UBC Community and Partners Publications

Total marine fisheries extractions by country in the Baltic Sea: 1950-present Rossing, Peter; Booth, Shawn; Zeller, Dirk 2010

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


FCRR_2010_18-1.pdf [ 6.89MB ]
JSON: 1.0074744.json
JSON-LD: 1.0074744+ld.json
RDF/XML (Pretty): 1.0074744.xml
RDF/JSON: 1.0074744+rdf.json
Turtle: 1.0074744+rdf-turtle.txt
N-Triples: 1.0074744+rdf-ntriples.txt
Original Record: 1.0074744 +original-record.json
Full Text

Full Text

 ISSN 1198-6727  Fisheries Centre Research Reports  2010 Volume 18 Number 1    TOTAL MARINE FISHERIES EXTRACTIONS BY COUNTRY IN THE BALTIC SEA:  1950-PRESENT    Fisheries Centre, University of British Columbia, Canada   TOTAL MARINE FISHERIES EXTRACTIONS BY COUNTRY IN THE BALTIC SEA:  1950-PRESENT     Edited by  Peter Rossing, Shawn Booth and Dirk Zeller                 Fisheries Centre Research Reports 18(1) 263 pages © published 2010 by  The Fisheries Centre, University of British Columbia  2202 Main Mall Vancouver, B.C., Canada, V6T 1Z4       ISSN 1198-6727   Fisheries Centre Research Reports 18(1) 2010   TOTAL MARINE FISHERIES EXTRACTIONS BY COUNTRY IN THE BALTIC SEA:  1950-PRESENT  Edited by Peter Rossing, Shawn Booth and Dirk Zeller   CONTENTS  Director’s Foreword .............................................................................................................. 1 Executive Summary ............................................................................................................... 3 Fisheries catches from the Baltic Sea Large Marine Ecosystem: 1950-2007 .......................... 7 Denmark’s marine fisheries catches in the Baltic Sea (1950-2007) ..................................... 39 Catch reconstruction for Estonia in the Baltic Sea from 1950–2007 ................................... 63 Baltic Sea fisheries catches for Finland (1950-2007) ........................................................... 85 Germany’s marine fisheries catches in the Baltic Sea (1950-2007) ................................... 107 Catch reconstruction for Latvia in the Baltic Sea from 1950-2007 ..................................... 127 Catch reconstruction for Lithuania in the Baltic Sea from 1950-2007............................... 145 Poland’s fisheries catches in the Baltic Sea (1950-2007) ................................................... 165 Russian fisheries catches in the Baltic Sea from 1950-2007 .............................................. 189 Sweden’s fisheries catches in the Baltic Sea (1950-2007) .................................................. 225   A Research Report from the Fisheries Centre, UBC and the Baltic Sea 2020 Foundation     Fisheries Centre Research Reports 18(1) 263 pages © Fisheries Centre, University of British Columbia, 2010  FISHERIES CENTRE RESEARCH REPORTS ARE ABSTRACTED IN THE FAO AQUATIC SCIENCES AND FISHERIES ABSTRACTS (ASFA) ISSN 1198-6727    Total marine fisheries extractions by country in the Baltic Sea: 1950-present, Rossing, Booth and Zeller 1   DIRECTOR’S FOREWORD  Illegal, Unreported and Unregulated (IUU) fisheries, which only a few years ago were considered to be mainly a problem for a limited number of regions or fisheries (e.g. Patagonian toothfish), have now been recognized as a substantial global issue of concern. The importance and negative impacts of IUU fisheries catches have gained increasing attention in the world‘s media, and also by fisheries scientists and managers. To account for IUU, catch reconstructions, such as those being conducted under the guidance of Dr. Dirk Zeller of the Sea Around Us Project, and documented in a previous Fisheries Centre Research Report (Vol. 15 (2), 2007) and in numerous peer-reviewed publications, show that fisheries statistics supplied by national, regional and international scientific and management agencies generally underestimate actual catches often by substantial margins. While historically perceived to be largely a developing country problem, the increasing prevalence of vessel apprehension illustrates that IUU catches are also prominent in some of the most developed countries of the world. The countries surrounding the Baltic Sea are predominantly members of the European Union, and have a long history of marine resource use in the Baltic Sea. Yet, as the study presented in this report illustrates, even these highly developed countries with their substantial resources and well established scientific, administrative and management institutions, have so far failed to address the data issues stemming from IUU in a transparent and comprehensive manner. The effort reported in this report, conducted through funding from the Baltic Sea 2020 Foundation (www.balticsea2020.org/), should contribute to more transparent and complete accounting of total catches for Baltic Sea fisheries, and may even serve as a blueprint for all other North- east Atlantic areas.  In general, obtaining a complete accounting of total catches (or removals) from the global ocean is fundamental to our ability to manage marine fishery resources sustainably for the benefits of both current and future generations. The work reported herein continues an effort by the Sea Around Us Project, through its catch reconstruction work, to provide such accounting.  Ussif Rashid Sumaila, Director UBC Fisheries Centre February 2010 2 Total marine fisheries extractions by country in the Baltic Sea: 1950-present, Rossing, Booth and Zeller   Total marine fisheries extractions by country in the Baltic Sea: 1950-present, Rossing, Booth and Zeller 3  EXECUTIVE SUMMARY  Illegal, Unreported and Unregulated (IUU) catches are a widespread source of underreporting in fisheries, including in the Baltic Sea, where reported landings from commercial fisheries have been the only form of landings officially recorded (although stock assessments try to account for some IUU catches). According to the European Court of Auditors, the incompleteness and unreliability of catch data have prevented the Total Allowable Catch (TAC) and quota system in the Baltic from functioning properly. In addition, the regulatory and institutional frameworks have guaranteed neither the exhaustiveness of data collection, nor the detection of inconsistencies during validation. Consequently, the International Council for the Exploration of the Sea (ICES) advisory processes related to managing Baltic Sea stocks have only partially satisfied the requirements of formal analysis. Policy makers therefore often underestimated the impact of fishing, contributing to the decline seen in some Baltic fisheries, threatening not only important commercial stocks like cod, but also ecosystem functions and the future economic viability of the fishing industry. This study estimated total fisheries catches by the countries bordering the Baltic Sea using a catch reconstruction approach for the period 1950-2007. There is a need for a better understanding of the impact of fisheries on marine ecosystems, by providing a comprehensive time series of total fisheries catches. This will contribute to more effective collaboration between the fishing industry, scientists, policy makers, NGOs and the general public.  