catches were subtracted from the ‗other species‘ category of the artisanal and subsistence catches.  Taxonomic composition  FAO data for Palau have poor taxonomic resolution until the 1990s. For the period 1950-1989, landings are mainly reported as the aggregated grouping, ‗miscellaneous marine fishes (MMF)‘. Perron (1983) presents a breakdown of the artisanal fishery by family, but these estimates were highly variable. Dalzell et al. (1996) has a partial breakdown of the 1992 artisanal fishery including an estimate for the crustacean fishery. Kitalong and Dalzell (1994) have a taxonomic breakdown by family for the years 1976-1990, with two separate breakdowns for the periods 1976-1983 and 1984-1990. Friedman et al. (2007) have the most detailed breakdown of the artisanal fishery including species-level composition for catches in four major fishing districts in Palau. Here, we used the breakdown in Kitalong and Dalzell (1994) for the period 1950-1990 and Friedman et al. (2007) for 2007 and 2008 (Table 2).  Table 3. Species composition of the families, which dominate the coastal fisheries of Palau based on Friedman et al. (2007). Family Taxonomy Proportion of catch by family Acanthuridae Acanthurus xanthopterus 0.22  Naso lituratus 0.34  Naso unicornis 0.21  Acanthuridae 0.23 Carangidae Caranx ignobilis 0.24  Caranx melampygus 0.43  Carangidae 0.33 Gerreidae Gerres macrosoma 0.97  Gerreidae 0.03 Haemulidae Plectorhinchus albovittatus 0.51  Plectorhinchus spp.  0.49 Holocentridae Holocentridae 1.00 Labridae Cheilinus undulatus 0.30  Choerodon anchorago 0.70 Lethrinidae Lethrinus harak 0.09  Lethrinus olivaceus 0.19  Lethrinus rubrioperculatus 0.06  Lethrinus xanthochilus 0.15  Lethrinidae 0.51 Lutjanidae Aprion virescens 0.09  Lutjanus bohar 0.15  Lutjanus gibbus 0.63  Lutjanidae 0.13 Mugilidae Mugilidae 1.00 Mullidae Mullidae 1.00 Scaridae Cetoscarus bicolor 0.17  Hipposcarus longiceps 0.45  Scarus ghobban 0.15  Scarus oviceps  0.11  Scaridae 0.12 Serranidae Plectropomus areolatus 0.22  Plectropomus leopardus 0.25  Epinephelus spp 0.44  Serranidae 0.09 Siganidae Siganus lineatus 0.57  Siganus spp. 0.43   Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  79   Using Friedman et al. (2007) we calculated the taxonomic composition using the weighted average of each fishing district as a proportion of the total catch in the study, and each area (lagoon, sheltered reef, outer reef) as a proportion of the total in each district. The proportion of each species in each region was then weighted according to the fishing area and district. The proportion each species represented in the total catch was taken as the sum of the proportion of the catch by weighted area. To compare this breakdown with that presented in Kitalong and Dalzell (1994), we grouped the species by family (see Table 3). In order to derive a complete time series between 1990 and 2007, we interpolated linearly between the Kitalong and Dalzell (1994) and Friedman (2007) estimates for each taxonomic family. The 1976 estimate was carried back, unaltered to 1950 and the 2007 estimate was carried forward unaltered to 2008.  Using Dalzell et al.’s (1996) estimate of crustacean catch for the early 1990s, we assumed 2% of artisanal catches were crustaceans for the period 1950-1990. This was incorporated into our artisanal sector taxonomic breakdown by subtracting 2% from the ‗others‘ category in the original breakdown. In 2007, we assumed a lower proportion of crustaceans (0.5%) in the artisanal breakdown as after 1990 crustacean categories begin to appear in the FAO data. The taxonomic breakdown with crustaceans incorporated was then applied to the unspecified portion of the artisanal catch (miscellaneous marine fish category in the FAO data). We assumed the same species breakdown for the subsistence catch. Sea cucumber catch which was estimated using a per capita rate was then subtracted from the ‗others‘ category for the artisanal and subsistence catch in all years.  For each family, a species-level breakdown was derived using Friedman‘s (2007) estimates (Table 3). This species breakdown was applied to each family group throughout the 1950- 2008 time period for the artisanal and subsistence fishery catches. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Figure 5.  Palau reconstructed  and FAO coastal catches 1950-2008. FAO reconstructed 0 2 4 6 8 10 12 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Large pelagic fish Baitfish Figure 4. Total catch of large pelagic species and the baitfish associated with the main gear used in this fishery: pole-and-line. 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Bait fish FAO Tuna Unreported subsistence Reported subsistence Figure 3. Palau total reconstructed catch including subsistence, artisanal, locally based tuna, baitfish, 1509-2008. 80 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller    RESULTS  Our estimate of the total reconstructed catch for Palau, which includes subsistence, artisanal, locally- based tuna fisheries and baitfish totalled 200,817 t for the 1950-2008 time period. This estimate was 43 % higher than the total landings presented by FAO on behalf of Palau for 1950-2008, which was estimated to be 140,483 t (Figure 3). The FAO reports 93,868 t of pelagic species, which were considered to be from the domestic offshore fisheries (Figure 4).  The remaining 49,609 t reported by the FAO were considered to be the coastal fisheries catches. Of the coastal catches presented by the FAO, 89% were considered to be from the commercial (reported artisanal) sector, while the remaining 11% were considered to be non-commercial (reported subsistence) catches (Figure 3). Reported artisanal catches totalled approximately 41,488 t and reported subsistence catches were just over 5,000 t (Figure 5). In addition to the reported subsistence, we estimated another 56,800 t of unreported subsistence (Figure 3). Subsistence catches were estimated to be almost 62,000 t for the 1950-2008 time period. Total coastal catches were estimated to be 103,480 t, 45% higher than the total reported coastal catches (Figure 5.)  Thus, total subsistence catches accounted for 60% of total coastal catches. Subsistence catches in 1950 represented 70% of the total coastal catch (total artisanal + total subsistence), decreasing to approximately 50% during the 1970s and 1980s and increasing again to approximately 65% in the recent period (1990s and 2000s; Figure 6).  The amount of baitfish caught for use in the pole-and-line fishery was estimated to be almost 3,500 t over the period 1964-1982 (Figure 4).   DISCUSSION  The total reconstructed catch for Palau from 1950-2008 was estimated to be 200,817 tonnes. This estimate is 43% higher than the landings reported by FAO on behalf of Palau for this same period (140,483 tonnes). The difference between these two estimates is mainly attributed to the addition of subsistence catches, which were predominantly unaccounted for in the reported data.  Our work showed that subsistence catches represented an average of 60% of the total domestic catch. For the Pacific Islands region, in general, Adams and Dalzell (1994) estimate that the subsistence catch can represent as much as 80% of the total catch. Currently, an estimated 87% of people are involved in the fishing industry in Palau (Palau International Coral Reef Center, unpublished data in Golbuu et al., 2005). In recent years there has been an increase in the amount of imported food in the Palauan diet (Johannes, 1981; Ota, 2006), but the results of this study show reliance on local reef fish remains high in Palau.  The subsistence catch has increased, since 1950, mostly attributed to increase in population size. The subsistence catch will likely continue rising due to a growing population (Anon., 2003a). The majority of the domestic supply of fish in the 1950s was from the subsistence sector (70%) and although the portion of the total domestic catch supplied by subsistence fisheries declined during the 1970s and 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 1950 1960 1970 1980 1990 2000 S u b si st e n c e  p o r ti o n  o f to ta l c o a st a l c a tc h  Figure 6. Subsistence portion of the total coastal catch, which includes artisanal and subsistence catches, 1950-2008. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  81   1980s,—likely due to an increase in commercial exploitation of fisheries resources—the subsistence sector continues to dominate domestic fisheries catches. The under-reporting of subsistence catches may adversely affect food security in Palau given the importance of this sector to the Palauan culture and diet. Nichols (1991) notes that one species in particular, the humphead wrasse (Cheilinus undulatus), is rarely sold due to a strong preference for this fish in the local diet. It is caught in unknown amounts and usually kept for home consumption. A lack of monitoring and reporting of species important to the subsistence fishery, such as C. undulatus, could lead to the silent extirpation of species from Palau‘s reefs.  The commercial offshore catches in Palau were well documented from 1937-1944 by the Japanese Office of the Chief of Naval Operations. The average catch during this time was approximately 3000 tonnes (Anon., 1944). During WWII, all commercial offshore fishing ceased. Catches prior to WWII, while being quite substantial, are not included as they fall outside the time period considered for this study. Recording of pelagic species, mostly skipjack, began in 1964. From 1964-1982 Van Camp Seafoods operated in Palau, and the catches are well documented in Lawson (1991).  Our estimate of total catches is likely still an under-representation of the actual annual catch due to the conservative approach employed in our estimations and the omission of the recreational fishery sector. The South Pacific Commission (1999) estimates roughly 8.3 t of billfish are taken annually in game fish tournaments in Palau. The major species taken in this fishery are blue marlin (Makaira nigricans), black marlin (Makaira indica), striped marlin (Tetrapturus audax), and sailfish (Istiophorus platypterus). Recreational fishing is on the rise in Palau as tourism increases (Nichols, 1991). Better documentation of this fishery is recommended for the conservation of billfish in Palau‘s EEZ.  Multiple studies have estimated Palau‘s offshore, artisanal, subsistence and invertebrate fisheries separately; however, this study is the first to combine all these sectors into one comprehensive estimate with a continuous time series from 1950-present. This study revealed some major deficiencies in the availability of fisheries data covering all sectors, with a particular scarcity of data on the subsistence sector, which could be argued is the sector most important to day to day existence of the Palauan people. Our comprehensive estimate of total marine fisheries catches for Palau will hopefully serve as a more realistic baseline from which fisheries management options and tradeoffs can be assessed.   ACKNOWLEDGMENTS  We thank the Sea Around Us Project, a scientific collaboration between the University of British Columbia and the Pew Environment Group. We would also like to thank Dr. Takashi Mita of the Global Collaboration Centre in Osaka University for graciously providing input for this study.  REFERENCES Anon. (1944) The Fishing Indistry of the Japanese mandated islands. United States Foreign Economic Administration Supply and Resources services, Office of the Economic Programs, Foreign Economic Administration for Occupied areas section, Cheif of Naval Operations, Washington, D.C. Anon. (1984) An Assessment of the skipjack and baitfish resources of Northern Marianna Islands, Guam, Palau, Federated States of Micronesia, and Marshall Islands. Skipjack Survey and Assessment Programme. South Pacfic Commission, Noumea, New Caledonia, xi+ 111 p. Anon. (2001) Profiles of Palau's Inshore Fisheries, 1989-1998. Palau Conservation Society, Korror, Palau. Anon. (2003a) 2000 Census of Population and Housing of the Republic of Palau. Office of Planning and Statistics, Government of Palau, Korror, Palau. Anon. (2003b) Community Consultations on Marine and Terrestrial Resources Uses. Palau Conservation Society, Korror, Palau. Bailey, K. and Williams, P.I. (1996) By-Catch and Discards in the Western Pacific Tuna Fisheries: A review of SPC Data holdings and literature. Oceanic Fisheries Programme. South Pacific Commission, Noumea, New Caledonia, vii+ 153 p. Chapman, M. (1987) Women's Fishing in Oceania. Human Ecology 15(3): 267-288.Dalzell, P., Adams, T. and Polunin, N. (1996) Coastal fisheries in the Pacific Islands. Oceanography and marine biology 34: 395-531. 82 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller   Clark, G.R. (2005) A 3000- year culture sequence from Palau, Western Micronesia. Asian Perspectives 44(2): 349- 380. Colbert, D. (2000) City, Seas and Storms: Managing Change in Pacific Island Economies. Adapting to Climate Change 4. World Bank, Washington, D.C, ix+ 119 p. Dalzell, P., T.J., A. and Polunin, N.V.C. (1996) Coastal Fisheries in the Pacific Islands. Oceanography and Marine Biology 34: 395-531. Friedmand, K., Kronen, M., Pinca, S., Lasi, F., Pakoa, K., Awira, R., Bonblin, P., Tardy, E., Chapman, L. and Margon, F. (2007) Palau Country Report: Profiles and Results from Surfey work at Ngarchelong, Ngatpang, Airai and Korror. . Pacific Regional Oceanic and Coastal Fisheries Development Programme. Secretariat of the Pacific Community, Nomeau, New Caledonia, xxxiv+ 411 p. Gillett, R. (2009) Fisheries in the economies of the Pacific Islands Countries and Territories. Asian Development Bank, Mandaluyong City, Philippines, 254 p. Gillett, R. (2010). Marine fishery resources of the Pacific Islands. FAO Fisheries and Aquaculture technical paper 537, Rome, 71 p. Gillett, R. (2011) Replacing Purse Seining with Pole-and-Line fishing in the central and Western Pacific: Some aspects of the Baitfish requirements. . Marine Policy 35(2): 148-154. Gillett, R. and Lightfoot, C. (2002) The Contribution of Fisheries to the economies of the Pacific Island Countries. Asian Development Bank, Manilla, Philippines. Johannes, R.E. (1981) Words of the Lagoon: Fishing and Marine Lore in the Palau District of Micronesia. University of California Press, Berkeley, California, 245 p. Kelleher, K. (2005) Discards in the World's Marine Fisheries: An Update. FAO Fisheries Technical Paper. Food and Agriculture Organization of the United Nations, Rome, 131 p. Kitalong, A. and Dalzell, P. (1994) A Preliminary Assessment of the Status of Inshore Coral Reef Fish Stocks in Palau. South Pacific Commission, Noumea, New Caledonia. Lawson, T. (1991) Status of Tuna Fisheries in the SPC Area During 1990, with annual catches since 1952. South Pacific Commission, Noumea, New Caledonia, 40 p. Mathews, E. and Oiterong, E. (1991) The Role of Women in the Fisheries of Palau. Master of Science, University of Oregon, Corvallis, Oregon. Nichols, P. (1991) Republic of Palau Marine Resource Profiles. Fisheries Development Section. Forum Fisheries Agency, Noumea, New Caledonia, iv+ 112 p. Ota, Y. (2006) Custom and Fishing: Cultural meaning and Social Relations of Pacific Fishing, Republic of Palau, Micronesia. PhD, University College of London, England, London. Pakoa, K., Lasi, F., Tardy, E. and Friedmand, K. (2009) The status of Sea Cucumbers exploited by Palau's Subsistence Fishery. Secretariat of the Pacific Community, Noumea, New Caledonia, vi+23 p. Perron, F., Nauro, A. and Patris, S. (1983) The Palau reef fish production study: A baseline study of the commercial reef fishing industry in Palau and a blueprint for the development of permanent fisheries managment system. . Division of Marine Resources, Korror, Palau. Preston, G. (1990) Inshore fishery resource management in Palau. South Pacific Commission, Noumea, New Caledonia. Yamashita, S. (2000) The Japanese Encounter with the South: Japanese Tourists in Palau. . The Contemporary Pacific 12(2): 26. Zeller, D., Booth, S. and Pauly, D. (2006) Fisheries Contributions to the Gross Domestic Product: Understanding small-scale fisheries in the pacific. Marine Resource Economics 21(4): 39. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  83    Appendix Table A1: FAO landings vs. total reconstructed catch for Palau, 1950-2008, in metric tonnes. Year FAO landings Total reconstructed catch 1950 300 863 1951 300 903 1952 300 942 1953 300 982 1954 400 1,021 1955 400 1,061 1956 400 1,100 1957 400 1,140 1958 500 1,167 1959 500 1,195 1960 500 1,222 1961 500 1,250 1962 600 1,277 1963 600 1,304 1964 1,866 2,542 1965 3,370 4,130 1966 3,386 4,174 1967 4,106 4,915 1968 5,756 6,621 1969 5,462 6,309 1970 8,882 9,811 1971 2,943 3,703 1972 2,319 3,118 1973 3,150 3,971 1974 7,608 8,588 1975 7,069 8,026 1976 6,136 7,051 1977 4,883 5,705 1978 10,602 11,647 1979 6,699 7,507 1980 7,516 8,423 1981 10,276 11,396 1982 5,053 5,874 1983 1,041 1,796 1984 1,037 1,808 1985 1,100 1,926 1986 1,100 2,002 1987 1,100 2,083 1988 1,100 2,132 1989 1,090 2,065 1990 1,076 2,201 1991 1,093 2,170 1992 1,271 2,447 1993 1,211 2,388 1994 1,086 2,436 1995 1,027 2,421 1996 1,000 2,512 1997 913 2,585 1998 952 2,640 1999 962 2,672 2000 1,097 2,807 2001 1,086 2,727 2002 1,030 2,714 2003 1,050 2,740 2004 1,081 2,685 2005 935 2,517 2006 969 2,473 2007 986 2,437 2008 1,008 2,492   AppendixTableA2: Total reconstructed catch (t) for Palau by major taxa.Others category represents 58 taxonomic groups including miscellaneous marine fishes. Year Katsuwonus pelamis Lethrinidae Hipposcarus longiceps Lutjanus gibbus Thunnus albacares Siganus lineatus Siganus spp. Others 1950 - 53 63 46 - 57 43 601 1951 - 55 66 48 - 60 45 629 1952 - 58 68 51 - 62 47 657 1953 - 60 71 53 - 65 49 684 1954 - 62 74 55 - 67 50 712 1955 - 65 77 57 - 70 52 739 1956 - 67 80 59 - 73 54 767 1957 - 70 83 61 - 75 56 794 1958 - 71 85 63 - 77 58 814 1959 - 73 87 64 - 79 59 833 1960 - 75 89 66 - 81 60 852 1961 - 76 91 67 - 83 62 871 1962 - 78 93 69 - 84 63 890 1963 - 80 95 70 - 86 64 909 1964 1,025 81 97 72 141 88 66 972 1965 2,497 83 99 73 173 90 67 1,048 1966 2,615 85 101 74 71 92 69 1,068 1967 3,354 84 100 74 52 91 68 1,090 1968 5,039 84 100 74 17 91 68 1,148 1969 4,629 84 99 73 133 90 68 1,133 1970 8,081 87 103 76 1 94 70 1,297 1971 2,133 90 108 79 10 98 73 1,112 1972 1,463 94 112 83 56 102 76 1,132 1973 2,309 94 111 82 41 101 76 1,157 1974 6,647 93 111 82 161 101 75 1,319 1975 5,971 93 111 82 298 101 75 1,297 1976 4,911 93 110 82 412 100 75 1,268 1977 3,592 93 110 81 420 100 75 1,234 1978 9,391 92 110 81 303 100 75 1,495 1979 5,687 92 110 81 1 100 75 1,361 1980 5,580 96 114 84 996 104 77 1,372 1981 6,931 99 117 87 2,480 107 80 1,495 1982 3,438 102 121 90 615 110 82 1,316 1983 75 105 125 92 - 114 85 1,199 1984 32 126 148 151 - 106 80 1,164 1985 82 130 152 156 15 110 82 1,199 1986 112 133 156 159 19 112 84 1,227 1987 139 136 160 164 22 115 86 1,260 1988 119 140 165 168 38 118 89 1,295 1989 72 141 166 169 5 119 89 1,303 1990 80 161 159 162 8 114 86 1,428 1991 - 168 163 169 - 119 89 1,460 1992 61 177 167 175 62 123 92 1,588 1993 - 189 171 182 39 127 95 1,583 1994 - 206 178 192 31 134 100 1,592 1995 - 220 178 193 3 135 101 1,594 1996 - 234 184 203 2 142 106 1,640 1997 - 253 195 217 1 151 113 1,654 1998 - 278 195 221 1 154 115 1,677 1999 - 283 196 225 1 156 117 1,694 2000 - 290 193 225 63 156 116 1,762 2001 - 296 187 220 41 152 114 1,721 2002 - 305 187 223 3 154 115 1,728 2003 - 317 186 225 19 155 116 1,733 2004 - 313 177 217 28 149 112 1,689 2005 - 317 172 214 0 147 110 1,557 2006 - 321 167 211 - 144 108 1,521 2007 - 325 163 209 - 143 107 1,491 2008 - 336 166 212 - 145 109 1,525  Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  85   RECONSTRUCTION OF SRI LANKA‘S FISHERIES CATCHES: 1950-20081  Devon O‘Mearaa, Sarah Harpera, Nishan Pererab and Dirk Zellera aSea Around Us Project, Fisheries Centre, University of British Columbia, 2202 Main Mall, Vancouver, BC, V6T 1Z4, Canada bLinnaeus University, SE 39182 Kalmar, Sweden d.omeara@fisheries.ubc.ca; s.harper@fisheries.ubc.ca; nishan.perera@lnu.se; d.zeller@fisheries.ubc.ca  ABSTRACT  Sri Lanka has a long history of reliance on the sea for the nutritional and economic well-being of its people. Fishing has long been an important industry and, while detailed fishing records exist dating back to the early 1900s, they are incomplete. In this study, we estimated total marine fisheries catches for the 1950-2008 time period by accounting for all fisheries sub-sectors and components and compared this to the reported landings as provided to FAO. Our total reconstructed catch which included commercial and subsistence catches, and discarded bycatch was estimated at almost 18 million tonnes over the 1950-2008 time period. This estimate was over 2 times larger than the total landings reported by Sri Lanka to the FAO. The majority of this discrepancy was due to catches from the subsistence sector and discarded bycatch associated with shrimp trawl fisheries. Improved monitoring of, and record-keeping for, these fisheries components is crucial to the longterm management of Sri Lanka‘s fisheries and to maintaining livelihoods and food security of the Sri Lankan people.   INTRODUCTION  The Democratic Socialist Republic of Sri Lanka is an island country southeast of India within the Bay of Bengal (Figure 1). The climate is tropical with seasonal monsoon and cyclones, but no upwelling. In 2009, the population was 20 million (Anon, 2009) with 32 percent living in coastal areas (UNEP, 2001). The Sri Lankan Exclusivity Economic Zone (EEZ) lies within FAO statistical area 57 (FAO, 2011).  The island was colonized by the Portuguese and the Dutch, but most influentially by the British. Sri Lanka, or ―Ceylon‖ as it was known prior to 1972, was a strategic military and trade link between West Asia and Southeast Asia. It acquired independence from the British Empire as the Dominion of Ceylon in 1948, just after World War II. In 1972, Ceylon became a republic and the name was changed back to the pre-colonial name: Sri Lanka (De Silva, 1981).  Attempts to record fisheries data in Sri Lanka may have begun during British rule; however, a rigorous island-wide attempt to estimate total landings did not start until after independence. Since 1910, general fisheries information was recorded by the resident marine biologist as part of an annual fisheries administration report. These reports included descriptions of traditional fisheries, destructive practices, fisheries regulations, results of test fisheries, policy changes, and financial record keeping; yet, information regarding landings on the island was incomplete (Pearson, 1911; 1922). By the 1930s, the importance of quantifying total landings was recognized, and by the 1940s, efforts to quantify landings were well underway with the appointment of 12 fisheries inspectors (FIs) within 20 fisheries districts. In the early 1950s, the number of FIs was increased to 24. The first comprehensive annual report of total landings was published in 1952 by the Department of Fisheries (DOF); the reports were, from then on, published annually (reviewed in Sivasubramaniam, 1997).  Records of landings in the 1950s focused mainly on the traditional practice of beach seining as it accounted for approximately 40% of total landings (Canagaratnam and Medcof, 1956). The use of the large beach seine, madella, began in the mid to late 1800s and continued to be the most commonly used traditional fishing techniques throughout the twentieth century (Alexander, 1977). Gillnetting began in the 195os, and eventually took over as the most widespread fishing method for small-scale fishers. Incidents of  1  Cite as: O‘Meara, D., Harper, S., Perera, N. and Zeller, D. (2011) Reconstruction of Sri Lanka‘s fisheries catches: 1950-2008. pp. 85- 96. In: Harper, S. and Zeller, D. (eds.) Fisheries catch reconstructions: Islands, Part II. Fisheries Centre Research Reports 19(4). Fisheries Centre, University of British Columbia [ISSN 1198-6727].  illegal dynamite fishing and fish poisoning were also reported. The DOF showed great interest at this time in test fisheries, with special attention to experimental dredging for pearl and windowpane oysters, as well as trawler surveys (Sivalingam, 1961).  Artisanal and traditional fisheries in the 1950s could not meet the island‘s domestic demand for marine fish. Thus, markets were supplemented with cheap imports of predominantly dried fish products from Pakistan, Japan, and India. Sri Lanka was not a large exporter of marine fish with the exception of a small market in Thailand and Singapore for shark fins, sea cucumbers and ornamental shells called ‗chanks‘ (Turbinella pyrum). Domestic marine fish production and export capacity were limited by poor infrastructure, most importantly the lack of ice and salt at landing sites, and inefficiencies attributed to the traditional nature of the fishery. In an attempt to improve upon traditional methods, the DOF imported nylon nets and implemented the development of a craft motorization program (Canagaratnam and Medcof, 1956). Subsidies for 11,000 outboard motors and the introduction of 17-23 foot fibre- reinforced plastic (FRP) boats were credited for the subsequent high annual growth rate of Sri Lanka‘s fisheries that lasted until the beginning of civil war in 1983 (RAPA, 1989). In the last few decades, there has been an effort to augment pelagic fisheries though government assistance to increase the number of multiday vessels capable of fishing offshore and in international waters.  With the aid of the FAO, statistical methods again improved in the 1970s with the removal of the position of statistical officer and the appointment of an additional 143 FIs, while a new sampling system was also adopted that utilized landing centers as primary sampling units, and boats as secondary sampling units. In 1981, the National Aquatic Resources Research and Development Agency (NARA) was established with the mandate to improve research and development, with an emphasis to better understand tuna biology and catch statistics by way of a collaborative effort with the Indo-Pacific Tuna Programme (IPTP), the Bay of Bengal Programme (BOBP), the Food and Agriculture Organization (FAO) and the Asian Development Bank (ADB) (Dayaratne and Maldeniya, 1996). Gillnetting, a practice that had become popular in the 1960s, continued as a favorite of Sri Lankan fishers and by the 1970s, was accountable for 60% of reported fisheries catches.  