To estimate the total catch (as opposed to reported landings) from 1950 to 2007, four IUU catch components were estimated for the nine coastal Baltic countries (Denmark, Estonia, Finland, Germany, Latvia, Lithuania, Poland, Russia and Sweden): ‗adjustments’ to ICES landings statistics (i.e., data source adjustments), ‗unreported landings‘, ‘discards’, and ‗recreational catches‘. ICES landings statistics were used as the ‗reported data’ baseline for our reconstruction, since they are the only publicly-available data, covering all taxa, fishing areas and countries in the Baltic Sea back to 1950. Estimated unreported landings and discards were applied to landings data on a taxonomic and country-specific basis. Discard estimates included ‗underwater‘ discards from actively-fishing trawl gear; ‗ghostfishing‘ due to lost or abandoned fishing gear; ‗boat-based’ discards, generally resulting from fishers‘ intentional behavior; and ‗seal- damaged’ discards representing catch lost because of seal damage. The inclusion of recreational catch estimates for each country allowed estimates of the likely total catch (as opposed to reported landings) to be derived. Information sources included a wide range of primary and secondary sources, including interviews and collaborations with fisheries experts in Baltic countries, peer-reviewed and grey literature, ICES online databases and publications, and national media sources. The philosophy behind reconstruction is to utilize all available data and information to derive ‗anchor points‘ of IUU catches in time and to use interpolations for time periods between anchor points. To account for some historical differences between countries, we grouped countries into ‗western‘ (Denmark, Finland, West Germany [1950-1990], Germany [after re-unification, 1991-2007] and Sweden) and ‗eastern‘ (former ‗eastern bloc‘: East Germany [1950- 1990], Estonia, Latvia, Lithuania, Poland and Russia). Overall, for the 1950-2007 period, our catch reconstruction estimated that IUU catches added approximately 30% to landings officially reported by ICES landings statistics (Figure 1). When catches peaked in 1997, total catches were likely 43% higher than reported landings, while for the 2000s (2000- 2007), total catches were on average 35% (i.e., about 280,000 t∙year-1) higher than reported landings. Comprehensive accounting of IUU catches provides an improved baseline for managing fisheries in the Baltic Sea, and aids the transition to ecosystem-based management.  0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 1950 1960 1970 1980 1990 2000 Cat ch (  t x 10  6 ) Year Total reconstructed catch ICES landings Figure 1. Total reconstructed catch (t) for the Baltic Sea, compared to reported landings from the ICES catch statistics database (1950-2007). 4 Total marine fisheries extractions by country in the Baltic Sea: 1950-present, Rossing, Booth and Zeller   Our results indicated that unreported landings were the greatest source of IUU catches. This highlighted significant management and enforcement issues, especially as related to unreported landings of cod, herring and sprat taken mainly by Denmark, Poland and Sweden. Unreported landings by all countries added 14% to officiallyreported landings for the entire period, and about 24% per year (i.e., 189,000 t∙year-1) since 2000. Unreported landings of cod were the most substantial, representing about 35% of unreported landings of all species since 2000 (Figure 2). A main contributor appears to have been unreported activities in Poland (see below). For 2000-2007, our reconstruction estimated that Poland was accountable for approximately 74% of all unreported landings of cod in the Baltic Sea. Unreported landings of herring by all countries were also significant, more recently (2000-2007) accounting for 33% per year of the unreported landings of all taxa (Figure 2).  Discards were also important. Overall, they added about 10% to reported landings. Denmark was found to be responsible for the majority of discards between 1950-2007, accounting for approximately 31% per year of total discards, while Germany, Poland and Sweden accounted for approximately 13% each. Discards were dominated by cod and herring, accounting for around 36% and 31% of total discards, respectively (1950-2007). More recently (2000-2007) cod accounted on average for 16% of total discards per year. Discards of flatfishes were also of concern as flatfish discards added 36% to the reported landings of flatfishes between 1950-2007. Since 2000, this has increased to 54%. Recreational catches totaled approximately 1.5 million tonnes (29,000 t∙year-1 since 2000) and comprised about 3% of our total catch reconstruction. Of the main commercial species, cod was the most important recreational target, accounting for 29% of all recreational catches since 2000, while generally non- commercial species such as European perch, Northern pike, whitefishes and sea trout accounted for 49% of recreational catches.  Adjustments to the ICES landings statistics, dealing solely with accounting problems and omissions of officially-reported data, added about 3% to ICES reported landings. The majority of these adjustments were country-specific accounting artifacts due to the dissolution of the USSR (see individual reports on Estonia, Latvia, Lithuania and Russia), rather than substantial missing data. National highlights for 2000-2007:  Denmark: The reconstructed catches were on average 33% higher than reported landings. Cod, herring and sprat accounted for nearly 70% of the difference between reported landings and total catch. Unreported landings and discards were the two biggest IUU components, accounting for 11% and 9% per year. An average discard rate of 43% made flatfishes the most underreported group, relative to flatfish reported landings. Estonia: The reconstructed catches were 25% higher than reported landings. Herring and sprat dominated Estonia‘s catches, accounting for 92% of the difference between reported landings and reconstructed total catches. While cod catches were small compared to other countries, cod were still underreported by around 100% relative to ICES reported landing. Herring and sprat accounted for 96% of the estimated total discards. For recent years, Estonia had good publicly-available data regarding discard and recreational catches. Finland: The reconstructed total of 135,000 t∙year-1 was 24% higher than the reported landings. Herring and sprat accounted for over 90% of the difference between reported landings and reconstructed total catches, as cod is of minor importance in Finland. Credit must be given to Finland‘s inclusion of recreational catches in the country‘s official landings statistics back to 1950. Germany: The reconstructed catches were 29% higher than reported landings. Cod, sprat and herring made up 81% of the difference between reported landings and reconstructed total catches. While flatfishes only represented around 7% of the total reconstructed catch, they accounted for 28% of total discards. Recreational cod catches were significant, equivalent to 19% of reconstructed cod 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 Catc h ( t  x 10  5 ) Year Herring Sprat Flatfish Others Cod Figure 2. Unreported landings (t) estimated for cod, herring, sprat, flatfish and 'others' for the Baltic Sea from 1950-2007. Total marine fisheries extractions by country in the Baltic Sea: 1950-present, Rossing, Booth and Zeller 5  Latvia: The reconstructed catches were more than 26% higher than reported landings. Sprat and herring accounted for 74% of the difference between reported landings and total reconstructed catches. Cod were the most underreported taxa, with approximately 88% of cod catches added as IUU catches, mainly due to discarding. Flatfishes were second-highest, with an annual average of 78% of catches added as IUU, mainly due to unreported landings.  Lithuania: The reconstructed catches were on average 44% higher than reported landings. Cod and sprat accounted for 61% and 21% of the total IUU, respectively. Cod represented, on average, 65% of the estimated unreported landings per year. Poland: The reconstructed catches were about 80% higher than reported landings. Cod, herring and sprat accounted for 93% of the difference between reported landings and reconstructed total catches. Unreported landings of cod, herring, and salmon were estimated to be substantial, on average equivalent to approximately 300%, 100%, and 55% of reported landing per year, respectively.  Russia: The reconstructed catches were over 25% higher than reported landings. On average, sprat, herring and cod accounted for 45%, 18% and 20% of total IUU catches per year. Access to data was generally a problem for Russia.  Sweden: The reconstructed catches were 20% larger than reported landings, despite herring and sprat catches (combined) being overreported by, on average 7% per year. Unreported landings were the largest component, averaging 23,000 t∙year-1 and accounting for about 12% per year of total catches. Of these, 98% were due to cod, herring, and sprat. Discards and recreational catches averaged 6% and 4% of total catches, respectively. Discarding was substantial for some species, e.g., flounder discards were more than 1.5 times larger than their reported landings.  The major problem encountered in our reconstruction efforts was a general lack of transparency in published fisheries data from ICES stock assessment working group reports (WGRs). WGRs publish aggregate estimates of IUU (including discards and unreported landings [unallocated catches]). It was not transparent which countries did, or did not contribute to these estimates, and in what proportions. The policy of ICES is to keep these data confidential, in an apparent effort to avoid political problems. We were also not granted access to the relational database ‗FishFrame‘, which contains Baltic-wide estimates for discards. Such non-transparency reduces accountability of managers, enforcement officers, and fishers, whose IUU catches are thus rarely exposed, and hinders the establishment of public trust. In 2012, the Common Fisheries Policy (CFP) is set to be renewed, which provides an opportunity to address some of the issues pertaining to accountability and transparency in fisheries data and practices. A high priority should be more accountability for total catches, as opposed to reported landings. For example, Denmark proposed an increase in allowable landings in return for the implementation of mandatory video monitoring to help eliminate discards and high-?grading. One hundred percent observer coverage as onboard and/or video coverage is the only real option to properly account for all catch activities in a transparent manner, as has been demonstrated successfully in other fisheries.    