Shortly after the establishment of NARA, civil war broke out between the Liberation Tigers of Tamil Eelam (LITE) and the Government of Sri Lanka (GoSL). The effect of the war on fisheries was considerable, especially in the north where restrictions (e.g., a ban on outboard motors greater than 40 hp, Maldeniya, 1997b) on fishers were put in place to prevent fuel and weapons from being illegally brought from India by the LITE. Additionally, the conflict led to the destruction of boats, gear, and infrastructure which included ice making facilities and highways important for fish transport to distant markets (Silucaithsam and Stokke, 2006). The northern fishing grounds, once responsible for producing over 40% of the country‘s reported landings, were the most productive and accessible fishing grounds in Sri Lanka due to the presence of a large continental shelf and a trawlable bottom (Engvall et al., 1977).  The 1990s saw an increase in reported landings due to improvements in the security situation in some areas of the north and the expansion of the fishing fleet offshore and internationally. By the 1990s, government officials recognized coastal resources were fully exploited, and efforts were shifted to expanding the potential of deep sea fisheries by providing boat and equipment subsidies (Mallikage, Figure 1. Map of Sri Lanka and ts Exclusive Economic Zone. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  87   2001). For billfish, this was attributed to improvements in gear and the expansion of fisheries into offshore and deep sea areas (Maldeniya et al., 1996).  Methods for improvement of catch statistics have been made in the 2000s, but overall, they remained the same since the changes made in 1981. The demand for marine fish has remained high, with a catch that was insufficient to meet demand. Despite the increase in multiday fishing vessels and other larger craft a large component of the marine fishing fleet continues to consist of small FRP boats with outboard motors as well as non-motorized traditional craft (FAO, 2006). The tsunami in December 2004 seriously affected 90% of the fishing community through losses of boats, fishing nets, housing, and lives. Eighty percent of fishing villages were completely destroyed, along with 12-14 fishing harbors (ITDG, 2005). Post-tsunami efforts to rebuild fisheries have resulted in an overabundance of fishing boats in some areas raising concerns for overfishing (Jayasuriya et al., 2005).  With the end of the civil war in 2010, efforts to increase fisheries production in the north were a high priority for the DOF. Growing domestic demand for seafood and the potential for substantial earnings from seafood exports appear to be the driving force behind current fisheries policy, with plans to double marine fisheries production in the future. Apart from increasing landings, offshore fisheries have been identified as a more viable source of high value export oriented species such as tuna. The lack of adequate offshore fishing capacity has been seen as a major obstacle to fisheries expansion, and there have been initiatives to allow commercial fishing by foreign vessels in exchange for access fees and prescribed landings in order to increase domestic fish supply (Anon., 2010).  Small-scale subsistence fisheries are often not considered when collecting fisheries statistics; however, they can constitute a large portion of actual catches (Zeller et al., 2007). The goal of this study was to more accurately quantify total marine fisheries catches, by taking into account all fisheries sub-sectors and components, including subsistence catch and discarded bycatch. The importance of fisheries to the livelihoods of Sri Lankan‘s, particularly coastal dwellers, requires a more comprehensive estimate and accounting of the true magnitude of fisheries extractions.   METHODS  Total marine fisheries catches were estimated using information obtained from national reports, independent studies, local experts and grey literature. Landings data presented by the FAO on behalf of Sri Lanka were compared to national landings data, and household surveys were used to estimate total demand for domestic seafood as compared to local supply. We also estimated discarded bycatch for the shrimp trawl and tuna longline fisheries. In this report we refer to ‘landings‘ as the amount of fish caught, brought to shore and recorded, while ‘catch‘ refers to the total amount of fish caught, and includes Illegal, Unreported and Unregulated (IUU) catches and discarded bycatch.  Population  Human population data were obtained for the 1950-1959 period from Populstat (www.populstat.info) and for the 1960-2008 time period from the World Bank (Anon, 2009). Population estimates were used to derive per capita marine supply and subsistence catch rates. The population of Sri Lanka has increased steadily from 7 million in 1950 to over 20 million in 2008 (Figure 2).   0 5 10 15 20 25 1950 1960 1970 1980 1990 2000 P o p u la ti o n  ( x 1 0 6 ) Year Figure 2: Human population trend for Sri Lanka. Data source: www.populstat.info and World Bank (Anon, 2009).  Commercial fisheries  Total commercial landings for Sri Lanka were available in nationally published reports as well as by the FAO; however, the national data contained a statistical error causing landings to be high for years prior to 1970 (Pathirana, 1972); landings reported to FAO and obtained from FAO FishStat were lower than nationally reported landings prior to 1970 (Figure 3). Therefore, it was assumed that the statistical error in the national data was accounted for and corrected in landings presented in FAO‘s FishStat. Landings presented by the FAO were also more complete from 1980-1990, where national landings data were sparse. FAO data for crustaceans were compared to prawn and lobster landings presented by NARA. For the 1994-2002 time period, prawn and lobster landings were used in place of the FAO‘s ‗miscellaneous marine crustaceans‘ grouping, as they were deemed to be a better representation of total crustacean catches (Figure 5). Marine crab fisheries, although known to occur in Sri Lanka, were assumed to be contained within a new, but smaller miscellaneous crustaceans category as no data was available to determine catch. With the exception of the amendment to crustaceans landings, the remainder of the FAO data was considered a good representation of commercial fisheries landings, both for the artisanal and industrial sectors. These landings were used as a baseline, to which we added components not accounted for in the officially reported data. Noteworthy are two non- fishery related events which are correlated with a noticeable decrease in landings over the time period considered; the beginning of the civil war in 1983 and the tsunami which occurred on December 26th, 2004.  Discards  Shrimp trawl fisheries are typically associated with considerable bycatch, which can either be landed or discarded at sea. A study in the late 1970s estimated bycatch associated with the shrimp fishery in two of Sri Lanka‘s main shrimp trawling grounds, Jaffna and Mannar (Subasinghe, 1981). The study provided estimates for both the landed and discarded components of the bycatch. Subasinghe (1981) presents discard rates for both areas, which gave an average rate of 10.2 kg of discards per kg of shrimp landed for 1979 (Discard rates for Mannar and Jaffina were 8.92 and 11.48 kg discarded per kg landed, respectively). These two regions were responsible for 60% of the commercial production of shrimp that year (Subasinghe, 1981; Saila, 1983). Therefore, we assumed that this discard rate was representative of Sri Lanka‘s shrimp trawl fisheries and applied the rate of 10.2 kg discards per kg of shrimp landed across the entire time period. Discards may have been even higher in earlier time periods due to greater benthic biomass and/or less storage capacity on vessels for non-target species; however, to remain conservative we held the discard rate constant back in time to 1950. For the recent time period, we carried the 1979 discard rate forward, unaltered, to 2008. This same study reported that over 80% of the discarded catch was silverbellies (Leiognathidae); we considered the remainder to be miscellaneous small pelagic fishes and miscellaneous sharks.  Depending on the type of gear used, bycatch is also of concern for tuna fisheries. The majority of tuna catches in Sri Lanka are skipjack tuna (Katsuwonus pelamis), representing roughly 60% of tuna catches and yellowfin tuna (Thunnus albacares), representing approximately 20% of the tuna catches. Tuna are predominantly caught using gillnets, although, longlines are becoming increasingly popular for catches Table 1. Estimated seafood consumption rates derived from the Department of Census and Statistics 2007 Household Income and Expenditure Survey. Year Per capita demand  (kg∙person-1∙year-1) 1981 19.39 1986 18.24 1991 14.64 2002 19.86 2005 24.12 2007 24.12 0 2 4 6 8 10 12 14 1950 1960 1970 C a tc h  ( t x  1 0 4 ) Year FAO National Figure 3: Comparison of landings data as presented by FAO and the national data source, indicating the statistical error in the national data, and its correction in data presented by FAO on behalf of Sri Lanka. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  89   aimed at the export market (Maldeniya, 1997b). Kelleher (2005) estimates discards by tuna longline in Sri Lankan waters to be 0.05%. Given that this was a very low discard rate, and given that we were unable to determine the portion of the tuna catch taken by longline, we did not estimate this component of the bycatch. As for bycatch associated with the tuna gillnet fisheries, information was also quite sparse. Due to the size of the nets used, incidental catch in the tuna gillnet fishery is mainly seerfish, billfish and shark. Given that these are marketable species, we assumed that the majority of the non-targeted catch for the tuna gillnet fishery was retained and that this portion of the catch was accounted for in the landings data.  Subsistence fisheries  We assumed that the subsistence component of small-scale fisheries was unaccounted for in the reported data. To estimate this component of the total catch, we calculated the island-wide marine seafood demand using per capita consumption data from the 2007 Department of Census and Statistics Household Income and Expenditure Survey (Anon, 2007); and compared this to the reported (commercial) landings presented by the FAO. We considered the difference between the supply of marine products for human consumption and the demand for seafood to be the subsistence catch.  The supply of marine products available for consumption by the local population was estimated as the commercial landings (FAO data) adjusted for imports and exports (W. Swartz, unpublished data, UBC Fisheries Centre). These adjusted landings were then converted to per capita supply rates using human population data.  To estimate marine demand, the per capita marine fish consumption was obtained from the 2007 Sri Lanka Department of Census and Statistics Household Income and Expenditure Survey (HIES). A detailed breakdown of per capita consumption of marine products was available for 2007 only. The per capita consumption of fish, which included aquaculture and freshwater products, was summarized in the 2007 survey for the years: 1981, 1986, 1991, 2002, and 2005. In order to remove aquaculture and freshwater consumption and calculate marine consumption, we assumed that the ratio of freshwater and aquaculture consumption to marine fish consumption remained the same over the entire survey period. This assumption resulted in a conservative estimate of per capita marine consumption as aquaculture and freshwater fish consumption have likely increased since the 1980s. However, in order to remain conservative, the amounts removed were assumed to be proportional to those in 2007. Conversion factors provided by the FAO for Indonesia (FAO, 2000) were used to convert product weight from the 2007 HIES into live weight. The resulting per capita seafood consumption rates for 1981, 1986 and 2007 were used as anchor points to derive a complete time series of consumption rates for the 1950-2008 study period (Table 1). We did not use the 1991 and 2002 estimates of per capita consumption since these points exactly matched FAO reported landings when they were multiplied by the human population. These points were likely estimates of per caput consumption (reported landings divided by the population) and hence left out 0 5 10 15 20 25 30 35 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 4 ) Year Artisanal Industrial Figure 4: Total commercial fisheries catches for Sri Lanka, separated by industrial and artisanal fisheries, 1950-2008.  Figure 5: Reported landings of marine crustaceans (light colour) and the additional estimated catches during the 1990s and early 2000s. 0 2 4 6 8 10 12 14 16 18 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 )3 Year Marine Crustaceans Add-on  of the analysis. We assumed that the consumption rate in 1950 was the same as that in the 1980s, and therefore carried the 1981 rate of 19.39 kg∙person-1∙year-1 back, unaltered to 1950. The 2007 estimate was carried forward to 2008. Years between anchor points were interpolated linearly. Finally, we subtracted the per capita marine supply (FAO landings adjusted for imports and exports) from the total per capita seafood demand to determine the per capita subsistence catch rate. Human population data were then used to convert per capita subsistence catch rates into total subsistence catch amounts. This calculation was not done for 2005 since, although the 2005 consumption estimate was thought to be reasonable, the reported landings were low due to the tsunami, which was likely the result of both fewer catches and poor reporting. The subsistence catch rate for the year following the tsunami (2005) was estimated by linear interpolation between the 2004 and 2006 subsistence catch rates and then was reduced by the same percent decline in catch (42%) as reported by the FAO for landings between the years 2004 and 2005. It is possible that subsistence was underestimated for anchor points following the beginning of civil conflict in 1983 as it is unlikely surveys included regions at war. The 2007 HIES states that Trincomalee and the Northern Province, known for high marine productivity and possibly higher per capita consumption, were not sampled in 2007 due to active conflict in these areas; consequently, it is likely the per capita consumption and hence the subsistence catch estimates are conservative.  Subsistence catches were assumed to be composed of small pelagic species (50%), demersal species (40%) and invertebrates such as crabs and cephalopods (10%). The small pelagic species caught were mainly clupeids and scombrids, with the most common species being Sardinella gibbosa, S. albella, Amblygaster sirm, A. clupeoides, Rastrelliger kanagurta, and Auxis thazard. Demersal species catches were mainly represented by Lethrinidae, Carangidae, Myliobatidae, Sciaenidae, Haemulidae, Leognathidae, and Acanthuridae (Canagaratnam and Medcof, 1956; Maldeniya, 1997a; MFAR, 2008). Industrial and artisanal catches were also improved for FAO ―crustaceans nei‖ utilizing assumptions based on Jayawardane et al. (2003). The species breakdowns for lobster and sea cucumbers were also improved based on local expert opinion (N. Perera, pers. obs., Linnaeus University)  Other IUU components  While catches of sea cucumbers and sharks are reported in the official landings data, they are likely under- estimates. Unreported catches of sea cucumbers and sharks are common in Sri Lankan waters; however, data on these were not readily available. Although we were unable to account for this unreported component as part of the reconstructed catch, it should be noted that IUU fishing is known to occur in Sri Lanka and should be further investigated (P. Ganapathiraju, pers. comm., UBC Fisheries Centre).  RESULTS  Commercial fisheries  Total marine fisheries catches by the commercial sector (artisanal and industrial) were estimated to be 8.4 million tonnes over the 1950-2007 time period (Figure 4). Catches in 1950 were approximately 20,000 t∙year-1 and increased steadily to over 300,000 t∙year-1 in 2004. This was followed by a substantial decrease in catches to around 15,000 tonnes in 2005, the year after the tsunami devastated Sri Lanka. Total commercial catches were composed of small- (artisanal) and large-scale (industrial) sectors, which represented 55% and 44%, respectively of the total commercial catch. The total commercial catch included over 50,000 tonnes of additional crustaceans, which were not represented in the reported landings as presented by FAO (Figure 5). Catches of marine crustaceans were estimated to be 320,000 tonnes for the period 1950-2008. These were mainly shrimp (75%) and lobster (9%), with the remainder being 0 20 40 60 80 100 120 140 160 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Year Figure 6: Reconstructed total subsistence catches for Sri Lanka, 1950-2008. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  91   miscellaneous marine crustaceans. Discards associated with the shrimp trawl fishery were estimated over the study period to be approximately 2.4 million tonnes (Figure 8).  Subsistence Fisheries  Total catches by the subsistence sector were estimated to be over 7 million tonnes from 1950-2008 (Figure 6). Subsistence catches remained relatively stable over the entire study period with an average annual catch of around 120,000 t·year-1 (Figure 8). A decrease in subsistence catches was observed for the late 1970s and early 1980s, but they increased again after that.  Total reconstructed catch  The total reconstructed catch of marine fisheries in Sri Lanka was estimated to be almost 18 million tonnes over the 1950- 2008 time period (Figure 7). This estimate of total catches was 2.13 times larger than the landings officially reported by Sri Lanka to the FAO. Reported landings, as presented by the FAO on behalf of Sri Lanka were 8.4 million tonnes. The subsistence catch represented 40% and discards represented 13% of the total estimated catch (Figure 8). The remainder of the total catch was from the artisanal (26%) and industrial (21%) sub-sectors of commercial fisheries. The estimate for commercial catch was almost entirely based on reported landings, while the subsistence and discards were entirely unreported components. Major contributing taxa in the reconstructed catch included silverbellies (Leiognathidae), skipjack tuna (Katsuwonus pelamis), herrings, sardines, and anchovies (Clupeoids), jacks (Carangidae), and yellowfin tuna (Thunnus albacares; Figure 9).  DISCUSSION  Total marine fisheries catches for Sri Lanka were estimated to be approximately 18 million tonnes over the 1950-2008 time period. This estimate was over 2 times larger than the landings reported by Sri Lanka to the FAO, which was approximately 8.4 million tonnes. This large discrepancy indicates a clear need for improvements in the collection and reporting of fisheries statistics in Sri Lanka. Our investigation into the fisheries of Sri Lanka revealed that information on subsistence fisheries, discarded bycatch and other IUU components was quite limited, even though these fisheries components contributed substantially to overall marine fisheries 0 10 20 30 40 50 60 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 4 ) Year Others Carangidae Thunnus albacares Katsuwonus pelamis Leiognathidae Clupeoids Clupeidae Scomber spp. Figure 9: Total reconstructed catches with main taxa caught. All other taxa (88 total) were grouped into ‗Others‘ category. Figure 8: Total reconstructed catches for Sri Lanka by component or fisheries sector, 1950-2008. 0 10 20 30 40 50 60 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 4 ) Year Subsistence Artisanal Discards Industrial 0 100 200 300 400 500 600 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Year Total reconstructed catch Supplied to FAO Figure 7: Total reconstructed catches compared to the data submitted by Sri Lanka to FAO, 1950-2008.  catches. Subsistence fisheries catches were the largest unreported component of the catch, and represented 40% of the total reconstructed catch.  Discards from the shrimp trawl fishery were also a substantial contributor (13%) to the total catch, and was unaccounted for in the official data. The high rate of discarding in Sri Lankan shrimp trawl fisheries has been attributed to the limited cold storage facilities on multi-day boats. Economically important species are often stored while other less valuable species are discarded. However, the majority of the bycatch consists of low-valued species of Leiognathidae, which are typically discarded (Subasinghe, 1981). Tuna longline fisheries, on the other hand, have a much lower discard rate (0.05%) according to Kelleher (2005). This low rate of discarding is thought to be due to fishers targeting and landing multiple species of high economic importance, thus reducing the amount of discarded fish (Kelleher, 2005). Beach seining in the early period (1950s) was reported to have few discards, with the exception of jellyfish which were known to seasonally clog nets (Canagaratnam and Medcof, 1956).  The year following the tsunami, reported landings were significantly lower, even though seafood consumption remained constant. Although DOF offices were badly damaged by the tsunami and efforts were directed into emergency measures rather than accounting for landings, it is likely catches also decreased, especially in small-scale and subsistence fisheries as they sustained a large amount of damage. An assessment of the impacts of the tsunami on coastal fishers suggested that fishing pressure may have initially decreased in 2005, but then increased to pre-tsunami levels caused by excessive replacement gear and vessels donated to local fishers as part of the relief effort (De Silva and Yamao, 2007).  Illegal, unreported and unregulated fisheries are known to occur in Sri Lankan waters. For example, the transshipments of shark fins caught in Sri Lanka‘s EEZ occur regularly. Most Sri Lankan vessels lack adequate refrigeration capacity and will therefore trade sacks of shark fins to foreign vessels at sea, which allows them to empty their hold and continue fishing while at sea. Such catches are not included in any reporting mechanism. Sri Lankan vessels also participate in the poaching of sharks and sea cucumbers, which are caught illegally outside of Sri Lanka‘s EEZ in the poorly regulated waters of Somalia, Madagascar, and the Seychelles, and are then landed in Sri Lanka. These catches are reported as domestic landings, but this is not necessarily the case. A lack of enforcement and proper reporting has allowed these IUU fisheries to continue and possibly expand over time (P. Ganapathiraju, pers. comm., UBC Fisheries Centre). Additionally, foreign vessels, in particular Indian vessels, engage in illegal fishing within Sri Lanka‘s EEZ and these catches are not reported for Sri Lanka.  Previous attempts to estimate the potential sustainable yield in Sri Lankan waters suggested harvest rates of 250,000 t∙year-1, with around 80,000 t allocated to demersal species catches and 170,000 t for pelagic species (RAPA, 1989). Our reconstructed catches indicate that this level was likely surpassed as far back as 1974. In this study we highlighted the lack of proper accounting for total fisheries catches, which in the case of the subsistence sector accounted for almost half of the domestic marine food supply. Without a realistic estimate of what is being extracted, fisheries are likely to be mismanaged and possibly overexploited. Although human and financial resources may not be available to establish and maintain in depth monitoring programs, regular surveys conducted every few years have been found to be very effective in estimating subsistence and small-scale catch in other developing countries (Brouwer et al., 1997; Zeller et al., 2006).  ACKNOWLEDGMENTS  We thank the Sea Around Us Project, a collaboration between the University of British Columbia and the Pew Environment Group. The present work was funded by FAO through the Bay of Bengal Large Marine Ecosystem Project (www.boblme.org/). We thank Rudolf Hermes (FAO-BOBLME) for critical comments.  REFERENCES  Alexander, P. (1977) Sea Tenure in Southern Sri Lanka. Ethnology 16 (3): 231-251. Anon (2007) Household income and expenditure survey. Sri Lanka Ministry of Finance and Planning- Department of Census and Statistics, Colombo, 117 p. Anon (2009) World Bank: World Databank. http://data.worldbank.org/country/sri-lanka Accessed: Nov 4, 2010. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  93   Anon. (2010) Saudi Arabia to fish in Sri Lankan waters. ColomboPage. http://www.colombopage.com/archive_10C/Nov28_1290968294KA.php. Brouwer, S.L., Mann, B.Q., Lamberth, S.J., Sauer, W.H.H. and Erasmus, C. (1997) A survey of the South African shore- angling fishery. South African Journal of Marine Science: 165-175. Canagaratnam, P. and Medcof, J.C. (1956) Bulletin No. 4: Ceylon's Beach Seine Fishery The Fisheries Research Station, Ceylon. Dayaratne, P. and Maldeniya, R. (1996) Status report on the development of tuna research and data collection National Aquatic Resources Agency. Colombo, Sri Lanka. ftp://ftp.fao.org/fi/CDrom/IOTC_Proceedings(1999- 2002)/files/proceedings/miscellaneous/ec/1996/EC604-01.pdf. Accessed: Jan 10, 2010. 