The Sea Around Us Project, February 2010 6 Total marine fisheries extractions by country in the Baltic Sea: 1950-present, Rossing, Booth and Zeller    Total marine fisheries extractions by country in the Baltic Sea: 1950-present, Rossing, Booth and Zeller 7  FISHERIES CATCHES FROM THE BALTIC SEA LARGE MARINE ECOSYSTEM: 1950-20071  Dirk Zeller, Shawn Booth, Sarah Bale, Peter Rossing, Sarah Harper and Daniel Pauly Sea Around Us Project, Fisheries Centre University of British Columbia, 2202 Main Mall, Vancouver, BC., V6T 1Z4, Canada; email: d.zeller@fisheries.ubc.ca; s.booth@fisheries.ubc.ca; s.bale@fisheries.ubc.ca; p.rossing@fisheries.ubc.ca; s.harper@fisheries.ubc.ca; d.pauly@fisheries.ubc.ca   ABSTRACT We estimated the total marine and brackish-water fisheries catches from the Baltic Sea Large Marine Ecosystem, from 1950-2007, taken by Denmark, Estonia, Finland, Germany, Latvia, Lithuania, Poland, Russia and Sweden. Using the ICES electronic catch database as the ‗reported data‘ baseline, we added several IUU (Illegal, Unreported and Unregulated) catch components: ‗adjustments‘ (data source adjustments to the reported landings), ‗unreported‘ landings (termed ‗unallocated‘ by ICES), estimates of ‗discards‘, and estimates of ‗recreational catches‘. This ‗reconstructed total catch‘ was 53.5 million tonnes from 1950-2007, which is approximately 30% higher than landings officially reported by ICES for the same period. For the period since 2000 (2000-2007) this difference was 35%. Our reconstruction is likely an underestimate, as we used minimum values based on conservative assumptions. However, a conservative estimate is still preferable to the current default assumption of ‗zero catch‘ in the absence of officially-reported IUU data. Currently, ICES does not disclose countries that do or do not submit estimates of these other fisheries sectors to ICES, and this lack of transparency prevents countries that are either not collecting or not reporting data from being held accountable by the public and other stakeholders on issues of public resource use.  INTRODUCTION The Baltic Sea, here defined as the Baltic Sea Large Marine Ecosystem (Figure 1; Sherman and Hempel, 2008), is a semi-enclosed sea situated in northern Europe and encompasses a surface area of approximately 394,000 km2 (www.seaaroundus.org). Nine coastal countries (Denmark, Estonia, Finland, Germany, Latvia, Lithuania, Poland, Russia and Sweden) surround the Baltic Sea and are involved in national, regional and international agencies to manage the sea and its resources. The Baltic Sea is a water body that consists of brackish water with large inputs of fresh water originating from rivers in the east and inputs of more saline water entering from the Atlantic Ocean through the Kattegat. The Baltic Sea is therefore often thought of as a large semi- enclosed estuary (Graneli et al., 1990). The watersheds‘ catchment encompasses an area approximately 4 times the surface area of the Baltic Sea (Thulin and Andrushaitis, 2003). The terrestrial landscape surrounding the Baltic Sea has a variety of human impacts that also influence                                                           1 Cite as: Zeller, D., Booth, S., Bale, S., Rossing, P., Harper, S., and Pauly, D. (2010) Fisheries catches from the Baltic Sea Large Marine Ecosystem: 1950-2007. pp. 7-38. In: Rossing, P., Booth, S., and Zeller, D. (eds.), Total marine fisheries extractions by country in the Baltic Sea: 1950-present. Fisheries Centre Research Reports 18 (1). Fisheries Centre, University of British Columbia, Canada [ISSN 1198-6727]. Figure 1. Map of the Baltic Sea with ICES subdivisions and surrounding countries (the Baltic Sea Large Marine Ecosystem consists of all subdivisions indicated).  RF Sweden Finland Russian Federation (RF) PolandGermany Latvia Lithuania Estonia Denmark 30 29 31 25 26 32 27 24 28-2 22 28-1 23 . 0 200 400 Km N8 Total marine fisheries extractions by country in the Baltic Sea: 1950-present, Rossing, Booth and Zeller   the physical and biological components of the sea. Currently, the human population inhabiting the drainage area of the Baltic Sea is approximately 85 million (this includes estimates from non-coastal states that are within the drainage area of the Baltic Sea) and it is estimated that 27 million people live within 50 km of the coastline (Hannerz and Destouni, 2006). Human activity has influenced the productivity of the Baltic Sea, with excessive inputs of nutrients contributing to eutrophication and algal blooms that cause large hypoxic/anoxic areas affecting the biological communities. Inputs of toxins from both point and non- point sources affect water quality, and there are ongoing studies detailing levels of pollutants in the sea‘s organisms (HELCOM, 2003).   The amount of salt- and fresh-water, and hence the salinity gradient, in part, determines the species composition of the aquatic ecosystem, which comprises marine, fresh water and diadromous species. The western portion near Denmark has the highest number of fish species (~100) while the north-eastern portion in the Gulf of Finland has only 20 fish species (Voipio, 1981); see Appendix Table A1 for a listing of taxa considered in this report.   From a fisheries perspective, salinity levels heavily impact whether the system is an Atlantic cod (Gadus morhua) or herring (Clupea harengus)/sprat (Sprattus sprattus) dominated system. Higher biomass and larger catches of cod from both the eastern and western stock have traditionally been considered to occur under more saline conditions, whereas herring and sprat biomass and catches increase during less saline conditions (HELCOM, 2003). Increasing salinity levels are associated with increased fluxes of saltwater entering through the Kattegat, which also results in increased oxygen levels. Higher salinity levels and the associated increased dissolved oxygen concentrations in the deep basins where cod spawn increases the survivorship of cod eggs (Nissling and Westin, 1991). Increased inflow of saltwater to the Baltic also leads to high population levels of marine copepods, the dominant food of pre-adult cod (Hammer et al., 2008). The International Council for the Exploration of the Sea (ICES) reports 153 taxa (including fish, mollusks, bivalves and crustaceans) being landed in fisheries, but cod, herring and sprat are the commercially most important species, accounting for over 90 per cent of reported landings. Fisheries catches and analysis have been previously documented for the Baltic Sea Large Marine Ecosystem (Heileman and Thulin, 2008) using analysis techniques documented in Pauly et al. (2008).  Fishing is known to also impact heavily on the resources and state of the Baltic Sea. Currently, sprat, Gulf of Riga herring, and cod are considered to be overfished in relation to fishing mortality and long-term yield (ICES, 2008a; 2009b), and this affects ecosystem functions and services. The decline in cod biomass since the 1980s (due to both decreases in habitat and excessive fishing mortality) has led to altered trophic relationships that affect the ecosystem. Declining abundance of cod and the increase in biomass of sprat and herring has led to an increase in hypoxic events due to trophic cascades (Österblom et al., 2007). Increased sprat and herring biomass result in increases predation on zooplankton, leaving less zooplankton biomass. Less zooplankton can cause an increase in phytoplankton/algae biomass, resulting in increased levels of eutrophication and hypoxia (Casini et al., 2008).   Information on total catches (in contrast to reported landings) from the Baltic Sea are not readily available, nor have they been comprehensively accounted for. ICES is the agency responsible for disseminating information about the state of the living resources in the Baltic Sea (based on information received from the individual countries), and provides scientific advice to governments and the international regulatory bodies that manage the Baltic Sea (ICES, 2009c). From 1973-2004, scientific advice from ICES, including recommended Total Allowable Catches (TACs) for cod, herring, sprat and salmon estimated through scientific stock assessment procedures, was taken under consideration by the International Baltic Sea Fishery Commission (IBSFC). The IBSFC members negotiated and considered socio-economic factors and political considerations, which generally resulted in higher TACs being allocated for the species covered under the Gdansk Convention (cod, herring, sprat and salmon) than stock assessments recommended. It is important to note that in some years agreements could not be reached and no TACs were agreed upon, leading to even higher fishing mortalities on these species, especially in the mid-1980s. Since 2005, with the changes in the membership of the European Union, TACs are now negotiated between the EU and Russia (Aps et al., 2007). Since 2006, the Baltic Sea Regional Advisory Council advises the member states of the EU and the European Commission on matters concerning the management of fisheries under the EU Common Fisheries Policy.   Changes in the fishing areas within the Baltic Sea where national fleets were allowed to operate have also changed during the time period considered here. In earlier time periods, countries claimed a 3 nm Total marine fisheries extractions by country in the Baltic Sea: 1950-present, Rossing, Booth and Zeller 9  territorial sea, which later increased to 12 nm. In 1978, Sweden became the first country in the Baltic to claim a 200 nm Exclusive Economic Zone (EEZ) under the provisions of the United Nations Law of the Sea (UNCLOS), but because of overlapping claims, the mid-line principle was used to settle claims. The changes brought about by the introduction of EEZs during the later part of the 1970s had the effect of shrinking the fishing areas of some countries (e.g., Denmark; Borberg, 1976). However, with the adoption of the Common Fisheries Policy by members of the EU in 1983, fishing fleets of member countries had access to each other‘s fishing areas (outside of the 12 nm territorial waters, unless fishery access agreements between individual countries were established). With EU membership expanding since 1983, more area of the Baltic Sea has come under EU management.  The officially reported fisheries data, as represented by ICES sources, are known to almost exclusively account for landings, not total catches. ICES stock assessment working group reports do provide some information and data on unallocated (unreported) catches and discards for some species, but unfortunately not in a transparent manner. The unallocated (unreported) catches from working group reports are presented as Baltic Sea-wide total amounts, and not by country, even though it is known that not all countries report these catches. Unfortunately, the default approach by the working groups is to substitute ‗zero‘ for those countries not presenting data for unallocated catches. This approach leads to under-estimation in this catch categories because there is no expansion (or substitution with estimates) methods used to account for countries not reporting. Further, the working group reports do not indicate which countries‘ data are included. This incomplete accounting in scientific stock assessment reports is apparently done for confidentiality reasons, but does not lead to a transparent and publicly accountable catch accounting system. It also hampers attempts to comprehensively assess the true nature of fisheries catches. Yet, to fully account for all catches, estimates of Illegal, Unreported and Unregulated catches (IUU), discards and recreational catches need to be assessed and included to better estimate likely total fisheries catches in the Baltic Sea. A further data source, presently called ‗Fishframe‘ (FishFrame, 2009) that contains information on discards by gear type, species, country and year, as well as some data on unallocated catches is available to authorized users. However, these data are also considered confidential, and access to these data was not given. This database is maintained by Denmark‘s National Institute of Aquatic Resources (DTU Aqua).   Additional data sources used include national data, published and grey literature case studies, unpublished reports, media sources and personal information based on communications and discussions with country- and region-specific experts from around the Baltic Sea region. Interestingly, many of the personal sources were very willing and keen to share their knowledge and information with us, but have expressed a clear preference for not being named, i.e., wanting to remain anonymous, usually out of concern about their perceived scientific standing, or concerns about their job security. Throughout this report, we treat such concerns seriously, and cite ‗anonymous source‘ for such material. We also endeavor to use such information in a manner so as not to make the original source apparent. However, the scientific and public community in Europe should consider it as a point of concern if scientists, environmental and fisheries experts are not willing to speak publicly on their knowledge and experience.  