4 p. De Silva, D. and Yamao, M. (2007) Effects of the tsunami on fisheries and coastal livelihood: a case study of tsunami- ravaged southern Sri Lanka. Disasters 31 (4): 386-404. De Silva, K.M. (1981) The V.O.C in Sri Lanka: The Last Phase 1767-1796. Pp 609. In A History of Sri Lanka. C. Hurst & Company Ltd. Engvall, L.O., Ratcliffe, C. and Scheepens, T.J. (1977) Assessment of problems and needs in marine small-scale fisheries. Ministry of Fisheries. Development of small-scale fisheries in Southwest Asia. ftp://ftp.fao.org/docrep/fao/006/ad739e/ad739e08.pdf  [Accessed: August 24, 2010]. Colombo, p1-9 p. FAO (2000) Conversion factors - landed weight to live weight. FAO Fisheries Circular No.847, Revision no. 1, Rome, 176 p. FAO (2006) Fishery Country Profile- Sri Lanka. fao.org/fi/oldsite/FCP/en/LKA/profile.htm. Accessed: June 24, 2010. FAO (2011) FAO major fishing areas. FAO Fisheries and Aquaculture Department. http://www.fao.org/fishery/area/search/en Accessed: Jan 24, 2011. ITDG (2005) Rebuilding fisheries livelihoods in Sri Lanka post-tsunami. Intermediate Technology Development Group (ITDG). http://practicalaction.org/docs/region_south_asia/rebuilding-fisheries-livelihoods- (draft).pdf. Accessed: Feb 14, 2011. Jayasuriya, S., Steele, P., Weerakoon, D., Knight-John, M. and Arunatilake, N. (2005) Post-tsunami recovery: issues and challenges in Sri Lanka. The Institute of Policy Studies (IPS) of Sri Lanka and the Asian Development Bank Institute (ADBI) and the Asian Economics Centre, University of Melbourne. http://www.ips.lk/news/newsarchive/2005/01122005_p_tsun/tsunami_recovery.pdf. Accessed: Feb 14, 2011. Jayawardane, P.A.A.T., Mclusky, D.S. and Tytler, P. (2003) Population bynamics of Metapenaeus dobsoni from the western coastal waters of Sri Lanka. Fisheries Management and Ecology 10: 179-189. Kelleher, K. (2005) Discards in the world's marine fisheries. FAO Fisheries Technical Paper 31-32, 111 p. Maldeniya, R. (1997a) The coastal fisheries of Sri Lanka: resources, exploitation, and managment. In Silvestre, G. and Pauly, D., (eds.), Status and Management of Tropical Coastal Fisheries in Asia. Asian Development Bank, Makati CIty. Maldeniya, R. (1997b) Small boat tuna longline fishery north-west coast of Sri Lanka. National Aquatic Resources Agency (NARA), Colombo. ftp://ftp.fao.org/fi/CDrom/IOTC_Proceedings(1999- 2002)/files/proceedings/miscellaneous/ec/1996/EC602-04.pdf. Accessed: Jan 27, 2011. Maldeniya, R., Dayaratne, P. and Amarasooriya, P.D.K.D. (1996) An analysis of billfish landings in the pelagic fisheries in Sri Lanka. National Aquatic Resources Agency, Colombo. ftp://ftp.fao.org/fi/CDrom/IOTC_Proceedings(1999- 2002)/files/proceedings/miscellaneous/ec/1996/EC602-31.pdf. Accessed: Jan 10, 2010. Mallikage, M. (2001) The effect of different cooling system on quality of pelagic species. Department of Fisheries and Aquatic Resources, Colombo, 34 p. MFAR (2008) Major marine fish types by commercial group. Ministry of Fisheries and Aquatic Resources. http://www.fisheries.gov.lk/Data/Fish%20Types.pdf. Accessed: January 27, 2011. Pathirana, W. (1972) Administration report of the acting director of fisheries for 1969-1970. Ceylon Government Press, Colombo. 146 p. Pearson, J. (1911) Report of the marine biologist for 1910-11. Ceylon administration reports: part IV- education, science and art., Colombo. Pearson, J. (1922) Report of the government marine biologist for 1922. Ceylon administration reports, Colombo. RAPA (1989) Marine fishery production in the Asia-Pacific Region- Sri Lanka. Regional office for Asia and the Pacific (RAPA) Food and Agricultural Organization (FAO) of the United Nations, Bangkok. Saila, S.B. (1983) Importance and assessment of discards in commercial fisheries. FAO Fisheries Circular 765: 62. Silucaithsam, A.S. and Stokke, K. (2006) Fisheries under fire: Impacts of war and challenges of reconstruction and development in Jaffina fisheries, Sri Lanka. Norsk Geografisk Tidsskrift- Norwegian Journal fo Geography 60: 240-248.  Sivalingam, S. (1961) The 1958 Pearl Oyster Fishery, Gulf of Mannar. Ceylon Fisheries Research Bulletin No 11. 28 p. Sivasubramaniam, K. (1997) One hundred years of fisheries management in Sri Lanka: lessons for the future. The department of Fisheries and Aquatic Resources, Colombo. 156p. Subasinghe, S. (1981) Fish By-catch...Bonus From the Sea- Sri Lanka. 141-142. UNEP (2001) Sri Lanka state of the environment report: coastal resources. United Nations Environment Programme. http://www.rrcap.unep.org/pub/soe/srilankasoe.cfm. Accessed: Nov 3rd, 2010. Zeller, D., Booth, S., Craig, P. and Pauly, D. (2006) Reconstruction of coral reef fisheries catches in American Samoa, 1950-2002. Coral Reefs 25: 144-152. Zeller, D., Booth, S., Gerald, D. and Pauly, D. (2007) Re-estimation of small-scale fishery catches for U.S. flag- associated island areas in the western Pacific: the last 50 years. Fishery Bulletin 105: 266-267. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  95    Appendix Table 1. FAO reported landings vs. total reconstructed catch for Sri Lanka, 1950-2008, in metric tonnes. Year FAO landings (t) Total reconstructed catch (t) 1950 20,622 151,813 1951 24,103 156,210 1952 24,709 158,421 1953 25,016 160,496 1954 26,433 166,998 1955 27,265 166,716 1956 32,702 176,616 1957 29,638 172,670 1958 35,737 185,640 1959 40,434 189,190 1960 50,775 198,438 1961 59,717 206,182 1962 71,137 227,544 1963 71,256 222,926 1964 87,796 259,676 1965 68,836 222,888 1966 72,083 230,049 1967 78,225 238,398 1968 93,080 245,640 1969 91,936 255,295 1970 83,855 242,466 1971 69,074 224,345 1972 85,438 243,631 1973 93,972 247,627 1974 96,608 264,940 1975 99,110 273,005 1976 122,870 302,695 1977 126,000 295,302 1978 142,768 319,436 1979 150,934 310,821 1980 167,594 320,543 1981 179,398 348,908 1982 184,664 374,713 1983 183,005 352,116 1984 137,909 302,847 1985 158,065 298,884 1986 159,437 298,798 1987 164,998 307,197 1988 175,347 320,621 1989 183,773 331,133 1990 159,173 305,788 1991 193,989 349,006 1992 189,939 360,916 1993 219,447 390,223 1994 240,307 381,993 1995 222,170 399,668 1996 242,031 411,686 1997 248,790 419,199 1998 259,746 443,709 1999 288,301 449,153 2000 284,314 483,307 2001 265,749 428,117 2002 271,927 432,235 2003 302,082 524,880 2004 303,168 533,482 2005 160,142 251,821 2006 239,292 432,512 2007 270,176 468,803 2008 285,028 503,501    Appendix Table 2. Total reconstructed catch (t) by major taxa for Sri Lanka, 1950-2008. Clupeoids include herrings, sardines, and anchovies. Others includes 86 taxa Year Leiognathidae Clupeoids Katsuwonus pelamis Clupeidae Thunnus albacares Carangidae Scomber spp. Others 1950 10,457 8,000 771 12,715 774 2,543 0 116,553 1951 10,511 8,000 890 12,807 1,150 2,561 0 120,290 1952 11,125 9,000 807 12,887 903 2,577 0 121,122 1953 11,748 7,100 723 12,983 655 3,697 0 123,591 1954 14,638 9,500 720 13,087 606 4,717 0 123,730 1955 13,537 11,600 717 13,137 557 4,027 0 123,140 1956 16,390 18,300 981 13,180 720 3,436 0 123,609 1957 15,303 14,100 1,245 13,253 883 3,351 0 124,536 1958 19,852 14,900 1,410 13,294 970 4,859 0 130,356 1959 18,749 17,000 1,576 13,341 1,055 5,968 0 131,502 1960 19,200 20,400 2,063 13,151 1,347 7,030 0 135,247 1961 18,094 28,400 2,551 13,192 1,639 6,838 0 135,467 1962 24,895 31,500 3,960 13,217 2,493 8,443 0 143,036 1963 21,509 22,400 5,369 13,228 3,348 9,846 0 147,226 1964 35,647 36,300 5,227 13,230 3,222 10,146 0 155,905 1965 23,203 24,200 5,084 13,224 3,096 6,545 3,000 147,536 1966 26,023 24,300 5,830 13,212 3,515 6,342 3,000 150,828 1967 27,706 20,000 6,576 13,190 3,931 8,638 3,000 158,357 1968 22,596 27,500 7,448 13,156 4,416 8,631 3,000 161,893 1969 30,482 21,700 8,322 13,105 4,901 8,621 4,000 168,164 1970 27,612 22,300 6,554 13,034 3,841 9,007 3,600 160,118 1971 25,861 18,400 4,785 12,942 2,783 7,588 5,100 151,986 1972 28,621 20,100 8,250 12,830 4,266 7,566 6,200 161,997 1973 26,281 20,600 9,919 12,700 5,244 8,440 4,900 164,443 1974 37,502 24,900 8,792 12,552 4,610 11,610 4,300 164,974 1975 42,447 32,530 6,937 12,387 3,771 8,637 7,994 166,294 1976 47,751 38,541 12,392 12,207 6,908 10,076 11,018 174,819 1977 41,644 46,278 11,583 12,010 5,806 11,192 9,179 166,788 1978 48,182 54,412 12,933 11,795 5,915 12,717 7,747 173,482 1979 37,957 59,276 9,692 11,557 6,555 12,440 13,388 173,344 1980 34,769 69,061 14,117 11,296 7,304 12,307 13,888 171,689 1981 47,292 64,479 15,196 11,151 8,068 12,796 12,906 189,926 1982 63,048 66,764 14,172 10,935 8,682 10,579 11,302 200,533 1983 49,864 70,971 14,649 10,705 9,264 10,726 15,518 185,937 1984 48,481 52,153 12,348 10,465 6,694 12,594 12,773 160,112 1985 33,183 27,682 13,699 10,217 7,160 10,139 13,000 196,803 1986 33,794 28,471 13,697 9,962 7,416 10,319 13,000 195,139 1987 34,848 29,460 14,442 10,108 7,785 10,638 13,000 199,916 1988 36,061 30,608 15,004 10,255 8,089 11,003 13,000 209,601 1989 36,590 31,064 16,500 10,400 8,727 11,165 13,000 216,687 1990 35,154 27,958 19,495 10,543 9,929 9,831 10,500 192,878 1991 40,141 33,426 21,990 10,683 11,934 11,112 12,000 219,721 1992 50,504 35,097 25,786 10,820 14,185 11,112 13,557 213,413 1993 49,542 37,379 29,692 10,951 16,478 13,068 10,854 233,113 1994 26,277 38,870 35,755 11,075 21,045 10,215 16,450 238,756 1995 48,498 49,785 33,915 11,190 16,499 9,148 17,642 230,633 1996 42,941 48,221 41,000 11,296 21,308 8,347 17,700 238,573 1997 42,681 47,200 50,012 11,393 27,094 9,179 20,000 231,641 1998 50,824 50,800 50,124 11,481 26,122 10,796 20,900 243,562 1999 35,840 51,370 64,316 11,562 32,767 10,992 21,350 242,307 2000 59,392 53,250 70,957 11,636 29,512 12,777 22,180 245,783 2001 35,461 49,270 66,692 11,698 26,522 12,290 16,760 226,183 2002 33,768 52,310 64,425 11,786 28,085 13,117 17,250 228,743 2003 79,929 56,390 75,146 11,954 34,425 17,331 17,760 249,705 2004 83,900 54,410 75,795 12,146 35,512 16,009 18,440 255,711 2005 21,620 24,870 44,938 6,489 24,887 7,248 9,680 121,769 2006 55,219 56,230 54,341 12,433 35,842 12,057 15,570 206,391 2007 59,779 63,520 73,240 12,524 32,998 13,885 16,290 212,857 2008 73,544 66,890 78,860 12,618 33,027 13,684 18,260 224,878  Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  97    FROM LOCAL TO GLOBAL: A CATCH RECONSTRUCTION OF TAIWAN‘S FISHERIES FROM 1950-20071 Daniel Kuo and Shawn Booth Sea Around Us Project, Fisheries Centre University of British Columbia, 2202 Main Mall, Vancouver, BC, V6T 1Z4, Canada danieltykuo@gmail.com; s.booth@fisheries.ubc.ca  ABSTRACT  Taiwan‘s coastal fisheries were considered overfished in the 1950s, and, as a result, Taiwan expanded its fisheries into offshore areas and to distant-waters beginning in 1959. Taiwan‘s fisheries catches within its own EEZ increased until 1980, and total catches were estimated to grow from approximately 95,000 tonnes in 1950 to 575,000 tonnes in 1980. However, since 1980, total catches taken by Taiwan within its own EEZ have been in decline and were estimated to be only 290,000 tonnes in 2007. Although the catches taken within the EEZ have been in decline, the distant-water fisheries (DWF) have expanded into all major oceans. By the 2000s, they were 3 to 5 times larger than those from its own EEZ. Total catches of the distant-water fleet were estimated to increase from 110,000 tonnes in 1959 to 1.5 million tonnes in 2007. However, the continued expansion of the DWF is not likely sustainable due to the increased pressure on fisheries stocks globally, and because of the large amount of fuel subsidies the DWF receives.  INTRODUCTION  Taiwan (―Republic of China‖) is an island country located off the southeast coast of the People‘s Republic of China in the South China Sea (Figure 1). The west coast of Taiwan is separated from China by the relatively shallow Taiwan Strait and its east coast lies along the much deeper Philippines Sea. Taiwan has always relied on the sea for resources, and its fisheries are well developed and provide an essential source of food and income for its people.  The population of Taiwan grew from approximately 7.5 million people in 1950 to approximately 23 million people in 2007 (Anon., 2009). The increasing population, and its seafood demand, the decline of near-shore resources, and the high level of pollution of inshore areas, have resulted in Taiwan‘s fisheries expanding into offshore waters and eventually to distant waters. Taiwan‘s diplomatic isolation has often caused problems during negotiations with neighbouring countries resulting in territorial disputes. Consequently, Taiwan‘s exclusive economic zone (EEZ) is highly disputed and frequently challenged (Figure 1; Chen, 2007).  1  Cite as: Kuo, D. and Booth, S. (2011) From local to global: a catch reconstruction of Taiwan‘s fisheries from 1950-2007. pp. 97-106. In: Harper, S. and Zeller, D. (eds.) Fisheries catch reconstructions: Islands, Part II. Fisheries Centre Research Reports 19(4). Fisheries Centre, University of British Columbia [ISSN 1198-6727]. Figure 1. Map of Taiwan and the exclusive economic zone (EEZ) it claims.   It is widely suspected that Taiwan, like a number of other countries, tolerates extensive Illegal, Unreported and Unregulated (IUU) fisheries, and thus the annual catches reported to the Food and Agriculture Organization (FAO) are minimum estimates that may omit significant catches. This report aims to provide a more accurate estimate of Taiwan‘s annual fisheries catches from 1950–2007 by including estimates of unreported catches and discards. Some catches, which contribute to the local or global supply of seafood, are often unreported for tax avoidance purposes, and are usually not included in assessments or quota considerations, which undermine fisheries management policies. Discards increase economic benefits for fishers by increasing profits, and are a form of fishing mortality that may not be accounted for, and therefore may lead to detrimental effects on exploited fish populations.  EEZ fisheries  Taiwan‘s fisheries have traditionally been considered as two distinct fisheries—the coastal and the ―distant-water fisheries‖. The coastal fisheries operate up to the 12 nm territorial sea boundary, whereas the distant-water fisheries operated past the 12 nm demarcation of the coastal fisheries.  However, Taiwan‘s definition of ‗distant-water‘ changed with the expansion of the fisheries into all of the major oceans beginning in 1959, using gears such as trawl, longline, and purse seine. Thus, the distant-water fisheries were re-categorized as those operating in the far seas, and the fisheries around the waters of Taiwan were re-defined as ‗coastal‘ and ‗offshore‘. Throughout this report we refer to the coastal and offshore fisheries as those operating in what is today considered to be within the Taiwanese exclusive economic zone (EEZ), and reserve the term distant-water fleet (DWF) for those fisheries operating outside the EEZ.  As an island country, Taiwan has extensively developed its fisheries for food security, but it has also used its fisheries to gain political influences.  After World War II, Taiwan was faced with a growing population, food shortages, and a heavily damaged fishing industry (Huang and Chuang, 2010). During that time, Taiwan‘s inshore fishing grounds were already considered to be overfished, and in order to meet domestic demand, Taiwan began to expand its fisheries into distant waters because of the declining fisheries resources within its local waters (Chen, 2007).  Distant water fisheries  The growth of the post-war Taiwanese distant-water fishery after WWII was dictated by urgent economical and political needs. During the post-war period, there was a push towards an ―Every Fisher Has His Boat‖ policy. As part of this policy, the government offered subsidies to encourage the purchase of fishing boats and to stimulate the launching of more boats (Chang et al., 2010). The government also attracted investors for the DWF with the support of the United States Agency for International Development (USAID) and the World Bank during 1953 – 1973 (Chang et al., 2010). Beginning in 1961, investments were directed towards Taiwan‘s DWF to support the production of large-scale longliners and trawlers (Chang et al., 2010). At the same time, the construction of small trawlers was suspended in 1967 because of the drastic increase of fishing boats in the coastal region and depletion of coastal fisheries resources.  During the mid-1970s to early 1980s, the DWF was affected by two international oil crises, which increased the costs of operation and resulted in decreased landings in some years (Chang et al., 2010). Also, many coastal nations declared EEZs starting in the late 1970s or early 1980s, and Taiwan‘s offshore fleets were prevented access to former areas of operation (Chen, 2007). Consequently, Taiwan began to focus on further expansions of the DWF resulting in an increase in landings from this fleet, while at the same time there was a decrease in landings from the offshore area of the Taiwanese EEZ.  From the mid-1980s to 1990s, the Taiwanese government continued to push for the growth of the DWF (Chang et al., 2010). In 1983, a policy was implemented to suspend the building of all small-scale fishing vessels (< 1000 GRT) in order to develop the large-scale distant-water vessels (Chang et al., 2010), i.e., high seas capable vessels. Beginning in the 1980s, the Taiwanese DWF also explored new gear types such as the squid drift net and tuna purse seine. The main fishery gear types consisted of tuna long line, purse seine, squid jig, saury stick held dip net, trawl, and large-scale tuna drift net fisheries (Chang et al., 2010). The main species targeted during this time were yellowfin tuna (Thunnus albacares), bigeye tuna (Thunnus obesus), albacore (Thunnus alalunga), skipjack (Katsuwonus pelamis), and argentine squid Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  99   (Illex argentinus) (Chang et al., 2010). The massive deployment of drift nets, however, was short-lived as drift net fishing was banned by the UN in 1992 (Chen, 2007).  From the mid-2000s to the present, due to increasing problems of IUU fishing and vessels operating under flags of convenience, the Taiwanese longline fishery was sanctioned by the International Commission for the Conservation of Atlantic Tunas (ICCAT) in 2005. This was devastating to the Taiwanese distant-water fisheries, since longline vessels targeting bigeye tuna had their quota substantially reduced and Taiwan also lost its Cooperating Status in ICCAT. The sanctions imposed by ICCAT can, however, also be seen as an opportunity for Taiwan to reform its fisheries policies (Chang et al., 2010).  Discards  Discarding is a common practice within the fishing industry, and unwanted fish that have been caught are simply thrown back to sea. Discards can include damaged fish, non-target species, spoiled fish, and undamaged target species that are discarded to make room for more valuable ones (i.e., ‗high grading‘). Much of the global discards currently are influenced by economic incentives, especially in DWFs. Depending on the markets of the fishing country, catches may be fully retained or largely discarded. For Taiwan, only a rough estimate of discards is possible, as neither the FAO nor the Taiwanese Fisheries Agency (FA) documents discards.   MATERIALS AND METHODS  FAO reports, on behalf of Taiwan, annual landings statistics through their fisheries capture database (Fishstat) since 1950. Internally, Taiwan has two sets of data relating to fisheries. National landing statistics by area and gear are available from 1959 to 2007 through Taiwan‘s Fisheries Agency (FA; www.fa.gov.tw) and a separate set of statistics are available from the Taiwanese Council of Agriculture, Executive Yuan that describe the food balance for the country from 1984 to 2007 (www.ttdais.gov.tw/view.php?catid=20242). The food balance statistics include the human population, domestic production for fish and seafood, imports, exports, and domestic supply, and are expressed in live weight equivalents. The sub-categories under the domestic production for fish and seafood include ‗fish‘, ‗shrimps and crabs‘, cephalopods, ‗shell fish‘, ‗others‘, and ‗dried (salted) fish‘.  A comparison between the national landings and FAO landings was conducted to determine the quality of data transfer from the Fisheries Agency (FA) to FAO. Freshwater fishes, corals, marine mammals, and seaweeds were removed from FAO landing statistics for the purpose of this comparison. Similarly, we excluded the freshwater fisheries and all aquaculture statistics from the FA national landings. The data from FA were also compared to the food balance data for the period 1984 – 2007 when published statistics are available.  The domestic production from the food balance data was converted into estimates of annual national catches by the following formula:  Catch = Domestic Production + Export - Import - Marine Culture - Freshwater Culture - Inland Catch             ...1)  The marine culture, freshwater culture, and inland catch statistics were taken from the FA‘s national landing statistics.  The annual catch determined from equation (1) was compared to the annual FAO and FA total landings to check for inter-agency discrepancies.  Unreported catches  We assume that discrepancies found between the food balance data and that supplied to the FAO via the FA represent unreported catches. For the years 1954–1983, when there are no data available from the FA, we use the average rate of unreported catches (~33%) reported from 1951 to 1953 (Sung, 1972) to estimate unreported catches. We further assume that the unreported catches are proportional to the spatially reported landings. Unreported catch rates were applied to the reported landings from the coastal, offshore and distant water fisheries. The unreported catches for each FAO area are distributed in proportion to the reported taxa, except in FAO area 61 and 71. For FAO 61, the ―sharks, rays, skate, etc, nei‖ category were  distributed in proportion to the species reported in Lawson (1997). The unreported catch rates for FAO area 71 were applied in proportion to the species reported for the Taiwanese distant water longline catch (Lawson, 1997).  Discards  The discard portion of the Taiwanese catch reconstruction is based on estimates, since no official records of Taiwanese discards exist. Discard rates were applied to the sum of the reported and unreported landings. Longline discard rates were based on data presented in Kelleher (2005; Table 1). The average rates of longline discards calculated from this publication were applied according to their reported areas and year. The discard rate of purse seine, trawl, squid jig, and nets are rates provided for other countries operating in the same area (Kelleher, 2005). We used reported discards by gear and target species as a proportion of retained catches by gear and target species to derive discard rates. We applied these derived discard rates to the sum of the reported and unreported landings to estimate discard amounts.  Table 1. Summary table of discard rates used for the various oceans and gear types of the Taiwan fisheries fleets. Year Ocean Gear Discard rate Source 1950-1997 Pacific longline 0.113 Mejuto et al. (1997) 1998-2007 Pacific longline 0.124 Mejuto et al. (1997) and Kelleher (2005) 1950-1994 Atlantic longline 0.160 Cramer 1998 and  Mejuto et al. (1997) 1995-2004 Atlantic longline 0.400 Kelleher (2005) 2005-2007 Atlantic longline 0.130 Mejuto et al. (1997) 1950-1989 Indian longline 0.340 ½ the 1990-1999 rate 1990-1999 Indian longline 0.680 Mejuto et al. (1997) 2000-2004 Indian longline 0.460 Mejuto et al. (1997) 2005-2007 Indian longline 0.390 Mejuto et al. (1997) 1950-2007 all purse seine 0.040 Amande et al. (2008) and Anon. (2003) 1950-2007 all shrimp trawl 0.970 Clucas (1997) 1950-2007 all demersal trawl 0.123 Kelleher (2005) 1950-2007 all squid jig 0.500 Kelleher (2005) 1950-2007 all other nets 0.014 Kelleher (2005)  The discard rates thus derived were applied spatially to each major fishery based on gear type. The longline and purse seine discards were assigned to both the EEZ and distant water fisheries.  Discards associated with squid jigging in the Pacific and net discards (excluding purse seines), which are almost exclusively used in coastal and offshore areas, were assigned to EEZ fisheries only. The longline and purse seine fisheries within the EEZ had discard rates applied in proportion to the reported landings by the two gear types using the FA‘s national statistics. For the DWF, the longline and purse seine discards were also assigned proportionally, based on reported landings by the two gear types. Discards from the trawl fishery and the squid fisheries in oceans other than the Pacific were assigned to the DWF.  An additional 279 t∙year-1 of whale shark catches are applied from 1990 to 2000, and 126 t∙year-1 applied from 2001 to 2007 to the offshore catches due to the unregulated nature of the whale shark fisheries (Chen et al., 1997; Chen and Phipps, 2002). Whale sharks are usually sold outside of the market system to avoid fees, and only two markets have recorded landings statistics for whale sharks (Chen et al., 1997).   RESULTS  The total reconstructed catches amount to 54.7 million tonnes from 1950-2007, as 0 200 400 600 800 1000 1200 1400 1600 1800 2000 1950 1960 1970 1980 1990 2000 C a tc h  ( x  1 0 3 t) Year Reported Unreported Discards Figure 2. Total reconstructed catches including reported landings, unreported catches and discards. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  101   compared to the landings reported to FAO of approximately 38.2 million tonnes. The difference of 16.5 million tonnes is estimates of IUU in the form of unreported catches and discards. Unreported catches are the dominant form of IUU accounting for 84% of the IUU catches assessed here (Figure 2).  Total unreported catches  The FAO landings statistics and the FA national landings statistics are largely indistinguishable (Figure 3), indicating that the data transfer between FA and the FAO was efficient. However, discrepancies in landings were found between 1950 and 1953, and from 1984 onwards using alternative sources. Catches for Taiwan from 1950 – 1953 were, on average, about 28% higher than the national FA statistics (Sung, 1972) and the discrepancies increased from approximately 11% in 1950 to 38% in 1953. Discrepancies between the food balance data and the FA data for the later time period (1984-2007) averaged 37%, and also showed an increasing trend through time (Figure 4). Over the entire period (1950-2007), unreported catches estimated from the discrepancies between the food balance data and the FA amount to 13.6 million tonnes. Although the yearly discrepancies were relatively small during the first 4 years (1950 -1953 average: ~25,000 t∙year-1), after 1983, these differences increased over time and range from 225,000 t∙year-1 to 567,000 t∙year-1.  Total discards  Our estimated total discard amounts were fairly constant from 1950 to 1959 and averaged 12,000 t∙year-1. However, with the expansion of the DWFs, discards increased and peaked at 80,000 tonnes in 1988, representing 5% of the total reconstructed catch. Total discards declined to approximately 47,000 tonnes in 2007 and represent 3% of the total reconstructed catches. From 1959 onwards, the average discard rate applied to reported landings and unreported catches was 5.3%. We estimate that between 1959 and 2007, the highest rate of 8% occurred in 1972, but has since dropped to 3% in 2007.    