The approach to retroactively estimate total catches uses a bottom-up approach to reconstruct catch time series (Zeller et al., 2007; Pauly et al., 2008). Such an approach often requires assumption-based inferences and interpolations, but is justified, despite data uncertainties, given the less acceptable alternatives that users of official data will interpret non-reported or missing data components as zero catches (Pauly et al., 1998). Estimates of total catches derived from catch reconstructions will clearly not be statistically ‗precise‘ in the sense of having small uncertainty. However, of importance here is the realization that, given our conservative approach to estimation, the estimates that will be derived are ‗less wrong‘ i.e., likely more ‗accurate‘ in the sense of being closer to the ‗true‘ value than the currently assumed ‗zero‘ catch substituted for ‗no data‘ by stock assessments.  The rational for fisheries catch reconstruction lies in creating a baseline of total catches rather than reported landings to better inform policy makers and the general public, and to contribute to the development of ecosystem-based fisheries management, which cannot be done without a comprehensive time-series of fisheries catches. It is hoped that by casting the net wide, and not relying on one set of data, that a better and more comprehensive picture will emerge on the likely total catches taken in the Baltic Sea over the last 50+ years. 10 Total marine fisheries extractions by country in the Baltic Sea: 1950-present, Rossing, Booth and Zeller   METHODS ICES maintains two different publicly accessible databases that provide time series data of fisheries statistics. The ‗ICES catch statistics database’ describes countries‘ reported landings by species (or higher taxonomic grouping), ICES statistical reporting area and year (in two parts: 1950-1972 and 1973-present). Throughout the present report, data from this database are referred to as ‗ICES landings statistics’. The second ICES database, officially called ‗ICES stock assessment results database’, describes countries‘ catches by stock and year, and presents data used by the ICES stock assessment working groups for their annual stock assessments. Hence, this database only presents data on cod, herring, sprat, brill (Scophthalmus rhombus), dab (Limanda limanda), turbot (Psetta maxima), plaice (Pleuronectes platessa) and flounder (Platichthys flesus) and no data on the remaining 145 taxa being caught. As we utilize some of these data, generally in conjunction with information taken directly from the ICES stock assessment working group reports (e.g., ICES Working Group on Baltic salmon and trout 2008 [ICES, 2008a]), we refer to the combined data source (i.e., ‗ICES stock assessment results database’ and ICES stock assessment working group reports) as ‗ICES stock assessment working group data’. In addition to the reported landings by stock, the ‗ICES stock assessment reports’ also contain some data on unallocated (unreported) catches of cod, salmon and Riga herring. Both databases are available electronically from ICES (ICES, 2009a) as are recent ICES stock assessment reports (ICES, 2009b).  As the aim of this study is to estimate total catches (as opposed to reported landings) taken in the Baltic Sea for the time period considered here, all fisheries components were estimated for the nine Baltic countries. The ‗reported data‘ baseline is represented by ‗ICES landings statistics‘. To this, we added four IUU components: ‗adjustments‘ to ICES landings statistics (i.e., data source adjustments to reported landings), unreported landings‘, ‗discards‘, and ‗recreational catches‘. ‗ICES landings statistics‘ formed the baseline for estimating total catches, and estimates of IUU catches were added to these. Discard estimates were applied to the sum of ICES landings, adjustments, and estimates of unreported catches. Recreational catch estimates were also made for each country, allowing estimates of total catches taken from the Baltic Sea from 1950-2007 to be derived. Catches reported by ICES on behalf of non-Baltic countries were presented here, but excluded from this catch reconstruction as they amounted to only 60,000 t over the entire time period considered. The basic methodological approach used here consisted of utilizing as wide a range of information and data sources as possible. Information and data for each catch component that were considered usable were transformed to create so-called ‗anchor points‘, and, when time series were lacking, expansion methods, including linear interpolations between anchor points in time, were done to create full time-series coverage for each catch component (Zeller et al., 2006; Zeller et al., 2007).  ICES landings statistics (ICES, 2009a)2 are available as two data sets covering the time periods 1950-1972 and 1973-2007, and describe the annual commercial landings data by each country for each taxon and statistical reporting area (Figure 1). ICES landings data were adjusted with information provided by our Baltic collaborators, from national sources, or from other ICES documents (e.g., ICES working group reports), and the adjustments plus ICES landings statistics represent what we consider to be the best estimate of reported landings. Unreported catches were generally transformed into anchor points by considering the available data as a proportion of corresponding reported landings per year (e.g., unreported landings data for cod for Denmark was related to the reported landings of cod by Denmark) in order to expand unreported catch estimates to time periods when data were lacking. The same method was used for discards; however, discard rates were applied to the sum of ICES landings statistics, adjustments, and unreported landings. In order to expand to time periods when there was no neighboring anchor point to interpolate to, an assumption-based approach was used. These methods allowed for a complete time series of catch data to be estimated for each catch component for each country that participates in Baltic Sea fisheries.  