0 10 20 30 40 50 60 70 1950 1960 1970 1980 1990 2000 D is cr ep a n cy  ( % ) Year Figure 4. Discrepancies (%) between the food balance data and reported landings. Dots are reported values and the solid line indicates an estimated value (mean of 1951-1953). 0 100 200 300 400 500 600 1950 1960 1970 1980 1990 2000 C a tc h  ( x  1 0 3 t) Year Offshore Offshore unreported Discards Coastal Coastal unreported Figure 5. Catches taken within the EEZ waters of Taiwan, including coastal (with catches considered here to be taken by the small-scale sector) and offshore fisheries, unreported catches and discards. 0 200 400 600 800 1000 1200 1400 1950 1960 1970 1980 1990 2000 L a n d in g s (x  1 0 3 t) Year FAO Taiwan Fisheries Agency Figure 3. National statistics as reported by FAO and by the Fisheries Agency of Taiwan.  EEZ fisheries  Fisheries in the EEZ of Taiwan, which consist of the coastal and offshore sectors, increased from the 1950s until the 1970s by approximately 100,000 t∙decade-1(Figure 5). Reported landings increased from 75,000 t∙year-1 in 1950 to 401,000 t∙year-1 in 1979. Annual catches within the EEZ peaked in 1980 at 407,000 t∙year-1, but have steadily declined and were only 190,000 t∙year-1 in 2007. The declines in landings from the EEZ waters have occurred in the offshore waters. Coastal fisheries, relying mostly on fixed nets, have been subject to overfishing since the 1950s (Chen, 2007) and apparently may also have been affected by the expansion of aquaculture. Coastal fisheries, roughly corresponding to small-scale fisheries, averaged reported landings of approximately 31,000 t∙year-1 between 1950 and 1979, and increased to average 47,000 t∙year-1 between 1980 and 2007 (Figure 5). Offshore reported landings increased from 55,000 tonnes in 1950, peaked in 1980 at about 370,000 tonnes, but have since declined to 135,000 tonnes in 2007 (Figure 5).  From 1950 to 2007, unreported catches within the EEZ were estimated to be over 5 million tonnes (Figure 5). Unreported catches increased from 8,400 tonnes in 1950, peaked at about 136,000 tonnes in 1980, and declined to 92,000 tonnes in 2007. Discards within the EEZ were estimated to be highest in the first decade due to the amount of trawling. Trawl landings declined from 89% to 33% of reported landings between 1950 and 1959, but due to the increased landings trawl discards were estimated to average 12,000 t∙year-1 between 1950 and 1958. With the expansion of fleets into distant waters beginning in 1959, we estimate that the discards within the EEZ waters declined, but those from DWF began to increase. Discards were estimated to average approximately 21,000 t∙year-1 from 1959-2007, but increased from 7,600 tonnes in 1959, peaked during the 1980s at 33,000 t∙year-1 before declining to 8,000 t∙year-1 in 2007 (Figure 5).  Distant water fisheries  Ine more recent times, distant-water fisheries represent the most significant component of Taiwanese fisheries. The DWF accounts for approximately 62% of the total reported Taiwanese landings from 1950 to 2007, and mainly targets tunas and squids using longline, purse seine and jig. Since 1985, the DWF have consistently been the largest, contributor to Taiwan‘s reported total landings. With decreased access to other neighbouring countries waters starting in the late-1970s due to the increasing declaration of EEZs, the DWF was rapidly developed in the 1980s, and reported landings from these fisheries during the 2000s are now 3 to 5 times larger than those from the EEZ in terms of annual reported landings (Figure 6).  Unreported catches from the DWF were estimated to contribute over 8.7 million tonnes to the total catches from 1959-2007 (Figure 6). Unreported catches increased from 26,000 tonnes in 1959 to 97,000 tonnes in 1971. Between 1972 and 1987, unreported catches were between 100,000 and 200,000 tonnes, and averaged 120,000 t∙year-1. Unreported catches have increased from 253,000 tonnes in 1988 to 477,000 tonnes in 2007. Discards from the DWF were estimated to have increased from 5,000 tonnes in 1959, peaked at 53,000 tonnes in 2002, but have since declined to approximately 39,000 tonnes in 2007.   DISCUSSION  From 1950-2007, the total reported landings were over 38 million tonnes. Reported landings increased from about 75,000 tonnes in 1950 to over 900,000 tonnes in 1987. Since 1988, annual reported landings have been approximately 1 million t∙year-1. Unreported catches from 1950-2007 were 13.7 million tonnes, 0 200 400 600 800 1000 1200 1400 1600 1950 1960 1970 1980 1990 2000 C a tc h  ( x  1 0 3 t) Year Reported Unreported Discards Figure 6. Total reconstructed catches for the DWF, including reported landings, unreported catches and discards. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  103   and increased from about 8,000 tonnes in 1950 to 568,000 tonnes in 2007. Discards rose from 10,000 tonnes in 1950 to 80,000 tonnes in 1988 and then began to decline, amounting to approximately 47,000 tonnes in 2007. As estimated here, our total reconstructed catches including reported landings, unreported catches and discards amount to 54.7 million tonnes and are 43% larger than reported landings. This estimation is probably conservative. While the unregulated portion of IUU catches in Taiwan may be small, the unreported illegal portion may be large, leaving out a potentially significant portion of IUU catches.  The unreported catches that we estimated, based on the food balance data, may represent unreported by- catch from the EEZ fisheries and DWF. With the exception of FAO fisheries statistical area 61, the only species reported from other fisheries statistical areas were tunas, marlins, squids, unidentified sharks, and unidentified marine fishes (―marine fishes nei‖). The ―marine fishes nei‖ category accounts for a small proportion of the overall catch statistics in FAO, and thus would be an underestimate because it doesn‘t account for all bycatch. It was not until 1991 that the government enforced an observer and vessel monitoring program to track fisheries catches (Chang et al., 2010). However, the fact that most of the reported DWF landings are high-value target species (e.g., tunas) makes it possible that the estimates of unreported catches reported here are mainly comprised of bycatch species.  The expansion of Taiwan‘s fishing fleet into offshore areas of the EEZ beginning in 1959 increased annual landings until many of Taiwan‘s neighbours declared EEZs, increasingly limiting the fishing areas accessible by Taiwan. The loss in catches from traditional fishing grounds caused a fairly rapid expansion of the DWF in the early years, but has since slowed. As estimated here, the total reconstructed DWF catches doubled from 113,000 tonnes to 251,000 tonnes in 8 years (1959-1966), doubled again from 251,000 tonnes to 500,000 tonnes in 7 years (1966-1972), but it took 17 years for DWF catches to double from 500,000 to 1 million tonnes (1972-1988). Total reconstructed DWF catches were estimated to be 1.5 million tonnes in 2007.  Taiwan‘s fisheries contribute to national food supply, and presently the DWF catches compensate for the declining catches from the Taiwanese EEZ. However, the DWFs are highly vulnerable to fluctuations in the energy sector in the form of rising fuel costs. Decreased catches were the result of the oil crisis in the 1970s and again in the 1980s. Fuel used by the DWF is subsidized by the Taiwanese government, and in the year 2000 the $US 0.09∙litre-1 subsidy amounted to $US 120 million (Sumaila et al., 2008). With increased pressure on fisheries stocks worldwide, this strategy is not likely to be sustainable, and the Taiwan DWF may thus experience difficulties expanding, or even maintaining their present catch volumes.   ACKNOWLEDGEMENTS  This is a contribution of the Sea Around Us Project, a scientific collaboration between the University of British Columbia and the Pew Environment Group.  REFERENCES  Amande, J.M., Ariz, J., Chassot, E., Chavance, P., Delgado, M.A. de, Gaertner, D., Murua, H., Pianet, R., and Ruiz, J. (2008). By-catch an d discards of the European purse seine tuna fishery in the Indian Ocean. Estimation and characteristics for the 2003-2007 period. IOTC-2008-WPEB-12. Ecosystem and by-catch wroking group. Bangkok, Thailand. 20-22 October 2008. 26, p. Anonymous (2003). Inter-American Tropical Tuna Commission Fishery Status Report No. 1. Tunas and billfishes in the eastern Pacific Ocean in 2002. 96, p. Anonymous (2009) Total Population, annual increase, and numbers and rates of natural increase, births and deaths. Department of Household Registration, M.O.I. Available at: www.ris.gov.tw/web_eng/eng_sta_hs.html [Accessed: August 2010]. Chang, S.-K., Liu, K.-Y. and Song, Y.-H. (2010) Distant water fisheries development and vessel monitoring system implementation in Taiwan--history and driving forces. Marine Policy 34: 541-548. Chen, C., Liu, K. and Joung, S. (1997) Preliminary Report on Taiwan's Whale Shark Fishery. TRAFFIC Bulletin 17: 53- 57. Chen, T. (2007) Taiwanese Offshore (Distant Water) Fisheries in Southeast Asia, 1936 - 1977. Doctor of Philosophy Dissertation, Murdoch University, Perth, Australia 368 p.  Chen, V. and Phipps, M. (2002) Management and Trade of Whale Sharks in Taiwan, TRAFFIC East Asia Report. Taipei, 1-25 p. Clucas, I. (1997) A study of the options for utilization of bycatch and discards from marine capture fisehries. FAO Fisheries Circular. No. 928. FAO, Rome, 59, p. Cramer, J. (1998) Pelagic longline billfish by-catch. Col. Vol. Sci. Pap. ICCAT, 47: 143-153. Huang, H.-W. and Chuang, C.-T. (2010) Fishing capacity management in Taiwan: experiences and prospects. Marine Policy 34: 70-76. Kelleher, K. (2005) Discards in the world's marine fisheries. An update. FAO Fisheries Technical Paper. No. 470. FAO, Rome, 131 p. Lawson, T. (1997) Estimation of Bycatch and Discards in Central and Western Pacific tuna Fisheries: Preliminary Results. South Pacific Commission, Nouméa, 12-30 p. Mejuto, García-Cortés, B., and Ramos-Cartelle, A.M. (2007) Preliminary approach to evaluate the importance of discards and other uses of billfish in the Spanish surface longline fishery carried out in differeent oceans betweeen 1993-2005. Col. Vol. Sci. Pap. ICCAT, 60(5): 1547-1554. Sumaila, U.R., Teh, L., Watson, R., and Pauly, D. (2008) Fuel price increase, subsidies, overcapacity, and resource sustainability. ICES Journal of Marine Science 65: 832-840. Sung, T.A. (1972) Taiwan Distant Waters Fishery [Partial photocopy of an internal report provided in confidence; in Chinese]. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  105   APPENDIX Appendix Table A1. Total Taiwanese catches (t) including coastal, offshore and DWF catches, and estimates of unreported catches and discards. Year Coastal Coastal unreported Offshore Offshore unreported DWF DWF unreported Discards 1950 20,885 2,313 54,297 6,082 n/a n/a 10,091 1951 25,291 7,293 58,146 16,768 n/a n/a 11,743 1952 29,696 9,991 61,365 20,645 n/a n/a 12,185 1953 30,250 11,372 65,194 24,509 n/a n/a 12,547 1954 30,804 10,276 69,023 23,025 11,672 3,894 12,085 1955 31,358 10,460 72,852 24,302 31,319 10,447 12,164 1956 31,911 10,645 76,681 25,579 42,582 14,205 12,256 1957 32,465 10,830 80,509 26,856 50,364 16,801 12,277 1958 33,019 11,015 84,338 28,134 65,177 21,742 12,578 1959 33,573 11,199 88,167 29,411 79,232 26,430 12,594 1960 31,761 10,595 91,170 30,413 88,763 29,610 14,811 1961 34,525 11,517 112,927 37,670 109,044 36,375 17,981 1962 35,214 11,747 128,176 42,757 120,505 40,198 19,403 1963 40,489 13,506 139,081 46,395 123,024 41,039 20,981 1964 35,597 11,875 156,419 52,178 129,641 43,246 22,057 1965 33,688 11,238 156,566 52,228 141,784 47,297 23,757 1966 28,615 9,545 167,718 55,948 170,469 56,865 28,230 1967 29,158 9,727 182,628 60,921 190,468 63,537 30,319 1968 27,985 9,335 204,064 68,072 242,674 80,952 39,559 1969 29,254 9,759 218,211 72,791 256,394 85,528 41,646 1970 30,338 10,120 232,541 77,571 274,857 91,687 45,363 1971 30,515 10,179 247,429 82,538 291,983 97,400 49,168 1972 27,383 9,134 238,833 79,670 344,309 114,855 62,451 1973 26,130 8,716 260,298 86,831 360,695 120,321 62,519 1974 25,565 8,528 238,499 79,559 314,568 104,934 51,703 1975 29,540 9,854 293,259 97,826 323,170 107,803 55,819 1976 32,732 10,919 314,452 104,895 322,634 107,625 58,122 1977 33,079 11,035 324,379 108,207 343,927 114,728 60,116 1978 32,024 10,683 355,094 118,453 329,542 109,929 58,025 1979 36,380 12,136 364,708 121,660 340,642 113,632 60,494 1980 35,645 11,891 370,906 123,727 350,744 117,002 59,177 1981 37,615 12,548 346,203 115,487 321,833 107,358 56,717 1982 39,445 13,158 345,471 115,243 316,514 105,583 57,166 1983 45,257 15,097 320,495 106,911 320,621 106,953 59,447 1984 49,650 14,840 334,131 99,868 369,453 110,425 61,066 1985 52,817 17,175 316,417 102,890 413,608 134,494 62,358 1986 55,087 20,169 306,179 112,101 463,361 169,650 69,506 1987 52,255 16,425 300,649 94,500 574,942 180,716 73,149 1988 47,439 17,205 308,114 111,743 697,246 252,868 80,051 1989 47,594 13,784 333,799 96,676 736,582 213,331 65,891 1990 46,162 14,669 292,670 93,005 768,945 244,357 62,047 1991 39,031 14,170 267,224 97,011 715,513 259,755 56,332 1992 43,201 10,892 280,792 70,797 738,388 186,173 50,352 1993 40,540 10,573 258,880 67,520 833,041 217,269 59,005 1994 43,496 14,822 242,553 82,655 680,082 231,753 53,388 1995 41,033 15,238 256,560 95,165 712,034 264,423 65,588 1996 40,576 13,202 256,933 83,599 669,567 217,860 64,938 1997 43,609 12,306 247,854 69,944 746,183 210,573 59,152 1998 39,911 14,689 209,721 77,188 841,794 309,825 67,427 1999 39,911 11,527 205,645 59,394 853,599 246,535 57,921 2000 44,016 14,701 169,520 56,620 879,804 293,857 64,585 2001 51,869 21,669 159,989 66,839 792,751 331,188 65,149 2002 49,669 23,090 185,939 86,441 806,549 374,956 67,028 2003 63,739 26,983 193,482 81,908 875,200 370,506 61,342 2004 56,290 26,822 197,722 94,213 725,849 345,863 61,952 2005 52,956 29,166 201,700 111,087 762,390 419,989 52,969 2006 54,381 32,809 154,873 93,436 758,058 457,343 50,032 2007 54,280 26,206 135,440 65,389 984,431 476,622 46,619    Appendix Table A2: Total reconstructed catch (t) for Taiwan within its EEZ, 1950-2007. Others category includes 92 taxa plus miscellaneous marine fishes grouping. Year Katsuwonus pelamis Elasmobranchii Shrimps and prawns Scomber japonicus Thunnus albacares Trichiurus lepturus Todarodes pacificus Saurida tumbil Others 1950 0 7,793 2,105 3,340 5,567 1,113 0 1,113 62,736 1951 0 10,416 2,436 5,208 6,510 1,302 0 1,302 77,286 1952 0 12,071 3,180 6,706 6,706 2,280 0 2,012 88,117 1953 0 13,144 3,216 6,918 6,918 2,352 0 2,075 95,421 1954 0 11,700 2,178 5,611 6,327 2,388 0 2,746 102,320 1955 0 11,447 3,095 7,563 6,132 3,679 0 8,994 98,201 1956 0 12,696 3,088 9,164 5,250 2,196 3,628 8,973 99,974 1957 0 12,680 2,765 6,983 4,870 2,848 827 9,004 110,829 1958 596 12,524 2,685 6,645 4,856 4,430 1,022 6,390 117,518 1959 481 10,908 3,167 7,058 4,652 3,529 2,566 8,502 121,652 1960 384 10,740 2,960 6,450 4,454 4,377 1,920 10,827 121,975 1961 1,143 11,026 3,315 5,868 4,649 6,554 3,201 9,222 151,830 1962 2,140 12,449 4,111 5,502 5,273 10,011 6,572 10,240 161,813 1963 1,496 11,081 5,996 8,582 4,803 8,739 7,716 14,093 177,174 1964 1,267 11,697 6,665 12,752 5,069 12,118 7,366 10,772 188,571 1965 839 12,782 9,164 12,125 5,490 12,430 6,787 15,328 178,996 1966 659 14,354 9,982 9,883 5,888 11,567 8,126 17,204 184,566 1967 453 16,930 12,624 6,498 6,575 14,130 5,365 15,642 204,665 1968 526 21,978 13,748 14,266 8,737 12,765 4,430 15,993 217,913 1969 536 22,525 19,928 17,002 10,178 9,956 5,208 15,624 230,056 1970 292 23,229 17,167 21,940 7,123 11,862 6,047 15,661 248,356 1971 577 23,676 18,323 16,118 5,607 12,338 3,780 15,761 275,201 1972 536 21,116 19,034 11,629 3,266 21,570 6,315 13,442 258,758 1973 1,497 20,561 22,678 5,064 14,061 20,067 14,433 13,167 271,445 1974 1,048 24,982 23,286 2,052 5,581 14,292 8,524 11,706 261,290 1975 1,575 38,465 35,359 2,735 10,106 14,876 13,377 14,733 300,222 1976 1,382 37,171 34,538 3,214 10,357 15,324 16,901 16,760 328,348 1977 1,962 35,664 35,432 6,061 14,470 14,754 12,600 17,996 339,301 1978 3,514 32,624 44,258 4,086 18,880 17,380 12,694 13,863 370,604 1979 2,554 29,720 47,963 4,489 21,237 11,838 23,602 18,177 376,904 1980 2,546 36,983 45,550 5,084 18,631 20,109 20,908 15,771 378,404 1981 1,872 30,727 39,428 15,529 15,626 16,344 24,570 10,975 358,096 1982 2,850 34,832 41,449 13,284 14,742 19,904 21,433 12,107 354,451 1983 5,605 30,954 42,896 15,501 13,784 17,098 11,875 10,956 340,893 1984 5,818 31,454 47,243 19,633 13,244 15,283 48,670 9,067 309,820 1985 12,241 39,056 52,032 23,308 13,454 14,144 17,167 7,993 312,356 1986 12,658 32,378 54,105 25,953 12,573 17,810 11,134 9,230 320,257 1987 13,909 30,018 44,525 20,472 16,236 11,582 12,576 6,503 310,804 1988 16,988 22,493 32,489 19,220 13,743 17,519 13,021 7,666 344,190 1989 26,106 26,216 29,816 18,246 10,007 23,684 23,350 4,468 332,690 1990 38,661 32,716 17,966 17,916 13,211 10,666 18,313 3,449 296,688 1991 32,881 32,123 14,713 15,392 13,192 9,457 26,799 2,892 272,749 1992 39,795 30,159 13,361 23,137 35,331 9,651 5,985 3,749 248,810 1993 49,823 23,912 8,394 21,226 36,165 8,725 5,279 3,445 225,435 1994 65,505 18,864 9,647 32,767 28,406 11,524 6,514 2,639 213,181 1995 81,935 21,428 14,481 32,530 27,565 9,737 13,361 4,051 209,083 1996 91,178 19,680 14,788 33,567 22,632 7,366 11,894 4,085 195,601 1997 58,985 18,503 13,199 28,238 33,512 5,348 7,424 1,916 212,235 1998 76,961 15,523 8,075 18,270 34,680 4,109 17,916 1,568 171,551 1999 66,962 19,803 7,909 23,998 27,530 4,962 5,961 1,628 163,373 2000 68,207 17,884 5,373 13,573 25,092 3,446 3,239 1,712 151,955 2001 70,155 14,463 6,334 12,839 27,285 4,668 2,492 1,188 167,347 2002 86,254 16,422 6,128 19,855 24,292 4,766 2,395 864 192,380 2003 62,978 23,402 611 28,636 21,259 4,633 256 975 229,432 2004 84,513 21,604 7,308 38,264 14,237 5,490 0 1,230 209,543 2005 69,930 21,173 12,847 43,292 18,771 4,716 0 1,281 228,895 2006 72,955 21,676 6,576 30,974 16,678 5,168 0 1,024 186,121 2007 60,188 16,789 4,655 21,673 12,604 3,706 0 818 164,776  Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  107    RECONSTRUCTION OF FISHERIES CATCHES FOR TOKELAU (1950-2009)1  Kyrstn Zylich, Sarah Harper, and Dirk Zeller Sea Around Us Project, Fisheries Centre, University of British Columbia, 2202 Main Mall, Vancouver, BC, V6T 1Z4, Canada k.zylich@fisheries.ubc.ca; s.harper@fisheries.ubc.ca; d.zeller@fisheries.ubc.ca   ABSTRACT  Total marine fisheries catches were estimated for Tokelau between 1950 and 2009. As there are no commercial fisheries in Tokelau, our estimate represents subsistence fisheries only. Subsistence catches were estimated using per capita consumption rates. These rates were either found in independent fisheries studies or estimated from dietary assessments. The subsistence catches total approximately 24,250 t over the 1950-2009 time period (399 t·year-1 in the 2000s), which is almost 4.3 times larger than the FAO reported landings. This report highlights the importance of accurate fisheries catch reporting, as it is required for proper fisheries management and, for small island countries such as Tokelau, it is essential for maintaining food security.  INTRODUCTION  Tokelau, a territory of New Zealand, is comprised of three atolls: Fakaofo, Nukunonu, and Atafu. They are located approximately 300 miles north of Western Samoa at 8°-10° S and 171°-173° W (Figure 1). The land area is only 12.2 km2, with an Exclusive Economic Zone (EEZ) of 319,031 km2 (www.seaaroundus.org). The atolls are situated with Atafu in the northwest, Nukunonu in the middle, and Fakaofo in the southeast. Although each atoll consists of many small islands and islets, the majority of the population for each atoll is confined to one main island or islet (van Pel, 1958). The atolls are all closed, leaving no deep water passes for ships to enter the lagoons (van Pel, 1958). Atafu is the smallest atoll, both in land area (2.5 km2) and lagoon size (19 km2). Nukunonu is the largest with about 5.5 km2 of land area and 109 km2 for lagoon coverage. Fakaofo falls in the middle with 3 km2 of land area, 59 km2 of lagoon coverage (Ono and Addison, 2009).  The Tokelau chain does, geographically, also consist of a fourth atoll, now known as Swains Island (Clanton, 2008). It is located 100 miles south of Fakaofo (Bertram and Watters, 1984) (Figure 1). The island was originally settled in 1400 by the Tokelauans and was given the name Olohega (Bertram and Watters, 1984). However, the atoll was officially annexed to the United States in 1925, and now forms part of American Samoa (Bertram and Watters, 1984; Clanton, 2008). Although there are many different  1  Cite as: Zylich, K., Harper, S. and Zeller, D. (2011) Reconstruction of fisheries catches for Tokelau (1950-2009). pp. 107-117. In: Harper, S. and Zeller, D. (eds.) Fisheries catch reconstructions: Islands, Part II. Fisheries Centre Research Reports 19(4). Fisheries Centre, University of British Columbia [ISSN 1198-6727] Figure 1. Map of Tokelau showing the three atolls, Atafu, Nukunono, and Fakaofo, and its EEZ, as well as Swains Island, a previous fourth atoll known as Olohega.  stories of what occurred with the early foreign explorers who came across Olohega, it is consistent that the eventual annexation to the United States occurred because of the Jennings family. In 1856, an Englishman reportedly sold his perceived ―claim of ownership‖ to an American, Eli Hutchinson Jennings Sr. (Bertram and Watters, 1984; Clanton, 2008). Jennings and his wife ran a copra plantation using Tokelauan labour (Clanton, 2008). In 1917, Tokelauan workers submitted a claim to Western Samoa complaining about the working conditions on Olohega (Bertram and Watters, 1984). However, the matter was referred over to the United States as officials in Apia stated they had no authority over the Jennings family (Bertram and Watters, 1984).  This led to questions of sovereignty of the island and seeing an opportunity to claim new land, the United States took possession. The island is still inhabited by a small Tokelauan population and many Tokelauans continue to consider it part of Tokelau, both historically and culturally (Bertram and Watters, 1984; Ono and Addison, 2009).  Tokelau was under British colonial rule (starting in 1889) and was incorporated with the Gilbert and Ellice Islands Colony in 1916 (Bertram and Watters, 1984). The current three atolls of Tokelau were brought under New Zealand administration in 1926. However, it was not until the Tokelau Act of 1948 that New Zealand took formal control over the atolls (Adams et al., 1995; Townend, 2007). Although New Zealand has been pushing for Tokelau to become an independent self-governing country, Tokelau has resisted at every step. In 1974, the New Zealand Ministry of Foreign Affairs was said to begin the ―process of administrative decolonization‖, even though it was known that Tokelau had firmly rejected the idea of self- government (Bertram and Watters, 1984). Since then, Tokelau voted in two referenda, first in 2006 and again in 2007, to determine whether or not they wish to become a self-governing nation. In both referenda they fell short of a two-thirds majority vote needed to change their political status (Hoëm, 2009).  Despite the Tokelauans best efforts to resist the new social and economic structure that the New Zealand government was trying to implement, there have been some radical changes to their way of life. Although tradition is still observed on the surface, the same sense of community and cooperation is not there (Hooper, 1985). This disruption is the result of New Zealand‘s attempts to introduce new programs and social order, which are counter to the Tokelauans expressed feelings (Bertram and Watters, 1984). A major change which had wide-spread effects was the introduction of the New Zealand public service lines in 1976, which were staffed by Tokelauans (Bertram and Watters, 1984; Hooper, 1985). This took a lot of able-bodied men out of the village labour force. Before this happened, the activities of these men were coordinated by the village council. Now that they are being employed by an outside entity there is less cohesiveness between the village and council (Hooper, 1985). Although this change led to less fisher-hours per week, the new influx of cash allowed more modern fishing gear to be purchased.  New aluminum dinghies (12-14 feet long with 15 to 25 horsepower outboard engines) have mostly replaced the traditional canoes used (some canoes with outboard motors are in use) (Hooper, 1985). These allow for more efficient fishing as it takes less time to manoeuvre the area and less human effort (Hooper, 1985).  As a fairly isolated community, Tokelau has been and continues to be very dependent on subsistence fishing activities. Almost everyone participates in some type of fishing activity (Chapman et al., 2005). Even though ship transport has increased significantly over the years, there is still heavy reliance on locally caught fish in the diet. On Atafu, the people not only engage in subsistence fishing, but also root- crop cultivation and fruit-tree harvesting (Ono and Addison, 2009). Previously, copra was a major commercial crop exported from the atolls and, at times, brought in decent revenues (Bertram and Watters, 1984), but not anymore (Ono and Addison, 2009). Tokelau is heavily subsidized by the New Zealand government, and beginning in 1976, many Tokelauans went to work in the public service lines with wages being paid by New Zealand (Hooper, 1985; Passfield, 1998). Tokelau does bring in a small amount of profit from the sale of handicrafts such as hats, bags, and wood carvings (Passfield, 1998; Ono and Addison, 2009).  Although fishing is extremely important to Tokelauans, catch amounts are not recorded or monitored. The FAO FishStat database, which provides time series data on marine fisheries landings from 1950 to present, is based predominantly on the national statistical data supplied by its member countries. Therefore, the quality of these data depends on the accuracy of the country‘s collection methods.  The objective of this study is to provide a complete time series estimate of the total marine fisheries catch of Tokelau from 1950-2009. Although there have been several studies in the past which have estimated Tokelau‘s fisheries catch for specific years, there has not been a comprehensive review of the trend over time. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  109     MATERIALS AND METHODS  Estimates were only made for subsistence catches, as there are no fishing practices in Tokelau which would qualify as commercial fishing. These estimates were based on reports from several different researchers. These reports consisted of a mixture of catch and consumption data, as well as information about eating and fishing habits, which were used to make assumptions when there were no data available. Available catch data were used, along with human population data, to estimate per capita consumption rates. Interpolations between data anchor points were used to estimate the catch rates for the entire study period (1950-2009). A taxonomic breakdown was then applied, which included categories by species and family.  Human population data  Our estimates for total catch amounts were computed from per capita consumption rates combined with human population data. Population data were obtained from the population statistics historical demography website (www.populstat.info [accessed July 8, 2011]). For most years after 1978, no population data were available. This can be attributed to the fact that in 1975 the New Zealand Ministry of Foreign Affairs assumed control of the islands and shortly after that they reverted from annual population censuses to having one estimate every five years (Bertram and Watters, 1984). Data for the years 1991, 1996, and 2001 were obtained from official Tokelau census data2 as well as the year 2006 (Anon., 2006). The population for 2007 was obtained from Gillett (2009) as this also provided a critical anchor point for seafood consumption for the same year. Data for 2008 and 2009 were linearly interpolated between the 2007 population and a 2011 estimate from the World Factbook3. For the remaining years when no data were available, a linear interpolation between years of known population was performed to give a complete time series of population data from 1950-2009 (Figure 2).  Subsistence fisheries  Several reports indicate the absence of commercial fisheries operating in Tokelau (Dalzell et al., 1996; Passfield, 1998). All fishing activities are therefore deemed to be subsistence. Subsistence fishing in Tokelau includes fishing for personal consumption, as gifts to send overseas, and for trade within the community.  Gillett (2009) estimated the 2007 total catch for Tokelau to be 375 tonnes. This estimate was obtained by analyzing estimates from several independent studies completed between 1977 and 1998, with adjustments for changes in transportation availability and population (Gillett, 2009). This estimate accounts for increased transport between Tokelau and Apia, Samoa, and thus includes 125 tonnes of export (Gillett, 2009). Using these figures and the current population, Gillett (2009) estimated a consumption rate of 214 kg∙person-1∙year-1, which was used as the 2007, 2008, and 2009 anchor point (Table 1).  2 http://www.spc.int/prism/country/tk/stats/Social/Population/age_sex_.htm [accessed June 8, 2011] 3 https://www.cia.gov/library/publications/the-world-factbook/geos/tl.html [accessed June 8, 2011] Table 1. Per capita consumption rates used to estimate total demand. Years Consumption Rate (kg/person/year) Source 1950 255 Assumption-baseda 1951-2006 - Linear interpolation 2007 214 Gillett (2009) 2008 214 Carried forward 2009 214 Carried forward a2007 estimate adjusted for increased domestically sourced fresh fish in 1950. Figure 2. Estimated population of Tokelau, 1950-2009. 0 200 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000 1950 1960 1970 1980 1990 2000 P o p u la ti o n Year   For the early time period (1950-1980s), the only contact that Tokelau had with the outside world was from a trading ship, which three or four times a year brought basic items such as flour, sugar, rice, kerosene, and tobacco to the atolls (van Pel, 1958). It was therefore assumed that they were not, at that time, receiving the canned meats and fish that were available later in the time period (Passfield, 1998). Given the percentages of total animal protein obtained from canned meats and fish in 1998 (Passfield, 1998), it was assumed that this amount of animal protein was supplied in the early period by fresh seafood. In this computation we assumed that the densities of all protein products were the same. The result of this assumption was a consumption rate in 1950 that was 41 kilograms higher than in 2007, (i.e., 255 kg∙person-1∙year-1) (Table 1). We then linearly interpolated between the 1950 assumed and the 2007 reported consumption rates. Using these consumption rates along with the population data, the total demand for fresh fish was calculated for the 1950-2009 time period.  Exports  In order to calculate total subsistence catch, ‗exports‘ needed to be added to what was being locally eaten by the Tokelauans. Although there are no formal exports, the people of Tokelau do send frozen and dried fish to friends and family in Samoa as gifts (Hooper, 1985; Passfield, 1998; Gillett, 2009). Passfield (1998) calculated exports for Fakaofo, and when extrapolated to all of Tokelau, they were estimated to be 15.4 tonnes per year. It is also known that it was around 1980 when the Tokelauans began sending frozen clams to Western Samoa a few times a year (Hooper, 1985). Therefore, the exports were set to zero from 1950 to 1979, and then linearly interpolated between zero tonnes in 1979 and 15.4 t in 1998. We interpolated again between the 1998 point and Gillett‘s (2009) point of 125 t for 2007-2009. Adding these values to the dietary demand for each year gives the estimate for annual subsistence catches, 1950-2009.  Catch composition  The Food and Agriculture Organization of the United Nations (FAO) reports the fish catch for Tokelau under two categories: ‗tuna-like fishes not elsewhere included‘ (nei) and ‗marine fishes nei‘. Also according to the FAO data, up until 1990 there were no tuna-like fish caught (except for one year of non-zero catch data in 1956). However, it is known that tuna- fishing has historically been a very important practice to the Tokelauans (Gillett and Toloa, 1987; Ono and Addison, 2009). In order to show a more complete picture of the catch composition, information was combined from several different sources. Gillett and Toloa (1987) gave us a basic composition by not only recording the species but also the percentage breakdown of the tuna and tuna-like fish. They also recorded the species information for the other pelagic fish, but only provided an overall percentage. Based on the household survey of Passfield (1998) and the observations of van Pel (1958), we broke down the inshore fish category into species and assigned percentages to both those and the species within the other pelagic category. Although marlins (Makaira spp.) have been known to be landed by Tokelauan fishers, they have not been included as a distinct taxon within the catch Table 2. Estimated catch composition for the subsistence fisheries of Tokelau. Prior to 1986, adjustments were made to account for the complete absence of Katsuwonus pelamis until 1980. Percentages were linearly interpolated between 1980 and 1986. (See ‗catch composition‘ in text). Taxa Catch (%) 1950-1980 1986-2009 Acanthocybium solandri 6.02 5.06 Acanthuridae 3.65 3.65 Elagatis bipinnulata 1.50 1.50 Selar crumenophthalmus 7.30 7.30 other Carangidae 7.30 7.30 Carcharhinidae 1.00 1.00 Chaetodontidae 1.46 1.46 Coryphaena hippurus 1.50 1.50 Cypselurus spp. 18.25 18.25 Grammatorcynus bilineatus 0.13 0.11 Gymnosarda unicolor 1.05 0.88 Istiophorus platypterus 0.50 0.50 Katsuwonus pelamis 0.00 3.52 Labridae 1.46 1.46 Lethrinidae 1.46 1.46 Lutjanidae Misc. invertebrates Misc. marine crustaceans 2.19 3.65 1.46 2.19 3.65 1.46 Misc. marine fishes 3.65 3.65 Mugilidae 3.65 3.65 Octopus spp. 1.46 1.46 Panulirus spp. 1.46 1.46 Scaridae 10.95 10.95 Serranidae 2.19 2.19 Sphyraena barracuda 0.50 0.50 Thunnus albacares 7.40 6.215 Thunnus obesus 7.40 6.215 Tridacna spp. 1.46 1.46 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  111   composition. Gillett (1985) reports Tokelauan fishers occasionally catch billfish, with sailfish (Istiophorus platypterus) being caught most commonly, while marlins are landed far less frequently than they used to be. Table (2) shows the complete breakdown of the subsistence catch for 1986 and onwards. Prior to 1986, the percentages of the various tuna species were adjusted to account for skipjack tuna (Katsuwonus pelamis) being absent from the waters of Tokelau from 1950-1980 (Table 2).  Skipjack tuna is subject to large fluctuations in abundance. For example, the Skipjack Survey and Assessment Programme which completed its surveys for Tokelau waters in 1978, spotted 71 schools in just 35 hours (Anon., 1983). By contrast, the United States Bureau of Commercial Fisheries spent approximately 47 hours of a research cruise looking for schools of skipjack in the waters of Tokelau and spotted none (Anon., 1983). A native Tokelauan communicated to the researchers of the Skipjack Survey that in a span of ten years they will probably see one or two extremely abundant years (Anon., 1983). Hooper (1985) described a span of 15-20 years with no skipjack sightings. This covers the period of 1950- 1970. In 1971, skipjack tuna reappeared but only for a span of about a week, after which they disappeared again (Hooper, 1985). In 1986, skipjack was recorded as part of the catch during a 12 week study (Gillett and Toloa, 1987). Taking into account all of this information, these fluctuations were adjusted for by assigning zero percent of the tuna catch to skipjack from 1950-1980. We renormalized the percentages for the other tuna species for this time period (Table 2). Skipjack tuna percentages were then linearly interpolated from zero in 1980 to 3.52% in 1986. The values for the other tuna species were also adjusted, as the percentage of skipjack increased.  Baitfish  Fisheries of Tokelau use handline and noose fishing techniques which require the use of baitfish. Baitfish fisheries in other parts of the South Pacific operate in parallel to the tuna pole-and-line fishery, often using the same vessels (Gillett, 2011). The baitfish required to catch a given amount of tuna varies depending on the type of bait species used, fishing method applied, and many other factors. Tokelau appears to be a unique case in all aspects of baitfish use. The method typically used in other countries does not apply for Tokelau due to topographical characteristics which prevent large vessels from entering the lagoon (Anon., 1983). Tokelauans use very traditional methods for tuna fishing which differ from the method that the tuna to baitfish ratio of 32:1 is based on (Gillett, 2011). Reports on Tokelauan use of baitfish suggest a range of species are used including flying fish (Cypselurus spp.), double-lined mackerel (Grammatorcynus bilineatus), garfish (Hemiramphus spp.), shortfin scad (Decapterus macrosoma), bigeye scad (Selar crumenophthalmus), and squirrelfish (Holocentridae) (Gillett, 1985). This variation in baitfish use may reflect the seasonal availability of certain species. Given these differences in baitfish use, we assume that the traditional methods used by the Tokelauans are more specific and efficient than the typical baitfish catch method for pole-and-line gear. Assuming that Tokelauan fishers use half as much baitfish as used for pole-and-line fishing, we modified Gillett‘s (2011) tuna to baitfish ratio to an adjusted 64:1. Tokelauans use baitfish to catch more than just tuna, therefore this ratio was applied to the following species which are all landed by the handline and noose methods: yellowfin tuna (Thunnus albacares), dogtooth tuna (Gymnosarda unicolor), wahoo (Acanthocybium solandri), rainbow runner (Elagatis bipinnulata), bigeye tuna (Thunnus obesus), indo-pacific sailfish (Istiophorus platypterus), sharks (Carcharhinidae), and barracuda (Sphyraena barracuda) (Gillett, 1985). Skipjack tuna was not included, as they are mainly caught with artificial lures (Gillett, 1985). Traditional methods made use of pearl-shell lures, however these have become rare and difficult to acquire (Hooper, 1985). Gillett (1985) found that Tokelauans are resourceful people and would even make use of a yellow, translucent handle of a screwdriver for a lure. Baitfish estimates for Tokelau were divided evenly between the six previously mentioned taxa of baitfish used. These estimates were not included in the subsistence catch composition and therefore the 3 taxa which are used solely for bait (Hemiramphus spp., Decapterus macrosoma, and Holocentridae) and not actually consumed in the diet are not listed in Table (2).   RESULTS  The reconstructed total catch for the period 1950-2009 was estimated at 24,255 t. This equates to approximately 4.3 times the total landings reported by the FAO on behalf of Tokelau for the same time period (Figure 3). Average annual catches peaked in the 1960s at approximately 460 t·year-1, with a low of 368 t·year-1 in the 1990s. Catches in the recent period (2000s) are estimated to be 399 t·year-1. Exports  0 50 100 150 200 250 300 350 400 450 500 1950 1960 1970 1980 1990 2000 C a tc h  ( t) Year Total reconstructed Supplied to FAO Figure 3. Total reconstructed fisheries catches for Tokelau compared to data supplied to FAO, 1950-2009. totalled 1,091 t over the study period, representing 4.5% of the total catch. However, these were not commercial exports, but rather gifts mainly sent to Samoa.  The total reconstructed catch was dominated by flying fish (Cypselurus spp.) with an estimated 4,425 t over the time period, representing approximately 18% of the total catch (Figure 4). Families Scaridae and Carangidae, as well as the species Selar crumenophthalmus (bigeye scad), Thunnus albacares (yellowfin tuna), Thunnus obesus (bigeye tuna), and Acanthocybium solandri (wahoo) also made up large portions of the catch, with approximately 44 t∙year-1, 29 t∙year-1, 30 t∙year-1, 28 t∙year-1, 28 t∙year-1, and 23 t∙year-1, respectively. The category ‗other taxa‘ includes the remaining 21 taxonomic groups of the subsistence catch (as per Table 2), as well as the other three baitfish taxa, as they represented smaller portions of the total catch. Together, they represent roughly 37% of the total catch. Baitfish alone represented only 0.4% of the total catch, with approximately 90 t being used over the entire time period.  DISCUSSION  The total reconstructed catch for Tokelau during the time period 1950- 2009 was over four times larger than the total catch reported by the FAO on behalf of Tokelau. The main reason for this difference is that there is no system in place for monitoring fisheries catches in Tokelau (Anon., 1983). According to Bertram and Watters (1984), the reports of the New Zealand Ministry of Foreign Affairs, which took over control of the islands in 1974-1975, show the lack of interest the Ministry had in the territory. Large discrepancies exist between the reported and reconstructed tuna catches. The FAO data suggest that there were zero tuna or tuna-like fish caught from 1950-1989, whereas the total tuna and tuna-like fish estimated from the reconstruction for the years of 1950-1989 represented 68% of the estimated tuna and tuna-like catch for the entire 1950-2009 study period. Also, the data supplied to FAO only represent two categories: ‗tuna-like fish nei‘ and ‗marine fish nei‘. In contrast, our reconstruction accounts for 31 taxa. Furthermore, while there are no invertebrate species taken into account in the FAO data, our reconstructed estimate suggests that invertebrates represent 9.5% of the total reconstructed catch.  Since all of Tokelau‘s catches are subsistence and our estimates are based on consumption rates, annual catch rates are subject to fluctuations which correspond to fluctuations in the population. There is a general trend of population decline from its peak in the 1960s until present. Countering the effect of this decline is the increased transport between Tokelau and Samoa which has led to an increase in informal exports. In 1980, Tokelauans began sending frozen seafood by ship to friends and family as gifts. The decrease in dietary demand, in combination with increasing amounts being sent overseas, has resulted in a relatively constant overall trend in catches.  0 50 100 150 200 250 300 350 400 450 500 1950 1960 1970 1980 1990 2000 C a tc h  ( t) Year Scaridae Selar crumenophthalmus other Carangidae Thunnus albacares Thunnus obesus Acanthocybium solandri Figure 4. Taxonomic breakdown of total reconstructed catch for Tokelau. The grouping ‗other taxa‘ represents 24 taxa (as listed in Table 2, plus 3 additional baitfish taxa). Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  113   Fishing is an activity that almost everyone in Tokelau takes part in. Fishing occurs off-shore, on the outer reef, and in the lagoons (van Pel, 1958). Typically, it is the men who fish on the outer reef, with women and children taking part in gathering and fishing in the lagoon and on the reef (Chapman, 1987; Ono and Addison, 2009). Chapman et al. (2005) surveyed the atolls to obtain an estimate of the gender distribution for various fishing activities, the percentage of time using different fishing techniques, and the percentage of people on the island who take part in fishing. Of the 75% of households interviewed, 99% take part in fishing activities.  An average of 12% of the fishing effort was performed by women, with most of this effort focused on reef gleaning and diving for clams. Women also participated, albeit to a lesser extent, in gillnetting and reef fishing (Chapman et al., 2005). Passfield (1998) found that each household spent approximately 14 person hours per week fishing.  Although the Tokelauans have incorporated newer, modern fishing gear into their practices, there are some areas where tradition still dominates. The handmade wooden canoes are still required for handling yellowfin, skipjack, marlin, or large sharks, as well as for a traditional Tokelauan method of noosing pāla (Acanthocybium solandri) (Hooper, 1985). Many traditional fishing methods such as lama hahave (fishing with a scoop net and torch) and takiulu (luring fish into a prepared noose) are still used. Tokelauans also still use the inati system to distribute large catches (particularly of skipjack tuna) or catches of sacred ‗fish‘ (turtles, marlin, and sailfish), so that everyone receives an equal share (Hooper, 1985; Ono and Addison, 2009). Use of spears and spear-guns are not a favoured form of fishing, as it is believed that they result in fish being afraid whenever people enter the water (Hooper, 1985). However, spear-guns have become more widely used in recent years (Ono and Addison, 2009). The main methods of fishing are trolling, reef gleaning, reef fishing, mid-water fishing, gillnetting, diving, bottom fishing, and bait fishing (Chapman et al., 2005). Today, as well as in the past, trolling for Katsuwonus pelamis (skipjack tuna) and Thunnus albacares (yellowfin tuna) is the most important type of fishing to the Tokelauan people (Ono and Addison, 2009).  There are no commercial fisheries in Tokelau. Although it is known that a portion of the catch is used for internal island trade or shipped to friends and family as gifts, these activities are not a commercial cash enterprise. The act of selling fish is viewed as offensive by the Tokelauan people (Hooper, 1985). Traditional systems make it almost impossible to sell fish, regardless of what the individual‘s views are. The people of Tokelau operate under the inati system, which essentially requires fishers to share large catches with the entire community. This takes away the motivation and possibility of catching fish to be sold for pers0nal financial gain (Passfield, 1998). Even if Tokelau abandoned these traditional systems, they do not have the resources to maintain a commercial fleet of their own. There are no ports for larger commercial vessels to dock and land large quantities of catch. The MV Tokelau, which transports goods and people to and from Tokelau,  must anchor offshore and have its passengers and cargo transported to land by a low aluminum barge with an outboard motor (Townend, 2009). The atolls are closed and only have about 15 natural depressions and a few blasted channels which small vessels can pass through into the lagoon (Gillett and Toloa, 1987). Therefore, given the topography of the atolls, it would not be easy to maintain a commercial fleet for Tokelau. In 1980, the United Nations Development Programme (UNDP) attempted to develop the Tokelauan fishing industry by giving each atoll a 29-foot catamaran which would be used to start a small artisanal fishery or to simply help increase their current subsistence efforts (Hooper, 1985). However, use of the catamarans required a large amount of village cooperation and this was difficult to organize with a large proportion of the population working in the public service sector (Hooper, 1985). The catamarans also have high operating costs (e.g., fuel and maintenance) which would require fish to be sold to recoup costs, which is something the Tokelauans frown upon (Hooper, 1985). Thus, the catamarans ended up being used for shipping people and goods across the lagoon (Hooper, 1985). Another reason that commercial fishing would not be profitable in Tokelau is that fishing beyond the subsistence level would possibly put too much pressure on the already limited fish stocks which would likely be detrimental to resource levels (van Pel, 1958; Passfield, 1998).  Overall it appears that there is little threat of overfishing in Tokelau. This can be attributed to the Tokelauans methods and views toward fishing. The Tokelauans are very aware of what is available in their waters and thus know what they can and cannot take. As well, appropriate measures would be taken if any declines were in fact observed (Hooper, 1985). The Tokelauans rely heavily on the fish, and know if overfished they will not have anything to eat for the next week, month, or year. Giant clams on the other hand, may be in jeopardy. According to an early report by van Pel (1958), Tridacna gigas were relatively abundant in the early period, whereas later on there is no mention of them. However, later reports mention two other clam species: Tridacna squamosa and Tridacna maxima (Passfield, 1998). According  to Hooper (1985), giant clams were regarded as an emergency resource that were only harvested when severe weather conditions prevented the catch of other fish. However, around 1980, the Tokelauans made use of the commercial freezers given to them by the UNDP and would load them up with frozen giant clams to be sent to friends and family in Western Samoa as gifts (Hooper, 1985). This occurred only a few times a year based on the shipping schedule (Hooper, 1985). Although this may have put pressure on stocks, it seems that the Tokelauans only had to change their technique and dive a little deeper if they wished to continue harvesting (Hooper, 1985). In a more recent study, there is mention of frequent harvesting of T. squamosa and T. maxima. However, it also stated that on Atafu there have been strict limitations put in place on the harvesting of these clams (Ono and Addison, 2009). Therefore, it seems that concern over these stocks has resulted in measures being put in place to prevent overharvesting.  Although there have not been many studies conducted on the fishing activities of Tokelau, it can be seen from the data that has been collected that even something as simple as a household survey once every five years could provide important information (Zeller et al., 2006; Gillett, 2009). Although creel surveys (i.e., interviews with fishers and surveillance of their catch) are too costly for small countries to complete every year, a survey completed once every five years would provide important data which could be interpolated between collection years (Zeller et al., 2007). If data were collected on the eating habits of Tokelauans, how much they export and import, how much of their seafood consumption is fresh, what types of fish they are catching, or what kind of gear they are using, this could be used to derive more comprehensive estimates of their annual catches (Zeller et al., 2007). Regional (e.g., SPC) and international agencies (e.g., FAO) should consider facilitating and establishing such data collection and utilization approaches, and the required technical and financial resources. For a small island country such as Tokelau, where people rely on the ocean for their sustenance, monitoring of fisheries removals is fundamental to maintaining national food security.   ACKNOWLEDGMENTS  We acknowledge the support of the Sea Around Us Project, a scientific collaboration between the University of British Columbia and the Pew Environment Group.  REFERENCES  Adams, T., Richards, A., Dalzell, P. and Bell, L. (1995) Research on fisheries in the pacific islands region. Forum Fisheries Agency and South Pacific Commision, Nouméa, New Caledonia, 79 p. Anon. (1983) An assessment of the skipjack and baitfish resources of Tokelau. South Pacific Commission, Nouméa, New Caledonia, 46 p. Anon. (2006) Tokelau 2006 census of population and dwellings - tabular report. Statistics New Zealand and the Office of the Council for the Ongoing Government of Tokelau 67 p. Bertram, I.G. and Watters, R.F. (1984) New Zealand and its small island neighbours: A review of New Zealand policy toward the Cook Islands, Niue, Tokelau, Kiribati and Tuvalu. Institute of Policy Studies, Wellington, New Zealand, 574 p. Chapman, L., Des Rochers, K. and Pelasio, M. (2005) Survey of fishing activities in Tokelau. SPC Fisheries Newsletter 115: 36-40. Chapman, M. (1987) Women's fishing in oceania. Human Ecology 15: 267-288. Clanton, A. (2008) The men who would be king: Forgotten challenges to U.S. sovereignty. UCLA Pacific Basin Law Journal 26: 1-50. Dalzell, P., Adams, T.J.H. and Polunin, N.V.C. (1996) Coastal fisheries in the Pacific islands. p. 395-531 In Ansell, A.D., Gibson, R.N. and Barnes, M., (eds.), Oceanography and Marine Biology: An Annual Review. UCL Press, London, UK. Gillett, R. (1985) Traditional tuna fishing in Tokelau. South Pacific Commision, Nouméa, New Caledonia, 52 p. Gillett, R. (2009) Fisheries in the economies of the pacific island countries and territories. Asian Development Bank, Mandaluyong City, Philippines, 521 p. Gillett, R. (2011) Replacing purse seining with pole-and-line fishing in central and Western Pacific: Some aspects of the baitfish requirements. Marine Policy 35: 148-154. Gillett, R. and Toloa, F. (1987) The importance of small-scale tuna fishing: a Tokelau case study. p. 177-190 In Doulman, D.J., (ed.) Tuna Issues and Perspectives in the Pacific Islands Region. Pacific Islands Development Program, Honolulu. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  115   Hoëm, I. (2009) 'Getting out from under': Leadership, conflict resolution and Tokelau migration. p. 191-202 In Lee, H. and Francis, S.T., (eds.), Migration and Transnationalism: Pacific Perspectives. ANU E Press, Canberra, Australia. Hooper, A. (1985) Tokelau fishing in traditional and modern contexts. p. 9-38 In Ruddle, K. and Johannes, R.E., (eds.), The Traditional Knowledge and Management of Coastal Systems in Asia and the Pacific. UNESCO, Jakarta, Indonesia. Ono, R. and Addison, J. (2009) Ethnoecology and Tokelauan fishing lore from Atafu Atoll, Tokelau. SPC Traditional Marine Resource Management and Knowledge Information Bulletin 26: 3-22. Passfield, K. (1998) A report of a survey of the marine resources of Fakaofo Atoll, Tokelau. Fisheries and Environmental Resource Consultants Ltd 31 p. Townend, A. (2007) Tokelau's 2006 referendum on self-government. New Zealand Journal of Public and International Law 5: 121-168. Townend, A. (2009) From a Tokotoko: pacific travels with professor Angelo. Victoria University of Wellington Law Review 39: 575-582. van Pel, H. (1958) A survey of fisheries in the Tokelau islands. South Pacific Commision, Nouméa, New Caledonia, 18 p. Zeller, D., Booth, S., Craig, P. and Pauly, D. (2006) Reconstruction of coral reef fisheries catches in American Samoa, 1950–2002. Coral Reefs 25: 144-152. Zeller, D., Booth, S., Davis, G. and Pauly, D. (2007) Re-estimation of small-scale fishery catches for U.S. flag associated island areas. Fisheries Bulletin 105.     Appendix Table A1: FAO landings vs. total reconstructed catch (in tonnes) for Tokelau, 1950-2009. Year FAO landings Reconstructed catch 1950 0.25 384.0 1951 0.25 408.5 1952 0.25 407.3 1953 0.25 406.1 1954 0.25 430.3 1955 0.25 429.1 1956 0.50 427.8 1957 0.25 451.7 1958 0.25 450.4 1959 0.25 449.1 1960 0.25 472.7 1961 0.25 471.3 1962 0.25 470.0 1963 0.25 468.6 1964 0.25 467.2 1965 0.25 465.8 1966 0.25 464.5 1967 0.25 463.1 1968 0.25 461.7 1969 0.25 411.9 1970 100.00 410.7 1971 100.00 409.4 1972 100.00 384.2 1973 100.00 383.0 1974 100.00 381.9 1975 100.00 380.7 1976 100.00 379.6 1977 100.00 378.4 1978 100.00 377.3 1979 100.00 376.1 1980 100.00 375.8 1981 100.00 379.3 1982 100.00 382.7 1983 100.00 386.2 1984 150.00 389.8 1985 150.00 393.1 1986 150.00 396.5 1987 200.00 391.5 1988 200.00 386.5 1989 200.00 381.6 1990 220.00 376.7 1991 231.00 371.8 1992 191.00 416.5 1993 200.00 393.6 1994 200.00 370.8 1995 200.00 358.0 1996 200.00 345.0 1997 200.00 345.3 1998 200.00 346.0 1999 200.00 357.8 2000 160.00 369.6 2001 150.00 388.8 2002 140.00 396.9 2003 130.00 405.0 2004 120.00 412.8 2005 110.00 420.9 2006 100.00 429.0 2007 90.00 376.7 2008 80.00 388.3 2009 70.00 399.6   Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  117   Appendix Table A2: Total reconstructed catch (in tonnes) for Tokelau (1950-2009) by major taxa. Others grouping includes 24 taxa. Year Cypselurus spp. Scaridae Selar crumenophthalmus Other Carangidae Thunnus albacares Thunnus obesus Acanthocybium solandri Others 1950 70.1 41.9 28.2 27.9 28.3 28.3 23.0 136.3 1951 74.5 44.5 30.0 29.7 30.1 30.1 24.5 145.0 1952 74.3 44.4 29.9 29.6 30.0 30.0 24.4 144.6 1953 74.1 44.3 29.8 29.5 29.9 29.9 24.4 144.2 1954 78.5 46.9 31.6 31.3 31.7 31.7 25.8 152.8 1955 78.3 46.8 31.5 31.2 31.6 31.6 25.7 152.3 1956 78.1 46.7 31.4 31.1 31.5 31.5 25.7 151.9 1957 82.4 49.3 33.1 32.8 33.3 33.3 27.1 160.4 1958 82.2 49.1 33.0 32.8 33.2 33.2 27.0 159.9 1959 81.9 49.0 33.0 32.7 33.1 33.1 26.9 159.4 1960 86.2 51.6 34.7 34.4 34.8 34.8 28.4 167.8 1961 86.0 51.4 34.6 34.3 34.7 34.7 28.3 167.3 1962 85.7 51.3 34.5 34.2 34.6 34.6 28.2 166.8 1963 85.5 51.1 34.4 34.1 34.5 34.5 28.1 166.4 1964 85.2 51.0 34.3 34.0 34.4 34.4 28.0 165.9 1965 85.0 50.8 34.2 33.9 34.3 34.3 28.0 165.4 1966 84.7 50.7 34.1 33.8 34.2 34.2 27.9 164.9 1967 84.5 50.5 34.0 33.7 34.1 34.1 27.8 164.4 1968 84.2 50.4 33.9 33.6 34.0 34.0 27.7 163.9 1969 75.1 44.9 30.2 29.9 30.4 30.4 24.7 146.2 1970 74.9 44.8 30.1 29.9 30.3 30.3 24.6 145.8 1971 74.7 44.7 30.0 29.8 30.2 30.2 24.6 145.4 1972 70.1 41.9 28.2 27.9 28.3 28.3 23.1 136.4 1973 69.9 41.8 28.1 27.9 28.2 28.2 23.0 136.0 1974 69.7 41.7 28.0 27.8 28.1 28.1 22.9 135.6 1975 69.5 41.5 27.9 27.7 28.1 28.1 22.8 135.2 1976 69.2 41.4 27.8 27.6 28.0 28.0 22.8 134.8 1977 69.0 41.3 27.8 27.5 27.9 27.9 22.7 134.3 1978 68.8 41.1 27.7 27.4 27.8 27.8 22.6 133.9 1979 68.6 41.0 27.6 27.3 27.7 27.7 22.6 133.5 1980 68.6 41.0 27.6 27.3 27.7 27.7 22.5 133.4 1981 69.2 41.4 27.8 27.6 27.2 27.2 22.2 136.8 1982 69.8 41.7 28.1 27.8 26.7 26.7 21.7 140.1 1983 70.5 42.1 28.3 28.1 26.2 26.2 21.3 143.5 1984 71.1 42.5 28.6 28.4 25.7 25.7 20.9 147.0 1985 71.7 42.9 28.8 28.6 25.1 25.1 20.4 150.4 1986 72.3 43.3 29.1 28.8 24.6 24.6 20.0 153.9 1987 71.4 42.7 28.7 28.5 24.3 24.3 19.7 151.9 1988 70.5 42.2 28.3 28.1 23.9 23.9 19.5 150.0 1989 69.6 41.6 28.0 27.8 23.6 23.6 19.2 148.1 1990 68.7 41.1 27.6 27.4 23.3 23.3 19.0 146.2 1991 67.8 40.6 27.3 27.0 23.0 23.0 18.7 144.3 1992 76.0 45.5 30.5 30.3 25.8 25.8 21.0 161.6 1993 71.8 42.9 28.9 28.6 24.4 24.4 19.8 152.7 1994 67.7 40.5 27.2 27.0 23.0 23.0 18.7 143.9 1995 65.3 39.1 26.2 26.0 22.2 22.2 18.1 138.9 1996 62.9 37.6 25.3 25.1 21.4 21.4 17.4 133.9 1997 63.0 37.7 25.3 25.1 21.4 21.4 17.4 134.0 1998 63.1 37.8 25.4 25.2 21.4 21.4 17.4 134.3 1999 65.3 39.0 26.2 26.0 22.2 22.2 18.0 138.8 2000 67.4 40.3 27.1 26.9 22.9 22.9 18.6 143.4 2001 70.9 42.4 28.5 28.3 24.1 24.1 19.6 150.9 2002 72.4 43.3 29.1 28.9 24.6 24.6 20.0 154.0 2003 73.9 44.2 29.7 29.5 25.1 25.1 20.4 157.1 2004 75.3 45.1 30.3 30.0 25.6 25.6 20.8 160.2 2005 76.8 45.9 30.9 30.6 26.1 26.1 21.2 163.3 2006 78.3 46.8 31.5 31.2 26.6 26.6 21.6 166.5 2007 68.7 41.1 27.6 27.4 23.3 23.3 19.0 146.2 2008 70.8 42.4 28.5 28.2 24.0 24.0 19.6 150.7 2009 72.9 43.6 29.3 29.1 24.8 24.8 20.2 155.1    Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  119   RECONSTRUCTING MARINE FISHERIES CATCHES FOR THE KINGDOM OF TONGA: 1950-20071  Patricia Sun, Sarah Harper, Shawn Booth and Dirk Zeller Sea Around Us Project, Fisheries Centre, University of British Columbia, 2202 Main Mall, Vancouver, BC, V6T 1Z4, Canada p.sun@fisheries.ubc.ca; s.harper@fisheries.ubc.ca; s.booth@fisheries.ubc.ca; d.zeller@fisheries.ubc.ca  ABSTRACT  Total marine fisheries catches were estimated for Tonga from 1950-2007 by reconstructing past catches and accounting for all fisheries sectors, including commercial and non-commercial components. Our catch reconstruction for Tonga estimated the total catch to be over 3.5 times larger than the landings reported to FAO over the study period. For recent period (2000s), this discrepancy has decreased with total catches averaging 5,600 t∙year-1, compared to 3,300 t∙year-1 reported by FAO on behalf of Tonga. The majority of catches that were unaccounted for in the reported landings were from the subsistence sector, which represented 70% of the total reconstructed catch. This illustrates the substantial under-representation of small-scale subsistence fisheries in the official statistics for Tonga. Small-scale fisheries play a key role in the socioeconomic framework and food security of Pacific Island countries such as Tonga and this needs greater recognition and better accounting.   INTRODUCTION  The Kingdom of Tonga is located between 15˚- 23.5˚S and 173˚- 177˚W in the South Pacific Ocean (Figure 1). Tonga consists of approximately 170 islands, of which 37 are inhabited (Zann 1994; Anon., 2010a), and has an Exclusive Economic Zone (EEZ) of approximately 665,000 km2 (www.seaaroundus.org).  The islands of Tonga are clustered into three groups: Tongatapu, Ha‘apai, and Vava‘u, which have a combined land area of about 747 km2 (Malm, 2009). About 70% of the population of Tonga resides on Tongatapu, which is also the location of the capital city, Nuku‘alofa. The islands of Ha‘apai and Vava‘u are less developed, have smaller populations and rely more on subsistence fishing and farming (Evans et al., 2003).  Prior to the Tongan constitution of 1875, fishing rights to nearshore marine areas were under control of community chiefs and belonged only to coastal people (Kronen et al., 2003). After 1875, there was an abolishment of exclusive fishing rights to particular marine areas and all people had the right to fish or gather marine resources (Kronen et al., 2003). This system is still in place today, with the exception of fish fences, live rock extraction,  1  Cite as: Sun, P., Harper, S., Booth, S. and Zeller, D. (2011) Reconstructing marine fisheries catches for the Kingdom of Tonga: 1950- 2007. pp. 119-130. In: Harper, S. and Zeller, D. (eds.) Fisheries catch reconstructions: Islands, Part II. Fisheries Centre Research Reports 19(4). Fisheries Centre, University of British Columbia [ISSN 1198-6727].  Figure 1. Map of Tonga and its Exclusive Economic Zone.  aquaculture, and marine protected areas (Kronen et al., 2003).  Tongan fisheries include both commercial and non-commercial sectors. The small-scale fisheries target reef-lagoon, pelagic and deep-slope species. While the majority of small-scale fishery catches are from reef areas, Tonga‘s deep-slope fishery has been considered one of most successful in the Pacific Island region (Dalzell et al., 1996). Traditionally, Tongans fished only inside the reef (Bell et al., 1994); however, in the early 1980s a deep-water fishery began, initially using small boats, but eventually expanding to include larger vessels (Mead, 1979; 1981). A large-scale commercial fishery for offshore pelagics started in the 1980s with one vessel, but grew to 25 vessels by the 2000s (Likiliki et al., 2005). A commercial lobster fishery began in the late 1960s and a sea cucumber (bêche-de-mer) fishery was established in the 1990s. Neither of these invertebrate fisheries expanded substantially. The open ocean fisheries were conducted mainly by men, while reef collecting activities (reef gleaning) involved mostly women and children (Malm, 2009).  Expansion of Tonga‘s commercial fisheries has been relatively unstable due to large fluctuations in stocks. For example, mullet stocks (Mugilidae) in Tongatapu have faced several major collapses, beginning in the 1960s when monofilament nets were introduced (SPC, 1988). Deepwater snapper (Lutjanidae) stocks on offshore slopes and seamounts have also been substantially depleted (Zann, 1994). The lobster fishery has gone through various boom and bust cycles since it began in the 1960s (Bell et al., 1994) and the bêche-de- mer fishery, which began in 1990, was deemed overfished by 1996 and has been under a ten-year moratorium since 1998 (Bell et al., 1994).  The commercial fisheries sector of Tonga makes up around 3% of the country‘s GDP (Gillet, 2009); however, subsistence farming and fishing are practiced by the majority of people (Zann, 1994). Though there has been a shift in recent years towards agriculture and imported foods in Tongatapu, fishing remains an important source of protein in the diet for many on the outer islands of Ha‘apai and Vava‘u.  Methods for subsistence fishing vary depending on the equipment that is readily available, and many fishers employ a combination of techniques. The most popular gear is the handline, followed by spear fishing and gillnetting (Kronen and Bender, 2007). Women and children use spears, traps, and may also participate in some types of group fishing. Men typically fish with spears, hooks, nets, and traps.  The objective of this study is to provide a comprehensive estimate of Tonga‘s total marine fisheries catches by accounting for commercial (industrial and artisanal) and non-commercial (subsistence) fisheries sectors. The Food and Agriculture Organization of the United Nations (FAO) reports marine fisheries landings, as supplied by each member country, in the form of time-series data, extending from 1950 to present, and can be found in the FAO FishStat database (www.fao.org/fishery/statistics/en). These data are based on national statistical data from each country, and can be subject to omissions depending on the method of statistical collection and the quality of data transfer. As such, much of the FAO landings data include only the commercial fisheries sector. Given the importance of fishing for subsistence in the coastal communities of a country such as Tonga, and its common under-representation in catch statistics (e.g., Zeller et al., 2006a), we will estimate this unreported fisheries component to improve our understanding of Tonga‘s use of marine fisheries resources using a catch reconstruction approach (Zeller et al., 2007). This study looked at all subsectors within Tonga‘s fisheries, including large-scale (i.e., offshore fishing), artisanal (i.e., reef, deep-slope, and near-shore fishing), and subsistence fishing in order to estimate total marine fisheries catches by Tonga from 1950-2010.   METHODS  Tongan fisheries focus on targeting reef and lagoon, deep-slope and pelagic species. There are both commercial and non-commercial components to the reef and lagoon fisheries, whereas the deep-slope and pelagic fisheries are predominantly commercial. Estimates of Tonga‘s commercial landings were obtained from FAO, whereas subsistence catch data and taxonomic information on small-scale fisheries were derived from household surveys, independent studies and from gray literature sources. Subsistence fisheries estimates were only available for some years; therefore, subsistence catch data were combined with human population data to derive per capita subsistence catch rates, which were then expanded to cover the entire study period. Our efforts at reconstructing Tonga‘s fisheries catches focused mainly on estimating catches for the small-scale sector and improving the taxonomic resolution of all sectors. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  121    Human Population data  Human population data for 1950 were obtained from populstat (www.populstat.info/), for 1956, 1966, 1976, 1986 and 1996 from the Statistic Department of Tonga (Fifita, 1996), and for 2006 from the Secretariat of the Pacific Community (Anon., 2010b). Linear interpolations were used to estimate the population between years of known data. Data from 1976 onwards provided a population breakdown for the main island (Tongatapu) and the outer islands (Ha‘apai, Vava‘u, and others). Prior to 1976 we extrapolated the ratio back to 1950 and applied it to the total population to derive the Tongatapu and outer Island populations for the earlier time period (Figure 2).  Subsistence fisheries  Most inhabitants of the outer island groups of Ha‘apai and Vava‘u participate in fisheries on some level. On the main island group, Tongatapu, there has been a shift in dietary preferences from fish as a major source of protein to mutton flaps, chicken pieces, and corned beef, which have become cheap alternatives (Finau et al., 1994; Gillett, 2009). Our estimates for subsistence fishing reflect a decrease on islands with increased urbanisation (e.g., Tongatapu) due to wider availability of other food alternatives. This is supported by Kronen et al. (2003), who found a decrease in fresh fish consumption with increased urbanization.  Reliable estimates for subsistence fisheries catches in the early time period were not readily available. Kent (1980) estimated that local fisheries production in 1977 (i.e., 1,100 t) supplied less than half of the populations seafood demand. Assuming the remaining demand was met through subsistence catch, we derived an estimated subsistence amount of 1,300 t for 1977 by assuming that artisanal catches supplied approximately 45% of the overall demand. An estimate by Dalzell et al. (1994), for the mid 1990s was 933 t. Both of these seemed low when considering that more recent estimates were more than double. The estimation techniques in these earlier accounts were not adequately described, and they seemed unrealistically low when converted to per capita subsistence catch rates (i.e., 10-14 kg∙person-1∙year-1). Furthermore, the FAO country profile for Tonga (www.fao.org/fishery/countrysector/FI-CP_TO/en [accessed June 2010]) suggests that these earlier subsistence estimates were in fact low. Therefore, we used Gillet and Lightfoot‘s (2002) estimate for 1997 of 2,863 t as the best estimate of subsistence catches. This 1997 subsistence estimate was converted to per capita subsistence catch rates for the main island and the outer islands using the proportion of subsistence catches from the main and outer islands (40% and 60% respectively) as presented by Lovell and Palaki (2002). The resulting per capita rates were: 17 kg∙person-1∙year-1 for the main island and 56 kg∙person-1∙year-1 for the outer islands for 1997. Halapua (1981 in Bell et al., 1994) estimated that in the Ha‘apai island group (i.e., outer islands), subsistence reef- lagoon fisheries constitute 70% of the total annual catch. We applied this breakdown to the 1997 catch for the outer islands using our estimated subsistence catch as 70% of the total catches to derive our artisanal catch amount (i.e., 56 kg∙person-1∙year-1 accounted for 70% of total catch rate for outer islands being approximately 80 kg∙person-1∙year-1). We assumed that in 1950 catches on the outer Islands were all subsistence and used our 1997 per capita rate for total catches of 80 kg∙person-1∙year-1 (i.e. domestic supply) as the 1950 subsistence catch rate. A linear interpolation was used to derive a time series from 1950 to 1997, and this trend was carried forward from 1997 to 2007. 0 20 40 60 80 100 120 1950 1960 1970 1980 1990 2000 P o p u la ti o n  ( x  1 0 3 ) Year Tongatapu Outer islands Figure 2. Tonga‘s population for the main island (Tongatapu) and outer islands combined (Ha‘apai, Vava‘u and others), 1950-2007. There has been a migration to the main island, Tongatapu, from the outer islands in the recent period. In the mid-2000s approximately 70% of the population resided on Tongatapu, whereas in the 1950s just over 50% resided on the main island.   We further assumed that the per capita subsistence catch rate for the main island (Tongatapu) in 1950 was the same as the outer islands in the late 1990s (56 kg∙person-1∙year-1), as commercial fisheries and food imports were less developed in the early time period and the entire population would have relied more heavily but not exclusively on subsistence catches.  In the later time period, as commercial fisheries developed, mainly around Tongatapu, reliance on subsistence fisheries diminished as the population derived more of their animal protein supply from commercial fisheries and from alternative non-seafood sources including imported food items. To derive a complete time series, we carried the 1950 rate (56 kg∙person-1∙year-1) forward unaltered to 1970, when the reef and lagoon fishery of Tongatapu started to show signs of overexploitation. From the 1970 anchor point of 56 kg∙person-1∙year-1 to the 1997 estimate of 17 kg∙person-1∙year-1 we interpolated linearly. From 1997 onward, the rate was decreased by 1% year-1, which is the same decrease observed in the subsistence rate for the outer islands.  When combined with commercial catches for the domestic market (i.e., artisanal catch), these subsistence catch estimates resulted in an average per capita fish consumption rate of 80 kg∙person-1∙year-1 in 1950, decreasing to 38 kg∙person-1∙year-1 by 2007.  Commercial fisheries data  Offshore pelagics: Catches for tuna presented in the FAO data were comparable to estimates presented in Likiliki et al. (2005) for the tuna longline fishery. For some years, the FAO provide more comprehensive estimates (Table 1). We therefore assumed that the FAO data for offshore pelagics (tuna and billfishes) best represented total catches for the industrial pelagics fishery. We assumed that catches of pelagic species for the domestic market by the small-scale sector, were not included in the landings presented by the FAO and were therefore estimated as a component of the artisanal catch (see taxonomic breakdown section).  Deep slope fishery: The deep slope fishery began in the 1980s and mainly targets snapper (Lutjanidae) and grouper (Serranidae). This fishery was developed to alleviate pressure on the reef and lagoon fish stocks, which by the 1980s were already under pressure from overexploitation (Ministry of Fisheries, 2007). Initially, these catches would have supplied the domestic market; however, once their value as an export commodity was established, most of the catch went to foreign markets (Ministry of Fisheries, 2007). Catches for the deep slope fishery were obtained from Bell et al. (1994) for the period 1986-1992, from the Ministry of Fisheries (2007) for the period 1992-2005, and from FAO Fishstat for 2006 and 2007. Estimates for the 1980-1987 period were derived by linear interpolation from assumed zero catches in 1980 to the first available anchor point in 1987. We assumed that these estimates included catches for both domestic and foreign (export) markets, although the majority would have been for export, and thus were not included in our calculations of total domestic supply (see Subsistence section).  Artisanal fishery: A report by Bell et al. (1994) presented estimates for the artisanal fishery of Tongatapu for 1987 and 1993. The 1993 estimate was based on catches from the two major landing sites, expanded to account for all of Tongatapu.  These estimates were made over a 10-month period (March-December), so catches were also expanded (using the monthly average over the 10-months) to represent catches for January and February. The 1987 and 1993 artisanal catch estimates were 823 t and 386 t, respectively. These included mullet (Mugilidae) catches, which were subsequently removed from the artisanal catch estimate and treated separately. We estimated artisanal catches for 1950 using our Tongatapu subsistence catch estimate for 1950 and the assumption that in 1950 subsistence catches made up 70% of the total catch, while artisanal catches made up the remaining 30%. This assumption was based on our estimated subsistence-artisanal breakdown for the outer islands in the 1990s. On the outer islands, we assumed that in 1950 there was no artisanal sector and that all catches were subsistence. For the later time period, we estimated artisanal catches using our subsistence estimate and the assumption that, in the 1990s on the Table 1. Comparison of tuna catches presented in FAO Fishstat and in Likilili et al. (2005) for the Tongan Longline fishery, which included albacore, bigeye, black marlin, skipjack, swordfish, tuna-like, and yellowfin catches. Year FAO  Likiliki et al. (2005) 1997 690 214 1998 870 193 1999 1129 327 2000 1269 931 2001 1836 1988 2002 1804 1647 2003 1126 1308 2004 517 373  Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  123   outer islands, artisanal catches represented 30% of the total catch. Artisanal catches for 1950, 1987 and 1993 for both main and outer islands were then converted to per capita artisanal catch rates and linear interpolations were done to derive a complete time series from 1950-1993. From 1993-2007 we, conservatively, held the artisanal catch rate constant.  Mullet: Traditionally, mullet were the most sought after species by the domestic market (Bell et al., 1994). Zann et al. (1994) estimated that in the 1960s, mullet represented 40% of all domestically marketed species. However, heavy exploitation of mullet stocks in the 1950s and 1960s lead to a substantial decline in mullet landings, which likely began in the 1970s (Zann et al., 1984) and continued to decline dramatically, with a near collapse in the 1990s (Bell et al., 1994).  FAO data do not present mullet catches as a separate category, but they may be accounted for as ‗miscellaneous marine fishes‘. Independent reports were obtained, which described the state of the mullet fishery and presented catches for some years (Bell et al., 1994; Kimura and Fa‘anunu, 1995). Bell et al. (1994) present mullet landings for Tongatapu of 140 t in 1987 and 3.2 t in 1993. We estimated that from 1950-1970, mullet represented 40% of artisanal catches for Tongatapu, which translates into a 1950 catch of almost 250 t and a 1970 catch amount of 364 t. A linear interpolation was done to estimate catches for the time period, 1970 to 1987 and 1987 to 1993. From 1994 onward we carried forward the 1993 catch of 3.2 t unaltered.  