The general methods presented here represent the default approach for each component for each country. However, whenever country specific data on any of the components were available, such specific data were utilized in preference to the default approach. Each individual country report (this volume) describes such details. To account for some political and economic differences between countries, we accounted for catch components by considering groupings of countries. Western countries were considered to include Denmark, Finland, West Germany (1950-1990) and Germany (after re-unification, 1991-2007), and Sweden; former eastern bloc countries included East Germany (1950-1990), Poland, Russia Federation,                                                           2 ICES refers to these data as ‗catch statistics‘.  Total marine fisheries extractions by country in the Baltic Sea: 1950-present, Rossing, Booth and Zeller 11  and the Baltic states of Estonia, Latvia and Lithuania. Non-Baltic countries that have had fishing access to the Baltic Sea as reported in ICES landings statistics include Faeroe Islands, Netherlands, Norway, Spain, England and Wales.  Illegal, Unreported and Unregulated (IUU) catches  Here, we consider that adjustments to ICES landings statistics, unreported landings, discards and recreational catches all form components of IUU catches. Each of these components was treated separately in both a taxon- and country specific manner when information was available, or a country‘s IUU components were estimated with assumed default methods, when information was not available.  Adjustments to ICES landings statistics  Adjustments to ICES landings statistics were taken from ICES stock assessment working group data (e.g., ICES, 2008a), the ‗ICES stock assessment results‘ database (ICES, 2009a), and nationally reported data. These adjustments were incorporated to present the best estimate of reported landings.  National data were used for some countries in the early time periods (1950s) to fill gaps from the first year of ICES reported data back to 1950 (i.e., some countries data time series in ICES reported landings statistics do not start until after 1950, even though data are reported in national documents). Landings reported by the ICES stock assessment working groups were generally used for cod and flatfishes, as the stock assessment working group data provided specific landings data for the eastern and western cod stocks (ICES, 2007; 2008a), while the stock assessment working group data for flatfishes (ICES, 2008a) provided a better taxonomic breakdown.  Unreported landings  The foundation for our estimates of unreported landings was guided by what ICES calls ‗unallocated‘ catches and which we refer to as unreported landings throughout this report. The ICES stock assessment working group data only contain Baltic-wide summary data of unreported landings for the western and eastern cod stocks, salmon, and Riga herring.3 Unreported landings of both cod stocks and salmon were converted into percentages of the Baltic-wide reported landings for salmon and each cod stock (as reported in the respective working group reports) to form anchor points. Unfortunately, the ICES source did not provide country-specific transparency in regards to unreported landings. The Baltic-wide percentage rates, which were adjusted if a country was known to not report unallocated landings to ICES (see cod below), were then applied to the sum of each country‘s ICES landings statistics and adjustments. As it is known that not all countries provide data on unreported landings (or even estimates thereof) to stock assessment working groups, the resulting unreported landings totals obtained through our approach are minimal estimates at least at the aggregated level.   Cod: Anchor points for unreported catches of western cod were available for 1993, 1994, 1996, 2004, and 2005 (Table 2.4.1 in ICES, 2007) as a total for all countries, while the eastern cod stock had anchor points developed for 1993-1996, and from 2000-2007 (Table 2.3.1 in ICES, 2008a). However, these anchor points were adjusted by excluding Sweden‘s reported landings from the total landings because Sweden does not report any unallocated cod landings to the ICES working groups (Persson, this vol.). Despite repeated requests to ICES to obtain information on which countries did not contribute data, we were not permitted access to that knowledge (Table 1). The adjusted rate was estimated as,                                                             3 For details regarding unreported catches of Riga herring, see reports on Estonia (Veitch et al., this volume) and Latvia (Rossing et al., this volume), which are the countries that exploit this stock. Table 1. Default anchor points (%) used for estimating unreported landings from the western cod stock in the Baltic Sea based on ICES (2007, Table 2.4.1). Dashes (-) indicate years of linear interpolation between anchor points.   Countries Year Western  Eastern  1950 5.00a 0.00b 1951-1979 - 0.00b 1980 20.10c 0.00b 1981-1990 - 0.00b 1991-1992 - - 1993 40.20 40.20 1994 39.64 39.64 1995 - - 1996 5.34 5.34 1997-1999 - - 2004 0.07 0.07 2005 0.04 0.04 2006-2007 0.04d 0.04d a assumption of  5% of reported landings; b assumption of zero unreported landings; c assumption of ½ the 1993 rate; d2005 value carried forward. 12 Total marine fisheries extractions by country in the Baltic Sea: 1950-present, Rossing, Booth and Zeller   Unreported % = unallocated catches /(total cod landings – Sweden‘s cod landings)  In order to account for historical and political differences, the former eastern bloc countries (East Germany, Estonia, Latvia, Lithuania, Poland, and the Russian Federation) were treated separately from western countries. The western countries, which already had market economies in 1950, were assumed to have unreported landings throughout the time period considered here. As there were no reports documenting unreported landings prior to 1993, we used an assumption-based approach to derive an anchor point for 1980 and 1950. For 1950, it was conservatively assumed that unreported landings were 5% of the reported landings, while for 1980, unreported landings were assumed to have been half the 1993 estimated rate (Table 2). For the former eastern bloc countries, with state controlled economies in the early years, it was assumed that there was no incentive to not report all landings because prices for different species were similar enough to discourage mis-reporting (R. Oeberst, pers. comm., Johann Heinrich von Thünen-Institut). Thus, unreported landings by eastern countries were conservatively set to zero from 1950 to 1990, and then linearly interpolated to 1993, the first year when estimates of unreported catches were available (Table 2).  