Taxonomic breakdown for artisanal and subsistence fisheries  Taxonomic composition for the reef, deep-slope, and coastal pelagic fisheries are presented in Dalzell et al. (1996) but only to the family level. Mead (1980) provided a similar breakdown for deep-slope and pelagic fisheries, which included both family and species details, but only for 1979. Bell et al. (1994) provided taxonomic information to the species level for the reef fishery for 1987 and 1993, for the deep-slope fishery (1986-1992), for the mullet fishery and for pelagic species. We used the species composition in Bell et al. (1994) as it provided the most comprehensive taxonomic breakdown by sector and species. Mullet catches  Table 2. Species composition of the reef fisheries catches in Tonga applied to both artisanal and subsistence catches. Source: Ministry of Fisheries Database, Inshore Fisheries Statistics (Bell et al., 1994). Family Species Tongan Name Catch (%) Acanthuridae Acanthurus spp. Pone 8.71 Acanthuridae Acanthurus lineatus Ponetuhi 0.50 Acanthuridae A. triostegus Manini 0.25 Acanthuridae Naso spp. Ume lei 0.54 Acanthuridae N. unicornis Ume 8.90 Balistidae Pseudobalistes fuscus Humu 1.53 Belonidae Tylosurus crocodilis Haku 0.10 Carangidae Caranx spp. Lupo 0.33 Diodontidae Diodon spp. Sokisoki 0.32 Ephippidae Platax pinnatus Sifisifi  0.07 Fistulariidae Fistularia spp. Totao 0.09 Gerreidae Gerres spp. Matu 2.01 Haemulidae Plectorhinchus spp. Fotu'a 0.97 Holocentridae Ostichthys spp. Ta'a 2.40 Holocentridae Sargocentron spp. Telekihi 0.19 Holocentridae Myripristis spp. Malau 0.70 Kyphosidae Kyphosus cinerascens Nue 0.50 Labridae Cheilinus undulatus Tangafa 0.13 Leiognathidae Leiognathus spp. Sipesipa 0.05 Labridae Thalassoma spp. Meai 0.87 Labridae Cheilinus spp. Lalafi 0.67 Lethrinidae Lethrinus atkinsoni Hoputu 1.78 Lethrinidae Lethrinus harak Tanutanu 11.53 Lethrinidae Lethrinus nebulosus Koango, Liki  2.58 Lethrinidae Lethrinus spp. Manga 1.33 Lutjanidae Lutjanus bohar Fangamea 0.07 Lutjanidae Lutjanus kasmira Fate 1.21 Lethrinidae Gymnocranius spp. Mu 1.23 Monocanthidae Aluterus spp. Papae 0.05 Mullidae Mulloidichthys spp. Vete 1.40 Mullidae Parupeneus spp. Tukuleia 3.82 Muraenidae Gymnothorax spp. Toke 0.14 Pomacentridae Abudefduf septemfasciatus Tukuku moana 0.68 Priacanthidae Priacanthus spp. Mataheheva 0.40 Scaridae Leptoscarus vaigiensis Ufu 7.67 Scaridae Scarus spp. Olomea, Pose 18.31 Scaridae Bolbometopon muricatum Menenga 0.22 Serranidae Epinephelus spp. Ngatala 5.84 Siganidae Siganus argenteus Ma'ava 6.11 Siganidae S. chrysospilos Pongongo 0.63 Siganidae S. spinus O 4.31 Sphyraenidae Sphyraena barracuda Ono 0.15 Sphyraenidae Sphyraena spp. Hapatu 0.45 Terapontidae Therapon jarbua Kavakava 0.27  were composed of Mugil cephalus (70%), Valamugil seheli (15%) and Liza spp. (15%). The species breakdown for the deep-slope fishery was Etelis coruscans (25.44%), Pristipomoides filamentosus (22.35%), Epinethelus septemfasciatus (13.98%), Etelis carbunculus (6.07%), Lethrinus chrysostomus (5.74%), Pristipomoides flavipinnis (3.90 %), Epinephelus morrhua (3.71%) and others (18.81%). The remainder of the artisanal catch was broken into reef-lagoon species (70%), sharks and rays (0.15%), other oceanic pelagics (20.1%), small pelagics (8.25%), and ‗miscellaneous marine fishes‘ (2.5%). To each of these categories, we applied the species breakdown presented in Bell et al. (1994). Reef fishery catches were assigned to 49 taxa from 25 families (Table 2). Catches of oceanic pelagics were composed of barracudas (Sphyraena spp.; 32%), marlin (Makaira nigricans, M. indica and Tetrapturus audax; 40%) and dolphinfish (Coryphaena hippurus; 28%).  Small pelagics consisted of 11 taxa from 5 families (Table 3). Tuna catches by the artisanal fishery were mainly skipjack (Katsuwonus pelamis; 76%), with smaller amounts of yellowfin tuna (Thunnus albacares; 20%) and little tuna (Euthynnus affinis; 3%). The taxonomic breakdown for sharks was derived from species presented in Bell et al. (1994). We assumed: grey reef shark (Carcharhinus amblyrhynchos; 20%), silvertip shark (C. albimarginatus; 20%), black-tip reef shark (C. melanopterus; 20%), great white shark (Carcharodon carcharias; 5%), hammerhead (Sphyrna lewini; 20%), mako shark (Isurus oxyrinchus; 5%) and tiger shark (Galeocerdo cuvieri; 10%). No species breakdown was available for ray catches. Thus 1/3 of the shark and ray allocation was assigned to rays  Reef gleaning, which targets mainly invertebrates, is known to be a significant contributor to subsistence fishing and is conducted mainly by women (Chapman, 1987; Malm, 2009). While invertebrate fisheries are likely of equal importance as finfish in the subsistence sector (Adams and Dalzell, 1994), quantitative information on the magnitude of invertebrate extractions for subsistence purposes was not readily available (Bell et al., 1994). Therefore, we derived the breakdown of invertebrate to finfish catches for the subsistence fisheries using an estimate given by Malm (2009) for the amount of invertebrates caught by women on the outer islands of Tonga in 1975 (11 kg whole weight∙household-1∙week-1), since almost all invertebrate fishing was conducted by women. We took 30% of the 11 kg∙household-1∙week-1 to estimate the meat portion of the invertebrate catch (60-70% shell weight [Kunatuba and Uwate, 1993 in Malm, 2009]) and assuming a household size of approximately 5 persons. We expanded the repored catch rate to cover 52 weeks (whole year). The resulting per capita invertebrate catch rate was 34 kg∙person-1∙year-1. We then compared the per capita catch rate for invertebrates to the total per capita subsistence catch rate for the outer islands in 1975 (67 kg∙person-1∙year-1) and used the resulting ratio to estimate the finfish component (50%) of the artisanal catch. We then applied this breakdown (50% invertebrates and 50% finfish) to subsistence catches throughout the study period. We used the artisanal sector, reef-species breakdown for the finfish component of the subsistence catch and estimated a breakdown for the invertebrate catch of 80% molluscs and 20% crustaceans.  Table 3. Species composition of artisanal catches of small pelagics. Source: Bell et al. (1994) Species Family  Catch (%) Stolephorus devisi Engraulidae 20.0 Atherinomorus lacunosus Atherininae 20.0 Spratelloides delicatulus Clupeidae 13.0 Selar crumenopthalmus Carangidae 11.0 Atule mate Carangidae 11.0 Hypoatherina ovalaua Atherinidae 9.2 Sardinella sirm Clupeidae 7.0 Herklotsichthys punctatus Clupeidae 6.0 Spratelloides gracilis Clupeidae 2.0 Scomberoides spp. Carangidae 0.4 Gazza minuta Leiognathidae 0.4 0 1 2 3 4 5 6 7 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Year as supplied to FAO Total reconstructed Figure 3. Total reconstructed catch for Tonga 1950-2007 compared to total landings as supplied to the FAO. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  125    RESULTS  Total reconstructed catch  Our estimate of total marine fisheries catches for Tonga, 1950-2007, which included total commercial catches and subsistence catch estimates, was over 323,000 t (Figure 3). This estimate of total marine fisheries catches for Tonga is over 3.5 times larger than the 94,753 t that were reported by the FAO on behalf of Tonga, with major discrepancies prior to the mid-late 1980s. For the more recent years, total reconstructed catches were on average 2 times larger than reported landings. The reconstruction also suggests that total catches peaked in the 1970s and 1980s and has declined since. This pattern is in contrast to the reported data which suggest catches increased until the early 2000 (Figure 3).  Our data suggest that the subsistence sector catches 3 times the amount of fish that the commercial sector lands, and accounts for 70% of total marine fisheries extractions (Figure 4). The time series of subsistence catches shows an increasing trend until the early 1970s, followed by a decrease throughout the remainder of the study period (Figure 4).  Catches for the artisanal sector (mullet and deep slope catches excluded) were estimated to be 54,309 t over the 1950-2007 time period (Figure 4). Of this total, 54% were from the main island of Tongatapu, while the remaining 46% were from the outer islands.  Mullet catches totalled approximately 250 t in 1950, increased to a peak of approximately 360 t∙year-1 in the late 1960s and then steadily decreased, with a dramatic decline in the late 1980s (Figure 4). We estimated that catches of mullet in the recent period (2000s) were about 3 t∙year-1.  The deep slope fishery caught an estimated 5,176 t, mainly of snapper and grouper, between 1980 and 2007 (Figure 4). The deep-slope fishery represented approximately 5% of all commercial sector catches.  Large tuna and billfish catches totalled 19,496 t from 1967-2007. Prior to 1967, catches of tuna and billfish were not reported by the FAO as a separate category.   DISCUSSION  Our total reconstructed catch estimate for Tongan marine fisheries for the period 1950-2010, was 3.5 times larger than the data reported to FAO on behalf of Tonga suggest. For more recent periods, this discrepancy declined to two times, i.e., 2,300 t∙year-1. Subsistence catches accounted for the largest unreported component of our reconstruction, representing 84% (3,200 t∙year-1 ) and 48% (2,600 t∙year-1) of the total catch for Tonga in 1950 and 2007, respectively, which is in line with the regional estimate for the South Pacific Islands of 80% (Zann and Vuki, 2000). Artisanal catches were estimated using independent data and were approximately 16% (623 t∙year-1) and 24% (1,317 t∙year-1) of the total catch for 1950 and 2007, respectively. Artisanal catch estimates likely did not include the portion of the catch that was sold directly to restaurants, hotels or at the roadside, as catch estimates from Bell et al. (1994) were 0 1 2 3 4 5 6 7 1950 1960 1970 1980 1990 2000 C a tc h  ( t x  1 0 3 ) Year Deep-slope Subsistence Artisanal Large Pelagics Mullet Figure 4. Total marine fisheries catches for Tonga, 1950-2007. Catches include mullet and deep-slope  and large pelagic fisheries in addition to estimates of artisanal and subsistence sectors.  based on surveys from fish market sales. Our estimate of artisanal catches may, therefore, have been on the low side. The remaining catch was from the large-scale pelagic fishery, which was zero in 1950 and 28% of the total catch in 2007.  Artisanal fisheries (i.e., small-scale commercial) are the main supplier of seafood to the domestic commercial market in Tonga. In 1950, we assumed that the outer islands had no commercially marketed seafood, and all seafood was sourced through subsistence fishing. However, our estimates of artisanal catches on the outer islands for the recent period were substantial, compared to catches on the main island of Tongatapu. This is likely due to the more recent transition on the outer islands towards a cash-based economy. More fish are likely sold, whereas previously, catches on the outer islands were mainly subsistence. Our results suggested that overall, the main island of Tongatapu landed 54% of the artisanal catch, while the outer islands landed the remaining 46%. Mead (1987) provided a similar estimate, with over half of the commercial catch being landed in Tongatapu.  The average per capita fish consumption rate was 80 kg∙person-1∙year-1 in 1950, decreasing to 38 kg∙person-1∙year-1 by 2007 when the commercial production for the domestic market (i.e., artisanal catch) was combined with the subsistence catch estimate. This is well within the range presented by Gillet and Lightfoot (2002) of between 14-102 kg∙person-1∙year-1 and for the recent period, close to the range given by Anon. (1993) of 20-50 kg∙person-1∙year-1. Lower consumption rates were presented in the literature (e.g., 23 kg∙person-1∙year-1 derived from Kimura and Fa'anunu [1995]); however, these were likely based solely on commercial production.  On the main island, there has been a shift in diet preference away from seafood toward cheaper, imported animal protein alternatives such as mutton flaps and corned beef (Finau et al., 1994). The reason for this shift may have been partly driven by a decline in reef-lagoon stocks due to overexploitation around Tongatapu, in addition to the increasing cost of local seafood. The proportionately smaller artisanal catches on the main island as compared to the outer islands may be the result of this decline in availability and accessibility of marine products on Tongatapu.  The large-scale commercial fisheries sector of Tonga expanded substantially in the 1980s, when profitable commercial tuna longline operations began (Bell et al., 1994). Most of the tuna catch is exported, making this fishery the highest contributor to fisheries-based revenue for Tonga. The domestic commercial tuna fishery expanded in the 1980s from one vessel to 33 by 2003 (Likiliki et al., 2005). According to Bell et al. (1994), tuna stocks in Tonga did not seem depleted at that time, and could support an expansion, if this fishery were given sufficient resources. Annual tuna catches consist of 70% albacore (Thunnus alalunga), 20% yellowfin (Thunnus albacares) and 10% big eye (Thunnus obesus; Likiliki et al., 2005). Severe declines in 2003 and 2004 catches due to the effects of El Nino resulted in all foreign vessels leaving Tonga in 2005 (Likiliki et al., 2005; Gillett 2009).  Mullet is an important food fish in the Tongan diet (Kimura and Fa‘anunu, 1995). Mullet catches began to decline in the 1970s, followed by dramatic declines in the 1980s and early 1990s due to the use of fish fences (Bell et al., 1994; Kimura and Fa‘anunu, 1995). Both of these sources agree that Mugil cephalus, which made up 70-75% of the mullet catch in the 1980s, was on the verge of local extinction by the 1990s.  This study illustrated the importance of the small-scale fisheries sector to Tonga, and the magnitude of under-representation of subsistence catches. Most Tongans, fish more to meet their daily food needs than for commercial purposes (Kronen, 2004); however, this sector is often overlooked in fisheries management. The subsistence sector accounts for a substantial portion of total marine fisheries removals and draws mainly from nearshore, reef resources. Gillet (2009) estimated that the contribution by the subsistence sector to Tonga‘s GDP to be roughly 3%, while the commercial sector contributes roughly 6%. Reef fisheries provide a crucial source of animal protein to the Tongan people, particularly the inhabitants of the outer islands where subsistence fisheries dominate. Thus, reef resources and subsistence fisheries are crucial to national food security. Our estimate of domestic seafood supply, based on artisanal and subsistence catches, ranged from 80 kg∙person-1∙year-1 in 1950 to 38 kg∙person-1∙year-1, in 2007, which is within the range of 14-102 kg∙person-1∙year-1 given in Gillet and Lightfoot (2002), but less than the 102 kg∙person-1∙year-1 presented by Finau et al. (1994). Our estimate of domestic seafood supply was on average higher than some previous estimates (Kent, 1980; Anon., 1993; Kunatuba and Uwate, 1983 in Kimura and Fa‘anunu, 1995); however, their lower consumption estimates were likely based only on Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  127   commercial production. There is agreement, however, that the majority (60-70%) of domestic production is from reef and lagoon resources (Bell et al., 1994; Zann and Vuki, 2000).  The health of the Tongan people relies heavily on the health of the reef ecosystem. Although parts of Tonga have switched to a cashed-based economy which relies less on subsistence fishing, alternative economic systems such as bartering still exist on the outer islands and should be considered in resource management decisions and policy development, particularly in terms of food security. For countries such as Tonga, who remain closely tied to the sea for their basic needs, proper accounting of marine fisheries extractions is paramount to their socio-economic stability.   ACKNOWLEDGEMENTS  We acknowledge the support of the Sea Around Us Project, a collaboration between the University of British Columbia and the Pew Environment Group.  REFERENCES  Anon. (1993). Tonga National Tourism Plan / prepared by Nicholas Clark and Associates for Government of Tonga and Asian Development Bank.  Nicholas Clark and Associates, Nuku'alofa. Available at: http://www.adb.org/Documents/Reports/Consultant/24244-TON/24244-08-TON-TACR.pdf [accessed May 21, 2010]. Anon. (2010a) CIA world fact book. Available from: www.cia.gov/library/publications/the-world- factbook/geos/tn.html [accessed: March 9, 2010] Anon. (2010b). SPC Statistics and Demography. Available from: http://www.spc.int/sdp/index.php?option=com_docman&task=cat_view&gid=28&Itemid=42 [accessed March 30, 2010] Adams, T.J.H. and P.J. Dalzell. 1994. Artisanal fishing. Pacific Islands Biodiversity Issues. Bell, L., Fa‘anunu, U., and Koloa, T. (1994) Fisheries resource profiles: Kingdom of Tonga. Forum Fisheries Agency. 197 pages Chapman, M. (1987) Women‘s fishing in Oceana. Human Ecology 15 (3): 267-288.  Evans, M., Sinclair, R.C., Fusimalohi, C., Viliami, L.A., and Freeman, M. (2003). Consumption of traditional versus imported foods in Tonga: implications for programs designed to reduce diet-related non-communicable diseases in developing countries. Ecology of Food and Nutrition. 42(2):153-176. FAO (2002) Available from: www.fao.org/fi/fcp/en/TON/profile.htm [accessed: July 24, 2007] Fifita, V. (1996). Statistics department Tonga. Available from: http://www.pmo.gov.to/tongastats/Population-Census- 1996.html [accessed March 30, 2010]. Finau, T.L., Udagawa, K., and Nakajo, N. (1994) Fish and meat consumption of Tongan people. Fisheries Research Bulletin of Tonga. 1:29-36. Gillett, R. (2009) Fisheries in the economies of the Pacific island countries and territories.  Andaluyong City, Philippines: Asian Development Bank. Gillett, R. and C. Lightfoot. (2002) The contribution of fisheries to the economies of Pacific island countries. Pacific studies series. Asian Development Bank. Forum Fisheries Agency and World Bank, Manila. Kent, G. (1980). The politics of Pacific Island Fisheries. Westview Press: Boulder ,USA. 191 p.Kimura, T., and Fa'anunu, U. (1995) Biological survey and resource management of mullet in Tonga. Inshore Fish. Mgmt. South Pacific Commission. pp.317-322. Kronen, M. (2004) Fishing for fortunes? A socio-economic assessment of Tonga‘s artisanal fisheries. Fisheries Research 70: 121-134. Kronen, M., Clua, E., McArdle, B., and Labrosse, P. (2003) Use and status of marine resources – a complex system of dependencies between man and nature: case studies from Tonga and Fiji, South Pacific. Noumea, New Caledonia: Secretariat of the Pacific Community. Likiliki, P.M., Matoto S.V., Fa'anunu, U. (2005) Tonga tuna fisheries status report. Western and Central Pacific Fisheries Commission. Lovell, E. R. and Palaki, A. (2002) National coral reef status report Tonga. pp 317-343.  In Kulbicki, M. - a 485 p. Malm, T. (2009) Women of the coral gardens: the significance of marine gathering in Tonga. SPC Traditional Marine Resource Management and Knowledge Information Bulletin #25. pp. 2-15. Mead, P.D. (1987). Deep Sea Fisheries Development Project. Report of the third visit to Tonga. South Pacific Commission, Noumea, New Caledonia. 49 pages. Minister of Fisheries. 1997. Report of the Minister for Fisheries. Government of Tonga. Ministry of Fisheries. 2007. Tonga national snapper & grouper fisheries management plan. SPC (1988) Twentieth regional technical meeting on fisheries: country statement – Tonga. WP.23.  Wilkinson, W. (1973). Fisheries developments in Tonga. South Pacific Islands Fisheries Newsletter No. 9, July 1973, South Pacific Commission , Noumea, pages 38-43, (Fish Lib, SPC Lib). Zann, L.P. (1994) The status of coral reefs in South Western Pacific islands. Marine Pollution Bulletin. 29: 52-61. Zann, L.P., Vuki, V.C. (2000) The Status and Management of Subsistence Fisheries in the South Pacific. Ocean Yearbook No. 14. The International Ocean Institute, The University of Chicago Press. Zeller, D., Booth, S., Craig, P. and Pauly, D. (2006a) Reconstruction of coral reef fisheries catches in American Samoa, 1950-2002. Coral Reefs 25: 144-152. Zeller, D., Booth, S., Davis, G. and Pauly, D. (2007) Re-estimation of small-scale fisheries catches for U.S. flag island areas in the Western Pacific: The last 50 years. Fishery Bulletin 105: 266-277. Zeller, D., Booth, S. and Pauly, D. (2006b) Fisheries contributions to GDP: Underestimating small‐scale fisheries in the Pacific. Marine Resource Economics 21: 355‐374. Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  129    Appendix Table A1. FAO landings vs. total reconstructed catch (in tonnes) for Tonga, 1950-2007, in metric tonnes. Year FAO landings Total reconstructed catch 1950 200 3,830 1951 200 3,944 1952 200 4,056 1953 200 4,167 1954 300 4,278 1955 300 4,387 1956 300 4,495 1957 300 4,647 1958 300 4,798 1959 300 4,947 1960 400 5,094 1961 400 5,240 1962 400 5,385 1963 400 5,528 1964 400 5,670 1965 500 5,811 1966 500 5,950 1967 581 6,110 1968 548 6,156 1969 555 6,240 1970 444 6,306 1971 533 6,297 1972 532 6,294 1973 626 6,282 1974 753 6,273 1975 928 6,259 1976 1,038 6,233 1977 1,216 6,160 1978 1,210 6,065 1979 2,000 5,988 1980 1,994 6,328 1981 2,093 6,247 1982 2,229 6,267 1983 2,365 6,218 1984 2,502 6,133 1985 2,690 6,070 1986 2,952 6,008 1987 2,724 5,608 1988 2,692 5,425 1989 2,664 5,245 1990 1,616 5,031 1991 1,915 4,890 1992 2,217 4,767 1993 2,282 4,645 1994 2,435 4,648 1995 2,530 4,596 1996 2,826 4,598 1997 2,763 4,570 1998 3,937 4,724 1999 4,029 4,930 2000 3,545 5,054 2001 4,332 5,499 2002 4,493 5,518 2003 4,107 4,815 2004 1,414 4,307 2005 1,632 4,468 2006 2,041 4,597 2007 1,927 4,599    Appendix Table A2. Total reconstructed catch (in tonnes) for Tonga by major taxa, 1950-2007. Others grouping includes 81 taxa. Year Misc. molluscs Scarus spp. Misc. crustaceans Lethrinus harak Naso unicornis Acanthurus spp. Leptoscarus vaigiensis Thunnus alalunga Others 1950 1,283 342 321 215 166 162 143 - 1,199 1951 1,317 352 329 222 171 168 147 - 1,239 1952 1,351 363 337 229 176 173 152 - 1,278 1953 1,384 373 345 235 181 178 156 - 1,317 1954 1,417 384 353 242 187 183 161 - 1,356 1955 1,450 394 361 248 192 188 165 - 1,394 1956 1,482 405 369 255 197 192 169 - 1,431 1957 1,530 419 381 264 204 199 175 - 1,483 1958 1,577 433 392 273 210 206 181 - 1,534 1959 1,623 447 404 281 217 213 187 - 1,584 1960 1,670 461 415 290 224 219 193 - 1,633 1961 1,715 474 426 299 230 226 199 - 1,682 1962 1,761 488 437 307 237 232 204 - 1,730 1963 1,806 501 448 316 244 238 210 - 1,778 1964 1,851 515 459 324 250 245 215 - 1,825 1965 1,895 528 470 332 256 251 221 - 1,871 1966 1,939 541 481 341 263 257 226 - 1,917 1967 1,964 549 487 345 267 261 230 - 1,980 1968 1,988 556 493 350 270 265 233 - 2,005 1969 2,011 564 498 355 274 268 236 - 2,031 1970 2,035 571 504 360 278 272 239 - 2,055 1971 2,029 574 502 361 279 273 240 - 2,058 1972 2,022 576 500 362 280 274 241 - 2,060 1973 2,013 577 498 363 281 275 242 - 2,060 1974 2,002 578 495 364 281 275 242 - 2,059 1975 1,991 579 492 365 282 276 243 - 2,058 1976 1,994 580 488 365 282 276 243 - 2,054 1977 2,003 575 480 362 280 274 241 - 2,035 1978 1,975 571 472 359 277 271 239  2,014 1979 1,946 566 463 356 275 269 237 - 1,993 1980 1,917 560 454 353 272 267 235 0 2,392 1981 1,887 555 446 350 270 264 232 0 2,405 1982 1,856 550 437 346 267 261 230 106 2,320 1983 1,825 544 427 342 264 259 228 143 2,298 1984 1,793 538 418 339 261 256 225 135 2,329 1985 1,760 531 409 335 258 253 222 174 2,317 1986 1,727 525 399 331 255 250 220 206 2,317 1987 1,694 518 389 326 252 246 217 252 2,566 1988 1,660 506 379 319 246 241 212 242 2,302 1989 1,625 493 369 311 240 235 206 195 2,218 1990 1,591 480 359 302 233 228 201 152 1,959 1991 1,556 466 349 293 226 222 195 171 1,984 1992 1,520 451 339 284 219 215 189 199 1,735 1993 1,432 436 329 274 212 207 182 231 1,677 1994 1,376 428 318 270 208 204 179 343 1,690 1995 1,319 421 308 265 205 200 176 379 1,693 1996 1,261 413 297 260 201 197 173 431 1,774 1997 1,203 406 286 256 197 193 170 493 1,811 1998 1,179 404 284 254 196 192 169 616 1,927 1999 1,154 402 282 253 195 191 168 801 2,027 2000 1,130 400 279 252 194 190 167 862 2,118 2001 1,129 398 277 250 193 189 167 1,268 2,483 2002 1,136 396 275 249 192 188 166 1,189 2,381 2003 1,138 394 272 248 191 187 165 611 2,340 2004 1,139 392 270 247 190 186 164 182 2,264 2005 1,120 390 268 245 189 185 163 283 2,362 2006 1,111 388 265 244 188 184 162 414 2,536 2007 1,104 386 263 243 188 184 162 390 2,604  Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  131   RECONSTRUCTION OF MARINE FISHERIES CATCHES FOR TUVALU (1950-2009)1 Kendyl Crawforda, Sarah Harperb and Dirk Zellerb a Department of Marine and Environmental Science, Hampton University Hampton, Virginia 23668, USA b Sea Around Us Project, Fisheries Centre, University of British Columbia 2202 Main Mall, Vancouver, V6T 1Z4, Canada kendyl.crawleycrawford@my.hamptonu.edu; s.harper@fisheries.ubc.ca; d.zeller@fisheries.ubc.ca  ABSTRACT Tuvalu‘s total marine fisheries catches within its EEZ were reconstructed for the years 1950 to 2009. This reconstruction accounts for officially un- and underreported catches of artisanal and subsistence fishery sectors as well as the baitfish used in the pole-and-line tuna fishery. FAO data were used in combination with data from fish markets, regional reports and consumption data. Total reconstructed catches were estimated to be 69,631 t over the six decades, which is approximately 5 times  larger than the amount reported by the FAO on behalf of Tuvalu (12,241 t). Total catches increased from 813 t·year-1 in 1950 to 1,607 t·year-1 by 2009. The majority of total catches were from the subsistence sector (87%). This investigation reveals the need for an improvement in the accounting of marine fishes catches by all fisheries sectors. Due to the heavy rates of fish consumption in Tuvalu, reliable estimation of catches and resulting resource management decisions will play a role in Tuvalu‘s future food security. INTRODUCTION Tuvalu is an archipelago in the South Pacific consisting of nine atolls; Nanumea, Niutao, Nui, Vaitupu, Nukufetau, Nukulaelae, Niulakita and Funafuti. Tuvalu is located at 8° 31' S, 179° 13' E, approximately halfway between Australia and Hawaii in the south central Pacific (Figure 1). The country‘s total land area of 26 km2 is tiny in comparison to its nearly 752,000 km2 Exclusive Economic Zone (EEZ) (www.seaaroundus.org). Tuvalu is critically vulnerable to sea level rise due to its low-lying topography, with most of the country less than 3m above sea level (Connell, 2003; Sauni and Fay-Sauni, 2005; Rayfuse, 2011; Stephen, 2011). The continental shelf off Tuvalu is minimal; there are patch and fringing barrier reefs immediately surrounded by 1,000 m depths. The inner lagoons provide the only significant shallow water areas (Sauni and Fay-Sauni, 2005). Formerly a British colony known as the Ellice Islands, Tuvalu gained its independence in October 1978. The country has been politically stable and its economy has grown from an initial Tuvalu Trust Fund investment of $ 27 million Australian Dollars (AU$)  in 1987 to AU$ 66 million in 2002 (Gemenne and Shen, 2009). As of 2002, the Gross Domestic Product (GDP) of Tuvalu was AU$ 26.9 million, of which 8.2% was from the artisanal fishing industry (Gillett, 2009). The Tuvaluan economy is considered traditional and predominantly non-cash (Sauni and Fay-Sauni, 2005). Marine products, wages and  1  Cite as: Crawford, K., Harper, S. and Zeller, D. (2011) Reconstruction of marine fisheries catches for Tuvalu (1950-2009). pp. 131- 143. In: Harper, S. and Zeller, D. (eds.) Fisheries catch reconstructions: Islands, Part II. Fisheries Centre Research Reports 19(4). Fisheries Centre, University of British Columbia [ISSN 1198-6727] Figure 1: Map of Tuvalu and its Exclusive Economic Zone.  remittances sent by family members working overseas - many on foreign fishing vessels - are considered to be the most important sources of income (Sauni and Fay-Sauni, 2005). In the past ten years, annual revenue from foreign fishing fleets has varied, providing between 5.5% and 36.7% of total government revenue and grants (Gillett, 2009). In 1999, Tuvalu received USD$ 5.9 million in foreign fishing access fees. This amount comprised 42.6% of its GDP (Gillett and Lightfoot, 2002). The revenue from foreign tuna fishing is approximately 14.6% of the total value of the fish caught (Gillett, 2009). With Tuvalu receiving such a comparatively small percentage of the value of its fish, the government continues in their attempt to develop commercial fisheries in its vast EEZ, and stimulate economic growth. However, the development of larger-scale commercial fisheries in Tuvalu is hindered by high costs, difficulty raising funds and a lack of infrastructure required for fleet maintenance and operations, processing, internal distribution and export (Gillett, 2002). The largest component of Tuvalu‘s fishing activities is subsistence, i.e., for direct consumption. More than 80% of domestic coastal catch in Tuvalu is produced by subsistence fishing (Gillett, 2010). Dalzell (1996) estimated that subsistence and artisanal fisheries make up 87% and 13%, respectively, of Tuvaluan coastal fisheries production. A recent Household Income and Expenditure Survey revealed that fishing contributes to 8% of personal income, after wages and remittances sent from overseas (Anon., 2006). Data on the artisanal sector is incomplete due to the high occurrence of informal bartering (Lambeth, 2000; Poulasi, 2008; WCPFC, 2009, 2010). Fishermen or their wives sell their catch either from home, the roadside using handcarts and ice chests or in small markets (Lambeth, 2000; Gay, 2010). As a result, catch data for the subsistence and artisanal fisheries are largely unknown. As more Tuvaluans try to make the change to working for an income, the artisanal sector is assumed to grow (Sokimi and Chapman, 2005).  