Atlantic salmon: Unreported landings of salmon are not presented in the ICES stock assessment results database; however, some information on unreported landings of salmon were available from ICES working group reports (ICES, 2008b). The working group reports Baltic-wide estimates of the mode, minimum and maximum of unreported landings for all Baltic countries combined from 1981- 2007. Countries whose reported landings data included recreational catches (Denmark, Finland, and Sweden; see Table 2.1.2 in ICES, 2008b) had anchor points derived in a slightly different manner than countries that did not report recreational catches.  Thus, we used two separate assumption-based approaches to estimate salmon IUU catches:  a) for Denmark, Finland, and Sweden, whose reported landings in Table 2.1.2 in ICES (2008b) included estimates of recreational catches, we utilized both the mode and the minimum value of unreported landings from the ICES salmon working group report (Table 2.1.2 in ICES, 2008b) to estimate country specific unreported salmon catches. The mode was used prior to countries reporting recreational catches, and after countries started to report recreational catches the minimum was used; and b) for all other countries, whose reported landings data in Table 2.1.2 in ICES (2008b) did not include recreational catches, we used the mode value of unreported landings from Table 2.1.2 in ICES (2008b) to estimate country specific unreported catches.  In keeping with the assumption-based approach, the former eastern bloc countries were assumed to report all landings, and thus, were conservatively assumed to have no unreported landings from 1950-1990 (Table 3). To account for the shift to market economies and the associated underreporting incentives, linear interpolations were made between 1990 and the first anchor point in 1993 based on the above described method for estimating unreported landings. Prior to applying this method to the western countries, a correction was applied to reported landings, as it was assumed that the eastern bloc countries reported all landings of salmon prior to 1990. For the earlier time periods, assumption-based approaches were used, but when unreported landings were first documented in ICES working group data, the estimates of unreported landings between 1950 and 1990 for western countries in each year were calculated as,  Unreported landings % = mode of unreported catches/[total reported landings – eastern bloc landings]  Table 2. Default anchor points (%) used for estimating unreported landings of eastern cod stocks in the Baltic Sea based on ICES (2007, Table 2.3.1). Dashes (-) indicate years of linear interpolation between anchor points.   Eastern cod Year Western countries Eastern countries 1950 5.0a 0.0b 1951-1979 - 0.0b 1980 31.1c 0.0b 1981-1990 - 0.0b 1991-1992 - - 1993 62.2 62.2 1994 103.0 103.0 1995 30.0 30.0 1996 10.0 10.0 1997-1999 - - 2000 46.0 46.0 2001 47.6 47.6 2002 46.6 46.6 2003 61.5 61.5 2004 52.9 52.9 2005 46.4 46.4 2006 47.9 47.9 2007 43.2 43.2 aassumption of 5% of reported landings;  bassumption of zero unreported catches;  cassumption of ½ the estimated rate for 1993. Total marine fisheries extractions by country in the Baltic Sea: 1950-present, Rossing, Booth and Zeller 13   In 1950, West Germany‘s unreported landings of salmon were assumed to be 5% of reported landings and linear interpolations were used to the first anchor point in 1981 (Table 4). After 1981, West Germany‘s unreported landings were based on the mode. Finland has reported its recreational catches since 1953, and in order to estimate unreported landings of salmon, we set the 1950 rate to 5% of ICES reported landings and linearly interpolated to the mode estimate in 1981, but used the minimum value from 1981- 2007. Sweden and Denmark started to report recreational catches in 1988 and 1998, respectively, and thus were treated differently than Germany and Finland. In these cases, it was also assumed that the 1950 unreported rate was 5%, and linear interpolations were carried out to the first unreported landings rate using the adjusted unreported landings percentage for each year i.e.,  Unreported landings % = min of unreported catches /[total reported landings– eastern bloc landings]   For Finland, Sweden and Denmark, the minimum estimate was used for the year that recreational catches (1981, 1988 and 1998, respectively) were included in the reports to ICES. Thus, for Finland, the 1950 estimate of unreported catches was 5% in 1950 and this rate was linearly interpolated to the mode-based rate for 1981. However, since Finland started to report its recreational catches in 1981, the estimate of unreported landings for the 1981-2007 time period were based on the minimum rate-based estimate of unreported catches (Table 4).  Table 4. Default anchor points (%) used for estimating unreported landings of Atlantic salmon in the Baltic Sea for western countries reporting recreational catches (Finland, Sweden and Denmark), and for Germany based on Table 2.1.2 in ICES (2008b). Dashes (-) indicate years of linear interpolation between anchor points.  Year Finland Sweden Denmark Germany Year Finland Sweden Denmark Germany 1950 5.0a 5.0a 5.0a 5.0a 1993 6.0b 5.9b 19.4c 19.4c 1951-1979 - - - - 1994 6.0b 5.9b 18.7c 18.7c 1980 19.9b 19.9c 19.9c 19.9c 1995 6.0b 6.2b 19.5c 19.5c 1981 6.0b 20.4c 20.4c 20.4c 1996 6.0b 6.4b 20.4c 20.4c 1982 6.0b 20.7c 20.7c 20.7c 1997 6.0b 6.5b 20.8c 20.8c 1983 7.0b 22.6c 22.6c 22.6c 1998 6.7b 6.7b 6.7b 20.1c 1984 6.0b 20.7c 20.7c 20.7c 1999 6.6b 6.6b 6.6b 20.4c 1985 6.0b 19.7c 19.7c 19.7c 2000 6.8b 6.8b 6.8b 19.9c 1986 7.0b 22.0c 22.0c 22.0c 2001 6.6b 6.6b 6.6b 20.4c 1987 6.0b 21.4c 21.4c 21.4c 2002 6.5b 6.5b 6.5b 20.5c 1988 7.0b 7.1b 22.2c 22.2c 2003 6.7b 6.7b 6.7b 20.1c 1989 7.0b 7.1b 22.6c 22.6c 2004 6.0b 6.0b 6.0b 20.6c 1990 7.0b 7.2b 24.4c 24.4c 2005 6.2b 6.2b 6.2b 20.7c 1991 6.0b 5.8b 19.5c 19.5c 2006 6.0b 6.0b 6.0b 22.2c 1992 6.0b 5.6b 20.1c 20.1c 2007 6.5b 6.5b 6.5b 21.4c a assumption of 5% of reported landing