The heavy dependence on fish for animal protein is evident through the unique Tuvaluan word ‗miti’ which describes a craving specifically for fish. In 2004 and 2005, the urban per capita fish consumption in Tuvalu was 68.8 kg·person-1∙year-1 with 97% of that amount being fresh fish. For rural areas, the per capita consumption was 147.4 kg∙person-1·year-1 (99% being fresh fish) (Gillett, 2009). Tuvalu‘s seafood consumption rate is among the highest in the world (Gillett and Lightfoot, 2002; Gillett et al., 2001). The island communities of Tuvalu are distinguished not only for their fishing ability, but also their rich knowledge of their environment (Gay, 2010). The close relationship between Tuvaluans and the ocean is readily apparent through their dependence on fish for food security. The island of Niutao, one of the main islands of Tuvalu, has one of the highest population to reef area densities in the region, with 246 people per km2 of reef (Adams et al., 1996). At the same time, on the atoll of Funafuti, there was a lower population to reef area density of 165 people per km2 of reef (Adams et al., 1996). However, Funafuti is now home to approximately 5,000 people, 47% of the country‘s population (Sauni et al., 2008). Even though the urban island has set up management measures such as the 33 km2 Funafuti conservation area, it is highly urbanized and as such faces problems such as sewage treatment and waste disposal. There is heavy fishing pressure from subsistence needs on this island with many people fishing after work and on the weekends (Sauni et al., 2008). It is estimated that 93% of households on Funafuti eat fresh fish that they catch and 70% of households eat fresh invertebrates that they catch (Sauni et al., 2008). Fish catch rates have increased when compared to estimates from previous years (Sauni and Fay-Sauni, 2005).  Thus there is concern for the sustainability of inshore resources in light of the increased fishing pressure and population growth on Funafuti (Adams et al., 1996; Gillett, 2002; Sauni and Fay-Sauni, 2005; Aylesworth and Campbell, 2009).  Fishing in Tuvalu uses a range of techniques including pole-and-line, trolling and reef gleaning, which are used to collect finfish, bivalves, crustaceans and other invertebrates in nearshore and offshore Tuvaluan waters. The fishing roles on Tuvalu, like many other Pacific Islands are divided by gender, with women mainly reef gleaning at low tide, and processing, and men fishing both inshore and offshore. The introduction of outboard engines on canoes in the 1960s and 1970s has made fishing much easier, and consequently, women have felt less of a need for their auxiliary fishing activities (Lambeth, 2000). However, when men are unable to fish because of the weather, women‘s collecting activities are vital (Chapman, 1987). While fishing techniques vary among the different islands, the main gears used are gillnetting, handlining, castnetting, pole-and-line and spearfishing. Handlines are used to catch demersal fish on the reef (Sauni and Fay-Sauni, 2005). On the outer reefs, spears, handlines, scoop nets and deep bottom methods such as deep-bottom droplining are used (Chapman and Cusack, 1990; Sauni et al., 2008). As of 1991, there were an estimated 200 motorized and 500 non-motorized vessels, most less than 10 m (Gillett, 2003). On the most populated island of Funafuti, there are 10 to 20 commercial vessels (4-5 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  133   0 2 4 6 8 10 12 1950 1960 1970 1980 1990 2000 P o p u la ti o n  (x  1 03 ) Year Nukufetau Vaitupu Funafuti All OthersNiutao Figure 2. Human population of Tuvalu by major islands, 1950- 2009. m) engaged in trolling for mainly skipjack and yellowfin tuna (Gillett, 2003) and some line fishing for reef species (Gillett, 2002).  Fisheries in Tuvalu are dominated taxonomically by skipjack (Katsuwonus pelamis) and yellowfin tuna (Thunnus albacares), which in 1978 made up half of the total fish catch (Gillett et al., 2001).  A more recent estimate suggests skipjack and yellowfin represent approximately 75% of total fish landings (Gillett, 2002). Other pelagic fish such as flying fish (Cypselurus spp.) also represent a substantial part of the catch. Flying fish, usually caught at night with the use of scoop nets and lights (Gillett, 2002), are commonly used as baitfish in the tuna pole-and-line fishery (Aylesworth and Campbell, 2009). Fish inhabiting the lagoon and reef habitats such as red snapper (Lutjanus gibbus) are also caught but make up a lesser portion of the catch. Bycatch is generally consumed, bartered or given away as a part of the fishery. In the early 1980s, the government established the National Fishing Corporation of Tuvalu (NAFICOT) tasked with the goal of developing industrial fisheries (Sokimi and Chapman, 2005; Gillett, 2011a). One of the roles of NAFICOT was to manage national fishing vessels. In 1982, Japan donated a pole-and-line vessel, in 1989, Japan donated 7 additional vessels (6 launches and 1 extension vessel), and in 2004, Korea donated two longliners. Community fishing centers (CFC) were developed in the 1990s with foreign aid to provide an income opportunity for fishers and also to redistribute the excess supply of fish to the urban population center of Funafuti through NAFICOT. The CFCs provide salting, drying and at times cold storage facilities. However, ice is not usually present in many of the CFCs or on local fishing vessels because of a lack of transportation infrastructure and water availability (Aylesworth and Campbell, 2009). Presently, most CFCs have fallen into disrepair and rely heavily on government subsidies. In 2009, NAFICOT went bankrupt when the government decided to cease financial support (Gay, 2010). This same year, a joint venture was established between Tuvalu and Taiwan; the first purse seiner flying the Tuvaluan flag, the FV Taumoana began fishing in August of that year in Tuvalu, FSM and Kiribati, landing a total of 4,877 t of tuna, most likely skipjack and yellowfin (WCPFC, 2010).  The purpose of this study is to reconstruct total marine fisheries catches by Tuvalu within its EEZ between 1950 and 2010 by accounting for all fishing sectors, as a baseline for the assessment of food security and resource availability. During the completion of this study, it was necessary to make assumptions in order to fill large gaps in data availability because the official reported data includes neither taxonomic specificity nor quantitative detail, especially for the pre-1980 time period. MATERIALS AND METHODS The Tuvalu Fisheries Department collects offshore commercial catch data for the national fleet. In addition, the department actively collects monthly reports from various fish markets reporting sales and purchasing information. In 2010, Tuvalu started a national catch database for their inshore fisheries (T. Poulasi, pers. comm., Tuvalu Fisheries Department). Unfortunately, data from 2010 was not yet available at the time of this study. Catch data other than for the national fleet for all years prior to 2010 were not collected by the national Fisheries Department (T. Poulasi, pers. comm., Tuvalu Fisheries Department). The artisanal data collected from fish market reports are limited, as they encompass only the small amount of catch sold within the markets (WCPFC, 2009), whereas much catch is sold informally. Data on subsistence fishery catches have not been collected at all. The national data are considered insufficient for the evaluation and monitoring of fishing activities (Chapman, 2004). However, recent attempts  have been made to improve data collection.  To provide the best picture of Tuvaluan fisheries, data from FAO FishStat, regional reports and independent assessments were accessed and used to develop data anchor points for the estimation of total catches. Interpolations between anchor points were used to derive a complete time series, using a catch reconstruction approach developed by Zeller et al. (2006; 2007). Human Population Data Population data were obtained in order to convert available per capita consumption rates into an estimate of overall demand. National population data were obtained from Populstat (www.populstat.info/ [date accessed: 20 July 2011]) prior to 1997, from the 2008 Biannual Statistical Report (Anon., 2008) for years 2002 to 2007 and from Index Mundi (www.indexmundi.com [date accessed: 20 July 2011]) for 2009. A linear interpolation was used between years of known population data in order to obtain a complete time series. Population data by island were obtained for Funafuti, Niutao, Nukufetau and Vaitupu from Populstat for 1985, 1987 and 1996, a Household Income and Expenditure Survey for 2005 (Anon., 2006) and City Population (www.citypopulation.de [date accessed: 20 July 2011]), for 1979, 1991 and 2002. Over the past 50 years, migration to Funafuti from the other islands has resulted in nearly 50% of the current total population residing on Funafuti, driven mainly by the availability of government jobs (Connell, 2003). The earliest population data found for Funafuti was for 1963 obtained from Populstat. The percentage of the total population which resided in Funafuti in 1963 was applied back in time to 1950. From 1964 to 2005, the population of Funafuti was determined through interpolation of data points. For Niutao, Nukufetau, Vaitupu and all others, the earliest year of data was 1979. Therefore, we calculated the proportion that each of these islands represented and applied these same proportions back to 1950. After 1979, interpolation was done between the island population data. From 2005 to 2010, all island populations were calculated based on 2005 percentages (Figure 2). Fishing in Tuvalu The FAO FishStat database was used as the official catch data for Tuvalu (Figure 3). However, FAO data are presented by FAO area and do not delineate the amount taken within EEZ areas. This report aims to reconstruct the catches taken by Tuvaluans within their waters; therefore, it was necessary to disaggregate the catch taken in Tuvalu‘s EEZ from that taken outside. A significant component of fishing recorded in the FAO data occurred outside of Tuvalu‘s EEZ. For example, in 1982, Tuvalu fished in a partnership with the Ika Corporation in Fijan waters (SPC, 1994). Tuvalu also fished in Fiji and Solomon Islands in 1987 and 1988 with a peak catch of 1,091 t from the Solomon Islands in 1988 (SPC, 1994; Sauni et al., 2008). In 2009, Tuvalu in a joint venture with the Fong Haur fishing company of Taiwan fished in the Federated States of Micronesia and in Kiribati (WCPFC, 2010). Due to the large proportion of fishing known to occur outside of Tuvalu‘s EEZ (SPC, 1994; Apinelu, 2004; Sauni et al., 2008; WCPFC, 2010), we assumed 90% of the FAO tuna catch (bigeye (Thunnus obesus), skipjack (Katsuwonus pelamis), yellowfin (Thunnus albacares) and tuna-like) was taken outside of Tuvalu‘s waters (Figure 3). Remaining FAO landings (ie., marine fishes nei) were assumed to have been from small-scale fisheries within Tuvalu‘s EEZ.    Figure 3. Catches presented by the FAO on behalf of Tuavlu and allocation of FAO data to the Tuvaluan EEZ based on assigning 90% of reported tuna catches  to be taken outside EEZ waters, 1950-2009. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1950 1960 1970 1980 1990 2000 C at ch  (t  x  1 03 ) Year Total FAO Allocated to EEZ Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  135   Small-scale fisheries  FAO FishStat presents landings of miscellaneous marine fishes from 1950 to 2009. Prior to 1982, FAO landings for Tuvalu are less than 0.5 t·year-1 and landings for tuna and tuna-like fishes during this period are zero. Catch amounts or rates from national or independent sources were not readily available for either the artisanal or subsistence sectors; therefore per capita consumption rates were used to determine the overall fresh fish demand for Tuvalu (Gillett and Lightfoot, 2002; Sauni et al., 2008; Gillett, 2009). This overall demand was compared to the reported supply to determine the magnitude of underreporting. For 2004, national fresh fish consumption in Tuvalu was calculated using consumption rates for the islands of Funafuti, Nukufetau, Niutao and Vaitupu (Sauni et al., 2008) (Table 1). For the remaining five islands not represented individually, here referred to as ‗all others‘, the average across the four individual islands (i.e., 151.0 kg·person-1·year-1) was used (Sauni et al., 2008). To derive a nation-wide consumption rate, a 2004 weighted average for all islands was calculated (145 kg·person-1·year-1). This rate was used for all of Tuvalu for 2004 and was carried forward, unaltered to 2010 (Gillett and Lightfoot, 2002; Sauni et al., 2008). The 1950 fresh fish consumption rate was assumed to be similar to the rate for the other islands in 2004 (see All others Table 1) with the addition of 5 kg to account for limited imports of protein alternatives available in the later period. The resulting per capita consumption rate for 1950 was 155.6 kg·person-1·year-1. A linear interpolation between these two anchor points between the 1950 and the 2004 per capita rates was used to complete the time series. The consumption rates were then combined with the human population data to estimate overall demand of fresh fish.  To determine the proportion of the total demand supplied by each sector, Dalzell‘s (1996) breakdown of 13% artisanal and 87% subsistence was used for 1996 to 2009. In 1950, we assumed 100% of the catch to be subsistence and interpolated to 87% subsistence in 1996. The annual artisanal production estimates were then checked to ensure that they exceeded the artisanal landing amounts given in annual reports by the Western and Central Pacific Fisheries Commission (Poulasi, 2008; WCPFC, 2009, 2010). This was necessary as the WCPFC reports reflect only the small fraction of catch landed and received by the CFCs. Fishers normally sell their catch directly to consumers; these transactions are not documented (Poulasi, 2008). Invertebrates The people of Tuvalu also consume invertebrates as part of their diet. However, unlike for fresh fish, there is no local commercial fishery for invertebrates (Sauni et al., 2008). Thus, an annual subsistence invertebrate demand was calculated with invertebrate consumption data from the Secretariat of the Pacific Community Coastal Fisheries Program (www.spc.int/coastfish/ [date accessed: 15 July 2011]) (Table 2). Similar to the fresh fish consumption data, invertebrate consumption data were available for Funafuti, Nukufetau, Niutao and Vaitupu. A national average per capita invertebrate consumption rate was calculated to represent the remaining islands. To determine the annual invertebrate demand, the same methodology was used as in determining total demand of fresh fish (see Small-scale fisheries). The weighted average consumption rate for invertebrates (3.8 kg·person-1·year-1) was used for 2004 and all years following. For 1950, the national average was used with an adjustment of 2 kg to account for the limited availability of alternate protein sources at that time (i.e., 5.8 kg·person-1·year-1). An interpolation was done between the 1950 (5.8 kg·person-1·year-1) and 2004 (3.8 kg∙person-1·year-1) anchor points and the 2004 rate was carried forward unaltered to 2010. Taxonomic breakdown A substantial portion of both the subsistence and artisanal catch consists of tuna (Gillett et al., 2001). A thorough review of the scientific and grey literature provided numerous estimates of the tuna component Table 1.  Fresh fish consumption rates per island for 2004 (Sauni et al., 2008). Island Fresh fish consumption rate (kg/person/year)a Funafuti 135.0 Nukufetau 117.8 Niutao 185.3 Vaitupu 162.5 All Others 150.6 aweighted average for all islands 145 kg∙person-1∙year-1. Table 2.  Invertebrate consumption rates per island 2004. Island Invertebrate consumption rate (kg/person/year) Funafuti 4.6 Nukufetau 5.6 Niutao 3.6 Vaitupu 0.9 All Others 3.7  of the Tuvaluan catch. Sauni (2008) presents annual tuna landings data for the 1970s of 350 t·year-1. This amount, however, is viewed to be slightly overestimated (Eginton and Mead, 1978).  Gillett (2001) reports that 50% of fish sold in Funafuti is tuna. The Pacific Island Fisheries Regional and Country Information report suggested that 75% of all fish landings are ocean species, mainly skipjack and yellowfin tuna (Gillett, 2002). Based on these information sources, an estimate of 300 t was used as an anchor point for 1975. This represents 32.5% of the total demand for 1975. We assumed that tuna consumption was similar in 1950, and allocated 32.5% of the total 1950 catch to tuna. The species composition for the tuna was derived from yearly reports provided by the Tuvaluan government to the Western and Central Pacific Fisheries Commission (Poulasi, 2008; WCPFC, 2009, 2010). The proportion of skipjack to yellowfin tuna represented in the catch were calculated by averaging annual catch data derived from fish market sales for the years 2003 to 2008. To account for other pelagic species, a small portion of the tuna catch (arbitrarily assigned as 10%) was allotted to miscellaneous large pelagic. The remaining 90% were assigned as 54% skipjack and 36% yellowfin tuna (Table 3). Catches were also broken down taxonomically according to fishing sector, either subsistence or artisanal. For the artisanal sector, species compositions given by Dalzell (1996) were used to taxonomically disaggregate all families aside from Scombridae (Appendix A1). The Etelinae subfamily (Family Lutjanidae) was divided into Etelis spp. and Pristipomoides spp. with each genus receiving half of the Etelinae percentage, to improve taxonomic resolution. Scombridae was excluded, because the tuna component of the artisanal fishery had already been calculated. The invertebrate and fresh fish demands were kept separate for the subsistence sector. The taxonomic breakdown for the subsistence fishery was determined through fish (Appendix Table A2) and invertebrate (Appendix Table A3) catch compositions available in a national report (Sauni et al., 2008). Species catch composition data were available for the islands of Funafuti, Niutao, Nukufetau and Vaitupu for several different habitats such as lagoon, outer reef, sheltered coastal reef and intertidal reef flat. Some species groups that composed a very small percentage of the total catch were grouped and represented at the family level. The invertebrate catch compositions were applied to the invertebrate catch and the fish compositions applied to the fish catch derived from the demand estimates.  Baitfish  Associated with the pole-and-line fishery for tuna is the use of baitfish, rarely accounted for in fisheries statistics. Baitfish fisheries often operate in parallel to the skipjack pole-and-line fisheries, utilizing the same vessels. Tuvalu‘s baitfish resources are extremely limited (Gillett, 2011b). In the past, Tuvalu‘s only pole-and-line fishing vessel, Te Tautai, was forced to fish outside of the Tuvaluan EEZ because of the low availability of baitfish (Gentle, 1991).In addition, baitfish resources are more variable around atolls than high islands (Anon., 1984), and atolls like Tuvalu have been less reliable sources of bait. A regional pole- and-line ratio of tuna to baitfish was presented by Gillett (2011b) as 32:1. The dominant species used as baitfish in Tuvalu are Spratelloides delicatulus, Archamia lineolata, Bregmaceros spp. and Atherinomorous lacunoaa (Anon., 1984).  The amount of baitfish used was calculated by taking the estimated tuna catch and applying the 32:1 tuna to baitfish ratio provided by Gillett (2011b) for the region and the taxonomic composition derived from the Tuna Programme (Anon., 1984) was applied (Table 4).  Table 4. Taxonomic breakdown of baitfish for Tuvalu. Source: (Anon., 1984) Group Taxon  Common name Proportion of catch (%) Clupeidae Spratelloides delicatulus Delicate round herring 92 Atherinidae Atherinomorus lacunosus Hardyhead silverside 2 Misc. marine fishes Misc. marine fishes Misc. marine fishes 6  Bêche-de-mer As sea cucumber are not a part of the Tuvaluan diet, they have traditionally received little interest by fisheries managers (Belhadjali, 1997). However, a small bêche-de-mer export industry has developed in Table 3. Taxonomic breakdown of tuna for Tuvalu.  Source: (Anon., 1984) Group Taxon  Common name Proportion of catch (%) Scombridae Katsuwonus pelamis Skipjack tuna 54 Scombridae Thunnus albacares Yellowfin tuna 36 Miscellaneous large pelagics Misc. pelagics Misc. large pelagics 10 Fisheries catch reconstructions: Islands, Part II. Harper and Zeller  137   Tuvalu and has become the island‘s main fishery export item since the late 1970s (Gay, 2010). The industry began after the Fisheries Department became the recipient of UN Development Program funding in 1978. The first export occurred in 1979, 1.8 t of bêche-de-mer sold to Fiji (Belhadjali, 1997). With widely varying production amounts, export continued from 1980 to 1983 (Belhadjali, 1997). Remaining stagnant for a decade, export restarted from 1993 through 1995 supplying Singapore and Fiji. In 2007, harvest in Nukufetau, Nukulaelae and Funafuti began once again with exports to Hong Kong. Only the islands of Funafuti and Nukufetau have suitable habitats for the most profitable sea cucumbers, therefore the sustainability of the bêche-de-mer fishery is of concern (Gay, 2010). The Fisheries Department does not require data on bêche-de-mer catch or exports to be submitted  (Gay, 2010). Export weights, nevertheless, were found for all years prior to 2007. Gillett (2009) provided a bêche-de-mer export value of AU$5000 for 2007. From 2007 to 2009, that same export value was used. The monetary value was converted to USD using 2007 exchange rates and the export amount was calculated using the average dollar value per dried kilogram from the years 1993 to 1995 (Belhadjali, 1997). In the processing of sea cucumbers, 90% of their body weight is lost (Dalzell et al., 1996). Thus, all dried weights were converted back to live weights with a conversion factor of ten to represent the bêche-de-mer catch as a component of the total reconstructed catch.   RESULTS  Total landings presented by the FAO on behalf of Tuvalu for the 1950- 2009 time period were 32,255 t, with essentially 0.25 t·year-1 reported until 1977, then fluctuating around 600 t·year-1 until the early 2000s when landings increased to 4,198 t by 2009.  In contrast, reconstructed data for Tuvaluan marine fishes catch taken within the EEZ were over 67,000 t since 1950, increasing from around 800 t·year-1  in 1950 to about 1,600 t·year-1 by 2009 (Figure 4). The artisanal sector accounted for just under 6,000 t of fish over the 1950-2009 period, whereas the subsistence sector accounted for over 61,000 t (Figure 5). The most commonly caught families in the subsistence sector were Lethrinidae and Serranidae. The most common species, however, were the blue sea chub (Kyphosus cinerascens), the humpback red snapper (Lutjanus gibbus) and the bluespot mullet (Valamugil seheli). Carangidae, Gempylidae and Lutjanidae were the most common families caught in the artisanal sector. FAO presented catches of tuna and tuna-like Table 5. Bêche-de-mer catch from 1950-2009. Year Catch (t) 1979 18.00 1980 8.05 1981 0.90 1982 1.99 1983-1992 0.00 1993 8.71 1994 36.78 1995 32.28 1996-2006 0.00 2007 3.24 2008 3.24 2009 3.24 Figure 4. Total reconstructed catch from 1950-2009 for Tuvalu compared to the subset of FAO data assigned to Tuvalu EEZ. 0 200 400 600 800 1000 1200 1400 1600 1800 1950 1960 1970 1980 1990 2000 C at ch  (t ) Year Reported FAO within EEZ Total Reconstructed 0 200 400 600 800 1000 1200 1400 1600 1800 1950 1960 1970 1980 1990 2000 C at ch  (t ) Year Subsistence Artisanal Figure 5. Total reconstructed EEZ catches presented by fishing sector, 1950-2009.  fishes from 1982 onward with a total of 2,223 t. Our reconstruction suggested that 12,438 t of tuna was caught between 1982 and 2009. From 1950 to 1981, 9,347 t of tuna were estimated as part of the reconstruction. A total of 680 t of baitfish were caught for use in the pole-and-line tuna fishery, 92% of which were delicate round herring (Spratelloides delicatulus). There were no inverte